6

A Review of the Safety and Effectiveness of Select Ingredients in Compounded Topical Pain Creams

The safety and effectiveness of each active pharmaceutical ingredient (API) in a compounded topical pain cream depends on two factors. First, the API should have a mechanism of action to treat pain, and second, the topical formulation must deliver the API to the site of action in an amount that is sufficient to achieve an effect but is also appropriate to be safe. In theory, topical APIs intended to treat local or regional pain act on nerves in the skin or in underlying muscles or joints by blocking nerve signals, reducing inflammation, relaxing muscle spasms, or increasing the effects of other substances (Cline and Turrentine, 2016; Leppert et al., 2018). However, distal responses to topical applications may also occur (Glinn et al., 2017; Leppert et al., 2018). Whether these systemic actions are intended or not, they are critically important considerations in the review of a compounded topical preparation’s safety and effectiveness profile.

Despite their many benefits, all pain medications—including U.S. Food and Drug Administration (FDA)-approved pain medications—have some potential for adverse effects or intolerance among certain populations of patients. For example, chronic use of nonsteroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen may cause gastrointestinal effects (e.g., ulcers, bleeding) as well as be linked to kidney and liver damage. Based on clinical studies and the product’s label, pharmacists and prescribing clinicians know to exercise certain cautions when treating certain populations with NSAIDs, including older populations, patients on anticoagulants or steroid drugs, and women in early pregnancy. The ongoing opioid epidemic in the United States serves as another cautionary example of the potential risks of pain medication. Inappropriate use of opioids in pain management plans

has contributed to a substantial increase in the rates of opioid addiction and consequential deaths from overdose (Solomon et al., 2010; Wehrwein, 2010).

Given this context, clear, evidence-based conclusions on the safety and effectiveness of pain medications helps to inform clinical guidelines for the use of pain medications and to mitigate risks of adverse effects in specific patients. Currently, little is known about the safety and effectiveness of compounded topical pain creams. In the context of the somewhat recent rise in the supply and demand of compounded preparations, this gap in knowledge creates a substantial public health concern.

This chapter addresses the committee’s charge to identify and analyze available scientific data relating to the ingredients used in compounded topical pain creams. There are several sections in this chapter. A total of 20 ingredients were selected using a process described in the next section, followed by a summary of the search strategy employed to identify relevant data. Following this introductory overview of the committee’s research approach, overall conclusions and the data on which they are based are tabularized in the subsequent section. Next, each of the 20 ingredients is separately discussed, with ingredients grouped by category. Of note, some published reports also include data on compounded preparations containing multiple active ingredients. The committee summarized these research findings in the final section of this chapter.

Primary data reviewed by the committee included safety, efficacy, and/or effectiveness studies in humans.¹ When available, the committee prioritized findings of randomized controlled trials (RCTs), followed by quasi-experimental designs and cohort studies. Select case reports are briefly summarized to inform the overall database for each ingredient. Effectiveness data are followed by pharmacokinetic and safety data for each ingredient, where available. Additionally, preclinical animal pharmacokinetic data are discussed for ingredients that lacked such evidence in humans.

___________________

¹ The current research study has a charge to evaluate the effectiveness, rather than the efficacy, of compounded topical pain creams. While similar, the terms are not synonymous. Efficacy refers to the therapeutic effect of a treatment under controlled conditions, while effectiveness refers to the therapeutic effect in “real-world” situations in which certain contextual measures (e.g., placebo effect) may not be strictly controlled and broad outcome measures (e.g., health-related quality of life) are considered. Effectiveness data alone may not be sufficient to inform conclusions regarding a treatment’s therapeutic effect (Ernst and Pittler, 2006; Kim, 2013). For the purposes of this report, the committee evaluated all relevant data produced by randomized controlled trials, nonrandomized clinical studies, case reports, and where applicable, preclinical studies to help address the study’s charge. As a result, many of the research findings discussed throughout the report assess outcomes related to the potential effectiveness and/or efficacy of compounded topical pain creams. Given its broader application to the body of research reviewed, the term effectiveness is used more generally across the report.

It is important to note that the committee’s review of the literature primarily focused on the application of nonsterile compounded topical creams to treat pain when applied to intact external skin.² The committee largely excludes discussions on wound care and mucosal membranes and cavities (e.g., mouth, eyes, vulva, vagina, anus, and nose); however, in select circumstances these findings are included where there is insufficient evidence to comment on outcomes of topical application on intact skin. In addition, evidence on oral administration is occasionally discussed to highlight potential side effects if the data show evidence of systemic absorption.³

RESEARCH APPROACH

Selection of Active Pharmaceutical Ingredients for Review

APIs evaluated by the committee are listed in Box 6-1; these APIs were chosen using a process described in Chapter 1. FDA introduced a list of 37 APIs that have been identified in common formulations of compounded topical pain creams, 10 of which were determined to be of high-priority interest and were selected for evaluation by the committee. Recognizing that this high-priority list was not a comprehensive list of ingredients used in compounded topical pain cream products, the committee chose 10 additional chemicals to ensure examples from various possible pain pathways in humans were represented. Specific choices were made to expand the groups of substances (e.g., inclusion of drugs with potential local and systemic anesthetic actions) and to include additional drugs from important groups (e.g., NSAIDs). It is important to note that the omission of a category, or mechanism, does not imply safety or effectiveness of drugs in that category when used in compounded topical pain creams.

Search Strategy Summary

The study committee conducted a literature search to identify a comprehensive body of evidence to inform its work. In coordination with one of the National Academies’ senior research librarians, the committee constructed a literature search strategy to identify a broad body of research evidence. An initial search queried six databases—Medline, Embase, PubMed, Scopus, ClinicalTrials.gov, and Toxnet—for content

___________________

² There is a growing literature on the overlap of signaling pathways between sensations related to itch and pain (Schmelz, 2019). As such, there are several APIs with indications for itch that may be also used to treat pain and vice versa. However, for the purposes of this report, the focus is on the effectiveness of the API to treat pain.

³ See Appendix G for an overview of the potential systemic side effects related to the use of the 20 APIs reviewed.

related to the safety and effectiveness of the 20 APIs used in topical products. This literature search included human, animal, and in vitro studies, but it was limited to peer-reviewed articles published in the English language. The search was not limited by date of publication, but editorials, commentaries, letters, and notes were excluded. A complete description of the search syntax used can be found in Appendix B. The search yielded 1,716 articles for the 10 APIs prioritized by FDA and 7,306 articles on the additional 10 APIs added by the committee with potential relevance to the committee’s charge.

An initial screen of the 9,022 articles was performed by National Academies staff to eliminate articles not relevant to the study’s scope. A total of 7,230 articles were removed for either (1) discussing only topical application to the eye, or (2) not discussing the treatment of pain in the title, keywords, or abstract of the article. After this first screen, the remaining 1,792 articles were assigned a type of evidence rating by National Academies staff based on study design, using a modified Cochrane scale to expand the nonrandomized trial categories (Harris et al., 2001; University of Oxford, 2009). Given the committee’s research questions, the scope of the literature search was limited to the topical application of any of the 20 ingredients to skin with the intent to treat pain. Committee members then screened all articles for content relevance, which resulted in a total of 169

articles.⁴ See Figure B-1 in Appendix B for a depiction of the flow of articles through the search and selection process. See also the Search Strategies section in Appendix B for the full list of specific terms used in the literature search efforts.⁵

After articles relevant to the committee’s task were identified, all systematic reviews, RCTs, and all clinical studies with a control group were reviewed for evidence regarding the safety and effectiveness of the drug applied topically to treat pain. Case reports, case series, and preclinical studies were discussed (where relevant) for APIs lacking the level of evidence provided by systematic reviews and controlled clinical studies.⁶ Because study design is not the only measure of quality evidence, the committee also considered evidence of methodological rigor detailed in the publications in its review of the evidence. Using the 2019 Cochrane Risk of Bias Assessment Tool (Sterne et al., 2019), the committee evaluated the risk of bias in all RCTs identified as relevant to the committee’s charge.⁷ In summary, out of the 22 RCTs that reviewed the effectiveness of individual APIs, 7 were determined to have a high risk of bias, 13 have a low risk of bias, and 2 were noted as having some concerns. Notably, out of the 11 RCTs that reviewed the effectiveness of multiple ingredients in compounded topical pain creams, 7 were determined to have a high risk of bias, 3 have a low risk of bias, and 1 was noted as having some concerns. The level of bias attributed to each RCT is included in footnotes within the summaries below.

___________________

⁴ Of note, these calculations represent the true number of articles resulting from the 20-ingredient literature search. There are a number of articles with relevance to multiple APIs that are represented more than once in the total. In addition, the committee identified many other published studies that have tangential relevance to the committee’s charge; however, only the studies with the most direct relevance were reviewed in this report. As such, the references for this chapter represent the most relevant data and are not an inclusive list of all of the articles reviewed.

⁵ In recognition of the limited peer-reviewed evidence to describe the safety and effectiveness of compounded topical pain creams, the committee also reviewed submitted resources from national stakeholders, such as the Professional Compounding Centers of America, to support its research efforts. These are available in the study’s Public Access Folder; see https://www8.nationalacademies.org/pa/managerequest.aspx?key=HMD-HSP-18-18 (accessed April 9, 2020).

⁶ Case series, case reports, and preclinical studies were considered to be low-tier evidence for informing the clinical effectiveness of APIs in topical compounded topical creams (Harris et al., 2001; University of Oxford, 2009).

⁷ A few additional RCTs are referenced within the summaries, but their research focus was considered to be out of scope.

SUMMARY OF RESEARCH FINDINGS

Topical Application of Single-Ingredient Compounded Topical Pain Creams

From its review of the evidence, the committee determined several key findings related to the safety and effectiveness of topical application of single-ingredient compounded pain cream preparations, including the following:

• There are very few well-designed, randomized, placebo-controlled research trials that have investigated the bioavailability, effectiveness, or safety of active pharmaceutical ingredients in compounded topical pain creams used to treat pain.
• There is preclinical data (for a limited number of ingredients) that suggest a potential for effectiveness to treat pain in humans.
• Based on the available evidence, the 20 ingredients reviewed demonstrated wide variability in their potential for skin absorption, which was found to be dependent on the drug characteristics, skin condition, and excipient(s) used.⁸
• If topical dosing is sufficient to provide centralized pain relief, adverse effects and drug interactions for systemic exposure are a concern.
• For many specific combinations of drugs and excipients, evidence is inadequate to determine whether the ingredients are absorbed at the local, regional, or systemic level, and there is very little data on the relative risk of adverse effects in any of those three categories.
• When used appropriately, little conclusive data support high risk of safety concerns for any of the 20 studied ingredients beyond local skin irritation; however, high levels of systemic absorption can have potentially life-threatening consequences particularly for preparations including ketamine, clonidine, and bupivacaine.

The committee details its review of the literature below; however, to provide clarity, the committee organized its major research findings into a table (see Table 6-1). This table includes short, descriptive summaries related to the research findings on the safety and effectiveness of the 20 APIs reviewed in this report. It is critical to note that in an effort to create such a table, a substantial amount of context and detail was sacrificed. To obtain a comprehensive appreciation of the body of evidence reviewed in

___________________

⁸ A significant limitation in a large portion of the body of evidence reviewed is the failure of studies to disclose the excipients or enhancers used in the compounded topical formulations. The absence of this information hinders a complete interpretation of the research results.

this report—including the pain conditions, measurements, and assessments, as well as the quality of the studies reviewed—it is critical that the reader reviews the detailed narrative text in the remainder of the chapter.

Topical Application of Multiple-Ingredient Compounded Topical Pain Creams

Given their popularity and use (see Chapter 2), the committee also reviewed evidence on the safety and effectiveness of compounded topical pain creams that contain combinations of multiple APIs. In summary, little information was found to inform the committee’s charge. Several RCTs, clinical trials, and case reports were identified, but overall there was a dearth of information on the performance of individual agents compared to combined agents, which limited the interpretability of the results. In addition, many of the articles lacked appropriate controls to address potential placebo effects. As discussed in Chapter 2, the placebo effect is a well-recognized experience in the health care field (IOM, 2011; NIA, 2020). Rubbing a cream on skin can be comforting, and the action itself can be therapeutic, an important confounder that can be addressed in well-designed, placebo-controlled trials.

An additional confounding factor in several of the multiagent studies is the presence of lidocaine in the compounded formulation. Lidocaine is already FDA approved clinically to induce topical anesthesia, and data were insufficient to determine whether effectiveness improved in combination with other ingredients. Importantly, no conclusions can be made on the committee’s review of the creams containing five to seven ingredients because of the consistent presence of lidocaine, the inclusion of APIs not evaluated individually by the committee, or the lack of appropriate controls. In regard to findings related to safety, the adverse effects were determined to be similar to those reported for the individual ingredients. A more detailed review of the committee’s findings on the safety and effectiveness of multiple-ingredient compounded topical pain creams can be found in the later sections of this chapter.

TABLE 6-1
Available Clinical Evidence for the Topical Application of SingleIngredient Compounded Pain Preparations

A REVIEW OF THE SAFETY AND EFFECTIVENESS OF SELECT ACTIVE PHARMACEUTICAL INGREDIENTS IN COMPOUNDED TOPICAL PAIN CREAMS

Alpha-2-Adrenergic Receptor Agonist: Clonidine

Summary

Clonidine, an alpha-2 and imidazoline receptor agonist, is typically used to treat hypertension, but it is also used to treat attention deficit hyperactivity disorder in children, and it is used as adjunct epidural therapy for severe cancer-related pain (Yasaei and Saadabadi, 2019). Although it does not have an FDA-approved indication for pain, studies related to the use of topical clonidine to treat pain largely focus on its efficacy for peripheral neuropathy.

Based on a small systematic review, RCTs, and observational studies there is some suggestive evidence for the efficacy of topical clonidine cream (0.1 percent to 0.2 percent concentration) to alleviate pain for diabetic neuropathy; however, the data are conflicting. No significant safety concerns are reported, but the area of application and serum levels have limited discussion in the studies. See Box 6-2 for a summary of research findings.

Overall, the committee reviewed four RCTs: two studies were assessed as having a low risk of bias (Campbell et al., 2012; Kiani et al., 2015a), and two studies raised some concerns of bias (BioDelivery Sciences International, 2017a,c; see Appendix B for more details). In addition to reviewing studies where clonidine alone was analyzed, the committee also included studies of clonidine used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Systematic review

A systematic review conducted by Wrzosek et al. (2015) assessed the efficacy of clonidine (alone) in topical creams to treat chronic neuropathic pain in adults and the frequency of adverse events related to use. The systematic review examined the outcomes of two studies, Campbell et al. (2009, 2012),⁹ determined by the authors to be of moderate to low quality. In these studies, the authors conducted a double-blind, randomized, placebo-controlled parallel study to evaluate the efficacy of topical clonidine 0.1 percent gel in diabetic patients (type 1 or 2) with painful peripheral neuropathy defined as an average numeric pain rating

___________________

⁹ Risk-of-bias assessment by committee (Campbell et al., 2012): Low (see Appendix B for more details). Note that, given its publication as an abstract, a risk-of-bias assessment was not conducted for Campbell et al. (2009).

scale (NPRS) ≥ 4 during the 7 days prior to treatment. Participants were randomized in blocks in order to stratify baseline pain severity. Participants applied a metered dose (0.65 mg clonidine) to each foot three times per day for 12 weeks. One hundred seventy-nine patients completed the study. No significant difference was noted in reduction in pain intensity between clonidine and placebo groups.

The authors also tested a pain response to capsaicin in a subset of participants. In the participants who experienced a pain response ≥ 2, the clonidine group experienced a mean decrease in NPRS of 2.6 compared to a reduction of 1.4 in the placebo group. Serum concentrations were obtained at regular visits. Specific pharmacokinetic results are not provided. However, the authors state that only two patients had serum clonidine concentrations greater than 200 pg/ml (therapeutic is > 1,000 pg/ml) and that these participants did not experience adverse effects attributable to clonidine. However, no details are provided. The gel used in these studies appears to be a proprietary compound from Arcion Therapeutics (2009).

Randomized controlled trials

A double-blind, randomized, placebo-controlled parallel study conducted by BioDelivery Sciences International evaluated the efficacy of topical clonidine 0.1 percent gel in diabetic patients (type 1 or 2) with painful peripheral neuropathy defined as an average NPRS of ≥ 4 during the 24 hours prior to treatment (BioDelivery Sciences

International, 2017a).¹⁰ A total of 130 subjects were enrolled in each group, and 117 completed the study in the clonidine group and 114 completed the study in the placebo group. The clonidine group applied a total dose of 3.9 mg clonidine to each foot each day for 12 weeks with an option to enroll in an open-label study following the treatment phase. The primary endpoint was reduction in NPRS. Clonidine was not superior to placebo in change in NPRS from baseline. In addition, there was no difference between the two groups in mean daily NPRS scores for worst pain intensity.

A double-blind, randomized, comparator-controlled parallel study evaluated the efficacy of topical clonidine 0.1 percent gel in diabetic patients (type 1 or 2) with painful peripheral neuropathy defined as an NPRS of ≥ 4 during the 24 hours prior to treatment (BioDelivery Sciences International, 2017c).¹¹ Gel is described as an aqueous gel formulation for topical use. Gel was applied to both feet three times per day, but the total amount of drug applied to each foot was not mentioned. A total of 138 patients were enrolled and evenly divided between the two groups; 58 in the clonidine group completed the study compared to 67 in the comparator group. Clonidine gel was not superior to the comparator. No significant difference occurred in mean reduction in NPRS or mean daily worst pain intensity NPRS scores between the two groups.

A randomized, double-blind study was conducted to evaluate the efficacy and safety of clonidine versus capsaicin in the treatment of painful diabetic neuropathy in patients with type 2 diabetes pain assessed using a visual analog scale (VAS) of ≥ 4 (Kiani et al., 2015a).¹² Clonidine was administered as a 0.1 percent gel with details about the preparation of the gel provided. Capsaicin was supplied as a 0.75 percent cream. Drugs were applied below the ankle to the feet three times per day for a period of 12 weeks. Participants were assessed at 4-week intervals. There was no difference between the groups in reduction in pain as assessed by the VAS. A total of 70 patients were allocated to the capsaicin group and 69 to the clonidine group; 30 participants dropped out of the capsaicin group and 16 from the clonidine group. No significant changes in blood pressure were noted in the clonidine group, of which 53 patients completed.

Clinical studies

An open-label study to assess the long-term use of clonidine 0.1 percent gel was conducted after the completion of the 12-week study noted above (BioDelivery Sciences International, 2017b). A total of 197 subjects were enrolled in the open-label study. It is not clear how many of the 197 subjects came from the previously mentioned study. Clonidine was

___________________

¹⁰ Risk-of-bias assessment by committee: Some concerns (see Appendix B for more details).

¹¹ Risk-of-bias assessment by committee: Some concerns (see Appendix B for more details).

¹² Risk-of-bias assessment by committee: Low (see Appendix B for more details).

applied in the same manner as the double-blind, placebo-controlled study. Only 47 subjects completed the study. The primary outcome was the summary of neuropathic pain symptom inventory documented at the last visit. The mean score was 34.4, which, based on supportive clinical guidance, is categorized as high pain symptom severity (Wong et al., 2019).

A pilot study was conducted in 17 patients with chronic orofacial pain who were diagnosed with neuralgia or neuropathy (Epstein et al., 1997) and given clonidine 0.2 mg/g in a cream base manufactured by Glaxo. Participants were instructed to apply the cream with their finger to the site of pain four times per day for 4 weeks. A VAS was used to determine level of pain. Out of 17 patients, 7 individuals reported improvement of 75 percent or greater, 8 reported no improvement, 1 reported 40 percent improvement, and 1 reported 10 percent improvement. In patients with neuralgia, 4 out of 7 reported improvement, and 6 out of 12 patients with neuropathic pain reported improvement.

Case report

Kopsky and Keppel Hesselink (2017) describe a 54-year-old woman with neuropathic pain in both feet reported as 8 out of 10 on the Numerical Rating Scale (NRS) following chemotherapy treatment despite being prescribed oral gabapentin 2,000 mg daily and oxycodone 20 to 30 mg daily. Pain was reduced to 3 by a test application of compounded topical baclofen 5 percent cream, which was said to be a greater reduction than with compounded clonidine 0.2 percent cream or compounded lidocaine 3 percent combined with isosorbide dinitrate 0.4 percent cream. Allodynia was still present but eliminated with ketamine 10 percent cream. Pain level increased over time from 3 to 7 on the NRS, which led to further treatment.

Dermal Penetration/Bioavailability

Randomized controlled trials

Campbell et al. (2009) reported that subjects with detectable clonidine concentrations after topical application of clonidine 0.1 percent or 0.2 percent gel achieved more pain relief than subjects who did not have detectable serum concentrations. Clonidine has an elimination half-life of 6–20 hours (PubChem, 2020a).

Campbell et al. (2012) did not provide specific pharmacokinetic results; however, the authors reported that participants’ serum levels of clonidine at 2 weeks were similar to those at 12 weeks. They also noted that generally, serum levels were below the level of detection (10 pg/mL) but that two outliers produced levels at 796 pg/mL and 315 pg/mL. For context, they note that the average plasma level of clonidine for treating hypertension is more than 1,000 pg/mL.

Safety and Adverse Effects

Randomized controlled trials

In Campbell et al. (2012) discussed above, no serious adverse events were noted in the clonidine group, but skin reactions were noted in the placebo group. In BioDelivery Sciences International (2017a), discussed above, more serious adverse events, which include cardiac, gastrointestinal, musculoskeletal, nervous system, respiratory, and vascular disorders, were noted in the clonidine group (12 versus 7 in the control group) while other nonserious adverse events, including peripheral edema, back pain, headache, and skin disorders were in the placebo group (58 versus 38). In BioDelivery Sciences International (2017c), a small number of serious adverse events were documented (clonidine group—2; comparator group—6) as well as other adverse effects (clonidine—4; comparator—2). Finally, in BioDelivery Sciences International (2017b), discussed above, serious adverse events were reported in 23 patients and other adverse events in 79 patients.¹³

From the FDA-approved label, serious adverse effects from oral administration of this API include increased suicidal thoughts and cardiac arrhythmias. See Table G-1 in Appendix G.

Anticonvulsant: Carbamazepine

Summary

Carbamazepine is an anticonvulsant drug that exerts its effect by blocking voltage-gated sodium channels, which results in neural membrane stabilization and decreased ectopic nerve discharges. In the United States, oral carbamazepine is FDA approved for the treatment of patients with focal onset seizures and generalized onset seizures. For neuropathic pain, it is FDA approved for the treatment of trigeminal neuralgia and glossopharyngeal neuralgia. It also has an FDA-approved indication for bipolar disorder (acute treatment of hypomania and mild moderate manic or mixed episodes). Some studies, mostly performed prior to gabapentin becoming available, suggest effectiveness in patients with painful diabetic neuropathy and possibly other neuropathic pain conditions.

Systemic carbamazepine has many side effects that include dose-dependent toxic effects that more commonly occur with rapid dosage increase, such as dizziness, drowsiness, nausea, ataxia, and blurred vision. Other potential side effects are impaired liver function and hyponatremia that require routine monitoring during therapy. Treatment with oral carbamazepine has also been associated with serious side effects: blood dyscrasias (bone marrow depression, including aplastic anemia, agranulocytosis,

___________________

¹³ One confirmed death during the course of the study, but cause of death was not disclosed.

leukopenia thrombocytopenia), lymphadenopathy, Stevens-Johnson syndrome, toxic epidermal necrolysis, drug reaction with eosinophilia and systemic symptoms, and others. Other dermatological side effects include pruritus and skin rash. Systemic carbamazepine also has many drug–drug interactions and is a strong CYP3A4 inducer that reduces the serum levels of many other medications.

Related to the topical application of carbamazepine, well-designed clinical studies that address the committee’s specific research interests are not available. The literature includes a few case studies either as single-ingredient compounded topical drug or in combination with other medications. The quality of the data does not allow any conclusion regarding effectiveness or safety. Information on the safety of topical carbamazepine is limited. The quality of the data does not allow any conclusion regarding effectiveness or safety. As described above, as an oral medication it has many systemic side effects including potentially severe drug reactions that affect the bone marrow, skin, and liver. There is a suggestion in preclinical and in vitro studies that topical carbamazepine is toxic to epidermal keratinocytes when applied in high concentration and with a long exposure. See Box 6-3 for a summary of research findings.

Effectiveness

There are no relevant clinical studies that examine the single-drug use of carbamazepine in compounded topical creams to treat pain when applied to intact skin.

Dermal Penetration/Bioavailability

In vitro studies

A study by Fourie et al. (2004) in Franz diffusion cells using human abdominal skin showed that modifications to the (very lipophilic) carbamazepine molecule can enhance skin permeation by increasing the aqueous solubility, in particular as N-methyl- and as N-(2-hydroxyethyl)carbamazepine. Of note, the excipients vary according to author, with no clear information about the effect on skin penetration or absorption (see Anonymous, 2008; Zhou et al., 2015). See Chapter 5 for a discussion on the critical role of excipients in drug delivery and absorption.

Safety and Adverse Events

In vitro studies

Al-Musawi et al. (2017) conducted a toxicity analysis of topical drugs (i.e., amitriptyline, carbamazepine, gabapentin) commonly used for the treatment of neuropathic orofacial pain. After examining the effects on keratinocytes in cell culture, the authors reported that carbamazepine was cytotoxic to skin and oral keratinocytes, demonstrating significant decrease in cellular viability and cell counts, at high concentrations (1.7 mM) and with long exposure (2 hours). The authors also reported that topical gabapentin was only minimally cytotoxic to skin or oral keratinocytes at high concentrations (5.54 mM) and long exposure (24 hours). Importantly, topical amitriptyline was reported as cytotoxic to skin and oral keratinocytes at both short (30 minutes) and long (24 hours) exposure times and at low (200 µM) and high concentrations (1.8 mM).

From the FDA-approved label, serious adverse effects from oral administration of this API include Stevens-Johnson syndrome, toxic epidermal necrolysis, atrioventricular block, syncope, and liver failure. See Table G-1 in Appendix G.

Anticonvulsant: Gabapentin

Summary

Gabapentin, while structurally related to gamma aminobutyric acid (GABA), is believed to exert its effect by binding to the α₂ò subunit of voltage-gated calcium channels on primary afferent neurons with the central nervous system (CNS). Gabapentin and the structurally similar medication

pregabalin are often labeled as “gabapentinoids,” with pregabalin binding to the same receptors as gabapentin, but with higher binding affinity. Both are believed to bind presynaptically to modulate calcium influx at the nerve terminals, thus inhibiting the release of excitatory neurotransmitters. Gabapentin is entirely excreted through the kidney and pregabalin predominately this way with almost no direct drug–drug interactions (PubChem, 2020b,d).

Oral gabapentin is an FDA-approved anticonvulsant. Gabapentin is commonly used orally, in part because it is relatively safe with few clinically relevant drug interactions. Most common side effects are somnolence, dizziness, and peripheral edema. Occasionally significant peripheral edema can lead to discontinuation, with complete reversal of symptoms with cessation of the medication (PubChem, 2020b).

In the United States, gabapentin is labeled as an adjunctive therapy in the treatment of focal seizures (with or without generalization), and it is FDA approved for the management of postherpetic neuralgia. Pregabalin has also labeled indications for fibromyalgia, painful diabetic neuropathy, and neuropathic pain associated with spinal cord injury (PubChem, 2020d). Studies and guidelines support the use of gabapentin in these conditions; for example, it is used as a first- or second-line therapy to treat painful diabetic neuropathy (Attal, 2010; Bril, 2011; Finnerup, 2013; NICE, 2013).

As outlined in the section below, well-designed clinical studies focusing on the safety and effectiveness on topical gabapentin are not available; therefore, definitive conclusions on its safety and effectiveness in this dosage form are not possible. The best single-drug studies are available for vulvodynia where case series suggested benefit, but a recent blinded and placebo-controlled crossover study failed to show benefit for gabapentin.¹⁴ Studies of urine concentration in patients exposed to topical gabapentin suggest very low to undetectable absorption rates. In vitro studies using Franz diffusion cells document absorption rates that vary greatly by excipient used. Information on safety is limited. Major adverse effects were only reported for single cases, where topical compounded creams also included other drugs, with toxicity not likely caused by gabapentin. Gabapentin in oral preparation is considered fairly safe, thus the risk of side effects from topical gabapentin appears low, even if there is some degree of systemic absorption. See Box 6-4 for a summary of research findings. (See the Multiagent Compounded Topical Pain Creams section of this chapter for further discussion of creams with gabapentin in combination with other APIs.)

___________________

¹⁴ Vulvodynia is outside the scope of this review as the vulva is not representative of normal skin. However, several studies have shown benefit of gabapentin in treating vulvodynia (e.g., Boardman et al., 2008; University of Rochester and Mae Stone Goode Foundation, 2015).

Effectiveness

Randomized controlled trials

There are no relevant RCTs that examine the single-drug use of gabapentin in compounded topical creams to treat pain when applied to intact skin. The best evidence for single-drug topical gabapentin is from studies in vulvodynia, which was outside the scope of the committee’s research focus.

Case reports and case series

Hiom et al. (2014) published a case series in patients with mixed neuropathic pain and reported benefit in 20 of 23 patients when treated with gabapentin 6 percent topically applied three times per day. Eleven of 23 patients reported at least 30 percent pain reduction, and 2 of 3 patients with postherpetic neuralgia reported reduction in pain by 60 percent and 57 percent.

Vadaurri (2008) provides a narrative review about topical medication in palliative care that includes two patients treated with a topical gabapentin cream, with gabapentin 5 percent compounded in pluronic lecithin organogel. Two patients described received gabapentin 5 percent in pluronic lecithin organogel with benefit reported with marked reduction in numeric pain scores within days. One patient had severe right leg pain (with no clear diagnosis provided) and used the cream as 2 mL twice daily. The other patient had painful neuropathy following chemotherapy and was started on 1 mL to each hand and foot twice daily.

Many case reports and case series include gabapentin as part of multidrug topical preparations that are summarized in the Multiagent Compounded Topical Pain Creams section of this chapter.

Preclinical studies

Animal studies evaluating the peripheral action of gabapentin include Ortega-Varela et al. (2007), who studied subcutaneous injection in the rat model to treat formalin-induced injury.

Dermal Penetration/Bioavailability

Clinical studies

Glinn et al. (2017) reported on urine drug concentrations in a large number of patients receiving gabapentin as oral and/or topical medication. Glinn et al. analyzed 29,139 urine specimens with gabapentin prescribed in an oral formulation with urine detection positive in 34 percent (9,948) positive. In 85 percent of positive results from patients prescribed oral gabapentin the concentration was above 10,000 ng/mL. In the 187 patients with gabapentin prescribed in the form of a cream, gabapentin was detected in 22 (12 percent). Of the positive cases, all but two had gabapentin present below 1,500 ng/mL. Two specimens showed gabapentin present greater than 10,000 ng/mL, likely owing to receiving gabapentin orally as well as topically, but that information was not included in the patient demographic information. Glinn et al. concluded that gabapentin topically results in very low detection in urine compared to oral administration. For the positive oral specimens, the mean urine concentration level was greater than 10,000 ng/mL. In contrast, for the positive topical specimens, the mean concentration was reported as 261 ng/mL.

The authors also analyzed amitriptyline, ketamine, and cyclobenzaprine urine drug concentration levels. They noted that in contrast to gabapentin and amitriptyline, ketamine and cyclobenzaprine were more readily detectable and with higher concentrations (as percentage of levels with oral dosing). Excipients varied across studies and absorption rates varied greatly between excipients.

Preclinical studies

Wang et al. (2013) reported on the percutaneous absorption of gabapentin in Franz diffusion cells with human trunk skin, as well as other model drugs. They compounded their preparations in two different bases, the “Versatile” cream and a reference base. They reported on the pharmacokinetic profile for each drug for up to 48 hours. For gabapentin, the peak absorption (flux rate) peaked at 4 hours and was sustained at a lower and gradually diminishing rate thereafter. As with the other drugs, the Versatile base formulation provided enhanced absorption compared to the reference.

Martin et al. (2017) reported on a variety of topical gabapentin preparations that included commercially available proprietary bases and enhancers. They report that topical gabapentin delivery varies greatly. They showed that gabapentin 6 percent with Carbopol hydrogels containing dimethyl sulfoxide (DMSO) or 70 percent ethanol, and a compounded gabapentin 10 percent in Lipoderm formulation, were able to facilitate permeation of the gabapentin molecule across human skin.

Bassani and Banov (2016) studied the skin absorption of two mixed compounded topical creams using cadaver skin by the Franz Finite Dose Model. The transdermal creams contained ketamine 5 percent, gabapentin 10 percent, clonidine 0.2 percent, and baclofen 2 percent. One cream was compounded using Lipoderm and the second was compounded using Lipoderm ActiveMax. All drugs were shown to penetrate into and through human cadaver trunk skin.

Bryson et al. (2014) reported an in vitro study of the transdermal penetration of gabapentin in Franz diffusion cells and an in vivo preclinical rodent study. The penetration of gabapentin was studied in Franz diffusion cells using porcine skin, with gabapentin compounded as 5 percent in two commonly used commercial bases (Lipobase, Lipoderm) and a standard poloxamer lecithin organogel. The penetration and retention of gabapentin was dependent on base and overall poor, with only microgram levels of gabapentin penetrating into the porcine skin over a 24-hour period. With the poloxamer lecithin organogel base, three to four times as much gabapentin was retained in the skin compared to Lipobase and Lipoderm bases.

For the in vitro study, the authors used gabapentin 5 percent and 1 percent that were compounded using Lipoderm as the base and tested in an in vivo preclinical rodent formalin pain model (hind paw). They documented reduced nociceptive behavior of paw flinches with subcutaneous injection of gabapentin and with 5 percent topical gabapentin gel applied to the ipsilateral hind paw prior to formalin administration. The authors noted evidence of systemic effect of gabapentin applied topically, as well as the local effect, as was shown by reduced formalin-induced behavior with gabapentin applied ipsilateral versus no effect when applied to the contralateral limb. (Note that studies by Bryson et al. had pharmaceutical industry support.)

Le Uyen et al. (2018) reported in vitro analysis of penetration across porcine skin for different gabapentin preparations. In their study, they compared gabapentin encapsulated elastic liposomes with compounded gabapentin-based pluronic lecithin organogel regarding their efficiency in transdermal delivery of gabapentin. Gabapentin released slowly from liposomes over 12 hours while it was rapidly released from pluronic lecithin organogel within 4 hours. After 24 hours, liposomes significantly accelerated the percutaneous penetration of gabapentin through the porcine skin leading to

higher cumulative drug concentrations (~98 percent of drug permeated) as compared to pluronic lecithin organogel (~55 percent of drug permeated).

Safety and Adverse Events

Studies listed for gabapentin as a single-drug compounded cream do not report side effects from the topical medication.

Case reports and case series

Pomerleau et al. (2014) describes a 23-year-old man with accidental overdose after rubbing a compounded topical with multiple drugs over his entire body. In this single case, safety concerns owing to high absorption were likely caused by the clonidine component of the compounded drug (documented by very high serum levels), rather than gabapentin.

In vitro studies

Al-Musawi et al. (2017), as discussed in the carbamazepine section above, conducted a toxicity analysis of commonly used topical drugs (i.e., amitriptyline, carbamazepine, gabapentin) commonly used for the treatment of neuropathic orofacial pain. After examining the effects on keratinocytes in cell culture, the authors reported that carbamazepine was cytotoxic to skin and oral keratinocytes, demonstrating significant decrease in cellular viability and cell counts, at high concentrations (1.7 mM) and with long exposure (2 hours). The authors also reported that topical gabapentin was only minimally cytotoxic to skin or oral keratinocytes at high concentrations (5.54 mM) and long exposure (24 hours). Importantly, topical amitriptyline was reported as cytotoxic to skin and oral keratinocytes at both short (30 minutes) and long (24 hours) exposure times and at low (200 µM) and high concentrations (1.8 mM).

From the FDA-approved label, serious adverse effects from oral administration of this API include Stevens-Johnson syndrome. See Table G-1 in Appendix G.

Anticonvulsants: Topiramate

Summary

Oral topiramate is FDA approved as an anticonvulsant, though it is not structurally related to existing anticonvulsants (Privitera, 1997). Topiramate has multiple pharmacological mechanisms of action, which led to its approval to treat migraine prophylaxis. The most notable mechanisms of action thought to be responsible for its reported effectiveness treating these conditions are its role as a sodium channel blocker, GABA-A receptor enhancer, and calcium channel inhibitor (Wiffen et al., 2013). Common

adverse drug reactions to oral dosage of topiramate include dizziness, cognitive disturbance, weight loss, and nausea (Sommer and Fenn, 2010).

No relevant clinical studies regarding the use of compounded topical topiramate were found, so no conclusions regarding its safety or effectiveness can be formed. From the FDA-approved label, serious adverse effects from oral administration of this API include erythema multiforme, Stevens-Johnson syndrome, and toxic epidermal necrolysis. See Table G-1 in Appendix G. Studies are needed as there are no data to support its use. See Box 6-5 for a summary of research findings.

Antispasmodic: Baclofen

Summary

Oral and topical baclofen is FDA approved to treat reversible spasticity. Baclofen is thought to be primarily a centrally acting drug, is a structural analogue of GABA, and an agonist of GABA_B receptors in the CNS with the greatest density in the dorsal horn of the spinal cord. By binding to these receptors, monosynaptic and polysynaptic spinal reflexes are inhibited, thus, reducing muscle tone and especially flexor spasms. Presynaptic binding reduces calcium influx while postsynaptic binding increases potassium efflux from the Ia afferent terminal, resulting in hyperpolarization and interruption of the action potential transmission (Elovic, 2001).

Evidence from animal studies suggests that a peripheral antinociceptive effect of baclofen may occur by a mechanism related to specific types of potassium channels (Reis and Durarte, 2006). See Box 6-6 for a summary of research findings.

In addition to reviewing studies where baclofen alone was analyzed, the committee also included studies of baclofen used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized controlled trials

There are no relevant RCTs that examine the single-drug use of baclofen in compounded topical creams to treat pain when applied to intact skin.

Case reports

Hesselink and Kopsky (2013) describe a 65-year-old woman with acromegaly and neuropathic pain in her legs. The patient reported pain as 9 on a 0–10 NRS and was initially treated with a compounded analgesic cream, based on baclofen 5 percent and told to apply 1 gram maximally three times per day. After 2 weeks, she reported reduction of her pain by more than 95 percent based on one application daily, but she complained of numbness in her legs. The dose was reduced to baclofen 2 percent, and she reported benefits with use only 1–3 times per week for 6 months without side effects.

Kopsky and Hesselink (2017) describe a 54-year-old woman with neuropathic pain in both feet reported as 8 out of 10 on the NRS following chemotherapy treatment despite being prescribed oral gabapentin 2,000 mg daily and oxycodone 20 to 30 mg daily. Pain was reduced to 3 by a test application of compounded topical baclofen 5 percent cream, which was said to be a greater reduction than with compounded clonidine 0.2 percent

cream or compounded lidocaine 3 percent combined with isosorbide dinitrate 0.4 percent cream. Allodynia was still present but eliminated with ketamine 10 percent cream. Pain level increased over time from 3 to 7 on the NRS, which led to further treatment.

Preclinical studies

Preclinical studies using animal pain models suggest that baclofen (2.5–10 mg kg^–1 doses) has antinociception activity in mice and rats when given systemically for a formalin test (Shafizadeh et al., 1997) and tail-flick test (Sabetkasai et al., 1999) or intrathecally for tail-flick and hot plate tests (Aran and Hammon, 1991), but there are no studies of topical application.

Dermal Penetration/Bioavailability

No relevant studies with baclofen alone were found in the literature; however, the committee included studies of baclofen used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Safety and Adverse Effects

No relevant studies with topical baclofen alone were found in the literature. From the FDA-approved label, serious adverse effects from oral administration of this API include gastrointestinal bleeding. See Table G-1 in Appendix G.

Antispasmodic: Cyclobenzaprine

Summary

Cyclobenzaprine is a tricyclic amine salt with structural similarities to the tricyclic antidepressants such as amitriptyline. Although it has sedative effects and has been investigated as an antidepressant, when taken orally its central action also relaxes skeletal muscle and is FDA approved for relief of musculoskeletal pain and muscle spasms (Kroenke et al., 2009; McNeil Consumer Healthcare, 2013). Off-label use has also included chronic pain states such as fibromyalgia (Bryson et al., 2015). Cyclobenzaprine depresses noradrenergic and serotonergic descending pathways from the brain, thereby inhibiting alpha and gamma motor neurons in the ventral horn of the spinal cord (Bryson et al., 2015; McNeil, 2013). Its action is therefore primarily within the CNS at the brain stem, although some effect at the spinal cord level may contribute to muscle relaxation (McNeil Consumer Healthcare, 2013). Cyclobenzaprine is highly bound to blood proteins, metabolized by

the liver to glucuronides, and eliminated primarily via urinary excretion, with an effective half-life after oral ingestion of 18 hours (8 to 37 hours measured in 18 subjects) and a plasma clearance of 0.7 L/min (McNeil Consumer Healthcare, 2013; Pesko, 1998). Cyclobenzaprine is also excreted in bile and undergoes enterohepatic circulation (McNeil Consumer Healthcare, 2013).

Cyclobenzaprine relieves muscular spasms of local origin through its centrally inhibition action, but it is ineffective for spasm resulting from CNS disease. No clear evidence supports a peripheral mechanism of action. The available literature is unclear on whether topical treatment will result in absorption of a sufficient systemic dose to result in pain relief. See Box 6-7 for a summary of research findings.

In addition to reviewing studies where cyclobenzaprine alone was analyzed, the committee also included studies of cyclobenzaprine used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized controlled trials

There are no relevant RCTs that examine the single-drug use of cyclobenzaprine in compounded topical creams to treat pain when applied to intact skin.

Clinical studies

A retrospective cohort study indicated that cyclobenzaprine as part of a compounded topical pain cream with four other drugs was more effective than Volatren gel; however, as with the clinical trials, the individual effect of cyclobenzaprine cannot be assessed (Somberg and Molnar, 2015a).

Preclinical studies

The effect of topical cyclobenzaprine formulations (1 or 5 percent in Lipoderm base) on nociception was studied using male hamsters (six animals per group) injected in the hind paw with formalin to produce pain (Bryson et al., 2015). In addition to the two cyclobenzaprine topical treatment groups, a control group received topical Lipoderm base and another group received a subcutaneous injection of 10 µg/kg cyclobenzaprine. Animals were treated with 0.5 mL of topical cream to the ventral and dorsal hind paw or subcutaneous injection 30 minute before receiving the injection of formalin. Pain reaction was scored as the number of paw flicks or holding the paw up. Topical application of cyclobenzaprine at either concentration did not reduce pain-related behaviors compared to the vehicle control. Subcutaneous injection (location not specifically stated) did reduce pain-related behaviors compared to the vehicle control and topical pain cream groups. The mechanism of action for oral administration of cyclobenzaprine in treating musculoskeletal pain or spasms involves a primary action at the brain stem level and not locally. It appears to be unknown whether local targets exist for pain relief by this compound. Thus, the available literature is unclear on whether topical treatment will result in sufficient systemic dose to result in pain relief.

Dermal Penetration/Bioavailability

Clinical studies

Urinary samples were analyzed for drug constituents (including cyclobenzaprine) in patients prescribed a compounded topical cream or oral medications with the same compounds (Glinn et al., 2017). Patients were adults with the largest fraction (26 percent) for the oral group being ages 51–60 years, and the largest fraction for the topical pain cream group (26 percent) was ages 31–40 years old. Pain creams tested typically contained one or more of 6 percent gabapentin, 2 percent cyclobenzaprine, 5 percent ketamine, or 5 percent amitriptyline by weight, and also frequently contained ketoprofen, baclofen, clonidine, or lidocaine, which were not analyzed. Of the 74 subjects in the topical application group tested for cyclobenzaprine and its major metabolite, norcyclobenzaprine, in urine, only 7 were positive for the former and 6 for the latter. Higher percentages tested positive for the group prescribed oral cyclobenzaprine (884 positive for cyclobenzaprine and 867 for norcyclobenzaprine with n = 9,036), and the urinary concentrations were much higher. Detected mean urinary concentrations in the topical group were 54 ng/mL and 57 ng/mL for cyclobenzaprine and norcyclobenzaprine, respectively, compared to mean detected urinary concentrations in the oral group of 383 ng/mL and 272 ng/mL, respectively. Glinn et al. (2017) noted that the low percentages of detections may be the result of patients taking their pain medication

only when needed and not frequently enough to result in more detectable urinary concentrations.

Preclinical studies

Dermal penetration of cyclobenzaprine was studied in vitro using a Franz diffusion cell model with 5 percent cyclobenzaprine in three different bases: Lipoderm, Lipobase, and standard poloxamer lecithin organogel. Cyclobenzaprine in each of the three bases showed dermal penetration. For formulations 1 and 2, penetration was rapid within the first 30 minutes to 1 hour, reaching approximately 76 to 84 percent of maximum levels (30 to 40 µg/cm²) within 1 hour. For formulation 3, penetration was initially slower but reached a maximum of 600 µg/cm³ after 25 hours. “Modest levels” were retained in the skin (formulation 1: 28.94; formulation 2: 55.07; formulation 3: 42.70 µg/g skin).

In animal model studies, Bryson et al. (2015) postulated that topical cyclobenzaprine (at 1 percent and 5 percent concentration) was ineffective in relieving local pain unlike subcutaneous injection (10 mcg/kg) because topical application did not result in a sufficiently high systemic concentration.

Safety and Adverse Events

Case studies

Cyclobenzaprine was reported to have caused an allergic skin reaction in a 39-year-old White male with degenerative disc disease and cervical radiculopathy who experienced substantial pain relief from application (Turrentine et al., 2015). The topical pain cream contained 10 percent ketamine, 5 percent diclofenac, 2 percent baclofen, 1 percent bupivacaine, 2 percent cyclobenzaprine, 6 percent gabapentin, 3 percent ibuprofen, and 3 percent pentoxifylline in Lipoderm ActiveMax cream base. However, after several weeks of use, a pruritic rash occurred at the site of application. Discontinuation of the cream cleared the rash with return of the pain. Subsequent patch testing of the individual components and the mixture implicated cyclobenzaprine. Removal of cyclobenzaprine from the original mixture resolved the allergic reaction.

From the FDA-approved label, serious adverse effects from oral administration of this API include cardiac dysrhythmia, heart block, myocardial infarction, and syncope. See Table G-1 in Appendix G.

Antispasmodic: Orphenadrine

Summary

Orphenadrine is an FDA-approved (oral, injection) skeletal muscle relaxant that is often used to relieve pain caused by muscle injuries (e.g.,

sprains). Orphenadrine binds and inhibits both histamine H1 receptors and NMDA receptors. It exhibits anticholinergic effects and may have a relaxing effect on skeletal muscle spasms (by central antimuscarinic action). The most common side effects of oral dosages include drowsiness, dry mouth, confusion, and visual disturbances. Orphenadrine also has potential for abuse, and fatal overdoses have been reported (NIDDK, 2017). No data were found on the effectiveness or safety of orphenadrine used topically. See Box 6-8 for a summary of research findings.

Effectiveness

There are no relevant clinical studies that examine the single-drug use of orphenadrine in compounded topical creams to treat pain when applied to intact skin.

Dermal Penetration/Bioavailability

In vitro studies

Diffusion of orphenadrine across human cadaver skin was evaluated using the Franz diffusion cell (Wang and Black, 2013). A 10 percent cream was formulated in Versatile cream base and compared to a reference cream (not defined). After 48 hours, a total of 0.00477 mg (0.191 percent) in the Versatile cream and 0.00326 mg (0.130 percent) in the reference cream were measured in the receiver compartment.

Safety and Adverse Effects

No data were found on the safety of orphenadrine alone in topical pain creams. From the FDA-approved label, serious adverse effects from oral administration of this API include palpitations and tachyarrhythmia. See Table G-1 in Appendix G.

Calcium Channel Antagonist: Nifedipine

Summary

Nifedipine is a dihydropyridine calcium channel blocking agent. It works by inhibiting the influx of extracellular calcium into myocardial cells and into vascular smooth muscle cells. Its pharmacologic actions include vasodilation, and it is used primarily as an antihypertensive and for angina. The most common adverse events reported in clinical trials include peripheral edema, headache, dizziness, constipation, and flushing (Snider et al., 2008). Nifedipine does not have an FDA-approved indication to treat pain.

Topical nifedipine has been used in the treatment of anal fissures and has demonstrated evidence of relative safety and effectiveness (see Golfam et al., 2010; Khaledifar et al., 2015; Perrotti et al., 2002; Salem et al., 2018); however, for this report, the committee largely excludes discussions on wound care and mucosal membranes and cavities (e.g., mouth, eyes, vulva, vagina, anus, and nose). Related to the report’s focus, there is insufficient evidence on the effectiveness or safety of nifedipine used topically on intact skin for the relief of pain. In fact, the committee’s search identified only a single retrospective study, which demonstrated no difference between compounded topical formulations (with or without nifedipine) in predicting efficacy in treating pain. See Box 6-9 for a summary of research findings.

In addition to reviewing studies on nifedipine alone, the committee also included studies of nifedipine used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized controlled trials

There are no relevant RCTs that examine the single-drug use of nifedipine in compounded topical creams to treat pain when applied to intact skin.

Retrospective clinical studies

A retrospective review of data from patient charts examined the influence of 2 percent nifedipine in treating inflammatory disorders (Somberg and Holnar, 2015b). In this study, the authors

applied two formulations both containing ketamine 10 percent, baclofen 2 percent, gabapentin 6 percent, amitriptyline 4 percent, bupivacaine 2 percent, and clonidine 0.2 percent, but one cream also contained 2 percent nifedipine (n = 205) and the other cream did not (n = 78). Both creams contained inactive ingredients of pentoxifylline and Tranilast (listed as an antiallergic agent used for treatment of inflammatory disorders). The authors determined no differences between formulations (with or without nifedipine) in predicting efficacy. The location of patient charts were not specified, and the authors had no affiliation with a medical institution.

Dermal Penetration/Bioavailability

No data were found on the dermal penetration and bioavailability of nifedipine alone in topical pain creams.

Safety and Adverse Effects

In Somberg and Molnar (2015b), the formulation with 2 percent nifedipine produced an adverse effect rate of 3.8 percent, as compared to a rate of 5.7 percent for the other cream. Adverse effects were minor to moderate and included skin irritation, burning, rash, flushing, and three cases of unspecified minor adverse effects. However, the authors did not differentiate between the two treatment groups with respect to the adverse effects.

From the FDA-approved label, serious adverse effects from oral administration of this API include myocardial infarction, ventricular arrhythmia, suicidal thoughts, and kidney damage. See Table G-1 in Appendix G.

Cannabinoid: Cannabidiol

Summary

Medicinal use of marijuana (Cannabis sativa) can be traced back 5,000 years for treatment of cramps and pain in China (Zou and Kumar, 2018). Of the more than 100 phytocannabinoids in Cannabis, cannabidiol (CBD) has attracted recent medicinal attention because of its antiepileptic,¹⁵ antinociceptive, anti-inflammatory, and antifibrotic drug properties, as well as its low intoxicating effects, unlike Δ⁹-tetrahydrocannabinol (THC)¹⁶ (Bruni et al., 2018; NASEM, 2017; VanDolah et al., 2019). Based on a few case and experimental animal studies, topically applied CBD oil appears to be absorbed locally, regionally, and systemically to reduce pain, inflammation, and fibrosis. Possible safety concerns from systemic absorption by other routes include CNS depression, liver enzyme elevation, and suicidal thoughts and actions. RCTs are needed to establish efficacy and safety for topical use of CBD in humans. Variable concentrations of THC in CBD oil is concerning (VanDolah et al., 2019). See Box 6-10 for a summary of research findings.

Effectiveness

Randomized controlled trials

There are no relevant clinical studies that examine the single-drug use of CBD in compounded topical creams to treat pain when applied to intact skin.¹⁷

A transdermal gel for regional and systemic delivery of CBD (Zynerba Pharmaceuticals) is in clinical development for treatment of epilepsy, developmental and epileptic encephalopathy, fragile-X syndrome, and osteoarthritis (Bruni et al., 2018).¹⁸ The manufacturer’s website notes that this product is currently experimental and not yet approved by FDA.

Case reports

Chelliah et al. (2018) reported on a case series of three children (6 months old, 3 years old, and 10 years old) using topical CBD alone

___________________

¹⁵ CBD has been approved by FDA as the active ingredient in an oral solution (EPIDIOLEX) for the treatment of rare forms of severe epilepsy, along with other medications (FDA, 2018; GW Biosciences, 2018). The mechanism of action for seizure reduction, however, is unknown and does not appear to involve the cannabinoid receptors (GW Biosciences, 2018). Currently, CBD is not FDA approved to treat pain.

¹⁶ THC is present at variable low levels in CBD, typically < 0.3 percent (VanDolah et al., 2019).

¹⁷ Preclinical and pilot studies using CBD and hemp oils have been conducted for treatment of inflammation and pain (VanDolah et al., 2019).

¹⁸ Compounded drugs prepared for investigational new drug trials are subject to current good manufacturing practice requirements under section 501(a)(2)(B) of the Federal Food, Drug, and Cosmetic Act.

or in combination with emu oil to treat painful lesions from epidermolysis bullosa (authors from the Stanford University Medical Center). Information on treatment and outcome appears to have been collected from parents with little specifics on amount or frequency applied. The topical CBD oil was reported to promote faster wound healing, less blistering, and reduction in pain for all three children, and one was weaned off oral opioid analgesics.

In a case series in the Netherlands, one case included topical CBD application in addition to sublingual CBD oil (20 mg/mL) with THC (13 mg/mL) to treat epidermolysis bullosa (Schraeder et al., 2019). A 64-year-old woman was started at sublingual doses of 0.5 mg CBD and 0.325 THC four times daily, which was increased stepwise to 2.5 mg CBD and 1.625 mg THC four times daily. By 3 months, she was able to wean off oxycodone except for dressing changes, and for the next 2 years she replaced topical morphine with 1 mg topical CBD and 0.65 mg THC applied daily to her painful heel and was weaned off amitriptyline. The CBD/THC oil applications also resulted in a moderate reduction in pruritus, with increased appetite as the only side effect.

Preclinical studies

Del Rio et al. (2018) administered the CBD aminoquinone derivative (VCE-004.3) by intraperitoneal injection or topically in a mouse model of fibrotic disease. After induction of subcutaneous fibrosis, CBD derivative was administered topically (250 µM; formulated with 7:3 polypropylene glycol:ethanol) and by injection (20 mg/kg) for 3 weeks

(location of fibrosis of topical application not specified). CBD reduced expression of fibrotic genes to below vehicle control levels, and was as effective as injected CBD in reducing skin fibrosis, although neither completely restored the decrease in the adipose tissue layer.

Hammell et al. (2016) applied CBD hydroalcoholic gel (1 or 10 percent CBD) to the shaved backs of rats after inducing arthritis of the knee by injections of Freund’s adjuvant. Of the applied CBD doses (0.6, 3.1, 6.2, or 62.3 mg/day for 4 days), the two higher doses equally reduced swelling (including immune markers of inflammation) and pain indicators in the knee by about 50 percent of vehicle control.

Lodzki et al. (2003) examined the anti-inflammatory effect of 100 mg CBD gel (3 percent weight concentration [w/w] CBD and 40 percent ethanol in a carbomer gel) applied to the abdomen and hip area of mice (abdomen and hip area) under an occluded patch for 19 hours before carrageenan injection in a paw. Paw thickness after this injection was about 0.1 mm in CBD treated mice compared to about 0.3 mm in injected controls without CBD.

Dermal Penetration/Bioavailability

Preclinical studies

Reported advantages of topically administered CBD are the potential for greater systemic absorption of this highly lipophilic drug with poor bioavailability by the oral route, while also avoiding extensive first-pass metabolism of ingested CBD (Bruni et al., 2018).¹⁹ CBD in a polyethylene gel was reported to readily cross the stratum corneum in rats, with some accumulation likely in the skin because this lipophilic compound cannot as readily cross the more aqueous dermis layer (Paudel et al., 2010). Lipophilic properties are also expected to result in dermal penetration through skin follicles with accumulation in sebaceous glands (Bruni et al., 2018).

Animal studies indicate the ability of topical CBD to reach systemic circulation. Hammell et al. (2016) reported CBD plasma concentrations after 4 days of topical application to the shaved backs of rats at doses of 0.62, 3.1, 6.2, and 62.3 mg/day were 3.8 ± 1.4, 17.5 ± 4.4, 33.3 ± 9.7, and 1,629.9 ± 379 ng/mL, respectively. Application of 100 mg of CBD to the abdomen and hip area of mice resulted in plasma concentrations of 0.68 to 1.07 µg/mL in the first 24 hours that afterward stabilized to 0.67 µg/mL for the duration of the study (72 hours) (Lodzki et al., 2003). After 24 hours, a reservoir of CBD was measured in abdominal (110 µg/cm²) and hip (37 µg/cm²) skin and in abdominal muscle (11.5 µg/g). Topical application of

___________________

¹⁹ Ingested CBD is metabolized mainly by the cytochrome P450 (CYP) 2C19 and CYP3A4, and UGT1A7, UGT1A9, and UGT2B7 isoenzymes primarily by the liver (GW Biosciences, 2018; Ujvary and Hanus, 2016).

CBD to the backs of guinea pigs under occluded patch resulted in a steady-state plasma concentration of 6.3 ng/mL beginning at 15.5 hours (Paudel et al., 2010). Plasma concentrations began to decline 6 hours after the patch was removed after 48 hours of application. Addition of a penetration enhancer increased plasma concentrations by 3.7 times.

Safety and Adverse Events

Preclinical studies

Hammell et al. (2016) reported that none of the CBD systemic doses in rats (up to 1.9 ± 379 ng/mL in plasma from topical application of 62.3 mg/day for 4 days) affected their behavior using the open-field exploratory behavior test.

Based on individual clinical response and tolerability, the maximum recommended maintenance dosage of oral administration of CBD is 10 mg/kg twice daily (20 mg/kg/day) (GW Biosciences, 2018). The most common side effects noted in clinical trials of this oral drug were sleepiness, decreased appetite, diarrhea, increased liver transaminase levels, tiredness and weakness, rash (hypersensitivity), insomnia, and infections (GW Biosciences, 2018). The approved drug label also notes nausea, vomiting, and fever and warns of suicidal thoughts or actions (about 1 in 500 people); and that adjustments of doses are necessary in patients with moderate or severe hepatic impairment because of an increase in exposure to CBD (2.5 to 5.2 times higher area under the curve) (GW Biosciences, 2018). Effects related to the gastrointestinal tract are likely not relevant for topical administration. However, variable amounts of THC (with psychogenic effects) in unregulated CBD oils is concerning for patient safety (VanDolah et al., 2019).

In terms of potential drug–drug interactions, according to the clinical trials of oral CBD, the effects of CBD on sleepiness may be increased by other drugs (e.g., clobazam) or alcohol (GW Biosciences, 2018). Coadministration of CBD with drugs or substances that affect key metabolic enzymes (CYP3A4 or CYP2C19, and possibly UGT1A7, UGT1A9, and UGT2B7) may affect its efficacy or potential for adverse effects. Moderate to strong inhibitors of these enzymes (e.g., valproate, keloconazole) will increase CBD plasma concentrations and increase the risk of adverse effects (e.g., liver transaminase elevations) (GW Biosciences, 2018). Alternatively, effectiveness may be reduced by coadministration of certain antiepilepsy drugs (e.g., carbamazepine, topiramate, phenytoin) or the antibiotic rifampin, which induces these enzymes and increases CBD metabolism (Alsherbiny and Li, 2019; GW Biosciences, 2018).

CBD can also alter the toxicity or efficacy of other drugs through inhibition of certain enzymes. For example, increases in the plasma concentration of diazepam and the active metabolite of clobazam have been reported

with coadministration of EPIDIOLEX (GW Biosciences, 2018). Studies of patients with treatment-resistant epilepsy have reported that coadministration of CBD altered serum levels of topiramate, rufinamide, clobazam, eslicarbazepine, and zonisamide (Alsherbiny and Li, 2019).

In general, however, topically administered CBD, which would be associated with less first-pass metabolism and lower peak plasma levels than by the oral route, may be less likely to have such metabolic enzyme interactions with other drugs. From the FDA-approved label, serious adverse effects from oral administration of this API include increased suicidal thoughts and increased liver enzymes. See Table G-1 in Appendix G.

Dissociative Anesthetic and NMDA Receptor Antagonist: Ketamine

Summary

Ketamine is an FDA-approved anesthetic agent (intravenous, intramuscular) that has dissociative properties. In addition to anesthetic properties it also has analgesic, anti-inflammatory, and antidepressant properties. It works by blocking the NMDA receptor and reducing release of the excitatory neurotransmitter glutamine. Ketamine has a complicated pharmacologic profile and also has actions on AMPA, GABA_A, muscarinic, nicotinic, mu and kappa opiate, dopamine, and various ion channels, among others (Jonkman et al., 2017; Vadivelu et al., 2016).

While its systemic effects are well demonstrated, there is no evidence that topical ketamine is superior to placebo for various neuropathic pain conditions in concentrations from 1 to 10 percent, or that it reduces allodynia or hyperalgesia. There are also no data about the pharmacokinetic properties of ketamine following dermal application.

The majority of pharmaceutical preparations of ketamine are a racemic mixture. The S(+) isomer (esketamine) is available as a nasal spray to treat depression. S(+) ketamine is the active enantiomer. Ketamine is metabolized in the liver. The primary active metabolite is norketamine, which has approximately 20–30 percent of the analgesic activity of parent ketamine (Mion et al., 2013). Ketamine has a half-life of approximately 2–3 hours, but it has an active metabolite, norketamine, that has a half-life up to 12 hours (Quibell et al., 2015). Ketamine is primarily excreted in the urine; however, less than 5 percent is excreted unchanged (Zanos et al., 2018). See Box 6-11 for a summary of research findings.

Overall, the committee reviewed six RCTs: two were assessed as a low risk of bias (Finch et al., 2009; Lynch et al., 2003) and four were assessed as a high risk of bias (de Barros et al., 2012; Lynch et al., 2005b; Mahoney et al., 2012; Pöyhiä and Pöyhiä and Vainio, 2006). (See Appendix B for more details.) In addition to reviewing studies where ketamine alone was analyzed, the

committee also included studies of ketamine used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized clinical trials

Six RCTs evaluated ketamine topically for neuropathic pain including postherpetic neuralgia, complex regional pain syndrome (CRPS) and diabetic peripheral neuropathy. Ketamine did not reduce pain any better than the comparator or placebo treatment. In Finch (2009), 20 consecutive patients with CRPS (type I or type II) who attended a small private pain center in Perth, Australia, were studied to determine whether ketamine had any sensory effect, particularly on allodynia.²⁰ Ketamine 10 percent in pluronic lecithin organogel (referred to as ketamine cream by the Professional Compounding Centers of America) was studied compared to the vehicle alone. Patient and investigators were blinded to the products. An investigator applied 0.5 mL cream to the most hyperalgesic

___________________

²⁰ Risk-of-bias assessment by the committee: Low (see Appendix B for more details).

dorsal aspect of the symptomatic limb (hand or foot) with 0.5 mL of the other applied to the healthy limb; sensory testing was conducted on each side of the forehead. This process was repeated with a reversal of the cream applied with at least 1 week between assessments.

The investigators conducted sensory testing before and 30 minutes after application of cream that included threshold to light touch, pressure–pain, punctate stimulation, light brushing, and thermal stimuli. Venous blood was obtained 1 hour after application in the first 10 patients and analyzed for ketamine and norketamine by high-performance liquid chromatography. Pain or touch threshold did not change after treatment with ketamine. However, ketamine did reduce allodynia in the symptomatic limb as well as pain invoked by pricking the skin in both the symptomatic and healthy limb. Allodynia and hyperanalgesia to various experimental stimuli was reported in the ipsilateral forehead, which in some cases was lessened with ketamine treatment of the symptomatic limb. Ketamine and its metabolite were not detected at a threshold of 0.7 mcg/L ketamine and 0.5 mcg/L norketamine.

In Mahoney (2012), ketamine 5 percent cream versus placebo (vehicle composed of primarily Aquaphor gel) was studied in 27 patients for the treatment of painful diabetic neuropathy in patients with either type 1 or type 2 diabetes.²¹ Methods were poorly described. Patients were included if they had at least a score of 8 on the Michigan Neuropathy Screening Instrument or a score of 3 on the physical examination instrument. The physical examination instrument was not described. Patients were randomly divided into the treatment or placebo arm, although the exact randomization process was not described. A compounding pharmacy prepared the ketamine cream and placebo and blinded the products to the investigators and patients. Participants applied 1 mL of cream three times per day to the entire foot distal to the ankle for a 1-month period of time. Seven pain characteristics (intensity, sharpness, hot, cold, dull, sensitive, and itchy) were assessed prior to study and after the study was completed using an 11-point Likert scale. Twenty-seven patients were enrolled, but 10 (37.0 percent) dropped out at what appears to be prior to the start of treatment and no reasons were stated for the withdrawal. Pain scores for all seven pain characteristics were reduced in both the ketamine and placebo group.

de Barros (2012) provides a short communication that describes a randomized, double-blind, placebo-controlled crossover study in 12 patients that investigated the role of s-ketamine 1 percent ointment in the treatment of postherpetic neuralgia.²² Details about the methods were limited. Twelve patients were enrolled in the study and applied the ointment to the site of

___________________

²¹ Risk-of-bias assessment by the committee: High (see Appendix B for more details).

²² Risk-of-bias assessment by the committee: High (see Appendix B for more details).

pain four times per day for 15 days, followed by a 7-day washout and then a second treatment exactly as the first treatment. Patients reported Numerical Verbal Scale pain scores (0–10). There was a significant reduction in pain scores for both groups over time. However, there was no difference between s-ketamine and placebo at reducing pain scores. Several patients experienced adverse skin reactions.

Pöyhiä and Vainio (2006) evaluated the ability of ketamine applied topically to reduce capsaicin-evoked hyperalgesia in healthy volunteers.²³ In one study nine healthy volunteers were assigned ketamine 5 percent versus placebo gel 1 mL 10 minutes prior to injection of capsaicin intradermally to the left forearm. Treatments were applied in a randomized, double-blind, and crossover manner with three test periods conducted with a 1-week interval between each. The testing included ketamine gel on the left forearm and placebo on the right; placebo on the left and ketamine on the right; or placebo on both sides assigned in a randomized fashion. Ketamine did not affect spontaneous burning pain but the intensity of the hyperalgesia was significantly reduced when ketamine was applied to the either forearm, suggesting a systemic effect.

Lynch et al. (2003, 2005b) evaluated the effects of topical ketamine, amitriptyline, or a combination of the two in the treatment of neuropathic pain. To minimize redundancy within the chapter, a review of these studies is described in the Multiagent Compounded Topical Pain Creams section below.

Clinical studies

In Rabi et al. (2016), ketamine 10 percent in Lipoderm was studied in five subjects with a spinal cord injury and neuropathic pain at an outpatient rehabilitation hospital. Subjects applied up to 4 grams to the site of maximum pain every 8 hours for 2 weeks. The patients’ reduction in pain scores from baseline was as follows: 63 percent, 43 percent, 25 percent, 14 percent, and 60 percent. No subject was able to reduce oral pain medications. No adverse effects were noted. Patients were noted to be very satisfied (n = 1), somewhat satisfied (n = 2), indifferent (n = 1), and somewhat unsatisfied (n = 1).

Retrospective clinical studies

Durham (2018) provided a retrospective chart review of 16 patients 18 years of age and older with complex regional pain syndrome who were treated with topical ketamine between May 2006 and April 2013 at an academic medical center specialty pain clinic. The authors state that nine different compounds were used by the patients, although the manuscript only lists eight including one product with just ketamine 6 percent alone. The other seven compounds used in the study

___________________

²³ Risk-of-bias assessment by the committee: High (see Appendix B for more details).

included ketamine 10 percent with 1–4 additional ingredients that included gabapentin (6–10 percent), ketoprofen 10 percent, lidocaine (3–5 percent), clonidine 0.2 percent, and baclofen 1 percent. Improvement was noted by eight subjects; seven subjects reported worsening of pain.

Quan (2003) provided a short communication where authors report on 23 patients with postherpetic neuralgia that were treated with a topical ketamine gel preparation of 0.5 percent between 1994 and 2002. There was some improvement: from severe to mild, n = 8; severe to moderate, n = 7; and no improvement, n = 8.

Case reports

Ushida et al. (2002) described the effectiveness of topical ketamine in five patients with complex regional pain syndrome type I (CRPS I) and in two patients with type II (CRPS II). With a concentration of ketamine of 0.5 percent to 0.85 percent applied three times per day, a decrease in allodynia and hyperalgesia was noted (measured by the VAS in four patients with acute early dystrophic stage of CRPS I). No pain relief was observed in the one patient with chronic atrophic stage CRPS I and in both patients with CRPS II.

Skavinski (2019) reported the use of ketamine gel to treat severe pain associated with decubitus ulcers on the heels of a 54-year-old woman with a complicated medical history and course of therapy, admitted to a palliative care service. Ketamine 5 percent gel did not work, but increasing the ketamine concentration to 10 percent and adding lidocaine (2 percent gel) resulted in a decrease in opioid requirement. Ultimately the patient was switched to a topical product with ketamine 15 percent alone with limited data on success.

In Hesselink and Kopsky (2013), the authors describe a case of a 72-year-old woman with severe pain, swelling, and a poor quality of life, diagnosed with CRPS I. She was treated with ketamine 10 percent cream and palmitoylethanolamide capsules. The swelling improved, and the patient was more mobile at 1 month and continued to improve at 2 months. It is not clear which of the two treatments contributed to her improvement.

In Gammaitoni et al. (2000), the authors describe five patients that applied doses ranging from 0.13 to 0.37 mg/kg of ketamine gel for neuropathic pain. Of the three patients with regional sympathetic dystrophy, two had no response and one patient noted a reduction in pain of 55 to 60 percent. One patient with postherpetic neuralgia reported a 63 percent reduction in pain, and one patient with a postlaminectomy syndrome radiculopathy described between 53 and 100 percent reduction in pain depending on the site.

Dermal Penetration/Bioavailability

In vitro studies

Percutaneous absorption of ketamine in Lipoderm and Lipoderm ActiveMax (Bassani and Banov, 2016) and Versatile compared to a reference gel (Wang and Black, 2013) was studied using human cadaver skin in Franz diffusion cell chambers. Ketamine penetrated into and through human cadaver skin in all cases. A few of the RCTs described above were able to detect ketamine and metabolite in some participants.

Safety and Adverse Events

Ketamine, when administered systemically, has the potential to cause adverse effects because of its actions on numerous receptors. Potential adverse effects include cognitive impairment, emergence reactions, elevated heart rate and blood pressure, hypertonic muscle movements, and nausea and vomiting.

Case reports

In a case report, a 35-year-old man who applied excessive amounts of a topical pain cream containing ketamine, baclofen, amitriptyline, lidocaine, and ketoprofen presented to the hospital unresponsive after a presumed seizure. He required endotracheal intubation and mechanical ventilation for 2.5 days. Ketamine was detected in his central spinal fluid by gas chromatography-mass spectrometry (Sigillito et al., 2003). In a second case (Cardis and Pasieca, 2016), a man in his 80s with Parkinson’s disease presented with altered mental status after applying a topical pain cream containing ketamine 10 percent, amitriptyline 5 percent, and lidocaine 5 percent to most of his upper body the day prior to presentation. A stroke was ruled out, and after persistent delirium and altered levels of consciousness, the history of the topical pain cream use was elicited. Amitriptyline, lidocaine, and ketamine and metabolites were detected in urine by mass spectroscopy. Amitriptyline, lidocaine, and ketamine concentration was 2,360 ng/mL (normal level, 0 ng/mL) (Cardis and Pasieca, 2016).

From the FDA-approved label, serious adverse effects from oral administration of this API include bradyarrhythmia, cardiac dysrhythmia, and respiratory depression. See Table G-1 in Appendix G.

Local Anesthetic: Bupivacaine

Summary

Bupivacaine hydrochloride is an FDA-approved amide local anesthetic (injection). Local anesthetics block the generation and propagation or conduction of action potentials in axons by reversible inhibition of ion flux through the voltage-dependent sodium channel (Berde, 1993; Hospira,

2011). Amide local anesthetics such as bupivacaine are metabolized to inactive byproducts in the liver via conjugation with glucuronic acid, so diseases that result in decreased liver blood flow or enzyme dysfunction decrease clearance of bupivacaine. Generally, bupivacaine is considered a long-acting local anesthetic. It has high protein binding capacity (95 percent) and low fetal/maternal ratio (0.2 to 0.4). Bupivacaine’s half-life in adults is 2.7 hours and in neonates 8.1 hours (Fresenius Kabi, 2018).

Although outside the scope of the current report, RCTs indicate liquid bupivacaine is effective in providing local anesthesia for suturing of superficial skin lacerations. In addition, low levels have been shown to be absorbed after topical application to damaged skin, but no data were found on dermal penetration through intact skin (Alvi et al., 1998; Jellish et al., 2018). See Box 6-12 for a summary of research findings.

In addition to reviewing studies where bupivacaine alone was analyzed, the committee also included studies of bupivacaine used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized controlled trials

There are no relevant clinical studies that examine the single-drug use of bupivacaine in compounded topical creams to treat pain when applied to intact skin. However, although outside the scope of the current report, there are several randomized clinical trials that demonstrated potential effectiveness of compounded topical bupivacaine in treating pain in patients after receiving skin grafts for a variety of surgical

conditions (Butler et al., 1993) and in suturing skin lacerations (Eidelman et al., 2005; Keyes et al., 1998; Kuhn et al., 1996; Smith et al., 1996).

Dermal Penetration/Bioavailability

Randomized controlled trials

No clinical studies were found on the dermal absorption or bioavailability of topical bupivacaine through intact skin. However, although outside of the committee’s research scope, the committee identified two RCTs that examined the pharmacokinetic properties of topical bupivacaine in damaged skin (i.e., burn patients). These studies, Alvi et al. (1998) and Jellish et al. (2018), determined that low levels of buvipicaine are absorbed and present in serum concentrations after topical application to damaged skin.

Safety and Adverse Effects

Clinical and preclinical studies

The systemic toxic effects of local anesthetics such as bupivacaine primarily affect the cardiovascular system and the CNS. As described in Alvi et al. (1998), local anesthetics at low dose have beneficial cardiac antidysrhythmic effects. However, higher bupivacaine concentrations are associated with CNS toxicity (e.g., seizures), myocardial depression, and vasodilation. The intravenous dose in monkeys causing toxicity is 4.3 mg/kg, and seizures occur in humans after large doses of bupivacaine at serum levels greater than 4 mcg/mL. Toxic levels of bupivacaine particularly affect the cardiac conduction system, depressing cardiac conduction and excitability, which may lead to atrioventricular block, ventricular arrhythmias, and cardiac arrest, which is sometimes fatal (Pfizer, 2012). A recent clinical report suggests that cardiovascular changes are more likely to occur after unintended intravascular injection of bupivacaine, so incremental dosing and ultrasound guidance for precise drug delivery are recommended to minimize, or lipid emulsion to treat, such complications (Waldinger et al., 2020).

Given the half-life of bupivacaine in adults is 2.7 hours and in neonates 8.1 hours, it is not surprising that cardiac arrest in patients receiving bupivacaine requires resuscitation for several hours to give enough time for the drug to ultimately decrease in the heart so the conduction system can return to normal. Cardiac and CNS collapse, and even death, have been reported in otherwise healthy individuals receiving the highest concentration of bupivacaine (0.75 percent) given by injection, hence the FDA black box warning that lower doses be used for obstetrical anesthesia and precautions (e.g., divided doses) taken for any nerve blocks with bupivacaine where systemic absorption is possible (Pfizer, 2012). Topical administration of bupivacaine on mucous membranes or nonintact skin could enhance systemic absorption and potentially result in significant toxicity.

From the FDA-approved label, serious adverse effects from oral administration of this API include cardiac arrest and respiratory depression. See Table G-1 in Appendix G.

Local Anesthetic (Short-Acting): Lidocaine

Summary

Lidocaine is an FDA-approved local anesthetic for topical or rectal treatment of postherpetic neuralgia. It is also used off-label to treat diabetic neuropathy and acute pain. Lidocaine is a nonselective, voltage-gated sodium channel inhibitor, affecting both the generation and conduction of nerve impulses. It stabilizes nerve membranes, reducing ectopic activity in damaged afferent pain receptors. Other effects on keratinocytes and immune cells, or activation of irritant receptors (TRPV1 and TRPA1), may also contribute to the analgesic effect of topical lidocaine (Sawynok and Liu, 2014).

FDA-approved creams, gels, foam sprays, and solutions containing lidocaine are most often used for short-term analgesia, such as before painful medical procedures or to treat cuts, burns, and insect bites, but it may also be used as a patch in chronic conditions. The concentration of lidocaine in these formulations is usually around 2 percent to 5 percent w/w (FDA, 2019a). Its usual use is for injection for dental analgesia and minor surgery, or infiltration into wounds, but it can also provide surface anesthesia when applied topically, such as a medicated plaster, gel, or spray. Lidocaine is readily absorbed from mucous membranes and through damaged skin, and from injection sites, but absorption through intact skin is poor. To be clinically useful as a topical agent, lidocaine must be formulated with a carrier to facilitate transfer across the skin. These products contain high concentrations of lidocaine because it crosses intact skin poorly.

The committee elected to review the effectiveness, absorption, and safety of topical lidocaine because it is commonly mixed with other priority ingredients. Uniquely, the committee determined that lidocaine can be considered a positive control for other topical ingredients because there is an FDA-approved lidocaine ointment (5 percent). The committee was unable to obtain the original full FDA review of the product. As such, the committee relied on data from FDA-submitted resources and approval documents for the FDA-approved lidocaine patch. In addition to reviewing studies where lidocaine alone was analyzed, the committee also included studies of lidocaine used in combination with other APIs. See Box 6-13 for a summary of research findings. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

FDA-approval of lidocaine gel

Submitted resources provided by FDA (2019b) informed the committee that 5 percent lidocaine ointment was first FDA approved under NDA 008048 on May 17, 1951. According to the Federal Register Notice (Vol. 35, No. 172, p. 14020-1) published on September 3, 1970, 5 percent Xylocaine (lidocaine) ointment (NDA 008048), marketed by AstraZeneca Pharmaceutical Products, Inc., was determined to be effective under the Drug Efficacy Study Implementation (DESI) program.²⁴ Specifically, 5 percent topical Xylocaine ointment was found to be effective “for production of anesthesia of accessible mucous membranes of the oro-pharynx” and possibly effective for the “control of itching, burning, and other unpleasant symptoms due to abrasions, herpes zoster, eczema, and similar conditions; hemorrhoids, fissures, postoperative anorectal conditions, nipple soreness and for anesthesia of unbroken skin.” In a follow-up Federal Register Notice (Vol. 43, No. 206, p. 49570-2), published on October 24, 1978, 5 percent Xylocaine ointment (RX; NDA 008048), marketed by AstraZeneca Pharmaceutical Products, Inc., was found to be effective for two additional indications “for use as an anesthetic lubricant for endotracheal intubation” and “for the temporary relief of pain and itching associated with minor burns and abrasions of the skin.”

___________________

²⁴ For more information on FDA’s DESI program, see https://www.fda.gov/drugs/enforcementactivities-fda/drug-efficacy-study-implementation-desi (accessed March 8, 2020).

Other clinical studies

In a Cochrane evaluation, a total of 12 small studies (508 patients) of modest quality that tested topical lidocaine against topical placebo for a number of weeks were evaluated (Derry et al., 2014). One study also tested a cream containing amitriptyline, which is an antidepressant. The 508 patients in the studies had different types of neuropathic pain, with pain after herpes zoster infection the most common. Six studies enrolled participants with moderate or severe postherpetic neuralgia, and the remaining studies enrolled different, or mixed neuropathic pain conditions, including trigeminal neuralgia and postsurgical or posttraumatic neuralgia. Four different formulations were used: 5 percent medicated patch, 5 percent cream, 5 percent gel, and 8 percent spray. Most studies used a crossover design, and two used a parallel-group design.

All studies reviewed in this evaluation were categorized as third-tier, very low-quality evidence. Specifically, the authors stated that these studies contained a small number of participants who were considered very likely to be biased or used outcomes of limited clinical usefulness, or both. From this group, all but one study indicated that topical lidocaine was more effective than placebo at providing some measure of pain relief. From the group of studies, only one multiple-dose study reported primary outcome of participants with ≥ 50 percent or ≥ 30 percent pain intensity reduction (Meier et al., 2003).

Dermal Penetration/Bioavailability

Clinical and preclinical studies

There is insufficient evidence on dermal penetration/bioavailability of FDA-approved topical lidocaine gel (5 percent). Based on resources provided by FDA (2019a), it is unlikely that in vivo or in vitro penetration studies were performed when this drug was brought for approval in the 1950s.²⁵

In the topical application of Lidoderm (lidocaine patch, 5 percent), the systemic exposure of lidocaine is minimal. In healthy volunteers, the absorption of lidocaine after 12 hours of topical application of three lidocaine patches, which contain 2,100 mg of lidocaine, was 3 ± 2 percent of applied dose (Endo Pharmaceuticals, 2015). Lidocaine does not cross intact skin well, and when applied as a patch, with steady controlled release of the drug, the amount of lidocaine that penetrates is enough to cause analgesia, but not anesthesia (Hong et al., 2016). In an in vitro lidocaine patch permeation study through excised rat skin, the results exhibited large unequal, within lot, variation in the amount of permeation (FDA, 1998).

___________________

²⁵ It was difficult for FDA to obtain and appropriately review the original approval information from the early 1950s (FDA, 2019a).

Likewise, a clinical study that examined the release of lidocaine from the patch applied to the chest area of volunteers, resulted in significant variability within subject; variation occurred between subject, within lot, and between lots (FDA, 1998).

Safety and Adverse Events

Clinical studies

Systemic adverse reactions following appropriate use of a lidocaine patch is unlikely, owing to the small dose absorbed; however, in circumstances of systemic administration, lidocaine is a potent antiarrhythmic drug, and at high doses it can cause CNS disturbances, such as psychosis, seizures, and at high concentration even death. The amount of lidocaine reaching the systemic circulation when the patch is used is low (of the order of 3 percent), which is well below therapeutic antiarrhythmic concentrations or toxic concentrations in patients with normal cardiac, renal, and hepatic function. However, toxicity could be reached if patients inappropriately use several lidocaine patches at once. Lidocaine is metabolized in the liver and excreted by the kidneys, so caution is also required when an individual has severe cardiac, renal, or hepatic impairment.

In the Cochrane evaluation of 12 small studies (508 patients) there was no clear evidence of an effect of topical lidocaine on the incidence of adverse events or withdrawals; however, localized skin reactions to the patch or carrier in the formulation have occurred (Derry et al., 2014).

NMDA Receptor Antagonist: Memantine

Summary

Oral memantine is an FDA-approved NMDA receptor antagonist and is used orally to treat dementia associated with Alzheimer’s disease. It is not approved to treat pain, but it has been suggested to reduce peripheral nociceptive activity when administered systemically or intrathecally (Davidson and Carlton, 1998). Adverse effects occurring in patients receiving memantine in these clinical studies include dizziness, confusion, agitation, vertigo, vomiting, and hypertension (PubChem, 2020c). From the FDA-approved label, serious adverse effects from oral administration of this API include cerebrovascular accident, seizures, and kidney failure. See Table G-1 in Appendix G. No clinical evidence was found on the effectiveness or safety of memantine use in compounded topical pain creams. See Box 6-14 for a summary of research findings.

Nonsteroidal Anti-Inflammatory Drug (NSAID): Meloxicam

Summary

Oral meloxicam is an FDA-approved NSAID that is a potent inhibitor of cyclo-oxygenase-2 (COX-2) and prostaglandin synthesis, used to treat arthritis-related pain. In vitro evaluation of NSAIDs (meloxicam) on cell signaling pathways in cultured synovial fluid indicates these agents may interfere with a natural regulatory signal that limits inflammation (Largo et al., 2004). The physiochemical properties of meloxicam indicate that this NSAID is not an ideal drug for delivery through the skin. Meloxicam is zwitterionic, and it is lipophilic with a high melting point and low solubility, which make it unsuitable for transdermal delivery (Ah et al., 2010). However, a variety of approaches have been investigated to enhance the topical delivery of meloxicam with promising outcomes in animal studies (Chen and Gao, 2016). Numerous in vitro and in vivo animal studies have demonstrated that meloxicam can be delivered into the skin and underlying tissues. No human clinical trials have been conducted to confirm the safety or effectiveness of meloxicam administrated in a compounded topical cream. See Box 6-15 for a summary of research findings.

Effectiveness

Randomized controlled trials

A review of randomized clinical trials provided insufficient evidence to compare topical NSAID therapies with oral delivery of the same agent, other topical treatments, or other treatments (Glass, 2006).

Preclinical

Meloxicam gel (1 percent w/w gel) showed lower analgesic effectiveness than piroxicam (0.5 percent w/w) and diclofenac (1 percent

w/w) gels in the rat writhing test and in formalin-induced phase 1 pain. Meloxicam gel provided greater protection than the comparator gels in formalin-induced phase 2 pain (Gupta et al., 2002).

Dermal Penetration/Bioavailability

Preclinical studies

Permeation through human cadaver skin in Franz diffusion cell chambers was reported as having a flux of 2.43 ± 0.47 mcg/cm²/h following application of 0.5 g meloxicam 0.3 percent gel formulation containing 2.5 percent hydroxypropylcellulose in 1:1:1 propylene glycol, ethanol, and water with the addition of 5 percent menthol as a penetration enhancer (Jantharaprapap and Stagni, 2007).

A comparison of oral and transdermal plasma and synovial fluid levels was conducted in six beagle dogs using liquid chromatography-tandem mass spectrometry. The oral dose given was 0.31 mg/kg. The meloxicam gel (1 percent meloxicam in carbopol 940) was formulated with the penetration enhancer, transcutol. The meloxicam gel was administered transdermally at 1.25 mg/kg. Relative bioavailability of the topical gel was 1.05 percent of the oral plasma concentration. The ratio of meloxicam in synovial fluid compared to the plasma concentration was higher for topical

administration than oral delivery. This result was attributed to direct penetration into the target tissue following topical administration (Yuan et al., 2009).

Other preclinical studies have examined the pharmacokinetics of meloxicam following topical delivery in rats (Chang et al., 2007; Gupta et al., 2002; Jain et al., 2008). Some of these studies, and others, have used advanced drug delivery formulations to improve penetration rates and systemic absorption of the meloxicam (Chang et al., 2007; Duangjit et al., 2014a,b; Huang et al., 2011; Jain et al., 2008; Machado et al., 2018; Ngawhirunpat et al., 2009; Zhang et al., 2009).

Safety and Adverse Effects

No human clinical data have been reported; however, several studies have evaluated the potential for skin irritation with meloxicam following topical (transdermal) delivery in animal models (Bachhav and Patravale, 2010; Jiang et al., 2018; Khurana et al., 2013a,b,c).

From the FDA-approved label, serious adverse effects from oral administration of this API include increased risk of cardiovascular events, gastrointestinal bleeding, and ulceration. See Table G-1 in Appendix G.

Nonsteroidal Anti-Inflammatory Drug (NSAID): Naproxen

Summary

Oral naproxen is an FDA-approved NSAID with analgesic, antiinflammatory, and antipyretic activities. The mechanism of action is unknown but involves inhibition of cyclo-oxygenase (COX-1 and COX-2), which leads to reduced prostaglandin synthesis. Well-designed clinical studies focusing on the effectiveness of topical naproxen are limited and often provide inconsistent results. For example, some studies indicate topical naproxen (1 percent to 10 percent gel) is more effective than placebo, while in others, naproxen was determined to be no more effective than placebo in acute conditions (Moore et al., 1998). Several studies lack strong methodological rigor. Studies reported very few adverse effects, with the exception of self-reported skin irritation or itching, and there is clinical evidence that naproxen exhibits dermal penetration when applied topically. Overall, the committee reviewed six RCTs: one study was assessed as having a high risk of bias (Cokmez et al., 2003), and five were assessed as having a low risk of bias (Baixauli et al., 1990; Eslamian et al., 2017; Montagna et al., 1990; Nadal et al., 1990; Thorling et al., 1990). See Box 6-16 for a summary of research findings. (See Appendix B for more details.)

Effectiveness

Systematic review

A systematic review by Moore et al. (1998, p. 333) examined “the effectiveness and safety of topical non-steroidal antiinflammatory drugs in acute and chronic pain conditions.” In a review of five placebo-controlled trials, the authors determined that topically applied naproxen did not show significant efficacy in acute pain conditions (e.g., soft tissue trauma, strains, and sprains). The efficacy of topically applied naproxen in chronic pain conditions was not described in the review.

Randomized controlled trials

In a nonblinded RCT by Cokmez et al. (2003) to prevent infusion phlebitis in Turkey, Naprosyn gel was applied to patients’ skin over the course of the vein proximal to the cannulas (n = 127).²⁶ This was compared to 10 mg/day transdermal glyceryl trinitrate (GTN) patches (n = 136) and controls (n = 123). The study found patients who received the Naprosyn gel developed less phlebitis compared to no treatment control and the GTN group (p < .05).

In a double-blind RCT by Eslamian et al. (2017) in Iran, the analgesic efficacy of 5 percent naproxen gel was tested for pain associated with orthodontic separator placement.²⁷ Thirty-four patients (11 males and 23 females) between the ages of 14 to 20 (mean: 16.88 years) were put in a split mouth design study. The 5 percent naproxen gel and placebo gel were applied immediately after spacer placement and every 8 hours after

___________________

²⁶ Risk-of-bias assessment by committee: High (see Appendix B for more details).

²⁷ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

for 3 days. Pain was rated using a VAS (0–100 in increments of 10) at 2 and 6 hours after placement, as well as at 10 am and 6 pm of the second, third, and seventh days after application of the gel. The gel was comprised of carbomer P934 gel-forming substance (50 g), preservatives (5 g methylparaben and 1 g propylaparaben), glycerin as humectant (400 ml), pH regulator (NaOH), and 10 g naproxen powder. Findings show the naproxen gel side had lower pain scores at all times compared with placebo (p < .001).

Montagna et al. (1990) conducted a single-blind study comparing 5 percent meclofenamic acid gel to 10 percent naproxen gel in patients (n = 40, 20 each group) with musculoskeletal disorders.²⁸ Evaluation was on pain, tenderness, swelling, and restriction of movement. Patients applied cream for 15 days (the amount and frequency are unclear). Meclofenamic acid gel outperformed naproxen gel by alleviating symptoms faster, although both were effective in controlling the symptoms and showed similarly good tolerance.

Nadal et al. (1990) performed a similar study comparing 10 percent naproxen gel to piketoprofen cream (ketoprofen).²⁹ Fifty patients (25 in each group) were treated for soft tissue lesions, and cream was applied every 12 hours, as and when required. The naproxen gel was considered more effective and more rapid in onset than piketoprofen cream. Both were tolerated well.

Baixauli et al. (1990) also did a comparative study between 10 percent naproxen gel and 10 percent ketoprofen gel in 30 patients with moderate to severe pain caused by acute soft tissue lesions.³⁰ There were 15 patients in the naproxen group (8 male, 7 female; mean age: 72.2 years) and 14 in the ketoprofen group (8 male, 6 female; mean age: 33.3 years). Three to 5 cm of gel was applied once every 12 hours as needed for 7 days. Naproxen gel and ketoprofen gel efficacy and tolerability were deemed comparable, except on deep palpation where naproxen gel had a significantly greater reduction in pain by the third day of treatment.

In a narrative review by Heyneman (1995), she discussed a double-blinded, randomized, and parallel study by Thorling et al. (1990) that compared 10 percent naproxen gel and a placebo in 120 patients with soft tissue injuries (mostly synovitis and tendinitis).³¹ There were several limitations to the study (dosage was not standardized, no adherence measure, some took acetaminophen 500 mg tablets, and measurement error), but the study did find the naproxen gel significantly reduced pain compared to placebo.

___________________

²⁸ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

²⁹ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

³⁰ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

³¹ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

Dermal Penetration/Bioavailability

Clinical studies

In the study by Attia (2009), transdermal administration of 1 percent naproxen sodium gel prepared without penetration enhancers, when compared with oral administration of sodium naproxen tablets, was significantly different on maximum blood concentration (lower), time to reach the peak blood concentration (the same), terminal elimination half-life (1 hour sooner), and area under the curve (higher).

There were no studies that evaluated the blood levels of naproxen resulting from topical versus oral administration of the drug. Only Attia (2009) discussed serum concentrations, which peaked at 1.3 µg/mL after approximately 8 hours after application of a 1 percent naproxen gel.

Safety and Adverse Effects

No severe adverse events were reported (Cokmez et al., 2003; Eslamian et al., 2017; Nadal et al., 1990; Thorling et al., 1990).

From the FDA-approved label, serious adverse effects from oral administration of this API include increased risk of cardiovascular events, gastrointestinal bleeding, and ulceration. See Table G-1 in Appendix G.

Opioid Agonists: Tramadol

Summary

Oral tramadol is an FDA-approved analgesic that is a weak agonist at the mu opioid receptor and also inhibits the uptake of norepinephrine and serotonin. It is considered a treatment for mild to moderate chronic pain, but not severe pain. Tramadol is available orally alone or in combination with acetaminophen in immediate release and extended release preparations. Onset of effects is within 1 hour with peak analgesic effects within 2 hours after immediate release preparation. Time to peak concentrations are 1–3 hours with immediate release preparations and 12 hours with extended-release preparations. Tramadol undergoes extensive metabolism by the liver and has an active metabolite, O-desmethyl-tramadol. Approximately 30 percent is excreted unchanged in the urine. The half-life of tramadol is approximately 6 hours, and the half-life of the active metabolite is approximately 7 hours (Ardakani and Rouini, 2007; Skinner et al., 2009). No data were found on the safety and effectiveness of topically applied tramadol. See Box 6-17 for a summary of research findings.

In addition to reviewing studies where tramadol alone was analyzed, the committee also included studies of tramadol used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized controlled trials

There are no relevant clinical studies that examine the single-drug use of tramadol in compounded topical creams to treat pain when applied to intact skin.

Dermal Penetration/Bioavailability

There is no evidence on the ability of tramadol to penetrate the skin or enter the blood stream.

Safety and Adverse Effects

There is no relevant evidence on the safety or adverse effects of compounded topical tramadol. If a formulation were developed that produced a systemic level after topical application, then systemic side effects should be considered. After oral administration side effects would be expected to be similar to other opioid analgesics, for which the most common adverse effects are skin (flushing and pruritus), gastrointestinal (constipation, nausea, vomiting), and neurologic (dizziness, drowsiness) (Khansari et al., 2013).

From the FDA-approved label, serious adverse effects from oral administration of this API include potential addiction, abuse, and misuse; respiratory depression; accidental ingestion; and sedation, coma, and death if used with benzodiazepines or alcohol. See Table G-1 in Appendix G.

Steroid: Dexamethasone

Summary

Dexamethasone is a potent glucocorticoid that has anti-inflammatory properties. It is FDA approved to be used topically in ophthalmic and otic preparations. It has been studied for application in the mouth and treatment of recurrent aphthous ulcers, but it has no approved indications to treat pain. No relevant evidence exists regarding effectiveness, absorption, or safety of dexamethasone alone or in combination with other active ingredients in compounded topical pain creams. From the FDA-approved label, serious adverse effects from oral administration of this API include cardiomyopathy, hyperglycemia, or pancreatitis. See Box 6-18 for a summary of research findings. See Table G-1 in Appendix G.

Tricyclic Antidepressant: Amitriptyline

Summary

Oral amitriptyline is an FDA-approved tricyclic antidepressant that was originally developed as a mood-regulating agent. It is thought to work by inhibiting the reuptake of norepinephrine and serotonin by presynaptic neuronal membranes in the CNS. At the spinal cord, amitriptyline exhibits ion-channel blocking effects on sodium, potassium, and NMDA channels. Norepinephrine, and sodium and NMDA channels, are involved in maintenance of some types of neuropathic pain. Off-label use of the oral formulation has been found to consistently reduce neuropathic forms of pain to a greater degree than comparison placebo groups (Moore et al.,

2015). However, there is very little evidence for its effectiveness when applied as a topical agent. Very low doses were detected in the blood stream after topical application. Adverse clinical effects included skin irritation, dryness, itching, and redness. One patient reported drowsiness as a side effect with use of 10 percent amitriptyline cream. See Box 6-19 for a summary of research findings.

Overall the committee reviewed six RCTs. Three were assessed as having a low risk of bias (Dualé et al., 2008; Kiani et al., 2015b; Lynch et al., 2003), and three were assessed as having a high risk of bias (Gerner et al., 2003; Ho et al., 2008; Lynch et al., 2005b). (See Appendix B for more details.) In addition to reviewing studies where amitriptyline alone was analyzed, the committee also included studies of amitriptyline used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized controlled trials

A double-blind, randomized, placebo-controlled crossover study evaluated the efficacy of topical 5 percent amitriptyline, 5 percent lidocaine, or placebo (compounded in pluronic lecithin organogel) in treating patients with neuropathic pain (Ho et al., 2008).³² Thirty-five patients (16 males, 19 females; mean age 57.4 years)

___________________

³² Risk-of-bias assessment by committee: High (see Appendix B for more details).

with postsurgical neuropathic pain, postherpetic neuralgia, or diabetic neuropathy with allodynia or hyperalgesia were assigned in random sequence to apply 3–5 mL of a cream twice daily for 1 week. The primary outcome measure was change in pain intensity (baseline versus posttreatment average pain) using a 0 to 100 mm VAS. No significant change in pain intensity was found with amitriptyline. Both lidocaine and placebo resulted in significantly (p < .05) reduced scores compared to amitriptyline.

The effectiveness of 2 percent amitriptyline in treating peripheral diabetic neuropathy was evaluated in 51 patients (17 males, 34 females; mean age 57.5 years) (Kiani et al., 2015b).³³ A matched group of 51 patients given 0.75 percent capsaicin served as controls. Patients applied the cream three times daily to their feet for 12 weeks. Pain was measured by the VAS with a 50 percent reduction considered a positive response. After 12 weeks, 43.1 percent of patients given amitriptyline and 37.3 percent of patients given capsaicin were considered responders, which was not statistically significant.

The sensory effects of amitriptyline diluted in water/isopropanol/glycerin solution were examined in 15 healthy young male volunteers given randomized treatments of 0 (vehicle), 25, 50, or 100 mM on different areas of the skin of the back (Dualé et al., 2008).³⁴ Total dose for the session was 82.4 mg. Saline and lidocaine–prilocaine cream served as negative and positive controls, respectively. Mechanical thresholds for touch and nociception, and thermal thresholds for cold, warm, and heat sensation were recorded for each area. Amitriptyline induced a mild increase (p ≤ .01) of the tactile and mechanical nociceptive thresholds at ≥ 50 mM, and all concentrations significantly decreased (p ≤ .01) cold and heat thresholds compared to controls. These effects were no longer significant after 4 hours.

Analgesic effects of amitriptyline were evaluated in 14 healthy volunteers (sex not stated) by application of 0.3 mL of 0, 10, 50, or 100 mmol/L solutions to the upper arm (Gerner et al., 2003).³⁵ The vehicle was water/isopropanol/glycerin, and pain was measured by poking with a blunt needle. The VAS was significantly reduced at concentrations ≥ 50 mmol/L for up to 4 hours after removal of the gauze.

Lynch et al. (2003, 2005b) evaluated the effects of topical ketamine, amitriptyline, or a combination of the two in the treatment of neuropathic pain. To minimize redundancy within the chapter, a review of these studies is described in the Multiagent Compounded Topical Pain Creams section below.

___________________

³³ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

³⁴ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

³⁵ Risk-of-bias assessment by committee: High (see Appendix B for more details).

Case reports

Three patients with neuropathic pain in the hands and feet were treated with 3 mL of 5 and 10 percent amitriptyline (Kopsky et al., 2012). All patients noted partial relief with the 5 percent cream and almost complete relief with the 10 percent cream. Two patients reported tiredness/drowsiness with 10 percent cream.

Another patient (42 years, male) was prescribed 5 percent amitriptyline, three times daily, for chronic severe pain in the ankle after a fall. After 1 month, his pain score decreased by 30 percent (Kopsky and Keppel Hesselink, 2011).

A 42-year-old male was given amitriptyline cream (150 mg/2 mL) for abdominal pain caused by Crohn’s disease (Scott et al., 1999). The cream was applied to the chest at bedtime, and the patient monitored for 6 weeks. The abdominal pain remained unchanged with treatment, although the patient reported his mood improved.

A 62-year-old woman with pain in the upper arms and left foot attributable to multiple sclerosis was prescribed 5 percent amitriptyline (Kopsky and Keppel Hesselink, 2012; Kopsky et al., 2012). Application of the cream to the upper forearms resulted in decreased pain in the foot and upper arms after several minutes delay. Blood levels were not measured.

Preclinical studies

Animal studies have also demonstrated a potential cutaneous effect. Two studies tested cutaneous nociception in male Sprague-Dawley rats (n = 6–8) before and after application of 0.3 mL containing amitriptyline to the shaved back for 3 hours under a patch. Test solutions were formulated in isopropyl alcohol/water or saline/glycerin. The cutaneous trunci muscle reflex response was measured via pinprick. A 2.5 percent solution resulted in complete block to the pinprick for 4.5 hours with complete recovery at 96 hours (Colvin et al., 2011). Concentration-related analgesic effect was observed with almost 100 percent at 500 mM, 90 percent at 100 mM, and 50 percent at 50 mM; no effects were seen after several days, 25 hours, and 10 hours, respectively. Rats given 500 mM developed redness and skin induration at the application site, which disappeared after several hours (Haderer et al., 2003).

Dermal Penetration/Bioavailability

Clinical studies

Five urine samples were collected to evaluate drug levels resulting from topical amitriptyline application (Glinn et al., 2017). Dosing and formulation information were not available. Amitriptyline was detected in two samples, both at < 25 ng/mL, while nortriptyline was not detected. The therapeutic level of amitriptyline is typically defined as between 70 and 220 ng/mL (Ulrich et al., 2001).

Amitriptyline was not detected (Limit of Detection = 2 ng/mL) in plasma samples from 15 healthy young male volunteers following application of 82.4 mg to the skin of the back (Dualé et al., 2008).

A 42-year-old male was given amitriptyline gel (150 mg/2 mL) for application to the chest once daily to treat abdominal pain caused by Crohn’s disease (Scott et al., 1999). After 9 days of treatment, blood levels were measured over 24 hours. The patient had previously been given amitriptyline 80 mg/day by intramuscular injection for 19 days resulting in a steady state level of 201 ng/mL. With application of the amitriptyline gel, blood levels ranged from 196.2 to 263.3 ng/mL.

Preclinical studies

Dermal penetration and deposition were measured using excised skin from the backs of female nude mice and pigs (Liu et al., 2016). A 15 mM amitriptyline solution in 20 percent propylene glycol was placed on the donor side of the skin for 24 hours. The amount deposited in the skin, the flux, and the total percent absorbed (skin + receptor) were 28.25 nmol/mg, 10.22 nmol/cm²/h, and 21 percent, respectively, for nude mouse skin and 11.76 nmol/mg, 22.53 nmol/cm²/h, and 14 percent, respectively, for pig skin. Using a 6.5 mM solution, the total absorbed was 33 percent and 19 percent for nude mouse and pig, respectively.

Safety and Adverse Effects

Clinical studies

One study reported in patients with peripheral diabetic neuropathy treated with 2 percent amitriptyline, 8.8 percent reported dryness and 4.4 percent had itching (Kiani et al., 2015b). Another clinical study reported redness was observed at the application site (upper arm) up to 6 hours after application in 10 out of 14 volunteers (50 mmol/L treatment) and 12 out of 14 volunteers (100 mmol/L treatment) (Gerner et al., 2003).

Drowsiness was reported in a patient treated with 10 percent cream for neuropathic pain in the feet (Kopsky et al., 2012).

From the FDA-approved label, serious adverse effects from oral administration of this API include increased suicidal thoughts or cardiac arrhythmias. See Table G-1 in Appendix G.

Tricyclic Antidepressant: Doxepin

Summary

Doxepin is an FDA-approved psychotherapeutic agent (oral, topical) from the class of dibenzoxepin tricyclic compounds. The mechanism of action of doxepin has not been confirmed. The current understanding is that doxepin prevents reuptake of norepinephrine into nerve terminals by

influencing activity in the synapse; this allows neurotransmitter activity to be prolonged. Norepinephrine is thought to be involved in the maintenance of some types of neuropathic pain, and this provides a rationale for off-label use of doxepin to treat certain types of pain.

Evidence from a few clinical studies and case reports suggest topical doxepin at 3.3 percent concentration as aqueous cream may be potentially effective for treating pain, including neuropathic pain and complex regional pain syndrome. Safety data indicate the potential for centrally mediated side effects when systemic absorption occurs, including drowsiness, rash, and headache. Topical applications also demonstrated occasional drowsiness and headache, as well as local effects such as itching and burning. The one relevant RCT reviewed by the committee was assessed as having a low risk of bias (McCleane, 2000). See Box 6-20 for a summary of research findings. (See Appendix B for more details.)

Effectiveness

Randomized controlled trials

One randomized, double-blind, placebo-controlled human study was conducted to assess the analgesic efficacy of topical administration of 3.3 percent doxepin HCl cream compared to 0.025 percent capsaicin cream and a combination cream of 3.3 percent doxepin and 0.025 percent capsaicin for treating chronic neuropathic pain

(McCleane, 2000).³⁶ Patients across the three comparative treatment groups (n = 95) produced similar analgesic effects with the combination treatment having fastest onset. Overall pain was unchanged in the placebo group.

No relevant clinical trials are listed in ClinicalTrials.gov; two trials are disclosed—one is for treating oral mucositis and the other is for treating pruritus.

Case reports and case series

A case report describes a 17-year-old female leukemia patient with mucormycosis infection that developed severe refractory peripheral neuropathy believed to be a side effect from treatment. The patient responded “dramatically and consistently” to topical treatment with 5 percent doxepin cream three times daily (Dworsky et al., 2017).

Another case report exists of a 32-year-old female patient with complex regional pain syndrome type 1 resulting from a wrist injury. Symptoms were reduced significantly after 2 weeks of twice-daily topical application of doxepin cream (McCleane, 2002).

Preclinical studies

Topical doxepin was significantly more effective than control (p < .05) in rat nociceptive pain (Gerner et al., 2006).

Dermal Penetration/Bioavailability

Preclinical studies

Permeation through human cadaver skin was reported as having a flux of 2.74 ± 0.14 µg/h cm² following application of doxepin formulated in a 3 percent (weight/volume) nanoemulsion (Sandig et al., 2013).

Safety and Adverse Events

Clinical trials

In one clinical trial, side effects were indicated as minor. Drowsiness occurred in four patients (9.7 percent) in the doxepin and two patients (5.5 percent) in the doxepin/capsaicin group, skin rash was reported in one patient (2.4 percent) with doxepin, headache was reported in one patient (2.8 percent) with doxepin/capsaicin, and itching was reported by two patients (4.9 percent) with doxepin. Twenty-seven patients (81 percent) in the capsaicin group, 22 patients (61 percent) in the doxepin/capsaicin group, and 4 patients (17 percent) in the doxepin group reported burning discomfort after application of cream (McCleane et al., 2000).

From the FDA-approved label, serious adverse effects from oral administration of this API include ventricular arrhythmia, thrombocytopenia, suicidal thoughts, or kidney damage. See Table G-1 in Appendix G.

___________________

³⁶ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

Vasodilator: Pentoxifylline

Summary

Pentoxifylline is a methylxanthine derivative that is used in micro-circulatory disorders as a vasoactive drug promoting blood flow via oral route of administration. In the United States, oral pentoxifylline is FDA approved for symptomatic treatment of patients with chronic occlusive peripheral vascular disorders of the extremities. In such patients pentoxifylline may give relief of signs and symptoms of impaired blood flow, such as intermittent claudication or trophic ulcers. It is nearly completely absorbed by oral administration with relatively few reported side effects. The primary side effects of oral administration have been malaise (1–3 percent), flushing (1–3 percent), dizziness/light-headedness (9.4 percent), headache (4.9 percent), nausea (14 percent), vomiting (3.4 percent), and abdominal discomfort, bloating, diarrhea, or dyspepsia (1–3 percent) (Cunha, 2016).

There is a well-designed clinical crossover randomized study demonstrating efficacy of compounded pentoxifylline on experimentally induced allodynia and pain induced by the injection of dermal capsaicin in humans (Ragavendran et al., 2016). There is also one carefully designed and conducted crossover study in rats demonstrating efficacy in experimentally induced neuropathic pain symptoms (Ragavendran et al., 2013). Finally, there is one ongoing randomized clinical trial of topical pentoxifylline with clonidine in patients with neuropathic pain by the same research group, but recruitment has been slow and results are not expected to available until the end of 2020 (Coderre et al., 2017). Other literature includes few case studies either as single-drug compounded topical drug or in combination with others (Carr et al., 1994; Paulis et al., 2015; Safaeian et al., 2016). The quality of the remaining studies does not allow any additional conclusions regarding effectiveness or safety. See Box 6-21 for a summary of research findings.

The one RCT reviewed by the committee was assessed as having a low risk of bias (Ragavendran et al., 2016). (See Appendix B for more details.) In addition to reviewing studies where pentoxifylline alone was analyzed, the committee also included studies of pentoxifylline used in combination with other APIs. (See the Multiagent Compounded Topical Pain Creams section of this chapter.)

Effectiveness

Randomized controlled trials

A randomized crossover study (Ragavendran et al., 2016) was conducted in 69 patients exposed to intradermal injection of capsaicin to induce allodynia and pain.³⁷ The evaluation of

___________________

³⁷ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

pentoxifylline alone involved application of pentoxifylline 5 percent in a vehicle of anhydrous ethanol (6.5 percent), polyethylene glycol 400 (20 percent), propylene glycol (53.5 percent), and oleyl alcohol (20 percent) topically to intact skin in the area of capsaicin-induced allodynia. The control was the vehicle alone. Dosing was controlled by a pump equipped with an actuator, which when pressed once was calibrated to deliver 0.5 ± 0.04 mL of solution used to cover approximately 1,200 mm² of the skin overlapping the area of capsaicin-induced hypersensitivity. The effect of pentoxifylline on VAS scores, dynamic mechanical allodynia, or punctate mechanical allodynia was no different from that of the inactive vehicle.

A forthcoming study by the same research group will examine clinical outcomes from a randomized clinical trial of topical pentoxifylline with clonidine in patients with neuropathic pain.

Preclinical studies

In rats with induced neuropathic pain via (1) chronic postischemia pain, (2) chronic constriction injury of the sciatic nerve, (3) diabetic neuropathy, or (4) chemotherapy-induced painful neuropathy, topical application of pentoxifylline and its active metabolite lisofylline were tested across a variety of concentrations (Ragavendran et al., 2013). An amount of 150 mg ± 2.7 mg of the ointment was used for all rat hind paw applications and compared to ointment base application to the ipsilateral paw. Pentoxifylline alone was tested at 0.6, 1.2, 2.5, and 5 percent w/w (n = 6), and lisofylline alone was tested at 0.063, 0.09, 0.125, and 0.25 percent w/w (n = 10). All of the rats underwent initial baseline paw

withdrawal threshold (PWT) assessment before application of the ointment followed by testing at 45–90 minutes postapplication. When tested alone, pentoxifylline only significantly elevated PWT at the highest dose tested (5 percent). Lisofylline alone significantly increased PWT at 0.09, 0.125, and 0.25 percent w/w.

In a separate study of rats (Laferriere et al., 2014) with induced hind paw chronic postischemia pain and allodynia, a topical application of lisofylline (an analogue of pentoxyfilline) and apraclonidine produced significant dose-dependent antiallodynic effects when compared with the vehicle/base in rats with chronic postischemia pain (n = 30), a finding not reproduced with control treatment of the nonpainful paw. Topical combination produced antiallodynic effects lasting up to 6 hours (n = 15).

Dermal Penetration/Bioavailability

In vitro studies

One study (Wang and Black, 2013) examined the dermal penetration of multiple products including pentoxifylline, showing better transdermal absorption with a novel proprietary product showing 2–3 times the penetration with the novel excipient. There is no clear interpretation of the values in a clinical sense.

Safety and Adverse Effects

Randomized controlled trials

Ragavendran et al. (2016) determined that at a dose of pentoxifylline 5 percent, no subject exhibited a drop in blood pressure greater than 15 percent to 30 percent. Additionally, no subjects reported any adverse events associated with the administration of active treatments or vehicle.

From the FDA-approved label, serious adverse effects from oral administration of this API include thrombocytopenia. See Table G-1 in Appendix G.

MULTIAGENT COMPOUNDED TOPICAL PAIN CREAMS

As summarized in the introduction of this chapter, with one exception (i.e., a clinical study of clonidine and pentoxifylline), little information was found to inform the effectiveness and safety of topical pain creams containing multiple chemicals. In addition, adverse events were similar to those reported for the individual ingredients. An important confounding factor identified in several of the studies described below is the presence of lidocaine in the compounded formulation. Lidocaine is used clinically, and is FDA approved, to induce topical analgesia. The methodological controls and outcome data in many of the articles reviewed below were insufficient to determine whether efficacy improved in combination with

other ingredients. As a result, no conclusions can be made on the committee’s review of the creams containing five to seven ingredients owing to the consistent presence of lidocaine, the inclusion of chemicals not evaluated individually by the committee, or the lack of appropriate controls.

Relevant data within the categories of effectiveness, dermal absorption/bioavailability, and safety and adverse effects are summarized where possible.

Two-Drug Combinations: Amitriptyline and Ketamine

Effectiveness

Randomized controlled trials and other clinical trials

Lynch et al. (2003) conducted a double-blind, placebo-controlled, four-way crossover pilot study in 20 patients to evaluate the safety, efficacy, and tolerability of ketamine, amitriptyline, or a combination of the two in the treatment of neuropathic pain.³⁸ Patients with postherpetic neuralgia, diabetic neuropathy, or postsurgical or posttraumatic neuropathic pain who had moderate to severe pain for 3 months or greater duration with allodynia and hyperalgesia were included in the study. Treatment included four different topical creams: amitriptyline 1 percent; ketamine 0.5 percent; amitriptyline 1 percent + ketamine 0.5 percent; or placebo (vehicle only). Subjects applied 5 mL of cream to the site of maximum pain four times daily for 2-day treatment periods. Pain was measured using the McGill Pain Questionnaire and a Likert scale measuring pain intensity “now” and “past 24 hours” and a VAS indicating pain relief (no relief to complete relief). Plasma concentrations of amitriptyline and ketamine and norketamine were measured. It is not clear whether there was a washout period between the four arms. All 20 patients completed the study, and analysis was conducted on 18 subjects who had complete data. There was no significant difference in the McGill Pain Questionnaire by drug type. There was a reduction over time of pain regardless of the treatment. Ketamine and norketamine concentrations were low overall and below detectable limits in 8 out of 11 subjects, with norketamine below limit of detection in 5 of those 8 patients.

Lynch et al. (2005b) conducted a randomized placebo-controlled study in 92 subjects to evaluate the efficacy of topical amitriptyline 2 percent versus ketamine 1 percent versus a combination of amitriptyline 2 percent/ketamine 1 percent for the treatment of neuropathic pain over a 3-week period in patients with mixed neuropathic pain.³⁹ The excipient was described as a moisturizing creamlike base. Patients with a diagnosis of

___________________

³⁸ Risk-of-assessment by the committee: Low (see Appendix B for more details).

³⁹ Risk-of-bias assessment by the committee: High (see Appendix B for more details).

postherpetic neuralgia, diabetic neuropathy, or postsurgical/posttraumatic neuropathic pain that was moderate to severe in nature and persisted for 3 months or more were eligible to participate. Subjects applied 4 mL of cream to site of maximum pain three times per day for 3 weeks. Average pain levels were assessed at baseline and at weeks 2 and 3. Outcome measures included measures of spontaneous pain using the numeric rating scale for pain intensity, the McGill Pain Questionnaire, and a measure of spontaneous pain that assesses sensory and affective dimensions of pain; perceived disability; and patient satisfaction. Plasma concentrations of amitriptyline, ketamine, and norketamine were measured after 2 weeks and 3 weeks of treatment. A total of 92 subjects were randomized to the four treatment arms, and 80 subjects completed the study. Reason for withdrawal was adverse events in five patients. All groups experienced a reduction in pain over time, but there was no difference between treatment groups with respect to pain measures, sensory measures, or perceived disability. All groups identified a moderate level of satisfaction. Detectable but low ketamine concentrations were found in three subjects. One patient using 2 percent amitriptyline cream for 3 weeks had blood levels of amitriptyline and nortriptyline of 92 ng/mL and 24 ng/mL, respectively; the study noted two other patients not given amitriptyline also had detectable blood levels (Lynch et al., 2005b). Blood levels in patients using the 1 percent cream were below the limit of detection, < 15.7 ng/mL (Lynch et al., 2005b). With longer treatment, pain scores after 2 to 12 months were 3.83–4.42 compared to a pretreatment score of 6.57 resulting in a 34–46 percent reduction in pain (Lynch et al., 2005a).

A cohort of participants from within Lynch et al. (2005b) were treated with 1 percent amitriptyline/0.5 percent ketamine cream for 3 weeks; they reported greater catastrophic thoughts and feelings when in pain, measured using the Pain Catastrophizing Scale, which was associated with less pain reduction with treatment and greater pain reduction with placebo (p < .05) (Sullivan et al., 2008).⁴⁰ Further evaluation of the individual responses showed that high scores on the measure of pain catastrophizing prospectively predicted poorer response to treatment. Fewer high-catastrophizing than low-catastrophizing individuals showed moderate or substantial reduction in pain ratings, and fewer high-catastrophizing than low-catastrophizing individuals achieved pain ratings below 4/10 (Mankovsky et al., 2012).

A phase 2 study evaluated the safety and efficacy of a combination of ketamine 2 percent and amitriptyline 4 percent (KA) in the treatment of cancer chemotherapy-induced peripheral neuropathy (CIPN) (Gewandter et al., 2014).⁴¹ Patients were recruited from multiple sites. The

___________________

⁴⁰ Risk-of-bias assessment by committee: High (see Appendix B for more details).

⁴¹ Risk-of-bias assessment by committee: High (see Appendix B for more details).

total population was 88 percent White, 8 percent African American, and 71 percent female. Participants were at least 1 month post chemotherapy completion and had an average 7-day pain, numbness, and tingling rating of ≥ 4. Subjects were randomized in blocks of four using computer-generated random numbers and stratified based on study site and two treatment groups (patients taking cancer chemotherapeutic taxane agents versus nontaxane agents).

Participants applied cream using a measuring device in an amount up to but not exceeding 4 g of KA cream two times per day to the area with pain, numbness, and tingling. Pain, numbness, and tingling diaries were completed for a 7-day period prior to the start of the study and after study enrollment at weeks 3 and 6. A measure of worst pain over the past 24 hours using numeric rating scale was assessed. A total of 462 subjects were enrolled (229 KA and 233 placebo). Two subjects were excluded from each group leaving 227 in KA and 231 in placebo groups whose data were analyzed on intent to treat basis. Four subjects withdrew from KA and 3 from placebo for skin-related reasons.

At the 6-week assessment, KA cream had no effect on CIPN scores (adjusted mean difference = –0.17, p = .363). In addition, there were no differences in scores for numbness or tingling in the hands and feet between the groups. Adverse events were reported in similar numbers in both groups (147 in KA group, 158 in placebo group), and when evaluated by class there was no significant difference between KA and placebo. The most common adverse events reported were hypertension, fatigue, dyspepsia, insomnia, headache, fever, musculoskeletal and connective tissue disorders, neuropathy, respiratory disorders, and a burning sensation or rash where applied.

Two randomized studies evaluated the effectiveness of 2 percent ketamine/4 percent amitriptyline cream in treating postherpetic neuralgia (84 males; 56 females) or diabetic peripheral neuropathy (65 males; 49 females) (EpiCept Corporation, 2008a,b).⁴² The cream was applied twice daily for 28 days, and pain intensity scores were compared to baseline and to controls (46–57 males; 30–55 females). Patients with postherpetic neuralgia had a significantly greater change (p = .044) in pain score with treatment compared to placebo. A slightly greater change (non-significant) in pain score was found for patients with diabetic peripheral neuropathy.

Retrospective clinical studies

Electronic medical records were reviewed to determine the effectiveness of 1–2 percent amitriptyline/0.5 percent ketamine for treatment of erythromelalgia (n = 36; 89 percent female) (Poterucha et al., 2013). Formulations were either in a pluronic lecithin

___________________

⁴² Risk-of-bias assessment by committee (EpiCept Corporation, 2008a,b): High (see Appendix B for more details).

organogel or in a moisturizing cream with application 1–6 times daily. Among the erythromelalgia patients 75 percent had improvement, 3 percent had complete relief, 39 percent noted substantial relief, and 33 percent had some relief, while 19 percent had no relief and 6 percent had worsening symptoms (Poterucha et al., 2013).

Case reports

Case reports describe moderate to complete pain/itch relief with 1 percent amitriptyline/0.5 percent ketamine for brachioradial pruritus (Poterucha et al., 2013), 5 percent amitriptyline/10 percent ketamine for ankle pain after a fall (Kopsky and Keppel Hesselink, 2011), and 2.5 percent amitriptyline/0.5 percent ketamine with severe proctodynia (Lehman and Sciallis, 2008). No adverse effects were reported in these cases.

Safety and Adverse Effects

Randomized controlled trials

As described above in Lynch et al. (2005b), in a randomized placebo-controlled study of 92 subjects to evaluate the efficacy of topical amitriptyline 2 percent versus ketamine 1 percent versus a combination of amitriptyline 2 percent/ketamine 1 percent for the treatment of neuropathic pain, adverse effects were reported in 30 percent of the subjects and were evenly distributed across treatment groups.⁴³ The most common adverse effects were minor skin irritation at the site of application. In Lynch et al. (2003), a double-blind, placebo-controlled, four-way crossover pilot study in 20 patients to evaluate the effect of ketamine, amitriptyline, or a combination of the two in the treatment of neuropathic pain, no subjects described severe adverse effects related to ketamine, although two subjects noted skin irritation and rash at the site of application.

Clinical studies

In patients treated with 2 percent ketamine/4 percent amitriptyline cream for postherpetic neuralgia (84 males; 56 females), seven reported vertigo while among those treated for diabetic peripheral neuropathy (65 males; 49 females), and two had pruritus and rash (EpiCent Corporation, 2008a,b). The percentage of patients with chemotherapy-induced peripheral neuropathy reporting adverse events was similar between the ketamine/amitriptyline treated group and the placebo group after 6 weeks (Gewandter et al., 2014).

Among 36 patients treated with 1–2 percent amitriptyline/0.5 percent ketamine for erythromelalgia, 1 showed reddening and 1 had a worsening of Raynaud phenomenon associated with erythromelalgia; no systemic effects were seen (Poterucha et al., 2013).

___________________

⁴³ Risk-of-bias assessment by the committee: High (see Appendix B for more details).

In vitro studies

As disused in the carbamazepine and gabapentin sections above, Al-Musawi et al. (2017) conducted a toxicity analysis of commonly used topical drugs (i.e., amitriptyline, carbamazepine, gabapentin) commonly used for the treatment of neuropathic orofacial pain. After examining the effects on keratinocytes in cell culture, the authors reported that carbamazepine was cytotoxic to skin and oral keratinocytes, demonstrating significant decrease in cellular viability and cell counts, at high concentrations (1.7 mM) and with long exposure (2 hours). The authors also reported that topical gabapentin was only minimally cytotoxic to skin or oral keratinocytes at high concentrations (5.54 mM) and long exposure (24 hours). Importantly, topical amitriptyline was reported as cytotoxic to skin and oral keratinocytes at both short (30 minutes) and long (24 hours) exposure times and at low (200 µM) and high concentrations (1.8 mM).

Two-Drug Combinations: Clonidine and Pentoxifylline

Effectiveness

Randomized clinical trials

A randomized, double-blind study in healthy volunteers evaluated whether clonidine in combination with pentoxifylline reduced pain after topical capsaicin. Participants were randomized to three different groups. Group 1 (n = 23): clonidine 0.04 percent and pentoxyfylline 2 percent was compared to placebo in a crossover design. Group 2 (n = 23): clonidine 0.1 percent and pentoxyfylline 5 percent were compared to placebo in a crossover design. Group 3 (n = 23): compared clonidine 0.1 percent to pentoxyfylline 5 percent. The high-dose clonidine/pentoxifylline group had a significant reduction in mean VAS scores compared to either drug alone or the vehicle (Ragavendran et al., 2016).⁴⁴

Three-Drug Combinations: Piroxicam, Lidocaine, and Cyclobenzaprine Hydrochloride (PLC Cream)

Effectiveness

Randomized controlled trials

Popescu et al. (2018) conducted an RCT involving 90 Romanian National rugby players (men and women, numbers of each not specified) who were older than 15 years, had musculoskeletal sports injuries without fracture or need for surgery, no therapy started in the past 7 days, and no use of corticosteroids.⁴⁵ Forty-seven rugby players applied a topical gel to the site of injury daily for 14 days containing piroxicam 5 mg/g, lidocaine 20 mg/g, and 5 mg/g cyclobenzaprine hydrochloride

___________________

⁴⁴ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

⁴⁵ Risk-of-bias assessment by committee: Some concerns (see Appendix B for more details).

(PLC), whereas 43 players applied a topical gel with only 10 mg/g piroxicam. The study design attempted to balance the use of ice and different analgesics between the two groups, although 45.6 percent of the PLC gel group used systemic analgesics versus 65.1 percent of the piroxicam gel group. Self-reported pain was surveyed on a VAS, 0–10, with movement and with pressure after application and on days 2, 7, 14.

Compared to the piroxicam gel control, participants using the PLC gel reported a greater decrease in pain sooner with the first application and thereafter on each measurement day. Pain alleviation with application lasted longer as well. On day 14 versus day 1, a 70 percent reduction was reported for the group using the PLC gel versus 45 percent for the group using piroxicam gel. The PLC gel group had 32 percent and 57 percent of participants who were pain free with movement and pressure, respectively, compared to 2.32 percent and 6.97 percent with piroxicam gel. No adverse side effects related to the treatment were reported for either of the two groups. The PLC gel was more effective than a higher dose piroxicam (10 mg/g) gel for pain in sports injuries in young adults. Limitations of this study for assessing the specific efficacy of cyclobenzaprine is that the PLC gel also contained a considerable amount of lidocaine, an anesthetic.

Another RCT in Romania involving 256 subjects investigated the use of the PLC topical gel (appears to be the same formulation as Popescu et al., 2018) for pain control in patients undergoing extracorporeal shock wave lithotripsy (ESWL) for renal or ureteral stones (Pricop et al., 2016).⁴⁶ The three study groups received either 1 g of the PLC topical gel applied to the treatment site at different times prior to ESWL or no PLC gel: Group A (number of women/men, 42/39) 30 minutes prior, Group B (60/30) 60 minutes prior, and Group C (46/39) no treatment. VAS pain scores of Groups A (3.76 + 1.03) and B (3.40 + 0.83) were less than C (5.38 + 1.46) (p < .0001 for both). Groups A and especially B also needed less rescue medication (tramadol) than C during the procedure. Application 60 minutes prior to treatment seems more optimal than 30 minutes based on less need for rescue medication: 7.8 percent for A, 4.7 percent for B, and 13.6 percent for C. Although application of the PLC gel before ESWL reduced pain perception and need for opioids, cyclobenzaprine was not evaluated separately and the study did not include a vehicle control/placebo.

Safety and Adverse Effects

Randomized controlled trials

Within the Pricop et al. (2016) RCT described above, no side effects or skin reactions were observed in the treated groups.

___________________

⁴⁶ Risk-of-bias assessment by committee: High (see Appendix B for more details).

Three-Drug Combinations: Baclofen, Amitriptyline, and Ketamine

Effectiveness

Randomized controlled trials

Patients with neuropathic pain, numbness, and/or tingling caused by chemotherapy-induced peripheral neuropathy were prescribed baclofen 10 mg/amitriptyline HCL 40 mg/ketamine 20 mg in a pluronic lecithin organogel (n = 101; 65 percent female; 94 percent White) or a placebo (n = 102; 59 percent female; 90 percent White) twice daily for 4 weeks (Barton et al., 2011).^47,48 The primary endpoint was change in the sensory neuropathy subscale as measured by the European Organization for Research and Treatment of Cancer QLQ-CIPN20 instrument from baseline to 4 weeks. The Brief Pain Inventory did not demonstrate significant differences nor did the Profile of Mood States; however, treated patients had significant improvement in the sensory (n = 75; p = .053) and motor subscales (n = 75; p = .021) of the CIPN20 owing to improvement in symptoms of tingling, cramping, and shooting/burning pain in the hands. Blood samples were drawn on a small subset of participants (n = 8) at the end of 4 weeks of treatment to evaluate systemic absorption by measuring concentrations of drugs and their metabolites. One of the four participants in the active drug arm had detectable concentrations of amitriptyline that were below therapeutic concentrations and no detectable ketamine or baclofen. Another participant had low therapeutic concentrations of baclofen but undetectable concentrations of amitriptyline and ketamine. The range of blood concentrations considered therapeutic were not described.

Safety and Adverse Effects

Clinical studies

Incidences of adverse events were similar between patients on baclofen 10 mg/amitriptyline HCL 40 mg/ketamine 20 mg gel and those on placebo (Barton et al., 2011).

Three-Drug Combinations: Ketamine, Amitriptyline, and Lidocaine

Effectiveness

Clinical studies

A 1 percent ketamine/2 percent amitriptyline/5 percent lidocaine gel was evaluated in 16 patients (88 percent female) with neuropathic pain from radiation-induced dermatitis (Uzaraga et al., 2012). Patients applied ~4 mL to the painful areas three times daily until 2 weeks

___________________

⁴⁷ Risk-of-bias assessment by committee: High (see Appendix B for more details).

⁴⁸ Treatment gel and placebo gel were compounded at Gateway Health Mart Pharmacy Laboratory in Bismarck, North Dakota.

after completion of radiotherapy. A reduction in scores for intensity, sharpness, burning, sensitivity, unpleasantness, and deepness was found 30 minutes after application; reduction in burning was maintained for 2 weeks posttreatment.

In a letter to the editor the authors retrospectively examined change in the numeric rating scale after topical application of ketamine 10 percent/amitriptyline 5 percent/lidocaine 5 percent for the treatment of chronic pruritus (Lee, 2017). This was a retrospective review of medical and pharmacy records of patients prescribed the topical product between September 1, 2013, and June 30, 2016. A total of 96 patients were identified. Average numeric rating scale reduced an average of 4.61 from 8.63 to 4.19 after treatment. Authors indicate 63 percent attributed relief directly to the topical product, although it is unclear what data elements were in the medical record that allowed for such a statement. Sixteen reported mild burning and redness.

Case reports

A 24-year-old male presenting with inflammatory linear verrucous epidermal nevus was prescribed 10 percent ketamine/5 percent amitriptyline/5 percent lidocaine in a Lipoderm base. After 6 weeks, one to three times daily, the patient had complete relief of pruritus and an itch rating of 0/10 down from 9.5/10 (Jaller and Yosipovitch, 2018).

Safety and Adverse Effects

Clinical studies

Of 16 patients treated with 1 percent ketamine/2 percent amitriptyline/5 percent lidocaine gel (88 percent female) for neuropathic pain from radiation-induced dermatitis, 3 had irritation at grade 1–2 and 5 reported fatigue at grade 1 (Uzaraga et al., 2012).

Cardis and Pasieka (2016) describe a man in his 80s who was treated with ketamine 10 percent/amitriptyline 5 percent/lidocaine 5 percent in Lipoderm for pruritus associated with atopic dermatitis. After applying to a small test area the patient gradually increased coverage of the cream to most of his upper body. He presented to the emergency department with slurred speech, ataxia, and altered mental status and was admitted to the hospital to rule out stroke. He had persistent nonfocal findings on neurologic examination, and the temporal relationship of his symptoms to the escalating dose of the cream was identified. The authors state that amitriptyline, lidocaine, ketamine, and their metabolites were detected in the urine by mass spectroscopy. A concentration of 2,360 ng/mL was reported, but it is unclear what this concentration represents. The patient improved over the next 2 weeks and was discharged on hospital day 17.

Greater Than Three Drug Combinations, Including Ketamine, Gabapentin, Clonidine, Lidocaine, Cyclobenzaprine, Baclofen, Ketoprofen, Diclofenac, Flurbiprofen, Tramadol, Bupivacaine, Amitriptyline, Nifedipine, Ibuprofen, Pentoxifylline, Imipramine, and/or Mefenamic Acid

Effectiveness

Randomized controlled trials

The safety and efficacy of three different compounded pain creams tailored for three types of pain (neuropathic, nociceptive, and mixed pain disorder) were evaluated in a clinical trial (Brutcher et al., 2019).⁴⁹ The randomized, double-blind, placebo-controlled, parallel-group clinical trial had a duration of 1 month of treatment with a compounded cream or a placebo cream. For those participants considered to have predominantly neuropathic pain (n = 133), the active cream contained 10 percent ketamine, 6 percent gabapentin, 0.2 percent clonidine, and 2 percent lidocaine. For those considered to have nociceptive pain (n = 133), the active cream contained 10 percent ketoprofen, 2 percent baclofen, 2 percent cyclobenzaprine, and 2 percent lidocaine. The active cream for those considered to have mixed pain (n = 133) contained 10 percent ketamine, 6 percent gabapentin, 3 percent diclofenac, 2 percent baclofen, 2 percent cyclobenzaprine, and 2 percent lidocaine. No significant change in average pain score on a 0–10 point numerical rating scale in the preceding week between the active cream and placebo cream in any of the three groups was found. Also, no change was found in secondary outcome measures including the SF-36 or in medication reduction. Of the 399 participants randomly assigned, 396 received treatment and 390 completed the study protocol.

Cohort studies

In a retrospective analysis by Somberg and Molnar (2015a) of a large cohort of patients (n = 2,177) with chronic pain, two topical pain creams were compared to Voltaren (diclofenac sodium) gel before and after treatment using the Visual Numeric Pain Intensity Scale. The study population was 37 percent White, 24 percent Hispanic, 18 percent African American, 7 percent Native American, 3 percent Asian, and 11 percent other ethnicity; mean age of 39 ± 9 years; with chronic extremity, joint, musculoskeletal, neuropathic, or other chronic pain conditions treated with the topical creams. Cream 1 contained flurbiprofen 20 percent, tramadol 5 percent, clonidine 0.2 percent, cyclobenzaprine 4 percent, and bupivacaine 3 percent. Cream 2 contained flurbiprofen 20 percent, baclofen 2 percent, clonidine 0.2 percent, gabapentin 10 percent, and lidocaine 5 percent. Pain intensity scores improved more in the two groups that

___________________

⁴⁹ Risk-of-bias assessment by committee: Low (see Appendix B for more details).

used compounded creams as compared to Voltaren gel as follows: In the cream 1 group (n = 1,141), the score fell from 8.44 ± 1.19 pretreatment to 5.33 ± 2.01 posttreatment, p < .001, and by 2.93 ± 1.58 in the cream 2 group (n = 527) from 8.42 ± 1.27 to 5.50 ± 1.96 posttreatment, p < .001, representing an average decrease in the score of 37 percent and 35 percent. The group using Voltaren gel (n = 509) had a pain score that decreased from 7.93 ± 0.81 to 6.44 ± 1.14, p < .001, or an average decrease of 19 percent, which was significantly less than the other two groups. Of note, there were several limitations in this retrospective analysis, including age differences between groups and in the baseline pain intensity of patients in the Voltaren cream 2 group. In addition, physicians were allowed to choose what medication and duration of medication was prescribed.

The same authors reviewed medical records to determine efficacy of a compounded topical cream containing six or seven drugs: both contained 10 percent ketamine/2 percent baclofen/6 percent gabapentin/4 percent amitriptyline/2 percent bupivacaine/0.2 percent clonidine; one cream contained 2 percent nifedipine. A total of 283 patients with diabetic neuropathy or other chronic pain (no further description), included 78 receiving the six-drug cream and 205 receiving the seven-drug cream. Pain scores decreased by 2.4 (35 percent; p < .001) with the six-drug cream and by 3.0 (40 percent; p < .001) with the seven-drug cream (Somberg and Molnar, 2015b).

Case reports

Alexander and Wynn (2007) described significant pain relief with a 6 percent gabapentin/2 percent amitriptyline/5 percent lidocaine/10 percent ketoprofen transdermal gel in a patient following gall bladder surgery.

In case reports by Safaeian et al. (2016) a topical cream with seven ingredients (diclofenac 5 percent, ibuprofen 3 percent, baclofen 2 percent, cyclobenzaprine 2 percent, bupivacaine 1 percent, gabapentin 6 percent, and pentoxifylline 1 percent), referred to as T7, was used for the treatment of radicular pain. The first patient was a 39-year-old man with 3 weeks of neck pain radiating into the left upper extremity and with 10/10 pain intensity on the NRS. He also complained of weakness, numbness, and tingling in the left arm. He was using an opiate prescription with no relief. A magnetic resonance imaging (MRI) test revealed severe stenosis at the left C6–7 neural foramen and Spurling’s test (a physical maneuver used to assess nerve root radicular pain) was positive on the left. He was prescribed 1–2 grams of T7 to the neck 3–4 times per day. He was also prescribed oral gabapentin and physical therapy, which he did not use. Pain was reduced by 4 months to 6/10 and symptoms improved.

The second patient was a 47-year-old woman with chronic low back pain on opiates and other medications with a history of an L5-S1 laminectomy and fusion presenting with acute back pain (10/10 on the

NRS). There was no significant change noted by MRI and she had a negative straight leg raise test. The patient was prescribed T7, and at 3-month follow-up, she reported pain of 7/10 and 30 percent improvement in symptoms. The third case was a 65-year-old man with a history of chronic low back and radicular pain who also reported relief of pain (5/10 to 3/10) with the compounded cream T7 at 4-month follow-up. In none of these three cases was adverse events reported.

In another case report using a different combination, a 78-year-old woman with refractory postherpetic neuralgia was treated with a compounded topical cream composed of gabapentin 6 percent, ketoprofen 10 percent, amitriptyline 2 percent, and lidocaine 5 percent (Hohmeier and Almon, 2015). She experienced some relief. A new topical cream with gabapentin 6 percent, ketoprofen 10 percent, lidocaine 5 percent, and ketamine 10 percent was prescribed in addition to an oral mucosal topical agent containing gabapentin 10 percent in Orabase. The patient noted further decreases in pain but not complete resolution.

Safety and Adverse Effects

Clinical studies

In the clinical study reported by Somberg and Molnar (2015b) described above, adverse events were reported in 5.7 percent of patients using a six-drug cream and in 3.8 percent of patients using a seven-drug cream including skin irritation, burning sensation at application site, rash, and flushing (Somberg and Molnar, 2015b).

Case reports

In Pomerleau et al. (2014), a 23-year-old man presented to the emergency department with altered mental status after he rubbed an unknown amount of a prescribed compounded topical cream all over his body. The cream contained clonidine 0.2 percent, gabapentin 6 percent, imipramine 3 percent, ketamine 10 percent, lidocaine 2 percent, and mefenamic acid 1 percent. On presentation to the emergency department he was bradycardic with a heart rate of 46 beats/minute, blood pressure 180/87 mmHg, respiratory rate of 21 breaths/minute, and a temperature of 95.6°F. Urine toxicology screen was positive for amphetamines, tetrahydrocannabinol, and tricyclic antidepressants. It was negative for methamphetamine, cocaine, phencyclidine, barbiturates, benzodiazepines, methadone, and opiates. His skin was decontaminated with soap and water. His mental status declined, his trachea was intubated to protect his airway, and he was transported to a tertiary care facility. Upon arrival at the second hospital, he remained bradycardic and hypertensive with a maximum blood pressure of 214/139 mm Hg. He was extubated on day 2 and made a full recovery. Serum concentrations were noted as follows: clonidine 5,200 ng/mL (therapeutic range 0.5–4.5 ng/mL), lidocaine and metabolite

not detected, imipramine 13 ng/mL, and desipramine < 10 ng/mL. No analysis for ketamine was performed.

Sigillito et al. (2003) present an abstract describing a 35-year-old man who presented to the hospital after suffering a seizure. He was unresponsive and had apparently lost all brainstem reflexes. His trachea was intubated, and he was ventilated for 2.5 days. His EEG demonstrated burst suppression. A urine drug screen was positive for benzodiazepines and tricyclic antidepressants. It was later determined that the patient had applied an excessive amount of a compounded cream used for chronic pain. It was estimated that his total exposure dose of ketamine was 900 mg, baclofen 900 mg, amitriptyline 360 mg, lidocaine 900 mg, and ketoprofen 1,800 mg. Ketamine was detected in cerebrospinal fluid by gas chromatography-mass spectrometry. He improved and was discharged on the fourth hospital day.

In a case report, a 22-month-old child developed toxicity consistent with clonidine (apnea and bradycardia) following the ingestion of a topical pain cream containing clonidine 0.2 percent, ketoprofen 7 percent, lidocaine 2.5 percent, prilocaine 2.5 percent, camphor 3 percent, menthol 3 percent, ethoxydiglycol 5 percent, and gabapentin 5 percent. Serum toxicology analysis assessed clonidine levels at 2.6 ng/ml. Given the API’s therapeutic range of 0.5–4.5 ng/mL, the authors noted it was unusual that the patient’s serum concentration of clonidine was comparably low. Neither camphor or ketoprofen were detected, and gabapentin concentration was < 1 mcg/mL (Cates et al., 2018).

In another case report, an 18-month-old was found unresponsive and brought to the emergency department after a parent applied “one pump” (one click) of a compounded topical pain cream to the child’s bottom to treat a diaper rash. In the hospital the child was bradycardic and hypotensive with decreased respiratory effort requiring endotracheal intubation and mechanical ventilation. The topical cream contained ketamine 100 mg, clonidine 2 mg, gabapentin 60 mg, mefenamic acid 10 mg, imipramine 30 mg, and lidocaine 10 mg/pump. The child improved with supportive care. A serum clonidine concentration was 9.2 ng/mL (reference range, 0.5–4.5 ng/mL) and a norketamine level of 41 ng/mL (reporting limit, > 20 ng/mL) (Sullivan et al., 2013).

REFERENCES

Alexander, K., and T. Wynn. 2007. Transdermal gel in the treatment of postoperative pain. International Journal of Pharmaceutical Compounding 11(3):181–184.

Alsherbiny, M. A., and C. G. Li. 2019. Medicinal cannabis-potential drug interactions. Medicines (Basel) 6(3).

Alvi, R., S. Jones, D. Burrows, W. Collins, E. P. McKiernan, R. P. Jones, and P. Bunting. 1998. The safety of topical anaesthetic and analgesic agents in a gel when used to provide pain relief at split skin donor sites. Burns 24(1):54–57.

Anonymous. 2008. Carbamazepine topical preparations. International Journal of Pharmaceutical Compounding 12(2):150.

Aran, S., and D. L. Hammond. 1991. Antagonism of baclofen-induced antinociception by intrathecal administration of phaclofen or 2-hydroxy-saclofen, but not delta-aminovaleric acid in the rat. Journal of Pharmacology and Experimental Therapeutics 257(1):360–368.

Arcion Therapeutics. 2009. Efficacy and safety study of ARC-4558 for management of pain associated with painful diabetic neuropathy. https://ClinicalTrials.gov/show/NCT00695565 (accessed December 17, 2019).

Ardakani, Y. H., and M. R. Rouini. 2007. Improved liquid chromatographic method for the simultaneous determination of tramadol and its three main metabolites in human plasma, urine and saliva. Journal of Pharmaceutical & Biomedical Analysis 44(5):1168–1173.

Attal, N., G. Cruccu, R. Baron, M. Haanpaa, P. Hansson, T. S. Jensen, and T. Nurmikko. 2010. EFNS guidelines on the pharmacological treatment of neuropathic pain: 2010 revision. European Journal of Neurology 17(9):e1113–e1188.

Attia, D. A. 2009. In vitro and in vivo evaluation of transdermal absorption of naproxen sodium. Australian Journal of Basic and Applied Sciences 3(3):2154–2165.

Bachhav, Y. G., and V. B. Patravale. 2010. Formulation of meloxicam gel for topical application: In vitro and in vivo evaluation. Acta Pharmaceutica 60(2):153–163.

Baixauli, F., F. Ingles, P. Alcantara, R. Navarrete, E. Puchol, and F. Vidal. 1990. Percutaneous treatment of acute soft tissue lesions with naproxen gel and ketoprofen gel. Journal of International Medical Research 18(5):372–378.

Barton, D. L., E. J. Wos, R. Qin, B. I. Mattar, N. B. Green, K. S. Lanier, J. D. Bearden, 3rd, J. W. Kugler, K. L. Hoff, P. S. Reddy, K. M. Rowland, Jr., M. Riepl, B. Christensen, and C. L. Loprinzi. 2011. A double-blind, placebo-controlled trial of a topical treatment for chemotherapy-induced peripheral neuropathy: NCCTG trial N06CA. Supportive Care in Cancer 19(6):833–841.

Bassani, A. S., and D. Banov. 2016. Evaluation of the percutaneous absorption of ketamine HCl, gabapentin, clonidine HCl, and baclofen, in compounded transdermal pain formulations, using the Franz finite dose model. Pain Medicine 17:230–238.

Berde, C. B. 1993. Toxicity of local anesthetics in infants and children. Journal of Pediatrics 122(5):S14–S20.

BioDelivery Sciences International. 2017a. Study of clonidine hydrochloride topical gel, 0.1% in the treatment of pain associated with diabetic neuropathy. https://ClinicalTrials.gov/show/NCT02068027 (accessed December 17, 2019).

BioDelivery Sciences International. 2017b. A safety study of clonidine hydrochloride topical gel, 0.1% in the treatment of painful diabetic neuropathy. https://ClinicalTrials.gov/ct2/show/NCT02355158 (accessed December 5, 2019).

BioDelivery Sciences International. 2017c. The efficacy and safety of clonidine hydrochloride topical gel, vs clonidine hydrochloride gel comparator to treat painful diabetic neuropathy. https://ClinicalTrials.gov/ct2/show/NCT02643251 (accessed December 5, 2019).

Boardman, L. A., A. S. Cooper, L. R. Blais, and C. A. Raker. 2008. Topical gabapentin in the treatment of localized and generalized vulvodynia. Obstetrics and Gynecology 112(3):579–585.

Bril, V., J. England, G. M. Franklin, M. Backonja, J. Cohen, D. Del Toro, E. Feldman, D. J. Iverson, B. Perkins, J. W. Russell, and D. Zochodne. 2011. Evidence-based guideline: Treatment of painful diabetic neuropathy: report of the American Academy of Neurology, the American Association of Neuromuscular and Electrodiagnostic Medicine, and the American Academy of Physical Medicine and Rehabilitation. Neurology 76(20):1758–1765.

Bruni, N., C. Della Pepa, S. Oliaro-Bosso, E. Pessione, D. Gastaldi, and F. Dosio. 2018. Cannabinoid delivery systems for pain and inflammation treatment. Molecules 23(10).

Brutcher, R. E., C. Kurihara, M. C. Bicket, P. Moussavian-Yousefi, D. E. Reece, L. M. Solomon, S. R. Griffith, D. E. Jamison, and S. P. Cohen. 2019. Compounded topical pain creams to treat localized chronic pain: A randomized controlled trial. Annals of Internal Medicine 170(5):309–318.

Bryson, E., S. Asbill, and S. Sweitzer. 2014. Skin permeation and antinociception of topical gabapentin formulations. International Journal of Pharmaceutical Compounding 18(6):504–511.

Bryson, E., R. Hartman, J. Arnold, G. Gorman, S. Sweitzer, and S. Asbill. 2015. Skin permeation and antinociception of compounded topical cyclobenzaprine hydrochloride formulations. International Journal of Pharmaceutical Compounding 19(2):161–166.

Butler, P. E. M., P. A. Eadie, D. Lawlor, G. Edwards, and M. McHugh. 1993. Bupivacaine and kaltostat reduces post-operative donor site pain. British Journal of Plastic Surgery 46(6):523–524.

Campbell, C., J. Campbell, W. Schmidt, K. Brady, and B. Stouch. 2009. Abstract: Topical clonidine gel reduces pain caused by diabetic neuropathy: Results of a multicenter, placebo-controlled clinical trial. The Journal of Pain 10(4):S55.

Campbell, C. M., M. S. Kipnes, B. C. Stouch, K. L. Brady, M. Kelly, W. K. Schmidt, K. L. Petersen, M. C. Rowbotham, and J. N. Campbell. 2012. Randomized control trial of topical clonidine for treatment of painful diabetic neuropathy. Pain 153(9):1815–1823.

Cardis, M. A., and H. B. Pasieka. 2016. Safety of topical neuromodulators for the treatment of pruritus. JAMA Dermatology 152(12):1390–1391.

Carr, M. E., Jr., K. Sanders, and W. M. Todd. 1994. Pain relief and clinical improvement temporally related to the use of pentoxifylline in a patient with documented cholesterol emboli—A case report. Angiology 45(1):65–69.

Cates, A. L., S. M. Wheatley, and K. D. Katz. 2018. Clonidine overdose in a toddler due to accidental ingestion of a compounding cream. Pediatric Emergency Care 34(4):e79–e81.

Chang, J. S., Y. H. Tsai, P. C. Wu, and Y. B. Huang. 2007. The effect of mixed-solvent and terpenes on percutaneous absorption of meloxicam gel. Drug Development and Industrial Pharmacy 33(9):984–989.

Chelliah, M. P., Z. Zinn, P. Khuu, and J. M. C. Teng. 2018. Self-initiated use of topical cannabidiol oil for epidermolysis bullosa. Pediatric Dermatology 35(4):e224–e227.

Chen, J., and Y. Gao. 2016. Strategies for meloxicam delivery to and across the skin: A review. Drug Delivery 23(8):3146–3156.

Cline, A. E., and J. E. Turrentine. 2016. Compounded topical analgesics for chronic pain. Dermatitis 27(5):263–271.

Coderre, T. J., Louise Alan Edward Foundation, and McGill University Health Center. 2017. The effects of topical treatment with clonidine + pentoxifylline in patients with neuropathic pain. https://ClinicalTrials.gov/show/NCT03342950 (accessed March 2, 2020).

Cokmez, A., S. Gur, H. Genc, S. Deniz, and E. Tarcan. 2003. Effect of transdermal glyceryl trinitrate and anti-inflammatory gel in infusion phlebitis. ANZ Journal of Surgery 73(10):794–796.

Colvin, A. C., C. F. Wang, M. A. Soens, A. A. Mitani, G. Strichartz, and P. Gerner. 2011. Prolonged cutaneous analgesia with transdermal application of amitriptyline and capsaicin. Regional Anesthesia and Pain Medicine 36(3):236–240.

Cunha, J. P. 2016. Trental (pentoxifylline) side effects. https://www.rxlist.com/trental-side-effects-drug-center.htm#overview (accessed February 28, 2020).

Davidson, E. M., and S. M. Carlton. 1998. Intraplantar injection of dextrorphan, ketamine or memantine attenuates formalin-induced behaviors. Brain Research 785(1):136–142.

de Barros, G. A., H. A. Miot, A. M. Braz, F. Ramos, and M. A. Borges. 2012. Topical (s)ketamine for pain management of postherpetic neuralgia. Brazilian Annals of Dermatology 87(3):504–505.

Del Rio, C., I. Cantarero, B. Palomares, M. Gomez-Canas, J. Fernandez-Ruiz, C. Pavicic, A. Garcia-Martin, M. Luz Bellido, R. Ortega-Castro, C. Perez-Sanchez, C. Lopez-Pedrera, G. Appendino, M. A. Calzado, and E. Munoz. 2018. VCE-004.3, a cannabidiol aminoquinone derivative, prevents bleomycin-induced skin fibrosis and inflammation through PPARGAMMA- and CB2 receptor-dependent pathways. British Journal of Pharmacology 175(19):3813–3831.

Derry, S., P. J. Wiffen, R. A. Moore, and J. Quinlan. 2014. Topical lidocaine for neuropathic pain in adults. Cochrane Database of Systematic Reviews (7):CD010958.

Dualé, C., J. Daveau, J. M. Cardot, A. Boyer-Grand, P. Schoeffler, and C. Dubray. 2008. Cutaneous amitriptyline in human volunteers: Differential effects on the components of sensory information. Anesthesiology 108(4):714–721.

Duangjit, S., Y. Obata, H. Sano, Y. Onuki, P. Opanasopit, T. Ngawhirunpat, T. Miyoshi, S. Kato, and K. Takayama. 2014a. Comparative study of novel ultradeformable liposomes: Menthosomes, transfersomes and liposomes for enhancing skin permeation of meloxicam. Biological and Pharmaceutical Bulletin 37(2):239–247.

Duangjit, S., B. Pamornpathomkul, P. Opanasopit, T. Rojanarata, Y. Obata, K. Takayama, and T. Ngawhirunpat. 2014b. Role of the charge, carbon chain length, and content of surfactant on the skin penetration of meloxicam-loaded liposomes. International Journal of Nanomedicine 9:2005–2017.

Durham, M. J., H.S. Mekhjian, J. A. Goad, M. Lou, M. Ding, S.H. Richeimer. 2018. Topical ketamine in the treatment of complex regional pain syndrome. International Journal of Pharmaceutical Compounding 22(2):172–175.

Dworsky, Z. D., R. Bennett, J. M. Kim, and D. J. Kuo. 2017. Severe medication-induced peripheral neuropathy treated with topical doxepin cream in a paediatric patient with leukaemia. BMJ Case Reports 2017:bcr2017219900.

Eidelman, A., J. M. Weiss, I. K. Enu, J. Lau, and D. B. Carr. 2005. Comparative efficacy and costs of various topical anesthetics for repair of dermal lacerations: A systematic review of randomized, controlled trials. Journal of Clinical Anesthesia 17(2):106–116.

Elovic, E. 2001. Principles of pharmaceutical management of spastic hypertonia. Physical Medicine and Rehabilitation Clinics of North America 12:793–816.

Endo Pharmaceuticals. 2015. Lidoderm (lidocaine patch 5%) label. https://www.accessdata.fda.gov/drugsatfda_docs/label/2015/020612s012lbl.pdf (accessed February 27, 2020).

EpiCent Corporation. 2008a. A comparison of EpiCept NP-1 topical cream vs. oral gabapentin in postherpetic neuralgia (PHN). https://ClinicalTrials.gov/ct2/show/NCT00475904 (accessed March 3, 2020).

EpiCent Corporation. 2008b. A study of the efficacy and safety of amitriptyline/ketamine topical cream in patients with diabetic peripheral neuropathy. https://ClinicalTrials.gov/ct2/show/NCT00476151 (accessed March 2, 2020).

Epstein, J. B., M. Grushka, and N. Le. 1997. Topical clonidine for orofacial pain: A pilot study. Journal of Orofacial Pain 11(4):346–352.

Ernst, E., and M. H. Pittler. 2006. Efficacy or effectiveness? Journal of Internal Medicine 260(5):488–490.

Eslamian, L., A. Kianipour, and S. A. R. Mortazavi. 2017. The analgesic efficacy of 5% naproxen gel for pain associated with orthodontic separator placement: A randomized double-blind controlled trial. Anesthesiology and Pain Medicine 7(2):e42708.

FDA (U.S. Food and Drug Administration). 1998. Lidoderm approval package, Appendix 2, part 2. https://www.accessdata.fda.gov/drugsatfda_docs/nda/99/20612_clinphrmr_P2.pdf (accessed February 27, 2020).

FDA. 2018. FDA approves first drug comprised on an active ingredient derived from marijuana to treat rare, severe forms of epilepsy. https://www.fda.gov/news-events/press-announcements/fda-approves-first-drug-comprised-active-ingredient-derived-marijuana-treat-rare-severe-forms (accessed December 17, 2019).

FDA. 2019a. Orange book: Approved drug products with therapeutic equivalence evaluations. https://www.fda.gov/drugs/drug-approvals-and-databases/approved-drug-products-therapeutic-equivalence-evaluations-orange-book (accessed December 17, 2019).

FDA. 2019b. Information on approval of FDA-approved lidocaine ointment. October 30. Available through the National Academies of Sciences, Engineering, and Medicine Public Access File. https://www8.nationalacademies.org/pa/managerequest.aspx?key=HMD-HSP-18-18 (accessed April 6, 2020).

Finch, P. M., L. Knudsen, and P. D. Drummond. 2009. Reduction of allodynia in patients with complex regional pain syndrome: A double-blind placebo-controlled trial of topical ketamine. Pain 146:18–25.

Finnerup, N. B., S. H. Sindrup, and T. S. Jensen. 2013. Management of painful neuropathies. Handbook of Clinical Neurology 115:279–290.

Fourie, L., J. C. Breytenbach, J. Du Plessis, C. Goosen, H. Swart, and J. Hadgraft. 2004. Percutaneous delivery of carbamazepine and selected n-alkyl and n-hydroxyalkyl analogues. International Journal of Pharmaceutics 279(1–2):59–66.

Fresenius Kabi. 2018. Sensorcaine (bupivacaine) label. https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/018304s049lbl.pdf (accessed February 25, 2020).

Gammaitoni, A., R. M. Gallagher, and M. Welz-Bosna. 2000. Topical ketamine gel: Possible role in treating neuropathic pain. Pain Medicine 1(1):97–100.

Gerner, P., G. Kao, V. Srinivasa, S. Narang, and G. K. Wang. 2003. Topical amitriptyline in healthy volunteers. Regional Anesthesia and Pain Medicine 28(4):289–293.

Gerner, P., V. Srinivasa, A. M. Zizza, Z. Y. Zhuang, S. Luo, D. Zurakowski, S. Eappen, and G. Wang. 2006. Doxepin by topical application and intrathecal route in rats. Anesthesia and Analgesia 102(1):283–287.

Gewandter, J. S., S. G. Mohile, C. E. Heckler, J. L. Ryan, J. J. Kirshner, P. J. Flynn, J. O. Hopkins, and G. R. Morrow. 2014. A phase III randomized, placebo-controlled study of topical amitriptyline and ketamine for chemotherapy-induced peripheral neuropathy (CIPN): A University of Rochester CCOP study of 462 cancer survivors. Support Care Cancer 22(7):1807–1814.

Glass, G. G. 2006. Osteoarthritis. Disease-a-Month 52(9):343–362.

Glinn, M. A., A. J. Lickteig, L. Weber, S. Recer, M. Salske, A. Harvey, B. Rappold, J. Stensland, and P. Bell. 2017. Urinary concentrations of topically administered pain medications. Journal of Analytical Toxicology 41(2):127–133.

Golfam, F., P. Golfam, A. Khalaj, and S. S. S. Mortaz. 2010. The effect of topical nifedipine in treatment of chronic anal fissure. Acta Medica Iranica 48(5):295–299.

Gupta, S. K., P. Bansal, R. K. Bhardwaj, J. Jaiswal, and T. Velpandian. 2002. Comparison of analgesic and anti-inflammatory activity of meloxicam gel with diclofenac and piroxicam gels in animal models: Pharmacokinetic parameters after topical application. Skin Pharmacology and Applied Skin Physiology 15(2):105–111.

GW Biosciences. 2018. Epidiolex (cannabidol) oral solution, CX label. https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/210365lbl.pdf (accessed December 17, 2019).

Haderer, A., P. Gerner, G. Kao, V. Srinivasa, and G. K. Wang. 2003. Cutaneous analgesia after transdermal application of amitriptyline versus lidocaine in rats. Anesthesia and Analgesia 96(6):1707–1710.

Hammell, D. C., L. P. Zhang, F. Ma, S. M. Abshire, S. L. McIlwrath, A. L. Stinchcomb, and K. N. Westlund. 2016. Transdermal cannabidiol reduces inflammation and pain-related behaviours in a rat model of arthritis. European Journal of Pain 20(6):936–948.

Harris, R. P., M. Helfand, S. H. Woolf, K. N. Lohr, C. D. Mulrow, S. M. Teutsch, D. Atkins, and Methods Work Group, Third US Preventive Services Task Force. 2001. Current methods of the US Preventive Services Task Force: A review of the process. American Journal of Preventive Medicine 20(Suppl 3):21–35.

Hesselink, J. M. K., and D. J. Kopsky. 2013. Treatment of chronic regional pain syndrome type 1 with palmitoylethanolamide and topical ketamine cream: Modulation of nonneuronal cells. Journal of Pain Research 6:239–245.

Heyneman, C. A. 1995. Topical nonsteroidal antiinflammatory drugs for acute soft tissue injuries. Annals of Pharmacotherapy 29(7–8):780–782.

Hiom, S., G. K. Patel, R. G. Newcombe, S. Khot, and C. Martin. 2014. Severe postherpetic neuralgia and other neuropathic pain syndromes alleviated by topical gabapentin. British Journal of Dermatology 173(1):300–302.

Ho, K. Y., B. K. Huh, W. D. White, C. C. Yeh, and E. J. Miller. 2008. Topical amitriptyline versus lidocaine in the treatment of neuropathic pain. Clinical Journal of Pain 24(1):51–55.

Hohmeier, K. C., and L. M. Almon. 2015. Topical and intranasal analgesic therapy in a woman with refractory postherpetic neuralgia. Case Reports in Medicine 2015:392874.

Hong, J. M., H. J. Lee, A. R. Cho, J. S. Baik, D. W. Lee, Y. T. Ji, K. C. Yoo, and H. K. Kim. 2016. Pretreatmet with 5% lidocaine patch reduces cannula-induced and propofol-induced pain: A randomized, double-blind, placebo-controlled study. Korean Journal of Anesthesiology 69(5):468–473.

Hospira Inc. 2011. Marcaine label. https://www.accessdata.fda.gov/drugsatfda_docs/label/2012/018692s015lbl.pdf (accessed December 17, 2019).

Huang, C. T., C. H. Tsai, H. Y. Tsou, Y. B. Huang, Y. H. Tsai, and P. C. Wu. 2011. Formulation optimization of transdermal meloxicam potassium-loaded mesomorphic phases containing ethanol, oleic acid and mixture surfactant using the statistical experimental design methodology. Journal of Microencapsulation 28(6):508–514.

IOM (Institute of Medicine). 2011. Relieving pain in America: A blueprint for transforming prevention, care, education, and research. Washington, DC: The National Academies Press.

Jain, S. K., Y. Gupta, A. Jain, and S. Amin. 2008. Elastic liposomes bearing meloxicam-betacyclodextrin for transdermal delivery. Current Drug Delivery 5(3):207–214.

Jaller, J. A., and G. Yosipovitch. 2018. Successful treatment of epidermal nevus-associated pruritus with topical ketamine-amitriptyline-lidocaine. Acta Dermato Venereologica 98(1):121–122.

Jantharaprapap, R., and G. Stagni. 2007. Effects of penetration enhancers on in vitro permeability of meloxicam gels. International Journal of Pharmaceutics 343(1–2):26–33.

Jiang, Q., J. Wang, P. Ma, C. Liu, M. Sun, Y. Sun, and Z. He. 2018. Ion-pair formation combined with a penetration enhancer as a dual strategy to improve the transdermal delivery of meloxicam. Drug Delivery and Translational Research 8(1):64–72.

Jonkman, K., A. Dahan, T. van de Donk, L. Aarts, M. Niesters, and M. van Velzen. 2017. Ketamine for pain. F1000Research 6, F1000 Faculty Rev-1711.

Keyes, P. D., J. M. Tallon, and J. Rizos. 1998. Topical anesthesia: Current indications, options, and evidence in the repair of uncomplicated lacerations. Canadian Family Physician 44(OCT):2152–2156.

Khaledifar, B., M. Y. A. Mahmoudi, and M. Mobasheri. 2015. A double-blind randomized trial comparing the effectiveness and safety of nifedipine and isosorbide dinitrate in chronic anal fissure. Malaysian Journal of Medical Sciences 22(5):42–49.

Khansari, M., M. Sohrabi, and F. Zamani. 2013. The useage of opioids and their adverse effects in gastrointestinal practice: A review. Middle East Journal of Digestive Diseases 5(1):5–16.

Khurana, S., P. M. Bedi, and N. K. Jain. 2013a. Preparation and evaluation of solid lipid nanoparticles based nanogel for dermal delivery of meloxicam. Chemistry and Physics of Lipids 175–176:65–72.

Khurana, S., N. K. Jain, and P. M. Bedi. 2013b. Development and characterization of a novel controlled release drug delivery system based on nanostructured lipid carriers gel for meloxicam. Life Sciences 93(21):763–772.

Khurana, S., N. K. Jain, and P. M. Bedi. 2013c. Nanoemulsion based gel for transdermal delivery of meloxicam: Physico-chemical, mechanistic investigation. Life Sciences 92(6–7):383–392.

Kiani, J., F. Sajedi, S. A. Nasrollahi, and F. Esna-Ashari. 2015a. A randomized clinical trial of efficacy and safety of the topical clonidine and capsaicin in the treatment of painful diabetic neuropathy. Journal of Research in Medical Sciences 20(4):359.

Kiani, J., S. A. Nasrollahi, F. Esna-Ashari, P. Fallah, and F. Sajedi. 2015b. Amitriptyline 2% cream vs. capsaicin 0.75% cream in the treatment of painful diabetic neuropathy (double blind, randomized clinical trial of efficacy and safety). Iranian Journal of Pharmaceutical Research 14(4):1263–1268.

Kim, S. Y. 2013. Efficacy versus effectiveness. Korean Journal of Family Medicine 34(4):227.

Kopsky, D. J., and J. M. Keppel Hesselink. 2011. Multimodal stepped care approach involving topical analgesics for severe intractable neuropathic pain in CRPS type I: A case report. Case Reports in Medicine 2011:319750.

Kopsky, D. J., and J. M. Keppel Hesselink. 2012. High doses of topical amitriptyline in neuropathic pain: Two cases and literature review. Pain Practice 12(2):148–153.

Kopsky, D. J., and J. M. Keppel Hesselink. 2017. Topical phenytoin for the treatment of neuropathic pain. Journal of Pain Research 10:469.

Kopsky, D. J., R. Liebregts, and J. M. Keppel Hesselink. 2012. Central neuropathic pain in a patient with multiple sclerosis treated successfully with topical amitriptyline. Case Reports in Medicine 2012:471835.

Kroenke, K., E. E. Krebs, and M. J. Bair. 2009. Pharmacotherapy of chronic pain: A synthesis of recommendations from systematic reviews. General Hospital Psychiatry 31(3):206–219.

Kuhn, M., S. O. P. Rossi, J. L. Plummer, and J. Raftos. 1996. Topical anaesthesia for minor lacerations: MAC versus TAC. Medical Journal of Australia 164(5):277–280.

Laferriere, A., R. Abaji, C. Y. M. Tsai, J. V. Ragavendran, and T. J. Coderre. 2014. Topical combinations to treat microvascular dysfunction of chronic postischemia pain. Anesthesia and Analgesia 118(4):830–840.

Largo, R., I. Díez-Ortego, O. Sanchez-Pernaute, M. J. López-Armada, M. A. Alvarez-Soria, J. Egido, and G. Herrero-Beaumont. 2004. Ep2/ep4 signalling inhibits monocyte chemo-attractant protein-1 production induced by interleukin 1β in synovial fibroblasts. Annals of the Rheumatic Diseases 63(10):1197–1204.

Le Uyen, M., S. Baltzley, and A. AlGhananeem. 2018. Gabapentin in elastic liposomes: Preparation, characterization, drug release, and penetration through porcine skin. International Journal of Pharmaceutical Compounding 22(6):498–503.

Lee, H. G., S. K. Grossman, R. Valdes-Rodriguez, F. Berenato, J. Korbutov, Y. H. Chan, M. J. Lavery, and G. Yosipovitch. 2017. Topical ketamine-amitriptyline-lidocaine for chronic pruritus: A retrospective study assessing efficacy and tolerability. Journal of the American Academy of Dermatology 76(4):760–761.

Lehman, J. S., and G. F. Sciallis. 2008. Effective use of topical amitriptyline hydrochloride 2.5% and ketamine hydrochloride 0.5% for analgesia in refractory proctodynia. Journal of Drugs in Dermatology 7(9):887–889.

Leppert, W., M. Malec-Milewska, R. Zajaczkowska, and J. Wordliczek. 2018. Transdermal and topical drug administration in the treatment of pain. Molecules 23(3):681.

Liu, K. S., Y. W. Chen, I. A. Aljuffali, C. W. Chang, J. J. Wang, and J. Y. Fang. 2016. Topically applied mesoridazine exhibits the strongest cutaneous analgesia and minimized skin disruption among tricyclic antidepressants: The skin absorption assessment. European Journal of Pharmaceutics and Biopharmaceutics 105:59–68.

Lodzki, M., B. Godin, L. Rakou, R. Mechoulam, R. Gallily, and E. Touitou. 2003. Cannabidiol-transdermal delivery and anti-inflammatory effect in a murine model. Journal of Controlled Release 93(3):377–387.

Lynch, M. E., A. J. Clark, and J. Sawynok. 2003. A pilot study examining topical amitriptyline, ketamine, and a combination of both in the treatment of neuropathic pain. Clinical Journal of Pain 19(5):323–328.

Lynch, M. E., A. J. Clark, J. Sawynok, and M. J. Sullivan. 2005a. Topical amitriptyline and ketamine in neuropathic pain syndromes: An open-label study. The Journal of Pain 6(10):644–649.

Lynch, M. E., A. J. Clark, J. Sawynok, and M. J. L. Sullivan. 2005b. Topical 2% amitriptyline and 1% ketamine in neuropathic pain syndromes: A randomized, double-blind, placebo-controlled trial. Anesthesiology 103(1):140–146.

Machado, T. C., A. B. Gelain, J. Rosa, S. G. Cardoso, and T. Caon. 2018. Cocrystallization as a novel approach to enhance the transdermal administration of meloxicam. European Journal of Pharmaceutical Sciences 123:184–190.

Mahoney, J. M., V. Vardaxis, J. L. Moore, A. M. Hall, K. E. Haffner, and M. C. Peterson. 2012. Topical ketamine cream in the treatment of painful diabetic neuropathy: A randomized, placebo-controlled, double-blind initial study. Journal of the American Podiatric Medical Association 102(3):178–183.

Mankovsky, T., M. Lynch, A. Clark, J. Sawynok, and M. J. Sullivan. 2012. Pain catastrophizing predicts poor response to topical analgesics in patients with neuropathic pain. Pain Research and Management 17(1):10–14.

Martin, C. J., N. Alcock, S. Hiom, and J. C. Birchall. 2017. Development and evaluation of topical gabapentin formulations. Pharmaceutics 9(3):31.

McCleane, G. J. 2000. Topical doxepin hydrochloride reduces neuropathic pain: A randomized, double-blind, placebo controlled study. Pain Clinic 12(1):47–50.

McCleane, G. 2002. Topical application of doxepin hydrochloride can reduce the symptoms of complex regional pain syndrome: A case report. Injury 33(1):88–89.

McNeil Consumer Healthcare. 2013. Flexeril (cyclobenzaprine HCL) tablets label. https://www.accessdata.fda.gov/drugsatfda_docs/label/2013/017821s051lbl.pdf (accessed December 17, 2019).

Meier, T., G. Wasner, M. Faust, T. Kuntzer, F. Ochsner, M. Hueppe, J. Bogousslavsky, and R. Baron. 2003. Efficacy of lidocaine patch 5% in the treatment of focal peripheral neuropathic pain syndromes: A randomized, double-blind, placebo-controlled study. Pain 106:151–158.

Mion, G., and T. Villevieille. 2013. Ketamine pharmacology: An update (pharmacodynamics and molecular aspects, recent findings). CNS Neuroscience & Therapeutics 19(6):370–380.

Montagna, C. G., L. Turroni, D. Martinelli, and M. C. Orlandini. 1990. Single-blind comparative study of meclofenamic acid gel versus naproxen gel in acute musculoskeletal disorders. Current Therapeutic Research—Clinical and Experimental 47(6):933–939.

Moore, R. A., M. R. Tramer, D. Carroll, P. J. Wiffen, and H. J. McQuay. 1998. Quantitative systematic review of topically applied non-steroidal anti-inflammatory drugs. British Medical Journal 316(7128):333–338.

Moore, R. A., S. Derry, D. Aldington, P. Cole, and P. J. Wiffen. 2015. Amitriptyline for neuropathic pain in adults. Cochrane Database of Systematic Reviews (7):CD008242.

Nadal, A., P. Barcelo, and F. Ingles. 1990. Naproxen (Naprosyn) gel as topical treatment for the acute soft tissue lesions. Clinical Trials Journal 27(4):250–257.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2017. The health effects of cannabis and cannabinoids: The current state of evidence and recommendations for research. Washington, DC: The National Academies Press.

Ngawhirunpat, T., P. Opanasopit, T. Rojanarata, P. Akkaramongkolporn, U. Ruktanonchai, and P. Supaphol. 2009. Development of meloxicam-loaded electrospun polyvinyl alcohol mats as a transdermal therapeutic agent. Pharmaceutical Development and Technology 14(1):70–79.

NIA (National Institute on Aging). 2020. Placebos in clinical trials. https://www.nia.nih.gov/health/placebos-clinical-trials (accessed March 28, 2020).

NICE (National Institute for Health and Care Excellence). 2013. Neuropathic pain—Pharmacological management. https://www.nice.org.uk/guidance/cg173/evidence/full-guideline-pdf-4840898221 (accessed February 21, 2020).

NIDDK (National Institute of Diabetes and Digestive and Kidney Diseases). 2017. LiverTox: Orphenadrine. https://www.ncbi.nlm.nih.gov/books/NBK548850 (accessed December 6, 2019).

Ortega-Varela, L. F., J. E. Herrera, N. L. Caram-Salas, H. I. Rocha-Gonzalez, and V. GranadosSoto. 2007. Isobolographic analyses of the gabapentin-metamizol combination after local peripheral, intrathecal and oral administration in the rat. Pharmacology 79(4):214–222.

Paudel, K. S., D. C. Hammell, R. U. Agu, S. Valiveti, and A. L. Stinchcomb. 2010. Cannabidiol bioavailability after nasal and transdermal application: Effect of permeation enhancers. Drug Development and Industrial Pharmacy 36(9):1088–1097.

Paulis, G., D. Barletta, P. Turchi, A. Vitarelli, G. Dachille, A. Fabiani, and R. Gennaro. 2015. Efficacy and safety evaluation of pentoxifylline associated with other antioxidants in medical treatment of Peyronie’s disease: A case-control study. Research and Reports in Urology 8:1–10.

Perrotti, P., A. Bove, C. Antropoli, D. Molino, M. Antropoli, A. Balzano, G. De Stefano, and F. Attena. 2002. Topical nifedipine with lidocaine ointment vs. active control for treatment of chronic anal fissure: Results of a prospective, randomized, double-blind study. Diseases of the Colon and Rectum 45(11):1468–1475.

Pesko, L. J. 1998. Gel relieves arthritis pain. American Druggist 215(May):56.

Pfizer. 2012. Marcaine, bupivacaine hydrochloride injection, USP label. https://www.accessdata.fda.gov/drugsatfda_docs/label/2012/018692s015lbl.pdf (accessed December 17, 2019).

Pomerleau, A. C., C. E. Gooden, C. R. Fantz, and B. W. Morgan. 2014. Dermal exposure to a compounded pain cream resulting in severely elevated clonidine concentration. Journal of Medical Toxicology 10(1):61–64.

Popescu, A. N., E. Cotenescu, B. Costandache, B. Gherghiceanu, and M. L. Todosi. 2018. The benefits and risks of a tricomponent gel containing piroxicam, lidocaine and cyclobenzaprine, in musculoskeletal injuries in athletes. Farmacia 66(3):541–547.

Poterucha, T. J., S. L. Murphy, M. D. P. Davis, P. Sandroni, R. H. Rho, R. A. Warndahl, and W. T. Weiss. 2013. Topical amitriptyline combined with ketamine for the treatment of erythromelalgia: A retrospective study of 36 patients at Mayo Clinic. Journal of Drugs in Dermatology 12(3):308–310.

Pöyhiä, R., and A. Vainio. 2006. Topically administered ketamine reduces capsaicin-evoked mechanical hyperalgesia. Clinical Journal of Pain 22(1):32–36.

Pricop, C., I. Negru, C. Ciuta, V. Jinga, A. Iliesiu, I. A. Checherita, L. Todosi, D. Radavoi, and M. Jinga. 2016. The efficacy of piroxicam/lidocaine/cyclobenzaprine hydrochloride topical gel in the pain management during extracorporeal shock wave lithotripsy (ESWL). Farmacia 64(5):757–762.

Privitera, M. D. 1997. Topiramate: A new antiepileptic drug. Annals of Pharmacotherapy 31(10):1164–1173.

PubChem. 2020a. Clonidine: Compound summary. https://pubchem.ncbi.nlm.nih.gov/compound/Clonidine (accessed February 21, 2020).

PubChem. 2020b. Gapapentin: Compound summary. https://pubchem.ncbi.nlm.nih.gov/compound/Gabapentin (accessed March 13, 2020).

PubChem. 2020c. Memantine: Compound summary. https://pubchem.ncbi.nlm.nih.gov/compound/4054 (accessed February 26, 2020).

PubChem. 2020d. Pregabalin: Compound summary. https://pubchem.ncbi.nlm.nih.gov/compound/5486971 (accessed March 13, 2020).

Quibell, R., M. Fallon, M. Mihalyo, R. Twycross, and A. Wilcock. 2015. Ketamine. Journal of Pain and Symptom Management 50(2):268–278.

Rabi, J., J. Minori, H. Abad, R. Lee, and M. Gittler. 2016. Topical ketamine 10% for neuropathic pain in spinal cord injury patients: An open-label trial. International Journal of Pharmaceutical Compounding 20(6):517–520.

Ragavendran, J. V., A. Laferriere, W. H. Xiao, G. J. Bennett, S. S. V. Padi, J. Zhang, and T. J. Coderre. 2013. Topical combinations aimed at treating microvascular dysfunction reduce allodynia in rat models of CRPS-I and neuropathic pain. The Journal of Pain 14(1):66–78.

Ragavendran, J. V., A. Laferriere, G. J. Bennett, M. A. Ware, W. Gandhi, K. Bley, P. Schweinhardt, and T. J. Coderre. 2016. Effects of topical combinations of clonidine and pentoxifylline on capsaicin-induced allodynia and postcapsaicin tourniquet-induced pain in healthy volunteers: A double-blind, randomized, controlled study. Pain 157(10):2366–2374.

Reis, G. M. L., and I. D. G. Duarte. 2006. Baclofen, an agonist at peripheral GABAB receptors, induces antinociception via activation of TEA-sensitive potassium channels. British Journal of Pharmacology 149(6):733–739.

Sabetkasai, M., S. Ahang, B. Shafaghi, and M. R. Zarrindast. 1999. Baclofen-induced antinociception and nicotinic receptor mechanism(s). BMC Pharmacology and Toxicology 85(5):247–251.

Safaeian, P., R. Mattie, M. Hahn, C. T. Plastaras, and Z. L. McCormick. 2016. Novel treatment of radicular pain with a multi-mechanistic combination topical agent: A case series and literature review. Anesthesiology and Pain Medicine 6(2):e33322.

Salem, A. E., E. A. Mohamed, H. M. Elghadban, and G. M. Abdelghani. 2018. Potential combination topical therapy of anal fissure: Development, evaluation, and clinical study. Drug Delivery 25(1):1672–1682.

Sandig, A. G., A. C. Campmany, F. F. Campos, M. J. Villena, and B. C. Naveros. 2013. Transdermal delivery of imipramine and doxepin from newly oil-in-water nanoemulsions for an analgesic and anti-allodynic activity: Development, characterization and in vivo evaluation. Colloids and Surfaces B: Biointerfaces 103:558–565.

Sawynok, J., and J. Liu. 2014. Contributions of peripheral, spinal, and supraspinal actions to analgesia. European Journal of Pharmacology 734(1):114–121.

Schmelz, M. 2019. Itch processing in the skin. Frontiers in Medicine 6:167.

Schrader, N. H. B., J. C. Duipmans, B. Molenbuur, A. P. Wolff, and M. F. Jonkman. 2019. Combined tetrahydrocannabinol and cannabidiol to treat pain in epidermolysis bullosa: A report of three cases. British Journal of Dermatology 180(4):922–924.

Scott, M. A., K. J. Letrent, K. L. Hager, and J. L. Burch. 1999. Use of transdermal amitriptyline gel in a patient with chronic pain and depression. Pharmacotherapy 19(2):236–239.

Shafizadeh, M., S. Semnanian, M. R. Zarrindast, and B. Hashemi. 1997. Involvement of gabab receptors in the antinociception induced by baclofen in the formalin test. General Pharmacology 28(4):611–615.

Sigillito, R. J., V. E. Tuckler, K. W. Van Meter, and J. Martinez. 2003. Near fatal accidental transdermal overdose of compounded ketamine, baclofen, amitriptyline, lidocaine, and ketoprofen: A case report. Journal of Clinical Toxicology 41(5):672.

Skavinski, K. A. 2019. Opioid-sparing effects of topical ketamine in treating severe pain from decubitus ulcers. Journal of Pain and Palliative Care Pharmacotherapy 32(2–3):170–174.

Skinner, D. J., J. Epstein, and M. Pappagallo. 2009. Chapter 69—Tramadol. In Current Therapy in Pain, edited by H. S. Smith. Philadelphia, PA: W.B. Saunders. Pp. 508–512.

Smith, G. A., S. D. Strausbaugh, C. Harbeck-Weber, B. J. Shields, J. D. Powers, and D. Hackenberg. 1996. Comparison of topical anesthetics without cocaine to tetracaine-adrenaline-cocaine and lidocaine infiltration during repair of lacerations: Bupivacaine-norepinephrine is an effective new topical anesthetic agent. Pediatrics 97(3):301–307.

Snider, M. E., D. S. Nuzum, and A. Veverka. 2008. Long-acting nifedipine in the management of the hypertensive patient. Vascular Health and Risk Management 4(6):1249–1257.

Solomon, D. H., J. A. Rassen, R. J. Glynn, K. Garneau, R. Levin, J. Lee, and S. Schneeweiss. 2010. The comparative safety of opioids for nonmalignant pain in older adults. Archives of Internal Medicine 170(22):1979–1986.

Somberg, J. C., and J. Molnar. 2015a. Retrospective evaluation on the analgesic activities of 2 compounded topical creams and voltaren gel in chronic noncancer pain. American Journal of Therapeutics 22(5):342–349.

Somberg, J. C., and J. Molnar. 2015b. Retrospective study on the analgesic activity of a topical (TT-CTAC) cream in patients with diabetic neuropathy and other chronic pain conditions. American Journal of Therapeutics 22(3):214–221.

Sommer, B. R., and H. H. Fenn. 2010. Review of topiramate for the treatment of epilepsy in elderly patients. Clinical Interventions in Aging 5:89–99.

Sterne, J., J. Savović, M. Page, R. Elbers, N. Blencowe, I. Boutron, C. Cates, H.-Y. Cheng, M. Corbett, S. Eldridge, H. Ma, S. Hopewell, A. Hróbjartsson, D. Junqueira, P. Jüni, J. Kirkham, T. Lasserson, T. Li, A. McAleenan, B. Reeves, S. Shepperd, I. Shrier, L. Stewart, K. Tilling, I. White, P. Whiting, and J. Higgins. 2019. ROB 2: A revised tool for assessing risk of bias in randomised trials. BMJ 366:I4898.

Sullivan, M. J. L., M. E. Lynch, A. J. Clark, T. Mankovsky, and J. Sawynok. 2008. Catastrophizing and treatment outcome: Differential impact on response to placebo and active treatment outcome. Contemporary Hypnosis 25(3–4):129–140.

Sullivan, R. W., M. Ryzewski, M. G. Holland, and J. M. Marraffa. 2013. Compounded ointment results in severe toxicity in a pediatric patient. Pediatric Emergency Care 29(11):1220–1222.

Thorling, J., B. Linden, R. Berg, and A. Sandahl. 1990. A double-blind comparison of naproxen gel and placebo in the treatment of soft tissue injuries. Current Medical Research and Opinion 12(4):242–248.

Turrentine, J. E., G. Marrazzo, and P. D. Cruz, Jr. 2015. Novel use of patch testing in the first report of allergic contact dermatitis to cyclobenzaprine. Dermatitis 26(1):60–61.

Ujvary, I., and L. Hanus. 2016. Human metabolites of cannabidiol: A review on their formation, biological activity, and relevance in therapy. Cannabis and Cannabinoid Research 1(1):90–101.

Ulrich, S., G. Northoff, C. Wurthmann, G. Partscht, U. Pester, H. Herscu, and F. P. Meyer. 2001. Serum levels of amitriptyline and therapeutic effect in non-delusional moderately to severely depressed in-patients: A therapeutic window relationship. Pharmacopsychiatry 34(1):33–40.

University of Oxford. 2009. Oxford Centre for Evidence-based Medicine—Levels of evidence. https://www.cebm.net/2009/06/oxford-centre-evidence-based-medicine-levels-evidence-march-2009 (accessed March 3, 2020).

University of Rochester and Mae Stone Goode Foundation. 2015. Novel topical therapies for the treatment of genital pain. https://ClinicalTrials.gov/show/NCT02099006 (accessed April 7, 2020).

Ushida, T., T. Tani, T. Kanbara, V. S. Zinchuk, M. Kawasaki, and H. Yamamoto. 2002. Analgesic effects of ketamine ointment in patients with complex regional pain syndrome type 1. Regional Anesthesia and Pain Medicine 27(5):524–528.

Uzaraga, I., B. Gerbis, E. Holwerda, D. Gillis, and E. Wai. 2012. Topical amitriptyline, ketamine, and lidocaine in neuropathic pain caused by radiation skin reaction: A pilot study. Supportive Care in Cancer 20(7):1515–1524.

Vadaurri, V. 2008. Topical treatment of neuropathic pain. International Journal of Pharmaceutical Compounding 12(3):183–190.

Vadivelu, N., E. Schermer, V. Kodumudi, K. Belani, R. D. Urman, and A. D. Kaye. 2016. Role of ketamine for analgesia in adults and children. Journal of Anaesthesiology, Clinical Pharmacology 32(3):298–306.

VanDolah, H. J., B. A. Bauer, and K. F. Mauck. 2019. Clinicians’ guide to cannabidiol and hemp oils. Mayo Clinic Proceedings 94(9):1840–1851.

Waldinger, R., G. Weinberg, and M. Gitman. 2020. Local anesthetic toxicity in the geriatric population. Drugs & Aging 31:1–9.

Wang, X., and L. Black. 2013. Ex vivo percutaneous absorption of ketamine, bupivacaine, diclofenac, gabapentin, orphenadrine, and pentoxifylline: Comparison of versatile cream vs. Reference cream. International Journal of Pharmaceutical Compounding 17(6):520–525.

Wehrwein, P. 2010. The safety of painkillers—Harvard health blog. https://www.health.harvard.edu/blog/the-safety-of-painkillers-20101220915 (accessed December 12, 2019).

Wiffen, P. J., S. Derry, M. P. Lunn, and R. A. Moore. 2013. Topiramate for neuropathic pain and fibromyalgia in adults. Cochrane Database of Systematic Reviews (8):CD008314.

Wong, M. L., L. Fleming, L. E. Robayo, and E. Widerström-Noga. 2019. Utility of the neuropathic pain symptom inventory in people with spinal cord injury. Spinal Cord 58:35–42.

Wrzosek, A., J. Woron, J. Dobrogowski, and J. Wordliczek. 2015. Topical clonidine for neuropathic pain. Cochrane Database of Systematic Reviews 2015(2):CD010967.

Yasaei, R., and A. Saadabadi. 2019. Clonidine. Treasure Island, FL: StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK459124 (accessed December 5, 2019).

Yuan, Y., X. Y. Chen, S. M. Li, X. Y. Wei, H. M. Yao, and D. F. Zhong. 2009. Pharmacokinetic studies of meloxicam following oral and transdermal administration in beagle dogs. Acta Pharmacologica Sinica 30(7):1060–1064.

Zanos, P., R. Moaddel, P. J. Morris, L. M. Riggs, J. N. Highland, P. Georgiou, E. F. R. Pereira, E. X. Albuquerque, C. J. Thomas, C. A. Zarate, Jr., and T. D. Gould. 2018. Ketamine and ketamine metabolite pharmacology: Insights into therapeutic mechanisms. Pharmacological Reviews 70(3):621–660.

Zhang, J. Y., L. Fang, Z. Tan, J. Wu, and Z. G. He. 2009. Influence of ion-pairing and chemical enhancers on the transdermal delivery of meloxicam. Drug Development and Industrial Pharmacy 35(6):663–670.

Zhou, Y., H. Cui, C. Shu, Y. Ling, R. Wang, H. Li, Y. Chen, T. Lu, and W. Zhong. 2015. A supramolecular hydrogel based on carbamazepine. Chemical Communications 51(83):15294–15296.

Zou, S., and U. Kumar. 2018. Cannabinoid receptors and the endocannabinoid system: Signaling and function in the central nervous system. International Journal of Molecular Sciences 19(3).