3

Factors Contributing to Patient Variability

In the workshop’s second session, panelists considered sources of outcome variability that are related to individual patients being treated and what steps can be taken to minimize the impact of these variables. Jennifer Elisseeff, the Morton Goldberg Professor at the Wilmer Eye Institute at Johns Hopkins University, discussed how translational research and clinical trials can provide a better understanding of the mechanisms of action and the potential sources of variability. Joseph Wu, the director of the Stanford Cardiovascular Institute, described the factors that contribute to variability in patient outcomes in cardiovascular stem cell clinical trials. Steve Badylak, a professor of surgery at the McGowan Institute for Regenerative Medicine at the University of Pittsburgh, showed how unacceptable variability can result in acceptable outcomes. Flagg Flanagan, the chief executive officer and chairman of the board of DiscGenics, discussed patient variability in a low back pain trial using an allogeneic cell therapy. The session was moderated by Brian Fiske, the senior vice president for research programs at The Michael J. Fox Foundation for Parkinson’s Research.

UNDERSTANDING THERAPEUTIC MECHANISMS

In order to develop good functional assays, Elisseeff said, it is important to understand a product’s therapeutic mechanism of action in vivo. The composition of cells and tissues is complex, and the translational process and clinical trials help to elucidate a product’s mechanisms of action.

Traditional cell and tissue engineering approaches involve using cells, biological signals (e.g., growth factors), and biomaterials (e.g., scaffolds), alone or in combination, to produce new tissue. The discovery and development of embryonic and adult stem cells energized the field of regenerative medicine, Elisseeff said. Stem cells have the ability to differentiate into many different types of tissues and have a significant capacity for proliferation. Elisseeff raised the question, however, of whether stem cells are a rate-limiting factor in repair, and whether they are the best target to promote tissue repair.

Elisseeff described two examples of how moving to clinical trials helped to identify what was most therapeutically relevant about a product. A clinical trial of hydrogels for cartilage repair found that the standard

paradigm, a three-dimensional matrix scaffold, was not serving as a template for new tissue repair. Rather, it was serving as a way to direct the endogenous wound healing process. This demonstrated that the standard template and scaffold approach might not always be the most relevant approach, she said. And a clinical study of soft tissue fillers for wrinkles presented a unique opportunity to retrieve a biomaterial implant after several months of implantation. In the safety study, the biomaterial was implanted in skin that was scheduled to be removed as part of a tummy tuck procedure. Interestingly, she said, there were tissue-specific immune responses to the biomaterial. This finding spurred her to expand the background information she collected on the immune system and transformed her approach to studying the therapeutic relevance of products. Elisseeff said she has moved from engineering tissue microenvironments toward understanding the tissue immune microenvironment in order to inform future immunoengineering efforts.

The Tissue Immune Environment and Biomaterials

The immune system responds to tissue trauma (e.g., surgery), Elisseeff said, and biomaterials can modify that immune response. Biomaterials are broadly classified as either biological or synthetic. Biological materials, such as the scaffolds derived from the extracellular matrix of tissues, can promote tissue repair.¹ Synthetic materials can provide greater physical control over their properties, but they can also elicit a foreign body response when implanted, leading to tissue fibrosis (the formation of excess tissue). Elisseeff said it is important to consider these factors, but added that biomaterials can also serve as a model for further study.

Elisseeff said that, historically, biomaterials researchers have focused on the innate immune response to biomaterials (including the number of neutrophils and macrophages, two types of white blood cells). She emphasized that the innate and adaptive (or acquired) immune systems are closely linked and said that there is much yet to be understood about the impact of the adaptive immune response on tissue repair and biomaterials responses, especially the role of T cells. She reminded participants that receptors on T cells bind specifically to antigens, which leads to differentiation and the release of a variety of cytokines. She then described some of the characteristics of three of the helper T cell subtypes that are involved in systemic impact, antigen specificity, and memory. Type 1 helper T cells, or Th1 cells, are involved in the immune response to tumors, intracellular pathogens, and self (autoimmunity). Type 2 helper T cells, or Th2 cells, secrete interleukin 4, or IL-4 (and other cytokines), and are involved in

___________________

¹ Extracellular matrix scaffolds are discussed further by Badylak.

allergic responses and immune responses to helminths (parasitic worms). Scientists are also interested in the role that Th2 cells may play in regeneration, Elisseeff said. Type 17 helper T cells, or Th17 cells, secrete interleukin 17 (IL-17) and are involved in autoimmunity and the immune response to extracellular pathogens. Elisseeff suggested that Th17 cells might also play a role in fibrosis and added that while the number of Th17 cells is proportionally small, they can exert substantial effects.

Although biological scaffolds used in tissue engineering are decellularized extracellular matrix, some intracellular proteins do remain, Elisseeff explained. These proteins signal when damage has occurred within the immune system, which can promote repair. When delivering cells, she said, it is important to consider not only the number of live cells being transplanted but also the number of dead cells being delivered and the immune stimulatory impact that those dead cells or remnants might have. Elisseeff and her colleagues found that IL-4-producing T cells (Th2 cells) in the surgical wounds or biomaterials were required to direct macrophages and to create an environment conducive to regeneration (Sadtler et al., 2016). The presence of Th2 cells was found to lead to repair, while the absence of Th2 cells led to a fibrotic response. Systemic changes following the biomaterial implant were also observed, including an increased size of lymph nodes. The impact of these changes is often neglected, she said.

The immune system is accessible for therapeutic manipulation, Elisseeff said. She suggested a regenerative approach that engineers immunotherapies to promote an immune environment that is receptive to the therapeutic product being delivered. Immune cells and the immune environment set the stage for later regenerative processes (e.g., vascularization, stem cell mobilization, tissue matrix growth).

The Tissue Immune Environment and Synthetic Materials

The immune environment is also involved in the foreign body response to synthetic materials. A fibrotic capsule often forms around medical implants such as breast implants, Elisseeff said, and gamma delta T cells and CD4-positive T cells (IL-17-producing T cells) are often found in the tissue around breast implants. She described how, in a mouse model, different types of scaffolds induced IL-17-producing T cells and senescent cells developed. Senescent (growth-arrested) cells are involved in aging, but they also play a role in tumor promotion and tissue repair. The senescence-associated secretory phenotype includes the secretion of a range of cytokines, and Elisseeff suggested that senescent cells could possibly serve as a communication bridge between the stroma (i.e., tissue) and the immune system. Specifically, senescent cells associated with such things as tissue damage, disease, aging, or stress could be communicating with and modulating

the immune system and affecting tissue repair. This process is a potential source of variability across patients, she said.

Understanding Mechanisms of Action

Elisseeff emphasized the importance of carefully evaluating clinical data and understanding the mechanisms of action in order to build the evidence base for scientific conclusions. Early on, small trials are an opportunity to learn about mechanisms, she said. She described one example, an anti-PD-1 checkpoint inhibitor study for the treatment of cancer, in which the first two patients in the first group responded to treatment, but the many patients who followed did not. An analysis of the data revealed that the responders had Lynch syndrome, while the non-responders did not. Had data from the non-responders come in first, Elisseeff said, that trial might have been halted. This illustrates what can be learned from early studies and shows the importance of keeping an open mind when results are not as anticipated, she said.

Leveraging Immune Cells to Promote Regeneration

Elisseeff briefly summarized an immune model of tissue and material responses. She reiterated that the immune response to biological scaffolds appears to involve Th2 cells and leads to tissue regeneration, while the immune response to synthetic scaffolds seems to involve Th17 cells and cell senescence and can lead to fibrosis. There are many patient systemic factors that can contribute to variability in the immune environment for regeneration, including, for example, infection, the microbiome, other injuries, a history of antigen exposure, age, and sex.

Applying these insights about the immune environment, Elisseeff is conducting a phase I clinical trial of an extracellular matrix product that is manufactured in a GMP facility and categorized as a biologic by FDA. The study is assessing cell migration over time and the biological activity of the scaffold and its impact on the immune environment.

A participant asked Elisseeff to expand on the potential impact of the immune response in clinical trials, such as in the response to sham controls or to off-target injection of cells that might be stimulating the immune system and the wound-healing response. The immune response is one probable mechanism for the wound healing that is observed in clinical trials, Elisseeff said. She mentioned research performed on the correlation of a patient’s systemic immune state with the speed of rehabilitation from an orthopedic surgery. For example, if one is fighting an infection at the same time as regeneration efforts are taking place, the speed of surgical healing will be slower. These are factors that are out of researchers’ and

clinicians’ control. Joint areas are particularly challenging as senescent cells are often not cleared, leading to chronic inflammation and fibrosis; conversely, she said, clearing those cells can promote healing. Results from several preclinical animal studies suggest that the limiting factor to healing is not the therapeutic cells but the immune environment, she added.

In closing, Elisseeff reiterated the need to think about the local immune environment that exists and that can be modified with biomaterial scaffolds in order to promote regeneration. It is important to think about the immune environment and how it can affect variability and to consider using the immune system as a therapeutic modality. Collecting more immunological data in clinical trials is also important, she said, and partnerships are needed among surgeons, immunologists, and industry to make the most of the clinical trials being conducted.

OUTCOME VARIABILITY IN CARDIOVASCULAR STEM CELL TRIALS

To date, Wu said, more than 7,000 cardiovascular patients have been treated with adult stem cells as participants in a wide range of randomized clinical trials of regenerative medicine products (Fernández-Avilés et al., 2017). So far, however, the data have shown neutral or minimal benefits, Wu said.

When considering how to design better trials, a key question is why there is such significant variability both across and within trials, Wu said, and he cited the FOCUS-CCTRN trial in which patients received autologous bone marrow cells for the treatment of chronic ischemic heart disease. Overall there was no improvement versus placebo across a set of cardiac clinical outcomes. Results for individual patients, however, were quite heterogeneous (Perin et al., 2012).

Possible Causes of Heterogeneity in Clinical Trials

Wu discussed five possible causes of this heterogeneity that he had observed from his perspective as a cardiologist: poor delivery, poor cellular engraftment, poor dosing, poor patient population, and poor cell type.

• Delivery of cells. Cardiologists deliver stem cells to the heart via injection through catheters, such as through intracoronary injections (into the coronary artery) or transendocardial injections (into the muscle), whereas cardiac surgeons open the thorax to inject stem cells into the heart. Wu pointed out that injection via catheter is semi-non-invasive compared with the invasive thoracic surgical injection of stem cells.

• Cellular engraftment. “Bad” or poor delivery of cells could lead to poor cellular engraftment. Wu referred to a study comparing transendocardial versus intracoronary injection of cells in patients with non-ischemic dilated cardiomyopathy (Vrtovec et al., 2013). Patients were injected with approximately 100 million radio-labeled cells, and the biodistribution was assessed by single-photon emission computed tomographic imaging. After 18 hours, cells delivered via intracoronary injection appeared to be concentrated in the liver and spleen, with minimal retention of infused cells in the heart. In contrast, transendocardial injection led to a much greater retention of injected cells in the heart. After 6 months, patients who had higher rates of retention of injected cells in the heart showed greater improvement in cardiac function. Wu said that the disposition of transplanted cells is not generally tracked, so it is usually not known where they end up in the body.
• Dosing. Clinical trials of cardiac stem cell therapies generally administer a one-time dose of 20 million to 200 million cells (Nguyen et al., 2016). About 95 percent of the cells injected are no longer in the heart within 24 to 48 hours, Wu said, as they leak out to end up in the spleen, liver, and kidney. For example, if 100 million cells are injected, only about 5 million cells are engrafted overall, and 99 percent of those cells will be dead within 4 to 6 weeks. The result is that about 50,000 cells are engrafted in the heart, which is less than 0.0001 percent of the total number of cells in the human heart (2 billion to 3 billion cardiac cells and 4 billion to 5 billion fibroblasts).
• Patient population. Preclinical studies of autologous stem cell infusions are generally performed in young, healthy mice, which, Wu said, are similar in terms of relative age to a 15-year-old human. By comparison, the typical patient in a cardiac clinical study is around age 70 and has multiple comorbidities (e.g., hypertension, diabetes, coronary artery disease, multiple stents, prior history of alcohol use or smoking). Stem cells are most active in early life and have decreasing functionality across the life span. As such, the autologous adult stem cells used in the clinical studies will be very different in quality from the stem cells used in the preclinical studies, Wu said.
• Cell type. Having a patient population with advanced age and multiple comorbidities can lead to poor cell type, as noted above. Adult stem cells from a 70-year-old with multiple diseases do not have ideal regenerative potential. Because of this, Wu said, there has been interest in transitioning to iPSCs derived from PBMCs (discussed by Svendsen in Chapter 2). Wu described this approach as “resetting the clock.”

Challenges and Opportunities for Cardiovascular Stem Cell Trials

“One size does not fit all” when it comes to treating heart disease, Wu said. Heart disease is complex. The cause can be genetic or acquired or secondary to other conditions. Correspondingly, treatment is not as simple as injecting stem cells. Furthermore, in clinical medicine, with some exceptions, chronic non-infectious diseases are generally not treated with a one-time dosing. In addition to these challenges, Wu listed several other issues facing those who conduct cardiovascular stem cell trials, including the fact that cardiac cells do not engraft well. The majority of cells injected into the heart die within 4 to 6 weeks; moreover, injecting too many cells into a patient can result in serious arrhythmia. In cases where improvement is seen as a result of paracrine factors, exosomes, or small molecules, it is important to consider the biological half-lives of the therapeutic agent in vivo. In particular, Wu said, the challenge is to determine how the particular pathway that is being targeted remains activated.

Wu briefly described several examples of how investigators are working to bring cardiovascular stem cell therapies to clinical trials. Menasché and colleagues injected human embryonic stem cell–derived cardiac myocytes during bypass surgery (Menasché et al., 2015). Sawa and colleagues are using cardiac cell sheets of iPSC-derived cardiac myocytes in patients with heart disease (Cyranoski, 2018).

Stanford University and the California Institute for Regenerative Medicine (CIRM) have been conducting a preclinical study of human embryonic stem cell–derived cardiomyocytes, working with FDA to develop a pilot trial in humans. Wu described the recent completion of the Stanford-CIRM preclinical IND-enabling studies, with an anticipated submission of an IND application to FDA in early 2019. Key issues to be addressed include the potential tumorigenicity and immunogenicity of the product and the costs of product development. There is a low risk of tumor formation, Wu said, but the risk must be made as close to zero as possible. Regarding immunogenicity, Wu said that embryonic stem cell–based therapies are allogeneic and will therefore require an immunotolerance regimen. This is a concern because the goal of stem cell therapy is to avoid the immunosuppressive drugs needed for a heart transplant. Wu suggested that many iPSC-based therapies would also be allogeneic because biotechnology companies must have a return on their investment, and it is not cost-effective to make autologous cells for every patient. Therefore, a question for further research is whether universal iPSC lines that lack human leukocyte antigens I and II could be developed to avoid immune rejection. The cost of product development is driven by the high cost of conducting studies in large animals (e.g., pigs) and by the need to demonstrate efficacy in randomized double-blinded studies that require large numbers of patients. Securing funding for these

studies is a challenge, Wu said, and it will be important to consider the cost–benefit analysis of competing therapies, such as adult stem cell therapy, standard medical therapy, or surgical mechanical devices.

Wu said that immunological data are being collected in cardiology trials, including those involving cytokine levels or CD4/CD8/TH17 populations in PBMC samples. In cardiology, many investigational products fail in clinical trials, he said, and, as a result, cardiology clinical trials are generally very large to allow the determination of the effect of the product in light of high rates of placebo effects. Still, Wu said, the efforts should be continued because valuable knowledge is being gained from these clinical trials.

Using Open Data for Therapeutic Development

In an effort to accelerate the discovery and translation of new therapeutics, the Stanford Cardiovascular Institute, with funding from CIRM and the National Institutes of Health, is working to create a biorepository of 1,000 iPSC lines from a diverse population of patients.² DNA sequencing will be performed on the cell lines, and the data will be linked to the Pharmacogenomics Knowledgebase (PharmGKB) to further understanding of how human genetic variation determines drug response phenotypes.³ These studies will be able to incorporate whole-genome sequencing, Wu said, as the price point of sequencing has dropped significantly. Data will also be linked to patient information in the Stanford Translational Research Integrated Database Environment. At the time of the workshop, the team had recruited more than 1,200 patients with a broad range of cardiovascular disease diagnoses and had generated iPSC lines for more than 800 of those patients. An important feature of the biobank, Wu said, is its open sharing resource policy. At the time of the workshop, the biobank had provided more than 2,200 vials to more than 200 investigators across the country.

ACCEPTABLE RESULTS FROM UNACCEPTABLE VARIABILITY

The field of regenerative medicine exists, Badylak said, because traditional approaches have not been sufficient and have not yielded new advancements. Regenerative medicine attempts to alter the body’s default response to injury. That default response, he said, is an inflammatory

___________________

² For more information on the Stanford Cardiovascular Institute, see http://med.stanford.edu/cvi.html (accessed January 3, 2019).

³ For more information on the Pharmacogenomics Knowledgebase, see https://www.pharmgkb.org (accessed January 3, 2019).

process that often results in scar tissue that prevents regeneration of the injured tissue. By definition then, regenerative medicine works by creating an abnormal event, promoting tissue regeneration in the face of the inflammatory response. As such, “normal” criteria or milestones should not be expected or demanded in the process of regenerative medicine, he suggested. Badylak agreed with the previous discussion about the value of embracing variability and said that there is much that can be learned from variability.

Ranges of Biological Variation

As discussed earlier, variability affecting regenerative engineering products can appear in the source material (e.g., cells, scaffolds), in the preclinical test results, and in the outcome measures. Badylak defined variation in biology as

He added that intrinsic biological variation contributes to the imprecision of measurements and must be considered when test results are interpreted (Bruce and Lapsley, 2014).

The interdisciplinary convergence of fields discussed by Ameer and Laurencin (see Chapter 1) needs to include the field of immunology, Badylak said. He also mentioned that studies of regenerating species, such as the axolotl, have shown that if the animals are depleted of macrophages, they cannot regenerate tissue. This demonstrates the importance of immunomodulation in regenerative medicine approaches, he said. Badylak added that he and Elisseeff had recently launched a new journal, Journal of Immunology and Regenerative Medicine, to address some of these issues.⁵ More immunological data should be collected in clinical trials because they would provide an opportunity to learn about the role of the immune system, Badylak said, agreeing with Elisseeff and Wu. For example, he asked, is a high white blood cell count an adverse event, or is it associated with healing? He noted that there are established ranges of acceptable biological variability for the results of laboratory tests such as a complete blood

___________________

⁴ See https://www.britannica.com/science/variation-biology (accessed December 16, 2018).

⁵ For more information on the Journal of Immunology and Regenerative Medicine, see https://www.journals.elsevier.com/journal-of-immunology-and-regenerative-medicine (accessed December 16, 2018).

count (CBC) or a blood chemistry profile. Different normal ranges are also established based on age and sex.

Extracellular Matrix Preparation and Contributions to Variability

Badylak briefly described his research on the extracellular matrix. “The matrix used to be thought of primarily as something that was structural,” he said, but now “it is understood that it is actually a reservoir of signaling molecules.” The extracellular matrix is a microenvironment created by secreted products of the cells, and it serves to facilitate communication among the cells (e.g., providing cues for migration, attachment, proliferation, spatial organization). Extracellular matrix is obtained by the decellularization of tissues and organs. However, as Elisseeff mentioned, intracellular proteins remain in the matrix after a tissue is decellularized. Badylak said that much of the variability in outcomes associated with some of the 100 or so different extracellular matrix products on the market can be attributed to poor decellularization.

Cells are intimately associated with the matrix, Badylak explained, and any change in the microenvironment of the matrix (e.g., mechanical force) will affect the cells. The cell will then react via intracellular pathways which alter the transcriptome and the subsequent secretome, which results in alterations to the matrix. Badylak said that this interaction between cells and their extracellular matrix is called dynamic reciprocity (Bornstein et al., 1982). He emphasized that cells change and respond to nearly everything in their microenvironment, including the manufacturing process they are subject to during the creation of regenerative engineering products. It is important to understand what those changes are, and whether they can be manipulated to achieve a therapeutic goal. The extracellular matrix is composed of more than 800 separate protein molecular components, Badylak said, and each extracellular matrix preparation is bound to be slightly different, which contributes to overall variability of the product. The question then is, Which molecular components are causing certain biologic effects of interest?

“Unacceptable” Variability with “Acceptable” Outcomes

Badylak offered two clinical examples of how the default tissue healing response can be changed, ultimately generating acceptable outcomes from an unacceptably variable set of circumstances. He said that preclinical studies in animals had helped to uncover how the implantation of biomaterial had led to the development of tendon, cartilage, or muscle rather than scar tissue. Echoing Elisseeff, he said that the immune response plays an important role in how the biomaterial is received by

the body and that the biomaterial has molecules that can modulate the immune response.

Volumetric Muscle Loss

Volumetric muscle loss is the traumatic or surgical loss of skeletal muscle that can result in muscle impairment (Grogan et al., 2011). This can happen, for example, as the result of trauma, tumor ablation, or disease. Muscle tissue can regenerate to some extent, but Badylak said that a loss of more than 20 percent leads to the formation of nonfunctional scar tissue.

Badylak described a study of 13 patients, 8 male and 5 female, ranging in age from 18 to 70 years. Each patient had experienced volumetric muscle loss, and their various losses had resulted from 8 different causes (e.g., sports-related injuries, combat-related injuries, accidents) and occurred at 7 different anatomic sites. On average, the patients had sustained their injury about 4.5 years prior to the study, and all had undergone numerous previous surgeries with unacceptable outcomes. For the study, scar tissue was surgically removed, and one of three different extracellular matrix products was implanted in contact with healthy tissue (Dziki et al., 2016). Imaging and analysis of biopsied tissue indicated that the implantation of the extracellular matrix had promoted new functional muscle fiber formation, Badylak said. Stem cells that were recruited to the site differentiated into muscle cells. Patient functional performance was also assessed, and, on average, patients showed at least 25 percent improvement in strength and function.

These positive outcomes were achieved despite what would conventionally be considered unacceptable variability in several aspects of the trial, Badylak said. “What is unacceptable variability [in this case]?” he asked.

Even with all of this variation, the results in all 13 patients appeared to be good.

Esophageal Cancer

Rates of esophageal cancer have been rising at an accelerated pace, Badylak said. Some patients with gastroesophageal reflux disease will progress to Barrett’s esophagus, and some of those will go on to develop cell

dysplasia and then cancer. The current standard of care for esophageal cancer is the removal of the esophagus, which can be a complicated and painful surgery. Half of the patients experience complications, many have a reduced quality of life due to changes in the stomach as well as related digestive issues, and mortality rates from the surgery range from 2 to 18 percent, Badylak said.

The incidence of precancerous Barrett’s esophagus is the same as the incidence of colon polyps. But while the standard treatment for a colon polyp is removal, the treatment for Barrett’s esophagus is surveillance and the use of drugs that reduce gastric acid secretion. The reason for the difference, Badylak explained, is that removal of the suspect tissue in the esophagus leads to the formation of scar tissue with strictures, or narrowed passages, meaning that the treatment can be worse than the disease in the early stages.

Badylak described a regenerative medicine study performed in patients with advanced esophageal conditions, all with non-operative comorbidities. Badylak and colleagues developed an endoscopic technique to remove the esophageal mucosa, completely denuding the inside of the esophagus. A procedure like this would normally lead to scarring and stricture “100 percent of the time within a couple of weeks,” he said. To prevent this and to foster regeneration, the doctors essentially “wallpapered” the remaining muscle layer of esophagus with a piece of extracellular matrix. A stent was inserted for about a week to hold the matrix in place (Badylak et al., 2011).

Biopsy specimens from the first five patients with mucosal carcinoma showed that the removal of the mucosa and implantation of extracellular matrix results in a remodeled esophageal mucosa consisting of a normal cell population. Badylak said that patients were released from the hospital after about 1 day, compared with 13 days after an esophagectomy. To date, the procedure has been done in 14 patients, and Badylak reported that there has been no recurrence of cancer and excellent outcomes overall.

As was the case with the first example, this study would have been considered to have unacceptable variability by conventional standards, Badylak said. Specifically, there were two different tissue sources, two manufacturing methods, two terminal sterilization methods, two molecular composition profiles, and two mechanical property files. And yet, Badylak said again, there were good results.

A participant observed that a high level of heterogeneity in a clinical trial could be acceptable if the trial was large enough to be able to develop some hypotheses about the role of the heterogeneity or if the treatment was robust enough that it had an effect despite the heterogeneity. Even though the trial was quite small, Badylak said, it was able to show that the esophageal scaffold was successful in regenerating tissue without scarring and strictures despite the presence of the variability in the study. The par-

ticipant asked how much variability is too much—so much that you end up “creating a non-learning situation.” All of the patients in the trial had comorbidities that disqualified them as candidates for surgical interventions and had no other options, Badylak said. Conducting a highly controlled study with a large number of patients would have been ideal, but this small study was set up to be a learning opportunity despite the high variability. Another participant observed that data in this field accrue slowly, as most of the trials discussed had small populations. She suggested that it would be valuable to develop data registries and collaborations within and across research centers. Badylak agreed and added that many of the patients trying these interventions have failed to respond to traditional therapies and have experienced many different attempts to cure their disease. It is challenging to try to account for all of those variables, he said, but there is much that can be learned by observing and embracing this variability.

Closing Remarks

The field has evolved over the past 25 years, Badylak said. For example, early efforts grew tissue outside the body for implantation, but it became clear that the cells did not survive, prompting researchers to consider paracrine effects. The field is advancing, and there is a need to embrace variability and to try combination approaches, he said. No one group has the full solution to the challenges that have been encountered in the field, Badylak said, but a coordinated effort across the regulatory, patient, and physician communities could help make progress in learning and trying new approaches.

ALLOGENEIC CELL THERAPY FOR DEGENERATIVE DISC DISEASE

DiscGenics is a clinical-stage biotechnology company developing cell therapies for patients with degenerative disc disease, Flanagan said. Low back pain is a global health problem and a leading cause of disability worldwide, affecting about one-quarter of the population. More than $100 billion is spent annually in the United States on the management of low back pain. It is associated with missed work, lost wages, and an addiction to pain medication (including opioids), and there are currently no curative treatments.

One of the major causes of low back and neck pain is degenerative disc disease. The discs between the vertebrae progressively break down, leading to inflammation, breakdown of extracellular matrix, and disc rupture. Flanagan said that degenerative disc disease is very difficult to treat. In the mild-to-moderate stages of disease (when disc height is reduced by about 50 percent) treatment is symptomatic (e.g., patients are given pain medication, physical therapy, steroid injections, ablation, chiropractic and alternative care). Once

in the moderate stages of disease, treatment is essentially “buying time” as degeneration continues, he said. There can occasionally be self-healing, but as the degeneration continues, the only treatment is spinal fusion surgery at a cost of more than $100,000. Flanagan added that once vertebrae are fused, there is increased pressure and wear on the adjacent areas.

Manufacturing an Allogeneic Disc Cell Therapy

DiscGenics has developed an allogeneic regenerative therapy using what it calls discogenic cells. The cells are derived from progenitor cells from adult donor disc tissue, which Flanagan noted is immune-privileged tissue (i.e., it does not mount an immune response to a foreign antigen). He said that the mechanism of action is hypothesized to be both regenerative and anti-inflammatory, and he added that thousands of doses can be manufactured from a single donor.

Flanagan described the two-part cell manufacturing process in this way: in the first part, donor tissue is collected, and intermediate cells are generated, analyzed for purity and potency, and cryopreserved, while the second stage of manufacturing involves culturing the intermediate cells, generating the discogenic cells, finalizing the formulation of the product, and cryopreserving the final allogeneic cell therapy product. These cells must pass a second checkpoint for potency, safety, identity, and purity. Product lots and subsets of lots are also tested upon release to determine their stability. Product manufactured with tissue from more than 20 donors has so far shown a very consistent release profile, Flanagan said, which suggests that the goal of producing a cryopreserved off-the-shelf, allogeneic cell therapy product is feasible. In vitro assays have shown the discogenic cells to be anti-inflammatory (in a two-way mixed leukocyte reaction) and able to produce extracellular matrix. The cells are also non-tumor-forming and robust through cryopreservation.

Translation to the Clinic

DiscGenics’s first product candidate is Injectable Disc Cell Therapy (IDCT).⁶ It is an allogeneic non-invasive cell therapy based on DiscGenics’s proprietary cell manufacturing and delivery vehicle technology. Flanagan said that IDCT is cost-effective relative to surgical fusion; furthermore, it is non-surgical and easy to administer in a treatment room, and it is available as an off-the-shelf frozen product.

___________________

⁶ For more information, see https://www.discgenics.com/research-and-development/#idct (accessed December 16, 2018).

Flanagan provided a brief overview of the preclinical safety and efficacy studies of IDCT in four different animal models. No toxicity or safety issues were identified, he said. Concerning the product’s efficacy, there was consistent bioactivity across all the donors, he said, and an improvement in disc height and a normalization of disc tissue were demonstrated.

The first clinical trial of IDCT has now been initiated, Flanagan reported, with the launch of a phase I/II trial in the United States in the first quarter of 2018.⁷ IDCT is being reviewed by FDA as a biologic. Flanagan said that DiscGenics worked very closely with FDA during the preclinical stage and received approval of the IND within 28 days. This is a prospective, multicenter, double-blind, vehicle- and saline-controlled safety and preliminary efficacy study of a single injection in 60 patients. In addition, a clinical trial notification has been filed with the Japanese regulatory authority to begin a clinical trial there.⁸

Patients enrolled in the U.S. study have painful, single-level lumbar degenerative disc disease (3 to 7 on a Modified Pfirrmann scale; pain for 6 months; unresponsive to conservative care for 3 months). A key design element of the study to help control variability, Flanagan said, was “finding the right patient.” To be enrolled, patients had to have single-level disc degeneration (many patients have multi-level disc degeneration). This was done because the primary endpoint used was a visual analog score of pain before and after treatment. To assess the impact of the treatment it was important to screen out patients with other sources of pain, he explained. Clinical screening and radiographic evaluation were done to exclude patients who also had adjacent-level disc pain, facet pain, radiculopathy, muscular pain, or obvious psychological issues associated with back pain.

Fiske asked whether the use of a placebo, which may have variable effects from patient to patient, can affect outcomes in a clinical trial. The clinical trial of IDCT includes a vehicle control and a saline control, Flanagan said. The saline injection is known to provide some clinical improvement in patients with degenerative disc disease for a 1- to 2-week period, and FDA stressed the importance of including a normal saline control. In other disc trials, he noted, there have been cases where the investigational product did not show improvement over the saline control. He hypothesized that the dual anti-inflammatory and regenerative mechanisms of action of IDCT would more than compensate for the normal saline effect

___________________

⁷ClinicalTrials.gov Identifier NCT03347708. See https://clinicaltrials.gov/ct2/show/NCT03347708 (accessed December 16, 2018).

⁸ The Japanese Pharmaceuticals and Medical Devices Agency approved DiscGenics’s clinical trial notification application for IDCT in December 2018. For more information, see https://www.discgenics.com/news-posts/2018/12/10/discgenics-receives-approval-from-japanese-pmda-to-begin-clinical-evaluation-of-idct (accessed January 14, 2019).

and would reduce pain over the long term (compared to the short-term effect of saline).

Another workshop participant suggested that it is important to better understand the heterogeneity of patient response, particularly from a genetic perspective. Anecdotal evidence suggests that some people generate tissue better than others, and a genome-wide association study to identify the genetic signatures of better tissue regenerators could help facilitate precision medicine, she added. DiscGenics attempted to conduct genetic and biomarker analysis as part of the 60-patient clinical trial, Flanagan said, but the cost was prohibitive. As a privately funded company, DiscGenics must focus on meeting milestones to demonstrate progress to its investors who are interested in a commercial solution, he said, and it must collect relevant data for submission to FDA. Fiske noted that conducting genetic analysis at scale would require bringing together many datasets, and he said it would be important to establish which outcome should be assessed in order to identify an association.

The next steps for IDCT, Flanagan said, are to scale up manufacturing while ensuring product consistency across donors and to identify additional patients for enrollment in the trial. He said the DiscGenics technology is unique in that cells collected from disc tissue are driven back to a progenitor state to then differentiate into disc tissue in a compartmentalized space. He referred to the cost and challenges of scale-up and manufacturing, but he said he was optimistic about bringing the product to market. Elisseeff cautioned that there is not yet evidence that the disc cells will make disc tissue in a human. She also clarified that the disc space is not fully immunoprivileged but rather is slow to clear foreign antigen. Cell death in that space results in senescent cells that the immune system does not clear. Injecting cells therapeutically, even if a large percentage of those cells die, can still induce healing and clearance of senescent cells if some cells survive. She reiterated the need to be open to the possibility of different mechanisms when analyzing the results of clinical trials, even if preclinical studies show the growth of the tissue. Flanagan agreed and added that their preclinical studies in animals offer hope that this could potentially be a curative option.

This page intentionally left blank.