Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
I Current Risks of Disease Transmission
Blood and Blood Components: How Safe Are They Todays Kenrad E. Nelson Control of transfi~sion-transmitted infections is pretty much of a success story, at least in the United States, although in the rest of the world the problem of transfi~sion-transmitted human immunodef~ciency virus (HIV) and other viruses such as hepatitis C virus (HCV) is still one of some magnitude. I want to review some of the data collected in the United States over the 10 or 12 years since the risk from HIV was recognized, then briefly describe some recent data I have collected in Thailand, and conclude by reviewing my own data and those of others on transmission of hepatitis and other diseases, viral and bacterial. Blood banks have used three broad strategies to control transfusion related infections. First and most important are efforts to exclude donors whose behaviors might put them at high risk for HIV infection or hepatitis, such as drug users, homosexual or bisexual men, or heterosexuals with high-risk sexual partners. Second is active recruitment of low-risk repeat donors. Seventy to 80 percent of donors in the United States are now repeat donors Finally, and what the public focuses on most, is serologic screening, which in fact may be the least important of the preventive measures. A paper published in Transfusion2 by Michael Busch from the Irwin Memorial Blood Bank in San Francisco showed that during the late 1970s and early 1980s, transfi~sion-transmitted HIV infections in San Francisco were a substantial problem. Busch's estimate is that roughly 5,000 people in San Francisco had transfusion transmitted HIV infections and that more than 2,000 developed acquired immunodeficiency syndrome (AIDS) from transfusions This chapter was originally presented to the Forum in January 1995, but was updated by the author in October 1996 to reflect some important new developments. 2Busch, MP, MJ Young, SM Samson, JW Mosley, JW Ward, HA Perkins (1991). Risk of human immunodeficiency virus (HIV) transmission by blood transfusions before the implementation of HIV-l antibody screening. The Transfusion Safety Study Group. Transfusion, 31(1): Al 1. 3
4 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK from a single blood bank. Expressed as prevalence of HIV among donors in 1982 and 1983, more than 1 percent of all donors were HIV positive. However, long before screening was begun in 1985, this had been reduced six- or sevenfold by excluding donors who had a history of male-male sex. After the serologic test was instituted, it was widely believed that with exclusion of high-risk donors and screening of all donations for antibodies to HIV, the blood supply was very safe. In 1988, however, John Ward and colleagues Tom the Centers for Disease Control (CDC) published a report in the New England Journal of Medicines describing 13 people who were apparently infected with HIV from screened blood and had subsequently developed AIDS. Some risk of transmission was obviously still present. Three strategies have been used to evaluate the residual risk from screened blood. The first is to followup the recipients, people who have been transfused with screened blood, to see whether or not they develop an HIV infection. That was the approach that we took at Johns Hopkins. The second approach has been to test the blood in the blood banks by a more sensitive technique such as DNA PCR (deoxyribonucleic acid polymerase chain reaction). This was done by Vyas and Busch in San Francisco. The third technique uses mathematical models to estimate the infectious window period prior to seroconversion and the probability that an infected donor would be in the window period. A study using the first of these methods involved three hospitals: my own institution (Johns Hopkins in Baltimore) and St. Luke's Episcopal and Texas Methodist, both in Houston.4 The study's short name, FACTS, stands for Frequency of Agents Communicable by Transfusion Study. The initial objective was to evaluate the effectiveness of HIV screening of the blood supply. Evaluating the risks of human T-lymphotropic virus (HTLV) I/II and hepatitis transmission by transfusion was added later. The patients were adult cardiac surgery patients operated on in one of these three hospitals. Evaluation of transfusion-related infections in this group had several advantages. First, they were very heavily transfused. Second, they had a very low risk for HIV infection by any other means. Third, over 90 3Ward, JW, SD Holmberg, JR Allen, DL Cohn, SE Critchley, SH Kleinman, BA Lenes, O Ravenholt, JR Davis, MG Quinn, et al. (1988). Transmission of human immunodeficiency virus (HIV) by blood transfusions screened as negative for HIV antibody. New England Journal of Medicine, 318(8): 473~78. 4Nelson, KE, JG Donahue, A Munoz, ND Cohen, PM Ness, A Teague, VA Stambolis, DH Yawn, B Callicott, H McAllister, et al. (1992). Transmission of retroviruses from seronegative donors by transfusion during cardiac surgery. A multicenter study of HIV-1 and HTLV-I/II infections. Annals of Internal Medicine, 117(7): 550559.
CURRENT RISKS OF DISEASE TRANSMISSION 5 percent were still alive 6 months after the surgery, and the follow-up was very good. Blood samples were obtained prior to the operation. During the initial hospitalization, medical records were reviewed. A postoperative sample was obtained 6 to 8 months after surgery, along with a questionnaire that asked about high-risk behavior for HIV or hepatitis infection. Evidence of hepatitis or other signs of a transfusion-related illness were sought, and any history of additional postoperative transfusions was obtained. Roughly 12,000 people enrolled in the study, of whom almost 80 percent or 9,294 were transfused. Very importantly, in this study we also followed the remaining 2,200 people who underwent a cardiac surgical procedure, which normally requires a transfusion, but who were not transfused. They served as an important internal control, particularly for hepatitis, because they can be used to estimate background rates of hepatitis in hospitalized cardiac surgical patients. A total of 120,000 units were transfused; on average, each transfused patient received 13 units. Among the recipients of these units, we found two people who seroconverted to HIV positive, i.e., 0.0017 percent, or a point estimate of 1 infection per 60,000 units transfused. Neither patient had risk behavior for HIV infection other than transfusion. All of the donors of the two seroconverters were located, and they included one who acknowledged male- to-male sexual relations, although he had denied it at the time he donated blood; this donor seroconverted after donating blood. We found another donor who seroconverted to HIV positive after donating blood given to the second HIV-positive cardiac case. Thus, we were fairly confident that both of these cases were transfusion acquired. Screening of donors for HTLV I antibodies was instituted in 1988, while our study was still under way. Therefore, we were able to compare directly the impact of donor screening on postoperative HTLV incidence. There were a total of seven transfusion-associated HTLV I or II infections, only one of which occurred after donor screening was begun. This was a patient with HTLV II infection. The point estimate for HTLV II positivity was 1/67,000 units. Subsequent data have shown that the HTLV I screening test is not quite as sensitive in the detection of lITLV II infection. In fact, there is now some debate about the wisdom of adding HTLV II-specif~c antigens to the HTLV I screening in order to improve the sensitivity of the serological screening for HTLV II. However, HTLV II is not as clearly associated with human illness as HTLV I. The second of the three approaches to estimating the risk of HIV transmission from HIV antibody-screened blood was taken by G.N. Vyas and colleagues at the Irwin Memorial Blood Center in San Francisco. They pooled blood samples from 50 donors and did cultures and a polymerase chain
6 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK reaction test (PCR) in an attempt to detect the presence of HIV directly.5 They identified one positive sample in the first pool, but none in 1,900 subsequent pools. Therefore, their estimate was 1 in 160,000 units (which is not statistically different Tom the 1/60,000 estimate of FACTS). Another approach to estimating the risk was recently reported by Lyle Petersen and colleagues at the CDC.6 They first identified repeat donors who had seroconverted. Then they did a look-back at outcomes for patients transfused with those donations. They found 561 repeat donors who had seroconverted and whose units had been transfused. Information was obtained on 182 recipients, of whom 36 had seroconverted to HIV positive since transfusion. When they modeled the interval between the last negative donation and the first positive donation to estimate when seroconversion might have occulted, the curve that best fit the data showed a median interdonation interval of 45 days for the donors whose antibody-screened blood led to seroconversion in recipients. Based upon this mathematical model, the group at CDC hypothesized that the median interval Mom the point at which a donor becomes infectious until the enzyme-linked immunosorbent assay (ELISA) for antibodies to HIV-1 and HIV-2 become positive was 45 days. In a recent study published in Transfusion, 7 Busch et al. investigated whether current screening procedures have significantly shortened this seronegative window period. This study used donors who were PCR positive but antibody negative and who were subsequently found to have seroconverted, i.e., became positive on tests for the presence of antibodies to HIV. The interval Tom PCR positivity to antibody positivity could be estimated with these data. The PCR-positive sample was then tested with a number of more sensitive tests for the detection of HIV antibodies, antigens, or nucleic acid that are now available. This study showed that one-third to one-half of these seroconverters could be detected earlier with the current, more sensitive ELISAs: 80 percent were RNA (ribonucleic acid) PCR positive, and roughly 60 percent were p24 antigen-positive. 5Busch, MP, BE Eble, H Khayam-Bashi, et al. (l991). Evaluation of screened blood donations for human immunodeficiency virus type 1 infection by culture and DNA amplification of pooled cells. New England Journal of Medicine, 325: 2-S. 6Petersen, LR, GA Satten, R Dodd, M Busch, S Kleinman, A Grindon, B Lenes (1994). Duration of time from onset of human immunodeficiency virus type 1 infectiousness to development of detectable antibody. The HIV Seroconversion Study Group. Transfusion, 34(4): 283-289. 7Busch, MP, LL Lee, GA Satten, DR Henrard, H Farzadegan, KE Nelson, S Read, RY Dodd, LR Petersen (1995). Time course of detection of viral and serologic markers preceding humar~ immunodeficiency virus type 1 seroconversion: Implications for screening of blood and tissue donors. Transfusion, 35(2): 91-97.
CURRENT RISKS OF DISEASE TRANSA'IISSION 7 Taking the time to onset of infectiousness as marked by PCR positivity, the second-generation antibody assays (ELISA) reduced the seronegative window period by 6 days, the third generation assays reduced the window by about 19 days, and PCR and p24 antigen assays cut it even further. At present the seronegative window period is probably about 20 to 25 days or shorter. Because that is an average, some people will have longer windows and some will have some shorter, but there has been a substantial reduction in the window period with newer assays. A study that is ongoing, the Retrovirus Epidemiology Donors Study (REDS), fimded by the National Heart, Lung and Blood Institute, began in 1989 and is to continue until 1998. Its purpose is to monitor the safety of the nation's blood supply by studying donors who test positive, using a case control methodology with seronegative donors as controls. Five blood centers are involved in REDS, and so far the study has included 2.3 million donations from almost 586,000 multiple donors during a three-year period. Again, using data from the multiple donors that seroconvert, one can estimate the incidence and the length of the window period. If the HIV-1 incidence is multiplied by the length of the window period in repeat donors, one can estimate the rate of false-seronegative donations during the window.8 In a paper that Michael Busch presented at the National Institute of Health (NIH) Consensus Development Conference in January 1995, he reported 33 cases of HIV infection in repeat donors in 822,000 person years, for an overall incidence rate of 4 per 100,000 person years.9 Using a window period estimate of 22 days yields an estimated risk of transfusion during the window period of 2.4 per million (1 in 416,000~. The HIV prevalence rates in first- time donors are higher, which could also mean that HIV incidence may be higher among first-time donors, and therefore they are at higher risk of being in the window period. Despite this possible underestimation of the risk using this method, the estimate is probably fairly accurate. We now have estimates of the risk of HIV infection Tom screened blood from four different studies. One was a follow-up of recipients in which the estimated risk was 1 in 60,000. A study with PCR estimated a risk of about 1 in 150,000. A mathematical model that was published in the New England ~Schreiber, GB, MP Busch, SH Kleinman, JJ Korelitz (1996). The risk of transfusion- transmitted viral infections. New England Journal of Medicine, 334: 1685-1690. 9Busch, MP (1995). Incidence of infectious disease markers in blood donors: Implications for residual risk of viral transmission by transfusion (abstract). NIH Consensus Development Conference on Infectious Disease Testing for Blood Transfusions, January 9-11, 1995, Bethesda, Maryland.
8 BLOOD AND PLOOD PRODUCTS: SAFETY AND RISK Journal of Medicine,~° using all the Red Cross data for the 2 or 3 years after screening began, estimated a risk at 1 in 153,000. Finally, the mathematical model based on the Red Cross data estimates a risk of transfi~sion-transmitted lIIV of 1 in 420,000 blood donors.)' The actual risk of HIV infection in HIV-seronegative donors is somewhat higher than the last figure, probably about 1 in 350,000, but other screening tests, namely the hepatitis B virus (HBV) core antibody test, the alanine aminotransferase (ALT) test, and the serologic test for syphilis, actually function as surrogate markers for HIV, eliminating some HIV-infected but antibody-negative donors because of positivity on these other screening tests. In fact, one of the reasons for the recent NIH Consensus Development Conference recommendation to retain the HBV core antibody test was for its value as a surrogate marker for HIV (as well as a direct marker for HBV infection). However, the HBV core antibody test was originally developed as a surrogate marker for hepatitis non-A, non-B. Nearly every year since 1985 new tests and screening procedures have been introduced in an effort to reduce the risk of transfi:sion-transmitted infection from blood and blood products. More effective serologic screening, recruiting more repeat donors, screening rigorously by questionnaire, and confidential unit exclusion all have been used to combat the risk of disease transmission and have resulted in a four- or fivefold decrease in the prevalence of HIV infection in donors. Some people feel, and the REDS data would support this, that perhaps p24 antigen screening of the donors would further reduce the residual small risk of transfusion-transmitted HIV. I would also like to describe some studies I have been involved with in a blood bank in the city of Chiang Mai, in northern Thailand, where the risk of infection from blood transfusion is very much higher than it is in the United States. In Thailand the rate of HIV-seropositive donors over the last six years has been about 3 to 4 percent. In this setting of a very high seroprevalence in blood donors, most of whom are neither male homosexuals nor intravenous drug users, the exclusion of these high-risk donors had much less impact than it has had in the United States. Therefore, it seemed reasonable to evaluate the effectiveness of screening donor blood for p24 antigen. In testing some 44,000 donors, we found 48 who were p24 antigen positive, 7 of whom did not have HIV-1 antibody and were neutralizable (and therefore were infected with HIV i°Cumming, PD, EL Wallace, JO Schorr, RY Dodd (1989). Exposure of patients to human immunodeficiency virus through the transfusion of blood components that test antibody-negative. New England Journal of Medicine, 321: 941-946. ~ ~Lackritz, EM, GA Satten, J Aberle-Grasse, et al. (1995). Estimated risk of transmission of the human immunodeficiency virus by screened blood in the United States. New England Journal of Medicine, 333: 1 721-1 725.
CURRENT RISKS OF DISEASE TRANSMISSION 9 1 but in the seronegative window period). Ire a two-year period, p24 antigen screening detected 7 more infected donors that would have been missed by standard antibody screening tests. In this study the ratio between antibody positivity (1519 donors) and antigen prevalence in antibody-negative donors (7 donors) was 1:217. If this ratio held true in the United States, where roughly 1 in 10,000 donors is antibody positive and there are about 13 million donors per year, we might exclude 6 additional HIV-infected donors with p24 antigen screening of all blood donations. The current estimate of the number of HIV-infected donors in the United States can be calculated from the risk data presented above. If the risk is 1 in 420,000 units and 13 million units per year are collected, then we might expect roughly 50 HIV infections per year. Reducing that number by 6 to 10 would constitute a 15 to 20 percent reduction. In Thailand the cost of p24 antigen testing is only a little over $4,000 for each transfusion infection prevented. The comparable figure in the United States would be somewhere in the $5 million to $10 million range to prevent one transfusion-transmitted HIV infection. There is a second important virus that has been known to be a transfusion risk for some time. In the late 1970s and early 1980s, after the advent of screening of blood donors for HBV infection and after testing for infection with hepatitis A virus was introduced, it became clear that a large number of transfi~sion-transmitted hepatitis cases still occurred. This newly recognized type of hepatitis was called non-A, non-B hepatitis, and the search for the responsible virus began. Several studies have examined whether testing donors for surrogate markers would prevent some cases of non-A, non-B hepatitis (now identified as being primarily due to infection with a third hepatitis virus, HCV).'2 Three studies showed roughly a twofold reduction in the risk of posttransfi~sion hepatitis associated with transfusion of blood tested and found to be negative Mach, RD, W Szmuness, JW Mosley, FB Hollinger, R Kahn, CE Stevens, VM Edwards, J Werch (1981). Serum alanine aminotransferase of donors in relation to the risk of non-A, non-B hepatitis in recipients: The transfusion-transmitted viruses study. New England Journal of Medicine, 304: 989-994. Stevens, CE, RD Aach, FB Hollinger, JW Mosley, W Szmuness, R Kahn, J Werch, VM Edwards, (1984). Hepatitis B virus antibody in blood donors and the occurrence of non-A, non-B hepatitis in transfusion recipients: An analysis of the transfusion- transmitted viruses study. Annals of Internal Medicine, 101: 733-738. Koziol, DC, PV Holland, DW Ailing, JC Melpolder, RE Solomon, RH Purcell, EM Hudson, FJ Shoup, H. Krakaven, HJ Alter (1986). Antibody to hepatitis B core antigen as a paradoxical marker for non-A, non-B hepatitis agents in donated blood. Annals of Internal Medicine, 104: 488~95. Aynard, JP, C Janot, S Gayet, C Guillemin, P Canton, P Gardner, F Streiff (1986). Post-transfusion non-A, non- B hepatitis after cardiac surgery: Prospective analysis of donor blood anti-HBc antibody as a predictive indicator of the occurrence of non-A, non-B hepatitis in recipients. Fox Sanguinis, 51: 23~238.
10 BLOOD AN19 BLOOD PRODUCTS: SAFETY AND RISK for HBV core antibodies (compared to the rates when HBV core antibody positive blood had been transfused. Within a couple of years of these studies, blood banks began testing for antibodies to HBV core antigen and for alanine aminotransferase (ALT), in part as a result of the great public concern about blood safety associated with the AIDS epidemic. The Johns Hopkins-Houston study of cardiac surgery patients, already described, was still in progress. Therefore, we were able to examine the efficacy of surrogate marker testing on the incidence of posttransfi~sion HCV infection as soon as the first-generation test for HCV became available. The study included three periods of donor screening: one soon after the start of HIV screening but before the start of surrogate marker (i.e., anti-HBV core and ALT) testing, which was begun in March 1985 and lasted until September 1986; a second period, between October 1986 and May 1990; and the third period, which began with the use of a specific test for donor antibodies to HCV in May 1990. Our data indicate that the risk of HCV infection associated with transfusion was substantial prior to screening foe surrogate markers or HCV antibodies.'3 There is about an 18-fold difference in the rate of HCV infection attributable to transfusion: 366 cases occurred in 9,821 patients who were transfused but only 5 in the 2,400 who were not. The decline in transfusion-associated HCV infection with screening of donors for surrogate markers was significant regardless of which generation ELISA was used to test recipients, as was the much larger decrease after donor testing for anti-HCV antibody. About twice as many seroconverters were detected by the second generation ELISA, but there were also more indeterminate tests with the confirmatory radioimmunoblot assay (RIBA). 14 Testing these indeterminate samples by PCR and a third generation RIBA showed that only about a third of them were actually infected with HCV. The second-generation assay is now used to screen donors, so the best estimate of the current risk is about 3 per 1 0,000. A recent paper by Blajchman and colleagues from Canada's reported a controlled study in which they assigned 4,500 patients to receive blood that either was or was not tested for surrogate markers (anti-HBV core antigen and ~3Donahue, JG, A Munoz, PM Ness, DE Brown, DH Yawn, HA McAlister, BA Reitz, KE Nelson (1992). The declining risk of post-transfusion hepatitis C virus infection. New England Journal of Medicine, 32 7: 369-373. Abelson, KE, F Ahmed, PM Ness, V Strumbolis, C Parniss, G Gosch, D Yawn, V McAlister (1993). Comparison of first and second generation ELISA screening tests in detecting HCV infections in transfused cardiac surgery patients. Transfusion, 33(5): 5116. ~5Blajchman, MA, SB Bull, SV Feinman (1995). Post-transfusion hepatitis: Impact of non-A, non-B hepatitis surrogate tests. Lancet, 345(8941): 21-25.
CURRENT RISKS OF DISEASE TRANSMISSION 11 ALT). Unlike the United States, Canada had not required surrogate marker screening on the basis of the data from the three studies cited above. However, while the study was in progress, anti-HCV testing was introduced throughout Canada. The results of this study generally confirmed the findings from our study. Prior to the institution of the specific test for HCV, roughly 20 seroconverters were seen per 1,000 units of untested blood transfused. Screening for surrogate markers reduced this rate to 5 per 1,000 units. After HCV testing was instituted there was still a small difference associated with surrogate marker screening of donors, but it was not significant. An important issue that blood banks are now facing is whether surrogate marker testing should be continued in the face of a sensitive and specific test for HCV. Most experts think that such testing is no longer justified by reduction of the risk of HCV. Indeed, in 1995 the NIH Consensus Development Conference advocated dropping ALT testing but continuing the use of anti-HBV core antigen screening because of its utility in reducing the risk of transfusion- transmitted HIV and HBV. The risk of transmitting HBV by transfusion is a continuing concern to those working to improve the safety of the blood supply. In fact, HBV was the first viral infection for which donors were screened. Screening of donors for HBV infection began when the hepatitis B surface antigen (HBsAg) test was licensed in 1972. Nevertheless, there have been several case reports of people who have been infected with HBV despite screening for HBsAg. These HBV infections among persons receiving HBsAg screened blood could have occurred for several reasons. In an HBV-infectious donor the HBsAg test could be negative because HBsAg was present only at a level below the sensitivity of the assay, it was bound to antibody as an immune complex, or the HBV strain was a mutant virus without a surface antigen detectable by the current screening test. The Hopkins-Houston cardiac surgery study allowed an estimate of the risk of transfusion-transmitted HBV and an assessment of the utility of other markers. To estimate the rate of transfusion-transmitted HBV, we screened transfused patients before and 6 months after their transfusion for antibodies to HBV core antigen. There was only about a twofold increase in the rate of incident HBV infections in the transfused patients compared to the incidence among those who were not transfused; the incidence of HBV infections among transfused patients was about tenfold lower per unit of blood than we found for HCV infections in the same study population. However, 39 patients seroconverted to HBV positivity after transfusion. Interestingly, the method of donor screening had a significant effect on the rates of posttransfi~sion HBV infections in this population. HBV infection rates started at about 0.048 percent prior to surrogate marker screening and fell only to 0.039 percent with non-A, non-B surrogate marker testing (anti-HBV
12 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK core antigen and ALT). After HCV testing was instituted, HBV seroconversion fell to a rate of 4 per 100,000 units transfused. These data suggest that HCV infection might be a surrogate for hepatitis B virus infections in donors. We studied this issue farther by identifying donors to the Baltimore/Washington, D.C. Red Cross who tested positive for hepatitis C and examining their sera for other markers of HBV infection. Surprisingly, only 36 percent of these HCV-positive donors were without other markers. Many had anti-HBV core antibodies, elevated ALT or both, suggesting that the algorithm for testing now in use may actually prevent some infections from viruses other than their target. Similarly, ALT testing may have improved the safety with regard to transmission of HBV. James Moseley did another analysis that supports our data. He found 15 recipients among patients enrolled in the Transfusion Transmitted Viruses Study of the 1970si7 who seroconverted to HBV positive when transfused with hepatitis B surface antigen-negative blood. Moseley then tested the original samples from these patients with current tests for hepatitis B surface antigen, HBV core antibody, and HBV-DNA PCR. Six of the 15 were core antibody positive. These data suggest that if we had tested donors for HBV core antibody in the middle and late 1970s, we might have prevented some transfusion-related HBV infections even though these donors had been found negative for hepatitis B surface antigen. The incidence of infection with HIV, HTLV, HCV, and HBV has very recently been determined for the 586,507 repeat donors in REDS. These data, together with the estimated duration of the seronegative window periods for these infections, were used to estimate the risk of transmission by transfusion between 1991 and 1993.~8 This study estimated the residual risks of transmission of these viruses from transfusion of screened blood as 1 in 493,000 donor units for HIV, 1 in 641,000 donor units for HTLV, 1 in 103,000 donor units for HCV, and 1 in 63,000 donor units for HBV. Another disease that has assumed increased importance as a transfi~sion- transmitted infection is Chagas' disease, or Trypanosoma cruzi infections. In much of South and Central America there are focal areas where people are Abelson, KE (1995). The risk of the transmission of retroviruses, hepatitis C virus and hepatitis B virus according to the methods of donor screening (abstract). NIH Consensus Development Conference on Infectious Disease Testing for Blood Transfusions, January 9-11, 1995, Bethesda, Maryland. ~7Mosley, JW, CE Stevens, RD Aach, FB Halligen, LT Mimmus, LR Solomon, CH Barbasn, GJ Nemo (1995). Donor screening for antibody to hepatitis B core antigen and hepatitis B virus infection in transfusion to recipients. Transfusion, 35: 5-12. '~Schreiber, GB, MP Busch, SH Kleinman, JJ Korelitz (1996). The risk oftransfusion- transmitted viral infections. New England Journal of Medicine, 334: 1685-1690.
CURRENT RISKS OF DISEASE TRANSMISSION 13 carriers of T. cruzi and can transmit the infection by blood transfusion. Four cases have now been reported in the United States and Canada. As a result, Roger Dodd and his group at the Red Cross'9 began a study in two blood banks, in Los Angeles and Miami, that were considered to have a donor population who might be at high risk for infection with T. cruzi. They screened all donors by questionnaire to ascertain whether they had ever resided for more than 4 weeks in an area where Chagas disease is endemic. Among 91,216 donors, 6,302 said yes, and their blood was tested for T. croci. Twenty-three were positive on an initial enzyme immunoassay, and 10 of them were confirmed to be positive by radioimmunoprecipitation assay. I would like to conclude by calling your attention to an important risk that has not, in my opinion, been adequately addressed, namely transfi~sion- transmitted bacterial infections.20 Many papers in the literature have reported on the transmission of Yersinia enterocolitica infection by transfusion Dom apparently healthy blood donors. Y. enterocolitica is one of a number of organisms that can cause asymptomatic bacteremia in donors and survive and replicate in the cold temperatures at which blood is stored. More recently there has been some focus on the risks associated with the transfusion of platelets, which must be stored at room temperature to preserve their Junction. In one study that evaluated 35,000 platelet transfusions, 37, or about 1 in 1000, were associated with febrile episodes. Ten of these episodes were shown to be bacteremic. The risk of bacteremia per unit of platelets transfused was 1 in 2,000. Contamination of platelets by staphylococcal and other organisms that are nonnal skin flora during the donation process is not uncommon. In the four or five days that the platelets may be stored at room temperature, there may be enough bacterial growth for the recipient to develop a significant infection after transfusion of these platelets even if visual inspection at the time of use uncovers no sign of bacterial growth. In order to evaluate a procedure to reduce this risk, Yomtovian and her colleagues at the Cleveland Clinic2i have cultured or done Gram stains in order to identify bacterial contamination of platelets. They then deferred transfusion of platelets found to be positive and found a significant decrease i9Dodd, RY (1995). Preventing transfusion-transmiKed Chagas disease: the American Red Cross approach (abstract). NIH Consensus Development Conference on Infectious Disease Testing for Blood Transfusions, January 9-11, 1995, Bethesda, Maryland. 20Blajchman, MA (1994). Transfusion-associated bacterial sepsis: The phoenix rises yet again. Transfusion, 34: 940-942. 2iYomtovian, R. HM Lazarus, LT Goodnough, NV Hirschler, AM Morrissey, ~ Jacobs (1993). A prospective microbiologic surveillance program to detect and prevent the transfusion of bacterially contaminated platelets. Transfusion, 33: 902-909.
14 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK in the rate of bacteremia in recipients of the culture- and Gram stain-negative platelets. However, another study using Gram status or culture to identify units of platelets at high risk of causing bacterial sepsis concluded that these techniques were poor greening methods because of their inadequte sensitivity and specificity.22 Clearly, more research needs to be done in order to develop sensitive, specific, and practical methods for reducing the risks of the transfusion transmission of bacterial infections, especially those associated with platelet transfusion. The occurrence of transfusion-transmitted HIV and a better understanding of the high frequency of chronic hepatitis C virus infections have obviously led to a greater appreciation of the potential importance of transfusion acquired infections. What will the fixture bring? It seems inevitable that we will discover new pathogens that are transmissible by the transfusion of blood products. Indeed, researchers have recently identified and sequenced a new flavivirus that is carried in the blood of approximately 1 percent of blood donors and is transmittable by transfusion.23 Although it has been named hepatitis G virus (HGV), preliminary clinical data suggest that individuals who have acquired HGV infections do not seem to develop chronic hepatitis despite chronic infection with HGV.24 The strains of HIV-1 that have caused the worldwide pandemic of AIDS have been designated as group M viruses. Another group of HIV-1 viruses have been identified that cause AIDS but show extensive genetic divergence from group M strains.25 These HIV-1 viruses have been designated group O viruses. The antibody response elicited by these group O strains is not 22Barrett, BB, JW Anderson, KC Anderson (1993). Strategies for the avoidance of bacterial contamination of blood components. Transfusion, 33: 228-234. 23Simons, IN, TP Leavy, GJ Dawson, TJ Pilot-Matis, AS Muerhoff, GG Schlauder, SM Desai, IK Mushahwar (1995). Isolation of novel virus-like sequences associated with human hepatitis. Nature Medicine 1: 564-569. Linnen, J. J Wages Jr, ZY Zhang-Keck, KE FIN, HZ Krawczynski, H Alter, E Koonin, M Gallagher, M Alter, S Hadzlyannis (1996). Molecular cloning and disease association of hepatitis G virus: A transfusion-transmissible agent. Science, 271: 505-508. Simons, IN, TJ Pilot-Matis, TP Leavy, GJ Dawson, SM Desai, GG Schlauder, AS Muerhoff, JC Erker, SJ Buijk, ML Chalmers (1995). Identification of two flavivirus-like genomes in the GB hepatitis agent. Proceedings of the National Academy of Sciences USA, 92: 3401-3405. 24Alter, HI, Y Nakatsuji, JW-K Shih, J Melpolder, K Kiyosawa, J Wages, J Kim (1996). Tranfusion-associated hepatitis G virus infection (abstract 120). Paper presented at the IX Triennial International Symposium on Viral Hepatitis and Liver Diseases, Rome, Italy, April 21-25. Alter, M, M Gallagher, T Morris, C Moyer, K Krawczyaski, Y Khadyakow, H Fields, J Kim, A Margolis (1996). Epidemiology of non A-non E hepatitis (abstract 119). Paper presented at the IX Triennial International Symposium on Viral Hepatitis and Liver Diseases, Rome, Italy, April 21-25. 25Gurtler, LG, PH Hauser, J Eberle (1994). A new subtype of human immunodeficiency virus type 1 (MVP-5180) from Cameroon. Journal of Virology, 68: 1581-1585.
CURRENT RISKS OF DISEASE TRANSMISSION 15 consistently detected by currently licensed ELISA kits.26 Most persons infected with group O viruses have been from West and Central Africa, especially Cameroon, Gabon, Nigeria, Niger, Senegal and Togo. However, one patient from Los Angeles and a French national have been found to be infected with HIV-1 group O strains.27 Several companies are developing ELISA screening tests that will detect both group O and group M strains of HIV-1, but none of these assays are currently licensed by FDA. These case suggest that there will be a continuing need for rapid development, evaluation, and licensure of new screening tests in order to maintain the safety of the blood supply. 26Loussert-Ajaka, I, TD Ly, ML Chaix (1994). HIV-1/HIV-2 seronegati~ity in HIV-1 subtype O infected patients. Lancer, 344: 1333-1334. 27Centers for Disease Control and Prevention (1996). Identification of HIV-1 group O infection 1996. Morbidity and Mortality Weekly Report, 45: 561-565.
Viral :Inactivation of Blood Products: A General Overview Bernard Horowitz Over the past decade, blood banking and blood processing procedures and the practice of transfusion medicine have changed substantially. Today, we are more aware of the dangers of blood transfusion and of steps to reduce if not eliminate these dangers. Blood donors are examined and questioned more closely than ever before in an attempt to eliminate donors who are more likely to harbor an infectious blood-borne virus. Every donation is tested by new and more sensitive blood tests, and in some cases blood screening tests are introduced even before their benefit is established. Donor histories and test results have been computerized, and the e~ror-prone manual transcription of critical information is being eliminated. Manufactured blood products are more highly purified than ever before, and purification procedures have been modified to more consistently reduce viral load. Virus inactivation technology is in widespread use in the preparation of coagulation factor concentrates, and validated virus inactivation methods are beginning to be applied to all blood protein solutions including immune globulins and fresh frozen plasma (FFP). One could not fathom introduction of a new blood protein product today if it was not virally inactivated. With respect to viral safety, the data are clear: the only way to achieve absolute safety is through viral inactivation, and numerous advantages accrue on adoption of virus inactivation processes. The window period of seronegativity will no longer be of concern, errors in testing or the inadvertent release of a blood unit that tests positive will no longer result in viral transmission, new viruses or new viral serotypes will be eliminated even before their presence is recognized, and tests for rare viruses need not be deployed. Nowhere can the value of virus inactivation be illustrated better than in the preparation of coagulation factor concentrates. Antihemophilic factor (AHF) concentrate and prothrombin complex concentrate manufactured without viral inactivation transmitted human immunodef~ciency virus (HIV), hepatitis 17
18 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK B virus (HBV), and hepatitis C virus (HCV) at high Dequency.28 As late as 1985, essentially every vial of these concentrates was contaminated by HCV. With the advent of viral inactivation, HIV transmission was virtually eliminated. For example, in the United States not a single documented case of HIV transmission has been associated with concentrate infusion since 1987. In fact, solvent/detergent- (S/D-) treated products are used in the preparation of approximately two-thirds of the world's plasma-derived coagulation factor concentrates, and more than 7 million doses have been infused without a single documented case of HIV, HBV, or HCV transmission (see Table 1~. TABLE 1 S/D-Treated Product Usage: 1985-March 1994 Product Units Doses (approx.) Factor VII 1.9 MU1,900 Factor VIIa 2.6 MU2,600 Factor VIII 6,085 MU6,085,000 Factor IX 353 MU353,000 Prothrombin complex 113 MU105,667 Fibrin glue 325,930 ml65,186 Fibrinogen 93,300 g23,300 IMIG & IVIG 1,266,245 g253,249 MAb IgM 2,697 vials2,697 Anti-D IgG 83,702 vials83,702 Plasma 789,479 units197,400 Sum 7,173,701 SOURCE: Horowitz, B. AM Prince, MS Horowitz, C Watklevicz (1993~. Viral safety of solvent-detergent treated blood products. In Brown F (ed), Virological Safety Aspects of Plasma Derivatives, Developments in Biological Standardization, 81: 147-161; updated with information on file. The commonly employed viral elimination procedures are: . . heat (pasteurization, dry heat, vapor), · solvent/detergent, · beta-propiolactone/ultraviolet light, acid. 2SHorowitz, B. MPJ Piet (1986). Transmission of viral diseases by plasma protein fractions. Plasma Therapy Transfusion Technology, 7: 503-513.
CURRENT RISKS OF DISEASE T~NSMISSION sodium thiocyanate filtration, extensive purification, and combinations. ~7 19 Each has distinctive features. S/D acts by disrupting the viral lipid envelope, and a 12-year history of safety with respect to enveloped viruses supports its use. Virus kill is rapid ~1 hour) and complete. Because action is directed toward lipids, nonenveloped viruses and proteins (except for lipoproteins) are unaffected, and S/D can be applied equally and predictably with a high rate of recovery to complex mixtures such as plasma and to highly purified protein solutions. Safety with respect to HBV, HCV, and HIV is supported by 13 independently run clinical trials. Methods of thermal inactivation are advantageous in that all classes of virus are potentially susceptible, although nonenveloped viruses tend to be heat stable. Because thermal inactivation methods are not inherently specific, means of stabilizing proteins while achieving excellent virus inactivation had to be identified. With pasteurization, proteins are stabilized by addition of high concentrations of low-molecular-weight solutes, especially sugars and amino acids. Although viruses are also stabilized, relatively good discrimination can be achieved, although at some cost in protein recovery. Using the duck HBV as a model, S/D treatment is more effective than pasteurization at killing HBV.29 Additionally, many nonenveloped viruses are also heat resistant. With dry heat, proteins are stabilized by reducing the moisture content, and process recovery can be high. A particular advantage of the dry heat method is that it can be performed on product in the final container, eliminating the possibility of posttreatment recontamination. With all other methods, recontamination is prevented by separating pre- and postvirus inactivation areas, equipment, and personnel. Nonetheless, despite these differences, each method has eliminated HIV transmission by pooled plasma products, and HBV and HCV transmission has either been eliminated or greatly reduced.30 More recently, the apparent transmission of hepatitis A virus (HAY) by an ion-exchange-purified, S/D-treated AHF concentrate in several European countries raised concerns about nonenveloped viruses, first because they are 29Long, Z. C-S Sun, EM White, B Horowitz, AF Sito (1993). Hepatitis B viral clearance studies using duck virus model. In Brown F (ed): Virological Safety Aspects of Plasma Derivatives. Developments in Biological Standardization, 81: 163-168. 30Horowitz, B (1991). Inactivation of viruses found with plasma proteins. In Goldstein, J (ed.), Biotechnology of Blood. Boston: Butterworth-Heinemann.
20 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK not inactivated by S/D treatment and second because they tend to be heat stable. Viruses in this class include HAV and parvovirus B 19. Consequently, manufacturers are examining newer viral elimination procedures in combination with established virucidal procedures. The advantage of combining methods that act by independent mechanisms is that both a broader spectrum and a higher quantity of viruses can be eliminated. As examples, antibody affinity purification validated as a virus removal method has been combined with either S/D or heat treatment; some products are now treated by both S/D and heat; other products have been processed through virus-removing filters that have been developed recently and added to existing processes. New methods of viral inactivation under exploration include the use of chaotropes such as sodium thiocyanate, short- wavelength ultraviolet light in the presence of antioxidants, microwave heating, extraction with supercritical fluids, and iodine. Given the extensive history of safety with respect to the principal viruses of concern achieved by currently employed methods, it seems likely that these techniques will supplement rather than replace existing processes. As an example, research at the New York Blood Center has shown that combining S/D with ultraviolet C irradiation kills a wide variety of viruses including HBV, HCV, HIV, HAV, and parvovirus (Table 2~. TABLE 2 Viral Elimination by Combination of S/D and Ultraviolet C Light (UVC)a Virus Viral Elimination (logo) SD UVC Sum VSV >6.5 4.4 >10.9 Sindbis >6.3 >6.0 >12.3 HBV >6.0 na >6.0 HCV >5.0 na >5.0 HIV >6.2 >5.6 >11.8 EMC 0 >5.6 >5.6 HAV oh >5.3 >5.3 Parvovirus 0 >5.0 >5.0 ° Abbreviations: VSV, vesicular stomatitus virus; EMC, encephalomypcarditis virus; na, not available h S/D enhances immune neutralization Thus, despite being prepared from plasma pools, today's coagulation factor concentrates have proven to be safe Tom transmission of HBV, HCV,
CURRENT RISKS OF DISEASE TRANSMISSION 21 and HIV. Virally inactivated concentrates are now safer than the individual units from which they were derived. Success with the sterilization of coagulation factor concentrates encourages research into the sterilization of blood components, i.e., FFP, red blood cell concentrates (RBCCs) and platelet concentrates. Before addressing the viral inactivation of blood components, we must ask if individual units of blood are already safe enough. It is my belief that the goal should be to reduce viral risk to 1 per 1 million or less, and that this goal can only be achieved through virus inactivation. Transfusion Plasma. Our experience with S/D encouraged us to develop S/D-treated plasma (SD-plasma) as a substitute for fresh frozen plasma (FFP). Briefly, units of FFP are combined, thawed, and treated with 1% tri~n- butyl~phosphate (TNBP) and 1% Triton X-100 at 30°C for four hours, the reagents removed by hydrophobic chromatography. The final product is then sterile filtered, frozen, and optionally, lyophilized. Viral inactivation has been extensively validated. Under these conditions of S/D treatment, the rate of VSV and Sindbis villas killing exceeds that observed with AHF concentrates, treated either with TNBP-cholate or TNBP-Tween. We have also shown that >106 infectious doses (IDA) of HBV, >105 IDso of HCV, and >1072 ID50 of HIV are killed and that >109' IDS of HAV are neutralized. Because of pooling, a dose of SD-plasma consistently has 30 times more anti-HAV antibody than a dose of intramuscular immune globulin, known to prevent the spread of HAV, and has approximately the same quantity of antiparvovirus antibody as a dose of intravenous immune globulin, reported to be effective in the therapy of parvovirus infections. The coagulation factor content resembles that of the start pool and is more consistent than that found in individual donor units. There is no evidence that coagulation factors are activated, and the level of other proteins is nonnal. Toxicology studies indicate that the tiny amounts of TNBP and Triton X-100 that remain are safe. SD-plasma has been extensively evaluated in the United States and Europe.3' In the United States, more than 20 clinical study sites took part. In our own studies, 93 patients were treated on 504 occasions with 1,334 units of SD-Plasma. This included the successful treatment of 37 surgical episodes and 75 bleeding episodes in patients who were congenitally coagulation factor deficient and 9 successful uses to reverse warfarin therapy in advance of surgery. In patients with chronic or acute thrombotic thrombocytopenic purpura, SD-plasma was just as good as FFP in stimulating an increase in platelet count. Formal viral safety studies indicate that virus has not been ~ _v. ____ _ . . . . . _ ~ A ~ __ _ 3~Pehta, JC (1994). Clinical studies with solvent detergent-treated plasma. Transfusion Medicine Audio Updates.
22 BLOOD AND BLOOD PRODUCTS: SAFETY AND RISK transmitted, and this conclusion is supported by published studies32 and the more than 1 million units infused in Europe to date. Blood Cell Concentrates. Sterilization of cellular products is more difficult than virus inactivation of blood protein products, because blood cells are more complex and fragile than proteins, and multiple viral forms are present, including cell-Dee virus, virus that is a&Brent to cell membranes, actively replicating virus, and latently infected cells. Nonetheless, because erythrocytes and platelets do not replicate, methods that modify membranes or nucleic acid may prove useful. Red Blood Cell Concentrates. Numerous methods have been investigated, including the use of beta-propiolactone, nitrogen mustards, aryl diol epoxides, ozone, and halogenated oxidizing agents, but the best results described thus far employ photodynamically active sensitizers and visible light. Early work by Matthews and coworkers 33 showed good virus kill with hematoporphyrin derivative. We have shown that by substituting phthalocyanine, which absorbs light where hemoglobin does not, virus kill is greatly improved.34 Phthalocyanines and other dyes, like methylene blue or sapphyrins, activate oxygen to its reactive forms. With phthalocyanine treatment of red cell concentrates, we have begun to analyze the complex reaction pathways through the addition of quenchers of reactive oxygen species. Some compounds like mannitol and glutathione will quench oxygen radicals, while other compounds like tryptophan and sodium azide principally quench singlet oxygen. Using this approach we have shown that virus kill is not mediated by oxygen radicals but is mediated by singlet oxygen. This finding has practical importance because we can enhance reaction specificity by including quenchers of radicals at the time of light exposure. Platelet Concentrates. Photodynamically active compounds such as those under evaluation in the treatment of RBCCs reduce platelet aggregation response to collagen and to other agonists, but encouraging results have been 32Inbal, A, O Epstein, D Blickstein, et al. (1993). Evaluation of solvent/detergent treated plasma in management of patients with hereditary and acquired coagulation disorders. Blood Coagulation and Fibrinolysis, 4: 599-604. 33Matthews, JL, IT Newman, F Sogandares-Bernal, et al. (1988). Photodynamic therapy of viral contaminants with the potential for blood banking applications. Transfusion, 28: 81-88. 34Horowitz, B. B Williams, S Rywkin, et al. (1991). Inactivation of viruses in blood with aluminum phthalocyanine derivatives. Transfusion, 31: 102-108.
CURRENT RISKS OF DISEASE TRANSMISSION 23 obtained with psoralen derivatives. Psoralens are naturally occulting furocoumarins found in many foods, and they have been used therapeutically since antiquity. The principal reaction of psoralens on exposure to long- wavelength ultraviolet light is the cross-linking of nucleic acids. The initial report on the treatment of platelets with psoralens came from Corash's laboratory.35 Treatment of an oxygen depleted platelet concentrate with 8'- methoxypsoralen and UVA irradiation was shown to inactivate >6.7 logo CFU of Escherichia coli, >6.9 logo CFU of Staphylococcus aureus, >7.3 logo PFU of phage fd, 2.5 logic PFU of phage R17, and 5.1 logic PFU of feline leukemia virus. When treatment was under deoxygenated conditions, platelet morphology, process recovery, and response to the aggregation agent A23187 were comparable to thopse of untreated controls. If not first deoxygenated, aggregation response was adversely affected. We have overcome the deoxygenation requirement by adding quenchers of active oxygen species.36 Additionally, new synthetic psoralens with increased reactivity with nucleic acids are being developed and may serve to enhance reaction specificity further. In conclusion, blood and blood products have never been safer. However, the public's continuing concern about the safety of the blood supply from viruses, and the differential safety profile between blood and other pharmaceuticals demand that we continually improve. The achievements of the past are laudable. Nonetheless, safety Tom viruses falls well short of a standard of less than one transmission per 1 million units transfused, a realistic goal that we believe the transfusion community should adopt. For single- donor blood products, improved screening systems may achieve this goal; however, screening systems alone will never eliminate the window of seronegativity, and screening tests cannot anticipate new viruses or viral serotypes. Pooled blood products that have been virally inactivated meet this standard for most viruses, and use of a second viral elimination procedure that complements the first one will further ensure the safety of these products. Incorporation of viral inactivation procedures into the manufacture of all blood products, including blood cell concentrates, overcomes the weaknesses of screening procedures, and the further development of virus inactivation methodologies should continue to be encouraged. 35Lin, A, GP Wiesehahn, PA Morel, ~ Corash (1989). Use of 8-methoxypsoralen and long wave-length ultraviolet radiation for decontamination of platelet concentrates. Blood, 74: 517-525. 36Margolis-Nunno, H. R Robinson, E Ben-Hur, B Horowitz (1994). Quencher enhanced specificity of psoralen photosensitized virus inactivation in platelet concentrates. Transfusion, 34: 802-810.