Blood Safety Summary - January 1999
DATE: February 8, 1999
TO: Interested Parties
FROM: Stephen D. Nightingale, M. D., Executive Secretary
Advisory Committee on Blood Safety and Availability
SUBJECT: Summary of January 28, 1999 Meeting
The Advisory Committee on Blood Safety and Availability met on January 28, 1999 at the Crown Plaza Hotel, 14th and K Streets, N. W., Washington, D. C. 20005.
The meeting was held to consider options for implementing a prior recommendation of the Advisory Committee. This prior recommendation was to expand the direct notification component of current hepatitis C virus (HCV) infection lookback efforts. This expansion would include recipients of blood from donors subsequently identified as repeat reactive by the single antigen enzyme immunoassay (EIA-1) screening test for HCV infection that was licensed in 1990.
Members present were Dr. Caplan; Mr. Allen; Drs. Aubuchon, Busch, Chamberland, Feigal, Goosby, Guerra, and Hoots; Ms. Jones; Drs. Kuhn and McCurdy; Ms. O'Connor; Drs.Penner and Piliavin; CAPT Rutherford; Drs. Schiff, Secundy, and Snyder; and Mr. Walsh. Also present were Dr. Davey, Consultant to the Committee; Dr. Epstein, Director of the Office of Blood Research and Review, Food and Drug Administration; CAPT Lawrence McMurtry, Deputy Executive Secretary to the Committee; Dr. Nightingale; and approximately 45 members of the public.
The meeting opened at 9:11 AM with the reading of a statement of the conflict of interest laws that governed the meeting. This was followed by welcoming remarks from the Chairman.
Dr. Harold Margolis of CDC then updated the Advisory Committee on his agency's general lookback programs for public notification of all those at risk of having acquired transfusion-associated HCV infection prior to July 1992. He noted that some of these campaigns would be directed at the public at large, while others would target specific patient groups. He cited a recent article in Parade Magazine as an example of general notification that was already under way. He described how focus groups had contributed to the design of CDC public education programs, and he announced that there would be a formal launch of these programs in March 1999.
Dr. Margolis discussed coordinating the content of the public health messages that CDC and other organizations would be promulgating on HCV infection. He estimated that the fiscal 1999 CDC budget for general HCV lookback would be roughly $1 million. He described plans to reach specific patient groups that were likely to have been transfused before July 1992, such as cardiac surgery patients. These plans include reaching these patients through their providers. He concluded by observing that reaching patients who do not have a regular health care provider would pose a substantial challenge.
Dr. Margolis' presentation was praised by Dr. Caplan and by several members of the Committee. Dr. Kuhn asked if a public mailing, similar to the HIV message sent by former Surgeon General Koop, was under consideration as a component of general notification. Dr. Margolis responded that the idea was still on the table, although CDC was currently focusing on how to reach the patients at risk more directly. Dr. Margolis also mentioned the need for allocating enough resources to the public sector to accommodate those who would seek testing, and he noted current plans of the CDC, FDA, and the Agency for Heath Care Policy and Research to evaluate the effectiveness of HCV lookback.
Dr. Michael Busch then summarized an extensive analysis that he, Dr. Leslie Tobler, and several colleagues had performed on the use of the signal to cutoff ratio of the EIA-1 screening test to improve retrospectively its performance as a diagnostic test between July 1990 and June 1991, the time before confirmatory tests became generally available. At Dr. Penner's request, Dr. Busch first reminded the Committee that
the EIA-1 used a single recombinant protein, called c100, that comprises only about 5 % of the HCV genome;
the first generation supplemental test to confirm repeat reactive EIA-1 tests, the Recombinant Immunoblot Assay (RIBA-1), contained the c100 protein, a 5-1-1 component of it, and a control band;
the second generation EIA-2 and RIBA-2 tests added core antigens c22 and c33 (blood specimens that react with only one of these antigens are considered indeterminate); and
the third generation EIA-3 and RIBA-3 tests added a non-structural antigen called NS-5.
(Implicit in Dr. Busch's presentation was the understanding that the donor screening tests under discussion for triggering HCV lookback are those tests performed when a donor subsequently returns to make a repeat donation, and not tests performed on the donor at the time of the original donation. If the tests performed at the time of the original donation had been positive, the original donation would have been declined. Lookback is triggered by the possibility that, at the time of the original donation, the donor may have been at an early stage of HCV infection in which there may have been HCV virus in the blood, but that antibodies to HCV - which is what the screening and supplemental tests detect - have not yet developed. Also implicit in Dr. Busch's presentation was the understanding that if the donor never returns to make a repeat donation, the screening tests that would trigger HCV lookback will never be performed.)
Dr. Busch presented data from the Blood Centers of the Pacific and the Kansas City Blood Center that showed about 80% of all EIA-1 repeat reactive donors were identified between July 1990 and June 1991. He concluded that a first-generation lookback would have to include unconfirmed as well as confirmed EIA-1 tests, because about 80% of the EIA-1 repeat reactive donations had been culled by EIA-1 screening before the first confirmatory test became available.
Dr. Busch acknowledged that Dr. Epstein had been the first to suggest use of the signal to cutoff ratio to refine the performance of the EIA-1 test. Dr. Busch presented a data set from the American Red Cross which showed that
- 62% of EIA-2 repeat reactive donors had a positive supplemental test;
- in the vast majority of those EIA-2 repeat reactive tests that were confirmed by a positive RIBA-2 supplemental test, the EIA-2 signal to cutoff ratio was greater than 3.5; and
- in the vast majority of the EIA-2 repeat reactive tests that were not confirmed because a RIBA-2 supplemental test was negative, the EIA-2 signal to cutoff ratio was less than 3.5.
He then presented similar Red Cross data for the EIA-3 screening test. Sixty-four percent of these EIA-3 repeat reactive donors were confirmed positive by RIBA-3. There was a similarly strong association between the signal to cutoff ratio and the result of the RIBA-3 test.
These correlations led Dr. Busch to collect data from four blood centers and compare the signal to cutoff ratios of a total of 3753 EIA-1 repeat reactive tests with the results of the RIBA-2 supplemental tests performed on these samples. Overall, 60% of the RIBA-2 supplemental tests in this sample were positive, 5% indeterminate, and 30% negative. Dr. Busch's data showed an equally strong correlation between the results of the RIBA-2 test and the signal to cutoff ratio. These data were presented to support the use of the signal to cutoff ratio as an estimate of what the result of a RIBA-2 test would have been if the RIBA-2 test had been available and performed on the sample.
Dr. Busch then discussed choice of a signal to cutoff ratio that would optimally distinguish an uninfected from an infected donor, using the expected result of a RIBA-2 test as a gold standard. If a signal to cutoff ratio of 2.5 was chosen to discriminate positive from negative EIA-1 tests, the following would be expected to occur:
- about 7% of the EIA-1 tests with a ratio below 2.5 would be confirmed positive by RIBA-2, and 90% would be expected to be negative by RIBA-2 (the remainder would be indeterminate).
- in contrast, about 75% of the EIA-1 tests above 2.5 would be confirmed positive by RIBA-2, and about 20% would be expected to be negative by RIBA-2 (again, the remainder would be indeterminate).
- The sensitivity of the EIA-1 test, using the signal to cutoff ratio of 2.5, would be 89%.
- The specificity of the EIA-1 test, using the signal to cutoff ratio of 2.5, would be 87%.
- The predictive value of a positive EIA-1 test, with positive defined as a signal to cutoff ratio over 2.5, would be 75%.
- The predictive value of a negative EIA-1 test, with negative defined as a signal to cutoff ratio less than 2.5, would be 92%.
Dr. Busch pointed out that a RIBA-2 indeterminate test on an EIA-1 repeat reactive screening test would be expected to have a lower predictive value than a RIBA-2 indeterminate test on an EIA-2 repeat reactive test. In fact, he found that only 11 of 69 (16%) of EIA-1 repeat reactive, RIBA-2 indeterminates were positive by RIBA-3, and that none of these RIBA-3 positive samples were positive by genomic amplification testing. On this basis, he concluded that notification based on EIA-1 repeat reactive, RIBA-2 indeterminate tests was not warranted.
Dr. Busch then presented genomic amplification test data to address the question of how many of the approximately 10% of donors with a signal to cutoff ratio below 2.5 but a positive RIBA-2 test would actually have HCV genomic material in their blood. He found that
- 90% of the RIBA-2 confirmed EIA-1 tests with a signal to cutoff ratio over 2.5 were positive for HCV RNA, whereas
- 50% of the RIBA-2 confirmed EIA-1 tests with a signal to cutoff ratio below 2.5 were positive for HCV RNA.
As a result, only half of the 10% of the EIA-1 confirmed positive samples that would not trigger lookback because their signal to cutoff ratio was less than 2.5 would be positive for HCV RNA.
Dr. Busch then compared the use of 2.5, 2.0, and 1.5 as the signal to cutoff ratio for the EIA-1 test. He found that
- with the 2.5 cutoff, 89% of those who would be RIBA-2 positive would be notified and 87% of those who would be RIBA-2 negative would not be notified (as noted above);
- with a 2.0 cutoff, 91% of those who would be RIBA-2 positive would be notified, and 78% of those who would be RIBA-2 negative would not be notified; and
- with a cutoff of 1.5, 95% of those who would be RIBA-2 positive would be notified, and 57% of those who would be RIBA-2 negative would not be notified.
Similar results would be obtained using the serum HCV RNA instead of RIBA-2 as the reference standard.
In the comments that followed, Dr. Busch's presentation was also praised by the Chairman and Committee members. Dr. Busch noted that one consequence of his recommendations would be to place additional responsibility on blood establishments to determine the signal to cutoff ratio of their EIA-1 tests. This would be more difficult than simply notifying transfusion services of all EIA-1 repeat reactive tests, regardless of their signal to cutoff ratio.
Dr. Epstein noted that FDA had already established a policy permitting the use of the investigational RIBA-2 test to guide lookback determinations in its March and September 1998 Guidances on HCV lookback. In regard to use of RIBA-1 or other supplemental tests, Dr. Epstein suggested that these be considered clinical diagnostic testing rather than donor screening for purposes of HCV lookback, and acceptable for clinical diagnostic testing if performed in a CLIA-certified laboratory.
After a break, Dr. Susan Galel described the experience of the Stanford University Blood Center with EIA-1 lookback triggered by a positive (investigational) RIBA-1 supplemental test. She described the circumstances at her institution that led her and her colleagues to institute direct notification of recipients of blood from donors who had subsequently tested repeat reactive by an EIA-1 test that was confirmed by the RIBA-1 supplemental test. Dr. Galel compared the performance of the RIBA-1 and RIBA-2 tests and found them to be similar. At Stanford,
- 109 EIA-1 repeat reactive tests had been evaluated by the RIBA-1; 37 were positive, 15 indeterminate, and 53 negative.
- 84 EIA-1 repeat reactive tests had been evaluated by the RIBA-2; 43 were positive, 8 indeterminate, and 33 negative.
Dr. Galel also cited two published studies which had found that 373 of 393 (95%) RIBA-1 positive samples were also RIBA-2 positive, and that only 11 of 431 (2.5%) of RIBA-1 negative samples were RIBA-2 positive.
Dr. Galel noted further that
- 82% of her patients who were EIA-1 repeat reactive and RIBA-1 positive were also positive for HCV RNA by PCR, whereas
- 1.5% of her EIA-1 repeat reactive and RIBA-2 negative patients were also positive for HCV RNA by PCR.
She concluded that the RIBA-1 was comparable to the RIBA-2 for the purpose of determining which EIA-1 repeat reactive donations should trigger direct notification in an HCV lookback effort.
In the public comment period that followed, Mr. Michael Hall, representing the American Liver Foundation, reiterated the previously expressed position of the Foundation in support of targeted lookback for those who received blood from donors who subsequently tested positive for HCV between 1990 and 1992. He also reiterated the American Liver Foundation's support for a broad based and well financed national general notification campaign, and for full funding of the research agenda outlined at the March 1997 National Institutes of Health Consensus Conference on HCV infection.
Ms. Jan Hamilton of the Hemophilia Federation of America reiterated her organization's prior support for extending the direct notification component of HCV lookback to recipients of blood from donors who subsequently tested positive for HCV between 1990 and 1992. Her prepared statement had raised several questions about the general notification component of the HCV lookback, and she acknowledged that most of them had been answered by Dr. Margolis.
Mr. Cory Dubin of the Committee of Ten Thousand reiterated his organization's support for an expanded lookback. He also appealed for a letter to go out to every home in America regarding the risks of HCV infection as a means of informing individuals who received transfusions before 1990 of their risk. In addition, he reiterated several concerns that he had previously brought to the attention of the Advisory Committee, including compensation for transfusion-associated injury, the long duration of the debate over HCV lookback, fiscal pressure on the blood industry, and trust between the consumers and the providers of blood products.
After lunch, Dr. Caplan reviewed the issue before the Committee. The use of the signal to cutoff ratios in diagnostic tests, and the relative merits of the various signal to cutoff ratios for the EIA-1 test presented by Dr. Busch, were discussed in detail. Dr. Busch estimated that, if direct notification was to be based on a signal to cutoff ratio of 2.5 or above, about 100,000 notifications would be triggered, about 10,000 individuals would be alive and be traced by the notification effort, and about half of these, or 5,000 individuals, would have been previously unaware of their potential HCV infection. Dr. Busch estimated that using a signal to cutoff ratio of 2.5 to trigger direct notification, as opposed to using simply an EIA-1 repeat reactive test to trigger direct notification, would prevent about 452 false positive notifications for every true positive notification that would be prevented. He also estimated that, if a signal to cutoff ratio of 1.5 rather than 2.5 were chosen, there would be approximately 200 additional true positive notifications and approximately 20,000 additional false positive notifications. He also reiterated a point made earlier by Mr. Dubin, which was that about two thirds of all those exposed to HCV via blood transfusion would not be reached by direct notification, because their exposure occurred before the introduction of the EIA-1 screening test in 1990.
Dr. Davey and Dr. Epstein commented on effort that lookback entailed. Dr. AuBuchon amplified these comments, and noted that, in his own most recent cost effectiveness analysis, lookback would cost about $1 million per year of life extended for the current second generation lookback, and about $2 million per year of life extended for a first generation lookback. Dr. AuBuchon then offered the following motion:
The Committee reaffirms its understanding that recipients of transfusions should be made aware of the risks of HCV transmission associated with their transfusion. The Committee also reaffirms its comprehension of the situation that overlapping, multifaceted efforts are required to convey this information effectively. At the same time, the Committee recognizes that health care resources for this effort are not unlimited, and any program will necessarily fail to reach all HCV-infected recipients. Furthermore, direct notification carries with it not only cost, but also negative psychosocial implications that gained increased prominence with reduced likelihood of infection transmission.
Therefore, the Committee recommends with respect to expanding HCV lookback to include Version 1.0 anti-HCV testing, that donors testing repeatedly reactive in Version 1.0 testing be included in a targeted lookback program, if the blood collection agency has a record of a positive result in any supplemental HCV assay, or if the blood collection agency has other accessible data to indicate that the donor is at higher risk of being truly infected with HCV.
Dr. Guerra seconded the motion.
Discussion centered around the action to be recommended if no supplemental data were available. Dr. Epstein noted that blood establishments were mandated to retain records of the signal to cutoff ratio of EIA-1 tests. However, Dr. Epstein was concerned that blood establishments might prefer to report all repeat reactive EIA-1 tests to transfusion services, rather than review the signal to cutoff ratios and report only those above a certain level. CAPT Rutherford stated that the Department of Defense would not wish to have its option of notification based solely on an EIA-1 repeat reactive test.
There was discussion whether to amend Dr. AuBuchon's motion in light of the above considerations or to begin again. Dr. Piliavin reminded the group that a motion could be withdrawn with the consent of the proposer and the second of the motion. Dr. AuBuchon and Dr. Guerra chose to withdraw the motion.
While an alternative motion was being crafted, Dr. Feigal responded to CAPT Rutherford's prior concern by noting that, in an integrated health care system such as the Armed Services, a transfusion service could choose to notify based solely on a repeat reactive EIA-1 test and work with its blood establishment to achieve this goal. However, this option would not be feasible in a diffuse civilian system, where multiple transfusion services are serviced by multiple blood establishments.
Dr. Epstein, on behalf of Dr. Davey, then proposed the following motion.
The Advisory Committee recommends that targeted lookback should be initiated based on a repeatedly reactive EIA-1 test result on a repeat donor unless a supplemental test result was performed and did not indicate significant risk of HCV infection or no supplemental test result is available, but the signal to cutoff ration of the repeatedly reactive EIA-1 test was less than 2.5.
Dr. Schiff seconded the motion.
Dr. Galel suggested than notification not be performed if subsequent testing of the donor indicated that HCV infection was not present. Dr. Penner stated his preference for a lower signal to cutoff ratio, but he indicated he felt that the ratio of 2.5 might under the circumstances be a reasonable accomodation. Dr. Piliavin called for a vote on the motion. Dr. Nightingale confirmed that Drs. Davey and Epstein had wished, in respose to Dr. Galel's suggestion, to append the phrase,
... or followup testing is negative.
to their motion; there was no objection to this addition.
There were 11 votes in favor of the motion as appended, none against, and two abstentions. The motion was approved.
Dr. Caplan then encouraged the Committee to consider a resolution of support for sufficient funding for public education and outreach. Dr. Miriam Alter of CDC noted that funds would also be needed by providers of testing and counseling. Subsequent discussion strongly supported these concepts. Dr. Caplan proposed the following motion, which after discussion was read as follows:
The Advisory Committee believes that, in light of the scope of the hepatitis C epidemic, current funding is inadequate for notification, testing, counseling, education, and therapy. We urge the Secretary to take immediate measures to increase funding to meet this major public health challenge. We further urge the Public Health Service to work with professional and private organizations to promulgate appropriate recommendations for testing, counseling, and therapy, and to secure additional resources for these purposes.
Dr. Penner seconded the motion. It was approved unanimously, with no abstentions.
Dr. Davey then proposed the following motion:
The Advisory Committee urges the Secretary to consider providing appropriate support and resources for blood centers and hospitals, both public and private, to conduct HCV lookback activities.
Dr. Hoots seconded the motion.
There was discussion about the various forms that appropriate support and resources could take. The motion was approved unanimously, with no abstentions.
A discussion was then begun about the length of time it had taken to bring the issue of hepatitis C lookback to closure. Dr. Caplan brought this discussion to its own closure by noting that he had allowed a "cathectic moment" so that all concerned could move forward, hand in hand, to a brighter future.
Dr. Caplan then reviewed several issues that might be brought before the Advisory Committee. These included leukodepletion and its associated costs, use of blood donated by individuals with hemochromatosis, the status of blood substitutes, and compensation for transfusion-associated injury. Dr. Feigal noted that almost all of these issues had been discussed by other committees, and that some of these committees had made recommendations that FDA would probably accept.
Dr. Feigal then suggested that the Advisory Committee might wish to review the adequacy of the donor supply, particularly in the light of various policy proposals that might limit future donations.
Dr. AuBuchon suggested that the Advisory Committee might wish to examine its role, and the role of others, in the governance of America's blood supply.
Mr. Walsh requested that the Advisory Committee maintain its interest in the ongoing shortages of plasma derivatives.
Dr. Guerra asked that the Advisory Committee examine demand as well as supply.
Dr. Caplan suggested examining altruism as a guarantor of both safety and availability of blood.
Dr. Busch raised the concern, in response to Dr. Feigal's earlier comment, that many issues - for example, leukodepletion or deferral of donors who had resided in Britain - that were within the scope of the Advisory Committee's charter might become policy before the Advisory Committee had the opportunity to review them. Dr. Busch specifically requested that these two issues be considered by the Advisory Committee.
Mr. Allen raised the issue of how cost should be considered in the Advisory Committee's deliberations. Dr. Feigal responded that the FDA is explicitly prohibited from considering cost in its own deliberations, even if that consideration was raised by a committee explicitly chartered to incorporate costs into its deliberations. Dr. Caplan took issue with Dr. Feigal's assertions. A brief, preliminary discussion of this complex issue followed.
Dr. Feigal suggested a joint meeting between the Advisory Committee and other committees, or a meeting among some members of various committees, to discuss perspectives and juristictions.
Dr. Caplan thanked all the members for their efforts. The meeting was adjourned at 3:22 PM.
This summary was approved by Dr. Caplan on February 5, 1998. The summaries, resolutions, and transcripts of this and all previous meetings of the Advisory Committee can be accessed at http://www.hhs.gov/bloodsafety.