192
anti-HCV tests should be as sensitive as possible to detect the maximum number of HCV carriers among blood donors, falsepositive results may be magnified by attempts to increase sensitivity. False-positive tests on blood donors cause unnecessary discard of units not infectious for HCV, plus indefinite deferral and unwarranted health concerns for volunteer donors.’ Using the two FDA-approved EIA-1 anti-HCV test systems, one from Ortho Diagnostics Systems (Raritan, NJ) and the other from Abbott Laboratories (Abbott Park, IL), we evaluated 5216 samples from US donors: 4012 in October 1990 and 1204 in February 1991. Although the starting material, a recombinant HCV protein (C-100),2 is the same for both licensed tests, the frequency of repeatably reactive (RR) results varied between them (Fig. 1). Each RR sample detected by either manufacturer’s kit was then tested by two supplemental methods: RIBA-I1 (second-generation recombinant immunoblot assay, Chiron/Ortho Diagnostic Systems, Raritan NJ) and HCV neutralization (Abbott Laboratories). With Abbott’s EIA test system, we found an anti-HCV RR rate of 0.92 percent, compared with 0.36 percent for the EIA of Ortho. Of the 48 RR samples from the Abbott test, 9 (19%) were RIBA-11-reactive or -indeterminate, and 11 (23%) neutralized. Eight (42%)of 19 RR samples from the Ortho test were RIBA-11-reactive or -indeterminate, and 7 (37%) of 19 neutralized (with two additional equivocal results) (Fig. 2). All nine RIBA-11-reactive samples also neutralized: seven were RR with EIA-1 from both manufacturers, one was RR with the Abbott test only, and one by the Ortho test only (this last sample was considered equivocal on neutralization, because the original test absorbance was
Ortho test system I
NEG
RR
I
NEG
I
3s
13 (.25%)
(-67%)
6 (.12%)
5162 (98.96%)
FIG. 1. Results of HCV EIA in 5216 samples from US blood donors.
/-A Abbott
1-
\ \
Ortho EIA-RR N=lQ
FIG. 2. Results of supplementary tests: N Neutralized; Abbott = and Ortho = %Y. @ RIBA reactive/indeterminate; Abbott = %S and Ortho = YIY.
less than the manufacturer’s cutoff for reactivity.) One RIBA11-indeterminatesample (with 3 reactivity at the C-22 band) was not detected by the Ortho test oust under the cutoff). Of the 13 EIA-1-RR samples that neutralized, including one of two with equivocal results, three (23%) were RIBA-II-nonreactive. The concordance between the anti-HCV EIA-1 screening results and supplementary assays is illustrated in Fig. 2. Abbott’s EIA-1 may be more sensitive than that of Ortho. The Abbott test identified two samples that were RIBA-IIreactive or -indeterminate (and that neutralized), plus two with neutralization only that were missed by the Ortho test; the Ortho test identified one sample that was RIBA-11-reactive and another that equivocally neutralized, both of which were missed by the Abbott test. If specificity equates with the reactivity of supplementary tests, then Ortho’s anti-HCV EIA-1 assay appears to be more ~ p e c i f i c .It~ should be noted that the supplemental tests are for research use only and are not yet considered confirmatory. Additional information is needed to establish that samples from blood donors found to be anti-HCV EIA-1RR and also reactive on one or more supplemental assays correlate with both HCV infectivity and transmission via blood transf~sions.~.~
+
BERNIEBETLACH,MT(ASCP) STEVEN ANDERSON,BS RICK RODRIGUEZ, BS KEN KURAMOTO,h4T KATHLEEN SAZAMA, MD, JD PAULV. HOLLAND,MD Sacramento Medical Foundation Center for Blood Research 1625 Stockton Blvd Sacramento, CA 95816
I
I
I
/
TRANSFUSION Vol. 32, No. 2-1992
LElTERS TO THE EDITOR
References 1. Public Health Service inter-agency guidelines for screening donors of blood, plasma, organs, tissues, and semen for evidence of hepatitis B and hepatitis C. Morbid Mortal Weekly Rep 1991;401-17. 2. Kuo G, Choo QL, Alter HJ, et al. An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 1989;244:362-4. 3. Mimrns L, Vallari D, Ducharme L, et al. Specificity of anti-HCV ELISA assessed by reactivity to three immunodominant HCV regions. Lancet 1990;336:1590-1. 4. Van der Poel CL, Reesink HW,Schaasberg W, et al. Infectivity of blood seropositive for hepatitis C virus antibodies. Lancet 1990;335:558-60. 5. Esteban JI, Gonzalez A, Hernandez J, et al. Evaluation of antibodies to hepatitis C virus in a study of transfusion-associatedhepatitis. N Engl J Med 1990;323:1107-12.
Prestorage white cell filtration of blood To the Editor: The recent editorial by Heal and Cohen’ provided a wellbalanced discussion of the potential benefits and hazards of inline, prestorage filtration of whole blood. I agree with those authors that discussions of the merits of filtration are often clouded by strong, but poorly documented claims made by commercial interests as to the benefits of filtration before blood component production. One potential benefit of prestorage filtration that deserves further study, but was not mentioned in the editorial, is that of improved quality control of the blood components that are
TRANSFUSION
1992-Vol. 32, No. 2
193
LETTERS TO THE EDITOR
meant to be white cell (WBC)-reduced. As stated by the authors, bedside filtration is technique-dependent. Depending on the training, interest, and available time of nursing staff, bedside filters may or may not be used appropriately. What is not clear is whether suboptimal technique leads to suboptimal WBC removal and, thereby, defeats the intended objective of filtration. A related quality issue is the use of third-generation filters for more blood components than recommended by the manufacturers. From my personal experience and that of others, it is not uncommon for a single filter to be used for more than 2 units of red cells, despite specific manufacturer instructions to the contrary. How does such incorrect use affect the effectiveness of WBC removal? Even though I am not advocating the prestorage filtration of blood, I believe the issues raised need careful examination. If data demonstrate that improper
use of bedside WBC filters significantly affects the number of WBCs removed, alternative filtration methods might be viewed more favorably.
MARKA. POPOVSKY,MD American Red Cross Blood Services Northeast Region Headquarters 180 Rustcraf! Road Dedharn, u4 02026
Reference 1. Heal JH, &hen HJ. Do white cells in stored blood components
reduce the likelihood of posttransfusion bacterial sepsis? (editorial). Transfusion 1991;31:581-3.
Submission of Letters Instructions for submission of letters can be found in the Detailed Instructions for Authors published on pages 92 to 97 of the January 1992 issue. Submit letters to: Peter D. Issltt, PhD, FIMLS, FIBlol, FRCPath Transfusion Service Duke University Medical Center, P.O. Box 2928 Durham, NC 27710. EDITOR'S NOTE: In order to permit timely publication of correspondence, the references have not been verified as they have been for articles appearing In Transfusion, and, therefore, the accuracy of cited references in Letters to the Editor Is the sole responsibility of the authors.
anti-HCV tests should be as sensitive as possible to detect the maximum number of HCV carriers among blood donors, falsepositive results may be magnified by attempts to increase sensitivity. False-positive tests on blood donors cause unnecessary discard of units not infectious for HCV, plus indefinite deferral and unwarranted health concerns for volunteer donors.’ Using the two FDA-approved EIA-1 anti-HCV test systems, one from Ortho Diagnostics Systems (Raritan, NJ) and the other from Abbott Laboratories (Abbott Park, IL), we evaluated 5216 samples from US donors: 4012 in October 1990 and 1204 in February 1991. Although the starting material, a recombinant HCV protein (C-100),2 is the same for both licensed tests, the frequency of repeatably reactive (RR) results varied between them (Fig. 1). Each RR sample detected by either manufacturer’s kit was then tested by two supplemental methods: RIBA-I1 (second-generation recombinant immunoblot assay, Chiron/Ortho Diagnostic Systems, Raritan NJ) and HCV neutralization (Abbott Laboratories). With Abbott’s EIA test system, we found an anti-HCV RR rate of 0.92 percent, compared with 0.36 percent for the EIA of Ortho. Of the 48 RR samples from the Abbott test, 9 (19%) were RIBA-11-reactive or -indeterminate, and 11 (23%) neutralized. Eight (42%)of 19 RR samples from the Ortho test were RIBA-11-reactive or -indeterminate, and 7 (37%) of 19 neutralized (with two additional equivocal results) (Fig. 2). All nine RIBA-11-reactive samples also neutralized: seven were RR with EIA-1 from both manufacturers, one was RR with the Abbott test only, and one by the Ortho test only (this last sample was considered equivocal on neutralization, because the original test absorbance was
Ortho test system I
NEG
RR
I
NEG
I
3s
13 (.25%)
(-67%)
6 (.12%)
5162 (98.96%)
FIG. 1. Results of HCV EIA in 5216 samples from US blood donors.
/-A Abbott
1-
\ \
Ortho EIA-RR N=lQ
FIG. 2. Results of supplementary tests: N Neutralized; Abbott = and Ortho = %Y. @ RIBA reactive/indeterminate; Abbott = %S and Ortho = YIY.
less than the manufacturer’s cutoff for reactivity.) One RIBA11-indeterminatesample (with 3 reactivity at the C-22 band) was not detected by the Ortho test oust under the cutoff). Of the 13 EIA-1-RR samples that neutralized, including one of two with equivocal results, three (23%) were RIBA-II-nonreactive. The concordance between the anti-HCV EIA-1 screening results and supplementary assays is illustrated in Fig. 2. Abbott’s EIA-1 may be more sensitive than that of Ortho. The Abbott test identified two samples that were RIBA-IIreactive or -indeterminate (and that neutralized), plus two with neutralization only that were missed by the Ortho test; the Ortho test identified one sample that was RIBA-11-reactive and another that equivocally neutralized, both of which were missed by the Abbott test. If specificity equates with the reactivity of supplementary tests, then Ortho’s anti-HCV EIA-1 assay appears to be more ~ p e c i f i c .It~ should be noted that the supplemental tests are for research use only and are not yet considered confirmatory. Additional information is needed to establish that samples from blood donors found to be anti-HCV EIA-1RR and also reactive on one or more supplemental assays correlate with both HCV infectivity and transmission via blood transf~sions.~.~
+
BERNIEBETLACH,MT(ASCP) STEVEN ANDERSON,BS RICK RODRIGUEZ, BS KEN KURAMOTO,h4T KATHLEEN SAZAMA, MD, JD PAULV. HOLLAND,MD Sacramento Medical Foundation Center for Blood Research 1625 Stockton Blvd Sacramento, CA 95816
I
I
I
/
TRANSFUSION Vol. 32, No. 2-1992
LElTERS TO THE EDITOR
References 1. Public Health Service inter-agency guidelines for screening donors of blood, plasma, organs, tissues, and semen for evidence of hepatitis B and hepatitis C. Morbid Mortal Weekly Rep 1991;401-17. 2. Kuo G, Choo QL, Alter HJ, et al. An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 1989;244:362-4. 3. Mimrns L, Vallari D, Ducharme L, et al. Specificity of anti-HCV ELISA assessed by reactivity to three immunodominant HCV regions. Lancet 1990;336:1590-1. 4. Van der Poel CL, Reesink HW,Schaasberg W, et al. Infectivity of blood seropositive for hepatitis C virus antibodies. Lancet 1990;335:558-60. 5. Esteban JI, Gonzalez A, Hernandez J, et al. Evaluation of antibodies to hepatitis C virus in a study of transfusion-associatedhepatitis. N Engl J Med 1990;323:1107-12.
Prestorage white cell filtration of blood To the Editor: The recent editorial by Heal and Cohen’ provided a wellbalanced discussion of the potential benefits and hazards of inline, prestorage filtration of whole blood. I agree with those authors that discussions of the merits of filtration are often clouded by strong, but poorly documented claims made by commercial interests as to the benefits of filtration before blood component production. One potential benefit of prestorage filtration that deserves further study, but was not mentioned in the editorial, is that of improved quality control of the blood components that are
TRANSFUSION
1992-Vol. 32, No. 2
193
LETTERS TO THE EDITOR
meant to be white cell (WBC)-reduced. As stated by the authors, bedside filtration is technique-dependent. Depending on the training, interest, and available time of nursing staff, bedside filters may or may not be used appropriately. What is not clear is whether suboptimal technique leads to suboptimal WBC removal and, thereby, defeats the intended objective of filtration. A related quality issue is the use of third-generation filters for more blood components than recommended by the manufacturers. From my personal experience and that of others, it is not uncommon for a single filter to be used for more than 2 units of red cells, despite specific manufacturer instructions to the contrary. How does such incorrect use affect the effectiveness of WBC removal? Even though I am not advocating the prestorage filtration of blood, I believe the issues raised need careful examination. If data demonstrate that improper
use of bedside WBC filters significantly affects the number of WBCs removed, alternative filtration methods might be viewed more favorably.
MARKA. POPOVSKY,MD American Red Cross Blood Services Northeast Region Headquarters 180 Rustcraf! Road Dedharn, u4 02026
Reference 1. Heal JH, &hen HJ. Do white cells in stored blood components
reduce the likelihood of posttransfusion bacterial sepsis? (editorial). Transfusion 1991;31:581-3.
Submission of Letters Instructions for submission of letters can be found in the Detailed Instructions for Authors published on pages 92 to 97 of the January 1992 issue. Submit letters to: Peter D. Issltt, PhD, FIMLS, FIBlol, FRCPath Transfusion Service Duke University Medical Center, P.O. Box 2928 Durham, NC 27710. EDITOR'S NOTE: In order to permit timely publication of correspondence, the references have not been verified as they have been for articles appearing In Transfusion, and, therefore, the accuracy of cited references in Letters to the Editor Is the sole responsibility of the authors.
