Risk Factors in Transmission of Non-A, Non-B Postt r an sfusio n H epa ti t i s The Role of Hepatitis B Antibody in Donor Blood M. E. CONRAD,R. G. KNODELL, E. L. BRADLEY, JR., E. P. F L A N N E R Y AN , D A. L. G I N S B E R G Front the University of Alabanio. Birniinghani. Alabania: the Univerrity of New Mexico School of Medicine, Albuquerque. New Me-rico; Letternian Army Medical Cenrer. San Francisco, Cali/ornia; and the George Wayhington University Medical Center. Washington. D.C.
Risk of developing icteric hepatitis in a transfusion study involving cardiac surgery patients was 0.2 per cent per unit of blood transfused with the ratio of icteric to anicteric cases being 1:4. Risk of developing hepatitis was proportional to the number of units transfused: one to four units, 4 per cent; six to ten units, 8 per cent; I I to 20 units, 19 per cent; > 21 units, 42 per cent. The prevalence of type B hepatitis was low (6 per cent), with the vast majority of patients being shown to have non-A non-B hepatitis. However, a greater incidence of hepatitis type B serologic events was observed among recipients of antiHBs positive blood than those transfused only with units not containing antibody (p = 0.04). A significantly greater incidence of non-A, non-B hepatitis was obsened among patients transfused with blood containing anti-HBs when compared with a group who received blood without antibody (p < 0.01). Caution should be exercised in interpretation of this difference because patients transfused with blood containing anti-HBs received significantly more units of blood. However, utilization of stepwise regression analysis to unconfound the two dependent variables suggest that the use of blood containing antiHBs increases the hepatitis risk (p = 0.06) although the number of units transfused was the more significant factor (p < 0.001). Additional data from carefully designed studies are needed to determine if donor blood containing anti-HBs significantly increases the risk of transmitting non-A non-B hepatitis.
EXCLUSION of commercial donor blood and blood donors with HBsAg has resulted in a Received for publication June 21, 1976; accepted August I I , 1976. These studies were supported by the United States Army Medical Research and Development Command and by National Institutes of Health grants 2YOlHB50008 and H L I 1310. Address reprint requests to: Division of Hematology and Oncology, University of Alabama, Birmingham A L 35294.
marked decrease in the occurence of type B hepatitis among transfusion recipients. Currently, the vast majority of cases of posttransfusion hepatitis seem to be caused by an unidentified virus o r group of viruses.R.'".'2 However, even ultransensitive methods of testing HBsAg such as radioimmunoassay do not eliminate all carriers of type B hepatitis from blood donor populations because immunological evidence of type B hepatitis transmission both with and without liver enzyme elevations is still observed among transfused recipients. I n addition, HBsAg screening does not appear to have markedly decreased the total overall incidence of posttransfusion hepatitis. I' In an attempt to find methods to further decrease the occurence of posttransfusion hepatitis, investigators have questioned whether blood containing anti-HBs should be used for t r a n S ~ u S j o n ~ ~ . ~ . ~ n . ~ ~ Exclusion ~ ~ 7 . ~ x . ~ ~ of this segment of the donor population would be a logical extention of currently accepted blood banking practices which prohibit the use of donors with a history of hepatitis because carriers of anti-HBs have immunologic evidence of previous exposure to hepatitis type B virus.5 O n e important theoretical reason for examining the incid e n c e of posttransfusion hepatitis among recipients of blood containing anti-HBs is t h e possibility that such units could contain undetectable quantities of HBsAg complexed
5 79 Transfurlon Nov.-Dec. 1977
Volume 17
Number 6
580
CONRAD ET AL
to antibodv which could disassociate following transfusion t o cause hepatitis B.20On the c o n t r a r y , anti-HBs containing transfusions might provide a protective effect against developing t y p e B hepatitis by passive immunization as has been observed in nontransfusion associated type B hepatitis studies when i m m u n e h u m a n s e r u m g a m m a globulin c o n t a i n i n g a n t i - H B s w a s administered prophylactically.1.9.19M o s t previous investigations of t h e effect of transfusing blood containing a n t i - H B s h a v e shown n e i t h e r significant deliterious nor salutory results in regard to transmission of either type B posttransfusion hepatitis or type B hepatitis immunological event^.^.^.^^.^^ H o w e v e r , in a recently reported study of t h e efficacy of imm u n e h u m a n s e r u m g a m m a globulin preparations in t h e prevention of posttransfusion he pa ti ti^'^, transfusion of anti-HBs containing blood appeared to be associated with a markedly increased incidence of nonA non-B posttransfusion hepatitis. I t was, however, emphasized t h a t this association w a s confounded by t h e increased blood requirement in t h e g r o u p who received antiHBs t r a n s f u s i o n s . I n t h i s p a p e r , f u r t h e r statistical analysis of t h e anti-HBs transfus i o n a l d a t a f r o m o u r e a r l i e r r e p o r t is presented which a t t e m p t s t o s e p a r a t e t h e two dependent transfusion variables and continues t o support a possible relationship between transfusion of donor units containing anti-HBs and t h e development of non-A non-B posttransfusion hepatitis.
Subjects, Methods, and Materials Description of the study design and methods has been presented previously.17 Briefly, subjects were patients admitted for cardiac surgery to either Walter Reed or Letterman Army Medical Centers between August, 1972 and December, 1974. These subjects were volunteers in a study to determine the effect of immune serum gamma globulin (ISG) preparations upon the occurrence of posttransfusion hepatitis. One-third of the patients were injected with a placebo solution
Transfusion July-Aug. 1977
whereas the remainder received ISG .preparations . intramuscularly prior to surgery and blood transfusion. Data suggesting an efficacy for the ISG preparations have been reported previously.I7 Blood used at surgery was obtained from volunteer donors through t h e A r m y Blood Program and the Irwin Memorial Blood Bank. Units were screened for HBsAg prior to transfusion by count ere1ect rophoresis while all-donor blood and patient sera were ultimately screened for purposes of the study by radioimmunoassay for both HBsAg (Ausria) and anti-HBs (Ausab). Positive HBsAg results were confirmed by specificity testing.I4 Investigation of serum specimens for antibody to hepatitis A virus was performed by immune electron microscopy in patients suspected of having hepatitis p o ~ t o p e r a t i v e l y .Serologic ~ studies to detect antibody to cytomegalovirus and Epstein-Barr virus were performed on serial serum specimens by complement fixation and immunofluorescent techniques, respectively.’2.1s Three patients had evidence of hepatitis B infection and t h r e e showed serologic changes consistent with cytomegalovirus infection. These patients were excluded from analysis in this report. No hepatitis patient showed serologic evidence for infection with either hepatitis A or Epstein-Barr virus. Patients were classified as having posttransfusion hepatitis if I ) preoperative liver enzymes were normal, 2) a serum transaminase (SGPT) greater than 2.5 times the upper limit of normal was found from 14 to 180 days postoperatively, and 3) no other cause for abnormal liver function tests was identified. All suspected cases of hepatitis were reviewed by an independent panel appointed by the National Heart and Lung Institote Policy Board for Hepatitis Trialsz6 which classified cases as “probable” and “possible.” Possible cases fulfilled the first two criteria cited above. It was believed that hepatitis was as likely an explanation as any other for the SGPT elevation, but that other potential causes could not be excluded. Results have been expressed in terms of probable cases and total hepatitis cases. Fisher’s exact test (two-tailed) was used to calculate statistical differences among groups with enumeration data.‘ The unpaired t test was utilized to calculate statistical differences for measurement data.6 Measurement data were normalized by either logarithmic or square root transformation as indicated in the text. Multiple classification data were examined by analysis of variance.25 Except as reported in the text, there were no statistically significant differences in the distribution of patients by age or sex in the various groups in this study (p > 0.25).
Volume 17
58 I
RISK FACTORS IN HEPATITIS
Numbcr4
Table 2. Relationship of Development of Serologic Evidence of Hepatitis Type 6 Exposure and the Detection of Anti-H6s in Transfused Units of Blood
Table 1. Relationship of Number of Donor Units to Occurrence of Non-A Non-6 Posttransfusion Hepatitis
Number of Units Transfused
Number of Subjects
Hepatitis
Patients with Hepatitis %
1 to5 6 t o 10 11 t o 2 0 21 + Totals
23 118 88 26 255
1 9 17 11 38
4.3 7.6 19.3 42.3 14.9
Recipient Response
Transfused Blood No Anti-HBs
Anti-HBs
Total
~
HBsAg or Anti-HBs No Serologic evidence of
3
10
13
121
121
242
124
131
255
exposure
Totals
Results Aliquots of all-donor blood were available for 255 patients who did not develop either type A or type B posttransfusion hepatitis based on serologic testing of serial serum samples; 124 patients received no anti-HBs positive transfusions while 13 1 received transfusions containing anti-HBs. Thirty-eight of these 255 subjects developed non-A non-B posttransfusion hepatitis, 7 of whom were icteric. The mean number of blood transfusions was I2 with a range of 1 to 55 units. Thus the risk of developing non-A non-B hepatitis in this study was 1.23 per cent per unit transfused and 0.23 per cent per unit if only icteric cases were considered. The risk of developing nonA non-B hepatitis increased in direct proportion to t h e number of units of blood which were transfused (Table I). Among the 255 subjects, there were 13 patients who showed serologic evidence of exposure to hepatitis type B even though they did not develop hepatitis (Table 2). Each of these subjects was transfused only with HBsAg negative units by radioimmunoassay. Ten of t h e 13 subjects received oFe or more units of blood containing anti-HBs and three did not (p = 0.04). This observation was not explained by significant differences in the number of units transfused in each group(mean, 12.3 versus 10.4 units).
The relationship of the pretransfusion serologic status and occurrence of non-A non-B type posttransfusion hepatitis in our patients is reported in Table 3. Patients with serologic evidence of antiHBs in their serum prior to transfusion were not provided protection against non-A non-B type posttransfusion hepatitis as would be anticipated if the two diseases were caused by immonologically different viruses. On the contrary, patients with anti-HBs had a slightly greater risk of developing non-A non-B type hepatitis (p = 0.02). This is not explained by differences in the percentage of patients who received units containing anti-HBs or the number of units of blood transfused in each group (mean, 12.3 versus 12.1 units). However, patients who had anti-HBs prior to transfusion were slightly older than patients who had no evidence of antibody (47.8 14.3 versus 43.5 ;t 14.7 years, p = 0.05). Comparison of the occurrence of non-A non-B hepatitis in patients who were transfused with one or more units of blood containing anti-HBs versus patients who received only blood which contained no detectible antibody showed a significantly greater number of patients in the former group (Table 4). This difference persisted whether only probable (p = 0.01) or probable plus possible cases (p = 0.004) were considered for analysis. All patients who developed icteric hepatitis were in the group which received blood containing anti-
*
Table 3. Relationship o f Pretransfusion Serologic Status and Occurrence of Non-A, Non-6 Posttransfusion Hepatitis Pretransfusion Serologic Status of Recipient Anti-HBs
No Anti-HBs
Number of Recipients
57
198
Serologic Status of Donor Blood
Number o f Recipients
Non-A, Non-B Hepatitis
Anti-HBs
31
10
N o Anti-HBs
26
4
100
ia
98
6
255
38
Anti-HBs No Anti-HBs
Totals
255
582
TronsfuGm July-Aug. 1911
C O N R A D ET AL
Table 5. Association of Anrl-nBs Blood Transfusion with Severity of Non-A, Non-B Posttransfusion Hepatitis
Table 4. Association of Anti-HBs Blood Transfusion with Occurrence of Non-A, Non-B Posttransfusion Hepatitis (255 Pa tien tsl Anti-HBs
No Anti-HBs
28 103 131
10 114 124
He pati tis No hepatitis .Totals
lcteric hepatitis Probable anicteric Possible anicteric Totals
HBs (Table 5 ) . That the differences observed were not related to the administration of gamma globulin preparations was suggested by sephrate analysis of data from the placebo-treated patients where non-A non-B hepatitis was also more common among the group who received anti-HBs containing transfusions (p = 0.03, Table 6). . A comparison of the number of transfused units of blood containing anti-H Bs and the occurrence of non-A non-B hepatitis in recipients is shown in Table 7. Among patients who received anti-HBs positive units of blood, the 28 hepatitis patients received a mean of 2.32 anti-HBs positive units whereas the 103 patients who did not develop hepatitis received only 1.79 units of anti-HBs positive blood. When these data were normalized by square root transformation, geometric means of 1.97 i 0.37 and 1.62 f 0.15 were observed in the hepatitis patients and nonhepatitis patients respectively. The difference between the two groups was significant at the p < 0.01 level. Even though there seemed to be a significant difference between the occurrence of non-A non-B hepatitis among patients transfused with blood containing anti-HBs versus those who received no blood containing anti-H Bs, these data were confounded by the differences in the total number of transfusions received by subjects in various groups in the study (Table 8). Since anti-HBs was detected in only 8 per cent of the total units o f donor blood, i t is not surprising that patients who received units of blood containing anti-HBs would be those who received more units of blood at surgery (14.2 f 9.4 versus 9.8 f 5.3) irrespective of whether o r not they developed hepatitis. Comparisons made with relationship to the number o f units of blood received and the occurrence or nonoccurrence of hepatitis irrespective o f whether
Anti-HBs
No Anti-HBs
7 7 14 28
0 3 7 10
the patient received anti-H Bs positive transfusions revealed non-A non-B hepatitis was associated with transfusion o f an increased number of units of blood (17.8 + 11.9 versus I 1 .0 f 6.6). Analyses of these data by stepwise regression elimination for dependent variables and analysis of variance tables revealed that the occurence of hepatitis could be best related statistically to the number of units of blood transfused ( p < 0.001) with less statistical assurance of a relationship between the development of hepatitis and transfusion with anti-HBs positive blood ( p = 0.06).
Discuss ion I t has been previously recognized that the incidence of posttransfusion hepatitis is greater among patients who receive blood from a large number of donors.2' I f only icterus is used as the criterion for detection of hepatitis, an incidence of approximately 0.2 per cent per unit transfused occurred with prospective follow-up of blood recipients in this study. When serial transaminase determinations are obtained at intervals after transfusion, a fivefold increase in detection of hepatitis was seen. These anicteric cases of posttransfusion hepatitis cannot be disregarded as clinically unimportant as these patients may develop chronic hepatic disease and represent a source of potential infection to persons with whom they have close personal c o n t a c t . T h e s e d a t a a t t e s t to t h e
Table 6. Association of Anti-HBs Blood Transfusion and Non-A Non-B Posttransfusion Hepatitis with Prophylaxis with Gamma Globulin or Placebo (255 Patients) Placebo (81)
Hepatitis
No Hepatitis Total
Low Titer 1SG (86)
High Titer 1SG (88)
Anti-HBs
No Anti-HBs
Anti-HBs
No Anti-HBs
Anti-HBs
No Anti-HBs
14 33 47
3 31 34
7
36
3 40 43
7 34 41
4 43 47
43
Volume 17 NumbcrJ
583
RISK FACTORS IN HEPATITIS
continued importance of posttransfusion hepatitis as a significant complication of blood transfusion even when volunteer donors who have been tested for HBsAg are utilized. Since development of hepatitis seems to increase in direct proportion to the number of donor units transfused, risk-to-benefit ratios should be considered by health personnel for each unit of blood transfused. During the previous decade, the majority of posttransfusion hepatitis cases were caused by hepatitis type B Since currently only approximately 10 per cent of blood recipients who develop hepatitis can be shown immunologically to have hepatitis BX.I6..”’laboratory testing of donor blood for HBsAg has been relatively effective in diminishing transmission of hepatitis B by blood transfusion. These latter cases seem to be due to an inability of even the most sensitive methods of screening for HBsAg to detect certain carriers. I n the present study, development of hepatitis B was seen in three patients and the de novo appearance of either HBsAg o r anti-HBs without evidence of liver dysfunction o c c u r r e d i n 13 additional patients who received blood which was negative for HBsAg by radioirnmunoassay.” This inability to detect all carriers of hepatitis B virus by HBsAg testing has led investigators to study whether blood containing anti-HBs
Table 7. Number of Units of Anti-HBs Blood Transfused versus Occurrence of Non-A Non-B Hepatitis Units Transfused Containing Anti-HBs
0 1 2 3 4 >4
Number of Patients
Patients with Hepatitis, %
l o ( 8.1) 14 (19.4) 8 (22.9) 2 (14.3) 1 (33.3) 3 (42.8)
124 72 35 14 3 7
was responsible for transmitting hepatitis B,2.:1.10.18.17.IX.24 It has been shown in vitro t h a t an e x c e s s of a n t i - H b s in s e r u m specimens will complex HBsAg so that the antigen cannot be detected immunologically.’” Theoretically, the antigen-antibody complex could disassociate after transfusion to produce hepatitis type B. Interestingly, 10 of the 13 patients who received only HBsAg negative transfusions but still developed a de n o w serologic response without liver enzyme elevations indicating exposure to hepatitis B virus got transfusions containing anti-HBs. This incidence of B serologic response in recipients of anti-H Bs containing transfusions was significantly greater than that seen in the group which received only anti-HBs negative blood ( p = 0.04). This would seem to
Table 8. Association of Anti-HB Blood Transfusion with Occurrence of Non-A Non-B Hepatitis and the Mean Number o f Units Transfused Anti-HBs
No Anti-HBs
Total
Hepatitis Arith. Geom.
(28patients) 20.4i 12.5 (2.35) 17.5 k 1.7(1.11)
(10patients) 10.4 t 6.4 (2.03) 8.7 A 1.9(1.24)
(38patients) 17.8f 11.9 (1.94) 14.62 1.9(1.11)
No Hepatitis Arith. Geom.
(103patients) 12.5 f 7.6 ( .75) 10.5 f 1.8(1.06)
(1 14 patients) 9.8 * 5.2 ( .49) 8.5 f 1.9 (1.05)
(217 patients) 11.0* 6.6( .45) 9.4 * 1.8 (1.04)
Total Arith. Geom.
(131 patients) 14.2t 9.4 ( .82) 11.8 f 1.8 (1.05)
(124patients) 9.8 f 5.3( .47) 8.5 t 1.8 (1.05)
NOTE: The association of transfusion of blood containing anti-HBs with the nccurence o f non-A non-B hepatitis i s shown above and compared with patients who received blood without antibody. The mean number of units of blood transfused intn patients in each group with one standard deviation from the arithmetic and geometric mean are shown. The numbers in parentheses are the standard error of the mean. The geometric mean was calculated using conversion into common logarithms.
Transfusion
C O N R A D ET A L Table 9. Published Reports of the Effect of Transfusion of Anti-HBs Positive Blood upon the Occurrence of Posttransfusion Hepatitis Ratio of Hepatitis Patients t o Transfused Subjectst References
Anti- H 6s Transfused
No Anti-HBs
Alter et a1.3 Goldfield11 Gocke and Panicklo Aach et a/.* Koretz et al.18 Current Study
2/37 5/28 4/35 131169 15/45 281131
0114 14/98 2211 33 81189 6/18 101124
~~
tData cited exclude cases of type B hepatitis from numerator and denominator.
s u p p o r t c e r t a i n of t h e d a t a r e c e n t l y published by Koretz et ui.” in which the transfusion of large amounts of anti-HBs was associated with an increased incidence of serologic events related to hepatitis type B. However, there is no convincing evidence that the exclusion of donors with anti-HBs will reduce the occurrence of clinically detectable type B hepatitis in transfusion recipients. There is also no convincing d a t a to support a protective effect for anti-HBs transfusions since a similar incidence of hepatitis type B has been seen among recipients transfused with anti-HBs positive and negat ive blood. p.3. l o I n this article we have examined the relationship to anti-HBs positive transfusions and the subsequent development of non-A, non-B hepatitis. It was postulated t h a t donors with immunologic evidence of exposure to one hepatitis agent might be more likely to harbor other hepatitis agents for which there a r e presently no serologic tests. Unlike previous studies (Table 9), a significantly greater incidence of posttransfusion hepatitis was seen among patients who received blood containing anti-HBs than among patients transfused only with units containing no antibody (p=0.004). Much of the statistical difference observed could be attributed to the transfusion of more units of blood to the group that received one or more units containing anti-HBs. This was not surprising since only 8 per cent of the units utilized contained anti-HBs and the chances
July-Aug. 1977
of receiving one of these units increase in proportion to the number of units transfused. H o w e v e r , while t h e n u m b e r of u n i t s transfused was the more highly significant risk factor in the occurrence of posttransfusion hepatitis, stepwise regression analysis provides suggestive evidence that transfusion of anti-HBs containing blood also placed patients in this study at a higher risk of developing non-A non-B posttransfusion hepatitis (p = 0.06). With one exception, the previous studies attesting to the safety of transfusion o f blood containing anti-HBs have had relatively few patients at risk (Table 9). Whether the difference in our results from those noted in Aach’s study’ might be explained by differences in the transfusion requirements for the various patient groups cannot be determined from the text. The possibility that anti-HBs may act as a m a r k e r in donor populations to indicate blood carrying an increased risk of transmitting non-A, non-B hepatitis is intriguing. It suggests that environments exposing individuals to hepatitis B also contain non-B hepatitis agents and that anti-HBs testing of donor blood might provide an additional means of diminishing transmission of posttransfusion hepatitis. However, it is evident that even if this hypothesis was convincingly substantiated, exclusion of donors with antiHBs would merely reduce transmission and not eliminate the hazard. Additional information is needed with regard to the possible deleterious effects of anti-H Bs containing transfusions since this antibody is found in 5 to 10 per cent of volunteer blood donors and loss of these donor units would constitute a significant problem to the national blood program. Finally, any study which examines the effect of transfusing blood containing anti-HBs must carefully consider the number of units received by each recipient before the conclusions of t h e study can be considered credible. References I.
A Cooperative Study: Prophylactic gamma globulin for prevention of endemic hepatitis. EfTects of US. gamma globulin upon the incidence of
Volume 17 Number4
2.
3.
4.
5.
6. 7.
8.
9.
10.
11.
12.
13. 14.
15.
16.
RISK FACTORS IN HEPATITIS
viral hepatitis and other infectious diseases in U.S. soldiers abroad. A r c h . I n t e r n . Med. 128:723, 1971. Aach, R. D., H. J. Alter, F. B. Hollinger, P. V. Holland, J. J. Lander, J. L. Melnick, and J. M. Weiler: Risk of transfusing blood containing antibody to hepatitis B surface antigen. Lancet 2:190. 1974. Alter, H. J., P. V. Holland, R . H. Purcell, J. J. Lander, S. M. Feinstone, A. G. Morrow, and P. J. Schmidt: Posttransfusion hepatitis after exclusion of commercial and hepatitis-B antigenpositivedonors. Ann. Intern. Med. 77:691, 1972. Batson, H. C.: An Introduction to Statistics in the Medical Sciences. Minneapolis, Minn., Burgess Publishing Company, 1956, p. 48. Barker, L. F., M. R. Peterson, N. R . Shulman, and R . Murray: Antibody responses in viral hepatitis, type B. JAMA 223:1005, 1973. Edwards, A. L.: Statistical Methods. New York, Holt, Reinhart and Winston, 1967, p. 200. Feinstone, S. M., .A. 2. Kapikian, and R. H. Purcell: Hepatitis A: Detection by immune electron microscopy of a viruslike antigen associated with acute illness. Science 182:1026, 1973. -,A. Z. Kapikian, R . H. Purcell, H. J. Alter, , and P. V. Holland: Transfusion-associated hepatitis not due to viral hepatitis type A or B. N. Engl. J. Med. 292:767, 1975. Ginsberg, A. L., M. E. Conrad, W. H. Bancroft, C. M. Ling, and L. R. Overby: Prevention of endemic HAA-positive hepatitis with gamma globulin: Use of a simple radioimmune assay to detect HAA. N. Engl. J . Med. 286562, 1972. Gocke, D. J., and J. M. Panick: Transfusion of blood containing anti-HBAg. In: Hepatitis and Blood Transfusion. G. N. Vyas, H. A. Perkins, and R. Schmid. Eds. New York, Grune and Stratton, 1972, p. 319. Goldfield, M.: Some epidemiological studies of transfusion-associated hepatitis. In: Transmissible Disease and Blood Transfusion. T. J. Greenwalt, and G. A. Jamieson. Eds. New York, Grune and Stratton, 1974, p. 141. Henle, G., and W. Henle: lmmunofluorescence in cells derived from Burkitt's lymphoma. J. Bacteriol. 91:1248, 1966. International Forum: Does blood with antibodies against hepatitis B carry an increased risk to transmit hepatitis? Vox Sang. 28:460, 1975. Irwin, G. R., A. M. Alien, W. H. Bancroft, M. Willhight, and P. K. Russell: Specificity and sensitivity of radioimmunoassay for hepatitis B antigen. Appl. Microbiol. 28:600, 1974. Kent, J. F., and E. H. Fife, Jr.: Precise standardization of reagents for complement fixation. Am. J.Trop. Med. 12:103, 1963. Knodell, R. G., M. E. Conrad, J . L. Deinstag, and C. J. Bell: Etiological spectrum of post-transfusion hepatitis. Gastroenterology 69:1278, 1975.
585
17. -,
18.
19.
20.
21.
22.
23.
24.
25.
26.
M. E. Conrad, A. L. Ginberg, C. J. Bell, and E. P. Flannery: Efficacy of prophylactic gamma-globulin i n preventing non-A non-B posttransfusion hepatitis. Lancet 1557. 1976. Koretz, R. L., L. R. Overby, and G. L. Gitnick: Post-transfusion hepatitis: The role of hepatitis B antibody. Gastroenterology 70556, 1976. Krugman, S., and J. P. Giles: Viral hepatitis type B (MS-2-strain): Further observations on natural history and prevention. N. Engl. J. Med. 288:755, 1973. Ling, C. M., and L. R. Overby: Prevalence of hepatitis B virus antigen as revealed by direct radioimmunoassay with 'lSl antibody. J. Irnmunol. 109:834, 1972. Mollison, P. L.: Blood Transfusion in Clinical Medicine, 5th ed. London, Blackwell Scientific, 1972, p. 609. Prince, A. M., B. Brotman, G. F. Grady, W. J. Kuhns, C . Hazzi, R. W. Levine, and S. J. Millian: Long incubation post-transfusion hepatitis without serological evidence of exposure to hepatitis B virus. Lancet 2:241, 1974. Prince, A. M., R. L. Hargrove, W. Szmuness, C. E. Cherubin, V. J. Fontana, and G. H. JetTries: Immunologic distinction between infectious and serum hepatitis. N. Engl. J. Med. 282:987, 1970. SeetT, L. B., H. J. Zimmerman, and E. C. Wright: V. A. Cooperative study of gamma globulin prophylaxis o f post-transfusion hepatitis. Gastroenterology 643393, 1973. Snedecor, G . W. and W. G. Cochran: Statistical Methods, 6th ed. Ames, Iowa, Iowa S t a t e University Press. 1967, p. 299. Surgenor, D. Mac N., T. C. Chalmers, M. E. Conrad, W. T. Friederwald, G . F. Grady, M . Hamilton, J. W. Mosely, .A. M. Prince, and J. M. Stengle: Clinical trials-of hepatitis B immune globulin. Development of policies and materials for the 1972-1975 studies sponsored by the National Heart and Lung Institute. N. Engl. J. Med. 293:1060, 1975.
M. E. Conrad, M.D., Division of Hematology and Oncology, University of Alabama, Birmingham, Alabama. R. G. Knodell, M.D., Division of Gastroenterology, University of New Mexico School of Medicine, Albuquerque, New Mexico. E. L. Bradley, Jr., Ph.D., Department of Biostatistics, University of Alabama, Birmingham, Alabama. E. P. Flannery, M.D., Hematology and Oncology Service, L e t t e r m a n Army Medical C e n t e r , S a n Francisco, California. A. L. Ginsberg, M.D., Division of Gastroenterology, George Washington University Medical Center, Washington, D.C.
Risk of developing icteric hepatitis in a transfusion study involving cardiac surgery patients was 0.2 per cent per unit of blood transfused with the ratio of icteric to anicteric cases being 1:4. Risk of developing hepatitis was proportional to the number of units transfused: one to four units, 4 per cent; six to ten units, 8 per cent; I I to 20 units, 19 per cent; > 21 units, 42 per cent. The prevalence of type B hepatitis was low (6 per cent), with the vast majority of patients being shown to have non-A non-B hepatitis. However, a greater incidence of hepatitis type B serologic events was observed among recipients of antiHBs positive blood than those transfused only with units not containing antibody (p = 0.04). A significantly greater incidence of non-A, non-B hepatitis was obsened among patients transfused with blood containing anti-HBs when compared with a group who received blood without antibody (p < 0.01). Caution should be exercised in interpretation of this difference because patients transfused with blood containing anti-HBs received significantly more units of blood. However, utilization of stepwise regression analysis to unconfound the two dependent variables suggest that the use of blood containing antiHBs increases the hepatitis risk (p = 0.06) although the number of units transfused was the more significant factor (p < 0.001). Additional data from carefully designed studies are needed to determine if donor blood containing anti-HBs significantly increases the risk of transmitting non-A non-B hepatitis.
EXCLUSION of commercial donor blood and blood donors with HBsAg has resulted in a Received for publication June 21, 1976; accepted August I I , 1976. These studies were supported by the United States Army Medical Research and Development Command and by National Institutes of Health grants 2YOlHB50008 and H L I 1310. Address reprint requests to: Division of Hematology and Oncology, University of Alabama, Birmingham A L 35294.
marked decrease in the occurence of type B hepatitis among transfusion recipients. Currently, the vast majority of cases of posttransfusion hepatitis seem to be caused by an unidentified virus o r group of viruses.R.'".'2 However, even ultransensitive methods of testing HBsAg such as radioimmunoassay do not eliminate all carriers of type B hepatitis from blood donor populations because immunological evidence of type B hepatitis transmission both with and without liver enzyme elevations is still observed among transfused recipients. I n addition, HBsAg screening does not appear to have markedly decreased the total overall incidence of posttransfusion hepatitis. I' In an attempt to find methods to further decrease the occurence of posttransfusion hepatitis, investigators have questioned whether blood containing anti-HBs should be used for t r a n S ~ u S j o n ~ ~ . ~ . ~ n . ~ ~ Exclusion ~ ~ 7 . ~ x . ~ ~ of this segment of the donor population would be a logical extention of currently accepted blood banking practices which prohibit the use of donors with a history of hepatitis because carriers of anti-HBs have immunologic evidence of previous exposure to hepatitis type B virus.5 O n e important theoretical reason for examining the incid e n c e of posttransfusion hepatitis among recipients of blood containing anti-HBs is t h e possibility that such units could contain undetectable quantities of HBsAg complexed
5 79 Transfurlon Nov.-Dec. 1977
Volume 17
Number 6
580
CONRAD ET AL
to antibodv which could disassociate following transfusion t o cause hepatitis B.20On the c o n t r a r y , anti-HBs containing transfusions might provide a protective effect against developing t y p e B hepatitis by passive immunization as has been observed in nontransfusion associated type B hepatitis studies when i m m u n e h u m a n s e r u m g a m m a globulin c o n t a i n i n g a n t i - H B s w a s administered prophylactically.1.9.19M o s t previous investigations of t h e effect of transfusing blood containing a n t i - H B s h a v e shown n e i t h e r significant deliterious nor salutory results in regard to transmission of either type B posttransfusion hepatitis or type B hepatitis immunological event^.^.^.^^.^^ H o w e v e r , in a recently reported study of t h e efficacy of imm u n e h u m a n s e r u m g a m m a globulin preparations in t h e prevention of posttransfusion he pa ti ti^'^, transfusion of anti-HBs containing blood appeared to be associated with a markedly increased incidence of nonA non-B posttransfusion hepatitis. I t was, however, emphasized t h a t this association w a s confounded by t h e increased blood requirement in t h e g r o u p who received antiHBs t r a n s f u s i o n s . I n t h i s p a p e r , f u r t h e r statistical analysis of t h e anti-HBs transfus i o n a l d a t a f r o m o u r e a r l i e r r e p o r t is presented which a t t e m p t s t o s e p a r a t e t h e two dependent transfusion variables and continues t o support a possible relationship between transfusion of donor units containing anti-HBs and t h e development of non-A non-B posttransfusion hepatitis.
Subjects, Methods, and Materials Description of the study design and methods has been presented previously.17 Briefly, subjects were patients admitted for cardiac surgery to either Walter Reed or Letterman Army Medical Centers between August, 1972 and December, 1974. These subjects were volunteers in a study to determine the effect of immune serum gamma globulin (ISG) preparations upon the occurrence of posttransfusion hepatitis. One-third of the patients were injected with a placebo solution
Transfusion July-Aug. 1977
whereas the remainder received ISG .preparations . intramuscularly prior to surgery and blood transfusion. Data suggesting an efficacy for the ISG preparations have been reported previously.I7 Blood used at surgery was obtained from volunteer donors through t h e A r m y Blood Program and the Irwin Memorial Blood Bank. Units were screened for HBsAg prior to transfusion by count ere1ect rophoresis while all-donor blood and patient sera were ultimately screened for purposes of the study by radioimmunoassay for both HBsAg (Ausria) and anti-HBs (Ausab). Positive HBsAg results were confirmed by specificity testing.I4 Investigation of serum specimens for antibody to hepatitis A virus was performed by immune electron microscopy in patients suspected of having hepatitis p o ~ t o p e r a t i v e l y .Serologic ~ studies to detect antibody to cytomegalovirus and Epstein-Barr virus were performed on serial serum specimens by complement fixation and immunofluorescent techniques, respectively.’2.1s Three patients had evidence of hepatitis B infection and t h r e e showed serologic changes consistent with cytomegalovirus infection. These patients were excluded from analysis in this report. No hepatitis patient showed serologic evidence for infection with either hepatitis A or Epstein-Barr virus. Patients were classified as having posttransfusion hepatitis if I ) preoperative liver enzymes were normal, 2) a serum transaminase (SGPT) greater than 2.5 times the upper limit of normal was found from 14 to 180 days postoperatively, and 3) no other cause for abnormal liver function tests was identified. All suspected cases of hepatitis were reviewed by an independent panel appointed by the National Heart and Lung Institote Policy Board for Hepatitis Trialsz6 which classified cases as “probable” and “possible.” Possible cases fulfilled the first two criteria cited above. It was believed that hepatitis was as likely an explanation as any other for the SGPT elevation, but that other potential causes could not be excluded. Results have been expressed in terms of probable cases and total hepatitis cases. Fisher’s exact test (two-tailed) was used to calculate statistical differences among groups with enumeration data.‘ The unpaired t test was utilized to calculate statistical differences for measurement data.6 Measurement data were normalized by either logarithmic or square root transformation as indicated in the text. Multiple classification data were examined by analysis of variance.25 Except as reported in the text, there were no statistically significant differences in the distribution of patients by age or sex in the various groups in this study (p > 0.25).
Volume 17
58 I
RISK FACTORS IN HEPATITIS
Numbcr4
Table 2. Relationship of Development of Serologic Evidence of Hepatitis Type 6 Exposure and the Detection of Anti-H6s in Transfused Units of Blood
Table 1. Relationship of Number of Donor Units to Occurrence of Non-A Non-6 Posttransfusion Hepatitis
Number of Units Transfused
Number of Subjects
Hepatitis
Patients with Hepatitis %
1 to5 6 t o 10 11 t o 2 0 21 + Totals
23 118 88 26 255
1 9 17 11 38
4.3 7.6 19.3 42.3 14.9
Recipient Response
Transfused Blood No Anti-HBs
Anti-HBs
Total
~
HBsAg or Anti-HBs No Serologic evidence of
3
10
13
121
121
242
124
131
255
exposure
Totals
Results Aliquots of all-donor blood were available for 255 patients who did not develop either type A or type B posttransfusion hepatitis based on serologic testing of serial serum samples; 124 patients received no anti-HBs positive transfusions while 13 1 received transfusions containing anti-HBs. Thirty-eight of these 255 subjects developed non-A non-B posttransfusion hepatitis, 7 of whom were icteric. The mean number of blood transfusions was I2 with a range of 1 to 55 units. Thus the risk of developing non-A non-B hepatitis in this study was 1.23 per cent per unit transfused and 0.23 per cent per unit if only icteric cases were considered. The risk of developing nonA non-B hepatitis increased in direct proportion to t h e number of units of blood which were transfused (Table I). Among the 255 subjects, there were 13 patients who showed serologic evidence of exposure to hepatitis type B even though they did not develop hepatitis (Table 2). Each of these subjects was transfused only with HBsAg negative units by radioimmunoassay. Ten of t h e 13 subjects received oFe or more units of blood containing anti-HBs and three did not (p = 0.04). This observation was not explained by significant differences in the number of units transfused in each group(mean, 12.3 versus 10.4 units).
The relationship of the pretransfusion serologic status and occurrence of non-A non-B type posttransfusion hepatitis in our patients is reported in Table 3. Patients with serologic evidence of antiHBs in their serum prior to transfusion were not provided protection against non-A non-B type posttransfusion hepatitis as would be anticipated if the two diseases were caused by immonologically different viruses. On the contrary, patients with anti-HBs had a slightly greater risk of developing non-A non-B type hepatitis (p = 0.02). This is not explained by differences in the percentage of patients who received units containing anti-HBs or the number of units of blood transfused in each group (mean, 12.3 versus 12.1 units). However, patients who had anti-HBs prior to transfusion were slightly older than patients who had no evidence of antibody (47.8 14.3 versus 43.5 ;t 14.7 years, p = 0.05). Comparison of the occurrence of non-A non-B hepatitis in patients who were transfused with one or more units of blood containing anti-HBs versus patients who received only blood which contained no detectible antibody showed a significantly greater number of patients in the former group (Table 4). This difference persisted whether only probable (p = 0.01) or probable plus possible cases (p = 0.004) were considered for analysis. All patients who developed icteric hepatitis were in the group which received blood containing anti-
*
Table 3. Relationship o f Pretransfusion Serologic Status and Occurrence of Non-A, Non-6 Posttransfusion Hepatitis Pretransfusion Serologic Status of Recipient Anti-HBs
No Anti-HBs
Number of Recipients
57
198
Serologic Status of Donor Blood
Number o f Recipients
Non-A, Non-B Hepatitis
Anti-HBs
31
10
N o Anti-HBs
26
4
100
ia
98
6
255
38
Anti-HBs No Anti-HBs
Totals
255
582
TronsfuGm July-Aug. 1911
C O N R A D ET AL
Table 5. Association of Anrl-nBs Blood Transfusion with Severity of Non-A, Non-B Posttransfusion Hepatitis
Table 4. Association of Anti-HBs Blood Transfusion with Occurrence of Non-A, Non-B Posttransfusion Hepatitis (255 Pa tien tsl Anti-HBs
No Anti-HBs
28 103 131
10 114 124
He pati tis No hepatitis .Totals
lcteric hepatitis Probable anicteric Possible anicteric Totals
HBs (Table 5 ) . That the differences observed were not related to the administration of gamma globulin preparations was suggested by sephrate analysis of data from the placebo-treated patients where non-A non-B hepatitis was also more common among the group who received anti-HBs containing transfusions (p = 0.03, Table 6). . A comparison of the number of transfused units of blood containing anti-H Bs and the occurrence of non-A non-B hepatitis in recipients is shown in Table 7. Among patients who received anti-HBs positive units of blood, the 28 hepatitis patients received a mean of 2.32 anti-HBs positive units whereas the 103 patients who did not develop hepatitis received only 1.79 units of anti-HBs positive blood. When these data were normalized by square root transformation, geometric means of 1.97 i 0.37 and 1.62 f 0.15 were observed in the hepatitis patients and nonhepatitis patients respectively. The difference between the two groups was significant at the p < 0.01 level. Even though there seemed to be a significant difference between the occurrence of non-A non-B hepatitis among patients transfused with blood containing anti-HBs versus those who received no blood containing anti-H Bs, these data were confounded by the differences in the total number of transfusions received by subjects in various groups in the study (Table 8). Since anti-HBs was detected in only 8 per cent of the total units o f donor blood, i t is not surprising that patients who received units of blood containing anti-HBs would be those who received more units of blood at surgery (14.2 f 9.4 versus 9.8 f 5.3) irrespective of whether o r not they developed hepatitis. Comparisons made with relationship to the number o f units of blood received and the occurrence or nonoccurrence of hepatitis irrespective o f whether
Anti-HBs
No Anti-HBs
7 7 14 28
0 3 7 10
the patient received anti-H Bs positive transfusions revealed non-A non-B hepatitis was associated with transfusion o f an increased number of units of blood (17.8 + 11.9 versus I 1 .0 f 6.6). Analyses of these data by stepwise regression elimination for dependent variables and analysis of variance tables revealed that the occurence of hepatitis could be best related statistically to the number of units of blood transfused ( p < 0.001) with less statistical assurance of a relationship between the development of hepatitis and transfusion with anti-HBs positive blood ( p = 0.06).
Discuss ion I t has been previously recognized that the incidence of posttransfusion hepatitis is greater among patients who receive blood from a large number of donors.2' I f only icterus is used as the criterion for detection of hepatitis, an incidence of approximately 0.2 per cent per unit transfused occurred with prospective follow-up of blood recipients in this study. When serial transaminase determinations are obtained at intervals after transfusion, a fivefold increase in detection of hepatitis was seen. These anicteric cases of posttransfusion hepatitis cannot be disregarded as clinically unimportant as these patients may develop chronic hepatic disease and represent a source of potential infection to persons with whom they have close personal c o n t a c t . T h e s e d a t a a t t e s t to t h e
Table 6. Association of Anti-HBs Blood Transfusion and Non-A Non-B Posttransfusion Hepatitis with Prophylaxis with Gamma Globulin or Placebo (255 Patients) Placebo (81)
Hepatitis
No Hepatitis Total
Low Titer 1SG (86)
High Titer 1SG (88)
Anti-HBs
No Anti-HBs
Anti-HBs
No Anti-HBs
Anti-HBs
No Anti-HBs
14 33 47
3 31 34
7
36
3 40 43
7 34 41
4 43 47
43
Volume 17 NumbcrJ
583
RISK FACTORS IN HEPATITIS
continued importance of posttransfusion hepatitis as a significant complication of blood transfusion even when volunteer donors who have been tested for HBsAg are utilized. Since development of hepatitis seems to increase in direct proportion to the number of donor units transfused, risk-to-benefit ratios should be considered by health personnel for each unit of blood transfused. During the previous decade, the majority of posttransfusion hepatitis cases were caused by hepatitis type B Since currently only approximately 10 per cent of blood recipients who develop hepatitis can be shown immunologically to have hepatitis BX.I6..”’laboratory testing of donor blood for HBsAg has been relatively effective in diminishing transmission of hepatitis B by blood transfusion. These latter cases seem to be due to an inability of even the most sensitive methods of screening for HBsAg to detect certain carriers. I n the present study, development of hepatitis B was seen in three patients and the de novo appearance of either HBsAg o r anti-HBs without evidence of liver dysfunction o c c u r r e d i n 13 additional patients who received blood which was negative for HBsAg by radioirnmunoassay.” This inability to detect all carriers of hepatitis B virus by HBsAg testing has led investigators to study whether blood containing anti-HBs
Table 7. Number of Units of Anti-HBs Blood Transfused versus Occurrence of Non-A Non-B Hepatitis Units Transfused Containing Anti-HBs
0 1 2 3 4 >4
Number of Patients
Patients with Hepatitis, %
l o ( 8.1) 14 (19.4) 8 (22.9) 2 (14.3) 1 (33.3) 3 (42.8)
124 72 35 14 3 7
was responsible for transmitting hepatitis B,2.:1.10.18.17.IX.24 It has been shown in vitro t h a t an e x c e s s of a n t i - H b s in s e r u m specimens will complex HBsAg so that the antigen cannot be detected immunologically.’” Theoretically, the antigen-antibody complex could disassociate after transfusion to produce hepatitis type B. Interestingly, 10 of the 13 patients who received only HBsAg negative transfusions but still developed a de n o w serologic response without liver enzyme elevations indicating exposure to hepatitis B virus got transfusions containing anti-HBs. This incidence of B serologic response in recipients of anti-H Bs containing transfusions was significantly greater than that seen in the group which received only anti-HBs negative blood ( p = 0.04). This would seem to
Table 8. Association of Anti-HB Blood Transfusion with Occurrence of Non-A Non-B Hepatitis and the Mean Number o f Units Transfused Anti-HBs
No Anti-HBs
Total
Hepatitis Arith. Geom.
(28patients) 20.4i 12.5 (2.35) 17.5 k 1.7(1.11)
(10patients) 10.4 t 6.4 (2.03) 8.7 A 1.9(1.24)
(38patients) 17.8f 11.9 (1.94) 14.62 1.9(1.11)
No Hepatitis Arith. Geom.
(103patients) 12.5 f 7.6 ( .75) 10.5 f 1.8(1.06)
(1 14 patients) 9.8 * 5.2 ( .49) 8.5 f 1.9 (1.05)
(217 patients) 11.0* 6.6( .45) 9.4 * 1.8 (1.04)
Total Arith. Geom.
(131 patients) 14.2t 9.4 ( .82) 11.8 f 1.8 (1.05)
(124patients) 9.8 f 5.3( .47) 8.5 t 1.8 (1.05)
NOTE: The association of transfusion of blood containing anti-HBs with the nccurence o f non-A non-B hepatitis i s shown above and compared with patients who received blood without antibody. The mean number of units of blood transfused intn patients in each group with one standard deviation from the arithmetic and geometric mean are shown. The numbers in parentheses are the standard error of the mean. The geometric mean was calculated using conversion into common logarithms.
Transfusion
C O N R A D ET A L Table 9. Published Reports of the Effect of Transfusion of Anti-HBs Positive Blood upon the Occurrence of Posttransfusion Hepatitis Ratio of Hepatitis Patients t o Transfused Subjectst References
Anti- H 6s Transfused
No Anti-HBs
Alter et a1.3 Goldfield11 Gocke and Panicklo Aach et a/.* Koretz et al.18 Current Study
2/37 5/28 4/35 131169 15/45 281131
0114 14/98 2211 33 81189 6/18 101124
~~
tData cited exclude cases of type B hepatitis from numerator and denominator.
s u p p o r t c e r t a i n of t h e d a t a r e c e n t l y published by Koretz et ui.” in which the transfusion of large amounts of anti-HBs was associated with an increased incidence of serologic events related to hepatitis type B. However, there is no convincing evidence that the exclusion of donors with anti-HBs will reduce the occurrence of clinically detectable type B hepatitis in transfusion recipients. There is also no convincing d a t a to support a protective effect for anti-HBs transfusions since a similar incidence of hepatitis type B has been seen among recipients transfused with anti-HBs positive and negat ive blood. p.3. l o I n this article we have examined the relationship to anti-HBs positive transfusions and the subsequent development of non-A, non-B hepatitis. It was postulated t h a t donors with immunologic evidence of exposure to one hepatitis agent might be more likely to harbor other hepatitis agents for which there a r e presently no serologic tests. Unlike previous studies (Table 9), a significantly greater incidence of posttransfusion hepatitis was seen among patients who received blood containing anti-HBs than among patients transfused only with units containing no antibody (p=0.004). Much of the statistical difference observed could be attributed to the transfusion of more units of blood to the group that received one or more units containing anti-HBs. This was not surprising since only 8 per cent of the units utilized contained anti-HBs and the chances
July-Aug. 1977
of receiving one of these units increase in proportion to the number of units transfused. H o w e v e r , while t h e n u m b e r of u n i t s transfused was the more highly significant risk factor in the occurrence of posttransfusion hepatitis, stepwise regression analysis provides suggestive evidence that transfusion of anti-HBs containing blood also placed patients in this study at a higher risk of developing non-A non-B posttransfusion hepatitis (p = 0.06). With one exception, the previous studies attesting to the safety of transfusion o f blood containing anti-HBs have had relatively few patients at risk (Table 9). Whether the difference in our results from those noted in Aach’s study’ might be explained by differences in the transfusion requirements for the various patient groups cannot be determined from the text. The possibility that anti-HBs may act as a m a r k e r in donor populations to indicate blood carrying an increased risk of transmitting non-A, non-B hepatitis is intriguing. It suggests that environments exposing individuals to hepatitis B also contain non-B hepatitis agents and that anti-HBs testing of donor blood might provide an additional means of diminishing transmission of posttransfusion hepatitis. However, it is evident that even if this hypothesis was convincingly substantiated, exclusion of donors with antiHBs would merely reduce transmission and not eliminate the hazard. Additional information is needed with regard to the possible deleterious effects of anti-H Bs containing transfusions since this antibody is found in 5 to 10 per cent of volunteer blood donors and loss of these donor units would constitute a significant problem to the national blood program. Finally, any study which examines the effect of transfusing blood containing anti-HBs must carefully consider the number of units received by each recipient before the conclusions of t h e study can be considered credible. References I.
A Cooperative Study: Prophylactic gamma globulin for prevention of endemic hepatitis. EfTects of US. gamma globulin upon the incidence of
Volume 17 Number4
2.
3.
4.
5.
6. 7.
8.
9.
10.
11.
12.
13. 14.
15.
16.
RISK FACTORS IN HEPATITIS
viral hepatitis and other infectious diseases in U.S. soldiers abroad. A r c h . I n t e r n . Med. 128:723, 1971. Aach, R. D., H. J. Alter, F. B. Hollinger, P. V. Holland, J. J. Lander, J. L. Melnick, and J. M. Weiler: Risk of transfusing blood containing antibody to hepatitis B surface antigen. Lancet 2:190. 1974. Alter, H. J., P. V. Holland, R . H. Purcell, J. J. Lander, S. M. Feinstone, A. G. Morrow, and P. J. Schmidt: Posttransfusion hepatitis after exclusion of commercial and hepatitis-B antigenpositivedonors. Ann. Intern. Med. 77:691, 1972. Batson, H. C.: An Introduction to Statistics in the Medical Sciences. Minneapolis, Minn., Burgess Publishing Company, 1956, p. 48. Barker, L. F., M. R. Peterson, N. R . Shulman, and R . Murray: Antibody responses in viral hepatitis, type B. JAMA 223:1005, 1973. Edwards, A. L.: Statistical Methods. New York, Holt, Reinhart and Winston, 1967, p. 200. Feinstone, S. M., .A. 2. Kapikian, and R. H. Purcell: Hepatitis A: Detection by immune electron microscopy of a viruslike antigen associated with acute illness. Science 182:1026, 1973. -,A. Z. Kapikian, R . H. Purcell, H. J. Alter, , and P. V. Holland: Transfusion-associated hepatitis not due to viral hepatitis type A or B. N. Engl. J. Med. 292:767, 1975. Ginsberg, A. L., M. E. Conrad, W. H. Bancroft, C. M. Ling, and L. R. Overby: Prevention of endemic HAA-positive hepatitis with gamma globulin: Use of a simple radioimmune assay to detect HAA. N. Engl. J . Med. 286562, 1972. Gocke, D. J., and J. M. Panick: Transfusion of blood containing anti-HBAg. In: Hepatitis and Blood Transfusion. G. N. Vyas, H. A. Perkins, and R. Schmid. Eds. New York, Grune and Stratton, 1972, p. 319. Goldfield, M.: Some epidemiological studies of transfusion-associated hepatitis. In: Transmissible Disease and Blood Transfusion. T. J. Greenwalt, and G. A. Jamieson. Eds. New York, Grune and Stratton, 1974, p. 141. Henle, G., and W. Henle: lmmunofluorescence in cells derived from Burkitt's lymphoma. J. Bacteriol. 91:1248, 1966. International Forum: Does blood with antibodies against hepatitis B carry an increased risk to transmit hepatitis? Vox Sang. 28:460, 1975. Irwin, G. R., A. M. Alien, W. H. Bancroft, M. Willhight, and P. K. Russell: Specificity and sensitivity of radioimmunoassay for hepatitis B antigen. Appl. Microbiol. 28:600, 1974. Kent, J. F., and E. H. Fife, Jr.: Precise standardization of reagents for complement fixation. Am. J.Trop. Med. 12:103, 1963. Knodell, R. G., M. E. Conrad, J . L. Deinstag, and C. J. Bell: Etiological spectrum of post-transfusion hepatitis. Gastroenterology 69:1278, 1975.
585
17. -,
18.
19.
20.
21.
22.
23.
24.
25.
26.
M. E. Conrad, A. L. Ginberg, C. J. Bell, and E. P. Flannery: Efficacy of prophylactic gamma-globulin i n preventing non-A non-B posttransfusion hepatitis. Lancet 1557. 1976. Koretz, R. L., L. R. Overby, and G. L. Gitnick: Post-transfusion hepatitis: The role of hepatitis B antibody. Gastroenterology 70556, 1976. Krugman, S., and J. P. Giles: Viral hepatitis type B (MS-2-strain): Further observations on natural history and prevention. N. Engl. J. Med. 288:755, 1973. Ling, C. M., and L. R. Overby: Prevalence of hepatitis B virus antigen as revealed by direct radioimmunoassay with 'lSl antibody. J. Irnmunol. 109:834, 1972. Mollison, P. L.: Blood Transfusion in Clinical Medicine, 5th ed. London, Blackwell Scientific, 1972, p. 609. Prince, A. M., B. Brotman, G. F. Grady, W. J. Kuhns, C . Hazzi, R. W. Levine, and S. J. Millian: Long incubation post-transfusion hepatitis without serological evidence of exposure to hepatitis B virus. Lancet 2:241, 1974. Prince, A. M., R. L. Hargrove, W. Szmuness, C. E. Cherubin, V. J. Fontana, and G. H. JetTries: Immunologic distinction between infectious and serum hepatitis. N. Engl. J. Med. 282:987, 1970. SeetT, L. B., H. J. Zimmerman, and E. C. Wright: V. A. Cooperative study of gamma globulin prophylaxis o f post-transfusion hepatitis. Gastroenterology 643393, 1973. Snedecor, G . W. and W. G. Cochran: Statistical Methods, 6th ed. Ames, Iowa, Iowa S t a t e University Press. 1967, p. 299. Surgenor, D. Mac N., T. C. Chalmers, M. E. Conrad, W. T. Friederwald, G . F. Grady, M . Hamilton, J. W. Mosely, .A. M. Prince, and J. M. Stengle: Clinical trials-of hepatitis B immune globulin. Development of policies and materials for the 1972-1975 studies sponsored by the National Heart and Lung Institute. N. Engl. J. Med. 293:1060, 1975.
M. E. Conrad, M.D., Division of Hematology and Oncology, University of Alabama, Birmingham, Alabama. R. G. Knodell, M.D., Division of Gastroenterology, University of New Mexico School of Medicine, Albuquerque, New Mexico. E. L. Bradley, Jr., Ph.D., Department of Biostatistics, University of Alabama, Birmingham, Alabama. E. P. Flannery, M.D., Hematology and Oncology Service, L e t t e r m a n Army Medical C e n t e r , S a n Francisco, California. A. L. Ginsberg, M.D., Division of Gastroenterology, George Washington University Medical Center, Washington, D.C.
