Antibody t o the Hepatitis B Surface Antigen in Immune Serum Globulin J. H. HOOFNAGLE, R. J. GERETY, AND L. F. BARKER From the Division of Blood and Blood Products. Bureau of Biologics, Food and Drug Administration, Bethesda, Maryland
A collection of 1,278 lots of immune serum globulin (ISG) prepared by 19 United States manufacturers between 1962 and 1974 were tested for the hepatitis B surface antigen ( H h A g ) and antibody (anti-HB,). Ten lots (0.8%), all of which were produced between 1962 and 1965 by two different manufacturers, were weakly positive for HB, Ag (by radioimmunoassay). Seven hundred and seven lots (55.3%)were positive for anti-HBS (by passive hemagglutination). In general, titers of antiHBs in lots of ISG were low, and the prevalence of antiHB, positive lots varied considerably among different manufacturers. ISG prepared from placental material was more commonly positive for anti-HB, than was ISG prepared from plasma. There was a striking overall increase in prevalence and titer of anti-HB, in ISG lots prepared during 1973 and 1974. This probably reflects the effect of elimination of strongly HhAg-positive plasma units with the onset of routine screening for HB,Ag which began in 1972.
WHILEthe effectiveness of immune serum globulin (ISG, Cohn fraction 11) in the protection against viral hepatitis, type A, has been well documented,12.21the efficacy of ISG in the prophylaxis of viral hepatitis, type B, remains in doubt.".18 In several studies, ISG has been shown to have little or no effect on the incidence and severity of disease after parenteral exposure to the hepatitis B v ~ ~ u s . ~ Recently, * " * ~ * however, ~~*~~ two studies of hepatitis occurring in endemic situations have shown a clear effect of ISG in reducing both the incidence of clinical hepatitis and the development of the chronic carrier state of the hepatitis B surface antigen (HB,Ag, Australia antigen).2.22Sensitive immunological assays for antibody to HB,Ag .(anti-HB,) have revealed that many lots of commercial ISG possess low titers of antiHB,.l8 These low levels of antibody to the surface antigen of the hepatitis B virus in
ISG have been cited as the cause of its poor efficacy in the prevention of type B hepatitis. ISG is prepared from pooled, human plasma as fraction I1 of the Cohn, coldethanol procedure. Unlike every other unpasteurized fraction from the Cohn procedure, ISG does not appear to transmit viral hepatitis, type B, even when prepared from plasma of proven infectivity.'" Testing of plasma fractions for HB,Ag has demonstrated that ISG made from strongly HB,Ag-positive plasma does not contain detectable HB,Ag.7*20This lack of HB,Ag reactivity of ISG correlates well with its documented freedom from infectivity. The availability of large numbers of lots of ISG made this study possible. The study was undertaken to assess the levels of anti-HB, and HB,Ag in commercially prepared lots of ISG manufactured in the United States over the past 12 years.
Received for publication December 9, 1974; accepted January 4. 1975.
Materials and Methods
Prior to release, a sample of every lot of ISG manufactured in the United States is submitted to the Bureau of Biologics of the Food and Drug Administration for potency, purity, and safety testing. Residual samples of these lots have been retained at 4C and were used for this study. Because of the large number of samples involved, a representative collection was chosen consisting of 1,278 lots of immune globulins submitted by 19 manufacturers from 1962 to October 1974. This represented approximately 20 percent of the total number of ISG lots submitted to the Bureau of Biologics during this time span. Materials from some manufacturers were underrepresented because of the lack of residual, stored samples. Data on the available lots consisted of type of immune globulin (immune serum globulin, measles, mumps, pertussis, poliomyelitis, tetanus, and
408 Transfusion sepP(.-oct.I975
Volume 15 Number 5
Volume IS Number S
409
ANTI-HB, IN ISG
vaccinia immune gloublin), manufacturer, lot number, source (whether placenta o r plasma derived), and date of submission to the Bureau of Biologics. An additional 12 I lots of ISG prepared by an ammonium sulfate precipitation method by two different manufacturers were tested but were analyzed separately. Samples were tested for HB,Ag by a solidphase radioimmunoassay (RIA)" using the room temperature procedure recommended by the manufacturer.* Samples were tested in duplicate. Those samples yielding counts per minute (cpm) in excess of 2. I times the negative control mean were considered to be reactive. All reactive samples were subject to specificity testing with human anti-HB, to inhibit the reaction." Only those samples whose reaction cpm were reduced by a t least 50 per cent by human anti-HB, and not by normal human serum were considered to be true, specific positives. Positive specimens were also tested by counterele~trophoresis~and reversed passive hemaggl~tination.**~ A selected number of positive samples were further tested by a more recently licensed, solid-phase radioimmunoassay*** and by a double antibody radioimmunoassay (kindly performed by Dr. Blaine Hollinger).O Samples were assayed for anti-HB, by passive hemagglutinationas using human, type 0 erythrocytes coated with the adw subtype of HB.Ag.t ISG lots were tested at an initial titer of 1.2 and were tested on two separate occasions. For analysis of anti-HB, prevalence, those lots giving reproducible titers of 1.8 or greater in the absence of agglutination of uncoated, control erythrocytes were considered positive for anti-HB,. Serial
*Ausria-I: Abbott Laboratories. **Auscell: Abbott Laboratories. ***Ausria-11: Abbott Laboratories. +Virgo Reagents: Electro-Nucleonics.
twofold dilutions in T A P buffer were made to determine anti-HB, end-point titers. Control sera with established titers were included weekly to insure the uniformity of passive hemagglutination reagents.
Results Among 1,278 lots of ISG prepared by theCohn fractionation method by 19 manufacturers between 1962 and 1974, only 10 (0.8%) were reproducibly positive for HB,Ag on R I A and could be confirmed as specific. All ten were weakly positive, yielding cpm from 2.1 to 6.0 times the negative control mean on the Ausria-1 RIA system. All ten were also positive by Ausria-I1 and by double antibody RIA. None of the ten were reactive by the less sensitive method of counterelectrophoresis, but five were reactive a t a titer of 1% o r greater on reversed passive hemagglutination (and inhibited to < 1:2 by human antiHB,). These ten HB,Ag-positive lots of ISG were manufactured between 1962 and 1965; nine by manufacturer B and one by manufacturer C. None of these ten possessed detectable anti-HB, by passive hemagglutination. Many other samples gave borderline results on RIA, but did not fit the criteria of being reproducibly positive and inhibited by human anti-HB,. Sixty-four samples (5 W ) gave reproducible, false positive reactions on R I A (not inhibited by human anti-HBs). No sample submitted during 1973 or 1974 was positive on RIA or gave false positive reactions. Among 1,278 lots of ISG prepared by the Cohn fractionation method between 1962 and 1974,707 (55.3%) were positive for anti-HB, a t a titer of 1:8 o r greater on passive hemagglutination. In Table 1, the prevalence of anti-HB, positive lots of ISG is tabulated by manufacturer and by year of submission. Considerable variation existed in the prevalence of anti-HB, positive lots between different manufacturers. Thus, among the six
Table 1. Prevalence of Anti-HE, Positive' Lots of Immune Globulins by Manufacturer and Year of Submission to the Bureau of Eioloaics ~
Manufacturers
A B C D E F Others Total
1964-1965 (%)
1966-1 967 ( %)
1968-1 969
1970-1 97 1
1972
(%I
(96)
(%I
1973-1 974 (%)
35 (9/26) 9 (3/35) 97 (35/36) 90 (26/29) 23 (9/40)
60 (24/40) 0 (0/42) 100 (40/40) 94 (17/18) 28 ( 10136)
24 W34) 0 (0140) 99 (35/40) 53 ( 17/32) 41 (16/39)
22 (2/9) 6 (2/35) 92 (36/39) 50 ( 14/28) 3 (1/29)
81 (47/58) 58 (129/224)
95 (19/20) 56 ( 1 10/196)
50 (112) 81 (17/21) 85 (1 7/20) 85 (1 7/20) 95 (19/20) 0 (0/11) 30 (3/10) 71 (74/104)
89 (34/38) 100 (19/19) 100 (27/27) 100 (18/18) 100 (34/34) 87 (34/39) 74 (26/35) 91 (192/210)
1962-1963
(%I
13 (2/16) 13 (5/40) 63 (20/32) 72 (21/29) 20 (6/30) 11 (3/27) 56 (14/25) 0 (0/1) 0 (0/19) 41 (76/186) 35 (55/159) 35 (71/199)
'A titer of 2 1:8 on passive hemagglutination was considered positive.
410
Transfusion
HOOFNAGLE, ET AL.
manufacturers of ISG from whom adequate numbers of ISG lots were available, the prevalence of anti-HB, positive lots varied at least fivefold during any one year up to 1973. For the entire 13 years, the prevalence of anti-HB, positive lots varied between manufacturers from less than 20 per cent (46/232: manufacturer B) to almost 90 per cent (210/234: manufacturer C). Some of the variation in prevalence of anti-HB, positive lots of ISG between manufacturers could be accounted for by source of the material. Manufacturers C and D,who demonstrated the highest prevalences of anti-HB, reactive lots, both used placental material as the major source of ISG. Among 339 lots of ISG submitted prior to 1972 which were derived totally or partially from placental material, 300 (88%) were anti-HB, positive, whereas among 625 lots derived from plasma only, 141 (22.6%)were anti-HB, positive. Also apparent in Table 1 is a general rise in the prevalence of anti-HB, positive lots in the years 1972 to 1974. Among 964 lots submitted prior to 1972, 441 (46%) were anti-HB, positive, whereas among 314 lots submitted from 1972 to 1974,266 (85%) were anti-HB, positive. For the years before 1972, the prevalence of anti-HB, positive lots ranged from 30 to 61 per cent, whereas during 1973 and 1974 the prevalence of positive lots was 97 and 89 per cent, respectively. The year 1972 appeared to be a transition period with 71 percent of lots tested being anti-HB, positive. This rise in prevalence of anti-HB, was seen among ISG lots of all manufacturers studied, but was most obvious in those manufacturers with low prevalences prior to 1972. All manufacturers (A, B, E, and F) not using placental material demonstrated low levels of anti-HB, prior to 1972 (total ranging from 5 to 36%) and high prevalences subsequent to 1972 (totals ranging from 87 to 100%). Titers of anti-HB, in lots of commercially pre-
Sew-Ocl. 1975
pared ISG also reflected a rise in anti-HB, during 1972 to 1974. In Table 2, the mean and range of titers of ISG lots is shown for the six major manufacturers according to two-year periods. This data was generated only on lots of standard immune serum globulin; specific immune globulins (measles, mumps, pertussis, poliomyelitis, tetanus, and vaccinia immune globulins) were excluded. Clearly, not only the prevalence of antiHB, positive lots but also the titer of anti-HBs in plasma-derived ISG lots rose during the period of 1972 to 1974. As was the case in the analysis of prevalence, the year 1972 appeared to be a transition period and was, therefore, analyzed alone in the presentation of results. Manufacturer B demonstrated the most dramatic rise in both prevalence and titer of anti-HB in lots of ISG. In the ten years prior to 1972, only ten of 172 lots of ISG (6%)were positive for antiHB, a t a titer of 1:8 or greater, and the mean titer of ISG lots was 1:2. During 1973 and 1974, 19 of 19 (100%)lots were positive for anti-HB, with an overall mean titer of 1512. Among 121 lots of ISG prepared by ammonium sulfate precipitation, 40 (33%) were found positive for HB,Ag by RIA. All 40 could be demonstrated to be specific positives. Immune electron microscopy on one lot revealed 20 nanometer, spherical, and tubular particles typical of HB,Ag positive serum. No lots prepared by this method had detectable anti-H&. No ISG lots produced by ammonium sulfate precipitate have been released in the United States since 1969.
Discussion Studies on the distribution of HB,Ag reactivity during Cohn, cold ethanol fractionation have shown that most HB,Ag is found in fractions 111 (thrombin) and IV
Table 2. Mean and Range' of Anti-HB, Titers in ISG Lots Submitted b y Six Manufacturers from Manufacturer
1961 to 1974
Year of Submission
1962-1963 1:4 (
A collection of 1,278 lots of immune serum globulin (ISG) prepared by 19 United States manufacturers between 1962 and 1974 were tested for the hepatitis B surface antigen ( H h A g ) and antibody (anti-HB,). Ten lots (0.8%), all of which were produced between 1962 and 1965 by two different manufacturers, were weakly positive for HB, Ag (by radioimmunoassay). Seven hundred and seven lots (55.3%)were positive for anti-HBS (by passive hemagglutination). In general, titers of antiHBs in lots of ISG were low, and the prevalence of antiHB, positive lots varied considerably among different manufacturers. ISG prepared from placental material was more commonly positive for anti-HB, than was ISG prepared from plasma. There was a striking overall increase in prevalence and titer of anti-HB, in ISG lots prepared during 1973 and 1974. This probably reflects the effect of elimination of strongly HhAg-positive plasma units with the onset of routine screening for HB,Ag which began in 1972.
WHILEthe effectiveness of immune serum globulin (ISG, Cohn fraction 11) in the protection against viral hepatitis, type A, has been well documented,12.21the efficacy of ISG in the prophylaxis of viral hepatitis, type B, remains in doubt.".18 In several studies, ISG has been shown to have little or no effect on the incidence and severity of disease after parenteral exposure to the hepatitis B v ~ ~ u s . ~ Recently, * " * ~ * however, ~~*~~ two studies of hepatitis occurring in endemic situations have shown a clear effect of ISG in reducing both the incidence of clinical hepatitis and the development of the chronic carrier state of the hepatitis B surface antigen (HB,Ag, Australia antigen).2.22Sensitive immunological assays for antibody to HB,Ag .(anti-HB,) have revealed that many lots of commercial ISG possess low titers of antiHB,.l8 These low levels of antibody to the surface antigen of the hepatitis B virus in
ISG have been cited as the cause of its poor efficacy in the prevention of type B hepatitis. ISG is prepared from pooled, human plasma as fraction I1 of the Cohn, coldethanol procedure. Unlike every other unpasteurized fraction from the Cohn procedure, ISG does not appear to transmit viral hepatitis, type B, even when prepared from plasma of proven infectivity.'" Testing of plasma fractions for HB,Ag has demonstrated that ISG made from strongly HB,Ag-positive plasma does not contain detectable HB,Ag.7*20This lack of HB,Ag reactivity of ISG correlates well with its documented freedom from infectivity. The availability of large numbers of lots of ISG made this study possible. The study was undertaken to assess the levels of anti-HB, and HB,Ag in commercially prepared lots of ISG manufactured in the United States over the past 12 years.
Received for publication December 9, 1974; accepted January 4. 1975.
Materials and Methods
Prior to release, a sample of every lot of ISG manufactured in the United States is submitted to the Bureau of Biologics of the Food and Drug Administration for potency, purity, and safety testing. Residual samples of these lots have been retained at 4C and were used for this study. Because of the large number of samples involved, a representative collection was chosen consisting of 1,278 lots of immune globulins submitted by 19 manufacturers from 1962 to October 1974. This represented approximately 20 percent of the total number of ISG lots submitted to the Bureau of Biologics during this time span. Materials from some manufacturers were underrepresented because of the lack of residual, stored samples. Data on the available lots consisted of type of immune globulin (immune serum globulin, measles, mumps, pertussis, poliomyelitis, tetanus, and
408 Transfusion sepP(.-oct.I975
Volume 15 Number 5
Volume IS Number S
409
ANTI-HB, IN ISG
vaccinia immune gloublin), manufacturer, lot number, source (whether placenta o r plasma derived), and date of submission to the Bureau of Biologics. An additional 12 I lots of ISG prepared by an ammonium sulfate precipitation method by two different manufacturers were tested but were analyzed separately. Samples were tested for HB,Ag by a solidphase radioimmunoassay (RIA)" using the room temperature procedure recommended by the manufacturer.* Samples were tested in duplicate. Those samples yielding counts per minute (cpm) in excess of 2. I times the negative control mean were considered to be reactive. All reactive samples were subject to specificity testing with human anti-HB, to inhibit the reaction." Only those samples whose reaction cpm were reduced by a t least 50 per cent by human anti-HB, and not by normal human serum were considered to be true, specific positives. Positive specimens were also tested by counterele~trophoresis~and reversed passive hemaggl~tination.**~ A selected number of positive samples were further tested by a more recently licensed, solid-phase radioimmunoassay*** and by a double antibody radioimmunoassay (kindly performed by Dr. Blaine Hollinger).O Samples were assayed for anti-HB, by passive hemagglutinationas using human, type 0 erythrocytes coated with the adw subtype of HB.Ag.t ISG lots were tested at an initial titer of 1.2 and were tested on two separate occasions. For analysis of anti-HB, prevalence, those lots giving reproducible titers of 1.8 or greater in the absence of agglutination of uncoated, control erythrocytes were considered positive for anti-HB,. Serial
*Ausria-I: Abbott Laboratories. **Auscell: Abbott Laboratories. ***Ausria-11: Abbott Laboratories. +Virgo Reagents: Electro-Nucleonics.
twofold dilutions in T A P buffer were made to determine anti-HB, end-point titers. Control sera with established titers were included weekly to insure the uniformity of passive hemagglutination reagents.
Results Among 1,278 lots of ISG prepared by theCohn fractionation method by 19 manufacturers between 1962 and 1974, only 10 (0.8%) were reproducibly positive for HB,Ag on R I A and could be confirmed as specific. All ten were weakly positive, yielding cpm from 2.1 to 6.0 times the negative control mean on the Ausria-1 RIA system. All ten were also positive by Ausria-I1 and by double antibody RIA. None of the ten were reactive by the less sensitive method of counterelectrophoresis, but five were reactive a t a titer of 1% o r greater on reversed passive hemagglutination (and inhibited to < 1:2 by human antiHB,). These ten HB,Ag-positive lots of ISG were manufactured between 1962 and 1965; nine by manufacturer B and one by manufacturer C. None of these ten possessed detectable anti-HB, by passive hemagglutination. Many other samples gave borderline results on RIA, but did not fit the criteria of being reproducibly positive and inhibited by human anti-HB,. Sixty-four samples (5 W ) gave reproducible, false positive reactions on R I A (not inhibited by human anti-HBs). No sample submitted during 1973 or 1974 was positive on RIA or gave false positive reactions. Among 1,278 lots of ISG prepared by the Cohn fractionation method between 1962 and 1974,707 (55.3%) were positive for anti-HB, a t a titer of 1:8 o r greater on passive hemagglutination. In Table 1, the prevalence of anti-HB, positive lots of ISG is tabulated by manufacturer and by year of submission. Considerable variation existed in the prevalence of anti-HB, positive lots between different manufacturers. Thus, among the six
Table 1. Prevalence of Anti-HE, Positive' Lots of Immune Globulins by Manufacturer and Year of Submission to the Bureau of Eioloaics ~
Manufacturers
A B C D E F Others Total
1964-1965 (%)
1966-1 967 ( %)
1968-1 969
1970-1 97 1
1972
(%I
(96)
(%I
1973-1 974 (%)
35 (9/26) 9 (3/35) 97 (35/36) 90 (26/29) 23 (9/40)
60 (24/40) 0 (0/42) 100 (40/40) 94 (17/18) 28 ( 10136)
24 W34) 0 (0140) 99 (35/40) 53 ( 17/32) 41 (16/39)
22 (2/9) 6 (2/35) 92 (36/39) 50 ( 14/28) 3 (1/29)
81 (47/58) 58 (129/224)
95 (19/20) 56 ( 1 10/196)
50 (112) 81 (17/21) 85 (1 7/20) 85 (1 7/20) 95 (19/20) 0 (0/11) 30 (3/10) 71 (74/104)
89 (34/38) 100 (19/19) 100 (27/27) 100 (18/18) 100 (34/34) 87 (34/39) 74 (26/35) 91 (192/210)
1962-1963
(%I
13 (2/16) 13 (5/40) 63 (20/32) 72 (21/29) 20 (6/30) 11 (3/27) 56 (14/25) 0 (0/1) 0 (0/19) 41 (76/186) 35 (55/159) 35 (71/199)
'A titer of 2 1:8 on passive hemagglutination was considered positive.
410
Transfusion
HOOFNAGLE, ET AL.
manufacturers of ISG from whom adequate numbers of ISG lots were available, the prevalence of anti-HB, positive lots varied at least fivefold during any one year up to 1973. For the entire 13 years, the prevalence of anti-HB, positive lots varied between manufacturers from less than 20 per cent (46/232: manufacturer B) to almost 90 per cent (210/234: manufacturer C). Some of the variation in prevalence of anti-HB, positive lots of ISG between manufacturers could be accounted for by source of the material. Manufacturers C and D,who demonstrated the highest prevalences of anti-HB, reactive lots, both used placental material as the major source of ISG. Among 339 lots of ISG submitted prior to 1972 which were derived totally or partially from placental material, 300 (88%) were anti-HB, positive, whereas among 625 lots derived from plasma only, 141 (22.6%)were anti-HB, positive. Also apparent in Table 1 is a general rise in the prevalence of anti-HB, positive lots in the years 1972 to 1974. Among 964 lots submitted prior to 1972, 441 (46%) were anti-HB, positive, whereas among 314 lots submitted from 1972 to 1974,266 (85%) were anti-HB, positive. For the years before 1972, the prevalence of anti-HB, positive lots ranged from 30 to 61 per cent, whereas during 1973 and 1974 the prevalence of positive lots was 97 and 89 per cent, respectively. The year 1972 appeared to be a transition period with 71 percent of lots tested being anti-HB, positive. This rise in prevalence of anti-HB, was seen among ISG lots of all manufacturers studied, but was most obvious in those manufacturers with low prevalences prior to 1972. All manufacturers (A, B, E, and F) not using placental material demonstrated low levels of anti-HB, prior to 1972 (total ranging from 5 to 36%) and high prevalences subsequent to 1972 (totals ranging from 87 to 100%). Titers of anti-HB, in lots of commercially pre-
Sew-Ocl. 1975
pared ISG also reflected a rise in anti-HB, during 1972 to 1974. In Table 2, the mean and range of titers of ISG lots is shown for the six major manufacturers according to two-year periods. This data was generated only on lots of standard immune serum globulin; specific immune globulins (measles, mumps, pertussis, poliomyelitis, tetanus, and vaccinia immune globulins) were excluded. Clearly, not only the prevalence of antiHB, positive lots but also the titer of anti-HBs in plasma-derived ISG lots rose during the period of 1972 to 1974. As was the case in the analysis of prevalence, the year 1972 appeared to be a transition period and was, therefore, analyzed alone in the presentation of results. Manufacturer B demonstrated the most dramatic rise in both prevalence and titer of anti-HB in lots of ISG. In the ten years prior to 1972, only ten of 172 lots of ISG (6%)were positive for antiHB, a t a titer of 1:8 or greater, and the mean titer of ISG lots was 1:2. During 1973 and 1974, 19 of 19 (100%)lots were positive for anti-HB, with an overall mean titer of 1512. Among 121 lots of ISG prepared by ammonium sulfate precipitation, 40 (33%) were found positive for HB,Ag by RIA. All 40 could be demonstrated to be specific positives. Immune electron microscopy on one lot revealed 20 nanometer, spherical, and tubular particles typical of HB,Ag positive serum. No lots prepared by this method had detectable anti-H&. No ISG lots produced by ammonium sulfate precipitate have been released in the United States since 1969.
Discussion Studies on the distribution of HB,Ag reactivity during Cohn, cold ethanol fractionation have shown that most HB,Ag is found in fractions 111 (thrombin) and IV
Table 2. Mean and Range' of Anti-HB, Titers in ISG Lots Submitted b y Six Manufacturers from Manufacturer
1961 to 1974
Year of Submission
1962-1963 1:4 (
