ARTICLES

Infection with Human Immunodeficiency Virus Type 1 (HIV-1) among Recipients of Antibody-Positive Blood Donations Elizabeth Donegan, MD; Maria Stuart, MPH; Joyce C. Niland, PhD; Henry S. Sacks, MD, PhD; Stanley P. Azen, PhD; Shelby L. Dietrich, MD; Cheryl Faucett, MS; Mary Ann Fletcher, PhD; Steven H. Kleinman, MD; Eva A. Operskalski, PhD; Herbert A. Perkins, MD; Johanna Pindyck, MD; Eugene R. Schiff, MD; Daniel P. Stites, MD; Peter A. Tomasulo, MD; James W. Mosley, MD; and the Transfusion Safety Group* Objective: To assess the incidence of human immunodeficiency virus type 1 (HIV-1) transmission by antibody (anti-HIV-l)-positive blood components, and to determine the immunologic and clinical course in HIV-1-infected recipients. Design and Subjects: We retrospectively tested approximately 200 000 donor blood component specimens stored in late 1984 and 1985 for anti-HIV-1, and we contacted recipients of positive specimens to determine their serologic status. They were compared with both recipients of HIV-1-negative transfusions and healthy (untransfused) controls. Subjects were seen at 3- to 6-month intervals for up to 4 years for clinical and immunologic evaluations. Measurements and Main Results: Of 133 recipients, 9 had other possible exposures. Excluding these cases, 111 of 124 (89.5%) were anti-HIV-1-positive (95% CI, 84.1% to 94.5%). The recipient's sex, age, underlying condition, and type of component did not influence infection rates. The cumulative risk for developing the acquired immunodeficiency syndrome (AIDS) within 38 months after transfusion was 13% (CI, 7.5% to 21.6%). At 36 ± 3 months after the index transfusion, seropositive recipients had lower counts of CD2+CDw26+, CD4+, CD4+CD29+, and CD4+CD45RA+ subsets and more CD8+I2+ lymphocytes than did recipients of anti-HIV-1-negative transfusions. The CD4+ and CD2+CDw26+ subsets changed the most rapidly. The absolute CD8+ count remained normal. Conclusions: Transfusion of anti-HIV-1-positive blood infected 90% of recipients. The rate of progression to AIDS within the first 38 months after infection was similar to that reported for homosexual men and hemophiliacs. Although most lymphocyte subset counts changed over time, CD8+ counts were constant.

Annals of Internal Medicine. 1990;113:733-739. From the University of California, San Francisco, California; University of Southern California, Los Angeles, California; Mount Sinai Medical Center, New York, New York; Huntington Memorial Hospital, Pasadena, California; the University of Miami, Miami, Florida; the American Red Cross Blood Services, Los Angeles-Orange Counties Region, Los Angeles, California, and South Florida Region, Miami, Florida; Irwin Memorial Blood Centers, San Francisco, California; and the New York Blood Center, New York, New York. For current author addresses, see end of text. *Members of the Transfusion Safety Study Group are listed in the Acknowledgments section at the end of text.

Oince transfusion-associated development of the acquired immunodeficiency syndrome (AIDS) was first recognized, the number of reported cases has increased steadily (1-3). Identification of transfusion-related asymptomatic human immunodeficiency virus type 1 (HIV-1) infection has depended almost entirely on lookback programs (4-6) that could not determine the donor's HIV-1 status at the time of unscreened donations. As a result, factors possibly influencing the incidence of transmission by HIV-1-contaminated transfusions have remained poorly defined. In addition, the course of HIV-1 infection has not been examined in a large group of prospectively followed blood recipients infected by cross-sectionally identified donors. Retrospective post-licensure testing of donations made before donor screening was routine (7) permitted us to observe 133 persons with known exposure to HIV-1-contaminated blood. We recruited 64 randomly selected recipients of anti-HIV-1-negative donations to control for the hematologic and immunologic effects of serious, long-term illness. Both groups were compared with healthy (untransfused) control subjects. In this report, we summarize our data about HIV-1 infection rates among recipients of anti-HIV-1-positive donations and examine recipient characteristics considered to be potentially important in determining HIV-1 infection rates among persons exposed through transfusion. A sufficient number of recipients has been followed for 4 years to permit survival analysis for that period; follow-up will continue for several years. In addition, data on hematologic indices and immunophenotypically defined lymphocyte subpopulations are adequate for statistical comparisons of these values in both seropositive and control recipients with those in healthy control subjects at the last visit within the interval of 36 ± 3 months after transfusion of the index unit. Finally, we looked at the rate with which laboratory values became abnormal among HIV-1-infected recipients compared with the rate among control recipients in the interval from 24 ± 3 to 36 ± 3 months after transfusion. Methods Identification and Recruitment of Recipients From September 1984 through early February 1985, we asked donors in New York, southern Florida, San Francisco, and Los Angeles to consent to anti-HIV-1 testing when a method became available. We collected approximately 200 000 repository specimens before the first enzyme-linked immunoassays (ELISAs) were licensed. © 1990 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19716/ by a University of California San Diego User on 06/05/2017

733

Table 1. Lymphocyte Safety Study

Subsets

Used in Evaluating

the Immune

Marker

in the

Transfusion

Function

CD2+ CD2+CDw26+ CD2+CDllb+ CD4+ CD4+CD29+ CD4+CD45RA+ CD8+ CD8+I2+ CD56+ CD20+ CD21 +

Pan T-cell marker, having sheep erythrocyte receptors (14) Activated T cells; CDw26 marks an alternative pathway of T-cell activation (15) Subset of cytolytic T cells (16) Total helper-inducer T cells (17) CD4+ cells that help B cells respond to antigen with immunoglobulin synthesis and interleukin-2 production (18) CD4+ cells that induce the activation of CD8 suppressor and cytotoxic cells (19) Suppressor-cytotoxic T cells (20) Suppressor T cells bearing HLA/DR antigen expressing class II antigen and suggestive of activation (21) Large granular lymphocytes with a natural killer cell potential (22, 23) Total peripheral B lymphocytes (24) Peripheral B cells (25)

After HIV-1 screening systems became available, we developed a protocol for uniform repository screening at each blood service and tested 196 681 serum specimens by ELISA (Electro-Nucleonics Inc., Columbia, Maryland). All repeatedly reactive serum specimens and a sample of nonreactive serum specimens were evaluated by Western immunoblot and radioimmunoprecipitation assay. Our summary is limited to data for recipients of donations for which all three tests were in agreement. The four blood services listed the hospitals to which antiHIV-1-positive components were distributed. The hospital blood banks then identified the recipients of these components and their physicians. If a recipient's physician gave permission, the recipient was asked to enroll in the study. Control recipients were randomly selected by tracing a subset of components from donations for which repository specimens were negative. Healthy control subjects were anti-HIV-1 -negative donors or were anti-HIV-1-negative sexual and household contacts of patients with congenital hematologic disorders seen as another part of the study. The sample size for healthy control subjects among the various tests ranged from 473 to 513 because of missing values for some indices. Observations Transfused subjects were seen at intervals of 3 to 6 months. Healthy control subjects were usually seen at 6-month intervals. At each visit, all subjects were interviewed and charts reviewed using standardized forms. Data were collected on medical and social history (including other possible exposures to HIV-1). The diagnosis of AIDS was made by the physician or clinic responsible for medical supervision, but study personnel reviewed the information to confirm that the case met the criteria of the Centers for Disease Control (8-10). The Coordinating Center Laboratory tested specimens from enrolled subjects for anti-HIV-1 by ELISA (Abbott Laboratories, Abbott Park, Illinois, or Genetic Systems Corporation, Seattle, Washington). Repeatedly reactive specimens were evaluated by Western immunoblot in one of three laboratories (11). Each clinical center carried out complete blood counts and did flow cytometric analyses of peripheral blood mononuclear cell subpopulations in whole blood by simultaneous two-color flow cytometry (Coulter EPICS-C, Hialeah, Florida) (12). All immunology laboratories used the same lots of monoclonal antibodies (Coulter Diagnostics, Hialeah, Florida) and standardized protocols. A quality control program monitored interand intralaboratory comparability (13). Table 1 lists the immunophenotypic markers covered in this report and indicates known or possible functions associated with their expression (14-25). Statistical Analyses For analysis of the course of HIV-1 infection, we omitted 13 anti-HIV-1-negative recipients of components from seropositive donations. Currently available information, including some 734

Status of Blood Recipients

from special studies, suggests that none of these 13 subjects was infected, but that possibility cannot be eliminated. Our analysis of HIV-1 progression also excluded recipients with potential HIV-1 exposures other than the index transfusion. In calculating component-specific rates, second exposures to antiHIV-1-positive components of the same type (2 subjects) have been ignored. In calculating confidence intervals (CIs) for the incidence of HIV-1 transmission, we used a logit transformation of the binomial distribution. For seropositive recipients, the rate of progression to AIDS was defined from the time of the index transfusion. The probability of having progressed to AIDS was calculated using the Kaplan-Meier product limit estimate of survival. Subjects without AIDS who did not have a visit after 30 June 1988 were censored from the analysis as of the date of their last visit before then. If their status through 30 June was known because of a later visit, they were censored as of 30 June to limit the analysis to the period for which the number of subjects observed was adequate for reliable definition of longer-term survival. Ninety-five percent CIs were computed using the Greenwood estimate of variance (26). We explored group differences in laboratory indices in two ways. First, we compared median values for the three study groups cross-sectionally. Second, we compared the odds of abnormal values in the seropositive recipients with those of control recipients longitudinally at three different time points. The cross-sectional comparison was made 36 ± 3 months after the index transfusion for the two groups of recipients and 24 ± 3 months after entry for healthy control subjects. This latter comparison was done at approximately the same time as laboratory testing of specimens from the former two groups. The comparison used an overall Kruskal-Wallis one-way analysis of variance of ranks (27) and then three pairwise Wilcoxon rank sum tests of median laboratory values. The trends in laboratory values for seropositive recipients relative to the control recipients were determined for three intervals, all measured from the date of the index transfusion: 24 ± 3 months, 30 ± 3 months, and 36 ± 3 months. For each interval, we calculated a seropositive recipient's risk for having an abnormal laboratory value relative to that risk for a control recipient. Cutoffs indicating abnormality were obtained from the distribution of laboratory values for healthy control subjects and defined to be the 25th percentile if the direction of abnormality was downward and the 75th percentile if the direction of abnormality was upward. Statistical significance of each relative risk was tested using a standard 2 x 2 chi-square test. Sample sizes for these comparisons varied because of missed visits or missing values. The median number of observations was 56, 69, and 71 for positive recipients and 39, 47, and 39 for recipient controls for the three intervals, respectively. Comparisons for hematologic and immunologic data were judged to be significant at the 0.05 level. Because of the multiplicity of comparisons and interconnections between some markers, however, results that are significant at the 0.05 level should be interpreted as trends. A Bonferroni adjustment of all 50 comparisons suggests that statistical significance be established at 0.001 = 0.05/50.

15 November 1990 • Annals of Internal Medicine • Volume 113 • Number 10

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19716/ by a University of California San Diego User on 06/05/2017

Results

Table 2. Characteristics

Incidence of Antibodies to Human Immunodeficiency Virus Type 1 among Recipients

Characteristic

From the 307 confirmed donations in 1984 and 1985, the four blood services distributed 548 components to approximately 180 hospitals. Hospitals informed the investigators of the identity or outcome (or both) of 453 recipients of 467 of these components. Two hundred and eighty-seven recipients of 293 components died before they could be contacted. Two received the index transfusion and died shortly thereafter. One unenrolled recipient refused participation because she was "too sick with AIDS" approximately 14 months after the index transfusion, and 2 additional unenrolled recipients died from AIDS after approximately 24 to 27 months. We identified 166 recipients who survived the condition for which they were transfused; 133 (80%) recipients of 135 components from 101 positive donors enrolled. Reasons for not participating included inability to locate the recipient, geographic distance, refusal of the patient's physician, and refusal of the patient. The enrollment period was from 6 to 36 months (median, 20 months) after the index transfusion. Anti-HIV-1 positivity was confirmed in 120 patients (90%). The cases of 9 of the 120 anti-HIV-1-positive recipients were excluded from the analysis. One had HIV-1 infection when the index transfusion was given; 3 had been treated for hemophilia; and 5 reported intravenous drug abuse. With these exclusions, the incidence at enrollment for patients with no other known exposure remained 89.5% (111 of 124) (CI, 84.1% to 94.9%). Transfusion of the index component was verified through all available records for the 13 seronegative recipients. Nineteen donors' components were given to 2 recipients, and 4 donors' components were given to from 3 to 7 recipients. The anti-HIV-1 status of recipients with common exposure to a donor was concordant in all instances. Two recipients who received a donation from 1 anti-HIV-1-positive donor both remained seronegative; all of the other 53 recipients with shared exposure were anti-HIV-1 positive when seen. Infection rates among the 124 evaluated recipients were not influenced by sex, age, underlying conditions,

of iRecipient Study• Groups* Seropositive Recipients

Age, n (%) < 10 y 10-59 y > 60 y

Mean ± SD Sex, n (%) Male Female Component transfused, n (%) Whole blood Packed red cells Washed red cells Platelets FFP/cryoprecipitate Reason for transfusion, n (%) Surgery Trauma Medicalt

(/!= I l l )

Control Recipients (n = 64)

7(6) 47 (42) 57 (51) 51.8 ± 22.1

10 (16) 17 (27) 37 (58) 49.9 ± 27.0

59 (53) 52 (47)

39 (61) 25 (39)

6(5) 71 (64) 0(0) 19 (17) 15 (14)

1(2) 41 (64) 1(2) 13 (20) 8(13)

50 (45) 10(9) 51 (46)

22 (34) 4(6) 38 (59)

* The P value for age is 0.03; for inean age ± SD, 0.62; for sex, 0.29; for component transfused, 0.48; amd for reason for tr;ansfusion, 0.23. FFP = fresh frozen plasma. t Malignancy, hematologic and reilal diseases, prematurity, and other medical reasons.

or the reason for the transfusion. They were also similar for exposures to whole blood (6 of 7), packed red cells (71 of 78), platelets (19 of 21), fresh frozen plasma (11 of 12), and cryoprecipitate (4 of 4). Neither of 2 recipients of washed red cells was anti-HIV-1 positive; no other recipient of components from those 2 donations was enrolled. Progression to the Acquired Immunodeficiency Syndrome From enrollment through the last visit before 1 July 1988, 26 (23%) of the 111 persons infected with HIV-1 through transfusion died; 9 (35%) of those who died are known to have developed AIDS, but death was attributed to other causes in 6 of them. Among the 85 living seropositive recipients, 3 (4%) are known to have AIDS. Thus, 12 (11%) developed AIDS within 4 years after the probable date of infection. Figure 1 shows the Kaplan-Meier plot of the cumulative probability of developing AIDS in this population. The risk was estimated to be 13% 38 months after transfusion (CI, 7.5% to 21.6%). Laboratory Values in Infected Recipients without the Acquired Immunodeficiency Syndrome

Figure 1. The Kaplan-Meier estimate of the probability of developing the acquired immunodeficiency syndrome {AIDS), plotted as a function of time since transfusion with the index donation (solid line). The broken lines show the 95% CIs.

We assessed the hematologic values and lymphocyte subpopulations of the anti-HIV-1-positive recipients whose HIV-1 infection had not progressed to AIDS. Because persons who acquire HIV-1 infection from transfusion have underlying conditions that may be associated with persisting abnormalities in some laboratory indices, we compared the distribution of values not only with that in healthy control subjects, but also with that in anti-HIV-1-negative control recipients. We identified 154 living recipients of components

15 November 1990 • Annals of Internal Medicine

• Volume 113 • Number 10

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19716/ by a University of California San Diego User on 06/05/2017

735

from seronegative repository donations. Sixty-six (fortythree percent) agreed to participate, of whom 2 were excluded because of anti-HIV-1 positivity or an indeterminate immunoblot result. For the 64 subjects whose data were analyzed, the intervals from the index antiHIV-1-negative transfusion to enrollment ranged from 11 to 35 months (median, 20 months). Chi-square tests showed that the seropositive and control recipients did not differ significantly in sex, type of component administered, or underlying disease for which transfusion was given (Table 2). The mean ages did not differ significantly, but the age distribution showed that a larger percentage of seropositive recipients (42%) than of control recipients (27%) were from 10 to 59 years of age. As an indicator of the chronicity and severity of underlying diseases, we used the frequency of transfusion and surgical procedures from entry until the last observation covered in the present analysis. Seventeen (fifteen percent) of seropositive recipients and 16 (25%) control recipients had blood transfusion on one or more occasions; the difference was not significant. Thirty-two (twenty-nine percent) of the seropositive recipients and 31 (48%) of the control recipients had interim major or minor operative procedures. This difference was significant by chi-square testing (P = 0.015). Table 3 shows median values for hematologic variables and immunophenotypically defined lymphocyte subpopulations 36 ± 3 months after the index transfusion for the two groups of recipients and the healthy control subjects. The control recipients differed from

the healthy control subjects in several indices. They had lower platelet and CD4+ and CD4+CD45RA+ lymphocyte counts. They also had increased CD8+I2+ lymphocyte counts and increased CD4+CD29+/CD4+ and CD8+I2+/CD8+ cell ratios. The CD4+/CD8+, CD4+CD29+/CD8+I2+, and CD4+CD45RA+/CD8+I2+ ratios were very significantly lower (P < 0.0001) in control recipients than in healthy control subjects. The seropositive recipients differed significantly from the healthy control subjects in almost all variables. The only exceptions were hemoglobin level, the absolute counts of CD2+CDllb+ and CD8+ cells, and the proportion of CD4+CD45RA+ cells among the CD4+ population. Seropositive recipients generally differed significantly from the control recipients in the same variables in which they differed from the healthy control subjects. Some differences, however, were less profound, as indicated by comparison of median values. Not differing significantly were platelet counts and the ratios of CD4+CD29+/CD4+, CD4+CD45RA+/CD4+, and CD20+CD21+/CD20+. In contrast to the modestly higher than normal counts for the CD4+CD29+ and CD56+ subsets among control recipients, both subsets were markedly lower than normal in seropositive recipients. Examination of the medians longitudinally (Table 4) from 24 ± 3 to 36 ± 3 months after the index transfusion indicated that the trend among seropositive recipients was generally toward lower counts compared with

Table 3. Laboratory Variables foiy Healthy \ Control Siibjects 24 ± 3 Months and Anti-HIV-1-Positive Recipien ts without AIDS 36 ± 3 MonU hs after the Variable

Hematologic indices Hemoglobin, g/dL Leukocytes, cells/uL Platelets, X / 0 7 / A L

Lymphocytes, cells/[xL Immunophenotypic subpopulations CD2+, cells/pL CD2+CDw26+, cells/pL CD2+CDw26+/CD2+, ratio C D 2 + C D l l b + , cells/)JLL CD4+, cells/uL CD4+CD29+, cells/pL CD4+CD29 +/CD4+, ratio CD4+CD45RA+, cells/>L CD4+CD45RA+/CD4+, ratio CD8+, cells/uL CD8+I2+, cells/'pL CD8+I2+/CD8+,ra//0 CD56+, cells/yL CD4+/CD8+,raf/

Infection with human immunodeficiency virus type 1 (HIV-1) among recipients of antibody-positive blood donations.

To assess the incidence of human immunodeficiency virus type 1(HIV-1) transmission by antibody (anti-HIV-1)-positive blood components, and to determin...
1MB Sizes 0 Downloads 0 Views