Weight gain and increased concentrations of receptor proteins for tumor necrosis factor after patients with symptomatic HIV infection received fortified nutrition support

OBJECTIVE

To determine whether certain nutrients and dietary factors act as modulators of the immune system and improve the nutritional status of immunocompromised patients.

DESIGN

Controlled, double-blind, crossover phase trials of the effects of a fortified formula in patients infected with the human immunodeficiency virus (HIV). Patients consumed a control formula for 4 months and a study formula for 4 months.

SUBJECTS

Ten men with symptomatic HIV infection who were following stable medication regimens and had no malignancies, mycobacteriosis, or additional virus infection requiring systemic treatment.

INTERVENTION

Formula fortified with alpha-linolenic acid (1.8 g/day), arginine (7.8 g/day), and RNA (0.75 g/day) and a standard formula.

MAIN OUTCOME MEASURES

Nutritional status determined by anthropometric, bioelectrical, biochemical, and dietary assessment; energy expenditure determined by indirect calorimetry; disease progression; CD4 lymphocyte counts; HIV p24 antigen plasma concentrations; tumor necrosis factor (TNF) receptor proteins; and compliance control parameters.

STATISTICAL ANALYSES PERFORMED

Student's t tests for paired and unpaired data.

RESULTS

Fortified nutrition resulted in a weight gain (+ 2.9 kg/4 months vs -0.5 kg/4 months with the control formula, P < .05), an incorporation of eicosaenoic acid into erythrocyte cell membranes (+ 47% of baseline values, P < .05), and increased plasma arginine concentrations (96.8 +/- 45.1 vs 51.8 +/- 20.9 mumol/L, P < .01). The serum concentrations of the soluble tumor necrosis factor receptor (sTNFR) proteins increased during the study period (sTNFR 55 = + 0.23 vs -0.40 ng/mL, P < .001; sTNFR 75 = + 0.90 vs -0.36 ng/mL, P < .01), whereas no changes in CD4+ lymphocyte counts were observed.

CONCLUSION

Increasing dietary intakes of n-3 polyunsaturated fatty acids, L-arginine, and RNA increased body weight, possibly by modulating the negative effects of TNF.

Anti-V3 loop spectrotype in HIV-infected individuals during zidovudine therapy

In order to investigate the role played by zidovudine (ZDV) as immune modulator, particularly on B-cell response, the anti-V3 loop spectrotype in 115 sera from 26 HIV-infected individuals was evaluated, prior to and during treatment with ZDV, by isoelectric focusing and reverse blotting (IEF-RB), a technique useful for indirectly measuring the activity and the number of B-cell clones. All 18 patients showing seroreactivity by IEFRB displayed a clear oligoclonal banding pattern, with no change in the spectrotype (i.e. new bands), in sequential analysis over the course of therapy. Only minor changes in band intensity were found, without any correlation with ZDV treatment or CD4+ cell count. In addition, among the sera reactive in spectrotypic analysis, the percentage of those with p24 antigen positivity was significantly lower than those with no detectable p24 antigen (19.8% vs 80.2%, respectively, p = < 0.0001, Fisher's exact test). In conclusion, it could not be demonstrated by IEFRB that there was any effect of ZDV on the activity and the number of anti-V3 specific B-cell clones. This data is in line with previous studies showing the constancy of anti-gp120 antibody spectrotype over the long course of the disease.

Effect of stavudine on human immunodeficiency virus type 1 virus load as measured by quantitative mononuclear cell culture, plasma RNA, and immune complex-dissociated antigenemia

The antiviral effect of stavudine (2', 3'-didehydro-3'-deoxythymidine) against human immunodeficiency virus (HIV) type 1 was measured in 15 HIV-infected patients at baseline and at weeks 4, 10, 22, 34, and 52 of therapy. Patients received 0.1, 0.5, 1.0, or 2.0 mg/kg/day of stavudine. At all time points examined during the 52 weeks of therapy, the median virus titers in peripheral blood mononuclear cells were decreased 1-2 logs, and median immune complex-dissociated antigen levels were reduced 37%-67% compared with baseline values. Plasma RNA content measured by polymerase chain reaction was reduced approximately 0.5 log from baseline median values at both time points examined (weeks 10 and 52). These data demonstrate that stavudine has a substantial and durable antiviral effect.

Serum HIV-1 RNA load to predict CD4+ T-cell depletion in asymptomatic patients

The temporal association between the increase in viral replication and the depletion in CD4+ T cells in HIV-1 infection is not yet clear. To investigate this phenomenon HIV-1 RNA was quantified in several frozen sera from 20 asymptomatic HIV-1 infected patients in the 2 years preceding CD4+ T cell depletion of 50% or more, and compared with 20 HIV-1 infected paired patients who were stable in the same period. In each group, no statistically significant variation in the mean HIV-1 RNA titre was found between the last checkup and the one 24 months earlier. The mean HIV-1 RNA titre was 10(3.86) copies/ml in the non-progressor group and 10(5.12) copies/ml in the progressor group. These data support the view that the quantity of circulating HIV-1 RNA is an early predictor of disease progression that is relatively constant during the asymptomatic period of HIV-1 infection.

[A perinatal diagnostic problem: patients serologically negative but infected with HIV]

Comprehensive HIV diagnosis of 142 patients was done by ELISA, IgG immunoblot, IgA immunoblot (in patients under 24 months), plasma antigen determination, viral isolation by culture and genome detection by polymerase chain reaction (PCR) amplification. Results showed that 14 patients (10%) with negative or indefinite results for anti-HIV-1 IgG antibody were in fact infected by the virus. Eleven of these patients were between 2 and 24 months of age, two between 4 and 6.5 years and one was 30 years old. Diagnosis was obtained by antigen positivity in four of them; by PCR amplification of peripheral mononuclear cells of proviral DNA in five of them; by PCR and immunofluorescence (IF) of cultured cells in four cases, and the last diagnosis was made by IgG immunoblot and IF of the viral culture. These cases pose a problem because of false negative HIV serology, particularly in patients under 24 months of age.

Serum anti-p24 antibody concentration has a predictive value on the decrease of CD4 lymphocyte count higher than acid-dissociated p24 antigen

We studied the prognostic value on the decrease of CD4 lymphocyte count of anti-p24 antibody (ab) titer and compared this value with that of polyclonal and monoclonal p24 (ag) titer before and after immune complex dissociation (ICD); 53 human immunodeficiency virus (HIV)-infected patients having CD4+ counts above 400/mm3 when first examined were followed up over a 3-year period including at least four visits; HIV disease progressors (n = 18) were defined as having CD4+ counts below 200/mm3 and non-progressors (n = 35) as having CD4+ counts still above 400/mm3 at the end of follow-up. Sera were collected at each visit and assayed for p24 ag with and without ICD and for anti-p24 ab titer. The mean anti-p24 ab titer of progressors and of non-progressors at entry was significantly different. A threshold of anti-p24 ab titer indicating a HIV progression was determined at 1/300. The proportion of individuals with an anti-p24 ab titer lower than 1/300 at least once during the study period was 34% in non-progressors and 94% in progressors. The difference between progressors and non-progressors at entry was significant with monoclonal p24 ag without ICD and more significant with monoclonal p24 ag after ICD. The marker having the highest predictive value was the anti-p24 ab titer, then monoclonal p24 ag with ICD, then polyclonal p24 ag with ICD. Anti-p24 ab is an earlier and stronger marker of the decrease of CD4 lymphocyte count than p24 ag even after ICD.

Prevalence of viremia in human immunodeficiency virus-infected patients with renal disease

BACKGROUND

The prevalence of viremia and its relationship to the pathogenesis of nephropathy in human immunodeficiency virus (HIV)-infected patients with renal disease is unknown. To assess the prevalence of plasma viremia in HIV-infected patients with chronic renal disease, we performed a cohort study in two urban university medical centers.

METHODS

Samples of blood from 11 HIV-infected patients with renal failure who were treated with hemodialysis were analyzed concurrently with control samples from three non-HIV-positive patients receiving hemodialysis treatment. Samples from four HIV-infected patients with chronic renal insufficiency were evaluated concurrently. Thirty-three HIV-infected patients with serum creatinine levels of less than 132 mumol/L (1.5 mg/dL), and trace or absent dipstick proteinuria served as controls for the population with renal disease. The patients infected with HIV were staged by CD4 cell counts and the presence of opportunistic infections. Blood samples were analyzed for plasma HIV p24 antigenemia by antigen capture enzyme-linked immunosorbent assay. Blood samples were analyzed for the presence of viremia by infection of normal stimulated peripheral blood mononuclear cell cultures with plasma samples and detection of HIV p24 antigen in culture supernatants.

RESULTS

Two of the 11 patients treated with hemodialysis had evidence of HIV p24 antigenemia, while seven of the 11 had evidence of plasma viremia. The proportion of hemodialysis patients with detectable antigenemia and viremia was similar to that in patients with chronic renal insufficiency. A significantly greater proportion of HIV-infected patients with renal disease had plasma viremia and antigenemia, compared with HIV-infected patients without renal disease. In logistic regression analysis, race, CD4 cell count (either on a continuous scale or dichotomized at 0.2 x 10(9)/L), and treatment with zidovudine were not significantly associated with the presence of plasma viremia, but patient age and the presence of renal disease were factors independently associated with viremia.

CONCLUSIONS

The similar proportions of HIV-infected patients with viremia in groups of patients with chronic renal insufficiency and with renal disease treated with hemodialysis suggest that dialysis treatment does not increase the prevalence of plasma viremia in HIV-infected patients with renal disease. The similar proportions of HIV-infected hemodialyzed patients and patients with chronic renal insufficiency with plasma viremia, and the greater prevalence of viremia in patients with renal disease compared with HIV-infected patients without clinical renal disease suggest that plasma viremia and renal dysfunction are related. Whether this represents a cause and effect relationship is unknown. The greater prevalence of viremia in HIV-infected patients with renal disease has implications for the pathogenesis of HIV-related renal diseases and for caregivers in clinical settings and dialysis units.

Serum 27E10 antigen: a new potential marker for staging HIV disease

MRP8 and MRP14 are myeloic related proteins expressed by most circulating and emigrated neutrophils and monocytes. Their composite molecule MRP8/14 (27E10 antigen) was shown to exhibit striking antimicrobial properties. The aim of the present study was to assess the value of MRPs as markers for detection of the different stages of HIV infection (Centres for Disease Control and Prevention, 1993). By employing the ELISA technique we measured serum concentrations of these proteins in samples from 122 HIV patients at the various stages of disease, and the results were compared with those for healthy controls. Serum levels of the heterodimeric molecule 27E10 were significantly increased (P < 0.001) in patients with CDC stages II and III, with the highest levels being in patients with stage III and acute ongoing opportunistic infections. For the single component MRP14, significantly raised levels (P < 0.05) were only found in HIV stage III individuals with acute clinical events. Similar associations were not found for MRP8 alone. Increase was not related to CD4+ cell count. There was a significant correlation between 27E10 antigen serum concentrations and levels of neopterin in patients with HIV stages II and III without acute concurrent illness. Patients being treated with Zidovudine showed no statistically significant variation in levels of 27E10 and its single components MRP8 and MRP14 compared with untreated patients. These findings suggest that elevation of MRP14 levels occurs in HIV+ individuals at later stages post-HIV infection, after the onset of opportunistic infections. 27E10 antigen is concluded to be a potential marker for the different stages of HIV disease.

[Detection of human immunodeficiency virus antigen both free and in immune complexes]

The aim of this work was to increase sensitivity in the detection of antigens from HIV-infected patients, through a process of immune complex dissociation without loss of antigenicity. 500 microliters of sera were mixed with 100 microliters of PEG 12%, stored one night in refrigerator, and centrifuged at 2000 g during 20 minutes. 200 microliters of buffer AcH/Ac- (pH 3.5) were added to the sediment, and incubated at 37 degrees C during one hour with periodic shaking. This was neutralized with 100 microliters of buffer TRIS/CIH (pH 8.6). The antigen was investigated in the original sample, supernatant and sediment. Samples of 105 patients with positive serology, confirmed by Western Blot following CDC criteria, were processed. The antigen was detected in 62 (59%) samples precipitated with PEG, but only 35 (33%) when conventional methods were used. Applying statistics X2: 13.97, P < 0.001, a highly significant association can be observed between PEG dissociation treatment and antigen detection. 27 negative sera by the standard method became positive in the whole sediment, and only 8 in the supernatant. In addition, 40 negative sera were processed, which had not become positive for the antigen by PEG treatment.

Prospective study of antigenemia, plasma viremia and lymphocytic viremia in HIV-infected hemophiliacs

A total of 186 blood samples from 24 HIV-1 seropositive hemophiliac patients, monitored every four months for 29 months, were investigated for the presence of viral antigen in plasma. In addition, peripheral blood mononuclear cells (PBMC) were cultured for HIV-1, using normal PBMC as a target for replication. Antigenemia was detected in 51% of the patients and from PBMC in 87.5% of the patients. The incidence of HIV isolation in asymptomatic patients (42.8%) was similar to that found in symptomatic patients (51.4%). Patients with opportunistic infections had a higher incidence of lymphocytic viremia (p < 0.05). Plasma viremia was closely associated (p < 0.05) with low CD4+ counts and infection progression. The persistence of antigenemia was also a marker of a poor clinical course. In treated patients, plasma viremia was the marker that better correlated with the clinical course, and it did not appear during the first nine months of therapy. Zidovudine doses of > 500 mg/day significantly lowered the appearance of antigenemia and lymphocytic viremia (p < 0.05).

Stability of free and complexed human immunodeficiency virus type 1 antigen at 4 degrees C and at room temperature

Free and immune-complex-dissociated (ICD) human immunodeficiency virus type 1 (HIV-1) antigenemias in serum specimens stored at room temperature (RT) and 4 degrees C for 1 to 35 days were evaluated. At all time points examined, there was no significant loss in detectable levels of ICD HIV-1 antigen at either RT or 4 degrees C. Free HIV-1 antigen was not stable in serum samples stored at RT for more than 2 days but was stable in samples stored at 4 degrees C for up to 4 days. Loss of free antigen occurred more rapidly in samples with high antigen content at baseline. Use of the ICD antigen assay allowed accurate quantitation of antigen in samples stored at RT or 4 degrees C for as long as 1 month.

Utility of various commercially available human immunodeficiency virus (HIV) antibody diagnostic kits for use in conjunction with efficacy trials of HIV-1 vaccines

There is a need for human immunodeficiency virus (HIV) screening assays which will distinguish uninfected HIV vaccine recipients from HIV-infected individuals. Commercial screening kits were used to test serum samples from low- and high-risk participants in clinical trials before and after immunization with various recombinant HIV type 1 (HIV-1) envelope glycoprotein 120 (gp120) candidate vaccines. All kits were 100% sensitive in detecting HIV infection. Both Murex Single Use Diagnostic System and United Biomedical, Inc., HIV type 1 or 2 (HIV-1/2) enzyme immunoassay (EIA) kits, which detect antibodies to HIV-1 gp41, were 98 to 100% specific when used to screen baseline or recombinant gp120-vaccinated populations as vaccine-induced antibodies to gp120 were nonreactive in these tests. The Abbott HIVAB HIV-1 EIA (lysate of whole infected cells, reactive with anti-gp120 antibodies) gave high levels of reactivity due to vaccine-induced antibodies and a high baseline rate of false positives (12 of 83) among nonvaccinated high-risk volunteers. Assays containing only gp41 and p24 solid-phase components are compatible with gp120-based vaccines but are unlikely to be useful in a similar role for vaccines containing gp160, gp41, or gp120 plus p24 antigens. Efficacy trials must be designed in concert with available diagnostic screening assays to avoid problems caused by vaccine-induced seroconversion in high-risk populations.

Major antigenic region on the integrase (IN) protein of human immunodeficiency virus type 1 determined by reactivity of human sera and a monoclonal antibody to IN protein

The gene encoding the integrase (IN) protein of human immunodeficiency virus type 1 (HIV-1) was expressed in vaccinia virus and Escherichia coli, and sera from 55 HIV-1-infected individuals were examined for immunoreactivity to the recombinant IN proteins by Western immunoblot. Approximately 98% (54 of 55) of the HIV-1-infected individuals showed reactivity to both the full-length IN protein of 32 kDa (IN32 protein) and the carboxy-terminal portion of the IN protein (IN17 protein). Serum samples from only 6 of the 54 antibody-positive individuals and a monoclonal antibody against the IN protein, 6F4, reacted with the amino-terminal portion of the IN protein (IN15 protein). The eight AIDS patients tested were seronegative to IN15 protein. The magnitude of reactivity to the recombinant IN proteins decreased slightly in the progression of the course of HIV-1 infection. These results suggest that a B-cell immunodominant epitope(s) on the IN protein is located on the C-terminal IN17 portion and that a minor epitope(s) recognizable by 6F4 and by rare patients is on the N-terminal IN15 portion.

[New diagnostic developments in infection serology]

New developments concerning the diagnostic of virus-transmitted diseases led to a strong improvement of virus security in hemotherapy. The introduction of additional virus antigens associated with the core region of the hepatitis C virus (HCV) resulted in an increase of sensitivity and especially of specificity of second-generation enzyme immuno assays (EIAs). The recombinant immunoblot (RIBA) with 4 synthetic antigens continues to be the most accepted confirmatory test in the field of HCV diagnostic. Beside the simultaneous detection of specific IgG and IgM, especially the HIV-p24 antigen test is discussed concerning HIV-1, -2 diagnostic in order to shorten the 'windows period' in the early time of infection. It is assumed that the HIV-p24 antigen test is not able to effectively improve virus security in areas with low HIV-1, -2 prevalence.

Mother-to-child HIV-1 transmission: quantitative assessment of viral burden as a diagnostic tool and prognostic parameter in HIV-1-infected children

Polymerase chain reaction was performed in 251 infants born to HIV-1-seropositive mothers to diagnose HIV-1 infection. Assay specificity was invariably > 95%, regardless of age at testing, while sensitivity ranged from 15% in neonates (within 48 h of birth) to > 95% in infants over 1 month of age. Evaluation of viral burden in 43 infected infants by means of quantitative DNA-PCR disclosed that the number of HIV-1 proviruses ranged from 5 to 947 per 100,000 peripheral blood mononuclear cells. Clinical follow-up demonstrated that a high viral burden was associated significantly with disease onset.

Therapeutic effect of an anti-HIV candidate vaccine

A chimpanzee infected with the HIV since 8 months and presenting regularly with antigenemia was inoculated with a candidate vaccine. It received 3 doses, one dose every 15 days. Thirty days after the third injection, we noted the disappearance of the HIV antigens in the serum and its persistence in the lysate of the cells. We noted also a strong precipitation reaction both in the tube and in the gel between the antibodies and the antigens released by the lysis of the cells. The analysis of this precipitate demonstrated that it was constituted of immune complexes in which the antibodies were of high affinity. At the 240th day after the third injection of the candidate vaccine we noted the disappearance of the HIV antigens in the lysate of cells as well. From these results, we conclude that the candidate vaccine we tested can elicit high affinity antibodies able to clear the HIV antigens and destroy the cells containing the HIV antigens probably with the help of specific killer cells.

High rate of nonspecific anti-hepatitis C reactivity amongst homosexual men in comparison with that found in other sexually active groups and blood donors. Viral Hepatitis and AIDS Study Group

OBJECTIVES

To investigate the concordance of anti-hepatitis C virus (anti-HCV) reactivity by a second-generation enzyme immunoassay (EIA-2) and by a four-antigen recombinant immunoblot assay (4-RIBA) in homosexual men, in comparison with that found in other sexually active groups and blood donors.

DESIGN

Prospective study.

SETTING

Tertiary referral centre, Seville, Spain.

SUBJECTS

A total of 1203 subjects were studied. Eight hundred and three were sexually active individuals: 547 female prostitutes, 88 heterosexual men who had frequent sexual intercourse with prostitutes, and 168 homosexual men. All of them denied blood transfusion and parenteral drug use. In addition, 400 voluntary blood donors were selected at random.

MAIN OUTCOME MEASURES

All serum samples were screened for anti-HCV by EIA-2 and repeatedly reactive sera were tested by 4-RIBA. Homosexual men were also screened for anti-human immunodeficiency virus (anti-HIV), hepatitis B virus (HBV) markers and gammaglobulin concentration. Finally, serum samples from homosexual men reactive for anti-HCV by EIA-2 were analyzed for HCV-RNA by polymerase chain reaction (PCR).

RESULTS

Concordance between EIA-2 and 4-RIBA in female prostitutes (71.4%), clients of prostitutes (75.0%), and blood donors (83.3%) was significantly higher than in homosexual men (38.8%) (P < 0.04). In this collective the concordance between 4-RIBA and PCR was 85.7% for positive cases and 88.8% for negative ones, and EIA-2 ratios in reactive sera were significantly higher in 4-RIBA confirmed cases (P < 0.0001). No correlation between false positive EIA-2 results and presence of HIV infection, HBV markers or hypergammaglobulinaemia was found in homosexual men by univariate analysis.

CONCLUSIONS

There is a high level of non-specific anti-HCV reactivity by EIA-2 amongst homosexual men in comparison with that found in other sexually active groups and blood donors. The true prevalence of HCV infection amongst homosexual men could be even lower than previously described.

Stable biological and antigenic characteristics of HIV-2SBL6669 in nonpathogenic infection of macaques

The purpose of the present study was to investigate if the biological and antigenic properties of human immunodeficiency virus type 2 change over time in cynomolgus macaques (Macaca fascicularis) experimentally infected with HIV-2SBL6669. Sequential virus isolates and serum samples were obtained during a 2-year period and studied in autologous neutralization assays. All six macaques studied seroconverted shortly after infection and remained healthy during the observation period. Virus could be isolated from all six animals during the first 100 days postinfection. Thereafter four monkeys became virus isolation negative, either permanently or transiently (two macaques each), whereas two macaques remained virus isolation positive during the entire observation period. Sequential reisolates from the macaques invariably replicated in HUT-78, U937-2, and Jurkat-tat cell lines, similarly to the HIV-2SBL6669 inoculum virus. The ability to produce neutralizing antibodies correlated with positive virus isolations, hence four macaques produced neutralizing antibodies against inoculum virus and sequential reisolates. Once the neutralizing antibody appeared, sequential reisolates obtained at both early and late time after infection were neutralized, indicating that the neutralizing epitopes of the virus are conserved in the infected animals over time. This is different from the pathogenic SIVsm infection in macaques or HIV-1 infection in humans, where emergence of neutralization resistant variants seems to be the rule. In contrast, in HIV-2-infected macaques the biological properties of the virus are stable and the neutralizing antibody response shows extensive cross-reactivity.

[The isolation and characteristics of monoclonal antibodies to recombinant proteins of HIV-1 and HIV-2--the env and gag gene products]

Ten hybridomas secreting monoclonal antibodies (Mab) against recombinant HIV-1 and HIV-2 antigens were produced (3 Mab anti-gag protein, 2 anti-env1, and 5 anti-env2). In the immunoblotting assay all the anti-gag Mabs reacted with HIV capsid protein p24, whereas one of them reacted also with p55 protein and with 7 other polypeptides. Another anti-gag Mab cross-reacted with the antigen of subpopulation of human peripheral blood lymphocytes. The third one interacted with the antigen of both HIV-1 and HIV-2. All the 10 Mabs interacted with natural HIV antigens and can be used for identification and differentiation of HIV-1 and HIV-2.

A rapid test for the detection of human immunodeficiency virus antibodies in cord blood

A commercially available rapid test (HIVCHEK) was compared with an enzyme-linked immunosorbent assay (ELISA) for identifying human immunodeficiency virus type 1 in the serum of newborn infants. Of 1309 cord blood samples tested, the HIVCHEK test detected all the true-positive samples detected by ELISA. Of the 35 samples with positive ELISA results, six had negative results on Western blot; only 1 of the 30 samples with positive HIVCHEK results had negative results on Western blot. Thus the HIVCHEK test can be used to facilitate the rapid identification of HIV-1 in the serum of newborn infants.

Early detection of IgA specific antibodies in HIV-1 infected children by peptide-ELISA and peptide time-resolved fluoro-immunoassay

The presence of specific IgA antibodies in sera from 25 infants born to HIV-1 seropositive mothers was investigated by peptide-ELISA and peptide time-resolved fluoro-immunoassay (TR-FIA). The infants had been monitored at different times after birth for clinical signs and/or symptoms of HIV-1 infection and for detection of HIV-1 in lymphocyte cultures. Serum samples had also been tested for HIV-1 IgG antibodies by commercial ELISA and Western blot and for p24 antigen. Eleven of 25 children were then identified as infected. IgA detection was performed after rProtein G treatment to remove interfering IgG. In the infected group, IgA specific antibodies to a synthetic peptide representing a highly conserved region of the transmembrane glycoprotein gp41 (env: 594-613) were detected in 27 (73%) out of 37 serum samples (9 of 11 children) by the peptide-ELISA test. IgA specific antibodies to the same peptide were found in 30 (81%) sera (9 of 11 children) by the peptide-TR-FIA. Specific HIV-1 IgA antibodies were detected as early as 2 months of age in serum samples from five out of seven children (71% sensitivity) using peptide-ELISA and from six out of seven (86% sensitivity) by peptide-TR-FIA. Conversely, IgA specific antibodies to HIV-1 were absent in two infected children as well as in the sera of all uninfected children tested during the follow up period. Since maternal IgA does not cross the placenta, IgA detection in the serum of the infant is indicative of HIV-1 infection.(ABSTRACT TRUNCATED AT 250 WORDS)