Detection of measles- and mumps-specific IgG antibodies in paired serum and oral fluid samples from Norwegian conscripts.

The aim of this study was to measure the seroprevalence to mumps in Norwegian conscripts belonging to the first children vaccination cohorts that had been offered two doses of MMR vaccine. The seroprevalence to mumps was 76% with the Microimmune assay and 85% with the Enzygnost assay. We also compared the performance of the Microimmune assay for detection of mumps- and measles-specific IgG antibodies in 340 paired serum and oral fluid samples from the conscripts and evaluated the effect of revaccination. Mumps-specific IgG antibodies were detected in only 61% of the oral fluids. In contrast, high levels of measles-specific IgG antibodies were detected in both the serum and oral fluid samples. Based on these results, we are only able to recommend the use of oral fluid for surveillance of measles in Norway. Our results may also indicate that the seroprevalence necessary to interrupt transmission of mumps has not been reached in vaccinated young adult Norwegians. Seroconversion was observed in all initially measles seronegative conscripts after revaccination, whereas 23 of 27 initially mumps seronegative conscripts failed to seroconvert.

Seroprevalence of measles among Norwegian military conscripts in 2004.

The study presented here was conducted in order to evaluate the impact of Norway's childhood immunization program against measles, which was implemented in 1969. In the study, the level of measles immunity was measured among 1,405 military conscripts belonging to the first childhood immunization cohorts that were offered two doses of the measles, mumps and rubella vaccine. The overall seroprevalence of measles antibodies in this cohort was 89.3%. Two commercially available antibody assays were used, and the discordance between the two assays was 10.5%. Similar levels of immunity to measles were detected in earlier studies of Norwegian conscripts belonging to different childhood immunization cohorts.

Oral spray immunization may be an alternative to intranasal vaccine delivery to induce systemic antibodies but not nasal mucosal or cellular immunity.

Sixty-five healthy adult volunteers were immunized four times at 1-week intervals with an inactivated whole-virus influenza vaccine based on the strain A/New Caledonia/20/99 (H1N1) without adjuvant. The vaccine was administered as nasal spray with a newly developed device to secure intranasal delivery (OptiMist, OptiNose AS, Oslo, Norway), as regular nasal spray, nasal drops or as an oral spray. Significant IgA-antibody responses in nasal secretions were induced in volunteers immunized intranasally but not after oral spray immunization. In saliva, IgA antibodies were only marginally amplified even after oral spray immunizations. At least 73% of the volunteers belonging to any group of vaccine delivery reached serum haemagglutination inhibition titres of 40 or higher, considered protective against influenza, after only two vaccine doses. Those who had the vaccine delivered intranasally also showed evidence from in vitro secretion of granzyme B that cytotoxic T cells had been stimulated. Although immunization with the breath-actuated OptiMist device and nasal drops were superior with respect to both mucosal and systemic immune responses, oral spray immunization might still be considered for studies of mucosal adjuvants that are not yet acceptable for intranasal use.

Immunoglobulin-A antibodies in upper airway secretions may inhibit intranasal influenza virus replication in mice but not protect against clinical illness.

Mice immunized intranasally with a formalin-inactivated A/PR/8/34 (H1N1) influenza whole virus vaccine adjuvanted with cholera toxin, outer membrane vesicles from group B meningococci or formalin-inactivated whole cell Bordetella pertussis were protected against replication of the homologous virus in the nasal cavity. Only some mice were protected against clinical illness measured as weight loss and lowered body temperature. All mice immunized subcutaneously with one-tenth the intranasal vaccine dose without adjuvant were protected against clinical illness but not against local mucosal viral replication. Replicating virus was primarily found in animals with low concentrations of immunoglobulin (Ig)-A antibodies in saliva regardless of concentrations of IgG antibodies in serum. Clinical illness was seen only in those with low serum antibodies regardless of antibody levels in saliva. Nonreplicating nasal vaccines may not be sufficiently protective unless they also have a substantial influence on systemic immunity.

Humoral immunity to viral and bacterial antigens in lymphoma patients 4-10 years after high-dose therapy with ABMT. Serological responses to revaccinations according to EBMT guidelines.

The aim of this study was to investigate the late effects of ABMT on the immune system with regard to protective humoral immunity against common antigens and responses to recall antigens (vaccines). The vaccines were given according to EBMT guidelines from 1995. The protocol included 35 patients with malignant lymphoma in CR 4-10 years after ABMT, and 35 controls. The results show that prior to ABMT the proportion of patients with protective immunity against poliomyelitis, tetanus and diphtheria was similar to that of controls. At study entry 4-10 years after ABMT, the proportion of patients with protective immunity against poliomyelitis and diphtheria was reduced, while all patients maintained protection against tetanus. A significant decrease in geometric mean antibody concentrations or titres was observed against all three antigens during this period. Serum levels of antibodies against different pneumococcal serotypes were lower in the patients than in the controls prior to vaccination. The responses to pneumococcal vaccination, which is considered to be a T cell-independent vaccine, were studied. Unlike controls, a minority of patients achieved protective levels of antibodies after a single vaccination. Despite persistent levels of protective antibodies in many patients post ABMT, secondary booster responses after one vaccination with T cell-dependent vaccines (tetanus, diphtheria and polio) were absent. In conclusion, this study shows that post ABMT, a full re-vaccination program was necessary to mount responses comparable to those observed after a single vaccination in controls.

Human T-cell lymphotropic virus testing of blood donors in Norway: a cost-effect model.

BACKGROUND: Human T-cell lymphotropic virus type I and II (HTLV-I and II) are human retroviruses that can be transmitted by transfusion of whole blood. An HTLV-I infection is associated with adult T-cell leukaemia (ATL) and with tropical spastic paraparesis (TSP). Antibody tests from 5.5 million European blood donors have shown that the HTLV prevalence is low, ranging from 0 to 0.02%. This paper examines costs and effects associated with the intervention of testing all new blood donors for HTLV. METHODS: A mathematical model was used to calculate the number of cases prevented by the intervention. For a given prevalence of HTLV in the blood donor population, the model calculates the number of recipients infected by transfusion, and the number of partners and offspring that will in turn be infected. The model then calculates the number of subjects with disease due to HTLV-I infection and the number of deaths from disease. From these numbers the measures of cost and effect are calculated. RESULTS: Testing all new blood donors for HTLV is calculated to cost US$ 9.2 million per life saved, or US$ 420,000 per quality adjusted life year gained by the intervention, when the HTLV prevalence among donors is 1 per 100,000. When the prevalence among donors is 10 per 100,000 the intervention will cost US$ 0.9 million per life saved, or US$ 41,000 per quality adjusted life year gained. The same analysis shows that testing blood donors for human immunodeficiency virus (HIV) saves money when the HIV prevalence among donors is above 0.7 per 100,000. CONCLUSION: For Norway, studies suggest a willingness to pay to save a statistical life of approximately US$ 1.2 million. The costs fall under this value when the number of infected persons is > or = 8 per 100,000 donors. The results are uncertain because of the uncertainty in HTLV infection and disease parameters.

[Hepatitis C virus among pregnant women in Norway--occurrence of antibodies and pregnancy outcome]. / Hepatitt C-virus blant gravide kvinner i Norge--forekomst av antistoffer og svangerskapsutfall.

BACKGROUND: Transmission of hepatitis C virus from mother to child is well documented. The prevalence of antibodies against hepatitis C virus among pregnant women in Norway is however, not known. The aim of this study was to estimate the maternal prevalence of antibodies against hepatitis C virus and to study the association between presence of antibodies and fetal death. MATERIAL AND METHODS: From a study of 35,940 pregnant women, a random sample of 970 women and all women with fetal death after 16 weeks of gestation (n = 283), were tested for antibodies against hepatitis C virus. RESULTS: 7 out of 970 women in the random sample (0.7%; 0.2-1.3%, 95% confidence interval) had antibodies against hepatitis C virus. The same prevalence (0.7%, 2 out of 283) was found among women with fetal death. INTERPRETATION: The prevalence of antibodies against hepatitis C virus among Norwegian women was unexpectedly high. Further research is necessary to understand the causes and implications of this observation.

[Hepatitis C virus infection among pregnant women and their children in Norway]. / Hepatitt C-virusinfeksjon hos gravide og barn i Norge.

Estimates indicate the births of 6 to 78 children vertically infected with hepatitis C virus infection each year in Norway. There is insufficient knowledge of the magnitude of this health problem and the National Institute of Public Health commissioned the authors to approach issues relating to vertical transmission of hepatitis C virus (HCV) infection in Norway. The risk of vertical transmission of HCV appears to be associated with the titre of the maternal viral load. Vertical transmission from nonviraemic mothers has not been demonstrated. No postexposure prophylaxis exists. There is a lack of association between vertical HCV transmission and delivery mode and no association with breast feeding. Universal screening for HCV infection among pregnant women is not recommended. Children born to women known to be HCV-positive should be followed up with antibody and polymerase chain reaction investigations in order to clarify their HCV status. More studies of HCV infection among pregnant women and their children in Norway are needed.

[HTLV examination of Norwegian blood donors]. / HTLV-undersøkelse av norske blodgivere.

Approximately one third of the Norwegian blood donor population has been tested for infection with human T-lymphotropic virus type I and II (HTLV-I/II). This study was initiated to provide an indication as to whether or not the Norwegian transfusion service should screen the entire donor population for HTLV I/II. No HTLV-I infections were found among the blood donors. One new donor was confirmed HTLV-II positive. This individual had previously used drugs intravenously. HTLV-I/II infection can be regarded as a marker for risk behaviour, and testing can be of significance in the quality assurance of the transfusion service. We recommend that the entire blood donor population be tested for HTLV-I/II infections, and thereafter only new donors. The benefit of this scheme should be evaluated in the future.

The impact of exposure group on the progression rate to acquired immunodeficiency syndrome. A comparison between intravenous drug users, homosexual men and heterosexually infected subjects.

The objective was to study the impact of exposure group on the progression rate to the acquired immunodeficiency syndrome (AIDS). 289 subjects in Oslo, Norway, infected with the human immunodeficiency syndrome (HIV) and without major clinical signs of HIV infection (102 intravenous drug users, 151 homosexual men and 36 heterosexually infected subjects) were recruited to the Oslo HIV Cohort Study from 1989 and followed until 1 January 1995. 15 (14.7%) of the intravenous drug users, 56 (37.1%) of the homosexual men and 5 (12.5%) of the heterosexually infected subjects developed AIDS during a mean time of 47 months (p < 0.001, log rank test). When controlling for possible confounding variables (age, number of CD4+ lymphocytes, antiviral therapy at study entry, gender and year of HIV diagnosis), the relative risk of AIDS progression was 2.2 [1.1-4.5, 95% confidence interval (CI)] for homosexual men and 0.5 (0.2-1.3, 95% CI) for heterosexually infected subjects as compared to intravenous drug users. In a subgroup with known time of seroconversion (n = 60), 47% (18/38) of the homosexual men, 20% (3/15) of the intravenous users and none (0/7) of the heterosexually infected subjects developed AIDS (p = 0.04, log rank test). The results suggest that homosexual men have more rapid progression to AIDS than intravenous drug users and heterosexually infected subjects.

Interferon-alpha 2b therapy in low-activity hepatitis C: a pilot study.

BACKGROUND: Many patients with chronic hepatitis C have long periods of normal or near-normal liver enzyme levels, even though histologic alterations have been confirmed. The recommendation today is not to treat this patient group. METHODS: In a pilot study 23 hepatitis C virus (HCV) RNA-positive patients with alanine aminotransferase (ALAT) levels less than 1.5 times upper normal limits for at least 6 months on more than three occasions and with histologic liver abnormalities compatible with chronic hepatitis C were treated with 3 MU of interferon-alpha 2b three times a week for 6 months. RESULTS: Nine patients (39%) became HCV RNA-negative in serum during treatment, but only two (8.7%) remained so after 6 months' follow-up. Significantly more patients with genotype other than type 1 became HCV RNA-negative than patients with genotype 1 during treatment (P = 0.005). CONCLUSIONS: Patients with low-activity chronic hepatitis C have a response to interferon-alpha treatment similar to that of patients with increased ALAT levels. Genotype seems to influence the rate of response.

Comparison of the sensitivity of four rapid assays for the detection of antibodies to HIV-1/HIV-2 during seroconversion.

OBJECTIVES: To compare the sensitivity of four rapid assays for the detection of antibodies against HIV-1 during early seroconversion. METHODS: Four rapid assays for the detection of antibodies to HIV-1/HIV-2 (SUDS HIV 1 + 2, TestPack HIV-1/HIV-2, HIV-SPOT and CAPILLUS HIV-1/HIV-2) were evaluated on 38 sera derived from 14 HIV-1 seroconverters, formerly tested with Abbott second- and third-generation HIV-1/HIV-2 enzyme immunoassay (EIA) and Diagnostic Biotechnology HIV blot 2.2. EIA-negative sera had also been investigated by HIV-1 antigen testing and polymerase chain reaction (PCR) analysis detecting HIV-1 proviral DNA. RESULTS: On 16 sera which were Abbott second-generation EIA negative, the SUDS assay rated highest with six positive results. Four sera were TestPack positive. No specimens in this group were HIV-SPOT positive or CAPILLUS positive. Of the remaining 22 sera (Abbott second- and third-generation EIA positive), all but two were positive in all tests. CONCLUSIONS: The study revealed differences in the sensitivities of the rapid assays in the early phase of seroconversion. The most sensitive assay, SUDS, was even more sensitive than the Abbott third-generation EIA, while TestPack and the EIA were equally sensitive. Based on these findings, we suggest the inclusion of these assays in the supplemental testing for detection of antibodies to HIV.

Co-infection with HIV-1/HTLV-II and the risk of progression to AIDS and death. The Oslo HIV Cohort Study Group.

Human T-cell lymphotropic virus (HTLV) types I and II were the first discovered human retroviruses. While HTLV-I has been clearly associated with disease, the health implications of HTLV-II infection are still unsettled. A prospective epidemiological study of 409 HIV-infected subjects of different transmission categories was performed to study the presence of HTLV-II antibodies, and whether HTLV-II antibodies are associated with the progression to AIDS and to death of any cause. Of 409 subjects, 30 (7.3%) were HTLV-II positive at study entry; 2 subjects seroconverted during follow-up. In the HTLV-II-positive group 2 were heterosexually HIV infected, 28 (of whom 2 were seroconverters) were IDUs and 2 were homosexual men. When controlling for transmission category, gender, age and CD4+ lymphocyte count at study entry, the relative risk of AIDS progression for the HTLV-II-positive group was 2.1 (0.8-5.1, 95% confidence interval (CI)) as compared to the HTLV-II-negative group. The adjusted relative risk of dying was 2.1 (1.0-4.3, 95% CI). When studying IDUs separately, the adjusted relative risk of AIDS progression was 2.3 (0.8-6.9, 95% CI) and the relative risk of dying was 2.0 (0.9-4.6, 95% CI). The results of this study suggest that HTLV-II is a cofactor in HIV disease progression. The number of HTLV-II-infected subjects, was, however, small, and insufficient control of confounding factors must be taken into consideration.

Maternal human immunodeficiency virus infection and pregnancy outcome.

OBJECTIVE: A longitudinal study to determine the natural history of HIV-1 infection in pregnancy, infancy and early childhood was carried out in Ndola, Zambia. DESIGN: Prospective study. SETTING: Kabushi and Chifubu clinics. SUBJECTS: A total of 965 women attending antenatal care were screened for anti-HIV antibodies using the Welcozyme test. All reactive sera were confirmed by Western Blot. One hundred and fifty seropositive pregnant women (cases) with their age and parity matched pregnant control (seronegative) were recruited into the study. They were followed up through delivery. MAIN OUTCOME MEASURE: personal characteristics, socio-economic and other risk factors. RESULTS: The prevalence of anti HIV-1 antibodies among the 965 women was 15.5pc. Results of baseline data between the two groups of women indicate significant differences (p < 0.05) in the following variables; marital status, outcome of last pregnancy, whether last child is still alive, history of herpes zoster, lymphadenopathy, dermatitis, oral thrush and mean haemoglobin level. There were no differences in the incidence of abortions, stillbirths and neonatal deaths. However, the mean birth weight of babies born out of seropositive women was significantly lower than babies of seronegative women. CONCLUSION: It is concluded that HIV-1 infection in pregnancy is associated with low birth weight.

PIP: Data on 130 HIV-1 infected pregnant women were compared with data on 150 HIV-1 negative pregnant women to determine the effect of HIV-1 infection on pregnancy outcomes. All the women were recruited while seeking prenatal services at Chifubu and Kabushi clinics in Ndola, Zambia, during 1991-1993. None of the HIV-1 infected women had AIDS. The HIV- 1 prevalence rate for the recruited pregnant women was 15.5%. The socioeconomic characteristics of the women in both suburbs were similar. Yet, pregnant women at Chifubu were more likely to be HIV-1 positive than those at Kabushi (p 0.001). The proximity to the border with Zaire and the higher inward and outward migration rates in Chifubu may contribute to the higher HIV-1 prevalence rate in Chifubu. HIV-1 infected women were more likely than controls to have a history of Herpes zoster, cervical lymphadenopathy, axillary lymphadenopathy, skin rash, and oral thrush (p 0.05). They were less likely than controls to be married, to have the outcome of their last birth be a live birth, and to have their last child still be alive (p = 0.01). HIV-1 pregnant women had a lower hemoglobin level and smaller newborns than controls (10.3 vs. 10.9 g % and 2.76 vs. 3.03 kg, respectively; p 0.03). When the researchers controlled for gestation, there was no difference in mean birth weights between the groups. Both groups had similar perinatal mortality outcomes (1 stillbirth each and 2 neonatal deaths each). The most significant finding is that HIV-1 infection in pregnancy contributes to low birth weight.

[Quality adjusted life years in assessment of preventive measures. Should blood donors be tested for HTLV-I/II infections?]. / Kvalitetsjusterte leveår i vurderingen av forebyggende helsearbeid. Bør blodgivere testes for HTLV-I/II-infeksjon?

In planning preventive health measures, quality adjusted life-years (QALYs) are useful as a measure of benefit. As an example, the question of whether blood donors should be routinely tested for antibodies to the Human T-lymphotropic viruses I and II (HTLV I/II) is analysed. A mathematical model was set up to describe the consequences, in terms of lost life-years and years with disease due to transfusion-mediated infection (if testing is not performed) or years with reduced quality of life (in the case of testing). These future outcomes were discounted and converted to QALYs. The cost per QALY is about NOK 2.33 million when the prevalence is 1 per 50,000 blood donors, and is reduced to 190,000 per QALY when the prevalence is 10 per 50,000. Using QALYs in evaluation of preventive medicine can be complicated, and calls for cooperation between epidemiologists and health economists.

[Human T-cell lymphotrophic virus type I and T-cell lymphoma]. / Humant T-cellelymfotropt virus type I og T-cellelymfom.

Human T-cell lymphotropic virus type I is an oncogenic retrovirus, endemic in Southwestern Japan, the Caribbean, some parts of Africa and Central and South America. The virus is etiologically associated with adult T-cell leukemia/lymphoma and a myelopathy called tropical spastic paraparesis or HTLV-I associated myelopathy. Transmission of the virus is almost identical to that of HIV. The latency period before onset of clinical symptoms can last from a few years (tropical spastic paraparesis) up to several decades (adult T-cell leukemia/lymphoma). Four different clinicopathological subtypes of the T-cell neoplasia are known, and in this article we describe two patients with the subtype lymphoma.

Phylogenetic relationship and geographic distribution of multiple human T-cell lymphotropic virus type II subtypes.

The current env-based subtyping of human T-cell lymphotropic virus type II (HTLV-II) identifies only two heterogenetic groups, HTLV-IIa and HTLV-IIb. To better understand the genetic diversity and phylogeny of HTLV-II, we examined the most divergent genomic region of HTLV-II, the long terminal repeat, by using restriction fragment length polymorphism (RFLP) and sequence analysis. Long terminal repeat sequences were amplified from peripheral blood mononuclear cells by PCR and digested with seven restriction endonucleases that differentiated HTLV-II into five HTLV-IIa (IIa0 to IIa4) and six HTLV-IIb (IIb0 to IIb5) restriction types, with HTLV-IIa0 and HTLV-IIb0 being prototypes for the MoT and NRA isolates, respectively. We examined 169 HTLV-II-infected samples, including 123 from blood donors and intravenous drug users (IDU) from the Americas, 16 from IDU from Europe, and 30 from Amerindians. Of the 169 samples, 109 (64.5%) were categorized as HTLV-IIa and 60 (35.5%) were categorized as HTLV-IIb. The predominant restriction types seen among the U.S. blood donors and U.S. IDU were IIa0 (68.7%) and IIb4 (10.4%). Four Spanish and seven Italian samples were IIb4, while five Norwegian samples were IIa2. Twelve Guaymi and all ten Seminole samples were single restriction types (IIb1 and IIb5, respectively), whereas the two Navajo and six Pueblo samples had a mixture of restriction types IIa0, IIa4, and IIb5. Of the HTLV-IIb restriction types observed in the U.S. non-Indians, 42.8% appear to have originated from the North Amerindian (IIb5), while 57.2% were similar to the European IIb4 restriction type. Sequences of 15 selected HTLV-II samples were determined and phylogenetically compared with 7 previously published HTLV-II LTR sequences. The derived topologies revealed three HTLV-IIa phylogroups (A-I to A-III) and four HTLV-IIb phylogroups (B-I to B-IV). Furthermore, the HTLV-IIa phylogroups appear to have evolved from the HTLV-IIb phylogroups. In the HTLV-IIa cluster, a Navajo (A-I) and a Brazilian (A-II) sequence formed separate phylogroups, while the remaining IIa sequences formed a single phylogroup (A-III). The four HTLV-IIb phylogroups were represented predominantly by a New York IDU (B-I), European IDU (B-II), North Amerindian and NRA (B-III), and Central Guaymi Indian (B-IV) sequence(s). Comparison of the phylogenetic data with the RFLP results revealed that results of the two methods correlated completely, demonstrating the ability of the RFLP method to predict the phylogroup of HTLV-II-infected samples accurately and quickly. GENBANK/U10258

Hepatitis C infection in patients with primary hypogammaglobulinemia after treatment with contaminated immune globulin.

BACKGROUND: In Scandinavia many patients with primary hypogammaglobulinemia contracted non-A, non-B hepatitis after intravenous treatment with an immune globulin product that was later found to contain a non-A, non-B hepatitis virus. METHODS: We studied the prevalence and clinical course of hepatitis C virus (HCV) infection in a group of 55 Norwegian patients with primary hypogammaglobulinemia and investigated its association with the use of contaminated immune globulin. We used the polymerase chain reaction to detect HCV RNA and performed HCV genotyping. We also analyzed the responses to treatment with interferon. RESULTS: Of 20 patients who received the contaminated immune globulin, 17 were seropositive for HCV RNA: In addition, 1 of 35 patients not exposed to the contaminated immune globulin was HCV RNA--positive. HCV genotype V was found in all 12 patients for whom genotyping was performed, but 8 patients also had genotype II or III, or both. All HCV RNA--positive patients had abnormal results on biochemical liver tests. All liver-biopsy specimens (from 15 patients) were abnormal, with portal inflammation, bile-duct damage, and focal necrosis. In six patients there was cirrhosis. Two patients died of liver failure. In 4 of the 10 patients treated with interferon there were complete, though transient, biochemical responses, but the follow-up biopsy specimens showed evidence of histologic progression. The poorest responses to interferon were among the patients with multiple HCV genotypes. All but one patient remained positive for HCV RNA: CONCLUSIONS: In patients with primary hypogammaglobulinemia there was a high rate of HCV infection after treatment with contaminated immune globulin. In these immunocompromised patients HCV infection has a severe and rapidly progressive course, and responses to interferon are poor.

[Risk of AIDS in the Oslo HIV-cohort study. A comparison between homosexual men, intravenous addicts and heterosexual persons]. / Risiko for AIDS i Oslo HIV-kohortstudien. En sammenlikning av homoseksuelle menn, intravenøse stoffmisbrukere og heteroseksuelle personer.

In order to study differences in risk of the development of AIDS in different groups of HIV infected subjects, 151 homosexual men, 110 intravenous drug users (IVDUs) and 36 heterosexually infected persons without major signs of HIV infection at entry to the study were enrolled in a cohort study. The mean follow-up time was 35 months. At the end of follow-up 40 subjects (13%) were diagnosed as having AIDS. This represented 20% (31/151) of the homosexual men, 7% (8/110) of the IVDUs and 3% (1/36) of the heterosexual subjects. The probability of being AIDS-free 36 months after entering the study was 0.88 (0.84-0.92, 95% CI) for the total study population, 0.83 (0.77-0.90) for the homosexual men, 0.92 (0.86-0.99) for IVDUs and 0.93 (0.91-1.0) for heterosexual subjects (p < 0.05, log rank test). In a Cox regression analysis, adjusting for CD4+ cell count at study entry, the relative risk of AIDS progression was 2.4 (1.1-5.2) for homosexual men and 0.3 (0.04-2.4) for heterosexual subjects, compared with IVDUs. The results demonstrate a higher risk of AIDS for homosexual men during the follow-up period.