Lower prevalence and incidence of HIV-1 syncytium-inducing phenotype among injecting drug users compared with homosexual men.

OBJECTIVE: To study the prevalence, incidence and predictive value for progression to AIDS of the HIV-1 syncytium-inducing (SI) phenotype in HIV-infected injecting drug users (IDU) compared with HIV-infected homosexual men. DESIGN: Two prospective cohort studies on HIV-1 infection among IDU and homosexual men. METHODS: HIV-infected IDU (n = 225) and homosexual men (n = 366) without AIDS were studied from March 1989 through December 1993. Data on laboratory markers, including the presence of SI variants, demographics, behavioural characteristics and clinical events were collected at every visit. RESULTS: At baseline, SI variants were detected in 4% of IDU and 17% of homosexual men. During the study period 18 IDU and 68 homosexual men switched from non-SI to SI phenotype (4-year cumulative incidence, 14.6 and 28.4%, respectively) before AIDS diagnosis. Among participants with a documented date of HIV infection the cumulative incidence of SI was lower among IDU than homosexual men (4-year cumulative incidence, 6.2 and 20.7%, respectively). At AIDS diagnosis, 21% of all AIDS cases among IDU had the SI phenotype compared with 54% among homosexual men. In both risk groups an accelerated CD4 decline was found after the non-SI-to-SI switch. The SI phenotype appeared to be a predictor of AIDS (multivariate relative hazard, 5.33), independent of CD4 cell count and p24 antigen at baseline. In the multivariate time-dependent analysis, the relative hazard of SI phenotype decreased considerably, which is consistent with the hypothesis that the effect of SI phenotype on progression to AIDS is mediated by CD4 cell count. CONCLUSION: The SI phenotype is associated with accelerated CD4 decline and progression to AIDS in both risk groups. The remarkable lower prevalence and incidence of the SI phenotype among IDU may implicate a difference in pathogenesis and natural history of HIV infection linked to transmission group.

Death from suicide and overdose among drug injectors after disclosure of first HIV test result.

OBJECTIVE: To investigate whether notifying injecting drug users (IDU) of their positive HIV serostatus contributes to suicide and overdose mortality risk. DESIGN: Members of a cohort of IDU, recruited since December 1985, who first learned their HIV serostatus after study entry but before December 1992, were studied for incidence of deaths due to suicide and overdose. METHODS: Incidence of mortality due to suicide/overdose was calculated from date of first HIV test result disclosure until the end of follow-up or diagnosis of AIDS. RESULTS: Eighty-six HIV-positive and 252 HIV-negative IDU were included with median follow-up times of 4.3 and 4.0 years, respectively. Seven suicides and 10 deaths from overdose were recorded. High suicide/overdose risk shortly after test result notification was not found among HIV-positive IDU. Only one out of eight HIV-positive IDU who died of suicide/overdose died within 6 months of first disclosure. The overall suicide/overdose mortality rate was higher for HIV-positive than HIV-negative IDU; the rate ratio being 2.46 [95% confidence interval (CI), 0.95-6.39] or 2.04 (95% CI, 0.77-5.39) after control for confounders. CONCLUSIONS: Notifying IDU of their positive HIV serostatus does not appear to lead to a sudden and substantial rise in suicide/overdose deaths. Although death from suicide/overdose is more common among HIV-positive than HIV-negative IDU, this difference is likely to result from factors other than test result disclosure. Therefore, provided that appropriate counselling is offered, we see no reason to discourage voluntary HIV test result notification for fear of inducing suicide in HIV-infected IDU.

Incidence and risk factors for acquisition of sexually transmitted diseases in heterosexuals with multiple partners.

BACKGROUND AND OBJECTIVES: The incidence of four STDs that are mainly transmitted during vaginal intercourse, and risk factors for acquisition during follow-up were studied in a cohort of heterosexuals with multiple partners in Amsterdam, the Netherlands. These heterosexuals with > or = 5 different sexual partners in the preceding 6 months were recruited from an STD clinic and participated voluntarily in a longitudinal HIV study between 1987 and 1991. STUDY DESIGN: Using survival techniques, the authors analyzed data on general characteristics, sexual behavior, reported and diagnosed STDs of 155 men and 234 women participating at least twice in the HIV study and returning regularly. RESULTS: The majority of women were prostitutes and the majority of men were clients of prostitutes. For men and women, the 2-year cumulative incidence were 2.7% (95% confidence interval [95% CI] 0.9-8.3) and 0.8% (95% CI 0.1-5.8), respectively, for syphilis, 8.5% (95% CI 3.9-18.0) and 5.3% (95% CI 2.6-10.8) for gonorrhea, 20.2% (95% CI 12.5-31.9) and 31.9% (95% CI 23.7-42.2) for Chlamydia trachomatis infection, and 24.5% (95% CI 17.3-34.2) for trichomoniasis (examination in women only). Among men, STD(s) diagnosed at entry was found to be an independent predictor of infection with any of these STDs during follow-up (relative hazard (RH) = 2.58, 95% CI 1.18-5.64). Among women, risk factors were age < or = 25 years (RH = 1.82, 95% CI 1.13-2.19) and current report of commercial unprotected vaginal contacts (RH = 2.19, 95% CI 1.11-4.13). Among women, trichomoniasis during follow-up was predicted by current report of commercial contacts (RH = 2.59, 95% CI 1.16-5.79). C. trachomatis infection in women was associated with native country outside the Netherlands (RH = 2.47, 95% CI 1.38-4.43), reported history of STDs in the 4 months before the study (RH = 2.92, 95% CI 1.39-6.14) and age < or = 25 years (RH = 1.90, 95% CI 1.06-3.40). CONCLUSIONS: This study shows that STDs were common among this group of heterosexuals, despite regularly participating in the HIV study in which consistent condom use was promoted. Prevention efforts should be intensified for men and women having commercial sexual contacts, young women, and individuals with a history of STDs.

The HIV prevalence among pregnant women in the Amsterdam region (1988-1991).

The objective of the study was to monitor the HIV prevalence in the years 1988-1991 among pregnant women in the Amsterdam region, visitors to an abortion clinic and 3 outpatient infertility clinics. All women attending these clinics were asked to participate in the study on a voluntary basis and were tested with informed consent. The women were questioned about risk-bearing behaviour of themselves and their sexual partner(s). In the period 1988-1991, of the 23,827 eligible pregnant women, 22,165 women participated (93.0%). Twenty-seven women were found to be positive for HIV antibodies (0.12%, 95% CI: 0.08%-0.17%), of whom twenty belonged to a known HIV risk group or had a partner who belonged to one of these groups and 7 women had no known HIV risk. Seventeen of the 27 women had a foreign nationality. The annual HIV prevalence among pregnant women was: 1988: 0.28%; 1989: 0.10%; 1990: 0.10%; 1991: 0.11%. In the years 1990 and 1991, of the 1,128 eligible women visiting the abortion clinic 953 (84.5%) were tested. Eleven women were HIV-seropositive (1.15%, 95% CI: 0.6%-2.0%), of whom 9 were from an AIDS endemic region, 1 woman had a partner from this region and 1 woman had no known HIV risk. Four African women had HIV-2 antibodies. At the 3 outpatient infertility clinics 1 woman was found to be HIV-positive (0.13%; 95% CI: 0.02-0.9). She had no other risk than a partner from an AIDS endemic area. In the Amsterdam region there was a steady and low HIV prevalence (0.1%) among pregnant women through the years 1988-1991. The prevalence in the abortion clinic was ten times higher. The program was able to detect possible high risk groups within the population. Migration and travelling can play an important role in the spread of HIV in the general heterosexual population.

High mortality among HIV-infected injecting drug users without AIDS diagnosis: implications for HIV infection epidemic modellers?

OBJECTIVE: To quantify mortality without AIDS diagnosis (pre-AIDS death) among HIV-infected injecting drug users (IDU) and to compare it with that among homosexual men. SETTING: Municipal Health Service in Amsterdam, The Netherlands. DESIGN: HIV-infected participants of a cohort of homosexual men (recruited since 1984) and a cohort of IDU (recruited since 1985) were followed up to investigate the occurrence of AIDS and pre-AIDS death up to 1 October 1992. Both seroprevalent and seroconverted participants were included. METHODS: Product-limit estimates of cumulative AIDS incidence and cumulative pre-AIDS mortality were calculated for a competing risks model. RESULTS: Of 455 HIV-infected homosexual men, 127 developed AIDS and two died without a diagnosis of AIDS during follow-up; 31 out of 279 HIV-infected IDU were diagnosed with AIDS and 33 died without an AIDS diagnosis. After 6.5 years of follow-up an estimated 43.7% of the homosexual men had been diagnosed with AIDS and 0.7% had died without an AIDS diagnosis, while 32.7% of the IDU had been diagnosed with AIDS and 19.8% had died without an AIDS diagnosis. CONCLUSIONS: After 6.5 years of follow-up one-fifth of a group of HIV-infected IDU had died without being diagnosed with AIDS. This phenomenon seriously limits the number of HIV-infected IDU that may ever be diagnosed with AIDS and should be considered by HIV infection epidemic modellers. The new case definitions for AIDS will limit the number of recorded cases of pre-AIDS death.

Trends in sexual behaviour and the incidence of sexually transmitted diseases and HIV among drug-using prostitutes, Amsterdam 1986-1992.

OBJECTIVE: To determine trends in commercial sexual risk behaviour among drug-using prostitutes in Amsterdam after local and national prevention campaigns began in 1987. DESIGN: A subgroup of 281 drug-using prostitutes was selected from a comprehensive, open cohort study of drug users in Amsterdam. METHODS: Cross-sectional (at intake) and within-person trends in sexual behaviour were evaluated. Person-time analysis was used to determine trends in incidence of sexually transmitted disease (STD). RESULTS: From 1986 to 1992, 'always' use of condoms in the 6 months preceding intake increased from 21 to 58% (P < 0.001) and the number of commercial contacts declined from 84 to 64 per month (P = 0.06). Consequently, the mean number of unprotected commercial contacts declined from 26 to nine per month (P < 0.001). Temporal trends within individual prostitutes from visits 1 to 8 (mean interval between visits, 4 months) showed a somewhat larger risk reduction than the cross-sectional trends. From 1989 to 1992, the STD incidence declined from 61 to 40 episodes per 100 person-years (P = 0.06). Prostitutes residing in Amsterdam for shorter time periods reported more STD and a relatively small core group reported the majority of STD episodes. At intake, when < 10% of all participants knew their HIV serostatus, HIV-positive prostitutes reported significantly lower levels of condom use and more STD than HIV-negatives, while, after learning their serostatus, levels of condom use and the STD-incidence were comparable. Prevalent and incident HIV infections were not observed among non-injecting prostitutes. CONCLUSIONS: Drug-using prostitutes have reduced sexual risk behaviour. There is evidence for the effectiveness of HIV testing and counseling. At present, new infections among clients of drug-using prostitutes are expected to occur, although not on a large scale. Since the incidence of STD continues to be high, it is important to monitor the spread of HIV among heterosexuals, including prostitutes and their clients, closely.

Progression of HIV infection among injecting drug users: indications for a lower rate of progression among those who have frequently borrowed injecting equipment.

OBJECTIVE: To study markers of progression in a cohort of HIV-infected intravenous drug users (IDU). DESIGN: A prospective epidemiologic study. SETTING AND PATIENTS: We studied progression of HIV infection among 126 IDU attending the Municipal Health Service in Amsterdam. MAIN OUTCOME MEASURES: Progression was defined as a decline of the CD4 cell count to < 200 x 10(6)/l on two consecutive follow-up visits or AIDS. RESULTS: Using Cox modelling, the following baseline variables were predictive of progression. Enhanced progression was associated with: age > 30 years [relative hazard (RH), 7.7 [95% confidence intervals (CI), 1.7-36.0]], core antibody negativity [RH, 5.3 (95% CI, 1.6-17.6)], CD4 cell count [for CD4 cells 350-500 x 10(6)/l, RH, 1.38 (95% CI, 0.37-5.16); for CD4 cells 200-350 x 10(6)/l, RH, 9.20 (95% CI, 2.73-31.05) compared with a CD4 count > 500 x 10(6)/l]. A lower rate of progression was associated with borrowing used injecting equipment. IDU who reported borrowing injecting equipment between 1980 and baseline 10-99 times or > 99 times had a RH of 0.44 (95% CI, 0.22-0.88) and 0.19 (95% CI, 0.03-0.37), respectively, compared with IDU who had borrowed < 10 times. p24 antigen positivity was more predictive than core antibody negativity in a model with time-dependent variables, the relative risk for p24 antigen-positive participants was 3.5 (95% CI, 1.3-9.3). Additional analysis of progression to AIDS in a larger group of IDU showed comparable results with regard to the effect of borrowing on progression. CONCLUSIONS: Our observation that those IDU who reported borrowing injecting equipment most frequently appeared to have the lowest rate of progression, corrected for some sources of potential confounding, requires further epidemiologic confirmation and extended laboratory studies since other sources of bias might have been present. Baseline CD4 count, age and core antibody or p24 antigen were predictive of progression in IDU. We wish to emphasize that our results do not imply that borrowing should be encouraged, but may have implications for our understanding of HIV pathogenesis.

Increasing morbidity without rise in non-AIDS mortality among HIV-infected intravenous drug users in Amsterdam.

OBJECTIVE: We determined pneumonia, endocarditis, hospitalizations and death rates in a group of HIV-infected and non-infected IVDU recruited at low-threshold methadone programmes in Amsterdam, The Netherlands to examine the influence of HIV infection on morbidity and non-AIDS mortality in intravenous drug users (IVDU) without AIDS. DESIGN: A prospective epidemiologic study among HIV-infected and non-infected IVDU in Amsterdam. SETTING: We analysed patients attending the Municipal Health Service of Amsterdam. PATIENTS: The patient group consisted of 197 HIV-seropositive IVDU and 193 HIV-seronegative IVDU as controls. MAIN OUTCOME MEASURES: We analysed incidence rates per person-year of bacterial pneumonia; hospitalizations and non-AIDS mortality; and relative risks for HIV-infected IVDU compared to non-infected IVDU. RESULTS: The incidence of bacterial pneumonia rose from 0.1 in 1986 to 0.29 in 1989 in HIV-infected IVDU. The overall relative risk for this group was 4.0 (95% confidence interval, 1.7-9.7) compared with non-infected controls. HIV-infected IVDU were more frequently hospitalized for bacterial pneumonia than non-infected IVDU (50 versus 13%). In contrast, non-AIDS mortality rates remained stable for both HIV-infected and non-infected IVDU. HIV seropositivity was not found to be an independent risk factor for non-AIDS mortality. CONCLUSION: We found a high and rising incidence of bacterial pneumonia among HIV-infected IVDU in Amsterdam without any consequential rise in non-AIDS mortality. This contrasts with reports from studies conducted in New York City, New York, USA, where bacterial pneumonia-related mortality has been found to increase markedly, coincident with the AIDS epidemic. Early detection of bacterial pneumonia and easy access to both inpatient and outpatient medical care may be important factors in preventing early death due to common bacterial pathogens in IVDU without AIDS.