Nicastri, Emanuelea; Angeletti, Claudioa; Palmisano, Luciab; Sarmati, Loredanac; Chiesi, Antoniob; Geraci, Andreab; Andreoni, Massimoc; Vella, Stefanob; the Italian Antiretroviral Treatment Group From the aNational Institute for Infectious Diseases, IRCCS L. Spallanzani, Rome bDepartment of Drug Research and Evaluation, Istituto Superiore di Sanità, Rome cDepartment of Public Health, University of Rome Tor Vergata, Italy.

The use of highly active antiretroviral therapy (HAART) has been associated with a decrease in morbidity and mortality in HIV-infected individuals [1-4]. Previous investigations found different rates of HIV disease progression and of virological and imunological response to antiretroviral therapy among HIV-infected women compared with men [5-11]. Particularly in early studies [7-9], clinical progression appeared to be more rapid among women than among demographically comparable men. This was attributed to the fact that women were less likely to be started on antiretroviral drugs or with prophylaxis for Pneumocystis carinii pneumonia than men with similar clinical profiles. The clinical course in HIV-infected women and men also differs because of frequent gynaecologic morbidity [10,11] and pregnancy [12], pretreatment anaemia [13] and older age at the time of HIV acquisition [14]. Many HIV-infected women are from underprivileged populations for whom discrimination, violence and other stressful conditions are common [6]. These women may be at particularly high risk of adverse health outcomes from HIV infection. Nevertheless, plasma HIV RNA levels are significantly lower in adult women than in men after adjustment for CD4 cell count, race and drug use within the prior 6 months [15,16]. Moreover, a few recent studies have shown no significant gender differences in timing of initiation of antiretroviral therapy; in clinical progression to nervous system diseases, AIDS and death; and in hospitalization and re-admission rates [16-23].

The present study assesses the clinical, virological and immunological differences by gender in a multicentre cohort study of HIV-infected Italian individuals during HAART who were enrolled between 1996 and 2002.

Objective: To assess gender differences in the long-term clinical, virological and immunological outcomes during highly active antiretroviral therapy (HAART).

Methods: This longitudinal observational multicentre study followed 2460 HIV-infected patients who had begun a protease inhibitor-based regimen for a median period of 43 months. Outcome measures were virological suppression (< 500 copies/ml), confirmed virological rebound after suppression, and death or new AIDS-defining illness (ADI).

Results: At baseline, 690 female patients (28.0%) had significantly lower age, higher prevalence of heterosexual contact and lower prevalence of intravenous drug use as risk factors for HIV infection compared with males. Furthermore, females had a lower number of AIDS-defining illnesses, higher CD4 cell counts and lower viral loads.

No gender differences were reported in terms of proportion of patients achieving viral suppression or exhibiting rebound after achieving viral suppression. Female patients experienced reduced clinical progression during follow-up compared with males (P = 0.008) by Kaplan-Meier analysis; however this difference was not significant in an adjusted analysis.

In a multivariate model, the interaction between gender and risk factor for HIV or viral load showed that female drug users and female patients with a baseline HIV RNA viral load of 104-105 copies/ml had a favourable clinical outcome compared with males (P = 0.035 and P = 0.015, respectively).

Conclusion: No differences were found between genders in terms of virological and immunological outcomes during long-term HAART. Nevertheless, a lower risk of clinical progression was reported among female patients with intermediate baseline viral load than in males.

Discussion TOP

Among PI- and NNRTI-naive patients treated with HAART, there were no differences by gender in terms of proportion of patients achieving viral suppression or exhibiting a confirmed viral rebound. However, during the 282-week study period, the Kaplan-Meyer estimates showed a lower number of new ADI or deaths in female patients treated with HAART than in males. Nevertheless, this difference was not significant at multivariate analysis. In the multivariate model considering the interaction between gender and risk factor for HIV infection or viral load, female drug users and female patients with a baseline viral load 104-105 copies/ml had a favourable clinical outcome, experiencing a lower risk of developing new ADI or death compared with males (P = 0.035 and P = 0.015, respectively).

Recently, in the HAART era, a clinical gender benefit has been described among female individuals with HIV/AIDS treated with antiretroviral regimens [16,24-26] whereas other studies have found no gender differences [22,27-29]. This better course of infection progression among female individuals has been linked to the greater frequency of clinic assessments in women than in men [25], and the higher risk of cardiovascular and cerebrovascular events among male patients during HAART [30]. Finally, HIV-infected women were reported to experience a fourfold lower rate of disease progression than men during zidovudine monotherapy, even if more severe toxicities occurred in women than in men during therapy with nucleoside reverse transcriptase inhibitors, including approximately a fourfold grater fat misdistribution and a disproportionately higher lactic acidosis [31-33]. Furthermore, the saquinavir concentration among HIV-infected patients treated with HAART including a saquinavir-boosted PI was significantly higher in females than in males and was associated with a stronger virological response in the female patients [34]. HIV-infected women had 2.3-fold higher zidovudine trisphosphate concentrations, 1.6-fold higher lamivudine trisphosphate concentrations and a faster virological suppression during HAART compared with males [35]. In our study, no gender difference was found with respect to time to virological suppression: this analysis was adjusted for confounding variables such as risk factors for HIV infection, previous ADI, baseline CD4 cell count and viral load, but not for adherence to drugs, adverse events and changes of antiretroviral therapy.

The multivariate analysis by age of the patients reaching plasma virological suppression at month 12 supports the conclusion that older HIV-1 patients have a better control of HIV replication. Few hypotheses to explain the relationship between older age and low plasma viral load during HAART are put forward in the specific literature on age [36,37]. Elderly individuals generally have higher economical income, higher antiretroviral drug adherence and a more stable social status than younger individuals. Specific studies need to be designed that would examine these likely determinant factors.

Our results might have been influenced by a few biases and confounding factors. First, our definition of virological response considered viral suppression as reduction to < 500 copies/ml, and this may have led us to underestimate the prognostic benefit of the virological response. Second, there was no analysis of the effect of switching antiretroviral therapies (i.e., from PI to NNRTI or to double PI) or of the drug exposure over the follow-up period. Third, the lack of data on body mass index, socio-professional status, behaviour, adherence and coinfection with hepatitis B or C virus are clear limitations in our cohort. Nevertheless, in the Italian National Health System, access to the diagnosis, care and treatment of individuals with HIV/AIDS is free of charge and so economic income is unlikely to represent a limit to access for medical care. Furthermore, data for adherence in the ICONA Italian cohort indicated no difference by gender in all patients and, particularly, in intravenous drug users [38]. Finally, lack of cause-specific mortality information is clearly a limitation in our analysis. Future analyses of clinical progression of HIV disease should distinguish among HIV-related events, treatment-related events and other events.

The Italian Antiretroviral Treatment Group will continue to monitor the implications of gender differences in baseline parameters and in the clinical progression of HIV disease during HAART, updating analyses at regular intervals. The favourable trend in long-term clinical outcome among female patients treated with HAART represent a very interesting issue and need further innovative research on the likely determinant factors. Drug interactions, resistant profiles, pharmacokinetics and therapeutic drug monitoring should be investigated in clinical studies to verify the likely different distribution of antiretroviral inhibitory quotients, adverse events and long-term adherence by gender.

The Italian Antiretroviral Treatment Group (IATG) is a national project that began in 1996. A total of 53 hospitals across Italy currently provide epidemiological data on HIV infection. A trained research assistant uses standardized procedures to collect prospectively clinical, laboratory and treatment data from medical records using specialized software provided by the Istituto Superiore di Sanità. Written informed consent was given by all participants in the project. The 2460 adult patients (³ 18 years of age) had not previously taken a protease inhibitor (PI) drug or a non-nucleoside reverse transcriptase inhibitor (NNRTI) and began a PI-based therapy between July 1996 and December 1998. The group was studied for a median follow-up period of 43 months; 245 (10.0%) were lost at follow-up. Plasma HIV RNA viral loads and CD4 T cell counts were recorded for all patients 0-3 months before starting therapy (referred to as baseline measurements). HIV viral load was measured at a local level using the HIV-1 reverse transcriptase polymerase chain reaction assay (Amplicor HIV-1 Monitor, Roche Diagnostic, Branchburg, New Jersey, USA), the in vitro signal amplification nucleic acid probe assay [VERSANT HIV-1 RNA 3.0 Assay (bDNA); Bayer Corp., Berkeley, California, USA], or the nucleic acid sequence-based amplification method (NASBA, Organon Teknika; Organon Teknika, Boxtel, The Netherlands).

Inclusion criteria for the study required a measurement of HIV-1 RNA viral load at baseline > 500 copies/ml and at least one test for HIV RNA viral load measurement in the first 12 months after starting the therapy.

AIDS-defining illnesses (ADI) were classified according to the 1993 revised clinical definition of the Centers for Disease Control and Prevention. In each centre, trained research assistants reviewed clinical reports to ascertain death and clinical events. The clinicians who determined the occurrences of clinical progression in the medical records were aware of the patients' immunological and virological status.

At baseline, the 690 female patients (28.0% of the 2460 enrolled patients) had significantly lower age, higher prevalence of heterosexual contact and lower prevalence of intravenous drug use as risk factors for HIV infection than male subjects. They were also reported to have higher previous exposure to antiretroviral therapy, higher CD4 cell counts and lower viral loads than male individuals.

Kaplan-Meier estimates showed no significant gender difference in terms of proportion of patients achieving viral suppression within 12 months after starting HAART (log rank test, P = 0.83). At month 12, 1629 patients (66.2%) reached an undetectable level of plasma viraemia (< 500 copies/ml). The multivariate model studied the effect of gender, adjusted for demographic and baseline clinical and virological parameters, on the proportion of patients achieving viral suppression within the first 12 months following initiation of HAART. The model failed to show any significant differences between gender and viral suppression [hazard ratio (HR) 1.04; 95% confidence interval (CI), 0.92-1.18; P = 0.53]; whereas, age per 10 years' increase (HR, 1.07; 95% CI, 1.01-1.14; P = 0.03), CD4 cell count > 350 × 106 cells/l (HR, 1.16; 95% CI, 1.02-1.32; P = 0.03), HIV RNA viral loads < 104 copies/ml (HR, 1.52; 95% CI, 1.32-1.74; P < 0.001) and HIV RNA viral loads 104-105 copies/ml (HR, 1.26; 95% CI, 1.12-1.41; P < 0.001) were independently associated with viral suppression.

During the follow-up, no significant gender differences were found in terms of proportion of patients exhibiting a confirmed viral rebound after achieving viral suppression. Overall, half of the patients had a viral rebound at the end of the follow-up period. The multivariate analysis showed that age per 10 years' increase (HR, 0.84; 95% CI, 0.75-0.94; P = 0.003) and HIV RNA viral load < 104 copies/ml (HR, 0.66; 95% CI, 0.53-0.83; P < 0.001) were protective parameters inversely associated to the viral rebound.

After 4 years of HAART, patients showed a progressive immune recovery, with an increase in median CD4 cell count of > 250 × 106 cells/l compared with baseline values. Female subjects maintained a higher CD4 cell count compared with male individuals during the entire follow-up period. Immune recovery was observed in patients either achieving or not achieving viral suppression within 12 months from the start of HAART (data not shown).

During the 282 weeks of follow-up, a total of 294 clinical events, including 194 new ADI and 100 deaths, occurred, with an incidence ratio of 3.9/100 person-years. The new ADI included, among others, the main events of oesophageal candidiasis (15.4%), Pneumocystis jirovecii pneumonia (12.5%), pulmonary tuberculosis (7.7%), Mycobacterium avium complex disseminated disease (7.7%). At Kaplan-Meier estimates, a significant gender difference was found in terms of proportion of patients experiencing a new ADI or death during the follow-up period (P = 0.008). Particularly, female patients experienced a reduced rate of clinical progression whether or not they achieved virological suppression within the first 12 months of HAART (P = 0.12 and P = 0.01, respectively) (data not shown). However, at multivariate analysis, the difference by gender was not significant (HR, 0.78; 95% CI, 0.58-1.05; P = 0.103). Previous ADI (HR, 1.48; 95% CI, 1.15-1.90; P = 0.002), CD4 cell count 200-350 × 106 cells/l (HR, 0.63; 95% CI, 0.46-0.86; P = 0.003), CD4 cell count > 350 × 106 cells/l (HR, 0.24; 95% CI, 0.15-0.41; P < 0.001) and HIV RNA viral load < 104 copies/ml (HR, 0.54; 95% CI, 0.37-0.80; P = 0.002) were independently associated with clinical progression. A significantly reduced risk of progression to ADI or death was found in patients infected via heterosexual or homosexual contacts compared with those infected by intravenous drug use (HR, 0.71; 95% CI, 0.52-0.96; P = 0.028, and HR, 0.65; 95% CI, 0.45-0.95; P = 0.027, respectively).

With regard to the interaction between HIV infection risk factor and gender on clinical outcome, a lower proportion of female intravenous drug users experienced clinical progression compared with male intravenous drug users (HR, 0.64; 95% CI, 0.42-0.97; P = 0.035). Furthermore, female patients with a baseline viral load 104-105 copies/ml presented approximately half the risk of clinical progression to a new ADI and death compared with male patients with the same baseline viral load (HR, 0.49; 95% CI, 0.28-0.87; P = 0.015). No significant gender difference in clinical progression was found with regard to baseline viral load > 105 copies/ml, baseline viral load < 104 copies/ml or CD4 cell count (data not shown).