Reduced incidence of kaposi's sarcoma and of systemic non‐hodgkin's lymphoma in HIV‐infected individuals treated with highly active antiretroviral therapy

CI, confidence interval; DMI-2, Dossier Médico-économique de I′Immunodéficience 2; HAART, highly active antiretroviral therapy; IR, incidence rate; KS, Kaposi's sarcoma; NHL, non-Hodgkin's lymphoma; PBL, primary brain lymphoma; RR, rate ratio. In contrast to the well-established reduction in the incidence of KS, the impact of HAART on the incidence of other AIDS-defining (i.e., NHL and invasive cervical cancer) or non-AIDS-defining (e.g., Hodgkin's disease) cancers has not been definitively ascertained.1 KS and NHL have different patterns of presentation, according to the HIV-related level of CD4 lymphopenia. Incidence rates of NHL tend to increase with decreasing levels of immunosuppression, whereas high rates of KS have been observed in either the early or the late phases of HIV infection.2 The introduction of HAART has not changed the level of CD4 lymphocytes at diagnosis of AIDS-related conditions, but the restoration of the immune system produced in HIV-infected individuals treated with HAART has made less common those illnesses that occur at advanced stages of immunodeficiency.3 This fact, thus, is likely to differentially influence the risk of KS or of NHL in HIV-infected individuals. Studies based on individual HAART data are relatively rare.4, 5, 6, 7 Their findings have consistently shown a marked decrease in PBLs and a 20–70% reduction in KS risk in HAART recipients. Regarding systemic NHL, nonstatistically significant decreased risks6, 8 or significantly reduced risks7 have been recorded for HAART-treated patients. A large overview of 23 prospective studies conducted from 1992 through 1999 in North America, Europe and Australia compared incidence rates of NHL in the pre-HAART (1992–1996) and the post-HAART (1997–1999) periods. Whereas significant declines were noted for PBL (–60%) and for immunoblastic lymphoma (–40%), a nonsignificant 20% increase was registered for Burkitt's lymphoma.1 In our study, data from the DMI-2 longitudinal database of HIV+ individuals followed up at Nice University Hospital, France, were analyzed to assess the impact of HAART on incidence rates of KS and systemic NHL. At first visit, information was collected on sociodemographic characteristics, HIV-exposure category, date of first HIV+ test, CD4+ cell count and clinical conditions. At each follow-up visit, scheduled at least every 6 months, data concerning HAART prescription (i.e., at least triple drug combination including or not protease inhibitors), clinical events and biologic markers were regularly recorded. Further details about the use of the DMI-2 database have previously been published.9 We evaluated 3,510 HIV-infected individuals followed up between January 1988 and October 2001. These persons were still alive at the beginning of the study period (i.e., as of January 1996, when HAART was first introduced in France) and were not diagnosed with KS or with NHL before the beginning of the study period. A person was considered positive for HAART history if he or she had started HAART during the study period (January 1996–October 2001) and was not diagnosed with KS or NHL before starting HAART. Individuals who developed KS or NHL before HAART initiation or within 60 days after starting HAART were considered as having a negative history of HAART. Person-years at risk for KS or NHL were computed from the date of enrollment to the date of death, the date of cancer diagnosis or the date of last follow-up. IRs per 1,000 person-years were calculated. Both crude RRs and RRs adjusted for age and HIV-transmission category (i.e., homosexual/bisexual men vs. others), with their 95% CIs, were computed to estimate cancer risk since the first treatment with HAART. Adjusted RRs were estimated using a Poisson regression model. Data management was performed using SPSS version 9 and data were analyzed using STATA version 7.0 (both from SPSS, Chicago, IL). Among the 3,510 HIV-infected individuals included in the present study, 2,589 (73.8%) had a positive history for HAART. Most patients were prescribed a combination therapy including protease inhibitors (88.3%), while 9.5% received a combination therapy with non-nucleoside reverse transcriptase inhibitors. Only 2.2% of patients received a combination therapy including Abacavir. Table I shows the main baseline characteristics of the study population, according to HAART history. The proportion of HAART-treated individuals was significantly higher among men (75.8%) than women (69.1%), in those aged 40 years or more and among homosexual/bisexual men (77.1%). Individuals with CD4+ counts below 200 at enrollment and those enrolled in 1996 or after were slightly more represented among those who received HAART. Overall, 14,013 person-years of follow-up were recorded among HIV-infected individuals with a positive history for HAART and 3,909 person-years among those with a negative history. Overall, 42 cases of KS and 36 cases of systemic NHL were diagnosed. None of these 3,510 HIV-infected individuals was diagnosed with PBL. After adjustment for age and HIV-transmission category, IRs for both KS and NHL were approximately 5 times lower among HIV-infected patients with a positive history for HAART (1.22/1,000 person-years for KS, 1.07/1,000 person-years for NHL) than IRs recorded among those with no history of HAART (RR = 0.22, 95% CI 0.12–0.41 for KS; RR = 0.24, 95% CI 0.12–0.47 for NHL) (Table II). Our findings are in agreement with recent data showing a statistically significant reduction of NHL risk in HIV+ patients treated with HAART.7 They strongly indicate that, in addition to KS risk, HAART may substantially reduce the risk of NHL in HIV+ individuals. Given the reductions in mortality of HIV-infected persons since the introduction of HAART in Western countries, systemic NHL may become more important as a cause of morbidity and mortality. Our findings, therefore, suggest that appropriate antiretroviral treatment of HIV infection may prevent such serious malignant complications of HIV infection. Yours sincerely, We thank Ms. M. Di Pasquale for editorial assistance. M.P. Carrieri, C. Pradier, P. Piselli, M. Pische, E. Rosenthal, P. Heudier, J.