Thrombocytopenia During Primary HIV-1 Infection Predicts the Risk of Recurrence During Chronic Infection

Le résumé fourni par la source

To the Editors: Thrombocytopenia is common among HIV-1–infected patients, and its prevalence ranges between 10% and 50% depending on the stage of the disease.1,2 The highest prevalence has been reported at the time of primary HIV-1 infection (PHI), reaching up to 50% according to historical data.3,4 These data have been, however, generated from small cohorts of symptomatic patients. The ANRS PRIMO cohort is one of the largest prospective cohorts of patients enrolled during or shortly after PHI.5 Here we assessed the prevalence of thrombocytopenia among patients presenting with PHI and the incidence and risk of recurrence during follow-up. We studied 957 patients enrolled between 1996 and 2009 (median follow-up of 37 months). PHI was classified at enrollment according to clinical symptoms as previously described.5 Thrombocytopenia was defined as a platelet count <150,000/mm3. Platelet count was measured at enrollment, at month 1 (M1), M3 and every 6 months thereafter. Fisher exact test for discrete variables and Wilcoxon-rank-sum test for continuous data were used to compare baseline characteristics of patients presenting with and those free from thrombocytopenia at enrolment. We further compared the incidence rates in thrombocytopenia after PHI between patients receiving combined antiretroviral therapy (cART) and those untreated, and according to the presence of thrombocytopenia during PHI, or not. Thrombocytopenia during follow-up was defined as 2 consecutive platelet measurement <150,000/mm3. Recurrence of thrombocytopenia during follow-up was defined as 2 consecutive platelet measurements <150,000/mm3 after having achieved 2 consecutive platelet measurements >150,000/mm3. We also used the more recent threshold of 100,000 platelets/mm3. A multivariate logistic regression was used to identify independent significant factors associated with thrombocytopenia at enrolment. Statistical analysis was performed with SAS software version 9.2 (SAS Institute Inc, Cary, NC) and STATA software (version 11.0 SE StataCorp 4905 Lakeway Drive, College Station, TX). Overall, 9.7% (n = 93) patients presented with thrombocytopenia at the time of PHI, of whom 22 (2.3%) had less than 100,000 platelets/mm3. No patient had an initial platelet count <50,000/mm3. Thrombocytopenia at enrollment was associated with a clinically severe PHI (P = 0.0002), a lower CD4 cell count (345 vs. 532 cells/mm3, P < 0.0001), a higher plasma HIV RNA (5.73 vs. 5.03 log10 copies/mL, P < 0.0001) and a higher intracellular HIV DNA (3.59 vs. 3.31 log10 copies/106 PBMC, P < 0.0001) (Table 1). There was no association with hepatitis B or C coinfection, cytomegalovirus coinfection, HIV-1 subtype, resistance mutations, or tropism (determined as previously described).6 Patients with initial thrombocytopenia were more likely to start early cART (66% received cART vs. 48.6% P = 0.001), logistic regression showed that cART initiation was most likely driven by lower CD4 cell count and higher plasma HIV RNA rather than thrombocytopenia itself. Duration of thrombocytopenia (time to reach 2 consecutive platelet counts >150,000/mm3) was significantly shorter in patients who initiated early cART during PHI (n = 61 patients, duration of thrombocytopenia 1.08 months) than those untreated (n = 32, duration of thrombocytopenia 1.58 months) (logrank P = 0.09; Wilcoxon P = 0.02). Of note, 13 patients with baseline thrombocytopenia never reached a platelet count >150,000/mm3 despite cART during a median follow-up of 30 months (median platelet count of 126,000/mm3 on the latest sample during follow-up). The 3-year risk of developing a new episode of thrombocytopenia during follow-up was 13.3% in patients with thrombocytopenia at baseline although it was 4.7% in patients with normal baseline platelet count (P < 0.01). When considering the patients who had 2 consecutive platelet measurements above 150,000/mm3 as from M3, incidence rate of thrombocytopenia during subsequent follow-up was 2.0/100 patient-years in those presenting with thrombocytopenia at baseline versus 0.6/100 patient-years in those with normal baseline platelet count (P = 0.03). When considering cART exposure, incidence of thrombocytopenia was 0.2/100 patient-years in treated patients versus 1.4/100 patient-years in untreated ones (P < 10−4). The prognostic role of baseline thrombocytopenia and the protective effect of cART during follow-up were found also when the definition of thrombocytopenia was set to 100,000 instead of 150,000 platelets/mm3.TABLE 1: Baseline Characteristics of Patients With (n = 93) and Without (n = 864) Thrombocytopenia at the Time of PHI in the ANRS PRIMO CohortUnlike historical data,3,4 we show that the prevalence of thrombocytopenia is low at the time of PHI (around 10%). The lower prevalence in our study compared with older ones might be related to the recent wider screening of PHI and the improvement of diagnosis tools, allowing an early diagnosis of PHI especially in paucisymptomatic and asymptomatic patients. Indeed, in the late 1990s, the vast majority of patients diagnosed at the time of PHI presented with severe symptoms,3,4 and we show here that thrombocytopenia is significantly associated with the clinical severity of PHI. Of note, no patient in our series had severe thrombocytopenia (<50,000 platelets/mm3) throughout follow-up. Thrombocytopenia in HIV-1–infected patients is multifactorial, including a direct effect of HIV-1 with a depressed platelet production due to HIV infection of megakaryocytes1,7 and peripheral destruction due to platelet-associated autoantibodies triggered by molecular mimicry between some HIV-1 peptides and membrane-associated platelet glycoproteins.8 Patients with a history of thrombocytopenia at the time of PHI had a 3-fold higher risk of experiencing a subsequent episode during chronic infection than those presenting with a normal platelet count at the time of PHI. Pre-cART data suggested that HIV-1–related severe thrombocytopenia during chronic infection was not associated with a higher risk of progression to AIDS.9 Recent data showed an association between platelets decline during chronic infection and the risk of brain injury in HIV-infected patients.10 Having a normal platelet count at the time of PHI, however, did not prevent from experiencing an episode of thrombocytopenia during chronic infection. The duration of thrombocytopenia was significantly shorter in patients initiating early cART than in those who remained untreated. Moreover, cART had a protective effect against the development of thrombocytopenia during follow-up as the incidence rate of thrombocytopenia was 7-time lower in patients exposed to cART than in the others. Several studies conducted in chronically HIV-infected patients have shown that cART interruption could lead to rapid decrease in platelet count, which recovered upon cART resumption.11–14 More generally, uncontrolled HIV-1 replication was associated with thrombocytopenia,15 the severity of which correlated with the high level of HIV RNA.16 In conclusion, our results suggest that the prevalence of thrombocytopenia is around 10% at the time of PHI, but the risk of recurrence is high during chronic infection. cART has a protective effect against the subsequent occurrence of thrombocytopenia. Altogether, these results strengthen the recommendations for wider and earlier screening of HIV-1 infection, earlier cART initiation, and avoiding treatment interruption.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.