517 
Predictive Factors of Thrombocytopenia in Patients Randomized to Intermittent or Continuous ART in the ANRS 106 Window Trial
Philippe Tangre*1, I Charreau1, M Lafaurie2, B Marchou3, C Katlama4, P Morlat5, A Polaert1, D Vittecoq6, J P Aboulker1, J M Molina2, and ANRS 106 Study Group
1INSERM SC10, Villejuif, France; 2Hosp St Louis, Paris, France; 3Hosp Purpan, Toulouse, France; 4Hosp Pitie Salpetriere, Paris, France; 5Hosp St Andre, Bordeaux, France; and 6Hosp Paul Brousse, Villejuif, France
Background: The benefit of ART on HIV-related
thrombocytopenia has been well reported. We addressed the predictive factors of
thrombocytopenia in the setting of an intermittent, time-guided ART treatment
trial.
Methods: In this sub-study of the ANRS-106 Window Trial,
403 HIV+ patients under stable HAART (CD4 count >450/mm3
and HIV RNA <200 copies/mL for at least 6 months) were
randomized to either continuous or intermittent (8 weeks off/8 weeks on)
therapy for 96 weeks. Cases of thrombocytopenia (platelet count <150x103/µL)
were reported throughout the study period. Only the first occurrence is
considered. Uni- and multivariate logistic regression
models were used to define predictors of thrombocytopenia.
Results:
Before or at week 0, 12 patients withdrew consent. Baseline
characteristics of the 391 patients included in the analysis were: 80% male; median (IQR) age 42 years (36, 48);
8% had AIDS and 1% a history thrombocytopenic purpura;
45 (12%) infected by either hepatitis B virus (HBV) or hepatitis C virus (HCV);
median CD4, 741/µL (605, 915); median platelet count, 243x103/µL (206,
283); 95% on a triple or more drug regimen, including in 68% zidovudine (AZT) and/or didanosine
(ddI). Thrombocytopenia (<150x103/µL) was reported in 69 patients:
50 on intermittent therapy (25.4%), and 19
on continuous therapy (9.8%) with a median time (IQR) of occurrence of 9 (8, 40)
and 40 (9, 74) weeks in the intermittent and continuous arms, respectively.
Significant platelet decrease (<50x103/µL) was observed in 11 (9 and
2 in the intermittent
and continuous arms, respectively), 2 of them (intermittent arm) having mild
hemorrhagic symptoms.
In univariate analysis, factors associated with
thrombocytopenia, were: intermittent
therapy strategy with odds ratio estimates (OR) and 95%CI 3.13 (1.77, 5.55),
history of thrombocytopenic purpura, OR 7.27 (1.19 to
44.38), baseline CD4 OR (by decrease of 100/µL) 1.25 (1.03 to 1.52) and low
baseline platelets counts (by decrease of 50x103/µL) OR 3.20 (2.24
to 4.58). No influence of age, CDC stage, AZT and/or ddI-based
regimen or viral hepatitis co-infection was observed. In multivariate analysis,
factors associated with increased risk of thrombocytopenia were intermittent
therapy strategy, OR 3.9 (2.01 to 7.57), p = 0.0003; history of thrombocytopenic purpura,
OR 28.04 (1.64 to 478.86), p = 0.02;
and a low platelet count, OR (by decrease of 50x103/µL) 3.40 (2.27
to 5.09), p = 0.0001.
Conclusions: Our study shows that in patients undergoing intermittent
therapy, there is a significant risk of thrombocytopenia in those who have of a
history of thrombocytopenic purpura and low baseline
platelet count. These patients should be discouraged from undertaking intermittent
therapy.
|
