Unsupervised treatment interruptions are associated with increased risk of AIDS or death

An Italian-English study of ‘unsupervised’ treatment interruptions (TIs), which are common in clinical practice, concludes that TIs of 12 weeks or more are associated with clinically significant increased risk of AIDS or death.

Antonella d’Arminio Monforte and colleagues looked at 2,832 patients of the ICONA cohort starting their first HAART. Only TIs of at least 12 weeks were considered.

Over a median follow-up of 148 weeks (IQR: 78–211) 553 patients (19.5%) interrupted HAART for ≥12 weeks and 52 patients (1.8%) interrupted treatment more than once. Patients were generally well and responding well to treatment with median CD4 and viral load at TI of 528 cells/mm3 (IQR: 320–776) and 2.61 log10 copies/ml (IQR: 1.90–3.90) respectively. Over the follow up period 184 patients experienced clinical progression (167 to AIDS and 17 deaths).

The rate of clinical events of patients on therapy was 1.9 per 100 person years (pys) (95% CI: 1.6–2.2) and during TI was 8.7 per 100 pys (95% CI: 6.2–11.8). Crude rate ratio off/on-therapy: 4.55 (95% CI: 3.14–6.48, P=0.0001). In Cox model, TI (RH=2.54 versus continued therapy; 95% CI: 1.54–4.20, P=0.0003), CDC stage B and C versus stage A (RH=1.76; P=0.007 and RH=2.34; P=0.0001), and higher pre-therapy viral load (RH=1.37 per log10 copies/ml higher; 95% CI: 1.12–1.66, P=0.002) were associated with higher risk of progression.

Patients with higher CD4 nadir (RH=0.89 per 100 cells/µl higher; P=0.03), with a greater latest CD4 increase on therapy (RH=0.76 per 100 cells/µl greater increase above nadir; P=0.0001) and with greater latest VL suppression (RH=0.72 per log10 copies/ml greater suppression below pre-therapy; P=0.0001) were at lower risk.

d’Arminio Monforte A, Cozzi-Lepri A, Murri R et al. The effect of HAART interruptions on clinical progression: evidence from ICONA cohort. 2nd IAS Conference on HIV Pathogenesis and Treatment, 13-16 July 2003, Paris. Abstract 145.