Abstract

INTRODUCTION: Since the beginning of the HIV epidemic in resource-rich countries, Pneumocystis jirovecii pneumonia (PjP) is one of the most frequent opportunistic AIDS-defining infections. The Collaboration of Observational HIV Epidemiological Research Europe (COHERE) has shown that primary Pneumocystis jirovecii Pneumonia (PjP) prophylaxis can be safely withdrawn in patients with CD4 counts of 100 to 200 cells/µL if plasma HIV-RNA is suppressed on combination antiretroviral therapy. Whether this holds true for secondary prophylaxis is not known, and this has proved difficult to determine due to the much lower population at risk. METHODS: We estimated the incidence of secondary PjP by including patient data collected from 1998 to 2015 from the COHERE cohort collaboration according to time-updated CD4 counts, HIV-RNA and use of PjP prophylaxis in persons >16 years of age. We fitted a Poisson generalized additive model in which the smoothed effect of CD4 was modelled by a restricted cubic spline, and HIV-RNA was stratified as low (<400), medium (400 to 10,000) or high (>10,000copies/mL). RESULTS: There were 373 recurrences of PjP during 74,295 person-years (py) in 10,476 patients. The PjP incidence in the different plasma HIV-RNA strata differed significantly and was lowest in the low stratum. For patients off prophylaxis with CD4 counts between 100 and 200 cells/µL and HIV-RNA below 400 copies/mL, the incidence of recurrent PjP was 3.9 (95% CI: 2.0 to 5.8) per 1000 py, not significantly different from patients on prophylaxis in the same stratum (1.9, 95% CI: 0.1 to 3.7). CONCLUSIONS: HIV viraemia importantly affects the risk of recurrent PjP. In virologically suppressed patients on ART with CD4 counts of 100 to 200/µL, the incidence of PjP off prophylaxis is below 10/1000 py. Secondary PjP prophylaxis may be safely withheld in such patients. While European guidelines recommend discontinuing secondary PjP prophylaxis only if CD4 counts rise above 200 cells/mL, the latest US Guidelines consider secondary prophylaxis discontinuation even in patients with a CD4 count above 100 cells/µL and suppressed viral load. Our results strengthen and support this US recommendation.