Abstract Body

Antiretroviral treatment during pregnancy has been effective in reducing perinatal HIV transmission. Some nucleos(t)ide reverse transcriptase inhibitors (NRTIs) display genotoxicity in vitro so we evaluated the incidence of cancer among children exposed to NRTIs in utero between 1990 and 2014 in the ANRS French Perinatal Cohort (EPF).

We updated the evaluation of cancer incidence among children exposed to NRTIs in the cohort by cross checking with the French national cancer registry. Associations between cancer risk and exposure to the various NRTIs were evaluated by univariate survival analysis and Cox proportional hazard models. Standardized incidence ratios (SIR) were used to compare risks with those for the general population.

Twenty-one cancers were identified in 15,163 children (median age: 9.9 years [interquartile range (IQR): 5.8-14.2]) exposed to at least one NRTI in utero between 1990 and 2014. Five children were exposed to zidovudine monotherapy, and 15 to various combinations, seven including didanosine. Didanosine was included in only 10% of the prescriptions but was associated with one third of the cancers. Compared with the 2281 children exposed to zidovudine monotherapy, the risk of cancer was higher in the 1461 children exposed to combinations including didanosine (adjusted HR = 3.0 [0.9-9.8]), but similar for all other NRTI combinations. The risk was specifically higher in children exposed during the first trimester, than never exposed, to didanosine (HR=5.5 [2.1-14.4]); it was not significantly associated with second or third trimester exposure to didanosine (HR=1.6 [0.2-12.2]).

Overall, the total number of cases was not significantly different from that expected for the general population (SIR = 0.8 [0.47-1.24]), but among those exposed to didanosine it was twice that expected (SIR = 2.5 [1.01-5.19]). Three cases of pineoblastoma, a very rare cancer, were observed, whereas 0.03 were expected; two of these cases were associated with didanosine exposure.

There are strong arguments to suggest that didanosine displays transplacental oncogenicity. Although these findings cannot be extrapolated to other NRTIs, they stress the need for comprehensive evaluation of the transplacental genotoxicity of this class of antiretrovirals.