Integrase strand transfer inhibitor (INSTI) use and cancer incidence


BACKGROUND: Limited data exist examining the association between cancer and INSTI use. We aimed to assess whether cumulative INSTI (dolutegravir, raltegravir, elvitegravir) exposure was associated with increased cancer risk.
METHODS: RESPOND participants INSTI-naïve at baseline (latest of local cohort enrolment or 1 January 2012) were followed until earliest of first cancer, final follow-up, or 1 October 2018. Negative binomial regression assessed associations between cancer incidence and time-updated cumulative INSTI exposure, adjusted for potential confounders (Figure). INSTI exposure was lagged by 6 months, accounting for possible confounding by indication and reverse causality.
RESULTS: Of 21,267 individuals included, 73% were male, 21% antiretroviral treatment (ART)-naïve, with median age 45 years (interquartile range [IQR] 37-52).
Overall, 9698 (46%) started INSTIs during follow-up. Median cumulative exposure on INSTIs was 26 months (IQR 13-40). During 106,637 person-years of follow-up (PYFU), there were 752 events (incidence rate [IR] 7.1/1000 PYFU [95% CI: 6.6-7.6]): 139 AIDS, 613 non-AIDS cancers. The commonest cancers were lung (11%), anal (9%) and non-Hodgkin's lymphoma (8%).
After adjustment, there was no association between cancer risk and lagged INSTI exposure (global p=0.69, Figure). Results were similar when stratified by age, excluding those with cancer prior to baseline, or analysing AIDS/non-AIDS cancers separately.
There was a significant interaction between INSTI exposure and ART-experience at baseline (interaction p=0.004). After adjustment, those ART-naïve (n=4502, 132 events) had a lower cancer incidence at >6 months vs no INSTI exposure (>0-6 months: IR ratio 0.78 [95% CI: 0.41-1.48]; >6 months: 0.51 [0.30-0.88]). Conversely, those ART-experienced with viral load <200 copies/mL (n=15147, 568 events) had a higher cancer incidence at >6 months vs no INSTI exposure (>0-6: 1.05 [0.69-1.60]; >6: 1.29 [1.01-1.65]).

CONCLUSIONS: In this large cohort, there was no association between lagged INSTI exposure and cancer risk, although this differed by prior ART exposure. Confounding by indication cannot be excluded.