| Contributors: |
Bruyand, Mathia; Ryom, Lene; Shepherd, Leah; Fatkenheuer, Gerd; Grulich, Andrew; Reiss, Peter; De Wit, Stéphane; Darminio Monforte, Antonella; Furrer, Hansjakob; Pradier, Christian; Lundgren, Jen; Sabin, Caroline; Powderly, B; Shortman, N; Moecklinghoff, C; Reilly, G; Franquet, X; Ryom, L; Sabin, Ca; Kamara, D; Smith, C; Phillips, A; Mocroft, A; Tverland, J; Mansfeld, M; Nielsen, J; Raben, D; Lundgren, Jd; Salbøl Brandt, R; Rickenbach, M; Fanti, I; Krum, E; Hillebregt, M; Geffard, S; Sundström, A; Delforge, M; Fontas, E; Torres, F; Mcmanus, H; Wright, S; Kjær, J; Sjøl, A; Meidahl, P; Helweg Larsen, J; Schmidt Iversen, J; Kirk, O; Reiss, P; Ross, M; Fux, Ca; Morlat, P; Moranne, O; Kesselring, Am; Kamara, Da; Weber, R; Pradier, C; Friis Møller, N; Kowalska, J; Sabin, C; Law, M; d'Arminio Monforte, A; Dabis, F; Bruyand, M; Bower, M; Fätkenheuer, G; Donald, A; Grulich, A; Zaheri, S; Gras, L; Prins, Jm; Kuijpers, Tw; Scherpbier, Hj; van der Meer, Jt; Wit, Fw; Godfried, Mh; van der Poll, T; Nellen, Fj; Lange, Jm; Geerlings, Se; van Vugt, M; Pajkrt, D; Bos, Jc; van der Valk, M |
| Description: |
BACKGROUND: The association between combination antiretroviral therapy (cART) and cancer risk, especially regimens containing protease inhibitors (PIs) or nonnucleoside reverse transcriptase inhibitors (NNRTIs), is unclear. METHODS: Participants were followed from the latest of D:A:D study entry or January 1, 2004, until the earliest of a first cancer diagnosis, February 1, 2012, death, or 6 months after the last visit. Multivariable Poisson regression models assessed associations between cumulative (per year) use of either any cART or PI/NNRTI, and the incidence of any cancer, non-AIDS-defining cancers (NADC), AIDS-defining cancers (ADC), and the most frequently occurring ADC (Kaposi sarcoma, non-Hodgkin lymphoma) and NADC (lung, invasive anal, head/neck cancers, and Hodgkin lymphoma). RESULTS: A total of 41,762 persons contributed 241,556 person-years (PY). A total of 1832 cancers were diagnosed [incidence rate: 0.76/100 PY (95% confidence interval: 0.72 to 0.79)], 718 ADC [0.30/100 PY (0.28-0.32)], and 1114 NADC [0.46/100 PY (0.43-0.49)]. Longer exposure to cART was associated with a lower ADC risk [adjusted rate ratio: 0.88/year (0.85-0.92)] but a higher NADC risk [1.02/year (1.00-1.03)]. Both PI and NNRTI use were associated with a lower ADC risk [PI: 0.96/year (0.92-1.00); NNRTI: 0.86/year (0.81-0.91)]. PI use was associated with a higher NADC risk [1.03/year (1.01-1.05)]. Although this was largely driven by an association with anal cancer [1.08/year (1.04-1.13)], the association remained after excluding anal cancers from the end point [1.02/year (1.01-1.04)]. No association was seen between NNRTI use and NADC [1.00/year (0.98-1.02)]. CONCLUSIONS: Cumulative use of PIs may be associated with a higher risk of anal cancer and possibly other NADC. Further investigation of biological mechanisms is warranted. |