D:A:D is an ongoing observational study of HIV cohorts in Europe, the United States and Australia. The smoking analysis included all cohort members without a cancer history. Follow-up began when a person entered D:A:D or January 2004 and continued to a first cancer diagnosis, death, the last study visit plus six months or February 2015. The investigators divided participants into current smokers, never-smokers, ex-smokers at baseline and smokers who stopped smoking during follow-up. The last group included five subgroups: those who had stopped smoking for less than one year, for one to two years, for two to three years, for three to five years or for more than five years. The D:A:D team used Poisson regression analysis to assess the impact of stopping smoking on incidence of (1) all cancers combined, (2) smoking-unrelated cancers, (3) smoking-related cancers excluding lung cancer and (4) lung cancer.
Related: New HIV Capsid Inhibitor Strongly Stops HIV Replication at Two Steps
The study group included 35,424 HIV-positive people, 72.5% of them men. At baseline, 53% had a viral load below 500 copies/mL, median CD4 count measured 444 cells/mm3 and median age stood at 40 years (interquartile range 34 to 46). Through a median follow-up of nine years, 1,980 cancers developed, including 1,251 smoking-unrelated cancers, 487 smoking-related cancers excluding lung cancer and 242 lung cancers.
Compared with people who never smoked, in people who quit incidence of all cancers was highest within the first year of quitting (adjusted rate ratio about 1.6), then fell significantly (P for trend < .01) to match the incidence in never-smokers after smokers had quit more than five years. The same type of analysis confirmed significant declines in relative incidence of smoking-unrelated cancers from less than one year after quitting to more than five years (P for trend = 0.04) and of smoking-related cancers excluding lung cancer (P for trend = .04). But lung cancer incidence remained elevated compared with never-smokers, standing more than 11 times higher within the first year of quitting and remaining about eight times higher among people who had quit more than five years (P for trend = .13).
The D:A:D investigators propose that persistently elevated lung cancer incidence after smoking stops in this HIV cohort "suggests that the oncogenic potential for smoking is not reversed for lung cancer in the time frame that we have investigated." They add that lung cancer incidence in HIV-negative people does drop consistently with longer time since quitting. While discouraging HIV-positive people from starting smoking and promoting cessation remain priorities, the researchers stress that "monitoring and awareness of lung cancer should continue in those who stop smoking." They advocate longer studies of HIV-positive people who quit smoking to learn "whether and when lung cancer incidence declines."
Mark Mascolini writes about HIV infection.