Cigarette smoking and risk of second primary cancer: a systematic review and meta-analysis

Approximately 5% of the population are living with a diagnosis of cancer. Recent improvements in survival following a diagnosis of cancer have led to an increase in second primary cancers (SPCs) worldwide. Their aetiology remains largely unknown with a large proportion believed to be related to modifiable lifestyle factors. We conducted a systematic review and meta-analysis of published data that evaluated an association between cigarette smoking and risk of SPC. Studies were identified by searching Medline, Web of Science, CINAHL (Cumulative Index to Nursing and Allied Health Literature) and Scopus databases through March 2021 using broad search criteria. A meta-analysis was performed to derive pooled relative risks (RRs) for SPC defined a priori as smoking-related based on current evidence (lung, upper aero-digestive tract, stomach, pancreas, colorectum, liver, kidney, ureter, bladder and acute myeloid leukaemia). Eleven cohort studies and ten case-control studies met the eligibility criteria for review. There was marked heterogeneity in methods used in terms of classification and timing of smoking, confounders adjusted for and duration of follow-up across the studies. Nine cohort and seven case-control studies classified smoking habits prior to diagnosis of first cancer while the remaining studies classified post-first cancer smoking habits. In a meta-analysis using six studies, an increased risk of smoking-related SPC was observed for both former (RR=1.42; 95% confidence interval (CI) 1.20-1.67) and current smoking (RR=2.76; 95% CI 2.29-3.33), compared with never smoking. The pooled RRs changed only slightly when studies which measured post-first cancer smoking were excluded. A two-fold increase in risk was observed for ever smoking compared with never smoking. In conclusion, there was evidence that smoking might increase the risk of SPC in cancer survivors. For better informed cancer survivorship practice guidelines, more studies are needed particularly of post-cancer smoking and for cancers not known to be caused by smoking.