Risks of second non-breast primaries following breast cancer in women: a systematic review and meta-analysis

Abstract

Background Second primary cancer incidence is rising among breast cancer survivors. We examined the risks of non-breast second primaries, in combination and at specific cancer sites, through a systematic review and meta-analysis. Methods We conducted a systematic search of PubMed, Embase, and Web of Science, seeking studies published by March 2022. We included studies that reported standardized incidence ratios (SIRs), with associated standard errors, assessing the combined risk of second non-breast primaries following breast cancer. We performed meta-analyses of combined second primary risks, stratifying by age, follow-up duration, and geographic region. We also assessed second primary risks at several specific sites, stratifying by age. The inverse variance method with DerSimonian–Laird estimators was used in all meta-analyses, assuming a random-effects model. Associated biases and study quality were evaluated using the Newcastle–Ottawa scale. Results One prospective and twenty-seven retrospective cohort studies were identified. SIRs for second non-breast primaries combined ranged from 0.84 to 1.84. The summary SIR estimate was 1.24 (95% CI 1.14–1.36, I2: 99%). This varied by age: the estimate was 1.59 (95% CI 1.36–1.85) when breast cancer was diagnosed before age 50, which was significantly higher than in women first diagnosed at 50 or over (SIR: 1.13, 95% CI 1.01–1.36, p for difference: < 0.001). SPC risks were also significantly higher when based on Asian, rather than European, registries (Asia—SIR: 1.47, 95% CI 1.29–1.67. Europe—SIR: 1.16, 95% CI 1.04–1.28). There were significantly increased risks of second thyroid (SIR: 1.89, 95% CI 1.49–2.38), corpus uteri (SIR: 1.84, 95% CI 1.53–2.23), ovary (SIR: 1.53, 95% CI 1.35–1.73), kidney (SIR: 1.43, 95% CI 1.17–1.73), oesophagus (SIR: 1.39, 95% CI 1.26–1.55), skin (melanoma) (SIR: 1.34, 95% CI 1.18–1.52), blood (leukaemia) (SIR: 1.30, 95% CI 1.17–1.45), lung (SIR: 1.25, 95% CI 1.03–1.51), stomach (SIR: 1.23, 95% CI 1.12–1.36) and bladder (SIR: 1.15, 95% CI 1.05–1.26) primaries. Conclusions Breast cancer survivors are at significantly increased risk of second primaries at many sites. Risks are higher for those diagnosed with breast cancer before age 50 and in Asian breast cancer survivors compared to European breast cancer survivors. This study is limited by a lack of data on potentially confounding variables. The conclusions may inform clinical management decisions following breast cancer, although specific clinical recommendations lie outside the scope of this review.

Risks of second primary cancers among 584,965 female and male breast cancer survivors in England: a 25-year retrospective cohort study

“…This is the largest published cohort study of its kind to date, with a longer mean follow-up than any other study of comparable size. 5 , 6 …”

“…Recent systematic reviews estimated female and male non-breast SPC risks following BC as 24% and 27% greater than population-level, 5 , 6 with greater relative increases in those diagnosed with BC below age 50. 5 , 6 Female thyroid, endometrial, ovary, kidney, oesophagus, melanoma, leukaemia, lung, stomach, and bladder SPCs risks were elevated, 5 whereas males were at increased thyroid, pancreatic, and colorectal SPC risks.…”

“…Recent systematic reviews estimated female and male non-breast SPC risks following BC as 24% and 27% greater than population-level, 5 , 6 with greater relative increases in those diagnosed with BC below age 50. 5 , 6 Female thyroid, endometrial, ovary, kidney, oesophagus, melanoma, leukaemia, lung, stomach, and bladder SPCs risks were elevated, 5 whereas males were at increased thyroid, pancreatic, and colorectal SPC risks. 6 These studies produced disparate risk estimates and relied predominantly on population-based cancer registries without linkages to other data, so were unable to reliably assess risk variation by sociodemographic factors, BC pathology, and BC treatment.…”

“…Participant data were linked to obtain comprehensive data on socioeconomic status, BC pathology, BC treatment, curative and prophylactic surgeries, and predominantly self-reported ethnicity. We aimed to estimate combined and site-specific relative and absolute SPC risks in both genders, and assess variability in risks by socioeconomic factors, tumour characteristics and treatments, at sites at particularly elevated risk following BC: the contralateral breast, 9 endometrium, 5 ovary, 5 and all non-breast sites combined. 5 , 6 …”

“…We aimed to estimate combined and site-specific relative and absolute SPC risks in both genders, and assess variability in risks by socioeconomic factors, tumour characteristics and treatments, at sites at particularly elevated risk following BC: the contralateral breast, 9 endometrium, 5 ovary, 5 and all non-breast sites combined. 5 , 6 …”