Risk of second breast cancer in female Hodgkin's lymphoma survivors: a meta-analysis

Background Women treated for Hodgkin's lymphoma (HL) have an elevated risk of developing second breast cancer (SBC) compared with the general population. We planned this meta-analysis to quantify the long-term risk of SBC and analyze the contributing risk factors among HL survivors. Methods According to predefined selection criteria, literature search identified 34 studies that were included in the analyses. Results After eliminating overlapping or duplicate data, 957 incidences of SBC were encountered in 24,505 females with HL over a median follow-up of 14.9 years. The medians: age at the diagnosis of HL, age at diagnosis of SBC, and latency since HL treatment to the development of SBC were 23.7, 35.0, and 17.7 years, respectively. The pooled relative risk (RR) of SBC was 8.23 (95% CI, 5.43-12.47, I.sup.2 = 96%), with a median absolute excess rate of 22.9 per 10,000 person-years. The RR was found inversely related to age at diagnosis of HL with the highest rate (68.7; [95%CI, 28.08-168.11], I.sup.2 = 79%), occurred in young patients ([less than or equai to] 15 years old), where the RR in older women ([greater than or equai to] 40 years old) was not significant (0.55; [95% CI, 0.09-3.52]). Analysis of RR by 5-year increments since the treatment of HL showed that the risk was highest after 15-19 years of latency (13.87; [95% CI, 7.91-24.30], I.sup.2 = 89%). Analysis of the effect of treatment modalities showed that the RR rates were (4.70; [95% CI, 3.28-6.75], I.sup.2 = 74%), (5.65; [95%CI, 2.94-10.88], I.sup.2 = 91%), and (1.19; [95% CI, 0.50-2.82], I.sup.2 = 65%), for radiotherapy (RT) only, combined RT and chemotherapy (CT), and CT only, respectively. To investigate the demonstrated heterogeneity, meta-regression analysis was performed when feasible. In most such analyses, the natural logarithm of RR was inversely associated with age at HL diagnosis. Conclusions We conclude that, the current meta-analysis provided the most recent comprehensive estimate of the risk of SBC in a broad-range of HL survivors. Younger age at diagnosis proved to be a dominant risk factor. The obtained results would serve providing breast cancer screening recommendations for HL survivors.