Risk of male breast cancer after Hodgkin lymphoma

Numerous studies have shown that female survivors of Hodgkin lymphoma (HL) treated with chest radiotherapy have an increased breast cancer (BC) risk.^1-4 The risk increases with higher radiation doses, larger volumes, and younger age at treatment.^4-9 Treatment-induced primary ovarian insufficiency appears to reduce radiation-associated risk.^5,10 Because male BC is a rare disease in the general population, very few studies assessed the risk of BC in male survivors of HL. Although several large HL cohort studies observed at least 1 case of male BC, we identified only 2 studies that quantified the risk of male BC after HL, both restricted to childhood survivors of HL.^11,12 So far, no studies provided treatment-specific risk estimates. Therefore, we assessed the treatment-specific risk of BC in males in a large cohort of Dutch survivors of HL with long-term and near complete follow-up.

Our cohort comprised 3077 male survivors of 5-year HL treated before age 51 years between 1965 and 2013 in 7 university and 13 general hospitals in the Netherlands. Detailed information regarding HL diagnosis, treatment modalities of primary and relapse treatment (ie, radiation fields, chemotherapy regimens, and number of cycles), and follow-up was collected from medical records as described previously.^13-16 Data on subsequent primary cancers were collected via review of medical records, responses to questionnaires sent to general practitioners (the response regarding subsequent malignancies was 96% complete until 1989),¹⁷ and record linkage with the Netherlands Cancer Registry (NCR) since 1989, when the NCR reached nationwide coverage.^18,19 

The time of being at risk was calculated starting at 5 years after the start of HL treatment and ended at date of the first BC diagnosis, death, or censoring (date of last vital status in case of linkage with NCR or date of last medical information otherwise), whichever occurred first. None of the male survivors of HL developed a ductal carcinoma in situ of the breast. Subsequent primary cancers other than BC were ignored in analyses. Expected numbers of BC incidences were calculated based on age-specific and calendar period–specific male BC incidence rates in the Dutch population multiplied by the corresponding number of at-risk person-years. From the observed and expected numbers of BCs, we calculated standardized incidence ratios (SIRs), absolute excess risks (AERs; expressed per 10 000 person-years), and the corresponding 95% confidence intervals (CIs).²⁰ Tests for heterogeneity and trends of SIR by age at HL treatment were performed within collapsed person-time in Poisson regression models.²¹ Cumulative 20-, 30-, and 40-year BC incidence was estimated in the presence of death as a competing risk. P < .05 was considered statistically significant. All statistical tests were performed using Stata statistical software, version 15 (StataCorp, College Station, TX).

After a median follow-up of 20 years (interquartile range [IQR], 14-26 years), we observed 8 males with BC in our cohort of survivors of HL (characteristics presented in Table 1). The median age at BC diagnosis was 55 years (IQR, 46-58 years), and the median interval between start of HL treatment and BC diagnosis was 27 (IQR, 21-29; range, 18-32) years. Compared with the general population, BC risk was 23.3-fold increased (95% CI, 10.1-46.0), representing 1.6 (95% CI, 0.7-3.3) excess BCs per 10 000 person-years (Table 2). BC risk decreased with older age at HL treatment (P_trend = .009); the SIR was highest for males treated aged <25 years (SIR 69.0; 95% CI, 14.2-201.6), corresponding to 1.6 excess cases per 10 000 person-years (95% CI, 0.3-4.9). Risk was not significantly increased in patients treated at age ≥35 years. SIRs remained increased up to 30 years after HL treatment. Regarding attained age at end of follow-up, we also observed the highest SIR among males while they were young (54.8; 95% CI, 1.4-305.6 for those aged <40 years). However, males with an attained age between 50 and 59 years had the highest AER (4.6; 95% CI, 1.4-10.9). However, numbers of events in subgroup analyses were small. The overall cumulative BC incidence was 0.1% (95% CI, 0.02-0.3) at 20 years of follow-up since HL treatment, 0.6% (95% CI, 0.2-1.1) at 30 years, and 0.7% (95% CI, 0.3-1.4) at 40 years. By the age 30 years, 0.04% of male survivors of HL had developed BC (95% CI, 0.004-0.2), and by the age of 60 years, this was 0.4% (95% CI, 0.2-0.8).

So far, data on BC risk in male survivors of HL are scarce. Recently, a systematic review on male BC after childhood cancer concluded that most reports on male BC concerned case reports and that risk measures were only reported in 5 cohort studies of pediatric cancer survivors.¹² The authors conducted an analysis in the PanCareSurFup childhood cancer cohort and reported a 35.8-fold increased BC risk (95% CI, 7.4-104.6, based on 3 cases) in male survivors of childhood HL compared with the general population, corresponding to an excess of 0.5 cases per 10 000 person-years (95% CI, 0.1-1.4). We observed a lower SIR but a higher AER, which can be explained by a difference in background risk because survivors of HL in our study had a median attained age of 50.9 years, compared with 33.6 years in the PanCareSurFup cohort. There was no difference in duration of follow-up (median, 20 years).

To our knowledge, our study is the first report to assess treatment-specific risks for male BC. Based on research in female survivors of HL, we know that chest radiotherapy, especially with large radiation volumes and high radiation doses, strongly increases BC risk.² However, treatment with alkylating chemotherapy reduces BC risk, which has been attributed to primary ovarian insufficiency. In our cohort, 7 of the 8 survivors of HL with BC were treated with chest radiotherapy (including axillary fields), and 5 of these men had received alkylating chemotherapy. In line with the findings in women, treatment with chest radiotherapy without alkylating chemotherapy yielded a strongly increased SIR (20.7; 95% CI, 2.5-74.8); the SIR after treatment with chest radiotherapy and alkylating chemotherapy was not significantly different (41.1; 95% CI, 13.4-96.0).

Recent data among female childhood cancer survivors show an increased BC risk after anthracycline-containing chemotherapy.^6,22-25 Four males with BC received anthracycline-containing chemotherapy for their HL, all of whom also received chest radiotherapy. We did not observe BC cases in patients who were treated with anthracyclines without chest radiotherapy. Treatment with chest radiotherapy and anthracycline-containing chemotherapy resulted in a SIR of 48.1 (95% CI, 13.1-123.1). Because of the limited number of cases, it was not possible to examine separate effects of chest radiotherapy, anthracyclines, and alkylating agents.

The majority of males were diagnosed with stage II BC. Almost all males were treated for BC with radiotherapy, chemotherapy, and/or endocrine therapy before or after surgery (Table 1). Remarkably, 3 males had already developed another malignancy before BC. Two males developed a malignancy after BC, of whom 1 also had another malignancy before (Table 1). After a median follow-up of 4.7 years (IQR, 2.6-6.3 years) since BC diagnosis, 2 males had died from BC.

In conclusion, this study shows a 23-fold increased BC risk in male survivors of HL up to 30 years after HL treatment compared with that in the general population. Because of the rarity of male BC in the general population, this high relative risk results in only 1.6 excess BC cases per 1000 patients followed up for 10 years. Although male BC after HL treatment is an uncommon event, with a 40-year cumulative incidence of 0.7%, clinicians should be aware of this strongly increased risk, particularly in patients who were treated with chest radiotherapy and had a young age at HL treatment. To ensure early diagnosis and treatment, clinicians should be alert to symptoms of BC in male survivors of HL.

The authors thank Rosemarie Wijnands for the data support at the Netherlands Cancer Institute.

This work was supported by the Dutch Cancer Society (NKI 2010-4720) (F.E.v.L.).

Contribution: S.d.V., I.M.K., B.M.P.A., and F.E.v.L. designed the study; C.P.M.J., S.E.R., J.M.R., M.R.N., Y.M.B., B.M.P.A., and the BETER consortium provided study patients; S.d.V., M.S., and F.E.v.L. collected and assembled data; S.d.V., I.M.K., M.S., B.M.P.A., and F.E.v.L. analyzed and interpreted data; S.d.V., I.M.K., and F.E.v.L. contributed to manuscript writing; and all authors approved the final manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

A complete list of the members of the BETER (Better care after Hodgkin lymphoma: Evaluation of long-term Treatment Effects and screening Recommendations) consortium appears in the supplemental Appendix.

Correspondence: Flora E. van Leeuwen, Department of Epidemiology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands; e-mail: f.v.leeuwen@nki.nl.