HL has been a successful model for the development of effective treatment approach in clinical oncology. Long-term survivors of that disease have also allowed better recognition and understanding of the late effects of therapy. In a large cohort of 25,305 women with HL, and with 957 incidences of SBC reported from North American and European institutions, the current meta-analysis quantified the risk of SBC. The overall pooled analysis showed that there was an approximate 9-fold increase in the risk of SBC incidence (pooled RR
8.23), and AER of 23 patients per 10,000 person-years.
Patients included in this meta-analysis developed HD at a median age of 23.7
years. There has been incongruity about the influence of younger age at HL diagnosis and the higher risk of SBC, where some studies have failed to prove that effect [39
]. The current meta-analysis clearly showed that younger age at HL diagnosis was associated with increased risk of SBC, and the risk remained after adjusting for other covariates. Moreover, we demonstrated that the RR of SBC for women who developed HL above the age of 40
years was not significant. It is presumed that the higher risk associated with young age at HD diagnosis is attributed to the effect of RT delivered at a time when breast tissue is proliferating.
RT used as the sole therapeutic modality was associated with a 5-fold (RR
4.70) increase in risk and a 14-fold (RR
14.08) increase among young (≤ 30
years of age). Almost all SBC arose within or at the margin of the RT field. The RR of combined RT and any CT was slightly higher than that associated with RT only (5.65 vs. 4.70). An even higher, was the RR associated with the combination of RT and alkylating CT (RR
6.59), thus, the potential protective effect of gonado-toxic alkylating CT was not demonstrated. Several studies showed an inverse association between the use of alkylating CT in HL and SBC risk [24
], however, other investigators reported increased risk [30
]. In this meta-analysis and based on data reported from three studies, the risk associated with the combination of RT and non-alkylating CT was not significant [37
], also found insignificant, was the risk related to the use of CT only.
Analysis of potential additional risk factors was limited due to lack of sufficient data and/or inability to compute missing information, therefore, cautions should be exerted in interpreting results. Two studies [24
], reported on the mantle field RT dose where there was only a small difference in RR between dose
Gy (5.99, and 6.13, respectively). While some studies showed that subjects with SBC were found to be significantly more likely to have received higher doses of mantle RT [52
], this observation was complicated by the fact that patients treated with higher radiation doses have had longer follow-up. Guibout et al. [46
], did not find a significant association between RT dose and SBC, suggesting that the increased risk after HL may indicate a specific susceptibility for developing SBC, or a particular susceptibility to radiation and/or chemotherapy, or both. Conversely, De Bruin et al. showed that the risk of SBC is related to the RT volume [55
], where mantle field irradiation was associated with a 2.7-fold increased risk of SBC compared with mediastinal irradiation alone. Besides, the meta-analysis reported by Franklin et al. showed a RR of 3 comparing extended field versus involved field RT [14
The reason for failing to show a convincing evidence of RT dose–response effect associated with SBC risk is at best divisive. There is evidence for a strongly linear radiation dose response, but only in the lower dose range (up to 5 or 10
]. It has been suggested that cell killing tends to decrease the carcinogenic effect of RT along an exponential curve at doses above 10
]. However, it is known that BC is a known complication of low-dose breast radiation [67
], thus BC may remain a risk among adolescent women who receive any dose of thoracic irradiation for HL.
Although new RT techniques and treatment strategies have the potential to reduce the future burden of late effects, nevertheless, we have shown that an even higher risk was reported in more recent years suggesting that there remains a significant cohort at an increased risk of SBC.
Pelvic RT was found to be associated with a protective effect as reported by De Bruin et al. [55
]. The same effect was also noted by Basu et al., where 3.4% of patients who developed SBC received pelvic RT as compared with 26.3% among those who did not [52
]. It is presumed that the protective effect of pelvic RT is attributed to the induction of premature menopause and the role played by hormone stimulation in RT-induced breast carcinogenesis [43
]. The influence of splenectomy on SBC risk has been controversial. While some studies reported a modest higher risk [71
], other studies failed to show that effect [30
Only one study examined the receptor status of SBC [57
]. The RR of ER-negative/PR-negative SBC was 66% higher than ER-positive/PR-positive SBC among 5-year HL survivors, and nearly two-fold higher among 15-year survivors. Conversely, other studies of small numbers of SBC patients have not found a significant variation in hormone-receptor status when compared with primary BC controls [72
]. While the incidence of hormone receptor-positive BC in the general population exceeds that of ER-negative/PR-negative BC, it is postulated, however, that young women treated for HL may experience premature ovarian failure related to HL therapy, and therefore, their hormonal BC risk factors may differ from those in the general population.
The present meta-analysis has several limitations. First, it is not possible to completely exclude the possibility that the HL itself carries with it an increased risk of second malignancy including SBC. Second, it is very difficult to quantify the possible effect of confounding factors such as lifestyle factors, personal risk, family history, etc. For example, Landgren et al. found increased RR (1.81) of breast cancer among HL patient with positive (vs. negative) family history of cancer [50
]. Third, the analyses showed significant heterogeneity in risk estimate, nevertheless, investigating heterogeneity using meta-regression technique showed the dominant role of age at HL diagnosis. Other limitations include the lack of comprehensive treatment data including information on RT dose and additional treatments, and the lack of sufficient data to model the protective effect of endogenous hormone ablation against the risk associated with exposure to exogenous hormones. Moreover, it is not clear if a similar magnitude of risk is to be expected in a different patient population where the incidence of sporadic BC is low. Finally, the meta-analysis lacks the analysis of SBC outcome. However, SBC incidence rather its mortality was the main objective of the meta-analysis. Moreover, not all studies reported on mortality, besides, analysis of SBC mortality would be confounded by the mortality from HL itself or its therapy-related effect, ascertainment of the cause of death, age of diagnosis of HL or SBC, and length of follow-up.