Second cancer risk following Hodgkin's disease: a 20-year follow-up study

Abstract

Purpose: To determine risk factors for the development of second primary cancers during long-term follow-up of patients with Hodgkin's disease (HD).

Patients and methods: We assessed the risk of second cancers (SCs) in 1939 HD patients, who were admitted to the Netherlands Cancer Institute (NKI; Amsterdam) or the Dr Daniel den Hoed Cancer Center (DDHK; Rotterdam) between 1966 and 1986. For 97% of the cohort, we obtained a medical status up to at least January 1989. The median follow-up duration of the patients was 9.2 years; for 17% of the patients, follow-up was longer than 15 years. For more than 98% of all second tumors, the diagnosis was confirmed through a pathology report.

Results: In all, 146 patients developed a SC, compared with 41.9 cases expected on the basis of incidence rates in the general population (relative risk [RR], 3.5; 95% confidence interval [CI], 2.9 to 4.1). The mean 20-year actuarial risk of all SCs was 20% (95% CI, 17% to 24%). Significantly increased RRs were observed for leukemia (RR, 34.7; 95% CI, 23.6 to 49.3), non-Hodgkin's lymphoma (NHL) (RR, 20.6; 95% CI, 13.1 to 30.9), lung cancer (RR, 3.7; 95% CI, 2.5 to 5.3), all gastrointestinal cancers combined (RR, 2.0; 95% CI, 1.2 to 3.1), all urogenital cancers combined (RR, 2.4; 95% CI, 1.4 to 3.7), melanoma (RR, 4.9; 95% CI, 1.6 to 11.3), and soft tissue sarcoma (RR, 8.8; 95% CI, 1.8 to 25.8). As compared with the general population, the cohort experienced an excess of 63 cancer cases per 10,000 person-years. Cox-model analysis indicated the following as significant risk factors for developing leukemia: first-year treatment with chemotherapy (CT), follow-up treatment with CT, age at diagnosis of HD greater than 40 years, splenectomy, and advanced stage. Patients treated with CT in the 1980s had a substantially lower risk of leukemia than patients treated in the 1970s (10-year actuarial risks of 2.1% and 6.4%, respectively; P = .07). Significant risk factors for NHL were older age, male sex, and combined modality treatment as compared with either modality alone. Risk of lung cancer was strongly related to radiotherapy (RT), while an additional role of CT could not be demonstrated. After more than 15 years of follow-up, women treated with mantle-field irradiation before age 20 years had a greater than forty-fold increased risk of breast cancer (P < .001).

Conclusion: While the long-term consequences of HD treatment as administered in the 1960s and 1970s are still evolving, it is promising that patients who received the new treatment regimens introduced in the 1980s have a much lower leukemia risk than patients treated in earlier years. Beginning 10 years after initial RT, the follow-up program of women who received mantle-field irradiation before age 30 years should routinely include breast palpation and yearly mammography.