Background: Treatment decisions for patients with limited-stage Hodgkin lymphoma (HL) involve trade-offs between combined modality therapy (CMT), which includes involved field radiation therapy (IFRT), versus ABVD alone. While CMT is associated with superior disease control, these gains must be balanced against late-effects associated with radiation, including a long-term increased risk of solid cancers. This balance may be particularly important for young women in whom IFRT would include portions of the breast, and in whom substantially increased relative risks (RR) of breast cancer have been reported. The use of a Markov decision model (MDM) may assist in evaluating the principles of this trade-off.
Methods: We developed a MDM to explore which strategy, CMT or ABVD alone, maximizes life expectancy among young women with limited-stage HL involving supraclavicular, axillary, and/or mediastinal adenopathy and whose IFRT would therefore include portions of the breast. Three base cases assessing women ages 20, 30 and 40 years at the time of HL diagnosis were evaluated. Published data on the RR of breast and other cancers following HL treatment, the risk of HL relapse with CMT and ABVD alone, expected survival with breast and other cancers, the risk of relapse following salvage HL therapy, the risk of leukemia after salvage HL therapy, and expected survival with secondary leukemia were used to populate the model. The RR of breast and other solid cancers was assumed to be negligible for the first 10 years after treatment and constant thereafter. The risk of HL relapse was assumed to be constant during the first five years post-treatment, and substantially lower thereafter. Uncertainty in critical variables was explored with deterministic sensitivity analysis.
Results: In our model, women age 20, 30, and 40 years at the time of HL diagnosis have a life expectancy of 56.1, 64.1, and 70.1 years with CMT, and 62.5, 66.1, and 69.9 years with chemotherapy alone. For women age 40 years at HL diagnosis, CMT was the favored strategy at all time points. For younger women, ABVD alone maximized life expectancy, but this benefit was manifest only > 20 years after HL diagnosis. Our model was robust in one-way sensitivity analyses across a plausible range of critical variables. In two-way analyses, the model was sensitive to extreme values for the relative risk of breast cancer post-CMT, the relative risk of other cancer post-CMT, and the risk of HL relapse following ABVD alone.
Conclusions: The optimum treatment strategy for young women with limited-stage, supra-diaphragmatic HL is likely influenced by age at diagnosis. Use of a MDM may assist in evaluating the trade-offs associated with current treatment options.
Disclosure: No relevant conflicts of interest to declare.