Pediatric Hodgkin lymphoma: trade-offs between short- and long-term mortality risks

Abstract

As pediatric Hodgkin lymphoma (HL) survival rates approach > 95%, treatment decisions are increasingly based on minimizing late effects. Using a model-based approach, we explored whether the addition of radiotherapy contributes to improved overall long-term survival. We developed a state-transition model to simulate the lifetime HL clinical course, and we compared 2 treatment strategies: chemotherapy alone (CT) and chemoradiotherapy (CRT). Data on HL relapse, late recurrence, and excess second cancer and cardiac late-effects mortality were estimated from the published literature and databases. Outcomes included conditional life expectancy, cause-specific mortality, and proportion alive at age 50. For a hypothetical cohort of HL patients (diagnosis age 15), conditional life expectancy was 57.2 years with CT compared with 56.4 years with CRT. Estimated lifetime HL mortality risk was 3.6% with CT versus 2.2% with CRT. In contrast, combined risk of excess late-effects mortality was lower for CT (1.8% vs 7.4% with CRT). Among those alive at age 50, only 9.2% of those initially treated with CT were at risk for radiation-related late effects (100% for CRT). Initial treatment with CT may be associated with longer average per-person life expectancy. These results support the need for careful consideration of the risk-benefit profile of radiation as frontline therapy in pediatric patients.

Figure 1

Model diagram. Patients diagnosed with HL who respond to initial chemotherapy receive either no additional treatment (CT strategy) or low-dose radiation (CRT strategy). Patients may relapse and receive salvage therapy. Patients then face risks for late recurrence and/or treatment-specific excess second cancer and cardiac mortality late effects. *Late-recurrence risk spans from 3 years (treatment completion) to 10 years since initial diagnosis. Cumulative dosage for COPP/ABV × 4: cyclophosphamide, 2400 mg/m²; vincristine, 5.6 mg/m; procarbazine, 2800 mg/m²; prednisone, 2240 mg/m²; doxorubicin, 140 mg/m²; bleomycin, 40 IU/m²; and vinblastine, 24 mg/m².

Figure 2

Comparison of cumulative overall mortality. This figure shows the cumulative mortality probability for the base case. Solid green line represents CT; dotted green lines, CRT; and solid black line, U S general population.

Figure 3

Two-way sensitivity analysis on short-term CT EFS and CRT risk for radiation-related late-effects mortality. These figures show 2-way sensitivity analyses on the probability of short-term CT EFS and CRT relative risk of radiation-related cardiac (A) and second cancer mortality (B). In both panels, the region where CT is preferred is indicated by the light green shaded area, and for CRT, the dark gray shaded area.

Figure 4

Two-way sensitivity analysis on CRT risk for radiation-related cardiac and second cancer mortality by various levels of short-term CT EFS. These figures show 2-way sensitivity analysis on CRT relative risk for radiation-related cardiac and second cancer mortality for the following levels of short-term CT EFS: probability = 0.82 (A), 0.87 (base case; B), and 0.90 (C). In each panel, the region where CT is preferred is indicated by the light green shaded area, and for CRT, the dark gray shaded area.