Several strategies are available for the initial treatment of advanced-stage Hodgkin lymphoma, but the optimal strategy in terms of cost-effectiveness is unclear. The aim of this study was to compare the quality-adjusted effectiveness and costs of five modern treatment options for transplantation-eligible patients with newly diagnosed advanced-stage Hodgkin lymphoma.
A Markov decision-analytic model was developed using a 20-year time horizon. Five of the most common treatment approaches were selected based on clinical experience and expert opinion: (1) six cycles of doxorubicin, bleomycin, vinblastine, dacarbazine (ABVD), including data from the HD2000 trial, Viviani and colleagues, and EORTC trial; (2) six cycles of bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone (BEACOPP; from the HD15 trial or PET-adapted as in the HD18 trial, two initial cycles of BEACOPP followed by four additional cycles for patients with a positive PET and either two or four additional cycles of BEACOPP for patients with a negative PET); (3) PET-adapted escalation (as in the RATHL trial, two cycles of standard ABVD chemotherapy followed by an additional four cycles of ABVD or AVD in PET-negative patients and four cycles of BEACOPP in PET-positive patients); (4) six cycles of brentuximab vedotin, doxorubicin, vinblastine, dacarbazine (A-AVD) or ABVD as in the Echelon-1 trial; and (5) PET-adapted de-escalation (as in the AHL2011 trial, two cycles of BEACOPP followed by PET2 scan; PET-positive patients received two additional BEACOPP cycles and PET-negative patients received two cycles of ABVD; at PET4, PET-negative patients completed two further cycles of either ABVD or BEACOPP depending on what they received after PET2, and PET-positive patients received salvage therapy). Note that all uses of BEACOPP in these strategies were BEACOPPescalated. The randomised groups of interest from these studies comprised 4255 patients enrolled between April, 2000, and January, 2016. Baseline probability estimates and utilities were derived from the included trials in addition to a systematic review of published studies. A Canadian public health payer’s perspective was considered (CAN$1=US$0·74) and adjusted for inflation for 2018. All costs and benefits were discounted by 1·5% per year because life-years now are more valuable than future potential life-years.
Probabilistic analyses (10 000 simulations) showed that, for a willingness-to-pay threshold of CAN$50 000, a PET-adapted de-escalation strategy based on AHL2011 was more cost-effective 87% of the time. This strategy had the highest number of life-years (14·6 years [95% CI 13·7-15·1]) and quality-adjusted life years (13·2 years [95% CI 10·2-14·4]), and the lowest direct costs ($53 129 [95% CI 31 914-94 446]) compared with the other treatment regimens. Sensitivity analyses showed that the model was robust to key variables, including probability of treatment-related mortality, relapse, frequency of secondary malignancy, death from secondary malignancy, and probability of infertility after BEACOPP.
Our results suggest that, when considering cost, effectiveness, and short and long-term toxicities, the preferred treatment strategy for patients with newly diagnosed advanced-stage Hodgkin lymphoma is the PET-adapted de-escalation regimen starting with BEACOPP and de-escalating to ABVD as appropriate. Although our findings do not provide an absolute best treatmentapproach for clinicians to follow for all patients, they can contribute to shared decision making between patients and treating physicians.