ABVD or BEACOPP for Advanced Hodgkin Lymphoma

The development of multiagent chemotherapy dramatically changed the prognosis of patients with advanced-stage Hodgkin lymphoma (HL). Although almost all of these patients died when treated with radiotherapy or single-agent chemotherapy, the fourdrug regimen MOPP (mustargen, oncovin, procarbazine, and prednisone) led to remission rates of more than 50% and an overall survival (OS) rate of more than 60%. Shortly thereafter, ABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine) as an alternative anthracycline-containing regimen was introduced in the treatment of HL. It took 20 years before ABVD was accepted as better than MOPP or MOPP-based hybrid variants. In the meantime, multiagent regimens such as ChlVPP (chlorambucil, vinblastine, procarbazine, and prednisolone) plus EVA (etoposide, vinblastine, and doxorubicin) and Stanford V were evaluated but failed to improve outcomes of patients with advanced-stage HL. The quest for a more effective regimen in HLwas the rationale for the GermanHodgkin StudyGroup (GHSG) to design BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone)more than 20 years ago. This regimen was developed using baseline (BEACOPPbaseline) or escalated doses (BEACOPPescalated) and subsequently compared with the GHSG standard at that time, COPP (cyclophosphamide, vincristine, procarbazine, and prednisone) –ABVD, in the HD9 trial. With a total of 1,195 randomly assigned patients and 5 years of follow-up, BEACOPPescalated was found to be significantly better than BEACOPPbaseline and COPP-ABVD. The 10-year update confirmed and extended the initial findings, demonstrating an improvement of 18% in tumor control between BEACOPPescalated and COPP-ABVD as well as an OS difference of 11%. Because BEACOPPescalated was also associated withmore hematologic toxicity, infections, infertility, and secondary leukemia, there has been some controversy about this regimen since the initial publication. Themajority of patients treated with BEACOPPescalated in HD9 developed grade 3 to 4 leucopenia and thrombocytopenia, and 22% had infectious complications. However, this did not translate into higher treatment-associated mortality with BEACOPPescalated, as compared with the other two arms of the HD9 trial, BEACOPPbaseline and COPP-ABVD. In the 10-year update, the higher response rates in patients treated with BEACOPPescalated resulted in significantly lower overall mortality (12%) as compared with those in patients treated with BEACOPPbaseline (19%) or COPP-ABVD (24%). With this clear improvement in efficacy, the GHSG HD12 follow-up trial for advanced-stage HL aimed at reducing overall toxicity of BEACOPP by employing four cycles of BEACOPPescalated followed by four cycles of BEACOPPbaseline (ie, 41 4). However, this approach failed. It was the next-generation trial in advanced-stage HL, HD15, in which more than 2,100 patients were randomly assigned, that demonstrated the reduction to six cycles of BEACOPPescalated followed by radiation administered to residual positron emission tomography (PET) –positive disease $ 2.5 cm led to a significantly improved and less toxic regimen. With the disappearance of potential alternatives such as Stanford V, the choice of the best treatment for advanced-stage HL increasingly focused on the question of ABVD or BEACOPP. Subsequently, collaborative groups directly compared ABVD with BEACOPPescalated in four prospective trials. 10-13 Instead of using eight cycles of BEACOPPescalated, as in HD9, three of these trials used 41 4, similar to the experimental arms of the HD12 trial; the Italian HD2000 instead used four cycles of BEACOPPescalated followed by two cycles of BEACOPPbaseline (4 1 2). In these four trials, a total of 1,227 patients were randomly assigned to 41 4 or 4 1 2 BEACOPP variants and ABVD. All four trials reported significant improvements in tumor control, with 5-year progression-free survival (PFS) gains ranging between 12% and 18% and OS differences of 4% to 8% favoring BEACOPP. In the article accompanying this editorial, Merli et al report an update of their 2009 publication of the Italian HD2000 trial. This three-arm randomized trial registered 307 patients with HL, of whom 12 were subsequently excluded. The initial article, with a median follow-up of 42 months, also published in Journal of Clinical Oncology, had shown a significantly better PFS with BEACOPP 4 1 2 as compared with ABVD (81% v 68% P 5 .038). In the 10-year update, with a median follow-up of 120 months, PFS with ABVD, BEACOPP 4 1 2, and the 10-drug COPP-EBV-CAD (cyclophosphamide, lomustine, vindesine, melphalan, prednisone, epidoxorubicin, vincristine, procarbazine, vinblastine, and bleomycin) regimen was 69%, 75%, and 76%, respectively, with corresponding OS rates of 85%, 84%, and 86%. A total of 13 secondary neoplasias were reported, one after treatment with ABVD, six after BEACOPP, and six after COPP-EBV-CAD. The authors concluded that with longer follow-up, they were no longer able to confirm the superiority of BEACOPP 4 1 2 over ABVD in terms of PFS and that this was mainly because of higher mortality from secondary malignancies. However, with fewer