The predictive role of interim PET after the first chemotherapy cycle and sequential evaluation of response to ABVD in Hodgkin's lymphoma patients-the Polish Lymphoma Research Group (PLRG) Observational Study

Background: Interim PET after two ABVD cycles (iPET2) predicts treatment outcome in classical Hodgkin's lymphoma. To test whether an earlier assessment of chemosensitivity would improve the prediction accuracy, we launched a prospective, multicenter observational study aimed at assessing the predictive value of iPET after one ABVD (iPET1) and the kinetics of response assessed by sequential PET scanning. Patients and methods: Consecutive patients with newly diagnosed classical Hodgkin's lymphoma underwent interim PET scan after one ABVD course (iPET1). PETs were interpreted according to the Deauville score (DS) as negative (-) (DS 1-3) and positive (+) (DS 4, 5). Patients with iPET1 DS 3-5 underwent iPET2. Results: About 106 early (I-IIA) and 204 advanced (IIB-IV) patients were enrolled between January 2008 and October 2014. iPET1 was (-) in 87/106 (82%) or (+) in 19/106 (18%) of early, and (-) in 133/204 (65%) or (+) in 71/204 (35%) of advanced stage patients, respectively. Twenty-four patients were excluded from response analysis due to treatment escalation. After a median follow-up of 38.2 (3.2-90.2) months, 9/102 (9%) early and 43/184 (23%) advanced patients experienced a progression-free survival event. At 36 months, negative and positive predictive value for iPET1 were 94% and 41% (early) and 84% and 43% (advanced), respectively. The kinetics of PET response was assessed in 198 patients with both iPETs. All 116 patients with iPET1(-) remained iPET2(-) (fast responders), 41/82 with IPET1(+) became iPET2(-) (slow responders), and the remaining 41 stayed iPET2(+) (non-responders); progression-free survival at 36 months for fast, slow and non-responders was 0.88, 0.79 and 0.34, respectively. Conclusion: The optimal tool to predict ABVD outcome in HL remains iPET2 because it distinguishes responders, whatever their time to response, from non- responders. However, iPET1 identified fast responders with the best outcome and might guide early treatment de-escalation in both early and advanced-stage HL.