P108: Integrating baseline circulating tumor DNA with interim PET improves outcome prediction in relapsed/refractory classical Hodgkin lymphoma

Figure 1: Cumulative event rate of retreatment according to partitioning analysis categories. Background: Reliable biomarkers for early identification of treatment failure for relapsed/refractory (r/r) cHL are lacking. In this scenario, circulating tumor DNA (ctDNA) profiling may guide rational treatment choice. Aim: We performed a retrospective analysis in r/r cHL patients treated with the BEGEV regimen to assess the predictive efficacy of baseline ctDNA quantification. Additionally, we evaluated whether integrating ctDNA genotyping with interim positron emission tomography (iPET) may improve outcome prediction. Methods: 54 patients with r/r cHL treated with 4 cycles of BEGEV followed by autologous (auto) or allogeneic (allo) stem cell transplantation (SCT) were included in the study. Response was assessed by iPET after 2 cycles; complete response was defined as a Deauville Score (DS) of 1–3. Blood samples collected at baseline were profiled by CAPP-Seq. We performed partitioning analysis to evaluate the predictive value of baseline ctDNA load combined with iPET, assessed by need for further therapy within 18 months from BEGEV initiation. Results: In response to first-line therapy, 65% of patients were refractory. BEGEV was administered as second-line therapy in 61% of patients whereas 39% received BEGEV beyond second line. After induction, 54% of patients underwent auto-SCT and 24% proceeded to allo-SCT. The median baseline ctDNA value reported as haploid genome equivalent per ml (hGE/ml) was 39 (range, 4–5086), with ctDNA detected in PD patients being significantly higher as compared to CR patients (P=.0002). 31% of patients were iPET positive (DS 4/5), while 69% were negative. Baseline ctDNA predicted need for retreatment with an accuracy rate of 81% (sensitivity 63%, specificity 89%). iPET predicted the need for retreatment with similar accuracy rate (83%) and specificity (87%), but with a higher sensitivity (75%). Integrating baseline ctDNA and iPET resulted in an increased predictive value (accuracy rate 89%, sensitivity 75%, specificity 95%). Based on the results of the partitioning analysis, patients with positive iPET and high baseline ctDNA (>31 hGE/ml) showed a significantly inferior 18-months treatment-free survival with 92,3% (95% CI 84.9–99.7) of patients in this category requiring additional treatment (p<0.0001) (Fig 1). Conclusion: Patients with high baseline ctDNA and positive iPET have a limited benefit from completing the BEGEV regimen. Early switch to non-chemotherapeutic agents should be evaluated in prospective trials.