End-of-Treatment Response Assessment after Frontline Therapy for Aggressive B-Cell Lymphoma: Landmark Comparison of a Singular PET/CT Scan Versus Ultrasensitive Circulating Tumor DNA

Background: Lymphoma response criteria rely on PET/CT scans at end of therapy (EOT) to determine complete response (CR), which is necessary for cure. ~25% of patients achieve <CR by response criteria, but PET/CT scans have imperfect specificity and the positive predictive value of a single PET/CT scan at EOT is only ~50% (Kostakoglu 2021). To overcome this limitation, clinicians use additional tests including tissue biopsies or repeat PET/CT scans prior to salvage therapy to adjudicate positive results, but it remains unclear how often this occurs. Circulating tumor DNA (ctDNA) has improved tumor specificity that may reduce reliance on additional tests. We addressed these questions within a clinical trial delivering 1L therapy to pts with aggressive B-cell lymphoma.

Methods: Pts were enrolled in a phase 2 study of DA-EPOCH-R or R-CHOP +/- acalabrutinib for treatment naïve DLBCL or high-grade B-cell lymphoma (HGBCL) [NCT04002947]. All pts were >18yrs old, stage II-IV, with adequate organ function. Pts completing induction therapy with available EOT PET/CT scans and EOT plasma samples were studied. PET/CT scans were interpreted by 2 nuclear medicine (NM) radiologists blinded to clinical outcomes using the 5-point Deauville Score (DS) with DS4/5 considered positive. Tissue biopsies and unplanned PET scans after the EOT PET/CT scan to determine remission status were recorded. Measurable residual disease (MRD) in plasma cfDNA was evaluated by Phased Variant Enrichment & Detection Sequencing (PhasED-Seq) at Foresight Diagnostics blinded to clinical outcomes. Tumor derived Phased Variants (PVs) were identified from baseline pretreatment plasma samples or baseline tumor tissue, with matched constitutional DNA from paired PBMCs used to censor germline variants and CHIP. PVs were used to assess EOT MRD. Results were reported as MRD positive when ctDNA levels exceeded an analytical detection threshold (~1:10⁶ cfDNA molecules) corresponding to 98% specificity. The prognostic utility of MRD was compared to standardized PET/CT scan at EOT to freedom-from-progression (FFP) and progression-free survival (PFS).

Results: 55 enrolled pts completing planned therapy had EOT PET/CT and MRD analyzed. 54 (98%) pts were successfully genotyped and comprised the study population. Median age was 61 (range 26-85), including 22 (41%) female, 44 (81%) with advanced stage, and 26 (48%) with IPI score ³3. Pathologic subtypes included 24 (44%) GCB, 21 (39%) non-GCB, 8 (15%) HGBL-DH, and 1 (2%) THRLBCL. 43 (80%) pts received DA-EPOCH-R and 11 (20%) pts received R-CHOP. Twenty-six (48%) pts received acalabrutinib with R-chemotherapy. After 26 months median follow-up, the 2-year FFP and PFS of the cohort was 89% and 84%. Fourteen (26%) pts had a positive EOT PET/CT scan and 40 (74%) were negative. The concordance rate between the 2 NM radiologists on positive scans was 100%. Only 2 of 14 patients with a positive EOT PET/CT had clinical progression requiring salvage therapy (14% PPV). Nine (64%) of these pts required additional tests to determine remission status including 5 repeat PET/CT scans and 4 tissue biopsies. The 2-year PFS for pts with a positive vs negative EOT PET/CT was 66% vs 90% (p=0.08, Fig.A). Twelve (22%) pts had detectable EOT MRD and 42 (78%) were undetectable. 7 pts with detectable EOT MRD had clinical progression requiring salvage therapy or have died. While no pt with undetectable EOT MRD progressed, 1 (2%) died in remission of sudden cardiac death. Of 14 pts with positive EOT PET/CT, 7 (50%) had undetectable ctDNA and none progressed, suggesting MRD testing could have potentially rendered additional tests unnecessary. Conversely, patients with positive PET/CTs who progressed invariably had detectable ctDNA at EOT, suggesting utility of MRD for adjudicating PET/CT results. The 2-yr PFS for pts with detectable vs undetectable EOT MRD was 33% vs 98% (p<0.0001, Fig.B). At EOT, the PFS hazard ratio for MRD was 35.7, as compared with 3.1 for PET/CT.

Conclusions: Response evaluations relying on PET/CT scans often require additional tests and procedures to assess remission status in aggressive B-cell lymphomas. Initiation of salvage therapy without adjudication in patients not achieving CR by EOT PET/CT scans will result in overtreatment. Given the higher specificity and PPV of EOT MRD, we propose an integrated response evaluation strategy combining these methods to reduce unnecessary additional tests and overtreatment.