AUTHORS
Hua-Jay Jeffery Cherng, MD, Beatriz Raposo Corradini, MSc2, Ajay Major, MD, MBA3, Cheng-Shiun Leu, PhD4, Seda S. Tolu, MD5, Jennifer E. Amengual, MD6, Sandra Close, PhD7, Jacob Chabon, PhD8, Hayley Widden, PhD9, David Kurtz, MD, PhD10, Barbara Pro, MD11
AFFILIATIONS
1 Columbia University Irving Medical Center, New York, New York, United States
2 Columbia University Irving Medical Center, New York, New York, United States
3 University of Colorado, Aurora, Colorado, United States
4 Columbia University Irving Medical Center, New York, New York, United States
5 Columbia University Irving Medical Center, New York, New York, United States
6 Columbia University Irving Medical Center, New York, New York, United States
7 Natera, Inc, Austin, Texas, United States
8 Natera, Inc, Austin, Texas, United States
9 Natera, Inc, Austin, Texas, United States
10 Natera, Inc, Austin, Texas, United States
11 Columbia University Irving Medical Center, New York, New York, United States
ABSTRACT
Introduction
Circulating tumor DNA (ctDNA)-based molecular residual disease (MRD) detection is a clinically validated method of response assessment in diffuse large B-cell lymphoma (DLBCL). Patients (pts) with undetected ctDNA-MRD by phased variant (PV) enrichment and detection sequencing (PhasED-Seq) after 2 cycles of frontline therapy experienced a 2-year PFS rate of 96% (Roschewski JCO 2025). Real-time ctDNA-MRD testing with PhasED-Seq to guide de-escalation of frontline DLBCL therapy has not previously been performed. The benefits to quality of life (QoL) of de-escalation in DLBCL have also not been described.
Methods
SHORTEN-ctDNA is an ongoing single-center investigator-initiated feasibility study (NCT06693830) for pts with newly diagnosed stage II-IV DLBCL receiving standard R-CHOP or pola-R-CHP. Whole blood samples were drawn on cycle 4 day 1 (C4D1) for real-time ctDNA-MRD testing. These samples, alongside biopsy tissue to identify pt-specific PVs, were analyzed with PhasED-Seq (CLARITY, Foresight). PET scans were performed on C4D15 (iPET4). Pts with undetected ctDNA-MRD on C4D1 and complete metabolic response (CMR) on iPET4 received rituximab alone for C5-6. All other pts continued standard therapy. Equivocal iPET4 response could be adjudicated with biopsy for de-escalation eligibility. Exploratory pt-reported outcomes (PRO) analyses were performed.
Results
We report on updated outcomes on the first 16 evaluable pts. Median age was 68 years (40-92), 13 (81%) were stage III-IV, 10 (63%) had elevated lactate dehydrogenase (LDH), 8 (50%) had ≥2 extranodal sites, and 5 (31%) had bulky tumor ≥7.5 cm. The international prognostic index (IPI) was ≥3 for 9 (56%). Four pts received R-CHOP (25%) and 12 pola-R-CHP (75%).
Of the 16 pts, 15/16 (94%) had pt-specific PVs identified, and those 15 had successful real-time ctDNA-MRD testing. Median time from sample collection to results reporting was 8 days (7-11). A median 10 mL plasma was used from each sample, enabling a median empirical sensitivity of 0.96 parts per million (0.38-5.21). The rate of undetected ctDNA-MRD on C4D1 was 10/15 (67%); of these 10 pts, 7 (70%) were stage IV, 6 (60%) had elevated LDH, 5 (50%) had ≥2 extranodal sites, 3 (30%) had bulky tumor, and 5 (50%) had IPI ≥3. There were no significant differences between pts by ctDNA-MRD status. iPET4 was equivocal for 4/10; these 4 pts had biopsies showing inflammation (3) or benign tissue (1) confirming CMR before de-escalation. Residual FDG-avid sites resolved in 3 and persisted in 1 by end-of-treatment (EOT). All 10 pts with undetected ctDNA-MRD at C4D1 de-escalated to rituximab for C5-6, and 5 pts with detected ctDNA-MRD and one with unsuccessful testing continued standard therapy. After a median follow-up of 9.8 months (2.8-18.6), all pts (de-escalated and standard) remain in remission. Additional analysis of follow-up ctDNA-MRD data from banked samples is ongoing and will be presented.
PRO tools were administered on C4D1 and at EOT: EORTC QLQ-C30 (global health status and single summary score, range 0-100; higher scores reflect better QoL), FACIT item GP5 (overall side-effect burden, score 0-4; higher scores reflect greater burden), and a 20-item PRO-CTCAE instrument with composite scores (range 0-3, absent to very severe) calculated for individual adverse events (AE).
From C4D1 to EOT, mean QLQ-C30 global health status scores were stable with de-escalation (75 to 75) but worsened with standard therapy (60 to 54). Mean QLQ-C30 summary scores showed a similar trend (de-escalation: 83 to 82; standard: 78 to 70). Mean FACIT GP5 scores were stable with de-escalation (1.3 to 1.1) but worsened with standard therapy (2.8 to 3.3). Rates of PRO-CTCAE composite scores ≥2 from C4D1 to EOT for de-escalated vs standard pts, respectively, were: dysgeusia (33% to 20% [de-escalated] vs 83% to 60% [standard]), appetite (22% to 20% vs 50% to 40%), constipation (11% to 10% vs 33% to 80%), pain (11% to 20% vs 50% to 40%), headache (0% to 20% vs 17% to 60%), and fatigue (11% to 30% vs 33% to 60%).
Conclusion
Results from the ongoing SHORTEN-ctDNA study suggest that real-time ctDNA-MRD assessment to de-escalate frontline therapy in DLBCL is feasible. Short-term outcomes for the two-thirds of pts with undetected ctDNA-MRD on C4D1 who stopped chemotherapy early are excellent. For pts who underwent de-escalation, QoL remained stable from a high baseline, side effect burden decreased over time, and pt-reported AEs at EOT were mild.
DISCLOSURES
Hua-Jay Jeffery Cherng, MD: Foresight Diagnostics - Honoraria; Natera - Honoraria; ADC Therapeutics - Research Funding; Natera - Consultancy (Includes expert testimony); Foresight Diagnostics - Consultancy (Includes expert testimony)