AUTHORS
Harris Ebrahim, Kristena Yossef, MD2, Christopher D'Angelo, MD, MS3, Matthew A. Lunning, DO, FACP4, Julie Vose, MD5, James Olen Armitage, MD6, Snegha Ananth, MD7
AFFILIATIONS
1 University of Nebraska Medical Center, Omaha, United States
2 University of Nebraska Medical Center, Omaha, United States
3 University of Nebraska Medical Center, Omaha, United States
4 University of Nebraska Medical Center, Omaha, United States
5 University of Nebraska Medical Center, Omaha, United States
6 University of Nebraska Medical Center, Omaha, United States
7 University of Nebraska Medical Center, Omaha, United States
ABSTRACT
Introduction:
Circulating tumor DNA (ctDNA)-based measurable residual disease (MRD) assessment is recommended by NCCN for equivocal end-of-treatment PET/CT in frontline diffuse large B-cell lymphoma (DLBCL) when biopsy is not feasible. Tumor-informed MRD assays require 4–6 weeks of upfront assay development before serial blood monitoring. Delays may compromise timely MRD availability, yet this workflow has not been systematically evaluated. We evaluated real-world adoption, workflow efficiency, operational barriers, and turnaround times following implementation of an institutional clonoSEQ®-based MRD workflow.
Methods:
We retrospectively identified newly diagnosed DLBCL patients treated at the University of Nebraska Medical Center between January 1, 2024, and December 31, 2025, following implementation of an institutional standard practice for upfront clonality testing. Patients were classified into MRD-attempted (clonality assay ordered) and MRD-not-attempted cohorts. We evaluated MRD utilization, assay success, workflow intervals, turnaround times, and MRD availability at clinically relevant decision points.
Results:
Among 112 eligible patients, MRD testing was attempted in 67 (60%) and not attempted in 45 (40%). Patients who underwent MRD testing lived closer to the institute than those without MRD testing (median 11.7 vs 72.5 miles; P<0.001), whereas inpatient versus outpatient diagnosis, diagnosis-to-treatment interval, biopsy location (outside vs same institution), and insurance status were similar between groups. Assay development succeeded in 60/67 (90%); failures included polyclonality (n=4), insufficient tissue (n=2), and QC failure (n=1).
Workflow intervals included: pre-analytical (clonality order to tissue submission), analytical (tissue submission to assay development), and post-analytical (MRD blood draw to result reporting). Median intervals were 6 (IQR, 3–9), 11 (IQR, 9–12), and 9 (IQR, 7–11) days, respectively. Tissue submission occurred within 14 days in 57/67 (85%) patients (51% within 7 days; 7% >28 days). Assay development was completed within 14 days in 59/60 (98%), and MRD reporting within 14 and 28 days in 48/54 (89%) and 54/54 (100%) patients, respectively.
Median brain-to-clonality interval, defined as the time from the initial lymphoma consultation to clonality order, capturing the earliest clinician-controlled step was 20 days (IQR,4–112; range,0–239). Only 6/67 (10%) orders were placed on the day of consultation, whereas 22/67 (36%) were placed >90 days later, potentially delaying MRD availability.
Among 49 patients with both end-of-treatment PET/CT and MRD assessment, the median interval between PET/CT and MRD result availability was 10 days (IQR 7–20; −13 to 155). Eight patients (16%) experienced delays >60 days, driven by prolonged pre-analytical intervals, including Brain-to-Clonality (median 121 days; range, 8–239) and tissue submission (median 128.5 days; range, 11–260), whereas laboratory turnaround remained constant (median 8.5 days, 7-15).
Among 102 patients undergoing end-of-treatment PET/CT, 29 (28%) had a non-complete metabolic response. Concurrent MRD was available in 15 (52%); all 4 patients with detectable MRD progressed, whereas none of the 11 with undetectable MRD had progressed after a median follow-up of 15 months, although interpretation is limited by the small sample size.
Conclusions:
Given the limited positive predictive value of end-of-treatment PET/CT and the frequency of non-complete metabolic responses, optimizing workflows to ensure timely MRD availability may be as important as improving assay performance. Our findings suggest that the principal opportunity to improve tumor-informed ctDNA implementation lies in the pre-analytical phase. Standardized pathology-driven submission of diagnostic tissue for baseline clonality assessment may increase MRD uptake while ensuring timely result availability across tumor-informed ctDNA assays. Preliminary outcomes suggest favorable prognostic discrimination of detectable versus undetectable MRD in patients with non-complete metabolic response, although longer follow-up is needed.
DISCLOSURES
Harris Ebrahim: No relevant conflicts of interest to declare.