Harnessing Ctdna to Predict Therapeutic Success in DLBCL: Insights from a Multicenter Prospective Study

Background:

Diffuse large B-cell lymphoma (DLBCL), the most prevalent aggressive form of non-Hodgkin lymphoma, presents challenges in post-treatment surveillance due to the relapse risk. The unique genetic profile of each patient's lymphoma, characterized by specific immunoglobulin (Ig) gene rearrangements, offers a potential tumor-specific marker for MRD detection. This prospective multicenter study aims to evaluate ctDNA analysis with Ig gene sequencing for MRD monitoring in DLBCL patients. By assessing the sensitivity and specificity of this approach in detecting MRD and predicting clinical outcomes, we initiated a “Multi-center Study of DLBCL MRD Surveillance with ctDNA-IgNGS” (MeDIG Study, ChiCTR2400087307) to provide insights into the potential of ctDNA-based MRD monitoring for personalized treatment strategies in DLBCL.

Methods:

This study enrolled untreated, newly diagnosed DLBCL patients and relapsed/refractory DLBCL patients who were evaluated for hematopoietic stem cell transplantation or CAR-T therapy. For this report, only the data from the newly diagnosed cohort will be presented. Tumor specimens taken prior to treatment identified patients' specific Ig VDJ rearrangement biomarkers. Peripheral blood were collected for MRD detection before each cycle of chemotherapy, at the end of treatment and every three months for 2 years post-treatment. Imaging evaluation (PET-CT or contrast-enhanced CT) were performed every 3-6 months. The NEO-MRD assay (NeoImmune, Shenzhen, China) employs a technique that involves multiplex PCR followed by NGS to detect and monitor rearrangements in the IgH, V-J, D-J, and IgK/L loci.

Results:

141 patients met the inclusion criteria and had eligible tissue samples included out of 201 patients enrolled. The median age of the evaluable cohort was 55 years, comprising 48.8% males and 56.4% with stages III-IV disease. 62.6% of the patients had IPI scores ranging from 0 to 2, and 38.7% were identified as the GCB subtype. Treatment regimens administered included RCHOP/RCHOP-based chemotherapy in 61% of patients, REDOCH/REDOCH-based chemotherapy in 39%,.

We identified trackable clones in 91% of analyzed tumor tissues, with a mean of three clones per sample, largely represented by the IGH and IGK loci. A significant majority of IGH clones, amounting to 82%, exhibited mutation frequencies exceeding 5%. There was a statistically significant correlation between IGH mutation rates and variables including patient age, tumor stage, and COO subtype (P<0.05). Furthermore, analysis of tumor clone evolution revealed more evolved tumor clones correlate with worse clinical outcome(P=0.025).

Stable tumor ctDNA was detected in 72% of pre-treatment baseline plasmas, and the ctDNA copy number in baseline plasma exhibits a significant correlation with mid-treatment response as assessed by CT. Plasma MRD levels after the first cycle of treatment were found to predict mid-therapy response with a sensitivity of 100% and a specificity of 54%, yielding a positive predictive value of 23% and a negative predictive value of 95%. For predicting end-of-treatment outcomes, the MRD levels demonstrated a sensitivity of 100%, specificity of 53%, PPV of 20%, and NPV of 100%. Second-cycle MRD levels predicted mid-treatment response (83% sensitivity, 73% specificity, 31% PPV, and 93% NPV) and end-of-treatment outcomes (60% sensitivity, 68% specificity, 20% PPV, and 93% NPV).

Conclusion:

This study demonstrates the potential of ctDNA analysis with Ig sequencingfor MRD monitoring in DLBCL. The high sensitivity and negative predictive value of ctDNA-based MRD detection for predicting both mid-therapy and end-of-treatment outcomes underscore its utility in personalized treatment strategies. Tracking specific Ig gene rearrangements allows real-time tumor response monitoring and relapse risk assessment. Although broader studies are needed , current data hints that ctDNA-based MRD monitoring might substantially upgrade post-treatment monitoring and could direct treatment approaches in DLBCL.

No relevant conflicts of interest to declare.