Quantitative Baseline Circulating Tumor DNA Levels Correlate with GM-CSF Response to Idiotype Vaccine in Untreated Mantle Cell Lymphoma

Background: Mantle cell lymphoma (MCL) is an incurable lymphoma that responds to various forms of immunotherapy. Previously, we reported that specific anti-idiotype tumor responses as measured by GM-CSF levels after vaccine were associated with overall survival (OS) and time to next treatment (TTNT) following DA-EPOCH-R in untreated MCL at 11 years of follow-up (Dunleavy ASCO 2012). Levels of GM-CSF prior to idiotype vaccine also were correlated with post-vaccine GM-CSF levels suggesting that pre-existing anti-tumor immune response predict idiotype vaccine responsiveness. High-throughput DNA sequencing methods can detect and quantify circulating tumor-specific DNA (ctDNA) in peripheral blood of MCL patients prior to therapy and can be monitored for clearance after therapy. We hypothesized that pre-treatment ctDNA levels may be a surrogate marker for ongoing host anti-tumor immunity. Here, we analyzed the relationship of pre- and post-treatment levels of ctDNA to predict GM-CSF response and clinical outcomes in untreated MCL.

Methods: DA-EPOCH-R was administered x 6, followed by 5 cycles of Id-vaccine beginning at least 12 weeks later in untreated MCL. Id protein was produced by hybridoma technology, conjugated to keyhole limpet hemocyanin (KLH), and administered together with GM-CSF x 5 over 6 months. Pre- and post-vaccine samples were tested in parallel to assess humoral and cellular immune responses. Pre-treatment formalin-fixed paraffin embedded tissue specimens were de-identified and sent for identification of tumor-specific clonotypes. Tumor DNA was amplified using locus-specific primer sets for the immunoglobulin heavy-chain locus (IGH) complete (IGH-VDJH), IGH incomplete (IGH-DJH), and immunoglobulin k locus (IGK). Amplified products were sequenced and tumor-specific clonotypes were quantitated in peripheral blood mononuclear cells before treatment as ctDNA and after treatment as MRD.

Results: Characteristics of all 26 pts: median age 57 (r 22-73), male sex 73%, PS 1 (0-2), MIPI (low-65%; interm-16%; high-19%), and blastoid 15%. Responses to DA-EPOCH-R: CR-92%, PR-8%. Immune analyses were performed in 24 pts; vaccine not produced in 1 pt and 1 pt progressed before immune analyses. Baseline tumor was available for clonotype calibration in 20 patients. With a potential follow-up of 14 years, the median OS of the entire cohort is 9.0 years and 10/26 (38%) patients are still alive. Pre-treatment levels of ctDNA were inversely correlated with normalized antitumor GM-CSF response (r=0.49, p=0.035) but not with white blood cell count (r=0.18, p=0.46), LDH (r=0.21, p=0.39) or MIPI scores (r=22, p=0.37). Patients with low pre-treatment ctDNA levels had longer OS than patients with high levels of ctDNA, but this did not reach statistical significance (11.01 y vs. 5.71 y, p=0.67). Eight of 20 of patients (40%) were MRD positive after DA-EPOCH-R despite achieving CR/CRu. Patients who achieved MRD-zero had a trend towards improved OS (NR vs. 8.2 years, p=0.15).

Conclusions: Lower pre-treatment ctDNA levels correlate with normalized GM-CSF response to idiotype vaccine and may be a surrogate marker of ongoing host anti-tumor immunity. In this small study, ctDNA levels did not correlate with traditional markers of tumor proliferation. There is a trend towards improvement in OS in patients who achieve MRD-zero. Future studies are underway to explore the biologic significance of pre- and post-treatment ctDNA levels in MCL.

This work was supported by the Intramural Research Program of NCI at NIH and Adaptive Biotechnologies, Inc.