Circulating Tumor DNA As a Predictor of Survival of DLBCL Patients in a Daily Practice

Introduction:

Circulating tumor DNA (ctDNA) could be used as a sensitive biomarker of treatment response and outcomes in patients (pts) with lymphomas including diffuse large B-cell lymphoma (DLBCL). However, there is still limited data on the prognostic significance of ctDNA in a real-world setting. We aimed to assess the prognostic significance of ctDNA levels at diagnosis, interim restaging, and end of treatment (EoT), and its correlation with response and outcome of pts treated in a daily practice.

Methods:

Pts with previously untreated systemic DLBCL who signed the informed consent with available baseline ctDNA and planned R-CHOP therapy were eligible for the study. Altogether 32 DLBCL pts entered the analysis. Out of them, 24 pts were evaluable for ctDNA at interim restaging (after 3 cycles), and 23 pts at the EoT. ctDNA plasma levels were assessed using a custom panel covering coding regions of 538 genes most frequently altered in lymphomas. DNA alterations were identified using next generation sequencing (NGS)-based methodology of CAPP-Seq (CAncer Personalized Profiling by deep Sequencing). Following standard NGS data pre-processing, DNA variants were identified using VarScan 2 using buccal swabs as a source of germline DNA for more streamline variants calling. To calculate the level of plasma ctDNA, average variant allele frequency was used to determine the ctDNA fraction of extracted cell free DNA. ctDNA levels are reported as human haploid genome equivalents (hGE) per ml of plasma. ctDNA levels at each timepoints as well as its dynamics were compared with baseline characteristics and response rates using PET/CT scan and evaluation based on Deauville score (overall response rate, ORR; complete remission rate, CR), and survival (progression-free survival, PFS; overall survival, OS).

Results:

Among 32 included pts, median age at diagnosis was 65.5 years (40-80), clinical stage III-IV was observed in 81%, more than one extranodal involved sites in 59%, PS ECOG 2-4 in 19%, elevated levels of serum LDH in 81%, International Prognostic Index (IPI) 3-5 in 60%, non-GCB DLBCL in 53%. Median number of administered R-CHOP cycles was 6 (range 4-8), with 81% of pts completing the intended treatment. At interim and EoT, the ORR/CR rates were 84%/25% and 88%/72%, respectively. Median follow-up of all pts was 3.5 years. The 2-year PFS was 69%, and the 2-year OS 84%.

At diagnosis (n = 32), median plasma ctDNA level was 1307 hGE/ml. Using the most sensitive threshold of 5000 hGE/ml, pts with lower ctDNA levels (n = 23) had significantly better EoT ORR (96% vs 67%, P = 0.026) in comparison to those with higher ctDNA levels (n = 9), as well as significantly improved PFS (2-year PFS 83% vs 33%, HR 0.27, P = 0.012) with a trend towards better OS (87% vs 67%, HR 0.36, P = 0.125). Multivariate analysis of dichotomized ctDNA levels together with IPI showed that the ctDNA level at diagnosis is an independent prognostic factor for PFS.

At interim (n = 24), 9 pts had undetectable minimal residual disease (MRD). The MRD negative pts presented with a trend towards improved PFS (89% vs 53% at 2 years, HR 0.30, P = 0.1007), with a significantly improved OS (100% vs 67% at 2 years, P = 0.034). Pts with > 3.0-log decrease of ctDNA from baseline (n = 15) had improved ORR (100% vs 67%, P = 0.042) in comparison to those with < 3.0-log decrease (n = 9).

At EoT (n = 23), 13 pts remained MRD positive, and 10 pts achieved MRD negativity with an improved PFS (90% vs 46% at 2 years, HR 0.26, P = 0.0650) and significantly better OS (100% vs 69% at 2 years, P = 0.032). Among pts who achieved metabolic CR (n 16), there were 9 with negative MRD and 7 remained MRD positive, but the PFS and OS was not statistically different for the subgroups of CR pts. Pts with > 3.0-log decrease of ctDNA (n = 14) from baseline presented with improved CR (86% vs 44%, P = 0.036) in comparison to those with < 3.0-log decrease (n = 9).

Conclusion:

Even on a limited number of real-world-treated DLBCL pts, ctDNA evaluation confirmed strong independent prognostic significance of ctDNA levels at the time of Dg and the clearance of ctDNA (MRD negative) as well as its dynamics clearly correlate with survival and the treatment response.

First two authors contributed equally. Supported by MH CR (NU21-03-00411, DRO - VFN00064165), National Institute for Cancer Research (EXCELES - LX22NPO5102), MEYS CR via Charles University (Cooperatio program - Biology, Hematology-Oncology, SVV 260637).

Vodicka:Hoffmann-La Roche: Consultancy, Honoraria, Speakers Bureau; AbbVie: Consultancy; SwixxBiopharma: Consultancy. Polgarova:Novartis: Consultancy, Honoraria; Gilead: Consultancy, Honoraria; Kyowa Kirin: Consultancy, Honoraria; Roche: Honoraria. Klener:Gilead: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Travel Support, Speakers Bureau; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Roche: Research Funding; Lilly: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Sobi: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Trneny:Gilead Sciences, Takeda, Bristol-Myers Squibb, Roche, Janssen, Abbvie, SOBI: Other: Travel, Accommodations, Expenses; Janssen, Gilead Sciences, Takeda, Bristol-Myers Squibb, Amgen, Abbvie, Roche, MorphoSys, Novartis, SOBI, Swixx BioPharma: Honoraria; Takeda, Bristol-Myers Squibb, Incyte, Abbvie, Amgen, Roche, Gilead Sciences, Janssen, MorphoSys, Novartis, Genmab, SOBI, Autolus, Caribou Biosciences: Consultancy.