A53-Adapted Decitabine-Containing R-CHOP in Newly Diagnosed Diffuse Large B-Cell Lymphoma Defined by LymphGen: Preliminary Results of a Prospective Proof-of-Concept Study

1. Introduction

Diffuse large B cell lymphoma (DLBCL) is a heterogeneous group of tumors characterized by broad clinical, morphological, immunophenotypic, and genetic diversity [1]. Modern molecular genetic classifications have enabled stratification of DLBCL into biologically discrete subtypes with differential sensitivity to therapy [2]. Of particular clinical relevance is the A53 subtype as defined by the LymphGen algorithm, whose key molecular hallmarks include inactivation of the TP53 gene (resulting from point mutations or deletions) along with associated genomic instability or aneuploidy [3].

The role of TP53 aberrations as a universal driver of tumor progression and therapeutic resistance is being actively investigated across various B cell neoplasms. Among lymphoproliferative disorders, the prognostic and predictive significance of these alterations has been most thoroughly characterized in chronic lymphocytic leukemia and mantle cell lymphoma. In these entities, TP53deletions and mutations are equally powerful adverse prognostic factors and are tightly linked to primary resistance to immunochemotherapy [4,5]. The molecular landscape of TP53 is remarkably heterogeneous: aberrations may manifest as biallelic disruptions, isolated del(17p), or point mutations, occurring in approximately 60%, 10%, and 30% of cases, respectively [6]. In the current era of targeted agents, the clinical value of identifying TP53 abnormalities has expanded beyond pure prognostication—it has become a predictive biomarker that guides the selection of optimal therapeutic strategies [7]. These observations have provided the rationale for an in depth exploration of TP53 aberrations in other aggressive B cell lymphomas. In DLBCL, loss of TP53function similarly drives primary resistance to standard R CHOP immunochemotherapy, defining a patient subset with an extremely poor prognosis [8,9].

The prognostic impact of TP53 aberrations in DLBCL has been convincingly validated in both retrospective and prospective studies [10,11]. Large cohort studies employing high throughput sequencing have identified TP53 mutations in 20–25% of patients, establishing them as an independent predictor of significantly inferior overall and event free survival [11,3]. Clinically, this failure stems from the inability of conventional regimens to overcome the chemoresistance conferred by TP53 gene aberrations [12]. The key molecular underpinnings of this phenomenon include abrogation of cell cycle checkpoint control, profound disruption of programmed cell death (apoptosis) pathways, and an impaired cellular response to the DNA damaging effects of cytotoxic agents [13]. Collectively, these alterations enable the tumor clone to maintain a high proliferative drive even under the intense cytotoxic stress induced by therapy.

Given the biological features of TP53-aberrant DLBCL outlined above, integration of epigenetic agents into first-line therapy represents a rational strategy to overcome resistance, a concept supported by both preclinical models and early-phase clinical studies in related disease settings [14,15,16]. Decitabine, a DNA methyltransferase inhibitor, can reactivate epigenetically silenced tumor suppressor genes, apoptosis regulators, and major histocompatibility complex molecules, thereby resensitizing malignant cells to cytotoxic insult [17]. An additional, immunologically relevant mechanism of decitabine involves the induction of viral mimicry. Demethylation of endogenous retroviral sequences triggers transcription of double stranded RNA, activates interferon dependent innate immune responses, and may substantially enhance rituximab mediated antibody dependent cellular cytotoxicity [14].

Despite the compelling and consistent biological and preclinical rationale, prospective clinical data on the efficacy and safety of hypomethylating agents combined with R CHOP in patients with TP53-aberrant DLBCL—particularly those with the confirmed A53 molecular subtype—remain extremely scarce. In this context, the present pilot, single center study was designed to assess the complete metabolic response rate and toxicity profile of the molecularly adapted DAC R CHOP regimen in patients with newly diagnosed DLBCL and a verified A53 subtype.

2. Materials and Methods

2.1. Study Design and Patients

The present study was designed as a prospective, single-center cohort analysis. All cases examined herein were originally enrolled in a prospective single-center clinical trial, the design of which was developed in full accordance with the principles of evidence-based medicine and international standards for Good Clinical Practice [18]. The study was conducted in strict accordance with the ethical principles of the Declaration of Helsinki. The protocol underwent independent ethics review and was approved by the local ethics committee of the Moscow Clinical Research Center 52 (Moscow Healthcare Department). Before any study procedures were initiated, all patients received detailed information regarding the objectives and potential risks and provided written informed consent for participation, collection of clinical and anamnestic data, and comprehensive molecular genetic profiling of tumor material.

The overall analytic cohort comprised 70 patients with newly diagnosed diffuse large B cell lymphoma. The diagnosis was confirmed by comprehensive morphological and immunohistochemical analysis according to the current criteria of the 5th edition of the World Health Organization classification of tumors of hematopoietic and lymphoid tissues [19]. Patients were enrolled prospectively and consecutively from September 2023 onward. Key eligibility criteria included: age between 18 and 80 years; Ann Arbor stage I–IV disease; Eastern Cooperative Oncology Group (ECOG) performance status 0–4. Additional considerations included clinical suitability for and feasibility of induction immunochemotherapy with the R CHOP regimen, as well as the absence of absolute contraindications to any component of the backbone R CHOP regimen or to the targeted agents planned for incorporation into the protocol (acalabrutinib, decitabine, and lenalidomide).

2.2. Molecular-Genetic Profiling

Comprehensive molecular genetic profiling of tumor tissue was performed in all patients strictly prior to the initiation of any specific therapy. Genomic DNA was extracted from tumor specimens that had been fixed in 10% neutral buffered formalin and embedded in paraffin (FFPE blocks), or from fresh lymph node biopsy samples, followed by standard quality control assessment of the quantity and integrity of the isolated nucleic acid.

To detect somatic TP53 mutations and concomitant genetic aberrations, we employed targeted next generation sequencing (NGS) using a customized 60 gene panel encompassing: ARID1A, B2M, BTG1, BTG2, CCND3, CD70, CD79B, CIITA, CREBBP, DDX3X, DTX1, DUSP2, EP300, EZH2, FAS, GNA13, IRF4, IRF8, KMT2D, MPEG1, MYD88, NOTCH1, NOTCH2, PIM1, PRDM1, SGK1, SOCS1, STAT3, STAT6, TBL1XR1, TET2, TNFAIP3, TNFRSF14, TP53, ZFP36L1, MTOR, NFKBIA, ETV6, ACTG1, OSBPL10, MYC, BCL2, BCL6, FOXO1, ATM, CD79A, PIK3CD, PTEN, CD5, CD58, CDKN2A, CDKN2B, ASXL1, KRAS, EPHB1, BCL10, PRKCB, PLCG2, CARD11, and MEF2B. Sequencing was carried out to a sufficient depth of coverage to ensure reliable detection of low frequency variants. The targeted sequencing panel incorporated a bioinformatic copy number variation (CNV) analysis module, which enabled verification of structural genomic rearrangements and assessment of genomic instability, both of which are essential for the accurate application of the LymphGen algorithm. Molecular subtyping of DLBCL was performed according to the LymphGen genomic classification algorithm, based on the spectrum of identified genomic aberrations. During bioinformatic analysis, samples were classified as TP53 mutated when pathogenic variants were detected at a variant allele frequency (VAF) of ≥5%, a threshold that allowed exclusion of sequencing artifacts and low level somatic mosaicism.

Based on the comprehensive molecular genetic findings, patients were stratified according to the LymphGen algorithm. Six patients (8.5%) were classified as the A53 molecular subtype; all of these patients carried pathogenic TP53 variants. The remaining 64 patients fell into other molecular DLBCL subtypes.

2.3. Treatment Protocol

The backbone therapeutic approach for the entire patient cohort was a molecularly adapted immunochemotherapy protocol designated R CHOP X, where “X” denotes targeted modifications based on the verified LymphGen subtype. Given the need for prompt initiation of specific antitumor therapy and the inherent technical timelines required for high throughput sequencing, all patients received the first induction cycle as standard R CHOP without the addition of targeted agents.

Starting from cycle 2, after molecular profiling results and mutation status had been obtained, the therapeutic strategy was adapted. To overcome anticipated chemoresistance in patients with TP53mutations (n=6), the hypomethylating agent decitabine was integrated into first line therapy; this experimental regimen was designated DAC R CHOP. Decitabine was administered at a dose of 10 mg/m² as intravenous infusions on days 1 5 of each 21 day cycle, immediately before administration of the backbone regimen components.

Patients without TP53 mutations (n=64) received, from cycle 2 onward, therapy according to the R CHOP X protocol, with the incorporation of acalabrutinib or lenalidomide, strictly in accordance with clinical stage and verified disease subtype. The planned total duration of induction therapy was 6 cycles at 21 day intervals (1 cycle of standard R CHOP followed by 5 cycles of molecularly adapted therapy) for both groups. Dose modifications and clinically justified delays of subsequent cycles were performed as indicated, based on hematologic and non hematologic toxicity as well as individual treatment tolerability.

2.4. Response Assessment and Endpoints

Tumor response monitoring, including interim (after cycles 2 and 4) and final efficacy assessments, was performed using positron emission tomography combined with computed tomography (PET/CT) with the radiotracer ¹⁸F fluorodeoxyglucose. Metabolic response was interpreted according to the Lugano 2014 criteria, with mandatory application of the visual 5 point Deauville scale (DS). Complete metabolic response (CMR) was defined as a Deauville score of 1 to 3.

The primary endpoint of the study was the CMR rate after completion of protocol therapy in the cohort of patients with TP53 mutations who received the DAC R CHOP regimen. Secondary endpoints included overall response rate (ORR, comprising complete and partial metabolic responses), as well as detailed evaluation of the safety profile and incidence of adverse events (AEs). Adverse events were recorded and graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE), version 5.0.

2.5. Statistical Analysis

Statistical processing of clinical and laboratory data was performed using the R computing environment (R Foundation for Statistical Computing). Categorical variables are presented as absolute numbers and relative frequencies (percentages). Continuous variables are described using medians (Me) and interquartile ranges (IQR).

Given the pilot nature of the study and the limited sample size of the target patient group (n=6), the statistical analysis was predominantly descriptive in nature.

3. Results

3.1. Clinical and Molecular Characteristics of Patients and Treatment Efficacy

From September 2023 onward, 70 patients with newly diagnosed DLBCL were consecutively enrolled in the study and underwent comprehensive molecular genetic profiling of tumor tissue prior to the initiation of specific therapy. Based on NGS analysis and subsequent classification according to the LymphGen algorithm, six patients (8.5%) were assigned to the A53 molecular subtype. The group comprised 3 men and 3 women; the median age was 65 years (IQR, 56–67). Advanced stage disease (III–IV by Ann Arbor) was diagnosed in 4 of 6 patients (67%). Four patients (67%) were classified as high risk according to the International Prognostic Index (IPI ≥3).

All identified TP53 variants were classified as pathogenic. Mutations were localized to exons 4, 6, and 7, affecting the DNA binding domain of the p53 protein. The median variant allele frequency (VAF) was 23.5% (IQR, 8%–40%; range, 5%–57%). The p.Arg248Trp mutation (c.742C>T, exon 7) was independently identified in two patients (patients 4 and 6).

A complete metabolic response was achieved in all 6 evaluable patients (100%; 95% CI, 54%–100%).

Individual clinical and molecular characteristics and treatment responses of patients with TP53mutations are presented in Table 1.

3.2. Toxicity Profile

The spectrum and frequency of grade III–IV adverse events are summarized in Table 2. Grade III–IV hematologic toxicity (anemia, neutropenia, and thrombocytopenia) occurred in all patients (100%). Febrile neutropenia, which developed in every case, was successfully managed with broad spectrum antibiotics and granulocyte colony stimulating factor (G CSF) support, thereby allowing treatment to continue without premature discontinuation. In all instances, febrile neutropenia episodes occurred after the second and fourth cycles of therapy, temporally coinciding with the period of decitabine integration into the treatment protocol. During induction therapy, two patients experienced episodes of severe gastrointestinal bleeding. These complications were deemed pathogenetically related to the underlying disease (specific tumor involvement of the gastric wall, tumor lysis), were successfully managed, and did not require premature discontinuation of the protocol. No treatment related deaths were observed.

4. Discussion

In this prospective pilot study, the molecularly adapted DAC R CHOP strategy in patients with newly diagnosed DLBCL and a verified A53 subtype was associated with a high rate of complete metabolic response and a manageable toxicity profile. A complete metabolic response was achieved in all 6 evaluable patients, suggesting potential clinical activity of this approach in a molecularly defined patient population with the A53 subtype. This finding is particularly noteworthy because the cohort comprised patients with an inherently unfavorable molecular profile. These results support the biological rationale for the combination under investigation, although confirmation in controlled studies with larger sample sizes is warranted. The toxicity profile of the regimen supports its consideration as a platform for further development and clinical evaluation.

TP53 gene mutations occupy a leading position in the overall landscape of genetic events in oncogenesis and are detectable in approximately half of all human malignancies. The protein encoded by this gene, p53—often referred to as the “guardian of the genome”—orchestrates the cellular response to genotoxic stress through transcriptional regulation of target genes that control cell cycle arrest, DNA repair, apoptosis, and cellular senescence [13]. Loss of p53 function resulting from point mutations, deletions, or epigenetic silencing eliminates a critical barrier to tumor progression and confers selective advantages to the tumor clone under cytotoxic stress. Nevertheless, the prognostic and predictive value of TP53 aberrations varies substantially by disease entity, molecular context, and the role of this genetic event within the hierarchy of oncogenic drivers in a given tumor. The lack of a clear understanding of whether TP53 mutation represents an initiating or a late acquired event in tumor evolution precludes consideration of TP53-mutated status as a universal adverse prognostic factor across all neoplasms.

Among lymphoproliferative disorders, the most compelling evidence base regarding the clinical significance of TP53 aberrations has been accumulated in chronic lymphocytic leukemia and mantle cell lymphoma. In these entities, TP53 mutations and deletions typically represent a late genetic event acquired during clonal evolution under therapeutic selective pressure, and their detection is associated with primary resistance to standard immunochemotherapy, early relapse, and poor prognosis [4,5]. In clinical practice, assessment of TP53 aberrations is mandatory both at initial diagnosis and at relapse in chronic lymphocytic leukemia and mantle cell lymphoma, and their verification serves as an indication for targeted agents and cellular therapy as alternatives to conventional cytotoxic regimens [20,21].

In aggressive B cell lymphomas, and particularly in DLBCL, the landscape is considerably more complex and nuanced, a reflection of the marked molecular heterogeneity that characterizes this group of tumors [1]. In contrast to chronic lymphocytic leukemia and mantle cell lymphoma, TP53mutations in DLBCL are frequently detectable at initial diagnosis and may represent either early initiating events or late acquired aberrations. The clinical significance of TP53 mutations in DLBCL is substantially modified by the molecular context of the tumor, including concomitant genetic events, cell of origin, and the presence or absence of biallelic gene inactivation [3,22]. This biological complexity has driven the incorporation into contemporary genomic classifications of a distinct molecular subtype, A53, which encompasses DLBCL cases with TP53 inactivation and associated features of genomic instability, thereby enabling more precise patient stratification and the assembly of biologically homogeneous cohorts for clinical studies [2].

Accumulating evidence has shown that TP53 aberrations in DLBCL are associated not only with adverse prognosis but also with the biology of early treatment resistance [8]. Among patients who fail to respond to first line therapy, TP53 mutations are identified at the highest frequency and frequently co occur with alterations in chromatin remodeling processes, including defects in ARID1A and other epigenetic modifiers [23,24]. In this context, our observations contribute to a growing body of evidence supporting the need for biologically guided modifications to first line therapy in patients with TP53 aberrant DLBCL. Unlike retrospective studies, the prospective application of the LymphGen algorithmic classification enabled the identification of a biologically defined A53 group, characterized not only by TP53 aberrations but also by concomitant features of genomic instability. This approach ensures a more homogeneous molecular population and enhances the biological interpretability of our findings compared with studies based solely on assessment of TP53 mutation status without consideration of the tumor molecular context [9]. Nevertheless, in the absence of randomized control, the observed clinical effect cannot be unequivocally attributed to the addition of decitabine, as it may reflect epigenetic sensitization, the intrinsic sensitivity of the selected subpopulation, or a combination of both.

The potential mechanisms of decitabine activity in the setting of TP53 mutations encompass several synergistic biological processes. Inhibition of DNA methyltransferases by decitabine induces reactivation of epigenetically silenced tumor suppressor genes and regulators of apoptosis, thereby partially restoring cell cycle control even in the presence of functional p53 inactivation [17]. This effect is complemented by demethylation of major histocompatibility complex class I molecules on the tumor cell surface, which potentiates their immunological recognition by the T cell system. Critically, induction of the viral mimicry phenomenon also plays a key role: demethylation of endogenous retroviral sequences (ERVs) with subsequent activation of interferon dependent signaling pathways significantly enhances rituximab mediated antibody dependent cellular cytotoxicity and activates innate antitumor immune responses [25,14]. The aggregate of these mechanisms provides a rational biological foundation for combining decitabine with rituximab containing immunochemotherapy specifically in the patient population with TP53 aberrant DLBCL [15]. At the same time, to confirm the engagement of these mechanisms within the framework of the DAC R CHOP protocol, direct translational studies incorporating mandatory biomarker analyses are required, including assessment of global demethylation levels, expression of ERV transcripts, and parameters of interferon responses in tumor tissue and peripheral blood of patients.

Alongside the biological rationale and the emerging data on the therapeutic potential of the DAC R CHOP regimen, tolerability is a critical determinant of its practical applicability. The toxicity profile of DAC R CHOP in this cohort was as expected and broadly consistent with the characteristics anticipated for anthracycline containing immunochemotherapy regimens. Hematologic toxicity was universal in nature but was effectively managed through mandatory primary prophylaxis with G CSF. The addition of decitabine to the R CHOP backbone resulted in a predictable and clinically meaningful increase in the rate of febrile neutropenia, reaching 100%. This degree of myelosuppression exceeds historical data for standard R CHOP [26,27] and reflects the expected synergism between the cytotoxic and hypomethylating components. Accordingly, the regimen is associated with a high rate of severe hematologic toxicity that necessitates mandatory primary G CSF prophylaxis and intensive clinical monitoring. Nevertheless, the complications that arose were effectively managed with broad spectrum antibacterial therapy and supportive measures. In no case did dose reductions or critical delays in therapy cycles become necessary, and no fatal infectious complications were recorded. The severe gastrointestinal bleeding events observed in 2 patients during induction therapy warrant particular attention. These complications were attributable to local lysis of tumor substrate within the gastric wall, consistent with published data on the risk of hemorrhagic manifestations during effective treatment in patients with baseline gastrointestinal involvement [28,29,30]. Importantly, both episodes were successfully managed in a specialized hospital setting without premature protocol discontinuation, underscoring the manageability of this adverse event under appropriate clinical monitoring.

The results of this study must be interpreted in light of several important limitations. The small size of the analytic cohort (n=6) is the principal limiting factor: it reduces the precision of the true response rate estimate and substantially constrains statistical power for the identification of response predictors and rare adverse events. The absence of a control group precludes definitive attribution of the observed clinical effect to the addition of decitabine and prevents exclusion of the contribution of inherent biological features of the selected subpopulation. The short follow up period precludes assessment of response durability and the impact of this strategy on long term survival outcomes. Biological heterogeneity within the TP53 mutant group, including variability in mutation location across functional protein domains and differing allelic burden, assumes particular significance in the context of a small sample, wherein the presence of biologically unique cases may disproportionately influence aggregate results [31,10]. Finally, the single center design limits the generalizability of our findings to the broader population of patients with the A53 subtype of DLBCL.

Notwithstanding these limitations, our data demonstrate the clinical feasibility and preliminary efficacy of the DAC R CHOP regimen in patients with the A53 subtype of DLBCL and provide a rationale for conducting multicenter randomized trials with adequate statistical power and an integrated translational component, aimed at confirming the contribution of epigenetic sensitization and identifying molecular biomarkers of response. Further development of molecularly adapted strategies for A53 variants of DLBCL emerges as a priority direction in overcoming primary resistance to immunochemotherapy and improving long term outcomes in this patient population.

5. Conclusions

Molecular genetic stratification of DLBCL using the LymphGen classification identifies the A53 subtype, which is associated with an extremely poor prognosis under standard therapy. In this prospective pilot study, the molecularly adapted DAC R CHOP regimen demonstrated marked antitumor activity, with complete metabolic response achieved in all patients and a clinically manageable toxicity profile. Despite the limitations imposed by the small sample size, absence of a control group, and short follow up, these findings support further clinical evaluation of this approach in larger prospective multicenter trials incorporating translational biomarkers of epigenetic sensitization.