Editor's Note: At a highly anticipated hematologic oncology conference, the Late-Breaking Abstract (LBA) session became a focal point. Co-chaired by Professor Kim Linton from the University of Manchester and Professor Gilles Salles from Memorial Sloan Kettering Cancer Center in New York, the session brought together top clinical and research experts from around the globe. One of the core topics of the conference was how to move beyond the twenty-year-old R-CHOP regimen in the front-line treatment of Diffuse Large B-cell Lymphoma (DLBCL). Dr. Eleonora Calabretta from Harvard University, on behalf of her collaborative team, presented a deep retrospective analysis based on POLARIX clinical trial data, offering revolutionary new insights into the precision treatment of DLBCL.POLARIX Trial: Breaking the Two-Decade Standoff in First-Line DLBCL Treatment
Over the past two decades, numerous clinical trials have attempted to improve the prognosis of DLBCL patients by adding new targeted agents to the R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) regimen, but none achieved their primary endpoints. The POLARIX study is the first Phase III randomized controlled trial to achieve a historic breakthrough. This study confirmed that replacing vincristine in the R-CHOP regimen with the antibody-drug conjugate (ADC) Polatuzumab Vedotin, forming the Pola-R-CHP regimen, significantly improves patients’ progression-free survival (PFS). Dr. Calabretta first reviewed in her report: “POLARIX trial data showed that the two-year PFS rate in the Pola-R-CHP treatment group reached 76.7%, significantly outperforming the R-CHOP group’s 70.2%, and recently published data further confirmed this advantage persists at five years.” Furthermore, an exploratory subgroup analysis initially suggested that this benefit might be more pronounced in the transcriptionally defined Activated B-cell like (ABC) subtype, laying the groundwork for subsequent deeper molecular subtyping studies.
DLBclass Classifier: Moving Beyond Traditional Subtyping for Precise Genomic Profiling
DLBCL is a malignant tumor with high genetic heterogeneity. To stratify it more precisely, Dr. Calabretta’s team previously developed a molecular classification system called “DLBclass.” This system categorizes DLBCL into five molecular subtypes (Cluster 1 to Cluster 5), each with unique genetic features and clinical prognoses.
“To achieve comprehensive genetic analysis, we must integrate all types of genetic alterations, including mutations, copy number variations, and structural variations,” emphasized Dr. Calabretta. The DLBclass classifier is based on this principle, employing an advanced neural network-based algorithm to analyze 163 genomic features, capable of accurately assigning each tumor sample to a specific molecular subtype, with an overall accuracy of up to 90%, and reaching 97% in high-confidence samples (approximately 75% of samples).
Among these, “Cluster 5,” which is enriched with ABC-like features, is particularly noteworthy. This subtype is characterized by increased BCL2 gene copy number and activating mutations in MYD88 and CD79B genes, and it exhibits the poorest prognosis when treated with R-CHOP, showing a unique propensity for extranodal involvement.
Core Finding: Pola-R-CHP Therapy Delivers Maximum Benefit to “Cluster 5” DLBCL Patients
The core objective of this study was to retrospectively analyze 407 patients from the POLARIX trial with available biopsy samples, using the DLBclass classifier to determine if the efficacy of the Pola-R-CHP regimen differed across various molecular subtypes. The study results clearly revealed the population that benefits most from Pola-R-CHP therapy. In “Cluster 5” DLBCL patients, the five-year PFS rate in the Pola-R-CHP group was as high as 70%, compared to only 42% in the R-CHOP group, showing an extremely significant difference in efficacy (hazard ratio [HR] = 0.38, p = 0.005). Dr. Calabretta noted: “This finding is particularly important because it indicates that among all subtypes, Cluster 5 patients treated with R-CHOP have the poorest prognosis. The addition of Pola-R-CHP appears to reverse this intrinsically poor prognosis.” In stark contrast, no statistically significant differences in PFS curves were observed between the two treatment regimens in the other four molecular subtypes.
To validate the robustness of this finding, the research team conducted a multivariate analysis, and the benefit for “Cluster 5” patients from Pola-R-CHP treatment remained significant (HR = 0.41) after adjusting for important clinical and biological factors such as the International Prognostic Index (IPI), age, and traditional cell-of-origin (ABC/GCB) classification. This result strongly demonstrates that the molecular subtyping provided by DLBclass has stronger predictive value than traditional ABC/GCB subtyping. Furthermore, the analysis also showed that within the “Cluster 5” subtype, the efficacy advantage of Pola-R-CHP was consistent regardless of whether patients carried CD79B or MYD88 mutations.
Expert Commentary and Outlook on Clinical Application
In the subsequent Q&A session, conference chair Professor Gilles Salles posed a key question: traditional ABC-classified patients account for approximately 33%, while “Cluster 5” patients in this study account for about 25%; how to explain the relationship between the two. Dr. Calabretta responded: “Cluster 5 is indeed an ABC-enriched subtype, but not all Cluster 5 patients are of the ABC type. Our classifier can go beyond the cell-of-origin labels to precisely identify the biological entities most sensitive to Pola-R-CHP.” Another chair, Professor Kim Linton, asked from a clinical practice perspective how doctors should select patients when genomic sequencing is not yet widely available. Dr. Calabretta opined: “As we move towards an era of treatment decisions based on molecular features, comprehensive and accurate molecular subtyping is crucial. The future direction will inevitably be to promote the integration of such advanced diagnostic technologies into clinical routine.” Regarding the question of Overall Survival (OS), the research team stated that although a statistically significant OS advantage has not yet been observed in the “Cluster 5” subtype, a positive trend towards extended survival has been shown.
Conclusion: Ushering in a New Era of Molecularly Driven DLBCL Treatment
The release of this significant study marks a transition in DLBCL treatment from a “one-size-fits-all” standardized approach to a new stage of individualized therapy based on precise molecular subtyping. DLBclass not only provides a powerful tool for understanding the biological complexity of DLBCL but, more importantly, demonstrates its immense potential as a predictive biomarker to precisely identify the patient group that will benefit most from the breakthrough Pola-R-CHP therapy. As Dr. Calabretta mentioned in her summary, these findings strongly support the use of DLBclass in designing molecularly driven clinical trials and guiding future clinical practice. This achievement is not only a deep excavation of the value of the POLARIX trial but also contributes valuable insights and direction to the global lymphoma research field.
Source of Contribution/Interview: Oncology Outlook – Oncology Times
