Editor’s Note: The 2025 European Society for Medical Oncology (ESMO) Annual Congress was grandly held in Berlin, marking another milestone in global oncology research. In the field of hematologic malignancies, several breakthrough studies drew significant attention — ranging from explorations into the mechanisms of CAR-T resistance and the search for new therapeutic breakthroughs, to the innovative development of dual-target therapies that promise more precise and effective treatment options; from precision drug discontinuation studies in multiple myeloma aimed at improving long-term survival and quality of life, to single-cell mapping that provides novel insights into hematologic disease mechanisms. These studies collectively reflect the relentless advancement of medical science and innovation, bringing new hope to the diagnosis and treatment of blood disorders. To spotlight these important breakthroughs and foster academic exchange, Oncology Frontier – Hematology News invited Professor Jiesong Wang from the Fujian Cancer Hospital to provide a professional interpretation of one of the key studies — discussing its innovation, clinical significance, and implications for future research and clinical practice.1243MO: Integrating Interim 18F-FDG PET-CT Analysis and NCCN-IPI for Risk Stratification in Diffuse Large B-Cell Lymphoma — A Multicenter Retrospective Study
Q1
This study integrates interim National Comprehensive Cancer Network International Prognostic Index (iNCCN-IPI) data with PET-CT biomarkers to construct the iPET-NCCN-IPI model. Compared with traditional prognostic assessment based solely on the Deauville score (DS), what are the key findings of this research?
Professor Jiesong Wang:
For diffuse large B-cell lymphoma (DLBCL), especially in interim prognostic evaluation after four cycles of treatment, the Lugano criteria based on the five-point Deauville scoring system (DS) of 18F-FDG PET-CT is widely used in clinical practice. However, this system has certain limitations — notably, a relatively high false-positive rate. To address this, our study integrated clinical and pathological characteristics with interim PET-CT parameters, including the Deauville score (DS), the change in total lesion glycolysis (ΔTLG), the change in maximum standardized uptake value (ΔSUVmax), and interim abdominal residual disease (iARD). These parameters were combined with NCCN-IPI to form a new prognostic assessment system — the iPET-NCCN-IPI model. Compared with using the Deauville five-point method or NCCN-IPI alone, this integrated model provides more accurate interim prognostic stratification for DLBCL patients. For instance, patients identified as low risk by this model had better outcomes than those classified as DS 1–3 (typically complete responders) under the conventional method, while patients categorized as high risk by the model had worse outcomes than those with DS 4–5 scores using the Deauville method.
Q2
If the model reclassifies patients with a Deauville score (DS) of 3 as high risk, how should clinicians interpret and manage this in practice? Would intensified therapy be recommended for these patients?
Professor Jiesong Wang:
Under current practice, patients with a DS score of 3 are generally considered to have achieved complete response (CR). However, our study’s model indicates that over half of these patients would actually be categorized as high risk. Survival curve analyses confirmed that patients identified as high risk by this model indeed had worse outcomes than those in the low-risk group. One of the most important contributions of this model is its ability to further stratify DS 3 patients — dividing them into distinct high- and low-risk subgroups. For those identified as high risk by the model, our results showed poorer prognosis. However, at this stage, it is premature to recommend treatment intensification based solely on this finding. This is because our study is a multicenter retrospective analysis, which cannot fully eliminate data bias. Our next step is to conduct prospective studies to validate the model’s applicability and methodological robustness. If consistent results are obtained in prospective cohorts, we may then consider intensified or alternative therapeutic approaches for these high-risk patients — such as incorporating small-molecule targeted agents based on molecular subtypes, or introducing bispecific antibodies or CAR-T therapy earlier in the treatment course. The goal is to improve outcomes and potentially achieve cures for patients who might otherwise be refractory — a direction that will guide our future research.
Q3
How practical is the model for use in real-world clinical settings? How can it be adjusted to maximize its clinical utility?
Professor Jiesong Wang:
The iPET-NCCN-IPI model is built primarily on parameters readily available in routine clinical practice — interim PET-CT metrics such as DS, ΔTLG, ΔSUVmax, and iARD, in combination with the NCCN-IPI score. These are standard, non-invasive indices that require no additional specialized testing, making the model highly feasible for everyday clinical implementation. Moreover, integrating molecular biology techniques — such as next-generation sequencing (NGS) or molecular profiling — with the iPET-NCCN-IPI model could further refine DLBCL risk stratification. However, technologies like circulating tumor DNA (ctDNA) testing and NGS are not yet standard clinical procedures and would significantly increase patient costs. Therefore, at the current stage, our model’s parameters can be easily obtained in standard clinical settings, ensuring high practicality and strong potential for clinical dissemination.
Q4
Based on your findings, what impact might this innovative model have on future clinical practice? What directions should future research take?
Professor Jiesong Wang:
Our most critical next step is to carry out prospective, multicenter validation studies to confirm the generalizability and predictive power of this model. If future data verify that the iPET-NCCN-IPI model provides precise and reliable prognostic stratification, it could enable clinicians to identify poor-prognosis patients earlier and implement timely, tailored interventions to improve outcomes. This model also holds potential value for adaptive treatment strategies, providing a framework for individualized and precision-guided therapy in DLBCL management. Ultimately, it could help shape a more refined, patient-centered approach to lymphoma treatment, improving both survival and quality of life.
Professor Jiesong Wang Department of Hematology Fujian Cancer Hospital
