Simple Three-Factor Score Predicts Who Can Finish R-CHOP Therapy in Older Lymphoma Patients

Standard-dose R-CHOP-based therapy remains a cornerstone of first-line treatment for diffuse large B-cell lymphoma, the most common aggressive lymphoma in adults, yet its medical feasibility varies substantially among older patients. A new study published in Annals of Hematology addresses a practical question that clinicians face at every pretreatment consultation: which older patients can actually start and finish the planned full course of standard-dose R-CHOP-based therapy, and which will be forced off it for medical reasons? To answer it, researchers developed and internally validated a simple clinical risk score built from routine pretreatment information.

The study was a single-center retrospective analysis of 117 patients aged 60 years or older with newly diagnosed diffuse large B-cell lymphoma. The authors defined medical non-feasibility as the inability to initiate or complete the planned course of standard-dose R-CHOP-based therapy for medical reasons. Among the 117 patients, 39, or 33.3 percent, were classified as medically non-feasible. That figure underscores how often the intended standard treatment cannot be delivered as planned in this age group, even before any discussion of tumor biology or response.

Using logistic regression, the team identified three pretreatment factors independently associated with medical non-feasibility: increasing age, an Eastern Cooperative Oncology Group performance status of 2 or worse, and lower serum albumin. All three are routinely recorded before treatment begins, which is what makes the resulting score practical. The full model showed good discrimination, with an area under the receiver operating characteristic curve of 0.897, and an optimism-corrected AUC of 0.888 after internal validation with 1,000 bootstrap resamples. A simplified score retaining only the three key variables performed almost as well, with an AUC of 0.893.

The bootstrap procedure matters because it addresses a common weakness of prediction models developed in small single-center datasets. By repeatedly resampling the study population and refitting the model, the researchers estimated how much of the apparent performance was an artifact of fitting the same data twice. The small gap between the apparent AUC of 0.897 and the optimism-corrected value of 0.888 suggests the model was not severely overfitted, although such internal validation cannot substitute for testing in an independent external cohort.

Beyond classification, the authors explored how treatment feasibility relates to outcomes. Patients classified as medically feasible had significantly better progression-free survival and overall survival than those classified as non-feasible, with a log-rank P value below 0.0001 for both endpoints. The authors frame this as an exploratory analysis within the study, and the association should be read carefully: feasibility classification reflects whether the planned therapy could be delivered, so the survival difference is not evidence that the score itself predicts tumor biology or treatment response.

The clinical logic behind the three predictors is straightforward. Performance status of 2 or worse captures reduced functional reserve, which in older adults often reflects comorbidity, frailty or both, and is a well-established determinant of whether intensive chemotherapy can be tolerated. Serum albumin is a routine laboratory marker that reflects nutritional state, inflammation and general physiological reserve, and low values have repeatedly been linked to poorer tolerance of cytotoxic therapy and worse outcomes in lymphoma. Increasing age compounds these risks, although the study treats it as a continuous factor rather than a simple cutoff.

The authors are explicit about the limits of their tool. In the abstract, they state that a simple score model based on age, ECOG performance status and serum albumin may help identify older patients with diffuse large B-cell lymphoma who warrant more comprehensive pretreatment assessment, but that it should not independently guide chemotherapy intensity before external validation. In other words, the score is a screening and triage aid, not a decision rule. A high score should trigger closer geriatric or multidimensional assessment, not an automatic reduction in treatment.

Several caveats follow from the study design. It was retrospective and conducted at a single center, so the patient mix, supportive care practices and treatment decisions may differ from those at other institutions. The cohort of 117 patients is modest, and the number of events, 39 non-feasible cases, limits how many variables the model could reliably support, which is one reason the authors deliberately built a parsimonious model. The definition of medical non-feasibility also depends on local clinical judgment about what counts as a medical reason, which may not be uniform across centers.

For clinicians, the practical message is that three numbers already collected in every pretreatment workup, age, performance status and albumin, can be combined into a score that separates patients likely to complete standard-dose R-CHOP-based therapy from those at high risk of medical non-feasibility. For older patients and their families, the finding helps explain why oncologists order comprehensive assessments before committing to full-dose therapy. The next step, external validation in independent and ideally multi-center cohorts, will determine whether this score can safely move from research tool to routine pretreatment guidance.

Subject of Research: A clinical risk score predicting medical feasibility of full-course standard-dose R-CHOP therapy in older patients with diffuse large B-cell lymphoma.

Article Title: Factors associated with the medical feasibility of completing full-course standard-dose R-CHOP-based therapy in older patients with diffuse large B-cell lymphoma: Development of a clinical risk score

Article References: Dong, B., Zhao, M., Bai, X., Liu, Y., Ding, B., Liu, Y., Zhang, X., Li, Y., Xi, Y., & Li, Z. (2026). Factors associated with the medical feasibility of completing full-course standard-dose R-CHOP-based therapy in older patients with diffuse large B-cell lymphoma: Development of a clinical risk score. Annals of Hematology. https://doi.org/10.1007/s00277-026-07283-8

Image Credits: AI Generated

DOI: 10.1007/s00277-026-07283-8

Keywords: diffuse large B-cell lymphoma, R-CHOP, older patients, clinical risk score, serum albumin, ECOG performance status, retrospective study, progression-free survival, overall survival, pretreatment assessment, Factors, associated