Waiting Longer for Lymphoma Treatment May Not Shorten Survival, Nationwide Study Finds

When it comes to cancer, the received wisdom is simple: the sooner treatment starts, the better. Every day of delay feels like time handed to the tumor. But a sweeping new analysis of German cancer registry data is challenging that intuition for one of the most common blood cancers, diffuse large B-cell lymphoma, and the findings are as counterintuitive as they are clinically important. Patients whose immunochemotherapy began later after diagnosis did not die more often; in fact, they tended to live longer than those treated within the first week.

The study, published in BMC Cancer by a team led by Susanne von Kroge of University Medical Center Hamburg-Eppendorf together with collaborators at the Robert Koch Institute and the Hamburg Cancer Registry, drew on harmonized nationwide registry data covering the years 2020 through 2023, with vital status follow-up extending to December 2024. From a total of 16,443 patients diagnosed with a first primary diffuse large B-cell lymphoma, 10,872 began systemic therapy within 56 days of diagnosis and formed the primary analysis cohort. The median age was 71 years, 44.2 percent of patients were female, and the median follow-up lasted about a year and a half.

The researchers divided patients into groups according to the diagnosis-to-treatment interval, a measure epidemiologists call the DTI, ranging from treatment in the first week after diagnosis to initiation between weeks five and eight. Three-year overall survival across these five groups ranged from roughly 63 percent to 75 percent. On its face, that spread might seem to suggest that timing matters enormously. But when the team applied statistical models, the direction of the association was the opposite of what the urgency narrative would predict.

Using Kaplan-Meier survival curves and Cox proportional hazards regression, the investigators found that compared with patients treated in week one, those who started later had progressively lower hazards of death. Patients beginning therapy in week three had a 28 percent lower hazard of dying from any cause, with a hazard ratio of 0.72 and a 95 percent confidence interval of 0.64 to 0.81. Those starting in week four fared better still, at a hazard ratio of 0.63, and patients treated between weeks five and eight showed a hazard ratio of 0.62, or a 38 percent reduction. The confidence intervals in every case excluded one, indicating the associations were statistically robust.

To a clinician, such a pattern immediately raises a red flag of confounding. Doctors do not assign treatment dates at random. Patients with the most aggressive, bulky, or symptomatic disease are rushed into the infusion chair within days, while those with more indolent-appearing disease can safely wait for complete staging, cardiac assessments, and fertility discussions. If sicker patients are treated fastest, then early treatment will appear to worsen survival even if timing itself has no causal effect, a classic example of what epidemiologists call confounding by indication, or in this context, the waiting-time paradox.

The German team addressed this head-on by adjusting their models for patient, tumor, and structural characteristics, most importantly the International Prognostic Index, or IPI, the standard five-factor score that combines age, disease stage, number of extranodal sites, performance status, and lactate dehydrogenase levels. Low-risk IPI strongly predicted better survival, with a hazard ratio of 0.24 compared with high-risk disease. Even after this adjustment, a residual association between longer diagnosis-to-treatment intervals and improved survival persisted, hinting that the interval itself might carry independent prognostic information in patients with less aggressive disease.

Perhaps the most revealing part of the study is what predicted delay in the first place. The researchers modeled which factors pushed treatment later and found that low-risk IPI scores added an average of 7.89 days to the interval, while older age added 0.62 days per standard deviation. Patients who survived the first year had, on average, started treatment 3.26 days later than those who did not. In other words, the patients who waited longest were systematically the ones with the least threatening disease, and the ones treated most urgently were those whose lymphoma demanded immediate action.

Missing IPI data posed a methodological challenge that the authors handled with unusual care. Rather than discarding incomplete records, they treated missing IPI as a separate category and ran both complete-case and multiple-imputation sensitivity analyses to test whether their conclusions depended on how missingness was handled. The consistency of the results across these approaches strengthens the central finding that early initiation is primarily a marker of adverse disease biology rather than a driver of survival in its own right.

The implications for health policy and clinical practice are subtle but significant. In many healthcare systems, time-to-treatment is used as a quality indicator, with targets and audits pressing hospitals to start cancer therapy faster. For solid tumors with validated screening pathways, such metrics can be meaningful. But this study suggests that for diffuse large B-cell lymphoma, raw speed from diagnosis to treatment is a poor proxy for quality of care. Penalizing centers for longer intervals could perversely reward the triage of less sick patients while doing nothing for those with aggressive disease, who are already treated within days.

The authors conclude that early treatment in newly diagnosed DLBCL is primarily driven by adverse disease characteristics, and that the diagnosis-to-treatment interval may function as an independent prognostic factor mainly in less aggressive disease. The study was funded by the Innovationsfonds zur Förderung von Versorgungsforschung under grant number 01VSF25057, and the underlying data came from German state cancer registries used under legal provisions permitting scientific analysis of non-identifiable records. For the roughly 16,000 Germans diagnosed each year with this lymphoma, and for oncologists everywhere who feel the pressure of the clock, the message is a nuanced one: the calendar matters less than the biology, and a few extra days of careful preparation may be exactly what the least aggressive tumors allow, and what the most aggressive ones cannot afford.

Subject of Research: The association between the diagnosis-to-treatment interval and survival in newly diagnosed diffuse large B-cell lymphoma

Article Title: Insights into the survival impact of delayed immunochemotherapy in newly diagnosed diffuse large B-cell lymphoma: results from a nationwide registry study

Article References: von Kroge, S., Steinhaus, N. J., Bokemeyer, C., Tetzlaff, F., Walk, H., Kusche, H., LYSED study group, Klotz, S., Obi, N., Knesebeck, O. V., Beernink, M., Gundler, C., Ückert, F., Schulze, K., Augustin, J., Semmelhaack, R., Sassenberg-Gummert, A., & Peters, F. (2026). Insights into the survival impact of delayed immunochemotherapy in newly diagnosed diffuse large B-cell lymphoma: results from a nationwide registry study. BMC Cancer, 26(1), Article 1144. https://doi.org/10.1186/s12885-026-17024-9

Image Credits: AI Generated

DOI: 10.1186/s12885-026-17024-9

Keywords: diffuse large B-cell lymphoma, diagnosis-to-treatment interval, treatment delay, immunochemotherapy, survival, International Prognostic Index, cancer registry, Cox regression, prognostic markers, waiting-time paradox, hematology, oncology