Short diagnosis-to-treatment interval is associated with high risk and poor outcomes in diffuse large B-cell lymphoma.

Abstract

Introduction: Recent evidence has demonstrated associations between adverse risk characteristics and survival outcomes in diffuse large B-cell lymphoma (DLBCL) patients with a shorter diagnosis-to-treatment interval (DTI) (Maurer et al, 2018). The imperative for exigent therapy in aggressive DLBCL presentations may select against the recruitment of such patients to clinical trials and introduce significant bias. We sought to confirm the relationship between DTI and risk status and to compare cohorts of patients receiving standard therapy versus a clinical trial as initial treatment. Method(s): Treatment and outcome data were collected for patients with newly diagnosed DLBCL (as per WHO criteria) treated at our institution either with standard therapy or an upfront clinical trial. Associations between DTI and clinical characteristics and outcomes were investigated. Overall survival (OS) and progression free survival (PFS) were modelled using Cox regression. Differences in patient and disease characteristics were analysed with the Mann-Whitney U test or Chi-square test. Result(s): 455 patients receiving non-trial therapy with reliably determined DTI were identified. An additional 21 patients received upfront treatment on a clinical trial. The median DTI of the non-trial treatment group was 14 days, compared to 20 days for the clinical trial group (p=0.031). 1 patient was treated on a clinical trial within 7 days of diagnosis (5%) vs 122 (26.8%) patients receiving non-trial treatment. Patients with a DTI of >7 days demonstrated superior OS, HR 1.555 [95%CI: 1.085-2.227], p=0.016, and PFS, HR 1.564 [95%CI: 1.15-2.129], p=0.004. DTI of <7 days was significantly associated with several established markers of poor prognosis in DLBCL, including elevated lactate dehydrogenase (p<0.001), and advanced Ann Arbor stage (III or IV) p=0.034. Other poor prognostic markers such as the revised international prognostic (RIPI) score, cell of origin, Eastern Cooperative Group (ECOG) score >2, age >60 years and presence of 'B' symptoms did not show significant associations with a DTI <7 days. In comparing the non-trial and trial groups, no statistically significant difference between the aforementioned characteristics was present. Conclusion(s): These findings confirm that patients requiring emergent treatment for DLBCL experience inferior OS and PFS and are more likely to have elevated LDH and advanced stage disease. The significantly higher median DTI in clinical trial participants suggests these patients are not representative of the general DLBCL cohort. Our findings raise further concerns for significant selection bias against poor risk patients treated on upfront clinical trials in DLBCL. Further work is needed to evaluate the extent of this bias and strategies to ensure patients requiring emergent treatment are included in future clinical trials is essential.