Abstract
INTRODUCTION: Diffuse large B cell lymphoma (DLBCL) is an aggressive form of non-Hodgkin lymphoma for which a cure is usually the therapeutic goal of optimal treatment. Using a large population-based cohort we sought to examine the factors associated with optimal DLBCL treatment and survival. METHODS: DLBCL cases were identified through the population-based Victorian Cancer Registry, capturing new diagnoses for two time periods: 2008-2009 and 2012-2013. Treatment was pre-emptively classified as 'optimal' or 'suboptimal', according to compliance with current treatment guidelines. Univariable and multivariable logistic regression models were fitted to determine factors associated with treatment and survival. RESULTS: Altogether, 1442 DLBCL cases were included. Based on multivariable analysis, delivery of optimal treatment was less likely for those aged ≥80 years (p < 0.001), women (p = 0.012), those with medical comorbidity (p < 0.001), those treated in a non-metropolitan hospital (p = 0.02) and those who were ex-smokers (p = 0.02). Delivery of optimal treatment increased between 2008-2009 and the 2012-2013 (from 60% to 79%, p < 0.001). Delivery of optimal treatment was independently associated with a lower risk of death (hazard ratio (HR) = 0.60 (95% confidence interval (CI) 0.45-0.81), p = 0.001). CONCLUSION: Delivery of optimal treatment for DLBCL is associated with hospital location and category, highlighting possible demographic variation in treatment patterns. Together with an increase in the proportion of patients receiving optimal treatment in the more recent time period, this suggests that treatment decisions in DLBCL may be subject to non-clinical influences, which may have implications when evaluating equity of treatment access. The positive association with survival emphasizes the importance of delivering optimal treatment in DLBCL.