SummaryCentral nervous system (CNS) prophylaxis for diffuse large B-cell lymphoma (DLBCL) is controversial with even less evidence in the era of R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone) chemotherapy. We reviewed the impact of CNS prophylaxis in DLBCL patients treated with R-CHOP at a tertiary care centre over a 7-year period. CNS prophylaxis was recommended for 'higher risk' patients and consisted of intrathecal methotrexate and/or high-dose methotrexate. Of 214 patients 12Á6% received CNS prophylaxis. With a median follow-up of 27 months, eight patients (3Á7%) developed CNS relapse (75% isolated to the CNS and 62Á5% as parenchymal brain disease) at a median time of 17 months. Patients who did not receive CNS prophylaxis had lower events (2Á7%) than those who did (11Á1%). Half of the CNS relapses occurred in testicular lymphoma patients, 75% of whom had received CNS prophylaxis. In multivariate analysis, testicular involvement was the only significant prognostic factor for CNS relapse (hazard ratio 33Á5, P < 0Á001). In conclusion, CNS relapse in DLBCL appears to present as a later, more isolated parenchymal event and at a lower rate in the rituximab era compared with historical data. R-CHOP may negate the need for CNS prophylaxis with the exception of testicular lymphoma.

SummaryThere is a paucity of data on the impact of diagnostic and treatment delays on outcomes in haematological malignancies, particularly in patients with diffuse large B-cell lymphoma (DLBCL). Our database of patients treated for DLBCL between 2002 and 2010 was interrogated. Univariate and multivariate analyses were performed to determine the relationship between sociodemographic or disease-specific variables and delays. Cox Regression analysis was used to discern the impact of delays on survival. Patients (n = 278) waited a median of 4 weeks before seeking medical attention. It took a median of 8 weeks for a non-haematology physician to diagnose DLBCL and refer to a haematologist. A median of 3 weeks elapsed between specialist consultation and chemotherapy initiation. In multivariate logistic regression analysis, bone marrow involvement [odds ratio (OR) = 0Á41, P = 0Á018], Charlson comorbidity index (OR = 1Á42, P = 0Á017) and urgent inpatient chemotherapy (OR = 0Á40, P = 0Á012) were associated with diagnostic delays >6 weeks. Lack of pathological diagnosis at the time of haematology referral was the only factor that independently predicted for treatment delays >4 weeks (OR = 8Á25, P < 0Á01). Diagnostic or treatment delays did not impact survival or progression-free survival. In conclusion, selected disease and patient-related factors are associated with delays in management of DLBCL, but do not impact outcomes.

In the pre-rituximab era, transformation of indolent B-cell lymphoma to diffuse large B-cell lymphoma (DLBCL) was associated with an extremely poor outcome and a median post-transformation survival ranging from 1 to 2 years. We evaluated the impact of rituximab-cyclophosphamide, adriamycin, vincristine, prednisone (R-CHOP) on the survival outcomes of transformed lymphoma compared with de novo DLBCL. Between 2002 and 2010, 317 DLBCL patients who were consecutively diagnosed and treated with R-CHOP were identified at our institution. Patients with transformed lymphoma were included if they had not previously received R-CHOP. Patient characteristics, treatment, and outcome data were retrospectively collected. Sixty patients (19 %) had transformed lymphoma of which 37 (62 %) had transformed from follicular lymphoma, 50 (83 %) were chemotherapy naïve, and 58 (96 %) were rituximab naïve at the time of treatment. With a median follow-up of 31.4 months, 231 patients achieved either complete response or complete response unconfirmed (73 %) with no significant difference between de novo DLBCL (n = 192, 75 %) and the transformed group (n = 39, 65 %) (P = 0.25). Six patients (15 %) relapsed in the transformed group at a median time to relapse of 29.3 months. The 2-year and 5-year overall survivals for all patients were 82 and 72 %, respectively. The overall and progression-free survivals for transformed lymphoma and de novo DLBCL were not statistically different (P = 0.45 and P = 0.38, respectively). With R-CHOP chemotherapy, the prognosis of transformed lymphoma in patients with minimal chemotherapy exposure for indolent disease is similar to that of de novo DLBCL.