In the last 10 years retrospective studies and randomized clinical trial showed an improvement of survival in patients with Diffuse Large B Cell Lymphoma (DLBCL) and Follicular Lymphoma (FL) treated with chemotherapy regimens containing rituximab (R). However, clinical trials based data refer to a very selective subgroup of patients that does not necessarily correspond to the general population. Therefore we analysed data available from the cancer registry of the province of Modena (MCR), Italy, where R was introduced as standard treatment of DLBCL starting from 2003. Aim of this study was to evaluate if the survival benefit of R treatment observed in clinical trials in patients with DLBCL was confirmed in a study population
From 1996 to 2008 we identified 662 metachronous DLBCL (ICDO3 histology codes 9678–9680 6984/3) that were divided in two groups: 345 patients diagnosed between 1996 and 2002 (before the use of R), and 317 from 2003 to 2008 (when R was routinely used in the province of Modena). Relative survival (RS) was estimated using the Hakulinen method, with the complete cohort approach and age-adjusted by Corazziari method. The relative risks (RR) by gender and year of diagnosis (1996–2002 vs 2003–2008) was estimated assuming Poisson distribution for excess deaths. Moreover, by means of incidence based mortality approach (IBM) we estimated the overall effect of introduction of the R with the rate mortality ratio (RMR) between the cohort 2004–2007 (last follow-up 2008, N=198) and 1998–2001 (last follow-up 2002, N=200), and the effect by gender and age at diagnosis (<70 and ≥70 years old). Mortality data were extracted from MCR using the ICD9 codes 200 and 202. For all estimates we reported the confidence intervals at 95% (95%IC).
The cohort consisted of 662 patients recorded in MCR in the period 1996–2008 (last data of follow-up is 31/12/2009) 51% of which were male and 47% ≥70 years old. The population of the province of Modena in 2008 was 677,896 (about 1.1% of Italian population). Overall, the RS in the period 2003–2008 and 1996–2002 was 61.0% (95%IC:53.8–67.9%) and 43.1% (95%IC: 37.0–49.1%), respectively (RR 0.63, P<0.001). The RS of male gender in 2003–2008 was 62.1% (95%IC: 51.6–71.8%) compared with 38.0% (95%CI: 29.8–46.6%) in the period 1996–2002 (RR=0.56, P<0.001); the RS for female was 58.7% (95%IC: 48.2–68.2%) in the period 2003–2008 and 48.1% (95%CI: 39.4–56.6%) in the period 1996–2002 (RR=0.71, P=0.047). Overall, from IBM analysis the RMR between period 2004–2008 vs 1998–2001 was 0.68 (95%CI: 0.50–0.94, P=0.018). From the same comparison for age <70, RMR was 0.56 (95%CI: 0.33–0.96, P=0.034) and for age ≥70 RMR was 0.74 (95%CI: 0.49–1.09, P=0.128). For male gender RMR was 0.71 (95%CI: 0.46–1.11, P=0.132) and for female was 0.65 (95%CI 0.41–1.03, P=0.068).
This population based study includes exclusively patients with DLBCL, many of which would not have met the trial inclusion criteria, thus reflecting the experience of everyday clinical practice in a homogenous geographic area. Our results show an improvement in survival after the introduction of R for the treatment of DLBCL, that confirms the data from clinical trials and the population based study of LH Sehn (1). Furthermore, using IBM method the RMR highlighted an overall decrease of mortality in the “R period” compared to the period “before R”. This is particularly evident for patients younger than 70 and, with a minor effect, in the female gender.
the study has been supported by “Associazione Angela Serra per la Ricerca sul Cancro”, Italy.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.