Incidence Clinical Characteristics and Survival of Non Splenic Marginal Zone Lymphomas: A Population Based Study from a Cancer Registry in Northern Italy

Purpose: The aim of the present study was to evaluate incidence, clinical characteristics and survival of non splenic marginal zone lymphomas (NS-MZL) diagnosed in the province of Modena (Italy) from 1997 to 2006.

Patients and Methods: Cases with a diagnosis of NS-MZL were identified from the population based archive of malignant lymphomas (ML) diagnosed in the province of Modena (Italy) from 1997 to 2006 obtained from the local cancer registry. ICD-O3 code 9699-3 was used to identify cases. A complete dataset including demographic, clinical data, treatment and survival data was defined and data collected by active chart review and, when possible by automatic searches. Annual Percentage Change (APC) of age-standardized rates was evaluated using a log-linear model.

Results: One hundred and fifty-three patients with a diagnosis of NS-MZL were identified representing 9 % of all B-NHL. Median age of cases was 67 years (range 16 to 97 years). Clinical presentation was purely nodal in 7% and purely extranodal in 65% of patients. Ann Arbor stage was III–IV in 37% of cases, bone marrow was involved in 26%. The most frequent involved extranodal site was gastrointestinal (GI) system (39.4%) followed by skin (16.8%) and salivary/lacrimal glands (9.6%); multiple extranodal sites were present in 17% of cases. World-age standardized incidence of NS-MZL was 1.4/1 (Males/Females) per 100.000 people. During the studied period we observed a trend for decreasing incidence for all cases (APC: -4.1% C.I.95%: -8.9%;1.1%). Incidence reduction was mostly due to a significant decrease in the incidence of GI NS-MZL (APC: -12.5%, C.I.95%: -19.2%; -5.3%) thus suggesting a positive effect of the extensive use of Helicobacter Pylori (HP) detection methods and of HP eradicating therapies. Looking at age and sex standardized rates we observed a bimodal age distribution confirming our previous finding (Luminari et al. Hematol Oncol 2007). This was more evident among males with a first peak of incidence around 30 years and a second around 70 years. A cut off value of 45 years was used to distinguish the two groups; patients younger than 45 were 24 (16%) and showed different clinical characteristics compared to older patients; these included higher frequency of males (75%vs 46%: P=0.014), stage I–II disease (83% vs 58%: P=0.034), skin involvement (61% vs 10%: P<0.001), and mediastinal involvement (17% vs 2%: P=0.012). At multivariate logistic regression analysis male gender (P=0.04), skin involvement (P<0.001) and mediastinal involvement (P=0.001) were independently correlated with age. After a median follow-up for alive patients of 67months (range 11 to 128), 50 patients died; lymphoma was the cause of death in 50% of cases. Five-year survival rate was 67%. At multivariate analysis elder age (> 60 years: HR 3.67: P=0.004), stage III–IV (HR 2.14: P=0.01), skin involvement (HR 0.24: P=0.05) and more recent diagnosis (years 2004–2006 vs others: HR 0.32 P=0.045) showed association with OS.

Conclusions: This population-based study demonstrated the feasibility and clinical usefulness of a population based cancer registry for studying rare diseases such as NS-MZL. In particular our study allowed us to show a decrease in the incidence of NS-MZL with GI involvement that we interpret to be an effect of the wide use of HP eradicating therapies. Moreover, among NS-MZL analyzed in the current study we identified a subgroup of young male patients with skin involvement and with excellent prognosis that may represent a distinct group of NS-MZL.