“Non-Hodgkin's Lymphomas and Diabetes of Type 2 in Elderly Patients: Impact On Management in the Real Life”.

Prevalence of patients (pts) suffering of non-Hodgkin's lymphoma (NHL) and diabetes is increasing. Previous investigations were mainly focused on possible association with an increased risk of NHL in pts with type 2 diabetes. In another way, it is well known that NHL patients (pts) with co-morbidity are treated less intensively and have worse prognosis. We propose a retrospective analysis of a homogeneous aggressive lymphoma cohort based on a multicentric regional network database to assess the cross-impact of diabetes and NHL in term of pts management and survival.

All pts with high-grade NHL cases, histologically proven, diagnosed between 2003 and 2008 in the Hematology Department of regional network HEMATOLIM, have been included. Diabetes of type 2 has been defined according to international criteria. Data about diabetes and lymphoma treatments, toxicity, relapse, progression or death, have been collected from the Database Structure Régionale de Référence sur les Lymphomes en Limousin. Statistical analysis had been done by StatView and SAS 9.1.2 softwares.

We included 251 NHL pts with aggressive NHL concerning 69.3% Diffuse Large B Cell Lymphoma (DLBCL) (n=174), 7.6% Mantle cell (n=19), 5.2% Peripheral T-cell (n=13), 5.2% Burkitt (n=13) and 12.7% others (anaplastic (n=9), Follicular grade 3B (n=8),T angioimmunoblastic (n=8), primary mediastinal NHL (n=3), nasal NK-T (n=2), T-cell-rich BCL (n=1), centroblastic BCL (n=1)). Among these NHL, 16.7% had type 2 diabetes (n=42). Sex ratio was 1 for diabetic pts (dp) versus (vs) 1.4 for non-diabetic pts (ndp), the average age was 65.1±15.8, 71.3±8.4 for dp vs 63.8±16.6 for ndp (p=0.005). All received first-line chemotherapy mainly Cyclophosphamide/Oncovin/ Adriamycine/Prednisone ± Rituximab (54.6%, n=137), with no difference between dp and ndp (respectively 59.5% vs 57.0%). Chemo-resistance was reported for 22.7% of pts (28.6% dp (n=12) and 21.5% ndp (n=45)). Chemotherapy-related toxicities were more frequent in dp (71.4%, n=30, vs 47.8%, n=100 ndp) (p=0.005), mainly fever and/or bleeding and infectious complications or back to hospitalization. Dose reductions were more frequent in dp, 31.0%, (n=13) vs ndp 12.0%, n=25) (p=0.0017) and adjustements in time-interval (54.8%, n=23, vs 38.8%, n=81 ndp) (p=0.05) of chemotherapy course. A logistic regression analysis showed that dose reductions and treatment-related toxicities were associated to diabetes status (p=0.015 and p=0.026, respectively). Complete remission (CR) was achieved for 66% ndp (n=138) vs 40.5% dp (n=17) with significant difference (p=0.0019). Relapses: no significant difference between both groups with 13.9% ndp (n=29) vs 19% dp (n=8). Mortality rate was significantly increased: 41.6 % ndp vs 59.5% dp (p=0.033), mainly due to NHL complications, 41.6% ndp (n=87) vs 59.5% dp (n=25) (p=0.033). With a median follow-up of 17 months (m) (range 0-72), median overall survival (OS) was not reached for ndp vs 12±9 m for dp (p=0.006), median event-free survival (EFS) was 41±6.2 m for ndp vs 11±4.3 m for dp (p=0.002). In a multivariate analysis, the differences disappeared on OS and EFS after adjustment on age. A logistic regression analysis showed an important increasing of complications, adjustment of doses, and CR but these chemotherapy-related toxicities had no impact on dp survival. The impact of chemotherapy mainly including corticosteroids on the long term diabetes status has been assessed at the end of NHL treatment, 35.7% increased oral medication to oral insulin or received an intensified dose of the antidiabetic treatment (9.5%, n=4). Some dp with oral treatment improved their glycemia (7.1%, n=3), and used only hygiene-dietary measures. Long term insulin treatment was introduced after glycemia disorders for 3.8% ndp (n=8).

Aggressive NHL were more frequently associated with type 2 diabetes (16%) in our regional cohort probably due to the study population median age around 68 years. A real impact of incidence of the chemotherapy-related toxicities, back to hospitalization, a less dose-intensity chemotherapy on dp CR has been significantly demonstrated. EFS and OS were not significant in multivariate analysis especially with adjustment on age. These results claim for a cautious management of diabetes at the initial assessment and during chemotherapy for decreasing complications and by an intensive patient education.

No relevant conflicts of interest to declare.