The Impact of Diabetes Mellitus As Comorbid Condition, in Patients with Lower-Risk Myelodysplastic Syndromes (MDS)— a Study of the European Leukemianet-Based MDS (EUMDS) Registry

Background: Myelodysplastic syndrome is mainly a disease of the elderly, while the prevalence of Diabetes Mellitus (DM) in the aged population is about 25%.

Aim: We sought to investigate the impact of DM as comorbid condition on the diagnosis of MDS.

Patients and methods: We investigated the clinical and laboratory features at time of MDS diagnosis of 431 DM patients, occurring among a total of 2114 MDS patients registered in the EUMDS registry, and compared them with the non-DM patients. Data were collected by 145 centers located in 16 European countries and Israel.

Results: The overall incidence of DM at registration visit was 23%. Higher frequency of DM was observed among MDS patients in Israel (40.3%), Portugal (26.3%) and the Netherlands (25.7%) and lower in Denmark (9.3%), Serbia (10.5%) and Romania (13.6%). Diabetic MDS patients were more commonly male (65.6% vs 60.2%, p=0.01), with higher frequency in Portugal (80%), United Kingdom (76.3%) and Greece (72.9%) and lower in Romania (33.3%), Germany (45.5%) and Israel (54.1%). Higher incidence of DM was encountered among patients with MDS-RA (24.3%, more prominently in the Netherlands, Israel and Spain) and lower among those with RCMD-RS (15.4%). Ring sideroblasts were less commonly encountered among DM patients (24.9% vs 29.5%, p: n.s.) and DM patients had a greater proportion of high MDS-CI risk than non-DM (9% vs 4.8%, p<0.001). The Performance Status (Karnofsky scale) was worse in DM than in non-DM patients ("able" persons 69.1% vs 79.4%, p<0.01). Moreover, DM patients were besides antidiabetic treatment more commonly receiving antiplatelet, anticoagulant and antihypertensive drugs, as well as proton pump inhibitors. No significant difference in the prognostic categorization distribution according to IPSS, WPSS or IPSS-R, between DM and non-DM patients was observed. DM patients were more anemic at diagnosis of MDS (Hb: 10.1±1.7 vs 10.3±1.9 g/dl, p=0.03), had lower MCV (95.5±9.6 vs 98.1±10.6 fl, p<0.01), higher WBC count (5.6±3.3 vs 5.0±3.0 x 10⁹/l, p<0.001), and higher absolute neutrophil count (3.3±2.5 vs 2.8±2.2 x 10⁹/l, p<0.001). No significant difference in baseline platelet counts and in the percentages of patients, exhibiting 0, 1, 2 or 3 cytopenias, between DM and non-DM patients was noted. However, DM patients had higher mean serum creatinine levels (102±65 vs 89±35 μmol/l, p<0.001) and among them there was a significant negative correlation between Hb and serum creatinine levels (r= -0.131, p=0.02). All three iron status parameters (serum iron, transferrin saturation and serum ferritin levels were significantly decreased among DM-patients, implying a possible influence of DM on iron metabolism. Serum EPO levels were not significantly different between the two groups, nevertheless, considering that diabetic MDS patients were more anemic, it could be assumed that they might exhibit somewhat blunted response, probably as a result of some degree of impairment of their renal function. Finally, patient-reported Quality of Life, using the EQ-5D tool was significantly lower among DM-, compared to non-DM patients (0.70±0.24 vs 0.75±0.22, p<0.001).

Conclusions: Diabetes Mellitus may modulate the common presenting features of MDS, and since some degree of normocytic, normochromic anemia is commonly associated with this disease, special attention is warranted to discriminate the real MDS patients among the diabetic patient population.