Population-based studies in Myelodysplastic syndromes : Prognostic scores, socioeconomic status, and therapy-related disease

Sammanfattning: The aim of this thesis was to expand the epidemiological knowledge of the haematological malignancy MDS and the related condition chronic myelomonocytic leukaemia (CMML). Using nationwide registers, the papers in this thesis address aspects of prognostication, comorbidity, socioeconomic status, and therapy-related disease, using a population-based approach.In paper I we validated the prognostic scoring systems WPSS, IPSS, and IPSS-R in a cohort of 1329 MDS patients. IPSS-R was the most effective scoring system, with the highest C-index of 0.74. The scoring systems were equally effective for therapy-related MDS (t-MDS) as they were for de novo MDS. In paper II we validated the scoring systems, IPSS-R, CPSS, MDAPS, and Mayo score and the comorbidity indices CCI, HCT-CI, and MDS-CI in a cohort of 337 patients with CMML. We concluded that CPSS is the most powerful scoring system. Among comorbidity indices, the CCI gave the most prognostic information. There was a strikingly high prevalence of autoimmune conditions affecting 25% of patients. In paper III we studied the effect of socioeconomic status in a cohort of 2945 patients with MDS. When adjusting for known prognostic factors, mortality was 50% higher in patients with the lowest income compared to those with the highest income and 40% higher among patients with the shortest education compared to those with the longest. Further, a lower socioeconomic status was associated with a reduced probability of receiving effective treatment and with a lower probability of a cytogenetic evaluation at diagnosis. In paper IV we studied t-MDS in a cohort of 2705 patients with MDS, of whom 16% had t-MDS. Patients with t-MDS had a shorter median survival as compared to de novo MDS (15.8 months versus 31.1 months). Previous treatment with either chemotherapy alone or in combination with radiation was associated with a shorter survival than treatment with radiation only. Having a non-malignant disease or a solid tumour as a primary disease was associated with a longer survival, compared with those with a haematological malignancy. IPSS-R and the WHO classification were effective in predicting survival in most subgroups of t-MDS. The t-MDS subgroup treated with radiation only was similar to patients with de novo MDS and should be regarded as having de novo MDS regarding prognostication and treatment.In summary, the findings in this thesis provide evidence for how to improve prognostication and expand knowledge on the patient and disease-specific characteristics leading to the diverse outcomes in MDS and CMML.