The myelodysplastic syndromes (MDS) have highly variable outcomes and prognostic scoring systems are important tools for risk assessment and to guide therapeutic decisions. However, few population-based studies have compared the value of the different scoring systems. With data from the nationwide Swedish population-based MDS register we validated the International Prognostic Scoring System (IPSS), revised IPSS (IPSS-R) and the World Health Organization (WHO) Classification-based Prognostic Scoring System (WPSS). We also present population-based data on incidence, clinical characteristics including detailed cytogenetics and outcome from the register. The study encompassed 1329 patients reported to the register between 2009 and 2013, 14% of these had therapy-related MDS (t-MDS). Based on the MDS register, the yearly crude incidence of MDS in Sweden was 2·9 per 100 000 inhabitants. IPSS-R had a significantly better prognostic power than IPSS (P < 0·001). There was a trend for better prognostic power of IPSS-R compared to WPSS (P = 0·05) and for WPSS compared to IPSS (P = 0·07). IPSS-R was superior to both IPSS and WPSS for patients aged ≤70 years. Patients with t-MDS had a worse outcome compared to de novo MDS (d-MDS), however, the validity of the prognostic scoring systems was comparable for d-MDS and t-MDS. In conclusion, population-based studies are important to validate prognostic scores in a 'real-world' setting. In our nationwide cohort, the IPSS-R showed the best predictive power.

In acute myeloid leukemia (AML) FLT3 internal tandem duplication (ITD) and nucleophosmin 1 (NPM1) mutations provide prognostic information with clinical relevance through choice of treatment, but the effect of age and sex on these molecular markers has not been evaluated. The Swedish AML Registry contains data on FLT3-ITD and NPM1 mutations dating to 2007, and 1570 adult patients younger than 75 years, excluding acute promyelocytic leukemia, had molecular results reported. Females more often had FLT3ITD and/or NPM1mut (FLT3ITD: female, 29%; male, 22% [P = .0015]; NPM1mut: female, 36%; male, 27% [P = .0001]), and more males were double negative (female, 53%; male, 64%; P &lt; .0001). Patients with FLT3ITD were younger than those without (59 vs 62 years; P = .023), in contrast to patients with NPM1mut (62 vs 60 years; P = .059). Interestingly, their prognostic effect had a strong dependence on age: FLT3ITD indicated poor survival in younger patients (&lt;60 years; P = .00003), but had no effect in older patients (60-74 years; P = .5), whereas NPM1mut indicated better survival in older patients (P = .00002), but not in younger patients (P = .95). In FLT3ITD/NPM1mut patients, the survival was less dependent on age than in the other molecular subsets. These findings are likely to have clinical relevance for risk grouping, study design, and choice of therapy.

The combination of eltrombopag with azacitidine in high-risk MDS patients is feasible and well tolerated. Improvements in platelet counts and the potential antileukaemic effect of eltrombopag should be explored in a randomised study.

No abstract

No abstract