Are We Undercounting MDS? An Analysis of Incidence Patterns of Myelodysplastic Syndromes in SEER 21 Regions: 2001-2016.

Background: Myelodysplastic syndromes (MDSs), a group of reportable malignancies in the Surveillance, Epidemiology, and End Results (SEER) Program since 2001, are poorly understood neoplasms. There have been several updates since they became reportable, with several changes introduced to cases diagnosed in 2010 and onwards. None have examined changes in patterns of MDS incidence over the long term, accounting for such changes. Objective: The objective of this analysis was to assess changes in incidence of MDS from 2001 to 2016 by demographic characteristics and histology, applying coding changes implemented in 2010. Methods: Incidence-SEER 21 region data for the 2001-2016 period were used to estimate incidence rates using SEER*Stat version 8.3.6. Cases were included that were diagnosed as MDS during this period having the following International Classification of Diseases for Oncology, 3rd edition (ICD-O-3) histology codes: 9980, 9982-9986, 9989, and 9991-9992. Annual incidence rates for the total population, as well as by demographic characteristics and histology, were estimated. All incidence rates were age adjusted using the 2000 US standard population (19 age groups; census P25-1130). Results: A total of 86,146 MDS cases were diagnosed during the 2001-2016 period, with an age-adjusted incidence rate of 4.7 cases per 100,000 population. The majority (~61%) were classified as MDS, unclassifiable (MDS-U, ICD-O-3: 9989). Annual rates steadily increased from 3.7 per 100,000 in 2001 to 5.6 per 100,000 in 2010, then declined to 3.8 per 100,000 in 2016, making an inverted V-shaped pattern. This pattern was observed for both sexes and all assessed racial and ethnic groups, as well as among the ≥65-year age groups. When the rates were assessed by histology, this pattern was observed for MDS-U, but not for other subtypes. Conclusion: MDS-U subtype dominates the observed trend in incident rates. The decline in rates since 2010 is most likely due to changes in coding and diagnostic criteria introduced in 2010. Further analysis is warranted to conclusively determine all factors leading to the changes observed.