Time trends in incidence and prevalence of immune thrombocytopenia: A nationwide cohort analysis

During the past decades, studies of basic frequency measures in primary immune thrombocytopenia (ITP) have emerged1, 2 emphasizing a large proportion of ITP patients are older. Some studies even suggest incidence rates increase over time, but that trend is yet to be explored past 2005.2, 3 For incidence rates, a recent literature review found that estimates of adult ITP varied from 1.9 to 3.9 per 100 000 person-years, were slightly higher in women compared to men, and increased with age.1 In Denmark, incidence rate peaked amongst persons over 60 years at 4.6 per 100 000 person-years.2 Likewise, knowledge of ITP prevalence is sparse. In 2009 point prevalence of ITP amongst adults in the UK was 4.5 per 100 000 persons,4 while the prevalence of chronic ITP (cITP) in the adult population of Denmark was 10.0 per 100 000 persons.5 In the United States, ITP prevalence estimates from the early 2000s ranged from 9.5 to 11.2 per 100 000 persons,6, 7 while cITP was 20.3 per 100 000 persons.8 Median age of diagnosis for ITP patients lies around 55 years and older patients are particularly at risk of a number of complications and co-morbidities such as haemorrhagic episodes,9, 10 as well as heightened risk of thromboembolism.11 Therefore management is more challenging where TPO-RAs may have contraindications and immunosuppressive treatments sometimes are poorly tolerated by older patients due to a more vulnerable state, leaving them susceptible to infection. In a recent study we showed that survival in cITP is improving,12 and alongside an ageing population, longer duration of disease and previous trends point towards increasing incidence and prevalence. Hence a contemporary update of frequency measures for adult ITP is necessary. Using a retrospective cohort design incidence and prevalence of ITP in the adult population (≥18 years) of Denmark was described from 1 January 1980 to 31 December 2016. All patients with ITP were identified using routinely updated Danish nationwide health registers, as elaborated on by Mannering et al.12 Primary and secondary ITP (pITP/sITP) diagnoses were defined using validated algorithms based on previously registered diagnoses13 and patients were deemed chronic if they had two ITP diagnoses with at least 12 months between, as agreed by an international working group.14 Incidence rates were calculated as 5-year averages, and point prevalence yearly using all living patients as of January 1st, using national census data for the Danish adult population. Cases before 1980 were unavailable. Data were analysed in STATA version 17. The cohort included 5443 patients with pITP and sITP (detailed in Table S1). After exclusion of sITP patients (n = 654), 36% of the remaining pITP patients died during follow up (Table 1). Descriptive statistics for the complete cohort are provided in Table S1. Both the incidence and prevalence of pITP and chronic pITP increased between 1980 and 2016, and pITP and chronic pITP incidence curves diverge over time (Figure 1). Five-year incidence of pITP increased from 1.01 per 100 000 person-years (95% confidence interval [CI] 0.87–1.16) in 1980–1984 to 5.71 (95% CI 5.23–6.23) in 2010–2014. Stratification of analyses by age at diagnosis displayed age differences in patients above and below 60 years. Throughout the whole study period, pITP and chronic pITP rates for age ≥60 years were over double that of the under 60 years. Prevalence for pITP and chronic pITP rise more steeply after 1994 (Figure 1). Five-year incidence rates were similar for men and women, and there was a female preponderance for prevalence (Figure S1). Our results show both the prevalence and incidence of pITP are increasing in the adult population, in particular at older ages. This is of interest as Denmark, as most of the western world, has an ageing population, with over 60 years of age being the fastest-growing sector of the population.15 Prevalence does not plateau as neither incidence, as shown here, nor duration of disease12 are fixed. Longer life expectancy for ITP patients contributing to increased prevalence may be attributed to improvement in treatment and/or earlier recognition, although further investigation is necessary to determine these relationships. Frederiksen and Schmidt2 showed an increase in both patients with high platelet counts and asymptomatic patients diagnosed in Denmark between 1973 and 1995, indicating earlier recognition of patients may be contributing to the increasing number of patients. Incidence of pITP, including chronic pITP, is both higher and increasing at a steeper rate for 60 years and over, compared to the under 60 years age group. As the population continues to age, it is expected the heightened incidence of pITP in old age will lead to overall increased prevalence of ITP, and the prevalence curves will likely begin to diverge as the over 60 years population increases. Older patients are at heightened risk of complications,9-11 indicating the burden of pITP on the healthcare system is likely to increase. Point prevalence estimates are underestimated early in the time period, due to use of new diagnoses only. A plateau is expected when all new cases account for total cases in the population if incidence is stable. However, as shown, incidence increases therefore prevalence continues to increase accordingly. The results of analyses on the chronic ITP population are susceptible to an unavoidable immortal time bias, as by definition 12 months is required between two diagnoses. The frailest patients who did not survive 12 months after first diagnoses are therefore unaccounted for. In summary, we provide contemporary estimates and show incidence and prevalence of pITP is increasing. These are necessary updates, due to improvements in the survival of ITP patients and large discrepancies in previous studies. Future studies should consider platelet count, to determine whether early diagnosis or over diagnosing is contributing to increasing prevalence and incidence. The study was designed by Henrik Frederiksen, Dennis Lund Hansen and Nikolaj Mannering, who also collected data. Dana Audrey Lawrie analysed data and wrote the paper, and all authors contributed to the revision of the article. Nikolaj Mannering received financial support from Novartis Healthcare and served as an Advisory Board member for Swedish Orphan Biovitrum (SOBI), outside this work. Henrik Frederiksen received financial support from Alexion, Gilead, Abbvie, Novartis and Janssen Pharmaceuticals, all outside this work. Dennis Lund Hansen has received research grants from Alexion and Novartis, and conference fee from EUSA Pharma outside this work. None of the authors declare any other conflicts of interest besides the above-mentioned. Register based research does not require Ethics Committee approval according to Danish law. All data were anonymised, and researchers did not have access to medical files. Not applicable. Not applicable. Not applicable. Data S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.