Myeloproliferative Neoplasms and Dementia Risk: A Population-Based Cohort Study

Dementia is a severely disabling condition that affected 50 million people and is likely to triple by 2050, thus placing a heavy burden on families and society [1]. Several dementia risk factors have been identified, including age, sex, and environmental and lifestyle factors together with genetic factors [2, 3]. Many of the known risk factors, such as obesity, diabetes, hypertension, and smoking, are potentially modifiable [2, 3].

Chronic inflammation is a potential key pathogenic factor in the development of dementia [4, 5], but many of its aspects and clinical implications are poorly understood. Philadelphia chromosome–negative chronic myeloproliferative neoplasms (MPNs) are chronic blood cancers caused by the somatic driver mutations JAK2V617F, CALR, and MPL in hematopoietic stem cells [6]. These specific gene mutations contribute to the inflammatory and thrombogenic state in patients diagnosed with MPNs, and the prediagnostic phase of MPNs may span as many as 15–20 years [7]. MPNs have been proposed as a human neuroinflammation model for studying the development of dementia [8]. To our knowledge, no population-based studies have examined associations between MPNs and dementia risk.

We conducted a cohort study to examine these associations in comparison with a general population comparison cohort. The study also used chronic lymphocytic leukemia (CLL) and chronic myelogenous leukemia (CML) as control diseases. CLL has a long clinical course, like MPNs, but chronic inflammation is not considered a driving force for CLL development and disease progression [9]. CML is another chronic myeloproliferative blood cancer caused by a specific reciprocal chromosome translocation (Philadelphia chromosome) and does not involve a chronic inflammatory state to the same extent as Philadelphia chromosome–negative MPNs [10].

We identified 9895 patients with MPNs, 9465 patients with CLL, and 1530 patients with CML (Figure S1a–c). The patients with MPNs included 3090 with ET, 3910 with PV, 410 with MF, and 2520 with unspecified MPN. Table 1 shows the baseline characteristics and follow-up information for the MPN, CLL, and CML cohorts. Patients with MPNs had more chronic inflammatory comorbidities than did their matched general population comparators, whereas patients with CLL did not (Table 1, Figure S2). Patients with CML were more likely to have comorbidities than were their comparison cohort comparators, but this difference was less pronounced than that observed in the MPN cohort (Table 1, Figure S2).

During the follow-up (median of 5 years for the MPN cohort and 7 years for the general population cohort), 520 (5.3%) people in the MPN cohort and 7070 (7.4%) people in the general population cohort were diagnosed with dementia, and 50.0% (n = 4945) of the MPN cohort and 31.4% (n = 30 110) of the general population cohort died (Table 1). Figure S3 shows the cause-specific cumulative incidence of dementia and death in these cohorts. Table S5 presents the incidence rate of dementia per 1000 person-years in the MPN, CLL, and CML cohorts and in their general population comparison cohorts. People diagnosed with MPNs had an increased risk of developing dementia compared with the general population comparison cohort (adjusted HR: 1.15, 95% CI: 1.04–1.27) (Figure 1, Table S5).

People diagnosed with CLL had a decreased dementia incidence (adjusted HR: 0.81, 95% CI: 0.72–0.90), whereas CML was associated with an increased dementia incidence (adjusted HR:1.14, 95% CI: 0.70–1.65) compared with their own general population comparison cohort (Figure 1, Table S5). However, our estimates for CML were imprecise because of the limited cohort size.

Baseline characteristics of patients with ET, PV, MF, and CU are shown in Table S6. Individuals with all MPN subtypes, particularly patients with PV, had more comorbidities than their matched unexposed comparators (Figure S4, Table S6). Patients with PV, MF, and CU tended to have elevated dementia risks, although the estimates for those with MF were imprecise (Figure 2, Table S7).

In first year of follow-up after diagnosis, patients with MPNs, CLL, and CML were more likely to be diagnosed with dementia than their general population comparison cohorts. Although the HR decreased for patients with MPNs after the first year, it remained above one (Figure 3, Table S8). Conversely, the elevated dementia risk for patients with CLL dropped below one after the first year of follow-up (Figure 3, Table S8). Patients with PV or CU, however, did not show an excess increase in HR during the first year of follow-up compared with the HRs with a longer follow-up (Figure 3, Table S9).

This large population-based cohort study showed that patients with MPNs had elevated dementia risk. The observed association was stronger in patients with PV than in patients with ET, and more pronounced in men with MPNs than in women.

All authors revised the manuscript and had final responsibility for the decision to submit for publication. Yuelian Sun, Hans Carl Hasselbalch, and Henrik Toft Sørensen designed and conceived the study. Katalin Veres performed the statistical analyses. Yuelian Sun drafted the manuscript. All authors interpreted the data, revised the manuscript, and approved the final manuscript.

The authors declare no conflicts of interest.