Risk of Osteoporosis and Osteoporotic Fractures in Patients with Myeloproliferative Neoplasms

Introduction: Clonal hematopoiesis (CH), which includes somatic mutations characteristic of myeloproliferative neoplasms (MPNs), has been linked to osteoclastogenesis and subsequent inflammatory bone loss. Osteoporosis is a common and preventable disease of older adults associated with increased risk of fractures and premature mortality. While myelodysplastic syndrome has been reported in adults with osteoporosis, the association of osteoporosis and MPNs is not known. We conducted a case-control study to examine the risk of osteoporosis and osteoporotic fractures in MPN patients. Methods: Using the Surveillance, Epidemiology, and End Results (SEER) cancer registry, we identified adults ≥65 years with a Philadelphia chromosome negative MPN diagnosis between 2006-2019. MPN subtypes included essential thrombocythemia (ET), polycythemia vera (PV), myelofibrosis (MF), and MPN-unclassified (MPN-U). A representative Medicare non-cancer control population from the same observation period was matched in a 3:1 ratio with MPN cases by age, sex, and location. In both cohorts, individuals were excluded if they had Health Maintenance Organization (HMO) coverage, or if they lacked continuous part A/B coverage during months -13 to -1 relative to the index diagnosis date. MPN cases were excluded if they had a concurrent hematologic and/or solid tumor neoplasm ±6 months of their index date. Descriptive statistics were used to evaluate baseline characteristics which included age, sex, race/ethnicity, diagnosis year, Charlson Comorbidity Index (CCI) score, and socioeconomic status (SES) defined in quintiles below national poverty line based on an individual's census tract. Presence of osteoporosis and osteoporotic fractures was defined by diagnosis codes ±6 months from index date. Multivariable logistic regression was performed to compare the odds of osteoporosis and/or osteoporotic fractures between adults with and without MPNs. Results: 18,171 MPN cases and 29,042 controls were included for analyses. The median age at MPN diagnosis was 77 years (IQR: 71-83). ET (45%) was most common, followed by PV (38%), and MF (13%). MPN patients were more likely to be <70 years (17% vs 13%), male (44% vs 42%), and White (89% vs 83%) compared to the controls (p<0.01 for all). A higher proportion of MPN patients were in the 1st and 2nd SES quintiles (i.e., lower poverty, 49%) vs 44% of the controls (p<0.01). Chronic comorbidities were more prevalent in MPN patients as determined by the CCI; 23% had a score ≥3 vs 10% in controls (p<0.01). More MPN patients had osteoporosis (10% vs 8%, p<0.01) and/or osteoporotic fractures (1% vs 0.2%, p<0.01). By MPN subtypes, these comorbidities were prevalent in 9%, 12%, 6%, and 11% of PV, ET, MF, and MPN-U patients respectively. After adjusting for all baseline characteristics in a multivariable logistic regression model, MPN diagnosis (OR=1.34, 95% CI, 1.25-1.44), age≥70 (OR=1.62, 95% CI, 1.44-1.82), female sex (OR=7.73, 95% CI, 6.97-8.59), Asian race (OR=2.19, 95% CI, 1.85-2.57), and CCI≥3 (OR=1.54, 95% CI, 1.40-1.70) were significant predictors of osteoporosis and/or osteoporotic fractures. In a separate model that additionally considered MPN subtype, increased risk of these comorbidities was observed with PV (OR=1.27, 95% CI, 1.15-1.40), ET (OR=1.44, 95% CI, 1.33-1.57), and MPN-U (OR=1.72, 95% CI, 1.32-2.22). Conversely, there was no significant association with MF (OR=0.997, 95% CI, 0.83-1.19). Conclusion: Patients with MPNs, especially PV, ET, and MPN-U, are more likely to have osteoporosis and/or osteoporotic fractures than adults without MPNs. This novel relationship adds to previous studies which have reported a significant association between comorbidities of “inflammaging” and myeloid malignancies. Interestingly, the increased odds of comorbid osteoporosis and related fractures were not observed for patients with MF. Prior investigations have implicated differential impact of MPN subtypes on bone phenotype. As opposed to low bone density and osteoporosis, increased trabecular volume and osteosclerosis have been described for MF. Further research on the distinct mechanisms that impact bone health in MPNs is warranted. In addition, whether therapies that target MPN driver mutations or modify inflammation could reduce disruption of the bone marrow stroma that disturb bone health remains an important question.