Subtype‐Specific Risk of Non‐Melanoma Skin Cancers in MPN Patients: An Expanded Analysis of the MPN ‐K Case–Control Database

Patients with polycythemia vera (PV), essential thrombocythemia (ET), and myelofibrosis (MF) are known to have a two- to three-fold increased risk of developing non-melanoma skin cancer (NMSC) compared with the general population 1-3. This heightened risk has been attributed to both the chronic inflammatory state inherent to myeloproliferative neoplasms (MPNs) and exposure to cytoreductive therapies, especially hydroxyurea (HU) and ruxolitinib 4, 5. Population-based studies consistently report standardized incidence ratios for NMSC in MPN patients ranging from 2.4 to 3.3 across various MPN subtypes 1, 6. In particular, the large MPN-K nested case–control study, involving 1881 patients, demonstrated significant associations between NMSC and exposure to HU (OR 2.28, 95% CI 1.15–4.51), ruxolitinib (OR 3.87, 95% CI 1.18–12.75), and combination cytoreductive therapy (OR 3.47, 95% CI 1.55–7.75) 4. Real-world cohort studies have further corroborated these findings, particularly highlighting ruxolitinib's association with an increased risk of squamous cell carcinoma (SCC) in PV and MF patients 5. Despite these observations, several important knowledge gaps remain. Most studies to date have analyzed NMSC as a single entity, potentially overlooking the distinct biological and clinical characteristics of basal cell carcinoma (BCC) and SCC 1, 4. Moreover, although BCC and SCC have markedly different metastatic potential and mortality in the general population, SCC carrying a metastasis rate of 1.2%–1.9% compared with approximately 0.004% for BCC, the comparative prognosis of these two skin cancers in MPN patients is yet to be fully characterized 7, 8. Additionally, the long-term clinical outcomes following BCC and SCC diagnoses in MPN patients, including survival, relapse, and disease progression, have not been thoroughly explored. Understanding the subtype-specific risks and outcomes is critical for optimizing treatment selection, developing surveillance strategies, and providing effective patient counseling. This is especially relevant for patients who develop skin cancers while undergoing HU or ruxolitinib therapy and who continue to require cytoreduction to control their myeloproliferative disease 4. In a prior study (MPN-K study 4), we reported the risk of NMSC in MPN patients treated with cytoreductive drugs, using data from a large international nested case–control study that included 1881 MPN patients. The study compared 647 cases of second cancers (SC) with 1234 MPN controls matched for sex, age at MPN diagnosis (±3 years), date of diagnosis, and disease duration. In this cohort, HU was associated with a two-fold increased risk of NMSC as a single entity (OR 2.28, 95% CI 1.15–4.51), but the study did not differentiate between BCC and SCC. This research letter aims to expand upon our previous findings by analyzing the MPN-K database to identify subtype-specific risk factors for BCC and SCC in MPN patients, with a particular focus on associations with cytoreductive therapies, including HU and ruxolitinib. Additionally, we sought to compare clinical characteristics and outcomes between MPN patients who developed BCC versus SCC, including relapse, thrombosis, disease progression, and mortality. Finally, we aimed to characterize the long-term clinical outcomes of MPN patients following cancer diagnosis to better inform clinical decision-making regarding surveillance and treatment strategies. Our analysis included 127 MPN patients who developed NMSC, consisting of 78 with BCC and 49 with SCC, along with 244 matched MPN controls who did not experience skin cancer (Table 1). In the BCC group, age at MPN diagnosis and sex distribution were similar between cases and controls. However, there was a significant difference in the distribution of MPN diagnoses (p = 0.048), with a lower proportion of PV among BCC cases and a higher frequency of post-ET and post-PV MF in this cohort. The distribution of driver mutations (JAK2, CALR, and MPL), abnormal karyotypes, cardiovascular risk factors, splenomegaly, and thrombosis history was comparable between cases and controls. Treatment exposure in the period between MPN diagnosis and skin cancer, including aspirin, antithrombotic therapy, HU, and ruxolitinib, did not differ significantly between the groups. In the SCC group, no significant differences were observed in demographic characteristics, MPN subtype, mutational profile, cytogenetics, cardiovascular risk factors, splenomegaly, thrombosis history, or treatment exposures. While HU exposure in the period between MPN diagnosis and skin cancer was numerically higher among SCC cases compared with controls (85.7% vs. 73.7%), this difference did not reach statistical significance (p = 0.099). HU exposure during the period between MPN diagnosis and skin cancer development was numerically higher in SCC cases compared with controls (85.7% vs. 73.7%), though this difference did not reach statistical significance (p = 0.099). In contrast, ruxolitinib exposure was significantly higher in SCC cases than controls (10.2% vs. 2.0%, p = 0.040). In multivariable analysis (Table 2), increasing age at MPN diagnosis was the only significant factor associated with the development of BCC (OR 1.29, 95% CI 1.07–1.55; p = 0.008). Cardiovascular risk factors, thrombotic events, and type of treatment were not significantly associated with BCC. While HU exposure was not associated with BCC (OR 1.88, 95% CI 0.85–4.16; p = 0.117), ruxolitinib demonstrated a trend toward a stronger association (OR 4.07, 95% CI 0.89–18.63; p = 0.071). In contrast, for SCC, HU exposure was associated with a three-fold increased risk (OR 3.04, 95% CI 1.00–9.29; p = 0.050), and ruxolitinib exposure had an even stronger association (OR 6.54, 95% CI 1.07–39.83; p = 0.042), though this estimate was of low precision. In the subgroup of 127 NMSC cases, a prospective analysis was also conducted. During a median follow-up of 9.2 years (IQR 5.9–14.0) after skin cancer diagnosis, thrombosis and bleeding occurred in 11.8% and 3.1% of patients, respectively, with no significant differences between the BCC and SCC groups (Table S1). However, a trend toward higher progression to overt MF was observed in SCC patients compared with those with BCC (18.4% vs. 7.7%; p = 0.091). Mortality was higher in SCC patients (26.5% vs. 12.8%, p = 0.051), and skin cancer relapse occurred in 14.3% of cases overall, more frequently in SCC (20.8% vs. 10.3%, p = 0.12). Most deaths were attributed to MPN progression or other non-skin cancer causes, leading to significantly worse overall survival in SCC compared with BCC (p = 0.028) (Figure 1A). Moreover, composite event-free survival (considering thrombosis, bleeding, progression to MF, skin cancer relapse, and death) was lower in SCC patients compared with BCC patients (p = 0.16) (Figure 1B). In summary, this analysis of the MPN-K nested case–control database provides valuable new insights into the distinct risks and outcomes associated with BCC and SCC in patients with MPNs. The study highlights important differences in treatment exposures. HU, a well-known photosensitizer, may contribute to increased skin cancer risk by impairing DNA repair mechanisms 6, while ruxolitinib has been shown to potentially compromise immune surveillance, thereby promoting the development of SCC 5, 9. Additionally, patients diagnosed with SCC exhibited significantly worse overall survival, suggesting a more aggressive clinical course and reinforcing the need for heightened vigilance and tailored management in this group 9. Although the retrospective design and relatively small sample size pose limitations to the generalizability of these findings, the matched case–control design helped control for important confounders, providing a robust foundation for the observed associations. Currently, ropeginterferon alfa-2b is the only alternative for patients requiring cytoreductive therapy, and a post-marketing observational study in this setting is ongoing 10. The authors have nothing to report. The authors have nothing to report. The authors have nothing to report. The authors declare no conflicts of interest. Aggregated data available by request. Patient-level data will not be shared. Table S1: Major outcomes after BCC and SCC. 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.