PS1468 IMPACT OF CYTOREDUCTIVE DRUGS ON SECOND CANCER IN MYELOPROLIFERATIVE NEOPLASMS

Background: The evidence that currently used cytoreductive agents may affect the risk of second cancer (SC) remains uncertain because patients with myeloproliferative neoplasms (MPN) may have an intrinsic propensity to develop new malignancies. Aims: We assessed the influence of exposure to cytoreductive drugs on the occurrence of SC in a large multicenter international nested case‐control study. Methods: Cases (n = 647) were polycythemia vera (PV, n = 216), essential thrombocythemia (ET, n = 317) and myelofibrosis (MF, n = 114) patients with SC, and controls (n = 1,234) were MPN patients who did not present or develop SC during a comparable observation period. For each case, up to 3 controls who were SC‐free at the index date (date of SC occurrence in cases) were matched by each center for sex, age at MPN diagnosis (± 5 years), date of MPN diagnosis (± 6 years) and MPN disease duration (± 3 years). Results: Cases The most frequent category of SC was represented by carcinoma (n = 426, 65.8%) with a trend towards an higher frequency in patients with ET and PV compared to MF (p = 0.079). In MF patients, carcinoma occurred closer to MPN diagnosis (p = 0.062). Hematological SC (HSC), non‐melanoma skin cancer (NMSC) and melanoma were 62, 127 and 32, respectively. In 25.8% of MF patients, HSC were synchronous with MPN diagnosis, and indolent non‐Hodgkin lymphoma and CLL were prevalent (10 out of 16 patients). Drug exposure Hydroxyurea : Overall, the percentage of cases treated with hydroxyurea (HU) was similar to that of controls (p = 0.793) both in monotherapy as well as in combination with other drugs (p = 0.310). In a cancer‐specific stratified multivariable model, HU showed a two‐fold higher risk of NMSC irrespective of line of treatment (OR = 2.28, 95% CI 1.15–4.51). Pipobroman : Cases with SC were significantly more exposed to this drug than controls (p = 0.018). The independent role of this drug in this association was found in multivariable analysis (OR = 2.10, 95% CI 1.09–4.06) particularly for NMSC (OR = 3.74, 95% CI 1.00 ‐ 14.01) Ruxolitinib : In 17 patients, this drug was given in first‐line monotherapy (n = 16 MF and n = 1 ET) and the exposure was greater in cases than in controls (p = 0.033) The association of ruxolitinib with SC was significantly higher in patients with NMSC whose risk was almost 4‐fold higher than in non‐exposed (OR = 3.87, 95% CI 1.18–12.75). IFN, busulfan, anagrelide : The proportion of patients treated with these drugs did not differ in comparison with controls. No association of these drugs with the risk of overall SC was observed in multivariable model. Impact of SC on MPN‐related outcomes After the occurrence of SC, 647 cases were followed with a median of 3.0 years (interquartile range: 1.1–5.3). Major thrombosis was recorded in 10.5%, corresponding 3.1% pts/year. Major bleeding, evolution in acute leukemia/myelofibrosis were documented in 5.3% and 5.7%, respectively. Summary/Conclusion: This nested case‐control study performed in rare diseases such as MPN, represents the most efficient way, among retrospective studies, to identify the association between drug exposure and outcomes. Further strength of this survey is the large number of patients recruited and followed up in a network of 30 European hematological centers.