66P PARPi and myeloid neoplasia: The Italian MITO-MaNGO experience based on a multicentric survey

The risk of PARPi related myeloid neoplasias (PrMN) in PARPi treated patients has been a growing concern. Morice's metanalyses, taking into consideration the trials and the World Health Organization (WHO) Vigibase, has shown a significant elevation of the risk (Peto OR 2,63 [95% CI 1,13–6,14], p=0,026) with no inter study heterogeneity. While few mono-centric studies have been published, data on a wider scale is lacking. Our aim was to gauge the incidence of PrMN in a real-life setting. A survey of 71 items was proposed to 17 MITO and MaNGO centers. Each center counted all patients ever treated with PARPi within the standard of care. Details on the choice of PARPi, line of treatment, length of therapy and BRCA mutational status were collected. Data cutoff was December 2023. A total of 2320 patients were collected (1254 BRCA mutated). Out of this number, 56 myeloid neoplasias were diagnosed, 35 MDS and 21 AML respectively (2.55%). Two patients had both MSD and AML. Thirty-two were BRCA 1 or 2 mutated, (2.5% of the total). Thirty-two had received Olaparib (resulting in an incidence of 2.5% in the total of patients treated with Olaparib at any line), 19 had received Niraparb (2%) and 4 (3.4%) had received Rucaparib respectively. The length of PARPi therapy before the diagnoses did not show a direct link between a longer exposure to PARPi and a higher risk of PrMN, with 7.4% of patients developing MSD or AML before six months and 20.4% after 6 to 12 months of maintenance therapy. Of all patients treated at each line, 0.52% developed a myeloid neoplasia after receiving PARPi in the first line, 4.2% in the 2°, 1.8% in the 3°, 10.8% in the 4° and 12.2% over the 4° line. Regarding the outcome there were 4 remissions, 4 partial responses, 8 progressions and 37 deaths. While still considered a rare collateral effect, PrMN have a much worse clinical outcome than non-therapy related MN which was confirmed. While PrMN did not seem more present in BRCA mutated patients than in BRCA wild type (2.5% vs 2.6%), the second, fourth and over the fourth lines of treatment had higher percentages of incidence. No difference in risk among PARPi was noted. As we move towards a better outcome for OC patients, it is paramount to identify higher risk cases and understand how to treat them accordingly.