Genomic characterization and detection of potential therapeutic targets for peritoneal mesothelioma in current practice

BAP1 is the most frequently mutated gene found in mesothelioma (pleural and peritoneal), with about 30–50% of cases harboring somatic mutations. (AACR GENIE and COSMIC, February 2022) [28, 29]. Also, a significant proportion of PeM patients might be affected by the so-called ‘BAP1 tumor predisposition syndrome’ (BAP1-TPDS), as they are carriers of a germline BAP1 mutation [30]. Besides a predisposition for mesothelioma, these patients are also commonly affected by BAP1-inactivated melanocytic tumors, uveal melanoma, cutaneous melanoma and renal cell carcinoma [31]. In line with other studies, we found oncogenic BAP1 mutations in 32% of tumors in the current cohort, of which two patients were known carriers of a BAP1 germline mutation [7, 10, 11]. BAP1 encodes for the tumor suppressor protein ‘ubiquitin carboxyl-terminal hydrolase,’ which plays a role in several cellular processes involved in oncogenesis [32]. Though there are currently no treatments directly targeting BAP1, there are therapies targeting molecular pathways in which BAP1 is involved. BAP1 is associated with BRCA1 activation, thereby playing a key role in homologous recombination repair (HRR) [32‐34]. Similar to ATM deficient tumors, BAP1 and BRCA1 deficient tumors might be susceptible to PARP inhibition and promising results have been reported in a phase 2 clinical trial [35]. However, in vitro results of sensitivity to PARP inhibition and its relationship to BAP1 status are inconsistent [36‐38]. Another potential target is EZH2, which is upregulated in BAP1 deficient tumors. A preclinical showed increased sensitivity to EZH2 inhibition in BAP1 deficient mice [39]. A phase 2 trial including 74 patients with BAP1 deficient mesothelioma treated patients with PeM with the EZH2 inhibitor tazemetostat as a monotherapy [40]. A disease control rate of 51% at twelve weeks and 25% at 24 weeks was reported, but no complete and only two partial responses were observed. These modest responses do not seem to be related to BAP1 deficiencies and biomarkers to predict the response to tazemetostat have not yet been identified. Due to its involvement in HRR, BAP1 has also been studied as a biomarker for response to chemotherapy. Wildtype BAP1 has been associated with sensitivity to gemcitabine treatment in mesothelioma cell lines, but this has not been validated in patients with PeM [41, 42].

Based on several mutations in NF2, protein kinase inhibitors everolimus and temsirolimus could be a potential treatment option for 16% of patients in our cohort. NF2 is a tumor suppressor gene that plays an important role in cell proliferation and survival [43, 44]. NF2 is involved in the mammalian target of rapamycin (mTOR) signaling pathway. Inactivating mutations of NF2 lead to cell cycle progression and cell proliferation [45, 46]. NF2 mutations are reported by previous studies in around 25% of cases of PeM [10, 11]. Some clinical studies and some preclinical evidence suggest that NF2 inactivation might be associated with response to mTOR inhibitors.[47, 48] Everolimus and temsirolimus are both mTOR inhibitors and have been approved by the FDA for the treatment of neuroendocrine tumors of the gastro-intestinal tract or lung, HER2/neu-negative breast cancer and renal cell carcinoma, among others. A phase 2 study in pleural mesothelioma only showed a 2% response rate to everolimus [49]. This study, however, did not stratify patients based on mutational status. Considering that only about 15% of mesothelioma cases show mutations in NF2, the response rate might be higher when only these patients are included. However, some studies suggest that combination treatment might be indicated [50, 51].

Mutations in ATM were present in two patients in our cohort (11%), but were reported in only 2% of the patients in the large cohort of Hiltbrunner et al. [11]. Although rare, patients with PeM and mutations in ATM could benefit from treatment with PARP inhibitors. ATM is located on chromosome 11 and codes for the ATM serine/threonine kinase protein. This protein plays a role in the HRR pathway, among others by p53 activation, which has an important role in cell cycle arrest and apoptosis [52]. Mateo et al. found that deleterious ATM mutations in metastatic prostate cancer were associated with good response to olaparib, a PARP inhibitor that is approved for the treatment of several solid tumors in the European Union [35, 53]. However, the same group found no survival benefit for castration resistant prostate cancer patients, but these findings were the result of an underpowered interim analysis [54]. For other malignancies, such as gastric-cancer and renal cell carcinoma, similar relations between ATM mutations and response to PARP inhibition have been reported [55, 56]. Fennell et al. performed a phase 2 trial, treating 26 mesothelioma patients (25 pleural, 1 peritoneal) with the PARP inhibitor rucaparib after at least one cycle of systemic chemotherapy. They found a disease control rate of 58% at twelve weeks and 23% at 24 weeks, while toxicity was limited [57]. They selected patients with BAP1 and/or BRCA1 deficient tumors, other key proteins in HRR. HRR deficient tumors, such as ATM inactivated tumors, might have similar responses to PARP inhibition.