Short Report

Pancreatic cancer (PC) is the fourth leading cause of cancer death with a five-year survival rate less than 10% [1]. Nearly 1 in 10 patients diagnosed with PC has familial clustering of PC, but only 10% could be attributed to high penetrance germline DNA alterations affecting DNA repair, mainly the homologous recombination (HR) pathway genes [2]. BRCA2 account for most inherited mutations in hereditary PC with a frequency ranging between 5 and 15% depending on population selection [2]. Germline aberrations affecting BRCA2 were also identified in sporadic cases [3]. Interest in germline HR-gene alterations has grown significantly in recent years with the advent of the poly ADP ribose polymerase inhibitors (PARPi). Tumors harboring HR pathway defect are highly responsive to PARPi due to synthetic lethality. This concept has been proven in clinical trials and this class of drugs have been approved in the treatment of patients with advanced PC harboring germline BRCA1/2 (gBRCA) mutation [4]. In vitro studies have also shown that the concept of synthetic lethality related to PARPi is not limited to BRCA1/2 and could be seen in tumors harboring defects in other HR genes [5]. Thus, there is some utility, in both cancer screening and treatment, to identifying other germline susceptibility aberrations in patients with PC mainly among the genes implicated in the HR pathway (BRCA and non-BRCA).

We retrospectively analyzed the data of 257 pancreatic adenocarcinoma patients who were treated in our institution and were referred to a germline DNA testing between 2017 and 2020. Patients were referred for testing based on national guidelines taking into account personal and family history of neoplasia (for example: Lynch syndrome related cancers and HR-pathway-gene related cancers). Their mean age at diagnosis was 60 years old (SD = 12.8). Fifty eight percent were female and 50% had a history of at least one first-degree relative with PC.

Germline DNA was extracted from peripheral blood and tested using a panel of 45 genes including among others HR genes (ATM-BRCA1-BRCA2-PALB2-BARD1-BRIP1-CHEK2-NBN- RAD51B- RAD51C-RAD51D) and MMR genes (MLH1-MSH2-MSH6-PMS2). The list of genes included in the panel differed between patients depending on the guidelines that were being followed at the time of each analysis. Patients have signed informed consent for the genetic analysis. Variant interpretation was based on Gustave Roussy database (internal database), Clinvar information and ACMG criteria for classification.

In total, 13.2% of patients (34/257) harbored germline PV/LPV. 4.5% (11/245) had BRCA2 alterations followed by ATM 4.4% (11/252), TP53 2.56% (1/39), CDKN2A 1.27% (3/236), BRCA1 1.2% (3/242), PALB2 0.85% (2/234), MSH6 0.43% (1/231), and MLH1 0.43% (1/232). Six of the 11 patients harboring germline alteration in ATM had a known family history of PC.

Overall, the results of this cohort reproduced what have been published in the literature concerning the frequency of gBRCA and non-BRCA HR-gene mutations. Among the latter group, ATM is the most frequently affected gene. The frequency of gATM deleterious mutations ranged between 3 and 5% in familial PC patients [6, 7]. It should be noted that nearly half of gATM patients in our cohort were sporadic PC. In two very large studies, ATM PV have been shown to confer a risk of PC 4 to 6 times compared to the general population [8,9,10]. Despite these findings, the role of gATM alterations in the current management of PC patients and their families is not well established yet, and the inclusion of ATM among the genes to be analyzed in patients with a family history of PC is still unclear. Nevertheless, 88% of the experts participating in the last international cancer of the pancreas screening consortium (CAPS) agreed to test for gATM (weak grade of recommendation = probably do it), and to use endoscopic ultrasound and biliary MRI for annual surveillance of mutation carriers with one PC affected first-degree relative [11]. Future prospective trials are needed to confirm the efficacy of this strategy.

In addition to their role in cancer risk, recent findings have suggested that ATM alteration may be of therapeutic significance in PC [12]. Pre-clinical studies have shown that ATM-mutant PC is highly sensitive to PARPi [13] but the clinical experience is still limited and trials addressing this issue are ongoing [14]. On the other hand, ATR-inhibitors are now being tested in clinical trial and showing even more promising results in ATM-deficient cells due to synthetic lethality [15,16,17].

In conclusion, our study showed that inherited ATM aberrations are as frequent as gBRCA2 PV/LPV. Treatment and prevention strategies in ATM patients and their families, however, are yet to be defined. Our cohort of 257 PC patients is an added value to the literature but the results should be interpreted with caution since patients included in this study were referred for germline testing based on national guidelines, taking into account personal and family history of neoplasia. The findings of this study could be included in future meta-analysis aiming at examining the role of ATM in PC susceptibility.