PALB2 as a familial gastric cancer gene: is the wait over?

Gastric cancer (GC) is the third leading cause of cancer-related mortality, accounting for >730,000 annual deaths worldwide¹. Prognosis is dismal as most gastric tumors are diagnosed in late stages where 5-year survival rates are <20%¹. While gastric tumors with intestinal histology are associated with a history of Helicobacter pylori-associated gastritis, the natural history of diffuse GCs is less understood¹. New tools for early GC detection that improve survival outcomes are therefore needed. Interestingly, about 1 in 10 GC patients report a family history of malignancy, suggesting an underlying genetic predisposition^1,2. Until recently, CDH1 was the only known familial GC gene³. CDH1 mutations are the cause of hereditary diffuse GC (HDGC) and this gene is routinely used for risk assessment and management^1,3. However, >40% HDGC families do not carry CDH1 mutations, suggest the existence of additional GC genes^4,5. In the last 3 years, we and others have suggested that mutations in homologous recombination (HR) repair genes, such as PALB2, likely explain a significant fraction of inherited GC ^4,6. In this issue, the study led by Eleanor Fewings and Marc Tischkowitz⁷, further strengthens the role of PALB2 in GC predisposition.

To identify new GC genes, Fewings et al¹ carried out whole exome sequencing in 28 HDGC families no pathogenic variants in CDH1 (CDH1-NPV). Exome data in CDH1-NPV families was used to prioritize candidate genes with loss of function (LOF) variants using interaction analyses. A cluster of DNA repair genes, which included known cancer genes such as MSH2 and PALB2, was identified as significantly enriched for LOFs. Subsequent analyses in the DNA repair cluster provided evidence for LOF variant co-segregation with disease in of the 28 CDH1-NPV families. One family had a frameshift PALB2 variant, one had LOFs in both ATR and NBN, 2 had RECQL5 mutations and 2 had LOF variants in MSH2, a known Lynch Syndrome (LS) gene⁸. Tumor analyses in the latter two MSH2 families did not reveal microsatellite instability or loss of mismatch repair proteins, LS hallmarks⁸, suggest that these variants were either non-pathogenic or that a new mechanism leads to GC in MSH2 families. To exclude a possible environmental cause, these families were evaluated for history of Helicobacter pylori infection, the strongest known GC risk factor¹, and most mutation carriers had previously tested negative for the bacteria. Fewings et al also performed combined analyses with data from the two previous reports^4,6 and found that PALB2 LOF variants are >7.5-fold more common in HDGC families than in the general population.

GC genetic studies have lagged behind those in other GI cancers such as colorectal cancer (CRC), where several susceptibility genes have been reported⁹. The study of GC genetics is difficult in part because the high mortality rates make family studies unfeasible. Furthermore, unlike CRC, GC is more commonly diagnosed in developing countries, where the research and practice of cancer genetics is more limited¹⁰. Fewings et al findings, therefore, should be commended. However, given the limited number of PALB2 families (only 10 have been reported), facilitating the translation of these findings to GC prevention will benefit from additional research. For example, should familial non-HDGC cases be tested for PALB2? our study⁶ found PALB2, suggesting that the PALB2 GC clinically different than HDGC.¹¹Further studies focusing on familial non-HDGC are therefore needed to understand if PALB2 mutations define a unique familial GC syndrome. Further, studies on other PALB2 cancer types families GC s^12,13. It is now important to identify which factors increase GC risk in such families. Both Fewings et al and our study⁶ found limited evidence for a modifier role of Helicobacter pylori in PALB2 mutation carriers but it is possible that risk could be affected by mutation hotpots, modifier genes and other unaccounted environmental factors. These questions clearly highlight the need for further research. This emerging body of data however provides a compelling case for offering PALB2 testing to CDH1-NPV families. Given the fact that gastric tumors of PALB2 carriers are likely to be HR repair deficient⁶ (as shown in our study), these individuals could potentially benefit from PARP inhibitor therapies¹⁴. 20 years after the discovery of CDH1, the emergence of PALB2 as a new familial GC gene has therefore dual importance for both prevention and treatment of GC.