Statistics from Altmetric.com
Oesophageal cancer is a deadly malignancy and is highly heterogeneous at the histological, geographical and molecular levels. It is broadly divided into adenocarcinomas and oesophageal squamous cell carcinomas (OSCCs). OSCCs are particularly prevalent in Asia and Africa where the disease has been linked to tobacco smoking, alcohol consumption and population-specific genetic risk variants.1 Clinically, oesophagectomy is the standard of care for curative therapy, usually proceeded by neoadjuvant chemotherapy/concurrent chemoradiotherapy, or in some cases definitive chemoradiotherapy without surgery. Unfortunately, the majority of patients with OSCC will relapse, and many either present too late or are too frail for curative treatment, resulting in an overall 5-year survival rate of 20%. Recent genomic studies have revealed a plethora of genetic alterations in OSCC, including mutations in TP53, PIK3CA and chromatin modifier genes.2–4 Robust approaches are thus needed for accurately subtyping patients with OSCC, to tailor existing therapies and to identify subtype-specific driver alterations that may highlight opportunities for novel interventions. Interestingly, several genomic alterations in OSCC are also observed in squamous carcinomas from other body sites, including the lung.
Previous studies applying molecular profiling to predict prognosis of patients with OSCC have suffered from limitations challenging their widespread clinical applicability.5 ,6 These include the lack of demonstrated validation in multiple independent cohorts from different patient populations, and a requirement to measure large numbers of genes to achieve accurate classification. Technical challenges also exist, particularly in …