Phenotypic subtypes associated with distinct therapeutic targets: Towards personalized medicine for patients with colorectal cancer


Session type:


Antonia Roseweir1,James Park1,Meera Patel1,Arfon Powell2,Paul Horgan1,Donald McMillan1,Joanne Edwards1
1University of Glasgow,2Cardiff University



Colorectal cancer (CRC) is the third most common cause of cancer death in the UK and current TNM-staging methods are suboptimal leading to a 60% 5-year survival rate. Therefore, there has been enormous effort to develop a prognostic classification of CRC. As a result, we recently proposed a novel histology-based classification of CRC, phenotypic subtypes, that group patients into four validated independently prognostic groups (immune, canonical, latent and stromal). The aim of the present study was to assess the predictive value of this classification by assessing protein expression levels of multiple cancer-associated pathways to determine distinct therapeutic targets for each group.


A cohort of 863 patients with stage I-III CRC were assessed for protein expression levels of previously studied clinical and pathological features, inflammatory features and signaling pathways to determine features distinct to each phenotypic subtype.


The immune subtype, which has the best prognosis, only showed associations with inflammatory characteristics including increased CD3+ and CD8+ lymphocyte infiltration (p<0.001) suggesting these patients may respond to immunotherapy. The stromal subtype, which has the worst prognosis, associated with increased metastasis (P<0.001) and increased STAT3 levels (p=0.040) suggesting JAK inhibitors may be promising in these patients. The canonical and latent subtypes had intermediate prognosis. The latent subtype associated with high IKKa(p=0,033) and low mannose phosphate isomerase (p=0.008) levels suggesting either IKKainhibitors or mannose may be potential treatments for this group. Finally, the canonical subtype had the most abundant elevated markers including SFK (p<0.001) and AKT (p=0.009) pathway members, therefore Dasatinib or BMK120 may be potential therapeutics within this group.


Phenotypic subtypes associated with distinct features for each subgroup and could be utilized as predictive biomarkers for potential therapeutic interventions targeting these features. Phenotypic subtypes represent a potential step towards precision medicine for patients with CRC.