Biology and Therapeutic Targets of Colorectal Serrated Adenocarcinoma; Clues for a Histologically Based Treatment against an Aggressive Tumor

Abstract

Serrated adenocarcinoma (SAC) is a tumor recognized by the WHO as a histological subtype accounting for around 9% of colorectal carcinomas. Compared to conventional carcinomas, SACs are characterized by a worse prognosis, weak development of the immune response, an active invasive front and a frequent resistance to targeted therapy due to a high occurrence of KRAS or BRAF mutation. Nonetheless, several high-throughput studies have recently been carried out unveiling the biology of this cancer and identifying potential molecular targets, favoring a future histologically based treatment. This review revises the current evidence, aiming to propose potential molecular targets and specific treatments for this aggressive tumor.

Keywords: angiogenesis; colorectal cancer; immune response; invasive front; molecular targets; serrated adenocarcinoma.

Figure 1

Differential histological features of histological subtypes of colorectal carcinoma. (A) Conventional carcinoma (CC) showing a cribriform gland pattern with basophilic cytoplasm, non-stratified nuclei, lobular dirty necrosis (centre) and lymphocytic infiltrates (upper-left corner). (B) Serrated adenocarcinoma (SAC) with typical vesicular stratified nuclei, eosinophilic cytoplasm and serrated lumen (left gland) with weak lymphocytic infiltration, tumor budding (black arrows) and desmoplastic stroma. (C) Colorectal carcinoma (CRC) with histological and molecular features of microsatellite instability (hmMSI-H) characterized by a medullar “solid” pattern and abundant intraepithelial tumor-infiltrating lymphocytes (black arrows) 20× original magnification. (Source: Authors).

Figure 2

Differentially enriched functions in the SAC and CC transcriptomes. (A) Venn diagram displaying the number of common genes up- or down-regulated in SAC and CC compared to normal adjacent mucosa. Differential expression analysis has been performed on normalized data with the Linear Models for Microarray Data (Limma) package by Bioconductor [29]. Using Limma, two comparisons were made between 5 normal and 12 CC samples and the same 5 normal and 13 SAC samples. (B) The Kegga method from Limma was used to perform this analysis, which obtained 49 pathways significantly enriched for shared differentially expressed genes in both groups: 26 pathways for CC (5 of them shared) and 22 pathways for SAC (3 of them shared). As expected, there were no shared pathways between the different genes of the groups CC and SAC. For the false discovery rate (FDR), the Benjamini and Hochberg method was used to get corrected p-values [30]. Differentially enriched functions in CC (C), including Wnt-signalling and in SAC (D), including those lifestyle-, neural-, immune-hypoxia-related (D) are shown. The VEGF signalling pathway was close to significance in SAC, with 7 differentially expressed genes (MAPKAPK3, VEGFA, PIK3R2, RAF1, PLA2G4C, PPP3CB and PRKCB) and some related hypoxia-associated pathways (AMPK and mTOR signalling). FSCN1 is differentially expressed in the comparison Normal vs. SAC, but it is not in Normal vs. CC.

Figure 3

Fascin1 expression in SAC-staining tumor-budding cells (black arrows) and abundant surrounding microvessels (white arrows) 20× original magnification. (Source: Authors).