1. Introduction
Colorectal cancer is a heterogeneous disease that can develop in different parts of the colon, with consequent differences in terms of risk factor, histological grade, tumor size and metastatic features [
1,
2]. Left-sided tumors (originating in the splenic flexure, descending colon, sigmoid colon, rectum, or one-third of the transverse colon) derive from the embryonic hindgut, whereas right-sided tumors (originating in the appendix, cecum, ascending colon, hepatic flexure, or two-thirds of the transverse colon) derive from the embryonic midgut [
3].
The different origins consequently lead to tumors with a different gene expression and mutation profile. In particular, right-sided tumors show a higher frequency of
BRAF mutation and microsatellite instability and more often occur in patients with a genetic predisposition to colorectal cancer (e.g., Lynch syndrome). Conversely, left-sided tumors are characterized by chromosomal instability and a gene expression profile involving the activation of the epidermal growth factor receptor (EGFR) pathway [
2,
4]. These differences result in different prognoses for the two tumor types, with right-sided tumors associated with poorer patient outcome [
2,
3,
4].
In addition to its prognostic relevance, there is evidence to suggest that tumor localization may be predictive of treatment efficacy with targeted agents, especially those directed against EGFR and vascular endothelial growth factor (VEGF) pathways [
3,
5,
6,
7,
8,
9,
10]. Although data on this specific topic are discordant due to the heterogeneity of the studies carried out, left-sided
RAS wild type (wt) tumors appear to be more responsive to EGFR inhibitors, possibly due to the higher frequency of
BRAF mutations in right-sided disease [
2,
7,
9,
11]. Results on the efficacy of bevacizumab (B) are even more conflicting, some studies finding no correlation with respect to tumor position [
12,
13] and others, conversely, reporting a link between the effectiveness of the monoclonal antibody and the side of the colon affected [
13,
14]. Among the latter, some authors found that left-sided or rectal tumors benefited more from B-based treatment [
14], whereas others observed that the drug prolonged progression-free (PFS) and overall survival (OS) in right- rather than left-sided tumors [
13].
Our extensive research into colorectal cancer revealed a correlation between different biomarkers involved in angiogenic and inflammatory processes and B efficacy [
15,
16,
17,
18,
19,
20], but we have never focused on these different markers in relation to tumor localization. We thus decided to investigate B efficacy, the distribution of a series of parameters involved in angiogenesis and inflammatory processes, and
RAS and
BRAF mutations in relation to tumor localization in a case series of metastatic colorectal cancer patients enrolled in the phase III multicenter, prospective, randomized “Italian Trial in Advanced Colorectal Cancer (ITACa)” trial [
21] The ITACa trial is registered on ClinicalTrials.gov (NCT01878422).
3. Discussion
The ITACa trial, a prospective multicenter randomized phase III study whose aim was to evaluate the improvement in PFS obtained by adding B to CT with respect to CT alone, reported similar PFS and OS in the CT and CT + B treatment arms [
21]. In the present study, we confirmed the lack of difference in survival between the CT + B and CT-only groups in a subgroup of ITACa patients. Of note, however, a benefit from the addition of B to CT was only seen in right-sided tumors, in agreement with results from the CALGB/SWOG 80405 trial in which treatment with cetuximab led to a clear improvement in PFS and OS in patients with left-sided
KRAS wt tumors, while B significantly improved survival parameters in those with right-sided lesions [
13]. In our study the benefit from B in right-sided tumors was evident for both PFS and OS but was only significant for PFS. However, given that ITACa trial patients who received first-line CT alone went on to receive second-line CT + B, it can be hypothesized that OS may have been impacted by the subsequent lines of therapy.
The distinct molecular features characterizing right- and left-sided tumors account for their different sensitivity to targeted drugs. First, the higher frequency of
BRAF and
RAS mutations found in right-sided tumors may explain the higher activity of EGFR inhibitors in left-sided lesions as these alterations represent well known mechanisms of resistance to the drug. However, EGFR inhibitor activity has been shown to be higher in left-sided tumors even when all
RAS wt patients are considered [
3,
13], suggesting that other molecular mechanisms such as higher EGFR ligand expression (epiregulin and amphiregulin) might be involved in left-sided lesions [
2,
22]. The reported higher frequency of
BRAF mutations in right-sided compared to left-sided tumors, also confirmed in this study, may partially explain the lower responsiveness of right-sided lesions to EGFR inhibitors. Another important factor is histological type. Mucinous cancer is a CRC subtype more frequently found in females and in the right colon. Although this tumor has been associated with poor outcome, its real clinical importance, especially with regard to its response to targeted agents, needs further investigation. There were only three cases of mucinous cancer in our patient population, surprisingly all left-sided. We do not believe this aspect could have influenced clinical results.
Our previous studies focusing on identifying biomarkers predictive of B efficacy indicated that a series of parameters might be associated with different activity of the drug. In particular, we found that specific
eNOS polymorphisms were correlated with significantly higher PFS and OS in the CT + B group [
15]. Interestingly, the present study revealed that
eNOS polymorphisms associated with higher B efficacy were more frequent in right-sided tumors, reflecting their higher sensitivity to the drug. However, this result was of borderline significance and the biological reason why these polymorphisms were more frequently associated with right-sided tumors remains to be clarified in larger patient population.
We observed a significant difference between right- and left-sided tumors in terms of baseline inflammatory index values. In particular, higher levels of SII, NLR and PLR were found in patients left-sided tumors with respect to right-sided ones. Our previous results showed that, in the overall ITACa case series, patients with low systemic inflammatory indexes (especially low NLR) benefited the most from the addition of B to CT in terms of PFS, suggesting that low systemic inflammatory indexes are associated with an increase in B activity [
18]. These results may help to explain the higher activity of B in right-sided tumors in which lower values of SII, NLR and PLR were present. Conversely, we found that circulating levels of biomarkers associated with angiogenesis, i.e., eNOS and EPHB4, were higher in right-sided tumors with respect to left-sided ones, reflecting more marked angiogenesis that may correlate with a greater B efficacy.
It has been demonstrated that tumors with microsatellite instability (MSI), known to be more frequent in right-sided disease, are associated with a higher cytotoxic T-cell infiltration and higher microvessel density, suggesting a higher angiogenic capacity of tumors with this localization [
23]. MSI-high CRC with long interspersed nucleotide element-1 (LINE-1) hypomethylation has been seen to have a poor prognosis [
24,
25], suggesting a complex biological interaction between MSI and LINE-1 hypomethylation. Moreover, differences in the mucosal microbiota of patients who develop right- or left-sided colorectal cancer may also help to explain the different angiogenic and inflammatory properties of the two types of lesions [
26]. These findings attest to the better outcome of patients with right-sided tumors treated with B-based treatment.