This study on a cohort of HPs and SSAs demonstrates, that HPs exhibit different immunophenotypes regarding to their localization (proximal vs. distal). The Immunohistochemical analysis included the following markers: AMACR, p16Ink, CK20, Ki67, BRAF, β-catenin and MLH1. We focussed on the expression of AMACR and p16Ink in HPs and SSAs with respect on lesion localisation since analysis of CK20, Ki67, BRAF, β-catenin and MLH1 either lacked significant results or resulted in unreliable data. All parameters analysed are involved in the serrated adenocarcinoma pathway. Besides the well known AMACR overexpression in prostate cancer , overexpression was also demonstrated in colorectal cancer . Moreover dysplastic epithelial cells in Barrett Oesophagus were shown to be associated with elevated AMACR levels .
Likewise SSAs, HPs have been shown to harbour specific genetic alterations such as somatic BRAF mutations and CPG island methylator phenotype (CIMP) [12, 13]. This suggests the progression of a subset of HPs towards SSAs. Thus one of our study aims was to identify potential biomarkers, that may allow more precise separation of HPs and SSAs and identify a subset of HPs that might qualify as precursor lesions of SSAs. However, the analysed biomarkers did not provide additional diagnostic information allowing a more distinct differentiation of HPs and SSAs. Recently Ki67 counting and distribution was shown to be different in HPs versus SSAs, with higher Ki67 values in SSAs . However, in our study we could not confirm this finding. This might be due to the fact, that in our study only a small number of SSAs were analysed. In addition in the study by Fujiimori et al. Ki67 evaluation was calculated using automated image processing software, whereas this was not the case in the present study.
Furthermore this study describes a significant overexpression of AMACR in left-sided distal HPs when compared to right-sided proximal HPs. Thus although morphological identically these lesions exhibit a varying AMACR immunophenotype. It is unlikely, that these distal HPs are at higher risk to progress to SSAs, since SSAs are typically located proximal in the right colon. Given the known function of AMACR regarding fat metabolism, it may be possible, that in the distal parts of the colon the mucosa is exposed to an increased amount of branched-chain fatty acids and fatty acid derivates.
Another possible explanation would be a specific supportive role of AMACR in the development of precursor lesions in the colorectal cancer carcinogenesis. This hypothesis is supported by recent data demonstrating increased AMACR expression in high grade dysplasia compared to low grade dysplasia in conventional adenomas .
Until now only few studies focussed on AMACR expression in gastrointestinal tumours. In a recent study on 1315 colorectal cancers AMACR overexpression was found to be associated with left-side tumour localisation in colorectal cancer . In detail, AMACR elevation was significantly associated with higher tumour differentiation grade (G1 and G2) and advanced tumour stage. In addition high AMACR expression levels were related to a tubular phenotype and less often to mucinous or signet cell carcinomas. These results presented by Marx et al. point towards a relevant role of AMACR expression at least in a subgroup of colorectal cancer and implicate a linkage of AMACR expression and site-related differences in metabolism/exposure to fatty acids.
In the present study in addition to differential expression of AMACR in HPs, we were able to detect significant differences in the expression of p16Ink4 in proximal and distal HPs. Similar to AMACR, p16Ink4 was overexpressed in distal HPs and significantly associated with AMACR overexpression. p16Ink4 is a well characterized protein with an important role in oncogene-induced cell aging [24, 25]. Upregulation of p16Ink4a was recently shown to function as an senescence barrier in the serrated route to colon cancer .
Taken together the overexpression of AMACR and p16Ink in distal HPs points toward a pathophysiological relevance of both these proteins. It is now necessary to identify in which processes AMACR and p16Ink in distal HPs are involved and if their dysregulation is implicated in the development of certain subgroups of adenomas.
Highly interesting in this context is a very recent study from Zhang et al. that analyses AMACR expression in normal mucosa, adenoma and colorectal carcinoma both on immunohistochemical and genetic level .
Zhang et al. reported AMACR negativity in normal colonic mucosa and tubular adenoma with low grade and intermediate dysplasia as well as in poorly differentiated carcinoma. In contrast AMACR overexpression was found in villous adenoma and high and moderately differentiated colon cancer. These results are in concordance with those of Marx et al. Using Laser-capture-microdissection Zhang et al. discovered different deletions in the AMACR promotor CpG Island, depending upon the underlying tissue: In normal colonic glands and tubular adenomas with low AMACR expression they detected a somatic double-deletion at CG3 and CG10, that was absent in villous adenomas and all colon cancers with variable AMACR levels. In contrast they identified a high prevalence (89%) of deletion of CG12-16 in moderately differentiated colon cancers with strong AMACR overexpression, whereas these deletions existed in only 14% of poorly differentiated colon cancer. This deletion of CG12-16 was shown to be a constitutional allele with a frequency of 43% in the general population.
Taken together, AMACR protein expression may be regulated by somatic or constitutional genetic alterations in AMACR Promoter CPG islands. Keeping the above mentioned findings in mind, it is now tempting to speculate, that distal HPs with increased AMACR expression might constitute precursor lesions of a pathway leading to the development of villous adenomas and/or subtypes of low grade colorectal cancers.