Search

Programmed cell death ligand 1 (PD-L1) on tumor cells is a significant prognostic biomarker for a number of malignancies, although less is known about the significance of PD-L1 positive immune cells in colon carcinoma. The purpose of this study is to evaluate the role of PD-L1 in a large cohort of colon carcinomas to identify patterns of PD-L1 expression in the tumor microenvironment and its correlation with other key immune subsets to better understand the impact of these immune cells. We assessed 1218 colon carcinomas on representative tissue microarray sections, gathered relevant clinicopathologic information, and performed immunohistochemical staining for mismatch repair proteins, CD8, CD163, LAG3, PD-L1, FoxP3, and BRAF V600E. We then performed automated quantification; manual quantification was used for PD-L1 tumor cells and immune cells. Dual PD-L1/PU.1 immunostain was also performed. The majority of PD-L1 positive cells expressed PU.1 thus representing tumor-associated macrophages. Based on the median number of PD-L1 positive immune cells (7.6/mm

Purpose:Perineural invasion (PNI) is associated with aggressive tumor behavior, recurrence, and metastasis, and can influence the administration of adjuvant treatment. However, standard histopathologic examination has limited sensitivity in detecting PNI and does not provide insights into its mechanistic underpinnings.Experimental Design:A multivariate Cox regression was performed to validate associations between PNI and survival in 2,029 patients across 12 cancer types. Differential expression and gene set enrichment analysis were used to learn PNI-associated programs. Machine learning models were applied to build a PNI gene expression classifier. A blinded re-review of hematoxylin and eosin (H&E) slides by a board-certified pathologist helped determine whether the classifier could improve occult histopathologic detection of PNI.Results:PNI associated with both poor overall survival [HR, 1.73; 95% confidence interval (CI), 1.27–2.36; P < 0.001] and disease-free survival (HR, 1.79; 95% CI, 1.38–2.32; P < 0.001). Neural-like, prosurvival, and invasive programs were enriched in PNI-positive tumors (Padj < 0.001). Although PNI-associated features likely reflect in part the increased presence of nerves, many differentially expressed genes mapped specifically to malignant cells from single-cell atlases. A PNI gene expression classifier was derived using random forest and evaluated as a tool for occult histopathologic detection. On a blinded H&E re-review of sections initially described as PNI negative, more specimens were reannotated as PNI positive in the high classifier score cohort compared with the low-scoring cohort (P = 0.03, Fisher exact test).Conclusions:This study provides salient biological insights regarding PNI and demonstrates a role for gene expression classifiers to augment detection of histopathologic features.

Supplementary Data Legends

Supplementary Data from Agrin Loss in Barrett's Esophagus-Related Neoplasia and Its Utility as a Diagnostic and Predictive Biomarker

Purpose:There is an unmet need for identifying novel biomarkers in Barrett's esophagus that could stratify patients with regards to neoplastic progression. We investigate the expression patterns of extracellular matrix (ECM) molecules in Barrett's esophagus and Barrett's esophagus–related neoplasia, and assess their value as biomarkers for the diagnosis of Barrett's esophagus–related neoplasia and to predict neoplastic progression.Experimental Design:Gene-expression analyses of ECM matrisome gene sets were performed using publicly available data on human Barrett's esophagus, Barrett's esophagus–related dysplasia, esophageal adenocarcinoma (ADCA) and normal esophagus. Immunohistochemical expression of basement membrane (BM) marker agrin (AGRN) and p53 was analyzed in biopsies of Barrett's esophagus–related neoplasia from 321 patients in three independent cohorts.Results:Differential gene-expression analysis revealed significant enrichment of ECM matrisome gene sets in dysplastic Barrett's esophagus and ADCA compared with controls. Loss of BM AGRN expression was observed in both Barrett's esophagus–related dysplasia and ADCA. The mean AGRN loss in Barrett's esophagus glands was significantly higher in Barrett's esophagus–related dysplasia and ADCA compared with non-dysplastic Barrett's esophagus (NDBE; P < 0.001; specificity = 82.2% and sensitivity = 96.4%). Loss of AGRN was significantly higher in NDBE samples from progressors compared with non-progressors (P < 0.001) and identified patients who progressed to advanced neoplasia with a specificity of 80.2% and sensitivity of 54.8%. Moreover, the combination of AGRN loss and abnormal p53 staining identified progression to Barrett's esophagus–related advanced neoplasia with a specificity and sensitivity of 86.5% and 58.7%.Conclusions:We highlight ECM changes during Barrett's esophagus progression to neoplasia. BM AGRN loss is a novel diagnostic biomarker that can identify patients with NDBE at increased risk of developing advanced neoplasia.

Supplementary Figure S1

Supplementary Table 1

Serrated adenocarcinoma (SAC), a recognised WHO variant of colonic adenocarcinoma, is the purported end-product of serrated neoplasia. However, the diagnosis of SAC is infrequently rendered, and little is known about its prognosis, immune microenvironment and molecular alterations.We assessed 903 consecutive colon carcinomas and recognised tumours with ≥ 5% (n = 77) serrated and ≥ 50% serrated patterns (n = 13). We assessed precursor polyps and synchronous polyps. We recorded demographic/clinical parameters, histological features and mismatch repair (MMR) status. We performed immunohistochemistry and quantification on tissue microarray for HLA class I/II proteins, B2MG, CD8, CD163, LAG3, FoxP3, PD-L1 and BRAF V600E.We identified ≥ 5% epithelial serration prevalence in 8.5% of cases and ≥ 50% epithelial serration prevalence in 1.4% of cases. Precursor lesions were present in 21.4% of cases; these were mostly tubular adenomas with two traditional serrated adenomas identified. SAC with ≥ 5% serrations exhibited lower numbers of CD8-positive lymphocytes (P = 0.002) and lower B2MG expression (P = 0.048), although neither value was significant at ≥ 50% serration threshold. There was no difference in HLA class I/II, or PD-L1 expression on tumour cells and no difference in PD-L1, LAG3, FoxP3 and CD163 expression on immune cells. There was no association with MMR status, or BRAFV600E relative to conventional adenocarcinoma. There was improved disease-specific survival on univariate (but not multivariate) analysis between carcinomas with serrated pattern and non-mucinous conventional colonic carcinomas at ≥ 5% epithelial serrations (P = 0.04).SAC category shows a limited impact on survival, and this phenotype may harbour a unique immunological milieu.

Mucinous adenocarcinoma (MAD), the most common subtype of colonic adenocarcinoma (CA), requires50% intratumoral mucin. There is limited data regarding the impact of MAD on key lymphocyte subsets and therapeutically critical immune elements. In this study we address: (1) the definition of MAD, (2) grading of MAD, and (3) the impact of MAD and extracellular mucin on intratumoral immune milieu. Estimation of the percentage of intratumoral mucin was performed by two pathologists. Tissue microarrays were stained for immune markers including CD8, CD163, PD-L1, FoxP3, β2 microglobulin, HLA class I, and HLA class II. Immunohistochemistry for BRAF V600E was performed. MMR status was determined on immunohistochemistry for MSH2, MSH6, MLH1, PMS2. Manual and automated HALO platforms were used for quantification. The 903 CAs included 62 (6.9%) MAD and 841 CA with ≤ 50% mucin. We identified 225 CAs with mucinous differentiation, defined by ≥10% mucin. On univariate analysis neither cut point, 50% (p = 0.08) and 10% (p = 0.08) mucin, correlated with disease-specific survival (DSS). There were no differences in key clinical, histological and molecular features between MAD and CA with mucinous differentiation. On univariate analysis of patients with MAD, tumor grade correlated with DSS (p = 0.0001) while MMR status did not (p = 0.86). There was no statistically significant difference in CD8 (P = 0.17) and CD163 (P = 0.05) positive immune cells between MAD and conventional CA. However, deficient (d) MMR MADs showed fewer CD8 (P = 0.0001), CD163 (P = 0.0001) and PD-L1 (P = 0.003) positive immune cells compared to proficient (p)MMR MADs, a finding also seen with at 10% mucin cut point. Although MAD does not impact DSS, this study raises the possibility that the immune milieu of dMMR MADs and tumors with =10% mucin may differ from pMMR MADs and tumors with10% mucin, a finding that may impact immune-oncology based therapeutics.