Interobserver, intraobserver, and interlaboratory variability in reporting pT4a colon cancer

Personalizing adjuvant therapy for patients with colorectal cancer

The current standard-of-care adjuvant treatment for patients with colorectal cancer (CRC) comprises a fluoropyrimidine (5-fluorouracil or capecitabine) as a single agent or in combination with oxaliplatin, for either 3 or 6 months. Selection of therapy depends on conventional histopathological staging procedures, which constitute a blunt tool for patient stratification. Given the relatively marginal survival benefits that patients can derive from adjuvant treatment, improving the safety of chemotherapy regimens and identifying patients most likely to benefit from them is an area of unmet need. Patient stratification should enable distinguishing those at low risk of recurrence and a high chance of cure by surgery from those at higher risk of recurrence who would derive greater absolute benefits from chemotherapy. To this end, genetic analyses have led to the discovery of germline determinants of toxicity from fluoropyrimidines, the identification of patients at high risk of life-threatening toxicity, and enabling dose modulation to improve safety. Thus far, results from analyses of resected tissue to identify mutational or transcriptomic signatures with value as prognostic biomarkers have been rather disappointing. In the past few years, the application of artificial intelligence-driven models to digital images of resected tissue has identified potentially useful algorithms that stratify patients into distinct prognostic groups. Similarly, liquid biopsy approaches involving measurements of circulating tumour DNA after surgery are additionally useful tools to identify patients at high and low risk of tumour recurrence. In this Perspective, we provide an overview of the current landscape of adjuvant therapy for patients with CRC and discuss how new technologies will enable better personalization of therapy in this setting.

Current Challenges and Controversies in Colorectal Carcinoma Pathologic Staging-A Practical Guide

The pathologic assessment of colorectal carcinoma specimens plays a crucial role in the therapeutic management of patients and disease prognostication. The TNM staging system is used globally and is a critical component of colorectal carcinoma pathology reporting. However, our experience informs us that there are significant variations in the assignment of the TNM stage, both between pathologists and between hospital centers. We identify several potential reasons for this, among them suboptimal gross and microscopic assessment of colorectal resection specimens and, later, nonuniformity in applying criteria set forth in pathologic TNM staging guidelines. In addition, some defining characteristics of the staging system remain poorly defined. We aim to enlist those issues with potential remedies to improve reproducibility and, therefore, multidisciplinary discussion.

Assessment of Elastic Laminal Invasion Contributes to an Objective pT3 Subclassification in Colon Cancer

The extent of tumor spread influences on the clinical outcome, and which determine T stage of colorectal cancer. However, pathologic discrimination between pT3 and pT4a in the eighth edition of the American Joint Committee on Cancer (AJCC)-TNM stage is subjective, and more objective discrimination method for deeply invasive advanced colon cancer is mandatory for standardized patient management. Peritoneal elastic laminal invasion (ELI) detected using elastic staining may increase the objective discrimination of deeply invasive advanced colon cancer. In this study, we constructed ELI study group to investigate feasibility, objectivity, and prognostic utility of ELI. Furthermore, pT classification using ELI was investigated based on these data. At first, concordance study investigated objectivity using 60 pT3 and pT4a colon cancers. Simultaneously, a multi-institutional retrospective study was performed to assess ELI's prognostic utility in 1202 colon cancer cases from 6 institutions. In the concordance study, objectivity, represented by κ, was higher in the ELI assessment than in pT classification. In the multi-institutional retrospective study, elastic staining revealed that ELI was a strong prognostic factor. The clinical outcome of pT3 cases with ELI was significantly and consistently worse than that of those without ELI. pT classification into pT3 without ELI, pT3 with ELI, and pT4a was an independent prognostic factor. In this study, we revealed that ELI is an objective method for discriminating deeply invasive advanced colon cancer. Based on its feasibility, objectivity, and prognostic utility, ELI can subdivide pT3 lesions into pT3a (without ELI) and pT3b (with ELI).

Data synthesis and adversarial networks: A review and meta-analysis in cancer imaging

Despite technological and medical advances, the detection, interpretation, and treatment of cancer based on imaging data continue to pose significant challenges. These include inter-observer variability, class imbalance, dataset shifts, inter- and intra-tumour heterogeneity, malignancy determination, and treatment effect uncertainty. Given the recent advancements in image synthesis, Generative Adversarial Networks (GANs), and adversarial training, we assess the potential of these technologies to address a number of key challenges of cancer imaging. We categorise these challenges into (a) data scarcity and imbalance, (b) data access and privacy, (c) data annotation and segmentation, (d) cancer detection and diagnosis, and (e) tumour profiling, treatment planning and monitoring. Based on our analysis of 164 publications that apply adversarial training techniques in the context of cancer imaging, we highlight multiple underexplored solutions with research potential. We further contribute the Synthesis Study Trustworthiness Test (SynTRUST), a meta-analysis framework for assessing the validation rigour of medical image synthesis studies. SynTRUST is based on 26 concrete measures of thoroughness, reproducibility, usefulness, scalability, and tenability. Based on SynTRUST, we analyse 16 of the most promising cancer imaging challenge solutions and observe a high validation rigour in general, but also several desirable improvements. With this work, we strive to bridge the gap between the needs of the clinical cancer imaging community and the current and prospective research on data synthesis and adversarial networks in the artificial intelligence community.

The measured distance between tumor cells and the peritoneal surface predicts the risk of peritoneal metastases and offers an objective means to differentiate between pT3 and pT4a colon cancer

Substantial variability exists in what pathologists consider as pT4a in colorectal cancer when tumor cells are within 1 mm of the free peritoneal surface. This study aimed to determine if the measured sub-millimeter distance between tumor cells and the free peritoneal surface would offer an objective means of stratifying patients according to the risk of developing peritoneal metastases. Histological slides of patients included in the COLOPEC trial, with resectable primary c/pT4N0-2M0 colon cancer, were centrally reassessed. Specific tumor morphological variables were collected, including distance from tumor to free peritoneal surface, measured in micrometers (µm). The primary outcome, 3-year peritoneal metastasis rate, was compared between four groups of patients stratified for relation of tumor cells to the peritoneum: 1) Full peritoneal penetration with tumor cells on the peritoneal surface, 2) 0-99 µm distance to the peritoneum, 3) 100-999 µm to the peritoneum, and 4) ≥1000 µm to the peritoneum, by using Kaplan-Meier analysis. In total, 189 cases were included in the present analysis. Cases with full peritoneal penetration (n = 89), 0-99 µm distance to the peritoneal surface (n = 34), 100-999 µm distance (n = 33), and ≥1000 µm distance (n = 33), showed significantly different 3-year peritoneal metastases rates of 25% vs 29% vs 6% vs 12%, respectively (Log Rank, p = 0.044). N-category did not influence the risk of peritoneal metastases in patients with a tumor distance beyond 100 µm, while only the N2 category seemed to result in an additive risk in patients with a distance of 0-99 µm. The findings of this study suggest that the measured shortest distance between tumor cells and the free peritoneal surface is useful as an objective means of stratifying patients according to the risk of developing peritoneal metastases. This simple measurement is practical and may help in providing a precise definition of pT4a. Trial registration: NCT02231086 (Clinicaltrials.gov).

Advances in the management of peritoneal malignancies

Peritoneal surface malignancies (PSMs) are usually associated with a poor prognosis. Nonetheless, in line with advances in the management of most abdominopelvic metastatic diseases, considerable progress has been made over the past decade. An improved understanding of disease biology has led to the more accurate prediction of neoplasia aggressiveness and the treatment response and has been reflected in the proposal of new classification systems. Achieving complete cytoreductive surgery remains the cornerstone of curative-intent treatment of PSMs. Alongside centralization in expert centres, enabling the delivery of multimodal and multidisciplinary strategies, preoperative management is a crucial step in order to select patients who are most likely to benefit from surgery. Depending on the specific PSM, the role of intraperitoneal chemotherapy and of perioperative systemic chemotherapy, in particular, in the neoadjuvant setting, is established in certain scenarios but questioned in several others, although more prospective data are required. In this Review, we describe advances in all aspects of the management of PSMs including disease biology, assessment and improvement of disease resectability, perioperative management, systemic therapy and pre-emptive management, and we speculate on future research directions.

Computer extracted features of nuclear morphology in hematoxylin and eosin images distinguish Stage II and IV colon tumors

We assessed the utility of quantitative features of colon cancer nuclei, extracted from digitized hematoxylin and eosin-stained whole slide images (WSIs), to distinguish between Stage II from Stage IV colon cancers. Our discovery cohort comprised 100 Stage II and Stage IV colon cancer cases sourced from the University Hospitals Cleveland Medical Center (UHCMC). We performed initial (independent) model validation on 51 (143) Stage II and 79 (54) Stage IV colon cancer cases from UHCMC (The Cancer Genome Atlas's Colon Adenocarcinoma, TCGA-COAD, cohort). Our approach comprised the following steps, (1) a fully convolutional deep neural network with VGG-18 architecture was trained to locate cancer on WSIs, (2) another deep-learning model based on Mask-RCNN with Resnet-50 architecture was used to segment all nuclei from within the identified cancer region, (3) a total of 26,641 quantitative morphometric features pertaining to nuclear shape, size, and texture were extracted from within and outside tumor nuclei, (4) a random forest classifier was trained to distinguish between Stage II and Stage IV colon cancers using the 5 most discriminatory features selected by the Wilcoxon rank-sum test. Our trained classifier using these top 5 features yielded an AUC of 0.81 and 0.78, respectively, on the held-out cases in UHCMC and TCGA validation sets. For 197 TCGA-COAD cases, the Cox-proportional hazards model yielded a hazard ratio of 2.20 (95% CI: 1.24-3.88) with a concordance index of 0.71 using only top-five features for risk stratification of overall survival. The Kaplan-Meier estimate also showed statistically significant separation between the low-risk and high-risk patients with a log-rank p-value of 0.0097. Finally, unsupervised clustering of the top-five features revealed that Stage IV colon cancers with peritoneal spread were morphologically more similar to Stage II colon cancers with no long-term metastases than Stage IV colon cancers with hematogenous spread. This article is protected by copyright. All rights reserved.

Pathology and Prognosis of Colonic Adenocarcinomas With Intermediate Primary Tumor Stage Between pT2 and pT3

CONTEXT.—

Primary tumor stage (pT) is an important prognostic indicator in colonic adenocarcinomas; however, cases that have no muscle fibers beyond the advancing tumor edge but also show no extension beyond the apparent outer border of muscularis propria (termed pT2int), have not been previously studied.

OBJECTIVE.—

To address the clinicopathologic characteristics and prognosis of pT2int tumors.

DESIGN.—

We recharacterized 168 colon carcinomas and compared pT2int cases to bona fide pT2 and pT3 tumors.

RESULTS.—

In side-by-side analysis, 21 pT2int cases diverged from 29 pT2 tumors only in terms of larger size (P = .03), but they were less likely to show high-grade (P = .03), lymphovascular (P < .001), and extramural venous invasion (P = .04); discontinuous tumor deposits (P = .02); lymph node involvement (P = .001); and advanced stage (P = .001), compared with 118 pT3 tumors. Combining pT2int with pT2 cases (versus pT3) was a better independent predictor of negative lymph nodes in multivariate analysis (P = .04; odds ratio [OR], 3.96; CI, 1.09-14.42) and absent distant metastasis in univariate analysis (P = .04), compared with sorting pT2int with pT3 cases (versus pT2). Proportional hazards regression showed that pT2 and pT2int cases together were associated with better disease-free survival compared with pT3 tumors (P = .04; OR, 3.65; CI, 1.05-12.70). Kaplan-Meier analysis demonstrated that when pT2int were grouped with pT2 tumors, they were significantly less likely to show disease progression compared with pT3 (P = .002; log-rank test) and showed a trend toward better disease-specific survival (P = .06), during a mean patient follow-up of 44.9 months.

CONCLUSIONS.—

These data support the conclusion that pT2int carcinomas have clinicopathologic characteristics and are associated with patient outcomes more closely aligned with pT2 rather than pT3 tumors.

Atypical Cells in Peritoneal Cleft: A Pitfall in Diagnosis of Colorectal Adenocarcinoma

Atypical cells in peritoneal clefts are usually either reactive mesothelial cells or pT4 colonic adenocarcinoma in colon specimen removed for primary colon cancer. However, rarely if ever are these atypical cells metastasis from other primary visceral malignancy due to "sac-like" anatomic structure of this area. We present a case where these atypical cells were determined to be metastasis of gynecological origin by judicious use of immunohistochemical stains. A final diagnosis of serous tubal intraepithelial carcinoma of right fallopian tube was diagnosed only after total abdominal hysterectomy bilateral salpingo-oophorectomy. To our knowledge, this is the first report of a serous tubal intraepithelial carcinoma presenting as stage 4 colonic adenocarcinoma. The importance of this interesting case is 2-fold. It highlights the peritoneal cleft as an anatomic region not often recognized or discussed as well as tumor presentation in this region. In addition, this example stresses the need for additional mesothelial markers in addition to WT-1 workup of atypical mesothelial proliferation.

T3 versus T4a staging challenges in deeply invasive colonic adenocarcinomas and correlation with clinical outcomes

Despite the latest 8th edition American Joint Committee on Cancer Staging Manual guidelines, disagreement still exists among pathologists regarding staging deeply invasive colonic adenocarcinomas ≤1 mm to the serosal surface. In this retrospective study, 151 untreated colonic adenocarcinomas staged initially as either pT3 or pT4a and with available 5-year follow-up data were retrieved and re-categorized: Group 1 (38 cases): pT4a with tumor at the serosa; Group 2 (49 cases): tumor ≤1 mm from the serosa, with intervening reactive fibrosis (40/49) or inflammation (9/49); Group 3 (64 cases): pT3 tumor >1 mm from the serosa. Clinical outcomes were analyzed. Groups 1 and 2 tumors showed significantly lower 5-year recurrence-free survival and lower overall survival rates (log-rank p < 0.001 for both), when compared with Group 3 tumors. Even after adjusting for adjuvant therapy and nodal metastases, the proportional hazards ratios for the risk of death (p < 0.001) and risk of recurrence (p = 0.005) showed significantly higher risk in Groups 1 and 2 compared with Group 3. The synchronous nodal (p = 0.012) and metachronous distant metastases (p = 0.004) were also significantly more in Groups 1 and 2 versus Group 3. Colonic adenocarcinomas ≤1 mm from the serosal surface behaved more akin to "bona fide" pT4a tumors at the serosal surface in our study with regards to clinical outcomes. We recommend these tumors be staged as pT4a rather than pT3, as supported by outcome data in our study. We hope this will also ensure reproducibility and consistency in staging these tumors across institutions.

Current Trends of Artificial Intelligence for Colorectal Cancer Pathology Image Analysis: A Systematic Review

Colorectal cancer (CRC) is one of the most common cancers requiring early pathologic diagnosis using colonoscopy biopsy samples. Recently, artificial intelligence (AI) has made significant progress and shown promising results in the field of medicine despite several limitations. We performed a systematic review of AI use in CRC pathology image analysis to visualize the state-of-the-art. Studies published between January 2000 and January 2020 were searched in major online databases including MEDLINE (PubMed, Cochrane Library, and EMBASE). Query terms included "colorectal neoplasm," "histology," and "artificial intelligence." Of 9000 identified studies, only 30 studies consisting of 40 models were selected for review. The algorithm features of the models were gland segmentation (n = 25, 62%), tumor classification (n = 8, 20%), tumor microenvironment characterization (n = 4, 10%), and prognosis prediction (n = 3, 8%). Only 20 gland segmentation models met the criteria for quantitative analysis, and the model proposed by Ding et al. (2019) performed the best. Studies with other features were in the elementary stage, although most showed impressive results. Overall, the state-of-the-art is promising for CRC pathological analysis. However, datasets in most studies had relatively limited scale and quality for clinical application of this technique. Future studies with larger datasets and high-quality annotations are required for routine practice-level validation.