Blood and lymphatic vessel invasion in pT1 colorectal cancer: an international concordance study

Risk assessment in pT1 colorectal cancer

Colorectal cancer (CRC) is a common malignancy worldwide and tumour stage is closely related to clinical outcome. A small but significant proportion of submucosal-invasive (ie, pT1) CRC are associated with regional lymph node metastases (LNM) and a worse prognosis. The likelihood of LNM in pT1 CRC needs to be balanced against the operative risk and costs of surgical resection when determining the best patient management. A wide range of histopathological and clinical factors may affect LNM risk in this setting. This script provides a comprehensive overview of the tumour and patient-associated features that have been linked to LNM risk in pT1 CRC. Some of the features are well established within the literature and are included in published guidelines, while others are novel and emerging in nature. Odds ratios for LNM that are associated with key predictive features are provided where appropriate, and published models developed as an aid to the calculation of LNM risk are discussed.

Use of artificial intelligence in the management of T1 colorectal cancer: a new tool in the arsenal or is deep learning out of its depth?

The field of artificial intelligence is rapidly evolving, and there has been an interest in its use to predict the risk of lymph node metastasis in T1 colorectal cancer. Accurately predicting lymph node invasion may result in fewer patients undergoing unnecessary surgeries; conversely, inadequate assessments will result in suboptimal oncological outcomes. This narrative review aims to summarize the current literature on deep learning for predicting the probability of lymph node metastasis in T1 colorectal cancer, highlighting areas of potential application and barriers that may limit its generalizability and clinical utility.

Whole slide image-based prediction of lymph node metastasis in T1 colorectal cancer using unsupervised artificial intelligence

OBJECTIVES

Lymph node metastasis (LNM) prediction for T1 colorectal cancer (CRC) is critical for determining the need for surgery after endoscopic resection because LNM occurs in 10%. We aimed to develop a novel artificial intelligence (AI) system using whole slide images (WSIs) to predict LNM.

METHODS

We conducted a retrospective single center study. To train and test the AI model, we included LNM status-confirmed T1 and T2 CRC between April 2001 and October 2021. These lesions were divided into two cohorts: training (T1 and T2) and testing (T1). WSIs were cropped into small patches and clustered by unsupervised K-means. The percentage of patches belonging to each cluster was calculated from each WSI. Each cluster's percentage, sex, and tumor location were extracted and learned using the random forest algorithm. We calculated the areas under the receiver operating characteristic curves (AUCs) to identify the LNM and the rate of over-surgery of the AI model and the guidelines.

RESULTS

The training cohort contained 217 T1 and 268 T2 CRCs, while 100 T1 cases (LNM-positivity 15%) were the test cohort. The AUC of the AI system for the test cohort was 0.74 (95% confidence interval [CI] 0.58-0.86), and 0.52 (95% CI 0.50-0.55) using the guidelines criteria (P = 0.0028). This AI model could reduce the 21% of over-surgery compared to the guidelines.

CONCLUSION

We developed a pathologist-independent predictive model for LNM in T1 CRC using WSI for determination of the need for surgery after endoscopic resection.

TRIAL REGISTRATION

UMIN Clinical Trials Registry (UMIN000046992, https://center6.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000053590).

Novel "resect and analysis" approach for T2 colorectal cancer with use of artificial intelligence

BACKGROUND AND AIMS

Because of a lack of reliable preoperative prediction of lymph node involvement in early-stage T2 colorectal cancer (CRC), surgical resection is the current standard treatment. This leads to overtreatment because only 25% of T2 CRC patients turn out to have lymph node metastasis (LNM). We assessed a novel artificial intelligence (AI) system to predict LNM in T2 CRC to ascertain patients who can be safely treated with less-invasive endoscopic resection such as endoscopic full-thickness resection and do not need surgery.

METHODS

We included 511 consecutive patients who had surgical resection with T2 CRC from 2001 to 2016; 411 patients (2001-2014) were used as a training set for the random forest-based AI prediction tool, and 100 patients (2014-2016) were used to validate the AI tool performance. The AI algorithm included 8 clinicopathologic variables (patient age and sex, tumor size and location, lymphatic invasion, vascular invasion, histologic differentiation, and serum carcinoembryonic antigen level) and predicted the likelihood of LNM by receiver-operating characteristics using area under the curve (AUC) estimates.

RESULTS

Rates of LNM in the training and validation datasets were 26% (106/411) and 28% (28/100), respectively. The AUC of the AI algorithm for the validation cohort was .93. With 96% sensitivity (95% confidence interval, 90%-99%), specificity was 88% (95% confidence interval, 80%-94%). In this case, 64% of patients could avoid surgery, whereas 1.6% of patients with LNM would lose a chance to receive surgery.

CONCLUSIONS

Our proposed AI prediction model has a potential to reduce unnecessary surgery for patients with T2 CRC with very little risk. (Clinical trial registration number: UMIN 000038257.).

Utility of artificial intelligence with deep learning of hematoxylin and eosin-stained whole slide images to predict lymph node metastasis in T1 colorectal cancer using endoscopically resected specimens; prediction of lymph node metastasis in T1 colorectal cancer

BACKGROUND

When endoscopically resected specimens of early colorectal cancer (CRC) show high-risk features, surgery should be performed based on current guidelines because of the high-risk of lymph node metastasis (LNM). The aim of this study was to determine the utility of an artificial intelligence (AI) with deep learning (DL) of hematoxylin and eosin (H&E)-stained endoscopic resection specimens without manual-pixel-level annotation for predicting LNM in T1 CRC. In addition, we assessed AI performance for patients with only submucosal (SM) invasion depth of 1000 to 2000 μm known to be difficult to predict LNM in clinical practice.

METHODS

H&E-stained whole slide images (WSIs) were scanned for endoscopic resection specimens of 400 patients who underwent endoscopic treatment for newly diagnosed T1 CRC with additional surgery. The area under the curve (AUC) of the receiver operating characteristic curve was used to determine the accuracy of AI for predicting LNM with a fivefold cross-validation in the training set and in a held-out test set.

RESULTS

We developed an AI model using a two-step attention-based DL approach without clinical features (AUC, 0.764). Incorporating clinical features into the model did not improve its prediction accuracy for LNM. Our model reduced unnecessary additional surgery by 15.1% more than using the current guidelines (67.4% vs. 82.5%). In patients with SM invasion depth of 1000 to 2000 μm, the AI avoided 16.1% of unnecessary additional surgery than using the JSCCR guidelines.

CONCLUSIONS

Our study is the first to show that AI trained with DL of H&E-stained WSIs has the potential to predict LNM in T1 CRC using only endoscopically resected specimens with conventional histologic risk factors.

Sampling and Reporting of Inflammatory Bowel Disease

Pathologists have an important and expanding role in the diagnosis and management of inflammatory bowel disease. This role includes the initial diagnosis of the disease, assessment of the response to treatment and the identification of short-term complications such as cytomegalovirus infection and long-term complications such as dysplasia. Furthermore, the assessment of resection specimens for complication of disease is important to determining the risk of subsequent disease or inflammation within an ileal pouch. Adequate sampling of the disease at endoscopy and from the surgical resection specimen is vital to determining the ultimate information that can be provided by the pathologist. This sampling is determined by the clinical scenario. Similarly, a standardized approach to reporting and synthesizing the histologic findings will improve patient management. This is best exemplified by the increasing interest in histologic activity indices, such as the Nancy index in ulcerative colitis, and in the standardized reporting for inflammatory bowel disease dysplasia recommended by the SCENIC international consensus.

Artificial Intelligence System to Determine Risk of T1 Colorectal Cancer Metastasis to Lymph Node

BACKGROUND & AIMS

In accordance with guidelines, most patients with T1 colorectal cancers (CRC) undergo surgical resection with lymph node dissection, despite the low incidence (∼10%) of metastasis to lymph nodes. To reduce unnecessary surgical resections, we used artificial intelligence to build a model to identify T1 colorectal tumors at risk for metastasis to lymph node and validated the model in a separate set of patients.

METHODS

We collected data from 3134 patients with T1 CRC treated at 6 hospitals in Japan from April 1997 through September 2017 (training cohort). We developed a machine-learning artificial neural network (ANN) using data on patients' age and sex, as well as tumor size, location, morphology, lymphatic and vascular invasion, and histologic grade. We then conducted the external validation on the ANN model using independent 939 patients at another hospital during the same period (validation cohort). We calculated areas under the receiver operator characteristics curves (AUCs) for the ability of the model and US guidelines to identify patients with lymph node metastases.

RESULTS

Lymph node metastases were found in 319 (10.2%) of 3134 patients in the training cohort and 79 (8.4%) of /939 patients in the validation cohort. In the validation cohort, the ANN model identified patients with lymph node metastases with an AUC of 0.83, whereas the guidelines identified patients with lymph node metastases with an AUC of 0.73 (P < .001). When the analysis was limited to patients with initial endoscopic resection (n = 517), the ANN model identified patients with lymph node metastases with an AUC of 0.84 and the guidelines identified these patients with an AUC of 0.77 (P = .005).

CONCLUSIONS

The ANN model outperformed guidelines in identifying patients with T1 CRCs who had lymph node metastases. This model might be used to determine which patients require additional surgery after endoscopic resection of T1 CRCs. UMIN Clinical Trials Registry no: UMIN000038609.

Lymphovascular invasion in early gastric cancer: impact of ancillary D2-40 and elastin staining on interobserver agreement

AIMS

Lymphatic invasion (LI) and venous invasion (VI) are the strongest risk factors for lymph node metastasis in patients with early gastric cancer, and may predict their prognosis. We aimed to investigate interobserver agreement among pathologists before and after adding ancillary staining for diagnosing LI and VI in this setting.

METHODS AND RESULTS

This retrospective study included 100 specimens of submucosal invasive gastric cancer from 100 patients treated using endoscopic resection. Three pathologists (expert, intermediate and trainee experience levels) independently evaluated individual LI and VI status using haematoxylin and eosin (H&E)-stained slides, and re-evaluated their decisions by reviewing corresponding D2-40-stained and elastin-stained slides. Interobserver agreement was assessed using κ statistics. With the ancillary D2-40 staining, there was an improved agreement for LI diagnoses between the expert and intermediate pathologist (H&E κ = 0.78, D2-40 κ = 0.85) and between the expert and trainee pathologist (H&E κ = 0.37, D2-40 κ = 0.56). With the ancillary elastin staining, there was an improved agreement for VI diagnoses between the expert and intermediate pathologists (H&E κ = 0.25, elastin κ = 0.63) and between the expert and trainee pathologists (H&E κ = 0.29, elastin κ = 0.45).

CONCLUSIONS

With both the ancillary D2-40 and elastin staining there was an improved interobserver agreement for LI and VI diagnoses, regardless of the pathologist's experience. This ancillary staining may be useful in improving the accuracy of LI and VI diagnoses, helping to improve the risk stratification of early gastric cancer patients.

Improvement of Care in Patients With Colorectal Cancer: Influence of the Introduction of Standardized Structured Reporting for Pathology

PURPOSE

The use of standardized structured reporting (SSR) can improve communication between cancer specialists, which might improve clinical care; however, there are no reliable data on whether the introduction of SSR is associated with improvements in clinical outcome.

PATIENTS AND METHODS

We performed a retrospective cohort study in the Netherlands, including all patients with colorectal cancer (CRC) from 2009 to 2014. As a reference, cohorts of 2007 and 2008 were included. Data from the Netherlands Cancer Registry were used and combined with data from the Dutch Pathology Registry (PALGA) and the Dutch ColoRectal Audit. We tested the preformulated hypothesis that use of SSR improves the care of patients with CRC by improving the completeness of the pathology reports, the quality of the pathology evaluation, and patient outcomes with respect to treatment and survival.

RESULTS

We included 72,859 patients with CRC (23.8% reference, 32.9% SSR, and 43.3% narrative reports). Use of SSR increased over time, which resulted in more complete pathology reports (95.8% v 89.8%; P < .001). Risk assessment in stage II colon cancer was more adequate and resulted in an increased delivery of adjuvant therapy in patients with SSR (19.6% v 15.1%; P = .001). Risk of death for patients in the SSR group was significantly lowered (corrected hazard ratio, 0.94; 95% CI 0.90 to 0.97).

CONCLUSION

We demonstrate that use of SSR improved patient care in those with CRC by providing more complete reports of higher quality, which had significant effects on the delivery of adjuvant therapy and patient outcomes.

Recommendation of long-term and systemic management according to the risk factors in rectal NETs patients

Rectal neuroendocrine tumors (NETs) are often found as small lesions, which can be treated by endoscopic resection. However, high risk cases with lymph node (LN) metastasis are indication of radical surgery. Furthermore, rectal NETs are often associated with late recurrences and/or multiple cancer development. Therefore, proper surgical indication and patients' management are required. We investigated the clinicopathological features of 79 rectal NET cases in order to elucidate risk factors for synchronous LN metastasis, recurrence, and multiple cancers. Recently, we reported that in pancreatic NET patients, a loss of heterozygosity (LOH) in PHLDA3 was associated with poorer prognosis, and that LOH of both PHLDA3 and MEN1 was frequently observed. Therefore, PHLDA3 and MEN1 LOH were also assessed in rectal NET patients for their association with clinicopathological features. Of the 79 patients, LN metastases were found in 12.7%, recurrences in 3.8%, and multiple cancers in 30.4% of the subjects. PHLDA3 and MEN1 LOH were found in 60.0% and 66.7% of the subjects, respectively. Lymphatic invasion and WHO classification 2010 were found to be independent risks for LN metastasis. There were three cases of recurrence, all of which occurred more than 3 years after resection and two of which exhibited LN metastasis. Older age and LOH in PHLDA3 were associated with the presence of multiple cancers. Long-term and systemic management of patients with rectal NETs is therefore recommended in accordance with these risk factors.

Pure Well-Differentiated Adenocarcinoma Is a Safe Factor for Lymph Node Metastasis in T1 and T2 Colorectal Cancer: A Pilot Study

Background and Aims

Various risk factors for lymph node metastasis (LNM) have been reported in colorectal T1 cancers. However, the factors available are insufficient for predicting LNM. We therefore investigated the utility of the new histological factor "pure well-differentiated adenocarcinoma" (PWDA) as a safe factor for predicting LNM in T1 and T2 cancers.

Materials and Methods

We reviewed 115 T2 cancers and 202 T1 cancers in patients who underwent surgical resection in our center. We investigated the rates of LNM among various clinicopathological factors, including PWDA. PWDA was defined as a lesion comprising only well-differentiated adenocarcinoma. The consistency of the diagnosis of PWDA was evaluated among two pathologists. In addition, 72 T1 cancers with LNM from 8 related hospitals over 10 years (2008-2017) were also analyzed.

Results

The rates of LNM and PWDA were 23.5% and 20.0%, respectively, in T2 cancers. Significant differences were noted between patients with and without LNM regarding lymphatic invasion (81.5% vs. 36.4%, p < 0.001), poor histology (51.9% vs. 19.3%, p = 0.008), and PWDA (3.7% vs. 25.0%, p = 0.015). The rates of LNM and PWDA were 8.4% and 36.1%, respectively, in T1 cancers. Regarding the 73 PWDA cases and 129 non-PWDA cases, the rates of LNM were 0.0% and 13.2%, respectively (p < 0.001). Among the 97 cases with lymphatic or venous invasion, the rates of LNM in 29 PWDA cases and 68 non-PWDA were 0% and 14.7%, respectively (p = 0.029). The agreement of the two pathologists for the diagnosis of PWDA was acceptable (kappa value > 0.5). A multicenter review showed no cases of PWDA among 72 T1 cancers with LNM.

Conclusions

PWDA is considered to be a safe factor for LNM in T1 cancer.

Neuroendocrine Tumors of the Large Intestine: Clinicopathological Features and Predictive Factors of Lymph Node Metastasis

A new histological classification of neuroendocrine tumors (NETs) was established in WHO 2010. ENET and NCCN proposed treatment algorithms for colorectal NET. Retrospective study of NET of the large intestine (colorectal and appendiceal NET) was performed among institutions allied with the Japanese Society for Cancer of the Colon and Rectum, and 760 neuroendocrine tumors from 2001 to 2011 were re-assessed using WHO 2010 criteria to elucidate the clinicopathological features of NET in the large intestine. Next, the clinicopathological relationship with lymph node metastasis was analyzed to predict lymph node metastasis in locally resected rectal NET. The primary site was rectum in 718/760 cases (94.5%), colon in 30/760 cases (3.9%), and appendix in 12/760 cases (1.6%). Patients were predominantly men (61.6%) with a mean age of 58.7 years. Tumor size was <10 mm in 65.4% of cases. Proportions of NET G1, G2, G3, and mixed adeno-neuroendocrine carcinoma (MANEC) were 88.4, 6.3, 3.9, and 1.3%, respectively. Of the 760 tumors, 468 were locally resected, and 292 were surgically resected with lymph node dissection. Rectal NET showed a higher proportion of NET G1, and colonic and appendiceal NET was more commonly G3 and MANEC. Of the 292 surgically resected cases, 233 NET G1 and G2 located in the rectum were used for the prediction of lymph node metastasis. Lymphatic and blood vessel invasion were independent predictive factors of lymph node metastasis. NET G2 cases showed more frequent lymph node metastasis than that seen in NET G1 cases, but this was not an independent predictor of lymph node metastasis. Of the 98 surgically resected cases <10 mm in size, we found 9 cases with lymph node metastasis (9.2%). All cases were NET G1, and eight of the nine cases were positive either for lymphatic invasion or blood vessel invasion. Using the WHO classification, we found NET in the large intestine showed a tumor-site-dependent variety of histological and clinicopathological features. Risk of lymph node metastasis in rectal NET was confirmed even in lesions smaller than 10 mm. Concordant assessment of vascular invasion will be required to estimate lymph node metastasis in small lesions.

12-Gene Recurrence Score Assay Stratifies the Recurrence Risk in Stage II/III Colon Cancer With Surgery Alone: The SUNRISE Study

PURPOSE

The 12-gene Recurrence Score assay has been validated in resected stage II colon cancer treated with or without chemotherapy and resected stage III disease treated with chemotherapy. This study evaluated the 12-gene Recurrence Score assay for stage II and III colon cancer without chemotherapy to reveal the natural course of recurrence risk in stage III disease.

METHODS

A cohort-sampling design was used. From 1,487 consecutive patients with stage II to III disease who had surgery alone, 630 patients were sampled for inclusion with a 1:2 ratio of recurrence and nonrecurrence. Sampling was stratified by stage (II v III). The assay was performed on formalin-fixed, paraffin-embedded primary cancer tissue. Association of the Recurrence Score result with recurrence-free interval (RFI) was assessed by using weighted Cox proportional hazards regression.

RESULTS

Overall, 597 of 630 patients were analyzable-247 patients had stage II, and 350 had stage III colon cancer. The continuous Recurrence Score was significantly associated with RFI after adjustment for disease stage (hazard ratio for a 25-unit increase in Recurrence Score, 2.05; 95% CI, 1.47 to 2.86; P < .001). With respect to prespecified subgroups, as defined by low (< 30), intermediate (30 to 40), and high (≥ 41) Recurrence Score risk groups, patients with stage II disease in the high-risk group had a 5-year risk of recurrence similar to patients with stage IIIA to IIIB disease in the low-risk group (19% v 20%), whereas patients with stage IIIA to IIIB disease in the high-risk group had a recurrence risk similar to that of patients with stage IIIC disease in the low-risk group (approximately 38%).

CONCLUSION

To our knowledge, this study provides the first validation of the 12-gene Recurrence Score assay in stage III colon cancer without chemotherapy and showed the heterogeneity of recurrence risks in stage III as well as in stage II colon cancer.

Optimizing the detection of venous invasion in colorectal cancer: the ontario, Canada, experience and beyond

Venous invasion (VI) is a well-established independent prognostic indicator in colorectal cancer (CRC). Its accurate detection is particularly important in stage II CRC as it may influence the decision to administer adjuvant therapy. The Royal College of Pathologists (RCPath) of the United Kingdom state that VI should be detected in at least 30% of CRC resection specimens. However, our experience in Ontario, Canada suggests that this (conservative) benchmark is rarely met. This article highlights the “Ontario experience” with respect to VI reporting and the key role that careful morphologic assessment, elastin staining and knowledge transfer has played in improving VI detection provincially and beyond.