Ki-67 Proliferation Index Assessment in Gastroenteropancreatic Neuroendocrine Tumors by Digital Image Analysis With Stringent Case and Hotspot Level Concordance Requirements

Utility of Whole Slide Imaging for Intraoperative Consultation: Experience of a Large Academic Center

CONTEXT.—

In the United States, review of digital whole slide images (WSIs) using specific systems is approved for primary diagnosis but has not been implemented for intraoperative consultation.

OBJECTIVE.—

To evaluate the safety of review of WSIs and compare the efficiency of review of WSIs and glass slides (GSs) for intraoperative consultation.

DESIGN.—

Ninety-one cases previously submitted for frozen section evaluation were randomly selected from 8 different anatomic pathology subspecialties. GSs from these cases were scanned on a Leica Aperio AT2 scanner at ×20 magnification (0.25 μm/pixel). The slides were deidentified, and a short relevant clinical history was provided for each slide. Nine board-certified general pathologists who do not routinely establish primary diagnoses using WSIs reviewed the WSIs using Leica Aperio ImageScope viewing software. After a washout period of 2-3 weeks, the pathologists reviewed the corresponding GSs using a light microscope (Olympus BX43). The pathologists recorded the diagnosis and time to reach the diagnosis. Intraobserver concordance, time to diagnosis, and specificity and sensitivity compared to the original diagnosis were evaluated.

RESULTS.—

The rate of intraobserver concordance between GS results and WSI results was 93.7%. Mean time to diagnosis was 1.25 minutes for GSs and 1.76 minutes for WSIs (P < .001). Specificity was 91% for GSs and 90% for WSIs; sensitivity was 92% for GSs and 92% for WSIs.

CONCLUSIONS.—

Time to diagnosis was longer with WSIs than with GSs, and scanning GSs and uploading the data to whole slide imaging systems takes time. However, review of WSIs appears to be a safe alternative to review of GSs. Use of WSIs allows reporting from a remote site during a public health emergency such as the COVID-19 pandemic and facilitates subspecialty histopathology services.

Clinicopathologic characteristics of neuroendocrine tumors with assessment by digital image analysis for Ki-67 index with a focus on the gastroenteropancreatic tract: a single-center study

OBJECTIVES

Neuroendocrine tumors (NETs) are a heterogeneous group of tumors that arise at various sites throughout the body. The gastroenteropancreatic (GEP) tract is the most common site of NETs. We investigated the clinicopathologic features of patients with GEP-NETs and the utility of digital image analysis, which was compared to eyeball estimation, a conventional method used to determine the Ki-67 labeling index.

METHODS

The clinicopathologic data of GEP-NET patients at Gangnam Severance Hospital from January 2008 to October 2019 were retrospectively analyzed. Each case was reclassified according to the 2019 World Health Organization classification system, to which the classification of grade 3 was added. Comparisons between eyeball estimation and the digital image analysis method for Ki-67 index assessment were performed by calculating Cohen's kappa (k) coefficient.

RESULTS

In total, 345 patients with GEP-NETs were enrolled. The mean age was 49.3 (range 13-79) years, with more male (61.1%) than female patients. The primary tumor sites were the rectum (70.1%), pancreas (12.5%), stomach (6.7%), and duodenum (5.8%). Overall, 298 (86.4%), 35 (10.1%), 2 (0.6%), and 10 (2.9%) patients exhibited grade 1, 2, and 3 and neuroendocrine carcinoma, respectively. Statistical analysis revealed that age > 50 years, tumor size > 2 cm, and presence of lymphovascular invasion, nodal metastasis, and distant metastasis were significantly associated with short overall survival. Additionally, 283 patients underwent digital image analysis of the Ki-67 index, and substantial agreement was found between the two methods (κ value: 0.765).

CONCLUSIONS

Eyeball estimation revealed non-inferior results compared with digital image analysis. Further research is needed to evaluate the possibility of using digital image analysis as an alternative analysis method.

Ki-67 assessment of pancreatic neuroendocrine neoplasms: Systematic review and meta-analysis of manual vs. digital pathology scoring

Ki-67 assessment is a key step in the diagnosis of neuroendocrine neoplasms (NENs) from all anatomic locations. Several challenges exist related to quantifying the Ki-67 proliferation index due to lack of method standardization and inter-reader variability. The application of digital pathology coupled with machine learning has been shown to be highly accurate and reproducible for the evaluation of Ki-67 in NENs. We systematically reviewed all published studies on the subject of Ki-67 assessment in pancreatic NENs (PanNENs) employing digital image analysis (DIA). The most common advantages of DIA were improvement in the standardization and reliability of Ki-67 evaluation, as well as its speed and practicality, compared to the current gold standard approach of manual counts from captured images, which is cumbersome and time consuming. The main limitations were attributed to higher costs, lack of widespread availability (as of yet), operator qualification and training issues (if it is not done by pathologists), and most importantly, the drawback of image algorithms counting contaminating non-neoplastic cells and other signals like hemosiderin. However, solutions are rapidly developing for all of these challenging issues. A comparative meta-analysis for DIA versus manual counting shows very high concordance (global coefficient of concordance: 0.94, 95% CI: 0.83-0.98) between these two modalities. These findings support the widespread adoption of validated DIA methods for Ki-67 assessment in PanNENs, provided that measures are in place to ensure counting of only tumor cells either by software modifications or education of non-pathologist operators, as well as selection of standard regions of interest for analysis. NENs, being cellular and monotonous neoplasms, are naturally more amenable to Ki-67 assessment. However, lessons of this review may be applicable to other neoplasms where proliferation activity has become an integral part of theranostic evaluation including breast, brain, and hematolymphoid neoplasms.

Automated assessment of Ki-67 proliferation index in neuroendocrine tumors by deep learning

The Ki‐67 proliferation index (PI) is a prognostic factor in neuroendocrine tumors (NETs) and defines tumor grade. Analysis of Ki‐67 PI requires calculation of Ki‐67‐positive and Ki‐67‐negative tumor cells, which is highly subjective. To overcome this, we developed a deep learning‐based Ki‐67 PI algorithm (KAI) that objectively calculates Ki‐67 PI. Our study material consisted of NETs divided into training (n = 39), testing (n = 124), and validation (n = 60) series. All slides were digitized and processed in the Aiforia^® Create (Aiforia Technologies, Helsinki, Finland) platform. The ICC between the pathologists and the KAI was 0.89. In 46% of the tumors, the Ki‐67 PIs calculated by the pathologists and the KAI were the same. In 12% of the tumors, the Ki‐67 PI calculated by the KAI was 1% lower and in 42% of the tumors on average 3% higher. The DL‐based Ki‐67 PI algorithm yields results similar to human observers. While the algorithm cannot replace the pathologist, it can assist in the laborious Ki‐67 PI assessment of NETs. In the future, this approach could be useful in, for example, multi‐center clinical trials where objective estimation of Ki‐67 PI is crucial.