Artificial Intelligence-Assisted Score Analysis for Predicting the Expression of the Immunotherapy Biomarker PD-L1 in Lung Cancer

Artificial Intelligence and Lung Pathology

This manuscript provides a comprehensive overview of the application of artificial intelligence (AI) in lung pathology, particularly in the diagnosis of lung cancer. It discusses various AI models designed to support pathologists and clinicians. AI models supporting pathologists are to standardize diagnosis, score PD-L1 status, supporting tumor cellularity count, and indicating explainability for pathologic judgements. Several models predict outcomes beyond pathologic diagnosis and predict clinical outcomes like patients' survival and molecular alterations. The manuscript emphasizes the potential of AI to enhance accuracy and efficiency in pathology, while also addressing the challenges and future directions for integrating AI into clinical practice.

Comprehensive Potential of Artificial Intelligence for Predicting PD-L1 Expression and EGFR Mutations in Lung Cancer: A Systematic Review and Meta-Analysis

OBJECTIVE

To evaluate the methodological quality and the predictive performance of artificial intelligence (AI) for predicting programmed death ligand 1 (PD-L1) expression and epidermal growth factor receptors (EGFR) mutations in lung cancer (LC) based on systematic review and meta-analysis.

METHODS

AI studies based on PET/CT, CT, PET, and immunohistochemistry (IHC)-whole-slide image (WSI) were included to predict PD-L1 expression or EGFR mutations in LC. The modified Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool was used to evaluate the methodological quality. A comprehensive meta-analysis was conducted to analyze the overall area under the curve (AUC). The Cochrane diagnostic test and I2 statistics were used to assess the heterogeneity of the meta-analysis.

RESULTS

A total of 45 AI studies were included, of which 10 were used to predict PD-L1 expression and 35 were used to predict EGFR mutations. Based on the analysis using the QUADAS-2 tool, 37 studies achieved a high-quality score of 7. In the meta-analysis of PD-L1 expression levels, the overall AUCs for PET/CT, CT, and IHC-WSI were 0.80 (95% confidence interval [CI], 0.77-0.84), 0.74 (95% CI, 0.69-0.77), and 0.95 (95% CI, 0.93-0.97), respectively. For EGFR mutation status, the overall AUCs for PET/CT, CT, and PET were 0.85 (95% CI, 0.81-0.88), 0.83 (95% CI, 0.80-0.86), and 0.75 (95% CI, 0.71-0.79), respectively. The Cochrane Diagnostic Test revealed an I2 value exceeding 50%, indicating substantial heterogeneity in the PD-L1 and EGFR meta-analyses. When AI was combined with clinicopathological features, the enhancement in predicting PD-L1 expression was not substantial, whereas the prediction of EGFR mutations showed improvement compared to the CT and PET models, albeit not significantly so compared to the PET/CT models.

CONCLUSIONS

The overall performance of AI in predicting PD-L1 expression and EGFR mutations in LC has promising clinical implications.

Imaging Analytics using Artificial Intelligence in Oncology: A Comprehensive Review

The present era has seen a surge in artificial intelligence-related research in oncology, mainly using deep learning, because of powerful computer hardware, improved algorithms and the availability of large amounts of data from open-source domains and the use of transfer learning. Here we discuss the multifaceted role of deep learning in cancer care, ranging from risk stratification, the screening and diagnosis of cancer, to the prediction of genomic mutations, treatment response and survival outcome prediction, through the use of convolutional neural networks. Another role of artificial intelligence is in the generation of automated radiology reports, which is a boon in high-volume centres to minimise report turnaround time. Although a validated and deployable deep-learning model for clinical use is still in its infancy, there is ongoing research to overcome the barriers for its universal implementation and we also delve into this aspect. We also briefly describe the role of radiomics in oncoimaging. Artificial intelligence can provide answers pertaining to cancer management at baseline imaging, saving cost and time. Imaging biobanks, which are repositories of anonymised images, are also briefly described. We also discuss the commercialisation and ethical issues pertaining to artificial intelligence. The latest generation generalist artificial intelligence model is also briefly described at the end of the article. We believe this article will not only enrich knowledge, but also promote research acumen in the minds of readers to take oncoimaging to another level using artificial intelligence and also work towards clinical translation of such research.

Revolutionizing Pathology with Artificial Intelligence: Innovations in Immunohistochemistry

Artificial intelligence (AI) is a reality of our times, and it has been successfully implemented in all fields, including medicine. As a relatively new domain, all efforts are directed towards creating algorithms applicable in most medical specialties. Pathology, as one of the most important areas of interest for precision medicine, has received significant attention in the development and implementation of AI algorithms. This focus is especially important for achieving accurate diagnoses. Moreover, immunohistochemistry (IHC) serves as a complementary diagnostic tool in pathology. It can be further augmented through the application of deep learning (DL) and machine learning (ML) algorithms for assessing and analyzing immunohistochemical markers. Such advancements can aid in delineating targeted therapeutic approaches and prognostic stratification. This article explores the applications and integration of various AI software programs and platforms used in immunohistochemical analysis. It concludes by highlighting the application of these technologies to pathologies such as breast, prostate, lung, melanocytic proliferations, and hematologic conditions. Additionally, it underscores the necessity for further innovative diagnostic algorithms to assist physicians in the diagnostic process.

A Pipeline for Evaluation of Machine Learning/Artificial Intelligence Models to Quantify Programmed Death Ligand 1 Immunohistochemistry

Immunohistochemistry (IHC) is used to guide treatment decisions in multiple cancer types. For treatment with checkpoint inhibitors, programmed death ligand 1 (PD-L1) IHC is used as a companion diagnostic. However, the scoring of PD-L1 is complicated by its expression in cancer and immune cells. Separation of cancer and noncancer regions is needed to calculate tumor proportion scores (TPS) of PD-L1, which is based on the percentage of PD-L1-positive cancer cells. Evaluation of PD-L1 expression requires highly experienced pathologists and is often challenging and time-consuming. Here, we used a multi-institutional cohort of 77 lung cancer cases stained centrally with the PD-L1 22C3 clone. We developed a 4-step pipeline for measuring TPS that includes the coregistration of hematoxylin and eosin, PD-L1, and negative control (NC) digital slides for exclusion of necrosis, segmentation of cancer regions, and quantification of PD-L1+ cells. As cancer segmentation is a challenging step for TPS generation, we trained DeepLab V3 in the Visiopharm software package to outline cancer regions in PD-L1 and NC images and evaluated the model performance by mean intersection over union (mIoU) against manual outlines. Only 14 cases were required to accomplish a mIoU of 0.82 for cancer segmentation in hematoxylin-stained NC cases. For PD-L1-stained slides, a model trained on PD-L1 tiles augmented by registered NC tiles achieved a mIoU of 0.79. In segmented cancer regions from whole slide images, the digital TPS achieved an accuracy of 75% against the manual TPS scores from the pathology report. Major reasons for algorithmic inaccuracies include the inclusion of immune cells in cancer outlines and poor nuclear segmentation of cancer cells. Our transparent and stepwise approach and performance metrics can be applied to any IHC assay to provide pathologists with important insights on when to apply and how to evaluate commercial automated IHC scoring systems.

A Deep Learning-Based Assay for Programmed Death Ligand 1 Immunohistochemistry Scoring in Non-Small Cell Lung Carcinoma: Does it Help Pathologists Score?

Several studies have developed various artificial intelligence (AI) models for immunohistochemical analysis of programmed death ligand 1 (PD-L1) in patients with non-small cell lung carcinoma; however, none have focused on specific ways by which AI-assisted systems could help pathologists determine the tumor proportion score (TPS). In this study, we developed an AI model to calculate the TPS of the PD-L1 22C3 assay and evaluated whether and how this AI-assisted system could help pathologists determine the TPS and analyze how AI-assisted systems could affect pathologists' assessment accuracy. We assessed the 4 methods of the AI-assisted system: (1 and 2) pathologists first assessed and then referred to automated AI scoring results (1, positive tumor cell percentage; 2, positive tumor cell percentage and visualized overlay image) for final confirmation, and (3 and 4) pathologists referred to the automated AI scoring results (3, positive tumor cell percentage; 4, positive tumor cell percentage and visualized overlay image) while determining TPS. Mixed-model analysis was used to calculate the odds ratios (ORs) with 95% CI for AI-assisted TPS methods 1 to 4 compared with pathologists' scoring. For all 584 samples of the tissue microarray, the OR for AI-assisted TPS methods 1 to 4 was 0.94 to 1.07 and not statistically significant. Of them, we found 332 discordant cases, on which the pathologists' judgments were inconsistent; the ORs for AI-assisted TPS methods 1, 2, 3, and 4 were 1.28 (1.06-1.54; P = .012), 1.29 (1.06-1.55; P = .010), 1.28 (1.06-1.54; P = .012), and 1.29 (1.06-1.55; P = .010), respectively, which were statistically significant. For discordant cases, the OR for each AI-assisted TPS method compared with the others was 0.99 to 1.01 and not statistically significant. This study emphasized the usefulness of the AI-assisted system for cases in which pathologists had difficulty determining the PD-L1 TPS.

Clinical Validation of Artificial Intelligence-Powered PD-L1 Tumor Proportion Score Interpretation for Immune Checkpoint Inhibitor Response Prediction in Non-Small Cell Lung Cancer

PURPOSE

Evaluation of PD-L1 tumor proportion score (TPS) by pathologists has been very impactful but is limited by factors such as intraobserver/interobserver bias and intratumor heterogeneity. We developed an artificial intelligence (AI)-powered analyzer to assess TPS for the prediction of immune checkpoint inhibitor (ICI) response in advanced non-small cell lung cancer (NSCLC).

MATERIALS AND METHODS

The AI analyzer was trained with 393,565 tumor cells annotated by board-certified pathologists for PD-L1 expression in 802 whole-slide images (WSIs) stained by 22C3 pharmDx immunohistochemistry. The clinical performance of the analyzer was validated in an external cohort of 430 WSIs from patients with NSCLC. Three pathologists performed annotations of this external cohort, and their consensus TPS was compared with AI-based TPS.

RESULTS

In comparing PD-L1 TPS assessed by AI analyzer and by pathologists, a significant positive correlation was observed (Spearman coefficient = 0.925; P < .001). The concordance of TPS between AI analyzer and pathologists according to TPS ≥50%, 1%-49%, and <1% was 85.7%, 89.3%, and 52.4%, respectively. In median progression-free survival (PFS), AI-based TPS predicted prognosis in the TPS 1%-49% or TPS <1% group better than the pathologist's reading, with the TPS ≥50% group as a reference (hazard ratio [HR], 1.49 [95% CI, 1.19 to 1.86] v HR, 1.36 [95% CI, 1.08 to 1.71] for TPS 1%-49% group, and HR, 2.38 [95% CI, 1.69 to 3.35] v HR, 1.62 [95% CI, 1.23 to 2.13] for TPS <1% group).

CONCLUSION

PD-L1 TPS assessed by AI analyzer correlates with that of pathologists, with clinical performance also being comparable when referenced to PFS. The AI model can accurately predict tumor response and PFS of ICI in advanced NSCLC via assessment of PD-L1 TPS.

Complete digital pathology transition: A large multi-center experience

INTRODUCTION

Transitioning from glass slide pathology to digital pathology for primary diagnostics requires an appropriate laboratory information system, an image management system, and slide scanners; it also reinforces the need for sophisticated pathology informatics including synoptic reporting. Previous reports have discussed the transition itself and relevant considerations for it, but not the selection criteria and considerations for the infrastructure.

OBJECTIVE

To describe the process used to evaluate slide scanners, image management systems, and synoptic reporting systems for a large multisite institution.

METHODS

Six network hospitals evaluated six slide scanners, three image management systems, and three synoptic reporting systems. Scanners were evaluated based on the quality of image, speed, ease of operation, and special capabilities (including z-stacking, fluorescence and others). Image management and synoptic reporting systems were evaluated for their ease of use and capacity.

RESULTS

Among the scanners evaluated, the Leica GT450 produced the highest quality images, while the 3DHistech Pannoramic provided fluorescence and superior z-stacking. The newest generation of scanners, released relatively recently, performed better than slightly older scanners from major manufacturers Although the Olympus VS200 was not fully vetted due to not meeting all inclusion criteria, it is discussed herein due to its exceptional versatility. For Image Management Software, the authors believe that Sectra is, at the time of writing the best developed option, but this could change in the very near future as other systems improve their capabilities. All synoptic reporting systems performed impressively.

CONCLUSIONS

Specifics regarding quality and abilities of different components will change rapidly with time, but large pathology practices considering such a transition should be aware of the issues discussed and evaluate the most current generation to arrive at appropriate conclusions.

ChatGPT as an aid for pathological diagnosis of cancer

Diagnostic workup of cancer patients is highly reliant on the science of pathology using cytopathology, histopathology, and other ancillary techniques like immunohistochemistry and molecular cytogenetics. Data processing and learning by means of artificial intelligence (AI) has become a spearhead for the advancement of medicine, with pathology and laboratory medicine being no exceptions. ChatGPT, an artificial intelligence (AI)-based chatbot, that was recently launched by OpenAI, is currently a talk of the town, and its role in cancer diagnosis is also being explored meticulously. Pathology workflow by integration of digital slides, implementation of advanced algorithms, and computer-aided diagnostic techniques extend the frontiers of the pathologist's view beyond a microscopic slide and enables effective integration, assimilation, and utilization of knowledge that is beyond human limits and boundaries. Despite of it's numerous advantages in the pathological diagnosis of cancer, it comes with several challenges like integration of digital slides with input language parameters, problems of bias, and legal issues which have to be addressed and worked up soon so that we as a pathologists diagnosing malignancies are on the same band wagon and don't miss the train.

Identifying Serum Small Extracellular Vesicle MicroRNA as a Noninvasive Diagnostic and Prognostic Biomarker for Ovarian Cancer

There remains a lack of effective and noninvasive methods for the diagnosis and prognosis prediction of epithelial ovarian carcinoma (EOC). Here, we investigated the possibility of serum-derived small extracellular vesicle (sEV) microRNAs (miRNAs) as potential biomarkers for distinguishing between benign and malignant adnexal masses and predicting the prognosis of EOC patients. A serum sEV miRNA model for identifying the EOC (sEVmiR-EOC) was successfully established in the training cohort. Furthermore, the sEVmiR-EOC model was confirmed in the testing cohort and validation cohort, demonstrating robust diagnostic accuracy. The sEVmiR-EOC model showed better performance than carbohydrate antigen 125 (CA125) in discriminating patients with stage I EOC from benign patients. Using EOC samples and follow-up data, we identified miR-141-3p and miR-200c-3p as potential prognostic predictors. Finally, we confirmed the change of the sEVmiR-EOC RiskScore between the preoperative and postoperative samples and found that the sEVmiR-EOC model could predict the prognosis of EOC patients.

Deep Learning for Lung Cancer Diagnosis, Prognosis and Prediction Using Histological and Cytological Images: A Systematic Review

Lung cancer is one of the deadliest cancers worldwide, with a high incidence rate, especially in tobacco smokers. Lung cancer accurate diagnosis is based on distinct histological patterns combined with molecular data for personalized treatment. Precise lung cancer classification from a single H&E slide can be challenging for a pathologist, requiring most of the time additional histochemical and special immunohistochemical stains for the final pathology report. According to WHO, small biopsy and cytology specimens are the available materials for about 70% of lung cancer patients with advanced-stage unresectable disease. Thus, the limited available diagnostic material necessitates its optimal management and processing for the completion of diagnosis and predictive testing according to the published guidelines. During the new era of Digital Pathology, Deep Learning offers the potential for lung cancer interpretation to assist pathologists' routine practice. Herein, we systematically review the current Artificial Intelligence-based approaches using histological and cytological images of lung cancer. Most of the published literature centered on the distinction between lung adenocarcinoma, lung squamous cell carcinoma, and small cell lung carcinoma, reflecting the realistic pathologist's routine. Furthermore, several studies developed algorithms for lung adenocarcinoma predominant architectural pattern determination, prognosis prediction, mutational status characterization, and PD-L1 expression status estimation.

Correlation between immunotherapy biomarker PD-L1 expression and genetic alteration in patients with non-small cell lung cancer

Programmed death-ligand 1 (PD-L1) has been widely used in immunotherapy evaluation of patients with non-small cell lung cancer (NSCLC). However, the effect is not particularly ideal, and the association between PD-L1 and genetic alterations requires more exploration. Here, we performed targeted next-generation sequencing and PD-L1 immunohistochemistry (IHC) testing for PD-L1 expression on both tumor cells (TCs) and tumor-infiltrating immune cells (ICs) in 1549 patients. Our studies showed that surgical method of resection was positively correlated with IC+, and a low tumor mutation burden (TMB) was negatively correlated with TC+. Furthermore, we found that EGFR was mutually exclusive with both ALK and STK11. In addition, the features between PD-L1 expression status and genomic alterations were characterized. These results suggest that clinical characteristics and molecular phenotypes are associated with PD-L1 expression signatures, which may provide novel insights for improving the efficiency of immune checkpoint inhibitors (ICIs) in immunotherapy.

Anesthesiologists' Perspective on the Use of Artificial Intelligence in Ultrasound-Guided Regional Anaesthesia in Terms of Medical Ethics and Medical Education: A Survey Study

OBJECTIVE

Controversy exists around the world as experts disagree on what artificial intelligence will imply for humanity in the future. Medical experts are starting to share perspectives on artificial intelligence with ethical and legal concerns appearing to prevail. The purpose of this study was to determine how anesthesiology and reanimation specialists in Turkey perceive the use of artificial intelligence in ultrasound-guided regional anesthetic applications in terms of medical ethics and education, as well as their perspectives on potential ethical issues.

MATERIALS AND METHODS

This descriptive and cross-sectional survey was conducted across Turkey between July 1 and August 31. Data were collected through an online questionnaire distributed by national associations and social media platforms. The questionnaire included questions about the descriptive features of the participants and the possible ethical problems that may be encountered in the use of artificial intelligence in regional anesthesia and 20 statements that were requested to be evaluated.

RESULTS

The average age of the 285 anesthesiologists who took part in the study was 42.00 ± 7.51, 144 of them were male, the average years spent in the field was 10.95 ± 7.15 years, 59.3% were involved in resident training, and 74.7% habitually used ultrasound guidance regional anesthetic applications. Of the participants, 80% thought artificial intelligence would benefit patients, 86.7% thought it would benefit resident training, 81.4% thought it would benefit post-graduate medical education, and 80.7% thought it would decrease complications in practice. There will be no ethical issues if sonographic data are captured anonymously, according to 78.25%, while 67% are concerned about who will be held accountable for inaccuracies.

CONCLUSION

The majority of anesthetists believe that using artificial intelligence in regional anesthetic applications will decrease complications. Although ethical concerns about privacy and data governance are low, participants do have ethical worries about "accountability for errors."

State of the Art of Pathologic and Molecular Testing

Advances in the treatment of non-small cell lung carcinoma have resulted in improved histologic classification and the implementation of molecular testing for predictive biomarkers into the routine diagnostic workflow. Over the past decade, molecular testing has evolved from single-gene assays to high-thoroughput comprehensive next-generation sequencing. Economic barriers, suboptimal turnaround time to obtain the results, and limited tissue available for molecular assays resulted in adoption of liquid biopsies (ctDNA) into clinical practice. Multiplex immunohistochemical/immunofluorescence assays evaluating tumor microenvironment together with the AI approaches are anticipated to translate from research into clinical care.

The Role of Genomics and Proteomics in Lung Cancer Early Detection and Treatment

Lung cancer is the leading cause of cancer-related death worldwide, with non-small-cell lung cancer (NSCLC) being the primary type. Unfortunately, it is often diagnosed at advanced stages, when therapy leaves patients with a dismal prognosis. Despite the advances in genomics and proteomics in the past decade, leading to progress in developing tools for early diagnosis, targeted therapies have shown promising results; however, the 5-year survival of NSCLC patients is only about 15%. Low-dose computed tomography or chest X-ray are the main types of screening tools. Lung cancer patients without specific, actionable mutations are currently treated with conventional therapies, such as platinum-based chemotherapy; however, resistances and relapses often occur in these patients. More noninvasive, inexpensive, and safer diagnostic methods based on novel biomarkers for NSCLC are of paramount importance. In the current review, we summarize genomic and proteomic biomarkers utilized for the early detection and treatment of NSCLC. We further discuss future opportunities to improve biomarkers for early detection and the effective treatment of NSCLC.

Tumor microenvironment features decipher the outperformance of neoadjuvant immunochemotherapy over chemotherapy in resectable non-small cell lung cancer

This study evaluated the efficacy of neoadjuvant immunochemotherapy (Io+Chemo) versus chemotherapy alone (Chemo) in resectable non–small cell lung cancer (NSCLC) in a real-world setting. The association of tumor immune microenvironment (TIME) with pathologic response to different neoadjuvant therapies was also explored.Stage I−III NSCLC patients who received Io+Chemo or Chemo alone followed by surgery were included in the study. Tumor tissues collected during surgery were subjected to TIME evaluation using multiplex immunohistochemistry to measure immune cell subsets, including T cells, B cells, NK cells, and macrophages. Fifty-five patients were included, including 24 treated with neoadjuvant Io+Chemo and 31 with Chemo alone. Io+Chemo induced significantly higher major pathologic response (MPR) (75.0% vs. 38.7%, P = 0.0133) and numerically better pathologic complete response (pCR) (33.3% vs. 12.9%, P = 0.1013) than Chemo. Compared with tumors with Chemo, tumors with Io+Chemo demonstrated a significantly higher ratio of M1 macrophage density in the tumor to that in the stroma (P = 0.0446), more abundant CD8⁺ cells in the stroma (P = 0.0335), and fewer PD-L1⁺CD68⁺ cells in both tumor and stroma. pCR/MPR patients displayed significantly higher density of CD3⁺, CD3⁺CD4⁺, CD20⁺, CD56 bright cell subsets and more tertiary lymphoid structures and significantly lower density of PD-L1⁺CD68⁺ and CD3⁺CD4⁺Foxp3⁺cells in the tumor or stroma. This study favored neoadjuvant Io+Chemo over Chemo and revealed the TIME features underlying the outperformance of Io+Chemo over Chemo.