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.

A Phase II Open-Label Randomized Clinical Trial of Preoperative Durvalumab or Durvalumab plus Tremelimumab in Resectable Head and Neck Squamous Cell Carcinoma.

PURPOSE: Clinical implications of neoadjuvant immunotherapy in patients with locally advanced but resectable head and neck squamous cell carcinoma (HNSCC) remain largely unexplored. PATIENTS AND METHODS: Patients with resectable HNSCC were randomized to receive a single dose of preoperative durvalumab (D) with or without tremelimumab (T) before resection, followed by postoperative (chemo)radiotherapy based on multidisciplinary discretion and 1-year D treatment. Artificial intelligence (AI)-powered spatial distribution analysis of tumor-infiltrating lymphocytes and high-dimensional profiling of circulating immune cells tracked dynamic intratumoral and systemic immune responses. RESULTS: Of the 48 patients enrolled (D, 24 patients; D+T, 24 patients), 45 underwent surgical resection per protocol (D, 21 patients; D+T, 24 patients). D±T had a favorable safety profile and did not delay surgery. Distant recurrence-free survival (DRFS) was significantly better in patients treated with D+T than in those treated with D monotherapy. AI-powered whole-slide image analysis demonstrated that D+T significantly reshaped the tumor microenvironment toward immune-inflamed phenotypes, in contrast with the D monotherapy or cytotoxic chemotherapy. High-dimensional profiling of circulating immune cells revealed a significant expansion of T-cell subsets characterized by proliferation and activation in response to D+T therapy, which was rare following D monotherapy. Importantly, expansion of specific clusters in CD8+ T cells and non-regulatory CD4+ T cells with activation and exhaustion programs was associated with prolonged DRFS in patients treated with D+T. CONCLUSIONS: Preoperative D±T is feasible and may benefit patients with resectable HNSCC. Distinct changes in the tumor microenvironment and circulating immune cells were induced by each treatment regimen, warranting further investigation.

An artificial intelligence-powered PD-L1 combined positive score (CPS) analyser in urothelial carcinoma alleviating interobserver and intersite variability.

AIMS: Immune checkpoint inhibitors targeting programmed death-ligand 1 (PD-L1) have shown promising clinical outcomes in urothelial carcinoma (UC). The combined positive score (CPS) quantifies PD-L1 22C3 expression in UC, but it can vary between pathologists due to the consideration of both immune and tumour cell positivity. METHODS AND RESULTS: An artificial intelligence (AI)-powered PD-L1 CPS analyser was developed using 1,275,907 cells and 6175.42 mm2 of tissue annotated by pathologists, extracted from 400 PD-L1 22C3-stained whole slide images of UC. We validated the AI model on 543 UC PD-L1 22C3 cases collected from three institutions. There were 446 cases (82.1%) where the CPS results (CPS ≥10 or <10) were in complete agreement between three pathologists, and 486 cases (89.5%) where the AI-powered CPS results matched the consensus of two or more pathologists. In the pathologist's assessment of the CPS, statistically significant differences were noted depending on the source hospital (P = 0.003). Three pathologists reevaluated discrepancy cases with AI-powered CPS results. After using the AI as a guide and revising, the complete agreement increased to 93.9%. The AI model contributed to improving the concordance between pathologists across various factors including hospital, specimen type, pathologic T stage, histologic subtypes, and dominant PD-L1-positive cell type. In the revised results, the evaluation discordance among slides from different hospitals was mitigated. CONCLUSION: This study suggests that AI models can help pathologists to reduce discrepancies between pathologists in quantifying immunohistochemistry including PD-L1 22C3 CPS, especially when evaluating data from different institutions, such as in a telepathology setting.

Inflamed immune phenotype predicts favorable clinical outcomes of immune checkpoint inhibitor therapy across multiple cancer types.

BACKGROUND: The inflamed immune phenotype (IIP), defined by enrichment of tumor-infiltrating lymphocytes (TILs) within intratumoral areas, is a promising tumor-agnostic biomarker of response to immune checkpoint inhibitor (ICI) therapy. However, it is challenging to define the IIP in an objective and reproducible manner during manual histopathologic examination. Here, we investigate artificial intelligence (AI)-based immune phenotypes capable of predicting ICI clinical outcomes in multiple solid tumor types. METHODS: Lunit SCOPE IO is a deep learning model which determines the immune phenotype of the tumor microenvironment based on TIL analysis. We evaluated the correlation between the IIP and ICI treatment outcomes in terms of objective response rates (ORR), progression-free survival (PFS), and overall survival (OS) in a cohort of 1,806 ICI-treated patients representing over 27 solid tumor types retrospectively collected from multiple institutions. RESULTS: We observed an overall IIP prevalence of 35.2% and significantly more favorable ORRs (26.3% vs 15.8%), PFS (median 5.3 vs 3.1 months, HR 0.68, 95% CI 0.61 to 0.76), and OS (median 25.3 vs 13.6 months, HR 0.66, 95% CI 0.57 to 0.75) after ICI therapy in IIP compared with non-IIP patients, respectively (p<0.001 for all comparisons). On subgroup analysis, the IIP was generally prognostic of favorable PFS across major patient subgroups, with the exception of the microsatellite unstable/mismatch repair deficient subgroup. CONCLUSION: The AI-based IIP may represent a practical, affordable, clinically actionable, and tumor-agnostic biomarker prognostic of ICI therapy response across diverse tumor types.

Discontinuation risk from adverse events: immunotherapy alone vs. combined with chemotherapy: a systematic review and network meta-analysis.

BACKGROUND: While immunotherapy combined with chemotherapy (Chemo-IO) is generally recognized for providing superior outcomes compared to monotherapy (mono-IO), it is associated with a higher incidence of treatment-related adverse events (TRAEs), which may lead to treatment discontinuation. In this study, we compared the rates of treatment discontinuation between mono-IO and Chemo-IO as first-line treatments for various solid tumors. METHODS: We systematically reviewed clinical trials from databases (PubMed, Embase, Cochrane Library, and an additional source) published from January 1, 2018, to July 10, 2023. We included phase III randomized controlled trials (RCTs) that utilized immunotherapy agents in at least one arm as first-line treatments for a variety of solid tumors. Data extraction followed the Preferred Reporting Items for Systematic Reviews (PRISMA) extension statement for network meta-analysis. A random effects model was used for the network meta-analysis, with the risk of bias assessed using the Cochrane risk-of-bias tool II. The primary outcomes encompassed treatment discontinuation rates due to TRAEs among patients who underwent immunotherapy, either alone or combined with chemotherapy, for various solid tumors. Pooled relative risks (RRs) with 95% confidence intervals (CIs) were calculated to compare between treatment groups. RESULTS: From 29 RCTs, a total of 21,677 patients and 5 types of treatment were analyzed. Compared to mono-IO, Chemo-IO showed a significantly higher rate of discontinuation due to TRAEs (RR 2.68, 95% CI 1.98-3.63). Subgroup analysis for non-small cell lung cancer (NSCLC) patients also exhibited a greater risk of discontinuation due to TRAEs with Chemo-IO compared to mono-IO (RR 2.93, 95% CI 1.67-5.14). Additional analyses evaluating discontinuation rates due to either treatment emergent adverse events (TEAEs) or AEs regardless of causality (any AEs) consistently revealed an elevated risk associated with Chemo-IO. CONCLUSIONS: Chemo-IO was associated with an elevated risk of treatment discontinuation not only due to TRAEs but also any AEs or TEAEs. Given that the treatment duration can impact clinical outcomes, a subset of patients might benefit more from mono-IO than combination therapy. Further research is imperative to identify and characterize this subset.

Phase III, Randomized Study of Atezolizumab Plus Bevacizumab and Chemotherapy in Patients With EGFR- or ALK-Mutated Non-Small-Cell Lung Cancer (ATTLAS, KCSG-LU19-04).

PURPOSE: In the treatment of non-small-cell lung cancer (NSCLC) with a driver mutation, the role of anti-PD-(L)1 antibody after tyrosine kinase inhibitor (TKI) remains unclear. This randomized, open-label, multicenter, phase III study evaluates the efficacy of atezolizumab plus bevacizumab, paclitaxel, and carboplatin (ABCP ) in EGFR- or ALK-mutated NSCLC that progressed before TKI therapy. MATERIALS AND METHODS: We compared the clinical efficacy of ABCP followed by maintenance therapy with atezolizumab plus bevacizumab with pemetrexed plus carboplatin or cisplatin (PC) followed by pemetrexed maintenance. The primary end point was progression-free survival (PFS). RESULTS: A total of 228 patients with activating EGFR mutation (n = 215) or ALK translocation (n = 13) were enrolled from 16 sites in the Republic of Korea and randomly assigned at 2:1 ratio to either ABCP (n = 154) or PC arm (n = 74). The median follow-up duration was 26.1 months (95% CI, 24.7 to 28.2). Objective response rates (69.5% v 41.9%, P < .001) and median PFS (8.48 v 5.62 months, hazard ratio [HR], 0.62 [95% CI, 0.45 to 0.86]; P = .004) were significantly better in the ABCP than PC arm. PFS benefit increased as PD-L1 expression increased, with an HR of 0.47, 0.41, and 0.24 for PD-L1 ≥1%, ≥10%, and ≥50%, respectively. Overall survival was similar between ABCP and PC arm (20.63 v 20.27 months, HR, 1.01 [95% CI, 0.69 to 1.46]; P = .975). The safety profile of the ABCP arm was comparable with that previously reported, with no additional safety signals, but higher rates of treatment-related adverse events were observed compared with the PC arm. CONCLUSION: To our knowledge, this study is the first randomized phase III study to demonstrate the clinical benefit of anti-PD-L1 antibody in combination with bevacizumab and chemotherapy in patients with EGFR- or ALK-mutated NSCLC who have progressed on relevant targeted therapy.

Artificial Intelligence in Breast Cancer Diagnosis and Personalized Medicine.

Breast cancer is a significant cause of cancer-related mortality in women worldwide. Early and precise diagnosis is crucial, and clinical outcomes can be markedly enhanced. The rise of artificial intelligence (AI) has ushered in a new era, notably in image analysis, paving the way for major advancements in breast cancer diagnosis and individualized treatment regimens. In the diagnostic workflow for patients with breast cancer, the role of AI encompasses screening, diagnosis, staging, biomarker evaluation, prognostication, and therapeutic response prediction. Although its potential is immense, its complete integration into clinical practice is challenging. Particularly, these challenges include the imperatives for extensive clinical validation, model generalizability, navigating the "black-box" conundrum, and pragmatic considerations of embedding AI into everyday clinical environments. In this review, we comprehensively explored the diverse applications of AI in breast cancer care, underlining its transformative promise and existing impediments. In radiology, we specifically address AI in mammography, tomosynthesis, risk prediction models, and supplementary imaging methods, including magnetic resonance imaging and ultrasound. In pathology, our focus is on AI applications for pathologic diagnosis, evaluation of biomarkers, and predictions related to genetic alterations, treatment response, and prognosis in the context of breast cancer diagnosis and treatment. Our discussion underscores the transformative potential of AI in breast cancer management and emphasizes the importance of focused research to realize the full spectrum of benefits of AI in patient care.

Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes as a predictive biomarker for axitinib in adenoid cystic carcinoma.

BACKGROUND: This study analyzed the predictive value of artificial intelligence (AI)-powered tumor-infiltrating lymphocyte (TIL) analysis in recurrent or metastatic (R/M) adenoid cystic carcinoma (ACC) treated with axitinib. METHODS: Patients from a multicenter, prospective phase II trial evaluating axitinib efficacy in R/M ACC were included in this study. H&E whole-side images of archival tumor tissues were analyzed by Lunit SCOPE IO, an AI-powered spatial TIL analyzer. RESULTS: Twenty-seven patients were included in the analysis. The best response was stable disease, and the median progression-free survival (PFS) was 11.1 months (95% CI, 9.2-13.7 months). Median TIL densities in the cancer and surrounding stroma were 25.8/mm2 (IQR, 8.3-73.0) and 180.4/mm2 (IQR, 69.6-342.8), respectively. Patients with stromal TIL density >342.5/mm2 exhibited longer PFS (p = 0.012). CONCLUSIONS: Cancer and stromal area TIL infiltration were generally low in R/M ACC. Higher stromal TIL infiltration was associated with a longer PFS with axitinib treatment.

Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes as a biomarker in locally advanced unresectable thymic epithelial neoplasm: A single-center, retrospective, longitudinal cohort study.

BACKGROUND: Thymic epithelial tumors (TET) are rare malignancies and lack well-defined biomarkers for neoadjuvant therapy. This study aimed to evaluate the clinical utility of artificial intelligence (AI)-powered tumor-infiltrating lymphocyte (TIL) analysis in TET. METHODS: Patients initially diagnosed with unresectable thymoma or thymic carcinoma who underwent neoadjuvant therapy between January 2004 and December 2021 formed our study population. Hematoxylin and eosin-stained sections from the initial biopsy and surgery were analyzed using an AI-powered spatial TIL analyzer. Intratumoral TIL (iTIL) and stromal TIL (sTIL) were quantified and their immune phenotype (IP) was identified. RESULTS: Thirty-five patients were included in this study. The proportion of patients with partial response to neoadjuvant therapy was higher in the group with nondesert IP in preneoadjuvant biopsy (63.6% vs. 17.6%, p = 0.038). A significant increase in both iTIL (median 22.18/mm2 vs. 340.69/mm2 , p < 0.001) and sTIL (median 175.19/mm2 vs. 531.02/mm2 , p = 0.004) was observed after neoadjuvant therapy. Patients with higher iTIL (>147/mm2 ) exhibited longer disease-free survival (median, 29 months vs. 12 months, p = 0.009) and overall survival (OS) (median, 62 months vs. 45 months, p = 0.002). Patients with higher sTIL (>232.1/mm2 ) exhibited longer OS (median 62 months vs. 30 months, p = 0.021). CONCLUSIONS: Nondesert IP in initial biopsy was associated with a better response to neoadjuvant therapy. Increased infiltration of both iTIL and sTIL in surgical specimens were associated with longer OS in patients with TET who underwent resection followed by neoadjuvant therapy.

Deep learning model improves tumor-infiltrating lymphocyte evaluation and therapeutic response prediction in breast cancer.

Tumor-infiltrating lymphocytes (TILs) have been recognized as key players in the tumor microenvironment of breast cancer, but substantial interobserver variability among pathologists has impeded its utility as a biomarker. We developed a deep learning (DL)-based TIL analyzer to evaluate stromal TILs (sTILs) in breast cancer. Three pathologists evaluated 402 whole slide images of breast cancer and interpreted the sTIL scores. A standalone performance of the DL model was evaluated in the 210 cases (52.2%) exhibiting sTIL score differences of less than 10 percentage points, yielding a concordance correlation coefficient of 0.755 (95% confidence interval [CI], 0.693-0.805) in comparison to the pathologists' scores. For the 226 slides (56.2%) showing a 10 percentage points or greater variance between pathologists and the DL model, revisions were made. The number of discordant cases was reduced to 116 (28.9%) with the DL assistance (p < 0.001). The DL assistance also increased the concordance correlation coefficient of the sTIL score among every two pathologists. In triple-negative and human epidermal growth factor receptor 2 (HER2)-positive breast cancer patients who underwent the neoadjuvant chemotherapy, the DL-assisted revision notably accentuated higher sTIL scores in responders (26.8 ± 19.6 vs. 19.0 ± 16.4, p = 0.003). Furthermore, the DL-assistant revision disclosed the correlation of sTIL-high tumors (sTIL ≥ 50) with the chemotherapeutic response (odd ratio 1.28 [95% confidence interval, 1.01-1.63], p = 0.039). Through enhancing inter-pathologist concordance in sTIL interpretation and predicting neoadjuvant chemotherapy response, here we report the utility of the DL-based tool as a reference for sTIL scoring in breast cancer assessment.

Papillary thyroid cancer immune phenotypes via tumor-infiltrating lymphocyte spatial analysis.

Standard-of-care treatment options provide an excellent prognosis for papillary thyroid cancers (PTCs); however, approximately 10% of cases are advanced PTCs, resulting in less than 50% 5-year survival rates. Understanding the tumor microenvironment is essential for understanding cancer progression and investigating potential biomarkers for treatment, such as immunotherapy. Our study focused on tumor-infiltrating lymphocytes (TILs), which are the main effectors of antitumor immunity and related to the mechanism of immunotherapy. Using an artificial intelligence model, we analyzed the density of intratumoral and peritumoral TILs in the pathologic slides of The Cancer Genome Atlas PTC cohort. Tumors were classified into three immune phenotypes (IPs) based on the spatial distribution of TILs: immune-desert (48%), immune-excluded (34%), and inflamed (18%). Immune-desert IP was mostly characterized by RAS mutations, high thyroid differentiation score, and low antitumor immune response. Immune-excluded IP predominantly consisted of BRAF V600E-mutated tumors and had a higher rate of lymph node metastasis. Inflamed IP was characterized by a high antitumor immune response, as demonstrated by a high cytolytic score, immune-related cell infiltrations, expression of immunomodulatory molecules (including immunotherapy target molecules), and enrichment of immune-related pathways. This study is the first to investigate IP classification using TILs in PTC through a tissue-based approach. Each IP had unique immune and genomic profiles. Further studies are warranted to assess the predictive value of IP classification in advanced PTC patients treated with immunotherapy.

Efficacy improvement in searching MEDLINE database using a novel PubMed visual analytic system: EEEvis.

PubMed is the most extensively used database and search engine in the biomedical and healthcare fields. However, users could experience several difficulties in acquiring their target papers facing massive numbers of search results, especially in their unfamiliar fields. Therefore, we developed a novel user interface for PubMed and conducted three steps of study: step A, a preliminary user survey with 76 medical experts regarding the current usability for the biomedical literature search task at PubMed; step B is implementing EEEvis, a novel interactive visual analytic system for the search task; step C, a randomized user study comparing PubMed and EEEvis. First, we conducted a Google survey of 76 medical experts regarding the unmet needs of PubMed and the user requirements for a novel search interface. According to the data of preliminary Google survey, we implemented a novel interactive visual analytic system for biomedical literature search. This EEEvis provides enhanced literature data analysis functions including (1) an overview of the bibliographic features including publication date, citation count, and impact factors, (2) an overview of the co-authorship network, and (3) interactive sorting, filtering, and highlighting. In the randomized user study of 24 medical experts, the search speed of EEEvis was not inferior to PubMed in the time to reach the first article (median difference 3 sec, 95% CI -2.1 to 8.5, P = 0.535) nor in the search completion time (median difference 8 sec, 95% CI -4.7 to 19.1, P = 0.771). However, 22 participants (91.7%) responded that they are willing to use EEEvis as their first choice for a biomedical literature search task, and 21 participants (87.5%) answered the bibliographic sorting and filtering functionalities of EEEvis as a major advantage. EEEvis could be a supplementary interface for PubMed that can enhance the user experience in the search for biomedical literature.

A Phase II Trial of S-1 and Oxaliplatin in Patients with Metastatic Breast Cancer Previously Treated with Anthracycline and Taxane (KCSG-BR07-03).

PURPOSE: This single-arm phase II trial investigate the efficacy and safety of S-1 plus oxaliplatin (SOX) in patients with metastatic breast cancer. Materials and Methods: Patients with metastatic breast cancer previously treated with anthracyclines and taxanes were enrolled. Patients received S-1 (40-60 mg depending on patient's body surface area, twice a day, day 1-14) and oxaliplatin (130 mg/m2, day 1) in 3 weeks cycle until disease progression or unacceptable toxicity. The primary endpoint was objective response rate (ORR) according to Response Evaluation Criteria in Solid Tumor 1.1. Secondary endpoints included time-to-progression (TTP), duration-of-response (DoR), overall survival (OS), and adverse events. RESULTS: A total of 87 patients were enrolled from 11 institutions in Korea. Hormone receptor was positive in 54 (62.1%) patients and six (6.9%) had human epidermal growth factor receptor 2-positive disease. Forty-eight patients (85.1%) had visceral metastasis and 74 (55.2%) had more than three sites of metastases. The ORR of SOX regimen was 38.5% (95% confidence interval [CI], 26.9 to 50.0) with a median TTP of 6.0 months (95% CI, 5.1 to 6.9). Median DoR and OS were 10.3 months (95% CI, 5.5 to 15.1) and 19.4 (95% CI, not estimated) months, respectively. Grade 3 or 4 neutropenia was reported in 28 patients (32.1%) and thrombocytopenia was observed in 23 patients (26.6%). CONCLUSION: This phase II study showed that SOX regimen is a reasonable option in metastatic breast cancer previously treated with anthracyclines and taxanes.

Therapeutic Implications of TGF-ß Pathway in Desmoid Tumor Based on Comprehensive Molecular Profiling and Clinicopathological Properties.

(1) Background: Desmoid tumors have a relatively high local failure rate after primary treatment using surgery and/or radiotherapy. Moreover, desmoid tumors recur at the primary site for many patients. An effective therapeutic strategy for the desmoid tumor is needed to maintain quality of life and prolong survival. (2) Method: First of all, we collected desmoid tumor tissues and investigated the status of protein expression for beta-catenin and alpha-SMA through immunohistochemistry. Then, we performed targeted sequencing and whole RNA sequencing. To compare the data with other cancer types, we used NGS data from sarcoma patients at Yonsei Cancer Center (YCC-sarcoma cohort, n = 48) and The Cancer Genome Atlas (TCGA, n = 9235). Secondly, we established the novel patient-derived preclinical models (n = 2) for the validation of treatment strategy. The same gene alteration of primary tissue was demonstrated. (3) Results: We discovered specific gene sets related to the TGF-ß signaling pathway. Moreover, we selected the combination treatment comprising TGF-ß inhibitor, vactosertib, and imatinib. In screening for the anti-proliferation effect, the combination treatment of TGF-ß inhibitor was more effective for tumor suppression than monotherapy. (4) Conclusion: We found preclinical indications that TGF-ß inhibitors could prove useful as a potential treatment for patients with desmoid tumors. Moreover, we could find some examples in clinical trials.

Artificial Intelligence-Powered Whole-Slide Image Analyzer Reveals a Distinctive Distribution of Tumor-Infiltrating Lymphocytes in Neuroendocrine Neoplasms.

Despite the importance of tumor-infiltrating lymphocytes (TIL) and PD-L1 expression to the immune checkpoint inhibitor (ICI) response, a comprehensive assessment of these biomarkers has not yet been conducted in neuroendocrine neoplasm (NEN). We collected 218 NENs from multiple organs, including 190 low/intermediate-grade NENs and 28 high-grade NENs. TIL distribution was derived from Lunit SCOPE IO, an artificial intelligence (AI)-powered hematoxylin and eosin (H&E) analyzer, as developed from 17,849 whole slide images. The proportion of intra-tumoral TIL-high cases was significantly higher in high-grade NEN (75.0% vs. 46.3%, p = 0.008). The proportion of PD-L1 combined positive score (CPS) ≥ 1 case was higher in high-grade NEN (85.7% vs. 33.2%, p < 0.001). The PD-L1 CPS ≥ 1 group showed higher intra-tumoral, stromal, and combined TIL densities, compared to the CPS < 1 group (7.13 vs. 2.95, p < 0.001; 200.9 vs. 120.5, p < 0.001; 86.7 vs. 56.1, p = 0.004). A significant correlation was observed between TIL density and PD-L1 CPS (r = 0.37, p < 0.001 for intra-tumoral TIL; r = 0.24, p = 0.002 for stromal TIL and combined TIL). AI-powered TIL analysis reveals that intra-tumoral TIL density is significantly higher in high-grade NEN, and PD-L1 CPS has a positive correlation with TIL densities, thus showing its value as predictive biomarkers for ICI response in NEN.

High-Resolution Profiling of Lung Adenocarcinoma Identifies Expression Subtypes with Specific Biomarkers and Clinically Relevant Vulnerabilities.

Lung adenocarcinoma (LUAD) is one of the most common cancer types and has various treatment options. Better biomarkers to predict therapeutic response are needed to guide choice of treatment modality and to improve precision medicine. Here, we used a consensus hierarchical clustering approach on 509 LUAD cases from The Cancer Genome Atlas to identify five robust LUAD expression subtypes. Genomic and proteomic data from patient samples and cell lines was then integrated to help define biomarkers of response to targeted therapies and immunotherapies. This approach defined subtypes with unique proteogenomic and dependency profiles. Subtype 4 (S4)-associated cell lines exhibited specific vulnerability to loss of CDK6 and CDK6-cyclin D3 complex gene (CCND3). Subtype 3 (S3) was characterized by dependency on CDK4, immune-related expression patterns, and altered MET signaling. Experimental validation showed that S3-associated cell lines responded to MET inhibitors, leading to increased expression of programmed death-ligand 1 (PD-L1). In an independent real-world patient dataset, patients with S3 tumors were enriched with responders to immune checkpoint blockade. Genomic features in S3 and S4 were further identified as biomarkers for enabling clinical diagnosis of these subtypes. Overall, our consensus hierarchical clustering approach identified robust tumor expression subtypes, and our subsequent integrative analysis of genomics, proteomics, and CRISPR screening data revealed subtype-specific biology and vulnerabilities. These LUAD expression subtypes and their biomarkers could help identify patients likely to respond to CDK4/6, MET, or PD-L1 inhibitors, potentially improving patient outcome. SIGNIFICANCE: Integrative analysis of multiomic and drug dependency data uncovers robust lung adenocarcinoma expression subtypes with unique therapeutic vulnerabilities and subtype-specific biomarkers of response.

A Phase II Study of Nivolumab plus Gemcitabine in Patients with Recurrent or Metastatic Nasopharyngeal Carcinoma (KCSG HN17-11).

PURPOSE: Although programmed death 1/programmed death ligand 1 (PD-1/PD-L1) inhibitors are promising agents for recurrent or metastatic nasopharyngeal carcinoma (NPC), PD-1/PD-L1 inhibitor monotherapy has shown modest efficacy. This study evaluated the efficacy and safety of nivolumab plus gemcitabine in patients with NPC who failed prior platinum-based chemotherapy. PATIENTS AND METHODS: This is a phase II, multicenter, open-label, single-arm study. Patients with recurrent or metastatic NPC received nivolumab 3 mg/kg and gemcitabine 1,250 mg/m2 every 2 weeks until disease progression or intolerable toxicity. The primary endpoint was progression-free survival (PFS). The secondary endpoints included objective response rate (ORR), overall survival (OS), and safety. To identify potential biomarkers, whole-exome sequencing, whole-transcriptome sequencing, and immune phenotype analysis based on Lunit SCOPE IO, an artificial intelligence-powered spatial tumor-infiltrating lymphocyte analyzer, were performed. RESULTS: Thirty-six patients were enrolled between June 2018 and June 2019. The ORR was 36.1% and disease control rate was 97.2%. With median follow-up of 22.0 months, median PFS was 13.8 months [95% confidence interval (CI), 8.6-16.8 months]. Median OS was not reached, and OS rate at 6 months was 97.0% (95% CI, 80.4%-99.6%). The grade ≥3 treatment-related adverse events were hypertension (2.8%) and anemia (2.8%). In multivariate analysis of mutation of chromatin modifier gene, tumor mutational burden (≥ 2.1 mut/Mb), and somatic copy-number alteration (SCNA) level, the group with high SCNA (> 3 points; HR, 7.0; 95% CI, 1.3-37.9; P = 0.02) had independently associated with poor PFS. Immune phenotype analysis showed that tumors with high proportion of immune-excluded immune phenotype was significantly correlated with poor PFS (HR, 4.4; 95% CI, 1.2-16.2; P = 0.018). CONCLUSIONS: Nivolumab plus gemcitabine showed promising efficacy with favorable toxicity profiles in patients with advanced NPC in whom platinum-based combination chemotherapy failed.

Artificial intelligence-powered programmed death ligand 1 analyser reduces interobserver variation in tumour proportion score for non-small cell lung cancer with better prediction of immunotherapy response.

BACKGROUND: Manual evaluation of programmed death ligand 1 (PD-L1) tumour proportion score (TPS) by pathologists is associated with interobserver bias. OBJECTIVE: This study explored the role of artificial intelligence (AI)-powered TPS analyser in minimisation of interobserver variation and enhancement of therapeutic response prediction. METHODS: A prototype model of an AI-powered TPS analyser was developed with a total of 802 non-small cell lung cancer (NSCLC) whole-slide images. Three independent board-certified pathologists labelled PD-L1 TPS in an external cohort of 479 NSCLC slides. For cases of disagreement between each pathologist and the AI model, the pathologists were asked to revise the TPS grade (<1%, 1%-49% and ≥50%) with AI assistance. The concordance rates among the pathologists with or without AI assistance and the effect of the AI-assisted revision on clinical outcome upon immune checkpoint inhibitor (ICI) treatment were evaluated. RESULTS: Without AI assistance, pathologists concordantly classified TPS in 81.4% of the cases. They revised their initial interpretation by using the AI model for the disagreement cases between the pathologist and the AI model (N = 91, 93 and 107 for each pathologist). The overall concordance rate among the pathologists was increased to 90.2% after the AI assistance (P < 0.001). A reduction in hazard ratio for overall survival and progression-free survival upon ICI treatment was identified in the TPS subgroups after the AI-assisted TPS revision. CONCLUSION: The AI-powered TPS analyser assistance improves the pathologists' consensus of reading and prediction of the therapeutic response, raising a possibility of standardised approach for the accurate interpretation.

Artificial Intelligence-Powered Hematoxylin and Eosin Analyzer Reveals Distinct Immunologic and Mutational Profiles among Immune Phenotypes in Non-Small-Cell Lung Cancer.

The tumor microenvironment can be classified into three immune phenotypes: inflamed, immune excluded, and immune-desert. Immunotherapy efficacy has been shown to vary by phenotype; yet, the mechanisms are poorly understood and demand further investigation. This study unveils the mechanisms using an artificial intelligence-powered software called Lunit SCOPE. Artificial intelligence was used to classify 965 samples of non-small-cell lung carcinoma from The Cancer Genome Atlas into the three immune phenotypes. The immune and mutational profiles that shape each phenotype using xCell, gene set enrichment analysis with RNA-sequencing data, and cBioportal were described. In the inflamed subtype, which showed higher cytolytic score, the enriched pathways were generally associated with immune response and immune-related cell types were highly expressed. In the immune excluded subtype, enriched glycolysis, fatty acid, and cholesterol metabolism pathways were observed. The KRAS mutation, BRAF mutation, and MET splicing variant were mostly observed in the inflamed subtype. The two prominent mutations found in the immune excluded subtype were EGFR and PIK3CA mutations. This study is the first to report the distinct immunologic and mutational landscapes of immune phenotypes, and demonstrates the biological relevance of the classification. In light of these findings, the study offers insights into potential treatment options tailored to each immune phenotype.

Artificial Intelligence-Powered Spatial Analysis of Tumor-Infiltrating Lymphocytes as Complementary Biomarker for Immune Checkpoint Inhibition in Non-Small-Cell Lung Cancer.

PURPOSE: Biomarkers on the basis of tumor-infiltrating lymphocytes (TIL) are potentially valuable in predicting the effectiveness of immune checkpoint inhibitors (ICI). However, clinical application remains challenging because of methodologic limitations and laborious process involved in spatial analysis of TIL distribution in whole-slide images (WSI). METHODS: We have developed an artificial intelligence (AI)-powered WSI analyzer of TIL in the tumor microenvironment that can define three immune phenotypes (IPs): inflamed, immune-excluded, and immune-desert. These IPs were correlated with tumor response to ICI and survival in two independent cohorts of patients with advanced non-small-cell lung cancer (NSCLC). RESULTS: Inflamed IP correlated with enrichment in local immune cytolytic activity, higher response rate, and prolonged progression-free survival compared with patients with immune-excluded or immune-desert phenotypes. At the WSI level, there was significant positive correlation between tumor proportion score (TPS) as determined by the AI model and control TPS analyzed by pathologists (P < .001). Overall, 44.0% of tumors were inflamed, 37.1% were immune-excluded, and 18.9% were immune-desert. Incidence of inflamed IP in patients with programmed death ligand-1 TPS at < 1%, 1%-49%, and ≥ 50% was 31.7%, 42.5%, and 56.8%, respectively. Median progression-free survival and overall survival were, respectively, 4.1 months and 24.8 months with inflamed IP, 2.2 months and 14.0 months with immune-excluded IP, and 2.4 months and 10.6 months with immune-desert IP. CONCLUSION: The AI-powered spatial analysis of TIL correlated with tumor response and progression-free survival of ICI in advanced NSCLC. This is potentially a supplementary biomarker to TPS as determined by a pathologist.