Summary
A study developed an AI-based system, ATLS-8, to quantify CD8+ tumor-infiltrating lymphocytes and assess their prognostic value in hepatocellular carcinoma patients undergoing liver resection. The system demonstrated improved prognostic accuracy compared to clinical variables alone, with higher CD8+ TIL density correlating with better overall survival.
Editors: Rouzan Karabakhtsian, MD, PhD, professor of pathology and director of the Women’s Health Pathology Fellowship, Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY; S. Emily Bachert, MD, associate pathologist, Brigham and Women’s Hospital, Boston; Amarpreet Bhalla, MD, assistant professor of pathology, Albert Einstein College of Medicine, Montefiore Medical Center; Divya Sharma, MD, associate professor, Department of Pathology and Laboratory Medicine, University of Cincinnati Medical Center; and Paula Toro, MD, gastrointestinal and hepatobiliary fellow, Cleveland Clinic.
AI-based tumor-infiltrating lymphocyte scoring system for assessing HCC prognosis with liver resection
March 2024—Tumor-infiltrating lymphocytes, particularly CD8+ TILs, are key prognostic markers in many cancers. However, their prognostic value in hepatocellular carcinoma (HCC) has been the focus of opposing research findings. The authors conducted a study to develop an AI-based system to quantify CD8+ TILs and assess their prognostic value for patients with HCC, given the heterogeneous outcomes in patients with HCC undergoing liver resection. They conducted a retrospective multicenter study of patients undergoing liver resection across three cohorts (N=514). They trained a deep neural network and a random forest model to segment tumor regions and locate CD8+ TILs in hematoxylin and eosin-stained and CD8-stained whole slide images. The authors developed the Automated CD8+ Tumor-infiltrating Lymphocyte Scoring (ATLS-8) system to qualify the CD8+ TILs and assess their prognostic value. In the discovery cohort, the five-year overall survival rates were 34.05 percent for the ATLS-8 high-risk group and 65.03 percent for the ATLS-8 low-risk group (hazard ratio [HR], 2.40; 95 percent confidence interval [CI], 1.37–4.19; p=.015). These findings were confirmed in validation cohort one, which had five-year overall survival rates of 28.57 percent for the high-risk group and 68.73 percent for the low-risk group (HR, 3.38; 95 percent CI, 1.27–9.02; p=.0098), and validation cohort two, which had overall survival rates of 59.26 percent for the high-risk group and 81.48 percent for the low-risk group (HR, 2.74; 95 percent CI, 1.05–7.15; p=.031). ATLS-8 improved the prognostic model based on clinical variables (C-index, 0.770 versus 0.757, 0.769 versus 0.727, and 0.712 versus 0.642, respectively, in the three cohorts). The authors concluded that in patients with HCC undergoing resection, higher CD8+ TIL density correlated with better overall survival based on ATLS-8 assessment. The authors determined that this system is a promising tool for advancing clinical immune microenvironment assessment and outcome prediction.
Chen Z, Xie T, Chen S, et al. AI-based tumor-infiltrating lymphocyte scoring system for assessing HCC prognosis in patients undergoing liver resection. JHEP Reports. 2025;7(2). doi.org/10.1016/j.jhepr.2024.101270
Correspondence: Dr. Cheng Lu at lucheng@gdph.org.cn
Evidence supporting unified assessment criteria for HER2 IHC in colorectal carcinoma
Human epidermal growth factor receptor 2 is an important biomarker for managing metastatic colorectal carcinoma. With FDA tumor-agnostic approval of fam-trastuzumab deruxtecan-nxki for treating unresectable or metastatic HER2-positive solid neoplasms, the interpretation of HER2 status in colorectal carcinoma (CRC) has become more important than ever. IHC with reflex in situ hybridization (ISH) is an accepted standard method of assessment. There are two sets of criteria for interpreting results—HER2 Amplification for Colorectal Cancer Enhanced Stratification (HERACLES) criteria and MyPathway criteria. The HERACLES criteria require ISH confirmation when IHC staining is 3+ in 10 percent to 49 percent of cells, whereas the MyPathway criteria mirror those for gastric HER2 assessment and do not recommend ISH confirmation in the previous scenario. The authors conducted a study to assess the prevalence of HER2 3+ heterogeneity and its association with ERBB2 copy number amplification to evaluate the necessity of ISH testing when IHC staining is 3+ in fewer than 50 percent of cells. Next-generation sequencing (NGS) of DNA (592-gene panel or whole exome sequencing) was performed for 13,208 CRC tumors submitted to Caris Life Sciences. HER2 (4B5) expression was tested using IHC. A subset of tumors were tested for ERBB2 amplification via chromogenic ISH or NGS (copy number amplification), or both. The χ2 test or Fisher exact test was applied, with P values adjusted for multiple comparisons (P< .05). Of the 13,208 CRCs with HER2 IHC, 87.4 percent (11,541) were negative for HER2 expression (3+ intensity and less than 10 percent tumor cell staining), and 11.2 percent (1,473) demonstrated at least low HER2 expression (1 to 2+ and 10 percent). Only 1.5 percent (194) of all tested tumors were positive or heterogeneously positive for HER2 overexpression (3+ and 10 percent). Of these, 14 percent (28 of 194) had heterogenous HER2 overexpression (3+ staining of 10 percent to 49 percent of cells). All 22 (100 percent) HER2-positive/heterogenous cases with successful ISH testing demonstrated amplification via ISH. Because the classification of tumors as HER2 positive/heterogenous using IHC correlated closely with ISH positivity, the study results suggest that ISH is unnecessary for CRCs with 3+ HER2 overexpression in 10 percent to 49 percent of neoplastic cells. The authors support use of the MyPathway system for scoring HER2, which can reduce laboratory expenses by eliminating ISH testing. The study also demonstrated that NGS determination of gene copy number less frequently correlated with IHC findings. Therefore, laboratories should be cautious when interpreting ERBB2 amplification using NGS in CRC without adding other testing methodologies. Given the frequency with which treating oncologists order IHC, ISH, and NGS interchangeably, the authors may advocate for upfront IHC testing with appropriate ISH reflex in the situation of 2+ staining in more than 10 percent of tumor cells for appropriately documenting HER2 status. Simultaneous NGS could be useful, but its results should not supersede IHC and ISH findings on final pathology reports.
Evans MG, Krause HB, Xiu J, et al. Evidence for unified assessment criteria of HER2 immunohistochemistry in colorectal carcinoma. Mod Pathol. 2025. doi.org/10.1016/j.modpat.2024.100654
Correspondence: Dr. Jaclyn F. Hechtman at jhechtman@carisls.com