Response to immunotherapy in gastric cancer is associated with microsatellite instability (or mismatch repair deficiency) and Epstein-Barr virus (EBV) positivity. We therefore aimed to develop and ...validate deep learning-based classifiers to detect microsatellite instability and EBV status from routine histology slides.
In this retrospective, multicentre study, we collected tissue samples from ten cohorts of patients with gastric cancer from seven countries (South Korea, Switzerland, Japan, Italy, Germany, the UK and the USA). We trained a deep learning-based classifier to detect microsatellite instability and EBV positivity from digitised, haematoxylin and eosin stained resection slides without annotating tumour containing regions. The performance of the classifier was assessed by within-cohort cross-validation in all ten cohorts and by external validation, for which we split the cohorts into a five-cohort training dataset and a five-cohort test dataset. We measured the area under the receiver operating curve (AUROC) for detection of microsatellite instability and EBV status. Microsatellite instability and EBV status were determined to be detectable if the lower bound of the 95% CI for the AUROC was above 0·5.
Across the ten cohorts, our analysis included 2823 patients with known microsatellite instability status and 2685 patients with known EBV status. In the within-cohort cross-validation, the deep learning-based classifier could detect microsatellite instability status in nine of ten cohorts, with AUROCs ranging from 0·597 (95% CI 0·522–0·737) to 0·836 (0·795–0·880) and EBV status in five of eight cohorts, with AUROCs ranging from 0·819 (0·752–0·841) to 0·897 (0·513–0·966). Training a classifier on the pooled training dataset and testing it on the five remaining cohorts resulted in high classification performance with AUROCs ranging from 0·723 (95% CI 0·676–0·794) to 0·863 (0·747–0·969) for detection of microsatellite instability and from 0·672 (0·403–0·989) to 0·859 (0·823–0·919) for detection of EBV status.
Classifiers became increasingly robust when trained on pooled cohorts. After prospective validation, this deep learning-based tissue classification system could be used as an inexpensive predictive biomarker for immunotherapy in gastric cancer.
German Cancer Aid and German Federal Ministry of Health.
Epstein–Barr virus positivity (EBV(+)) and high‐microsatellite instability (MSI‐H) have been identified as molecular subgroups in gastric carcinoma. The aim of our study was to determine the ...prognostic and predictive relevance of these subgroups in the context of platinum/5‐fluorouracil (5‐FU) based preoperative chemotherapy (CTx). Additionally, we investigated the clinical relevance of the low‐MSI (MSI‐L) phenotype. We analysed 760 adenocarcinomas of the stomach or the gastro‐oesophageal junction encompassing 143 biopsies before CTx and 617 resected tumours (291 without and 326 after CTx). EBV was determined by PCR and in situ hybridisation for selected cases. MSI was analysed by PCR using five microsatellite markers and classified as MSI‐H and MSI‐L. Frequencies of EBV(+), MSI‐H and MSI‐L in the biopsies before CTx were 4.2, 10.5 and 4.9% respectively. EBV(+) or MSI‐H did not correlate with response, but MSI‐L was associated with better response (p = 0.011). In the resected tumours, frequencies of EBV(+), MSI‐H and MSI‐L were 3.9, 9.6 and 4.5% respectively. Overall survival (OS) was significantly different in the non‐CTx group (p = 0.014). Patients with EBV(+) tumours showed the best OS, followed by MSI‐H. MSI‐L was significantly associated with worse OS (hazard ratio HR, 2.21; 95% confidence interval CI, 1.21–4.04, p = 0.01). In the resected tumours after CTx, MSI‐H was also associated with increased OS (HR, 0.54; 95% CI, 0.26–1.09, p = 0.085). In multivariable analysis, molecular classification was an independent prognostic factor in the completely resected (R0) non‐CTx group (p = 0.035). In conclusion, MSI‐H and EBV(+) are not predictive of response to neoadjuvant platinum/5‐FU based CTx, but they are indicative of a good prognosis. In particular, MSI‐H indicates a favourable prognosis irrespective of treatment with CTx. MSI‐L predicts good response to CTx and its negative prognostic effect for patients treated with surgery alone suggests that MSI‐L might help to identify patients with potentially high‐benefit from preoperative CTx.
Background
Recently, we presented Stroma AReactive Invasion Front Areas (SARIFA) as a new histomorphologic negative prognostic biomarker in gastric cancer. It is defined as direct contact between ...tumor cells and fat cells. The aim of this study was to further elucidate the underlying genomic, transcriptional, and immunological mechanisms of the SARIFA phenomenon.
Methods
To address these questions, SARIFA was classified on H&E-stained tissue sections of three cohorts: an external cohort (
n
= 489, prognostic validation), the TCGA-STAD cohort (
n
= 194, genomic and transcriptomic analysis), and a local cohort (
n
= 60, digital spatial profiling (whole transcriptome) and double RNA in situ hybridization/immunostaining of cytokines).
Results
SARIFA status proved to be an independent negative prognostic factor for overall survival in an external cohort of gastric carcinomas. In TCGA-STAD cohort, SARIFA is not driven by distinct genomic alterations, whereas the gene expression analyses showed an upregulation of FABP4 in SARIFA-positive tumors. In addition, the transcriptional regulations of white adipocyte differentiation, triglyceride metabolism, and catabolism were upregulated in pathway analyses. In the DSP analysis of SARIFA-positive tumors, FABP4 and the transcriptional regulation of white adipocyte differentiation were upregulated in macrophages. Additionally, a significantly lower expression of the cytokines IL6 and TNFα was observed at the invasion front.
Conclusions
SARIFA proves to be a strong negative prognostic biomarker in advanced gastric cancer, implicating an interaction of tumor cells with tumor-promoting adipocytes with crucial changes in tumor cell metabolism. SARIFA is not driven by tumor genetics but is very likely driven by an altered immune response as a causative mechanism.
The activity of the epidermal growth factor receptor (EGFR)-directed monoclonal antibody cetuximab combined with oxaliplatin/leucovorin/5-fluorouracil (FUFOX) was assessed in first-line metastatic ...gastric and oesophago-gastric junction (OGJ) cancer in a prospective phase II study showing a promising objective tumour response rate of 65% and a low mutation frequency of KRAS (3%). The aim of the correlative tumour tissue studies was to investigate the relationship between EGFR gene copy numbers, activation of the EGFR pathway, expression and mutation of E-cadherin, V600E BRAF mutation and clinical outcome of patients with gastric and OGJ cancer treated with cetuximab combined with FUFOX.
Patients included in this correlative study (n = 39) were a subset of patients from the clinical phase II study. The association between EGFR gene copy number, activation of the EGFR pathway, abundance and mutation of E-cadherin which plays an important role in these disorders, BRAF mutation and clinical outcome of patients was studied. EGFR gene copy number was assessed by FISH. Expression of the phosphorylated forms of EGFR and its downstream effectors Akt and MAPK, in addition to E-cadherin was analysed by immunohistochemistry. The frequency of mutant V600E BRAF was evaluated by allele-specific PCR and the mutation profile of the E-cadherin gene CDH1 was examined by DHPLC followed by direct sequence analysis. Correlations with overall survival (OS), time to progression (TTP) and overall response rate (ORR) were assessed.
Our study showed a significant association between increased EGFR gene copy number (≥ 4.0) and OS in gastric and OGJ cancer, indicating the possibility that patients may be selected for treatment on a genetic basis. Furthermore, a significant correlation was shown between activated EGFR and shorter TTP and ORR, but not between activated EGFR and OS. No V600E BRAF mutations were identified. On the other hand, an interesting trend between high E-cadherin expression levels and better OS was observed and two CDH1 exon 9 missense mutations (A408V and D402H) were detected.
Our finding that increased EGFR gene copy numbers, activated EGFR and the E-cadherin status are potentially interesting biomarkers needs to be confirmed in larger randomized clinical trials.
Multicentre clinical study with the European Clinical Trials Database number 2004-004024-12.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Background
Computational pathology uses deep learning (DL) to extract biomarkers from routine pathology slides. Large multicentric datasets improve performance, but such datasets are scarce for ...gastric cancer. This limitation could be overcome by Swarm Learning (SL).
Methods
Here, we report the results of a multicentric retrospective study of SL for prediction of molecular biomarkers in gastric cancer. We collected tissue samples with known microsatellite instability (MSI) and Epstein–Barr Virus (EBV) status from four patient cohorts from Switzerland, Germany, the UK and the USA, storing each dataset on a physically separate computer.
Results
On an external validation cohort, the SL-based classifier reached an area under the receiver operating curve (AUROC) of 0.8092 (± 0.0132) for MSI prediction and 0.8372 (± 0.0179) for EBV prediction. The centralized model, which was trained on all datasets on a single computer, reached a similar performance.
Conclusions
Our findings demonstrate the feasibility of SL-based molecular biomarkers in gastric cancer. In the future, SL could be used for collaborative training and, thus, improve the performance of these biomarkers. This may ultimately result in clinical-grade performance and generalizability.
The molecular characterization of endometrial endometrioid adenocarcinomas has provided major advances in its prognostic stratification. However, risk assessment of microsatellite instability (MSI) ...and copy-number (CN)-low cases remains a challenge. Thus, we aimed to identify tissue-based morphologic biomarkers that might help in the prognostic stratification of these cases. Histomorphologic parameters (WHO grading, tumor budding (TB), tumor-stroma ratio (as a quantitative description of stromal desmoplasia), tumor-infiltrating lymphocytes (TIL), "microcystic, elongated, fragmented" (MELF) pattern) were analyzed in resection specimens of the TCGA-UCEC cohort (
= 228). For each quantitative parameter, a two-tiered system was developed utilizing systematically determined cutoffs. Associations with survival outcomes were calculated in univariate and multivariate analysis and validated in two independent cohorts. In MSI tumors, only TB remained an independent prognostic factor. TB (≥3 buds/high-power field) was associated with inferior outcomes and with lymph node metastases. The prognostic significance of TB was confirmed in two validation cohorts. For CN-low tumors, established grading defined by the WHO was independently prognostic with inferior outcomes for high-grade tumors. The evaluation of TB might help in identifying MSI-patients with unfavorable prognosis who, e.g., could benefit from lymphadenectomy. WHO-based grading facilitates independent prognostic stratification of CN-low endometrioid adenocarcinomas. Therefore, we propose the utilization of TB and WHO-based grading, two tissue-based and easy-to-assess biomarkers, in MSI/CN-low endometrial carcinomas for improved clinical management.
Even though the post-operative outcome varies greatly among patients with nodal positive colon cancer (UICC stage III), personalized prediction of systemic disease recurrence is currently ...insufficient. We investigated in a retrospective setting whether genetic and immunological biomarkers can be applied for stratification of distant metastasis occurrence risk.
Eighty four patients with complete resection (R0) of stage III colon cancer from two clinical centres were analysed for genetic biomarkers: microsatellite instability, oncogenic mutations in KRAS exon2 and BRAF exon15, expression of osteopontin and the metastasis-associated genes SASH1 and MACC1. Tumor-infiltrating CD3 and CD8 positive T-cells were quantified by immunocytochemistry. Results were correlated with outcome and response to 5-FU based adjuvant chemotherapy, using Cox's proportional hazard models and integrative two-step cluster analysis.
Distant metastasis risk was significantly correlated with oncogenic KRAS mutations (p = 0.015), expression of SASH1 (p = 0.016), and the density of CD8-positive T-cells (p = 0.007) in Kaplan-Meier analysis. Upon multivariate Cox-regression analysis, KRAS mutation (p = 0.008) and density of CD8-positive TILs (p = 0.009) were retained as prognostic parameters for metachronous distant metastasis. Integrative two-step cluster analysis was used to combine all genetic markers, allowing stratification of patient subgroups. Post-operative distant metastasis risk ranged from 31% (low-risk) to 41% (intermediate), and 57% (high-risk) (p = 0.032). Increased expression of osteopontin (p = 0.019) and low density of CD8-positive T-cells (p = 0.043) were significantly associated with unfavourable response to 5-FU.
Integrative biomarker analysis allows stratification of stage III colon cancer patients for the risk of metastatic disease recurrence and may indicate response to 5-FU. Thus, biomarker analysis might facilitate the use of adjuvant therapy for high risk patients.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
We aimed to investigate patients with gastric/gastro-esophageal adenocarcinomas for sex- and age-specific differences regarding overall survival (OS) and response to neoadjuvant chemotherapy (CTx) ...under consideration of tumor specific molecular subtypes. Overall, 717 patients were analyzed, including 426 patients treated with and 291 treated without neoadjuvant CTx. Microsatellite instability (MSI) and Epstein-Barr virus positivity (EBV+) were determined previously. Females demonstrated a significantly increased OS (
0.035), particularly in the subgroup treated with CTx (
0.054). No significant differences regarding age were found. In the molecular subgroups, no sex-related differences were observed in the non-CTx group. However in the CTx group, females with MSI-high (H) tumors showed the best OS (
0.043), followed by the male MSI-H (
0.198) and female MSS (
0.114) compared to the male MSS group as reference. The interaction between sex and MSI in this patient group was noticeable (
0.053) and was included as a relevant factor in multivariable analyses. In conclusion, our results show an effect of sex on OS in gastric/gastro-esophageal cancer specifically for patients treated with neoadjuvant CTx. The superior survival of women with MSI-H tumors after neoadjuvant CTx implies that combined consideration of these factors could contribute to an individualized treatment of the patients.
We aimed to determine the clinical and prognostic relevance of allelic imbalance (AI) of the major histocompatibility complex (MHC) class I genes, encompassing the human leukocyte antigen (HLA) class ...I and beta-2 microglobulin (B2M) genes, in the context of neoadjuvant platinum/fluoropyrimidine chemotherapy (CTx). Biopsies before CTx were studied in 158 patients with adenocarcinoma of the stomach or gastroesophageal junction. The response was histopathologically evaluated. AI was detected by multiplex PCRs analysis of four or five microsatellite markers in HLA and B2M regions, respectively. AI with no marker was significantly associated with response or survival. However, subgroup analysis revealed differences. AI at marker D6S265, close to the HLA-A gene, was associated with an obvious increased risk in responding (HR, 3.62; 95% CI, 0.96-13.68,
= 0.058) but not in non-responding patients (HR, 0.92; 95% CI, 0.51-1.65,
= 0.773). Markers D6S273 and D6S2872 showed similar results. The interaction between AI at D6S265 and response to CTx was significant in a multivariable analysis (
= 0.010). No associations were observed for B2M markers. Our results underline the importance of intact neoantigen presentation specifically for responding patients and may help explain an unexpectedly poor survival of a patient despite significant tumor regression after neoadjuvant platinum/fluoropyrimidine CTx.
We investigated the prognostic and predictive impact of p53 expression for gastric cancer (GC) patients treated without or with preoperative chemotherapy (CTx) and its relationship with specific ...molecular GC subtypes. Specimens from 694 GC patients (562 surgical resection specimens without or after CTx, 132 biopsies before CTx) were analyzed by p53 immunohistochemistry. High (H) and low (L) microsatellite instability (MSI) and Epstein–Barr virus positivity were determined previously. Our results show that aberrant p53 expression was a negative prognostic factor in uni- and multivariable analysis in the resection specimens cohort (each p < 0.01). Subgroup analysis showed the strongest prognostic effect for patients with distally located tumors or no CTx treatment. In the biopsy cohort before CTx, p53 did not predict response or survival. p53 expression was significantly different among the molecular subtypes in surgical resection and bioptic specimens with strong association of altered p53 with MSI-L. Patients with MSI-H and aberrant p53 showed the worst survival in the biopsy cohort. In conclusion, the prognostic impact of p53 in GC differs according to tumor localization and CTx. Altered p53 is characteristic for MSI-L, and the p53 status in biopsies before CTx delineates MSI-H subtypes with inverse prognostic impact.
PDF
