Comprehensive assessment of immune context and immunotherapy response via noninvasive imaging in gastric cancer.

BACKGROUNDThe tumor immune microenvironment can provide prognostic and therapeutic information. We aimed to develop noninvasive imaging biomarkers from computed tomography (CT) for comprehensive evaluation of immune context and investigate their associations with prognosis and immunotherapy response in gastric cancer (GC).METHODSThis study involved 2,600 patients with GC from 9 independent cohorts. We developed and validated 2 CT imaging biomarkers (lymphoid radiomics score [LRS] and myeloid radiomics score [MRS]) for evaluating the IHC-derived lymphoid and myeloid immune context respectively, and integrated them into a combined imaging biomarker [LRS/MRS: low(-) or high(+)] with 4 radiomics immune subtypes: 1 (-/-), 2 (+/-), 3 (-/+), and 4 (+/+). We further evaluated the imaging biomarkers' predictive values on prognosis and immunotherapy response.RESULTSThe developed imaging biomarkers (LRS and MRS) had a high accuracy in predicting lymphoid (AUC range: 0.765-0.773) and myeloid (AUC range: 0.736-0.750) immune context. Further, similar to the IHC-derived immune context, 2 imaging biomarkers (HR range: 0.240-0.761 for LRS; 1.301-4.012 for MRS) and the combined biomarker were independent predictors for disease-free and overall survival in the training and all validation cohorts (all P < 0.05). Additionally, patients with high LRS or low MRS may benefit more from immunotherapy (P < 0.001). Further, a highly heterogeneous outcome on objective response ​rate was observed in 4 imaging subtypes: 1 (-/-) with 27.3%, 2 (+/-) with 53.3%, 3 (-/+) with 10.2%, and 4 (+/+) with 30.0% (P < 0.0001).CONCLUSIONThe noninvasive imaging biomarkers could accurately evaluate the immune context and provide information regarding prognosis and immunotherapy for GC.

Biology-guided deep learning predicts prognosis and cancer immunotherapy response.

Substantial progress has been made in using deep learning for cancer detection and diagnosis in medical images. Yet, there is limited success on prediction of treatment response and outcomes, which has important implications for personalized treatment strategies. A significant hurdle for clinical translation of current data-driven deep learning models is lack of interpretability, often attributable to a disconnect from the underlying pathobiology. Here, we present a biology-guided deep learning approach that enables simultaneous prediction of the tumor immune and stromal microenvironment status as well as treatment outcomes from medical images. We validate the model for predicting prognosis of gastric cancer and the benefit from adjuvant chemotherapy in a multi-center international study. Further, the model predicts response to immune checkpoint inhibitors and complements clinically approved biomarkers. Importantly, our model identifies a subset of mismatch repair-deficient tumors that are non-responsive to immunotherapy and may inform the selection of patients for combination treatments.

Non-invasive tumor microenvironment evaluation and treatment response prediction in gastric cancer using deep learning radiomics.

The tumor microenvironment (TME) plays a critical role in disease progression and is a key determinant of therapeutic response in cancer patients. Here, we propose a noninvasive approach to predict the TME status from radiological images by combining radiomics and deep learning analyses. Using multi-institution cohorts of 2,686 patients with gastric cancer, we show that the radiological model accurately predicted the TME status and is an independent prognostic factor beyond clinicopathologic variables. The model further predicts the benefit from adjuvant chemotherapy for patients with localized disease. In patients treated with checkpoint blockade immunotherapy, the model predicts clinical response and further improves predictive accuracy when combined with existing biomarkers. Our approach enables noninvasive assessment of the TME, which opens the door for longitudinal monitoring and tracking response to cancer therapy. Given the routine use of radiologic imaging in oncology, our approach can be extended to many other solid tumor types.

Cancer immunotherapy response prediction from multi-modal clinical and image data using semi-supervised deep learning.

BACKGROUND AND PURPOSE: Immunotherapy is a standard treatment for many tumor types. However, only a small proportion of patients derive clinical benefit and reliable predictive biomarkers of immunotherapy response are lacking. Although deep learning has made substantial progress in improving cancer detection and diagnosis, there is limited success on the prediction of treatment response. Here, we aim to predict immunotherapy response of gastric cancer patients using routinely available clinical and image data. MATERIALS AND METHODS: We present a multi-modal deep learning radiomics approach to predict immunotherapy response using both clinical data and computed tomography images. The model was trained using 168 advanced gastric cancer patients treated with immunotherapy. To overcome limitations of small training data, we leverage an additional dataset of 2,029 patients who did not receive immunotherapy in a semi-supervised framework to learn intrinsic imaging phenotypes of the disease. We evaluated model performance in two independent cohorts of 81 patients treated with immunotherapy. RESULTS: The deep learning model achieved area under receiver operating characteristics curve (AUC) of 0.791 (95% CI 0.633-0.950) and 0.812 (95% CI 0.669-0.956) for predicting immunotherapy response in the internal and external validation cohorts. When combined with PD-L1 expression, the integrative model further improved the AUC by 4-7% in absolute terms. CONCLUSION: The deep learning model achieved promising performance for predicting immunotherapy response from routine clinical and image data. The proposed multi-modal approach is general and can incorporate other relevant information to further improve prediction of immunotherapy response.

Noninvasive imaging evaluation of peritoneal recurrence and chemotherapy benefit in gastric cancer after gastrectomy: a multicenter study.

BACKGROUND: Peritoneal recurrence (PR) is the predominant pattern of relapse after curative-intent surgery in gastric cancer (GC) and indicates a dismal prognosis. Accurate prediction of PR is crucial for patient management and treatment. The authors aimed to develop a noninvasive imaging biomarker from computed tomography (CT) for PR evaluation, and investigate its associations with prognosis and chemotherapy benefit. METHODS: In this multicenter study including five independent cohorts of 2005 GC patients, the authors extracted 584 quantitative features from the intratumoral and peritumoral regions on contrast-enhanced CT images. The artificial intelligence algorithms were used to select significant PR-related features, and then integrated into a radiomic imaging signature. And improvements of diagnostic accuracy for PR by clinicians with the signature assistance were quantified. Using Shapley values, the authors determined the most relevant features and provided explanations to prediction. The authors further evaluated its predictive performance in prognosis and chemotherapy response. RESULTS: The developed radiomics signature had a consistently high accuracy in predicting PR in the training cohort (area under the curve: 0.732) and internal and Sun Yat-sen University Cancer Center validation cohorts (0.721 and 0.728). The radiomics signature was the most important feature in Shapley interpretation. The diagnostic accuracy of PR with the radiomics signature assistance was improved by 10.13-18.86% for clinicians ( P <0.001). Furthermore, it was also applicable in the survival prediction. In multivariable analysis, the radiomics signature remained an independent predictor for PR and prognosis ( P <0.001 for all). Importantly, patients with predicting high risk of PR from radiomics signature could gain survival benefit from adjuvant chemotherapy. By contrast, chemotherapy had no impact on survival for patients with a predicted low risk of PR. CONCLUSION: The noninvasive and explainable model developed from preoperative CT images could accurately predict PR and chemotherapy benefit in patients with GC, which will allow the optimization of individual decision-making.

Multimodality radiomics analysis based on [18F]FDG PET/CT imaging and multisequence MRI: application to nasopharyngeal carcinoma prognosis.

OBJECTIVES: To determine whether radiomics models developed from 2-deoxy-2-[18F]fluoro-D-glucose ([18F]FDG) PET/CT combined with multisequence MRI could contribute to predicting the progression-free survival (PFS) of nasopharyngeal carcinoma (NPC) patients. METHODS: One hundred thirty-two NPC patients who underwent both PET/CT and MRI scanning were retrospectively enrolled (88 vs. 44 for training vs. testing). For each modality/sequence (i.e., PET, CT, T1, T1C, and T2), 1906 radiomics features were extracted from the primary tumor volume. Univariate Cox model and correlation analysis were used for feature selection. A multivariate Cox model was used to establish radiomics signature. Prognostic performances of 5 individual modality models and 12 multimodality models (3 integrations × 4 fusion strategies) were assessed by the concordance index (C-index) and log-rank test. A clinical-radiomics nomogram was built to explore the clinical utilities of radiomics signature, which was evaluated by discrimination, calibration curve, and decision curve analysis (DCA). RESULTS: The radiomics signatures of individual modalities showed limited prognostic efficacy with a C-index of 0.539-0.664 in the testing cohort. Different fusion strategies exhibited a slight difference in predictive performance. The PET/CT and MRI integrated model achieved the best performance with a C-index of 0.745 (95% CI, 0.619-0.865) in the testing cohort (log-rank test, p < 0.05). Clinical-radiomics nomogram further improved the prognosis, which also showed satisfactory discrimination, calibration, and net benefit. CONCLUSIONS: Multimodality radiomics analysis by combining PET/CT with multisequence MRI could potentially improve the efficacy of PFS prediction for NPC patients. KEY POINTS: • Individual modality radiomics models showed limited performance in prognosis evaluation for NPC patients. • Combined PET, CT and multisequence MRI radiomics signature could improve the prognostic efficacy. • Multilevel fusion strategies exhibit comparable performance but feature-level fusion deserves more attention.

Non-invasive CT imaging biomarker to predict immunotherapy response in gastric cancer: a multicenter study.

BACKGROUND: Despite remarkable benefits have been provided by immune checkpoint inhibitors in gastric cancer (GC), predictions of treatment response and prognosis remain unsatisfactory, making identifying biomarkers desirable. The aim of this study was to develop and validate a CT imaging biomarker to predict the immunotherapy response in patients with GC and investigate the associated immune infiltration patterns. METHODS: This retrospective study included 294 GC patients who received anti-PD-1/PD-L1 immunotherapy from three independent medical centers between January 2017 and April 2022. A radiomics score (RS) was developed from the intratumoral and peritumoral features on pretreatment CT images to predict immunotherapy-related progression-free survival (irPFS). The performance of the RS was evaluated by the area under the time-dependent receiver operating characteristic curve (AUC). Multivariable Cox regression analysis was performed to construct predictive nomogram of irPFS. The C-index was used to determine the performance of the nomogram. Bulk RNA sequencing of tumors from 42 patients in The Cancer Genome Atlas was used to investigate the RS-associated immune infiltration patterns. RESULTS: Overall, 89 of 294 patients (median age, 57 years (IQR 48-66 years); 171 males) had an objective response to immunotherapy. The RS included 13 CT features that yielded AUCs of 12-month irPFS of 0.787, 0.810 and 0.785 in the training, internal validation, and external validation 1 cohorts, respectively, and an AUC of 24-month irPFS of 0.805 in the external validation 2 cohort. Patients with low RS had longer irPFS in each cohort (p<0.05). Multivariable Cox regression analyses showed RS is an independent prognostic factor of irPFS. The nomogram that integrated the RS and clinical characteristics showed improved performance in predicting irPFS, with C-index of 0.687-0.778 in the training and validation cohorts. The CT imaging biomarker was associated with M1 macrophage infiltration. CONCLUSION: The findings of this prognostic study suggest that the non-invasive CT imaging biomarker can effectively predict immunotherapy outcomes in patients with GC and is associated with innate immune signaling, which can serve as a potential tool for individual treatment decisions.

Noninvasive imaging of the tumor immune microenvironment correlates with response to immunotherapy in gastric cancer.

The tumor immune microenvironment (TIME) is associated with tumor prognosis and immunotherapy response. Here we develop and validate a CT-based radiomics score (RS) using 2272 gastric cancer (GC) patients to investigate the relationship between the radiomics imaging biomarker and the neutrophil-to-lymphocyte ratio (NLR) in the TIME, including its correlation with prognosis and immunotherapy response in advanced GC. The RS achieves an AUC of 0.795-0.861 in predicting the NLR in the TIME. Notably, the radiomics imaging biomarker is indistinguishable from the IHC-derived NLR status in predicting DFS and OS in each cohort (HR range: 1.694-3.394, P < 0.001). We find the objective responses of a cohort of anti-PD-1 immunotherapy patients is significantly higher in the low-RS group (60.9% and 42.9%) than in the high-RS group (8.1% and 14.3%). The radiomics imaging biomarker is a noninvasive method to evaluate TIME, and may correlate with prognosis and anti PD-1 immunotherapy response in GC patients.

PAK6 promotes homologous-recombination to enhance chemoresistance to oxaliplatin through ATR/CHK1 signaling in gastric cancer.

Chemoresistance remains the primary challenge of clinical treatment of gastric cancer (GC), making the biomarkers of chemoresistance crucial for treatment decision. Our previous study has reported that p21-actived kinase 6 (PAK6) is a prognostic factor for selecting which patients with GC are resistant to 5-fluorouracil/oxaliplatin chemotherapy. However, the mechanistic role of PAK6 in chemosensitivity remains unknown. The present study identified PAK6 as an important modulator of the DNA damage response (DDR) and chemosensitivity in GC. Analysis of specimens from patients revealed significant associations between the expression of PAK6 and poorer stages, deeper invasion, more lymph node metastases, higher recurrence rates, and resistance to oxaliplatin. Cells exhibited chemosensitivity to oxaliplatin after knockdown of PAK6, but showed more resistant to oxaliplatin when overexpressing PAK6. Functionally, PAK6 mediates cancer chemoresistance by enhancing homologous recombination (HR) to facilitate the DNA double-strand break repair. Mechanistically, PAK6 moves into nucleus to promote the activation of ATR, thereby further activating downstream repair protein CHK1 and recruiting RAD51 from cytoplasm to the DNA damaged site to repair the broken DNA in GC. Activation of ATR is the necessary step for PAK6 mediated HR repair to protect GC cells from oxaliplatin-induced apoptosis, and ATR inhibitor (AZD6738) could block the PAK6-mediated HR repair, thereby reversing the resistance to oxaliplatin and even promoting the sensitivity to oxaliplatin regardless of high expression of PAK6. In conclusion, these findings indicate a novel regulatory mechanism of PAK6 in modulating the DDR and chemoresistance in GC and provide a reversal suggestion in clinical decision.

Predicting peritoneal recurrence and disease-free survival from CT images in gastric cancer with multitask deep learning: a retrospective study.

BACKGROUND: Peritoneal recurrence is the predominant pattern of relapse after curative-intent surgery for gastric cancer and portends a dismal prognosis. Accurate individualised prediction of peritoneal recurrence is crucial to identify patients who might benefit from intensive treatment. We aimed to develop predictive models for peritoneal recurrence and prognosis in gastric cancer. METHODS: In this retrospective multi-institution study of 2320 patients, we developed a multitask deep learning model for the simultaneous prediction of peritoneal recurrence and disease-free survival using preoperative CT images. Patients in the training cohort (n=510) and the internal validation cohort (n=767) were recruited from Southern Medical University, Guangzhou, China. Patients in the external validation cohort (n=1043) were recruited from Sun Yat-sen University Cancer Center, Guangzhou, China. We evaluated the prognostic accuracy of the model as well as its association with chemotherapy response. Furthermore, we assessed whether the model could improve the ability of clinicians to predict peritoneal recurrence. FINDINGS: The deep learning model had a consistently high accuracy in predicting peritoneal recurrence in the training cohort (area under the receiver operating characteristic curve [AUC] 0·857; 95% CI 0·826-0·889), internal validation cohort (0·856; 0·829-0·882), and external validation cohort (0·843; 0·819-0·866). When informed by the artificial intelligence (AI) model, the sensitivity and inter-rater agreement of oncologists for predicting peritoneal recurrence was improved. The model was able to predict disease-free survival in the training cohort (C-index 0·654; 95% CI 0·616-0·691), internal validation cohort (0·668; 0·643-0·693), and external validation cohort (0·610; 0·583-0·636). In multivariable analysis, the model predicted peritoneal recurrence and disease-free survival independently of clinicopathological variables (p<0·0001 for all). For patients with a predicted high risk of peritoneal recurrence and low survival, adjuvant chemotherapy was associated with improved disease-free survival in both stage II disease (hazard ratio [HR] 0·543 [95% CI 0·362-0·815]; p=0·003) and stage III disease (0·531 [0·432-0·652]; p<0·0001). By contrast, chemotherapy had no impact on disease-free survival for patients with a predicted low risk of peritoneal recurrence and high survival. For the remaining patients, the benefit of chemotherapy depended on stage: only those with stage III disease derived benefit from chemotherapy (HR 0·637 [95% CI 0·484-0·838]; p=0·001). INTERPRETATION: The deep learning model could allow accurate prediction of peritoneal recurrence and survival in patients with gastric cancer. Prospective studies are required to test the clinical utility of this model in guiding personalised treatment in combination with clinicopathological criteria. FUNDING: None.

Radiomics signature based on computed tomography images for the preoperative prediction of lymph node metastasis at individual stations in gastric cancer: A multicenter study.

BACKGROUND: Specific diagnosis and treatment of gastric cancer (GC) require accurate preoperative predictions of lymph node metastasis (LNM) at individual stations, such as estimating the extent of lymph node dissection. This study aimed to develop a radiomics signature based on preoperative computed tomography (CT) images, for predicting the LNM status at each individual station. METHODS: We enrolled 1506 GC patients retrospectively from two centers as training (531) and external (975) validation cohorts, and recruited 112 patients prospectively from a single center as prospective validation cohort. Radiomics features were extracted from preoperative CT images and integrated with clinical characteristics to construct nomograms for LNM prediction at individual lymph node stations. Performance of the nomograms was assessed through calibration, discrimination and clinical usefulness. RESULTS: In training, external and prospective validation cohorts, radiomics signature was significantly associated with LNM status. Moreover, radiomics signature was an independent predictor of LNM status in the multivariable logistic regression analysis. The radiomics nomograms revealed good prediction performances, with AUCs of 0.716-0.871 in the training cohort, 0.678-0.768 in the external validation cohort and 0.700-0.841 in the prospective validation cohort for 12 nodal stations. The nomograms demonstrated a significant agreement between the actual probability and predictive probability in calibration curves. Decision curve analysis showed that nomograms had better net benefit than clinicopathologic characteristics. CONCLUSION: Radiomics nomograms for individual lymph node stations presented good prediction accuracy, which could provide important information for individual diagnosis and treatment of gastric cancer.

Radiographical assessment of tumour stroma and treatment outcomes using deep learning: a retrospective, multicohort study.

BACKGROUND: The tumour stroma microenvironment plays an important part in disease progression and its composition can influence treatment response and outcomes. Histological evaluation of tumour stroma is limited by access to tissue, spatial heterogeneity, and temporal evolution. We aimed to develop a radiological signature for non-invasive assessment of tumour stroma and treatment outcomes. METHODS: In this multicentre, retrospective study, we analysed CT images and outcome data of 2209 patients with resected gastric cancer from five independent cohorts recruited from two centres (Nanfang Hospital of Southern Medical University [Guangzhou, China] and Sun Yat-sen University Cancer Center [Guangzhou, China]). Patients with histologically confirmed gastric cancer, at least 15 lymph nodes harvested, preoperative abdominal CT available, and complete clinicopathological and follow-up data were eligible for inclusion. Tumour tissue was collected for patients in the training cohort (321 patients), internal validation cohort one (246 patients), and external validation cohort one (128 patients). Four stroma classes were defined according to the protein expression of α-smooth muscle actin and periostin assessed by immunohistochemistry. The primary objective was to predict the histologically based stroma classes by using preoperative CT images. We trained a deep convolutional neural network model using the training cohort and tested the model in the internal and external validation cohort one. We evaluated the model's association with prognosis in the training cohort, two internal, and two external validation cohorts and compared outcomes of patients who received or did not receive adjuvant chemotherapy. FINDINGS: The deep-learning model achieved a high diagnostic accuracy for assessing tumour stroma in both internal validation cohort one (area under the receiver operating characteristic curve [AUC] 0·96-0·98]) and external validation cohort one (AUC 0·89-0·94). The stromal imaging signature was significantly associated with disease-free survival and overall survival in all cohorts (p<0·0001). The predicted stroma classes remained an independent prognostic factor adjusting for clinicopathological variables including tumour size, stage, differentiation, and Lauren histology. In patients with stage II or III disease in predicted stroma classes one and two subgroups, patients who received adjuvant chemotherapy had improved survival compared with those who did not (in those with stage II disease hazard ratio [HR] 0·48 [95% CI 0·29-0·77], p=0·0021; and in those with stage III disease HR 0·70 [0·57-0·85], p=0·00042). However, in the other two subgroups adjuvant chemotherapy was not associated with survival and might even be detrimental in the predicted stroma class 4 subgroup (HR 1·48 [1·08-2·03], p=0·013). INTERPRETATION: The deep-learning model could allow for accurate and non-invasive evaluation of tumour stroma from CT images in gastric cancer. The radiographical model predicted chemotherapy outcomes and could be used in combination with clinicopathological criteria to refine prognosis and inform treatment decisions of patients with gastric cancer. FUNDING: None.

Prediction of local recurrence and distant metastasis using radiomics analysis of pretreatment nasopharyngeal [18F]FDG PET/CT images.

OBJECTIVES: To develop a radiomics signature to predict locoregional recurrence (LR) and distant metastasis (DM), as extracted from pretreatment 2-deoxy-2-[18F]fluoro-D-glucose ([18F]FDG) positron emission tomography/X-ray computed tomography (PET/CT) images in locally advanced nasopharyngeal carcinoma (NPC). MATERIALS AND METHODS: Eighty-five patients with Stage III-IVB NPC underwent pretreatment [18F]FDG PET/CT scans and received radiotherapy or chemoradiotherapy. 53 of them achieved disease control, and 32 of them failed after treatment (15: LR, 17: DM). A total of 114 radiomic features were extracted from PET/CT images. For univariate analysis, Wilcoxon test and Chi-square test were used to compare median values of features between different treatment outcomes and predict the risk of treatment failure, respectively. For multivariate analysis, all features were grouped into clusters based on Pearson correlation using hierarchical clustering, and the representative feature of each cluster was chosen by the Relief algorithm. Then sequential floating forward selection (SFFS) coupled with a support vector machine (SVM) classifier were used to derive the optimized feature set in terms of the area under receiver operating characteristic (ROC) curve (AUC). The performance of the model was evaluated by leave-one-out-cross-validation, fivefold cross-validation, tenfold cross-validation. RESULTS: Twenty features had significant differences between disease control and treatment failure. NPC patients with values of Compactness1, Compactness2, Coarseness_NGTDM or SGE_GLGLM above the median as well as patients with values of Irregularity, RLN_GLRLM or GLV_GLSZM below the median, showed a significant (p < 0.05) higher risk of treatment failure (about 50% vs. 25%). The derived radiomics signature consisted of 5 features with the highest AUC value of 0.8290 (sensitivity: 0.8438, specificity: 0.7736) using leave-one-out-cross-validation. CONCLUSION: Locoregional recurrence (LR) and DM of locally advanced NPC can be predicted using radiomics analysis of pretreatment [18F]FDG PET/CT. The SFFS feature selection coupled with SVM classifier can derive the optimized feature set with correspondingly highest AUC value for pretreatment prediction of LR and/or DM of NPC.

Noninvasive Prediction of Occult Peritoneal Metastasis in Gastric Cancer Using Deep Learning.

Importance: Occult peritoneal metastasis frequently occurs in patients with advanced gastric cancer and is poorly diagnosed with currently available tools. Because the presence of peritoneal metastasis precludes the possibility of curative surgery, there is an unmet need for a noninvasive approach to reliably identify patients with occult peritoneal metastasis. Objective: To assess the use of a deep learning model for predicting occult peritoneal metastasis based on preoperative computed tomography images. Design, Setting, and Participants: In this multicenter, retrospective cohort study, a deep convolutional neural network, the Peritoneal Metastasis Network (PMetNet), was trained to predict occult peritoneal metastasis based on preoperative computed tomography images. Data from a cohort of 1225 patients with gastric cancer who underwent surgery at Sun Yat-sen University Cancer Center (Guangzhou, China) were used for training purposes. To externally validate the model, data were collected from 2 independent cohorts comprising a total of 753 patients with gastric cancer who underwent surgery at Nanfang Hospital (Guangzhou, China) or the Third Affiliated Hospital of Southern Medical University (Guangzhou, China). The status of peritoneal metastasis for all patients was confirmed by pathological examination of pleural specimens obtained during surgery. Detailed clinicopathological data were collected for each patient. Data analysis was performed between September 1, 2019, and January 31, 2020. Main Outcomes and Measures: The area under the receiver operating characteristic curve (AUC) and decision curve were analyzed to evaluate performance in predicting occult peritoneal metastasis. Results: A total of 1978 patients (mean [SD] age, 56.0 [12.2] years; 1350 [68.3%] male) were included in the study. The PMetNet model achieved an AUC of 0.946 (95% CI, 0.927-0.965), with a sensitivity of 75.4% and a specificity of 92.9% in external validation cohort 1. In external validation cohort 2, the AUC was 0.920 (95% CI, 0.848-0.992), with a sensitivity of 87.5% and a specificity of 98.2%. The discrimination performance of PMetNet was substantially higher than conventional clinicopathological factors (AUC range, 0.51-0.63). In multivariable logistic regression analysis, PMetNet was an independent predictor of occult peritoneal metastasis. Conclusions and Relevance: The findings of this cohort study suggest that the PMetNet model can serve as a reliable noninvasive tool for early identification of patients with clinically occult peritoneal metastasis, which will inform individualized preoperative treatment decision-making and may avoid unnecessary surgery and complications. These results warrant further validation in prospective studies.

Development and Validation of a Deep Learning CT Signature to Predict Survival and Chemotherapy Benefit in Gastric Cancer: A Multicenter, Retrospective Study.

OBJECTIVE: We aimed to develop a deep learning-based signature to predict prognosis and benefit from adjuvant chemotherapy using preoperative computed tomography (CT) images. BACKGROUND: Current staging methods do not accurately predict the risk of disease relapse for patients with gastric cancer. METHODS: We proposed a novel deep neural network (S-net) to construct a CT signature for predicting disease-free survival (DFS) and overall survival in a training cohort of 457 patients, and independently tested it in an external validation cohort of 1158 patients. An integrated nomogram was constructed to demonstrate the added value of the imaging signature to established clinicopathologic factors for individualized survival prediction. Prediction performance was assessed with respect to discrimination, calibration, and clinical usefulness. RESULTS: The DeLIS was associated with DFS and overall survival in the overall validation cohort and among subgroups defined by clinicopathologic variables, and remained an independent prognostic factor in multivariable analysis (P< 0.001). Integrating the imaging signature and clinicopathologic factors improved prediction performance, with C-indices: 0.792-0.802 versus 0.719-0.724, and net reclassification improvement 10.1%-28.3%. Adjuvant chemotherapy was associated with improved DFS in stage II patients with high-DeLIS [hazard ratio = 0.362 (95% confidence interval 0.149-0.882)] and stage III patients with high- and intermediate-DeLIS [hazard ratio = 0.611 (0.442-0.843); 0.633 (0.433-0.925)]. On the other hand, adjuvant chemotherapy did not affect survival for patients with low-DeLIS, suggesting a predictive effect (Pinteraction = 0.048, 0.016 for DFS in stage II and III disease). CONCLUSIONS: The proposed imaging signature improved prognostic prediction and could help identify patients most likely to benefit from adjuvant chemotherapy in gastric cancer.

Radiomics Nomogram for Prediction of Peritoneal Metastasis in Patients With Gastric Cancer.

Objective: The aim of this study is to evaluate whether radiomics imaging signatures based on computed tomography (CT) could predict peritoneal metastasis (PM) in gastric cancer (GC) and to develop a nomogram for preoperative prediction of PM status. Methods: We collected CT images of pathological T4 gastric cancer in 955 consecutive patients of two cancer centers to analyze the radiomics features retrospectively and then developed and validated the prediction model built from 292 quantitative image features in the training cohort and two validation cohorts. Lasso regression model was applied for selecting feature and constructing radiomics signature. Predicting model was developed by multivariable logistic regression analysis. Radiomics nomogram was developed by the incorporation of radiomics signature and clinical T and N stage. Calibration, discrimination, and clinical usefulness were used to evaluate the performance of the nomogram. Results: In training and validation cohorts, PM status was associated with the radiomics signature significantly. It was found that the radiomics signature was an independent predictor for peritoneal metastasis in multivariable logistic analysis. For training and internal and external validation cohorts, the area under the receiver operating characteristic curves (AUCs) of radiomics signature for predicting PM were 0.751 (95%CI, 0.703-0.799), 0.802 (95%CI, 0.691-0.912), and 0.745 (95%CI, 0.683-0.806), respectively. Furthermore, for training and internal and external validation cohorts, the AUCs of radiomics nomogram for predicting PM were 0.792 (95%CI, 0.748-0.836), 0.870 (95%CI, 0.795-0.946), and 0.815 (95%CI, 0.763-0.867), respectively. Conclusions: CT-based radiomics signature could predict peritoneal metastasis, and the radiomics nomogram can make a meaningful contribution for predicting PM status in GC patient preoperatively.

Machine Learning Methods for Optimal Radiomics-Based Differentiation Between Recurrence and Inflammation: Application to Nasopharyngeal Carcinoma Post-therapy PET/CT Images.

PURPOSE: To identify optimal machine learning methods for radiomics-based differentiation of local recurrence versus inflammation from post-treatment nasopharyngeal positron emission tomography/X-ray computed tomography (PET/CT) images. PROCEDURES: Seventy-six nasopharyngeal carcinoma (NPC) patients were enrolled (41/35 local recurrence/inflammation as confirmed by pathology). Four hundred eighty-seven radiomics features were extracted from PET images for each patient. The diagnostic performance was investigated for 42 cross-combinations derived from 6 feature selection methods and 7 classifiers. Of the original cohort, 70 % was applied for feature selection and classifier development, and the remaining 30 % used as an independent validation set. The diagnostic performance was evaluated using area under the ROC curve (AUC), test error, sensitivity, and specificity. Furthermore, the performance of the radiomics signatures against routine features was statistically compared using DeLong's method. RESULTS: The cross-combination fisher score (FSCR) + k-nearest neighborhood (KNN), FSCR + support vector machines with radial basis function kernel (RBF-SVM), FSCR + random forest (RF), and minimum redundancy maximum relevance (MRMR) + RBF-SVM outperformed others in terms of accuracy (AUC 0.883, 0.867, 0.892, 0.883; sensitivity 0.833, 0.864, 0.831, 0.750; specificity 1, 1, 0.873, 1) and reliability (test error 0.091, 0.136, 0.150, 0.136). Compared with conventional metrics, the radiomics signatures showed higher AUC values (0.867-0.892 vs. 0.817), though the differences were not statistically significant (p = 0.462-0.560). CONCLUSION: This study identified the most accurate and reliable machine learning methods, which could enhance the application of radiomics methods in the precision of diagnosis of NPC.

Subregional Radiomics Analysis of PET/CT Imaging with Intratumor Partitioning: Application to Prognosis for Nasopharyngeal Carcinoma.

PURPOSE: This work aims to identify intratumoral habitats with distinct heterogeneity based on 2-deoxy-2-[18F]fluro-D-glucose positron emission tomography (PET)/X-ray computed tomography (CT) imaging, and to develop a subregional radiomics approach to predict progression-free survival (PFS) in patients with nasopharyngeal carcinoma (NPC). PROCEDURES: In total, 128 NPC patients (85 vs. 43 for primary vs. validation cohorts) who underwent pre-treatment PET/CT scan were enrolled retrospectively. Each tumor was partitioned into several phenotypically consistent subregions based on individual- and population-level clustering. For each subregion, 202 radiomics features were extracted to construct imaging biomarker for prognosis via Cox's proportional hazard model combined with forward stepwise feature selection. Relevance of imaging biomarkers and clinicopathological factors were assessed by multivariate Cox regression analysis and Spearman's correlation analysis. To investigate whether imaging biomarkers could provide complementary prognosis information beyond existing predictors, a scoring system was further developed for risk stratification and compared with AJCC staging system. RESULTS: Three subregions (denoted as S1, S2, and S3) were discovered with distinct PET/CT imaging characteristics in the two cohorts. The prognostic performance of imaging biomarker S3 outperformed the whole tumor (C-index, 0.69 vs. 0.58; log-rank test, p < 0.001 vs. p = 0.552). Imaging biomarker S3 and AJCC stage were identified as independent predictors (p = 0.011 and 0.042, respectively) after adjusting for clinicopathological factors. The scoring system outperformed the traditional AJCC staging system (log-rank test, p < 0.0001 vs. p = 0.0002 in primary cohort and p = 0.0021 vs. p = 0.0277 in validation cohort, respectively). CONCLUSIONS: Subregional radiomics analysis of PET/CT imaging has the potential to predict PFS in patients with NPC, which also provides complementary prognostic information for traditional predictors.

Radiomics Analysis of PET and CT Components of PET/CT Imaging Integrated with Clinical Parameters: Application to Prognosis for Nasopharyngeal Carcinoma.

PURPOSE: To investigate the prognostic performance of radiomics features, as extracted from positron emission tomography (PET) and X-ray computed tomography (CT) components of baseline 2-deoxy-2-[18F]fluoro-D-glucose ([18F]FDG) PET/CT images and integrated with clinical parameters, in patients with nasopharyngeal carcinoma (NPC). PROCEDURES: One hundred twenty-eight NPC patients (85 vs. 43 for training vs. validation), containing a subset of 86 patients with local-regional advanced stage, were enrolled. All patients underwent pretreatment PET/CT scans (mean follow-up time 24 ± 14 months). Three thousand two hundred seventy-six radiomics features extracted from PET or CT components and 13 clinical parameters were used to predict progression-free survival (PFS). Univariate analysis with Benjamini-Hochberg false discovery rate (FDR) correction was first used to screen significant features, and redundant features with Spearman's correlation > 0.8 were further eliminated. Then, seven multivariate models involving PET features and/or CT features and/or clinical parameters (denoted as clinical, PET, CT, clinical + PET, clinical + CT, PET + CT and clinical + PET + CT) were constructed by forward stepwise multivariate Cox regression. Model performance was evaluated by concordance index (C-index). RESULTS: Sixty patients encountered events (28 recurrences, 17 metastases, and 15 deaths). Six clinical parameters, 3 PET features, and 14 CT features in training cohort and 4 clinical parameters, 10 PET features, and 4 CT features in subset of local-regional advanced stage were significantly associated with PFS. Combining PET and/or CT features with clinical parameters showed equal or higher prognostic performance than models with PET or CT or clinical parameters alone (C-index 0.71-0.76 vs. 0.67-0.73 and 0.62-0.75 vs. 0.54-0.75 for training and validation cohorts, respectively), while the prognostic performance was significantly improved in local-regional advanced cohort (C-index 0.67-0.84 vs. 0.64-0.77, p value 0.001-0.059). CONCLUSION: Radiomics features extracted from the PET and CT components of baseline PET/CT images provide complementary prognostic information and improved outcome prediction for NPC patients compared with use of clinical parameters alone.

Robustness versus disease differentiation when varying parameter settings in radiomics features: application to nasopharyngeal PET/CT.

OBJECTIVES: To investigate the impact of parameter settings as used for the generation of radiomics features on their robustness and disease differentiation (nasopharyngeal carcinoma (NPC) versus chronic nasopharyngitis (CN) in FDG PET/CT imaging). METHODS: We studied 106 patients (69/37 NPC/CN, pathology confirmed), and extracted 57 radiomics features under different parameter settings. Robustness was assessed by the intra-class correlation coefficient (ICC). Logistic regression with leave-one-out cross validation was used to generate classification probabilities, and diagnostic performance was assessed by the area under the receiver operating characteristic curve (AUC). RESULTS: Varying averaging strategies and symmetry, 4/26 GLCM features showed poor range of pairwise ICCs of 0.02-0.98, while depicting good AUCs of 0.82-0.91. Varying distances, 5/26 GLCM features showed ICCs of 0.82-0.99 while corresponding AUCs were 0.52-0.91. 6/13 GLRLM features showed both high AUC (0.81-0.89) and high ICC (0.85-0.99) regarding to averaging strategies. 7/13 GLSZM features showed AUCs of 0.81-0.90 while having ICCs of 0.01-0.99 under different neighbourhoods. 2/5 NGTDM features showed AUCs of 0.81-0.85 while having ICCs of 0.19-0.89 for different window sizes. Differentiating a subset of NPC (stages I-II) form CN, both SumEntropy and SZLGE achieved significantly higher AUCs than metabolically active tumour volume (AUC: 0.91 vs. 0.72, p<0.01). CONCLUSIONS: Radiomics features depicting poor absolute-scale robustness regarding to parameter settings can still lead to good diagnostic performance. As such, robustness of radiomics features should not be overemphasized for removal of features towards assessment of clinical tasks. For differentiating NPC from CN, some radiomics features (e.g. SumEntropy, SZLGE, LGZE) outperformed conventional metrics. KEY POINTS: • Poor robustness did not necessarily translate into poor differentiation performance. • Absolute-scale robustness of radiomics features should not be overemphasized. • Radiomics features SumEntropy, SZLGE and LGZE outperformed conventional metrics.