Incremental value of PET primary lesion-based radiomics signature to conventional metabolic parameters and traditional risk factors for preoperative prediction of lymph node metastases in gastric cancer.

BACKGROUND: Precise preoperative prediction of lymph node metastasis (LNM) is crucial for optimal diagnosis and treatment in patients with gastric cancer (GC), in which existing imaging methods have certain limitations. We hypothesized that PET primary lesion-based radiomics signature could provide incremental value to conventional metabolic parameters and traditional risk indicators in predicting LNM in patients with GC. METHODS: This retrospective study was performed in 127 patients with GC who underwent preoperative PET/CT. Basic clinical data and PET conventional metabolic parameters were collected. Radiomics signature was constructed by the least absolute shrinkage and selection operator algorithm (LASSO) logistic regression. Based on the postoperative histological results, the patients were divided into LNM group and non-lymph node metastasis (NLNM) group. Receiver-operating characteristic (ROC) was used to evaluate the discriminatory ability of Radiomics score (Rad-score) for predicting LNM and determine whether adding Rad-score to PET conventional metabolic parameters and traditional risk factors could improve the predictive value in LNM. The Integrated discrimination improvement (IDI) and net reclassification improvement (NRI) were calculated to further confirm the incremental value of Rad-score for predicting LNM in GC. RESULTS: The LNM group had higher Rad-score than NLNM group [(0.35 (-0.13-0.85) vs. -0.61 (-1.92-0.18), P < 0 .001)]. After adjusted for gender, age, BMI, and FBG, multivariable logistic regression analysis illustrated that Rad-score (OR: 6.38, 95% CI: 2.73-14.91, P < 0.0001) was independent risk factors for LNM in GC. Adding PET conventional parameters to traditional risk factors increased the predictive value of LNM in GC (AUC 0.751 vs 0.651, P = 0.02). Additional inclusion of Rad-score to conventional metabolic parameters and traditional risk indicators significantly improved the AUC (0.882 vs 0.751; P = 0.006). Bootstrap resampling (times = 500) was used for internal verification, 95% confidence interval (CI) was 0.802-0.948, with the sensitivity equaled to 89.5%, and positive predictive value (PPV) was 93.5%. When Rad-score was added to conventional metabolic parameters and traditional risk indicators, net reclassification improvement (NRI) was 0.293 (P = 0.0040) and integrated discrimination improvement (IDI) was 0.293 (P = 0.0045). CONCLUSION: In GC patients, PET Radiomics signature of the primary lesion-based was significantly associated with LNM and could improve the prediction of LNM above PET conventional metabolic parameters and traditional risk factors, which could provide incremental value for individual diagnosis and treatment of GC.

18F-FDG PET/CT-based radiomics nomogram for the preoperative prediction of lymph node metastasis in gastric cancer.

Objective: Lymph node metastasis (LNM) is not only one of the important factors affecting the prognosis of gastric cancer but also an important basis for treatment decisions. The purpose of this study was to investigate the value of the radiomics nomogram based on preoperative 18F-deoxyglucose (FDG) PET/CT primary lesions and clinical risk factors for predicting LNM in gastric cancer (GC). Methods: We retrospectively analyzed radiomics features of preoperative 18F-FDG PET/CT images in 224 gastric cancer patients from two centers. The prediction model was developed in the training cohort (n = 134) and validated in the internal (n = 59) and external validation cohorts (n = 31). The least absolute shrinkage and selection operator (LASSO) regression was used to select features and build radiomics signatures. The radiomics feature score (Rad-score) was calculated and established a radiomics signature. Multivariate logistic regression analysis was used to screen independent risk factors for LNM. The minimum Akaike's information criterion (AIC) was used to select the optimal model parameters to construct a radiomics nomogram. The performance of the nomogram was assessed with calibration, discrimination, and clinical usefulness. Results: There was no significant difference between the internal verification and external verification of the clinical data of patients (all p > 0.05). The areas under the curve (AUCs) (95% CI) for predicting LNM based on the 18F-FDG PET/CT radiomics signature in the training cohort, internal validation cohort, and external validation cohort were 0.792 (95% CI: 0.712-0.870), 0.803 (95% CI: 0.681-0.924), and 0.762 (95% CI: 0.579-0.945), respectively. Multivariate logistic regression showed that carbohydrate antigen (CA) 19-9 [OR (95% CI): 10.180 (1.267-81.831)], PET/CT diagnosis of LNM [OR (95% CI): 6.370 (2.256-17.984)], PET/CT Rad-score [OR (95% CI): 16.536 (5.506-49.660)] were independent influencing factors of LNM (all p < 0.05), and a radiomics nomogram was established based on those factors. The AUCs (95% CI) for predicting LNM were 0.861 (95% CI: 0.799-0.924), 0.889 (95% CI: 0.800-0.976), and 0.897 (95% CI: 0.683-0.948) in the training cohort, the internal validation cohort, and the external validation cohort, respectively. Decision curve analysis (DCA) indicated that the 18F-FDG PET/CT-based radiomics nomogram has good clinical utility. Conclusions: Radiomics nomogram based on the primary tumor of 18F-FDG PET/CT could facilitate the preoperative individualized prediction of LNM, which is helpful for risk stratification in GC patients.

Radiomics model based on preoperative 18F-fluorodeoxyglucose PET predicts N2-3b lymph node metastasis in gastric cancer patients.

OBJECTIVE: The aim of the study was to construct and validate 18F-fluorodeoxyglucose (18F-FDG) PET-based radiomics nomogram and use it to predict N2-3b lymph node metastasis in Chinese patients with gastric cancer (GC). METHODS: A total of 127 patients with pathologically confirmed GC who underwent preoperative 18F-FDG PET/CT imaging between January 2014 and September 2020 were enrolled as subjects in this study. We use the LIFEx software to extract PET radiomic features. A radiomics signature (Rad-score) was developed with the least absolute shrinkage and selection operator algorithm. Then a prediction model, which incorporated the Rad-score and independent clinical risk factors, was constructed and presented with a radiomics nomogram. Receiver operating characteristic (ROC) analysis was used to assess the performance of Rad-score and the nomogram. Finally, decision curve analysis (DCA) was applied to evaluate the clinical usefulness of the nomogram. RESULTS: The PET Rad-score, which includes four selected features, was significantly related to pN2-3b (all P < 0.05). The prediction model, which comprised the Rad-score and carcinoembryonic antigen (CEA) level, showed good calibration and discrimination [area under the ROC curve: 0.81(95% confidence interval: 0.74-0.89), P < 0.001)]. The DCA also indicated that the prediction model was clinically useful. CONCLUSION: This study presents a radiomics nomogram consisting of a radiomics signature based on PET images and CEA level that can be conveniently used for personalized prediction of high-risk N2-3b metastasis in Chinese GC patients.

Relationship between total lesion glycolysis of primary lesions based on 18F-FDG PET/CT and lymph node metastasis in gastric adenocarcinoma: a cross-sectional preliminary study.

OBJECTIVES: We explored the relationship between lymph node metastasis (LNM) and total lesion glycolysis (TLG) of primary lesions determined by 18fluoro-2-deoxyglucose PET/computed tomography (18F-FDG PET/CT) in patients with gastric adenocarcinoma, and evaluated the independent effect of this association. METHODS: This retrospective study included 106 gastric adenocarcinoma patients who were examined by preoperative 18F-FDG PET/CT imaging between April 2016 and April 2020. We measured TLG of primary gastric lesions and evaluated its association with LNM. Multivariate logistic regression and a two-piece-wise linear regression were performed to evaluate the relationship between TLG of primary lesions and LNM. RESULTS: Of the 106 patients, 75 cases (71%) had LNM and 31 cases (29%) did not have LNM. Univariate analyses revealed that a per-SD increase in TLG was independently associated with LNM [odds ratio (OR) = 2.37; 95% confidence interval (CI), 1.42-3.98; P = 0.0010]. After full adjustment of confounding factors, multivariate analyses exhibited that TLG of primary lesions was still significantly associated with LNM (OR per-SD: 2.20; 95% CI, 1.16-4.19; P = 0.0164). Generalized additive model indicated a nonlinear relationship and saturation effect between TLG of primary lesions and LNM. When TLG of primary lesions was <23.2, TLG was significantly correlated with LNM (OR = 1.26; 95% CI, 1.07-1.48; P = 0.0053), whereas when TLG of primary lesions was ≥ 23.2, the probability of LNM was greater than 60%, gradually reached saturation effect, as high as 80% or more. CONCLUSIONS: In this preliminary study, there were saturation and segmentation effects between TLG of primary lesions determined by preoperative 18F-FDG PET/CT and LNM. When TLG of primary lesions was ≥ 23.2, the probability of LNM was greater than 60%, gradually reached saturation effect, as high as 80% or more. TLG of primary lesions is helpful in the preoperative diagnosis of LNM in patients with gastric adenocarcinoma.

Tumor-to-liver standard uptake ratio using fluorine-18 fluorodeoxyglucose positron emission tomography computed tomography effectively predict occult lymph node metastasis of non-small cell lung cancer patients.

OBJECTIVES: We aimed to investigate predictive factors of occult lymph node metastasis and to explore the diagnostic value of various standardized uptake value (SUV) parameters using fluorine-18 fluorodeoxyglucose (F-FDG) positron emission tomography computed tomography (PET/CT) in predicting occult lymph node metastasis of clinical N0 non-small cell lung cancer patients. METHODS: We retrospectively analyzed PET/computed tomography parameters of tumor and clinical data of 124 clinical N0 non-small cell lung cancer patients who underwent both preoperative F-FDG PET/computed tomography and anatomical pulmonary resection with systematic lymph node dissections. The SUVmax, SUVmean, metabolic total volume, and total lesion glycolysis of the primary tumor was automatically measured on the PET/computed tomography workstation. Standardized uptake ratio (SUR) were derived from tumor standardized uptake value divided by blood SUVmean (B-SUR) or liver SUVmean (L-SUR), respectively. RESULTS: According to postoperative pathology, 19 (15%) were diagnosed as occult lymph node metastasis among 124 clinical N0 non-small cell lung cancer patients. On univariate analysis, carcinoembryonic antigen, cytokeratin 19 fragment, lobulation, and all PET parameters were associated with occult lymph node metastasis. The area under the receiver operating characteristic curve, sensitivity, and negative predictive value of L-SURmax were the highest among all PET parameters (0.778, 94.7%, and 98.4%, respectively). On multivariate analysis, carcinoembryonic antigen, cytokeratin 19 fragment, and L-SURmax were independent risk factors for predicting occult lymph node metastasis. Compared to L-SURmax alone and the combination of carcinoembryonic antigen and cytokeratin 19 fragment, the model consisting of three independent risk factors achieved a greater area under the receiver operating characteristic curve (0.901 vs. 0.778 vs. 0.780, P = 0.021 and 0.0141). CONCLUSIONS: L-SURmax showed the most powerful predictive performance than the other PET parameters in predicting occult lymph node metastasis. The combination of three independent risk factors (carcinoembryonic antigen, cytokeratin 19 fragment, and L-SURmax) can effectively predict occult lymph node metastasis in clinical N0 non-small cell lung cancer patients.