Utility of texture analysis on T2-weighted MR for differentiating tumor deposits from mesorectal nodes in rectal cancer patients, in a retrospective cohort

Predicting tumor deposits in rectal cancer: a combined deep learning model using T2-MR imaging and clinical features

BACKGROUND

Tumor deposits (TDs) are associated with poor prognosis in rectal cancer (RC). This study aims to develop and validate a deep learning (DL) model incorporating T2-MR image and clinical factors for the preoperative prediction of TDs in RC patients.

METHODS AND METHODS

A total of 327 RC patients with pathologically confirmed TDs status from January 2016 to December 2019 were retrospectively recruited, and the T2-MR images and clinical variables were collected. Patients were randomly split into a development dataset (n = 246) and an independent testing dataset (n = 81). A single-channel DL model, a multi-channel DL model, a hybrid DL model, and a clinical model were constructed. The performance of these predictive models was assessed by using receiver operating characteristics (ROC) analysis and decision curve analysis (DCA).

RESULTS

The areas under the curves (AUCs) of the clinical, single-DL, multi-DL, and hybrid-DL models were 0.734 (95% CI, 0.674-0.788), 0.710 (95% CI, 0.649-0.766), 0.767 (95% CI, 0.710-0.819), and 0.857 (95% CI, 0.807-0.898) in the development dataset. The AUC of the hybrid-DL model was significantly higher than the single-DL and multi-DL models (both p < 0.001) in the development dataset, and the single-DL model (p = 0.028) in the testing dataset. Decision curve analysis demonstrated the hybrid-DL model had higher net benefit than other models across the majority range of threshold probabilities.

CONCLUSIONS

The proposed hybrid-DL model achieved good predictive efficacy and could be used to predict tumor deposits in rectal cancer.

CRITICAL RELEVANCE STATEMENT

The proposed hybrid-DL model achieved good predictive efficacy and could be used to predict tumor deposits in rectal cancer.

KEY POINTS

• Preoperative non-invasive identification of TDs is of great clinical significance. • The combined hybrid-DL model achieved good predictive efficacy and could be used to predict tumor deposits in rectal cancer. • A preoperative nomogram provides gastroenterologist with an accurate and effective tool.

Multiregional-based magnetic resonance imaging radiomics model for predicting tumor deposits in resectable rectal cancer

PURPOSE

To establish and validate an integrated model incorporating multiregional magnetic resonance imaging (MRI) radiomics features and clinical factors to predict tumor deposits (TDs) preoperatively in resectable rectal cancer (RC).

METHODS

This study retrospectively included 148 resectable RC patients [TDs+ (n = 45); TDs- (n = 103)] from August 2016 to August 2022, who were divided randomly into a testing cohort (n = 45) and a training cohort (n = 103). Radiomics features were extracted from the volume of interest on T2-weighted images (T2WI) and diffusion-weighted images (DWI) from pretreatment MRI. Model construction was performed after feature selection. Finally, five classification models were developed by support vector machine (SVM) algorithm to predict TDs in resectable RC using the selected clinical factor, single-regional radiomics features (extracted from primary tumor), and multiregional radiomics features (extracted from the primary tumor and mesorectal fat). Receiver-operating characteristic (ROC) curve analysis was employed to assess the discrimination performance of the five models. The AUCs of five models were compared by DeLon's test.

RESULTS

The training and testing cohorts included 31 (30.1%) and 14 (31.1%) patients with TDs, respectively. The AUCs of multiregional radiomics, single-regional radiomics, and the clinical models for predicting TDs were 0.839, 0.765, and 0.793, respectively. An integrated model incorporating multiregional radiomics features and clinical factors showed good predictive performance for predicting TDs in resectable RC (AUC, 0.931; 95% CI, 0.841-0.988), which demonstrated superiority over clinical model (P = 0.016), the single-regional radiomics model (P = 0.042), and the multiregional radiomics model (P = 0.025).

CONCLUSION

An integrated model combining multiregional MRI radiomic features and clinical factors can improve prediction performance for TDs and guide clinicians in implementing treatment plans individually for resectable RC patients.

A nomogram for preoperative differentiation of tumor deposits from lymph node metastasis in rectal cancer: A retrospective study

The objective is to develop and validate a combined model for noninvasive preoperative differentiating tumor deposits (TDs) from lymph node metastasis (LNM) in patients with rectal cancer (RC). A total of 204 patients were enrolled and randomly divided into 2 sets (training and validation set) at a ratio of 8:2. Radiomics features of tumor and peritumor fat were extracted by using Pyradiomics software from the axial T2-weighted imaging of MRI. Rad-score based on extracted Radiomics features were calculated by combination of feature selection and the machine learning method. Factors (Rad-score, laboratory test factor, clinical factor, traditional characters of tumor on MRI) with statistical significance were integrated to build a combined model. The combined model was visualized by a nomogram, and its distinguish ability, diagnostic accuracy, and clinical utility were evaluated by the receiver operating characteristic curve (ROC) analysis, calibration curve, and clinical decision curve, respectively. Carbohydrate antigen (CA) 19-9, MRI reported node stage (MRI-N stage), tumor volume (cm3), and Rad-score were all included in the combined model (odds ratio = 3.881 for Rad-score, 2.859 for CA19-9, 0.411 for MRI-N stage, and 1.055 for tumor volume). The distinguish ability of the combined model in the training and validation cohorts was area under the summary receiver operating characteristic curve (AUC) = 0.863, 95% confidence interval (CI): 0.8-0.911 and 0.815, 95% CI: 0.663-0.919, respectively. And the combined model outperformed the clinical model in both training and validation cohorts (AUC = 0.863 vs 0.749, 0.815 vs 0.627, P = .0022, .0302), outperformed the Rad-score model only in training cohorts (AUC = 0.863 vs 0.819, P = .0283). The combined model had highest net benefit and showed good diagnostic accuracy. The combined model incorporating Rad-score and clinical factors could provide a preoperative differentiation of TD from LNM and guide clinicians in making individualized treatment strategy for patients with RC.

Extramural Venous Invasion and Tumor Deposit at Diffusion-weighted MRI in Patients after Neoadjuvant Treatment for Rectal Cancer

Background Diffusion-weighted (DW) imaging is useful in detecting tumor in the primary tumor bed in locally advanced rectal cancer (LARC) after neoadjuvant therapy, but its value in detecting extramural venous invasion (EMVI) and tumor deposit is not well validated. Purpose To evaluate diagnostic accuracy and association with patient prognosis of viable EMVI and tumor deposit on DW images in patients with LARC after neoadjuvant therapy using whole-mount pathology specimens. Materials and Methods This retrospective study included patients who underwent neoadjuvant therapy and surgery from 2018 to 2021. Innovative five-point Likert scale was used by two radiologists to independently evaluate the likelihood of viable EMVI and tumor deposit on restaging DW MRI scans in four axial quadrants (12 to 3 o'clock, 3 to 6 o'clock, 6 to 9 o'clock, and 9 to 12 o'clock). Diagnostic accuracy was assessed at both the per-quadrant and per-patient level, with whole-mount pathology as the reference standard. Weighted κ values for interreader agreement and Cox regression models for disease-free survival and overall survival analyses were used. Results A total of 117 patients (mean age, 56 years ± 12 [SD]; 70 male, 47 female) were included. Pathologically proven viable EMVI and tumor deposit was detected in 29 of 117 patients (25%) and in 44 of 468 quadrants (9.4%). Per-quadrant analyses showed an area under the receiver operating characteristics curve of 0.75 (95% CI: 0.68, 0.83), with sensitivity and specificity of 55% and 96%, respectively. Good interreader agreement was observed between the radiologists (κ = 0.62). Per-patient analysis showed sensitivity and specificity of 62% and 93%, respectively. The presence of EMVI and tumor deposit on restaging DW MRI scans was associated with worse disease-free survival (hazard ratio [HR], 5.6; 95% CI: 2.4, 13.3) and overall survival (HR, 8.9; 95% CI: 1.6, 48.5). Conclusion DW imaging using the five-point Likert scale showed high specificity and moderate sensitivity in the detection of viable extramural venous invasion and tumor deposits in LARC after neoadjuvant therapy, and its presence on restaging DW MRI scans is associated with worse prognosis. Published under a CC BY 4.0 license. Supplemental material is available for this article. See also the editorial by Méndez and Ayuso in this issue.

Distinguishing mesorectal tumor deposits from metastatic lymph nodes by using diffusion-weighted and dynamic contrast-enhanced magnetic resonance imaging in rectal cancer

OBJECTIVES

This study aimed to identify whether apparent diffusion coefficient (ADC) values and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters are helpful in distinguishing mesorectal tumor deposits (TD) from metastatic lymph nodes (MLN) in rectal cancer (RC).

METHODS

Thirty patients (59 lesions, including 30 TD and 29 MLN) with RC who underwent pretreatment-MRI between February 2016 and August 2018 were enrolled. The morphological features, ADC values, and semi-quantitative parameters of DCE-MRI, including relative enhancement (RE), maximum enhancement (ME), maximum relative enhancement (MRE), time to peak (TTP), wash-in rates (WIR), wash-out rates (WOR), brevity of enhancement (BRE), and area under the curve (AUC) were measured on lesions (TD or MLN) and RC. The parameters were compared between TD and MLN, tumor with and without TD group by using Fisher's exact test, independent-samples t-test, and Mann-Whitney U test. The ratio (lesion-to-tumor) of the parameters was compared between TD and MLN. Receiver operating characteristic curve analysis and binary logistic regression analysis were used to assess the diagnostic ability of single and combined metrics for distinguishing TD from MLN.

RESULTS

The morphological features, including size, shape, and border, were significantly different between TD and MLN. TD exhibited significantly lower RE, MRE, RE-ratio, MRE-ratio, ADC_min-ratio, and ADC_mean-ratio than MLN. RE-ratio showed the highest AUC (0.749) and accuracy (77.97%) among single parameters. The combination of DCE-MRI and DWI parameters together showed higher diagnostic efficiency (AUC = 0.825).

CONCLUSIONS

Morphological features, ADC values, and DCE-MRI parameters can preoperatively help distinguish TD from MLN in RC.

KEY POINTS

• DWI and DCE-MRI can facilitate early detection and distinguishing mesorectal TD (tumor deposits) from MLN (metastatic lymph nodes) in rectal cancer preoperatively. • TD has some specific morphological features, including relatively larger size, lower short- to long-axis ratio, irregular shape, and ill-defined border on T2-weighted MR images in rectal cancer. • The combination of ADC values and semi-quantitative parameters of DCE-MRI (RE, MRE) can help to improve the diagnostic efficiency of TD in rectal cancer.

Radiomics for differentiating tumor deposits from lymph node metastasis in rectal cancer

BACKGROUND

Tumor deposits (TDs) are not equivalent to lymph node (LN) metastasis (LNM) but have become independent adverse prognostic factors in patients with rectal cancer (RC). Although preoperatively differentiating TDs and LNMs is helpful in designing individualized treatment strategies and achieving improved prognoses, it is a challenging task.

AIM

To establish a computed tomography (CT)-based radiomics model for preoperatively differentiating TDs from LNM in patients with RC.

METHODS

This study retrospectively enrolled 219 patients with RC [TDs⁺LNM^- (n = 89); LNM⁺ TDs^- (n = 115); TDs⁺LNM⁺ (n = 15)] from a single center between September 2016 and September 2021. Single-positive patients (i.e., TDs⁺LNM^- and LNM⁺TDs^-) were classified into the training (n = 163) and validation (n = 41) sets. We extracted numerous features from the enhanced CT (region 1: The main tumor; region 2: The largest peritumoral nodule). After deleting redundant features, three feature selection methods and three machine learning methods were used to select the best-performing classifier as the radiomics model (Rad-score). After validating Rad-score, its performance was further evaluated in the field of diagnosing double-positive patients (i.e., TDs⁺LNM⁺) by outlining all peritumoral nodules with diameter (short-axis) > 3 mm.

RESULTS

Rad-score 1 (radiomics signature of the main tumor) had an area under the curve (AUC) of 0.768 on the training dataset and 0.700 on the validation dataset. Rad-score 2 (radiomics signature of the largest peritumoral nodule) had a higher AUC (training set: 0.940; validation set: 0.918) than Rad-score 1. Clinical factors, including age, gender, location of RC, tumor markers, and radiological features of the largest peritumoral nodule, were excluded by logistic regression. Thus, the combined model was comprised of Rad-scores of 1 and 2. Considering that the combined model had similar AUCs with Rad-score 2 (P = 0.134 in the training set and 0.594 in the validation set), Rad-score 2 was used as the final model. For the diagnosis of double-positive patients in the mixed group [TDs⁺LNM⁺ (n = 15); single-positive (n = 15)], Rad-score 2 demonstrated moderate performance (sensitivity, 73.3%; specificity, 66.6%; and accuracy, 70.0%).

CONCLUSION

Radiomics analysis based on the largest peritumoral nodule can be helpful in preoperatively differentiating between TDs and LNM.

Predictive and Prognostic Assessment Models for Tumor Deposit in Colorectal Cancer Patients With No Distant Metastasis

Background

More and more evidence indicated that tumor deposit (TD) was significantly associated with local recurrence, distant metastasis (DM), and poor prognosis for patients with colorectal cancer (CRC). This study aims to explore the main clinical risk factors for the presence of TD in CRC patients with no DM (CRC-NDM) and the prognostic factors for TD-positive patients after surgery.

Methods

The data of patients with CRC-NDM between 2010 and 2017 were extracted from the Surveillance, Epidemiology, and End Results (SEER) database. A logistic regression model was used to identify risk factors for TD presence. Fine and Gray’s competing-risk model was performed to analyze prognostic factors for TD-positive CRC-NDM patients. A predictive nomogram was constructed using the multivariate logistic regression model. The concordance index (C-index), the area under the receiver operating characteristic (ROC) curve (AUC), and the calibration were used to evaluate the predictive nomogram. Also, a prognostic nomogram was built based on multivariate competing-risk regression. C-index, the calibration, and decision-curve analysis (DCA) were performed to validate the prognostic model.

Results

The predictive nomogram to predict the presence of TD had a C-index of 0.785 and AUC of 0.787 and 0.782 in the training and validation sets, respectively. From the competing-risk analysis, chemotherapy (subdistribution hazard ratio (SHR) = 0.542, p < 0.001) can significantly reduce CRC-specific death (CCSD). The prognostic nomogram for the outcome prediction in postoperative CRC-NDM patients with TD had a C-index of 0.727. The 5-year survival of CCSD was 17.16%, 36.20%, and 63.19% in low-, medium-, and high-risk subgroups, respectively (Gray’s test, p < 0.001).

Conclusions

We constructed an easily predictive nomogram in identifying the high-risk TD-positive CRC-NDM patients. Besides, a prognostic nomogram was built to help clinicians identify poor-outcome individuals in postoperative CRC-NDM patients with TD. For the high-risk or medium-risk subgroup, additional chemotherapy may be more advantageous for the TD-positive patients rather than radiotherapy.

A Six-microRNA Signature Nomogram for Preoperative Prediction of Tumor Deposits in Colorectal Cancer

Purpose

Tumor deposits (TDs) are acknowledged negative prognostic factors in colorectal cancer (CRC), and their pathogenesis remains a puzzle. This study aimed to construct and validate a nomogram available for preoperative TDs prediction in CRC patients.

Patients and Methods

Patients from the Surveillance, Epidemiology, and End Results (SEER) and the cancer genome atlas (TCGA) databases were randomly divided into training and validation sets according to the sample size ratio of 7:3. Univariate logistic regression was performed for identifying differentially expressed microRNAs between TDs and non-TDs. Nomograms for TDs prediction were developed from the multivariate logistic regression model with least absolute shrinkage and selection operator and were validated internally in terms of accuracy, calibration, and clinical utility. Based on the target genes, pathways tightly associated with TDs were selected using enrichment analysis.

Results

Six clinicopathologic factors and expressions of six microRNAs (miR-614, miR-1197, miR-4770, miR-3136, miR-3173, and miR-4636) differed significantly between TDs and non-TDs CRC patients from the SEER and TCGA training sets. We compared potential prediction discrimination between two nomograms: a clinicopathologic nomogram and a six-microRNA signature nomogram. The six-microRNA signature nomogram revealed better accuracy than the clinicopathologic one for TDs prediction (AUC values of 0.96 and 0.93 in the validation cohort). The calibration plots and decision curve analysis demonstrated that the six-microRNA signature nomogram had better validity and a greater prognostic benefit versus the clinicopathologic one for TDs prediction. Calcium signaling pathways were closely associated with roles of the six microRNAs in TDs of CRC patients.

Conclusion

The six-microRNA signature nomogram can be used as an efficient tool for preoperative TDs prediction in CRC patients.