L. Shen

Validation of a rectal cancer outcome prediction model with a cohort of Chinese patients

The risk of local recurrence (LR), distant metastases (DM) and overall survival (OS) of locally advanced rectal cancer after preoperative chemoradiation can be estimated by prediction models and visualized using nomograms, which have been trained and validated in European clinical trial populations. Data of 277 consecutive locally advanced rectal adenocarcinoma patients treated with preoperative chemoradiation and surgery from Shanghai Cancer Center, were retrospectively collected and used for external validation. Concordance index (C-index) and calibration curves were used to assess the performance of the previously developed prediction models in this routine clinical validation population. The C-index for the published prediction models was 0.72 ± 0.079, 0.75 ± 0.043 and 0.72 ± 0.089 in predicting 2-year LR, DM and OS in the Chinese population, respectively. Kaplan-Meier curves indicated good discriminating performance regarding LR, but could not convincingly discriminate a low-risk and medium-risk group for distant control and OS. Calibration curves showed a trend of underestimation of local and distant control, as well as OS in the observed data compared with the estimates predicted by the model. In conclusion, we externally validated three models for predicting 2-year LR, DM and OS of locally advanced rectal cancer patients who underwent preoperative chemoradiation and curative surgery with good discrimination in a single Chinese cohort. However, the model overestimated the local control rate compared to observations in the clinical cohort. Validation in other clinical cohorts and optimization of the prediction model, perhaps by including additional prognostic factors, may enhance model validity and its applicability for personalized treatment of locally advanced rectal cancer.

Reproducibility with repeat CT in radiomics study for rectal cancer

Purpose To evaluate the reproducibility of radiomics features by repeating computed tomographic (CT) scans in rectal cancer. To choose stable radiomics features for rectal cancer. Results Volume normalized features are much more reproducible than unnormalized features. The average value of all slices is the most reproducible feature type in rectal cancer. Different filters have little effect for the reproducibility of radiomics features. For the average type features, 496 out of 775 features showed high reproducibility (ICC ≥ 0.8), 225 out of 775 features showed medium reproducibility (0.8 > ICC ≥ 0.5) and 54 out of 775 features showed low reproducibility (ICC < 0.5). Methods 40 rectal cancer patients with stage II were enrolled in this study, each of whom underwent two CT scans within average 8.7 days. 775 radiomics features were defined in this study. For each features, five different values (value from the largest slice, maximum value, minimum value, average value of all slices and value from superposed intermediate matrix) were extracted. Meanwhile a LOG filter with different parameters was applied to these images to find stable filter value. Concordance correlation coefficients (CCC) and inter-class correlation coefficients (ICC) of two CT scans were calculated to assess the reproducibility, based on original features and volume normalized features. Conclusions Features are recommended to be normalized to volume in radiomics analysis. The average type radiomics features are the most stable features in rectal cancer. Further analysis of these features of rectal cancer can be warranted for treatment monitoring and prognosis prediction.

MO-DE-207B-09: A Consistent Test for Radiomics Softwares.

PURPOSEnThe purpose of this study is to investigate the consistency of the features extracted by different radiomics software.nnnMETHODSnCT sets of 212 patients with rectum tumor and three existing radiomics feature extraction tools (IBEX from MD_Anderson, RADIOMICS from Maastro, and a MATLAB based code-set from our institution) were enrolled in this study. Among thousands of features that were extracted by three softwares, 273 (between our codes and RADIOMICS), 51 (between IBEX and RADIOMICS) and 34 (between our codes and IBEX) feature pairs were proven to be conjugated according to feature definition. As for each matched feature pair, a Spearmans rank correlation test was conducted to measure the correlation of the feature pair yielded by different feature extraction tools. Furthermore, each of three datasets were standardized by z-score method and undergone a clustering analysis with NMF, respectively. Finally, the consistent of clustering was verified by a chi-square test.nnnRESULTSnThe consistent between IBEX and RADIOMICS (34 out of 54 features coefficient of correlation were above 0.9) were better than consistent between these two tools and our codes, from FDSCC (9 out of 33 and 9 out of 34 features coefficient of correlation were above 0.9, respectively). One of the causes we expected to be reasonable is the different GLCM definition among the algorithms in three softwares. We used 2D GLCM features, while IBEX used 2.5D and RADIOMICS used 3D, respectively. On the other hand, the most consistent clustering was between datasets yielded by IBEX and RADIOMICS, with an accuracy of 0.87 and a p-value (Chi-sqr test) of 4.5e-27. Meanwhile, the consistence accuracy of clustering between our codesets and other software was about 0.7.nnnCONCLUSIONnThis work indicated the potential inconsistency between different software has little impact on the clustering outcome. Additional attention should be paid when using different software in radiomics research.

SU-F-R-33: Can CT and CBCT Be Used Simultaneously for Radiomics Analysis

PURPOSEnTo investigate whether CBCT and CT can be used in radiomics analysis simultaneously. To establish a batch correction method for radiomics in two similar image modalities.nnnMETHODSnFour sites including rectum, bladder, femoral head and lung were considered as region of interest (ROI) in this study. For each site, 10 treatment planning CT images were collected. And 10 CBCT images which came from same site of same patient were acquired at first radiotherapy fraction. 253 radiomics features, which were selected by our test-retest study at rectum cancer CT (ICC>0.8), were calculated for both CBCT and CT images in MATLAB. Simple scaling (z-score) and nonlinear correction methods were applied to the CBCT radiomics features. The Pearson Correlation Coefficient was calculated to analyze the correlation between radiomics features of CT and CBCT images before and after correction. Cluster analysis of mixed data (for each site, 5 CT and 5 CBCT data are randomly selected) was implemented to validate the feasibility to merge radiomics data from CBCT and CT. The consistency of clustering result and site grouping was verified by a chi-square test for different datasets respectively.nnnRESULTSnFor simple scaling, 234 of the 253 features have correlation coefficient ρ>0.8 among which 154 features haveρ>0.9 . For radiomics data after nonlinear correction, 240 of the 253 features have ρ>0.8 among which 220 features have ρ>0.9. Cluster analysis of mixed data shows that data of four sites was almost precisely separated for simple scaling(p=1.29 * 10-7 , χ2 test) and nonlinear correction (p=5.98 * 10-7 , χ2 test), which is similar to the cluster result of CT data (p=4.52 * 10-8 , χ2 test).nnnCONCLUSIONnRadiomics data from CBCT can be merged with those from CT by simple scaling or nonlinear correction for radiomics analysis.

Can tumor regression grade influence survival outcome in ypT3 rectal cancer

PurposeLocally advanced rectal cancer (LARC) patients achieving ypT3 status following neoadjuvant chemoradiation are considered to have poor response with minimal downstaging. However, residual cancer cell amounts vary in the subserosa/perirectal fat. Tumor regression grading (TRG) is an evaluation method based on the proportion of fibrosis and residual cancer cells. The aim of this study is to assess the influence of TRG in ypT3 rectal cancer patients who received neoadjuvant chemoradiation.MethodsWe retrospectively reviewed 325 LARC patients who received neoadjuvant chemoradiation and surgery. TRG scores were recorded by two independent pathologists. Among these patients, 143 were staged as ypT3. We analyzed TRG and other clinicopathological factors and their relationship with survival outcome including overall survival (OS) and disease-free survival (DFS).ResultsAmong 143 ypT3 patients, 44 (30.8xa0%) were TRG1, 84 (58.7xa0%) were TRG2 and 15 (10.5xa0%) were TRG3. Seventy-nine (55.3xa0%) of these patients had metastatic lymph nodes. In univariate analysis, TRG was not associated with DFS (TRG2 vs TRG1, Pxa0=xa00.852; TRG3 vs TRG1, Pxa0=xa00.593) or OS (TRG2 vs TRG1, Pxa0=xa00.977; TRG3 vs TRG1, Pxa0=xa00.665). Palliative surgery (HR 3.845; 95xa0% CI 1.670–8.857; Pxa0=xa00.002) and metastatic lymph nodes after surgery (HR 5.894; 95xa0% CI 1.142–3.48; Pxa0=xa00.015) were significantly associated with decreased DFS, while palliative surgery was the only factor associated with worse OS (HR 6.011; 95xa0% CI 2.150–16.810; Pxa0=xa00.001). Palliative surgery (HR 3.923; 95xa0% CI 1.696–9.073; Pxa0=xa00.001) and metastatic lymph nodes (HR 2.011; 95xa0% CI 1.152–3.512; Pxa0=xa00.014) also showed prognostic significance for DFS in multivariate analysis.ConclusionsResidual cancer cells evaluated by TRG score after neoadjuvant chemoradiation do not influence survival outcome in ypT3 rectal cancer patients. However, lymph node status is a significant prognostic factor in ypT3 patients.

SU‐D‐207B‐01: Radiomics Feature Reproducibility From Repeat CT Scans of Patients with Rectal Cancer

PURPOSEnTo evaluate the reproducibility of radiomics features by repeating computed tomographic (CT) scans in rectal cancer. To choose stable radiomics features for rectal cancer.nnnMETHODSn40 rectal cancer patients were enrolled in this study, each of whom underwent two CT scans within average 8.7 days (5 days to 17 days), before any treatment was delivered. The rectal gross tumor volume (GTV) was distinguished and segmented by an experienced oncologist in both CTs. Totally, more than 2000 radiomics features were defined in this study, which were divided into four groups (I: GLCM, II: GLRLM III: Wavelet GLCM and IV: Wavelet GLRLM). For each group, five types of features were extracted (Max slice: features from the largest slice of target images, Max value: features from all slices of target images and choose the maximum value, Min value: minimum value of features for all slices, Average value: average value of features for all slices, Matrix sum: all slices of target images translate into GLCM and GLRLM matrices and superpose all matrices, then extract features from the superposed matrix). Meanwhile a LOG (Laplace of Gauss) filter with different parameters was applied to these images. Concordance correlation coefficients (CCC) and inter-class correlation coefficients (ICC) were calculated to assess the reproducibility.nnnRESULTSn403 radiomics features were extracted from each type of patients medical images. Features of average type are the most reproducible. Different filters have little effect for radiomics features. For the average type features, 253 out of 403 features (62.8%) showed high reproducibility (ICC≥0.8), 133 out of 403 features (33.0%) showed medium reproducibility (0.8≥ICC≥0.5) and 17 out of 403 features (4.2%) showed low reproducibility (ICC≥0.5).nnnCONCLUSIONnThe average type radiomics features are the most stable features in rectal cancer. Further analysis of these features of rectal cancer can be warranted for treatment monitoring and prognosis prediction.

SU-E-J-256: Predicting Metastasis-Free Survival of Rectal Cancer Patients Treated with Neoadjuvant Chemo-Radiotherapy by Data-Mining of CT Texture Features of Primary Lesions

Purpose: The purpose of this study is to investigate the relationship between computed tomographic (CT) texture features of primary lesions and metastasis-free survival for rectal cancer patients; and to develop a datamining prediction model using texture features. Methods: A total of 220 rectal cancer patients treated with neoadjuvant chemo-radiotherapy (CRT) were enrolled in this study. All patients underwent CT scans before CRT. The primary lesions on the CT images were delineated by two experienced oncologists. The CT images were filtered by Laplacian of Gaussian (LoG) filters with different filter values (1.0–2.5: from fine to coarse). Both filtered and unfiltered images were analyzed using Gray-level Co-occurrence Matrix (GLCM) texture analysis with different directions (transversal, sagittal, and coronal). Totally, 270 texture features with different species, directions and filter values were extracted. Texture features were examined with Student’s t-test for selecting predictive features. Principal Component Analysis (PCA) was performed upon the selected features to reduce the feature collinearity. Artificial neural network (ANN) and logistic regression were applied to establish metastasis prediction models. Results: Forty-six of 220 patients developed metastasis with a follow-up time of more than 2 years. Sixtyseven texture features were significantly different in t-test (p<0.05) between patients with and without metastasis,morexa0» and 12 of them were extremely significant (p<0.001). The Area-under-the-curve (AUC) of ANN was 0.72, and the concordance index (CI) of logistic regression was 0.71. The predictability of ANN was slightly better than logistic regression. Conclusion: CT texture features of primary lesions are related to metastasisfree survival of rectal cancer patients. Both ANN and logistic regression based models can be developed for prediction.«xa0less