Construction and validation of a deep learning prognostic model based on digital pathology images of stage III colorectal cancer.

BACKGROUND: TNM staging is the main reference standard for prognostic prediction of colorectal cancer (CRC), but the prognosis heterogeneity of patients with the same stage is still large. This study aimed to classify the tumor microenvironment of patients with stage III CRC and quantify the classified tumor tissues based on deep learning to explore the prognostic value of the developed tumor risk signature (TRS). METHODS: A tissue classification model was developed to identify nine tissues (adipose, background, debris, lymphocytes, mucus, smooth muscle, normal mucosa, stroma, and tumor) in whole-slide images (WSIs) of stage III CRC patients. This model was used to extract tumor tissues from WSIs of 265 stage III CRC patients from The Cancer Genome Atlas and 70 stage III CRC patients from the Sixth Affiliated Hospital of Sun Yat-sen University. We used three different deep learning models for tumor feature extraction and applied a Cox model to establish the TRS. Survival analysis was conducted to explore the prognostic performance of TRS. RESULTS: The tissue classification model achieved 94.4 % accuracy in identifying nine tissue types. The TRS showed a Harrell's concordance index of 0.736, 0.716, and 0.711 in the internal training, internal validation, and external validation sets. Survival analysis showed that TRS had significant predictive ability (hazard ratio: 3.632, p = 0.03) for prognostic prediction. CONCLUSION: The TRS is an independent and significant prognostic factor for PFS of stage III CRC patients and it contributes to risk stratification of patients with different clinical stages.

Neoadjuvant Chemotherapy With Oxaliplatin and Fluoropyrimidine Versus Upfront Surgery for Locally Advanced Colon Cancer: The Randomized, Phase III OPTICAL Trial.

PURPOSE: The role of neoadjuvant chemotherapy (NAC) in colon cancer remains unclear. This trial investigated whether 3 months of modified infusional fluorouracil, leucovorin, and oxaliplatin (mFOLFOX6) or capecitabine and oxaliplatin (CAPOX) as NAC could improve outcomes in patients with locally advanced colon cancer versus upfront surgery. PATIENTS AND METHODS: OPTICAL was a randomized, phase III trial in patients with clinically staged locally advanced colon cancer (T3 with extramural spread into the mesocolic fat ≥5 mm or T4). Patients were randomly assigned 1:1 to receive six preoperative cycles of mFOLFOX6 or four cycles of CAPOX, followed by surgery and adjuvant chemotherapy (NAC group), or immediate surgery and the physician's choice of adjuvant chemotherapy (upfront surgery group). The primary end point was 3-year disease-free survival (DFS) assessed in the modified intention-to-treat (mITT) population. RESULTS: Between January 2016 and April 2021, of the 752 patients enrolled, 744 patients were included in the mITT analysis (371 in the NAC group; 373 in the upfront surgery group). At a median follow-up of 48.0 months (IQR, 46.0-50.1), 3-year DFS rates were 82.1% in the NAC group and 77.5% in the upfront surgery group (stratified hazard ratio [HR], 0.74 [95% CI, 0.54 to 1.03]). The R0 resection was achieved in 98% of patients who underwent surgery in both groups. Compared with upfront surgery, NAC resulted in a 7% pathologic complete response rate (pCR), significantly lower rates of advanced tumor staging (pT3-4: 77% v 94%), lymph node metastasis (pN1-2: 31% v 46%), and potentially improved overall survival (stratified HR, 0.44 [95% CI, 0.25 to 0.77]). CONCLUSION: NAC with mFOLFOX6 or CAPOX did not show a significant DFS benefit. However, this neoadjuvant approach was safe, resulted in substantial pathologic downstaging, and appears to be a viable therapeutic option for locally advanced colon cancer.

Texture analysis of apparent diffusion coefficient maps: can it identify nonresponse to neoadjuvant chemotherapy for additional radiation therapy in rectal cancer patients?

Background: Neoadjuvant chemotherapy (NCT) alone can achieve comparable treatment outcomes to chemoradiotherapy in locally advanced rectal cancer (LARC) patients. This study aimed to investigate the value of texture analysis (TA) in apparent diffusion coefficient (ADC) maps for identifying non-responders to NCT. Methods: This retrospective study included patients with LARC after NCT, and they were categorized into nonresponse group (pTRG 3) and response group (pTRG 0-2) based on pathological tumor regression grade (pTRG). Predictive texture features were extracted from pre- and post-treatment ADC maps to construct a TA model using RandomForest. The ADC model was developed by manually measuring pre- and post-treatment ADC values and calculating their changes. Simultaneously, subjective evaluations based on magnetic resonance imaging assessment of TRG were performed by two experienced radiologists. Model performance was compared using the area under the curve (AUC) and DeLong test. Results: A total of 299 patients from two centers were divided into three cohorts: the primary cohort (center A; n = 194, with 36 non-responders and 158 responders), the internal validation cohort (center A; n = 49, with 9 non-responders) and external validation cohort (center B; n = 56, with 33 non-responders). The TA model was constructed by post_mean, mean_change, post_skewness, post_entropy, and entropy_change, which outperformed both the ADC model and subjective evaluations with an impressive AUC of 0.997 (95% confidence interval [CI], 0.975-1.000) in the primary cohort. Robust performances were observed in internal and external validation cohorts, with AUCs of 0.919 (95% CI, 0.805-0.978) and 0.938 (95% CI, 0.840-0.985), respectively. Conclusions: The TA model has the potential to serve as an imaging biomarker for identifying nonresponse to NCT in LARC patients, providing a valuable reference for these patients considering additional radiation therapy.

Intravoxel Incoherent Motion Diffusion-weighted Magnetic Resonance Imaging combined with Texture Analysis in Predicting the Histological Grades of Rectal Adenocarcinoma.

PURPOSE: To evaluate the predictive value of 3.0T MRI Intravoxel Incoherent motion diffusion-weighted magnetic resonance imaging (IVIM-DWI) combined with texture analysis (TA) in the histological grade of rectal adenocarcinoma. METHODS: Seventy-one patients with rectal adenocarcinoma confirmed by pathology after surgical resection were collected retrospectively. According to pathology, they were divided into a poorly differentiated group (n=23) and a moderately differentiated group (n=48). The IVIM-DWI parameters and TA characteristics of the two groups were compared, and a prediction model was constructed by multivariate logistic regression analysis. ROC curves were plotted for each individual and combined parameter. RESULTS: There were statistically significant differences in D and D* values between the two groups (P < 0.05). The three texture parameters SmallAreaEmphasis, Median, and Maximum had statistically significant differences between groups (P = 0.01, 0.004, 0.009, respectively). The logistic regression prediction model showed that D*, the median, and the maximum value were significant independent predictors, and the AUC of the regression prediction model was 0.860, which was significantly higher than other single parameters. CONCLUSION: 3.0T MRI IVIM-DWI parameters combined with TA can provide valuable information for predicting the histological grades of rectal adenocarcinoma one week before the operation.

Attention mechanism based multi-sequence MRI fusion improves prediction of response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer.

BACKGROUND: Accurate prediction of response to neoadjuvant chemoradiotherapy (nCRT) is very important for treatment plan decision in locally advanced rectal cancer (LARC). The aim of this study was to investigate whether self-attention mechanism based multi-sequence fusion strategy applied to multiparametric magnetic resonance imaging (MRI) based deep learning or hand-crafted radiomics model construction can improve prediction of response to nCRT in LARC. METHODS: This retrospective analysis enrolled 422 consecutive patients with LARC who received nCRT before surgery at two hospitals. All patients underwent multiparametric MRI scans with three imaging sequences. Tumor regression grade (TRG) was used to assess the response of nCRT based on the resected specimen. Patients were separated into 2 groups: poor responders (TRG 2, 3) versus good responders (TRG 0, 1). A self-attention mechanism, namely channel attention, was applied to fuse the three sequence information for deep learning and radiomics models construction. For comparison, other two models without channel attention were also constructed. All models were developed in the same hospital and validated in the other hospital. RESULTS: The deep learning model with channel attention mechanism achieved area under the curves (AUCs) of 0.898 in the internal validation cohort and 0.873 in the external validation cohort, which was the best performed model in all cohorts. More importantly, both the deep learning and radiomics model that applied channel attention mechanism performed better than those without channel attention mechanism. CONCLUSIONS: The self-attention mechanism based multi-sequence fusion strategy can improve prediction of response to nCRT in LARC.

A Longitudinal MRI-Based Artificial Intelligence System to Predict Pathological Complete Response After Neoadjuvant Therapy in Rectal Cancer: A Multicenter Validation Study.

BACKGROUND: Accurate prediction of response to neoadjuvant chemoradiotherapy is critical for subsequent treatment decisions for patients with locally advanced rectal cancer. OBJECTIVE: To develop and validate a deep learning model based on the comparison of paired MRI before and after neoadjuvant chemoradiotherapy to predict pathological complete response. DESIGN: By capturing the changes from MRI before and after neoadjuvant chemoradiotherapy in 638 patients, we trained a multitask deep learning model for response prediction (DeepRP-RC) that also allowed simultaneous segmentation. Its performance was independently tested in an internal and 3 external validation sets, and its prognostic value was also evaluated. SETTINGS: Multicenter study. PATIENTS: We retrospectively enrolled 1201 patients diagnosed with locally advanced rectal cancer who underwent neoadjuvant chemoradiotherapy before total mesorectal excision. Patients had been treated at 1 of 4 hospitals in China between January 2013 and December 2020. MAIN OUTCOME MEASURES: The main outcome was the accuracy of predicting pathological complete response, measured as the area under receiver operating curve for the training and validation data sets. RESULTS: DeepRP-RC achieved high performance in predicting pathological complete response after neoadjuvant chemoradiotherapy, with area under the curve values of 0.969 (0.942-0.996), 0.946 (0.915-0.977), 0.943 (0.888-0.998), and 0.919 (0.840-0.997) for the internal and 3 external validation sets, respectively. DeepRP-RC performed similarly well in the subgroups defined by receipt of radiotherapy, tumor location, T/N stages before and after neoadjuvant chemoradiotherapy, and age. Compared with experienced radiologists, the model showed substantially higher performance in pathological complete response prediction. The model was also highly accurate in identifying the patients with poor response. Furthermore, the model was significantly associated with disease-free survival independent of clinicopathological variables. LIMITATIONS: This study was limited by its retrospective design and absence of multiethnic data. CONCLUSIONS: DeepRP-RC could be an accurate preoperative tool for pathological complete response prediction in rectal cancer after neoadjuvant chemoradiotherapy. UN SISTEMA DE IA BASADO EN RESONANCIA MAGNTICA LONGITUDINAL PARA PREDECIR LA RESPUESTA PATOLGICA COMPLETA DESPUS DE LA TERAPIA NEOADYUVANTE EN EL CNCER DE RECTO UN ESTUDIO DE VALIDACIN MULTICNTRICO: ANTECEDENTES:La predicción precisa de la respuesta a la quimiorradioterapia neoadyuvante es fundamental para las decisiones de tratamiento posteriores para los pacientes con cáncer de recto localmente avanzado.OBJETIVO:Desarrollar y validar un modelo de aprendizaje profundo basado en la comparación de resonancias magnéticas pareadas antes y después de la quimiorradioterapia neoadyuvante para predecir la respuesta patológica completa.DISEÑO:Al capturar los cambios de las imágenes de resonancia magnética antes y después de la quimiorradioterapia neoadyuvante en 638 pacientes, entrenamos un modelo de aprendizaje profundo multitarea para la predicción de respuesta (DeepRP-RC) que también permitió la segmentación simultánea. Su rendimiento se probó de forma independiente en un conjunto de validación interna y tres externas, y también se evaluó su valor pronóstico.ESCENARIO:Estudio multicéntrico.PACIENTES:Volvimos a incluir retrospectivamente a 1201 pacientes diagnosticados con cáncer de recto localmente avanzado y sometidos a quimiorradioterapia neoadyuvante antes de la escisión total del mesorrecto. Eran de cuatro hospitales en China en el período entre enero de 2013 y diciembre de 2020.PRINCIPALES MEDIDAS DE RESULTADO:Los principales resultados fueron la precisión de la predicción de la respuesta patológica completa, medida como el área bajo la curva operativa del receptor para los conjuntos de datos de entrenamiento y validación.RESULTADOS:DeepRP-RC logró un alto rendimiento en la predicción de la respuesta patológica completa después de la quimiorradioterapia neoadyuvante, con valores de área bajo la curva de 0,969 (0,942-0,996), 0,946 (0,915-0,977), 0,943 (0,888-0,998), y 0,919 (0,840-0,997) para los conjuntos de validación interna y las tres externas, respectivamente. DeepRP-RC se desempeñó de manera similar en los subgrupos definidos por la recepción de radioterapia, la ubicación del tumor, los estadios T/N antes y después de la quimiorradioterapia neoadyuvante y la edad. En comparación con los radiólogos experimentados, el modelo mostró un rendimiento sustancialmente mayor en la predicción de la respuesta patológica completa. El modelo también fue muy preciso en la identificación de los pacientes con mala respuesta. Además, el modelo se asoció significativamente con la supervivencia libre de enfermedad independientemente de las variables clinicopatológicas.LIMITACIONES:Este estudio estuvo limitado por el diseño retrospectivo y la ausencia de datos multiétnicos.CONCLUSIONES:DeepRP-RC podría servir como una herramienta preoperatoria precisa para la predicción de la respuesta patológica completa en el cáncer de recto después de la quimiorradioterapia neoadyuvante. (Traducción-Dr. Felipe Bellolio ).

Development and Validation of a Post-Radiotherapy Prediction Model for Bowel Dysfunction After Rectal Cancer Resection.

BACKGROUND & AIMS: The benefit of radiotherapy for rectal cancer is based largely on a balance between a decrease in local recurrence and an increase in bowel dysfunction. Predicting postoperative disability is helpful for recovery plans and early intervention. We aimed to develop and validate a risk model to improve the prediction of major bowel dysfunction after restorative rectal cancer resection with neoadjuvant radiotherapy using perioperative features. METHODS: Eligible patients more than 1 year after restorative resection following radiotherapy were invited to complete the low anterior resection syndrome (LARS) score at 3 national hospitals in China. Clinical characteristics and imaging parameters were assessed with machine learning algorithms. The post-radiotherapy LARS prediction model (PORTLARS) was constructed by means of logistic regression on the basis of key factors with proportional weighs. The accuracy of the model for major LARS prediction was internally and externally validated. RESULTS: A total of 868 patients reported a mean LARS score of 28.4 after an average time of 4.7 years since surgery. Key predictors for major LARS included the length of distal rectum, anastomotic leakage, proximal colon of neorectum, and pathologic nodal stage. PORTLARS had a favorable area under the curve for predicting major LARS in the internal dataset (0.835; 95% CI, 0.800-0.870, n = 521) and external dataset (0.884; 95% CI, 0.848-0.921, n = 347). The model achieved both sensitivity and specificity >0.83 in the external validation. In addition, PORTLARS outperformed the preoperative LARS score for prediction of major events. CONCLUSIONS: PORTLARS could predict major bowel dysfunction after rectal cancer resection following radiotherapy with high accuracy and robustness. It may serve as a useful tool to identify patients who need additional support for long-term dysfunction in the early stage. CLINICALTRIALS: gov, number NCT05129215.

China special issue on gastrointestinal tumors-Radiological features of pathological complete response in mismatch repair deficient colorectal cancer after neoadjuvant PD-1 blockade: A post hoc analysis of the PICC phase II trial.

Neoadjuvant programmed cell death protein 1 (PD-1) blockade exhibits promising efficacy in patients with mismatch repair deficient (dMMR) colorectal cancer (CRC). However, discrepancies between radiological and histological findings have been reported in the PICC phase II trial (NCT03926338). Therefore, we strived to discern radiological features associated with pathological complete response (pCR) based on computed tomography (CT) images. Data were obtained from the PICC trial that included 36 tumors from 34 locally advanced dMMR CRC patients, who received neoadjuvant PD-1 blockade for 3 months. Among the 36 tumors, 28 (77.8%) tumors achieved pCR. There were no statistically significant differences in tumor longitudinal diameter, the percentage change in tumor longitudinal diameter from baseline, primary tumor sidedness, clinical stage, extramural venous invasion status, intratumoral calcification, peritumoral fat infiltration, intestinal fistula and tumor necrosis between the pCR and non-pCR tumors. Otherwise, tumors with pCR had smaller posttreatment tumor maximum thickness (median: 10 mm vs 13 mm, P = .004) and higher percentage decrease in tumor maximum thickness from baseline (52.9% vs 21.6%, P = .005) compared to non-pCR tumors. Additionally, a higher proportion of the absence of vascular sign (P = .003, odds ratio [OR] = 25.870 [95% CI, 1.357-493.110]), nodular sign (P < .001, OR = 189.000 [95% CI, 10.464-3413.803]) and extramural enhancement sign (P = .003, OR = 21.667 [2.848-164.830]) was observed in tumors with pCR. In conclusion, these CT-defined radiological features may have the potential to serve as valuable tools for clinicians in identifying patients who have achieved pCR after neoadjuvant PD-1 blockade, particularly in individuals who are willing to adopt a watch-and-wait strategy.

DeAF: A multimodal deep learning framework for disease prediction.

Multimodal deep learning models have been applied for disease prediction tasks, but difficulties exist in training due to the conflict between sub-models and fusion modules. To alleviate this issue, we propose a framework for decoupling feature alignment and fusion (DeAF), which separates the multimodal model training into two stages. In the first stage, unsupervised representation learning is conducted, and the modality adaptation (MA) module is used to align the features from various modalities. In the second stage, the self-attention fusion (SAF) module combines the medical image features and clinical data using supervised learning. Moreover, we apply the DeAF framework to predict the postoperative efficacy of CRS for colorectal cancer and whether the MCI patients change to Alzheimer's disease. The DeAF framework achieves a significant improvement in comparison to the previous methods. Furthermore, extensive ablation experiments are conducted to demonstrate the rationality and effectiveness of our framework. In conclusion, our framework enhances the interaction between the local medical image features and clinical data, and derive more discriminative multimodal features for disease prediction. The framework implementation is available at https://github.com/cchencan/DeAF.

CT-based deep learning model for the prediction of DNA mismatch repair deficient colorectal cancer: a diagnostic study.

BACKGROUND: Stratification of DNA mismatch repair (MMR) status in patients with colorectal cancer (CRC) enables individual clinical treatment decision making. The present study aimed to develop and validate a deep learning (DL) model based on the pre-treatment CT images for predicting MMR status in CRC. METHODS: 1812 eligible participants (training cohort: n = 1124; internal validation cohort: n = 482; external validation cohort: n = 206) with CRC were enrolled from two institutions. All pretherapeutic CT images from three dimensions were trained by the ResNet101, then integrated by Gaussian process regression (GPR) to develop a full-automatic DL model for MMR status prediction. The predictive performance of the DL model was evaluated using the area under the receiver operating characteristic curve (AUC) and then tested in the internal and external validation cohorts. Additionally, the participants from institution 1 were sub-grouped by various clinical factors for subgroup analysis, then the predictive performance of the DL model for identifying MMR status between participants in different groups were compared. RESULTS: The full-automatic DL model was established in the training cohort to stratify the MMR status, which presented promising discriminative ability with the AUCs of 0.986 (95% CI 0.971-1.000) in the internal validation cohort and 0.915 (95% CI 0.870-0.960) in the external validation cohort. In addition, the subgroup analysis based on the thickness of CT images, clinical T and N stages, gender, the longest diameter, and the location of tumors revealed that the DL model showed similar satisfying prediction performance. CONCLUSIONS: The DL model may potentially serve as a noninvasive tool to facilitate the pre-treatment individualized prediction of MMR status in patients with CRC, which could promote the personalized clinical-making decision.

Clinical significance of neoadjuvant chemotherapy for locally advanced colorectal cancer patients with deficient mismatch repair: possibly residual value in the era of immunotherapy.

Background: Deficient mismatch repair (dMMR) or microsatellite instability is one of the well-established molecular biomarkers in colorectal cancer (CRC). The efficiency of neoadjuvant chemotherapy (NAC) in locally advanced colorectal cancer (LACC) patients with dMMR is unclear. Objectives: We assessed the tumor response and clinical outcome in LACC patients with dMMR received NAC. Design: Retrospective, single-center analysis. Methods: From 2013 to 2018, a total of 577 LACC patients with dMMR who underwent radical surgery were identified. Among them, 109 patients who received adjuvant chemotherapy were further screened out for analysis. According to whether receiving NAC or not, 109 patients were divided into two groups with the purpose of retrospectively analyzing their characteristics, treatment, and survival results, especially the 5-year disease-free survival (DFS) and 5-year overall survival. Results: Baseline characteristics were matched between the two groups. One of 40 patients in NAC group recurred, while 13 of 69 patients in non-NAC group recurred. Univariate and multivariate analyses showed that NAC (hazard ratio: 0.115; 95% confidence interval: 0.015-0.897; p = 0.039) was independent influence factor for DFS. In NAC group, there were 13/40 (32.5%) patients for tumor regression grade 1 and 27/40 (67.5%) patients converted clinical positive N-stage into negative N-stage. Conclusion: In this study, NAC was associated with better tumor downstaging and longer 5-year DFS in LACC patients with dMMR. Consequently, NAC might be an additional treatment choice when it comes to such patients in the future.

Unsupervised Learning Composite Network to Reduce Training Cost of Deep Learning Model for Colorectal Cancer Diagnosis.

Deep learning facilitates complex medical data analysis and is increasingly being explored in colorectal cancer diagnostics. However, the training cost of the deep learning model limits its real-world medical utility. In this study, we present a composite network that combines deep learning and unsupervised K-means clustering algorithm (RK-net) for automatic processing of medical images. RK-net was more efficient in image refinement compared with manual screening and annotation. The training of a deep learning model for colorectal cancer diagnosis was accelerated by two times with utilization of RK-net-processed images. Better performance was observed in training loss and accuracy achievement as well. RK-net could be useful to refine medical images of the ever-expanding quantity and assist in subsequent construction of the artificial intelligence model.

Locally advanced rectal mucinous adenocarcinoma: is preoperative radiation necessary?

Background: Neoadjuvant chemoradiotherapy is recommended for locally advanced rectal cancer, allowing preoperative down-staging of the primary tumor to facilitate complete surgical removal. However, further investigation is warranted for identifying whether radiotherapy is necessary for rectal mucinous adenocarcinoma (RMAC). Thus, this study was designed to explore the relationship between mFOLFOX6 with or without preoperative radiotherapy and therapeutic efficacy in locally advanced RMAC. Methods: A total of 81 patients were retrospectively enrolled, with MRI-defined clinical stage II/III RMAC received neoadjuvant treatment with mFOLFOX6 alone (group A) or mFOLFOX6 plus radiation (group B), followed by total mesorectal excision. Tumor down-staging and tumor response were assessed based on post-treatment MRI-defined radiographical and pathological findings. Follow-up data were retrieved, and the Kaplan-Meier curve was used to determine the relationship between the 3-year disease-free survival (DFS) and overall survival (OS) in the two groups. Results: There were no significant differences in the clinical baseline characteristics of patients between group A and group B. The sphincter preservation rate in group B was 60.9%, higher than in group A (20.0%) (P=0.031). The rate of pathological complete response (pCR) was 14.0% in group B, while no patients had pCR in group A (P=0.029), and the tumor response rate in group B was higher than in group A (52.0% vs. 16.1%, P=0.001). The 3-year probability of OS in group A and B was 77.4% and 72.0% (P=0.509), and 3-year DFS was 58.1% and 56.0% (P=0.592), respectively. Conclusions: Neoadjuvant mFOLFOX6-based chemoradiotherapy could be a promising therapeutic option for patients with RMAC, which was associated with a high rate of pCR and sphincter preservation in comparison to treated with mFOLFOX6 alone.

Anatomical variations of the branches from left colic artery and middle colic artery at splenic flexure.

BACKGROUND: Variations of the vasculature at splenic flexure by left colic artery (LCA) and middle colic artery (MCA) remain ambiguous. OBJECTIVES: This study aim to investigate the anatomical variations of the branches from LCA and MCA at splenic flexure area. METHODS: Using ultra-thin CT images (0.5-mm thickness), we traced LCA and MCA till their merging site with paracolic marginal arteries through maximum intensity projection (MIP) reconstruction and computed tomography angiography (3D-CTA). RESULTS: A total of 229 cases were retrospectively enrolled. LCA ascending branch approached upwards till the distal third of the transverse colon in 37.6%, reached the splenic flexure in 37.6%, and reached the lower descending colon in 23.1%, and absent in 1.7% of the cases. Areas supplied by MCA left branch and aMCA were 33.2%, 44.5% and 22.3% in the proximal, middle and distal third of transverse colon of the cases, respectively. The accessory MCA separately originated from the superior mesenteric artery was found in 17.9% of the cases. Mutual correlation was found that, when the LCA ascending branch supplied the distal transverse colon, MCA left branch tended to feed the proximal transverse colon; when the LCA ascending branch supplied the lower part of descending colon, MCA left branch was more likely to feed the distal third of transverse colon. CONCLUSIONS: Vasculature at splenic flexure by LCA and MCA varied at specific pattern. This study could add more anatomical details for vessel management in surgeries for left-sided colon cancer.

External rectal prolapse: abdominal or perineal repair for men? A retrospective cohort study.

BACKGROUND: External rectal prolapse is a relatively rare disease, in which male patients account for a minority. The selection of abdominal repair or perineal repair for male patients has rarely been investigated. METHODS: Fifty-one male patients receiving abdominal repair (laparoscopic ventral rectopexy) or perineal repair (Delorme or Altemeier procedures) at the Sixth Affiliated Hospital of Sun Yat-sen University (Guangzhou, China) between March 2013 and September 2019 were retrospectively analysed. We compared the recurrence, complication rate, post-operative defecation disorder, length of stay, and quality of life between the abdominal and perineal groups. RESULTS: Of the 51 patients, 45 had a complete follow-up, with a median of 48.5 months (range, 22.8-101.8 months). A total of 35 patients were under age 40 years. The complication rate associated with abdominal repair was less than that associated with perineal repair (0% vs 20.7%, P = 0.031) and the recurrence rate was also lower (9.5% vs 41.7%, P = 0.018). Multivariate analysis showed that perineal repair (odds ratio, 9.827; 95% confidence interval, 1.296-74.50; P = 0.027) might be a risk factor for recurrence. Moreover, only perineal repair significantly improved post-operative constipation status (preoperative vs post-operative, 72.4% vs 25.0%, P = 0.001). There was no reported mortality in either of the groups. No patient's sexual function was affected by the surgery. CONCLUSIONS: Both surgical approaches were safe in men. Compared with perineal repair, the complication rate and recurrence rate for abdominal repair were lower. However, perineal repair was better able to correct constipation.

Development and Validation of an Image-based Deep Learning Algorithm for Detection of Synchronous Peritoneal Carcinomatosis in Colorectal Cancer.

OBJECTIVE: The aim of this study was to build a SVM classifier using ResNet-3D algorithm by artificial intelligence for prediction of synchronous PC. BACKGROUND: Adequate detection and staging of PC from CRC remain difficult. METHODS: The primary tumors in synchronous PC were delineated on preoperative contrast-enhanced computed tomography (CT) images. The features of adjacent peritoneum were extracted to build a ResNet3D + SVM classifier. The performance of ResNet3D + SVM classifier was evaluated in the test set and was compared to routine CT which was evaluated by radiologists. RESULTS: The training set consisted of 19,814 images from 54 patients with PC and 76 patients without PC. The test set consisted of 7837 images from 40 test patients. The ResNet-3D spent only 34 seconds to analyze the test images. To increase the accuracy of PC detection, we have built a SVM classifier by integrating ResNet-3D features with twelve PC-specific features (P < 0.05). The ResNet3D + SVM classifier showed accuracy of 94.11% with AUC of 0.922 (0.912-0.944), sensitivity of 93.75%, specificity of 94.44%, positive predictive value (PPV) of 93.75%, and negative predictive value (NPV) of 94.44% in the test set. The performance was superior to routine contrast-enhanced CT (AUC: 0.791). CONCLUSIONS: The ResNet3D + SVM classifier based on deep learning algorithm using ResNet-3D framework has shown great potential in prediction of synchronous PC in CRC.

Neoadjuvant PD-1 blockade with toripalimab, with or without celecoxib, in mismatch repair-deficient or microsatellite instability-high, locally advanced, colorectal cancer (PICC): a single-centre, parallel-group, non-comparative, randomised, phase 2 trial.

BACKGROUND: PD-1 blockade is highly effective in patients with mismatch repair-deficient or microsatellite instability-high metastatic colorectal cancer. The role of single-agent PD-1 blockade in the neoadjuvant setting for resectable mismatch repair-deficient or microsatellite instability-high colorectal cancer remains unclear. We investigated the efficacy and safety of PD-1 blockade with toripalimab, with or without the COX-2 inhibitor celecoxib, as neoadjuvant treatment for mismatch repair-deficient or microsatellite instability-high, locally advanced, colorectal cancers. METHODS: The PD-1 Inhibitor in Microsatellite Instability Colorectal Cancer (PICC) trial was a single-centre, open-label, parallel-group, non-comparative, randomised, phase 2 study undertaken at the Sixth Affiliated Hospital of Sun Yat-sen University (Guangzhou, China). Eligible patients were aged 18-75 years, had histologically confirmed mismatch repair-deficient or microsatellite instability-high colorectal cancer, had clinical stage T3-T4 or any T with lymph node positivity (N+), Eastern Cooperative Oncology Group performance score of 0 or 1, and adequate haematological, hepatic, and renal function. Participants were randomly assigned (1:1), without any stratification or balanced blocking, to receive toripalimab 3 mg/kg intravenously on day 1, with or without celecoxib 200 mg orally twice daily from day 1 to 14 of each 14-day cycle, for six cycles before surgical resection. Adjuvant treatment with toripalimab with or without celecoxib was permitted at the investigators' discretion. The primary endpoint was the proportion of patients with pathological complete response, defined as tumours without any viable tumour cells in the resected primary tumour sample and all sampled regional lymph nodes. All efficacy and safety analyses were assessed in the modified intention-to-treat population, which included all patients who were randomly assigned to treatment and who received at least one dose of toripalimab. This trial is registered with ClinicalTrials.gov, NCT03926338, and is ongoing. FINDINGS: Between May 1, 2019, and April 1, 2021, 53 patients were screened, of whom 34 were randomly assigned to either the toripalimab plus celecoxib group (n=17) or the toripalimab monotherapy group (n=17). As of data cutoff (Aug 10, 2021), median follow-up was 14·9 months (IQR 8·8-17·0). All patients received study treatment and underwent surgical resection; there were no treatment-related surgical delays. All 34 patients had an R0 resection (>1 mm resection margin). 15 of 17 patients (88% [95% CI 64-99]) in the toripalimab plus celecoxib group and 11 of 17 patients (65% [38-86]) in the toripalimab monotherapy group had a pathological complete response. All patients continued to receive adjuvant toripalimab with or without celecoxib for a total perioperative duration of 6 months and were alive and free of recurrence at data cutoff. During neoadjuvant treatment, ten (59%) patients in the toripalimab plus celecoxib group and ten (59%) in the toripalimab monotherapy group had grade 1-2 treatment-related adverse events. Only one (3%) of 34 patients, who was in the toripalimab plus celecoxib group, had a grade 3 or higher treatment-related adverse event during the neoadjuvant phase, which was grade 3 increased aspartate aminotransferase levels. In the adjuvant phase, only one (3%) of 34 patients, who was in the toripalimab monotherapy group, had a grade 3 or higher treatment-related adverse events, which was grade 3 increased aspartate aminotransferase and alanine aminotransferase levels. INTERPRETATION: Neoadjuvant toripalimab with or without celecoxib could be a potential therapeutic option for patients with mismatch repair deficient or microsatellite instability-high, locally advanced, colorectal cancer. This treatment was associated with a high pathological complete response rate and an acceptable safety profile, which did not compromise surgery. Longer term follow-up is needed to assess effects on survival-related endpoints. FUNDING: The National Key R&D Program of China, the National Natural Science Foundation of China, and the Chinese Society of Clinical Oncology-Junshi Biosciences Oncology Immunity Research. TRANSLATION: For the Chinese translation of the abstract see Supplementary Materials section.

Correlation Between the Distance to Mesorectal Fascia and Prognosis of cT3 Rectal Cancer: Results of a Multicenter Study From China.

BACKGROUND: The cT3 substage criteria based on extramural depth of tumor invasion in rectal cancer have several limitations. OBJECTIVE: This study proposed that the distance between the deepest tumor invasion and mesorectal fascia on pretherapy MRI can distinguish the prognosis of patients with cT3 rectal cancer. DESIGN: This is a cohort study. SETTING: This study included a prospective, single-center, observational cohort and a retrospective, multicenter, independent validation cohort. PATIENT: Patients who had cT3 rectal cancer with negative mesorectal fascia undergoing neoadjuvant chemoradiotherapy followed by radical surgery were included in 4 centers in China from January 2013 to September 2014. INTERVENTION: Baseline MRI with the distance between the deepest tumor invasion and mesorectal fascia, extramural depth of tumor invasion, and mesorectum thickness were measured. MAIN OUTCOME MEASURES: The cutoff of the distance between the deepest tumor invasion and mesorectal fascia was determined by time-dependent receiver operating characteristic curves, supported by a 5-year progression rate from the prospective cohort, and was then validated in a retrospective cohort. RESULTS: There were 124 and 274 patients included in the prospective and independent validation cohorts. The distance between the deepest tumor invasion and mesorectal fascia was the only predictor for cancer-specific death (HR, 0.1; 95% CI, 0.0-0.7) and was also a significant predictor for distant recurrence (HR, 0.4; 95% CI, 0.2-0.9). No statistically significant difference was observed in prognosis between patients classified as T3a/b and T3c/d. LIMITATIONS: The sample size is relatively small, and the study focused on cT3 rectal cancers with a negative mesorectal fascia. CONCLUSIONS: A cutoff of 7 mm of the distance between the deepest tumor invasion and mesorectal fascia on baseline MRI can distinguish cT3 rectal cancer from a different prognosis. We recommend using the distance between the deepest tumor invasion and mesorectal fascia on baseline MRI for local and systemic risk assessment and providing a tailored schedule of neoadjuvant treatment. See Video Abstract at http://links.lww.com/DCR/B682.CORRELACIÓN ENTRE LA DISTANCIA DE LA FASCIA MESORRECTAL Y EL PRONÓSTICO DEL CÁNCER DE RECTO cT3: RESULTADOS DE UN ESTUDIO MULTICÉNTRICO DE CHINAANTECEDENTES:Los criterios de subestadificación cT3 basados en la profundidad extramural de invasión tumoral en el cáncer de recto tienen varias limitaciones.OBJETIVO:Este estudio propuso que la distancia entre la invasión tumoral más profunda y la fascia mesorrectal en la resonancia magnética preterapia puede distinguir el pronóstico de los pacientes con cT3.DISEÑO:Estudio de cohorte.ENTORNO CLINICO:El estudio incluyó una cohorte observacional, prospectiva, unicéntrica, y una cohorte de validación retrospectiva, multicéntrica e independiente.PACIENTE:Se incluyeron pacientes con cáncer de recto cT3 con fascia mesorrectal negativa sometidos a quimio-radioterapia neoadyuvante seguida de cirugía radical en cuatro centros de China desde enero de 2013 hasta septiembre de 2014.INTERVENCIÓN:Imágenes de resonancia magnética de referencia fueron medidas con la distancia entre la invasión tumoral más profunda y la fascia mesorrectal; la profundidad extramural de la invasión tumoral y el grosor del mesorrecto.PRINCIPALES MEDIDAS DE VALORACION:El límite de la distancia entre la invasión tumoral más profunda y la fascia mesorrectal se determinó mediante curvas características operativas del receptor dependientes del tiempo y se apoyó en la tasa de progresión a 5 años de la cohorte prospectiva, y luego se validó en una cohorte retrospectiva.RESULTADOS:Se incluyeron 124 y 274 pacientes en la cohorte de validación prospectiva e independiente, respectivamente. La distancia entre la invasión tumoral más profunda de la fascia mesorrectal fue el único predictor de muerte específica por cáncer (Hazard ratio: 0.1, 95% CI, 0,0-0,7); y también fue un predictor significativo de recurrencia distante Hazard ratio: 0,4, 95% CI, 0,2-0,9). No se observaron diferencias estadísticamente significativas en el pronóstico entre los pacientes clasificados como T3a/b y T3c/d.LIMITACIONES:El tamaño de la muestra es relativamente pequeño y el estudio se centró en los cánceres de recto cT3 con fascia mesorrectal negativa.CONCLUSIONES:Un límite de 7 mm de distancia entre la invasión tumoral más profunda y la fascia mesorrectal en la resonancia magnética de referencia puede distinguir el cáncer de recto cT3 de diferentes pronósticos. Recomendamos la distancia entre la invasión tumoral más profunda y la fascia mesorrectal en la resonancia magnética de referencia para la evaluación del riesgo local y sistémico, proporcionando un programa personalizado de tratamiento neoadyuvante. Consulte Video Resumen en http://links.lww.com/DCR/B682. (Traducción- Dr. Francisco M. Abarca-Rendon).

Better prognostic determination of cT3 rectal cancer through measurement of distance to mesorectal fascia: A multicenter study.

OBJECTIVE: To forward the magnetic resonance imaging (MRI) based distance between the deepest tumor invasion and mesorectal fascia (DMRF), and to explore its prognosis differentiation value in cT3 stage rectal cancer with comparison of cT3 substage. METHODS: This was a retrospective, multicenter cohort study including cT3 rectal cancer patients undergoing neoadjuvant chemoradiotherapy followed by radical surgery from January 2013 to September 2014. DMRF and cT3 substage were evaluated from baseline MRI. The cutoff of DMRF was determined by disease progression. Multivariate cox regression was used to test the prognostic values of baseline variables. RESULTS: A total of 804 patients were included, of which 226 (28.1%) developed progression. A DMRF cutoff of 7 mm was chosen. DMRF category, the clock position of the deepest position of tumor invasion (CDTI) and extramural venous invasion (EMVI) were independent predictors for disease progression, and hazard ratios (HRs) were 0.26 [95% confidence interval (95% CI), 0.13-0.56], 1.88 (95% CI, 1.33-2.65) and 1.57 (95% CI, 1.13-2.18), respectively. cT3 substage was not a predictor for disease progression. CONCLUSIONS: The measurement of DMRF value on baseline MRI can better distinguish cT3 rectal cancer prognosis rather than cT3 substage, and was recommended in clinical evaluation.

Comparison of outcomes between neoadjuvant imatinib and upfront surgery in patients with localized rectal GIST: An inverse probability of treatment weighting analysis.

BACKGROUND AND OBJECTIVES: This study aimed to compare outcomes between neoadjuvant imatinib and upfront surgery in patients with localized rectal gastrointestinal stromal tumors (GIST) patients. METHODS: Eighty-five patients with localized rectal GIST were divided into two groups: upfront surgery ± adjuvant imatinib (Group A, n = 33) and the neoadjuvant imatinib + surgery + adjuvant imatinib (Group B, n = 52). Baseline characteristics between groups were controlled for with inverse probability of treatment weighting (IPTW) adjusted analysis. RESULTS: The response rate to neoadjuvant imatinib was 65.9%. After the IPTW-adjusted analysis, patients who underwent neoadjuvant therapy had better distant recurrence-free survival (DRFS) and disease-specific survival (DSS) compared with those who underwent upfront surgery (5-year DRFS 97.8 vs. 71.9%, hazard ratio [HR], 0.15; 95% CI, 0.03-0.87; p = 0.03; 5-year DSS 100 vs. 77.1%; HR, 0.11; 95% CI, 0.01-0.92; p = 0.04). While no significant association was found between overall survival (OS) and treatment groups (p = 0.07), 5-year OS was higher for the neoadjuvant group than upfront surgery group (97.8% vs. 71.9%; HR, 0.2; 95% CI, 0.03-1.15). CONCLUSIONS: In patients with localized rectal GIST, neoadjuvant imatinib not only shrunk the tumor size but also decreased the risk of metastasis and tumor-related deaths when compared to upfront surgery and adjuvant imatinib alone.