Development and external validation of a nomogram for predicting the effect of tumor size on survival of patients with perihilar cholangiocarcinoma

Machine Learning Model to Predict Early Recurrence in Patients with Perihilar Cholangiocarcinoma Planned Treatment with Curative Resection: A Multicenter Study

BACKGROUND & AIMS

Early recurrence is the leading cause of death for patients with perihilar cholangiocarcinoma (pCCA) after surgery. Identifying high-risk patients preoperatively is important. This study aimed to construct a preoperative prediction model for the early recurrence of pCCA patients planned treatment with curative resection.

METHODS

This study ultimately enrolled 400 pCCA patients after curative resection in five hospitals between 2013 and 2019. They were randomly divided into training (n=300) and testing groups (n=100) at a ratio of 3:1. Associated variables were identified via LASSO regression. Four machine learning models were constructed: support vector machine (SVM), random forest (RF), logistic regression, and K-nearest neighbors (KNN). The predictive ability of the models was evaluated via receiver operating characteristic (ROC) curves, precision-recall curve (PRC) curves, and decision curve analysis (DCA). KaplanMeier survival curves were drawn for the high/low-risk population.

RESULTS

Five factors, CA19-9, tumor size, total bilirubin, hepatic artery invasion, and portal vein invasion, were selected by LASSO regression. In both the training and testing groups, the ROC curve (AUC: 0.983 vs 0.952) and the PRC (0.981 vs 0.939) showed that RF was the best. The cutoff value for distinguishing high- and low-risk patients was 0.51. KM survival curves revealed that in both groups, there was a significant difference in RFS between high- and low-risk patients (P<0.001).

CONCLUSION

This study used preoperative variables from a large, multicenter database to construct a machine learning model that could effectively predict the early recurrence of pCCA patients planned treatment with curative resection and help clinicians make better treatment decisions.

A personalized prognostic model for long-term survival in patients with intrahepatic cholangiocarcinoma: a retrospective cohort study

Purpose

This study aimed to determine the optimal cutoff points for age and tumor size of patients with intrahepatic cholangiocarcinoma (ICC) and to establish and verify a predictive nomogram of overall survival at 1, 3, and 5 years.

Methods

From the SEER (Surveillance, Epidemiology, and End Results) database, 1,325 ICC patients were selected and randomly divided into training and testing cohorts at a 7:3 ratio. Using the X-tile software, age and tumor size were classified into 3 subgroups: ≤61, 62-74, and ≥75 years and ≤35, 36-55, and ≥56 mm. Subsequently, univariate and multivariate Cox regression analyses were performed using the R software in the training cohort to determine independent risk factors, compile the prediction nomogram, and verify it with the testing cohort findings.

Results

The C-indexes of the new prediction nomograms in the training and testing cohorts were 0.738 (95% confidence interval [CI], 0.718-0.758) and 0.750 (95% CI, 0.72-0.78), respectively. Furthermore, the areas under the 1-, 3-, and 5-year receiver operating characteristic (ROC) curves based on the nomogram were 0.792, 0.853, and 0.838, respectively, higher than the ROC based on the 7th and 8th editions of the American Joint Cancer Commission (AJCC) staging system.

Conclusion

This study established and verified a prognostic nomogram that improved the accuracy of the 1-, 3-, and 5-year survival predictions for ICC patients, compared with that based on the 7th and 8th editions of the AJCC staging system, and can help clinicians make personalized survival predictions.

Development and validation of a novel preoperative clinical model for predicting lymph node metastasis in perihilar cholangiocarcinoma

BACKGROUD

We aimed to develop a novel preoperative nomogram to predict lymph node metastasis (LNM) in perihilar cholangiocarcinoma (pCCA) patients.

METHODS

160 pCCA patients were enrolled at Lihuili Hospital from July 2006 to May 2022. A novel nomogram model was established to predict LNM in pCCA patients based on the independent predictive factors selected by the multivariate logistic regression model. The precision of the nomogram model was evaluated through internal and external validation with calibration curve statistics and the concordance index (C-index). Receiver operating characteristic (ROC) curve and decision curve analysis (DCA) were used to evaluate and determine the clinical utility of the nomogram.

RESULTS

Multivariate logistic regression demonstrated that age (OR = 0.963, 95% CI: 0.930-0.996, P = 0.030), CA19-9 level (> 559.8 U/mL vs. ≤559.8 U/mL: OR = 3.162, 95% CI: 1.519-6.582, P = 0.002) and tumour diameter (OR = 1.388, 95% CI: 1.083-1.778, P = 0.010) were independent predictive factors of LNM in pCCA patients. The C-index was 0.763 (95% CI: 0.667-0.860) and 0.677 (95% CI: 0.580-0.773) in training cohort and validation cohort, respectively. ROC curve analysis indicated the comparative stability and adequate discriminative ability of nomogram. The sensitivity and specificity were 0.820 and 0.652 in training cohort and 0.704 and 0.649 in validation cohort, respectively. DCA revealed that the nomogram model could augment net benefits in the prediction of LNM in pCCA patients.

CONCLUSIONS

The novel prediction model is useful for predicting LNM in pCCA patients and showed adequate discriminative ability and high predictive accuracy.

Development and validation of a clinical prediction model for the risk of distal metastasis in intrahepatic cholangiocarcinoma: a real-world study

BACKGROUND

Cholangiocarcinoma (CCA) is a highly malignant and easily metastatic bile duct tumor with poor prognosis. We aimed at studying the associated risk factors affecting distal metastasis of CCA and using nomogram to guide clinicians in predicting distal metastasis of CCA.

METHODS

Based on inclusion and exclusion criteria, 345 patients with CCA were selected from the Fifth Medical Center of Chinese PLA General Hospital and were divided into distal metastases (N = 21) and non-distal metastases (N = 324). LASSO regression models were used to screen for relevant parameters and to compare basic clinical information between the two groups of patients. Risk factors for distal metastasis were identified based on the results of univariate and multivariate logistic regression analyses. The nomogram was established based on the results of multivariate logistic regression, and we drawn the corresponding correlation heat map. The predictive accuracy of the nomogram was evaluated by receiver operating characteristic (ROC) curves and calibration plots. The utility of the model in clinical applications was illustrated by applying decision curve analysis (DCA), and overall survival(OS) analysis was performed using the method of Kaplan-meier.

RESULTS

This study identified 4 independent risk factors for distal metastasis of CCA, including CA199, cholesterol, hypertension and margin invasion, and developed the nomogram based on this. The result of validation showed that the model had significant accuracy for diagnosis with the area under ROC (AUC) of 0.882 (95% CI: 0.843-0.914). Calibration plots and DCA showed that the model had high clinical utility.

CONCLUSIONS

This study established and validated a model of nomogram for predicting distal metastasis in patients with CCA. Based on this, it could guide clinicians to make better decisions and provide more accurate prognosis and treatment for patients with CCA.

A meta-analysis of prognostic factors for early recurrence in perihilar cholangiocarcinoma after curative-intent resection

BACKGROUND

Perihilar cholangiocarcinoma (pCCA) is a type of cancer that has a high rate of recurrence after curative-intent surgery, with about half of all recurrences occurring within the first year. The primary aim of this study was to identify prognostic factors (PFs) for early recurrence (ER, within 12 months) after surgery.

METHODS

Systematic searching was conducted from database inception to September 28th, 2022, with duplicate independent review and data extraction. Data on eight predefined PFs were collected, and meta-analysis was performed on PFs for ER, summarized using forest plots.

RESULTS

The study enrolled 11 studies comprising 2877 patients. In the risk-of-bias assessment, seven studies were rated as low risk and four as moderate risk. More than 34.3% (95% confidence interval [CI], 26.1-42.5%) of the patients experienced ER after curative-intent pCCA resection. Of the PFs, vascular invasion (HR, 2.41; 95% CI, 1.47-3.95; OR, 1.60; 95% CI, 1.17-2.18), lymph node metastases (HR, 2.54; 95% CI, 1.92-3.37; OR, 4.26; 95% CI, 2.40-7.57), and R1 resection (HR, 3.27; 95% CI, 1.81-5.92; OR, 2.40; 95% CI, 1.36-4.22) were associated with an increased hazard for ER. The combined OR values also showed that tumor size, poor tumor differentiation, and perineural invasion were linked to an elevated risk of ER, but all of them had apparent heterogeneity.

CONCLUSION

These findings from the review could be used to plan surveillance of ER and guide post-operative individualized management in pCCA. Furthermore, prospective studies are needed to explore more prognostic factors for ER of pCCA.

Establishment and validation of a prognostic nomogram for extrahepatic cholangiocarcinoma

Simple summary

Accurately estimate the prognosis of patients with ECCA is important. However, the TNM system has some limitations, such as low accuracy, exclusion of other factors (e.g., age and sex), and poor performance in predicting individual survival risk. In contrast, a nomogram-based clinical model related to a comprehensive analysis of all risk factors is intuitive and straightforward, facilitating the probabilistic analysis of tumor-related risk factors. Simultaneously, a nomogram can also effectively drive personalized medicine and facilitate clinicians for prognosis prediction. Therefore, we construct a novel practical nomogram and risk stratification system to predict CSS in patients with ECCA.

Background

Accurately estimate the prognosis of patients with extrahepatic cholangiocarcinoma (ECCA) was important, but the existing staging system has limitations. The present study aimed to construct a novel practical nomogram and risk stratification system to predict cancer-specific survival (CSS) in ECCA patients.

Methods

3415 patients diagnosed with ECCA between 2010 and 2015 were selected from the SEER database and randomized into a training cohort and a validation cohort at 7:3. The nomogram was identified and calibrated using the C-index, receiver operating characteristic curve (ROC), and calibration plots. Decision curve analysis (DCA), net reclassification index (NRI), integrated discrimination improvement (IDI) and the risk stratification were used to compare the nomogram with the AJCC staging system.

Results

Nine variables were selected to establish the nomogram. The C-index (training cohort:0.785; validation cohort:0.776) and time-dependent AUC (>0.7) showed satisfactory discrimination. The calibration plots also revealed that the nomogram was consistent with the actual observations. The NRI (training cohort: 1-, 2-, and 3-year CSS:0.27, 0.27,0.52; validation cohort:1-,2-,3-year CSS:0.48,0.13,0.34), IDI (training cohort: 1-, 2-, 3-year CSS:0.22,0.18,0.16; validation cohort: 1-,2-,3-year CSS:0.18,0.16,0.17), and DCA indicated that the established nomogram significantly outperformed the AJCC staging system (P<0.05) and had better recognition compared to the AJCC staging system.

Conclusions

We developed a practical prognostic nomogram to help clinicians assess the prognosis of patients with ECCA.

Modification of the eighth AJCC/UICC staging system for perihilar cholangiocarcinoma: An alternative pathological staging system from cholangiocarcinoma-prevalent Northeast Thailand

Aim

This study aims to improve the classification performance of the eighth American Joint Committee on Cancer (AJCC) staging system for perihilar cholangiocarcinoma (pCCA) by proposing the Khon Kaen University (KKU) staging system developed in cholangiocarcinoma-prevalent Northeast Thailand.

Method

Four hundred eighty-eight patients with pCCA who underwent partial hepatectomy between 2002 and 2017 at the Srinagarind Hospital, Faculty of Medicine, Khon Kaen University, Thailand, were included. Overall survival (OS) related to clinicopathological features was analyzed using the Kaplan–Meier method. Logrank test was performed in univariate analysis to compare OS data of clinicopathological features to determine risk factors for poor survival. Significant features were further analyzed by multivariate analysis (Cox regression) to identify prognostic factors which were then employed to modify the eighth AJCC staging system.

Results

Multivariate analysis showed that growth pattern (HR = 4.67–19.72, p &lt; 0.001), moderately and poorly differentiated histological grades (HR = 2.31–4.99, p &lt; 0.05 and 0.001, respectively), lymph node metastasis N1 and N2 (HR = 1.37 and 2.18, p &lt; 0.05 and 0.01, respectively), and distant metastasis (HR = 2.11, p &lt; 0.001) were independent factors when compared to their respective reference groups. There was a clear separation of patients with pCCA into KKU stage: I [OS = 116 months (mo.)], II (OS = 46 mo.), IIIA (OS = 24 mo.), IIIB (11 mo.), IVA (OS = 7 mo.), and IVB (OS = 6 mo.).

Conclusion

The new staging system was based on the incorporation of growth patterns to modify the eighth AJCC staging system. The classification performance demonstrated that the KKU staging system was able to classify and distinctly separate patients with pCCA into those with good and poor outcomes. It was also able to improve the stratification performance and discriminative ability of different stages of pCCA classification better than the eighth AJCC staging system. Hence, the KKU staging system is proposed as an alternative model to augment the accuracy of survival prognostication and treatment performance for patients with pCCA.

Development and Validation of a Prognostic Model to Predict Recurrence-Free Survival After Curative Resection for Perihilar Cholangiocarcinoma: A Multicenter Study

Background

Recurrence is the main cause of death in perihilar cholangiocarcinoma (pCCA) patients after surgery. Identifying patients with a high risk of recurrence is important for decision-making regarding neoadjuvant therapy to improve long-term outcomes.

Aim

The objective of this study was to develop and validate a prognostic model to predict recurrence-free survival (RFS) after curative resection of pCCA.

Methods

Patients following curative resection for pCCA from January 2008 to January 2016 were identified from a multicenter database. Using random assignment, 70% of patients were assigned to the training cohort, and the remaining 30% were assigned to the validation cohort. Independent predictors of RFS after curative resection for pCCA were identified and used to construct a prognostic model. The predictive performance of the model was assessed using calibration curves and the C-index.

Results

A total of 341 patients were included. The median overall survival (OS) was 22 months, and the median RFS was 14 months. Independent predictors associated with RFS included lymph node involvement, macrovascular invasion, microvascular invasion, maximum tumor size, tumor differentiation, and carbohydrate antigen 19-9. The model incorporating these factors to predict 1-year RFS demonstrated better calibration and better performance than the 8th American Joint Committee on Cancer (AJCC) staging system in both the training and validation cohorts (C-indexes: 0.723 vs. 0.641; 0.743 vs. 0.607).

Conclusions

The prognostic model could identify patients at high risk of recurrence for pCCA to inform patients and surgeons, help guide decision-making for postoperative adjuvant therapy, and improve survival.

Predicting Disease-Specific Survival for Patients With Primary Cholangiocarcinoma Undergoing Curative Resection by Using a Decision Tree Model

Background

The aim of this study was to derive and validate a decision tree model to predict disease-specific survival after curative resection for primary cholangiocarcinoma (CCA).

Method

Twenty-one clinical characteristics were collected from 482 patients after curative resection for primary CCA. A total of 289 patients were randomly allocated into a training cohort and 193 were randomly allocated into a validation cohort. We built three decision tree models based on 5, 12, and 21 variables, respectively. Area under curve (AUC), sensitivity, and specificity were used for comparison of the 0.5-, 1-, and 3-year decision tree models and regression models. AUC and decision curve analysis (DCA) were used to determine the predictive performances of the 0.5-, 1-, and 3-year decision tree models and AJCC TNM stage models.

Results

According to the fitting degree and the computational cost, the decision tree model derived from 12 variables displayed superior predictive efficacy to the other two models, with an accuracy of 0.938 in the training cohort and 0.751 in the validation cohort. Maximum tumor size, resection margin, lymph node status, histological differentiation, TB level, ALBI, AKP, AAPR, ALT, γ-GT, CA19-9, and Child-Pugh grade were involved in the model. The performances of 0.5-, 1-, and 3-year decision tree models were better than those of conventional models and AJCC TNM stage models.

Conclusion

We developed a decision tree model to predict outcomes for CCA undergoing curative resection. The present decision tree model outperformed other clinical models, facilitating individual decision-making of adjuvant therapy after curative resection.

Prognostic nomogram for predicting cancer-specific survival in patients with resected hilar cholangiocarcinoma: a large cohort study

BACKGROUND

The aim of the study was to establish and validate a novel prognostic nomogram of cancer-specific survival (CSS) in resected hilar cholangiocarcinoma (HCCA) patients.

METHODS

A training cohort of 536 patients and an internal validation cohort of 270 patients were included in this study. The demographic and clinicopathological variables were extracted from the Surveillance, Epidemiology and End Results (SEER) database. Univariate and multivariate Cox regression analysis were performed in the training cohort, followed by the construction of nomogram for CSS. The performance of the nomogram was assessed by concordance index (C-index) and calibration plots and compared with the American Joint Committee on Cancer (AJCC) staging systems. Decision curve analysis (DCA) was applied to measure the predictive power and clinical value of the nomogram.

RESULTS

The nomogram incorporating age, tumor size, tumor grade, lymph node ratio (LNR) and T stage parameters was with a C-index of 0.655 in the training cohort, 0.626 in the validation cohort, compared with corresponding 0.631, 0.626 for the AJCC 8th staging system. The calibration curves exhibited excellent agreement between CSS probabilities predicted by nomogram and actual observation in the training cohort and validation cohort. DCA indicated that this nomogram generated substantial clinical value.

CONCLUSIONS

The proposed nomogram provided a more accurate prognostic prediction of CSS for individual patients with resected HCCA than the AJCC 8th staging system, which might be served as an effective tool to stratify resected HCCA patients with high risk and facilitate optimizing therapeutic benefit.

Nomogram for Predicting Overall Survival and Assessing the Survival Benefit of Adjuvant Treatment in pT1-2N0M0 Triple-Negative Breast Cancer: A Surveillance, Epidemiology, and End Results-Based Study

Background

Accurate survival prediction of triple-negative breast cancer (TNBC) is essential in the decision-making of adjuvant treatment. The aim of this prospective study was to develop a nomogram that predicts overall survival and assists adjuvant treatment formulation.

Methods

A total of 16,977 patients with pT1-2N0M0 TNBC between 2010 and 2015 from the SEER database were enrolled. Independent prognostic factors associated with overall survival (OS) were identified using univariate and multivariate Cox regression hazards method and utilized to compose the nomogram. The survival benefit of adjuvant treatment on OS were analyzed after stratification by nomogram sum-score.

Results

Patients were randomized 7:3 into the training and validation cohorts. Multivariate analysis revealed that age at diagnosis, grade, tumor size, laterality, and mastectomy type were independent prognostic factors of OS and were integrated to develop a nomogram for predicting prognosis. Patients were stratified into 3 prognostic subgroups according to the sum-score of our nomogram. There were no significant differences found in OS between surgery alone and other adjuvant treatment strategies in low risk group. In moderate risk group, patients receiving chemotherapy or the combination of chemotherapy and radiotherapy showed better OS than those receiving surgery alone or radiotherapy alone. For patients in high risk group, the combination of chemotherapy and radiotherapy could maximally improve the overall survival rate of patients.

Conclusion

A novel nomogram for OS prediction and risk stratification in patients with pT1-2N0M0 TNBC was developed. This cohort study reveals the prognostic roles of different adjuvant treatment strategies in subgroups, which may provide a reference for the decision-making of postoperative treatment, eventually improving prognosis for individual patients.

Survival Analysis in Gastrointestinal Neuroendocrine Carcinoma With Bone Metastasis at Diagnosis

Purpose

Gastrointestinal neuroendocrine carcinoma (NEC) with bone metastasis is rarely reported. The purpose of this study is to explore the prognosis and risk factors of such patients.

Patients and Methods

We retrospectively reviewed patients diagnosed as gastrointestinal NECs with bone metastasis at diagnosis from 2010 to 2016 by using the Surveillance, Epidemiology and End Results (SEER) database. Predictors of overall survival (OS) and cancer-specific survival (CSS) were analyzed by univariable and multivariable Cox analyses. Kaplan–Meier plots were constructed to show the correlation between independent predictors and survival.

Results

A total of 330 gastrointestinal NEC patients with bone metastasis at diagnosis were included for analysis. Over half of patients were male and older than 60 years old. The most common primary site of gastrointestinal NEC with bone metastasis was the pancreas. The prognosis of gastrointestinal NEC with bone metastasis (3-year OS and CSS rates: 16.7 and 17.0%) was very poor. On Cox multivariable analysis, age over 60 years old, no surgery, and lung metastasis were independent predictors of decreased OS and CSS.

Conclusions

We identified three independent factors associated with prognosis among gastrointestinal NEC patients with bone metastasis, namely age, surgery, and lung metastasis. For younger gastrointestinal NEC patients with bone metastasis, surgical resection of primary tumors as well as actively treating lung metastasis might be useful for prolonging survival.

Machine-learning radiomics to predict early recurrence in perihilar cholangiocarcinoma after curative resection

BACKGROUND AND AIMS

Up to 40%-65% of patients with perihilar cholangiocarcinoma (PHC) rapidly progress to early recurrence (ER) even after curative resection. Quantification of ER risk is difficult and a reliable prognostic prediction tool is absent. We developed and validated a multilevel model, integrating clinicopathology, molecular pathology and radiology, especially radiomics coupled with machine-learning algorithms, to predict the ER of patients after curative resection in PHC.

METHODS

In total, 274 patients who underwent contrast-enhanced CT (CECT) and curative resection at 2 institutions were retrospectively identified and randomly divided into training (n = 167), internal validation (n = 70) and external validation (n = 37) sets. A machine-learning analysis of 18,120 radiomic features based on multiphase CECT and 48 clinico-radiologic characteristics was performed for the multilevel model.

RESULTS

Comprehensively, 7 independent factors (tumour differentiation, lymph node metastasis, pre-operative CA19-9 level, enhancement pattern, A-Shrink score, V-Shrink score and P-Shrink score) were built to the multilevel model and quantified the risk of ER. We benchmarked the gain in discrimination with the area under the curve (AUC) of 0.883, superior to the rival clinical and radiomic models (AUCs 0.792-0.805). The accuracy (ACC) of the multilevel model was 0.826, which was significantly higher than those of the conventional staging systems (AJCC 8th (0.641), MSKCC (0.617) and Gazzaniga (0.581)).

CONCLUSION

The radiomics-based multilevel model demonstrated superior performance to rival models and conventional staging systems, and could serve as a visual prognostic tool to plan surveillance of ER and guide post-operative individualized management in PHC.