Development and validation of a nomogram for predicting the overall survival of patients with gastroenteropancreatic neuroendocrine neoplasms

Gastroenteropancreatic neuroendocrine neoplasms: current development, challenges, and clinical perspectives

Neuroendocrine neoplasms (NENs) are highly heterogeneous and potentially malignant tumors arising from secretory cells of the neuroendocrine system. Gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) are the most common subtype of NENs. Historically, GEP-NENs have been regarded as infrequent and slow-growing malignancies; however, recent data have demonstrated that the worldwide prevalence and incidence of GEP-NENs have increased exponentially over the last three decades. In addition, an increasing number of studies have proven that GEP-NENs result in a limited life expectancy. These findings suggested that the natural biology of GEP-NENs is more aggressive than commonly assumed. Therefore, there is an urgent need for advanced researches focusing on the diagnosis and management of patients with GEP-NENs. In this review, we have summarized the limitations and recent advancements in our comprehension of the epidemiology, clinical presentations, pathology, molecular biology, diagnosis, and treatment of GEP-NETs to identify factors contributing to delays in diagnosis and timely treatment of these patients.

Development and Validation of a Novel Machine Learning Model to Predict the Survival of Patients with Gastrointestinal Neuroendocrine Neoplasms

INTRODUCTION

Well-calibrated models for personalized prognostication of patients with gastrointestinal neuroendocrine neoplasms (GINENs) are limited. This study aimed to develop and validate a machine-learning model to predict the survival of patients with GINENs.

METHODS

Oblique random survival forest (ORSF) model, Cox proportional hazard risk model, Cox model with least absolute shrinkage and selection operator penalization, CoxBoost, Survival Gradient Boosting Machine, Extreme Gradient Boosting survival regression, DeepHit, DeepSurv, DNNSurv, logistic-hazard model, and PC-hazard model were compared. We further tuned hyperparameters and selected variables for the best-performing ORSF. Then, the final ORSF model was validated.

RESULTS

A total of 43,444 patients with GINENs were included. The median (interquartile range) survival time was 53 (19-102) months. The ORSF model performed best, in which age, histology, M stage, tumor size, primary tumor site, sex, tumor number, surgery, lymph nodes removed, N stage, race, and grade were ranked as important variables. However, chemotherapy and radiotherapy were not necessary for the ORSF model. The ORSF model had an overall C index of 0.86 (95% confidence interval, 0.85-0.87). The area under the receiver operation curves at 1, 3, 5, and 10 years were 0.91, 0.89, 0.87, and 0.80, respectively. The decision curve analysis showed superior clinical usefulness of the ORSF model than the American Joint Committee on Cancer Stage. A nomogram and an online tool were given.

CONCLUSION

The machine learning ORSF model could precisely predict the survival of patients with GINENs, with the ability to identify patients at high risk for death and probably guide clinical practice.

Rachel score: a nomogram model for predicting the prognosis of lung neuroendocrine tumors

BACKGROUND

Lung NET, classified in typical carcinoids (TC) and atypical carcinoids (AC), are highly heterogeneous in their biology and prognosis. The histological subtype and TNM stage are well-established prognostic factors for lung NET. In a previous work by our group, we demonstrated a significant impact of laterality on lung NET survival outcomes.

MATERIALS AND METHODS

We developed a nomogram that integrates relevant prognostic factors to predict lung NET outcomes. By adding the scores for each of the variables included in the model, it was possible to obtain a prognostic score (Rachel score). Wilcoxon non-parametric statistical test was applied among parameters and Harrell's concordance index was used to measure the models' predictive power. To test the discriminatory power and the predictive accuracy of the model, we calculated Gonen and Heller concordance index. Time-dependent ROC curves and their area under the curve (AUC) were used to evaluate the models' predictive performance.

RESULTS

By applying Rachel score, we were able to identify three prognostic groups (specifically, high, medium and low risk). These three groups were associate to well-defined ranges of points according to the obtained nomogram (I: 0-90, II: 91-130; III: > 130 points), providing a useful tool for prognostic stratification. The overall survival (OS) and progression free survival (PFS) Kaplan-Meier curves confirmed significant differences (p < 0.0001) among the three groups identified by Rachel score.

CONCLUSIONS

A prognostic nomogram was developed, incorporating variables with significant impact on lung NET survival. The nomogram showed a satisfactory and stable ability to predict OS and PFS in this population, confirming the heterogeneity beyond the histopathological diagnosis of TC vs AC.

Combined the surgery, radiation, and chemotherapy for predicting overall survival in patients with gastroenteropancreatic neuroendocrine tumors

BACKGROUND

Over the last few decades, the annual global incidence of gastroenteropancreatic neuroendocrine tumors (GEP-NETs) has steadily increased. Because of the complex and inconsistent treatment of GEP-NETs, the prognosis of patients with GEP-NETs is still difficult to assess. The study aimed to construct and validate the nomograms included treatment data for prediction overall survival (OS) in GEP-NETs patients.

METHODS

GEP-NETs patients determined from the Surveillance, Epidemiology, and End Results (SEER)-13 registry database (1992-2018) and with additional treatment data from the SEER-18 registry database (1975-2016). In order to select independent prognostic factors that contribute significantly to patient survival and can be included in the nomogram, multivariate Cox regression analysis was performed using the minimum value of Akaike information criterion (AIC) and we analyzed the relationship of variables with OS by calculating hazard ratios (HRs) and 95% CIs. In addition, we also comprehensively compared the nomogram using to predict OS with the current 7th American Joint Committee on Cancer (AJCC) staging system.

RESULTS

From 2004 to 2015, a total of 42,662 patients at diagnosis years with GEP-NETs were determined from the SEER database. The results indicated that the increasing incidence of GEP-NETs per year and the highest incidence is in patients aged 50-54. After removing cases lacking adequate clinicopathologic characteristics, the remaining eligible patients (n=7,564) were randomly divided into training (3,782 patients) and testing sets (3,782 patients). In the univariate analysis, sex, age, race, tumor location, SEER historic stage, pathology type, TNM, stage, surgery, radiation, chemotherapy, and CS tumor size were found to be significantly related to OS. Ultimately, the key factors for predicting OS were determined, involving sex, age, race, tumor location, SEER historic stage, M, N, grade, surgery, radiation, and chemotherapy. For internal validation, the C-index of the nomogram used to estimate OS in the training set was 0.816 (0.804-0.828). For external validation, the concordance index (C-index) of the nomogram used to predict OS was 0.822 (0.812-0.832). In the training and testing sets, our nomogram produced minimum AIC values and C-index of OS compared with AJCC stage. Decision curve analysis (DCA) indicated that the nomogram was better than the AJCC staging system because more clinical net benefits were obtained within a wider threshold probability range.

CONCLUSION

A nomogram combined treatment data may be better discrimination in predicting overall survival than AJCC staging system. We highly recommend to use our nomogram to evaluate individual risks based on different clinical features of GEP-NETs, which can improve the diagnosis and treatment outcomes of GEP-NETs patients and improve their quality of life.

A model for predicting the overall survival of gastroenteropancreatic neuroendocrine neoplasms after surgery

BACKGROUND

The increase in the incidence of gastroenteropancreatic neuroendocrine tumors (GEP-NENs) and refined morphological imaging techniques have led to a rise in the number of patients undergoing surgery. However, there is still a paucity of objective, clinically reliable and personalized tools to evaluate patient prognosis.

MATERIALS AND METHODS

We identified patients from the Surveillance, Epidemiology, and End Results (SEER) database who underwent surgery for GEP-NEN from 1975 to 2018. The predictors associated with OS were investigated by Multivariate Cox proportional hazards (PHs) regression analysis in the primary cohort; a prognostic nomogram was then built based on the multivariate analysis results. The performance of the nomogram was assessed by Harrell's concordance index (C-index) and calibration curve and compared with the eighth edition of the American Joint Committee on Cancer (AJCC) staging system.

RESULTS

A total of 45,889 patients were enrolled in our study; 32,321 were included in the primary cohort, and 13,568 were included in the validation cohort. A nomogram incorporating Age, Differentiation, M staging, and AJCC staging was subsequently built based on the multivariate analysis. The C-index (0.833 for the primary cohort and 0.845 for the validation cohort) and calibration curves indicated good discriminative ability and calibration of the nomogram. Further analysis demonstrated that the nomogram had superior discriminatory ability than the AJCC staging system (C-index= 0.706).

CONCLUSION

The proposed nomogram showed excellent prediction with good calibration and discrimination, which can be used to make well-informed and individualized clinical decisions regarding the clinical management of GEP-NENs.

Development and validation of a novel nomogram for predicting survival rate in pancreatic neuroendocrine neoplasms

BACKGROUND

Over the past decades, the incidence and prevalence of pancreatic neuroendocrine neoplasms (pNENs) have steadily increased. However, accurate prediction of the prognosis and treatment of this condition are currently challenging. This study aims to develop and validate a personalized nomogram to predict the survival of patients with pNENs.

MATERIALS AND METHODS

A total of 9739 patients with pNENs were downloaded from the Surveillance, Epidemiology, and End Results (SEER) database. Subsequently, the patients were randomly assigned to a derivation cohort (n = 6874) and a validation cohort (n = 2865). The survival of patients was assessed using the Cox proportional hazards (PHs) regression analysis. Then, the nomogram that predicted 3-and 5-year survival rates were developed in the derivation cohort. Further, the predictive performance of the nomogram was evaluated through discrimination and calibration.

RESULTS

The Cox regression analysis revealed that age, differentiation, the extent of tumor, M staging, and surgery were independent prognostic predictors for pNENs. The nomogram showed superior discrimination capability than AJCC staging in both derived and validation cohorts (C-index: 0.874 versus 0.721 and 0.833 versus 0.721). The calibration curves showed that the practical and predicted survival rates effectively coincided, specifically for the 3-year survival rate.

CONCLUSION

Our nomogram is a valuable tool for the prediction of the survival rate for patients with pNENs; this may promote individualized prognostic evaluation and treatment.