Risk factors of liver metastasis from advanced pancreatic adenocarcinoma: a large multicenter cohort study

Predicting liver metastasis in pancreatic neuroendocrine tumors with an interpretable machine learning algorithm: a SEER-based study

Background

Liver metastasis is the most common site of metastasis in pancreatic neuroendocrine tumors (PaNETs), significantly affecting patient prognosis. This study aims to develop machine learning algorithms to predict liver metastasis in PaNETs patients, assisting clinicians in the personalized clinical decision-making for treatment.

Methods

We collected data on eligible PaNETs patients from the Surveillance, Epidemiology, and End Results (SEER) database for the period from 2010 to 2021. The Boruta algorithm and the Least Absolute Shrinkage and Selection Operator (LASSO) were used for feature selection. We applied 10 different machine learning algorithms to develop models for predicting the risk of liver metastasis in PaNETs patients. The model's performance was assessed using a variety of metrics, including the area under the receiver operating characteristic curve (AUC), the area under the precision-recall curve (AUPRC), decision curve analysis (DCA), calibration curves, accuracy, sensitivity, specificity, F1 score, and Kappa score. The SHapley Additive exPlanations (SHAP) were employed to interpret models, and the best-performing model was used to develop a web-based calculator.

Results

The study included a cohort of 7,463 PaNETs patients, of whom 1,356 (18.2%) were diagnosed with liver metastasis at the time of initial diagnosis. Through the combined use of the Boruta and LASSO methods, T-stage, N-stage, tumor size, grade, surgery, lymphadenectomy, chemotherapy, and bone metastasis were identified as independent risk factors for liver metastasis in PaNETs. Compared to other machine learning algorithms, the gradient boosting machine (GBM) model exhibited superior performance, achieving an AUC of 0.937 (95% CI: 0.931-0.943), an AUPRC of 0.94, and an accuracy of 0.87. DCA and calibration curve analyses demonstrate that the GBM model provides better clinical decision-making capabilities and predictive performance. Furthermore, the SHAP framework revealed that surgery, N-stage, and T-stage are the primary decision factors influencing the machine learning model's predictions. Finally, based on the GBM algorithm, we developed an accessible web-based calculator to predict the risk of liver metastasis in PaNETs.

Conclusion

The GBM model excels in predicting the risk of liver metastasis in PaNETs patients, outperforming other machine learning models and providing critical support for developing personalized medical strategies in clinical practice.

Development and validation of a nomogram for a prognostic model for resected pancreatic ductal adenocarcinoma

BACKGROUND

Pancreatic ductal adenocarcinoma (PDAC) is a highly malignant tumor. Surgical resection is the most promising therapeutic strategy for PDAC, and how to improve the survival rate remains a vital key point. This study aimed to establish and validate a nomogram for predicting the prognosis of resected PDAC.

METHODS

A total of 174 patients with PDAC who underwent surgical resection at Lanzhou University Second Hospital and the First Affiliated Hospital of Zhengzhou University from January 2012 to July 2022 were enrolled. The clinicopathological characteristics and survival data were analyzed by R software (version 4.1.3). Univariate and multivariate Cox regression analyses were used to analyze the effects of clinicopathological characteristics on overall survival (OS).

RESULTS

Multivariate Cox regression showed that carbohydrate antigen 19-9 (CA19-9) ≥ 476 U/mL, carbohydrate antigen 125 (CA125) ≥ 32 U/mL, fasting blood glucose (FBG) < 6.86 mmol/L, aspartate aminotransferase (AST) ≥ 107 U/L, positive surgical margin, and more than 4 cycles of postoperative chemotherapy were independent prognostic factors for OS. Patients were divided into the high-risk and low-risk groups based on the median risk score calculated by multivariate Cox regression analysis. Kaplan-Meier survival curves revealed that the 5-year survival rates of the high-risk and low-risk groups in the training cohort were 5.79% and 24.3%, respectively, and those in the validation cohort were 0 and 19.0%, respectively (P < 0.05). Receiver operating characteristic (ROC) curve analysis revealed that area under the ROC curve (AUC) of the risk score in the training set and the validation set were 0.855 and 0.838, respectively. The C-indexes of the nomogram in the training set and validation set were 0.788 (95% CI: 0.745-0.831) and 0.773 (95% CI: 0.718-0.828), respectively.

CONCLUSIONS

We developed a nomogram that predicts OS in patients with resected PDAC, and the validation results showed that the nomogram model had a strong predictive ability. Particularly, FBG < 6.86 mmol/L and more than 4 cycles of postoperative chemotherapy can predict better OS of PDAC after surgery.

Predictive biomarkers of pancreatic cancer metastasis: A comprehensive review

This review provides a comprehensive overview of predictive biomarkers associated with metastasis in pancreatic cancer (PC), one of the most aggressive malignancies characterized by late-stage diagnosis and poor prognosis. Metastasis, particularly to the liver, lungs, and lymph nodes, significantly worsens patient outcomes by compromising organ function and promoting disease progression. Reliable biomarkers for predicting and detecting metastasis at early stages are critical for improving survival rates and guiding personalized therapies. This paper highlights both general and specific biomarkers, including genetic mutations, protein expression changes, and carbohydrate tumor markers such as CA19-9. Immunological factors, including PD-L1, inflammatory cytokines, and chemokines, further influence the metastatic process within the tumor microenvironment (TME). Specific biomarkers play pivotal roles in promoting metastasis through mechanisms such as epithelial-to-mesenchymal transition (EMT), tumor microenvironment remodeling, and immune evasion. Emerging markers such as circulating tumor cells (CTCs) and volatile organic compounds (VOCs) offer promising non-invasive tools for metastasis detection and monitoring. This review not only consolidates existing knowledge but also highlights the mechanisms through which specific biomarkers facilitate metastasis. Despite recent progress, challenges such as biomarker standardization, technical variability, and clinical validation remain, and addressing these hurdles is essential for integrating predictive biomarkers into clinical practice. Ultimately, this review contributes to advancing early detection strategies, personalized treatment options, and improved prognosis for PC patients.

Research trends in the field of the gut-brain interaction: Functional dyspepsia in the spotlight – An integrated bibliometric and science mapping approach

Objectives

This study aims to perform a bibliometric analysis of functional dyspepsia (FD), which includes visualizing bibliographic information, in order to identify prevailing study themes, topics of interest, contributing journals, countries, institutions, and authors as well as co-citation patterns.

Methods

The Web of Science™ Core Collection Database was used to retrieve all peer-reviewed scientific publications related to FD research. The validated search terms were entered into the “title” and “author keywords” fields, and the results were sorted by publication year from 2006 to 2022. There were no restrictions on language. On 12 February 2023, a manual export of the complete metadata for each original publication and review article was performed. CiteSpace was used to reveal co-authorship, publication, and co-citation patterns to find prominent authors, organizations, countries, and journals in FD research as well as to identify author keywords with strong citation bursts, which could indicate an emerging research area. VOSviewer was used to build the co-occurrence indicator (co-word) to identify the main author keywords on which previous studies focused and to induce clustered scientific landscape for two consecutive periods to identify intriguing areas for future research.

Results

A search of the database retrieved 2,957 documents. There was a wave-like pattern in the number of publications until 2017, after which there was a spike in publication volume. The USA, China, and Japan provided the majority of contributions. In terms of institution, Mayo Clin, Univ Newcastle, and Katholieke Univ Leuven were found to be the prolific institutions. Additionally, the results indicate that eastern Asian researchers contributed significantly to the global knowledge of literature that led other countries; however, Canada, the USA, Australia, England, and Germany were found to have the highest degree of betweenness centrality. Nicholas J. Talley, Jan Tack, Gerald Holtmann, Michael Camilleri, Ken Haruma, and Paul Moayyedi occupied the top positions based on productivity and centrality indicators. Six thematic clusters emerged (Helicobacter pylori infection; pathophysiological mechanisms of FD; extraintestinal co-morbidities and overlap syndromes associated with FD; herbal medicine in FD; diabetic gastroparesis; and dietary factors in FD). “Acupuncture,” “duodenal eosinophilia,” “gut microbiota,” and others were among the author keywords with rising prevalence.

Conclusion

In FD research, eastern Asian countries have established themselves as major contributors with the highest publishing productivity; however, research has primarily been driven by North America, Europe, and Australia, where cooperation is generally more active and highly influential scientific results are produced. Our analysis suggests that increased investments, training of human resources, improved infrastructures, and expanded collaborations are essential to improving the quality of FD research in Asia. The emerging author keyword analysis suggests that eosinophil-mast cell axis, gut microbiota, mental disorders, and acupuncture are the key areas that attract researchers’ attention as future research boulevards. There is a highly skewed distribution of research output across Asia, with most focus on complementary and alternative medicine (CAM) coming from Chinese, Japanese, and South Korean centers. However, CAM remains an underexplored area of research in the context of FD, and it deserves greater research efforts in order to obtain quality scientific evidence. Furthermore, we propose that the research framework of CAM should not be limited to dysmotility; rather, it could be interpreted within a more holistic context that includes the brain-gut-microbiota axis, as well as novel concepts such as duodenitis, increased mucosal permeability, and infiltration and activation of eosinophils and mast cells, among others. Overall, we provided bibliometrics-based overviews of relevant literature to researchers from different backgrounds and healthcare professionals to provide an in-depth overview of major trends in FD research.

Clinical risk factors of carbohydrate antigen-125, cytokeratin fragment 19, and neuron-specific enolase in liver metastases from elderly lung cancer patients

Objective: Lung cancer is a common malignant tumor characterized by challenging detection and lack of specificity in clinical manifestations. To investigate the correlation of tumor markers in the serum with liver metastasis and prognosis of lung cancer.

Methods: A total of 3,046 elderly lung cancer patients were retrospectively studied between September 1999 and July 2020. Divided into liver metastasis group and non-liver metastasis group. We compared a series of serum biomarkers between the two groups of elderly patients to predict the prognosis in patients with lung cancer by fluorescence in situ hybridization (FISH), advanced flow cytometry (FCM) and multi tumor marker protein chip, including tumor markers in the serum included alkaline phosphatase (ALP), serum calcium, hemoglobin (HB), alpha-fetoprotein (AFP), carcinoembryonic antigen (CEA), neuron-specific enolase (NSE), cytokeratin fragment 19 (Cyfra21-1), carbohydrate antigen-125 (CA-125), carbohydrate antigen-153 (CA-153), carbohydrate antigen-199 (CA-199), and free prostate specific antigen (free PSA). We used binary logistic regression analysis to determine risk factors, and used receiver operating curve (ROC) analysis to evaluate the diagnostic value of liver metastases in elderly patients with lung cancer.

Results: The proportion of lung cancer in the liver metastasis group was higher than that observed in the non-liver metastases group. The expression levels of CA-125, Cyfra21-1, and NSE in the liver metastasis group of lung cancer were significantly higher than those reported in the non-liver metastases group (p &lt; 0.05). ROC curve analysis shows that the area under the curve of CA-125, Cyfra21-1, and NSE are 0.614, 0.616 and 0.608, respectively. The sensitivity and specificity of CA-125 were 45.70% and 76.20%, the sensitivity and specificity of Cyfra21-1 were 60.10% and 57.10%, and the sensitivity and specificity of NSE were 44.10% and 75.00%, respectively.

Conclusion: High levels of CA-125, Cyfra21-1, and NSE in the serum may be associated with liver metastasis in elderly patients with lung cancer. CA-125 and NSE are factors influencing the prognosis of elderly patients with liver metastasis of lung cancer.

Risk Assessment of Liver Metastasis in Pancreatic Cancer Patients Using Multiple Models Based on Machine Learning: A Large Population-Based Study

Background

A more accurate prediction of liver metastasis (LM) in pancreatic cancer (PC) would help improve clinical therapeutic effects and follow-up strategies for the management of this disease. This study was to assess various prediction models to evaluate the risk of LM based on machine learning algorithms.

Methods

We retrospectively reviewed clinicopathological characteristics of PC patients from the Surveillance, Epidemiology, and End Results database from 2010 to 2018. The logistic regression, extreme gradient boosting, support vector, random forest (RF), and deep neural network machine algorithms were used to establish models to predict the risk of LM in PC patients. Specificity, sensitivity, and receiver operating characteristic (ROC) curves were used to determine the discriminatory capacity of the prediction models.

Results

A total of 47,919 PC patients were identified; 15,909 (33.2%) of which developed LM. After iterative filtering, a total of nine features were included to establish the risk model for LM based on machine learning. The RF showed the most promising results in the prediction of complications among the models (ROC 0.871 for training and 0.832 for test sets). In risk stratification analysis, the LM rate and 5-year cancer-specific survival (CSS) in the high-risk group were worse than those in the intermediate- and low-risk groups. Surgery, radiotherapy, and chemotherapy were found to significantly improve the CSS in the high- and intermediate-risk groups.

Conclusion

In this study, the RF model constructed could accurately predict the risk of LM in PC patients, which has the potential to provide clinicians with more personalized clinical decision-making recommendations.

Evolving pancreatic cancer treatment: From diagnosis to healthcare management

The prognosis of pancreatic ductal adenocarcinoma is still the worst among solid tumors. In this review, a panel of experts addressed the main unanswered questions about the clinical management of this disease, with the aim of providing practical decision support for physicians. On the basis of the evidence available from the literature, the main topics concerning pancreatic cancer are discussed: the diagnosis, as the need for a pathological characterization and the role for germ-line and somatic molecular profiling; the therapeutic management of resectable disease, as the role of upfront surgery or neoadjuvant chemotherapy, the post-operative restaging and the optimal timing foradjuvant chemotherapy, the management of the borderline resectable and locally advanced disease; the metastatic disease and the role of surgery for the management of patients with isolated metastasis and the use of biomarkers of metastatic potential; the role of supportive care and the healthcare management of pancreatic ductal adenocarcinoma.

Development and validation of a radiomics nomogram to discriminate advanced pancreatic cancer with liver metastases or other metastatic patterns

BACKGROUND

Patients with advanced pancreatic cancer (APC) and liver metastases have much poorer prognoses than patients with other metastatic patterns.

OBJECTIVE

This study aimed to develop and validate a radiomics model to discriminate patients with pancreatic cancer and liver metastases from those with other metastatic patterns.

METHODS

We evaluated 77 patients who had APC and performed texture analysis on the region of interest. 58 patients and 19 patients were allocated randomly into the training and validation cohorts with almost the same proportion of liver metastases. An independentsamples t-test was used for feature selection in the training cohort. Random forest classifier was used to construct models based on these features and a radiomics signature (RS) was derived. A nomogram was constructed based on RS and CA19-9, and was validated with calibration plot and decision curve. The prognostic value of RS was evaluated by Kaplan-Meier methods.

RESULTS

The constructed nomogram demonstrated good discrimination in the training (AUC = 0.93) and validation (AUC = 0.81) cohorts. In both cohorts, patients with RS > 0.61 had much poorer overall survival than patients with RS < 0.61.

CONCLUSIONS

This study presents a radiomics nomogram incorporating RS and CA19-9 to discriminate patients who have APC with liver metastases from patients with other metastatic patterns.

Prognostic stratification based on a novel nomogram for left-sided pancreatic adenocarcinoma after surgical resection: a multi-center study

Left-sided pancreatic adenocarcinoma (LPAC) has a poorer prognosis and has some distinct features compared to cancer of pancreatic head. A reliable model to predict the prognosis of LPAC following surgery is needed in clinical practice. Our study included 231 patients with resected LPAC from 3 Chinese pancreatic disease centers. Cox-regression analysis was conducted to identify independent risk factors of LAPC. Then we established a nomogram and performed C-index, receiver operating characteristic curve, calibration plot and decision curve analysis to assess its discrimination and calibration. As a result, CA19-9, surgical margin, tumor differentiation, lymph node metastasis, and postoperative adjuvant chemotherapy were identified as significant prognostic factors. Based on these predictors, a novel nomogram was constructed. The nomogram achieved high C-indexes in the training cohort (0.805) and validation cohort (0.719), which were superior than the AJCC-8 staging system and other nomograms. The area under curve of the nomogram for predicting patients survival at 1-, 2-, and 3-year in training cohort were more than 0.8. Kaplan-Meier survival curve for the subgroups stratified based on the nomogram showed a better separation than the AJCC-8 stage I, II, III, indicating a superior ability of risk stratification for our model. In summary, we constructed a nomogram which showed a better predictive ability for patients' survival with LPAC after surgical resection than the AJCC staging system and other predictive models. Our model would be helpful to discriminate high-risk LPAC and facilitate clinical decision making.

Development and validation of a simple-to-use nomogram to predict liver metastasis in patients with pancreatic neuroendocrine neoplasms: a large cohort study

Background

Liver metastasis is an important prognostic factor for pancreatic neuroendocrine neoplasms (pNENs), but the relationship between the clinical features of patients with pNEN and liver metastasis remains undetermined. The aim of this study was to establish and validate an easy-to-use nomogram to predict liver-metastasis in patients with pNEN.

Methods

We obtained the clinicopathologic data of 2960 patients with pancreatic neuroendocrine neoplasms from the Surveillance, Epidemiology and End Results (SEER) database between 2010 and 2016. Univariate and multivariate logistic regression were done to screen out independent influencing factors to establish the nomogram. The calibration plots and the area under the receiver operating characteristic curve (AUC) were used to evaluate the performance of nomogram. Decision curve analysis (DCA) was applied to compare the novel model with the conventional predictive methods.

Results

A total of 2960 patients with pancreatic neuroendocrine neoplasms were included in the study. Among these, 1974 patients were assigned to the training group and 986 patients to the validation group. Multivariate logistic regression identified, tumor size, grade, other site metastasis, T stage and N stage as independent risk factors. The calibration plot showed good discriminative ability in the training and validation groups, with C-indexes of 0.850 for the training cohort and 0.846 for the validation cohort. The AUC values were 0.850 (95% CI 0.830–0.869) and 0.839 (95% CI 0.812–0.866), respectively. The nomogram total points (NTP) had the potential to stratify patients into low risk, medium risk and high risk (P < 0.001). Finally, comparing the nomogram with traditional prediction methods, the DCA curve showed that the nomogram had better net benefit.

Conclusions

Our nomogram has a good ability to predict liver metastasis of pancreatic neuroendocrine neoplasms, and it can guide clinicians to provide suitable prevention and treatment measures for patients with medium- and high-risk liver metastasis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-021-01685-w.

Topically applicated curcumin/gelatin-blended nanofibrous mat inhibits pancreatic adenocarcinoma by increasing ROS production and endoplasmic reticulum stress mediated apoptosis

BACKGROUND

Pancreatic adenocarcinoma (PDAC) is one of the most fatal malignancies. Surgical resection supplemented by chemotherapy remains the major therapeutic regimen, but with unavoidable resistance and systemic toxic reaction. Curcumin is a known safe natural compound that can effectively eliminate pancreatic adenocarcinoma cells in vitro, making it a promising candidate for substitution in subsequent chemotherapy. However, due to its extremely low bioavailability caused by its insolubility and circular elimination, curcumin had an unexpectedly modest therapeutic effect in clinical trials.

RESULTS

Here, we electrospun curcumin/gelatin-blended nanofibrous mat to largely improve curcumin's bioavailability by local controlled-release. With characterization by scanning electron microscopy, fluorescence microscopy, Fourier transform infrared spectroscopy, X-ray diffraction and high-performance liquid chromatography, it was revealed that curcumin was uniformly dispersed in the fiber of the mats with nanoscopic dimensions and could be continuously released into the surrounding medium for days. The cancer inhibitory effects of nano-curcumin and underlying mechanisms were further explored by assays using pancreatic adenocarcinoma cell and experiments using xenograft model. The results showed the released nano-curcumin could effectively inhibit pancreatic adenocarcinoma cell proliferation not only in vitro, but more importantly in vivo. This cytotoxic effect of nano-curcumin against pancreatic adenocarcinoma was achieved through provoking the production of intracellular reactive oxygen species and activating endoplasmic reticulum stress, which leads to enhanced cell apoptosis via decreased phosphorylation of signal transducer and activator of transcription 3.

CONCLUSIONS

Clinically, curcumin/gelatin-blended nanofibrous mat could be a promising, secure, efficient and affordable substitutional agent for the elimination of residual cancer cells after tumor resection. Moreover, our strategy to obtain curcumin released from nanofibrous mat may provide a universally applicable approach for the study of the therapeutic effects and molecular mechanisms of other potential medicines with low bioavailability.

The diversity between curatively resected pancreatic head and body-tail cancers based on the 8th edition of AJCC staging system: a multicenter cohort study

BACKGROUND

To our knowledge, there are no studies to systematically compare the detailed clinical significance between curatively resected pancreatic head (ph) and body-tail (pbt) ductal adenocarcinoma based on the new 8th edition of AJCC staging system (8th AJCC stage) that was just applied in clinical practice in 2018.

METHODS

Three hundred fifty-one patients with curatively resected pancreatic adenocarcinoma (PC) from three center hospitals were entered into this multicenter cohort study.

RESULTS

Increasing tumor size (P < 0.001), T stage (T1 + T2 vs T3 + T4, P = 0.003), frequent postoperative liver metastasis (PLM) (P = 0.002) and 8th AJCC stage (IA to VI, P < 0.001; I + II vs III + IV, P = 0.002) were closely associated with the progression of pbt cancers compared with that in ph cancer patients. Moreover, tumor size≥3 cm (P = 0.012), 8th AJCC stage (III + IV) (P = 0.025) and PLM (P = 0.010) were identified as independent risk factors in pbt cancers in logistic analysis. Patients with pbt cancers had a significantly worse overall survival compared with ph cancer patients (P = 0.003). Moreover, pbt was an independent unfavorable factor in multivariate analysis (P = 0.011). In addition to lymph nodes metastasis, 8th AJCC stage, vascular invasion and PLM, increasing tumor size and advanced T stage were also closely associated with the poor prognosis in 131 cases of pbt cancer patients compared with Ph cancer patients.

CONCLUSION

Pbt, as an independent unfavorable factor for the prognosis of PC patients, are much more aggressive than that in ph cancers according to 8th AJCC staging system. 8th AJCC staging system are more comprehensive and sensitive to reflect the malignant biology of pbt cancers.

Development and validation of a nomogram to predict liver metastasis in patients with pancreatic ductal adenocarcinoma: a large cohort study

Background: Few studies have explored the relationship between clinicopathological factors of patients with pancreatic ductal adenocarcinoma (PDAC) and liver metastasis. The aim of this study was to develop and validate a nomogram to predict liver metastasis in patients with PDAC. Patients and methods: Patients diagnosed with PDAC between 2004 and 2015 from the Surveillance, Epidemiology, and End Results (SEER) database were retrospectively collected. The nomogram was established based on a logistic regression model. The precision of the nomogram was evaluated and compared using concordance index (C-index), and the area under receiver operating characteristic curve (AUC). The clinical use of nomogram was evaluated by making use of a decision curve analysis (DCA). Results: A total of 12,644 eligible patients, which were randomly divided into training (n=9,483) and validation cohorts (n=3,161), were included in this study. The nomograms, which were established on the basis of independent predictors, were well calibrated, and demonstrated good discriminative ability, with C-indexes of 0.784 for the training cohort and 0.790 for validation cohort. The values of AUC for training and validation cohort were 0.792 and 0.800, respectively. When other sites of distant metastases were included into this predictive system, the new predictive model demonstrated a better discriminative ability and greater net benefit in predicting liver metastasis in patients with PDAC in both the training and validation cohorts. Conclusion: Nomograms were constructed to predict liver metastasis in patients with PDAC. Validation revealed excellent discrimination and calibration of the nomograms, suggesting that the nomograms were well calibrated and could serve to improve the prediction of the risks of liver metastasis which can be used to guide the management of patients with PDAC.

A modified M-stage classification based on the metastatic patterns of pancreatic neuroendocrine neoplasms: a population-based study

BACKGROUND

The present study aims to improve the M-stage classification of pancreatic neuroendocrine neoplasms (pNENs).

METHODS

Two thousand six hundred sixty six pNENs were extracted from the Surveillance, Epidemiology, and End Results database to explore the metastatic patterns of pNENs. Metastatic patterns were categorized as single, two, or multiple (three or more) distant organ metastasis. The mean overall survival and hazard rate of different metastatic patterns were calculated by Kaplan-Meier and Cox proportional hazards models, respectively. The discriminatory capability of the modified M-stage classification was evaluated by Harrell's concordance index.

RESULTS

The overall survival time significantly decreased with an increasing number of metastatic organs. In addition, pNENs with only liver metastasis had better prognosis when compared to other metastatic patterns. Thus, we modified the M-stage classification (mM-stage) as follows: mM₀-stage, tumor without metastasis; mM₁-stage, tumor only metastasized to liver; mM₂-stage, tumor metastasized to other single distant organ (lung, bone, or brain) or two distant organs; mM₃-stage, tumor metastasized to three or more distant organs. Harrell's concordance index showed that the modified M-stage classification had superior discriminatory capability than both the American Joint Committee on Cancer (AJCC) and the European Neuroendocrine Tumor Society (ENETS) M-stage classifications.

CONCLUSIONS

The modified M-stage classification is superior to both AJCC and ENETS M-stage classifications in the prognosis of pNENs. In the future, individualized treatment and follow-up programs should be explored for patients with distinct metastatic patterns.