Novel machine-learning prediction tools for overall survival of patients with chondrosarcoma: Based on recursive partitioning analysis.

BACKGROUND: Chondrosarcoma (CHS), a bone malignancy, poses a significant challenge due to its heterogeneous nature and resistance to conventional treatments. There is a clear need for advanced prognostic instruments that can integrate multiple prognostic factors to deliver personalized survival predictions for individual patients. This study aimed to develop a novel prediction tool based on recursive partitioning analysis (RPA) to improve the estimation of overall survival for patients with CHS. METHODS: Data from the Surveillance, Epidemiology, and End Results (SEER) database were analyzed, including demographic, clinical, and treatment details of patients diagnosed between 2000 and 2018. Using C5.0 algorithm, decision trees were created to predict survival probabilities at 12, 24, 60, and 120 months. The performance of the models was assessed through confusion scatter plot, accuracy rate, receiver operator characteristic (ROC) curve, and area under ROC curve (AUC). RESULTS: The study identified tumor histology, surgery, age, visceral (brain/liver/lung) metastasis, chemotherapy, tumor grade, and sex as critical predictors. Decision trees revealed distinct patterns for survival prediction at each time point. The models showed high accuracy (82.40%-89.09% in training group, and 82.16%-88.74% in test group) and discriminatory power (AUC: 0.806-0.894 in training group, and 0.808-0.882 in test group) in both training and testing datasets. An interactive web-based shiny APP (URL: https://yangxg1209.shinyapps.io/chondrosarcoma_survival_prediction/) was developed, simplifying the survival prediction process for clinicians. CONCLUSIONS: This study successfully employed RPA to develop a user-friendly tool for personalized survival predictions in CHS. The decision tree models demonstrated robust predictive capabilities, with the interactive application facilitating clinical decision-making. Future prospective studies are recommended to validate these findings and further refine the predictive model.

Recursive partitioning analysis for survival stratification and early imaging prediction of molecular biomarker in glioma patients.

BACKGROUND: Glioma is the most common primary brain tumor with high mortality and disability rates. Recent studies have highlighted the significant prognostic consequences of subtyping molecular pathological markers using tumor samples, such as IDH, 1p/19q, and TERT. However, the relative importance of individual markers or marker combinations in affecting patient survival remains unclear. Moreover, the high cost and reliance on postoperative tumor samples hinder the widespread use of these molecular markers in clinical practice, particularly during the preoperative period. We aim to identify the most prominent molecular biomarker combination that affects patient survival and develop a preoperative MRI-based predictive model and clinical scoring system for this combination. METHODS: A cohort dataset of 2,879 patients was compiled for survival risk stratification. In a subset of 238 patients, recursive partitioning analysis (RPA) was applied to create a survival subgroup framework based on molecular markers. We then collected MRI data and applied Visually Accessible Rembrandt Images (VASARI) features to construct predictive models and clinical scoring systems. RESULTS: The RPA delineated four survival groups primarily defined by the status of IDH and TERT mutations. Predictive models incorporating VASARI features and clinical data achieved AUC values of 0.85 for IDH and 0.82 for TERT mutations. Nomogram-based scoring systems were also formulated to facilitate clinical application. CONCLUSIONS: The combination of IDH-TERT mutation status alone can identify the most distinct survival differences in glioma patients. The predictive model based on preoperative MRI features, supported by clinical assessments, offers a reliable method for early molecular mutation prediction and constitutes a valuable scoring tool for clinicians in guiding treatment strategies.

Evaluation of a novel model incorporating serological indicators into the conventional TNM staging system for nasopharyngeal carcinoma.

BACKGROUND: Non-anatomical factors significantly affect treatment guidance and prognostic prediction in nasopharyngeal carcinoma (NPC) patients. Here, we developed a novel survival model by combining conventional TNM staging and serological indicators. METHODS: We retrospectively enrolled 10,914 eligible patients with nonmetastatic NPC over 2009-2017 and randomly divided them into training (n = 7672) and validation (n = 3242) cohorts. The new staging system was constructed based on T category, N category, and pretreatment serological markers by using recursive partitioning analysis (RPA). RESULTS: In multivariate Cox analysis, pretreatment cell-free Epstein-Barr virus (cfEBV) DNA levels of >2000 copies/mL [HROS (95 % CI) = 1.78 (1.57-2.02)], elevated lactate dehydrogenase (LDH) levels [HROS (95 % CI) = 1.64 (1.41-1.92)], and C-reactive protein-to-albumin ratio (CAR) of >0.04 [HROS (95 % CI) = 1.20 (1.07-1.34)] were associated with negative prognosis (all P < 0.05). Through RPA, we stratified patients into four risk groups: RPA I (n = 3209), RPA II (n = 2063), RPA III (n = 1263), and RPA IV (n = 1137), with 5-year overall survival (OS) rates of 93.2 %, 86.0 %, 80.6 %, and 71.9 % (all P < 0.001), respectively. Compared with the TNM staging system (eighth edition), RPA risk grouping demonstrated higher prognostic prediction efficacy in the training [area under the curve (AUC) = 0.661 vs. 0.631, P < 0.001] and validation (AUC = 0.687 vs. 0.654, P = 0.001) cohorts. Furthermore, our model could distinguish sensitive patients suitable for induction chemotherapy well. CONCLUSION: Our novel RPA staging model outperformed the current TNM staging system in prognostic prediction and clinical decision-making. We recommend incorporating cfEBV DNA, LDH, and CAR into the TNM staging system.

Glasgow prognostic score for assessing the efficacy of whole-brain radiation therapy in cases of recursive partitioning analysis class 2 and class 3 multiple brain metastases: a retrospective study.

PURPOSE: Whole-brain radiotherapy (WBRT) may not be beneficial for patients with brain metastases (BMs). The Glasgow Prognostic Score (GPS) is a suggested prognostic factor for malignancies. However, GPS has never been assessed in patients with BMs who have undergone WBRT. The purpose of this study was to determine whether GPS can be used to identify subgroups of patients with BMs who have a poor prognosis, such as recursive partitioning analysis (RPA) Class 2 and Class 3, and who will not receive clinical prognostic benefits from WBRT. MATERIALS AND METHODS: A total of 180 Japanese patients with BMs were treated with WBRT between May 2008 and October 2015. We examined GPS, age, Karnofsky Performance Status (KPS), RPA, graded prognostic assessment (GPA), number of lesions, tumor size, history of brain surgery, presence of clinical symptoms, and radiation doses. RESULTS: The overall median survival time (MST) was 6.1 months. seventeen patients (9.4%) were alive more than 2 years after WBRT. In univariate analysis, KPS ≤ 70 (p = 0.0066), GPA class 0-2 (p = 0.0008), > 3 BMs (p = 0.012), > 4 BMs (p = 0.02), patients who received ≥ 3 Gy per fraction (p = 0.0068), GPS ≥ 1 (p = 0.0003), and GPS ≥ 2 (p = 0.0009) were found to significantly decrease the MST. Patients who had brain surgery before WBRT (p = 0.036) had a longer survival. On multivariate analysis, GPS ≥ 1 (p = 0.008) was found to significantly decrease MST. CONCLUSION: Our results suggest that GPS ≥ 1 indicates a poor prognosis in patients undergoing WBRT for intermediate and poor prognosis BMs.

Development and Validation of a Predictive Model for Acute Kidney Injury in Sepsis Patients Based on Recursive Partition Analysis.

BACKGROUND: Sepsis-associated acute kidney injury (SA-AKI) is a critical condition with significant clinical implications, yet there is a need for a predictive model that can reliably assess the risk of its development. This study is undertaken to bridge a gap in healthcare by creating a predictive model for SA-AKI with the goal of empowering healthcare providers with a tool that can revolutionize patient care and ultimately lead to improved outcomes. METHODS: A cohort of 615 patients afflicted with sepsis, who were admitted to the intensive care unit, underwent random stratification into 2 groups: a training set (n = 435) and a validation set (n = 180). Subsequently, a multivariate logistic regression model, imbued with nonzero coefficients via LASSO regression, was meticulously devised for the prognostication of SA-AKI. This model was thoughtfully rendered in the form of a nomogram. The salience of individual risk factors was assessed and ranked employing Shapley Additive Interpretation (SHAP). Recursive partition analysis was performed to stratify the risk of patients with sepsis. RESULTS: Among the panoply of clinical variables examined, hypertension, diabetes mellitus, C-reactive protein, procalcitonin (PCT), activated partial thromboplastin time, and platelet count emerged as robust and independent determinants of SA-AKI. The receiver operating characteristic curve analysis for SA-AKI risk discrimination in both the training set and validation set yielded an area under the curve estimates of 0.843 (95% CI: 0.805 to 0.882) and 0.834 (95% CI: 0.775 to 0.893), respectively. Notably, PCT exhibited the most conspicuous influence on the model's predictive capacity. Furthermore, statistically significant disparities were observed in the incidence of SA-AKI and the 28-day survival rate across high-risk, medium-risk, and low-risk cohorts (P < .05). CONCLUSION: The composite predictive model, amalgamating the quintet of SA-AKI predictors, holds significant promise in facilitating the identification of high-risk patient subsets.

Clinical utility of preoperative pan-immune-inflammation value (PIV) for prognostication in patients with esophageal squamous cell carcinoma.

BACKGROUND: Several researches have shown that pan-immune-inflammation value (PIV) is related to cancer prognosis in recent years. In esophageal squamous cell carcinoma (ESCC), nevertheless, the prognostic impact of PIV remains unclear. The present study sought to investigate the prognostic impact of preoperative PIV in ESCC with radical resection. METHODS: The data of 294 ESCC patients who received radical resection were retrospectively analyzed. Based on analyzing the non-linear relationship between PIV and cancer-specific survival (CSS), the optimal cutoff value for PIV was calculated by the restricted cubic spline (RCS) model. Cox proportional hazards regression was carried out to identify the prognostic factors. A risk stratification model was established by recursive partitioning analysis (RPA). The performance of the RPA-based model was assessed by the decision curve analysis (DCA) and receiver operating characteristic (ROC). RESULTS: The RCS visualized the non-linear relationship between PIV and CSS (P < 0.0001). Then patients were then divided into high and low groups based on the optimal threshold of 308.2. The 5-year CSS (17.7 % vs. 48.3 %, P < 0.001) was significantly worse in patients with high PIV than those in the low group. Subgroup analyses confirmed that patients with low PIV also achieved better 5-year survival at different pathological tumor node metastasis (pTNM) stages (pTNM I: P = 0.022; pTNM II: P = 0.001; pTNM III: P = 0.011). PIV served as an independent prognostic factor of CSS (hazard ratio = 1.983, P < 0.001). A new staging involving three risk groups with significantly different CSS was developed using RPA algorithms based on pTNM and PIV. Compared with the pTNM classification, the RPA-based model exhibited significantly superior performance for prognostication. CONCLUSION: The present study confirmed the prognostic impact of PIV in ESCC who treated with radical resection. PIV was associated with the tumor stage and prognosis, which might be useful in the preoperative assessment of ESCC.

Construction of refined staging classification systems integrating FIGO/T-categories and corpus uterine invasion for non-metastatic cervical cancer.

BACKGROUND: To investigate the prognostic value of corpus uterine invasion (CUI) in cervical cancer (CC), and determine the necessity to incorporate it for staging. METHODS: A total of 809 cases of biopsy-proven, non-metastatic CC were identified from an academic cancer center. Recursive partitioning analysis (RPA) method was used to develop the refined staging systems with respect to overall survival (OS). Internal validation was performed by using calibration curve with 1000 bootstrap resampling. Performances of the RPA-refined stages were compared against the conventional FIGO 2018 and 9th edition TNM-stage classifications by the receiver operating characteristic curve (ROC) and decision curve analysis (DCA). RESULTS: We identified that CUI was independently prognostic for death and relapse in our cohort. RPA modeling using a two-tiered stratification by CUI (positive and negative) and FIGO/T-categories divided CC into three risk groupings (FIGO I'-III'/T1'-3'), with 5-year OS of 90.8%, 82.1%, and 68.5% for proposed FIGO stage I'-III', respectively (p ≤ 0.003 for all pairwise comparisons), and 89.7%, 78.8%, and 68.0% for proposed T1'-3', respectively (p < 0.001 for all pairwise comparisons). The RPA-refined staging systems were well validated with RPA-predicted OS rates showed optimal agreement with actual observed survivals. Additionally, the RPA-refined stages outperformed the conventional FIGO/TNM-stage with significantly higher accuracy of survival prediction (AUC: RPA-FIGO vs. FIGO, 0.663 [95% CI 0.629-0.695] vs. 0.638 [0.604-0.671], p = 0.047; RPA-T vs. T, 0.661 [0.627-0.694] vs. 0.627 [0.592-0.660], p = 0.036). CONCLUSION: CUI affects the survival outcomes in patients with CC. Disease extended to corpus uterine should be classified as stage III/T3.

Relevance of the Updated Recursive Partitioning Analysis (U-RPA) Classification in the Contemporary Care of Patients with Brain Metastases.

Patients with brain metastases (BMETS) need information about the prognosis and potential value of treatment options to make informed therapeutic decisions, but tools to predict survival in contemporary practice are scarce. We propose an Updated Recursive Partitioning Analysis (U-RPA) instrument to predict survival and benefit from brain-directed treatment (BDT) of contemporary patients. This was a retrospective analysis of patients with BMETS treated between 2017 and 2019. With survival as the primary endpoint, we calculated the U-RPA and generated estimates using Kaplan-Meier curves and hazard ratios. Of 862 eligible patients, 752 received BDT and 110 received best supportive care (BSC). Median overall survival with BDT and BSC was 9.3 and 1.3 months, respectively. Patients in RPA class 1, 2A, 2B and 3 who underwent BDT had median survival of 28.1, 14.7, 7.6 and 3.3 months, respectively. The median survival for patients in RPA 3 who received BDT (n = 147), WBRT (n = 79) and SRS (n = 54) was 3.3, 2.9 and 4.1 months, respectively. The U-RPA defines prognosis estimates, independent of tumor type and treatment modality, which can assist to make value-based care treatment decisions. The prognosis for patients in U-RPA class 2B and 3 remains poor, with consideration for early palliative care involvement in these cases.

Prognostic model on overall survival in elderly nasopharyngeal carcinoma patients: a recursive partitioning analysis identifying pre-treatment risk stratification.

BACKGROUND: We aimed to evaluate the optimal management for elderly patients with nasopharyngeal carcinoma (NPC) with intensity-modulated radiotherapy (IMRT). METHODS: A total of 283 elderly patients with NPC diagnosed from 2015 to 2019 were enrolled in the study. Overall survival (OS) was the primary endpoint. Univariate and multivariate Cox regression analyses were preformed to identify potential prognostic factors. The recursive partitioning analysis (RPA) was used for risk stratification. Kaplan-Meier survival curves were applied to evaluate the survival endpoints, and log-rank test was utilized to assess differences between groups. The prognostic index (PI) was constructed to further predict patients' prognosis displayed by nomogram model. The area under the receiver operating characteristic (ROC) curves (AUC) and the calibration curves were applied to assess the effectiveness of the model. RESULTS: Based on RPA-based risk stratification, we demonstrated that elderly NPC patients who were treated with IC followed by RT had similar OS as those with induction chemotherapy (IC) combined with concurrent chemoradiotherapy (CCRT) in the middle- (stage I-III and pre-treatment EBV > 1840 copies/ml) and high-risk groups (stage IVA). IMRT alone may be the optimal treatment option for the low-risk group (stage I-III with pre-treatment EBV ≤ 1840 copies/ml). We established an integrated PI which was indicted with stronger prognostic power than each of the factors alone for elderly NPC patients (The AUC of PI was 0.75, 0.80, and 0.82 for 1-, 3-, 5-year prediction of OS, respectively). CONCLUSION: We present a robust model for clinical stratification which could guide individual therapy for elderly NPC patients.

Clinical utility of pretreatment serum squamous cell carcinoma antigen for prognostication and decision-making in patients with early-stage cervical cancer.

Background: To investigate the prognostic role of pretreatment squamous cell carcinoma antigen (SCCA) in early-stage cervical cancer (CC). Methods: We enrolled 487 cases of pathology-proven early-stage [International Federation of Gynecology and Obstetrics (FIGO) I/II] squamous or adenosquamous CC that were treated from 2012 to 2015. Restricted cubic splines (RCS) with a full Cox regression model were used to evaluate the association between SCCA levels and survival outcomes. Recursive partitioning analysis (RPA) was used to construct a risk stratification model for overall survival (OS). The performance of the RPA-based model was assessed using a receiver operating characteristic (ROC) curve. Results: RCS analysis revealed an association between SCCA and OS and disease-free survival (DFS); SCCA ⩾2.5 ng/mL was robust for risk discrimination in our cohort. SCCA had an interaction effect with FIGO classification: Patients with FIGO I and SCCA ⩾2.5 ng/mL overlapped with those with FIGO II and SCCA < 2.5 ng/mL for OS [hazard ratio, 1.04 (95% confidence interval (CI): 0.49-2.24), p = 0.903] and DFS [1.05 (0.56-1.98), p = 0.876]. RPA modeling incorporating SCCA (<2.5 ng/mL and ⩾2.5 ng/mL) and FIGO classification divided CC into three prognostic groups: RPA I, FIGO stage I, and SCCA < 2.5 ng/mL; RPA II, FIGO stage I, and SCCA ⩾ 2.5 ng/mL, or FIGO stage II and SCCA < 2.5 ng/mL; and RPA III, FIGO stage II, and SCCA ⩾ 2.5 ng/mL; with 5-year OS of 94.0%, 85.1%, and 73.5%, respectively (p < 0.001). ROC analysis confirmed that the RPA model outperformed the FIGO 2018 stage with significantly improved accuracy for survival prediction [area under the curve: RPA versus FIGO, 0.663 (95% CI: 0.619-0.705] versus 0.621 (0.576-0.664), p = 0.045]. Importantly, the RPA groupings were associated with the efficacy of treatment regimens. Surgery followed by adjuvant treatment had a higher OS (p < 0.01) and DFS (p = 0.024) than other treatments for RPA III, whereas outcomes were comparable among treatment regimens for RPA I-II. Conclusion: Herein, the role of SCCA for prognostication was confirmed, and a robust clinicomolecular risk stratification system that outperforms conventional FIGO classification in early-stage squamous and adenosquamous CC was presented. The model correlated with the efficacy of different treatment regimes.

Identification of Atezolizumab Plus Bevacizumab Prognostic Index via Recursive Partitioning Analysis in HCC: The ABE Index.

BACKGROUND/AIM: The purpose of this study was to ascertain a novel prognostic index via recursive partitioning analysis (RPA) in hepatocellular carcinoma (HCC) patients being treated with the combination of atezolizumab plus bevacizumab (ABE) in first-line setting. PATIENTS AND METHODS: A total of 784 patients with HCC were included in the analysis. RESULTS: RPA identified three groups of patients: high-risk [Child-Pugh B (CP-B) patients; CP-A and Albumin-Bilirubin (ALBI)-2 patients; CP-A and ALBI-1 patients with macrovascular invasion (MVI), and alpha-fetoprotein (α-FP) ≥400 ng/ml]; intermediate-risk [CP-A and ALBI-1 patients with aspartate aminotransferase (AST) normal value (NV), and αFP ≥400 ng/ml, but without MVI; CP-A and ALBI-1 patients with AST increased value (IV), and neutrophil-lymphocyte ratio (NLR) ≥3, but without MVI]; low-risk (CP-A and ALBI-1 patients with AST NV, and αFP <400 ng/ml, but without MVI; CP-A and ALBI-1 patients with AST IV, and NLR <3, but without MVI; CP-A and ALBI-1 patients with MVI, and αFP <400 ng/ml). Overall survival was 7.0 months in high-risk patients (20.8%), 14.2 months in intermediate-risk patients (19.1%), and 22.5 months in low-risk patients (60.1%). CONCLUSION: The ABE index allows for easy stratification of HCC patients treated with the combination of ABE in first-line setting.

Outcomes of single brain metastasis treated with gamma knife stereotaxic radiosurgery(GKSR). Our experience on 103 cases.

Background: Brain metastases (BM) occur in the natural course of malignant tumors in 18-40% of cases. Their management has changed considerably over the past decade thanks to the advent of Gamma knife Stereotactic Radiosurgery (GKSR). Objective: We report our experience on Single Brain metastasis treated with (GKSR). Methods: Patients treated by Gamma Knife stereotaxic radiosurgery (GKSR) in our institution between 2009 and 2021 for Single BM were recorded retrospectively. Results: A total of 103 patients (n = 52; 50.5% females) were included, with a mean age of 56.33 ± 11.33. Breast (n = 39, 37.9%) and lung (n = 36, 35%) were the common original location for the primary tumors. GKSR alone without prior surgery, radiotherapy, or chemotherapy was achieved in 81.5% (n = 84). Thirteen patients (15.1%) progressed in BM volume while finding the appearance of de novo BM in 5 (5.8%) patients. The median percentage of tumor control after radiosurgery treatment was 70% (IQR: 65-78) and only 26.2% (n = 27) of patients had > 80% tumor control and stability over the median follow-up time of 5 (95% CI, 4-6) months. We found only two cases of radionecrosis (1.9%). The median survival time was 5.21 (IQR, 3-8) months. Retreatment, recursive partitioning analysis (RPA) class, and tumor stability influenced the overall survival of BM respectively (Hazard Ratio adjust (HRa)= 5.610,p = 0.045; HRa= 6.133,p = 0.031; HRa= 22.463, p = 0.036). Conclusion: Stereotaxic Radiosurgery provides good results in terms of Overall survival with fewer neurocognitive disorders.RPA class and tumor control (stability) influenced the overall survival of single BM.

Establishment of a risk stratification model based on the combination of post-treatment serum squamous cell carcinoma antigen levels and FIGO stage of cervical cancer for treatment and surveillance decision-making.

OBJECTIVE: To develop a risk stratification model based on the International Federation of Gynecology and Obstetrics (FIGO) staging combined with squamous cell carcinoma antigen (SCC-Ag) for the classification of patients with cervical squamous cell carcinoma (CSCC) into different risk groups. METHODS: We retrospectively reviewed the data of 664 women with stage IIA-IVB CSCC according to the 2018 FIGO staging system who received definitive radiotherapy from March 2013 to December 2017 at the department of radiation oncology of Sun Yat-sen University Cancer Center. Cutoff values for continuous variables were estimated using receiver operating characteristic curve analysis. Using recursive partitioning analysis (RPA) modeling, overall survival was predicted based on the prognostic factors determined via Cox regression analysis. The predictive performance of the RPA model was assessed using the consistency index (C-index). Intergroup survival differences were determined and compared using Kaplan-Meier analysis and the log-rank test. RESULTS: Multivariate Cox regression analysis identified post-treatment SCC-Ag (< 1.35 ng/mL and > 1.35 ng/mL; hazard ratio (HR), 4.000; 95% confidence interval (CI), 2.911-5.496; P < 0.0001) and FIGO stage (II, III, and IV; HR, 2.582, 95% CI, 1.947-3.426; P < 0.0001) as the independent outcome predictors for overall survival. The RPA model based on the above prognostic factors divided the patients into high-, intermediate-, and low-risk groups. Significant differences in overall survival were observed among the three groups (5-year overall survival: low vs. intermediate vs. high, 91.3% vs. 76.7% vs. 29.5%, P < 0.0001). The predictive performance of the RPA model (C-index, 0.732; 95% CI, 0.701-0.763) was prominently superior to that of post-treatment SCC-Ag (C-index, 0.668; 95% CI, 0.635-0.702; P < 0.0001) and FIGO stage (C-index, 0.663; 95% CI, 0.631-0.695; P < 0.0001). CONCLUSIONS: The RPA model based on FIGO staging and post-treatment SCC-Ag can predict the overall survival of patients with CSCC, thereby providing a guide for the formulation of risk-adaptive treatment and individualized follow-up strategies.

Proposed prognostic subgroups and facilitated clinical decision-making for additional locoregional radiotherapy in de novo metastatic nasopharyngeal carcinoma: a retrospective study based on recursive partitioning analysis.

BACKGROUND: The high heterogeneity of de novo metastatic nasopharyngeal carcinoma (dmNPC) makes its prognosis and treatment challenging. We aimed to accurately stage dmNPC and assess the patterns of treatment strategies for different risk groups. METHODS: The study enrolled a total of 562 patients, 264 from 2007 to 2013 in the training cohort and 298 from 2014 to 2017 in the validation cohort. Univariate and multivariate Cox regression analyses were conducted to determine the independent variables for overall survival (OS). Recursive partitioning analysis (RPA) was applied to establish a novel risk-stratifying model based on these variables. RESULTS: After pairwise comparisons of OS, three risk groups were generated: low-risk (involved lesions ≤ 4 without liver involvement), intermediate-risk (involved lesions ≤ 4 with liver involvement or involved lesions > 4 with Epstein-Barr virus (EBV)-DNA < 62,000 copies/ml), and high-risk (involved lesions > 4 with EBV-DNA > 62,000 copies/ml). The 3-year OS rate differed significantly between groups (80.4%, 42.0%, and 20.4%, respectively, all P < 0.05). Adding locoregional intensity-modulated radiotherapy (LRRT) followed by palliative chemotherapy (PCT) resulted in a significant OS benefit over PCT alone for the low- and intermediate-risk groups (P = 0.0032 and P = 0.0014, respectively). However, it provided no survival benefits for the high-risk group (P = 0.6). Patients did not benefit from concurrent chemotherapy during LRRT among the three subgroups (P = 0.12, P = 0.13, and P = 0.3, respectively). These results were confirmed with the validation cohort. CONCLUSIONS: The novel RPA model revealed superior survival performance in subgroup stratification and could facilitate more effective treatment strategies for dmNPC.

A modified mTNM staging system based on lymph node ratio for colon neuroendocrine tumors: A recursive partitioning analysis.

Background: In the current tumor-lymph node-metastasis (TNM) staging system for colon neuroendocrine tumors, lymph node status is divided into N1 and N0. An assessment of the lymph node ratio (LNR) and a proposal for a modified mTNM staging system were the objectives of this study. Methods: Selecting the optimal cut-off value of LNR was done using X-tile. A Cox regression model and the Kaplan-Meier method were performed to calculate patient cancer-specific survival in the Surveillance, Epidemiology and End Results cohort. Recursive partitioning analysis was used to improve TNM staging. Results: The study included 674 patients. The current TNM staging system showed inadequate discriminatory power between stage I and stage II patients (p = 0.088). The optimal cut-off value was determined as 0.6 for LNR. Based on multivariate Cox regression analysis, the modified mN classification could be classified into mN 0 (LNR = 0.00), mN 1 (LNR = 0.01-0.60), and mN 2 (LNR > 0.60), and was found to be an independent factor affecting prognosis (p < 0.001). Using the American Joint Committee on Cancer T and modified mN classifications, the modified mTNM system was constructed, and it exhibited better prognostic discriminatory power ability than the traditional TNM system (C-index: 0.587 vs. 0.665). Conclusions: Our study determined that LNR is a prognostic factor in colon NET patients. In addition, to more accurately assess the prognosis of colon NET patients, we proposed a modified mTNM staging system.

Selection and validation of chemotherapy beneficiaries among elderly nasopharyngeal carcinoma (NPC) patients treated with intensity-modulated radiation therapy (IMRT): a large real-world study.

PURPOSE: Using real-world evidence, this study aimed to identify elderly nasopharyngeal carcinoma (NPC) patients who would benefit from chemotherapy. METHODS AND MATERIALS: 1714 elderly NPC patients between April 2007 and December 2017 were identified. Recursive partitioning analysis (RPA) was used to generate risk-stratified outcomes. Prognostic factors were performed for individual comparisons of different risk groups to assess chemotherapy benefits. RESULTS: The median follow-up was 59.3 (0.39-170.09) months. Epstein Barr virus (EBV) DNA and T stage were included in the RPA-generated risk stratification, categorizing patients into a good-prognosis group (EBV DNA ≤ 4000 copies/mL & T1-2), and a poor-prognosis group (EBV DNA ≤ 4000 copies/mL & T3-4 and EBV DNA > 4000 copies/mL & any T). Overall survival (OS) was significantly higher in the good-prognosis group compared with the training set (HR = 0.309, 95% CI 0.184-0.517; P < 0.001), and validated in the testing set (HR = 0.276, 95% CI 0.113-0.670; P = 0.002). In the poor-prognosis group, a significantly improved OS for chemoradiotherapy (CRT) compared with RT alone was observed (HR = 0.70, 95% CI 0.55-0.88; P = 0.003). Patients who received induction chemotherapy (IC) + concurrent chemoradiotherapy (CCRT) and CCRT had a significantly improved OS compared with RT alone (IC + CCRT vs. RT alone: P = 0.002; CCRT vs. RT alone: P = 0.008) but not in the IC + RT group (P = 0.306). The 5-year OS for CRT versus RT-alone with ACE-27 scores of 0, 1 and 2 were 76.0% versus 70.0% (P = 0.014), 80.5% versus 68.2% (P = 0.150) and 58.5% versus 62.2% (P = 0.490), respectively; for those aged 60-64, 65-70 and ≥ 70 years old they were 80.9% versus 75.9% (P = 0.068), 73.3% versus 63.4% (P = 0.270) and 64.8% versus 67.1% (P = 0.820), respectively. CONCLUSIONS: For elderly NPC patients a simple screening cutoff for chemotherapy beneficiaries might be EBV DNA < 4000 copies/ml & T3-4 and EBV DNA ≥ 4000 copies/ml & any T, but not for those > 70 years old and with an ACE-27 score > 1. IC + CCRT and CCRT were effective forms of chemotherapy.

Is postoperative radiotherapy an essential treatment for nonmetastatic lymphoepithelial carcinoma of the salivary gland?

BACKGROUND: Lymphoepithelial carcinoma of salivary gland (LECSG) is a rare malignant tumor. Whether postoperative radiotherapy (PORT) can improve locoregional control and which patients can benefit from PORT are unknown. This study aimed to evaluate the role of PORT and provide individualized suggestions for postoperative therapy in patients with LECSG. PATIENTS AND METHODS: We retrospectively reviewed patients with nonmetastatic LECSG who underwent surgery with or without PORT. Recursive partitioning analysis (RPA) was performed to categorize the patients and predict progression-free survival (PFS). RESULTS: A total of 223 patients were included, 34 (15.2%) received surgery alone, whereas the remaining 189 (84.8%) underwent PORT in the initial treatment. Although patients in the PORT group were with advanced T stage and N stage, the PORT group had an advantage over the non-PORT group on 1-year, 5-year and 10-year PFS and locoregional control (LRC). PORT was an independent prognostic factor for PFS and LRC. Furthermore, compared with T stage and N stage, the size of the primary tumor and the number of positive lymph nodes were better prognostic predictors. The RPA model was generated according to the endpoint of PFS and categorized patients into 3 prognostic groups: low-risk (maximum diameter of primary lesion (≤3 cm) and number of positive lymph nodes (≤2)), intermediate-risk (maximum diameter of primary lesion (>3 cm) and number of positive lymph nodes (≤2)), and high-risk (number of positive lymph nodes (>2)), with corresponding 5-year PFS rates of 90.0%, 75.0%, and 51.0%, respectively. Significant improvement in PFS was observed in the PORT group among intermediate-risk (P = 0.000) and high-risk patients (P = 0.000). CONCLUSIONS: PORT was shown to be a positive prognostic factor for PFS and LRC of LECSG. PORT was an essential treatment especially for patients with >3 cm maximum diameter of primary lesion and/or >2 positive lymph nodes.

An Exploratory Study of Refining TNM-8 M1 Categories and Prognostic Subgroups Using Plasma EBV DNA for Previously Untreated De Novo Metastatic Nasopharyngeal Carcinoma.

(1) Background: NPC patients with de novo distant metastasis appears to be a heterogeneous group who demonstrate a wide range of survival, as suggested by growing evidence. Nevertheless, the current 8th edition of TNM staging (TNM-8) grouping all these patients into the M1 category is not able to identify their survival differences. We sought to identify any anatomic and non-anatomic subgroups in this study. (2) Methods: Sixty-nine patients with treatment-naive de novo M1 NPC (training cohort) were prospectively recruited from 2007 to 2018. We performed univariable and multivariable analyses (UVA and MVA) to explore anatomic distant metastasis factors, which were significantly prognostic of overall survival (OS). Recursive partitioning analysis (RPA) with the incorporation of significant factors from MVA was then performed to derive a new set of RPA stage groups with OS segregation (Set 1 Anatomic-RPA stage groups); another run of MVA was performed with the addition of pre-treatment plasma EBV DNA. A second-round RPA with significant prognostic factors of OS identified in this round of MVA was performed again to derive another set of stage groups (Set 2 Prognostic-RPA stage groups). Both sets were then validated externally with an independent validation cohort of 67 patients with distant relapses of their initially non-metastatic NPC (rM1) after radical treatment. The performance of models in survival segregation was evaluated by the Akaike information criterion (AIC) and concordance index (C-index) under 1000 bootstrapping samples for the validation cohort; (3) Results: The 3-year OS and median follow-up in the training cohort were 36.0% and 17.8 months, respectively. Co-existence of liver-bone metastases was the only significant prognostic factor of OS in the first round UVA and MVA. Set 1 RPA based on anatomic factors that subdivide the M1 category into two groups: M1a (absence of co-existing liver-bone metastases; median OS 28.1 months) and M1b (co-existing liver-bone metastases; median OS 19.2 months, p = 0.023). When pre-treatment plasma EBV DNA was also added, it became the only significant prognostic factor in UVA (p = 0.001) and MVA (p = 0.015), while co-existing liver-bone metastases was only significant in UVA. Set 2 RPA with the incorporation of pre-treatment plasma EBV DNA yielded good segregation (M1a: EBV DNA ≤ 2500 copies/mL and M1b: EBV DNA > 2500 copies/mL; median OS 44.2 and 19.7 months, respectively, p < 0.001). Set 2 Prognostic-RPA groups (AIC: 228.1 [95% CI: 194.8−251.8] is superior to Set 1 Anatomic-RPA groups (AIC: 278.5 [254.6−301.2]) in the OS prediction (p < 0.001). Set 2 RPA groups (C-index 0.59 [95% CI: 0.54−0.67]) also performed better prediction agreement in the validation cohort (vs. Set 1: C-index 0.47 [95% CI: 0.41−0.53]) (p < 0.001); (4) Conclusions: Our Anatomic-RPA stage groups yielded good segregation for de novo M1 NPC, and prognostication was further improved by incorporating plasma EBV DNA. These new RPA stage groups for M1 NPC can be applied to countries/regions regardless of whether reliable and sensitive plasma EBV DNA assays are available or not.

The number of cycles of adjuvant chemotherapy in stage III and high-risk stage II rectal cancer: a nomogram and recursive partitioning analysis.

OBJECTIVE: The prognostic role of the number of cycles of adjuvant chemotherapy (ACT) after total mesorectal excision in stage III and high-risk stage II rectal cancer is unknown. As a result of this, our study was designed to assess the effect of the number of cycles of ACT on the prediction of cancer-specific survival. METHODS: Four hundred patients that were diagnosed as stage III and high-risk stage II rectal cancer from January 2012 to January 2018 and who had received total mesorectal excision were enrolled in this study. A nomogram incorporating the number of cycles of ACT was also developed in this study. For internal validation, the bootstrap method was used and the consistency index was used to evaluate the accuracy of the model. The patients were stratified into risk groups according to their tumor characteristics by recursive partitioning analysis. RESULTS: We found that the risk of death was decreased by 26% (HR = 0.74, 95% CI: 0.61-0.89, P = 0.0016) with each increasing ACT cycle. The N stage, positive lymph node ratio (PLNR), carcinoembryonic antigen, neutrophil-to-lymphocyte ratio, and the number of cycles of ACT were chosen and entered into the nomogram model. Recursive partitioning analysis-based risk stratification revealed a significant difference in the prognosis in rectal cancer patients with high-risk, intermediate-risk, and low-risk (3-year cancer-specific survival: 0.246 vs. 0.795 vs. 0.968, P < 0.0001). Seven or more cycles of ACT yielded better survival in patients with PLNR ≥ 0.28 but not in patients with PLNR < 0.28. CONCLUSION: In conclusion, the nomogram prognosis model based on the number of cycles of ACT predicted individual prognosis in rectal cancer patients who had undergone total mesorectal excision. These findings further showed that in patients with PLNR ≥ 0.28, no fewer than 7 cycles of ACT are needed to significantly reduce the patient's risk of death.

Sensory Capacities and Eating Behavior: Intriguing Results from a Large Cohort of Italian Individuals.

Eating behavior (EB) is a complex system influenced by many factors, but an undisputed role is played by the senses. In this work, we examined the effect of the sensory capacities on EB in 1152 Italian adult individuals. After administering a questionnaire on EB and assessing sensory performance through standard audiometric, olfactory, and taste tests, the prevalence of reduced sensory capacities (RSCs) and the correlation with selected risk factors were calculated. Regression models, structural equation modelling, and conditional recursive partitioning were used to investigate the relationship between variables. Around 70% of the subjects show reduced capacities in at least one sense, with taste being the most prevalent (55.21%). Male sex, aging, and low educational level are risk factors for RSCs. The increased number of senses with reduced capacities is a predictor of diminished food adventurousness and lower liking for vegetables, fish, and alcoholic beverages, while reduced capacities (RCs) in taste is a predictor of lower liking for alcoholic beverages and sweets. Overall, in addition to providing an overall picture of RSCs in Italian samples, our study reveals the association of RSCs with EB variables. This finding could have a relevant role in influencing individuals' dietary habits and, therefore, health status.