Recursive Partitioning Research Articles

DecTree: a physics-based geochemical surrogate for surface complexation of uranium on clay

Abstract. Geochemistry is usually the computational bottleneck in coupled reactive transport simulations, which hampers the complexity of the systems and of the processes they can investigate. In recent years, promising speedups have been obtained by substituting the numerical solution of geochemical models with approximated surrogates borrowed from artificial intelligence and machine learning (AI/ML). In the framework of the DONUT/EURAD project a set of benchmarks were defined to assess the performance and the accuracy of different surrogate approaches in settings relevant to the safety assessment of nuclear waste repositories, such as the surface complexation and exchange of U(VI) on clay. In this context, this work introduces am original surrogate modelling approach based on recursive partitioning of parameter space, which exploits prior domain knowledge for the training. The surrogate, which can be represented as a decision tree, hence the DecTree name, performs dimensionality reduction by identifying functional relationships between outputs and input variables using a straightforward non-monotonic extension of the Spearman's rank correlation coefficient. New predictions are then interpolated from the partitioned training data. Applied to a low-dimensional geochemical model, DecTree shows virtually no training time and excellent accuracy, ensuring a throughput of around 500 000 predictions per second on a single CPU core.

SURG-21. DEVELOPMENT AND VALIDATION OF A CLINICAL RISK SCORE FOR POSTOPERATIVE OUTCOME IN NEWLY DIAGNOSED GLIOBLASTOMA: A REPORT OF THE RANORESECT GROUP

Abstract BACKGROUND Following resection or biopsy, patients with newly diagnosed glioblastoma frequently enter clinical trials. Nuanced risk assessment is warranted to reduce imbalances between study arms. Here, we aim to make use of our RANO classification for extent of resection (I) to analyze the interactive effects of residual tumor with clinical and molecular factors on outcome and (II) to define a postoperative risk assessment tool. METHODS The RANO resect group retrospectively compiled an international, seven-center training cohort of patients with newly diagnosed glioblastoma. The combined associations of residual tumor with molecular or clinical factors and survival were analyzed, and recursive partitioning analysis was performed for risk modeling. The resulting model was prognostically verified in a separate external validation cohort. RESULTS Our training cohort compromised 1003 patients with newly diagnosed IDH-wildtype glioblastoma, including 744 patients who underwent adjuvant radiochemotherapy (TMZ/RT→TMZ). Residual tumor, MGMT promotor methylation status, age, and KPS were prognostic of overall survival and forwarded into regression tree analysis. By combining eleven terminal nodes and individually weighting the prognostic factors, an additive scale (range, 0-9 points) integrating these four variables distinguished patients with low (0-2 points), intermediate (3-5 points), and high risk (6-9 points) for poor postoperative outcome (median overall survival: 24 vs. 16 vs. 6 months; p<0.01). The presence of the three risk groups was confirmed in the subgroups of patients treated with or without RT/TMZ→TMZ. The prognostic value of the risk score was verified in a external single-center validation cohort of newly diagnosed glioblastoma (n = 258, p<0.01). Compared to previously postulated models, goodness-of-fit measurements were superior for our score. CONCLUSIONS The novel RANO risk score serves as an easy-to-use, yet highly prognostic tool for postoperative patient stratification prior to further therapy. The score may serve to guide patient management and reduce imbalances between study arms in prospective trials.

EPID-19. INTERACTION BETWEEN GERMLINE AND SOMATIC GENETICS TO ELUCIDATE RISK OF ADULT DIFFUSE GLIOMA

Abstract Acquired tumor alterations are part of the WHO diagnosis of glioma. However, germline predisposition to these alterations is unknown. GWAS in European and Asian populations identified germline variants in 30 regions, many in or near genes that also acquire alterations, suggesting that interactions between inherited and acquired genetics are involved in glioma development. To evaluate such interactions, we used 3686 glioma cases and 1889 controls. A case-control design and multinomial logistic regression were used to identify germline variants associated with development of TERT-promoter mutant gliomas, generating two interesting observations. First, TERT germline variants were associated with predisposition to IDHwt glioma that have a TERT promoter mutation, but not with IDH-mutant codeleted gliomas that have a TERT promoter mutation. Second, two novel regions were observed in this more homogeneous subgroup of TERT mutated tumors: variants in TXNDC11 (OR=2.66, p=5.7x10-8) and RUVBL2 (OR=3.1, p=5.8x10-8). High tumor expression of TXNDC11 is associated with poor prognosis in glioma and RUVBL2 interacts with TERT to affect telomerase activity. Similar analyses are being performed on additional clinically relevant acquired alterations. In parallel, case-case GWAS analyses were performed to evaluate germline variants associated with other tumor alterations. Comparing IDH-mutant versus IDHwt stratified by rs55705857 genotype, the most significant variants were PHLDB1 (p=1.2x10-13) and D2HGDH (p=2.1x10-10), highlighting the importance of these two variants independent of CCDC26 for the development of IDH-mutant glioma. As a parallel approach, we utilized recursive partitioning and regression trees (rpart) to classify IDH mutation status using the 30 known variants. The top four branches of the regression tree included CCDC26, D2HGDH, PHLDB1, and MAML2, as well as two Asian variants (RAB27A, CYP4F12). This highlights the importance of utilizing variants discovered in non-European populations in analyses. Overall, understanding these interactions are necessary to develop therapeutics, stratify clinical trials, and to develop relevant tumor models.

Development and validation of a clinical risk model for postoperative outcome in newly diagnosed glioblastoma: a report of the RANO resect group.

Following surgery, patients with newly diagnosed glioblastoma frequently enter clinical trials. Nuanced risk assessment is warranted to reduce imbalances between study arms. Here, we aimed (I) to analyze the interactive effects of residual tumor with clinical and molecular factors on outcome and (II) to define a postoperative risk assessment tool. The RANO resect group retrospectively compiled an international, seven-center training cohort of patients with newly diagnosed glioblastoma. The combined associations of residual tumor with molecular or clinical factors and survival were analyzed, and recursive partitioning analysis was performed for risk modeling. The resulting model was prognostically verified in a separate external validation cohort. Our training cohort compromised 1003 patients with newly diagnosed isocitrate dehydrogenase-wildtype glioblastoma. Residual tumor, O6-methylguanine DNA methyltransferase (MGMT) promotor methylation status, age, and postoperative KPS were prognostic for survival and incorporated into regression tree analysis. By individually weighting the prognostic factors, an additive score (range, 0-9 points) integrating these four variables distinguished patients with low (0-2 points), intermediate (3-5 points), and high risk (6-9 points) for inferior survival. The prognostic value of our risk model was retained in treatment-based subgroups and confirmed in an external validation cohort of 258 patients with glioblastoma. Compared to previously postulated models, goodness-of-fit measurements were superior for our model. The novel RANO risk model serves as an easy-to-use, yet highly prognostic tool for postoperative patient stratification prior to further therapy. The model may serve to guide patient management and reduce imbalances between study arms in prospective trials.

Development of an ultrasound-based clinical decision rule to rule-out diverticulitis

The concern for diverticulitis often leads to the use of computed tomography (CT) scans for diagnosis. We aim to develop an ultrasound-based clinical decision rule (CDR) to confidently rule-out the disease without requiring a CT scan. We analyzed data from a prospective study of adult emergency department (ED) patients with suspected diverticulitis who underwent both bedside ultrasound (US) and CT. Patient history, physical examination, laboratory findings, and US results were used to create a CDR via a recursive partitioning model designed to prioritize sensitivity, with a loss matrix heavily penalizing false negatives. We calculated the test characteristics for this CDR (TICS-Rule) and assessed the potential reduction in CT scans and ED length of stay. Data from 149 patients (84 female; mean age 58 ± 16) were used to develop the TICS-Rule. The final model integrates US diagnosis of simple and complicated diverticulitis, along with variables of heart rate, age, history of diverticulosis, vomiting, and leukocytosis. Negative US results and a heart rate below 100 effectively excluded diverticulitis. The sensitivity increased from 54.5% (32.2–75.6) in the US alone to 100% (84.6–100%) for complicated diverticulitis in the model. The TICS-Rule missed no cases of complicated diverticulitis but one case of simple diverticulitis. The median time from ED greeting to US interpretation was 103 min (IQR 62–169), compared to 285 min (IQR 229–372) for CT. The TICS-Rule uses a combination of negative US and heart rate less thanQ1 100 to exclude diverticulitis without the need for a CT scan. Integration of the TICS-Rule offers a promising enhancement to clinical decision-making while reducing both CT use and ED length of stay.

Establishment of a Novel Risk Stratification System Integrating Clinical and Pathological Parameters for Prognostication and Clinical Decision‐Making in Early‐Stage Cervical Cancer

ABSTRACTBackgroundHighly heterogeneity and inconsistency in terms of prognosis are widely identified for early‐stage cervical cancer (esCC). Herein, we aim to investigate for an intuitional risk stratification model for better prognostication and decision‐making in combination with clinical and pathological variables.MethodsWe enrolled 2071 CC patients with preoperative biopsy‐confirmed and clinically diagnosed with FIGO stage IA‐IIA who received radical hysterectomy from 2013 to 2018. Patients were randomly assigned to the training set (n = 1450) and internal validation set (n = 621), in a ratio of 7:3. We used recursive partitioning analysis (RPA) to develop a risk stratification model and assessed the ability of discrimination and calibration of the RPA‐derived model. The performances of the model were compared with the conventional FIGO 2018 and 9th edition T or N stage classifications.ResultsRPA divided patients into four risk groups with distinct survival: 5‐year OS for RPA I to IV were 98%, 95%, 85.5%, and 64.2%, respectively, in training cohort; and 99.5%, 93.2%, 85%, and 68.3% in internal validation cohort (log‐rank p < 0.001). Calibration curves confirmed that the RPA‐predicted survivals were in good agreement with the actual survivals. The RPA model outperformed the existing staging systems, with highest AUC for OS (training: 0.778 vs. 0.6–0.717; internal validation: 0.772 vs. 0.595–0.704; all p < 0.05), and C‐index for OS (training: 0.768 vs. 0.598–0.707; internal validation: 0.741 vs. 0.583–0.676; all p < 0.05). Importantly, there were associations between RPA groups and the efficacy of treatment regimens. No obvious discrepancy was observed among different treatment modalities in RPA I (p = 0.922), whereas significant survival improvements were identified in patients who received adjuvant chemoradiotherapy in RPA II–IV (p value were 0.028, 0.036, and 0.024, respectively).ConclusionWe presented a validated novel clinicopathological risk stratification signature for robust prognostication of esCC, which may be used for streamlining treatment strategies.

Brief Report: Alternative Pulmonary Function Measures of Emphysema in People With HIV.

People with HIV (PWH) have nearly twice the risk of emphysema than people without HIV. This risk, which has been associated with HIV-mediated changes in the lung immune environment and more extensive radiographic emphysema, may result in different patterns of airflow limitation on pulmonary function testing (PFT) than those traditionally used in people without HIV. In this prospective cohort of PWH in Atlanta, Georgia, we analyzed PFT and chest computed tomography data from July 2013 through June 2018. After comparing the prevalence of PFT measures of airflow limitation for those with and without radiographic emphysema, we used binary recursive partitioning to identify PFT measures that differentiated between PWH with and without radiographic emphysema. Among 167 PWH who had both PFT and computed tomography data during the study period, 89 (53%) had radiographic emphysema. Those with radiographic emphysema were more likely to have airflow limitations on PFTs. Recursive partitioning identified partitions at a forced expiratory volume in 1 second to forced vital capacity ratio (FEV1/FVC) of 0.78 and a residual volume of 116% predicted. These partitions enabled the identification of 84 (94%) PWH with radiographic emphysema, in contrast to the traditional diagnostic criteria of an FEV1/FVC ratio of 0.7, which only identified 49 (55%) of those with radiographic emphysema. Emphysema in PWH may have different patterns of airflow limitation on PFTs that are not adequately captured by traditional diagnostic criteria. Future studies can seek to validate these findings and determine optimal thresholds for diagnosing HIV-associated emphysema.

Impact of age on surgical outcomes for world federation of neurosurgical societies grade I and II aneurysmal subarachnoid haemorrhage: a novel prognostic model using recursive partitioning analysis

This study aimed to evaluate age as a prognostic factor and develop a comprehensive prognostic model for patients undergoing clipping surgery for World Federation of Neurosurgical Societies (WFNS) grade I/II aneurysmal subarachnoid haemorrhage (SAH). We retrospectively investigated 188 patients with WFNS grade I/II SAH who underwent microsurgical clipping at our institute between December 2010 and January 2020. The data of 176 patients (75 with grade I and 101 with grade II) were analysed. Data on patient demographics, aneurysm characteristics, SAH factors, surgical details, and clinical outcomes were collected. Prognostic factors were assessed using bivariate and multivariable logistic regression analyses, and recursive partitioning analysis. Favourable outcomes (mRS 0–2) were observed in 76% of patients. Age, a significant negative prognostic factor in multivariable analysis (odds ratio 0.55, 95% confidence interval 0.40–0.76, p < 0.001), was cutoff at 70 years by the receiver operating characteristic curve. Patients aged ≤ 70 years had significantly better outcomes than those aged > 70 years (84% vs. 46%, respectively; p < 0.001). Epileptic seizures were significantly associated with poor outcomes in older adults (p < 0.001). A prognostic model (favourable, intermediate, and poor) based on age and postoperative adverse events showed significantly different outcomes between age groups (p < 0.001). Age was a stronger prognostic factor than WFNS grading for patients with grade I/II SAH undergoing microsurgical clipping. For patients aged ≤ 70 years, precise microsurgeries with fewer complications were associated with favourable outcomes beyond WFNS grade. For older patients, postoperative intensive seizure management may prevent poor outcomes.

The role of surgery in recurrent local cerebral metastases: a multi-institutional retrospective analysis.

Local recurrent brain metastases are defined as lesions that recur in the brain at the same site after a previous local therapy. In patients already submitted to surgery, a second operation may be potentially challenging due to scar formation, infiltration of cerebral vessels or eloquent brain areas and local effect of previous radiotherapy. The aim of this study is to retrospectively review the results and complications of a second surgical treatment in a series of local recurrent lesions and to review the literature on this topic. 37 patients submitted to surgery for a local, histologically confirmed, recurrent brain metastases between 2000 and 2022 were retrospectively analyzed with respect to the following parameters: age, histology, anatomic location, time to recurrence, previous radiotherapy, size of recurrent tumors, preoperative and postoperative Karnofsky Performance Status(KPS) score, recursive partitioning analysis (RPA) class and graded prognostic assessment (GPA) score, surgery-related complications and the presence of further cerebral metastases. Overall survival (OS) was calculated using the Kaplan-Meier method. A multivariate Cox proportional hazard model was developed using stepwise regression (forwards selection) with predictive variables that were significant in the univariate analyses. A significant improvement of post-operative KPS status was obtained after second surgery. At multivariate analysis better results in terms of OS were achieved in patients with a pre-operative KPS ≥ 70 and in patients who had received radiotherapy after the initial surgery. No significant postoperative complications related to previous treatments were observed. Surgical resection of local recurrent brain metastases may improve patients ́ neurologic conditions allowing more time for systemic therapies to act with a low incidence of surgery-related morbidity and mortality. However, careful patient selection with a fair pre-operative clinical status seems mandatory to achieve the best post-operative results, since uniform treatment-paradigms cannot be established yet, due to the highly heterogeneous patient cohort.

Efficacy and safety of bevacizumab combined with temozolomide in the treatment of glioma a systematic review and meta-analysis of clinical trials.

Glioma is the most common malignant brain tumor in neurosurgery. Bevacizumab (BEV) is a monoclonal antibody that inhibits tumors by inhibiting vascular endothelial growth factor and reducing tumor angiogenesis. To evaluate the efficacy and safety of BEV combined with temozolomide (TMZ) in glioma, we performed a meta-analysis. PubMed, Embase, The Cochrane Library, and Web of Science databases were searched for randomized controlled trials comparing survival outcomes between TMZ combined with BEV and TMZ alone as well as cohort studies were included in our study. The primary outcome measures analyzed were overall survival (OS) and progression-free survival (PFS). A total of six randomized controlled trials and four cohort studies with a total of 2515 patients were included in our meta-analysis. The results of meta-analysis suggested that there were no significant improvements in overall survival, but the combination of TMZ and BEV prolonged progression-free survival, improved overall response rate (ORR), and increased the incidence of some adverse reactions, compared with TMZ alone. Subgroup analysis suggested sex, recursive partitioning analysis (RPA) grade, MGMT gene status and radiotherapy combination did not affect the improvement of OS with the combination of the two drugs, and RPA grade did not affect the improvement of PFS with the combination of the two drugs. The combination of TMZ and BEV can improve PFS as well as ORR in patients and has no benefit on OS. At the same time, the adverse reactions during the combination of the two drugs were acceptable.

Predicting Functional Outcome After Nonoperative Treatment of Proximal Humeral Fractures Involving the Surgical Neck

BackgroundIn nonoperative treated proximal humeral fractures (PHF), uncertainty remains regarding functional outcome. Therefore, the aim was to identify predictors of functional outcome following nonoperative treatment of PHF and to develop predictive models. MethodsAdults with a nonoperatively treated PHF involving the surgical neck were followed for 1 year. Radiographic parameters included fracture configuration, displacement, bone quality, and the critical shoulder angle (CSA). The neck-shaft angle (NSA) and humeral head offset assessed displacement. The greater tuberosity index (GTI) addressed greater tuberosity (GT) displacement relative to the articular surface while the impingement index addressed GT-displacement relative to the acromion. Multivariate regression models determined predictors of impaired function as measured by Constant Score (CS), flexion, and external rotation (ER). Recursive partitioning created a decision tree model over dichotomized functional outcome that combined flexion and ER (good = flexion ≥ 120° and ER ≥ 40°). ResultsIn 272 consecutive patients (mean age: 66 years, 69% women and 31% men), mean functional outcomes were a CS of 74 points, 138° flexion, and 50° ER. Older age was a predictor for impaired CS, flexion, and ER (p < 0.001), while sex only influenced CS (p = 0.040). Every 10% increase in GTI assessed on the Y-view explained a decrease of 3 CS points, 7° of flexion, and 5° of ER (p < 0.001). Every 10° of varus angulation assessed on anterior-posterior (ap) views in internal rotation resulted in a decrease of 1 CS point and 4° of flexion (p ≤ 0.004). Subgroups of combined important GT-displacement (GTI ≥ 1.15) with age > 76 years or < 119° varus angulation demonstrated the worst function. Linear prediction models estimated flexion, ER, and CS with a mean difference of 10°, 3°, and 3 points, respectively. The decision tree model predicted good function with 80% accuracy (positive predictive value = 81%; negative predictive value = 78%). ConclusionDemographic and radiographic predictors were identified allowing for accurate functional prognosis of nonoperatively treated PHF involving the surgical neck. A combined assessment of the ap-view in internal rotation and Y-view is sufficient for accurate function prediction. The identified subgroups resulting in good or impaired function, and the predictive models may be useful for patient counseling and guidance of treatment-related expectations.

Low body mass index patients have worse outcomes after mechanical thrombectomy

BackgroundThere is evidence that frailty is an independent predictor of worse outcomes after stroke. Similarly, although obesity is associated with a higher risk for stroke, there are multiple reports describing...

Comparative study of machine learning approaches integrated with genetic algorithm for IVF success prediction.

IVF is a widely-used assisted reproductive technology with a consistent success rate of around 30%, and improving this rate is crucial due to emotional, financial, and health-related implications for infertile couples. This study aimed to develop a model for predicting IVF outcome by comparing five machine-learning techniques. The research approached five prominent machine learning algorithms, including Random Forest, Artificial Neural Network (ANN), Support Vector Machine (SVM), Recursive Partitioning and Regression Trees (RPART), and AdaBoost, in the context of IVF success prediction. The study also incorporated GA as a feature selection method to enhance the predictive models' robustness. Findings demonstrate that AdaBoost, particularly when combined with GA feature selection, achieved the highest accuracy rate of 89.8%. Using GA, Random Forest also demonstrated strong performance, achieving an accuracy rate of 87.4%. Genetic Algorithm significantly improved the performance of all classifiers, emphasizing the importance of feature selection. Ten crucial features, including female age, AMH, endometrial thickness, sperm count, and various indicators of oocyte and embryo quality, were identified as key determinants of IVF success. These findings underscore the potential of machine learning and feature selection techniques to assist IVF clinicians in providing more accurate predictions, enabling tailored treatment plans for each patient. Future research and validation can further enhance the practicality and reliability of these predictive models in clinical IVF practice.

Smooth Sigmoid Surrogate (SSS): An Alternative to Greedy Search in Decision Trees

Greedy search (GS) or exhaustive search plays a crucial role in decision trees and their various extensions. We introduce an alternative splitting method called smooth sigmoid surrogate (SSS) in which the indicator threshold function used in GS is approximated by a smooth sigmoid function. This approach allows for parametric smoothing or regularization of the erratic and discrete GS process, making it more effective in identifying the true cutoff point, particularly in the presence of weak signals, as well as less prone to the inherent end-cut preference problem. Additionally, SSS provides a convenient means of evaluating the best split by referencing a parametric nonlinear model. Moreover, in many variants of recursive partitioning, SSS can be reformulated as a one-dimensional smooth optimization problem, rendering it computationally more efficient than GS. Extensive simulation studies and real data examples are provided to evaluate and demonstrate its effectiveness.

Heterogeneity in the association between social support and mental distress in old-age retirees – a computational approach using longitudinal cohort data

BackgroundMental distress among retirees and older people is a severe public health challenge, and information on new risk groups is needed. This study aims to identify subgroups of old-age retirees with varying associations between low social support and mental distress by applying model-based recursive partitioning (MOB).MethodsWe used the Helsinki Health Study follow-up survey data of old-age retired former municipal sector employees of the City of Helsinki, Finland. Phase 1 data were collected in 2000–2002, when all participants were employed, Phase 2 in 2007, Phase 3 in 2012, Phase 4 in 2017, and Phase 5 in 2022 (n = 4,466, 81% women). Social support and covariates were measured at each Phase 1–5 and the outcome, mental distress (Depression Anxiety Stress Scales [DASS-21]) was measured at a single occasion, during Phase 5. The three subscales and the common factor of general distress were analysed separately. An approach rooted in computational statistics was used to investigate risk factor heterogeneity in the association of low social support and mental distress. MOB combines decision trees with regression analysis to identify subgroups with the most significant heterogeneity among risk factors.ResultsMedian (IQR) general distress score from DASS-21 was 5.7 (3.0, 9.0), while Social Support Questionnaire number-score (SSQN) was 1.5 (1.15, 2.05). The primary effect modifier for the association between social support and general distress was education (p < 0.001). Those with high education had a different association of low social support and general distress than those with low or medium education. Additionally, the subgroup with low and medium education had a significant effect modification for age (p = 0.01). For the association between low social support and depressive symptoms, the moderating effect of education was dependent on gender, as men with medium–high education had the weakest association, while for women with medium–high education the association was strongest.ConclusionsOur results suggest that stratification by sociodemographic variables is justifiable when investigating risk factors of mental distress in old-age retirees. The incongruent association of low social support and depressive symptoms in men with medium–high education compared to women with medium–high education is a promising target for confirmatory research.

Patterns of Care for Brain Metastases in Asia: A Real-World Survey Conducted by the Federation of Asian Organizations for Radiation Oncology.

To report the patterns of care for brain metastases (BMs) in the Federation of Asian Organizations for Radiation Oncology (FARO). Overall, 37 questions were prepared. The survey was conducted online using Google Forms, and the URL was distributed to members of the FARO research committee. Radiation oncologists associated with FARO responded to the questionnaire between May 2023 and June 2023, and their answers were analyzed. Responses were received from 32 radiation oncologists in 13 countries participating in FARO. Twenty-six physicians (81.3%) were affiliated with academic centers, and 22 (68.8%) were able to perform stereotactic radiosurgery (SRS) or fractionated stereotactic radiotherapy (fSRT) for BMs at their institution. The most typically used prognostic index for BM was the recursive partitioning analysis classification (17 physicians, 53.1%). The maximum number of BMs indicated for SRT/SRS was ≤three (11 physicians, 34.4%), whereas eight (25.0%) physicians answered for 6-10 BMs. The maximum size of BMs considered for SRS/fSRT was ≤3 cm (14 physicians, 43.8%), whereas nine (28.1%) answered that SRS/fSRT was preferred if the maximum size was >4 cm. When whole-brain radiotherapy (RT) was indicated, hippocampal avoidance and memantine usage were limited to 50.0% and 25.0% of patients, respectively. The most typical RT modality after BM resection was SRS/fSRT alone, regardless of whether the margin was positive (19 physicians, 59.4%) or negative (13 physicians, 40.6%). We report the survey results of the patterns of care for BMs in the FARO. This survey was conducted only among a limited number of FARO members. Since many respondents were affiliated with relatively large-scale academic centers, large-scale surveys, including community hospitals, are warranted for future initiatives.

Interpretable subgroup learning-based modeling framework: Study of diabetic kidney disease prediction.

Complex diseases, like diabetic kidney disease (DKD), often exhibit heterogeneity, challenging accurate risk prediction with machine learning. Traditional global models ignore patient differences, and subgroup learning lacks interpretability and predictive efficiency. This study introduces the Interpretable Subgroup Learning-based Modeling (iSLIM) framework to address these issues. iSLIM integrates expert knowledge with a tree-based recursive partitioning approach to identify DKD subgroups within an EHR dataset of 11,559 patients. It then constructs separate models for each subgroup, enhancing predictive accuracy while preserving interpretability. Five clinically relevant subgroups are identified, achieving an average sensitivity of 0.8074, outperforming a single global model by 0.1104. Post hoc analyses provide pathological and biological evidence supporting subgroup validity and potential DKD risk factors. The iSLIM surpasses traditional global model in predictive performance and subgroup-specific risk factor interpretation, enhancing the understanding of DKD's heterogeneous mechanisms and potentially increasing the adoption of machine learning models in clinical decision-making.

Pathological response following neoadjuvant immune checkpoint inhibitors in patients with hepatocellular carcinoma: a cross-trial, patient-level analysis

Neoadjuvant use of immune checkpoint inhibitors (ICIs) before liver resection results in pathological tumour regression in patients with hepatocellular carcinoma. We aimed to describe the characteristics of pathological responses after preoperative ICI therapy for hepatocellular carcinoma and to evaluate the association between the depth of tumour regression and relapse-free survival. In this cross-trial, patient-level analysis, we performed a pooled analysis of data from patients with hepatocellular carcinoma receiving ICI therapy before liver resection as part of a global collaborative consortium (NeoHCC) of five phase 1 and 2 clinical trials and standardised observational protocols conducted in 12 tertiary referral centres across the USA, UK, and Taiwan. Eligible patients were adults (aged ≥18 years) diagnosed with hepatocellular carcinoma by tissue core biopsy before treatment initiation, a Liver Imaging Reporting and Data System score of 5 on imaging, or both, with an Eastern Cooperative Oncology Group performance status score of 0-1, and no extrahepatic spread or previous ICI treatment. Pathological response was measured as the percentage of non-viable tumour in the resected surgical specimen, with major pathological response corresponding to at least 70% tumour regression and pathological complete response corresponding to 100% tumour regression. We correlated pathological response with radiological overall response using RECIST criteria (version 1.1) and relapse-free survival, and evaluated the threshold of tumour regression that could be optimally associated with relapse-free survival. At data cutoff on Jan 31, 2024, 111 patients were included in the study, of whom data on pathological response were available for 104 (94%) patients. Patients received treatment from Oct 5, 2017, to Nov 15, 2023, mostly ICI combinations (76 [69%]), for a median of 1·4 months (IQR 0·7-2·9). 87 (78%) patients were men and 24 (22%) were women. Most patients had underlying viral chronic liver disease (73 [66%]) and Barcelona Clinic Liver Cancer stage A hepatocellular carcinoma (61 [55%]), without portal vein thrombosis (87 [78%]). We observed major pathological response in 33 (32%) patients and pathological complete response in 19 (18%) patients. Radiological overall response was associated with major pathological response, with 23 (74%) of 31 patients with radiological response showing major pathological response compared with ten (14%) of 73 patients without radiological response (p<0·0001). However, ten (30%) of 33 major pathological responses were not predicted by radiological response. After a median follow-up of 27·2 months (95% CI 22·3-32·1), median relapse-free survival for the whole cohort was 43·6 months (95% CI 28·3-not evaluable). Relapse-free survival was significantly longer in patients with major pathological response than in those who did not have a major pathological response (not reached [95% CI not evaluable-not evaluable] vs 28·3 months [12·8-43·8]; hazard ratio 0·26 [0·10-0·66]; p=0·0024) and in patients with pathological complete response than in those who did not have a pathological complete response (NR [95% CI not evaluable-not evaluable] vs 32·8 months [15·0-50·5]; 0·19 [0·05-0·78]; p=0·010). Unbiased recursive partitioning of the cohort for the risk of relapse, death, or both identified a threshold of 90% as the optimal cutoff of pathological tumour regression to predict improved relapse-free survival. The extent of tumour regression following neoadjuvant ICI therapy could identify patients with improved relapse-free survival following liver resection. The threshold of at least 90% tumour regression should be validated for its surrogate role for relapse-free survival in phase 3 randomised controlled trials. None.

Therapy Intensity Outweighs the Prognostic Importance of the Timing of Chemoradiotherapy in Newly Diagnosed Glioblastoma Patients.

To investigate the significance of the timing of chemoradiotherapy together with clinical and laboratory features in newly diagnosed glioblastoma. Clinical and laboratory parameters of 209 patients with glioblastoma potentially influencing overall (OS) and progression-free (PFS) survival were analyzed in univariable and multivariable models. On univariable analyses, Karnofsky performance status (p<0.001), recursive partitioning analysis (RPA) class (p<0.001), O6-methylguanine-DNA methyltransferase (MGMT)-status (p<0.001), extent of resection (p<0.001), radiotherapy dose (p=0.01), and the number of adjuvant temozolomide (TMZ) cycles (p<0.001) were significantly associated with OS. Additionally, MGMT-status (p<0.001), extent of resection (p=0.03), surgical site infections (p=0.02), and the number of adjuvant TMZ cycles (p<0.001) were significantly associated with PFS. Multivariable analysis identified radiotherapy dose as the only independent predictor (p=0.049) of OS. MGMT-status (p=0.02) and the number of adjuvant TMZ cycles (p<0.001) were independent predictors of PFS. The timing of chemoradiotherapy did not play a prognostic role. For OS, the radiotherapy dose, and for PFS, MGMT-status and the number of adjuvant TMZ cycles were identified as independent prognostic factors.

Nonlinear association between atherogenic index of plasma and chronic kidney disease: a nationwide cross-sectional study

BackgroundThe interplay between metabolic disorders and chronic kidney disease (CKD) has been well-documented. However, the connection between CKD and atherogenic index of plasma (AIP) remains understudied. This research delves into the correlation between these two factors, aiming to shed new light on their potential association.MethodsThe relationship between AIP and CKD was evaluated using a weighted multivariate logistic regression model, and the curvilinear relationship between AIP and CKD was explored through smooth curve fitting. We engaged a recursive partitioning algorithm in conjunction with a two-stage linear regression model to precisely determine the inflection point. By conducting stratified analyses, the heterogeneity within subpopulations was explored.ResultsIn the regression model that accounted for all covariates, ORs (95% CI) for the association between CKD and AIP were 1.12 (0.91, 1.36), indicating no significant association between AIP and CKD. However, sensitivity analyses suggested that the relationship between them may be non-linear. Smooth curve analysis confirmed the non-linear relationship between AIP and CKD, identifying an inflection point at -0.55. Below this threshold, AIP exhibited a significant inverse correlation with CKD. Conversely, above this threshold, a pronounced positive correlation was detected. Stratified analyses elucidated that a non-linear association between AIP and CKD was observed among female participants and those aged 50 and above.ConclusionWe found a curvilinear relationship between chronic kidney disease and atherogenic index of plasma.