Prognostic factors differ according to KRAS mutational status: A classification and regression tree model to define prognostic groups after hepatectomy for colorectal liver metastasis

Survival Tree Provides Individualized Estimates of Survival After Lung Transplant

INTRODUCTION

Identifying contributors to lung transplant survival is vital in mitigating mortality. To enhance individualized mortality estimation and determine variable interaction, we employed a survival tree algorithm utilizing recipient and donor data.

METHODS

United Network Organ Sharing data (2000-2021) were queried for single and double lung transplants in adult patients. Graft survival time <7 d was excluded. Sixty preoperative and immediate postoperative factors were evaluated with stepwise logistic regression on mortality; final model variables were included in survival tree modeling. Data were split into training and testing sets and additionally validated with 10-fold cross validation. Survival tree pruning and model selection was based on Akaike information criteria and log-likelihood values. Estimated survival probabilities and log-rank pairwise comparisons between subgroups were calculated.

RESULTS

A total of 27,296 lung transplant patients (8175 single; 19,121 double lung) were included. Stepwise logistic regression yielded 47 significant variables associated with mortality. Survival tree modeling returned six significant factors: recipient age, length of stay from transplant to discharge, recipient ventilator duration post-transplant, double lung transplant, recipient reintubation post-transplant, and donor cytomegalovirus status. Eight subgroups consisting of combinations of these factors were identified with distinct Kaplan-Meier survival curves.

CONCLUSIONS

Survival trees provide the ability to understand the effects and interactions of covariates on survival after lung transplantation. Individualized survival probability with this technique found that preoperative and postoperative factors influence survival after lung transplantation. Thus, preoperative patient counseling should acknowledge a degree of uncertainty given the influence of postoperative factors.

Role of sex in liver tumor occurrence and clinical outcomes: A comprehensive review

Clinical research on sex-based differences in the manifestations, pathophysiology, and prevalence of several diseases, including those affecting the liver, has expanded considerably in recent years. Increasing evidence suggests that liver diseases develop, progress, and respond to treatment differently depending on the sex. These observations support the concept that the liver is a sexually dimorphic organ in which estrogen and androgen receptors are present, which results in disparities between men and women in liver gene expression patterns, immune responses, and the progression of liver damage, including the propensity to develop liver malignancies. Sex hormones play protective or deleterious roles depending on the patient's sex, the severity of the underlying disease, and the nature of precipitating factors. Moreover, obesity, alcohol consumption, and active smoking, as well as social determinants of liver diseases leading to sex-related inequalities, may interact strongly with hormone-related mechanisms of liver damage. Drug-induced liver injury, viral hepatitis, and metabolic liver diseases are influenced by the status of sex hormones. Available data on the roles of sex hormones and gender differences in liver tumor occurrence and clinical outcomes are conflicting. Here, we critically review the main gender-based differences in the molecular mechanisms associated with liver carcinogenesis and the prevalence, prognosis, and treatment of primary and metastatic liver tumors.

Clinical Prediction Models for Prognosis of Colorectal Liver Metastases: A Comprehensive Review of Regression-Based and Machine Learning Models

Colorectal liver metastasis (CRLM) is a disease entity that warrants special attention due to its high frequency and potential curability. Identification of "high-risk" patients is increasingly popular for risk stratification and personalization of the management pathway. Traditional regression-based methods have been used to derive prediction models for these patients, and lately, focus has shifted to artificial intelligence-based models, with employment of variable supervised and unsupervised techniques. Multiple endpoints, like overall survival (OS), disease-free survival (DFS) and development or recurrence of postoperative complications have all been used as outcomes in these studies. This review provides an extensive overview of available clinical prediction models focusing on the prognosis of CRLM and highlights the different predictor types incorporated in each model. An overview of the modelling strategies and the outcomes chosen is provided. Specific patient and treatment characteristics included in the models are discussed in detail. Model development and validation methods are presented and critically appraised, and model performance is assessed within a proposed framework.

Non-tumor-related prognostic factors for immunotherapy-chemotherapy or immunotherapy alone as first-line in advanced non-small cell lung cancer (NSCLC)

Besides programmed death ligand 1 (PD-L1) expression, rapid, cost-effective and validated scores or models are critical for the prognosis and prediction of patients received immune checkpoint inhibitors (ICIs). In this retrospective study, 182 patients with NSCLC receiving ICIs from 2015 to 2022 were divided 1:1 into a training cohort and a validation cohort. We identified a score established by three factors and analyzed the prognostic implications by Kaplan-Meier approach (Log rank test) and time-dependent receiver operating characteristic (ROC) analyses. A non-tumor-related score (NTRS) was established that could be used as a prognostic factor (HR 2.260, 95% CI 1.559-3.276, P < 0.001 in training cohort; HR 2.114, 95% CI 1.493-2.994, P < 0.001 in validation cohort) and had a high time-dependent ROC for overall survival (OS) (AUC 0.670-0.782 in training cohort; AUC 0.682-0.841 in validation cohort). PD-L1 (1-49%) and NTRS (score = 0, 1, 2, 3) combination significantly improved the assessment of patients' OS and progress-free survival (PFS), which was statistically different in training cohorts (P < 0.001 for OS, 0.012 for PFS) and validation cohorts (P = 0.01 for OS, < 0.001 for PFS). The NTRS provided a better assessment of durable clinical benefit (DCB) compared to PD-L1 expression (P = 0.009 vs. 0.232 in training cohort; P = 0.004 vs. 0.434 in validation cohort). NTRS may help improve prognosis stratification of patients receiving ICIs in first-line NSCLC and may be combined with tumor-related parameters.

Does RAS Status Increase the Prevalence of Positive Resection Margin in Colorectal Liver Metastasis? A Systematic Review and Meta-Analysis

BACKGROUND

Colorectal liver metastasis has a high incidence, and RAS oncogene mutation status carries significant prognostic information. We aimed to assess whether RAS-mutated patients present more or less frequently with positive margins in their hepatic metastasectomy.

METHODS

We performed a systematic review and meta-analysis of studies from PubMed, Embase, and Lilacs databases. We analyzed liver metastatic colorectal cancer studies, which included information on RAS status and had surgical margin analysis of the liver metastasis. Odds ratios were computed using a random-effect model due to anticipated heterogeneity. We further performed a subanalysis limited to studies that included only patients with KRAS instead of all-RAS mutations.

RESULTS

From the 2,705 studies screened, 19 articles were included in the meta-analysis. There were 7,391 patients. The prevalence of positive resection margin was not significantly different between patients carrier vs non-carrier for the all-RAS mutations (OR .99; 95% CI 0.83-1.18; P = .87), and for only KRAS mutation (OR .93; 95% CI 0.73-1.19; P = .57).

CONCLUSIONS

Despite the strong correlation between colorectal liver metastasis prognosis and RAS mutation status, our meta-analysis's results suggest no correlation between the RAS status and the prevalence of positive resection margins. The findings contribute to a better understanding of the RAS mutation's role in the surgical resections of colorectal liver metastasis.

Isolated Pancreatic Metastases of Renal Cell Carcinoma-Clinical Particularities and Seed and Soil Hypothesis

A meta-analysis of 1470 isolated pancreatic metastases of renal cell carcinoma revealed, that, in addition to the unusual exclusive occurrence of pancreatic metastases and the favourable treatment results, the isPMRCC is characterised by further peculiarities of the clinical course: The lack of prognostic significance of volume and growth rate dependent risk factors and the independence of treatment results from standard or local resections. As an explanation for all these peculiarities, according to today's knowledge, a strong acting seed and soil mechanism can serve, which allows embolized tumour cells to grow to metastases only in the pancreas, and prevents them definitively or for years in all other organs. The good prognosis affects not only isolated PM, but also multi-organ metastases of the RCC, in which the additional occurrence of PM is also associated with a better prognosis. Genetic studies revealed specific changes in cases of PM of RCC: Lack of loss of 9p21.3 and 14q31.2, which are otherwise specific gene mutations at the onset of generalization, a low weight genome instability index, i.e., high genetic stability, and a low rate of PAB1 and a high rate of BPRM1 alterations, which signal a more favourable course. The cause of pancreatic organotropism in isPMRCC is still unclear, so only those factors that have been identified as promoting organotropism in other, more frequent tumour entities can be presented: Formation of the pre-metastatic niche, chemokine receptor-ligand mechanism, ability to metabolic adaptation, and immune surveillance.

A Scoping Review of the Use of Machine Learning in Health Economics and Outcomes Research: Part 2-Data From Nonwearables

OBJECTIVES

Despite the increasing interest in applying machine learning (ML) methods in health economics and outcomes research (HEOR), stakeholders face uncertainties in when and how ML can be used. We reviewed the recent applications of ML in HEOR.

METHODS

We searched PubMed for studies published between January 2020 and March 2021 and randomly chose 20% of the identified studies for the sake of manageability. Studies that were in HEOR and applied an ML technique were included. Studies related to wearable devices were excluded. We abstracted information on the ML applications, data types, and ML methods and analyzed it using descriptive statistics.

RESULTS

We retrieved 805 articles, of which 161 (20%) were randomly chosen. Ninety-two of the random sample met the eligibility criteria. We found that ML was primarily used for predicting future events (86%) rather than current events (14%). The most common response variables were clinical events or disease incidence (42%) and treatment outcomes (22%). ML was less used to predict economic outcomes such as health resource utilization (16%) or costs (3%). Although electronic medical records (35%) were frequently used for model development, claims data were used less frequently (9%). Tree-based methods (eg, random forests and boosting) were the most commonly used ML methods (31%).

CONCLUSIONS

The use of ML techniques in HEOR is growing rapidly, but there remain opportunities to apply them to predict economic outcomes, especially using claims databases, which could inform the development of cost-effectiveness models.

Prognostic factors of survival and a new scoring system for liver resection of colorectal liver metastasis

BACKGROUND

Hepatic resection has become the preferred treatment of choice for colorectal liver metastasis (CLM) patients.

AIM

To identify the prognostic factors and to formulate a new scoring system for management of CLM.

METHODS

Clinicopathologic and long-term survival data were analyzed to identify the significant predictors of survival by univariate and multivariate analyses with the Cox model. A clinical score was constructed based on the analysis results.

RESULTS

Three factors of worse overall survival were identified in the multivariate analysis. They were number of liver metastases ≥ 5, size of the largest liver lesion ≥ 4 cm, and the presence of nodal metastasis from the primary tumor. These three factors were chosen as criteria for a clinical risk score for overall survival. The clinical score highly correlated with median overall survival and 5-year survival (P = 0.002).

CONCLUSION

Priority over surgical resection should be given to the lowest score groups, and alternative oncological treatment should be considered in patients with the highest score.

Prognostic and Therapeutic Implications of Tumor Biology in Colorectal Liver Metastases

Prognostic models allow clinicians to predict survival outcomes, facilitate patient-physician discussions, and identify subgroups with potentially distinct prognoses [...].

Prognostic impact of neoadjuvant chemotherapy in patients with synchronous colorectal liver metastasis: A propensity score matching comparative study

BACKGROUND

Synchronous colorectal liver metastasis (SCRLM) is at an advanced tumor stage and requires multidisciplinary treatments. Neoadjuvant chemotherapy (NAC) is thought to be an effective treatment modality, but its prognostic impact is still unclear.

MATERIALS AND METHODS

Patients with resectable SCRLM presented to eight university hospitals between 2007 and 2017 were retrospectively reviewed. Propensity score matching (PSM) was performed to adjust baseline characteristics between patients who received NAC with those who underwent up-front hepatectomy. The prognostic impact of NAC was then evaluated.

RESULTS

The cohort comprised of 320 patients: 151 patients received NAC and the remaining 169 patients underwent up-front hepatectomy. After a 1:1 ratio of PSM, 102 patients per group were selected. Within the PSM cohort, 66% patients had multiple liver tumors, with 15% having five or more liver tumors. The median survival (95% confidence interval) periods for patients with and without NAC in the PSM cohort were 88.5 (68.4 - not reached) and 84.2 (52.1 - not reached) months, respectively (P = 0.51). On multivariate analysis, the postoperative events in these patients including operative complications and use of adjuvant chemotherapy after hepatectomy were prognostic factors with hazards (95% confidence interval) being 1.88 (1.18-2.98) and 0.65 (0.42-1.01), respectively.

CONCLUSION

This PSM study was restricted to patients with SCRLM and relatively advanced tumor stagings. NAC did not show any significant prognostic impact. While operative complications had a significant prognostic impact, use of adjuvant chemotherapy after hepatectomy had only a marginal prognostic impact. Reconsideration of indications for NAC is needed.

Assessment of appropriate body mass index cut-off points for long-term mortality among ST-elevation myocardial infarction survivors in Asian population using machine learning algorithm

Low body mass index (BMI) is a predictor of adverse events in patients with ST-elevated myocardial infarction (STEMI) in Western countries. Because the average BMI of Asians is significantly lower than that of the Western population, the appropriate cut-off BMI value and its role in long-term mortality are unclear in Asian patients. Between January 2006 and December 2017, 1215 patients who underwent percutaneous coronary intervention (PCI) for acute STEMI and were alive at discharge (mean age, 67.7 years; male, 75.4%) were evaluated. The cut-off BMI value, which could predict all-cause mortality within 10 years, was detected using a survival classification and regression tree (CART) model. The causes of death according to the BMI value were evaluated in each group. Based on the CART model, the patients were divided into three groups (BMI < 18 kg/m²: 54 patients, 18 kg/m² ≤ BMI ≤ 20 kg/m²: 109 patients, and BMI > 20 kg/m²: 1052 patients). The BMI decreased with age; with an increased BMI, patients with dyslipidemia, diabetes mellitus, and smoking habit increased. During the study period (median, 4.9 years), 194 patients (26.8%) died (cardiac death, 59 patients; non-cardiac death, 135 patients). All-cause mortality was more frequent as the BMI decreased (BMI < 18 kg/m²; 72.8%, 18 kg/m² ≤ BMI ≤ 20 kg/m²; 40.5%, and BMI > 20 kg/m²; 22.8%; log-rank p < 0.001). Non-cardiac deaths were more frequent than cardiac deaths in all groups, and the dominance of non-cardiac death was highest in the lowest BMI group. Cut-off BMI values of 18 kg/m² and 20 kg/m² can predict long-term mortality after PCI in Asian STEMI survivors, whose cut-off value is lower than that in the Western populations. The main causes of death in this cohort differed according to the BMI values.

The current understanding on the impact of KRAS on colorectal cancer

KRAS (kirsten rat sarcoma viral oncogene) is a member of the RAS family. KRAS mutations are one of most dominant mutations in colorectal cancer (CRC). The impact of KRAS mutations on the prognosis and survival of CRC patients drives many research studies to explore potential therapeutics or target therapy for the KRAS mutant CRC. This review summarizes the current understanding of the pathological consequences of the KRAS mutations in the development of CRC; and the impact of the mutations on the response and the sensitivity to the current front-line chemotherapy. The current therapeutic strategies for treating KRAS mutant CRC, the difficulties and challenges will also be discussed.

Resection of Colorectal Liver Metastasis: Prognostic Impact of Tumor Burden vs KRAS Mutational Status

BACKGROUND

The prognostic impact of colorectal liver metastasis (CRLM) morphologic characteristics relative to KRAS mutational status after hepatic resection remains ill defined.

STUDY DESIGN

Patients undergoing hepatectomy for CRLM between 2001 and 2018 were identified using an international multi-institutional database. Tumor burden score (TBS) was defined as distance from origin on a Cartesian plane that incorporated maximum tumor size (x-axis) and number of lesions (y-axis). Impact of TBS on overall survival (OS) relative to KRAS status (wild type [wtKRAS] vs mutated [mutKRAS]) was assessed.

RESULTS

Among 1,361 patients, the median number of metastatic lesions was 2 (interquartile range [IQR] 1-3), and median size of the largest metastatic lesion was 3.0 cm (IQR 2.0-5.0 cm), resulting in a median TBS of 4.1 (IQR 2.8-6.1); KRAS status was wtKRAS (n = 420, 30.9%), mutKRAS (n = 251, 18.4%), and unknown (n = 690, 50.7%). Overall median and 5-year OS were 49.5 months (95%CI 45.2-53.8) and 43.2%, respectively. In examining the entire cohort, TBS was associated with long-term prognosis (5-year OS, low TBS: 49.4% vs high TBS: 36.7%), as was KRAS mutational status (5-year OS, wtKRAS: 48.2% vs mutKRAS: 31.1%; unknown KRAS: 44.0%)(both p < 0.01). Among patients with wtKRAS tumors, TBS was strongly associated with improved OS (5-year OS, low TBS: 59.1% vs high TBS: 38.4%, p = 0.002); however, TBS failed to discriminate long-term prognosis among patients with mutKRAS tumors (5-year OS, low TBS: 37.4% vs high TBS: 26.7%, p = 0.19). In fact, patients with high TBS/wtKRAS CRLM had comparable outcomes to patients with low TBS/mutKRAS tumors (5-year OS, 38.4% vs 37.4%, respectively; p = 0.59). On multivariable analysis, while TBS was associated with OS among patients with wtKRAS CRLM (hazard ratio 1.43, 95%CI 1.02-2.00; p = 0.03), TBS was not an independent predictor of survival among patients with mutKRAS CRLM (HR 1.36, 95%CI 0.92-1.99; p = 0.12).

CONCLUSIONS

While TBS was associated with survival among patients with wtKRAS tumors, CRLM morphology was not predictive of long-term outcomes among patients with mutKRAS CRLM.