Inpatient Satisfaction With Surgical Resident Care After Elective General and Oncologic Surgery.

OBJECTIVE: To investigate inpatient satisfaction with surgical resident care. BACKGROUND: Surgical trainees are often the primary providers of care to surgical inpatients, yet patient satisfaction with surgical resident care is not well characterized or routinely assessed. METHODS: English-speaking, general surgery inpatients recovering from elective gastrointestinal and oncologic surgery were invited to complete a survey addressing their satisfaction with surgical resident care. Patients positively identified photos of surgical senior residents and interns before completing a modified version of the Consumer Assessment of Healthcare Providers and Systems Surgical Care Survey (S-CAHPS). Adapted S-CAHPS items were scored using the "top-box" method. RESULTS: Ninety percent of recruited patients agreed to participate (324/359, mean age=62.2, 50.3% male). Patients were able to correctly identify their seniors and interns 85% and 83% of the time, respectively ( P =0.14). On a 10-point scale, seniors had a mean rating of 9.23±1.27 and interns had a mean rating of 9.01±1.49 ( P =0.14). Ninety-nine percent of patients agreed it was important to help in the education of future surgeons. CONCLUSIONS: Surgical inpatients were able to recognize their resident physicians with high frequency and rated resident care highly overall, suggesting that they may serve as a willing source of feedback regarding residents' development of core competencies such as interpersonal skills, communication, professionalism, and patient care. Future work should investigate how to best incorporate patient evaluation of surgical resident care routinely into trainee assessment to support resident development.

Risk of suicide among stroke survivors in the United States.

BACKGROUND: Stroke is the largest cause of disability and the 5th leading cause of death in the United States. Suicide is the 12th leading cause of death in the United States. However, little is known about the risk of suicide among people with a prior stroke. OBJECTIVES: Using Multiple Cause of Death data (1999-2020) from the Centers for Disease Control and Prevention WONDER database, we examined via cross-sectional analysis the risk of suicide among survivors of stroke as compared to the general U.S. population and among subgroups within the United States. METHODS: We assessed disparities in suicide rate among patients with stroke stratified by sex, race, urbanization levels, and census regions using the CDC WONDER multiple cause of death database. Standardized mortality rates were calculated to compare the suicide rate of stroke patients with the rates among demographic-matched cohorts and the general United States population. RESULTS: As compared to the general population, stroke survivors had an elevated risk of suicide. Black stroke survivors had a lower rate of suicide as compared to the general population, while White stroke survivors and those in nonmetropolitan areas had an elevated risk compared to the general population. CONCLUSION: There was a slightly elevated risk of suicide among people with a prior stroke in the United States. This risk may be elevated among White people and among people living in nonmetropolitan areas.

Identification of Specific Educational Targets to Improve the Student Surgical Clerkship Experience.

BACKGROUND: This study describes the relationship between medical student perception of surgery, frequency of positive surgery clerkship activities, and overall surgical clerkship experience. METHODS: Medical students at four academic hospitals completed pre- and post-clerkship surveys assessing 1) surgery clerkship activities/experiences and 2) perceptions of surgery during the 2017-2018 academic year. RESULTS: Ninety-one percent of students completed both a pre- and post-clerkship survey (n = 162 of 179). Student perception of surgery significantly improved across the clerkship overall (P < 0.0001) and for 7 of 21 specific items. Eighty-six percent of students agreed that the clerkship was a meaningful experience. Sixty-six percent agreed that the operating room was a positive learning environment. Multivariable logistic regression identified one-on-one mentoring from a resident (OR [95% CI] = 2.12 [1.11-4.04], P = 0.02) and establishing a meaningful relationship with a surgical patient (OR = 2.21 [1.12-4.37], P = 0.02) as activities predictive of student agreement that the surgical clerkship was meaningful. Making an incision (OR = 2.92 [1.54-5.56], P = 0.001) and assisting in dissection (OR = 1.67 [1.03-2.69], P = 0.035) were predictive of student agreement that the operating room was a positive learning environment. Positive student perception of surgery before the clerkship was associated with increased frequency of positive clerkship activities including operative involvement (r = 0.26, P = 0.001) and relationships with surgical attendings (r = 0.20, P = 0.01), residents (r = 0.41, P < 0.0001), and patients (r = 0.24, P = 0.003). CONCLUSIONS: Interventions to improve surgery clerkship quality should target enhancing student relationships with residents and surgical patients as well as providing opportunity for student operative involvement beyond just suturing. In addition, fostering positive perceptions of surgery in the preclinical period may increase meaningfulness and experience with the later surgery clerkship.

Feasibility and Perceived Usefulness of Using Head-Mounted Cameras for Resident Video Portfolios.

BACKGROUND: There is limited guidance on how to longitudinally administer simulation materials or to incorporate video recordings into assessment portfolios of simulated surgical skills. MATERIALS AND METHODS: We launched a longitudinal weekly simulation curriculum for PGY1-PGY3 surgical residents based on the ACS/APDS Curriculum. Residents underwent monthly objective structured assessment of technical skills (OSATS) while wearing head-mounted cameras. Videos of OSATS performance accrued into individual online video portfolios. Residents were surveyed about their attitudes toward video recording. RESULTS: Twenty-seven general surgical residents participated, completing 161 OSATS encompassing 11 distinct skills and generating 258 videos of simulated skills performance. The overall survey response rate was 88%. Residents viewed the curriculum favorably overall, and 36.4% of residents accessed their videos. Of those who did not watch their videos, 78.6% cited not having enough time, whereas 28.6% did not think the videos would be useful. Over 95% of surveyed residents expressed interest in having a video library of attending-performed procedures, 59.1% were interested in having their own operations recorded, and 45.5% were interested in video-based coaching. CONCLUSIONS: Residents viewed longitudinal administration of the ACS/APDS Curriculum positively. Although video recording in simulation is feasible, resident interest may be higher for intraoperative recordings than for simulated skills.

Disparities in the risk of heart disease death among people with chronic lower respiratory disease.

BACKGROUND: Heart disease remains the leading cause of death in the United States, while chronic lower respiratory diseases (CLRD) are the sixth leading cause of death. Patients with CLRD have been shown to have an elevated risk of heart disease death. However, less is known regarding how this risk varies across demographic groups. METHODS: We used the Multiple Cause of Death database from the Centers for Disease Control Wide-ranging ONline Data for Epidemiologic Research to obtain 1999-2020 information on deaths with heart disease as a primary cause of death and CLRD as a contributing cause. We calculated age-adjusted mortality rates (AAMR) over time and for demographic subgroups. RESULTS: During 1999-2020, there were 1,178,048 heart disease deaths related to CLRD among people aged 45+. The AAMR for CLRD-associated heart disease deaths was 45.713 per 100,000 people. AAMR was highest among those aged 65+ (108.56 per 100,000). Elevated rates were seen among males (AAMR ratio = 1.744, 95% CI: 1.741-1.748), people living in the Midwest (AAMR ratio = 1.196, 95% CI: 1.190-1.202), and among people in rural areas (AAMR ratio = 1.309, 95% CI: 1.304-1.313) compared to their corresponding counterparts. Between 1999 and 2004 and 2016-2020 rates decreased among all demographic subgroups, except for among people aged 45-64, among whom deaths increased (AAMR ratio = 1.016, 95% CI: 1.003-1.030). CONCLUSION: Rates of CLRD-associated heart disease deaths have declined over time, but significant disparities remain. Enhanced interventions particularly among older people (65+), people living in rural areas, people living in the Midwest, and men may reduce CLRD-associated heart disease deaths in the United States.

From Machine Learning to Patient Outcomes: A Comprehensive Review of AI in Pancreatic Cancer.

Pancreatic cancer is a highly aggressive and difficult-to-detect cancer with a poor prognosis. Late diagnosis is common due to a lack of early symptoms, specific markers, and the challenging location of the pancreas. Imaging technologies have improved diagnosis, but there is still room for improvement in standardizing guidelines. Biopsies and histopathological analysis are challenging due to tumor heterogeneity. Artificial Intelligence (AI) revolutionizes healthcare by improving diagnosis, treatment, and patient care. AI algorithms can analyze medical images with precision, aiding in early disease detection. AI also plays a role in personalized medicine by analyzing patient data to tailor treatment plans. It streamlines administrative tasks, such as medical coding and documentation, and provides patient assistance through AI chatbots. However, challenges include data privacy, security, and ethical considerations. This review article focuses on the potential of AI in transforming pancreatic cancer care, offering improved diagnostics, personalized treatments, and operational efficiency, leading to better patient outcomes.

Patient satisfaction with women vs men surgical interns and senior residents.

BACKGROUND: Patient satisfaction is critical for referrals and reimbursement of surgical faculty but remains poorly characterized for residents. We investigated whether patient evaluations of surgical trainees vary by resident gender. METHODS: Surgical inpatients evaluated surgical resident care postoperatively after positively identifying trainees. Evaluations (Consumer Assessment of Healthcare Providers and Systems Surgical Care Surveys (S-CAHPS)) were scored by the "top-box" method, stratified by training level, and compared between women and men residents. RESULTS: Ninety-one percent of patients participated (n â€‹= â€‹324/357). Patients recognized women interns less than men (75.0 â€‹% vs 87.2 â€‹%, p â€‹= â€‹0.01). S-CAHPS scores for women vs men interns were equivalent except for spending sufficient time with patients (75.6 â€‹% vs 88.0 â€‹%, p â€‹= â€‹0.02). For senior residents, there was no difference in patient recognition of women vs men (83.9 â€‹% vs 85.2 â€‹%, p â€‹= â€‹0.91) or in any S-CAHPS scores (p â€‹> â€‹0.05). CONCLUSIONS: Gendered differences in patient evaluations of surgical trainees exist for interns but resolve by senior years. Future work should explore how patient evaluations can support trainee development while ensuring patients recognize the role of surgical residents regardless of gender.

Wedge Resection versus Stereotactic Body Radiation Therapy for Non-Small Cell Lung Cancer Tumors ≤8 mm.

The objective of this study was to evaluate the overall survival of patients with ≤8 mm non-small cell lung cancer (NSCLC) who undergo wedge resection versus stereotactic body radiation therapy (SBRT). Kaplan-Meier analysis, multivariable Cox proportional hazards modeling, and propensity score-matched analysis were performed to evaluate the overall survival of patients with ≤8 mm NSCLC in the National Cancer Database (NCDB) from 2004 to 2017 who underwent wedge resection versus patients who underwent SBRT. The above-mentioned matched analyses were repeated for patients with no comorbidities. Patients who were coded in the NCDB as having undergone radiation because surgery was contraindicated due to patient risk factors (e.g., comorbid conditions, advance age, etc.) and those with a history of prior malignancy were excluded from analysis. Of the 1505 patients who had NSCLC ≤8 mm during the study period, 1339 (89%) patients underwent wedge resection, and 166 (11%) patients underwent SBRT. In the unadjusted analysis, multivariable Cox modeling and propensity score-matched analysis, wedge resection was associated with improved survival when compared to SBRT. These results were consistent in a sensitivity analysis limited to patients with no comorbidities.

Using Renin Activity to Guide Mineralocorticoid Receptor Antagonist Therapy in Patients with Low Renin and Hypertension.

BACKGROUND: Mineralocorticoid receptor antagonists (MRAs) are often empirically used for patients with low-renin hypertension (LRH) or probable primary aldosteronism (PA) who decline surgery. However, the optimal approach to MRA therapy is unknown. Studies have shown that a rise in renin is an effective biomarker of prevention of cardiovascular complications of PA. This study aimed to determine whether empiric MRA therapy in patients with LRH or probable PA targeting unsuppressed renin is associated with a decrease in blood pressure and/or proteinuria. METHODS: Retrospective single-center cohort study from 2005 to 2021 included adults with LRH or probable PA (renin activity <1.0 ng/ml/h and detectable aldosterone levels). All patients were empirically treated with an MRA, targeting renin ≥1.0 ng/ml/h. RESULTS: Out of 39 patients studied, 32 (82.1%) achieved unsuppressed renin. Systolic and diastolic blood pressure decreased from 148.0 and 81.2 to 125.8 and 71.6 mm Hg, respectively (P < 0.001 for both). Similar blood pressure reductions were seen whether patients had high (>10 ng/dl) or low (<10 ng/dl) aldosterone levels. The majority (24/39; 61.5%) of patients had at least one baseline anti-hypertensive medication stopped. Among the six patients who had detectable proteinuria and albumin-to-creatinine (ACR) measurements post-treatment, the mean ACR decreased from 179.0 to 36.1 mg/g (P = 0.03). None of the patients studied had to completely stop treatment due to adverse reactions. CONCLUSIONS: Empiric MRA therapy in patients with LRH or probable PA targeting unsuppressed renin can safely and effectively improve blood pressure control and reduce proteinuria.

Role of artificial intelligence in risk prediction, prognostication, and therapy response assessment in colorectal cancer: current state and future directions.

Artificial Intelligence (AI) is a branch of computer science that utilizes optimization, probabilistic and statistical approaches to analyze and make predictions based on a vast amount of data. In recent years, AI has revolutionized the field of oncology and spearheaded novel approaches in the management of various cancers, including colorectal cancer (CRC). Notably, the applications of AI to diagnose, prognosticate, and predict response to therapy in CRC, is gaining traction and proving to be promising. There have also been several advancements in AI technologies to help predict metastases in CRC and in Computer-Aided Detection (CAD) Systems to improve miss rates for colorectal neoplasia. This article provides a comprehensive review of the role of AI in predicting risk, prognosis, and response to therapies among patients with CRC.

Suicide Risk Among Patients With Heart Disease and Heart Failure.

Heart disease is the leading cause of death in the United States (US). Suicide is the 12th leading cause of death. However, little is known about the risk of suicide in patients with heart disease and heart failure. Using Multiple Cause of Death data from the Centers for Disease Control and Prevention (CDC) Wide-ranging ONline Data for Epidemiologic Research (WONDER) database, we used a cross-sectional analysis to examine the risk of suicide in patients with heart disease and heart failure as compared with the general US population and in subgroups within the US. We assessed suicide rate in patients with heart disease and heart failure in the US population as a whole and stratified by race, time, urbanization levels, and census regions using the CDC WONDER Multiple Cause of Death database. Standardized mortality rates were calculated as observed deaths divided by expected deaths. As compared with the general population, patients with heart disease and heart failure had an elevated risk of suicide. This was true across racial and geographic subgroups. There was an elevated risk of suicide in patients with heart disease and heart failure in the United States. For heart disease, there were particular elevations in the Western US, and there was a particular elevation in Black Americans compared with the age-matched population.

Primary and Metastatic Lung Cancer: Rationale, Indications, and Outcomes of Thermal Ablation.

The widespread use of imaging as well as the efforts conducted through screening campaigns has dramatically increased the early detection rate of lung cancer. Historically, the management of lung cancer has heavily relied on surgery. However, the increased proportion of patients with comorbidities has given significance to less invasive therapeutic options like minimally invasive surgery and image-guided thermal ablation, which could precisely target the tumor without requiring general anesthesia or a thoracotomy. Thermal ablation is considered low-risk for lung tumors smaller than 3 cm that are located in peripheral lung and do not involve major blood vessels or airways. The rationale for ablative therapies relies on the fact that focused delivery of energy induces cell death and pathologic necrosis. Image-guided percutaneous thermal ablation therapies are established techniques in the local treatment of hepatic, renal, bone, thyroid and uterine lesions. In the lung, and specifically in the setting of metastatic disease, the 3 main indications for lung ablation are to serve as (1) curative intent, (2) as a strategy to achieve a chemo-holiday in oligometastatic disease, and (3) in oligoprogressive disease. Following these premises, the current paper aims to review the rationale, indications, and outcomes of thermal ablation as a form of local therapy in the treatment of primary and metastatic lung disease.

Role of Machine Learning-Based CT Body Composition in Risk Prediction and Prognostication: Current State and Future Directions.

CT body composition analysis has been shown to play an important role in predicting health and has the potential to improve patient outcomes if implemented clinically. Recent advances in artificial intelligence and machine learning have led to high speed and accuracy for extracting body composition metrics from CT scans. These may inform preoperative interventions and guide treatment planning. This review aims to discuss the clinical applications of CT body composition in clinical practice, as it moves towards widespread clinical implementation.

Potential Immunotherapy Targets for Liver-Directed Therapies, and the Current Scope of Immunotherapeutics for Liver-Related Malignancies.

Liver cancer, including hepatocellular carcinoma and intrahepatic cholangiocarcinoma, is increasing in incidence and mortality across the globe. An improved understanding of the complex tumor microenvironment has opened many therapeutic doors and led to the development of novel pharmaceuticals targeting cellular signaling pathways or immune checkpoints. These interventions have significantly improved tumor control rates and patient outcomes, both in clinical trials and in real-world practice. Interventional radiologists play an important role in the multidisciplinary team given their expertise in minimally invasive locoregional therapy, as the bulk of these tumors are usually in the liver. The aim of this review is to highlight the immunological therapeutic targets for primary liver cancers, the available immune-based approaches, and the contributions that interventional radiology can provide in the care of these patients.

Suicide among Cancer Patients: Current Knowledge and Directions for Observational Research.

Cancer is a major public health concern associated with an increased risk of psychosocial distress and suicide. The reasons for this increased risk are still being characterized. The purpose of this study is to highlight existing observational studies on cancer-related suicides in the United States and identify gaps for future research. This work helps inform clinical and policy decision-making on suicide prevention interventions and ongoing research on the detection and quantification of suicide risk among cancer patients. We identified 73 peer-reviewed studies (2010-2022) that examined the intersection of cancer and suicide using searches of PubMed and Embase. Overall, the reviewed studies showed that cancer patients have an elevated risk of suicide when compared to the general population. In general, the risk was higher among White, male, and older cancer patients, as well as among patients living in rural areas and with lower socioeconomic status. Future studies should further investigate the psychosocial aspects of receiving a diagnosis of cancer on patients' mental health as well as the impact of new treatments and their availability on suicide risk and disparities among cancer patients to better inform policies.

Primary clear cell adenocarcinoma of the lung: a national analysis.

Background: Primary clear cell adenocarcinoma of the lung (CCAL) is a rare form of lung cancer with poorly understood clinical features. We sought to investigate the clinicopathological characteristics and independent prognostic factors of primary CCAL. Methods: Overall survival (OS) of patients with CCAL in the National Cancer Database (NCDB) from 2004 to 2017 was compared to lung adenocarcinoma using Kaplan-Meier analysis, multivariable Cox proportional hazards modeling, and propensity score matching. Independent prognostic indicators for patients with CCAL were determined using multivariable Cox proportional hazards analysis. Results: A total of 1,396 CCAL and 462,360 lung adenocarcinoma patients were included in our analysis. When compared to patients diagnosed with lung adenocarcinoma, those diagnosed with CCAL were more likely to be younger, white, reside farther from a hospital, have higher Charlson/Deyo comorbidity condition (CDCC) scores, have private insurance, have T1, N0, M0 status. In unadjusted analysis, patients with CCAL had better survival than those with lung adenocarcinoma, although no significant differences in survival were found between the two groups with multivariable Cox proportional hazards and propensity score-matched analyses. Conclusions: In this national analysis, we found that the clinicopathological characteristics of CCAL are distinct from those of lung adenocarcinoma, but CCAL is not itself an independent predictor of survival after multivariable adjustment or propensity score-matched analysis.

Open, Video- and Robot-Assisted Thoracoscopic Lobectomy for Stage II-IIIA Non-Small Cell Lung Cancer.

BACKGROUND: This study compares the short- and long-term outcomes of open vs robotic vs video-assisted thoracoscopic surgery (VATS) lobectomy for stage II-IIIA non-small-cell lung cancer (NSCLC). METHODS: Outcomes of patients with stage II-IIIA NSCLC (excluding T4 tumors) who received open and minimally invasive surgery (MIS) lobectomy in the National Cancer Database from 2010 to 2017 were assessed using propensity score-matched analysis. RESULTS: A propensity score-matched analysis of 4652 open and 4652 MIS patients demonstrated a decreased median length of stay associated with MIS compared with open lobectomy (5 vs 6 days; P < .001). There were no significant differences in 30-day mortality, 30-day readmission, or overall survival between the open and MIS groups. A propensity score-matched analysis of 1186 VATS and 1186 robotic patients showed that compared with VATS, the robotic approach was associated with no significant differences in 30-day mortality, 30-day readmission, and overall survival. However, the robotic group had a decreased median length of stay compared with VATS (4 vs 5 days; P < .001). The conversion rate was also significantly lower for robotic compared with VATS lobectomy (8.9% vs 15.9%, P < .001). CONCLUSIONS: No significant differences were found in long-term survival between open and MIS lobectomy and between VATS and robotic lobectomy for stage II-IIIA NSCLC. However, the MIS approach was associated with a decreased length of stay compared with the open approach. The robotic approach was associated with decreased length of stay and decreased conversion rate compared with the VATS approach.

Deaths Due to COVID-19 in Patients With Cancer During Different Waves of the Pandemic in the US.

Importance: With the ongoing relaxation of guidelines to prevent COVID-19 transmission, particularly in hospital settings, medically vulnerable groups, such as patients with cancer, may experience a disparate burden of COVID-19 mortality compared with the general population. Objective: To evaluate COVID-19 mortality among US patients with cancer compared with the general US population during different waves of the pandemic. Design, Setting, and Participants: This cross-sectional study used data from the Center for Disease Control and Prevention's Wide-Ranging Online Data for Epidemiologic Research database to examine COVID-19 mortality among US patients with cancer and the general population from March 1, 2020, to May 31, 2022. The number of deaths due to COVID-19 during the 2021 to 2022 winter Omicron surge was compared with deaths during the preceding year's COVID-19 winter surge (when the wild-type SARS-CoV-2 variant was predominant) using mortality ratios. Data were analyzed from July 21 through August 31, 2022. Exposures: Pandemic wave during which the wild-type variant (December 2020 to February 2021), Delta variant (July 2021 to November 2021), or Omicron variant (December 2021 to February 2022) was predominant. Main Outcomes and Measures: Number of COVID-19 deaths per month. Results: The sample included 34 350 patients with cancer (14 498 females [42.2%] and 19 852 males [57.8%]) and 628 156 members of the general public (276 878 females [44.1%] and 351 278 males [55.9%]) who died from COVID-19 when the wild-type (December 2020-February 2021), Delta (July 2021-November 2021), and winter Omicron (December 2021-February 2022) variants were predominant. Among patients with cancer, the greatest number of COVID-19 deaths per month occurred during the winter Omicron period (n = 5958): at the peak of the winter Omicron period, there were 18% more deaths compared with the peak of the wild-type period. In contrast, among the general public, the greatest number of COVID-19 deaths per month occurred during the wild-type period (n = 105 327), and at the peak of the winter Omicron period, there were 21% fewer COVID-19 deaths compared with the peak of the wild-type period. In subgroup analyses by cancer site, COVID-19 mortality increased the most, by 38%, among patients with lymphoma during the winter Omicron period vs the wild-type period. Conclusions and Relevance: Findings of this cross-sectional study suggest that patients with cancer had a disparate burden of COVID-19 mortality during the winter Omicron wave compared with the general US population. With the emergence of new, immune-evasive SARS-CoV-2 variants, many of which are anticipated to be resistant to monoclonal antibody treatments, strategies to prevent COVID-19 transmission should remain a high priority.

An Investigation of Cancer-Directed Surgery for Different Histologic Subtypes of Malignant Pleural Mesothelioma.

BACKGROUND: The role of cancer-directed surgery in the treatment of stage I-IIIA malignant pleural mesothelioma (MPM) by histologic subtypes remains controversial. The objective of this study was to evaluate the survival of the different histologic subtypes for stage I-IIIA MPM stratified by cancer-directed surgery and nonoperative management. RESEARCH QUESTION: How is the histologic subtype, clinical stage, and use of cancer-directed surgery for MPM associated with overall survival? STUDY DESIGN AND METHODS: Overall survival of patients with stage I-IIIA epithelioid, sarcomatoid, and biphasic MPM in the National Cancer Database from 2004 through 2017 who underwent cancer-directed surgery (ie, surgery with or without chemotherapy or radiation) or chemotherapy with or without radiation (nonoperative management) was evaluated using Kaplan-Meier analysis, multivariable Cox proportional hazards analysis, and propensity score-matched analysis. RESULTS: Of 2,285 patients with stage I-IIIA MPM who met inclusion criteria, histologic subtype was epithelioid in 71% of patients, sarcomatoid in 12% of patients, and biphasic in 17% of patients. Median survival was 20 months in the epithelioid group, 8 months in the sarcomatoid group, and 13 months in the biphasic group (P < .01). Among patients who underwent surgery, median survival was 25 months in the epithelioid group, 8 months in the sarcomatoid group, and 15 months in the biphasic group (P < .01). In multivariable Cox proportional hazards analyses, surgery was associated with improved survival in the epithelioid group (P < .01) but not in the sarcomatoid (P = .63) or biphasic (P = .21) groups. These findings were consistent in propensity score-matched analyses for each MPM histologic type. INTERPRETATION: In this national analysis, cancer-directed surgery was found to be associated with improved survival for stage I-IIIA epithelioid MPM, but not for biphasic or sarcomatoid MPM.

Machine Learning Improves the Prediction of Responses to Immune Checkpoint Inhibitors in Metastatic Melanoma.

Pretreatment LDH is a standard prognostic biomarker for advanced melanoma and is associated with response to ICI. We assessed the role of machine learning-based radiomics in predicting responses to ICI and in complementing LDH for prognostication of metastatic melanoma. From 2008-2022, 79 patients with 168 metastatic hepatic lesions were identified. All patients had arterial phase CT images 1-month prior to initiation of ICI. Response to ICI was assessed on follow-up CT at 3 months using RECIST criteria. A machine learning algorithm was developed using radiomics. Maximum relevance minimum redundancy (mRMR) was used to select features. ROC analysis and logistic regression analyses evaluated performance. Shapley additive explanations were used to identify the variables that are the most important in predicting a response. mRMR selection revealed 15 features that are associated with a response to ICI. The machine learning model combining both radiomics features and pretreatment LDH resulted in better performance for response prediction compared to models that included radiomics or LDH alone (AUC of 0.89 (95% CI: [0.76-0.99]) vs. 0.81 (95% CI: [0.65-0.94]) and 0.81 (95% CI: [0.72-0.91]), respectively). Using SHAP analysis, LDH and two GLSZM were the most predictive of the outcome. Pre-treatment CT radiomic features performed equally well to serum LDH in predicting treatment response.