Applying AI and Guidelines to Assist Medical Students in Recognizing Patients With Heart Failure: Protocol for a Randomized Trial.

BACKGROUND: The integration of artificial intelligence (AI) into clinical practice is transforming both clinical practice and medical education. AI-based systems aim to improve the efficacy of clinical tasks, enhancing diagnostic accuracy and tailoring treatment delivery. As it becomes increasingly prevalent in health care for high-quality patient care, it is critical for health care providers to use the systems responsibly to mitigate bias, ensure effective outcomes, and provide safe clinical practices. In this study, the clinical task is the identification of heart failure (HF) prior to surgery with the intention of enhancing clinical decision-making skills. HF is a common and severe disease, but detection remains challenging due to its subtle manifestation, often concurrent with other medical conditions, and the absence of a simple and effective diagnostic test. While advanced HF algorithms have been developed, the use of these AI-based systems to enhance clinical decision-making in medical education remains understudied. OBJECTIVE: This research protocol is to demonstrate our study design, systematic procedures for selecting surgical cases from electronic health records, and interventions. The primary objective of this study is to measure the effectiveness of interventions aimed at improving HF recognition before surgery, the second objective is to evaluate the impact of inaccurate AI recommendations, and the third objective is to explore the relationship between the inclination to accept AI recommendations and their accuracy. METHODS: Our study used a 3 × 2 factorial design (intervention type × order of prepost sets) for this randomized trial with medical students. The student participants are asked to complete a 30-minute e-learning module that includes key information about the intervention and a 5-question quiz, and a 60-minute review of 20 surgical cases to determine the presence of HF. To mitigate selection bias in the pre- and posttests, we adopted a feature-based systematic sampling procedure. From a pool of 703 expert-reviewed surgical cases, 20 were selected based on features such as case complexity, model performance, and positive and negative labels. This study comprises three interventions: (1) a direct AI-based recommendation with a predicted HF score, (2) an indirect AI-based recommendation gauged through the area under the curve metric, and (3) an HF guideline-based intervention. RESULTS: As of July 2023, 62 of the enrolled medical students have fulfilled this study's participation, including the completion of a short quiz and the review of 20 surgical cases. The subject enrollment commenced in August 2022 and will end in December 2023, with the goal of recruiting 75 medical students in years 3 and 4 with clinical experience. CONCLUSIONS: We demonstrated a study protocol for the randomized trial, measuring the effectiveness of interventions using AI and HF guidelines among medical students to enhance HF recognition in preoperative care with electronic health record data. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/49842.

Frailty is associated with worse cognitive functioning in older adults.

Introduction: Frailty and impaired cognitive functioning often co-occur in older adults and are associated with adverse health outcomes. However, their relationship is unclear. This study sought to examine the association of frailty status with cognitive functioning in older adults. Method: The study population consisted of 2,296 older adults aged ≥60 from the National Health and Nutrition Examination Survey 2011-2014. Frailty status was measured based on the Fried Phenotype and the participants were categorized into three groups- robust, pre-frailty, and frailty. Cognitive functioning was measured using the Consortium to Establish a Registry for Alzheimer's Disease Word Learning subtest (CERAD-WL) immediate and delayed recall tests, the Animal Fluency test (AFT), and the Digit Symbol Substitution Test (DSST). Test-specific and global cognition z-scores were calculated. Multinomial linear regression models were constructed to examine the association between frailty status (reference: robust) and test-specific and global cognition z-scores. Multiple linear regression models were used to examine the relationship between the number of frailty dimensions and test-specific and global cognition z-scores. All models controlled for age, race/ethnicity, education, total cholesterol level, and systolic blood pressure. Results: About half of the participants (median age 68 years) were female (49.9%) and non-Hispanic White (48.7%). A quarter (23.3%) of the participants completed some college and above. Multinominal linear regression showed that compared with participants who were robust, those with frailty had worse DSST [ß = -0.234, 95% confidence interval (CI): -0.391, -0.078, P = 0.003] and global cognition z scores (ß = -0.129, 95% CI -0.233, -0.025, P = 0.02). Multiple linear regression model showed that the number of frailty dimensions was significantly associated with decreased the DSST (ß = -0.065, 95% CI -0.103, -0.026, P = 0.001) and global cognition z-scores (ß= -0.034, 95% CI -0.06, -0.009, P = 0.009). Conclusion: Frailty is associated with worse processing speed, sustained attention, working memory, and global cognition in older adults. Prevention and treatment of frailty in older adults may help protect their cognitive functioning. Further, clinicians should consider assessing cognitive functioning, especially processing speed, sustained attention, and working memory, among frail older patients, which may allow early identification and interventions of cognitive impairment.

Automatic Extraction of Medication Mentions from Tweets-Overview of the BioCreative VII Shared Task 3 Competition.

This study presents the outcomes of the shared task competition BioCreative VII (Task 3) focusing on the extraction of medication names from a Twitter user's publicly available tweets (the user's 'timeline'). In general, detecting health-related tweets is notoriously challenging for natural language processing tools. The main challenge, aside from the informality of the language used, is that people tweet about any and all topics, and most of their tweets are not related to health. Thus, finding those tweets in a user's timeline that mention specific health-related concepts such as medications requires addressing extreme imbalance. Task 3 called for detecting tweets in a user's timeline that mentions a medication name and, for each detected mention, extracting its span. The organizers made available a corpus consisting of 182 049 tweets publicly posted by 212 Twitter users with all medication mentions manually annotated. The corpus exhibits the natural distribution of positive tweets, with only 442 tweets (0.2%) mentioning a medication. This task was an opportunity for participants to evaluate methods that are robust to class imbalance beyond the simple lexical match. A total of 65 teams registered, and 16 teams submitted a system run. This study summarizes the corpus created by the organizers and the approaches taken by the participating teams for this challenge. The corpus is freely available at https://biocreative.bioinformatics.udel.edu/tasks/biocreative-vii/track-3/. The methods and the results of the competing systems are analyzed with a focus on the approaches taken for learning from class-imbalanced data.

Multicenter Validation of Natural Language Processing Algorithms for the Detection of Common Data Elements in Operative Notes for Total Hip Arthroplasty: Algorithm Development and Validation.

BACKGROUND: Natural language processing (NLP) methods are powerful tools for extracting and analyzing critical information from free-text data. MedTaggerIE, an open-source NLP pipeline for information extraction based on text patterns, has been widely used in the annotation of clinical notes. A rule-based system, MedTagger-total hip arthroplasty (THA), developed based on MedTaggerIE, was previously shown to correctly identify the surgical approach, fixation, and bearing surface from the THA operative notes at Mayo Clinic. OBJECTIVE: This study aimed to assess the implementability, usability, and portability of MedTagger-THA at two external institutions, Michigan Medicine and the University of Iowa, and provide lessons learned for best practices. METHODS: We conducted iterative test-apply-refinement processes with three involved sites-the development site (Mayo Clinic) and two deployment sites (Michigan Medicine and the University of Iowa). Mayo Clinic was the primary NLP development site, with the THA registry as the gold standard. The activities at the two deployment sites included the extraction of the operative notes, gold standard development (Michigan: registry data; Iowa: manual chart review), the refinement of NLP algorithms on training data, and the evaluation of test data. Error analyses were conducted to understand language variations across sites. To further assess the model specificity for approach and fixation, we applied the refined MedTagger-THA to arthroscopic hip procedures and periacetabular osteotomy cases, as neither of these operative notes should contain any approach or fixation keywords. RESULTS: MedTagger-THA algorithms were implemented and refined independently for both sites. At Michigan, the study comprised THA-related notes for 2569 patient-date pairs. Before model refinement, MedTagger-THA algorithms demonstrated excellent accuracy for approach (96.6%, 95% CI 94.6%-97.9%) and fixation (95.7%, 95% CI 92.4%-97.6%). These results were comparable with internal accuracy at the development site (99.2% for approach and 90.7% for fixation). Model refinement improved accuracies slightly for both approach (99%, 95% CI 97.6%-99.6%) and fixation (98%, 95% CI 95.3%-99.3%). The specificity of approach identification was 88.9% for arthroscopy cases, and the specificity of fixation identification was 100% for both periacetabular osteotomy and arthroscopy cases. At the Iowa site, the study comprised an overall data set of 100 operative notes (50 training notes and 50 test notes). MedTagger-THA algorithms achieved moderate-high performance on the training data. After model refinement, the model achieved high performance for approach (100%, 95% CI 91.3%-100%), fixation (98%, 95% CI 88.3%-100%), and bearing surface (92%, 95% CI 80.5%-97.3%). CONCLUSIONS: High performance across centers was achieved for the MedTagger-THA algorithms, demonstrating that they were sufficiently implementable, usable, and portable to different deployment sites. This study provided important lessons learned during the model deployment and validation processes, and it can serve as a reference for transferring rule-based electronic health record models.

Improvement of base editors and prime editors advances precision genome engineering in plants.

Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR associated protein (Cas)-mediated gene disruption has revolutionized biomedical research as well as plant and animal breeding. However, most disease-causing mutations and agronomically important genetic variations are single base polymorphisms (single-nucleotide polymorphisms) that require precision genome editing tools for correction of the sequences. Although homology-directed repair of double-stranded breaks (DSBs) can introduce precise changes, such repairs are inefficient in differentiated animal and plant cells. Base editing and prime editing are two recently developed genome engineering approaches that can efficiently introduce precise edits into target sites without requirement of DSB formation or donor DNA templates. They have been applied in several plant species with promising results. Here, we review the extensive literature on improving the efficiency, target scope, and specificity of base editors and prime editors in plants. We also highlight recent progress on base editing in plant organellar genomes and discuss how these precision genome editing tools are advancing basic plant research and crop breeding.

Improving Early Identification of Significant Weight Loss Using Clinical Decision Support System in Lung Cancer Radiation Therapy.

PURPOSE: Early identification of patients who may be at high risk of significant weight loss (SWL) is important for timely clinical intervention in lung cancer radiotherapy (RT). A clinical decision support system (CDSS) for SWL prediction was implemented within the routine clinical workflow and assessed on a prospective cohort of patients. MATERIALS AND METHODS: CDSS incorporated a machine learning prediction model on the basis of radiomics and dosiomics image features and was connected to a web-based dashboard for streamlined patient enrollment, feature extraction, SWL prediction, and physicians' evaluation processes. Patients with lung cancer (N = 37) treated with definitive RT without prior RT were prospectively enrolled in the study. Radiomics and dosiomics features were extracted from CT and 3D dose volume, and SWL probability (≥ 0.5 considered as SWL) was predicted. Two physicians predicted whether the patient would have SWL before and after reviewing the CDSS prediction. The physician's prediction performance without and with CDSS and prediction changes before and after using CDSS were compared. RESULTS: CDSS showed significantly better prediction accuracy than physicians (0.73 v 0.54) with higher specificity (0.81 v 0.50) but with lower sensitivity (0.55 v 0.64). Physicians changed their original prediction after reviewing CDSS prediction for four cases (three correctly and one incorrectly), for all of which CDSS prediction was correct. Physicians' prediction was improved with CDSS in accuracy (0.54-0.59), sensitivity (0.64-0.73), specificity (0.50-0.54), positive predictive value (0.35-0.40), and negative predictive value (0.76-0.82). CONCLUSION: Machine learning-based CDSS showed the potential to improve SWL prediction in lung cancer RT. More investigation on a larger patient cohort is needed to properly interpret CDSS prediction performance and its benefit in clinical decision making.

Radiation-Induced Skin Dermatitis: Treatment With CamWell® Herb to Soothe® Cream in Patients With Head and Neck Cancer Receiving Radiation Therapy.

BACKGROUND: Radiation-induced skin dermatitis (RISD) is a common outcome experienced by adult patients with head and neck cancer (HNC) who have undergone radiation therapy. There is no standardized recommended agent for the prevention or management of RISD. OBJECTIVES: The primary objective of this study was to retrospectively evaluate for effectiveness of a botanical topical agent, CamWell® Herb to Soothe® cream, on RISD. METHODS: 112 patients with HNC undergoing radiation therapy self-reported their RISD topical skin care agent during treatment as standard of care, CamWell used prophylactically, or CamWell use started after the first week of treatment. The primary endpoint was impact of RISD on the patient, as measured by mean Skindex-16 score throughout treatment. Measures were completed weekly. FINDINGS: The mean Skindex score was statistically significantly lower for the prophylactic group than for the standard-of-care group. CamWell may have played a role in managing RISD when compared to standard-of-care agents.

Early Elevation of Fibrosis-4 Liver Fibrosis Score Is Associated With Adverse Outcomes Among Patients With Coronavirus Disease 2019.

BACKGROUND: Limited prior data suggest that preexisting liver disease is associated with adverse outcomes among patients with coronavirus disease 2019 (COVID-19). Fibrosis-4 (FIB-4) is a noninvasive index of readily available laboratory measurements that represents hepatic fibrosis. We evaluated the association between FIB-4 at the early stage of infection and COVID-19 outcomes. METHODS: FIB-4 was evaluated at admission in a cohort of 267 patients admitted with early-stage COVID-19 confirmed through reverse-transcription polymerase chain reaction assay. Hazard of ventilator use and of high-flow oxygen was estimated using Cox regression models controlled for covariates. Risks of progression to severe disease and of death/prolonged hospitalization were estimated using multivariable logistic regression models. RESULTS: Forty-one (15%) patients progressed to severe disease, 36 (14%) required high-flow oxygen support, 10 (4%) required mechanical ventilator support, and 1 died. FIB-4 between 1.45 and 3.25 was associated with a greater than 5-fold (95% confidence interval [CI], 1.2-28) increased hazard of high-flow oxygen use, a greater than 4-fold (95% CI, 1.5-14.6) increased odds of progression to severe disease, and an over 3-fold (95% CI, 1.4-7.7) increased odds of death or prolonged hospitalization. FIB-4 >3.25 was associated with a greater than 12-fold (95% CI, 2.3-68. 7) increased hazard of high-flow oxygen use and an over 11-fold (95% CI, 3.1-45) increased risk of progression to severe disease. All associations were independent of sex, number of comorbidities, and inflammatory markers (D-dimer, C-reactive protein). CONCLUSIONS: FIB-4 at the early-stage of COVID-19 had an independent and dose-dependent association with adverse outcomes during hospitalization. FIB-4 provided significant prognostic value for estimating adverse outcomes among COVID-19 patients.

Longitudinal Trends of Financial Toxicity in Patients With Lung Cancer: A Prospective Cohort Study.

BACKGROUND: Cancer therapy is associated with severe financial burden. However, the magnitude and longitudinal patient relationship with financial toxicity (FT) in the initial course of therapy is unclear. METHODS: Patients with stage II-IV lung cancer were recruited in a prospective longitudinal study between July 2018 and March 2020. FT was measured via the validated COmprehensive Score for financial Toxicity (COST) at the time of cancer diagnosis and at 6-month follow-up (6MFU). 6MFU data were compared with corresponding baseline data. A lower COST score indicates increased financial hardship. RESULTS: At the time of analysis, 215 agreed to participate. Subsequently, 112 patients completed 6MFU. On average, slightly more FT was observed at diagnosis compared with 6MFU (median COSTbase 25 v COST6M 27; P < .001); however, individual patients experienced large changes in FT. At 6MFU, 27.7% of patients had made financial sacrifices to pay for treatment but only 4.5% refused medical care based on cost. Median reported out-of-pocket (OOP) costs for the initial 6 months of cancer treatment was $2,496 (range, $0-25,900). Risk factors for FT at diagnosis were unique from risk factors at 6MFU. Actual OOP expenses were not correlated with FT; however, inability to predict upcoming treatment expenses resulted in higher FT at 6MFU. DISCUSSION: FT is a pervasive challenge during the initiation of lung cancer treatment. Few patients are willing to sacrifice medical care regardless of the cost. Risk factors for FT evolve, resulting in unique interventional targets throughout therapy.

A multi-institutional phase 2 trial of stereotactic body radiotherapy in the treatment of bone metastases in pediatric and young adult patients with sarcoma.

BACKGROUND: Metastasectomy is standard of care for pediatric patients with metastatic sarcoma with limited disease. For patients with unresectable disease, stereotactic body radiotherapy (SBRT) may serve as an alternative. Herein, the authors report the results of a prospective, multi-institutional phase 2 trial of SBRT in children and young adults with metastatic sarcoma. METHODS: Patients aged >3 years and ≤40 years with unresected, osseous metastatic nonrhabdomyosarcoma sarcomas of soft tissue and bone were eligible. Patients received SBRT to a dose of 40 Gray (Gy) in 5 fractions. Local control (LC), progression-free survival (PFS), and overall survival (OS) were calculated using the Kaplan-Meier method. RESULTS: Fourteen patients with a median age of 17 years (range, 4-25 years) were treated to 37 distinct metastatic lesions. With a median follow-up of 6.8 months (30.5 months in surviving patients), the Kaplan-Meier patient-specific and lesion-specific LC rates at 6 months were 89% and 95%, respectively. The median PFS was 6 months and the median OS was 24 months. In a post hoc analysis, PFS (median, 9.3 months vs 3.7 months; log-rank P = .03) and OS (median not reached vs 12.7 months; log-rank P = .02) were improved when all known sites of metastatic disease were consolidated with SBRT compared with partial consolidation. SBRT was well tolerated, with 2 patients experiencing grade 3 toxicities. CONCLUSIONS: SBRT achieved high rates of LC in pediatric patients with inoperable metastatic nonrhabdomyosarcoma sarcomas of soft tissue and bone. These results suggest that the ability to achieve total consolidation of metastatic disease with SBRT is associated with improved PFS and OS.

Patient-Reported Outcome Measures and Dosimetric Correlates for Early Detection of Acute Radiation Therapy-Related Esophagitis.

BACKGROUND: We investigate the time to and clinical factors associated with patient-reported difficulty swallowing in lung cancer patients treated with radiation therapy (RT). METHODS: Between October 2016 and October 2019, lung cancer patients treated with conventionally fractionated RT at a tertiary cancer center were identified. Weekly, patients reported difficulty swallowing (patient-reported outcome version of the Common Terminology Criteria for Adverse Events [PRO-CTCAE] v.1: 0-none, 1-mild, 2-moderate, 3-severe, 4-very severe). Physicians graded dysphagia (CTCAE v.4: 0-none, 1-symptoms without altered intake, 2-symptomatic; altered eating/swallowing, 3-severely altered eating/swallowing, 4-life-threatening consequences, 5-death). Tumor-related difficulty swallowing was not recorded at baseline; thus, patients reporting ≥moderate symptoms ≤7 days of RT start were excluded. We evaluated the time to new patient reports of ≥moderate difficulty swallowing and CTCAE grade 2+ dysphagia and development over time using the cumulative incidence method. Multivariable logistic regression evaluated associations between clinical factors, esophageal V60, and development of esophageal symptoms. RESULTS: Of the 200 patients identified: median age was 69 years, 52% were male, and 89% had stage III+ disease. Patients received a median of 63 Gy with chemotherapy (91.5%). At least moderate difficulty swallowing during RT was reported by 76 of 200 patients (38%); clinicians rated dysphagia as altering oral intake or worse for 26 of 200 (13%). Median time to first report of symptoms was 21 days (interquartile ratio [IQR], 18-34.5) for the 76 patients who reported ≥moderate symptoms and 33 days (IQR, 24-42) in the 26 patients whose provider reported grade 2+ dysphagia. The 30-day incidence of patient-reported ≥moderate swallowing difficulty and provider grade 2+ dysphagia was 26% (95% CI: 20%-32%) and 6% (95% CI: 3%-9%), respectively. Esophageal V60 >7 % was the clinical factor most associated with patient-reported ≥moderate esophageal symptoms (odds ratio 6.1, 95% CI: 3.0-12.3). CONCLUSIONS: Patients report at least moderate difficulty swallowing more often and earlier than providers report grade 2+ dysphagia. Esophageal V60 ≥7% was most associated with development of moderate severity or worse patient-reported swallowing difficulty.

Consolidative Radiotherapy in Oligometastatic Lung Cancer: Patient Selection With a Prediction Nomogram.

BACKGROUND: Patients with stage IV oligometastatic (≤ 3 sites) non-small-cell lung cancer have a progression-free survival (PFS) and overall survival benefit when all sites of metastatic disease and the primary tumor are treated radically with consolidative radiotherapy (cRT). However, the optimal selection of patients most likely from cRT is yet to be defined. PATIENTS AND METHODS: Patients with metastatic non-small-cell lung cancer treated with definitive radiotherapy to all metastatic sites and primary tumor (2008-2019) were retrospectively identified. Univariable Cox proportional-hazards model was used to compare outcomes with demographic and clinical characteristics. A predictive nomogram model for selection of patients most likely to benefit from cRT was constructed. RESULTS: There were 91 patients identified with a total of 114 metastases treated. Median PFS from the start of cRT was 10.9 months (95% confidence interval [CI], 8.1-16.6), while the median survival time was 37.0 months (95% CI, 31.3-NR). On univariable modeling, patients with squamous histology (hazard ratio, 4.16; 95% CI, 1.99-8.71; P < .001) and those treated with non-stereotactic body radiotherapy hypofractionated therapy (hazard ratio, 5.43; 95% CI, 2.10-14.01; P < .001) had worse overall survival, while patients with targetable mutations (hazard ratio, 0.49; 95% CI, 0.25-0.98; P = .04) had a longer survival. Using a predictive nomogram model, patients with a solitary site of metastasis, targetable mutations, intracranial disease, and metachronous timing of oligometastases had a larger PFS benefit from cRT. CONCLUSION: cRT is associated with favorable outcomes in PFS and overall survival. These results may aid in patient counseling, selection for aggressive local therapy, and stratification in future prospective clinical trials.

Isolated progression of metastatic lung cancer: Clinical outcomes associated with definitive radiotherapy.

BACKGROUND: Progressive, metastatic non-small cell lung cancer (NSCLC) often requires the initiation of new systemic therapy. However, in patients with NSCLC that is oligoprogressive (≤3 lesions), local radiotherapy (RT) may allow for the eradication of resistant microclones and, therefore, the continuation of otherwise effective systemic therapy. METHODS: Patients treated from 2008 to 2019 with definitive doses of RT to all sites of intracranial or extracranial oligoprogression without a change in systemic therapy were identified. Radiographic progression-free survival (rPFS) and time to new therapy (TNT) were measured. Associations between baseline clinical and treatment-related variables were correlated with progression-free survival via Cox proportional hazards modeling. RESULTS: Among 198 unique patients, 253 oligoprogressive events were identified. Intracranial progression occurred in 51% of the patients, and extracranial progression occurred in 49%. In the entire cohort, the median rPFS was 7.9 months (95% CI, 6.5-10.0 months), and the median TNT was 8.8 months (95% CI, 7.2-10.9 months). On adjusted modeling, patients with the following disease characteristics were associated with better rPFS: better performance status (P = .003), fewer metastases (P = .03), longer time to oligoprogression (P = .009), and fewer previous systemic therapies (P = .02). Having multiple sites of oligoprogression was associated with worse rPFS (P < .001). CONCLUSIONS: In select patients with oligoprogression, definitive RT is a feasible treatment option to delay the initiation of next-line systemic therapies, which have more limited response rates and efficacy. Further randomized prospective data may help to validate these findings and identify which patients are most likely to benefit.

Propensity score adjusted analysis of patients with isolated locoregional recurrence versus de novo locally advanced NSCLC treated with definitive therapy.

OBJECTIVES: Definitive intent treatment of isolated locoregional recurrence (iLR) for non-small cell lung cancer (NSCLC) is becoming more common. This study explores outcomes associated with the definitive local treatment of iLR and compares these outcomes to newly diagnosed locally advanced NSCLC (LA-NSCLC) patients. MATERIALS AND METHODS: Patients with NSCLC treated with curative therapy between 2008 and 2019 at a tertiary academic institution were screened for iLR treated with subsequent definitive salvage therapy. Progression free survival (PFS), time to distant metastasis (TTDM), and overall survival (OS) were calculated via Kaplan-Meier methodology. Clinical outcomes were compared to a separate group of patients with de novo LA-NSCLC after adjusting for propensity score (PS). RESULTS: Sixty five cases of definitively salvaged iLR were compared to 302 patients with de novo LA-NSCLC. Most patients were treated with chemoradiotherapy (83.1% in iLR, 74.5% in LA-NSCLC). The median PFS, TTDM, and OS for the iLR cohort was 16.7 months (95% CI: 9.6-24.7), 35.8 months (95% CI: 17.1-NR), and 49.5 months (95% CI: 30.1-NR), respectively. After adjusting for PS, the iLR group was no different from the LA-NSCLC group in risk for progression (HR 0.78, 95% CI: 0.53-1.16, p = 0.22), distant metastasis (HR 0.81, 95% CI: 0.52-1.27, p = 0.36), or death (HR 0.90, 95% CI: 0.47-1.73, p = 0.75). Patterns of failure did not different significantly between groups. In the iLR cohort, patients with older age (HR 1.06, 95 CI: 1.01-1.10, p = 0.01) had a higher risk of death on multivariate analysis. CONCLUSION: To our knowledge, this is the first report that compares the definitive treatment of iLR to de novo LA-NSCLC. When treated with definitive local therapy, patients with iLR had no difference in clinical outcomes from de novo LA-NSCLC. The use of curative local therapy according to a LA-NSCLC paradigm is advisable in patients with iLR of NSCLC for whom definitive therapy is feasible.

Hospitalization and definitive radiotherapy in lung cancer: incidence, risk factors and survival impact.

BACKGROUND: Unplanned hospitalization during cancer treatment is costly, can disrupt treatment, and affect patient quality of life. However, incidence and risks factors for hospitalization during lung cancer radiotherapy are not well characterized. METHODS: Patients treated with definitive intent radiation (≥45 Gy) for lung cancer between 2008 and 2018 at a tertiary academic institution were identified. In addition to patient, tumor, and treatment related characteristics, specific baseline frailty markers (Charlson comorbidity index, ECOG, patient reported weight loss, BMI, hemoglobin, creatinine, albumin) were recorded. All cancer-related hospitalizations during or within 30 days of completing radiation were identified. Associations between baseline variables and any hospitalization, number of hospitalizations, and overall survival were identified using multivariable linear regression and multivariable Cox proportional-hazards models, respectively. RESULTS: Of 270 patients included: median age was 66.6 years (31-88), 50.4% of patients were male (n = 136), 62% were Caucasian (n = 168). Cancer-related hospitalization incidence was 17% (n = 47), of which 21% of patients hospitalized (n = 10/47) had > 1 hospitalization. On multivariable analysis, each 1 g/dL baseline drop in albumin was associated with a 2.4 times higher risk of any hospitalization (95% confidence interval (CI) 1.2-5.0, P = 0.01), and baseline hemoglobin ≤10 was associated with, on average, 2.7 more hospitalizations than having pre-treatment hemoglobin > 10 (95% CI 1.3-5.4, P = 0.01). After controlling for baseline variables, cancer-related hospitalization was associated with 1.8 times increased risk of all-cause death (95% CI: 1.02-3.1, P = 0.04). CONCLUSIONS: Our data show baseline factors can predict those who may be at increased risk for hospitalization, which was independently associated with increased mortality. Taken together, these data support the need for developing further studies aimed at early and aggressive interventions to decrease hospitalizations during treatment.

Spatial Radiation Dose Influence on Xerostomia Recovery and Its Comparison to Acute Incidence in Patients With Head and Neck Cancer.

PURPOSE: Radiation-induced xerostomia is one of the most prevalent symptoms during and after head and neck cancer radiation therapy (RT). We aimed to discover the spatial radiation dose-based (voxel dose) importance pattern in the major salivary glands in relation to the recovery of xerostomia 18 months after RT, and to compare the recovery voxel dose importance pattern to the acute incidence (injury) pattern. METHODS AND MATERIALS: This study included all patients within our database with xerostomia outcomes after completion of curative intensity modulated RT. Common Terminology Criteria for Adverse Events xerostomia grade was used to define recovered versus nonrecovered group at baseline, between end of treatment and 18 months post-RT, and beyond 18 months, respectively. Ridge logistic regression was performed to predict the probability of xerostomia recovery. Voxel doses within geometrically defined parotid glands (PG) and submandibular glands (SMG), demographic characteristics, and clinical factors were included in the algorithm. We plotted the normalized learned weights on the 3-dimensional PG and SMG structures to visualize the voxel dose importance for predicting xerostomia recovery. RESULTS: A total of 146 head and neck cancer patients from 2008 to 2016 were identified. The superior region of the ipsilateral and contralateral PG was the most influencial for xerostomia recovery. The area under the receiver operating characteristic curve evaluated using 10-fold cross-validation for ridge logistic regression was 0.68 ± 0.07. Compared with injury, the recovery voxel dose importance pattern was more symmetrical and was influenced by lower dose voxels. CONCLUSIONS: The superior portion of the 2 PGs (low dose region) are the most influential on xerostomia recovery and seem to be equal in their contribution. The dissimilarity of the influence pattern between injury and recovery suggests different underlying mechanisms. The importance pattern identified by spatial radiation dose and machine learning methods can improve our understanding of normal tissue toxicities in RT. Further external validation is warranted.

Multi-view radiomics and dosiomics analysis with machine learning for predicting acute-phase weight loss in lung cancer patients treated with radiotherapy.

We propose a multi-view data analysis approach using radiomics and dosiomics (R&D) texture features for predicting acute-phase weight loss (WL) in lung cancer radiotherapy. Baseline weight of 388 patients who underwent intensity modulated radiation therapy (IMRT) was measured between one month prior to and one week after the start of IMRT. Weight change between one week and two months after the commencement of IMRT was analyzed, and dichotomized at 5% WL. Each patient had a planning CT and contours of gross tumor volume (GTV) and esophagus (ESO). A total of 355 features including clinical parameter (CP), GTV and ESO (GTV&ESO) dose-volume histogram (DVH), GTV radiomics, and GTV&ESO dosiomics features were extracted. R&D features were categorized as first- (L1), second- (L2), higher-order (L3) statistics, and three combined groups, L1 + L2, L2 + L3 and L1 + L2 + L3. Multi-view texture analysis was performed to identify optimal R&D input features. In the training set (194 earlier patients), feature selection was performed using Boruta algorithm followed by collinearity removal based on variance inflation factor. Machine-learning models were developed using Laplacian kernel support vector machine (lpSVM), deep neural network (DNN) and their averaged ensemble classifiers. Prediction performance was tested on an independent test set (194 more recent patients), and compared among seven different input conditions: CP-only, DVH-only, R&D-only, DVH + CP, R&D + CP, R&D + DVH and R&D + DVH + CP. Combined GTV L1 + L2 + L3 radiomics and GTV&ESO L3 dosiomics were identified as optimal input features, which achieved the best performance with an ensemble classifier (AUC = 0.710), having statistically significantly higher predictability compared with DVH and/or CP features (p < 0.05). When this performance was compared to that with full R&D-only features which reflect traditional single-view data, there was a statistically significant difference (p < 0.05). Using optimized multi-view R&D input features is beneficial for predicting early WL in lung cancer radiotherapy, leading to improved performance compared to using conventional DVH and/or CP features.

Exploring the Relationship of Radiation Dose Exposed to the Length of Esophagus and Weight Loss in Patients with Lung Cancer.

PURPOSE: We investigate whether esophageal dose-length parameters (Ldose) can robustly predict significant weight loss-≥5% weight loss during radiation therapy (RT) compared with the weight before RT-in patients with lung cancer treated with definitive intent. METHODS AND MATERIALS: Patients with lung cancer treated with conventionally fractionated RT between 2010 and 2018 were retrospectively identified. LFdose and LPdose, the length of full- and partial-circumferential esophagus receiving greater than a threshold dose in Gy, respectively, were created. Multivariate logistic regression examined the associations between individual Ldose and weight loss after adjusting for clinical parameters and correcting for multiple comparisons. Ridge logistic regression examined the relative importance of Ldose compared with dose-volume (Vdose), mean dose (Dmean), and clinical parameters in determining weight loss. Univariate logistic regression examined the unadjusted probability of weight loss for important Ldose parameters. RESULTS: Among the 214 patients identified, median age was 66.9 years (range, 31.5-88.9 years), 50.5% (n = 108) were male, 68.2% (n = 146) had stage III lung cancer, median RT dose was 63 Gy (range, 60-66 Gy), and 88.3% (n = 189) received concurrent chemotherapy. Esophagus lengths receiving high full-circumferential (LF50-LF60) and high partial-circumferential doses (LP60) were associated with significant weight loss (P ≤ .05). LF65 and LP65 reached near significance (P = .06 and .053, respectively). LF65 > LF60 > LP65 were the most important dose parameters in determining weight loss compared with other Ldose, Vdose, and Dmean parameters. CONCLUSIONS: Esophageal Ldose parameters are an efficient way of interpreting complex dose parameters in relation to weight loss toxicity among patients with lung cancer receiving definitive RT.

Predicting acute radiation induced xerostomia in head and neck Cancer using MR and CT Radiomics of parotid and submandibular glands.

PURPOSE: To analyze baseline CT/MR-based image features of salivary glands to predict radiation-induced xerostomia 3-months after head-and-neck cancer (HNC) radiotherapy. METHODS: A retrospective analysis was performed on 266 HNC patients who were treated using radiotherapy at our institution between 2009 and 2018. CT and T1 post-contrast MR images along with NCI-CTCAE xerostomia grade (3-month follow-up) were prospectively collected at our institution. CT and MR images were registered on which parotid/submandibular glands were contoured. Image features were extracted for ipsilateral/contralateral parotid and submandibular glands relative to the location of the primary tumor. Dose-volume-histogram (DVH) parameters were also acquired. Features were pre-selected based on Spearman correlation before modelling by examining the correlation with xerostomia (p < 0.05). A shrinkage regression analysis of the pre-selected features was performed using LASSO. The internal validity of the variable selection was estimated by repeating the entire variable selection procedure using a leave-one-out-cross-validation. The most frequently selected variables were considered in the final model. A generalized linear regression with repeated ten-fold cross-validation was developed to predict radiation-induced xerostomia at 3-months after radiotherapy. This model was tested in an independent dataset (n = 50) of patients who were treated at the same institution in 2017-2018. We compared the prediction performances under eight conditions (DVH-only, CT-only, MR-only, CT + MR, DVH + CT, DVH + CT + MR, Clinical+CT + MR, and Clinical+DVH + CT + MR) using the area under the receiver operating characteristic curve (ROC-AUC). RESULTS: Among extracted features, 7 CT, 5 MR, and 2 DVH features were selected. The internal cohort (n = 216) ROC-AUC values for DVH, CT, MR, and Clinical+DVH + CT + MR features were 0.73 ± 0.01, 0.69 ± 0.01, 0.70 ± 0.01, and 0.79 ± 0.01, respectively. The validation cohort (n = 50) ROC-AUC values for DVH, CT, MR, and Clinical+DVH + CT + MR features were 0.63, 0.57, 0.66, and 0.68, respectively. The DVH-ROC was not significantly different than the CT-ROC (p = 0.8) or MR-ROC (p = 0.4). However, the CT + MR-ROC was significantly different than the CT-ROC (p = 0.03), but not the Clinical+DVH + CT + MR model (p = 0.5). CONCLUSION: Our results suggest that baseline CT and MR image features may reflect baseline salivary gland function and potential risk for radiation injury. The integration of baseline image features into prediction models has the potential to improve xerostomia risk stratification with the ultimate goal of truly personalized HNC radiotherapy.

Applying Non-Homogeneous Dose Optimization to Improve Conventionally Fractionated Radiation Plan Quality in Patients with Non-Small Cell Lung Cancer.

PURPOSE: Nonhomogeneous dose optimization (NHDO) is exploited in stereotactic body radiation therapy (SBRT) to increase dose delivery to the tumor and allow rapid dose falloff to surrounding normal tissues. We investigate changes in plan quality when NHDO is applied to inverse-planned conventionally fractionated radiation therapy (CF-RT) plans in patients with non-small cell lung cancer. METHODS AND MATERIALS: Patients with near-central non-small cell lung cancer treated with CF-RT in 2018 at a single institution were identified. CF-RT plans were replanned using NHDO techniques, including normalizing to a lower isodose line, while maintaining clinically acceptable normal tissue constraints and target coverage. Tumor control probabilities were calculated. We compared delivered CF-RT plans using homogenous dose optimization (HDO) versus NHDO using Wilcoxon signed-rank tests. Median values are reported. RESULTS: Thirteen patients were replanned with NHDO techniques. Planning target volume coverage by the prescription dose was similar (NHDO = 96% vs HDO = 97%, P = .3). All normal-tissue dose constraints were met. NHDO plans were prescribed to a lower-prescription isodose line compared with HDO plans (85% vs 97%, P = .001). NHDO increased mean dose to the planning target volume (73 Gy vs 67 Gy), dose heterogeneity, and dose falloff gradient (P < .03). NHDO decreased mean dose to surrounding lungs, esophagus, and heart (relative reduction of 6%, 14%, and 15%, respectively; P < .05). Other normal tissue objectives improved with NHDO, including total lung V40 and V60, heart V30, and maximum esophageal dose (P < .05). Tumor control probabilities doubled from 31.6% to 65.4% with NHDO (P = .001). CONCLUSIONS: In select patients, NHDO principles used in SBRT optimization can be applied to CF-RT. NHDO results in increased tumor dose, reduction in select organ-at-risk dose objectives, and better maintenance of target coverage and normal-tissue constraints compared with HDO. Our data demonstrate that principles of NHDO used in SBRT can also improve plan quality in CF-RT.