1
|
Leng JX, Carpenter DJ, Huang C, Qazi J, Arshad M, Mullikin TC, Reitman ZJ, Kirkpatrick JP, Floyd SR, Fecci PE, Chmura SJ, Hong JC, Salama JK. Determinants of Symptomatic Intracranial Progression After an Initial Stereotactic Radiosurgery Course. Adv Radiat Oncol 2024; 9:101475. [PMID: 38690297 PMCID: PMC11059392 DOI: 10.1016/j.adro.2024.101475] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/28/2023] [Accepted: 02/04/2024] [Indexed: 05/02/2024] Open
Abstract
Purpose Clinical and imaging surveillance of patients with brain metastases is important after stereotactic radiosurgery (SRS) because many will experience intracranial progression (ITCP) requiring multidisciplinary management. The prognostic significance of neurologic symptoms at the time of ITCP is poorly understood. Methods and Materials This was a multi-institutional, retrospective cohort study from 2015 to 2020, including all patients with brain metastases completing an initial course of SRS. The primary outcome was overall survival (OS) by presence of neurologic symptoms at ITCP. OS, freedom from ITCP (FF-ITCP), and freedom from symptomatic ITCP (FF-SITCP) were assessed via Kaplan-Meier method. Cox proportional hazard models tested parameters impacting FF-ITCP and FF-SITCP. Results Among 1383 patients, median age was 63.4 years, 55% were female, and common primaries were non-small cell lung (49%), breast (15%), and melanoma (9%). At a median follow-up of 8.72 months, asymptomatic and symptomatic ITCP were observed in 504 (36%) and 194 (14%) patients, respectively. The majority of ITCP were distant ITCP (79.5%). OS was worse with SITCP (median, 10.2 vs 17.9 months, P < .001). SITCP was associated with clinical factors including total treatment volume (P = .012), melanoma histology (P = .001), prior whole brain radiation therapy (P = .003), number of brain metastases (P < .001), interval of 1 to 2 years from primary and brain metastasis diagnosis (P = .012), controlled extracranial disease (P = .042), and receipt of pre-SRS chemotherapy (P = .015). Patients who were younger and received post-SRS chemotherapy (P = .001), immunotherapy (P < .001), and targeted or small-molecule inhibitor therapy (P < .026) had better FF-SITCP. Conclusions In this cohort study of patients with brain metastases completing SRS, neurologic symptoms at ITCP is prognostic for OS. This data informs post-SRS surveillance in clinical practice as well as future prospective studies needed in the modern management of brain metastases.
Collapse
Affiliation(s)
- Jim X. Leng
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - David J. Carpenter
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
- Department of Radiation Oncology, Wellstar Paulding Hospital, Hiram, Georgia
| | - Christina Huang
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - Jamiluddin Qazi
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - Muzamil Arshad
- Department of Radiation and Cellular Oncology, University of Chicago Medical Center, Chicago, Illinois
| | - Trey C. Mullikin
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - Zachary J. Reitman
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - John P. Kirkpatrick
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
- Department of Neurosurgery, Duke University Medical Center, Durham, North Carolina
| | - Scott R. Floyd
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - Peter E. Fecci
- Department of Neurosurgery, Duke University Medical Center, Durham, North Carolina
| | - Steven J. Chmura
- Department of Radiation and Cellular Oncology, University of Chicago Medical Center, Chicago, Illinois
| | - Julian C. Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, California
- Bakar Computational Health Sciences Institute, University of California San Francisco, San Francisco, California
- Joint Program in Computational Precision Health, University of California, San Francisco, California and University of California, Berkeley, California
| | - Joseph K. Salama
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
- Radiation Oncology Clinical Service, Durham VA Health Care System, Durham, North Carolina
| |
Collapse
|
2
|
James B Yu Md Mhs Fastro, Hong JC. AI Use in Prostate Cancer: Potential Improvements in Treatments and Patient Care. Oncology (Williston Park) 2024; 38:208-209. [PMID: 38776517 DOI: 10.46883/2024.25921021] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/25/2024]
Abstract
Artificial intelligence use in prostate cancer encompasses 4 main areas including diagnostic imaging, prediction of outcomes, histopathology, and treatment planning.
Collapse
|
3
|
Natesan D, Eisenstein EL, Thomas SM, Eclov NCW, Dalal NH, Stephens SJ, Malicki M, Shields S, Cobb A, Mowery YM, Niedzwiecki D, Tenenbaum JD, Palta M, Hong JC. Health Care Cost Reductions with Machine Learning-Directed Evaluations during Radiation Therapy - An Economic Analysis of a Randomized Controlled Study. NEJM AI 2024; 1:10.1056/aioa2300118. [PMID: 38586278 PMCID: PMC10997376 DOI: 10.1056/aioa2300118] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 04/09/2024]
Abstract
BACKGROUND Machine learning (ML) may cost-effectively direct health care by identifying patients most likely to benefit from preventative interventions to avoid negative and expensive outcomes. System for High-Intensity Evaluation During Radiation Therapy (SHIELD-RT; NCT04277650) was a single-institution, randomized controlled study in which electronic health record-based ML accurately identified patients at high risk for acute care (emergency visit or hospitalization) during radiotherapy (RT) and targeted them for supplemental clinical evaluations. This ML-directed intervention resulted in decreased acute care utilization. Given the limited prospective data showing the ability of ML to direct interventions cost-efficiently, an economic analysis was performed. METHODS A post hoc economic analysis was conducted of SHIELD-RT that included RT courses from January 7, 2019, to June 30, 2019. ML-identified high-risk courses (≥10% risk of acute care during RT) were randomized to receive standard of care weekly clinical evaluations with ad hoc supplemental evaluations per clinician discretion versus mandatory twice-weekly evaluations. The primary outcome was difference in mean total medical costs during and 15 days after RT. Acute care costs were obtained via institutional cost accounting. Physician and intervention costs were estimated via Medicare and Medicaid data. Negative binomial regression was used to estimate cost outcomes after adjustment for patient and disease factors. RESULTS A total of 311 high-risk RT courses among 305 patients were randomized to the standard (n=157) or the intervention (n=154) group. Unadjusted mean intervention group supplemental visit costs were $155 per course (95% confidence interval, $142 to $168). The intervention group had fewer acute care visits per course (standard, 0.47; intervention, 0.31; P=0.04). Total mean adjusted costs were $3110 per course for the standard group and $1494 for the intervention group (difference in means, $1616 [95% confidence interval, $1450 to $1783]; P=0.03). CONCLUSIONS In this economic analysis of a randomized controlled, health care ML study, mandatory supplemental evaluations for ML-identified high-risk patients were associated with both reduced total medical costs and improved clinical outcomes. Further study is needed to determine whether economic results are generalizable. (Funded in part by The Duke Endowment, The Conquer Cancer Foundation, the Duke Department of Radiation Oncology, and the National Cancer Institute of the National Institutes of Health [R01CA277782]; ClinicalTrials.gov number, NCT04277650.).
Collapse
Affiliation(s)
- Divya Natesan
- Department of Radiation Oncology, University of North Carolina, Chapel Hill, NC
- Department of Radiation Oncology, Duke University, Durham, NC
| | | | - Samantha M Thomas
- Department of Biostatistics and Bioinformatics, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| | | | - Nicole H Dalal
- Department of Radiation Oncology, Duke University, Durham, NC
| | | | - Mary Malicki
- Department of Radiation Oncology, Duke University, Durham, NC
| | - Stacey Shields
- Department of Radiation Oncology, Duke University, Durham, NC
| | - Alyssa Cobb
- Department of Radiation Oncology, Duke University, Durham, NC
| | - Yvonne M Mowery
- Department of Radiation Oncology, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| | - Donna Niedzwiecki
- Department of Biostatistics and Bioinformatics, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| | | | - Manisha Palta
- Department of Radiation Oncology, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| | - Julian C Hong
- Department of Radiation Oncology, University of California, San Francisco, San Francisco
- Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco
- UCSF-UC Berkeley Joint Program in Computational Precision Health, San Francisco, San Francisco
| |
Collapse
|
4
|
Friesner ID, Feng J, Kalnicki S, Garg M, Ohri N, Hong JC. Machine Learning-Based Prediction of Hospitalization During Chemoradiotherapy With Daily Step Counts. JAMA Oncol 2024:2816984. [PMID: 38546697 PMCID: PMC10979356 DOI: 10.1001/jamaoncol.2024.0014] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/10/2023] [Accepted: 09/21/2023] [Indexed: 04/01/2024]
Abstract
Importance Toxic effects of concurrent chemoradiotherapy (CRT) can cause treatment interruptions and hospitalizations, reducing treatment efficacy and increasing health care costs. Physical activity monitoring may enable early identification of patients at high risk for hospitalization who may benefit from proactive intervention. Objective To develop and validate machine learning (ML) approaches based on daily step counts collected by wearable devices on prospective trials to predict hospitalizations during CRT. Design, Setting, and Participants This study included patients with a variety of cancers enrolled from June 2015 to August 2018 on 3 prospective, single-institution trials of activity monitoring using wearable devices during CRT. Patients were followed up during and 1 month following CRT. Training and validation cohorts were generated temporally, stratifying for cancer diagnosis (70:30). Random forest, neural network, and elastic net-regularized logistic regression (EN) were trained to predict short-term hospitalization risk based on a combination of clinical characteristics and the preceding 2 weeks of activity data. To predict outcomes of activity data, models based only on activity-monitoring features and only on clinical features were trained and evaluated. Data analysis was completed from January 2022 to March 2023. Main Outcomes and Measures Model performance was evaluated in terms of the receiver operating characteristic area under curve (ROC AUC) in the stratified temporal validation cohort. Results Step counts from 214 patients (median [range] age, 61 [53-68] years; 113 [52.8%] male) were included. EN based on step counts and clinical features had high predictive ability (ROC AUC, 0.83; 95% CI, 0.66-0.92), outperforming random forest (ROC AUC, 0.76; 95% CI, 0.56-0.87; P = .02) and neural network (ROC AUC, 0.80; 95% CI, 0.71-0.88; P = .36). In an ablation study, the EN model based on only step counts demonstrated greater predictive ability than the EN model with step counts and clinical features (ROC AUC, 0.85; 95% CI, 0.70-0.93; P = .09). Both models outperformed the EN model trained on only clinical features (ROC AUC, 0.53; 95% CI, 0.31-0.66; P < .001). Conclusions and Relevance This study developed and validated a ML model based on activity-monitoring data collected during prospective clinical trials. Patient-generated health data have the potential to advance predictive ability of ML approaches. The resulting model from this study will be evaluated in an upcoming multi-institutional, cooperative group randomized trial.
Collapse
Affiliation(s)
- Isabel D. Friesner
- Bakar Computational Health Sciences Institute, University of California, San Francisco
| | - Jean Feng
- Bakar Computational Health Sciences Institute, University of California, San Francisco
- Department of Epidemiology and Biostatistics, University of California, San Francisco
| | - Shalom Kalnicki
- Department of Radiation Oncology, Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York
| | - Madhur Garg
- Department of Radiation Oncology, Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York
| | - Nitin Ohri
- Department of Radiation Oncology, Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York
| | - Julian C. Hong
- Bakar Computational Health Sciences Institute, University of California, San Francisco
- Department of Radiation Oncology, University of California, San Francisco
| |
Collapse
|
5
|
Lin H, Ni L, Phuong C, Hong JC. Natural Language Processing for Radiation Oncology: Personalizing Treatment Pathways. Pharmgenomics Pers Med 2024; 17:65-76. [PMID: 38370334 PMCID: PMC10874185 DOI: 10.2147/pgpm.s396971] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/23/2023] [Accepted: 01/29/2024] [Indexed: 02/20/2024] Open
Abstract
Natural language processing (NLP), a technology that translates human language into machine-readable data, is revolutionizing numerous sectors, including cancer care. This review outlines the evolution of NLP and its potential for crafting personalized treatment pathways for cancer patients. Leveraging NLP's ability to transform unstructured medical data into structured learnable formats, researchers can tap into the potential of big data for clinical and research applications. Significant advancements in NLP have spurred interest in developing tools that automate information extraction from clinical text, potentially transforming medical research and clinical practices in radiation oncology. Applications discussed include symptom and toxicity monitoring, identification of social determinants of health, improving patient-physician communication, patient education, and predictive modeling. However, several challenges impede the full realization of NLP's benefits, such as privacy and security concerns, biases in NLP models, and the interpretability and generalizability of these models. Overcoming these challenges necessitates a collaborative effort between computer scientists and the radiation oncology community. This paper serves as a comprehensive guide to understanding the intricacies of NLP algorithms, their performance assessment, past research contributions, and the future of NLP in radiation oncology research and clinics.
Collapse
Affiliation(s)
- Hui Lin
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA
- UC Berkeley-UCSF Graduate Program in Bioengineering, University of California, Berkeley and San Francisco, San Francisco, CA, USA
| | - Lisa Ni
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA
| | - Christina Phuong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA
| | - Julian C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, USA
- Joint Program in Computational Precision Health, University of California, Berkeley and San Francisco, Berkeley, CA, USA
| |
Collapse
|
6
|
Sabbagh A, Tilki D, Feng J, Huland H, Graefen M, Wiegel T, Böhmer D, Hong JC, Valdes G, Cowan JE, Cooperberg M, Feng FY, Mohammad T, Shelan M, D'Amico AV, Carroll PR, Mohamad O. Multi-institutional Development and External Validation of a Machine Learning Model for the Prediction of Distant Metastasis in Patients Treated by Salvage Radiotherapy for Biochemical Failure After Radical Prostatectomy. Eur Urol Focus 2024; 10:66-74. [PMID: 37507248 DOI: 10.1016/j.euf.2023.07.004] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/16/2023] [Revised: 05/30/2023] [Accepted: 07/13/2023] [Indexed: 07/30/2023]
Abstract
BACKGROUND Up to 40% of patients with prostate cancer may develop biochemical recurrence after surgery, with salvage radiation therapy (SRT) being the only curative option. In 2016, Tendulkar et al. (Contemporary update of a multi-institutional predictive nomogram for salvage radiotherapy after radical prostatectomy. J Clin Oncol 2016;34:3648-54) published a nomogram to predict distant metastasis in a cohort of patients treated with SRT with pre-SRT prostate-specific antigen (PSA) of 0.5 ng/ml after radical prostatectomy. In modern practice, SRT is delivered at lower PSA values. OBJECTIVE To train and externally validate a machine learning model to predict the risk of distant metastasis at 5 yr in a contemporary cohort of patients receiving SRT. DESIGN, SETTING, AND PARTICIPANTS We trained a machine learning model on data from 2418 patients treated with SRT at one institution, with a median PSA value of 0.27 ng/ml. External validation was done in 475 patients treated at two different institutions. Patients with cM1, pN1, or pT4 disease were excluded, as were patients with PSA >2 ng/ml or PSA 0, and patients with radiation dose <60 or ≥80 Gy. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS Model performance was assessed using calibration and time-dependent area under the receiver operating curve (tAUC). RESULTS AND LIMITATIONS Our model had better calibration and showed improved discrimination (tAUC = 0.72) compared with the Tendulkar model (tAUC = 0.60, p < 0.001). The main limitations of this study are its retrospective design and lack of validation on patients who received hormone therapy. CONCLUSIONS The updated model can be used to provide more individualized risk assessments to patients treated with SRT at low PSA values, improving decision-making. PATIENT SUMMARY Up to 40% of patients with prostate cancer may develop biochemical recurrence after surgery, with salvage radiation therapy as the only potentially curative option. We trained and validated a machine learning model using clinical and surgical data to predict a patient's risk of distant metastasis at 5 yr after treatment. Our model outperformed the reference tool and can improve clinical decision-making by providing more personalized risk assessment.
Collapse
Affiliation(s)
- Ali Sabbagh
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA
| | - Derya Tilki
- Department of Urology, University Hospital Hamburg-Eppendorf, Hamburg, Germany; Martini-Klinik Prostate Cancer Center, University Hospital-Hamburg-Eppendorf, Hamburg, Germany
| | - Jean Feng
- Department of Epidemiology and Biostatistics, University of California San Francisco, San Francisco, CA, USA
| | - Hartwig Huland
- Martini-Klinik Prostate Cancer Center, University Hospital-Hamburg-Eppendorf, Hamburg, Germany
| | - Markus Graefen
- Martini-Klinik Prostate Cancer Center, University Hospital-Hamburg-Eppendorf, Hamburg, Germany
| | - Thomas Wiegel
- Department of Radio Oncology, University Hospital Ulm, Ulm, Germany
| | - Dirk Böhmer
- Department of Radiation Oncology, Charité University Hospital, Berlin, Germany
| | - Julian C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA
| | - Gilmer Valdes
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA
| | - Janet E Cowan
- Department of Urology, University of California San Francisco, San Francisco, CA, USA
| | - Matthew Cooperberg
- Department of Epidemiology and Biostatistics, University of California San Francisco, San Francisco, CA, USA; Department of Urology, University of California San Francisco, San Francisco, CA, USA
| | - Felix Y Feng
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA; Department of Urology, University of California San Francisco, San Francisco, CA, USA
| | | | - Mohamed Shelan
- Department of Radiation Oncology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
| | - Anthony V D'Amico
- Department of Radiation Oncology, Brigham and Women's Hospital and Dana Farber Cancer Institute, Boston, MA, USA
| | - Peter R Carroll
- Department of Urology, University of California San Francisco, San Francisco, CA, USA
| | - Osama Mohamad
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA, USA; Department of Urology, University of California San Francisco, San Francisco, CA, USA.
| |
Collapse
|
7
|
Mayo CS, Feng MU, Brock KK, Kudner R, Balter P, Buchsbaum JC, Caissie A, Covington E, Daugherty EC, Dekker AL, Fuller CD, Hallstrom AL, Hong DS, Hong JC, Kamran SC, Katsoulakis E, Kildea J, Krauze AV, Kruse JJ, McNutt T, Mierzwa M, Moreno A, Palta JR, Popple R, Purdie TG, Richardson S, Sharp GC, Satomi S, Tarbox LR, Venkatesan AM, Witztum A, Woods KE, Yao Y, Farahani K, Aneja S, Gabriel PE, Hadjiiski L, Ruan D, Siewerdsen JH, Bratt S, Casagni M, Chen S, Christodouleas JC, DiDonato A, Hayman J, Kapoor R, Kravitz S, Sebastian S, Von Siebenthal M, Bosch W, Hurkmans C, Yom SS, Xiao Y. Operational Ontology for Oncology (O3): A Professional Society-Based, Multistakeholder, Consensus-Driven Informatics Standard Supporting Clinical and Research Use of Real-World Data From Patients Treated for Cancer. Int J Radiat Oncol Biol Phys 2023; 117:533-550. [PMID: 37244628 PMCID: PMC10741247 DOI: 10.1016/j.ijrobp.2023.05.033] [Citation(s) in RCA: 3] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/07/2022] [Revised: 05/17/2023] [Accepted: 05/19/2023] [Indexed: 05/29/2023]
Abstract
PURPOSE The ongoing lack of data standardization severely undermines the potential for automated learning from the vast amount of information routinely archived in electronic health records (EHRs), radiation oncology information systems, treatment planning systems, and other cancer care and outcomes databases. We sought to create a standardized ontology for clinical data, social determinants of health, and other radiation oncology concepts and interrelationships. METHODS AND MATERIALS The American Association of Physicists in Medicine's Big Data Science Committee was initiated in July 2019 to explore common ground from the stakeholders' collective experience of issues that typically compromise the formation of large inter- and intra-institutional databases from EHRs. The Big Data Science Committee adopted an iterative, cyclical approach to engaging stakeholders beyond its membership to optimize the integration of diverse perspectives from the community. RESULTS We developed the Operational Ontology for Oncology (O3), which identified 42 key elements, 359 attributes, 144 value sets, and 155 relationships ranked in relative importance of clinical significance, likelihood of availability in EHRs, and the ability to modify routine clinical processes to permit aggregation. Recommendations are provided for best use and development of the O3 to 4 constituencies: device manufacturers, centers of clinical care, researchers, and professional societies. CONCLUSIONS O3 is designed to extend and interoperate with existing global infrastructure and data science standards. The implementation of these recommendations will lower the barriers for aggregation of information that could be used to create large, representative, findable, accessible, interoperable, and reusable data sets to support the scientific objectives of grant programs. The construction of comprehensive "real-world" data sets and application of advanced analytical techniques, including artificial intelligence, holds the potential to revolutionize patient management and improve outcomes by leveraging increased access to information derived from larger, more representative data sets.
Collapse
Affiliation(s)
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Dan Ruan
- University of California, Los Angeles
| | | | | | | | | | | | | | | | | | | | | | | | | | | | - Sue S Yom
- University of California, San Francisco
| | | |
Collapse
|
8
|
Waters MR, Aneja S, Hong JC. Unlocking the Power of ChatGPT, Artificial Intelligence, and Large Language Models: Practical Suggestions for Radiation Oncologists. Pract Radiat Oncol 2023; 13:e484-e490. [PMID: 37598727 DOI: 10.1016/j.prro.2023.06.011] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/21/2023] [Revised: 06/28/2023] [Accepted: 06/29/2023] [Indexed: 08/22/2023]
Abstract
Recent advances in artificial intelligence (AI), such as generative AI and large language models (LLMs), have generated significant excitement about the potential of AI to revolutionize our lives, work, and interaction with technology. This article explores the practical applications of LLMs, particularly ChatGPT, in the field of radiation oncology. We offer a guide on how radiation oncologists can interact with LLMs like ChatGPT in their routine clinical and administrative tasks, highlighting potential use cases of the present and future. We also highlight limitations and ethical considerations, including the current state of LLMs in decision making, protection of sensitive data, and the important role of human review of AI-generated content.
Collapse
Affiliation(s)
- Michael R Waters
- Department of Radiation Oncology, Washington University School of Medicine, St. Louis, Missouri
| | - Sanjay Aneja
- Department of Radiation Oncology, Yale School of Medicine, New Haven, Connecticut
| | - Julian C Hong
- Department of Radiation Oncology and Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, California.
| |
Collapse
|
9
|
Huang CC, Qazi JJ, Leng JX, Carpenter DJ, Natarajan BD, Arshad M, Schultz O, Moravan MJ, Mullikin TC, Reitman ZJ, Kirkpatrick JP, Floyd SR, Chmura SJ, Hong JC, Salama JK. Pretreatment Clinical Parameters Associated with Intracranial Progression Burden Following an Initial Stereotactic Radiosurgery Course in a Multi-Institutional Brain Metastases Cohort. Int J Radiat Oncol Biol Phys 2023; 117:e109-e110. [PMID: 37784644 DOI: 10.1016/j.ijrobp.2023.06.887] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/04/2023]
Abstract
PURPOSE/OBJECTIVE(S) While brain metastasis (BM) velocity is a valuable prognostic metric at time of intracranial progression (ICP), pre-SRS risk factors for post-SRS high-burden intracranial progression (ICP) remain poorly characterized. We hypothesized that pre-SRS clinical parameters are associated with subsequent high-burden (ICP), defined as either ≥5 (ICP5) or new/progressive ≥11 BMs (ICP11). MATERIALS/METHODS All patients completing an initial SRS course for BMs at two institutions from 1/2015-12/2020 were retrospectively identified. Patients with prior whole brain radiation therapy (WBRT) and/or BM resection were eligible. Demographic and clinical parameters were collected. ICP was defined as any radiographic concern for distant and/or in-field progression per multidisciplinary consensus. Overall survival (OS) and freedom from ICP were estimated via the Kaplan Meier method. Cox models assessed association between parameters and freedom from ICP5 and ICP11. RESULTS We identified 1383 patients completed SRS, with a median follow up of 8.7 months. Patients were 54.8% female, 45.6% with KPS ≥90, and a median of 63.4 years old. Primary tumor types included non-small cell lung (48.7%), breast (14.7%), and melanoma (8.5%). 46.9% had oligometastatic disease (≤5 metastatic foci: including BMs) at SRS, and 53.4% underwent SRS for >1 BM. 10.3% of patients had undergone prior WBRT and 26.1% surgical resection. 555 patients (40.1%) experienced ICP following SRS, of whom 72.6% had 1-4, 11.5% had 5-10, and 15.9% had ≥11 new/progressive BMs. Among patients with ICP, 6-month freedom from ICP was 35.5% (95% CI: 31.1-40.5%) for those with 1-4 BMs at time of ICP, 29.7% (95% CI: 20.4-43.3%) for 5-10 BMs, and 20.5% (95% CI: 13.5-30.1%) for ≥11 BMs (p = 0.016). Respective 12-month OS rates were 56.8% (95% CI: 52.1-61.9%), 46.0% (95% CI: 35.1-60.1%), and 38.7% (95% CI: 29.4-50.9%; p<0.001). Neurologic symptoms at time of ICP were observed in 21.1% of patients with 1-4 BMs, 28.1% with 5-10 BMs, and 50.0% with new/progressive ≥11 BMs (p<0.001). On multivariable analysis, superior freedom from high-burden ICP was associated with the following pre-SRS parameters: oligometastatic burden (ICP5: HR 0.68, 95% CI: 0.47-0.99; ICP11: 0.59; 95% CI: 0.36-0.97), no prior immunotherapy (ICP11: HR 0.57, 95% CI: 0.34-0.57), and a single BM at time of initial SRS (1 vs 2 BM, ICP 5: HR 0.51, 95% CI: 0.31-0.82; ICP11: HR 0.45, 95% CI: 0.24-0.84), while primary tumor type was not associated with ICP5 or ICP11. CONCLUSION Pre-SRS parameters including polymetastatic burden, prior receipt of immunotherapy, and >1 BM were associated with post-SRS high-burden ICP. High burden ICP developed earlier following SRS completion and was associated with higher rates of neurologic decline and inferior OS.
Collapse
Affiliation(s)
- C C Huang
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - J J Qazi
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - J X Leng
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - D J Carpenter
- Department of Radiation Oncology, Duke University Cancer Center, Durham, NC
| | - B D Natarajan
- Department of Medicine, University of North Carolina, Chapel Hill, NC
| | - M Arshad
- Department of Radiation and Cellular Oncology, University of Chicago, Chicago, IL
| | - O Schultz
- Department of Radiation Oncology, University of Chicago, Chicago, IL
| | - M J Moravan
- Department of Radiation Oncology, Washington University in St. Louis, St. Louis, MO
| | - T C Mullikin
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | | | - J P Kirkpatrick
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC; Department of Neurosurgery, Duke University Medical Center, Durham, NC
| | - S R Floyd
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - S J Chmura
- Department of Radiation and Cellular Oncology, University of Chicago Medical Center, Chicago, IL
| | - J C Hong
- University of California, San Francisco, Bakar Computational Health Sciences Institute, San Francisco, CA; Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| | - J K Salama
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC; Durham VA Health Care System, Durham, NC
| |
Collapse
|
10
|
Qazi JJ, Leng JX, Huang CC, Carpenter DJ, Natarajan BD, Arshad M, Schultz O, Moravan MJ, Mullikin TC, Reitman ZJ, Kirkpatrick JP, Floyd SR, Chmura SJ, Hong JC, Salama JK. Multi-Institutional Outcomes Following Stereotactic Radiosurgery for Gastrointestinal Brain Metastases. Int J Radiat Oncol Biol Phys 2023; 117:e146-e147. [PMID: 37784725 DOI: 10.1016/j.ijrobp.2023.06.962] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/04/2023]
Abstract
PURPOSE/OBJECTIVE(S) Outcomes following stereotactic radiosurgery (SRS) for gastrointestinal (GI) brain metastases (BM) are poorly defined. We analyzed our multi-institutional database of SRS patients, comparing outcomes between GI and non-GI BM patients after SRS. MATERIALS/METHODS We retrospectively identified all patients completing an initial SRS course across two institutions from 1/2015-12/2020. Demographic and clinical parameters were manually captured. Intracranial progression (ICP) was defined as any concern on post-SRS imaging for recurrence determined by multidisciplinary consensus. Overall survival (OS) and freedom from ICP (FFICP) were estimated via Kaplan Meier models. Cox proportional hazard models were used to assess associations between ICP and parameters. RESULTS Among 1383 total patients completing SRS for BM, 102 (7.4%) had GI BM. Among these, 46 (45.1%) were of colorectal (CRC) and 34 (33.3%) esophageal origin. Other GI sites (21.6%) included anal, pancreatic, gastric, GI of unknown origin, and hepatocellular carcinoma. Median follow up was 8.7 mos. GI BM patients were more likely to be younger (mean 59.1 vs 63.5 yrs, p = 0.001), male (56.9% vs 44.3%, p = 0.014 ), have more extracranial metastases (mean 1.9 vs 1.6, p = 0.003), have received systemic therapy (73.5% vs 63.9%, p = 0.049) or resection of BM (45.1% vs 25.0%, p < 0.001) prior to SRS, have larger planned target volumes of all BMs (mean 20.3 ccs vs 15.0 ccs, p = 0.013), and were less likely to receive whole brain radiation therapy (WBRT) prior to SRS (3.9% vs 10.8%, p = 0.028) or systemic therapy after SRS (54.9% vs 68.9%, p = 0.004). Among GI patients, median OS was 28.2 mos (95% CI 16.5-35.3), with no significant differences between GI and non-GI patients (p = 0.220) or among GI subgroups (CRC vs other GI: p = 0.731; esophageal vs other GI: p = 0.478). Median FFICP was significantly worse for GI patients (6.2 mos, 95% CI 4.0-9.6 mos) than for non-GI patients (12.4 mos, 95% CI 10.8-13.9 mos; p = 0.004). After accounting for age, sex, performance status, number of irradiated BMs, extracranial disease burden, extracranial disease control, interval from primary cancer diagnosis to BM diagnosis, resection status, receipt of prior WBRT, and receipt of post-SRS systemic therapy, GI origin was significantly associated with worse FFICP (HR 1.50, 95% CI 1.15-2.02, p = 0.007). FFICP was not significantly different between GI subgroups, with CRC and esophageal patients demonstrating median times to ICP of 5.0 mos (95% CI 3.4-9.6) and 7.2 mos (95% CI 2.7-14.1), respectively. Only 2 GI patients (2.0%) had ICP at site of prior SRS. CONCLUSION Across a modern, multi-institutional SRS cohort comparing GI to non-GI primary patients, BMs of GI origin demonstrated inferior FFICP to those of non-GI origin. OS did not vary significantly across GI and non-GI cases. Among GI subtypes, no significant differences were identified across FFICP or OS. These data may help inform treatment decisions and post-SRS surveillance.
Collapse
Affiliation(s)
- J J Qazi
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - J X Leng
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - C C Huang
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - D J Carpenter
- Department of Radiation Oncology, Duke University Cancer Center, Durham, NC
| | - B D Natarajan
- Department of Medicine, University of North Carolina, Chapel Hill, NC
| | - M Arshad
- Department of Radiation and Cellular Oncology, University of Chicago, Chicago, IL
| | - O Schultz
- Department of Radiation Oncology, University of Chicago, Chicago, IL
| | - M J Moravan
- Department of Radiation Oncology, Washington University in St. Louis, St. Louis, MO
| | - T C Mullikin
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - Z J Reitman
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - J P Kirkpatrick
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC; Department of Neurosurgery, Duke University Medical Center, Durham, NC
| | - S R Floyd
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - S J Chmura
- Department of Medicine, University of North Carolina, Chapel Hill, NC
| | - J C Hong
- University of California, San Francisco, Bakar Computational Health Sciences Institute, San Francisco, CA; Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| | - J K Salama
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC; Durham VA Health Care System, Durham, NC
| |
Collapse
|
11
|
Chang JH, Lin A, Singer L, Mohamad O, Chan J, Friesner I, Zack T, Ashraf-Ganjouei A, Boreta L, Gottschalk A, Braunstein SE, Park CC, Hong JC. Identifying Common Topics in Patient Portal Messages with Unsupervised Natural Language Processing. Int J Radiat Oncol Biol Phys 2023; 117:e460-e461. [PMID: 37785473 DOI: 10.1016/j.ijrobp.2023.06.1657] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/04/2023]
Abstract
PURPOSE/OBJECTIVE(S) Patient portal messaging is an increasingly important form of communication between patients and medical providers. This has become particularly relevant in oncology, where patients undergo intense longitudinal treatments that require frequent communication regarding symptoms, appointments, and diagnostic results. The rise in the volume of these messages has significantly increased the workload of medical providers and consequent physician burn-out. Natural language processing (NLP), particularly transformer-based models, may offer an automated approach to characterize the content of patient messages and improve message triage and routing. In this study, we employed a state-of-the-art language model (Bidirectional Encoder Representations from Transformers; BERT) to identify data-derived categories of representative topics from real-world data thereby providing basic information to build an appropriate routing system. MATERIALS/METHODS Patient-generated portal messages sent to a messaging pool for a single institution radiation oncology department from 2014 to 2023 were extracted. BERTopic, an NLP-based topic modeling technique based on BERT was optimized for topic modeling of patient messages. Uniform Manifold Approximation and Projection (UMAP) was used to reduce dimensionality and visualize topic relationships across messages. The BERTopic-identified topic categories were subsequently labeled manually by one of the physician investigators. Differences of number of messages over time were assessed using t-tests. RESULTS A total of 47,492 messages were retrieved. The average number of messages per month from a single patient ranged from 1 to 18 (median 1.67, interquartile range 1.0-2.4). The total volume of patient messages showed a ten-fold increase over the study period, with 101 messages per month sent in 2014 and 999 messages per month in 2022 (p<0.001). BERTopic initially identified 35 topics whose relationships and degrees of overlap were visualized by UMAP. Due to physician-identified similarities, these topics were reduced into 13 categories. The most frequent topic category was messages about laboratory tests or imaging studies: 24.3%, followed by messages expressing appreciation: 18.9%, scheduling discussions: 15.6%, symptom-related messages: 11%, and treatment-related messages: 10.7%. CONCLUSION Patient portal messages sent to a single institution radiation oncology department have increased dramatically in volume since implementation, corresponding to a broader national trend. NLP successfully identified common subject themes across patient messages, many of which are related to scheduling. This presents potential opportunities to apply NLP to automate message routing in the future.
Collapse
Affiliation(s)
- J H Chang
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA; Department of Radiation Oncology, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea, Republic of (South) Korea
| | - A Lin
- University of California San Francisco, Department of Hematology and Oncology, San Francisco, CA
| | - L Singer
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| | - O Mohamad
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - J Chan
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - I Friesner
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA; University of California, San Francisco, Bakar Computational Health Sciences Institute, San Francisco, CA
| | - T Zack
- University of California San Francisco, San Francisco, CA
| | - A Ashraf-Ganjouei
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA; University of California, San Francisco, Bakar Computational Health Sciences Institute, San Francisco, CA
| | - L Boreta
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - A Gottschalk
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - S E Braunstein
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - C C Park
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - J C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA; University of California, San Francisco, Bakar Computational Health Sciences Institute, San Francisco, CA
| |
Collapse
|
12
|
Qian AS, Friesner I, Chen JJ, Boreta L, Braunstein SE, Hong JC. Natural Language Processing Identification of Symptoms in Emergency Department Visits in Patients Receiving Radiation. Int J Radiat Oncol Biol Phys 2023; 117:S144. [PMID: 37784369 DOI: 10.1016/j.ijrobp.2023.06.558] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/04/2023]
Abstract
PURPOSE/OBJECTIVE(S) Patients undergoing radiotherapy (RT) for cancer often require emergency department (ED) attention with possible hospitalization. Designing strategies to mitigate hospital admissions requires understanding the causal symptoms to tailor interventional strategies. Natural language processing (NLP) has previously been shown to accurately identify documented symptoms and may help characterize factors contributing to admission. The objective of this study was to use NLP to identify documented symptoms during ED visits and their association with subsequent hospital admission of patients undergoing RT. MATERIALS/METHODS A de-identified, single tertiary-care institution cohort of patients who received radiation between 2013 and 2022 was identified from the electronic health record using International Classification of Disease (ICD) and Current Procedural Terminology (CPT) codes. We applied a previously validated clinical Text Analysis and Knowledge Extraction System (cTAKES)-based NLP pipeline to extract Common Terminology Criteria for Adverse Events (CTCAE) encoded symptoms from ED encounter clinical notes. Chi-squared testing was used to compare demographics, and logistic regression was used to identify symptoms associated with subsequent admission from ED visits. RESULTS We identified 14,007 patients who received radiation, of whom 270 (1.9%) experienced 302 ED visits during their radiation course. 141 (46.7%) of ED visits resulted in an admission. Among patients with an ED encounter, there were no differences in admission rates based on primary language (p = 0.771), sex (p = 0.824), marital status (p = 0.753), race (p = 0.222), or age (p = 0.123). In admitted patients, the top 5 symptoms were pain (94.3%), nausea (92.1%), vomiting (73.7%), constipation (70.9%), and weakness (63.8%). In patients who did not require admission, the most common symptoms were pain (84.5%), nausea (67.1%), vomiting (47.2%), headache (36.6%), and weakness (35.4%). The 10 symptoms most associated with admission from the ED based on logistic regression were malaise (OR 21.7, [95% CI 10.1 - 51.0]), lethargy (19.1, [8.5 - 51.3]), flushing (15.7, [8.6 - 30.4]), agitation (12.4, [3.5 - 78.7]), somnolence (10.3, [4.7 - 25.9]), fall (8.5, [3.7 - 23.2]), fatigue (7.8, [4.6 - 13.4]), constipation (6.9, [4.2 - 11.6]), nausea (5.8, [3.0 - 12.2]), and swelling (5.4, [3.3 - 9.1]). CONCLUSION Admitted and non-admitted ED patients with cancer seen in the ED during radiotherapy are documented to experience similar symptoms, but certain symptoms are associated with a higher risk of hospital admission. NLP may offer a mechanism for early, automated identification to facilitate supportive interventions for patients at high risk for admission during radiotherapy.
Collapse
Affiliation(s)
- A S Qian
- University of California, San Francisco, San Francisco, CA
| | - I Friesner
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - J J Chen
- University of California, San Francisco, San Francisco, CA
| | - L Boreta
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - S E Braunstein
- University of California San Francisco, Department of Radiation Oncology, San Francisco, CA
| | - J C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| |
Collapse
|
13
|
Chen WS, Tuchayi AM, Witztum A, Carroll P, Small E, Feng FY, Hope T, Hong JC. Utility of PSMA PET Guided Metastasis-Directed Radiotherapy. Int J Radiat Oncol Biol Phys 2023; 117:e372. [PMID: 37785268 DOI: 10.1016/j.ijrobp.2023.06.2473] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/04/2023]
Abstract
PURPOSE/OBJECTIVE(S) Metastasis-directed radiotherapy (MDT) is becoming a mainstay in the management of oligometastatic prostate cancer (PCa), and PSMA-PET is currently the most sensitive imaging modality for PCa metastasis screening. The efficacy of MDT guided by PSMA-PET imaging has not yet been well characterized. Moreover, the optimal role of androgen deprivation therapy (ADT) in the context of MDT is not known. We sought to assess the efficacy of PSMA PET-guided MDT in patients with metastatic PCa treated with and without ADT. MATERIALS/METHODS This is a single institutional retrospective study of patients diagnosed with metastatic prostate cancer by PSMA-PET imaging who were treated with MDT. Biochemical progression was defined as a PSA increase of ≥ 25% and ≥ 2 ng/mL if PSA was ≥ 2 ng/mL at time of initiating salvage treatment, or a PSA increase of ≥ 25% if PSA was < 2 ng/mL at time of salvage treatment. Survival analyses were performed using the Kaplan-Meier method with log-rank testing for significance. Cumulative incidence analyses were performed with Gray's testing for significance. Adverse event data were assessed per CTCAE v5 guidelines. RESULTS A total of 196 PSMA PET-avid lesions from 101 patients were irradiated with stereotactic body radiotherapy (SBRT). Median time from prior definitive locoregional therapy to MDT was 6.2 years. 79 patients had hormone-sensitive PCa (HSPC) and 22 patients had castration-resistant PCa (CRPC) at time of MDT. 47 of 79 (59%) patients with HSPC received ADT along with MDT, and 20 of the 47 patients received augmented ADT. 25 of the 32 (78%) HSPC patients receiving MDT without ADT had undergone at least one prior course of ADT, and none had castrate levels of testosterone at time of MDT with a median testosterone level of 341 ng/dl. With a median follow-up of 22.4 months, 5 of 196 lesions (2.6%) demonstrated radiographic progression. 2-year cumulative incidence of progression from HSPC to CRPC was 11% in patients who received ADT at time of MDT and 35% in those who did not (P = 0.027). Median biochemical progression free survival of patients with CRPC, HSPC treated without ADT, and HSPC treated with ADT following MDT was 5.4, 7.6, and 43.9 months respectively (P<0.0001). 2-year overall survival of the abovementioned groups was 72.2%, 100%, and 97.5% respectively (P<0.001). No Grade 3-5 adverse effects were observed. CONCLUSION MDT guided by PSMA-PET imaging is well-tolerated and delays biochemical progression in patients with CRPC and HSPC, with a greater effect observed in patients also receiving ADT.
Collapse
Affiliation(s)
- W S Chen
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| | - A Moradi Tuchayi
- Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA
| | - A Witztum
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| | - P Carroll
- Department of Urology, University of California San Francisco, San Francisco, CA
| | - E Small
- Division of Hematology and Oncology, Department of Medicine, University of California San Francisco, San Francisco, CA
| | - F Y Feng
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| | - T Hope
- Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA
| | - J C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA
| |
Collapse
|
14
|
Sabbagh A, Washington SL, Tilki D, Hong JC, Feng J, Valdes G, Chen MH, Wu J, Huland H, Graefen M, Wiegel T, Böhmer D, Cowan JE, Cooperberg M, Feng FY, Roach M, Trock BJ, Partin AW, D'Amico AV, Carroll PR, Mohamad O. Development and External Validation of a Machine Learning Model for Prediction of Lymph Node Metastasis in Patients with Prostate Cancer. Eur Urol Oncol 2023; 6:501-507. [PMID: 36868922 DOI: 10.1016/j.euo.2023.02.006] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/13/2022] [Revised: 01/10/2023] [Accepted: 02/03/2023] [Indexed: 03/05/2023]
Abstract
BACKGROUND Pelvic lymph node dissection (PLND) is the gold standard for diagnosis of lymph node involvement (LNI) in patients with prostate cancer. The Roach formula, Memorial Sloan Kettering Cancer Center (MSKCC) calculator, and Briganti 2012 nomogram are elegant and simple traditional tools used to estimate the risk of LNI and select patients for PLND. OBJECTIVE To determine whether machine learning (ML) can improve patient selection and outperform currently available tools for predicting LNI using similar readily available clinicopathologic variables. DESIGN, SETTING, AND PARTICIPANTS Retrospective data for patients treated with surgery and PLND between 1990 and 2020 in two academic institutions were used. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS We trained three models (two logistic regression models and one gradient-boosted trees-based model [XGBoost]) on data provided from one institution (n = 20267) with age, prostate-specific antigen (PSA) levels, clinical T stage, percentage positive cores, and Gleason scores as inputs. We externally validated these models using data from another institution (n = 1322) and compared their performance to that of the traditional models using the area under the receiver operating characteristic curve (AUC), calibration, and decision curve analysis (DCA). RESULTS AND LIMITATIONS LNI was present in 2563 patients (11.9%) overall, and in 119 patients (9%) in the validation data set. XGBoost had the best performance among all the models. On external validation, its AUC outperformed that of the Roach formula by 0.08 (95% confidence interval [CI] 0.042-0.12), the MSKCC nomogram by 0.05 (95% CI 0.016-0.070), and the Briganti nomogram by 0.03 (95% CI 0.0092-0.051; all p < 0.05). It also had better calibration and clinical utility in terms of net benefit on DCA across relevant clinical thresholds. The main limitation of the study is its retrospective design. CONCLUSIONS Taking all measures of performance together, ML using standard clinicopathologic variables outperforms traditional tools in predicting LNI. PATIENT SUMMARY Determining the risk of cancer spread to the lymph nodes in patients with prostate cancer allows surgeons to perform lymph node dissection only in patients who need it and avoid the side effects of the procedure in those who do not. In this study, we used machine learning to develop a new calculator to predict the risk of lymph node involvement that outperformed traditional tools currently used by oncologists.
Collapse
Affiliation(s)
- Ali Sabbagh
- Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA
| | - Samuel L Washington
- Department of Urology, University of California-San Francisco, San Francisco, CA, USA; Department of Epidemiology and Biostatistics, University of California-San Francisco, San Francisco, CA, USA
| | - Derya Tilki
- Martini-Klinik Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany; Department of Urology, University Hospital Hamburg-Eppendorf, Hamburg, Germany
| | - Julian C Hong
- Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA
| | - Jean Feng
- Department of Epidemiology and Biostatistics, University of California-San Francisco, San Francisco, CA, USA
| | - Gilmer Valdes
- Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA
| | - Ming-Hui Chen
- Department of Statistics, University of Connecticut, Storrs, CT, USA
| | - Jing Wu
- Department of Computer Science and Statistics, University of Rhode Island, Kingston, RI, USA
| | - Hartwig Huland
- Martini-Klinik Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany
| | - Markus Graefen
- Martini-Klinik Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany
| | - Thomas Wiegel
- Department of Radio Oncology, University Hospital Ulm, Ulm, Germany
| | - Dirk Böhmer
- Department of Radiation Oncology, Charité University Hospital, Berlin, Germany
| | - Janet E Cowan
- Department of Urology, University of California-San Francisco, San Francisco, CA, USA
| | - Matthew Cooperberg
- Department of Urology, University of California-San Francisco, San Francisco, CA, USA; Department of Epidemiology and Biostatistics, University of California-San Francisco, San Francisco, CA, USA
| | - Felix Y Feng
- Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA; Department of Urology, University of California-San Francisco, San Francisco, CA, USA
| | - Mack Roach
- Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA
| | - Bruce J Trock
- Division of Epidemiology, Brady Urological Institute, Johns Hopkins Medical Institution, Baltimore, MD, USA
| | - Alan W Partin
- Department of Urology, Brady Urological Institute, Johns Hopkins Medical Institution, Baltimore, MD, USA
| | - Anthony V D'Amico
- Department of Radiation Oncology, Brigham and Women's Hospital and Dana Farber Cancer Institute, Boston, MA, USA
| | - Peter R Carroll
- Department of Urology, University of California-San Francisco, San Francisco, CA, USA
| | - Osama Mohamad
- Department of Radiation Oncology, University of California-San Francisco, San Francisco, CA, USA; Department of Urology, University of California-San Francisco, San Francisco, CA, USA.
| |
Collapse
|
15
|
Leng JX, Huang CC, Qazi JJ, Carpenter DJ, Natarajan BD, Arshad M, Ferreira M, Schultz O, Moravan MJ, Mullikin TC, Reitman ZJ, Kirkpatrick JP, Floyd SR, Salama AKS, Fecci P, Chmura SJ, Hong JC, Salama JK. Clinical Outcomes Following an Initial Stereotactic Radiosurgery Course for Brain Metastases from Melanoma. Int J Radiat Oncol Biol Phys 2023; 117:e128. [PMID: 37784684 DOI: 10.1016/j.ijrobp.2023.06.924] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/04/2023]
Abstract
PURPOSE/OBJECTIVE(S) Brain metastases (BM) are common in melanoma patients. The effect of gene mutations is not well characterized since first-line metastatic therapy has shifted from chemotherapy (CHT) to molecularly targeted therapies (TT) and immunotherapy (IO). We report outcomes of melanoma BM patients stratified by molecular subtype and pre-stereotactic radiosurgery (SRS) systemic therapy. MATERIALS/METHODS We identified all patients completing an initial SRS course for BM at two institutions between 1/2015 and 12/2020. Patients who had prior WBRT and/or resection were eligible. Demographic and clinical parameters were collected, along with melanoma tumor molecular characteristics. Intracranial progression (ICP) was defined as any radiographic distant and/or in-field progression per multidisciplinary consensus. Overall survival (OS) and freedom from ICP (FFICP) were estimated via the Kaplan Meier method. RESULTS From a total of 1383 SRS BM patients, we identified 118 (8.5%) with melanoma. Median follow up was 8.7 months, median age 64 years (IQR 51-72), 81% had cutaneous origin, and 55% had a KPS of 90-100. Molecular subtypes included BRAF (45%), NRAS (9.3%), and c-KIT (3.4%). Overall, 61% received IO prior to SRS, while 25% and 9.3% received TT and CHT prior to SRS respectively. 60% of patients harboring a mutation received IO as first line therapy, 10% received TT, and 30% received both TT and IO prior to SRS. BRAFmut patients more likely to have received TT prior to SRS (43% vs 9.2%, p<0.001) compared to BRAFwt patients. Median OS was 9.7 months (95% CI 7.8-13) and was not significantly different from non-melanoma patients (p = 0.6). Median FFICP was worse for melanoma patients (5.9 mos, 95% CI 3.5-8.5) than non-melanoma patients (8.96 mos, 95% CI 8.2-9.7, p = 0.009). A total of 72 ICP events occurred, with 56 (77.8%) distant ICP cases, 3 (4.2%) in-field ICP, and 13 (18%) ICP events that were radionecrosis (RN) only. RN was associated with the presence of a targetable mutation (18% vs 2%, p = 0.006) and receipt of TT pre-SRS (36% vs 9.8%, p = 0.001). BRAFmut patients had significantly worse FFICP (3.8 mos, 95% CI 3.0-6.8) compared to BRAFwt patients (8.5 mos, 95% CI 5.8-30.2, p = 0.006), although median OS was not significantly different (9.6 mos, 95% CI 6.9-16 vs 10.7 mos, 95% CI 6.7-15.5, p = 0.8). NRASmut was associated with better FFICP (29 mos, 95% CI 2.94-NA, p = 0.02). CONCLUSION In this modern, multi-institutional cohort of SRS patients, melanoma BM patients had worse FFICP compared to non-melanoma BM patients, and BRAFmut patients had worse FFICP than BRAFwt patients. RN was associated with mutational status and receipt of TT pre-SRS. OS did not vary significantly across groups. This analysis may help inform systemic therapy decisions and future genomic studies for patients with BMs from melanoma.
Collapse
Affiliation(s)
- J X Leng
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - C C Huang
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - J J Qazi
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - D J Carpenter
- Department of Radiation Oncology, Duke University Cancer Center, Durham, NC
| | - B D Natarajan
- Department of Medicine, University of North Carolina, Chapel Hill, NC
| | - M Arshad
- Department of Radiation and Cellular Oncology, University of Chicago, Chicago, IL
| | - M Ferreira
- Duke University Medical Center, Durham, NC
| | - O Schultz
- University of Chicago Pritzker School of Medicine, Chicago, IL
| | - M J Moravan
- Department of Radiation Oncology, Washington University in St. Louis, St. Louis, MO
| | - T C Mullikin
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | | | - J P Kirkpatrick
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | - S R Floyd
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC
| | | | - P Fecci
- Duke University Medical Center, Department of Neurosurgery, Durham, NC
| | - S J Chmura
- Department of Medicine, University of North Carolina, Chapel Hill, NC
| | - J C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, CA; University of California, San Francisco, Bakar Computational Health Sciences Institute, San Francisco, CA
| | - J K Salama
- Department of Radiation Oncology, Duke University Medical Center, Durham, NC; Durham VA Health Care System, Durham, NC
| |
Collapse
|
16
|
Sinha S, Holmgren AJ, Hong JC, Rotenstein LS. Ctrl-C: a cross-sectional study of the electronic health record usage patterns of US oncology clinicians. JNCI Cancer Spectr 2023; 7:pkad066. [PMID: 37688578 PMCID: PMC10555739 DOI: 10.1093/jncics/pkad066] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/22/2023] [Revised: 08/03/2023] [Accepted: 08/31/2023] [Indexed: 09/11/2023] Open
Abstract
Despite some positive impact, the use of electronic health records (EHRs) has been associated with negative effects, such as emotional exhaustion. We sought to compare EHR use patterns for oncology vs nononcology medical specialists. In this cross-sectional study, we employed EHR usage data for 349 ambulatory health-care systems nationwide collected from the vendor Epic from January to August 2019. We compared note composition, message volume, and time in the EHR system for oncology vs nononcology clinicians. Compared with nononcology medical specialists, oncologists had a statistically significantly greater percentage of notes derived from Copy and Paste functions but less SmartPhrase use. They received more total EHR messages per day than other medical specialists, with a higher proportion of results and system-generated messages. Our results point to priorities for enhancing EHR systems to meet the needs of oncology clinicians, particularly as related to facilitating the complex documentation, results, and therapy involved in oncology care.
Collapse
Affiliation(s)
- Sumi Sinha
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA, USA
| | - A Jay Holmgren
- Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
- Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA, USA
| | - Julian C Hong
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA, USA
- Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA, USA
| | - Lisa S Rotenstein
- Department of Medicine, Brigham and Women’s Hospital, Boston, MA, USA
| |
Collapse
|
17
|
Carpenter DJ, Leng J, Arshad M, Giles W, Kirkpatrick JP, Floyd SR, Chmura SJ, Salama JK, Hong JC. Intracranial and Extracranial Progression and Their Correlation With Overall Survival After Stereotactic Radiosurgery in a Multi-institutional Cohort With Brain Metastases. JAMA Netw Open 2023; 6:e2310117. [PMID: 37099292 PMCID: PMC10134007 DOI: 10.1001/jamanetworkopen.2023.10117] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/09/2022] [Accepted: 03/13/2023] [Indexed: 04/27/2023] Open
Abstract
Importance Clinical trials for metastatic malignant neoplasms are increasingly being extended to patients with brain metastases. Despite the preeminence of progression-free survival (PFS) as a primary oncologic end point, the correlation of intracranial progression (ICP) and extracranial progression (ECP) events with overall survival (OS) is poorly understood for patients with brain metastases following stereotactic radiosurgery (SRS). Objective To determine the correlation of ICP and ECP with OS among patients with brain metastases completing an initial SRS course. Design, Setting, and Participants This multi-institutional retrospective cohort study was conducted from January 1, 2015, to December 31, 2020. We included patients who completed an initial course of SRS for brain metastases during the study period, including receipt of single and/or multifraction SRS, prior whole-brain radiotherapy, and brain metastasis resection. Data analysis was performed on November 15, 2022. Exposures Non-OS end points included intracranial PFS, extracranial PFS, PFS, time to ICP, time to ECP, and any time to progression. Progression events were radiologically defined, incorporating multidisciplinary clinical consensus. Main Outcomes and Measures The primary outcome was correlation of surrogate end points to OS. Clinical end points were estimated from time of SRS completion via the Kaplan-Meier method, while end-point correlation to OS was measured using normal scores rank correlation with the iterative multiple imputation approach. Results This study included 1383 patients, with a mean age of 63.1 years (range, 20.9-92.8 years) and a median follow-up of 8.72 months (IQR, 3.25-19.68 months). The majority of participants were White (1032 [75%]), and more than half (758 [55%]) were women. Common primary tumor sites included the lung (757 [55%]), breast (203 [15%]), and skin (melanoma; 100 [7%]). Intracranial progression was observed in 698 patients (50%), preceding 492 of 1000 observed deaths (49%). Extracranial progression was observed in 800 patients (58%), preceding 627 of 1000 observed deaths (63%). Irrespective of deaths, 482 patients (35%) experienced both ICP and ECP, 534 (39%) experienced ICP (216 [16%]) or ECP (318 [23%]), and 367 (27%) experienced neither. The median OS was 9.93 months (95% CI, 9.08-11.05 months). Intracranial PFS had the highest correlation with OS (ρ = 0.84 [95% CI, 0.82-0.85]; median, 4.39 months [95% CI, 4.02-4.92 months]). Time to ICP had the lowest correlation with OS (ρ = 0.42 [95% CI, 0.34-0.50]) and the longest median time to event (median, 8.76 months [95% CI, 7.70-9.48 months]). Across specific primary tumor types, correlations of intracranial PFS and extracranial PFS with OS were consistently high despite corresponding differences in median outcome durations. Conclusions and Relevance The results of this cohort study of patients with brain metastases completing SRS suggest that intracranial PFS, extracranial PFS, and PFS had the highest correlations with OS and time to ICP had the lowest correlation with OS. These data may inform future patient inclusion and end-point selection for clinical trials.
Collapse
Affiliation(s)
- David J. Carpenter
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - Jim Leng
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - Muzamil Arshad
- Department of Radiation Oncology, University of Chicago Medical Center, Chicago, Illinois
| | - Will Giles
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - John P. Kirkpatrick
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
- Department of Neurosurgery, Duke University Medical Center, Durham, North Carolina
| | - Scott R. Floyd
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
| | - Steven J. Chmura
- Department of Radiation Oncology, University of Chicago Medical Center, Chicago, Illinois
| | - Joseph K. Salama
- Department of Radiation Oncology, Duke University Medical Center, Durham, North Carolina
- Radiation Oncology Clinical Service, Durham Veterans Affairs Health Care System, Durham, North Carolina
| | - Julian C. Hong
- Department of Radiation Oncology, University of California, San Francisco
- Bakar Computational Health Sciences Institute, University of California, San Francisco
- Joint Program in Computational Precision Health, University of California, San Francisco, and University of California, Berkeley
| |
Collapse
|
18
|
Kwon DH, Shakhnazaryan N, Shui D, Hong JC, Mohamad O, de Kouchkovsky I, Borno HT, Bose R, Chou J, Desai A, Fong L, Friedlander TW, Koshkin VS, Aggarwal RR, Feng FY, Hope TA, Small EJ. Serial stereotactic body radiation therapy for oligometastatic prostate cancer detected by novel PET-based radiotracers. Urol Oncol 2023; 41:145.e7-145.e15. [PMID: 36435709 DOI: 10.1016/j.urolonc.2022.10.025] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/15/2022] [Revised: 10/21/2022] [Accepted: 10/28/2022] [Indexed: 11/25/2022]
Abstract
BACKGROUND Radiopharmaceuticals, including Ga-68-prostate specific membrane antigen (PSMA)-11 and F-18-Fluciclovine, are increasingly used to inform therapies for prostate cancer (CaP). Stereotactic body radiation therapy (SBRT) to PET-detected oligometastatic CaP has been shown to improve progression free survival (PFS) and delay androgen deprivation therapy (ADT) compared to observation. For men who subsequently develop oligorecurrent CaP, outcomes following second SBRT are unknown. METHODS A retrospective cohort study was conducted. Eligibility criteria included patients with oligometastatic (1-5 lesions) CaP detected on PSMA or Fluciclovine PET who underwent 2 consecutive SBRT courses to tracer-avid sites. Data on stage, tracer type, concurrent systemic therapy, and prostate-specific antigen (PSA) responses for first SBRT (SBRT1) and second SBRT (SBRT2) were collected. Outcomes included PSA decline ≥50% (PSA50), PFS after SBRT2, and ADT initiation or intensification-free survival after SBRT2. Factors potentially associated with PSA50 after SBRT2 was evaluated with multivariable logistic regression. Factors potentially associated with PFS and ADT initiation/intensification-free survival after SBRT2 were evaluated with separate multivariable Cox proportional-hazards models. RESULTS Twenty-five patients were identified. At SBRT2, oligorecurrence was detected on PSMA and Fluciclovine PET in 17 (68%) and 8 (32%) patients, respectively. Fifteen (60%) patients had castration-sensitive disease and 10 (40%) had castration-resistant disease. After SBRT2, 16 (64%) achieved a PSA50 response, median PFS was 11.0mo, and median ADT initiation/intensification-free survival was 23.2mo. On multivariable analysis, maximum percent change in PSA after SBRT1 (OR 0.94, 95%CI 0.88-0.99, P = 0.046) and concurrent change in systemic therapy (OR 21.61, 95%CI 1.12-417.9, P = 0.042) were associated with PSA50 responses after SBRT2. PSA50 response after SBRT1 was associated with improved PFS (HR 0.36, 95%CI 0.00-0.42, P = 0.008) and ADT initiation/intensification-free survival (HR 0.07, 95%CI 0.01-0.68, P = 0.021) after SBRT2. From SBRT1 to last follow-up (median 48 months), 7 (28%) patients remained ADT-free. CONCLUSIONS Serial SBRT for oligometastatic CaP detected on PSMA or Fluciclovine PET is feasible and can achieve PSA declines, with or without systemic therapy. Degree of biochemical response to first SBRT warrants further study as a potential predictor of PSA response, PFS, and ADT initiation/intensification-free survival following a subsequent SBRT course. This preliminary evidence provides rationale for larger, prospective studies of this strategy.
Collapse
Affiliation(s)
- Daniel H Kwon
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA.
| | - Nonna Shakhnazaryan
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - David Shui
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Julian C Hong
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA; Department of Radiation Oncology, University of California, San Francisco, CA; Bakar Computational Health Sciences Institute, University of California, San Francisco, CA
| | - Osama Mohamad
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA; Department of Radiation Oncology, University of California, San Francisco, CA; Department of Urology, University of California, San Francisco, CA
| | - Ivan de Kouchkovsky
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Hala T Borno
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Rohit Bose
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Jonathan Chou
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Arpita Desai
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Lawrence Fong
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Terence W Friedlander
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Vadim S Koshkin
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Rahul R Aggarwal
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| | - Felix Y Feng
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA; Department of Radiation Oncology, University of California, San Francisco, CA
| | - Thomas A Hope
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA; Department of Radiology and Biomedical Imaging, University of California, San Francisco, CA
| | - Eric J Small
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA; Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA
| |
Collapse
|
19
|
Hong JC, Patel P, Eclov NCW, Stephens SJ, Mowery YM, Tenenbaum JD, Palta M. Healthcare provider evaluation of machine learning-directed care: reactions to deployment on a randomised controlled study. BMJ Health Care Inform 2023; 30:bmjhci-2022-100674. [PMID: 36764680 PMCID: PMC9923272 DOI: 10.1136/bmjhci-2022-100674] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/01/2022] [Accepted: 01/28/2023] [Indexed: 02/12/2023] Open
Abstract
OBJECTIVES Clinical artificial intelligence and machine learning (ML) face barriers related to implementation and trust. There have been few prospective opportunities to evaluate these concerns. System for High Intensity EvaLuation During Radiotherapy (NCT03775265) was a randomised controlled study demonstrating that ML accurately directed clinical evaluations to reduce acute care during cancer radiotherapy. We characterised subsequent perceptions and barriers to implementation. METHODS An anonymous 7-question Likert-type scale survey with optional free text was administered to multidisciplinary staff focused on workflow, agreement with ML and patient experience. RESULTS 59/71 (83%) responded. 81% disagreed/strongly disagreed their workflow was disrupted. 67% agreed/strongly agreed patients undergoing intervention were high risk. 75% agreed/strongly agreed they would implement the ML approach routinely if the study was positive. Free-text feedback focused on patient education and ML predictions. CONCLUSIONS Randomised data and firsthand experience support positive reception of clinical ML. Providers highlighted future priorities, including patient counselling and workflow optimisation.
Collapse
Affiliation(s)
- Julian C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, California, USA .,Bakar Computational Health Sciences Institute, University of California San Francisco, San Francisco, California, USA.,Joint Program in Computational Precision Health, UCSF-UC Berkeley, San Francisco, California, USA
| | - Pranalee Patel
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Neville C W Eclov
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Sarah J Stephens
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Yvonne M Mowery
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA,Department of Head and Neck Surgery & Communication Sciences, Duke University, Durham, North Carolina, USA
| | - Jessica D Tenenbaum
- Department of Biostatistics and Bioinformatics, Duke University, Durham, North Carolina, USA
| | - Manisha Palta
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| |
Collapse
|
20
|
Zack T, Dhaliwal G, Geha R, Margaretten M, Murray S, Hong JC. A Clinical Reasoning-Encoded Case Library Developed through Natural Language Processing. J Gen Intern Med 2023; 38:5-11. [PMID: 36071325 PMCID: PMC9849536 DOI: 10.1007/s11606-022-07758-0] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 01/20/2022] [Accepted: 07/29/2022] [Indexed: 01/22/2023]
Abstract
IMPORTANCE Case reports that externalize expert diagnostic reasoning are utilized for clinical reasoning instruction but are difficult to search based on symptoms, final diagnosis, or differential diagnosis construction. Computational approaches that uncover how experienced diagnosticians analyze the medical information in a case as they formulate a differential diagnosis can guide educational uses of case reports. OBJECTIVE To develop a "reasoning-encoded" case database for advanced clinical reasoning instruction by applying natural language processing (NLP), a sub-field of artificial intelligence, to a large case report library. DESIGN We collected 2525 cases from the New England Journal of Medicine (NEJM) Clinical Pathological Conference (CPC) from 1965 to 2020 and used NLP to analyze the medical terminology in each case to derive unbiased (not prespecified) categories of analysis used by the clinical discussant. We then analyzed and mapped the degree of category overlap between cases. RESULTS Our NLP algorithms identified clinically relevant categories that reflected the relationships between medical terms (which included symptoms, signs, test results, pathophysiology, and diagnoses). NLP extracted 43,291 symptoms across 2525 cases and physician-annotated 6532 diagnoses (both primary and related diagnoses). Our unsupervised learning computational approach identified 12 categories of medical terms that characterized the differential diagnosis discussions within individual cases. We used these categories to derive a measure of differential diagnosis similarity between cases and developed a website ( universeofcpc.com ) to allow visualization and exploration of 55 years of NEJM CPC case series. CONCLUSIONS Applying NLP to curated instances of diagnostic reasoning can provide insight into how expert clinicians correlate and coordinate disease categories and processes when creating a differential diagnosis. Our reasoning-encoded CPC case database can be used by clinician-educators to design a case-based curriculum and by physicians to direct their lifelong learning efforts.
Collapse
Affiliation(s)
- Travis Zack
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, 94158, USA.
| | - Gurpreet Dhaliwal
- San Francisco VA Medical Center, San Francisco, CA, USA
- Department of Medicine, University of California, San Francisco, CA, USA
| | - Rabih Geha
- San Francisco VA Medical Center, San Francisco, CA, USA
- Department of Medicine, University of California, San Francisco, CA, USA
| | - Mary Margaretten
- Division of Rheumatology, Department of Medicine, University of California, San Francisco, CA, USA
| | - Sara Murray
- Division of Hospital Medicine, Department of Medicine, University of California, San Francisco, CA, USA
| | - Julian C Hong
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, 94158, USA
- Department of Radiation Oncology, University of California, San Francisco, CA, USA
| |
Collapse
|
21
|
Ragavan MV, Legaspi N, LaLanne A, Hong JC, Small EJ, Borno HT. Analysis of Serious Adverse Event Reporting for Patients Enrolled in Cancer Clinical Trials During the COVID-19 Pandemic. JAMA Oncol 2022; 8:1849-1851. [PMID: 36301577 PMCID: PMC9614670 DOI: 10.1001/jamaoncol.2022.4919] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/24/2022]
Abstract
This cross-sectional study assesses whether a change occurred in reporting of serious adverse events for patients in oncology clinical trials in the US during the COVID-19 pandemic.
Collapse
Affiliation(s)
- Meera V. Ragavan
- Division of Hematology and Oncology, Department of Medicine, University of California, San Francisco, San Francisco
| | - Nichole Legaspi
- Division of Hematology and Oncology, Department of Medicine, University of California, San Francisco, San Francisco
| | - Alyssa LaLanne
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco
| | - Julian C. Hong
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco,Department of Radiation Oncology, University of California, San Francisco, San Francisco,Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco
| | - Eric J. Small
- Division of Hematology and Oncology, Department of Medicine, University of California, San Francisco, San Francisco,Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco
| | - Hala T. Borno
- Division of Hematology and Oncology, Department of Medicine, University of California, San Francisco, San Francisco,Trial Library, San Francisco, California
| |
Collapse
|
22
|
Abou Ziki MD, Taoutel R, Hong JC, Podell DN. Severe extra-glandular involvement and pleural effusions complicating primary Sjogren's syndrome: a case report. J Med Case Rep 2022; 16:374. [PMID: 36253840 PMCID: PMC9578189 DOI: 10.1186/s13256-022-03557-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/03/2021] [Accepted: 07/10/2022] [Indexed: 11/25/2022] Open
Abstract
Background Sjogren’s syndrome, an autoimmune disease of the exocrine glands, results in keratoconjunctivitis sicca, xerostomia, and dental caries. It is often overlooked, considered by clinicians to be a benign disease. However, it can cause life-threatening extra-glandular complications that affect multiple organ systems. Case presentation Here we present a 78-year-old Caucasian woman with a history of primary Sjogren’s syndrome (pSS) whose symptoms of keratoconjunctivitis sicca were managed managed conservatively. She was evaluated for sub-acute shortness of breath. Imaging showed severe bronchiectasis with features of lymphocytic interstitial pneumonia. She also had exudative bilateral pleural effusions and skin ulcers, likely vasculitic in origin. The workup was significant for anti-Ro antibody, pancytopenia, hypocomplementia, cryoglobulinemia and monoclonal gammopathy, all of which reflect disease severity. Although there was no evidence of malignancy, she developed B-cell non-Hodgkin lymphoma during follow-up. Conclusions Primary Sjogren’s syndrome can result in severe multi-organ disease. Pleural effusions are a rare complication of pSS, with only ten cases reported in the literature over the last 30 years, and tend to respond well to steroids. Prognostic biomarkers for disease severity include hypocomplementia, cryoglobulinemia, monoclonal gammopathy, and hypergammaglobulinemia. In this report we review the literature and the management of the disease.
Collapse
Affiliation(s)
- Maen D Abou Ziki
- Department of Medicine, Yale University School of Medicine, 300 George St, New Haven, CT, 06520, USA.
| | - Roy Taoutel
- Department of Medicine, Lankenau Medical Center, Wynnewood, PA, 19096, USA
| | - Julian C Hong
- Department of Medicine, Yale University School of Medicine, 300 George St, New Haven, CT, 06520, USA
| | - David N Podell
- Department of Medicine, Yale University School of Medicine, 300 George St, New Haven, CT, 06520, USA
| |
Collapse
|
23
|
Phuong C, Lin AM, Friesner I, Ni L, Aggarwal RR, Borno H, Koshkin VS, Desai A, Friedlander TW, Fong L, Bose R, Chou J, Rodvelt TJ, Mohamad O, Wong AC, Feng FY, Small EJ, Hong JC. Reliability of real-world data for diagnosis of metastatic prostate cancer. J Clin Oncol 2022. [DOI: 10.1200/jco.2022.40.28_suppl.397] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
397 Background: Real-world data (RWD) is playing an increasingly important role in cancer research. Surrogate endpoints such as metastasis-free survival play an important role in prostate cancer research, leading to interest in its computational extraction, typically with use of International Classification of Disease (ICD) metastatic codes. While prior studies have suggested that ICD codes are valid for identification of patients (pts) with metastatic prostate cancer (MPC), delays in coding may impact their accuracy. The objective of this informatics-based study is to quantify the time delay between diagnosis of MPC and entry of ICD MPC-related code and its interaction with changing institutional healthcare processes. Methods: A single institutional EHR data warehouse was queried to identify a random sample of 100 pts with MPC diagnosis based on ICD codes (ICD10 C79 or ICD9 198.5) from 2013-2021 who were also seen in the genitourinary medical oncology program (GUMOP). Of note, in 6/2018, the GUMOP adopted EHR-specific MPC visit diagnosis identifiers (Dx ID) to improve MPC coding during clinic independent of ICD codes typically used by RWD researchers. Thus, the study cohort was designed to include pts whose first follow up after being diagnosed with MPC was before (n = 50) or after (n = 50) Dx ID implementation. Date of first MPC ICD code entry at any point in the EHR was compared against true date of MPC, based on physician review of definitive imaging or pathology. Data analysis was performed with Wilcox Signed rank test, bivariate analyses, and multivariable linear regression. Covariates included modality of diagnosis confirmation and timing with Dx ID implementation. Results: One hundred pts with MPC ICD coded in the EHR were included, with 29 pts diagnosed by PSMA PET and 71 by conventional imaging. Median time from true MPC diagnosis to first subsequent clinic follow up was < 1 month (IQR 0-2), while median time from true MPC diagnosis to entry of ICD MPC-related code was longer at 4mo (IQR 0-15). 5 pts had C79 applied for N1 disease and 10 pts for work-up of biochemical recurrence. On multivariable analysis of potential factors affecting time interval to MPC ICD entry, Dx ID implementation (b = -6.5 mo [95% CI -1.8 to -11.2], p = 0.007) and non-PSMA based diagnosis (b = -5.7 mo [95% CI -0.5 to -10.8], p = 0.03) were independently associated with shorter time to ICD coding. In subset analysis of the cohort after Dx ID implementation, use of both ICD and Dx ID to identify pts with MPC reduced the median time from true MPC diagnosis to EHR coding (1mo, IQR 0-6.3) compared to ICD alone (2mo, IQR 0-8) (p = 0.003). Conclusions: Timing of MPC ICD entry is highly variable and may carry biases derived from healthcare processes, including data entry and diagnostic testing. This may be improved with EHR workflow interventions. It is essential to have domain knowledge of clinical coding practices to improve information retrieval and recognize potential limitations and biases.
Collapse
Affiliation(s)
| | - Amy M. Lin
- University of California, San Francisco, San Francisco, CA
| | - Izzy Friesner
- University of California, San Francisco, San Francisco, CA
| | - Lisa Ni
- University of California, San Francisco, San Francisco, CA
| | | | - Hala Borno
- University of California, San Francisco, San Francisco, CA
| | | | - Arpita Desai
- University of California, San Francisco, San Francisco, CA
| | | | - Lawrence Fong
- University of California, San Francisco, San Francisco, CA
| | - Rohit Bose
- University of California, San Francisco, San Francisco, CA
| | - Jonathan Chou
- University of California, San Francisco, San Francisco, CA
| | | | - Osama Mohamad
- University of California, San Francisco, San Francisco, CA
| | | | - Felix Y Feng
- University of California, San Francisco, San Francisco, CA
| | - Eric Jay Small
- University of California, San Francisco, San Francisco, CA
| | - Julian C. Hong
- University of California, San Francisco, San Francisco, CA
| |
Collapse
|
24
|
Hong JC, Eclov NCW, Stephens SJ, Mowery YM, Palta M. Implementation of machine learning in the clinic: challenges and lessons in prospective deployment from the System for High Intensity EvaLuation During Radiation Therapy (SHIELD-RT) randomized controlled study. BMC Bioinformatics 2022; 23:408. [PMID: 36180836 PMCID: PMC9526253 DOI: 10.1186/s12859-022-04940-3] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/14/2022] [Accepted: 09/16/2022] [Indexed: 12/02/2022] Open
Abstract
Background Artificial intelligence (AI) and machine learning (ML) have resulted in significant enthusiasm for their promise in healthcare. Despite this, prospective randomized controlled trials and successful clinical implementation remain limited. One clinical application of ML is mitigation of the increased risk for acute care during outpatient cancer therapy. We previously reported the results of the System for High Intensity EvaLuation During Radiation Therapy (SHIELD-RT) study (NCT04277650), which was a prospective, randomized quality improvement study demonstrating that ML based on electronic health record (EHR) data can direct supplemental clinical evaluations and reduce the rate of acute care during cancer radiotherapy with and without chemotherapy. The objective of this study is to report the workflow and operational challenges encountered during ML implementation on the SHIELD-RT study. Results Data extraction and manual review steps in the workflow represented significant time commitments for implementation of clinical ML on a prospective, randomized study. Barriers include limited data availability through the standard clinical workflow and commercial products, the need to aggregate data from multiple sources, and logistical challenges from altering the standard clinical workflow to deliver adaptive care. Conclusions The SHIELD-RT study was an early randomized controlled study which enabled assessment of barriers to clinical ML implementation, specifically those which leverage the EHR. These challenges build on a growing body of literature and may provide lessons for future healthcare ML adoption. Trial registration: NCT04277650. Registered 20 February 2020. Retrospectively registered quality improvement study.
Collapse
|
25
|
Huilgol YS, Adler‐Milstein J, Ivey SL, Hong JC. Opportunities to use electronic health record audit logs to improve cancer care. Cancer Med 2022; 11:3296-3303. [PMID: 35348298 PMCID: PMC9468426 DOI: 10.1002/cam4.4690] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/22/2021] [Revised: 01/21/2022] [Accepted: 03/10/2022] [Indexed: 12/11/2022] Open
Abstract
The rapid adoption of electronic health records (EHRs) has created extensive repositories of digitized data that can be used to inform improvements in care delivery, processes, and patient outcomes. While the clinical data captured in EHRs are widely used for such efforts, EHRs also capture audit log data that reflect how users interact with the EHR to deliver care. Automatically collected audit log data provide a unique opportunity for new insights into EHR user behavior and decision‐making processes. Here, we provide an overview of audit log data and examples that could be used to improve oncology care and outcomes in four domains: diagnostic reasoning and consumption, care team collaboration and communication, patient outcomes and experience, and provider burnout/fatigue. This data source could identify gaps in performance and care, physician uptake of EHR features that enhance decision‐making, and integration of data trends for oncology. Ensuring researchers and oncologists are familiar with the data's potential and developing the data engineering capacity to utilize this rich data source, will expand the breadth of research to improve cancer care.
Collapse
Affiliation(s)
- Yash S. Huilgol
- UC Berkeley‐UCSF Joint Medical Program University of California Berkeley California USA
- School of Medicine University of California San Francisco California USA
| | - Julia Adler‐Milstein
- School of Medicine University of California San Francisco California USA
- Center for Clinical Informatics and Improvement Research (CLIIR) University of California San Francisco California USA
| | - Susan L. Ivey
- UC Berkeley‐UCSF Joint Medical Program University of California Berkeley California USA
- School of Public Health University of California Berkeley California USA
| | - Julian C. Hong
- Bakar Computational Health Sciences Institute University of California San Francisco California USA
- Department of Radiation Oncology University of California San Francisco California USA
| |
Collapse
|
26
|
Kwon DH, Cadena J, Nguyen S, Chan KHR, Soper B, Gryshuk AL, Hong JC, Ray P, Huang FW. COVID-19 outcomes in patients with cancer: Findings from the University of California health system database. Cancer Med 2022; 11:2204-2215. [PMID: 35261195 PMCID: PMC9110901 DOI: 10.1002/cam4.4604] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/24/2021] [Revised: 01/13/2022] [Accepted: 01/19/2022] [Indexed: 01/10/2023] Open
Abstract
Background The interaction between cancer diagnoses and COVID‐19 infection and outcomes is unclear. We leveraged a state‐wide, multi‐institutional database to assess cancer‐related risk factors for poor COVID‐19 outcomes. Methods We conducted a retrospective cohort study using the University of California Health COVID Research Dataset, which includes electronic health data of patients tested for severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) at 17 California medical centers. We identified adults tested for SARS‐CoV‐2 from 2/1/2020–12/31/2020 and selected a cohort of patients with cancer. We obtained demographic, clinical, cancer type, and antineoplastic therapy data. The primary outcome was hospitalization within 30d after the first positive SARS‐CoV‐2 test. Secondary outcomes were SARS‐CoV‐2 positivity and severe COVID‐19 (intensive care, mechanical ventilation, or death within 30d after the first positive test). We used multivariable logistic regression to identify cancer‐related factors associated with outcomes. Results We identified 409,462 patients undergoing SARS‐CoV‐2 testing. Of 49,918 patients with cancer, 1781 (3.6%) tested positive. Patients with cancer were less likely to test positive (RR 0.70, 95% CI: 0.67–0.74, p < 0.001). Among the 1781 SARS‐CoV‐2‐positive patients with cancer, BCR/ABL‐negative myeloproliferative neoplasms (RR 2.15, 95% CI: 1.25–3.41, p = 0.007), venetoclax (RR 2.96, 95% CI: 1.14–5.66, p = 0.028), and methotrexate (RR 2.72, 95% CI: 1.10–5.19, p = 0.032) were associated with greater hospitalization risk. Cancer and therapy types were not associated with severe COVID‐19. Conclusions In this large, diverse cohort, cancer was associated with a decreased risk of SARS‐CoV‐2 positivity. Patients with BCR/ABL‐negative myeloproliferative neoplasm or receiving methotrexate or venetoclax may be at increased risk of hospitalization following SARS‐CoV‐2 infection. Mechanistic and comparative studies are needed to validate findings.
Collapse
Affiliation(s)
- Daniel H Kwon
- Division of Hematology/Oncology, Department of Medicine, University of California San Francisco, California, USA
| | - Jose Cadena
- Computational Engineering, Engineering Directorate, Lawrence Livermore National Laboratory, California, USA
| | - Sam Nguyen
- Computational Engineering, Engineering Directorate, Lawrence Livermore National Laboratory, California, USA
| | - Kwan Ho Ryan Chan
- Computational Engineering, Engineering Directorate, Lawrence Livermore National Laboratory, California, USA.,Department of Electrical and Computer Engineering, Johns Hopkins University, Baltimore, Maryland, USA
| | - Braden Soper
- Center for Applied Scientific Computing, Computing Directorate, Lawrence Livermore National Laboratory, California, USA
| | - Amy L Gryshuk
- Physical and Life Sciences Directorate, Lawrence Livermore National Laboratory, California, USA
| | - Julian C Hong
- Helen Diller Family Comprehensive Cancer Center, University of California San Francisco, San Francisco, California, USA.,Bakar Computational Health Sciences Institute, University of California San Francisco, California, USA.,Department of Radiation Oncology, University of California San Francisco, California, USA
| | - Priyadip Ray
- Computational Engineering, Engineering Directorate, Lawrence Livermore National Laboratory, California, USA
| | - Franklin W Huang
- Division of Hematology/Oncology, Department of Medicine, University of California San Francisco, California, USA.,Helen Diller Family Comprehensive Cancer Center, University of California San Francisco, San Francisco, California, USA.,Bakar Computational Health Sciences Institute, University of California San Francisco, California, USA
| |
Collapse
|
27
|
Shui D, Borno H, Bose R, Chou J, Desai A, Fong L, Friedlander TW, Huang FW, Koshkin VS, de Kouchkovsky I, Hong JC, Mohamad O, Feng FY, Aggarwal RR, Hope TA, Small EJ, Kwon DH. Serial stereotactic body radiation therapy for oligometastatic prostate cancer (PCa) detected by positron emission tomography (PET) imaging. J Clin Oncol 2022. [DOI: 10.1200/jco.2022.40.6_suppl.109] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
109 Background: Radiopharmaceuticals, including Ga-68-prostate specific membrane antigen (PSMA)-11, F-18-fluciclovine, and choline C-11, are increasingly used to stage and inform therapies for PCa. Stereotactic body radiation therapy (SBRT) to PET-detected oligometastatic PCa has been shown to improve progression free survival (PFS) compared to observation. However, for men who subsequently develop oligorecurrent disease, outcomes following second SBRT are unknown. Methods: A retrospective, single-center, cohort study was conducted. Pts were identified through electronic health records. Inclusion criteria included pts with oligometastatic (1-5 lesions) PCa detected on PSMA, fluciclovine, or choline C-11 PET who underwent 2 consecutive courses of SBRT to tracer-avid oligometastatic disease between 7/2013 and 7/2021. Exclusion criteria included presence of visceral metastases and pure small cell neuroendocrine PCa. Data on stage, tracer type, concurrent systemic therapy, and prostate-specific antigen (PSA) responses for first SBRT (SBRT1) and second SBRT (SBRT2) were collected. Outcomes included PSA decline of ≥50% (PSA50), ≥90% (PSA90), and PSA-PFS. SBRT2 outcomes were compared based on change of concurrent systemic therapy with SBRT2 (e.g., addition of abiraterone or anti-androgen withdrawal) and PSA50 to SBRT1 using Fisher’s exact text and Wilcoxon rank sum test, respectively. Results: A total of 12 pts met eligibility criteria. At SBRT1, 10 (83%) pts had hormone-sensitive PCa (HSPC) and 2 (17%) had castration-resistant PCa (CRPC). For PET tracers, 7 (58%) used PSMA, 4 (33%) fluciclovine, and 1 (8%) choline. After SBRT1, 12 pts (100%) had a PSA decline, 8 (67%) had a PSA50 response, and 6 (50%) a PSA90 response. Median PSA PFS after SBRT1 was 30mo (95%CI 9-65mo). Six (50%) SBRT1 pts had a concurrent change in systemic therapy. At SBRT2, 8 (67%) pts had HSPC and 4 (33%) had CRPC; 7 (58%) used PSMA and 5 (42%) fluciclovine. After SBRT2, 12 (100%) pts had a PSA decline, 8 (67%) had a PSA50 response, and 8 (67%) a PSA90 response. After SBRT2, median PSA PFS was 23mo (95%CI 12-35mo). Among 7 pts who had a concurrent change in systemic therapy with SBRT2, all (100%) had a PSA50 response; among 5 who did not (4 of whom did not receive any systemic therapy), 1 (20%) had a PSA50 response (P=0.01). Among 8 pts who had a PSA50 response to SBRT1, 7 (88%) had one to SBRT2; among 4 who did not have a PSA50 response to SBRT1, 1 (25%) had one to SBRT2 (P=0.01). No complications related to SBRT were documented. Conclusions: Serial SBRT for oligometastatic PCa detected on fluciclovine, PSMA, or choline PET is feasible and can achieve PSA declines independent of systemic therapy. PSA responses were greater when systemic therapy was changed. This preliminary evidence of benefit, based on PSA responses and PSA PFS, provides rationale for larger, prospective studies of serial SBRT for oligometastatic PCa.
Collapse
Affiliation(s)
- David Shui
- University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | - Hala Borno
- University of California, San Francisco Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | - Rohit Bose
- University of California, San Francisco, San Francisco, CA
| | - Jonathan Chou
- University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | - Arpita Desai
- University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | - Lawrence Fong
- University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | - Terence W. Friedlander
- University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | | | | | - Ivan de Kouchkovsky
- University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | | | - Osama Mohamad
- University of California, San Francisco, San Francisco, CA
| | - Felix Y Feng
- Department of Urology, University of California, San Francisco, CA
| | | | - Thomas A Hope
- University of California San Francisco, Department of Radiology and Biomedical Imaging, San Francisco, CA
| | - Eric Jay Small
- UCSF Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | | |
Collapse
|
28
|
Stephens SJ, Lloyd MR, Hong JC, Mehta T, James TA, Blitzblau R, Recht A, Spiegel DY. Abstract P3-04-04: Multi-institutional perspective on screening mammography and breast cancer stage at diagnosis during the COVID-19 pandemic. Cancer Res 2022. [DOI: 10.1158/1538-7445.sabcs21-p3-04-04] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/16/2022]
Abstract
Abstract
Background: During the SARS-CoV-2 pandemic in 2020, the use of routine screening mammography (SM) and diagnostic mammography (DM) was limited for several months in order to reduce patient exposure and redeploy medical personnel. Previous studies suggest such delays result in more late-stage breast cancer diagnoses. We hypothesized that this impact would vary between institutions depending on regional variations in shutdown periods and the ability and willingness of patients to resume screening. Methods: Patients diagnosed with invasive breast cancers from 2016-2020 were identified using the Beth Israel Deaconess Medical Center (BIDMC) and the Duke University Medical Center (DUMC) cancer registries. Rates of mammography were ascertained from billing data. Baseline patient characteristics, demographics, and clinical information were gathered and cross-referenced with the electronic medical record. Late-stage was defined as Anatomic Stage III-IV disease (AJCC 8th edition). Chi-squared analysis was used to examine monthly distributions in stage at presentation for diagnosis in 2016-2019 compared to in 2020 at each institution. Results: There were 5907 patients diagnosed with invasive breast cancer between 2016-2019 (1597 at BIDMC and 4310 at DUMC) and 1075 in 2020 (333 and 742, respectively). Mammography was limited from 3/16/20-6/8/20 at BIDMC and from 3/16/20-4/20/20 at DUMC. There were fewer SM at each institution during their respective shutdown periods in 2020 than in the same months in 2019: BIDMC 1713 versus 8566 (80% reduction) and at DUMC 1649 versus 5698 (71% reduction). Following the pandemic shutdown, SM volume increased in July-December 2020 compared to July-December 2019 (108% at BIDMC and 116% at DUMC). The proportion of patients diagnosed with late-stage disease at BIDMC was greater in 2020 than in 2016-2019, at 12.6% and 6.6%, respectively (p < 0.001); 86% of late-stage diagnoses and 68% of all diagnoses in 2020 at BIDMC occurred from July-December following the initial shutdown period. The proportion of patients diagnosed with late-stage disease at DUMC in these two cohorts were 14.3% in 2020 and 16.2%% in 2016-2019 (p = 0.1); 50% of late-stage diagnoses and 51% of all diagnoses in 2020 at DUMC occurred in the period following the initial shutdown from July-December. Conclusion: We identified variation between two large academic medical centers in the impact of the SARS-CoV-2 pandemic shutdown on the proportion of late-stage breast cancer diagnoses. These dissimilar outcomes may be the result of differences in referral patterns as well as regional differences in the approach to SM during the pandemic. In particular, a shorter closure time and substantial increase in SM volume following the initial shutdown period in the Southeast region may have prevented an increase in late-stage diagnoses. Further information and analysis may help suggest additional strategies to minimize adverse effects of reduced cancer screening in future public-health emergencies.
Table 1.Proportion of SM in 2020 compared to 2019 and proportion of late disease per month in 2020BIDMC 2020JanFebMarAprMayJunJulAugSepOct NovDecSM, %96109480034104106112105113112Late Disease, n(%)0 (0)2 (10)1 (3)0 (0)0 (0)3 (21)1 (7)9 (21)10 (15)10 (19)5 (12)1 (8)DUMC 2020JanFebMarAprMayJunJulAugSepOctNovDecSM, %565860040123111109127116120119Late Disease, n(%)12 (11)13 (19)14 (19)4 (18)4 (12)6 (11)12 (13)10 (13)6 (10)12 (19)7 (13)5 (13)
Citation Format: Sarah J Stephens, Maxwell R Lloyd, Julian C Hong, Tejas Mehta, Ted A James, Rachel Blitzblau, Abram Recht, Daphna Y Spiegel. Multi-institutional perspective on screening mammography and breast cancer stage at diagnosis during the COVID-19 pandemic [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-04-04.
Collapse
Affiliation(s)
| | | | - Julian C Hong
- University of California San Francisco, San Francisco, CA
| | - Tejas Mehta
- Beth Israel Deaconess Medical Center, Boston, MA
| | - Ted A James
- Beth Israel Deaconess Medical Center, Boston, MA
| | | | - Abram Recht
- Beth Israel Deaconess Medical Center, Boston, MA
| | | |
Collapse
|
29
|
Chen WC, Boreta L, Braunstein SE, Rabow MW, Kaplan LE, Tenenbaum JD, Morin O, Park CC, Hong JC. Association of mental health diagnosis with race and all-cause mortality after a cancer diagnosis: Large-scale analysis of electronic health record data. Cancer 2022; 128:344-352. [PMID: 34550601 PMCID: PMC8738115 DOI: 10.1002/cncr.33903] [Citation(s) in RCA: 6] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/18/2021] [Revised: 05/19/2021] [Accepted: 05/20/2021] [Indexed: 01/17/2023]
Abstract
BACKGROUND Disparity in mental health care among cancer patients remains understudied. METHODS A large, retrospective, single tertiary-care institution cohort study was conducted based on deidentified electronic health record data of 54,852 adult cancer patients without prior mental health diagnosis (MHD) diagnosed at the University of California, San Francisco between January 2012 and September 2019. The exposure of interest was early-onset MHD with or without psychotropic medication (PM) within 12 months of cancer diagnosis and primary outcome was all-cause mortality. RESULTS There were 8.2% of patients who received a new MHD at a median of 197 days (interquartile range, 61-553) after incident cancer diagnosis; 31.0% received a PM prescription; and 3.7% a mental health-related visit (MHRV). There were 62.6% of patients who were non-Hispanic White (NHW), 10.8% were Asian, 9.8% were Hispanic, and 3.8% were Black. Compared with NHWs, minority cancer patients had reduced adjusted odds of MHDs, PM prescriptions, and MHRVs, particularly for generalized anxiety (Asian odds ratio [OR], 0.66, 95% CI, 0.55-0.78; Black OR, 0.60, 95% CI, 0.45-0.79; Hispanic OR, 0.72, 95% CI, 0.61-0.85) and selective serotonin-reuptake inhibitors (Asian OR, 0.43, 95% CI, 0.37-0.50; Black OR, 0.51, 95% CI, 0.40-0.61; Hispanic OR, 0.79, 95% CI, 0.70-0.89). New early MHD with PM was associated with elevated all-cause mortality (12-24 months: hazard ratio [HR], 1.43, 95% CI, 1.25-1.64) that waned by 24 to 36 months (HR, 1.18, 95% CI, 0.95-1.45). CONCLUSIONS New mental health diagnosis with PM was a marker of early mortality among cancer patients. Minority cancer patients were less likely to receive documentation of MHDs or treatment, which may represent missed opportunities to identify and treat cancer-related mental health conditions.
Collapse
Affiliation(s)
- William C Chen
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
| | - Lauren Boreta
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
| | - Steve E Braunstein
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
| | - Michael W Rabow
- Department of Internal Medicine, Division of Palliative Medicine, and Department of Urology, University of California San Francisco, California
| | - Lawrence E Kaplan
- Department of Psychiatry, University of California San Francisco, California
| | | | - Olivier Morin
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
| | - Catherine C Park
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
| | - Julian C Hong
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
- Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA
| |
Collapse
|
30
|
Borno HT, Kim MO, Hong JC, Yousefi S, Lin A, Tolstykh I, Zhang S, McKay RR, Harismendy O, Cinar P, Rugo H, Koshkin VS, Rabow M, Wang C, Bailey A, Small EJ. OUP accepted manuscript. Oncologist 2022; 27:398-406. [PMID: 35348771 PMCID: PMC9074994 DOI: 10.1093/oncolo/oyac038] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/08/2021] [Accepted: 01/07/2022] [Indexed: 11/21/2022] Open
Abstract
Background The risks associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its associated illness, coronavirus disease 2019 (COVID-19), among patients with a cancer diagnosis have not been fully characterized. This study leverages data from a multi-institutional cohort study, the University of California Cancer COVID Consortium, to evaluate outcomes associated with SARS-CoV-2 infection among patients with cancer. Methods Clinical data were collected from March to November 2020 and included patient demographics, cancer history and treatment, SARS-CoV-2 exposure and testing, and COVID-19 clinical management and outcomes. Multivariate ordinal logistic regression permitting unequal slopes was used to evaluate the impact of demographic, disease, and treatment factors on SARS-CoV-2 related hospitalization, intensive care unit (ICU) admission, and mortality. Findings Among all evaluated patients (n = 303), 147 (48%) were male, 118 (29%) were older adults (≥65 years old), and 104 (34%) were non-Hispanic white. A subset (n = 63, 21%) had hematologic malignancies and the remaining had solid tumors. Patients were hospitalized for acute care (n = 79, 26%), ICU-level care (n = 28, 9%), or died (n = 21, 7%) due to COVID-19. Patients with ≥2 comorbidities were more likely to require acute care (odds ratio [OR] 2.09 [95% confidence interval (CI), 1.23-3.55]). Cough was identified as a significant predictor of ICU hospitalization (OR 2.16 [95% CI, 1.03-4.57]). Importantly, mortality was associated with an active cancer diagnosis (OR 3.64 [95% CI, 1.40-9.5]) or advanced age (OR 3.86 [95% CI, 1.2-12.44]). Interpretation This study observed that patients with active cancer or advanced age are at an increased risk of death from COVID-19. These study observations can inform risk counseling related to COVID-19 for patients with a cancer diagnosis.
Collapse
Affiliation(s)
- Hala T Borno
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
- Corresponding author: Hala T. Borno, MD 550 16th Street, 6th Floor, Box 3211 San Francisco, CA 94158.
| | - Mi-Ok Kim
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
- Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA
| | - Julian C Hong
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA, USA
- Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA, USA
| | - Sasha Yousefi
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA, USA
- Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA, USA
| | - Amy Lin
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
| | - Irina Tolstykh
- Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA
| | - Sylvia Zhang
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
| | - Rana R McKay
- Department of Medicine, University of California, San Diego, San Diego, CA, USA
- Moores Cancer Center, University of California, San Diego, San Diego, CA, USA
| | - Olivier Harismendy
- Moores Cancer Center, University of California, San Diego, San Diego, CA, USA
- Division of Biomedical Informatics, Department of Medicine, University of California, San Diego, San Diego, CA, USA
| | - Pelin Cinar
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
| | - Hope Rugo
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
| | - Vadim S Koshkin
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
| | - Maya Rabow
- College of Science, Northeastern University, Boston, MA, USA
| | | | - Adina Bailey
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
| | - Eric J Small
- Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
- Helen Diller Family Comprehensive Cancer Center, San Francisco, CA, USA
| |
Collapse
|
31
|
Garriga M, Feng FY, Lee WR, Hong JC. Early salvage versus adjuvant therapy for treatment of prostate cancer following prostatectomy. BMJ Evid Based Med 2021; 26:e8. [PMID: 33361287 DOI: 10.1136/bmjebm-2020-111592] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Accepted: 12/02/2020] [Indexed: 11/04/2022]
Affiliation(s)
- Meera Garriga
- School of Medicine, University of California San Francisco, San Francisco, California, USA
| | - Felix Y Feng
- Department of Radiation Oncology, University of California San Francisco, San Francisco, California, USA
| | - W Robert Lee
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Julian C Hong
- Department of Radiation Oncology, University of California San Francisco, San Francisco, California, USA
- Bakar Computational Health Sciences Institute, University of California San Francisco, San Francisco, California, USA
| |
Collapse
|
32
|
Affiliation(s)
- Julian C Hong
- Department of Radiation Oncology, University of California, San Francisco
- Bakar Computational Health Sciences Institute, University of California, San Francisco
| |
Collapse
|
33
|
Miao K, Dahle J, Yousefi S, Buchake B, Kaur P, Odisho AY, Cinar P, Hong JC. Machine learning-based approach to the risk assessment of potentially preventable outpatient cancer treatment-related emergency care and hospitalizations. J Clin Oncol 2021. [DOI: 10.1200/jco.2020.39.28_suppl.333] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
333 Background: Patients undergoing outpatient infusion chemotherapy for cancer are at risk for potentially preventable, unplanned acute care in the form of emergency department (ED) visits and hospital admissions. This can impact outcomes, patient decisions, and costs to the patient and healthcare system. To address this need, the Centers for Medicare & Medicaid Services developed the Chemotherapy Measure (OP-35). Recent randomized controlled data indicate that electronic health record (EHR)-based machine learning (ML) approaches accurately direct supportive care to reduce acute care during radiotherapy. As this may extend to systemic therapy, this study aims to develop and evaluate ML approaches to predict the risk of OP-35 qualifying, potentially preventable acute care within 30 days of infusional systemic therapy. Methods: This study included data from UCSF cancer patients receiving infusional chemotherapy from July 1, 2017, to February 11, 2021, (total 7,068 patients over 84,174 treatments). The data incorporated into the ML included 430 EHR-derived variables, including cancer diagnosis, therapeutic agents, laboratory values, vital signs, medications, and encounter history. Three ML approaches were trained to predict an OP-35 acute care risk following a systemic therapy infusion with least absolute shrinkage selection operator (LASSO), random forest, and gradient boosted trees (GBT; XGBoost) approaches. The models were trained on a subset (75% of patients; before October 12, 2019) of the dataset and validated on a mutually exclusive subset (25% patients; after October 12, 2019) based on the receiver operating characteristic (ROC) curves and calibration plots. Results: There were 1,651 total acute care visits (244 ED visits and 1,407 ED visits converted into hospitalization); 1,310 infusions included a qualifying acute care visit (200 with ED visits only, 0 direct hospital admissions, and 1,110 with both ED visit and hospitalization). Each ML approach demonstrated good performance in the internal validation cohort, with GBT (AUC 0.805) outpacing the random forest (0.750) and LASSO logistic regression (0.755) approaches. Visualization of calibration plots verified concordance between predicted and observed rates of acute care. All three models shared patient age and days elapsed since last treatment as important contributors. Conclusions: EHR-based ML approaches demonstrate high predictive ability for OP-35 qualifying acute care rates on a per-infusion basis, identifying 30-day potentially preventable acute care risk for patients undergoing chemotherapy. Prospective validation of these models is ongoing. Early prediction can facilitate interventional strategies which may reduce acute care, improve health outcomes, and reduce costs.
Collapse
Affiliation(s)
- Kevin Miao
- University of California, San Francisco, San Francisco, CA
| | - Justice Dahle
- University of California, San Francisco, San Francisco, CA
| | | | - Bilwa Buchake
- University of California, San Francisco, San Francisco, CA
| | - Parambir Kaur
- University of California, San Francisco, San Francisco, CA
| | | | - Pelin Cinar
- University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | | |
Collapse
|
34
|
Ragavan MV, LaLanne A, Skafel A, Hong JC, Odisho AY, Yousefi S, Small EJ, Borno H. Evaluating changes in “good safety monitoring” for cancer clinical trial participants during the COVID-19 pandemic. J Clin Oncol 2021. [DOI: 10.1200/jco.2020.39.28_suppl.217] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
217 Background: Comprehensive and frequent safety monitoring is an essential component of clinical trial conduct to accurately characterize potential toxicities of a study drug and to minimize potential harm to study participants. The COVID-19 pandemic substantially impacted the delivery of cancer care with reduced frequency of overall and in-person visits. We hypothesized that reporting of serious adverse events (SAEs) occurring on clinical trials may have been impacted by these care delivery changes. The current study evaluated pandemic-related changes in the frequency of safety monitoring for cancer patients (pts) enrolled on a clinical trial and identified predictors of SAE reporting before and during the pandemic. Methods: This study included all adult cancer pts enrolled in interventional therapeutic clinical trials at an academic cancer center between 1/1/2019 and 12/30/2020. In this analysis, the "pre-pandemic" period was defined as the time between 1/1/19 and 3/14/20, and the pandemic period between 3/15/20 and the data cutoff date of 12/30/2020. SAE was defined as a grade 3 or grade 4 adverse event (AE) as reported by the trial. Demographic characteristics of pts, visit type (virtual vs in-person), and frequency of SAE reporting were summarized pre-pandemic and during the pandemic. A multivariate logistic regression model was employed to identify predictors of SAE reporting, with the outcome defined as report of at least one SAE from the time pts went on study until the data cutoff date. Covariates included age, gender, race (white vs. non-white), having at least one virtual visit, and enrollment on a trial before versus during the pandemic. Results: This study included 190 pts; 138 (73%) enrolled on trial pre-pandemic and 52 (27%) enrolled during the pandemic. During-pandemic participants were more likely to be older than pts enrolled pre-pandemic, but otherwise the groups were similar in terms of race and gender. Overall, 78 pts (41%) reported an SAE. Among pre-pandemic enrollees, 50% reported at least one SAE, compared to 17% among during-pandemic enrollees. In the multivariate logistic regression model, only enrolling on trial pre-pandemic was associated with a higher likelihood of reporting at least one SAE. Visit type (virtual vs. in-person) was not recorded in over half of during-pandemic patient encounters. Conclusions: There was a significant decline in frequency of SAE reporting during the COVID-19 pandemic. While having at least one virtual visit was not a significant predictor of SAE reporting in the multivariate regression model, our analysis may underrepresent the association of virtual visits and SAE reporting. As the number of virtual visits is expected to stay high post-pandemic, further work is needed to characterize the association of virtual visits and SAE reporting to ensure ongoing adequate safety monitoring for clinical trial patients.
Collapse
Affiliation(s)
- Meera Vimala Ragavan
- Division of Hematology/Oncology, Department of Medicine, University of California San Francisco, San Francisco, CA
| | - Alyssa LaLanne
- UCSF Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | - Andrea Skafel
- UCSF Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | | | | | | | - Eric Jay Small
- UCSF Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| | - Hala Borno
- University of California, San Francisco Helen Diller Family Comprehensive Cancer Center, San Francisco, CA
| |
Collapse
|
35
|
Affiliation(s)
- Julian C Hong
- Department of Radiation Oncology, University of California, San Francisco
- Bakar Computational Health Sciences Institute, University of California, San Francisco
| | - Atul J Butte
- Bakar Computational Health Sciences Institute, University of California, San Francisco
- Department of Pediatrics, University of California, San Francisco
| |
Collapse
|
36
|
Abstract
This cohort study of patients with advanced cancer who received palliative radiotherapy within 30 days of death assesses models of prognostic criteria for providing radiotherapy at the end of life and compares outcomes with similar prognostic tools.
Collapse
Affiliation(s)
- Susan Y. Wu
- Division of Radiation Oncology, University of Texas MD Anderson Cancer Center, Houston
| | - Emily Yee
- Department of Radiation Oncology, University of California, San Francisco
| | | | - Shannon E. Fogh
- Department of Radiation Oncology, University of California, San Francisco
| | - Lauren Boreta
- Department of Radiation Oncology, University of California, San Francisco
| | | | - Julian C. Hong
- Department of Radiation Oncology, University of California, San Francisco
| |
Collapse
|
37
|
Sinha S, Garriga M, Naik N, McSteen BW, Odisho AY, Lin A, Hong JC. Disparities in Electronic Health Record Patient Portal Enrollment Among Oncology Patients. JAMA Oncol 2021; 7:935-937. [PMID: 33830178 DOI: 10.1001/jamaoncol.2021.0540] [Citation(s) in RCA: 23] [Impact Index Per Article: 7.7] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/14/2022]
Affiliation(s)
- Sumi Sinha
- Department of Radiation Oncology, University of California, San Francisco
| | - Meera Garriga
- University of California School of Medicine, San Francisco
| | - Nishali Naik
- Department of Radiation Oncology, University of California, San Francisco
| | | | - Anobel Y Odisho
- Department of Urology, University of California, San Francisco
| | - Amy Lin
- Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco
| | - Julian C Hong
- Department of Radiation Oncology, University of California, San Francisco.,University of California School of Medicine, San Francisco.,Bakar Computational Health Sciences Institute, University of California, San Francisco
| |
Collapse
|
38
|
Natesan D, Thomas SM, Eisenstein E, Eclov N, Dalal N, Stephens SJ, Malicki M, Shields S, Cobb A, Mowery YM, Niedzwiecki D, Tenenbaum J, Palta M, Hong JC. Impact of machine learning-directed on-treatment evaluations on cost of acute care visits: Economic analysis of SHIELD-RT. J Clin Oncol 2021. [DOI: 10.1200/jco.2021.39.15_suppl.1509] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
1509 Background: SHIELD-RT was a randomized controlled quality improvement study (NCT03775265) that implemented electronic health record-based machine learning (ML) to direct supplemental visits for high risk (HR) patients undergoing radiotherapy (RT). Acute care visits (ER visits or hospitalizations) were reduced from 22% to 12%. We evaluated the costs associated with acute visits in this study. Methods: Patients who initiated RT between 1/7/19 and 6/30/19 at a single institution were evaluated by a ML algorithm to identify HR courses (>10% risk of acute visit during RT). HR patients were randomized to standard weekly (S) or intervention of twice weekly (TW) evaluation during RT. Cost data associated with acute visits were obtained and compared between patients who underwent S or TW evaluations. Missing cost data were imputed using disease related groups (DRGs). Mean costs (standard deviation) were compared between arms with non-parametric Wilcoxon Rank Sum tests. Results: 311 HR courses were identified and randomized to either S (n=157) or TW (n=154) evaluations during RT. 85 patients (S: 51; TW: 34) had 121 distinct acute care visits (S: 74; TW: 47). Patients in the TW evaluation arm had fewer hospitalizations (29 vs 41) and ER visits (18 vs 33) than those in the S arm. There were fewer acute visits per patient in the TW arm (0.34) compared to S arm (0.49). Actual cost data was available for 102 visits at our institution, and imputed for 19 outside hospital visits. Mean cost associated with acute visits was lower in the TW arm ($1939, SD $5912) compared with the S arm ($4002, SD $11568; p=0.03). Differences in mean cost between arms are presented in the table. Conclusions: ML-directed evaluations for HR patients undergoing RT resulted in decreased costs of ER visits and hospitalizations. Costs were decreased across revenue centers, with the largest difference related to inpatient room costs. Future analyses will incorporate intervention costs, which are currently bundled with RT reimbursement.[Table: see text]
Collapse
Affiliation(s)
- Divya Natesan
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | - Samantha M. Thomas
- Duke University Medical Center, Department of Biostatistics and Bioinformatics, Durham, NC
| | | | - Neville Eclov
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | | | - Sarah J. Stephens
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | - Mary Malicki
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | - Stacey Shields
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | - Alyssa Cobb
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | - Yvonne Marie Mowery
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | - Donna Niedzwiecki
- Duke University Medical Center, Department of Biostatistics and Bioinformatics, Durham, NC
| | | | - Manisha Palta
- Duke University Medical Center, Department of Radiation Oncology, Durham, NC
| | | |
Collapse
|
39
|
Abstract
528 Background: During the SARS-CoV-2 pandemic, routine screening mammography (SM) was stopped and diagnostic mammography (DM) was limited for several months across the United States in order to reduce patient exposure and redeploy medical personnel. We hypothesized that this delay would result in patients presenting with later-stage disease following the initial shutdown. Methods: Patients diagnosed with invasive breast cancers from 2016-2020 were identified using the Beth Israel Deaconess Medical Center Cancer Registry. Baseline patient characteristics, demographics, and clinical information were gathered and cross-referenced with our electronic medical record. Late-stage disease was defined as initial anatomic stage III-IV disease in the AJCC 8th edition staging system. The control cohort consisted of patients diagnosed from 2016-2019; patients diagnosed in 2020 were the test cohort. Chi-squared analysis was used to compare monthly distributions in stage at diagnosis between the control and test cohorts. Multivariate analysis was performed using a logistic regression model. Results: There were 1597 patients diagnosed with invasive breast cancer between 2016-2019 and 333 in 2020. Median age at diagnosis was 60 years; 99% were female, and 69.1% were white. Mammography was limited from 3/16/20-6/8/20, with 90% reduction in volume during this time. The number of screening studies performed in March, April, May, and June of 2020 were 987, 1, 4, and 721 compared to 2042, 2141, 2241, and 2142 in 2019. The volume of new diagnoses per month decreased substantially during the shutdown (see table). The proportion of patients diagnosed with late-stage disease was 6.6% in the control cohort compared to 12.6% in the 2020 test cohort (p < 0.001); 92.9% of late-stage diagnoses in 2020 occurred from June to December following the shutdown period. On multivariate analysis, year of diagnosis (2020 vs 2016-2019; OR = 4.25 95% CI 0.035-0.095, p < 0.001), lower income (<200% of the federal poverty level; OR = 2.73 95% CI 0.016-0.099, p = 0.006) and increased Charlson Comorbidity Index (OR = 12.01 95% CI 0.037-0.052, p < 0.001) were associated with later stage at diagnosis. Conclusions: Patients were more likely to be diagnosed with late-stage breast cancer following the global shutdown due to the SARS-CoV-2 pandemic. Patients with lower income and medical comorbidities were disproportionately affected. These data raise significant concerns regarding the impact of SARS-CoV-2 on cancer diagnoses and long-term outcomes, especially in vulnerable patient populations.[Table: see text]
Collapse
Affiliation(s)
| | | | | | - Ted A. James
- Beth Israel Deaconess Medical Center, Boston, MA
| | - Tejas Mehta
- Beth Israel Deaconess Medical Center, Boston, MA
| | - Abram Recht
- Beth Israel Deaconess Medical Center, Boston, MA
| | - Daphna Spiegel
- Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA
| |
Collapse
|
40
|
Hong JC, Hauser ER, Redding TS, Sims KJ, Gellad ZF, O'Leary MC, Hyslop T, Madison AN, Qin X, Weiss D, Bullard AJ, Williams CD, Sullivan BA, Lieberman D, Provenzale D. Characterizing chronological accumulation of comorbidities in healthy veterans: a computational approach. Sci Rep 2021; 11:8104. [PMID: 33854078 PMCID: PMC8046765 DOI: 10.1038/s41598-021-85546-2] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/13/2020] [Accepted: 12/14/2020] [Indexed: 12/13/2022] Open
Abstract
Understanding patient accumulation of comorbidities can facilitate healthcare strategy and personalized preventative care. We applied a directed network graph to electronic health record (EHR) data and characterized comorbidities in a cohort of healthy veterans undergoing screening colonoscopy. The Veterans Affairs Cooperative Studies Program #380 was a prospective longitudinal study of screening and surveillance colonoscopy. We identified initial instances of three-digit ICD-9 diagnoses for participants with at least 5 years of linked EHR history (October 1999 to December 2015). For diagnoses affecting at least 10% of patients, we calculated pairwise chronological relative risk (RR). iGraph was used to produce directed graphs of comorbidities with RR > 1, as well as summary statistics, key diseases, and communities. A directed graph based on 2210 patients visualized longitudinal development of comorbidities. Top hub (preceding) diseases included ischemic heart disease, inflammatory and toxic neuropathy, and diabetes. Top authority (subsequent) diagnoses were acute kidney failure and hypertensive chronic kidney failure. Four communities of correlated comorbidities were identified. Close analysis of top hub and authority diagnoses demonstrated known relationships, correlated sequelae, and novel hypotheses. Directed network graphs portray chronologic comorbidity relationships. We identified relationships between comorbid diagnoses in this aging veteran cohort. This may direct healthcare prioritization and personalized care.
Collapse
Affiliation(s)
- Julian C Hong
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA. .,Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA, USA. .,Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA, USA.
| | - Elizabeth R Hauser
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA.,Department of Biostatistics and Bioinformatics, Duke University, Durham, NC, USA
| | - Thomas S Redding
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA
| | - Kellie J Sims
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA
| | - Ziad F Gellad
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA.,Department of Medicine, Duke University, Durham, NC, USA
| | - Meghan C O'Leary
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA
| | - Terry Hyslop
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA.,Department of Biostatistics and Bioinformatics, Duke University, Durham, NC, USA
| | - Ashton N Madison
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA
| | - Xuejun Qin
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA.,Department of Biostatistics and Bioinformatics, Duke University, Durham, NC, USA
| | - David Weiss
- Cooperative Studies Program Coordinating Center, Perry Point VA Medical Center, Perry Point, MD, USA
| | - A Jasmine Bullard
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA
| | - Christina D Williams
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA.,Department of Medicine, Duke University, Durham, NC, USA
| | - Brian A Sullivan
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA.,Department of Medicine, Duke University, Durham, NC, USA
| | - David Lieberman
- VA Portland Health Care System, Portland, OR, USA.,Oregon Health and Science University, Portland, OR, USA
| | - Dawn Provenzale
- Cooperative Studies Program Epidemiology Center-Durham, Durham VA Health Care System, Durham, NC, USA. .,Department of Medicine, Duke University, Durham, NC, USA.
| |
Collapse
|
41
|
Spiegel DY, Boyer MJ, Hong JC, Williams CD, Kelley MJ, Salama JK, Palta M. Survival Advantage With Adjuvant Chemotherapy for Locoregionally Advanced Rectal Cancer: A Veterans Health Administration Analysis. J Natl Compr Canc Netw 2021; 18:52-58. [PMID: 31910388 DOI: 10.6004/jnccn.2019.7329] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/04/2019] [Accepted: 06/11/2019] [Indexed: 11/17/2022]
Abstract
BACKGROUND Adjuvant chemotherapy (AC) after chemoradiation (CRT) and surgery for locoregionally advanced rectal cancer (LARC) is a standard of care in the United States. This study examined the role, optimal regimen, and duration of AC using data from the largest integrated health system in the United States. PATIENTS AND METHODS Using the Veterans Affairs Central Cancer Registry, patients with stage II-III rectal cancer diagnosed in 2001 through 2011 who received neoadjuvant CRT and surgery with or without AC were identified. Kaplan-Meier analysis, log-rank tests, and propensity score (PS) adjustment analysis were used to assess survival. RESULTS A total of 866 patients were identified; 417 received AC and 449 did not (observation [OBS] group). Median follow-up was 109 months. Median disease-specific survival (DSS) was not reached. Six-year DSS was 73.7%; 79.5% for the AC group versus 68.0% for the OBS group. PS-matched analysis for DSS favored AC (P=.0002). Median overall survival (OS) was 90.8 months. Six-year OS was 56.7%; 64.3% for AC versus 49.6% for OBS. In PS-matched analysis, median OS was 117.4 months for AC and 74.3 months for OBS (P<.0001). A DSS advantage was seen when comparing ≥4 months with <4 months of AC (P=.023). No difference in DSS or OS was seen with single-agent versus multiagent AC. CONCLUSIONS In this population of patients with LARC treated with neoadjuvant CRT and surgery, OS and DSS were improved among those treated with AC versus OBS. DSS benefits were seen with ≥4 months of AC. No additional benefit was observed with multiagent therapy. In the absence of phase III data, these findings support the use of AC for LARC.
Collapse
Affiliation(s)
- Daphna Y Spiegel
- Department of Radiation Oncology, Duke University, Durham, North Carolina.,Department of Radiation Oncology, Beth Israel Deaconess Medical Center, Boston, Massachusetts; and
| | - Matthew J Boyer
- Department of Radiation Oncology, Duke University, Durham, North Carolina
| | - Julian C Hong
- Department of Radiation Oncology, Duke University, Durham, North Carolina
| | - Christina D Williams
- Cooperative Studies Program Epidemiology Center-Durham, Durham Veterans Administration Medical Center
| | - Michael J Kelley
- Division of Medical Oncology, Department of Medicine, Duke University, and.,Division of Hematology-Oncology, Medical Service, Durham Veterans Administration Medical Center, Durham, North Carolina
| | - Joseph K Salama
- Department of Radiation Oncology, Duke University, Durham, North Carolina
| | - Manisha Palta
- Department of Radiation Oncology, Duke University, Durham, North Carolina
| |
Collapse
|
42
|
Alvarez JB, Bibault JE, Burgun A, Cai J, Cao Z, Chang K, Chen JH, Chen WC, Cho M, Cho PJ, Cornish TC, Costa A, Dekker A, Drukker K, Dunn J, Eminaga O, Erickson BJ, Fournier L, Gambhir SS, Gennatas ED, Giger ML, Halilaj I, Harrison AP, He B, Hong JC, Jin D, Jin MC, Jochems A, Kalpathy-Cramer J, Kapp DS, Karimzadeh M, Karnes W, Lambin P, Langlotz CP, Lee J, Li H, Liao JC, Lin AL, Lin RY, Liu Y, Lu L, Magnus D, McIntosh C, Miao S, Min JK, Neill DB, Oermann EK, Ouyang D, Peng L, Phene S, Poirot MG, Quon JL, Ranti D, Rao A, Raskar R, Rombaoa C, Rubin DL, Samarasena J, Seekins J, Seetharam K, Shearer E, Sibley A, Singh K, Singh P, Sordo M, Suraweera D, Valliani AAA, van Wijk Y, Vepakomma P, Wang B, Wang G, Wang N, Wang Y, Warner E, Welch M, Wong K, Wu Z, Xing F, Xing L, Yan K, Yan P, Yang L, Yeom KW, Zachariah R, Zeng D, Zhang L, Zhang L, Zhang X, Zhou L, Zou J. List of contributors. Artif Intell Med 2021. [DOI: 10.1016/b978-0-12-821259-2.00035-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/26/2022]
|
43
|
Lin AL, Chen WC, Hong JC. Electronic health record data mining for artificial intelligence healthcare. Artif Intell Med 2021. [DOI: 10.1016/b978-0-12-821259-2.00008-9] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/23/2022]
|
44
|
Hong JC, Fairchild AT, Tanksley JP, Palta M, Tenenbaum JD. Natural language processing for abstraction of cancer treatment toxicities: accuracy versus human experts. JAMIA Open 2020; 3:513-517. [PMID: 33623888 PMCID: PMC7886534 DOI: 10.1093/jamiaopen/ooaa064] [Citation(s) in RCA: 14] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/31/2020] [Revised: 10/26/2020] [Accepted: 10/30/2020] [Indexed: 12/29/2022] Open
Abstract
Objectives Expert abstraction of acute toxicities is critical in oncology research but is labor-intensive and variable. We assessed the accuracy of a natural language processing (NLP) pipeline to extract symptoms from clinical notes compared to physicians. Materials and Methods Two independent reviewers identified present and negated National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) v5.0 symptoms from 100 randomly selected notes for on-treatment visits during radiation therapy with adjudication by a third reviewer. A NLP pipeline based on Apache clinical Text Analysis Knowledge Extraction System was developed and used to extract CTCAE terms. Accuracy was assessed by precision, recall, and F1. Results The NLP pipeline demonstrated high accuracy for common physician-abstracted symptoms, such as radiation dermatitis (F1 0.88), fatigue (0.85), and nausea (0.88). NLP had poor sensitivity for negated symptoms. Conclusion NLP accurately detects a subset of documented present CTCAE symptoms, though is limited for negated symptoms. It may facilitate strategies to more consistently identify toxicities during cancer therapy.
Collapse
Affiliation(s)
- Julian C Hong
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, California, USA.,Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, California, USA.,Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Andrew T Fairchild
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Jarred P Tanksley
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Manisha Palta
- Department of Radiation Oncology, Duke University, Durham, North Carolina, USA
| | - Jessica D Tenenbaum
- Department of Biostatistics and Bioinformatics, Duke University, Durham, North Carolina, USA
| |
Collapse
|
45
|
Hong JC, Eclov NCW, Dalal NH, Thomas SM, Stephens SJ, Malicki M, Shields S, Cobb A, Mowery YM, Niedzwiecki D, Tenenbaum JD, Palta M. System for High-Intensity Evaluation During Radiation Therapy (SHIELD-RT): A Prospective Randomized Study of Machine Learning–Directed Clinical Evaluations During Radiation and Chemoradiation. J Clin Oncol 2020; 38:3652-3661. [DOI: 10.1200/jco.20.01688] [Citation(s) in RCA: 26] [Impact Index Per Article: 6.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/25/2023] Open
Abstract
PURPOSE Patients undergoing outpatient radiotherapy (RT) or chemoradiation (CRT) frequently require acute care (emergency department evaluation or hospitalization). Machine learning (ML) may guide interventions to reduce this risk. There are limited prospective studies investigating the clinical impact of ML in health care. The objective of this study was to determine whether ML can identify high-risk patients and direct mandatory twice-weekly clinical evaluation to reduce acute care visits during treatment. PATIENTS AND METHODS During this single-institution randomized quality improvement study (ClinicalTrials.gov identifier: NCT04277650 ), 963 outpatient adult courses of RT and CRT started from January 7 to June 30, 2019, were evaluated by an ML algorithm. Among these, 311 courses identified by ML as high risk (> 10% risk of acute care during treatment) were randomized to standard once-weekly clinical evaluation (n = 157) or mandatory twice-weekly evaluation (n = 154). Both arms allowed additional evaluations on the basis of clinician discretion. The primary end point was the rate of acute care visits during RT. Model performance was evaluated using receiver operating characteristic area under the curve (AUC) and decile calibration plots. RESULTS Twice-weekly evaluation reduced rates of acute care during treatment from 22.3% to 12.3% (difference, −10.0%; 95% CI, −18.3 to −1.6; relative risk, 0.556; 95% CI, 0.332 to 0.924; P = .02). Low-risk patients had a 2.7% acute care rate. Model discrimination was good in high- and low-risk patients undergoing standard once-weekly evaluation (AUC, 0.851). CONCLUSION In this prospective randomized study, ML accurately triaged patients undergoing RT and CRT, directing clinical management with reduced acute care rates versus standard of care. This prospective study demonstrates the potential benefit of ML in health care and offers opportunities to enhance care quality and reduce health care costs.
Collapse
Affiliation(s)
- Julian C. Hong
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
- Bakar Computational Health Sciences Institute, University of California, San Francisco, San Francisco, CA
- Department of Radiation Oncology, Duke University, Durham, NC
| | | | - Nicole H. Dalal
- Department of Medicine, University of California, San Francisco, San Francisco, CA
| | - Samantha M. Thomas
- Department of Biostatistics and Bioinformatics, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| | | | - Mary Malicki
- Department of Radiation Oncology, Duke University, Durham, NC
| | - Stacey Shields
- Department of Radiation Oncology, Duke University, Durham, NC
| | - Alyssa Cobb
- Department of Radiation Oncology, Duke University, Durham, NC
| | - Yvonne M. Mowery
- Department of Radiation Oncology, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| | - Donna Niedzwiecki
- Department of Biostatistics and Bioinformatics, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| | | | - Manisha Palta
- Department of Radiation Oncology, Duke University, Durham, NC
- Duke Cancer Institute, Duke University, Durham, NC
| |
Collapse
|
46
|
Affiliation(s)
- John Kang
- Department of Radiation Oncology, University of Washington, Seattle
| | - Olivier Morin
- Department of Radiation Oncology, University of San Francisco, San Francisco, California
| | - Julian C Hong
- Department of Radiation Oncology, University of San Francisco, San Francisco, California.,Bakar Computational Health Sciences Institute, University of San Francisco, San Francisco, California
| |
Collapse
|
47
|
Wang JX, Boreta L, Yee E, Braunstein SE, Hong JC. Patient characteristics associated with time-to-consult for inpatient palliative radiotherapy. J Clin Oncol 2020. [DOI: 10.1200/jco.2020.38.29_suppl.27] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
27 Background: Delayed inpatient radiotherapy (RT) consultation can result in delayed treatment, extended hospitalization, and increased costs. There are limited data characterizing populations that may have longer time-to-consult (TTC). Methods: A single institutional electronic health record deidentified corporate data warehouse was used to identify adult inpatients with a consultation to radiation oncology between 2014-2019. TTC was defined as time from admission to placement of consultation. Multivariate linear regression was used to examine adjusted associations for factors including age, diagnosis, admission date, admitting service, and patient-reported sex, race and ethnicity to TTC. Continuous variables were normalized to generate standardized beta coefficients (B). Results: A total of 856 admissions with radiation oncology consult were identified. Median TTC was 21 hours (interquartile range [IQR] 5-69). Median age was 61 (IQR 49-68) and 51% of patients were male. Most patients were white (58%) and non-Hispanic or Latino (80%). The most common admission primary diagnoses were brain metastasis (14%), bone metastasis (10%), and primary brain neoplasm (9%), Most common admitting services were Neurosurgery (49%), Hospital Medicine (22%), Malignant Hematology (8%), and Gynecologic Oncology (8%). Primary brain neoplasms (vs brain metastases B = 64; P < .001), other non-metastatic diagnoses (vs brain metastases B = 45; P < .001) and admission on the malignant hematology service (vs neurosurgery; B = 37; P = .03) were associated with longer TTC. Patient demographic characteristics (age sex, race, ethnicity), and admission date did not have significant associations with TTC. Longer TTC was correlated with longer hospitalization (Pearson’s corr = 0.48). Conclusions: No clear demographic disparities in inpatient RT consultation were identified. Certain diagnoses and services were associated with longer TTC, potentially related to clinical practice. Increased TTC was associated with longer hospital length of stay.
Collapse
Affiliation(s)
- Jon X Wang
- University of California, San Francisco, CA
| | - Lauren Boreta
- University of California San Francisco, San Francisco, CA
| | - Emily Yee
- University of California San Francisco, San Francisco, CA
| | | | | |
Collapse
|
48
|
Fairchild AT, Tanksley JP, Tenenbaum JD, Palta M, Hong JC. Interrater Reliability in Toxicity Identification: Limitations of Current Standards. Int J Radiat Oncol Biol Phys 2020; 107:996-1000. [DOI: 10.1016/j.ijrobp.2020.04.040] [Citation(s) in RCA: 9] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/13/2020] [Revised: 04/13/2020] [Accepted: 04/24/2020] [Indexed: 10/24/2022]
|
49
|
Jacobs CD, Carpenter DJ, Hong JC, Havrilesky LJ, Sosa JA, Chino JP. Radiation Records in the National Cancer Database: Variations in Coding and/or Practice Can Significantly Alter Survival Results. JCO Clin Cancer Inform 2020; 3:1-9. [PMID: 31050906 DOI: 10.1200/cci.18.00118] [Citation(s) in RCA: 14] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
PURPOSE The aim of the current work was to quantify internally inconsistent and anomalous radiation therapy (RT) data in the National Cancer Database (NCDB) and determine their association with overall survival (OS) using node-positive uterine cancer as a test clinical scenario. MATERIALS AND METHODS We identified all NCDB participants with International Federation of Gynecology and Obstetrics stage IIIC1 to IIIC2 uterine cancer treated with hysterectomy and adjuvant RT between 1998 and 2012. Variables that were reviewed to identify anomalous data included RT site, modality, dose, fractions, timing, duration, and stage. We used χ2 testing to associate anomalous data with reporting facility and demographic variables. OS was estimated using the Kaplan-Meier method and comparison between cohorts was performed using the log-rank test. Univariable and multivariable Cox proportional hazards regression analyses were performed. RESULTS Of the 14,298 analyzed participants, 2,288 (16.0%) had one or more anomalous data entry, 538 (3.8%) likely because of an incomplete RT course. χ2 testing suggested differences in anomalous data prevalence by reporting facility type (P = .0007), geographic region (P < .001), distance from participants' homes (P < .001), diagnosis year (P < .001), and location of RT relative to reporting facility (P = .0038). Five-year OS in those with one or more anomalous data entry was 51.3% versus 58.0% for those without anomalous data (P < .001), and anomalous data remained significantly associated with OS on multivariable analysis. After excluding insufficient, excessive, or unknown total RT dose, anomalous data were no longer significant on multivariable analysis. CONCLUSION The overwhelming majority of RT data within the NCDB seem to be appropriate for the clinical scenario. Nevertheless, approximately one eighth of participants in this test clinical scenario had adjuvant RT data that were internally inconsistent or outside generously defined norms. The presence of anomalous RT data was significantly associated with compromised OS, an effect not observed after correcting for total RT dose.
Collapse
Affiliation(s)
| | | | | | | | - Julie A Sosa
- University of California, San Francisco, San Francisco, CA
| | | |
Collapse
|
50
|
Chen W, Boreta L, Braunstein SE, Hong JC. Prediction of mental health disorder onset and impact on emergency visits following a cancer diagnosis. J Clin Oncol 2020. [DOI: 10.1200/jco.2020.38.15_suppl.2041] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022] Open
Abstract
2041 Background: Cancer patients are at increased risk of mental and emotional distress. The aim of this study is to investigate risk factors and timing of mental health disorder (MHD) onset following a cancer diagnosis, and evaluate its impact on emergency visits. Methods: All patients with a new onset diagnosis of malignancy ( ICD-10 codes C00-C97, with conversion of ICD-9 codes) were identified from an institutional de-identified electronic health data warehouse. Demographic data, Charlson comorbidity index excluding cancer, mortality, and time to onset of a new MHD diagnosis ( ICD-10 codes F00-F99) and emergency visits were extracted and used to calculate rates and Cox-model hazard ratios. A predictive logistic model of MHD was tested on an internal hold-out sample (25%). Results: A total of 110,306 patients with 338,208 person-years of follow up were identified with a new diagnosis of cancer from February 1980 to July 2019, of which 95,474 (86.5%) had no prior diagnosis of MHD. Actuarial rates of new MHD among previously MHD-free patients were 8.1% at 6 months, and 14.1% and 20.8% at 2 and 5 years. Median time to onset of MHD was fastest among head and neck cancer (57 days, HR 2.32 [2.1-2.6]), urinary organ cancer (94 days, HR 2.21 [2.0-2.4]), and lung and thoracic cancers (99 days, HR 2.47 [2.2-2.7]), compared to skin neoplasms (987 days, HR 1.0). Median time to onset was less than one year for all malignancies except for skin neoplasms and male genital cancers (840 days). Male sex, older age, Charlson score, divorce or legal separation, self-identification of a gender-neutral partner, African American or American Indian race, Hispanic ethnicity, current or former smoking status, and self-identification as Christian were associated with higher risk of MHD onset, while married status and native Hawaiian or Pacific Islander race were protective. A logistic model predicted new MHD with an AUROC of 0.72. Onset of new MHD was associated with greater rates of emergency visit (HR 1.92 [1.8-2.0], adjusted for cancer type and Charlson score), and patients with new MHD who experienced an emergency visit had a mean of 3.75+/-0.03 (SEM) total emergency visits versus 2.65+/-0.02 (p < 0.0001). Finally, onset of new MHD was associated with greater mortality even after adjusting for age, Charlson score and cancer type (HR 1.29, [1.23-1.35]). Conclusions: Onset of new mental health diagnosis after a cancer diagnosis was correlated with greater rates of emergency visits and mortality. Cancer patients with risk factors identified here may benefit from increased social and mental health support.
Collapse
Affiliation(s)
- William Chen
- UCSF Department of Radiation Oncology, San Francisco, CA
| | - Lauren Boreta
- University of California San Francisco, San Francisco, CA
| | | | - Julian C. Hong
- Department of Radiation Oncology, University of California, San Francisco, San Francisco, CA
| |
Collapse
|