Approach to machine learning for extraction of real-world data variables from electronic health records.

Background: As artificial intelligence (AI) continues to advance with breakthroughs in natural language processing (NLP) and machine learning (ML), such as the development of models like OpenAI's ChatGPT, new opportunities are emerging for efficient curation of electronic health records (EHR) into real-world data (RWD) for evidence generation in oncology. Our objective is to describe the research and development of industry methods to promote transparency and explainability. Methods: We applied NLP with ML techniques to train, validate, and test the extraction of information from unstructured documents (e.g., clinician notes, radiology reports, lab reports, etc.) to output a set of structured variables required for RWD analysis. This research used a nationwide electronic health record (EHR)-derived database. Models were selected based on performance. Variables curated with an approach using ML extraction are those where the value is determined solely based on an ML model (i.e. not confirmed by abstraction), which identifies key information from visit notes and documents. These models do not predict future events or infer missing information. Results: We developed an approach using NLP and ML for extraction of clinically meaningful information from unstructured EHR documents and found high performance of output variables compared with variables curated by manually abstracted data. These extraction methods resulted in research-ready variables including initial cancer diagnosis with date, advanced/metastatic diagnosis with date, disease stage, histology, smoking status, surgery status with date, biomarker test results with dates, and oral treatments with dates. Conclusion: NLP and ML enable the extraction of retrospective clinical data in EHR with speed and scalability to help researchers learn from the experience of every person with cancer.

Replication of Real-World Evidence in Oncology Using Electronic Health Record Data Extracted by Machine Learning.

Meaningful real-world evidence (RWE) generation requires unstructured data found in electronic health records (EHRs) which are often missing from administrative claims; however, obtaining relevant data from unstructured EHR sources is resource-intensive. In response, researchers are using natural language processing (NLP) with machine learning (ML) techniques (i.e., ML extraction) to extract real-world data (RWD) at scale. This study assessed the quality and fitness-for-use of EHR-derived oncology data curated using NLP with ML as compared to the reference standard of expert abstraction. Using a sample of 186,313 patients with lung cancer from a nationwide EHR-derived de-identified database, we performed a series of replication analyses demonstrating some common analyses conducted in retrospective observational research with complex EHR-derived data to generate evidence. Eligible patients were selected into biomarker- and treatment-defined cohorts, first with expert-abstracted then with ML-extracted data. We utilized the biomarker- and treatment-defined cohorts to perform analyses related to biomarker-associated survival and treatment comparative effectiveness, respectively. Across all analyses, the results differed by less than 8% between the data curation methods, and similar conclusions were reached. These results highlight that high-performance ML-extracted variables trained on expert-abstracted data can achieve similar results as when using abstracted data, unlocking the ability to perform oncology research at scale.

Exposure to US Cancer Drugs With Lack of Confirmed Benefit After US Food and Drug Administration Accelerated Approval.

This cross-sectional study evaluates patient exposure to oncology drugs withdrawn from the US Food and Drug Administration (FDA) Accelerated Approval program.

Considerations for the Use of Machine Learning Extracted Real-World Data to Support Evidence Generation: A Research-Centric Evaluation Framework.

A vast amount of real-world data, such as pathology reports and clinical notes, are captured as unstructured text in electronic health records (EHRs). However, this information is both difficult and costly to extract through human abstraction, especially when scaling to large datasets is needed. Fortunately, Natural Language Processing (NLP) and Machine Learning (ML) techniques provide promising solutions for a variety of information extraction tasks such as identifying a group of patients who have a specific diagnosis, share common characteristics, or show progression of a disease. However, using these ML-extracted data for research still introduces unique challenges in assessing validity and generalizability to different cohorts of interest. In order to enable effective and accurate use of ML-extracted real-world data (RWD) to support research and real-world evidence generation, we propose a research-centric evaluation framework for model developers, ML-extracted data users and other RWD stakeholders. This framework covers the fundamentals of evaluating RWD produced using ML methods to maximize the use of EHR data for research purposes.

Real-world Overall Survival Using Oncology Electronic Health Record Data: Friends of Cancer Research Pilot.

In prior work, Friends of Cancer Research convened multiple data partners to establish standardized definitions for oncology real-world end points derived from electronic health records (EHRs) and claims data. Here, we assessed the performance of real-world overall survival (rwOS) from data sets sourced from EHRs by evaluating the ability of the end point to reflect expected differences from a previous randomized controlled trial across five data sources, after applying inclusion/exclusion criteria. The KEYNOTE-189 clinical trial protocol of platinum doublet chemotherapy (chemotherapy) vs. programmed cell death protein 1 (PD-1) in combination with platinum doublet chemotherapy (PD-1 combination) in first-line nonsquamous metastatic non-small cell lung cancer guided retrospective cohort selection. The Kaplan-Meier product limit estimator was used to calculate 12-month rwOS with 95% confidence intervals (CIs) in each data source. Cox proportional hazards models estimated hazard ratios (HRs) and associated 95% CIs, controlled for prognostic factors. Once the inclusion/exclusion criteria were applied, the five resulting data sets included 155 to 1,501 patients in the chemotherapy cohort and 36 to 405 patients in the PD-1 combination cohort. Twelve-month rwOS ranged from 45% to 58% in the chemotherapy cohort and 44% to 68% in the PD-1 combination cohort. The adjusted HR for death ranged from 0.80 (95% CI: 0.69, 0.93) to 1.15 (95% CI: 0.71, 1.85), controlling for age, gender, performance status, and smoking status. This study yielded insights regarding data capture, including ability of real-world data to precisely identify patient populations and the impact of criteria on end points. Sensitivity analyses could elucidate data set-specific factors that drive results.

The Friends of Cancer Research Real-World Data Collaboration Pilot 2.0: Methodological Recommendations from Oncology Case Studies.

The purpose of this study was to evaluate the potential collective opportunities and challenges of transforming real-world data (RWD) to real-world evidence for clinical effectiveness by focusing on aligning analytic definitions of oncology end points. Patients treated with a qualifying therapy for advanced non-small cell lung cancer in the frontline setting meeting broad eligibility criteria were included to reflect the real-world population. Although a trend toward improved outcomes in patients receiving PD-(L)1 therapy over standard chemotherapy was observed in RWD analyses, the magnitude and consistency of treatment effect was more heterogeneous than previously observed in controlled clinical trials. The study design and analysis process highlighted the identification of pertinent methodological issues and potential innovative approaches that could inform the development of high-quality RWD studies.

ACA Medicaid expansion association with racial disparity reductions in timely cancer treatment.

OBJECTIVES: Racial disparities in cancer care and outcomes remain a societal challenge. Medicaid expansion through the Affordable Care Act was intended to improve health care access and equity. This study aimed to assess whether state Medicaid expansions were associated with a reduction in racial disparities in timely treatment among patients diagnosed with advanced cancer. STUDY DESIGN: This difference-in-differences study analyzed deidentified electronic health record-derived data. Patients aged 18 to 64 years with advanced or metastatic cancers diagnosed between January 1, 2011, and January 31, 2019, and receiving systemic therapy were included. METHODS: The primary end point was receipt of timely treatment, defined as first-line systemic therapy starting within 30 days after diagnosis of advanced or metastatic disease. Racial disparity was defined as adjusted percentage-point (PP) difference for Black vs White patients, adjusted for age, sex, practice setting, cancer type, stage, insurance marketplace, and area unemployment rate, with time and state fixed effects. RESULTS: The study included 30,310 patients (12.3% Black race). Without Medicaid expansion, Black patients were less likely to receive timely treatment than White patients (43.7% vs 48.4%; adjusted difference, -4.8 PP; P < .001). With Medicaid expansion, this disparity was diminished and lost significance (49.7% vs 50.5%; adjusted difference, -0.8 PP; P = .605). The adjusted difference-in-differences estimate was a 3.9 PP reduction in racial disparity (95% CI, 0.1-7.7 PP; P = .045). CONCLUSIONS: Medicaid expansion was associated with reduced Black-White racial disparities in receipt of timely systemic treatment for patients with advanced or metastatic cancers.

Real-world treatment patterns and outcomes of abemaciclib for the treatment of HR+, HER2- metastatic breast cancer.

OBJECTIVE: This retrospective observational study described baseline characteristics, real-world treatment patterns, and outcomes among patients with metastatic breast cancer treated with abemaciclib in the United States. METHODS: De-identified electronic health record-derived data were used to describe patients who began abemaciclib treatment on or after 30 June 2016 and ≥4 months before data cutoff (31 December 2018). Real-world response (rwR) and real-world progression assessments were abstracted from clinical documentation. Descriptive statistics were used to calculate the real-world best response. The Kaplan-Meier method estimated real-world time to first response (rwTTFR) and real-world progression-free survival (rwPFS). RESULTS: The median age of 118 female patients at abemaciclib initiation was 66.5 years (interquartile range, 57.0, 73.0). The breakdown of patients who received abemaciclib in first, second, third, or later lines was 28.8%, 21.2%, 20.3%, and 29.7%, respectively. Patients received abemaciclib as monotherapy (12.7%) or in combination with endocrine therapy: fulvestrant (59.3%); aromatase inhibitor (22.9%); aromatase inhibitor and fulvestrant (5.1%). There were 68 patients (57.6%) with ≥1 rwR assessment: 41.2% with a real-world complete response or real-world partial response. Median rwTTFR was 3.6 months (95% confidence interval, 3.5, 5.2). Twelve-month rwPFS probability was 61.7%. CONCLUSIONS: This study represents utilization and outcomes associated with abemaciclib approximately 1 year following FDA approval. Treatment patterns demonstrated heterogeneity and, as in clinical trials, patients appeared to benefit from abemaciclib treatment in the real world. More research investigating outcomes associated with abemaciclib treatment is needed, with larger samples and longer follow-up to enable closer evaluation by subgroup, regimen, and line of therapy.

First-line immune checkpoint inhibitor use in cisplatin-eligible patients with advanced urothelial carcinoma: a secular trend analysis.

Aim: Standard first-line treatment of advanced urothelial cell carcinoma involves cisplatin-based chemotherapy, with carboplatin or immune checkpoint inhibitor therapy (ICI) reserved for cisplatin-ineligible individuals. Methods: Using a large de-identified electronic health record-derived database of patients with advanced urothelial cell carcinoma in the USA, we examined trends in utilization of first-line systemic therapies in cisplatin-eligible patients from 1 January 2015 to 31 March 2018. Results: Among 1181 cisplatin-eligible patients, the quarterly proportion who received first-line ICI increased from 1 to 42% (ptrend <0.001), while the proportion who received cisplatin-based chemotherapy decreased from 53 to 33% (ptrend = 0.018). Patients receiving ICI were older than those receiving cisplatin (median age: 75 vs 68). Conclusion: Our analysis suggests rising off-label ICI use in cisplatin-eligible individuals, potentially because of ICI's favorable toxicity profile.

Trends in Checkpoint Inhibitor Therapy for Advanced Urothelial Cell Carcinoma at the End of Life: Insights from Real-World Practice.

Several immune checkpoint inhibitor therapies (CPIs) have been approved to treat metastatic urothelial cell carcinoma (mUC). Because of the favorable toxicity profile of CPI compared with chemotherapy, oncologists may have a low threshold to prescribe CPI to patients near the end of life. We evaluated trends in initiation of end-of-life systemic therapy in 1,637 individuals in the Flatiron Health Database who were diagnosed with mUC between 2015 and 2017 and who died. Rates of systemic therapy initiation in the last 30 and 60 days of life were 17.0% and 29.8%, respectively. The quarterly proportion of patients who initiated CPI within 60 days of death increased from 1.0% to 23% during the study period (p trend < .001). After CPI approval, end-of-life CPI initiation significantly increased among patients with poor performance status (p trend = .020) and did not significantly change among individuals with good performance status. The quarterly proportion of patients who initiated any systemic therapy at the end of life doubled (17.4% to 34.8%) during the study period, largely explained by increased CPI use. These findings suggest a dramatic rise in CPI use at the end of life in patients with mUC, a finding that may have important guideline and policy implications.

pAKT Expression and Response to Sorafenib in Differentiated Thyroid Cancer.

Sorafenib has an antitumor activity in patients with radioactive iodine-refractory differentiated thyroid carcinoma (RAIR-DTC). Prior research has implicated signaling through the MAPK and AKT/PI3K pathways in the progression of DTC. To assess whether the activity of these pathways is predictive of response to sorafenib, we retrospectively studied molecular tumor markers from these two pathways from a phase 2 study of sorafenib in RAIR-DTC. Tumor samples from 40 of 53 DTC subjects obtained prior to initiation of sorafenib were immunostained with DAB-labeled antibodies to phospho-AKT (pAKT), phospho-ERK (pERK), and phospho-S6 (pS6). BRAFV600E genetic mutation analysis was performed on all samples. Expression levels and mutational status were compared to response and progression-free survival (PFS) for each patient. Low tumor expression of nuclear pAKT was associated with partial response to sorafenib (p < 0.01). Patients with nuclear pAKT expression that was below the median for our sample were more than three times as likely to have a partial response as patients with equal to or above median expression. There was no correlation between tumor expression of nuclear pERK or pS6 and response. Endothelial cell and pericyte expression of pERK, pAKT, and pS6 were not predictive of response. There was no correlation between BRAFV600E mutation status and partial response. No correlation was observed between either the expression of pAKT, pERK, or pS6, or the presence of the BRAFV600E mutation, and PFS. In conclusion, lower tumor expression of nuclear pAKT was associated with higher rate of response to sorafenib. This observation justifies evaluation of combination therapy with sorafenib and an inhibitor of the PI3K/AKT signaling pathway in RAIR-DTC.

Dietary patterns and self-reported associations of diet with symptoms of inflammatory bowel disease.

BACKGROUND: There are insufficient data to make firm dietary recommendations for patients with inflammatory bowel disease (IBD). Yet patients frequently report that specific food items influence their symptoms. In this study, we describe patients' perceptions about the benefits and harms of selected foods and patients' dietary patterns. METHODS: CCFA Partners is an ongoing internet-based cohort study of patients with IBD. We used a semi-quantitative food frequency questionnaire to measure dietary consumption patterns and open-ended questions to elicit responses from patients about food items they believe ameliorate or exacerbate IBD. We categorized patients into four mutually exclusive disease categories: CD without an ostomy or pouch (CD), UC without an ostomy or pouch (UC), CD with an ostomy (CD-ostomy), and UC with a pouch (UC-pouch). RESULTS: Yogurt, rice, and bananas were more frequently reported to improve symptoms whereas non-leafy vegetables, spicy foods, fruit, nuts, leafy vegetables, fried foods, milk, red meat, soda, popcorn, dairy, alcohol, high-fiber foods, corn, fatty foods, seeds, coffee, and beans were more frequently reported to worsen symptoms. Compared to CD patients, CD-ostomy patients reported significantly greater consumption of cheese (odds ratio [OR] 1.56, 95 % CI 1.03-2.36), sweetened beverages (OR 2.14, 95 % CI 1.02-1.03), milk (OR 1.84, 95 % CI 1.35-2.52), pizza (OR 1.57, 95 % CI 1.12-2.20), and processed meats (OR 1.40; 95 % CI 1.04-1.89). CONCLUSIONS: Patients identified foods that they believe worsen symptoms and restricted their diet. Patients with ostomies ate a more liberal diet. Prospective studies are needed to determine whether diet influences disease course.

Changes in vitamin D and parathyroid hormone metabolism in incident pediatric Crohn's disease.

BACKGROUND: Prior studies of vitamin D metabolism in Crohn's disease (CD) did not include controls or examine changes following diagnosis. This study examined associations among 25-hydroxyvitamin D [25(OH)D], 1,25-dihydroxyvitamin D [1,25(OH)(2)D], and parathyroid hormone (PTH) levels in incident pediatric CD, compared with controls, and following diagnosis. METHODS: Serum vitamin D and PTH were measured at diagnosis (n = 78), 6, 12, and a median of 43 months (n = 52) later in CD participants, and once in 221 controls. Multivariate regression was used to examine baseline associations and quasi-least squares regression to assess subsequent changes. RESULTS: At diagnosis, 42% of CD participants were 25(OH)D-deficient (<20 ng/mL). The odds ratio for deficiency was 2.1 (95% confidence interval [CI]: 1.1, 3.9; P < 0.05) vs. controls, adjusted for age, race, and season. 1,25(OH)(2)D was lower in CD vs. controls (P < 0.05), adjusted for 25(OH)D, tumor necrosis factor alpha (TNF-α), and PTH. TNF-α was associated with lower 1,25(OH)(2)D (P < 0.05), and the positive association between PTH and 1,25(OH)(2)D in controls was absent in CD (interaction P = 0.02). Among participants with 25(OH)D <30 ng/mL, CD was associated with lower PTH (P < 0.05) vs. controls. Following diagnosis, 25(OH)D and 1,25(OH)(2)D improved (P < 0.001). At the final visit, 3% were 25(OH)D-deficient, PTH was no longer low relative to 25(OH)D, and 1,25(OH)(2)D was significantly elevated (P < 0.001) compared with controls. CONCLUSIONS: Incident CD was associated with 25(OH)D and 1,25(OH)2D deficiency and a relative hypoparathyroidism that resolved following diagnosis. Inflammatory cytokine suppression of PTH and renal 1-α-hyroxylase may contribute to these alterations.

Silicone and polyurethane tunneled infusion catheters: a comparison of durability and breakage rates.

PURPOSE: To examine the overall durability and breakage rates of dual-lumen silicone catheters in comparison with power-injectable dual-lumen polyurethane catheters. MATERIALS AND METHODS: Patients who received a 10-F dual-lumen silicone catheter or 9.5-F dual-lumen polyurethane catheter between January 2002 and July 2009 were identified through a quality assurance database. Medical records were reviewed retrospectively. A total of 117 silicone and 94 polyurethane catheters were identified in 192 patients. Reasons for catheter placement and removal were recorded, as were cases of breakage and repairs. Catheter durability was compared; survival analysis was also performed. RESULTS: Breakage occurred in nine of 117 silicone catheters (8%) and none of 94 polyurethane catheters (P = .005). Most catheters were placed for malignancy (162 of 211; 77%); nonmalignant indications such as total parenteral nutrition accounted for 49 out of 211 catheters (23%). The mean silicone catheter dwell time was 99 days (11,612 total catheter-days), and the mean polyurethane catheter dwell time was 78 days (7,362 total catheter-days). There was no significant difference in overall duration of function (ie, survival) between silicone and polyurethane catheters (P = .12). The infection rates were 3.6 per 1,000 catheter-days for silicone catheters and 3.5 per 1,000 catheter-days for polyurethane catheters (P value not significant). CONCLUSIONS: There were fewer catheter fractures with the polyurethane catheter compared with the silicone catheter, although there was no difference in the total access site service interval for the two catheter types.

Frequency and clinical correlations of granulomas in children with Crohn disease.

BACKGROUND: Granulomas are pathognomonic findings of Crohn disease (CD); however, their occurrence and clinical significance are not well characterized. Our aim was to determine the frequency and distribution of granulomas in untreated and treated patients with CD and their relation to age and disease severity. PATIENTS AND METHODS: Records from patients with CD undergoing colonoscopy with terminal ileum biopsy over 7 years were reviewed. Clinical information and laboratory, pathology, and radiology results were recorded. The frequency and distribution of granulomas were determined. RESULTS: From 376 patients with CD, 75% underwent concurrent esophagogastroduodenoscopy and colonoscopy. Of those, 65% (184/282) were untreated. Granulomas were identified in 48% (136/282) of all patients and in 61% (112/184) of untreated patients and 24.5% (24/98) of treated patients (P < 0.0005). The upper tract and terminal ileum biopsies were essential to the identification of 42% of patients with granulomas. The presence of granulomas at diagnosis was related to anti-Saccharomyces cerevisiae antibodies, hypoalbuminemia, perianal disease, and gastritis at presentation (P = 0.03, P = 0.008, P = 0.03, and P = 0.001), respectively, and to perianal disease and infliximab treatment at the latest visit (P = 0.02 and P = 0.01), respectively. Granulomas were not related to age, sex, ethnicity, weight and height z scores, hemoglobin, C-reactive protein, erythrocyte sedimentation rate, CARD15/NOD2 mutations, abdominal surgery, or stricturing or fistulizing disease. CONCLUSIONS: Granulomas were identified in 61% of fully investigated pediatric patients with CD at diagnosis, including a substantial proportion of patients in whom colonoscopy to the cecum would have been insufficient for diagnosis. Granulomas were more frequent in untreated patients (P < 0.0005), and their prevalence was not affected by age. The presence of granulomas at diagnosis was associated with perianal disease, gastritis, hypoalbuminemia, anti-Saccharomyces cerevisiae antibodies, and infliximab treatment.