A matching-based machine learning approach to estimating optimal dynamic treatment regimes with time-to-event outcomes.

Observational data (e.g. electronic health records) has become increasingly important in evidence-based research on dynamic treatment regimes, which tailor treatments over time to patients based on their characteristics and evolving clinical history. It is of great interest for clinicians and statisticians to identify an optimal dynamic treatment regime that can produce the best expected clinical outcome for each individual and thus maximize the treatment benefit over the population. Observational data impose various challenges for using statistical tools to estimate optimal dynamic treatment regimes. Notably, the task becomes more sophisticated when the clinical outcome of primary interest is time-to-event. Here, we propose a matching-based machine learning method to identify the optimal dynamic treatment regime with time-to-event outcomes subject to right-censoring using electronic health record data. In contrast to the established inverse probability weighting-based dynamic treatment regime methods, our proposed approach provides better protection against model misspecification and extreme weights in the context of treatment sequences, effectively addressing a prevalent challenge in the longitudinal analysis of electronic health record data. In simulations, the proposed method demonstrates robust performance across a range of scenarios. In addition, we illustrate the method with an application to estimate optimal dynamic treatment regimes for patients with advanced non-small cell lung cancer using a real-world, nationwide electronic health record database from Flatiron Health.

Keeping a track on leptomeningeal disease in non-small cell lung cancer: A single-institution experience with CNSideTM.

Background: Leptomeningeal disease (LMD) is a devastating complication for patients with advanced cancer. Diagnosis and monitoring the response to therapy remains challenging due to limited sensitivity and specificity of standard-of-care (SOC) diagnostic modalities, including cerebrospinal fluid (CSF) cytology, MRI, and clinical evaluation. These hindrances contribute to the poor survival of LMD patients. CNSide is a CLIA-validated test that detects and characterizes CSF-derived tumor cells and cell-free (cf) DNA. We performed a retrospective analysis on the utility of CNSide to analyze CSF obtained from advanced non-small cell lung cancer (aNSCLC) patients with suspected LMD treated at the Huntsman Cancer Institute in Salt Lake City, UT. Methods: CNSide was used to evaluate CSF from 15 patients with aNSCLC. CSF tumor cell quantification was performed throughout treatment for 5 patients. CSF tumor cells and cfDNA were characterized for actionable mutations. Results: In LMD-positive patients, CNSide detected CSF tumor cells in 88% (22/25) samples versus 40% (10/25) for cytology (matched samples). CSF tumor cell numbers tracked response to therapy in 5 patients where CNSide was used to quantify tumor cells throughout treatment. In 75% (9/12) of the patients, genetic alterations were detected in CSF, with the majority representing gene mutations and amplifications with therapeutic potential. The median survival for LMD patients was 16.1 m (5.2-NR). Conclusions: We show that CNSide can supplement the management of LMD in conjunction with SOC methods for the diagnosis, monitoring response to therapy, and identifying actionable mutations unique to the CSF in patients with LMD.

Real-world comparison of survival outcomes with cisplatin versus carboplatin in patients with limited-stage small-cell lung cancer.

INTRODUCTION: Limited-stage small-cell lung cancer (LS-SCLC) is potentially curable with concurrent chemoradiation (CRT). Cisplatin is the preferred platinum for the chemotherapy backbone in national guidelines. Unfortunately, many LS-SCLC patients are elderly, with comorbidities and poor performance status (PS), which preclude the use of cisplatin. Carboplatin may be a suitable alternative. This analysis evaluates the overall survival (OS) and time to next treatment (TTNT) in LS-SCLC patients receiving concurrent CRT by platinum use. MATERIALS AND METHODS: The study included LS-SCLC patients in the Flatiron Health nationwide de-identified electronic health record-derived database who received CRT in 2013-2019 with follow-up through May 2020. TTNT and OS were compared using both unadjusted and inverse propensity-weighted Cox proportional hazards models. RESULTS: This study included patients treated with carboplatin (n = 600) or cisplatin (n = 572) in combination with etoposide and radiation. Cisplatin patients were younger, had a shorter time from diagnosis to radiation, and had less kidney disease. In an unadjusted analysis, median overall survival (mOS) was greater in the cisplatin group than the carboplatin group with mOS of 22.3 months vs. 19.2 months and Hazard Ratio (HR) of 0.83 (p = 0.01). In the inverse propensity-weighted analysis, this difference was no longer significant (HR 0.93, p = 0.37). No differences were seen in TTNT. CONCLUSION: When balancing on key clinical factors, we observed no statistical difference in OS or TTNT by platinum choice in real-world LS-SCLC patients treated with CRT.  Although observational, the results from this large data set are consistent with the hypothesis that either cisplatin or carboplatin is an appropriate therapy regardless of health status.

Survival outcomes for patients with T3N0M0 squamous cell carcinoma of the glottis treated with definitive radiation alone versus chemoradiation.

BACKGROUND: Given the poor lymphatics of the glottis, we evaluated omission of chemotherapy in patients treated definitely for T3N0M0 squamous cell carcinoma (SCC) of the glottis. METHODS: We performed survival analysis of patients with T3N0M0 SCC of the glottis identified in the National Cancer Database treated with radiation alone versus chemoradiation. RESULTS: A total of 3785 patients were identified. Patients age ≥70 and those with comorbidities were less likely to receive chemotherapy (odds ratio [OR] 0.30, 95% CI [0.25-0.37] and 0.48 [0.31-0.76], respectively). Five-year OS was lower in patients treated with radiation versus chemoradiation (33.8% [30.3%-37.2%] vs. 58.0% [55.8%-60.0%]). In patients <70 with no comorbidities this difference persisted (51.0% [44.5%-57.0%] versus 66.7% [64.0%-69.3%]). CONCLUSION: Overall survival was higher in patients treated with chemoradiation compared to radiation alone, even when controlling for age and comorbidities. Radiotherapy with chemotherapy omission is not appropriate in patients with T3N0M0 SCC of the glottis.

A Real-World Analysis of the Use of Systemic Therapy in Malignant Pleural Mesothelioma and the Differential Impacts on Overall Survival by Practice Pattern.

INTRODUCTION: Malignant pleural mesothelioma (MPM) is an aggressive malignancy that affects older adults with frequent comorbidities, making real-world treatment decisions challenging. This study compares the overall survival (OS) of patients with MPM by physician's choice of first-line (1L) platinum chemotherapy (PC), second-line (2L) immunotherapy versus chemotherapy, and by receipt of maintenance therapy (MT). METHODS: The study included patients diagnosed with advanced MPM in the Flatiron Health electronic health record-derived database who initiated PC with pemetrexed in the 1L setting between 2011 and 2019. Patients in the 2L therapy analysis received single-agent chemotherapy versus immunotherapy after the progression of disease from our 1L cohort. Patients in the MT cohort were identified on the basis of continued receipt of pemetrexed with or without bevacizumab after dropping PC at prespecified intervals. The OS of patients by choice of 1L PC, 2L immunotherapy versus chemotherapy, and receipt of MT was summarized by means of Kaplan-Meier survival estimates and compared in the context of propensity score matching weighted analyses. RESULTS: In propensity score matching weighting analysis from 2065 patients with MPM, there was no evidence of an OS difference by choice of 1L PC (hazard ratio [HR] = 1.08, 95% confidence interval [CI]: 0.89-1.31, p = 0.43), suggestive evidence of an OS difference by choice of 2L immunotherapy versus chemotherapy (HR = 0.68, 95% CI: 0.42-1.08; p = 0.10), and no evidence of an OS difference by receipt of MT (HR = 0.92, 95% CI: 0.72-1.16, p = 0.46). CONCLUSIONS: Using real-world, propensity score-matched weighted analysis of MPM, we found there was no difference in OS by choice of 1L PC, 2L immunotherapy or chemotherapy, or by receipt of MT.

Dynamic Prediction of Near-Term Overall Survival in Patients with Advanced NSCLC Based on Real-World Data.

Patients with terminal cancers commonly receive aggressive and sub-optimal treatment near the end of life, which may not be beneficial in terms of duration or quality of life. To improve end-of-life care, it is essential to develop methods that can accurately predict the short-term risk of death. However, most prediction models for patients with cancer are static in the sense that they only use patient features at a fixed time. We proposed a dynamic prediction model (DPM) that can incorporate time-dependent predictors. We apply this method to patients with advanced non-small-cell lung cancer from a real-world database. Inverse probability of censoring weighted AUC with bootstrap inference was used to compare predictions among models. We found that increasing ECOG performance status and decreasing albumin had negative prognostic associations with overall survival (OS). Moreover, the negative prognostic implications strengthened over the patient disease course. DPMs using both time-independent and time-dependent predictors substantially improved short-term prediction accuracy compared to Cox models using only predictors at a fixed time. The proposed model can be broadly applied for prediction based on longitudinal data, including an estimation of the dynamic effects of time-dependent features on OS and updating predictions at any follow-up time.

Role of CTLA Inhibition in Management of Non-Small Cell Lung Cancer.

PURPOSE OF REVIEW: The use of single-agent or combination immunotherapy strategies has revolutionized the management of patients with non-small cell lung cancer (NSCLC). Here, we review the current role for CTLA-4 inhibitors in early-stage resectable NSCLC, unresectable stage III NSCLC, and in metastatic NSCLC. RECENT FINDINGS: Immunotherapy agents alone, or in combination with chemotherapy, represent the new standard of care for the management of metastatic squamous and non-squamous NSCLC without driver mutations. Combination CTLA-4 and PD-1/L1 inhibitors can be efficacious, particularly in tumor mutation burden (TMB) high tumors, providing a chemotherapy-free strategy for metastatic patients. Early signals from neoadjuvant trials suggest a benefit for combination CTLA-4 and PD-1 inhibitions prior to surgery, with improved rates of major pathologic response (MPR). The role for CTLA4 inhibitors is currently unknown in the adjuvant and unresectable stage III setting, although clinical trials are ongoing to evaluate this approach. There is a growing role for CTLA-4 inhibition in the neoadjuvant and metastatic settings for patients with NSCLC. Biomarker selection in ongoing clinical trials will be crucial to guide patient selection for CTLA4 inhibitor therapy. Combination strategies with PD-1/L1 inhibition have demonstrated the greatest efficacy to date.

Adoption of Patient-Generated Health Data in Oncology: A Report From the NCCN EHR Oncology Advisory Group.

BACKGROUND: Collecting, monitoring, and responding to patient-generated health data (PGHD) are associated with improved quality of life and patient satisfaction, and possibly with improved patient survival in oncology. However, the current state of adoption, types of PGHD collected, and degree of integration into electronic health records (EHRs) is unknown. METHODS: The NCCN EHR Oncology Advisory Group formed a Patient-Reported Outcomes (PRO) Workgroup to perform an assessment and provide recommendations for cancer centers, researchers, and EHR vendors to advance the collection and use of PGHD in oncology. The issues were evaluated via a survey of NCCN Member Institutions. Questions were designed to assess the current state of PGHD collection, including how, what, and where PGHD are collected. Additionally, detailed questions about governance and data integration into EHRs were asked. RESULTS: Of 28 Member Institutions surveyed, 23 responded. The collection and use of PGHD is widespread among NCCN Members Institutions (96%). Most centers (90%) embed at least some PGHD into the EHR, although challenges remain, as evidenced by 88% of respondents reporting the use of instruments not integrated. Forty-seven percent of respondents are leveraging PGHD for process automation and adherence to best evidence. Content type and integration touchpoints vary among the members, as well as governance maturity. CONCLUSIONS: The reported variability regarding PGHD suggests that it may not yet have reached its full potential for oncology care delivery. As the adoption of PGHD in oncology continues to expand, opportunities exist to enhance their utility. Among the recommendations for cancer centers is establishment of a governance process that includes patients. Researchers should consider determining which PGHD instruments confer the highest value. It is recommended that EHR vendors collaborate with cancer centers to develop solutions for the collection, interpretation, visualization, and use of PGHD.

Comparing adjuvant radiation to adjuvant chemoradiation in postsurgical p16+ oropharyngeal carcinoma patients with extranodal extension or positive margins.

BACKGROUND: Adjuvant guidelines in surgically resected p16+ oropharyngeal carcinoma (OPC) with positive surgical margins (PSM) or extranodal extension (ENE) are based on randomized controlled trials predating p16 status. It remains unclear if adjuvant chemotherapy is necessary in p16+ patients with these features. METHODS: The National Cancer Database was used to identify cases of nonmetastatic p16+ OPC diagnosed from 2010 to 2017. Patients treated with surgical resection followed by adjuvant radiation (aRT) or adjuvant chemoradiation (aCRT) were eligible for analysis. RESULTS: A total of 14 071 patients were eligible for analysis. Overall survival (OS) was not statistically different between aRT and aCRT in patients with PSM (hazard ratio (HR) 0.84, 95% confidence interval (CI) 0.56-1.28), ENE (HR 0.93, 95% CI 0.69-1.27) or both (HR 0.73, 95% CI 0.41-1.31). CONCLUSIONS: In patients with p16+ OPC with ENE, PSM, or both, adding chemotherapy to aRT was not associated with improved OS.

The role of gut microbiome in modulating response to immune checkpoint inhibitor therapy in cancer.

Immunotherapy has led to a paradigm shift in the treatment of several cancers. There have been significant efforts to identify biomarkers that can predict response and toxicities related to immune checkpoint inhibitor (ICPI) therapy. Despite these advances, it has been challenging to tease out why a subset of patients benefit more than others or why certain patients experience immune-related adverse events (irAEs). Although the immune-modulating properties of the human gut bacterial ecosystem are yet to be fully elucidated, there has been growing interest in evaluating the role of the gut microbiome in shaping the therapeutic response to cancer immunotherapy. Considerable research efforts are currently directed to utilizing metagenomic and metabolic profiling of stool microbiota in patients on ICPI-based therapies. Dysbiosis or loss of microbial diversity has been associated with a poor treatment response to ICPIs and worse survival outcomes in cancer patients. Emerging data have shown that certain bacterial strains, such as Faecalibacterium that confer sensitivity to ICPI, also have a higher propensity to increase the risk of irAEs. Additionally, the microbiome can modulate the local immune response at the intestinal interface and influence the trafficking of bacterial peptide primed T-cells distally, influencing the toxicity patterns to ICPI. Antibiotic or diet induced alterations in composition of the microbiome can also indirectly alter the production of certain bacterial metabolites such as deoxycholate and short chain fatty acids that can influence the anti-tumor tolerogenesis. Gaining sufficient understanding of the exact mechanisms underpinning the interplay between ICPI induced anti-tumor immunity and the immune modulatory role gut microbiome can be vital in identifying potential avenues of improving outcomes to cancer immunotherapy. In the current review, we have summarized and highlighted the key emerging data supporting the role of gut microbiome in regulating response to ICPIs in cancer.

Real World Characterization of Advanced Non-Small Cell Lung Cancer in Never Smokers by Actionable Mutation Status.

BACKGROUND: Non-small cell lung cancer (NSCLC) in never-smokers (NS) is vastly different from those with a smoking history in terms of etiology, driver mutations, and immunotherapy responsiveness. This study compares the real-world overall survival (OS) of NSCLC patients by smoking history and mutation status. METHODS: The study included 30,310 advanced or metastatic NSCLC patients in the Flatiron Health EHR-derived database who received biomarker testing results (EGFR, ALK, ROS1, and BRAF), and initiated therapy between 2011 and 2017, with follow up through June 2018. OS by smoking and driver mutation groups was summarized via Kaplan-Meier survival estimates, and compared in the context of a multivariate Cox proportional hazard model. RESULTS: OS differed by smoking and driver-mutation categories (adjusted and stratified P< .001). The median OS for wild-type (WT) smoking patients was 9.6 months, for mutated (MT) smokers was 19.4 months (adjusted and stratified hazard ratio [HR] relative to WT smokers 0.65; 95% CI 0.60-0.71), for WT NS was 15.1 months (HR 0.78; 95% CI 0.73-0.83 relative to WT smokers), and for MT NS was 25.5 months (HR 0.52; 95% CI 0.48-0.58 relative to WT smokers). CONCLUSION: NS with NSCLC survived longer than those with smoking history, in both groups of WT and mutation-positive patients. Findings highlight that in NSCLC patients, a history of never smoking may have similar effect on hazard of death as the presence of an actionable mutation. Taken together, differences in heredity, mutations, and biologic history suggest that NS lung cancer is a distinct clinical entity and must be managed accordingly.

Management of Postabortion Complications for the Emergency Medicine Clinician.

Although induced abortion is generally a safe outpatient procedure, many patients subsequently present to the emergency department, concerned about a postabortion complication. It is helpful for emergency physicians to understand the medications and procedures used in abortion care in the United States to effectively and efficiently triage and treat women presenting with potential complications from an abortion. Furthermore, because many states are experiencing increased abortion restrictions that limit access to care, emergency medicine physicians may encounter more patients presenting after self-managed abortions, which presents additional challenges. This article reviews the epidemiology and background of abortion care, including the range of symptoms and adverse effects that are within the scope of an uncomplicated procedure. This review also offers a comprehensive overview of management of abortion complications, including algorithms for more common complications and descriptions of less common but more severe adverse events. The article concludes with a recognition of the social stigma and legal regulations unique to abortion care.

Prognostic Significance of Patient-Reported Outcomes in Cancer.

PURPOSE: Performance status (PS), an established prognostic surrogate of cancer survival, is a physician-synthesized metric of patient symptoms and mobility that is prone to bias and subjectivity. The National Cancer Institute (NCI) Patient-Reported Outcomes Measurement Information System-Cancer (PROMIS-Ca) Bank, a patient-centric patient-reported outcome (PRO) evaluation of physical function (PF), fatigue, depression, anxiety, and pain, shares subject matter with PS and, therefore, may also be prognostic while eliminating physician interpretation. METHODS: Patients at Huntsman Cancer Institute were assessed using the NCI PROMIS-Ca Bank. Using tablets at routine office visits, PF, fatigue, depression, anxiety, and pain scores were collected from patients with advanced melanoma, non-small-cell lung cancer, colorectal cancer, and breast cancer. A PRO score collected at a single time point within 6 months of metastatic diagnosis for each patient was merged with curated clinical outcome data. The association of PROs, overall survival (OS), and hospitalization-free survival (HFS) were assessed in multivariable analysis that included sex and cancer type. RESULTS: Two hundred eighty-two complete sets of patient data were available for analysis. All 5 PRO domains were strongly prognostic of OS and HFS. While the PRO domains were interrelated with moderate to strong correlations (0.40-0.79), multivariable regression suggested that PF was most strongly associated with the clinical outcomes of OS (P < .001) and HFS (P < .001). CONCLUSION: NCI PROMIS-Ca PROs may be prognostic of both cancer survival and likelihood of hospitalization. Future prospective studies are needed for all major prognostic factors to fully understand the independent prognostic value of PROs.

Treatment of Lung Cancer Patients With Actionable Mutations in the Intensive Care Unit.

BACKGROUND: Patients with advanced-stage non-small-cell lung cancer (NSCLC) have high mortality rates in the intensive care unit (ICU). Although the benefit of chemotherapy for hematologic malignancies in the ICU has previously been explored, few data exist regarding the use of targeted therapy for NSCLC in such settings. The primary objective of the present study was to report our experience with the use of targeted therapy in patients with NSCLC in the ICU. MATERIALS AND METHODS: We performed a single-institution, retrospective medical record review. The eligibility criteria included patients with NSCLC with targetable mutations who had received tyrosine kinase inhibitors (TKIs) in the ICU. Cases were identified by queries of our institution's information warehouse database and pharmacy dispensary records from 2010 to 2015. RESULTS: All 9 patients who had received TKIs in the ICU had acute respiratory failure. Three patients were successfully extubated after initiating TKI therapy, although 1 required later tracheostomy. TKI therapy stabilized another patient's refractory disseminated intravascular coagulation. The remaining 5 patients showed no measurable clinical improvement and were transitioned to comfort care. The overall ICU mortality rate was 56%. CONCLUSION: Patients with metastatic NSCLC requiring mechanical ventilation have high mortality rates. Cytotoxic chemotherapy is generally contraindicated for poor performance status patients. However, targeted TKI therapy should be considered, given its proven efficacy and few systemic side effects. We recommend the empiric use of targeted therapy for NSCLC patients with suspected and/or known actionable mutations presenting with multifactorial respiratory failure to the ICU, with aggressive determination of the mutation status if not known.