Variation in US acute ischemic stroke treatment by hospital regions: limited endovascular access despite evidence

BACKGROUND

Although national organizations recognize the importance of regionalized acute ischemic stroke (AIS) care, data informing expansion are sparse. We assessed real-world regional variation in emergent AIS treatment, including growth in revascularization therapies and stroke center certification. We hypothesized that we would observe overall growth in revascularization therapy utilization, but observed differences would vary greatly regionally.

METHODS

A retrospective cross-sectional analysis was carried out of de-identified national inpatient Medicare Fee-for-Service datasets from 2016 to 2019. We identified AIS admissions and treatment with thrombolysis and endovascular thrombectomy (ET) with International Classification of Diseases, 10th Revision, Clinical Modification codes. We grouped hospitals in Dartmouth Atlas of Healthcare Hospital Referral Regions (HRR) and calculated hospital, demographic, and acute stroke treatment characteristics for each HRR. We calculated the percent of hospitals with stroke certification and AIS cases treated with thrombolysis or ET per HRR.

RESULTS

There were 957 958 AIS admissions. Relative mean (SD) growth in percent of AIS admissions receiving revascularization therapy per HRR from 2016 to 2019 was 13.4 (31.7)% (IQR -6.1-31.7%) for thrombolysis and 28.0 (72.0)% (IQR 0-56.0%) for ET. The proportion of HRRs with decreased or no difference in ET utilization was 38.9% and the proportion of HRRs with decreased or no difference in thrombolysis utilization was 32.7%. Mean (SD) stroke center certification proportion across HRRs was 45.3 (31.5)% and this varied widely (IQR 18.3-73.4%).

CONCLUSIONS

Overall growth in AIS treatment has been modest and, within HRRs, growth in AIS treatment and the proportion of centers with stroke certification varies dramatically.

Case report: A novel biallelic FTO variant causing multisystem anomalies with severe epilepsy, widening the spectrum of FTO syndrome

The fat mass and obesity-associated gene (FTO) codes for a DNA/RNA demethylase. Pathological variants in this gene are rare, with only three reports in the literature, all with mutations in the catalytic domain. We report the first biallelic human variant in fat mass and obesity-associated gene (c.287G>C, p.Arg96Pro/R96P) outside the catalytic site, causing numerous abnormalities across multiple organ systems, affecting respiratory, cardiovascular, and neurological function. Biochemical assays of cells with the patient's variant were performed to further quantify the effect of the variant on function. Loss-of-function resulting from the patient's R96P missense variant was demonstrated with in vitro biochemical characterization of demethylase activity, resulting in a 90% reduction in function of the fat mass and obesity-associated protein compared to wild-type. Our findings demonstrate a novel fat mass and obesity-associated gene non-catalytic site variant with a unique patient phenotype of bilateral multifocal epilepsy and multisystem congenital anomalies.

Risk, Determinants, and Pharmacologic Treatment of Depression Following Acute Ischemic Stroke

Background and Purpose: We assessed risk and determinants of new-onset depression in acute ischemic stroke (AIS) patients of all ages and no known history of depression. Additionally, we assessed patterns of post-stroke depression (PSD) treatment with pharmacotherapy. Methods: Retrospective cohort study of de-identified Marketscan Commercial Claims and Encounters and Medicare Supplemental and Coordination of Benefits Datasets for adults age ≥18 years admitted with AIS from July 1, 2016-July 1, 2017. We created Kaplan-Meier curves of cumulative risk of PSD up to 1.5 years following index AIS admission. We performed Cox regression to report hazard ratios for determinants of PSD up to 1.5 years following AIS. We summarized proportions treated with pharmacotherapy and identified the most commonly prescribed medications. Results: Of 8089 AIS patients, 1059 were diagnosed with PSD. At 1 year, cumulative risk of PSD was 13.4% (standard error .4) and 15.3% (standard error .5) at 1.5 years. History of anxiety was most strongly associated with PSD and discharge home least. Among those with PSD, 68.8% were prescribed an antidepressant and 8.4% an antipsychotic. The most commonly prescribed antidepressant was sertraline (28.5%). Conclusions: Among AIS patients of all ages, there is a persistently elevated cumulative risk of new diagnosis of PSD in the 1.5 years following AIS. Of the >2/3 treated with an antidepressant, sertraline was most commonly prescribed. Screening and treatment strategies for PSD require further study.

Risk and Predictors of Depression Following Acute Ischemic Stroke in the Elderly

OBJECTIVE

We sought to comprehensively evaluate predictors of poststroke depression (PSD) in the United States and to compare PSD to post-myocardial infarction (MI) depression to determine whether ischemic stroke uniquely elevates risk of depression.

METHODS

This is a retrospective cohort study of 100% deidentified inpatient, outpatient, and subacute nursing Medicare data from 2016 to 2017 for US patients ≥65 years of age from July 1, 2016, to December 31, 2017. We calculated Kaplan-Meier unadjusted cumulative risk of depression up to 1.5 years after the index admission. We performed Cox regression to report the hazard ratio for diagnosis of depression up to 1.5 years after stroke vs MI and independent predictors of PSD, and we controlled for patient demographics, comorbid conditions, length of stay, and acute stroke interventions.

RESULTS

In fully adjusted models, patients with stroke (n = 174,901) were ≈50% more likely than patients with MI (n = 193,418) to develop depression during the 1.5-year follow-up period (Kaplan-Meier cumulative risk 0.1596 ± 0.001 in patients with stroke vs 0.0973 ± 0.000778 in patients with MI, log-rank p < 0.0001). History of anxiety was the strongest predictor of PSD, while discharge home was most protective. Female patients, White patients, and patients <75 years of age were more likely to be diagnosed with depression after stroke.

CONCLUSIONS

Despite the similarities between MI and stroke, patients with stroke were significantly more likely to develop depression. There were several predictors of PSD, most significantly history of anxiety. Our findings lend credibility to a stroke-specific process causing depression and highlight the need for consistent depression screening in all patients with stroke.

Abstract 22: Risk and Predictors of Depression Following Acute Ischemic Stroke in the Elderly: Comparison With Acute Myocardial Infarction

Introduction: Post-stroke depression (PSD) occurs commonly following stroke and is associated with worse outcomes and higher mortality. Previous research has not identified consistent predictors of PSD, and debate remains about whether PSD differs from other types of depression, including depression following other ischemic vascular events.

Objective: We sought to comprehensively evaluate predictors of PSD in the US population and compare the hazard of developing PSD to post-myocardial infarction (MI) depression.

Methods: Retrospective cohort study of 100% de-identified inpatient, outpatient, and subacute nursing Medicare data from 2016-2017 for US patients aged ≥65 years from July 1, 2016 to December 31, 2017. We calculated Kaplan-Meier unadjusted cumulative risk of depression up to 1.5 years following index admission. We performed Cox regression to report the hazard ratio for diagnosis of depression up to 1.5 years post-stroke vs. MI, as well as independent predictors of PSD, and controlled for patient demographics, comorbidities, length of stay and acute stroke interventions.

Results: In fully adjusted models, stroke patients (n=174,901) were approximately 50% more likely than MI patients (n=193,418) to develop depression during the 1.5-year follow-up period (Kaplan-Meier cumulative risk 0.1596 ± 0.001 in stroke patients versus 0.0973 ± 0.000778 in MI patients, log-rank p<0.0001). History of anxiety was the strongest predictor of PSD, while discharge home was most protective. Female patients, White patients, and patients younger than 75 years were more likely to be diagnosed with depression post-stroke.

Conclusions: Despite the similarities between MI and stroke, patients who suffer from stroke were significantly more likely to develop depression. There were several predictors of post-stroke depression, most significantly history of anxiety. Our findings lend credibility to a stroke-specific process causing depression and highlight the need for consistent depression screening in all stroke patients.

Abstract 21: Sex Differences in Post-Stroke Depression in the Elderly

Introduction: Post-stroke depression (PSD) occurs in approximately one-third of ischemic stroke patients. However, there is conflicting evidence on sex differences in PSD.

Objective: We sought to assess sex differences in risk and time course of PSD in US ischemic stroke (IS) patients. We hypothesized that women are at greater risk of PSD than men, and that a greater proportion of women experience PSD in the acute post-stroke phase.

Methods: Retrospective cohort study of 100% de-identified data for US Medicare beneficiaries ≥65 years admitted for ischemic stroke from July 1, 2016 to December 31, 2017. We calculated Kaplan-Meier unadjusted cumulative risk of depression, stratified by sex, up to 1.5 years following index admission. We performed Cox regression to report the hazard ratio (HR) for diagnosis of depression up to 1.5 years post-stroke in males vs. females, adjusting for patient demographics, comorbidities, length of stay, and acute stroke interventions.

Results: Female stroke patients (n=90,474) were 20% more likely to develop PSD than males (n=84,427) in adjusted models. Cumulative risk of depression was consistently elevated for females throughout 1.5 years of follow-up (0.2055 [95% CI 0.2013-0.2097] vs. 0.1690 [95% CI 0.1639-0.1741] (log-rank p<0.0001). HR for PSD in females vs. males remained significant in fully adjusted analysis at 1.20 (95% CI 1.17-1.23, p<0.0001).

Conclusions: Over 1.5 years of follow-up, female stroke patients had significantly greater hazard of developing PSD, highlighting the need for long-term depression screening in this population and further investigation of underlying reasons for sex differences.

A High-Throughput Immune-Oncology Screen Identifies EGFR Inhibitors as Potent Enhancers of Antigen-Specific Cytotoxic T-lymphocyte Tumor Cell Killing

We developed a screening assay in which luciferized ID8 expressing OVA was cocultured with transgenic CD8+ T cells specifically recognizing the model antigen in an H-2b-restricted manner. The assay was screened with a small-molecule library to identify compounds that inhibit or enhance T cell-mediated killing of tumor cells. Erlotinib, an EGFR inhibitor, was the top compound that enhanced T-cell killing of tumor cells. Subsequent experiments with erlotinib and additional EGFR inhibitors validated the screen results. EGFR inhibitors increased both basal and IFNγ-induced MHC class-I presentation, which enhanced recognition and lysis of tumor cell targets by CD8+ cytotoxic T lymphocytes. The ID8 cell line was also transduced to constitutively express Cas9, and a pooled CRISPR screen, utilizing the same target tumor cell/T-cell assay, identified single-guide (sg)RNAs targeting EGFR that sensitized tumor cells to T cell-mediated killing. Combination of PD-1 blockade with EGFR inhibition showed significant synergistic efficacy in a syngeneic model, further validating EGFR inhibitors as immunomodulatory agents that enhance checkpoint blockade. This assay can be screened in high-throughput with small-molecule libraries and genome-wide CRISPR/Cas9 libraries to identify both compounds and target genes, respectively, that enhance or inhibit T-cell recognition and killing of tumor cells. Retrospective analyses of squamous-cell head and neck cancer (SCCHN) patients treated with the combination of afatinib and pembrolizumab demonstrated a rate of clinical activity exceeding that of each single agent. Prospective clinical trials evaluating the combination of an EGFR inhibitor and PD-1 blockade should be conducted.

Abstract 4935: High-throughput immune-oncology screen identifies EGFR inhibitors as potent enhancers of CTL antigen-specific tumor cell killing

As immune checkpoint blocking antibodies increasing become foundational therapies for the treatment of cancer, there is a pressing need to identify compounds that synergize with checkpoint blockade as the basis of combinatorial treatment regimens. We have developed a screening assay in which a luciferized tumor cell line expressing a model antigen is co-cultured with a transgenic CD8+ T cell specifically recognizing the model antigen in a H-2b-restricted manner. The target tumor cell/T cell assay was screened with a small molecule library to identify compounds that inhibit or enhance T cell-mediated killing of tumor cells in an antigen-dependent manner. The EGFR inhibitor Erlotinib was the top hit that enhanced T cell killing of tumor cells. Subsequent experiments with Erlotinib and additional EGFR inhibitors validated the screen result. EGFR inhibitors increase both basal and IFN-γ-induced antigen processing and presentation of MHC class-I, which enhanced recognition and lysis by CD8+ cytotoxic T lymphocytes. The tumor cell line was also transduced to constitutively express Cas9, and a pooled CRISPR screen utilizing the same target tumor cell/T cell assay identified sgRNAs targeting EGFR as sensitizing tumor cells to T cell-mediated killing. Combination of PD-1 blockade with EGFR inhibition showed significant synergistic efficacy in the MC38 syngeneic colon cancer model that was superior to PD-1 blockade or EGFR inhibition alone, further validating EGFR inhibitors as immunomodulatory agents that enhance PD-1 checkpoint blockade. This novel target tumor cell/T cell assay can be screened in high-throughput with small molecule libraries and genome-wide CRISPR/Cas9 libraries to identify both compounds AND target genes, respectively, that enhance or inhibit T cell recognition and killing of tumor cells.

Citation Format: Patrick H. Lizotte, Troy Luster, Megan E. Cavanaugh, Luke J. Taus, Abha Dhaneshwar, Naomi Mayman, Aaron Yang, Mark Bittinger, Paul Kirschmeier, Nathanael S. Gray, David A. Barbie, Pasi A. Janne. High-throughput immune-oncology screen identifies EGFR inhibitors as potent enhancers of CTL antigen-specific tumor cell killing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 4935.