Citations

Background: The focus of most existing phenotyping algorithms based on electronic health record (EHR) data has been to accurately identify cases and non-cases of specific diseases. However, a more challenging task is to accurately identify disease incidence, as identifying the first occurrence of disease is more important for efficient and valid clinical and epidemiological research. Moreover, stroke is a challenging phenotype due to diagnosis difficulty and common miscoding. This task generally requires utilization of multiple types of EHR data (e.g., diagnoses and procedure codes, unstructured clinical notes) and a more robust algorithm integrating both natural language processing and machine learning. In this study, we developed and validated an EHR-based classifier to accurately identify stroke incidence among a cohort of atrial fibrillation (AF) patients Methods: We developed a stroke phenotyping algorithm using International Classification of Diseases, Ninth Revision (ICD-9) codes, Current Procedural Terminology (CPT) codes, and expert-provided keywords as model features. Structured data was extracted from Rochester Epidemiology Project (REP) database. Natural Language Processing (NLP) was used to extract and validate keyword occurrence in clinical notes. A window of ±30 days was considered when including/excluding keywords/codes into the input vector. Frequencies of keywords/codes were used as input feature sets for model training. Multiple competing models were trained using various combinations of feature sets and two machine learning algorithms: logistic regression and random forest. Training data were provided by two nurse abstractors and included validated stroke incidences from a previously established atrial fibrillation cohort. Precision, recall, and F-score of the algorithm were calculated to assess and compare model performances. Results: Among 4,914 patients with atrial fibrillation, 1,773 patients were screened. 3,141 patients had no stroke-related codes or keywords and were presumed to be free of stroke during follow-up. Among the screened patients, 740 had validated strokes and 1,033 did not have a stroke based on review of the EHR by trained nurse abstractors. The best performing stroke incidence phenotyping classifier utilized Keywords+ICD-9+CPT features using a random forest classifier, achieving a precision of 0.942, recall of 0.943, and F-score of 0.943. Conclusion: In conclusion, we developed and validated a stroke algorithm that performed well for identifying stroke incidence in an enriched population (AF cohort), which extends beyond the typical binary case/non-case stroke identification problem. Future work will involve testing the generalizability of this algorithm in a general population.

Abstract Background While motility disorders may evolve over time, there is scant guidance around the role of repeat high-resolution esophageal manometry (HRM). Given the invasive nature of HRM and the implications on financial cost and patient discomfort, it is obvious that the emphasis should be on minimizing unnecessary repeat examinations. However, there are no recommendations around indications or timing of repeat HRM. Aims We aimed to determine the outcomes in patients who underwent repeat manometry and look for predictors of progression to achalasia or major motility disorder. Methods Consecutive reports from HRM studies performed between Aug 2013 – May 2017 were retrospectively analyzed. All patients with ≥ 2 HRM studies were included. Studies without a Chicago classification diagnosis were excluded. Chi-squared analysis was performed to determine if initial HRM diagnosis was associated with change in diagnosis on follow-up HRM. Initial and follow-up manometric parameters were compared with paired T-tests. Binary logistic regression analysis was performed to look for predictors of progression to achalasia or major motility disorder. Results 134 patients underwent ≥ 2 HRM studies. Initial diagnoses were IEM (45 patients [33.6%], EGJOO (34 [25.4%], absent peristalsis (18 [13.4%], achalasia (11 [8.2%], DES (4 [3.0%]), and JH (3 [2.2%]; 29 (14.2%) of patients had a normal HRM. 109 (81.3%) patients underwent 2 HRM, 18 (13.4%) 3 HRM, 4 (3%) 4 HRM, and 3 (2.2%) 5 HRM. The final follow-up HRM occurred after a median 496 [80 – 1823] days. 72 (53.7%) of patients had no change from their initial diagnosis. Patients with an initial diagnosis of DES were significantly more likely to have a change in diagnosis on the final follow-up (3 normal:1 IEM) (p = .043). No other classes reached significance. Patients with IEM had a significantly higher mean DCI (395.1 [0 - 3248] vs 790.8 [0 – 10715.0], p = .006) and IRP (4.5 [-10.4 – 14.2] vs [6.6 [-6.2 – 21.0], p = .017) on their follow-up HRM. 4 patients without achalasia (3 EGJOO:1 IEM) on their index HRM had a diagnosis of achalasia on their final HRM. The median IRP in non-achalasia patients with a diagnosis of achalasia on final HRM (22.3 [8.4 – 30.7] was significantly higher than those without a diagnosis of achalasia on final HRM (6.6 [-10.4 – 39.8]) (p = .013); however no manometric criteria or initial HRM diagnoses predicted progression to achalasia or major motility disorder on binary logistic regression analysis. Conclusions In most patients, repeat manometry did not change the manometric diagnosis. Patients with DES were significantly likely to have their diagnosis change with repeat HRM, and most of these patients had normalization of their HRM. Manometric parameters in IEM appear to improve over time. This finding could reflect interval therapy, or shed some light on the natural history of this disease. Funding Agencies None

Background Encapsulating peritoneal sclerosis (EPS) is a rare complication of peritoneal dialysis. The causes of EPS are not well defined and are likely multifactorial. A suitable animal model would facilitate research into the pathophysiology and treatment of EPS. Methods We developed a helper-dependent adenovirus that expresses both green fluorescent protein (GFP) and active transforming growth factor-beta (TGF-β1; HDAdTGF-β1). Mice were administered HDAdTGF-β1 via intraperitoneal injection and the response was compared with mice administered either first-generation adenovirus expressing TGF-β1 (AdTGF-β1) or control adenovirus (AdGFP). Results HDAdTGF-β1-treated mice continued to express the GFP reporter transgene to day 74, the end of the observation period. Transgene expression lasted less than 28 days in the animals treated with first-generation adenoviruses. Animals treated with first-generation AdTGF-β1 demonstrated submesothelial thickening and angiogenesis at day 7, with almost complete resolution by day 28. The HDAdTGF-β1-treated mice demonstrated progressive peritoneal fibrosis with adhesion formation and encapsulation of bowels. Weight gain was significantly reduced in animals treated with HDAdTGF-β1 compared to both the control-treated animals and the AdTGF-β1-treated animals. Inflammation was not a major component of the fibroproliferative response. Conclusions Peritoneal administration of a first-generation AdTGF-β1 leads to transient gene expression, resulting in a resolving fibrotic response and histology similar to that seen in simple peritoneal sclerosis. Prolonged TGF-β1 expression induced by the helper-dependent HDAdTGF-β1 led to changes in peritoneal morphology resembling EPS. This suggests that TGF-β1 may be a contributing factor in both simple peritoneal sclerosis and EPS. This model will be useful for elucidation of the mechanism of EPS and evaluation of potential treatment.

337 Background: Current risk stratification schema have limited prognostic performance in predicting outcome within National Comprehensive Cancer Network (NCCN) high-risk to very high-risk prostate cancer. Methods: Two multicenter high-risk cohorts were used for training (n = 214) and validation (n = 151) of a novel RNA microarray-based integrated clinical-genomic Classifier Optimized for Outcome in High-risk Prostate cancer (COOHP) to classify patients as COOHP favorable high-risk, standard high-risk, or very high-risk. Cox analysis was used to model metastasis-free survival (MFS), prostate cancer-specific survival (PCSS), and overall survival (OS). Model performance was compared to prior sub-classification systems using time-dependent c-indices. Results: Among NCCN high/very high-risk patients in the training cohort, 11% were classified as COOHP favorable high-risk, 70% as COOHP standard high-risk, and 18% as COOHP very high-risk. Patients with COOHP favorable high-risk disease had better rates of 5-year MFS compared to those with COOHP standard high-risk disease (94% vs 76%, hazard ratio [HR] 0.10, p = 0.02), and patients with COOHP very high-risk disease had worse 5-year MFS compared to those with COOHP standard high-risk disease (34% vs 76%, HR 3.5, p < 0.0001). Similarly, patients with COOHP very high-risk disease had worse 10-year PCSS compared to those with COOHP standard high-risk disease (36% vs 82%, HR 4.4, p < 0.0001). The c-indices for 5-year MFS and 10-year PCSS in the training cohort were 0.80 and 0.74, significantly improved compared to prior clinical and clinical-genomic risk stratification systems (0.62-0.69 for 5-year MFS and 0.56-0.63 for 10-year PCSS). These results were consistent in the validation cohort, where 5-year MFS significantly varied among the three COOHP subgroups (100% vs 89% vs 79%, p = 0.020), as did 10-year OS (100% vs 71% vs 53%, p = .040). Conclusions: A clinical-genomic risk stratification system specifically designed to discriminate prognosis in high-risk prostate cancer better identified favorable high-risk and very high-risk subsets of disease compared to prior clinical and clinical-genomic stratification systems.

149 Background: Patients with metastatic hormone-sensitive prostate cancer (mHSPC) have a variable response to ADT and some benefit from the addition of docetaxel or an androgen signaling pathway inhibitor. We found that loss of the epigenetic regulator KDM5D is associated with more aggressive prostate cancer (PC). We sought to determine whether MYBL2 which is regulated by KDM5D mediates this effect. MYBL2 is a transcription factor which controls key genes (e.g. FOXM1) and increases cell cycle progression and survival. Methods: AR expressing hormone-sensitive cell lines, LNCaP and LAPC4 were used. Motif analysis and CHiPseq of LNCaP with and without siKDM5D was performed and impact of modulation of KDM5D and MYBL2 in both cells on cell survival was assessed. Gene expression profiling (GEP) data assessed MYBL2’s association with KDM5D levels in localized disease (publicly available data) and mHSPC (Decipher whole Affymetrix platform on CHAARTED samples). Results: Silencing KDM5D increased H3K4me3 and increased MYBL2 expression. GEP showed a strong negative correlation between KDM5D and MYBL2 in patients with localized PC (-0.66; TCGA) but not from primary prostate cancer tissue with mHSPC (-0.03; CHAARTED). Cells with low KDM5D and high MYBL2 were androgen independent and more resistant to docetaxel. In TCGA, patients with high MYBL2 had a higher rate of relapse from localized disease. In patients with metastatic disease (CHAARTED) low MYBL2 was associated with a better overall survival (OS) on multivariable analysis when treated with ADT or ADT + docetaxel. Conclusions: Low MYBL2 is associated with a longer OS with ADT alone and ADT and docetaxel independent of clinical variables. Patients with high MYBL2 expression had better OS with ADT plus docetaxel compared with patients with high MYBL2 treated with ADT alone.[Table: see text]

139 Background: C may enhance response to immune checkpoint inhibitors by promoting an immune-permissive microenvironment. COSMIC-021 (NCT03170960), a multinational phase 1b study, is evaluating the combination of C with A in various solid tumors. We report interim results from Cohort 6 in mCRPC. Methods: Eligible pts were required to have radiographic progression in soft tissue after enzalutamide and/or abiraterone, measurable disease, and an ECOG PS of 0 or 1. Prior chemotherapy for mCSPC was permitted. Pts received C 40 mg PO qd and A 1200 mg IV q3w. CT/MRI scans were performed q6w for 52w and q12w thereafter. Bone scans were performed q12w. The primary endpoint is ORR per RECIST 1.1. Other endpoints include safety, ORR per irRECIST, duration of response (DOR), PFS, and OS. Results: As of Oct 2019, 44 mCRPC pts were enrolled with a median follow-up of 10.6 mo (range 3.4+, 17.9). Median age was 70 y (range 49, 90), 50% had ECOG PS 1, 34% had visceral metastases, and 61% had extrapelvic lymph node metastases. 27% of pts had prior docetaxel and 52% had ≥2 prior novel hormonal therapies. The most common any grade TEAEs were fatigue (57%), nausea (48%), decreased appetite (45%), diarrhea (39%), PPE (32%), and vomiting (32%). One Grade 5 TRAE of dehydration was reported in a 90 y/o. Median duration of treatment was 5.3 mo. The ORR per RECIST 1.1 among all pts was 32% (2 CRs [4.5%] and 12 PRs [27%]); 21 (48%) pts had SD giving a disease control rate of 80% in all pts. One pt with initial PD per RECIST 1.1 had an irPR per irRECIST. ORR per RECIST 1.1 was 33% in 36 pts with high-risk clinical features (visceral and/or extrapelvic lymph node metastases). Median DOR for all pts with response per RECIST 1.1 was 8.3 mo (range 1.38+, 9.76+). In 12 responders with at least 1 post-baseline PSA evaluation, 8 (67%) had a PSA decline ≥50%. Conclusions: The combination of C+A had a tolerable safety profile and demonstrated clinically meaningful activity with durable responses in men with mCRPC. Given the encouraging activity in these pts, especially in those with visceral and/or extra pelvic lymph node metastases, further evaluation of C+A in men with mCRPC is being pursued. Clinical trial information: NCT03170960.

Abstract Background Human epidermal growth factor receptor ( EGFR ) is an oncogenic gene and one of top targets of precision therapy in lung cancer with EGFR mutations. Although there are many reports for some individual cancers, comprehensive profiling of EGFR mutations, overexpression, amplification, DNA methylation, and their clinical associations across many different cancers simultaneously was not available. This study aimed to fill the gap and provide insights to the alteration spectrum of EGFR and its therapeutic and prognostic implications. Methods The Cancer Genome Atlas (TCGA) datasets for 32 cancer types involving 11,314 patients were analyzed for alterations (mutations and amplification/deletion), abnormal expression and DNA methylation in EGFR gene. Mutation frequency, genomic location distribution, functional impact, and clinical targeted therapy implication were compared among different cancer types, and their associations with patient survival were analyzed. Results EGFR alteration frequency, mutation sites across functional domains, amplification, overexpression, and DNA methylation patterns differed greatly among different cancer types. The overall mutation frequency in all cancers combined was relatively low. Targetable mutations, mainly in lung cancer, were primarily found in the Pkinase_Tyr domain. Glioblastoma multiforme had the highest rate of alterations, but it was dominated by gene amplification and most mutations were in the Furin‐like domain where targeted therapy was less effective. Low‐grade glioma often had gene amplification and increased EGFR expression which was associated with poor outcome. Colon and pancreatic adenocarcinoma had very few EGFR mutations; however, high EGFR expression was significantly associated with short patient survival. Squamous cell carcinoma regardless of their sites (the head and neck, lung, or esophagus) exhibited similar characteristics with an alteration frequency of about 5.0%, was dominated by gene amplification, and had increased EGFR expression generally associated with short patient survival. DNA methylation was highly associated with EGFR expression and patient outcomes in some cancers. Conclusions EGFR aberration type, frequency, distribution in functional domains, and expression vary from cancer to cancer. While mutations in the Pkinase_Tyr domain are more important for treatment selection, increased expression from amplification or deregulation affects more tumor types and leads to worse outcome, which calls for new treatment strategies for these EGFR‐ driven tumors.

Background Systemic AL amyloidosis is a clonal plasma cell disorder in which amyloid fibrils are deposited in tissues and organs, leading to multi-system organ dysfunction. The most frequently involved organs are the heart and kidney (individually or together), with advanced cardiac involvement conferring particularly poor outcomes. Achievement of hematologic response and improved organ function result in better outcomes. There are currently no approved treatments for AL amyloidosis; multiple myeloma (MM) treatment strategies are used for these pts. Active, tolerable treatment options specific for AL amyloidosis are needed. The oral proteasome inhibitor (PI) ixazomib (ixa) is active and approved in combination with lenalidomide (L)-dexamethasone (Dex) for the treatment of MM pts who have received ≥1 prior therapy. Methods RRAL pts with measurable disease and major organ involvement (cardiac/renal) who required treatment after 1-2 prior therapies (and were not refractory to prior PI therapy) were randomized to ixa (4.0 mg, d 1, 8, 15) plus Dex (20 mg, d 1, 8, 15, 22) or physician's choice (Dex alone or plus melphalan [M], cyclophosphamide [C], thalidomide [T], or lenalidomide [L]) in 28-d cycles until disease progression or unacceptable toxicity (or best response plus 2 cycles or maximum 18 mos therapy/600 mg total dose for MDex). Randomization was stratified by cardiac risk stage, relapsed vs refractory disease, and prior PI exposure. Primary endpoints were 1) overall hematologic response rate (ORR) centrally adjudicated, and 2) death or vital organ deterioration at 2 yrs. Key secondary endpoints were overall survival (OS) and hematologic complete response (CR) rate; other secondary endpoints included hematologic/vital organ progression-free survival (PFS), time to vital organ deterioration or mortality, duration of hematologic response (DOR), and safety. Results 168 pts were randomized to ixa-Dex (n=85) or physician's choice (n=83; 47 LDex, 24 MDex, 10 CDex, 2 TDex); median age was 65 (range 38-84) vs 66 (33-82) yrs, 60% vs 55% were male, 56% vs 63% had cardiac and 66% vs 58% had renal involvement (33% vs 23% had both) at initial diagnosis (plus 9% vs 12% liver, 12% vs 18% gastrointestinal tract, and 11% vs 10% peripheral nerve involvement), 47% vs 47% had prior bortezomib, and 47% vs 37% had prior transplant. Median time since diagnosis was 14.7 vs 15.9 mos. Hematologic responses were seen in 45 (53%) vs 42 (51%) pts receiving ixa-Dex vs physician's choice (odds ratio 1.10 [95% CI 0.60-2.01], p=0.762). Higher CR rates were seen with ixa-Dex vs physician's choice (26% vs 18%). Overall survival, overall/hematologic/vital organ PFS, time to vital organ deterioration/death, DOR, time to treatment failure, and time to subsequent therapy data all favored pts treated with ixa-Dex vs physician's choice (Figure). Vital organ response rates were 36% in the ixa-Dex arm vs 11% with physician's choice (cardiac response rate: 18% vs 5%; renal response rate: 28% vs 7%). At data cut-off, pts had received a median treatment duration of 11.7 vs 4.9 mos with ixa-Dex vs physician's choice, and 21% vs 6% of pts remained on treatment. Grade ≥3 adverse events (AEs) were seen in 59% vs 56% of pts, including 33% vs 41% with drug-related grade ≥3 AEs, 45% vs 33% had serious AEs, 25% vs 20% had AEs resulting in discontinuation, and there were 6% vs 5% on-study deaths. AEs of clinical importance included diarrhea (34% vs 30%), rash (33% vs 20%), cardiac arrhythmias (25% vs 15%), nausea (24% vs 14%), pneumonia (22% vs 16%), and peripheral neuropathy (20% vs 15%). Common (≥5% overall) grade ≥3 AEs were fatigue (9% vs 9%), anemia (2% vs 10%), cardiac failure, dyspnea (each 6% vs 4%), peripheral edema, and pneumonia (each 5% vs 5%). Conclusions Treatment with ixa-Dex significantly prolonged duration of composite survival and vital organ function, PFS, and time to subsequent therapy vs physician's choice. Moreover, ixa-Dex resulted in an improved CR rate and DOR and, although the primary endpoint of hematologic response was not met, all clinically relevant time-to-event endpoint data favored ixa-Dex vs physician's choice. Ixa-Dex was generally well tolerated and associated with a doubling of treatment duration vs physician's choice; no new safety signals were seen. TOURMALINE-AL1 is the first phase 3 trial in RRAL to show significant outcome improvements, suggesting ixa-Dex represents a new option for RRAL pts, who have limited access to therapies. Disclosures Dispenzieri: Akcea: Consultancy; Intellia: Consultancy; Alnylam: Research Funding; Celgene: Research Funding; Takeda: Research Funding; Pfizer: Research Funding; Janssen: Consultancy. Kastritis:Amgen: Honoraria, Research Funding; Janssen: Honoraria, Research Funding; Takeda: Honoraria; Pfizer: Honoraria; Prothena: Honoraria; Genesis: Honoraria. Wechalekar:Takeda: Honoraria; GSK: Honoraria; Celgene: Honoraria; Amgen: Research Funding; Janssen-Cilag: Honoraria. Schönland:Prothena: Membership on an entity's Board of Directors or advisory committees, Research Funding; Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding; Takeda: Membership on an entity's Board of Directors or advisory committees, Research Funding; Medac: Other: Travel Grant. Sanchorawala:Proclara: Consultancy, Honoraria; Caelum: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees; Janssen: Research Funding; Prothena: Research Funding; Celgene: Research Funding; Takeda: Research Funding. Landau:Pfizer: Membership on an entity's Board of Directors or advisory committees; Caelum: Membership on an entity's Board of Directors or advisory committees; Prothena: Membership on an entity's Board of Directors or advisory committees; Amgen: Research Funding; Takeda: Membership on an entity's Board of Directors or advisory committees, Research Funding; Karyopharm: Consultancy, Honoraria; Celgene: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees. Suzuki:Takeda: Honoraria; BMS: Honoraria, Research Funding; Ono: Research Funding; Celgene: Honoraria; Janssen: Honoraria. Comenzo:Takeda: Research Funding; Caelum: Consultancy, Membership on an entity's Board of Directors or advisory committees; Janssen: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Karyopharm: Research Funding; Prothena Biosciences: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Sanofi-Aventis: Membership on an entity's Board of Directors or advisory committees; Unum: Membership on an entity's Board of Directors or advisory committees, Research Funding; Myself: Patents & Royalties: Patent 9593332, Pending 20170008966. Berg:Millennium Pharmaceuticals, Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Ltd: Employment, Patents & Royalties. Liu:Millennium Pharmaceuticals, Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited: Employment. Faller:Phoenicia Biosciences: Equity Ownership; Briacell Pharmaceuticals: Equity Ownership; Millennium Pharmaceuticals, Inc., a wholly owned subsidiary of Takeda Pharmaceutical Company Limited: Employment; Viracta Pharmaceuticals: Equity Ownership; Boston University: Employment. Off Label Disclosure: Investigation of the oral proteasome inhibitor ixazomib in combination with dexamethasone versus physician's choice (of which there are no approved treatment options) for the treatment of relapsed/refractory primary systemic amyloidosis.