A Multimodal Data Analysis Approach for Targeted Drug Discovery Involving Topological Data Analysis (TDA).

In silico drug discovery refers to a combination of computational techniques that augment our ability to discover drug compounds from compound libraries. Many such techniques exist, including virtual high-throughput screening (vHTS), high-throughput screening (HTS), and mechanisms for data storage and querying. However, presently these tools are often used independent of one another. In this chapter, we describe a new multimodal in silico technique for the hit identification and lead generation phases of traditional drug discovery. Our technique leverages the benefits of three independent methods-virtual high-throughput screening, high-throughput screening, and structural fingerprint analysis-by using a fourth technique called topological data analysis (TDA). We describe how a compound library can be independently tested with vHTS, HTS, and fingerprint analysis, and how the results can be transformed into a topological data analysis network to identify compounds from a diverse group of structural families. This process of using TDA or similar clustering methods to identify drug leads is advantageous because it provides a mechanism for choosing structurally diverse compounds while maintaining the unique advantages of already established techniques such as vHTS and HTS.

Physician sentiment toward artificial intelligence (AI) in colonoscopic practice: a survey of US gastroenterologists.

Background and study aims Early studies have shown that artificial intelligence (AI) has the potential to augment the performance of gastroenterologists during endoscopy. Our aim was to determine how gastroenterologists view the potential role of AI in gastrointestinal endoscopy. Methods In this cross-sectional study, an online survey was sent to US gastroenterologists. The survey included questions about physician level of training, experience, and practice characteristics and physician perception of AI. Descriptive statistics were used to summarize sentiment about AI. Univariate and multivariate analyses were used to assess whether background information about physicians correlated to their sentiment. Results Surveys were emailed to 330 gastroenterologists nationwide. Between December 2018 and January 2019, 124 physicians (38 %) completed the survey. Eighty-six percent of physicians reported interest in AI-assisted colonoscopy; 84.7 % agreed that computer-assisted polyp detection (CADe) would improve their endoscopic performance. Of the respondents, 57.2 % felt comfortable using computer-aided diagnosis (CADx) to support a "diagnose and leave" strategy for hyperplastic polyps. Multivariate analysis showed that post-fellowship experience of fewer than 15 years was the most important factor in determining whether physicians were likely to believe that CADe would lead to more removed polyps (odds ratio = 5.09; P  = .01). The most common concerns about implementation of AI were cost (75.2 %), operator dependence (62.8 %), and increased procedural time (60.3 %). Conclusions Gastroenterologists have strong interest in the application of AI to colonoscopy, particularly with regard to CADe for polyp detection. The primary concerns were its cost, potential to increase procedural time, and potential to develop operator dependence. Future developments in AI should prioritize mitigation of these concerns.

Yield of biliary stent cytology: Is it time to think lean?

Background and study aims During evaluation of pancreaticobiliary strictures, it is common practice to send biliary stents for cytologic analysis. However, in recent years, complementary tissue acquisition techniques ranging from cholangioscopy to fine-needle biopsy have improved the ability to acquire tissue and diagnose malignancy. Data are limited on the current diagnostic yield and cost effectiveness of biliary stent analysis. Patients and methods We performed a retrospective study of all pancreaticobiliary stents sent for analysis in a tertiary care academic medical center from June 2013 to September 2016. Patient demographics, stent information, and final diagnosis history were collected through chart review. Costs were determined using published reimbursement rates for Medicare. Results Two hundred thirty-one stents from 175 patients were sent for cytologic analysis during the study period. Of the 62 stents obtained from patients ultimately diagnosed with malignancy, only one (1.6 %) had positive cytology for malignant cells, while the others were acellular/non-diagnostic (2/62, 3.2 %), negative (48/62, 77.4 %), or atypical (11/62, 17.7 %). The sensitivity of stent cytology for diagnosis of malignancy was 1.6 % (1/62). No cases were identified in which stent cytology changed clinical management. From a payer perspective, the mean estimated cost for each stent cytologic analysis is greater than $ 70.00. Conclusions While stent cytologic analysis is a common clinical practice, the diagnostic yield and cost effectiveness of the practice must be reevaluated. With the rise of newer diagnostic technologies such as digital cholangioscopy and endoscopic ultrasound-guided fine-needle biopsy, it may be time to "think lean" and acknowledge a sunset for biliary stent cytology.

Artificial intelligence in gastrointestinal endoscopy: The future is almost here.

Artificial intelligence (AI) enables machines to provide unparalleled value in a myriad of industries and applications. In recent years, researchers have harnessed artificial intelligence to analyze large-volume, unstructured medical data and perform clinical tasks, such as the identification of diabetic retinopathy or the diagnosis of cutaneous malignancies. Applications of artificial intelligence techniques, specifically machine learning and more recently deep learning, are beginning to emerge in gastrointestinal endoscopy. The most promising of these efforts have been in computer-aided detection and computer-aided diagnosis of colorectal polyps, with recent systems demonstrating high sensitivity and accuracy even when compared to expert human endoscopists. AI has also been utilized to identify gastrointestinal bleeding, to detect areas of inflammation, and even to diagnose certain gastrointestinal infections. Future work in the field should concentrate on creating seamless integration of AI systems with current endoscopy platforms and electronic medical records, developing training modules to teach clinicians how to use AI tools, and determining the best means for regulation and approval of new AI technology.

Albumin and Neutrophil-Lymphocyte Ratio (NLR) Predict Survival in Patients With Pancreatic Adenocarcinoma Treated With SBRT.

PURPOSE: To determine if pretreatment nutritional status and inflammatory markers correlate with survival in patients with locally advanced pancreatic adenocarcinoma treated with stereotactic body radiotherapy (SBRT). MATERIALS AND METHODS: We retrospectively reviewed 208 patients with newly diagnosed, locally advanced pancreatic adenocarcinoma treated with SBRT at our institution from 2002 to 2014. Laboratory values were collected before SBRT, including hemoglobin, platelets, albumin, red blood cell, white blood cell, neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio, and tumor markers CA 19-9 and CEA. Patients were followed every 3 months with computed tomography (CT) and/or positron emission tomography-CT imaging to monitor for local recurrence and overall survival (OS). RESULTS: Median follow-up after SBRT was 7.5 months (interquartile range, 4.6 to 12.0 mo) for all patients. Median OS for patients with NLR>5 compared with NLR≤5 was 6.9 and 8.5 months, respectively (P=0.0057). On univariate analysis, receipt of chemotherapy (P=0.05, hazard ratio [HR]=0.69), increased albumin (P=0.002, HR=0.64), increased red blood cell (P=0.05, HR=0.75), increased lymphocyte count (P=0.002, HR=0.66), decreased CEA (P=0.01, HR=0.96), and NLR≤5 (P=0.01, HR=0.65) correlated with improved OS. On multivariate analysis, higher albumin (P=0.03, HR=0.70), receipt of chemotherapy (P=0.007, HR=0.56), and NLR≤5 (P=0.02, HR=0.66) correlated with better survival. CONCLUSIONS: Preradiotherapy low albumin levels and NLR>5 correlate with decreased survival in patients with locally advanced pancreatic adenocarcinoma treated with SBRT, indicating the prognostic value of systemic inflammatory markers (such as NLR) and a role of nutritional supplementation to improve outcomes in these patients. Further investigation is warranted.

A Three-Step Letter Advance Directive Procedure to Facilitate Patient-Proxy Alignment in Advance Care Planning.

Background: Little is known about the extent of alignment between hematopoietic stem cell transplant (HSCT) patients and their healthcare proxies with respect to advance care planning (ACP). Aim: To determine if a structured three-step process using the letter advance directive (LAD) could (1) allow for the differences in opinion between patient-proxy dyads to surface and (2) help bridge preexisting discordance about specific treatment choices. Design: Blinded to each other, the HSCT patient (LAD-1) and proxy (LAD-2) each completed the LAD (step 1). They unmasked, compared LAD-1 and LAD-2, and discussed their choices (step 2). They completed a final letter directive (LAD-3) by consensus (step 3). Settings/Participants: Convenience sample of eighty dyads (patient and proxy) at a regional HSCT referral center. Results: The mean patient-proxy concordance was 72.9% for the 12 questions in the LAD. Wanting to be pain free at the end of life was the statement with the most amount of agreement (88.75% in LAD-1, 91.25% in LAD-2, and 90% in LAD-3). Patient-proxy dyads had notable discordance related to specific treatments. The highest discordance was related to ventilator support (46.3% of patients refused it, while 58.8% of proxies refused on behalf of the patient). Overall, proxies were more likely than patients to opt in for dialyses and hospice care but more likely to opt out for cardiac resuscitation and sedation to palliate refractory symptoms. On open discussion, patient-proxy discordance mostly resolved in favor of the patient. Conclusions: The ACP process should allow for patient-proxy differences to surface, facilitate a discussion about the granular details with the goal of reaching consensus. Our three-step approach using the LAD is an effective way to identify areas of patient-proxy concordance and discordance about specific treatment preferences. A structured patient-proxy discussion using the LAD helped reconcile discordance and most often in favor of a patient's original wishes.

Decaying relevance of clinical data towards future decisions in data-driven inpatient clinical order sets.

OBJECTIVE: Determine how varying longitudinal historical training data can impact prediction of future clinical decisions. Estimate the "decay rate" of clinical data source relevance. MATERIALS AND METHODS: We trained a clinical order recommender system, analogous to Netflix or Amazon's "Customers who bought A also bought B..." product recommenders, based on a tertiary academic hospital's structured electronic health record data. We used this system to predict future (2013) admission orders based on different subsets of historical training data (2009 through 2012), relative to existing human-authored order sets. RESULTS: Predicting future (2013) inpatient orders is more accurate with models trained on just one month of recent (2012) data than with 12 months of older (2009) data (ROC AUC 0.91 vs. 0.88, precision 27% vs. 22%, recall 52% vs. 43%, all P<10-10). Algorithmically learned models from even the older (2009) data was still more effective than existing human-authored order sets (ROC AUC 0.81, precision 16% recall 35%). Training with more longitudinal data (2009-2012) was no better than using only the most recent (2012) data, unless applying a decaying weighting scheme with a "half-life" of data relevance about 4 months. DISCUSSION: Clinical practice patterns (automatically) learned from electronic health record data can vary substantially across years. Gold standards for clinical decision support are elusive moving targets, reinforcing the need for automated methods that can adapt to evolving information. CONCLUSIONS AND RELEVANCM: Prioritizing small amounts of recent data is more effective than using larger amounts of older data towards future clinical predictions.

Does radiotherapy still have a role in unresected biliary tract cancer?

The benefits of radiotherapy for inoperable biliary tract cancer remain unclear due to the lack of randomized data. We evaluated the impact of radiotherapy on survival in elderly patients using the SEER-Medicare database. Patients in the SEER-Medicare database with inoperable biliary tract tumors diagnosed between 1998 and 2011 were included. We used multivariate logistic regression to evaluate factors associated with treatment selection, and multivariate Cox regression and propensity score matching to evaluate treatment selection in relation to subsequent survival. Of the 2343 patients included, 451 (19%) received radiotherapy within 4 months of diagnosis. The use of radiotherapy declined over time, and was influenced by receipt of chemotherapy and patient age, race, marital status, poverty status, and tumor stage and type. Median survival was 9.3 (95% CI 8.7-9.7) months among patients who did not receive radiation and 10.0 (95% CI 9.1-11.3) months among those who received radiation, conditional on having survived 4 months. In patients who received chemotherapy (n = 1053), receipt of radiation was associated with improved survival, with an adjusted hazard ratio of 0.82 (95% 0.70-0.97, P = 0.02). In patients who did not receive chemotherapy (n = 1290), receipt of radiation was not associated with improved survival, with an adjusted hazard ratio of 1.09 (95% 0.91-1.30, P = 0.34). Propensity-scored matched analyses showed similar results. Despite the survival benefit associated with the addition of radiotherapy to chemotherapy, the use of radiation for unresectable biliary tract cancers has declined over time.

Quantitative Analysis of (18)F-Fluorodeoxyglucose Positron Emission Tomography Identifies Novel Prognostic Imaging Biomarkers in Locally Advanced Pancreatic Cancer Patients Treated With Stereotactic Body Radiation Therapy.

PURPOSE: To identify prognostic biomarkers in pancreatic cancer using high-throughput quantitative image analysis. METHODS AND MATERIALS: In this institutional review board-approved study, we retrospectively analyzed images and outcomes for 139 locally advanced pancreatic cancer patients treated with stereotactic body radiation therapy (SBRT). The overall population was split into a training cohort (n=90) and a validation cohort (n=49) according to the time of treatment. We extracted quantitative imaging characteristics from pre-SBRT (18)F-fluorodeoxyglucose positron emission tomography, including statistical, morphologic, and texture features. A Cox proportional hazard regression model was built to predict overall survival (OS) in the training cohort using 162 robust image features. To avoid over-fitting, we applied the elastic net to obtain a sparse set of image features, whose linear combination constitutes a prognostic imaging signature. Univariate and multivariate Cox regression analyses were used to evaluate the association with OS, and concordance index (CI) was used to evaluate the survival prediction accuracy. RESULTS: The prognostic imaging signature included 7 features characterizing different tumor phenotypes, including shape, intensity, and texture. On the validation cohort, univariate analysis showed that this prognostic signature was significantly associated with OS (P=.002, hazard ratio 2.74), which improved upon conventional imaging predictors including tumor volume, maximum standardized uptake value, and total legion glycolysis (P=.018-.028, hazard ratio 1.51-1.57). On multivariate analysis, the proposed signature was the only significant prognostic index (P=.037, hazard ratio 3.72) when adjusted for conventional imaging and clinical factors (P=.123-.870, hazard ratio 0.53-1.30). In terms of CI, the proposed signature scored 0.66 and was significantly better than competing prognostic indices (CI 0.48-0.64, Wilcoxon rank sum test P<1e-6). CONCLUSION: Quantitative analysis identified novel (18)F-fluorodeoxyglucose positron emission tomography image features that showed improved prognostic value over conventional imaging metrics. If validated in large, prospective cohorts, the new prognostic signature might be used to identify patients for individualized risk-adaptive therapy.

Acridine Derivatives as Inhibitors of the IRE1α-XBP1 Pathway Are Cytotoxic to Human Multiple Myeloma.

Using a luciferase reporter-based high-throughput chemical library screen and topological data analysis, we identified N-acridine-9-yl-N',N'-dimethylpropane-1,3-diamine (DAPA) as an inhibitor of the inositol requiring kinase 1α (IRE1α)-X-box binding protein-1 (XBP1) pathway of the unfolded protein response. We designed a collection of analogues based on the structure of DAPA to explore structure-activity relationships and identified N(9)-(3-(dimethylamino)propyl)-N(3),N(3),N(6),N(6)-tetramethylacridine-3,6,9-triamine (3,6-DMAD), with 3,6-dimethylamino substitution on the chromophore, as a potent inhibitor. 3,6-DMAD inhibited both IRE1α oligomerization and in vitro endoribonuclease (RNase) activity, whereas the other analogues only blocked IRE1α oligomerization. Consistent with the inhibition of IRE1α-mediated XBP1 splicing, which is critical for multiple myeloma cell survival, these analogues were cytotoxic to multiple myeloma cell lines. Furthermore, 3,6-DMAD inhibited XBP1 splicing in vivo and the growth of multiple myeloma tumor xenografts. Our study not only confirmed the utilization of topological data analysis in drug discovery but also identified a class of compounds with a unique mechanism of action as potent IRE1α-XBP1 inhibitors in the treatment of multiple myeloma. Mol Cancer Ther; 15(9); 2055-65. ©2016 AACR.

Identification of Doxorubicin as an Inhibitor of the IRE1α-XBP1 Axis of the Unfolded Protein Response.

Activation of the IRE1α-XBP1 branch of the unfolded protein response (UPR) has been implicated in multiple types of human cancers, including multiple myeloma (MM). Through an in silico drug discovery approach based on protein-compound virtual docking, we identified the anthracycline antibiotic doxorubicin as an in vitro and in vivo inhibitor of XBP1 activation, a previously unknown activity for this widely utilized cancer chemotherapeutic drug. Through a series of mechanistic and phenotypic studies, we showed that this novel activity of doxorubicin was not due to inhibition of topoisomerase II (Topo II). Consistent with its inhibitory activity on the IRE1α-XBP1 branch of the UPR, doxorubicin displayed more potent cytotoxicity against MM cell lines than other cancer cell lines that have lower basal IRE1α-XBP1 activity. In addition, doxorubicin significantly inhibited XBP1 activation in CD138(+) tumor cells isolated from MM patients. Our findings suggest that the UPR-modulating activity of doxorubicin may be utilized clinically to target IRE1α-XBP1-dependent tumors such as MM.

Single- versus multifraction stereotactic body radiation therapy for pancreatic adenocarcinoma: outcomes and toxicity.

PURPOSE: We report updated outcomes of single- versus multifraction stereotactic body radiation therapy (SBRT) for unresectable pancreatic adenocarcinoma. METHODS AND MATERIALS: We included 167 patients with unresectable pancreatic adenocarcinoma treated at our institution from 2002 to 2013, with 1-fraction (45.5% of patient) or 5-fraction (54.5% of patients) SBRT. The majority of patients (87.5%) received chemotherapy. RESULTS: Median follow-up was 7.9 months (range: 0.1-63.6). The 6- and 12-month cumulative incidence rates (CIR) of local recurrence for patients treated with single-fraction SBRT were 5.3% (95% confidence interval [CI], 0.2%-10.4%) and 9.5% (95% CI, 2.7%-16.2%), respectively. The 6- and 12-month CIR with multifraction SBRT were 3.4% (95% CI, 0.0-7.2%) and 11.7% (95% CI, 4.8%-18.6%), respectively. Median survival from diagnosis for all patients was 13.6 months (95% CI, 12.2-15.0 months). The 6- and 12- month survival rates from SBRT for the single-fraction group were 67.0% (95% CI, 57.2%-78.5%) and 30.8% (95% CI, 21.9%-43.6%), respectively. The 6- and 12- month survival rates for the multifraction group were 75.7% (95% CI, 67.2%-85.3%) and 34.9% (95% CI, 26.1%-46.8%), respectively. There were no differences in CIR or survival rates between the single- and multifraction groups. The 6- and 12-month cumulative incidence rates of gastrointestinal toxicity grade ≥3 were 8.1% (95% CI, 1.8%-14.4%) and 12.3% (95% CI, 4.7%-20.0%), respectively, in the single-fraction group, and both were 5.6% (95% CI, 0.8%-10.5%) in the multifraction group. There were significantly fewer instances of toxicity grade ≥2 with multifraction SBRT (P=.005). Local recurrence and toxicity grade ≥2 were independent predictors of worse survival. CONCLUSIONS: Multifraction SBRT for pancreatic cancer significantly reduces gastrointestinal toxicity without compromising local control.