SEARCH · bioRxiv Science

Results for “Clinical Trials”

Search indexed bioRxiv preprints in genomics, neuroscience, cell biology and bioinformatics. Read source abstracts and check manuscript versions; preprints are not peer reviewed.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2Linked to original sources

bioRxiv2018

Real Time PCR for the Evaluation of Treatment Response in Clinical Trials of Adult Chronic Chagas Disease: Usefulness of Serial Blood Sampling and qPCR Replicates.

This work evaluated a serial blood sampling procedure to enhance the sensitivity of duplex real time PCR (qPCR) for baseline detection and quantification of parasitic loads and post-treatment identification of failure in the context of clinical trials for treatment of chronic Chagas disease, namely DNDi-CH-E1224-001 (NCT01489228) and MSF-DNDi PCR sampling optimization study (NCT01678599). Patients from Cochabamba (N= 294), Tarija (N = 257), and Aiquile (N= 220) were enrolled. Three serial blood samples were collected at each time-point, and qPCR triplicates were tested per sample. The first two samples were collected during the same day and the third one seven days later.\n\nA patient was considered PCR positive if at least one qPCR replicate was detectable. Cumulative results of multiple samples and qPCR replicates enhanced the proportion of pre-treatment sample positivity from 54.8 to 76.2%, 59.5 to 77.8%, and 73.5 to 90.2% in Cochabamba, Tarija, and Aiquile cohorts, respectively and increased cumulative detection of treatment failure from 72.9 to 91.7%, 77.8 to 88.9%, and 42.9 to 69.1% for E1224 low, short, and high dosage regimes, respectively; and from 4.6 to 15.9% and 9.5 to 32.1% for the benznidazole (BZN) arm in the DNDi-CH-E1224-001 and MSF-DNDi studies, respectively. The monitoring of patients treated with placebo in the DNDi-CH-E1224-001 trial revealed fluctuations in parasitic loads and occasional non-detectable results. This serial sampling strategy enhanced PCR sensitivity to detecting treatment failure during follow-up and has the potential for improving recruitment capacity in Chagas disease trials which require an initial positive qPCR result for patient admission.

microbiology↗

bioRxiv2018

Phenotypes associated with genes encoding drug targets are predictive of clinical trial side effects

Biomedical scientists face major challenges in developing novel drugs for unmet medical needs. Only a small fraction of early drug programs progress to the market, due to safety and efficacy failures, despite extensive efforts to predict drug and target safety as early as possible using a variety of assays in vitro and in preclinical species. In principle, characterizing the effect of natural variation in the genes encoding drug targets should present a powerful alternate approach to predict not only whether a protein will be an effective drug target, but also whether a protein will be an inherently safe drug target, while avoiding the challenges of translating biology from experiments in non-human species. We have embarked on a retrospective analysis, demonstrating for the first time a statistical link between the organ systems involved in genetic syndromes of drug target genes and the organ systems in which side effects are observed clinically. Across 1,819 drugs and 21 organ system phenotype categories analyzed, drug side effects are more likely to occur in organ systems where there is genetic evidence of a link between the drug target and a phenotype involving that organ system, compared to when there is no such genetic evidence (30.0% vs 19.2%; OR = 1.80). Conversely, we find that having genetic evidence of a Mendelian syndrome involving a drug target in which a certain organ system is unaffected decreases the likelihood that side effects will manifest in that organ system, relative to having no informative syndrome (18.5% vs 20.2%; OR = 0.89). We find a relationship between genetics and side effects even when controlling for known confounders such as drug delivery route and indication. We highlight examples where genetics of drug targets could have anticipated side effects observed during clinical trials. This result suggests that human genetic data should be routinely used to predict potential safety issues associated with novel drug targets. This may lead to selection of better targets, appropriate monitoring of putative side effects early in development, reduction of the use of preclinical animal experiments, and ultimately increased success of molecules. Furthermore, deeply phenotyping human knockouts will be critically important to understand the full spectrum of effects that a new drug may elicit.

genetics↗

bioRxiv2018

An Assessment of Skin Lesion Measurement Techniques for Use in Clinical Trials of Acute Bacterial Skin and Skin Structure Infections

BackgroundLimited data are available to support a reproducible measurement technique that could be used to assess the response of a skin lesion associated with a bacterial infection to antibacterial therapy.\n\nMethodsThis multicenter, observational study enrolled patients with a major cutaneous abscess, a traumatic wound or surgical site infection, or a cellulitis. The primary objective was to characterize the intra- and inter-observer variability inherent in measuring the size of the erythema associated with the presenting skin infection. At least two observers made ruler measurements of the infection site with the length of the infection measured as the longest dimension of the erythematous area and width measured as the largest dimension perpendicular to the longest length. Intra- and inter-observer variability was determined by the intraclass correlation coefficient (ICC). Photographs, tracings, and thermal imaging were also performed.\n\nResultsThe intra-observer ICC (95% CI) for lesion area as measured by ruler was 0.999 (0.998, 0.999), suggesting only a very small amount of the variability was due to measurement error. The difference in mean lesion area measurements by the same observer was <1%. The inter-observer ICC (95% CI) for lesion area as measured by ruler was 0.990 (0.981, 0.995) suggesting that the results between observers were also highly reliable.\n\nConclusionsMeasurement of infection area as defined by erythema and measured by ruler shows excellent intra- and inter-observer reliability and can be used in future clinical trials of acute bacterial skin infections.

pathology↗

bioRxiv2015

Diagnostic Assessment of Osteosarcoma Chemoresistance Based on Virtual Clinical Trials

Osteosarcoma is the most common primary bone tumor in pediatric and young adult patients. Successful treatment of osteosarcomas requires a combination of surgical resection and systemic chemotherapy, both neoadjuvant (prior to surgery) and adjuvant (after surgery). The degree of necrosis following neoadjuvant chemotherapy correlates with the subsequent probability of disease-free survival. Tumors with less than 10% of viable cells after treatment represent patients with a more favorable prognosis. However, being able to predict early, such as at the time of the pre-treatment tumor biopsy, how the patient will respond to the standard chemotherapy would provide an opportunity for more personalized patient care. Patients with unfavorable predictions could be studied in a protocol, rather than a standard setting, towards improving therapeutic success. The onset of necrotic cells in osteosarcomas treated with chemotherapeutic agents is a measure of tumor sensitivity to the drugs. We hypothesize that the remaining viable cells, i.e., cells that have not responded to the treatment, are chemoresistant, and that the pathological characteristics of these chemoresistant tumor cells within the osteosarcoma pre-treatment biopsy can predict tumor response to the standard-of-care chemotherapeutic treatment. This hypothesis can be tested by comparing patient histopathology samples before, as well as after treatment to identify both morphological and immunochemical cellular features that are characteristic of chemoresistant cells, i.e., cells that survived treatment. Consequently, using computational simulations of dynamic changes in tumor pathology under the simulated standard of care chemotherapeutic treatment, one can couple the pre- and post-treatment morphological and spatial patterns of chemoresistant cells, and correlate them with patient clinical diagnoses. This procedure, that we named Virtual Clinical Trials, can serve as a potential predictive biomarker providing a novel value added decision support tool for oncologists.

Systems Biology↗

bioRxiv2018

Comparative efficacy and acceptability of non-surgical brain stimulation for the acute treatment of adult major depressive episodes: A systematic review and network meta-analysis of 113 randomised clinical trials

Background: Non-surgical brain stimulation techniques have been applied as tertiary treatments in major depression. However, the relative efficacy and acceptability of individual protocols is uncertain. Our aim was to estimate the comparative clinical efficacy and acceptability of non-surgical brain stimulation for the acute treatment of major depressive episodes in adults.\n\nMethods: Embase, PubMed/MEDLINE and PsycINFO were searched up until May 8, 2018, supplemented by manual searches of bibliographies of recent reviews and included trials. We included clinical trials with random allocation to electroconvulsive therapy (ECT), repetitive transcranial magnetic stimulation (rTMS), accelerated TMS (aTMS), priming TMS (pTMS), deep TMS (dTMS), theta burst stimulation (TBS), synchronised TMS (sTMS), magnetic seizure therapy (MST) or transcranial direct current stimulation (tDCS) protocols or sham. Data were extracted from published reports and outcomes were synthesised using pairwise and network random-effects meta-analysis. Primary outcomes were response (efficacy) and all-cause discontinuation (acceptability). We computed odds ratios (OR) with 95% confidence intervals (CI). Remission and continuous post-treatment depression severity scores were also examined.\n\nResults: 113 trials (262 treatment arms) randomising 6,750 patients (mean age = 47.9 years; 59% female) with major depressive disorder or bipolar depression met our inclusion criteria. In terms of efficacy, 10 out of 18 treatment protocols were associated with higher response relative to sham in network meta-analysis: bitemporal ECT (OR=8.91, 95%CI 2.57-30.91), high-dose right-unilateral ECT (OR=7.27, 1.90-27.78), pTMS (OR=6.02, 2.21-16.38), MST (OR=5.55, 1.06-28.99), bilateral rTMS (OR=4.92, 2.93-8.25), bilateral TBS (OR=4.44, 1.47-13.41), low-frequency right rTMS (OR=3.65, 2.13-6.24), intermittent TBS (OR=3.20, 1.45-7.08), high-frequency left rTMS (OR=3.17, 2.29-4.37) and tDCS (OR=2.65, 1.55-4.55). Comparing active treatments, bitemporal ECT and high-dose right-unilateral ECT were associated with increased response. All treatment protocols were at least as acceptable as sham treatment.\n\nConclusion: We found that non-surgical brain stimulation techniques constitute viable alternative or add-on treatments for adult patients with major depressive episodes. Our findings also highlight the need to consider other patient and treatment-related factors in addition to antidepressant efficacy and acceptability when making clinical decisions; and emphasize important research priorities in the field of brain stimulation.\n\nO_LSTTreatment abbreviationsC_LSTECT = Electroconvulsive Therapy O_LIBF ECT = bifrontal ECT (1)\nC_LIO_LIBT ECT = bitemporal ECT (2)\nC_LIO_LIRUL ECT= right unilateral ECT O_LIH-RUL = high-dose RUL ECT (3)\nC_LIO_LILM-RUL = low to moderate-dose RUL ECT (4)\nC_LI\nC_LI\nrTMS = repetitive Transcranial Magnetic Stimulation O_LIHF-L rTMS = high-frequency rTMS of the left DLPFC (5)\nC_LIO_LIHF-R rTMS = high-frequency rTMS of the right DLPFC (6)\nC_LIO_LILF-R rTMS = low-frequency rTMS of the right DLPFC (7)\nC_LIO_LILF-L rTMS = low-frequency rTMS of the left DLPFC (8)\nC_LIO_LIBL rTMS = bilateral rTMS of the DLPFC (9)\nC_LI\ndTMS = deep Transcranial Magnetic Stimulation (10)\npTMS = priming Transcranial Magnetic Stimulation (11)\naTMS = accelerated Transcranial Magnetic Stimulation (12)\nsTMS = synchronised Transcranial Magnetic Stimulation (13)\nTBS = Theta Burst Stimulation O_LIiTBS = intermittent TBS of the left DLPFC (14)\nC_LIO_LIcTBS = continuous TBS of the right DLPFC (15)\nC_LIO_LIblTBS = bilateral TBS of the DLPFC (16)\nC_LI\nMST = Magnetic Seizure Therapy (17)\ntDCS = transcranial Direct Current Stimulation (18)\n\n\nKey points\n\nQuestion: What is the clinical efficacy and acceptability of non-surgical brain stimulation protocols for the acute treatment of major depressive episodes in adults?\n\nFindings: In this network meta-analysis, 10 out of 18 treatment protocols were associated with higher response rates relative to sham, most notably bitemporal and high-dose right unilateral electroconvulsive therapy. All treatment protocols were at least as acceptable as sham treatment.\n\nMeaning: Non-surgical brain stimulation techniques constitute viable alternative or add-on treatment strategies for adult patients with major depressive episodes.

neuroscience↗

bioRxiv2015

Sequential Multiple Assignment Randomized Trials: An Opportunity for Improved Design of Stroke Reperfusion Trials

Background: Modern clinical trials in stroke reperfusion fall into two categories: alternative systemic pharmacological regimens to alteplase and \"rescue\" endovascular approaches using targeted thrombectomy devices and/or medications delivered directly for persistently vessel occlusions. Clinical trials in stroke have not evaluated how initial pharmacological thrombolytic management might influence subsequent rescue strategy. A sequential multiple assignment randomized trial (SMART) is a novel trial design that can test these dynamic treatment regimens and lead to treatment guidelines which more closely mimic practice.\n\nAim: To characterize a SMART design in comparison to traditional approaches for stroke reperfusion trials.\n\nMethods: We conducted a numerical simulation study that evaluated the performance of contrasting acute stroke clinical trial designs of both initial reperfusion and rescue therapy. We compare a SMART design where the same patients are followed through initial reperfusion and rescue therapy within one trial to a standard phase III design comparing two reperfusion treatments and a separate phase II futility design of rescue therapy in terms of sample size, power, and ability to address particular research questions.\n\nResults: Traditional trial designs can be well powered and have optimal design characteristics for independent treatment effects. When treatments, such as the reperfusion and rescue therapies, may interact, commonly used designs fail to detect this. A SMART design, with similar sample size to standard designs, can detect treatment interactions.\n\nConclusions: The use of SMART designs to investigate effective and realistic dynamic treatment regimens is a promising way to accelerate the discovery of new, effective treatments for stroke.

Scientific Communication and Education↗

bioRxiv2015

Phase i trials in melanoma: A framework to translate preclinical findings to the clinic

AbstractWe present a, mathematical model driven, framework to implement virtual or imaginary clinical trials (phase i trials) that can be used to bridge the gap between preclinical studies and the clinic. The trial implementation process includes the development of an experimentally validated mathematical model, generation of a cohort of heterogeneous virtual patients, an assessment of stratification factors, and optimization of treatment strategy. We show the detailed process through application to melanoma treatment, using a combination therapy of chemotherapy and an AKT inhibitor, which was recently tested in a phase 1 clinical trial. We developed a mathematical model, composed of ordinary differential equations, based on experimental data showing that such therapies differentially induce autophagy in melanoma cells. Model parameters were estimated using an optimization algorithm that minimizes differences between predicted cell populations and experimentally measured cell numbers. The calibrated model was validated by comparing predicted cell populations with experimentally measured melanoma cell populations in twelve different treatment scheduling conditions. By using this validated model as the foundation for a genetic algorithm, we generated a cohort of virtual patients that mimics the heterogeneous combination therapy responses observed in a companion clinical trial. Sensitivity analysis of this cohort defined parameters that discriminated virtual patients having more favorable versus less favorable outcomes. Finally, the model predicts optimal therapeutic approaches across all virtual patients.\n\nOne Sentence SummaryWe propose a computational framework to implement phase i trials (virtual/imaginary yet informed clinical trials) in cancer, using an experimentally calibrated mathematical model of melanoma combination therapy, that can readily capture observed heterogeneous clinical outcomes and be used to optimize future clinical trial design.

Cancer Biology↗

bioRxiv2018

Cell-free DNA profiling of metastatic prostate cancer reveals microsatellite instability, structural rearrangements and clonal hematopoiesis

BackgroundThere are multiple existing and emerging therapeutic avenues for metastatic prostate cancer, with a common denominator, which is the need for predictive biomarkers. Circulating tumor DNA (ctDNA) has the potential to cost-efficiently accelerate precision medicine trials to improve clinical efficacy and diminish costs and toxicity. However, comprehensive ctDNA profiling in metastatic prostate cancer to date has been limited.\n\nMethodsA combination of targeted- and low-pass whole genome sequencing was performed on plasma cell-free DNA and matched white blood cell germline DNA in 364 blood samples from 217 metastatic prostate cancer patients.\n\nResultsctDNA was detected in 85.9% of baseline samples, correlated to line of therapy and was mirrored by circulating tumor cell enumeration of synchronous blood samples. Comprehensive profiling of the androgen receptor (AR) revealed a continuous increase in the fraction of patients with intra-AR structural variation, from 15.4% during first line mCRPC therapy to 45.2% in fourth line, indicating a continuous evolution of AR during the course of the disease. Patients displayed frequent alterations in DNA repair deficiency genes (18.0%). Additionally, the microsatellite instability phenotype was identified in 3.81% of eligible samples ([≥]0.1 ctDNA fraction). Sequencing of non-repetitive intronic- and exonic regions of PTEN, RB1 and TP53 detected biallelic inactivation in 47.5%, 20.3% and 44.1% of samples with [≥]0.2 ctDNA fraction, respectively. Only one patient carried a clonal high-impact variant without a detectable second hit. Intronic high-impact structural variation was twice as common as exonic mutations in PTEN and RB1. Finally, 14.6% of patients presented false positive variants due to clonal hematopoiesis, commonly ignored in commercially available assays.\n\nConclusionsctDNA profiles appear to mirror the genomic landscape of metastatic prostate cancer tissue and may cost-efficiently provide somatic information in clinical trials designed to identify predictive biomarkers. However, intronic sequencing of the interrogated tumor suppressors challenge the ubiquitous focus on coding regions and is vital, together with profiling of synchronous white blood cells, to minimize erroneous assignments which in turn may confound results and impede true associations in clinical trials.

genomics↗

bioRxiv2018

Postmarketing commitments for novel drugs and biologics approved by the US Food and Drug Administration: a cross-sectional analysis

BackgroundPostmarketing commitments are clinical studies that drug sponsors agree to conduct at the time of FDA approval, but which are not required by statute or regulation. The objective of this study was to determine the characteristics, completion, and dissemination of postmarketing commitments agreed upon by sponsors at first FDA approval.\n\nMethodsWe performed a cross-sectional analysis of postmarketing commitments for new drugs and biologics approved 2009-2012. Using public FDA documents, ClinicalTrials.gov, and Scopus, we determined postmarketing commitments and their characteristics known at the time of FDA approval; number of postmarketing commitments subject to reporting requirements, for which FDA is required to make study status information available to the public (\"506B studies\"), and their statuses; and rates of registration and results reporting on ClinicalTrials.gov and publication in peer-reviewed journals for all clinical trials, with follow-up through July 2018.\n\nResultsAmong 110 novel drugs and biologics approved by the FDA between 2009-2012, 61 (55.5%) had at least one postmarketing commitment at the time of first approval. Of 331 total postmarketing commitments, 271 (81.9%) were non-human subjects research, predominantly chemistry, manufacturing, and controls studies; 49 (14.8%) were clinical trials (33 new and 16 ongoing trials for which follow-up results would be reported). Study descriptions for the new clinical trials often lacked information to establish study design features. Of the 89 (26.9%) 506B studies subject to public reporting requirements, of which 42 were clinical trials, 59 (66.3%) did not have an up-to-date status provided by FDA. Nearly all new clinical trials (28 of 31, 90.3%) were registered on ClinicalTrials.gov; of the 23 registered trials that were completed or terminated, 22 (95.7%) had reported results. Only half (14 of 29, 48.3%) of completed or terminated clinical trials, registered or unregistered, were published in peer-reviewed journals. Conclusions: The majority of postmarketing commitments agreed to by sponsors at the time of FDA approval for novel drugs and biologics approved between 2009-2012 were chemistry, manufacturing, and controls studies. While only 15% were clinical trials, these trials were nearly always registered with reported results on ClinicalTrials.gov. However, despite FDA public reporting requirements, up-to-date study status information was often unavailable for 506B studies.

epidemiology↗

bioRxiv2017

Systematic interrogation of diverse Omic data reveals interpretable, robust, and generalizable transcriptomic features of clinically successful therapeutic targets

Target selection is the first and pivotal step in drug discovery. An incorrect choice may not manifest itself for many years after hundreds of millions of research dollars have been spent. We collected a set of 332 targets that succeeded or failed in phase III clinical trials, and explored whether Omic features describing the target genes could predict clinical success. We obtained features from the recently published comprehensive resource: Harmonizome. Nineteen features appeared to be significantly correlated with phase III clinical trial outcomes, but only 4 passed validation schemes that used bootstrapping or modified permutation tests to assess feature robustness and generalizability while accounting for target class selection bias. We also used classifiers to perform multivariate feature selection and found that classifiers with a single feature performed as well in cross-validation as classifiers with more features (AUROC=0.57 and AUPR=0.81). The two predominantly selected features were mean mRNA expression across tissues and standard deviation of expression across tissues, where successful targets tended to have lower mean expression and higher expression variance than failed targets. This finding supports the conventional wisdom that it is favorable for a target to be present in the tissue(s) affected by a disease and absent from other tissues. Overall, our results suggest that it is feasible to construct a model integrating interpretable target features to inform target selection. We anticipate deeper insights and better models in the future, as researchers can reuse the data we have provided to improve methods for handling sample biases and learn more informative features. Code, documentation, and data for this study have been deposited on GitHub at https://github.com/arouillard/omic-features-successful-targets.\n\nAUTHOR SUMMARYDrug discovery often begins with a hypothesis that changing the abundance or activity of a target--a biological molecule, usually a protein--will cure a disease or ameliorate its symptoms. Whether a target hypothesis translates into a successful therapy depends in part on the characteristics of the target, but it is not completely understood which target characteristics are important for success. We sought to answer this question with a supervised machine learning approach. We obtained outcomes of target hypotheses tested in clinical trials, scoring targets as successful or failed, and then obtained thousands of features (i.e. properties or characteristics) of targets from dozens of biological datasets. We statistically tested which features differed between successful and failed targets, and built a computational model that used these features to predict success or failure of targets in clinical trials. We found that successful targets tended to have more variable mRNA abundance from tissue to tissue and lower average abundance across tissues than failed targets. Thus, it is probably favorable for a target to be present in the tissue(s) affected by a disease and absent from other tissues. Our work demonstrates the feasibility of predicting clinical trial outcomes from target features.

bioinformatics↗

bioRxiv2018

Reverse Engineered Virtual Patient Populations as Surrogates for Real Patient-Level Data

ObjectivesTo demonstrate a new method for generating virtual, individual-level data by testing it on a known clinical trial population. DesignVirtualization of aggregate data from a clinical trial. SettingVirtual Participants936,100 virtual patients InterventionsNone Main Outcomes MeasuresOdds ratios for adverse outcomes in virtual patient populations compared to clinical trial participants. MethodsThe replicate engineered virtual patient populations (RE-ViPPs) method, based on aggregate cross-tabulated categorical population data, does not require access to individual-level data. Using sequential regression combined with randomization, it generates virtual individual patients to comprise populations that, on average, closely resemble the real population in question. The method is validated by applying it to aggregated data from the seminal SPRINT trial, which compared intensive versus standard blood pressure treatment goals on major adverse cardiovascular events. ResultsThe method yields virtual populations, each with 9361 patients, faithfully mimicking the real SPRINT participants. Multiple logistic regression on 100 such populations shows that factors with the highest odds ratios for the primary event are, in descending order, past clinical cardiovascular disease, age [≥] 75, chronic kidney disease, high non-HDL, and smoking history. Intensive blood pressure treatment, the trials intervention, had an odds ratio of 0.74 [0.63-0.87]. On all these measures, the 100 RE-ViPPs mirrored the real SPRINT participants, including the intensive therapy result (actual SPRINT odds ratio: 0.74 [0.62-0.88]). ConclusionsClinical data dissemination has limitations. The most coveted data is descriptive at the individual level but comes with significant cost, effort, and time. There is potential for privacy breaches, and the open-data movement has progressed slowly due to data-ownership concerns. RE-ViPPs closely matched the true SPRINT population. Applied to trials, registries, and databases, RE-ViPPs could reduce open-data burdens by encouraging dissemination of aggregate cross-tabulated real data that allow investigators to generate and measure virtual patients.

bioinformatics↗

bioRxiv2018

Generating perfusion maps from structural optical coherence tomography with artificial intelligence

Despite advances in artificial intelligence (AI), its application in medical imaging has been burdened and limited by expert-generated labels. We used images from optical coherence tomography angiography (OCTA), a relatively new imaging modality that measures retinal blood flow, to train an AI algorithm to generate flow maps from standard optical coherence tomography (OCT) images, exceeding the ability and bypassing the need for expert labeling. Deep learning was able to infer flow from single structural OCT images with similar fidelity to OCTA and significantly better than expert clinicians (P < 0.00001). Our model allows generating flow maps from large volumes of previously collected OCT data in existing clinical trials and clinical practice. This finding demonstrates a novel application of AI to medical imaging, whereby subtle regularities between different modalities are used to image the same body part and AI is used to generate detailed inferences of tissue function from structure imaging.

neuroscience↗

bioRxiv2016

Designing fecal microbiota transplant trials that account for differences in donor stool efficacy

1Fecal microbiota transplantation (FMT) is a highly effective intervention for patients suffering from recurrent Clostridium difficile, a common hospital-acquired infection. FMTs success as a therapy for C. difficile has inspired interest in performing clinical trials that experiment with FMT as a therapy for other conditions like inflammatory bowel disease, obesity, diabetes, and Parkinsons disease. Results from clinical trials that use FMT to treat inflammatory bowel disease suggest that, for at least one condition beyond C. difficile, most FMT donors produce stool that is not efficacious. The optimal strategies for identifying and using efficacious donors have not been investigated. We therefore examined the optimal Bayesian response-adaptive strategy for allocating patients to donors and formulated a computationally-tractable myopic heuristic. This heuristic computes the probability that a donor is efficacious by updating prior expectations about the efficacy of FMT, the placebo rate, and the fraction of donors that produce efficacious stool. In simulations designed to mimic a recent FMT clinical trial, for which traditional power calculations predict ~100% statistical power, we found that accounting for differences in donor stool efficacy reduced the predicted statistical power to ~9%. For these simulations, using the heuristic Bayesian allocation strategy more than quadrupled the statistical power to ~39%. We use the results of similar simulations to make recommendations about the number of patients, number of donors, and choice of clinical endpoint that clinical trials should use to optimize their ability to detect if FMT is effective for treating a condition.

Bioinformatics↗

bioRxiv2016

A community-based collaboration to build prediction models for short-term discontinuation of docetaxel in metastatic castration-resistant prostate cancer patients

BackgroundDocetaxel has a demonstrated survival benefit for metastatic castration-resistant prostate cancer (mCRPC). However, 10-20% of patients discontinue docetaxel prematurely because of toxicity-induced adverse events, and managing risk factors for toxicity remains an ongoing challenge for health care providers and patients. Prospective identification of high-risk patients for early discontinuation has the potential to assist clinical decision-making and can improve the design of more efficient clinical trials. In partnership with Project Data Sphere (PDS), a non-profit initiative facilitating clinical trial data-sharing, we designed an open-data, crowdsourced DREAM (Dialogue for Reverse Engineering Assessments and Methods) Challenge for developing models to predict early discontinuation of docetaxel\n\nMethodsData from the comparator arms of four phase III clinical trials in first-line mCRPC were obtained from PDS, including 476 patients treated with docetaxel and prednisone from the ASCENT2 trial, 598 patients treated with docetaxel, prednisone/prednisolone, and placebo in the VENICE trial, 526 patients treated with docetaxel, prednisone, and placebo in the MAINSAIL trial, and 528 patients treated with docetaxel and placebo in the ENTHUSE 33 trial. Early discontinuation was defined as treatment stoppage within three months due to adverse treatment effects. Over 150 clinical features including laboratory values, medical history, lesion measures, prior treatment, and demographic variables were curated and made freely available for model building for all four trials. The ASCENT2, VENICE, and MAINSAIL trial data sets formed the training set that also included patient discontinuation status. The ENTHUSE 33 trial, with patient discontinuation status hidden, was used as an independent validation set to evaluate model performance. Prediction performance was assessed using area under the precision-recall curve (AUPRC) and the Bayes factor was used to compare the performance between prediction models.\n\nResultsThe frequency of early discontinuation was similar between training (ASCENT2, VENICE, and MAINSAIL) and validation (ENTHUSE 33) sets, 12.3% versus 10.4% of docetaxel-treated patients, respectively. In total, 34 independent teams submitted predictions from 61 different models. AUPRC ranged from 0.088 to 0.178 across submissions with a random model performance of 0.104. Seven models with comparable AUPRC scores (Bayes factor [≤]; 3) were observed to outperform all other models. A post-challenge analysis of risk predictions generated by these seven models revealed three distinct patient subgroups: patients consistently predicted to be at high-risk or low-risk for early discontinuation and those with discordant risk predictions. Early discontinuation events were two-times higher in the high-versus low-risk subgroup and baseline clinical features such as presence/absence of metastatic liver lesions, and prior treatment with analgesics and ACE inhibitors exhibited statistically significant differences between the high- and low-risk subgroups (adjusted P < 0.05). An ensemble-based model constructed from a post-Challenge community collaboration resulted in the best overall prediction performance (AUPRC = 0.230) and represented a marked improvement over any individual Challenge submission. A\n\nFindingsOur results demonstrate that routinely collected clinical features can be used to prospectively inform clinicians of mCRPC patients risk to discontinue docetaxel treatment early due to adverse events and to the best of our knowledge is the first to establish performance benchmarks in this area. This work also underscores the \"wisdom of crowds\" approach by demonstrating that improved prediction of patient outcomes is obtainable by combining methods across an extended community. These findings were made possible because data from separate trials were made publicly available and centrally compiled through PDS.

bioinformatics↗

bioRxiv2018

Effects and Feasibility of Hyperthermic Baths for Patients with Depressive Disorder: A Randomized Controlled Clinical Pilot Trial

BackgroundEvaluation of efficacy, safety and feasibility of hyperthermic baths (HTB; head-out-of-water-immersion in 40{degrees}C), twice a week, compared to a physical exercise program (PEP; moderate intensity aerobic exercises) in moderate to severe depression.\n\nMethodSingle-site, open-label randomized controlled 8-week parallel-group pilot study at an university outpatient clinic as part of usual depression care. Medically stable outpatients with depressive disorder (ICD-10: F32/F33) as determined by the 17-item Hamilton Depression Rating Scale (HAM-D) score [≥]18 and a score [≥]2 on item 1 (Depressed Mood) were randomly assigned to receive either two sessions of HTB or PEP per week (40-45 min) provided by two trained doctoral students. An independent biometric center used computer-generated tables to allocate treatments. Primary outcome measure was the change in HAM-D total score from baseline (T0) to the 2-week time point (T1). Linear regression analyses, adjusted for baseline values, were performed to estimate intervention effects on an intention-to-treat (ITT) principle.\n\nFindings45 patients (HTB n = 22; PEP n = 23) were randomized and analyzed according to ITT (mean age = 48.4 years, SD = 11.3, mean HAM-D score = 21.7, SD = 3.2). Baseline-adjusted mean difference was 4.3 points in the HAM-D score in favor of HTB (p<0.001). This improvement was achieved after two weeks. Compliance with the intervention and follow-up was far better in the HTB group (2 vs 13 dropouts). There were no treatment-related serious adverse events. Main limitation: the number of dropouts in the PEP group (13 of 23) was far higher than in other trials investigating exercise in depression (18.1 % dropouts).\n\nConclusionsHTB seems to be a fast-acting, safe and easy accessible method leading to clinically relevant improvement in depressive disorder after two weeks; it is also suitable for persons who have problems performing exercise training.\n\nTrial registrationGerman Clinical Trials Register (DRKS) with the registration number DRKS00011013 (registration date 2016-09-19) before onset of the study.

clinical trials↗

bioRxiv2017

A Low-Cost Multiplex Biomarker Assay Stratifies Colorectal Cancer Patient Samples into Clinically-Relevant Subtypes

Previously, we classified colorectal cancers (CRCs) into five CRCA subtypes with different prognoses and potential treatment responses, using a 786-gene signature. We merged our subtypes and those described by five other groups into four consensus molecular subtypes (CMS) that are similar to CRCA subtypes. Here we demonstrate the analytical development and application of a custom NanoString platform-based biomarker assay to stratify CRC into subtypes. To reduce costs, we switched from the standard protocol to a custom modified protocol (NanoCRCA) with a high Pearson correlation coefficient (>0.88) between protocols. Technical replicates were highly correlated (>0.96). The assay included a reduced robust 38-gene panel from the 786-gene signature that was selected using an in-laboratory developed computational pipeline of class prediction methods. We applied our NanoCRCA assay to untreated CRCs including fresh-frozen and formalin-fixed paraffin-embedded (FFPE) samples (n=81) with matched microarray or RNA-Seq profiles. We further compared the assay results with CMS classification, different platforms (microarrays/RNA-Seq) and gene-set classifiers (38 and 786 genes). NanoCRCA classified fresh-frozen samples (n=39; not including those showing a mixture of subtypes) into all five CRCA subtypes with overall high concordance across platforms (89.7%) and with CMS subtypes (84.6%), independent of tumour cellularity. This analytical validation of the assay shows the association of subtypes with their known molecular, mutational and clinical characteristics. Overall, our modified NanoCRCA assay with further clinical assessment may facilitate prospective validation of CRC subtypes in clinical trials and beyond.\n\nNovelty and ImpactWe previously identified five gene expression-based CRC subtypes with prognostic and potential predictive differences using a 786-gene signature and microarray platform. Subtype-driven clinical trials require a validated assay suitable for routine clinical use. This study demonstrates, for the first time, how molecular CRCA subtype can be detected using NanoString Technology-based biomarker assay (NanoCRCA) suitable for clinical validation. NanoCRCA is suitable for analysing FFPE samples, and this assay may facilitate patient stratification within clinical trials.

cancer biology↗

bioRxiv2018

Predicting onset, progression, and clinical subtypes of Parkinson disease using machine learning

BackgroundThe clinical manifestations of Parkinson disease are characterized by heterogeneity in age at onset, disease duration, rate of progression, and constellation of motor versus nonmotor features. Due to these variable presentations, counseling of patients about their individual risks and prognosis is limited. There is an unmet need for predictive tests that facilitate early detection and characterization of distinct disease subtypes as well as improved, individualized predictions of the disease course. The emergence of machine learning to detect hidden patterns in complex, multi-dimensional datasets provides unparalleled opportunities to address this critical need.\n\nMethods and FindingsWe used unsupervised and supervised machine learning approaches for subtype identification and prediction. We used machine learning methods on comprehensive, longitudinal clinical data from the Parkinson Disease Progression Marker Initiative (PPMI) (n=328 cases) to identify patient subtypes and to predict disease progression. The resulting models were validated in an independent, clinically well-characterized cohort from the Parkinson Disease Biomarker Program (PDBP) (n=112 cases). Our analysis distinguished three distinct disease subtypes with highly predictable progression rates, corresponding to slow, moderate and fast disease progressors. We achieved highly accurate projections of disease progression four years after initial diagnosis with an average Area Under the Curve of 0.93 (95% CI: 0.96 {+/-} 0.01 for PDvec1, 0.87 {+/-} 0.03 for PDvec2, and 0.96 {+/-} 0.02 for PDvec3). We have demonstrated robust replication of these findings in the independent validation cohort.\n\nConclusionsThese data-driven results enable clinicians to deconstruct the heterogeneity within their patient cohorts. This knowledge could have immediate implications for clinical trials by improving the detection of significant clinical outcomes that might have been masked by cohort heterogeneity. We anticipate that machine learning models will improve patient counseling, clinical trial design, allocation of healthcare resources and ultimately individualized clinical care.

neuroscience↗

bioRxiv2017

Large-scale phenome-wide association study of PCSK9 loss-of-function variants demonstrates protection against ischemic stroke

PCSK9 inhibitors are a potent new therapy for hypercholesterolemia and have been shown to decrease risk of coronary heart disease. Although short-term clinical trial results have not demonstrated major adverse effects, long-term data will not be available for some time. Genetic studies in large well-phenotyped biobanks offer a unique opportunity to predict drug effects and provide context for the evaluation of future clinical trial outcomes. We tested association of the PCSK9 loss-of-function variant rsll591147 (R46L) in a hypothesis-driven 11 phenotype set and a hypothesis-generating 278 phenotype set in 337,536 individuals of British ancestry in the United Kingdom Biobank (UKB), with independent discovery (n = 225K) and replication (n = 112K). In addition to the known association with lipid levels (OR 0.63) and coronary heart disease (OR 0.73), the T allele of rs11591147 showed a protective effect on ischemic stroke (OR 0.61, p = 0.002) but not hemorrhagic stroke in the hypothesis-driven screen. We did not observe an association with type 2 diabetes, cataracts, heart failure, atrial fibrillation, and cognitive dysfunction. In the phenome-wide screen, the variant was associated with a reduction in metabolic disorders, ischemic heart disease, coronary artery bypass graft operations, percutaneous coronary interventions and history of angina. A single variant analysis of UKB data using TreeWAS, a Bayesian analysis framework to study genetic associations leveraging phenotype correlations, also showed evidence of association with cerebral infarction and vascular occlusion. This result represents the first genetic evidence in a large cohort for the protective effect of PCSK9 inhibition on ischemic stroke, and corroborates exploratory evidence from clinical trials. PCSK9 inhibition was not associated with variables other than those related to low density lipoprotein cholesterol and atherosclerosis, suggesting that other effects are either small or absent.

genetics↗