Human and Machine Intelligence Together Drive Drug Repurposing in Rare Diseases.

Repurposing is an increasingly attractive method within the field of drug development for its efficiency at identifying new therapeutic opportunities among approved drugs at greatly reduced cost and time of more traditional methods. Repurposing has generated significant interest in the realm of rare disease treatment as an innovative strategy for finding ways to manage these complex conditions. The selection of which agents should be tested in which conditions is currently informed by both human and machine discovery, yet the appropriate balance between these approaches, including the role of artificial intelligence (AI), remains a significant topic of discussion in drug discovery for rare diseases and other conditions. Our drug repurposing team at Vanderbilt University Medical Center synergizes machine learning techniques like phenome-wide association study-a powerful regression method for generating hypotheses about new indications for an approved drug-with the knowledge and creativity of scientific, legal, and clinical domain experts. While our computational approaches generate drug repurposing hits with a high probability of success in a clinical trial, human knowledge remains essential for the hypothesis creation, interpretation, "go-no go" decisions with which machines continue to struggle. Here, we reflect on our experience synergizing AI and human knowledge toward realizable patient outcomes, providing case studies from our portfolio that inform how we balance human knowledge and machine intelligence for drug repurposing in rare disease.

Leveraging Human Genetics to Identify Safety Signals Prior to Drug Marketing Approval and Clinical Use.

INTRODUCTION: When a new drug or biologic product enters the market, its full spectrum of side effects is not yet fully understood, as use in the real world often uncovers nuances not suggested within the relatively narrow confines of preapproval preclinical and trial work. OBJECTIVE: We describe a new, phenome-wide association study (PheWAS)- and evidence-based approach for detection of potential adverse drug effects. METHODS: We leveraged our established platform, which integrates human genetic data with associated phenotypes in electronic health records from 29,722 patients of European ancestry, to identify gene-phenotype associations that may represent known safety issues. We examined PheWAS data and the published literature for 16 genes, each of which encodes a protein targeted by at least one drug or biologic product. RESULTS: Initial data demonstrated that our novel approach (safety ascertainment using PheWAS [SA-PheWAS]) can replicate published safety information across multiple drug classes, with validated findings for 13 of 16 gene-drug class pairs. CONCLUSIONS: By connecting and integrating in vivo and in silico data, SA-PheWAS offers an opportunity to supplement current methods for predicting or confirming safety signals associated with therapeutic agents.

Topical versus systemic antimicrobial therapy for treating mildly infected diabetic foot ulcers: a randomized, controlled, double-blinded, multicenter trial of pexiganan cream.

BACKGROUND: Topical antimicrobial therapy of infected diabetic foot ulcers can focus on the wound and avoid the adverse effects of systemic anti-infective agents. We compared the efficacy of outpatient treatment using an investigational topical antimicrobial peptide, pexiganan acetate cream, with the efficacy of systemic therapy using an oral fluoroquinolone antibiotic, ofloxacin, for mildly infected diabetic foot ulcers. METHODS: In 2 consecutive, double-blind, controlled trials (study 303 and study 304), we randomized diabetic patients with a mildly infected diabetic foot ulcer to receive the active topical agent or active oral antibiotic, plus a respective inactive placebo. The primary outcome of interest was clinical cure or improvement of the infection. Secondary outcomes included eradication of wound pathogens and wound healing, which was documented by a semiquantitative scoring system. RESULTS: Overall, 835 patients were randomized; those in each treatment arm were similar with regard to demographic and clinical characteristics. Although study 303 failed to demonstrate equivalence, study 304 and the combined data for the 2 trials demonstrated equivalent results (within the 95% confidence interval) for topical pexiganan and oral ofloxacin in clinical improvement rates (85%-90%), overall microbiological eradication rates (42%-47%), and wound healing rates. The incidence of worsening cellulitis (2%-4%) and amputation (2%-3%) did not differ significantly between treatment arms. Bacterial resistance to ofloxacin emerged in some patients who received ofloxacin, but no significant resistance to pexiganan emerged among patients who received pexiganan. CONCLUSIONS: Topical pexiganan might be an effective alternative to oral antibiotic therapy in treating diabetic patients with a mildly infected foot ulcer, and might reduce the risk of selecting antimicrobial-resistant bacteria.

Motivation for Launching a Cancer Metastasis Inhibition (CMI) Program.

Metastatic cancers impose significant burdens on patients, affecting quality of life, morbidity, and mortality. Even during remission, microscopic metastases can lurk, but few therapies directly target tumor cell metastasis. Agents that interfere with this process would represent a new paradigm in cancer management, changing the 'waiting game' into a time of active prevention. These therapies could take multiple forms based on the pathways involved in the metastatic process. For example, a phenome-wide association study showed that a single nucleotide polymorphism in the gene TBXA2R is associated with increased metastasis in multiple primary cancers (P = 0.003), suggesting clinical applicability of TBXA2R antagonists. Emerging data related to the role of platelets in metastasis are concordant with our sense that these pathways present significant opportunities for therapeutic development. However, before real progress can be made toward clinical targeting of the metastatic process, foundational work is needed to define informative measures of critical elements such as circulating tumor cells and tumor DNA, and circulatory vs. lymphatic spread. These challenges require an expansion of team science and composition to obtain competitive funding. At our academic medical center, we have implemented a Cancer Metastasis Inhibition (CMI) program investigating this approach across multiple cancers.

Up-Regulation of Interleukin-9 and the Interleukin-9-Associated Calcium-Activated Chloride Channel hCLCA1 in Nasal Mucosa Following In Vivo Allergen Challenge.

: Interleukin (IL)-9 is a pleiotropic T helper 2-type cytokine that has been shown to be up-regulated in allergic airway disease, including asthma. IL-9 has been demonstrated to be a potent stimulus for the production and secretion of mucus from airway epithelial cells via induction of a calcium-activated chloride channel, hCLCA1. The objective of this study was to investigate the expression of IL-9 and hCLCA1 following allergen challenge in the nasal mucosa of allergic rhinitis patients. Nasal biopsies were obtained from allergic rhinitis patients out of allergen season both before (baseline) and after local antigen challenge with either ragweed or diluent (control). Immunohistochemistry and in situ hybridization were used to assess IL-9 protein and hCLCA1 messenger ribonucleic acid. Eosinophils and T cells were detected using immunohistochemistry. IL-9 and hCLCA1 were very low at baseline, and expression was significantly up-regulated following ragweed challenge. Whereas the number of eosinophils increased after allergen challenge, T-cell counts did not change significantly. The results of this study demonstrate the relationship between specific allergen challenge and expression of both IL-9 and hCLCA1, suggesting a possible mechanism for the increased production of mucus from airway epithelial cells in allergic rhinitis.

When Enough Is Enough: Decision Criteria for Moving a Known Drug into Clinical Testing for a New Indication in the Absence of Preclinical Efficacy Data.

Many animal models of disease are suboptimal in their representation of human diseases and lack of predictive power in the success of pivotal human trials. In the context of repurposing drugs with known human safety, it is sometimes appropriate to conduct the "last experiment first," that is, progressing directly to human investigations. However, there are not accepted criteria for when to proceed straight to humans to test a new indication. We propose a specific set of criteria to guide the decision-making around when to initiate human proof of principle without preclinical efficacy studies in animal models. This approach could accelerate the transition of novel therapeutic approaches to human applications.

Accelerating Precision Drug Development and Drug Repurposing by Leveraging Human Genetics.

The potential impact of using human genetic data linked to longitudinal electronic medical records on drug development is extraordinary; however, the practical application of these data necessitates some organizational innovations. Vanderbilt has created resources such as an easily queried database of >2.6 million de-identified electronic health records linked to BioVU, which is a DNA biobank with more than 230,000 unique samples. To ensure these data are used to maximally benefit and accelerate both de novo drug discovery and drug repurposing efforts, we created the Accelerating Drug Development and Repurposing Incubator, a multidisciplinary think tank of experts in various therapeutic areas within both basic and clinical science as well as experts in legal, business, and other operational domains. The Incubator supports a diverse pipeline of drug indication finding projects, leveraging the natural experiment of human genetics.

A phase I/IIA trial of continuous five-day infusion of squalamine lactate (MSI-1256F) plus carboplatin and paclitaxel in patients with advanced non-small cell lung cancer.

PURPOSE: Squalamine is an antitumor agent that has been shown to have antiangiogenic activity in animal models. This Phase I/IIA study was designed to assess the safety, clinical response, and pharmacokinetics of squalamine when administered as a 5-day continuous infusion in conjunction with standard chemotherapy every 3 weeks in patients with stage IIIB (pleural effusion) or stage IV non-small cell lung cancer. EXPERIMENTAL DESIGN: Patients with chemotherapy-naive non-small cell lung cancer were treated with escalating doses of squalamine in combination with standard doses of paclitaxel and carboplatin. Paclitaxel and carboplatin were administered on day 1, followed by squalamine as a continuous infusion on days 1-5, every 21 days. RESULTS: A total of 45 patients were enrolled (18 patients in the Phase I dose escalation arm and 27 in the Phase IIA arm). The starting dose of squalamine was 100 mg/m(2)/day and escalated to 400 mg/m(2)/day; two of three patients at 400 mg/m(2)/day had dose-limiting toxicity that included grade 3/4 arthralgia, myalgia, and neutropenia. On the basis of safety and toxicity, 300 mg/m(2)/day was selected as the Phase II dose of squalamine in this combination regimen. An additional 27 patients (a total of 33) were enrolled according to the protocol treatment schema at 300 mg/m(2)/day. There was no pharmacokinetic evidence of drug interactions for the combination of squalamine, carboplatin, and paclitaxel. Forty-three patients were evaluable for response. Partial tumor responses were observed in 12 (28%) of these patients; an additional 8 evaluable patients (19%) were reported to have stable disease. For all of the patients treated, the median survival was 10.0 months; and 1-year survival was 40%. CONCLUSIONS: The combination of squalamine given continuously daily for 5 days, with paclitaxel and carboplatin given on day 1, is well tolerated. Patient survival data and the safety profile of this drug combination suggests that the use of squalamine given at its maximum tolerated dose with cytotoxic chemotherapy should be explored further as a potentially effective therapeutic strategy for patients with stage IIIB or IV non-small cell lung cancer.

A Phase I and pharmacokinetic study of squalamine, an aminosterol angiogenesis inhibitor.

PURPOSE: The purpose of this study was to assess the feasibility and characterize the pharmacokinetics of squalamine administered as a continuous i.v. infusion daily for 5 days every 3 weeks. EXPERIMENTAL DESIGN: Patients with advanced solid malignancies were treated with escalating doses of squalamine as a 5-day continuous i.v. infusion every 3 weeks. Doses were initially escalated in 100% increments from a starting dose of 6 mg/m(2)/day, with a single patient treated at each dose level until moderate toxicity was observed, at which time additional patients were treated. RESULTS: Thirty-three patients were treated with 73 courses of squalamine at 13 dose levels ranging from 6 to 700 mg/m(2)/day. Hepatotoxicity, characterized by brief, asymptomatic elevations in transaminases and hyperbilirubinemia, was the principal dose-limiting toxicity of squalamine. At 700 mg/m(2)/day, two of three patients developed grade 4 hyperbilirubinemia, which precluded further dose escalation. At 500 mg/m(2)/day, one of seven patients experienced dose-limiting grade 4 hyperbilirubinemia and grade 3 neurosensory changes, which resolved soon after treatment. Squalamine pharmacokinetics were dose-proportional. At 500 mg/m(2)/day, the mean (percentage coefficient of variation) clearance, half-life, and volume of distribution of squalamine were 2.67 liters/h/m(2) (85%), 9.46 h (81%), and 36.84 liters/m(2) (124%), respectively, and steady-state concentrations [20.08 micro g/ml (13%)] were well above those that inhibit angiogenesis in preclinical models. CONCLUSIONS: At the recommended Phase II dose of 500 mg/m(2)/day, squalamine is well tolerated and results in plasma concentrations at least an order of magnitude higher than those required for prominent antiangiogenic effects in preclinical studies.

Increased expression of the calcium-activated chloride channel hCLCA1 in airways of patients with obstructive chronic bronchitis.

BACKGROUND: Interleukin (IL)-9 and its effect on enhancing the human calcium-activated chloride channel 1 (hCLCA1) expression have been shown to induce mucin production. Increased expression of hCLCA1 may, in turn, contribute to mucus overproduction in chronic obstructive pulmonary disease (COPD) with a chronic bronchitis (CB) phenotype. OBJECTIVE: To determine the expression of IL-9, IL-9 receptor (IL-9R), hCLCA1 and mucoglycoconjugates in COPD. METHODS: Bronchial biopsies were obtained from six patients with obstructive CB and six healthy control subjects. IL-9, IL-9R and hCLCA1 expression were detected using immunohistochemistry. Additionally, in situ hybridization was performed to determine the expression of hCLCA1 messenger RNA. Mucin production was assessed using periodic acid-Schiff staining. RESULTS: There was a significantly higher number of IL-9 immunoreactive cells in the submucosa of patients with COPD than that of healthy control subjects (P<0.05). Also, a significant increase in the expression of IL-9R, hCLCA1 (protein and messenger RNA) and mucin (periodic acid-Schiff-positive cells) was noted in the bronchial epithelium of patients with COPD compared the control subjects (P<0.05). CONCLUSION: Increased expression of IL-9, IL-9R and hCLCA1 in the bronchial mucosa of patients with obstructive CB suggests that mucus overproduction in this disease may be, at least in part, due to hCLCA1.

Increased expression of interleukin-9, interleukin-9 receptor, and the calcium-activated chloride channel hCLCA1 in the upper airways of patients with cystic fibrosis.

OBJECTIVES/HYPOTHESIS: Mucus overproduction is commonly found in airway disease in patients with cystic fibrosis. Interleukin-9 (IL-9) has been shown to mediate airway hyper-responsiveness and mucus overproduction. Recently, the calcium-activated chloride channel hCLCA1 has been described to be upregulated by IL-9 and has been thought to regulate the expression of soluble gel-forming mucins. We sought to examine the expression of IL-9, interleukin-9 receptor (IL-9R), and hCLCA1 in the upper airway of patients with cystic fibrosis in comparison to healthy control subjects and to demonstrate the relationship of IL-9, IL-9R, and hCLCA1 expression with mucus production. STUDY DESIGN: Prospective design. METHODS: Biopsy samples from nasal polyps of four patients with cystic fibrosis, nasal mucosa of six patients with cystic fibrosis, sinus mucosa of eight patients with cystic fibrosis, and nasal mucosa of six healthy control subjects were stained with periodic acid-Schiff (PAS) to identify mucus glycoconjugates. IL-9, IL-9R, and hCLCA1 expression was determined by immunocytochemical study. RESULTS: We demonstrated significant increases in IL-9, IL-9R, and hCLCA1 immunoreactivity in the mucosa of patients with cystic fibrosis compared with that found in control subjects (P <.05). There were no significant differences between the different locations (nasal polyps, nasal mucosa, and sinus mucosa) in the patient group (P >.05). We also observed a significant increase in the number of mucus-producing cells in biopsy specimens from patients with cystic fibrosis in comparison to control subjects. A positive correlation was found between hCLCA1-positive cells and IL-9-positive cells (correlation coefficient [r] = 0.79, P <.05) or IL-9R-positive cells (r = 0.92, P <.05). Moreover, a positive correlation was also present between PAS-positive (mucus-producing) cells and hCLCA1-positive cells (r = 0.64, P <.05) or IL-9R-positive cells (r = 0.64, P <.05). CONCLUSIONS: Increased expression of IL-9 and IL-9R, as well as upregulation of hCLCA1, in mucus-overproducing epithelium of patients with cystic fibrosis supports the hypothesis that IL-9 contributes to mucus overproduction in cystic fibrosis. Expression of hCLCA1 may also be responsible, in part, for the overproduction of mucus. These preliminary findings suggest that hCLCA1 might be an interesting new therapeutic target to control mucus overproduction in airway disease in patients with cystic fibrosis.

Squalamine lactate reduces choroidal neovascularization in a laser-injury model in the rat.

PURPOSE: To determine if systemically administered squalamine lactate, a novel aminosterol with antineoplastic and antiangiogenic activity, inhibits the development of experimental choroidal neovascularization membranes (CNVMs) induced by laser trauma in a rat model. METHODS: Twenty anesthetized male Brown-Norway rats received a series of 8 krypton red laser lesions per eye (647 nm, 0.05 second, 50 microm, 150 mW). One half the animals received an intraperitoneal injection of squalamine and the other one half received an injection of 5% dextrose in water, all performed in a masked fashion. Fundus photography and fluorescein angiography were performed at postlaser treatment days 14 and 28, and ocular tissues were processed for light microscopic examination following euthanasia of the rats on postlaser treatment day 28. RESULTS: Although fundus photography and fluorescein angiography yielded no statistically significant quantitative differences between the two groups, histologic analysis of the lesion sites revealed a partial but statistically significant reduction of experimental CNVM development in the squalamine-treated population. In particular, the squalamine-treated eyes (n = 20) demonstrated lesions (n = 149) with a mean CNVM thickness +/- SD of 47 +/- 11 microm, as compared with the control eyes (n = 20) that had lesions (n = 142) with a mean CNVM thickness +/- SD of 63 +/- 14 microm (P < 0.001). CONCLUSION: Systemically administered squalamine lactate partially reduced choroidal neovascular membrane development induced by laser trauma in this animal model. In conjunction with other existing and developing therapies, this agent may have a potential role in the treatment of human CNVM formation. Further study of squalamine lactate for treatment of neovascular eye disease is warranted.

Effect of squalamine on iris neovascularization in monkeys.

PURPOSE: To investigate the effect of squalamine, an antiangiogenic aminosterol, in an experimental model of iris neovascularization. METHODS: Iris neovascularization was created in cynomolgus monkeys by occluding retinal veins with an argon laser and inducing persistent hypotony with a central corneal suture. Twenty-four eyes were treated in three groups. In Group 1, four eyes were injected intravitreally with 3 microg/0.1 mL squalamine and four eyes with balanced saline solution (controls) immediately after vein occlusion (day 1); injections were repeated every 3 days for 3 weeks. In Group 2, 1 mg/kg squalamine was administered with intravenous infusion in dextrose 5% in four animals; four control animals received only dextrose. Infusions began on day 1 and were repeated every 3 days for 3 weeks. In Group 3, after development of iris neovascularization on day 7, 1 mg/kg squalamine was injected systemically in four animals; four control animals received dextrose 5%. Monkeys were examined by slit-lamp biomicroscopy and underwent color photography and fluorescein angiography. RESULTS: Group 1: All eyes, treated and control, developed intense and persistent rubeosis iridis. Group 2: Two of the four treated eyes in this group developed minimal iris neovascularization; the other two had no iris neovascularization. All four control eyes developed intense, persistent iris neovascularization. Group 3: All eyes developed extensive rubeosis iridis; iris neovascularization regressed in all four treated eyes after squalamine injections. Two of four treated eyes retained minimal iris neovascularization; two showed complete regression of rubeosis iridis. Rubeosis iridis completely regressed in two of the four control eyes; the remaining two control eyes had intense, persistent iris neovascularization. CONCLUSIONS: Intravitreally injected squalamine did not affect the development of iris neovascularization; however, systemic squalamine injection inhibited the development of iris neovascularization and caused partial regression of new vessels in a primate model.