Artificial intelligence in oncology: From bench to clinic.

In the past few years, Artificial Intelligence (AI) techniques have been applied to almost every facet of oncology, from basic research to drug development and clinical care. In the clinical arena where AI has perhaps received the most attention, AI is showing promise in enhancing and automating image-based diagnostic approaches in fields such as radiology and pathology. Robust AI applications, which retain high performance and reproducibility over multiple datasets, extend from predicting indications for drug development to improving clinical decision support using electronic health record data. In this article, we review some of these advances. We also introduce common concepts and fundamentals of AI and its various uses, along with its caveats, to provide an overview of the opportunities and challenges in the field of oncology. Leveraging AI techniques productively to provide better care throughout a patient's medical journey can fuel the predictive promise of precision medicine.

A machine learning and network framework to discover new indications for small molecules.

Drug repurposing, identifying novel indications for drugs, bypasses common drug development pitfalls to ultimately deliver therapies to patients faster. However, most repurposing discoveries have been led by anecdotal observations (e.g. Viagra) or experimental-based repurposing screens, which are costly, time-consuming, and imprecise. Recently, more systematic computational approaches have been proposed, however these rely on utilizing the information from the diseases a drug is already approved to treat. This inherently limits the algorithms, making them unusable for investigational molecules. Here, we present a computational approach to drug repurposing, CATNIP, that requires only biological and chemical information of a molecule. CATNIP is trained with 2,576 diverse small molecules and uses 16 different drug similarity features, such as structural, target, or pathway based similarity. This model obtains significant predictive power (AUC = 0.841). Using our model, we created a repurposing network to identify broad scale repurposing opportunities between drug types. By exploiting this network, we identified literature-supported repurposing candidates, such as the use of systemic hormonal preparations for the treatment of respiratory illnesses. Furthermore, we demonstrated that we can use our approach to identify novel uses for defined drug classes. We found that adrenergic uptake inhibitors, specifically amitriptyline and trimipramine, could be potential therapies for Parkinson's disease. Additionally, using CATNIP, we predicted the kinase inhibitor, vandetanib, as a possible treatment for Type 2 Diabetes. Overall, this systematic approach to drug repurposing lays the groundwork to streamline future drug development efforts.

Distribution of manganese and other biometals in flatiron mice.

Flatiron (ffe) mice display features of "ferroportin disease" or Type IV hereditary hemochromatosis. While it is known that both Fe and Mn metabolism are impaired in flatiron mice, the effects of ferroportin (Fpn) deficiency on physiological distribution of these and other biometals is unknown. We hypothesized that Fe, Mn, Zn and/or Cu distribution would be altered in ffe/+ compared to wild-type (+/+) mice. ICP-MS analysis showed that Mn, Zn and Cu levels were significantly reduced in femurs from ffe/+ mice. Bone deposits reflect metal accumulation, therefore these data indicate that Mn, Zn and Cu metabolism are affected by Fpn deficiency. The observations that muscle Cu, lung Mn, and kidney Cu and Zn levels were reduced in ffe/+ mice support the idea that metal metabolism is impaired. While all four biometals appeared to accumulate in brains of flatiron mice, significant gender effects were observed for Mn and Zn levels in male ffe/+ mice. Metals were higher in olfactory bulbs of ffe/+ mice regardless of gender. To further study brain metal distribution, (54)MnCl2 was administered by intravenous injection and total brain (54)Mn was measured over time. At 72 h, (54)Mn was significantly greater in brains of ffe/+ mice compared to +/+ mice while blood (54)Mn was cleared to the same levels by 24 h. Taken together, these results indicate that Fpn deficiency decreases Mn trafficking out of the brain, alters body Fe, Mn, Zn and Cu levels, and promotes metal accumulation in olfactory bulbs.

The Missing Pieces of Artificial Intelligence in Medicine.

Stakeholders across the entire healthcare chain are looking to incorporate artificial intelligence (AI) into their decision-making process. From early-stage drug discovery to clinical decision support systems, we have seen examples of how AI can improve efficiency and decrease costs. In this Opinion, we discuss some of the key factors that should be prioritized to enable the successful integration of AI across the healthcare value chain. In particular, we believe a focus on model interpretability is crucial to obtain a deeper understanding of the underlying biological mechanisms and guide further investigations. Additionally, we discuss the importance of integrating diverse types of data within any AI framework to limit bias, increase accuracy, and model the interdisciplinary nature of medicine. We believe that widespread adoption of these practices will help accelerate the continued integration of AI into our current healthcare framework.

Localized immunotherapy via liposome-anchored Anti-CD137 + IL-2 prevents lethal toxicity and elicits local and systemic antitumor immunity.

Immunostimulatory agonists such as anti-CD137 and interleukin (IL)-2 have elicited potent antitumor immune responses in preclinical studies, but their clinical use is limited by inflammatory toxicities that result upon systemic administration. We hypothesized that by rigorously restricting the biodistribution of immunotherapeutic agents to a locally accessible lesion and draining lymph node(s), effective local and systemic antitumor immunity could be achieved in the absence of systemic toxicity. We anchored anti-CD137 and an engineered IL-2Fc fusion protein to the surfaces of PEGylated liposomes, whose physical size permitted dissemination in the tumor parenchyma and tumor-draining lymph nodes but blocked entry into the systemic circulation following intratumoral injection. In the B16F10 melanoma model, intratumoral liposome-coupled anti-CD137 + IL-2Fc therapy cured a majority of established primary tumors while avoiding the lethal inflammatory toxicities caused by equivalent intratumoral doses of soluble immunotherapy. Immunoliposome therapy induced protective antitumor memory and elicited systemic antitumor immunity that significantly inhibited the growth of simultaneously established distal tumors. Tumor inhibition was CD8(+) T-cell-dependent and was associated with increased CD8(+) T-cell infiltration in both treated and distal tumors, enhanced activation of tumor antigen-specific T cells in draining lymph nodes, and a reduction in regulatory T cells in treated tumors. These data suggest that local nanoparticle-anchored delivery of immuno-agonists represents a promising strategy to improve the therapeutic window and clinical applicability of highly potent but otherwise intolerable regimens of cancer immunotherapy. Cancer Res; 73(5); 1547-58. ©2012 AACR.

Generation of effector memory T cell-based mucosal and systemic immunity with pulmonary nanoparticle vaccination.

Many pathogens infiltrate the body and initiate infection via mucosal surfaces. Hence, eliciting cellular immune responses at mucosal portals of entry is of great interest for vaccine development against mucosal pathogens. We describe a pulmonary vaccination strategy combining Toll-like receptor (TLR) agonists with antigen-carrying lipid nanocapsules [interbilayer-crosslinked multilamellar vesicles (ICMVs)], which elicit high-frequency, long-lived, antigen-specific effector memory T cell responses at multiple mucosal sites. Pulmonary immunization using protein- or peptide-loaded ICMVs combined with two TLR agonists, polyinosinic-polycytidylic acid (polyI:C) and monophosphoryl lipid A, was safe and well tolerated in mice, and led to increased antigen transport to draining lymph nodes compared to equivalent subcutaneous vaccination. This response was mediated by the vast number of antigen-presenting cells (APCs) in the lungs. Nanocapsules primed 13-fold more T cells than did equivalent soluble vaccines, elicited increased expression of mucosal homing integrin α4ß7⁺, and generated long-lived T cells in both the lungs and distal (for example, vaginal) mucosa strongly biased toward an effector memory (T(EM)) phenotype. These T(EM) responses were highly protective in both therapeutic tumor and prophylactic viral vaccine settings. Together, these data suggest that targeting cross-presentation-promoting particulate vaccines to the APC-rich pulmonary mucosa can promote robust T cell responses for protection of mucosal surfaces.