SCIENCE

Despite advances in high-throughput screening, combinatorial chemistry, databanks, and computational models, drug R&D remains expensive and slow—often taking over a decade. Pharma companies spend nearly $90B annually on preclinical research and trials, yet about 90% of drugs that look effective in pre-human studies fail in human trials. A key reason is that many preclinical efficacy methods don’t faithfully recapitulate in vivo microenvironments, driving failed trials and major time and cost burdens.

We developed a patented “Lab-on-a-Brane” (LOB) that better recreates in vivo tissue microenvironments, including barrier and transport functions, enabling organ–capillary interface models. We expanded the platform to support an air–liquid interface (e.g., lung microvasculature), then extended it to a “tumor-train” to model migration and invasion. Finally, we transformed the approach into a scalable, clinically relevant ex vivo Simple Microchamber Array Technology (SMART) that can concurrently assess multiple regimens directly on patient tissue.

This patent covers a human-relevant, ex vivo drug response screening and prediction method that uses viable patient-derived tumor tissue with intact extracellular matrix (ECM) to evaluate cancer therapeutic candidates outside the patient. The workflow emphasizes standardized collection and storage to preserve viability, preparation for analysis, and 3D microarchitecture and tumor microenvironment characterization of both ECM and cellular fractions to inform correlation and engineering of tissue proxies. Therapeutic candidates are then assessed by direct exposure on live or engineered tissue, enabling functional response evaluation in a setting designed to better reflect tumor biology than simplified models. Overall, the method supports a more rigorous, translational path for human-relevant NAMs based drug screening and personalized evaluation while retaining the complexity of intact tumor tissue.

With continued support from BCRFA, this project focuses on characterizing the controlled delivery of multiple isolated fluid streams to spatially distinct regions of intact breast cancer core biopsies within POET®. The work emphasizes rigorous validation of localized exposure, demonstrating that different agents can be applied to defined areas of the same biopsy while preserving native tissue architecture. By enabling spatially resolved, parallel testing within a single patient sample, the study strengthens the scientific basis for multiplexed efficacy assessment in ex vivo breast tumor tissue. This capability also supports more efficient use of limited biopsy material while generating richer, region-specific response data from each specimen.

Simple Microchamber Array Technology (SMART) is our patented biomimetic microchamber array device and set of methods for multiplexed exposure of intact biological samples to an array of fluids in parallel using microchannels and open-top microchambers. The architecture enables simultaneous evaluation of multiple compounds or conditions on a single intact tissue sample, supporting side-by-side comparison while preserving tissue microarchitecture. After exposure, samples can be characterized for response, viability, and related phenotypes to support comparative assessment across conditions. By compressing many test conditions into a single run, SMART is designed to increase throughput, conserve scarce patient material, and accelerate R&D iteration for human-relevant evaluation workflows.

This issued CerFlux patent builds on our Lab-on-a-Brane innovation to deliver a configurable, human-relevant way to model barrier and transport behavior under realistic flow. The invention uses modular cassettes, such as an engineered scaffold positioned between fluidic chambers, integrated into a closed-loop recirculating system where flow and pressure can be precisely tuned to approximate in vivo-like dynamics. Because the cassettes can be scaled and reconfigured, including multi-cassette arrangements, the platform supports a range of R&D workflows where interfaces matter, from transport and PK-style studies to head-to-head compound evaluation in physiologically meaningful conditions. This architecture is designed to help teams generate more decision-ready data by bringing controllable physiology into an experimental format that is practical to deploy and iterate.

In this preprint and accompanying NIH 3D Exchange repository, we address a real-world translational challenge exposed by COVID-19: severe supply-chain disruptions that created critical shortages in personal protective equipment (PPE), even at well-prepared medical centers. The work focused on developing a pragmatic strategy for emergency response manufacturing: instead of 3D-printing entire large volume PPE (often too slow), use additive manufacturing to rapidly produce small, low-cost adapters that repurpose existing components. As a case study, we describe a simple, inexpensive adaptation of elastomeric half-mask respirators for emergency clinical use in high-risk settings, highlighting how fast, locally deployable engineering solutions can help buffer healthcare systems against future supply shocks.

Breast cancer is highly heterogeneous, yet systemic therapy is still too often selected using generalized clinical factors rather than patient-specific evidence, leaving many patients exposed to rounds of ineffective treatment and avoidable toxicity. Through our BCRFA-supported work, CerFlux is advancing its Personalized Oncology Efficacy Test (POET®)to help bridge this gap by testing a patient’s own tumor tissue ex vivo against multiple therapeutics in parallel, before treatment begins. The goal is straightforward: identify which options are most likely to work for an individual tumor - and which are unlikely to benefit - so care teams can move faster toward effective therapy while reducing the burden of failed treatment. By bringing actionable efficacy insight closer to the start of care, this approach is designed to improve decision-making, conserve precious time, and support more personalized treatment pathways.

Pancreatic ductal adenocarcinoma (PDAC) remains one of the deadliest cancers, and clinicians still lack predictive tools to determine which systemic therapy is most likely to benefit an individual patient, often leading to trial-and-error care and unnecessary toxicity. In this grant-supported effort, CerFlux is advancing its Personalized Oncology Efficacy Test (POET®) to help match patients to the right regimen before treatment begins. Building on a patented biomimetic in-vitro platform for pharmacologic transport and pancreatic tumor microtissue modeling, the project aims to identify both effective and ineffective options for each patient in advance of clinical decision-making. The work is strengthened through a commercial–academic collaboration with The Ohio State University’s James Comprehensive Cancer Center, including planned machine learning to generate a quantitative “POET Score” that helps rank therapies for each patient.

This JMIRx Med short paper examines whether per-capita COVID-19 testing can mislead policy decisions by ignoring population density (how closely people live and interact). Using publicly reported data for all 67 Alabama counties, the authors compare tests-per-capita against cases and find only a weak relationship (reported correlation r=0.28), while new cases concentrate more heavily in denser areas even when those areas have relatively lower testing when viewed through a density lens. The paper argues that density-agnostic reporting can create a false sense of securityand recommends realigning testing allocation toward higher-density regions to better match transmission risk.