Where Einstein Meets Edison

Real Solutions from an Artificial Liver – MIT startup facilitates pre-clinical development with a unique liver model

Real Solutions from an Artificial Liver – MIT startup facilitates pre-clinical development with a unique liver model

Nov 7, 2010

The cost of developing a new drug is an often quoted, but rarely consistent, statistic.  Cited amounts range from hundreds of millions to several billion dollars.  It all depends on sectors, time frames and specific associated costs, but the bottom line is that the number is astronomical.  More than a few people have weighed in to suggest policy changes or new financing structures that address the problem[i].  MIT scientists are likely to be part of a multifaceted solution as they pioneer new techniques in drug discovery and, equally importantly, in pre-clinical models.

Pre-clinical studies allow researchers to screen and evaluate pharmaceutical candidates before advancing compounds into more expensive and risky clinical trials.  An effective pre-clinical model can help reduce expense and minimize risk on both ends of the drug development timeline; two critical goals for pharmaceutical and biotechnology companies, who are acutely aware of the substantial cost of developing new drugs.  Coupled with the economic downturn and – for a slew of large pharmaceutical companies – an impending patent cliff, the sunk costs and lost time of a failed drug development program are crippling, often forcing companies to turn to mergers, acquisition or constriction (read “restructuring”) as ways to assuage stock holders.  If a new candidate is identified correctly and manages to advance through clinical trials, it emerges on the other side of the drug pipeline to face an increasingly stringent Food & Drug Administration (FDA).  Accidents, like the highly damaging incident when Merck’s FDA-approved analgesic drug, Vioxx, was found to be associated with heart attacks in a number of patients[ii], have caused the FDA to be very cautious in their review process, frequently requesting additional clinical trials beyond the traditional three phases or just flat out rejecting drug candidates, much to the chagrin of otherwise optimistic companies and investors.  Scientific advances in pre-clinical drug candidate screening may help improve the low odds of success and reduce costs by identifying potential pitfalls earlier in the pipeline.

During the research and pre-clinical stages of drug development, companies use benchmark assays to predict adverse events at later stages in the process.  More predictive assays allow scientists and their corporate leadership to more accurately select the right candidates to move forward.  One of the most critical screens is for liver toxicity, a chronic threat that can arise at any point in development.  Dose-limiting liver toxicities stemming from either the drug itself or its metabolic products are major reasons why many drug candidates fail to progress.  The standard in vivo markers indicative of potential toxicity are alanine aminotransferase (ALT) and aspartate aminotransferase (AST), liver enzymes that are normally confined to liver cells but are released into the blood stream at abnormally high levels when those cells are damaged.  Unfortunately, these markers are not always perfectly predictive and, as a result, drug-induced liver toxicities are far too common in late-stage development and marketed drugs[iii], resulting in money wasted on clinical trials and costly, damaging recalls[iv].  Obviously, it is better to detect this type of toxicity earlier in the process, so scientists try to improve their predictive capabilities using in vitro models of the liver designed to mimic the human organ.

Hepregen, an MIT spin-out based in Medford, MA, is developing a new in vitro tool that is designed to enhance the selective attrition of new drug candidates.  Their core technology is based on the ability to micro-pattern tissue culture substrates that enhance and localize the growth of different cell types found in the liver.  Using soft polydimethylsiloxane molds patterned with small circular holes, researchers cover a normal tissue culture surface and then treat the exposed surface with collagen[v].  The result is a surface polka-dotted with circular collagen islands.  When the templates are removed, the tissue culture plate is seeded with primary hepatocytes and fibroblasts to create a two-dimensional culture system that closely mimics in vivo structure, genotype and metabolism.  The patterned primary hepatocytes grow exclusively on the collagen islands, retaining viability and structure for several weeks.  This is a significantly longer window than that of liver slices, which survive only a day.  Hepregen validated their model with various measures including tests for albumin secretion and expression quantification of a variety of phase I and phase II metabolic enzymes to ensure that the cultured cells were capable of activity similar to those in vivo.  Compounds known to induce liver toxicity in vivo were also tested in the system and hepatocyte cell death or changes in morphology were observed at varying concentrations, proving that cytotoxicity in the model system correlated with clinical toxicity.  Two-dimensional cultures, like the Hepregen approach, are an important advancement for pre-clinical liver toxicity models because they are more ideal for scaling to industrial high-throughput applications than three-dimensional tissue engineered scaffold alternatives[vi] or liver slices, but fail to achieve quite the same level of in vivo therapeutic consistency.  All of the methods are also plagued by the necessity of primary hepatocytes, which are often not readily available.  The need for primary hepatocytes severely limits scalability which is the primary distinguishing feature of this technology.  If they can achieve adoption before other liver toxicity models reach market, then Hepregen may have an upper hand, otherwise they may need to find a way to achieve similar physiological mimicry with a sustainable cell source.  Hepregen was initially funded by a $5M Series A from Battelle Ventures and has subsequently received several development grants from the Food & Drug Administration, the National Institute of Health, and the National Science Foundation.  Continued renewal of non-dilutive funding resources such as these grants will allow Hepregen to continue optimization of their platform technology.

            As pharmaceutical and biotechnology companies worldwide continue to feel the increased financial strain of cost-sensitive medical insurance payers, skeptical venture capitalists and investors, and shrinking patent windows, the need for advanced technologies that prove presciently predictive of late-stage failures will continue to grow.  Hepregen’s patterned liver culture model is an exciting advancement in this field and may become a valuable tool for therapeutics companies wishing to more rigorously interrogate lead candidates prior to pre-clinical animal testing.  An in vitro liver culture model designed to effectively identify ‘bad apples’, or toxic drug candidates, sooner rather than later allows a great number of lead drugs to be discarded prior to full-blown clinical trials, saving millions in trial costs.  The pharmaceutical and biotech research and development world is full of small molecules and proteins that could be the next blockbuster but that may also fail after millions in sunk costs.  Companies like Hepregen are working on tools to aid the search for truly promising candidates, hopefully driving down the price of those miracle cures by illuminating problems before money is wasted on advancing drugs through clinical trials when they never should have been there in the first place.


[i] McArdle M.  No Refills.  The Atlantic (July/August 2010) http://www.theatlantic.com/magazine/archive/2010/07/no-refills/8133/1/

[ii] Official Vioxx Settlement. http://www.officialvioxxsettlement.com/

[iii] Ballet, F. Hepatotoxicity in drug development: detection, significance and solutions. J. Hepatol. 26 (1997)  26-36

[iv] Kaplowitz N.  Idiosyncratic drug hepatotoxicity.  Nat. Rev. Drug Discov. 4 (2005) 489-499

[v] Khetani SR, Bhatia SN.  Microscale culture of human liver cells for drug development.  Nat. Biotech. 26 (2007) 120-126

[vi] Schmelzer E, Triolo F, Turner ME, et al.  Three-dimensional perfusion bioreactor culture supports differentiation of human fetal liver cells.  Tissue Eng Part A.  16 (2010) 2007-2016

Christopher M. Pirie, PhD

About

Scientist & Co-founder, Manus Biosynthesis";s:15:"profile_teambio";s:432:"Dr. Pirie is a scientist and co-founder at Cambridge, MA based Manus Biosynthesis. Chris earned his PhD in Biological Engineering from MIT in 2011 where his thesis research was focused on therapeutic protein engineering and he was a recipient of the NSF Graduate Research Fellowship. Also while at MIT he served on the Institute Committee on Intellectual Property and served as a writer and editor for the Entrepreneurship Review.";s:11:"profile_bio";s:1019:"He received his Bachelors degree, cum laude with college honors, in Bioengineering from the University of Washington where he studied intracellular drug delivery. In addition to his academic research he has served as a student representative to the Institute Committee on Intellectual Property where he helps guide MIT policy on important IP issues like open source publishing and patent reform. While at Washington he served on the Bioengineering Curriculum Committee pushing for improved course work in molecular transport and mathematical modeling. Born in the San Francisco Bay Area, he grew up in Seattle and now lives in Cambridge. As a life sciences editor and writer for MITER he sources primarily research focused articles with tangential, intellectual excursions into ideas on venture capital and technology translation. Chris enjoys sailing the Charles River, reading classic literature and astrophysics books, traveling to places other people only read about, and BASE jumping off the Green Building.