Tag Archives: diagnostics

Exosomes: The Little Vesicles That Could

A Boston Biotech Watch guest post by Dima Ter-Ovanesyan*

When I mention to other biologists that I work on exosomes, I am used to getting blank stares. Exosomes, also sometimes called microvesicles, are small lipid vesicles secreted by all cells. Like many dark corners of biology, the exosome field is a province of a few experts and is still largely unknown to the mainstream. But thanks to some exciting early results and a long list of potential medical applications, exosomes are beginning to move out of the shadows and into the light. Exosomes were the subject of two sessions at the annual American Association of Cancer Research (AACR)  meeting earlier this month. In January, the first-ever conference on exosomes brought together more than two hundred researchers from around the world. And now even the popular press is picking up on the applications of exosomes to RNA interference (RNAi) drug delivery and diagnostics in fields ranging from cancer to diabetes.

The concept of cells budding off small particles was actually first mentioned in Charles Darwin’s The Variation of Plants and Animals Under Domestication in 1868 and exosomes were first observed by electron microscopy in the 1980s. For twenty years after that, exosomes were thought by many to be nothing more than “cellular trash bags” that dump proteins deemed to have outlived their “use-by” date. This changed a few years ago, though, when a number of independent studies showed that exosomes actually contain not only protein but also RNA. This discovery opened the possibility of using the RNA in exosomes floating around in bodily fluids to learn all kinds of secrets about the cells that release them.

Charles Darwin

There first (as usual)

I learned about the diagnostic potential of exosomes in early 2009 when a venture capitalist asked my opinion on a business plan. The company, called Exosome Diagnostics, was being spun out of Massachusetts General Hospital (MGH) to develop new diagnostics based on analyzing the RNA in exosomes isolated from blood and urine. At the time, I was an undergraduate at MIT working on microRNAs. After reading the business plan, the patent filing, and the underlying scientific publication, I called the VC back. “This could be HUGE,” I told him. “We should get in on this,” he replied, and promptly forgot all about it. I did not. After graduation, I opted to learn more about exosomes through a brief stint at the Curie Institute in Paris. I then joined Exosome Diagnostics, which by then was flush with a $20M Series A.

The company’s platform is based on the work done at MGH, which showed that mutant RNA transcripts derived from key genes can be detected in exosomes released by cancer cells. Johan Skog and colleagues had shown that by isolating exosomes from blood and looking at the RNA inside, they could tell whether a patient’s tumor contained mutant EGF Receptor (EGFR) – establishing the proof of principle for exosomes as companion diagnostics. Oncologists could use them to peer into the genetics of tumor and decide whether a patient would be a good candidate for EGFR-inhibiting drugs such as Tarceva® and Iressa®. As pharma companies move increasingly towards targeted cancer therapies coupled to companion diagnostics (think PLX-4032, Plexxikon’s BRAF inhibitor for melanoma, or Crizotinib, Pfizer’s ALK inhibitor for lung cancer), the appeal of a blood or urine test instead of a biopsy to analyze specific mutations is not hard to see. This is particularly advantageous in cancers where obtaining a biopsy is difficult, like brain cancer or lung cancer.

I think that companion diagnostics in cancer are just the tip of the iceberg for exosomes in diagnostics. Several recent studies have shown that different levels of certain RNAs in exosomes and other vesicles isolated from blood correlate with different disease states. In other words, exosomes have RNA “signatures” for different diseases. One recently published study , for example, showed that a specific microRNA in vesicles derived from diabetics’ blood was elevated compared to vesicles taken from non-diabetics. Amazingly, the researchers also showed that this biomarker could potentially be used to identify patients who will get diabetes before any clinical symptoms occur. Although the test in the paper was not yet sensitive enough for the clinic, the results raise the intriguing possibility of using RNA signatures to predict disease, not just diagnose it.

Several reports at the January exosome conference highlighted diagnostic applications of profiling exosomal RNA in different diseases:

  1. Signatures of mRNA isolated from exosomes in the blood could be used to classify brain tumors based on aggressiveness.
  2. A specific microRNA isolated from exosomes in the cerebrospinal fluid (CSF) was found in brain trauma patients but not in healthy controls.
  3. Certain microRNAs isolated from exosomes in the blood of pregnant women could be used to predict premature births.

Although larger studies will be needed to confirm these effects, I imagine that, in the not too distant future, exosomes will join the list of proteins and metabolites currently profiled during routine blood draws. Through methods such as high-throughput sequencing, the RNA inside exosomes (dare I say “exo-transcriptome”?) will be analyzed and will prove spectacularly useful in helping physicians assess and track patient health.

Exosome Micrograph

I spy a novel class of biomarkers (Image courtesy Johan Skog and Casey Maguire, Massachusetts General Hospital)

Exciting as the potential of exosomes is in diagnostics, you may be wondering what they actually do. Are the different RNA profiles in exosomes more than a correlation? Do exosomes actually contribute to disease? This was an intense topic of debate at the exosome conference, and, after listening to four days of presentations on the topic, I would venture to say that we still really don’t know. The fact that exosomes contain RNA raises the extremely intriguing possibility that cells throughout the body communicate by sending each other little “packages” of RNA, and the RNA can then perhaps act in the recipient cells – be translated into protein in the case of mRNA or repress the expression of other genes in the case of microRNAs. And, if this is the case, one could imagine hijacking this pathway to deliver therapeutic RNAs and thereby overcome an enormous roadblock.

The main obstacle to realizing the tremendous potential of RNAi therapeutics is the challenge of delivery. Whether or not exosomes are actually used to transfer RNA between cells remains to be proven, but results presented at the exosome conference and recently published by a group at Oxford made the first attempt at this approach. Led by Matthew Wood, the group loaded modified exosomes with an siRNA designed to knock down the BACE1 gene implicated in Alzheimer’s Disease.  Although they used healthy mice as models, the researchers demonstrated proof-of-principle by showing reduced levels of BACE1 in the brain. Of course, it would be reassuring to know what exosomes actually do in the body before injecting them into patients, especially given that exosome-based drugs would be complex biologics composed of several different proteins, not just lipid vesicles. As we learn more about the biology of exosomes, however, I could imagine scientists designing exosome-mimicking particles with the minimum necessary components to carry the RNA therapeutic to its designated site in the body.

As evidenced at the exosome sessions at the AACR meeting, there is increasing interest in exosomes secreted by cancer cells. Cancer exosomes have been found to contain oncogenic mRNA and proteins, and it is thus tempting to speculate that exosomes may have a role in modifying the tumor microenvironment and helping the cancer spread. If true, one could even imagine using therapeutics to specifically target cancer exosomes. Unlike with the use of exosomes for diagnostics, however, the use of exosomes for therapeutics will require scientists to uncover some of the fundamental biology of what exosomes are and how they function. Luckily, as evidenced by the excitement at the exosome conference, there are at least two hundred of them up for the challenge.

# # #

*Dima Ter-Ovanesyan (dimatero@gmail.com) graduated from MIT in 2010 with a Bachelor of Science in Biology. At MIT, he worked on RNAi screens in cancer with Michael Hemann and microRNA targeting with Chris Burge and David Bartel. He was also an associate at CBT Advisors. He currently works at Exosome Diagnostics.

5 Comments

Filed under Uncategorized

Medicine Gets Personal – But How Do VCs Make Money?

Boston Biotech Watch has been keeping a close eye on three big trends and their impact on VC deal-making: real-world applications of genetic data, personalized medicine and health care reform. Can startups use genetic data to drive down drug costs? To what extent will genetics become the high-value gatekeeper for future pharma industry success? And will VCs be able to exit from companies in this sector quickly enough to reap outsized returns?

Judging from the VC activity in the space, some venture investors apparently think that strong exits are likely. What a radical departure! Right up until the early years of this decade, “diagnostics” was a dirty word in biotech venture circles. Most diagnostics deals smelled bad to most VCs whether the deals were sample-prep focused (like Cytyc, which was a massive success) or cancer biomarker repositories like DiaDexus, a high-profile joint venture between SmithKline Beecham and Incyte that raised $102.5 million in 2000, is still privately held and, despite one commercial test for coronary disease that finally achieved Medicare reimbursement in 2007, does not appear to have provided much – if any – of a VC return.

It has long been a VC maxim that “you could wait forever for the US health care system to move in a more rational direction” and that therefore VCs had to do deals that were consistent with the existing models no matter how broken these models were. Cynicism was rewarded, idealism punished.

Yet suddenly the United States appears to be on the verge of the largest health care reform (HCR) in its history and, perhaps not surprisingly, what feels like dozens of deals related to diagnostics, genetics and HCR have begun to materialize. The deals reflect many different ways of looking at the personalized medicine opportunity (see Tables 1 and 2).

Boston Biotech Watch recently attended the sixth “Personalized Medicine Conference” at Harvard and did some additional reading and research. This, along with proprietary information from CBT Advisors serves as basis for this snapshot. Our goals here are threefold:

(1) To explain – with examples – what sorts of companies are getting funded;
(2) To disclose the rationale driving the deals for some of the key investors in the space; and
(3) To hold up one recent high-profile deal, Generation Health, as the sort that other investors were clamoring (mostly without success) to get into.

Partners HealthCare Center for Personalized Medicine and Genomics logo

Just judging by the attendance at this high-quality conference, put on annually by the Partners HealthCare Center for Personalized Genetic Medicine (PCPGM) as well as Harvard Business School (HBS), the field is gaining momentum. More than 600 participants registered, compared to just 237 at the inaugural conference in 2005.

Our breakdown of VC deals in the personalized medicine space follows in Tables 1 and 2 below. Why are VCs convinced – despite such a negative history for investing in diagnostics – that personalized medicine is where the big money will be? Try “tenfold growth,” a squishy yet thought-provoking projection included in the December, 2009, report entitled “The New Science of Personalized Medicine” by PriceWaterhouseCoopers (PWC). Even allowing for the typical hyperbole associated with such reports, there is apparently more money than ever to be made from genetics, genomics, diagnostics, theranostics and related technologies.

Business model Company Market status Indication Technology VCs in Amount raised Exit
Content – algorithm Genomic Health Commercial Breast & other cancers Biomarkers + algorithm Kleiner Perkins, Versant & others $103M total IPO in 2005
Box Handylab Commercial Hospital infections Rapid DNA assay Arboretum, Ardesta, Dow Ventures, DuPont Ventures, EDF, Lurie, SBV, Wolverine $46M total Trade sale to Becton Dickin-son 2009 for $275 million
Technology platform (+content) GeneOhm Commercial Staph & other ID Rapid DNA assay CB Health, Domain, CHL, Kaiser Permanente, QuestMark,
Posco
Raised $26M Series C in Jan. ’05 Trade sale to Becton Dickin-son 2006 for $255 million

Table 1: Diagnostics and genetic testing companies from which top-tier VCs have exited

San Francisco-based venture capitalist Dion Madsen, a Managing Director at Physic Ventures, affirmed the newfound VC enthusiasm for personalized medicine when Boston Biotech Watch paid him a December visit. Physic is one of many VCs looking hard at the diagnostics space and one of the few to have diagnostics as a mandate. The firm’s tagline is “Investing in Keeping People Healthy.” So Madsen is an especially apt guide to the promise and the pitfalls of the space.

Ahead of a shift to test-prompted care
VC dealmakers usually like to tell themselves that they are just ahead of a paradigm shift and this field is no exception. The idea that genomic information is useful for drug discovery and clinical testing is starting to “percolate” through pharma, Madsen said, and is already leading to better drug design. But the use of genetic information related to the individual patient, for example in the form of genetic-based diagnostic tests, he said, is “only just beginning.”

Behind the big numbers is a firm conviction that payers in the US healthcare system (insurers and government programs like Medicare) will actually come to rely upon and reward molecular and scientific information instead of simply succumbing to ever more expensive marketing campaigns by pharmaceutical, biotech and medical device companies.

Comparative Effectiveness compares treatments

If I take them all, will they cancel each other out?

There is very little in the current package of health reform bills being negotiated in both houses of the US Congress that deals with molecular testing. The closest that HCR comes is in mandating a relatively modest $1 billion for so-called “Comparative Effectiveness” (CE) funding which is meant to determine which therapeutic regimes – be they surgeries, implantable devices, dietary regimes or drugs – are actually working in contrast to the traditional approach of casting each and every clinical trial in the form of a validation or rejection of a single new medication or device. Still, for Madsen, the CE trend is a friend. “Comparative Effectiveness is already a reality,” Madsen said. “That card has been turned.”

Physic has developed four simple criteria – they fit on one side of a sheet of notebook paper – that characterized “doable deals” in the personalized medicine space. For Physic, an investment must be:

1. Actionable – it informs a decision around treatment, preventive action or behavior

2. Cost-effective

3. Based on validated science; and

4. Clinically meaningful.

To pick a widely publicized group of companies that, in our view, fail on “actionability,” consider the consumer genomics companies 23andme, Navigenics and Knome. These companies have won some high-profile backers – 23andme, for example, has Google as a key investor. “What 23and me and DNA Direct are doing is really interesting,” Madsen said, “[it is] just ahead of its time.” These services – which have been dubbed “recreational genomics” – are not actionable enough, he said, for them to be good VC investments. Madsen: “The utility of learning every base pair is very low.”

Genomic Health: A Pioneer, Yes, But a Replicable One?

Historically, only a handful of VC-backed diagnostics companies have managed to fulfill Physic’s criteria and make their investors money. Genomic Health (NASDAQ: GHDX) is perhaps the most prominent of these. The company raced from its first institutional funding to Medicare reimbursement in just five years and pulled off a successful IPO in 2005. In the meantime, its single marketed test – an algorithm-based test
OncotypeDX

called OncotypeDX for guiding breast cancer therapy – now earns more than $140 million in annual revenue. It helps physicians choose treatments that are on the extreme end of the cost spectrum – a $3,500 test that can allow patients – and payers – to avoid bills of $30,000 or more for chemotherapy. That value proposition – along with Genomic Health’s compelling retrospective data – convinced Medicare and other insurers to agree to reimburse the test beginning in January, 2006.

But OncotypeDX is an imperfect example in several ways: First and foremost, not many therapies cost $30,000, so very few tests will be reimbursed at $3,000 or more. Second, FDA has signaled that tests based on algorithms like OncotypeDX will require a greater degree of validation in the future. (How much tougher the regulatory regime will be is likely to become clear in mid-2010, when FDA issues its long-awaited guidelines for so-called “IVD MIA” tests – in vitro diagnostic multivariate assays.) And finally, the return on the $103 million invested in Genomic Health before the IPO was probably more like 3x than the usual 6-8x that VCs consider a “home run.”

Brook Byers

Brook Byers, Kleiner Perkins’ diagnostics VC visionary (Image Justin A. Knight)

Genomic Health was a Kleiner Perkins deal and the other two “DX” companies in which Kleiner invested, CardioDx (founded 2004) and XDx (2000), have apparently not made it to big revenues or VC exits nearly so quickly. Indeed, both are still privately held. One East Coast VC to whom Boston Biotech Watch spoke said, “Yes, CardioDx has found a potentially relevant market opportunity, but they had to do a 4,000-patient study.” CardioDx is reported to be raising money at a lofty valuation.

Business model Company Marketing Status Indication(s) Technology VCs in Most recent financing
Content On-Q-Ity R&D Monitoring of cancer progression via DNA repair biomarkers Biomarkers, microfluidics Mohr Davidow, Bessemer, Physic, Northgate, Atlas $26M Series A Dec. 09
Content Artemis R&D Prenatal diagnostics Microfluidics Mohr Davidow, Alloy, Sutter Hill $9M in Oct. ’09
Technology platform (+content) T2 Biosystems R&D Not announced POCD – nanoparticle MR assay Flagship, Polaris, Flybridge, Partners Healthcare and In-Q-Tel $10.8M Series B Aug. ’08
Long-range disease prediction & risk assessment Tethys Bioscience R&D Diabetes Blood test; panel of biomarkers Aeris, Kleiner Perkins, Mohr Davidow, Intel Capital Raised $25M Series D Nov. 09
“Genetics Benefit Manager” Generation Health One corp. partnership announced All genetic tests esp. in high-value treatment areas Evaluate tests for payers; bridge payers, providers, patients Highland Capital $5M Series A Nov. 08, Deal with CVS-Caremark Nov. ’09

Table 2: Private diagnostics and genetic testing companies in which VCs have invested


Table 2: Private diagnostics and genetic testing companies in which VCs have invested

Among the still-private companies identified in the CBT Advisors screen (see Table 2 for examples), several are looking for ways to capture content and use it to provide immediate value to patients and payers. We consider these the “content” companies. Genomic Health, CardioDx and XDx all fall into this category. These companies run the gamut of indications, with existing plays in cancer (many including Genomic Health, Genomic Vision, Precision Therapeutics, Claros, MTM Labs and On-Q-Ity, which will be discussed further along in this post); cardiovascular disease (CardioDx, XDx), rheumatology and inflammation (Crescendo), diabetes (Tethys) and the ever-popular (and close-to-market) infectious disease, particularly point-of-care tests for nosocomial infections (Opgen, Progentech, Curetis, AdvanDx).

Another group has developed a proprietary technology that either grabs the content (e.g. the microfluidics of Artemis Health, a prenatal diagnostics company) or that prepares it for analysis (Handylab, acquired in October by Becton Dickinson for a reported $275 million). Some technologies do both (T2 Biosystems, a Boston-area Polaris investment based on technology from the prolific Robert Langer lab at MIT). We consider these to be “box” or “sample prep” companies although some of course are also offering unique content.

Recently Physic Ventures acted on its strategy and put its money into a Boston-area diagnostics startup, On-Q-Ity, that meets all four criteria. Like Genomic Health, On-Q-Ity (from Oncology + Quality + Clarity) will provide actionable information in the form of decision support to physicians treating cancer patients. . The validated science consists of (1) biomarkers found in tumor cells that determine their level of progression and therefore the advisability of treating patients at a particular moment; and (2) assays that determine susceptibility to specific chemotherapeutic agents based on mutations in the genes involved in DNA repair. As with the “box+content” companies, On-Q-Ity not only has the rights to these biomarkers and mutation assays but also a proprietary microfluidics technology that is able in principle to pluck circulating tumor cells out of the bloodstream even when these cells are quite rare. The company then applies the two technologies, yielding an unprecedented snapshot of both “treatment response and tumor cell composition … at a molecular level,” Madsen said. Both cost-effectiveness and clinical validity will have to be determined by clinical trial, presumably done prospectively.

On-Q-Ity’s management is something of a dream team. The CEO, Mara Aspinall, was the long-time president of Genzyme’s genetic testing division, which under her leadership developed and commercialized many new tests. Aspinall, who is also on the board of one of Massachusetts’ largest health insurers (Blue Cross Blue Shield of Massachusetts) has about the best track record imaginable for a genetic testing company CEO. In her spare time, she serves as a lecturer in health care policy at Harvard Business School.

In our view, On-Q-Ity scores highest on the first criterion, actionability. As we will address again when we get to Generation Health, oncology diagnostics are already high-value due to the high cost of treatment. In an article on personalized medicine published in 2007 by Aspinall and her HBS colleague Richard Hamermesh, she identified five cancer indications (pancreatic, liver and so on) in which patients typically have low one-year survival and therefore “do not have time to spare” for traditional, “trial-and-error” medicine. If On-Q-Ity can use biomarkers to inform physicians when to treat aggressively or even which chemotherapeutic agents to deploy, then its tests will undoubtedly be reimbursed at or perhaps even above the levels seen for OncotypeDX.

The wild card for On-Q-Ity is the level of validation that will be demanded by FDA and payers. Madsen said that even in the honeymoon phase following the investment, “We are still struggling with, do we need a prospective trial? If so, how do we design it?” These demanding constituencies – FDA, payers, oncologists, cancer patients – will, it seems to us, insist on such a trial. As Madsen put it, “How do you tell an oncologist not to treat a patient with the standard of care? This is our challenge.”

Even when a company meets all of Physic’s criteria, the road may still be uncomfortably long. After all, these companies – like CardioDx and its 4,000-patient study, not to mention DiaDexus and its single approved test – are all attempting to achieve validation under the “old” criteria. How soon can HCR change that?

Generation Health: “The Consumer Reports of Genetics”
These struggles are what make Generation Health stand out. GenHealth

Generation Health logo
seemed to be the darling of the Personalized Medicine Conference and VC firms have been “pounding down the doors” to get in, according to a couple of top-tier VCs (the only announced VC investor is Highland, which made a first institutional investment in the company in 2008, though rumor has it that a second Boston-area fund has joined the syndicate).

GenHealth has the potential to be a high-flyer because it stands in a far different corner of the health care system – next to the payer. GenHealth intends to “help employers and other health care payors manage medical costs and improve their employees’ and members’ health by assuring optimal utilization of genetic testing.” To do this, according to its web site, it will perform three tasks:

• Establish a rational basis for covering or excluding genetic tests based on clinical validity and utility;
• Negotiate discounted rates for tests; and
• Identify patients who would benefit from testing through analysis of medical and pharmacy claims.

These activities would make GenHealth a “filter” for insurance companies and employers. Madsen dubbed them “the “Consumer Reports of Genetics” – a company perceived to be a fair arbiter of the value of genetic tests. “We’ve seen other companies such as DNA Direct do this for HMOs and payers including some we know very well,” Madsen said. “But no other company can do it to the extent that Generation Health would. GenHealth will have better decision-making data,” presumably from aggregating anonymized data across insurers or analyzing claims. In November, 2009, GenHealth signed its first public collaboration with CVS Caremark, a pharmacy benefit manager that already has a pharmacogenomics program. (No surprise about the identity of the first deal partner – CVS Caremark’s Chief Medical Officer Troyen Brennan sits on GenHealth’s board).

What gets VCs excited about GenHealth is its ability not only to take advantage of HCR but to actually participate in it by driving down health care costs and increasing use of gatekeeping genetic tests. GenHealth styles itself a “Genetic Benefit Manager [GBM],” analogous to the Pharmacy Benefit Managers (“PBMs”) Medco and the like – a company where GenHealth founding CEO Per Lofberg served as chairman from 1993 to 2000.

Raju Kucherlapati, Harvard professor and Personalized Medicine Conference founder

Raju Kucherlapati, Harvard professor and Personalized Medicine Conference founder (Image Justin A. Knight)


The discussion about GenHealth’s business stimulated one of the more interesting exchanges of the conference. PCPGM founder and conference organizer Raju Kucherlapati asked CVS Caremark’s Brennan exactly how many tests CVS Caremark is already reimbursing for or including in its decision-making process about providing pharmacy benefits. Brennen did not answer the question, but he did say that the first inroads are in “high-cost disease.” If a treatment costs $100,000 a patient, for example, and a test costs $1,000, even one patient being safely spared the treatment more than pays for the cost of the test for many patients.

“Right now [our testing] is limited to a series of cancer diagnostics,” said Brennen. “Like most PBMs, we operate a specialty pharmacy with high-cost medications and that is where we do the most genetic testing,” he added. In the non-specialty areas, there is not yet “reasonable evidence” for incorporating it into practice, although testing is more prevalent there than it was five years ago. However, the amount spent on testing is expected to grow quickly, he said, because “at Caremark, we will be leaders in cost reduction. That is why it is important for us to incorporate genetic tests. We want to stay away from provider-driven modes,” that is, to pay for care that matters to the patient.

(At the same moment as this edition of Boston Biotech Watch went up on the morning of December 21, 2009, CVS Caremark announced that it was taking “an increased ownership interest” in Generation Health. The press release quoted CVS Caremark Chairman Tom Ryan as saying that, with the additional investment, CVS Caremark is “accelerating our commitment to personalized medicine and making genomic benefit management an integral part of our PBM offering.” Indeed, in the same announcement, CVS Caremark named GenHealth CEO Per Lofberg as the President of the company’s PBM business; GenHealth co-founder will become its new CEO. Although the release said that GenHealth will continue to operate as an independent business, “offering a full range of GBM services to health care payors,” it was left unclear how free GenHealth would be to do strategic deals with other PBMs. Terms were not disclosed.)

Meanwhile, new genetic tests keep pouring in to payers at a rate of what feels like “100 a week,” said Madsen. Each new test faces the traditional gauntlet of long-term, prospective studies before it can start making investors money. So not only are new tests needed but also, as Aspinall and Hamermesh described in their 2007 article, better regulatory and reimbursement regimes. Until these key pieces are in place, most VC deals in the space will be vulnerable to the cash and momentum drain of drawn-out prospective testing.

# # #

Disclaimer: CBT Advisors has worked with Precision Therapeutics and Genomic Vision. When he was a venture capitalist, Steve Dickman was part of a team that invested in Precision Therapeutics.

4 Comments

Filed under Uncategorized