When the Target Market is Right, Prognosis for Diagnostic Start-Ups is Bright

Posted by Gabor Bethlendy

Mar 27, 2014 10:49:00 PM


Gabor Bethlendy, Rina Wolf, and Arindam Bhattacharjee, PhD


Gabor Bethlendy and Arindam Bhattacharjee, Parabase Genomics, Boston, MA, USA

Rina Wolf, XiFin, San Diego, CA


Much attention has been paid to personalized medicine in the past ten years. Most of this frenzy has been created by large drug and companion diagnostic companies to drive their value proposition and support inflated pricing on drugs. With the Affordable Care Act that gravy train is going to take a severe slow down in the coming years as it forces hospitals into Accountable Care Organizations (ACOs) and refocuses care from fee-for-service to a value-based model. Some have argued that ACOs will have a dramatic impact on diagnostics, but these generalizations overlook the fact that ACOs are focused on Medicaid and Medicare populations and the jury is still out on their effectiveness. The emphasis will move towards prevention and moving patients to lower cost settings and reducing costs by preventing the use of expensive over utilized imaging and other repetitive services. This bodes well for some very specific diagnostic niches.

While it is true there have been some reports from SVB that point to the lack luster performance of diagnostic investments. However, this and other reports on the prognosis of diagnostics miss the mark as they all focus on Oncology, CNS, Anti-Infectives, which are markedly different than the diagnosis of more reachable markets in Mendelian disorders where the big diagnostic labs do not play.

7 Reasons Why Neonatal Diagnostics is a Good Move

  1. Diagnostic science can be less risky than therapeutics. It is likely that investments in diagnostics have been less than stellar because the focus has mostly been on high volume low margin common disorders. Much like the shift to interest in rare disease therapeutics in recent years, focusing on monogenetic disorders can significantly reduce the risk profile for diagnostics. The fact is that the fundamental science behind predicting diagnostic outcomes is very different for monogenic disorders than common diseases like heart disease or diagnosing cancer. Of course like any field it is a moving target but the return on investment for the patient is much higher. Unlike complex disorders, where many different genes and environment interactions lead to disease, Mendelian or monogenic disease present with simpler 1:1 correspondence between a causal variation in a single or few genes, and is easier to diagnose at the molecular level. The germline DNA is mostly monotonic and its analysis is not affected by analyte in the body fluid, which is a huge challenge for other tests.

             The term diagnostics is typically used to refer to a test used as a last resort for confirmatory testing after a clinical diagnosis has been established. Screening on the other hand is the other extreme of finding a disease before signs or symptoms are apparent, and traditionally has been a smaller market. One exception is newborn screening where only 30-50 disorders are screened after birth in almost every newborn. However, the remaining ~ 3500 Mendelian diseases are not screened but disproportionately affect newborns and infants to a large extent. Thus many newborn admittants in the hospital have a Mendelian disease without clear indication and only symptoms. They are stabilized and released, lost to follow-up or may die without ever knowing what was the cause of their illness. The testing of previous years, in both diagnostics and screening, were done mostly using multiple tests like x-rays, MRIs or biochemical approaches to rule-in or rule-out a cause. For many of these diseases we have done screening for decades using Sanger sequencing. The new molecular diagnostic approaches using next generation sequencing is the next step where we can parallel process these tests at the same price or lower and can make a difference here and produce a better outcome than just confirmatory diagnosis by assisting the clinician in making a clinical diagnosis and avoiding medical sequelae. Biochemical and other approaches in diagnostics are a ‘short sighted’ test approaches, as they do not inform the severity of the disease or the precise molecular cause of the disease. This ignorance doesn’t help the family in the long run, and they may go on to have a second child with the same disease.

            Until now society and the industry has depended heavily on therapeutics to solve human suffering, but moving forward that does not need to be the case given the tools that are available. These newer diagnostics also have the potential to not only inform, but also reduce incidences.  Specifically, when it comes to recessive diseases, incidences are avoidable. It is not hard or complex to understand this. A good example of this is the impact on the Ashkenaji Jews (AJ) since the 1970s where such testing has resulted in a 90% reduction in Tay Sachs, making this now mostly a non-AJ disease.

  1.  Diagnostics are not necessarily more capital intensive while commercialization and profitability can be reached early. Conventional wisdom tells us that large validation studies are required to bring these tests to the market, but this is more true for complex or multifactorial disorders like those targeted by such tests as OncotypeDx, CardioDx and others where thousands of samples are normally validated. In Rare Disorders (which are Mendelian in nature) one will never find 2,600 cases, and it is possible that sample sizes as low as 20 per condition may suffice, if the studies are credible. Here calling mutations correctly requires very low coverage using next generation sequencing which also eliminates the need for time consuming Sanger validation. Truth is technically speaking
  2. Value-based pricing is possible in the segments that are less sensitive to third party payors. Traditionally companies have focused on large market segments hoping to win a smaller slice of a big pie. That top down thinking is attractive to large companies as they have the existing channels and instrumentation, but often prove a disaster for start-ups to enter. Moreover, they exist in low cost outpatient settings where many large service providers compete, mostly on price.  At the end of the day how many companies can be experts in cancer, diabetes and cardio and still offer a value? In this economic climate Hospitals are looking for tests that can provide an early diagnosis to reduce the average length of stay and moving patients to lower cost settings.  The fact is 34% of genetic disorder admissions account for 50% of the total hospital charges, and themean length of stay are 40% longer than those without an underlying genetic condition. Most of these are neonatal or pediatric cases that require diagnosis and treatment starting with the first week of life. What is the clinical actionability of an exome or genome sequence when the test takes 12 to 16 weeks and the average length of stay for the majority of neonates is about 10 days? Long turn around times like this leave little clinical or economic value for anyone in the care and purchasing cycle. Value can be delivered to, and therefore higher pricing can be extracted from, Hospitals by offering tests that have an immediate impact in the very first few weeks of life, not weeks or months later.
  3. Regulatory uncertainty exists, but mostly due to confusion regarding reimbursement vs. payment strategies. The validation process for a laboratory developed (LDT) test is very clear, and it is possible to obtain CLIA’88 CAP approval in the span of a few months. The confusion comes when investors talk about reimbursement as the sine qua non of the business model, when often a shift in focus on developing an install base in segments not sensitive to third-party payers can mitigate this risk. One has to wonder what the large drug and diagnostic companies with FDA approved drugs (and companion diagnostics to go with them) and 510k cleared devices would gain by supporting the 80% of the rest of the market that are LDTs encroaching on their turf? The fact is that the 1984 Orphan Drug Act provides a lot of latitude to companies making Orphan drugs as well as diagnostic tools. Since the entire rare genetic testing market is run under CLIA it is unlikely it will be regulated to the extent that would do damage to mostly children who get these tests who happen to be one of the most expensive part of the hospital’s cost.
  4. Reimbursement is a mess, but that is not always relevant. We have to remember that although what the Centers for Medicare & Medicaid Services (CMS) does affects all of healthcare, stand alone reimbursement is most relevant for outpatient visits. For testing performed during anModern_Medical_Team in-patient stay those costs are born by the Hospital itself under Diagnostic Related Group (DRG) for Medicare cases (of which babies are not part of) or by the hospital itself for younger patient populations. In theory, one could argue that even non-Medicare cases are impacted by APS-DRG, but even there these are driven by the 3,000 or so Major Complications and Comorbid Conditions (MCC) that when present have been shown to increase the average length of stay by at least one day.  Of these metabolic and other conditions (that are Mendelian in nature), and which are very prevalent in NICUs, are just one example of how a fast turn-around test for an early on-set condition in neonates would get paid by institutions as it would help reduce length of stay and drive to a more definitive and actionable diagnosis. Under an inpatient stay, even under the Affordable Care Act, the driving force for economic and clinical test utility is whether the testing can reduce the average length of stay.  This is something that can be shown through a smaller case controlled study. Additionally, studies both from academic institutions and Geisinger have shown both clinical utility for un-diagnosed genetic disorders and economic benefits of using whole exome sequencing to reduce the diagnostic delays due to slow singleton gene testing. Even with their excessive turn around times, these exome level tests are already being reimbursed for major academic institutions as well as several diagnostic service providers. It makes sense that a panel that would be more targeted, faster and have a more immediate impact on patient care would get favorable review from payors as they would also be less expensive and require less add-on testing. CPT code “stacking” of analytes represented by Tier I codes is expected to work for the rest of 2014 and these complex exomes, genome and panel tests are likely to fit under the newly announced AMA codes. It is true, there are uncertainties, but none that are not manageable given the right scenario.
  5. Confusing the costs of next generation sequencing (NGS) in the research setting with pricing and utility in diagnostics. It should be no surprise that maturing NGS instrument vendors are lowering prices in order to boost reagent sales especially in this climate of NIH funding cuts. But does the Holy Grail of the $1,000 genome translate to clinical practice when and where it really matters? This cost does not take into consideration the intellectual property and pre and post analytical effort that is required for a clinically meaningful result. It only takes talking to a handful of Genetic Counselors, physicians and third party payors to figure out that a 12 to 16 week turn-around exome test that is likely going to cost $9,000/sample in the clinic is not likely to get widely adopted tomorrow, never for screening nor diagnostics. This is even ignoring the fact that in certain cases, one has to do family analysis to make a good call. After all that, you still have to validate with Sanger.  Insurance does not pay for the healthy family members even if it improves the results. For this and other reasons, for the foreseeable future, exomes and genomes are relegated to the longer tail of clinical utilization that have very low impact on changing patient care and therefore unlikely to get paid for in any significant numbers to make them a viable commercial play. To mitigate this the focus should be on panels with fast turn-around times that change patient care while patients are still in the hospital.
  6. Using the wrong comps to make the investment decision.Today’s companies in Life Sciences are different from before when everyone did drugs and biochemical assays. It is very likely that comparisons to OncotypeDx, Pac Bio and other box-type diagnostic starts are the wrong model to use when considering a diagnostic investment. More recent start-ups like Foundation Medicine, Verinata, Natera, Invitae are more relevant as each have built not boxes or drugs, but rather an expertise in a specific areas for understanding different disease classes. Unfortunately this is the thinking today, but there are emerging markets with very distinct unmet needs that are readily available for those who see them.


As we move into the next phase of translating genomics to practice, hospitals will look for ways to move patients to lower cost settings. Focusing on opportunities that enable that to happen, outside of the traditional mix of complex disorder, start-ups can mitigate much of the risk in the diagnostic space. Establishing beachheads in segments that are less sensitive to third party payors will be critical as long as one can show clinical utility, which is a given in this space. Investments need to focus on ways to move to the bedside in years rather than decades, as is the current system of thinking. We have to consider that families often do not consternate over spending $1,000 on a TV, yet expect insurance to pay for a genetic test that could help their child.  Can we educate them enough so they know the difference they can make for their child and future generations so as to pay out-of-pocket? Lastly, tools are not enough to win in diagnostics. The company has to deliver something the physician and/or healthcare professional will have less and less time for as genomics becomes more a regular part of medical diagnostics and practice; a timely report that allows them to make sense of all those new and potentially never seen before, private mutations that are not in the public databases. The future of automated diagnostics is here, and it is certainly much easier for simple genetic diseases than cancer or therapeutics.