CEO SUMMARY: There will be an expanding role for innovative clinical labs as healthcare moves forward on its path toward personalized medicine. However, to capitalize on this opportunity, pathology groups and clinical labs will need to beef up their information systems. They will also need to recruit lab staff who are skilled in interpreting molecular and genetic test data expressly to advise and consult with referring physicians. One experienced industry consultant predicts that oncology will be the first medical specialty to make extensive use of molecular and genetic testing in this way.
GENETIC TESTING IS POISED to transform medicine and foster unprecedented growth in molecular diagnostics. At the same time, most clinical labs and pathology groups are unprepared to deal with the informatics overload that is heading their way.
That is the prediction of Katherine Tynan, Ph.D., President of Tynan Consulting, LLC, in San Carlos, California. Her firm provides strategic business planning for diagnostic and pharmaceutical companies that are developing molecular and companion diagnostics, including product development and pricing, reimbursement, and market entry strategies.
“It takes less time now for advances in molecular diagnostics to find acceptance in clinical use,” observed Tynan. “This spring the American College of Medical Genetics issued a practice guideline suggesting that whole exome sequencing and whole genome sequencing be used for the diagnosis of idiopathic pediatric cases.
“As such testing moves into the clinical front line, labs will see a potential avalanche of data coming at them,” she noted. Tynan was speaking at THE DARK REPORT’S Executive War College in New Orleans in May. “This will require a significant response by clinical laboratories,” she added.
“To make clinical use of the vast amounts of data generated by these types of diagnostic tests, labs will need to do two things,” she advised. “First, labs must hire specialists in molecular medicine so that they can advise and consult with treating physicians on the results of these tests. Second, labs will need to invest in more robust information systems to store and analyze the data produced by molecular and genetic tests.
“Many pathologists may be aware that a large managed care company recently assessed how much money would be spent on molecular diagnostics in the next five to 10 years,” Tynan said. “UnitedHealthcare’s report, UnitedHealth Center for Health Reform and Modernization 2012, noted that spending on molecular diagnostic services is currently in the range of $6 billion to $8 billion per year,” she said. “The report predicts such spending will climb to $15 billion to $25 billion by 2021.
“Pathologists and molecular scientists will be at ground zero in this trend,” predicted Tynan. “As molecular medicine grows, one major driver in this spending will be the expanded use of tumor genome profiling. This trend is being driven by relatively cheap sequencing and the biological understanding that cancer, for the most part, is an acquired somatic genetic disease. Deriving value from this spending will require clinical labs to re-evaluate workflow and how they interact with referring physicians.
“As they re-evaluate workflow, clinical labs and pathology groups will recognize the need for two significant investments,” commented Tynan. “First will be the need to invest in more robust information systems.
“That is because the existing information technology (IT) infrastructure used by many clinical labs and pathology groups today will struggle under the avalanche of data generated by molecular and genetic testing,” she added. “Fortunately, information technology is getting faster, better, cheaper—and it will continue to do so.
“Second, and of greater importance, is the need for labs offering molecular diagnostics to invest in more sophisticated clinical expertise,” continued Tynan. “Few of the 17,000 pathologists working today have been trained in molecular pathology.
“That is why demand for pathologists who are board-certified in molecular pathology will outstrip the supply for many years to come,” she said. “Laboratory medicine must respond with different ways to encourage more lab professionals at all levels of certification to train and work in molecular pathology and genetic testing.
“Another problem exacerbates these two deficiencies,” stated Tynan. “The business models are not defined in this space. This is true for clinicians delivering gene-based medical services as well as the pathology laboratories that provide them with molecular and genetic testing.
“In addition, financial incentives and reimbursement are not yet in alignment,” she declared. “As a member of the Economic Advisory Committee for the Association of Molecular Pathology, I can say that workable solutions for each of these significant issues have yet to be developed.
“There are several reasons why the financial incentives are not aligned,” said Tynan. “Many types of molecular tests come with a very high cost, as do some targeted therapies. Payers have yet to address this new clinical use of expensive companion diagnostic tests with expensive therapies. And the future will only see an increase in the number of high-priced genetic tests and targeted therapies.
“But if patients can’t afford one of these expensive interventions now, how can they possibly afford a cocktail that includes several of these expensive medications?” asked Tynan “This question will be particularly problematic if combinations of targeted treatments are required earlier in the care cycle to minimize resistance or if the treatment of cancer makes it a chronic disease, as many experts believe will be the case.
Opportunities For Labs
“Along with these significant challenges, however, will be opportunities for pathologists and clinical lab professionals,” she added. “Within the next three to seven years, labs will need to hire expert intermediaries who can interpret and integrate the data produced by genetic and molecular testing. One positive aspect of this development is that we are approaching an inflection point where pathologists could play a significant role in changing how medicine is practiced.”
Tynan was very specific in her recommendations about the skills and capabilities that clinical labs and pathology groups should be developing. “Good decision support tools are years away,” she observed. “Thus, in the short term, labs will rely on expert intermediaries who can take that genetic and molecular information and derive the valuable clinical insights that help the referring physicians.
“At the same time, labs also need to give their IT departments the capabilities to handle and combine numerical, morphological, molecular, and image data so this information can be presented in a single report,” Tynan explained. “Oncologists cannot do this for us.
Oncologists Are Overloaded
“In fact, it’s already incredibly difficult for oncologists to keep up with developments in the literature,” she observed. “It will be high value-added for pathologists and molecular geneticists to deliver that information to oncologists in a digestible format so they can make treatment decisions for their patients.
“Having explained the challenges that labs face as genetic medicine and molecular diagnostics become more sophisticated, it will be helpful to remember how laboratory medicine got to this point,” noted Tynan. “Current treatment protocols for HIV offer a useful example.
“It is now common that, during the course of treatment for an HIV-positive patient, a sample will be collected and sent to the lab for sequencing,” commented Tynan. “Medications will be adjusted based purely on the genotype of the virus. And, as the virus acquires resistance to certain therapeutic drugs, the patient’s mix of different medications will be changed appropriately.
“In many ways, the treatment model for patients with cancer will be similar to how we treat patients with HIV,” continued Tynan. “Genetic insights now give us deeper knowledge about the taxonomy of different types of cancer. In turn, this allows us to transform treatment based on the taxonomy of disease.
“You will see a shift in the conversation about breast and prostate cancer, for example,” she noted. “This will come with better understanding of the specific pathways that are disrupted, along with the specific molecular drivers that drive those particular cancers. After all, cancer is an acquired somatic genetic disease.
Cancer Is Complex Disease
“The problem with cancer is that it is far more complex than many other indications,” she said. “Disease heterogeneity is enormous, and this is reflected in complete sequencing data of supposedly discrete cancer states,” explained Tynan as she showed data from 25 ovarian cancer patients. “This creates challenges in how to interpret the information produced from genetic and molecular tests. “That development has a practical consequence,” she added. “It means that any pathologist or molecular medicine physician working on a case will need to add a data integration step to the workflow in the lab. The laboratory professional will need to do data integration before communicating results to the treating physicians.
“It means that any pathologist or molecular medicine physician working on a case will need to add a data integration step to the workflow in the lab.”
“To illustrate this trend, let me provide you with the hypothetical example of a female patient with non-small cell lung cancer,” said Tynan. “This example will give pathologists and clinical lab managers a sense of how cancer treatment is evolving. It will also help me describe how genetic medicine and molecular testing is going to require laboratories to integrate service lines and use information technology in more intense ways.
“Assume the patient relapses 12 months after her initial diagnosis of cancer,” noted Tynan. “She was being treated with epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs). When the oncologist believes her therapy needs to be changed, he or she will ask, ‘what is the best course of action?’
“When treating this patient today, for obvious reasons the oncologist has few options,” commented Tynan. “First, many cancers are poorly understood and most treatment guidelines are based on tumor tissue of origin or histopathology.
“Second, lab medicine has a limited number of molecular diagnostic tests directed toward targeted therapies,” she said. “Third, the standard of care is population-based medicine—the ‘one size fits all’ principle that has been the foundation of medicine for decades.
Multiple Limiting Factors
“Despite these limiting factors,” continued Tynan, “our female patient actually is fortunate because her non-small cell lung cancer has been identified with an EGFR mutation. That makes her eligible for a TK inhibitor. But now her disease has relapsed and she’s entering into a period of uncertainty.
“She likely has a multitude of physicians and connection points to the healthcare system,” added Tynan. “This huge care team is deciding which tests to order, how those tests should be interpreted, and how those tests should be integrated into her course of treatment.
“These are complex issues,” she stated. “Additional complexity comes from the fact that the patient has multiple physicians and there are arbitrary discipline boundaries between them.
“In this example, when the original biopsy was taken, it was sent to the anatomic pathology laboratory where it was reviewed by the pathologist,” she stated. “Then it was sent to the clinical pathology lab where the patient’s materials had to be marked, micro-dissected and go through the process of being tested for an EGFR mutation.
“Questions that arise include: Should we re-test the original biopsy for additional mutations? Can we request a new biopsy? Are there blood-based markers that might inform this next treatment decision analysis?” noted Tynan. “These questions demonstrate how the treatment of cancer will be more complex, along with the need to manage the patient on a longer term basis, like someone with a chronic disease.
“I would argue that discipline boundaries in medicine today are becoming less rele- vant because all these steps described above are part of one field—that of molecular medicine,” observed Tynan.
“The point is that this current model is unsustainable and that’s where the opportunity lies for pathologists,” she emphasized. “The iterative testing we do today involves moving samples around and issuing individual reports at various stages.
“Without a sophisticated level of data integration, we pile cost upon cost,” continued Tynan. “Current processes and workflows are not particularly helpful for moving the treatment decision forward.
“Consider how different this patient’s care will be in three to five years,” she said. “Now a rising number of patients—such as this woman—will benefit from more comprehensive molecular diagnostics that repurpose existing and emerging therapies. We’ll have a deeper understanding of molecular pathways that drive confidence in those treatment recommendations.
“The combination of mutations and other test findings will be very specific to the individual patient,” continued Tynan. “Some of these markers will be actionable and some may not be actionable. The pathologists working with these markers will be aided by very small, clinical studies that will group these patients together.
“This is where information systems become critically important,” emphasized Tynan. “It will be incumbent upon us to prospectively track outcomes in these patients because the groups will be narrowly defined based on their molecular profiles. That’s the reality of personalized medicine.
“Here is the important change in laboratory medicine,” emphasized Tynan. “Going forward, therapy based on tissue of origin will give way to therapy based on a particular pathway or combination of pathways disrupted.
“That is why, in the age of personalized medicine, we need to retain that information for future data mining,” she said. “This is the information that helps us understand how the diagnostic testing and the course of treatment benefited individual patients.
Looking Ahead 10 Years
“In seven to 10 years, we will have a deeper understanding of our patient’s non-small cell lung cancer and her physician will make more precise decisions about her care,” she continued.
“This care will be more precise because we will no longer have a one-size-fits-all approach,” she said. “Molecular data from multiple biomarkers will allow pathologists to prepare a list of treatment options.
“Remember that smaller cohorts of patients will generate the rules that physicians will use to prescribe an expanded selection of targeted drugs,” observed Tynan. “Clinical labs will need to offer the range of diagnostic tests required to support these therapeutic decisions.
“Expect to see studies produce a growing number of defined molecular signatures that physicians will use to monitor each patient’s disease as she/he progresses through treatment,” she said. “Testing for molecular signatures expands the lab’s role and creates an ongoing clinical relationship with the referring physician.
“Also at this point in the future, I anticipate there will be knowledge databases tracking inputs and outcomes,” said Tynan. “There will also be automated interpretation and decision making systems.
“At the 10-year mark, if we don’t have systems such as IBM’s Watson helping us, then we’re never going to keep up,” she confided. “The rate at which the information is being generated in this field far exceeds our ability to read the literature and stay up to date with it.
“To summarize, laboratories should view these developments in the following time frames,” advised Tynan. “During the short-term—meaning three to five years—a deeper understanding of biological pathways will occur that drives new treatment options.
“In turn, this new knowledge will give physicians more confidence in their treatment recommendations,” she commented. “Growing numbers of patients will benefit from diagnostics that repurpose existing and emerging therapies. Using these therapies in combinations will become the standard of care. Oncology will be the first medical specialty where these approaches become well-established.
“In the longer term—say seven to 10 years—both labs and clinicians will have access to richer sets of data on patient populations,” continued Tynan. “These data will be generated from patients treated by targeted therapeutic interventions.
Two Changes for Labs
“Finally, to serve this evolution in personalized medicine, clinical laboratories and pathology groups will need to make two changes discussed earlier,” she stated. “The first is to build a more robust informatics capability to analyze and manage vast quantities of molecular and genetic data and track patient outcomes.
“The second is to hire molecular pathologists and molecular scientists to work with clinicians to interpret the data and help them identify therapeutic options,” Tynan concluded. “As this process moves forward, it is likely that one shift in laboratory medicine will be to use individual molecular signatures to track disease and support less-invasive patient monitoring.”
Predictions Are that Oncology Will Evolve To Manage Cancer Like a Chronic Disease
CLINICAL LABS AND PATHOLOGY GROUPS are positioned to play ever more important roles as personalized medicine becomes a reality, particularly in oncology. That’s the prediction of Katherine Tynan, Ph.D., President of Tynan Consulting.
“Personalized medicine will expand the role of laboratory professionals,” she stated. “One role will be the classic opportunity to assist oncologists in diagnosing disease. Another role will be in helping oncologists to manage the long-term care of patients dealing with a chronic disease.
“If we’re truly successful with personalized medicine, then cancer will become a manageable chronic disease,” explained Tynan. “In this scenario, if pathologists are not engaged in working with oncologists in molecular diagnostics, then those pathologists will not be involved in the lifetime treatment and the lifetime testing of that patient.
“This will occur because it will be necessary for laboratories to identify the individual molecular signatures specific to each patient’s disease,” she added. “The tip of this iceberg today is circulating tumor cells (CTCs). However, disease markers that are even more specific are on the horizon.
“Oncologists will want to run tests to check those signatures every few months to monitor patient response and disease progression. If your lab is not involved in establishing that molecular signature, your lab won’t be involved in the ongoing testing required to provide future care to that patient.
“Progress can be seen in published studies,” noted Tynan. “One example is an article from the New England Journal of Medicine about a group of physicians and other providers at the Johns Hopkins School of Medicine (NEJM 364:4 Jan 27 2011).
“The authors described how individual molecular signatures were identified at the time of disease diagnosis, then were used to monitor disease over time,” she explained. “This approach allowed them to get early insight into disease recurrence. This knowledge was used to change that therapeutic intervention and bring the disease back down to a more manageable state.
“Similar progress is being made in other translational medicine programs,” continued Tynan. “One study published in Translational Medicine (Sequist et al., Sci Transl Med 3 75ra26 (2011)) described how a handful of patients carrying specific EGFR mutations acquired resistance via two independent pathways. One pathway was an additional EGFR mutation which interfered with binding of the drug to the receptor. A second pathway were mutations in PIK3CA that resulted in a morphological shift to a mesenchymal cell type that responded well to chemotherapy.
“In each case, by virtue of having the molecular definition of disease, the researchers could alter the treatment and reduce the tumor burden,” she explained. “Technically, it wasn’t a stasis on the disease but rather resistance or insensitivity developed to a drug. As was reported, the tumor could flip back and forth between being sensitive or insensitive to TKI’s.
“These examples show the progress clinicians are making in treating these diseases,” noted Tynan. “At the same time, this progress is opening new doors for pathologists and laboratory scientists to provide greater clinical value to physicians.”
Advances in Molecular Testing Give Labs New Ways to Support Physicians, Patients
PATHOLOGISTS SHOULD KEEP IN MIND that the clinical services Katherine Tynan discussed in these pages are already available.
“Currently, targeted oncology gene panels are available at many academic medical centers,” she said. “While the cost of this care is high, costs for sequencing and instruments are coming down rapidly. Reagent costs are already relatively low.
“Therefore, we will see more oncologists requesting this type of testing so that they can make ongoing treatment decisions,” added Tynan. “One academic medical center has a personalized oncology pilot program that takes about 27 days to complete. They have a sequencing tumor board that discusses and interprets the findings in the context of the patient’s clinical presentation.
“Pathologists who believe such work is prohibitive because of high cost will be surprised to learn that this work-up can now be done for $3,600—a dollar amount that is well within the range of many tests in clinical use today,” emphasized Tynan.
“The cost of sequencing is no longer a significant barrier when you consider that the typical work-up cost for a leukemia patient is $3,400,” she concluded. “The cost of interpretation and data integration in a typical hospital outside of the major academic medical centers is the big unknown. This is one opportunity for local pathologists.”