Lessons from oncology drive precision medicine across clinical specialties

Assaf Halevy, Founder and CEO of 2bPrecise, LLC

There is little question that oncology has served as a vanguard for precision medicine in the United States. Genetic and genomic data have influenced cancer risk assessment, diagnosis and treatment for more than 25 years.

As we enter a new decade, however, it has become clear that oncology is less a vanguard than a beachhead. Rather than a steadily advancing force, we should view precision oncology as supplying us with experiential insights and evidence-based practices that will equip the industry to make rapid advances into other clinical specialties. In short, precision medicine in general is still in its early stages, preparing to move to the next level.

From cancer risk assessment to targeted therapies

Healthcare leaders often mark the beginning of the present precision medicine era as 1994, when researchers identified mutations¹ in what has become known as the BRCA1 gene in populations of women with family histories of breast and/or ovarian cancer spanning multiple generations. A year later, similar correlations were associated with the BRCA2 gene. Subsequently, studies revealed more than 50 cancer syndromes² affected by mutations in specific genes. These breakthroughs enabled physicians to more accurately identify patients at increased risk for these typically deadly diseases, and initiate appropriate preventative, screening and diagnostic activities.

Additional research in this new era of cancer genome sequencing (somatic) has driven precision medicine even deeper into healthcare. The Cancer Genome Anatomy Project (CGAP) was first funded in 1997 to document the sequences of RNA transcripts in tumor cells. As technology improved, the CGAP eventually added the study of gene expression profiles of cancerous, precancerous and normal tissues to its charge. And, as reported in 2006,³ sequencing was first performed on tumor tissues found in breast and colorectal tumors. With greater understanding of genetic influences and the dynamic behavior of tumor cells, researchers were able to make significant advances in clinical care.

In 2017, pembrolizumab became the first tumor-agnostic drug approved by the FDA.  Since then, several more have entered the market, including nivolumab and larotrectinib. The trend towards targeted molecular therapy⁴  is expected to accelerate in the upcoming years as evidenced by, for example, combination therapies that target specific mechanisms of action simultaneously.

The impact of these discoveries has been significant, contributing in no small degree to the 27 percent drop in cancer mortality in the U.S. over the past 25 years. Because of improved risk identification, cancer may be diagnosed at earlier stages when treatment may be more efficacious. Reduced reliance on trial-and-error approaches means patients respond to therapy sooner and with fewer side effects. Patients can be spared unnecessary and debilitating treatments altogether, as demonstrated by TAILORx researchers finding that not all post-surgical breast cancer patients require chemotherapy.⁵

Expanding value in multiple clinical specialties

With this mounting evidence, scientists and patient advocates, recognizing the potential for genetics and genomics to drive dimensional change across healthcare, are pressing for similar advances in expanded clinical areas. Researchers are continuing to discover genetic/ genomic linkages to diseases at a marked pace: genetic/genomic factors impact no fewer than 2,000 medical conditions,⁶ as noted by the NIH. Further, it was reported in 2018 that 75,000 genetic/genomic tests were on the market with 10 new tests added each day.⁷ Bio-pharmacy and life sciences companies have begun to leverage molecular insights selectively to develop targeted therapies and render these advances actionable.

While it is difficult to predict definitively which clinical areas will adopt precision medicine most quickly, a review of current literature indicates escalating interest in the disciplines of rare diseases (which include conditions such as  cystic fibrosis and cerebral palsy), cardiology, maternal-fetal medicine, pediatrics, neurology and behavioral health.

During its 2019 Building Precision Medicine Summit, KLAS Research collected information from health system executives about the directions their precision medicine programs were taking:⁸

• Mayo Clinic has established a Center for Individualized Medicine (CIM), with a focus on rare diseases and Leadership notes the CIM just completed preemptive pharmacogenomic testing of 10,000 patients, with results made available in the EHR.
• The Geisinger MyCode Community Health Initiative (a biorepository) supports screening for more than 60 genes that meet rigorous specifications for clinical actionability. With a focus on interdisciplinary research and care, MyCode involves participation from the system’s Obesity Institute and the Autism and Developmental Medicine Institute, among others.
• Nationwide Children’s, through its Institute for Genomic Medicine (dedicated to both scientific research and clinical practice), prioritizes undiagnosed conditions and rare diseases in its pediatric population. Tests are ordered in a variety of clinical areas, notably in neurology and gastroenterology.
• Rady Children’s Hospital similarly tests critically ill pediatric patients. During the KLAS event, leaders shared that the facility has sequenced more than 750 children and about one-third of those have received a definitive diagnosis. For approximately 25 percent of those children, the diagnosis resulted in a change in their medical management.

Barriers and accelerators to enterprise-wide precision medicine

As this important and necessary movement accelerates, healthcare leaders must acknowledge that, as precision medicine in practice becomes a standard of care for certain cancer genotypes, precision medicine is gaining acceptance in other therapeutic areas across many organizations.

In a paper produced by McKinsey & Company, “Precision Medicine: Opening the Aperture,” authors noted that – as in other areas of practice – precision medicine pivots on the transformation of information from data (e.g., genetic sequencing) to insights  (e.g., clinical guidelines) and on to actions such as therapeutic choice (e.g., determination of a care plan).⁹

Following on the scientific advances and  the subsequent development of actionable therapeutic options, the realization of the next evolutionary phase – which we are already experiencing today – lies squarely with care providers that must bridge the final mile between science to daily practice.

During the 2019 KLAS Building Precision Medicine Summit, attendees shared that the greatest motivators for fully leveraging precision medicine were saving lives (64 percent) and improving quality of life (16 percent).¹⁰ They also noted that among the greatest obstacles were concerns about

1. reimbursement,
2. the need to better educate both providers and patients, and
3. the lack of technology infrastructure to help them manage genomic data to ensure it is meaningful and actionable.

Reimbursement

Many providers express concern  about  the cost of testing and who will ultimately be held financially responsible. An organization that has its own in-house molecular laboratory bears the expense of processing and analyzing genomic tests. Even when the facility outsources testing to a reference lab, however, healthcare providers remain concerned about the cost. If fees for external labs are not reimbursed, they pass the cost on to the practice’s patients.

Unfortunately, reimbursement for genetic/ genomic testing lags far behind the rate of scientific discovery. For providers – and, more importantly, patients – to enjoy the full benefit of precision medicine, payers need systematic ways of evaluating genetic tests for reimbursement. Without this information, insurers cannot properly assess how to pay for genetic tests.

The National Human Genome Research Institute (NHGRI)¹¹  notes payers find it difficult to develop  appropriate policies because they are not able to easily evaluate what type of genetic test was performed, whether the test was appropriate to perform and whether the test is scientifically valid. A contributing factor, the NHGRI says, is that fewer than 200 CPT codes exist for the many available genetic tests. This means that there is no straightforward way to bill for many tests or for payers to identify what genetic tests were given. In addition, payers struggle to keep up with the volume of new genetic and next-generation sequencing tests being developed. Finally, they cite the paucity of data available to help them evaluate the economics of genetic testing.

The reimbursement tide, however, is turning. The Centers for Medicare and Medicaid  Services (CMS) is watching the Federal Drug Administration (FDA) as it investigates and approves growing numbers of targeted therapies. Since the FDA’s approval of pembrolizumab in 2017, a multiple of additional therapies have been approved – including olaparib (Lynparza), for example, which is prescribed for patients with advanced or recurrent ovarian cancer and is broadly covered by Medicare Part D and Medicare Advantage plans.

Similarly, the nation’s largest payer, UnitedHealthcare, issued a notable policy decision last August which generated significant buzz in the industry and is widely regarded as a bellwether for other payers in the industry.¹² UnitedHealthcare began covering testing so physicians could match their patients to antidepressant and anti-anxiety medications most likely to work for them based on their genetic profiles. In its rationale for the policy change, UnitedHealthcare cited the GUIDED study (among others) as support for its new coverage decision regarding antidepressants, as reviewed in the Jan. 4, 2019, issue of the Journal of Psychiatric Research.¹³

The largest study of its kind to date, GUIDED included more than 1,100 patients with depression. The study showed that individuals in the GeneSight (a pharmacogenomics multi- gene panel) cohort had a 50 percent higher rate of remission at week eight, a 30 percent higher rate of response and an 11 percent greater improvement in symptoms compared  to those in a treatment-as-usual group. In other words, scientific evidence demonstrated the positive impact of the test and supported economic  value due to faster and better treatment response – prompting UnitedHealthcare’s significant policy change.

In short, payers appear to be following a clear and steady – but cautious – maturity curve in expanding reimbursement for genetic tests and indicated therapies. As this trend continues while, concurrently, the costs of tests diminish, it is likely the reimbursement barrier will gradually disappear.

Education

Many providers still find themselves on the uphill side of the precision medicine learning curve. For most primary care and specialty physicians, education in genomics occurred very early in their training – and then only superficially.

An internet search uncovers no lack of articles exploring, and to some degree lamenting, the lack of provider preparedness. In a 2015 Clinical Genetics article¹⁴ about whole genome sequencing, for example, the authors stated, “Surveys have found that physicians of all types often lack genomic literacy and frequently feel unprepared to use or respond to even single gene testing, especially primary care physicians.” An article appearing June 2019 in the Journal of Personalized Medicine¹⁵ likewise underscored this concern: “Despite reported consumer enthusiasm for genetic testing and national initiatives focusing on precision medicine, several studies have reported that knowledge among healthcare providers, including PCPs (primary care physicians), is lacking. Exposure to genetics/genomics or precision medicine in medical curricula is also limited to the initial years, and then dropping precipitously. For practicing providers, opportunities  to engage with and learn about genomic medicine may be limited.”

Playing catch-up, physicians also strain under escalating patient demand for genomic information. FDA approval of, and heavy commercial promotion around, direct-to- consumer (DTC)  genomic tests (like those from 23andMe, Ancestry and Cologuard) have prompted growing numbers of patients to arrive at medical appointments with questions about undergoing testing or with DTC results in hand. They expect their providers – often PCPs – to be able to advise them, help them understand what results mean and incorporate these new insights into their ongoing care and treatment.

In short, providers need educate not only themselves, but their patients as well. All too often, patients do not fully understand test results – and may even come to erroneous conclusions. Commonly, for instance, patients who test positive for a BRCA mutation assume the result means they will most certainly get breast cancer, instead of understanding it indicates only a higher risk for developing the disease. Conversely, those whose tests are negative may believe they are safe from cancer and forego important screenings. Patients need professional clinical guidance to understand that science may discover as-of-yet-unknown mutations correlating to cancer and that non-genetic factors may also impact their level of risk.

“In other words, scientific evidence demonstrated the positive impact of the test and supported economic value due to faster and better treatment response – prompting UnitedHealthcare’s significant policy change.”

Technology infrastructure

Interestingly, one of the most significant challenges to effective use of genetic/genomic data within clinical settings is lack of information technology (IT). As healthcare leaders struggle to get their arms around the voluminous scientific evidence that supports clinical utility of genomics, they often overlook how they are going to gather, store, manage and share data sets resulting from this emerging field of medicine. In addition, leaders must ensure the IT system can provide protections for privacy and security of their organization and its patients. Industry experts note that issues impacting the sharing of de-identified data must be resolved quickly. Genetic profiles and their interplay with other data points (e.g., efficacy of and outcomes from specific immunotherapies) hold great potential to speed further advances, such as the development of new treatments.

Making relevant data available to genetic scientists – in a protected and appropriate environment – to support research and development is an important consideration. It is likely precision medicine will follow a trajectory similar to that of clinical interoperability, which continues to grow through health information exchanges. It is likely data sharing in precision medicine will evolve across two dimensions:

• Through technology and Precise interoperable care will rely upon health IT platforms developed for adding genomic findings to the minimal data set in a shared medical record based on patient consent.
• Through content and Genomic networks and research-based communities have already begun to share so they can improve their understanding and improve outcomes. Consider research surrounding rare diseases. A researcher working in isolation may be limited to an “n of 1” scenario. Data sharing among colleagues, however, could increase that to an “n of 3,” even for the rarest diseases. Treatment- based precise care insights, based on shared data, will follow this wave.

As noted, oncology is well recognized as the standard bearer for clinical use of genetic/ genomic data. However, the obstacles these specialists have faced when incorporating this information into clinical decision-making are less well-known. One medical oncologist at the Seidman Cancer Center at University Hospitals in Cleveland recently outlined the challenges he and his colleagues face:

• Results from sequencing cancer tissue (somatic results) often are returned to the oncologist in large
• Staff would manually “scrape” the reported genetic variants into a text-based format, which would then be saved in the summary section of the pathology
• Future reference to test results (e.g., when the physician wanted to share the result with a genetic counselor) was also an irritant, this oncologist said. Because the results were not stored discretely in the EHR, time was invested in a paper chase to relocate the document.

• In some instances, oncologists are required to log into a web-based portal provided by the commercial lab that processed the genomic test to review results. Most oncologists work with multiple labs,  each with its own forte as well as its own username and password. Physicians needed to remember who processed which test, as well as those specific log-in credentials.

Unfortunately, these challenges will only be repeated and compounded as more specialties adopt precision medicine. With a wider selection of genetic/genomic tests being ordered across a broad range of departments in a healthcare organization, the influx of results must be consumed and managed so they can be made available – easily and efficiently – to whoever needs them within the typical clinical workflow.

Two additional significant issues can be addressed by information technology:

• The first is governance around the ordering of genetic/genomic Currently, tests are being ordered in pockets throughout an organization. In many cases, leadership is not even aware of which clinicians are leveraging these tests. As more and more payers implement policies around reimbursement for genetic tests (and organizations want to negotiate contract rates), the organizations will need insight into ordering practices, so they can ensure individual providers are compliant and consistent.
• The second involves the use of genomic insights over While an individual’s genetic profile rarely changes during the course a lifetime (rendering test result useful long term), phenotypical status (i.e., symptoms, developing conditions) is highly variable. In addition, as noted earlier, scientific research and therapeutic application in genetics/genomics is progressing at a dizzying pace. Providers must be able to factor in these dynamic discoveries into clinical decision-making, within the context of previous genetic/ genomic testing.

Some industry thought leaders assume EHR vendors will lead the way in solving these challenges. However, it is unlikely EHRs offer a total solution. For example, while some EHRs accept genetic/genomic reports into the system as scanned documents, results are rarely saved as discreet data and therefore cannot be integrated with other clinical information or analyzed for purposes such as cohort identification to support population health initiatives. In addition, EHR functionality was not originally developed to accommodate genomic data – which is returned in a unique vocabulary.

Even standard EHR features like family history documentation are not robust enough to collect and display insights that can help identify which patients could benefit from genomic testing for potential heritable diseases.

While industry experts agree that genetic/ genomic data must be integrated with patient clinical information and be presented with the provider’s typical workflow, they acknowledge that EHRs most likely will play a supportive, rather than a leadership, role in accomplishing this task. Instead, independent, interoperable solutions that can integrate results from multiple molecular labs, render the data meaningful in the clinical context and provide single-click access at the point of care are the likely solution. An enterprise-wide solution likewise prevents genomic data from being locked into a discrete software system such as a specialty-specific precision medicine solution, instead liberating and liquidating the information to be used along the care continuum, across specialties and in diverse care settings.

The need to keep pace with scientific genomic knowledgebases, evolving treatment regimens across multiple specialties, and evidence of treatment efficacy and outcomes also requires IT solutions more sophisticated than an EHR. Even now, developers are investigating how best to incorporate artificial intelligence and machine learning to advance delivery of truly “precise” healthcare. One possible avenue is developing an information capsule summary that can:

• Isolate a specific patient, with complete clinical-genomic information including medical history;
• Apply up-to-date research findings contained in specialized knowledgebases;
• Integrate and link relevant therapeutic options, and present known, statistically supported outcomes;
• Present similar cohorts of patients, along with therapies, clinical status and ultimate

It is not difficult to understand how a provider could use this extraordinary set of information to consider and select a care plan tailored specifically to the patient’s genomic profile and clinical presentation, based on the latest scientific evidence and outcomes achieved in similar patients.

Lessons Learned

Lessons learned in oncology can fuel adoption  of precision medicine across a broader spectrum of diseases and conditions. Market forces are evolving to provide  a fertile environment for this new standard of care, as evidenced by greater flexibility within reimbursement policies, consumer interest and advances in “bench- to-bedside” research. Increasingly, healthcare leaders are adopting strategies to initiate and sustain precision medicine programs by ensuring providers and patients receive the critical information and education. Finally, and key to adoption of precision medicine by an organization system-wide, these same leaders are obliged to learn about options to make sure they have the processes, educational resources, workflows, infrastructure (e.g., IT), and ecosystem in place to promote success for years to come. JoPM

Assaf Halevy is founder and CEO of 2bPrecise, LLC, leading an international team dedicated to bridging the final mile between the science of genomics and making that data useful at the point of care.

 

References

1. Questions About the BRCA1 and BRCA2 Gene Study and Breast Cancer; https://www.genome.gov/10000940/brca1brca2-study-faq, June 2012
2. The genetics of cancer, https://www.cancer.gov/about-cancer/causes-prevention/genetics, Oct 2017
3. The Consensus Coding Sequences of Human Breast and Colorectal Cancers, https://science.sciencemag.org/content/314/5797/268, Oct 2006
4. Tumor-Agnostic Treatment for Cancer: An Expert Perspective, https://cancer.net/blog/2018-12/tumor-agnostic-treatment- cancer-expert-perspective, Dec 2018
5. TAILORx trial finds most women with early breast cancer do not benefit from chemotherapy, https://www.cancer.gov/news-events/press-releases/2018/tailorx-breast-cancer-chemotherapy, June 2018
6. NIH Research Portfolio Online Reporting Tools (RePORT), https://report.nih.gov
7. Genetic Test Availability And Spending: Where Are We Now? Where Are We Going?, https://pubmed.ncbi.nlm.nih.gov/29733704/, May 2018
8. BuildPM19 Implementation Guide, KLAS Research; March 2019
9. The precision medicine revolution, from oncology to large polygenic diseases, Precision Medicine: Opening the Aperture, 51-70. McKinsey&Co. 2019
10. BuildPM19 Implementation Guide, KLAS Research; March 2019
11. Coverage and Reimbursement of Genetic Tests, https://www.genome.gov/about-genomics/policy-issues/Coverage-Reimbursement-of-Genetic-Tests, Aug 2019
12. UnitedHealthcare To Cover Genetic Testing for Precision Medicine in Depression, Anxiety, https://www.clinicalomics.com/topics/molecular-dx-topic/unitedhealthcare-to-cover-genetic-testing-forprecision-medicine-in-depression-anxiety, Aug 2019
13. Does obtaining CYP2D6 and CYP2C19 pharmacogenetic testing predict antidepressant  response or adverse drug reactions?https://www.sciencedirect.com/science/article/abs/pii/S0165178118318523?via%3Dihub
14. Are physicians prepared for whole genome sequencing? a qualitative analysis? https://onlinelibrary.wiley.com/doi/abs/10.1111/cge.12626, June 2015
15. Primary Care Physicians’ Knowledge, Attitudes, and Experience with Personal Genetic Testing, https://www.ncbi.nlm.nih.gov/pmc/, June 2019