Choosing a Precision Medicine Solution
5 Questions Clinicians and Researchers Should Ask Up Front
Clinicians and researchers are understandably apprehensive when choosing a precision medicine solution. Who can blame them?
In addition to establishing whether the solution can effectively combine all relevant health determinants about a patient and treat a population’s unique circumstances, there are a number of role-specific needs to consider before fully investing in the solution-adoption process.
To enter that process properly, I’d like to suggest five essential questions these prospective users should ask themselves right from the start.
1. Do I need the ability to reference genomic data?
The size and complexity of genomic data can create a real issue for the clinician or researcher who wants to warehouse and handle it. But if your answer to this question is yes (or potentially yes), then ask yourself whether there’s a benefit from handling all of the other clinical data types, too. To this day, there are still uncertainties around how genes should impact decisions, so responsible clinicians and researchers should still compare genomic data with traditional clinical data in order to truly be practicing precision medicine.
For example, consider a gene variant for Alzheimer’s disease—by itself, what does that even mean? All that it tells us is that a patient has a variant of a gene that has been linked, to some extent, to the development of a specific type of Alzheimer’s disease. This may be enough information for some purposes, but certainly isn’t enough information to base a clinical decision on. (On a side note, exciting research looking at the clinical potential of coupling genetic variants, disease phenotypes, and multi-scale brain models has recently shown the potential advantages of combining multiple types of information from multiple sources to better understand the patient and their condition.) Combined with a range of standard clinical data and the emerging results of new investigations, knowledge of the variant’s existence in a patient could well contribute to a clinician’s precisely focused decision-making process, prompting them to begin therapeutic or preventative action. Putting the patient on a genomic-variant registry is another important step toward enabling the records of patients with very specific conditions and circumstances to be easily retrieved, and it will ultimately ensure that those patients will benefit from new treatments as soon as they’re available.
2. Will I focus on research or clinical use, or perhaps both?
For researchers, any loose end matters, so they’ll want a fully comprehensive view of all data that can possibly be collected. Does it include all of the genomic data that they’ll need in the course of their work, and does it have reliable continuity over many years? Contrast these needs with those of the clinician, for whom a fully integrated view of the relevant data is key: in addition to all pertinent genomic data, they’ll want full patient histories, exam findings, lab results, and more to inform their assessments and care plans.
For example, a pulmonologist using a precision medicine solution to do research would want a complete lung cancer genomic data set to inform their work, while a clinical pulmonologist using that same solution would want their work informed by a complete lung cancer genomic data set integrated with a portfolio of high-quality, patient-specific data.
3. How might I use the solution in clinical trials?
With tens of thousands of publicly available genomic sequences of varying importance and impact, a prospective user choosing a precision medicine solution for clinical trials must not only ensure that their genomic data sets are rigorously scored, but that they’re actually appropriate and presentable. They need high quality clinical and geo-localization data for every patient in order to determine whether those patients should be included in the trial.
For example, one of the largest U.S. clinical/genomics trials registered with the clinical trials database, and sponsored by the National Cancer Institute, has enrolled 3,000 patients to study gene expression of lymphoma, leukemia, and multiple myeloma. In this trial, DNA sequencing methods are being used to analyze base changes in the genome of the cancer cells. While there are several existing reports that have described the sequencing of whole genomes from a few patients, the greater number of cases in this trial will allow researchers to identify biologically relevant patterns in humans. In these types of newly emerging trials, using a precision medicine solution to analyze and compare detailed genetic information with detailed clinical information is the right approach. That solution would help the aforementioned prospective user to accurately and efficiently exclude and/or include subjects from the data set that the user will ultimately present for review, thereby raising the quality of its results.
4. Will I need to place all the data in a raw data store and make it available as different use cases come along?
If a researcher is running tests for, say, atrial fibrillation, she’ll need the ability to store and link to all of the clinical data in a non-fixed manner so that she can modify the disease’s data model and pull out relevant information on an ad hoc basis. If, in a few years, there’s a breakthrough discovery in the genetic sequence that links to the disease, she’ll be able to refer to that raw data and re-run the original tests. This is a postmodern version of a traditional practice: researchers have long kept freezers full of blood samples from test-trial subjects so that when a new test comes along, they can run the samples and draw new conclusions from the results. Today, the raw data store represents an immense collection of such “freezers,” a repository for data that’s been processed to the best of our current abilities but wisely reserved for another day. For a clinician, the advantage is more straightforward. For example, if a new drug for migraines debuts, they might go through their database, identify ten current patients with difficult-to-treat migraines, contact them, and make recommendations.
5. Will I need analytics to examine an entire population for a specific issue?
A clinician can use a precision medicine solution to analyze genomic data and identify a clearly defined group in their patient population that’s at risk for, say, diabetes, and then intervene by making specific lifestyle recommendations to that group that are backed by data, not just the same old generalized advice (i.e., imprecision medicine) that may or may not apply to this specific patient, yet must be given in the absence of the information. Similarly, a researcher working for a payer could determine which specific members of a diabetic population might benefit from, say, one of the new DPP-4 inhibitors (relatively expensive type 2 diabetes agents), notify or simply add to the registry all of the other members whose genetics inhibit any possible benefits from that drug, and prevent the latter group—and their payers’ wallets—from investing further in a futile treatment.
While precision medicine solutions offer unique value propositions to countless healthcare professionals, the broad theme stressing the importance of crunching all of a subject’s known clinical data and making precise decisions for groups with common characteristics should be a starting point for the conversation, not an end. The needs of the roles of those who will be using the solution should be taken into account at every step of the adoption process, not merely in the spirit of due diligence, but in the interest of the population those roles are looking to serve. It’s the healthcare professional who asks these and similar questions—and truly reflects on their answers in the weeks and months leading up to their buying decision—that does best by the patient and by themselves, providing a true value-added service that doesn’t come standard on any platform today.
This article was originally published on Orion Health and is republished here with permission.