Clinical utility of your genome
No evidence is no proof
One of the most frequently-used arguments against direct-to-consumer or direct-to-provider (as in a doctor acting on a consumer’s behalf), DNA sequencing tests is the much-touted lack of the demonstrated clinical utility of such tests. I always somewhat scoffed at this argument, because I saw it as a form of a “chicken-and-egg” type of situation. I obviously agree that consumers need to be protected, and not sold bogus tests that do not provide any benefit. However, on an individual basis, the clinical utility of genome sequencing (the entire DNA sequence), exome sequencing (of all the genes in the genome), or the sequencing of preselected group of genes (referred to as targeted gene panel), has been demonstrated copious number of times. But it is true that it has not been demonstrated on a population scale. However, one cannot bemoan the lack of utility if there is not even data yet to state if there are such benefits or not. Before we can use such arguments, tests on a sample of the population have to be done in order to figure out if genome sequencing is beneficial, or potentially harmful. Obviously I am leaning heavily towards the former, although I agree that we should take every precaution necessary to guard against the latter.
The good news is that scientific data has been trickling in which begins to provide clues to the clinical utility of genome sequencing. This data shows that the use of genomic technologies has the potential to be a cost-effective means of diagnosing individuals on a broader scale, as it can provide information faster, which translates into a more effective means of treatment, and saves money by avoiding additional expensive testing.
Clinical utility peeking in
An early publication that caught my attention was one in which the Association for Molecular Pathology Economic Affairs Committee investigated the cost and value of three specific genomic sequencing procedures, by analyzing multiple laboratories. These included the use of a targeted gene panel in advanced non-small-cell lung cancer, another targeted gene panel in sensorineural hearing loss, as well as exome sequencing in undiagnosed neurodevelopmental disorders. This was the first such publication I found that was getting to the bottom of this, so I devoured it completely. Such information appears to be truly sparse (at least to my investigating eyes), so such research is always of great value.
Let’s briefly look at the presented findings, and start with lung cancer. This is the most common cancer in the world, and the leading cause of death by cancer, accounting for nearly 20%. If there is a clinical utility we should all wish to observe, it would definitely be in this area. What was observed was that sequencing specific genes (up to 50), as compared to the conventional approach, helped to direct patients towards targeted therapy by an additional 7% (from 6% to 13%), massively decreased the number of patients with a non-targeted therapy from 83% to 20%, and instead, directed far more people towards clinical trials by identifying promising investigational genetic alterations (increase from 4% to 54%). The overall economic impact per million lives was equated to a savings of nearly $3 million. This looks like a pretty impressive demonstration to me for the beneficial use of genomic technology in patient care.
Case number two was studying sensorineural hearing loss using a panel of more than 50 genes to seek diagnosis as compared to the current standard care. The diagnostic yield improved from 25% to 36%. This translated to $0.25 million savings per million patients based on the average test cost of $1500. Still looking good!
Finally, there was the neurodevelopmental disorders in children. This was of particular interest, because this approach used exome sequencing towards obtaining the diagnosis. Two different scenarios were investigated, where additional tests were added to look for large-scale genomic structural changes as well. The genome sequence can be very malleable to large structural changes, and these can include deletions, duplications, the inversion of information, or a wacky combination of all of the above. As can be imagined, these can result in a great impact, including disease development.
The authors investigated the use of such a test first, followed by the exome sequencing, or vice versa. In both cases, the diagnosis was improved over the conventional approach by 30% to 40%. Assuming that the average exome test cost is $2500, the first approach per million patients resulted in a $1.3 million dollar savings, and in the second scenario (in which the exome was being used first), resulted in nearly $1 million dollars in additional savings. However, if the cost of exome test was at $1500, in both cases this would result in more than $10 million dollars in cost savings! And that is not unreasonable to expect for an exome cost currently, even in a clinical setting.
So the verdict is in for these stated populations of patients, and there was definitely great data delivered! But I also delved into whatever references for similar such studies were provided in that article, and they are presented below.
Saved money, enhanced quality of life, better treatment access - what's not to like?
The second example looked specifically at how the testing of multiple genes versus the testing of a single gene could affect the therapy selection in patients with metastatic melanoma. While the authors used real test results, the economic impact of therapy - as well as the impact on patient outcome - was modelled. How people were affected was measured in quality-adjusted life years (QALY) where 1 QALY is equivalent to one year in perfect health, while 0 QALY means you are dead. Anything in between the two indicates imperfect health and quality of life. Oh, and you can have negative QALY values too, meaning outcomes that would be considered worse than death. Not something I would want to ever ponder about, but I am sure some people know what that means. Like Joker? Or Prometheus?
The authors showed that the gene sequencing panel strategy cost $8943US less per patient, increased QALYs by 0.0174 per patient, and these results were 90% likely under almost any consideration in the model. So just like in the example of lung cancer patients above, broader genetic investigation was able to guide more people towards improved metastatic melanoma targeted treatment, or clinical trials. This might not sound like much, but if you were to extend this to the entire population, where melanoma is one of the most common cancers (8900 patients diagnosed in the US each year), now we are talking of an annual savings of $79.6 million US and a gain of 155 QALYs. That was just taking the test and drug costs into consideration, so if other medical costs were to be included, the impact would likely be even more significant.
The final example is the one that captivated me the most. This study looked into the use of either whole genome or whole exome (only genes in the genome), sequencing in undiagnosed children and infants with neurodevelopmental disorders. We are talking of a broad category that includes intellectual disability, developmental delay, autism, and so forth. 100 families were investigated, with 119 children, where parent-child trios were sequenced to improve the chances of a diagnosis. As the parents were healthy and asymptomatic, one would suggest that a child had to obtain mutations affecting the same genes from each parent to produce the disease. By being able to follow what genetic information was passed on from each parent to a child, then any suspect data that is observed in the child but also seen in only one parent, is likely not a cause of the observed symptoms in the child. You can also try to find alterations in the genome in the children that is not seen in either parent. Families received immediate whole genome sequencing based on the acute severity of symptoms of the child, or if the offspring were still infants. The rest received whole exome sequencing.
The data presented was most convincing. A definitive molecular diagnosis was established for 45 of the families. That included 40% of families with children with non-acute disorder and 73% of families with acutely ill infants. This is not the shocking information yet! For the non-acute children, there averaged 13 prior other examinations with these test costs averaging $19,000 (in the range of $3,248 to $55,321 for the tests only, and, once again, omitting any other health care costs)!
This means that, taking the observed diagnostic efficiency into account, genomic sequencing of a family trio is cost-effective up to the $7640 investment spent on their DNA testing. Even for the infants, they averaged 7 prior diagnostic tests with a mean cost of $9500. Imagine the stress burden placed on such an early life. The especially encouraging observation is that the molecular diagnosis for these infants was delivered very early on in their lives (with a median of 50 days!), and for nearly 50% of all diagnosed families, the diagnosis altered the patient management of their condition.
Another revelation that might be shocking is that 51% of the diagnoses included de novo mutations, meaning they were not inherited from the parents, but arose spontaneously. However, a high incidence of de novo events leading to neurodevelopmental complications appears to be a hallmark of such disorders.
Modern holistic definition of clinical utility
One can argue that these assessments do not really investigate the clinical utility which should be evidenced by the observed benefits from treatment. By such strict criteria, the time span to collect such information would be quite long (the entire lifetime of individuals studied), but here is where the definition of utility should be examined. According to a recently published definition by the American College of Medical Genetics and Genomics, who are the authoritative body of medical genomics guidelines, the ACMG defined clinical utility on a much broader spectrum of societal benefit. In their elegantly crafted statement, “clinical utility of genetic testing and services should take into account effects on diagnostic or therapeutic management, implications for prognosis, health and psychological benefits to patients and their relatives, and economic impact on health-care systems. We believe that clinical utility must also take into account the value a diagnosis can bring to the individual, the family, and society in general.”
I am a big fan of this approach because, indeed, clinical utility should not be defined solely on whether an individual will be successfully treated, as additional benefits can become evident just from obtaining a successful diagnosis. The easiest one that instantly jumps to mind is the ending of further unnecessary testing if a diagnosis is obtained. As evidenced in the examples provided, such diagnostic quests can have an immense impact on the resources and people involved, beyond the patient alone, who obviously deserves the greatest sympathy. Another benefit is that other family members can be tested if necessary.
The benefits to an individual and their family can extend further in more subtle ways as well, such as proper reproductive planning based on accurate risk assessment. The knowledge of the conditions involved can alleviate the distress of uncertainty, can allow for improved preparations towards the future, or lead towards appropriate support, all of which can offer valuable psychological benefits. As the ACMG states, when treating an afflicted individual, it is the interests of the entire family that should be regarded in clinical considerations.
With that in mind, the clinical utility as demonstrated in these examples becomes even more obvious, and it is exciting to see that support for the adoption of genomic medicine on a wider scale is beginning to trickle in. Nevertheless, these notable examples are still case-specific, and it does not mean that they translate to the broader general population, the majority of which is asymptomatic. However, it does start painting the picture, that if the technology was used at the right time with the right person, the clinical utility becomes evident. Now the big question will be if that utility would persist on a whole population level, at least in terms of the financial investment needed, if every person was theoretically tested to identify high-risk individuals. There are now institutions, and even countries, that are betting yes to that question.
And so is Merogenomics, which is why such genomic information is studied everyday, and why Merogenomics can provide assistance to those individuals who seek access to genomic sequencing technologies.
This article has been produced by Merogenomics Inc. and edited by Kerri Bryant. Reproduction and reuse of any portion of this content requires Merogenomics Inc. permission and source acknowledgment. It is your responsibility to obtain additional permissions from the third party owners that might be cited by Merogenomics Inc. Merogenomics Inc. disclaims any responsibility for any use you make of content owned by third parties without their permission.
Products and Services Promoted by Merogenomics Inc.