The use of broad-panel, next-generation sequencing (NGS) proved to be more cost-effective than single-gene testing in patients with non-squamous non-small cell lung cancer, according to a study presented during the recent American Society of Clinical Oncology (ASCO) virtual meeting.
In this exclusive MedPage Today video, the study’s lead author, Nathan A. Pennell, MD, PhD, of the Cleveland Clinic’s Taussig Cancer Center, explains how much of a difference in value his model showed between choosing to run a narrow genomics panel or broader NGS.
Following is a transcript of his remarks:
The background is that in the United States right now, there are seven different genetic alterations that are found in patients with adenocarcinoma of the lung or non-squamous non-small cell lung cancer, that have FDA-approved therapies that are guideline driven and should be tested for in every patient with advanced disease. However, there is a disconnect in terms of testing to identify these, and the most common strategy actually that’s done is to use what’s called single-gene tests, looking at more common genetic alterations like EGFR mutations and ALK gene fusions.
And in fact, studies have shown that we do pretty well with testing for EGFR and ALK — probably between 70% to approaching 90%, depending on which study you look at, of patients who are getting tested for these two. But the minute you drop down to say a third genetic marker, like ROS1, or BRAF, and now we have NTRK and MET and RET gene fusions, we’re now talking about probably less than 20% of patients are being tested for these.
And there’s a lot of reasons for this: You run out of tissue, a lot of places aren’t aware quite that they need to test for this. Really the best way to test for now for at least seven markers is to use something called next-gen sequencing, a broad multiplex test that can look for all of the alterations at the same time.
So, what we wanted to do with this model is basically illustrate the value and the cost of doing broad NGS, for all of these markers on all eligible patients in the United States hypothetically, compared to the current model of really just looking for EGFR and ALK. And using data on basically reimbursed costs for single-gene tests versus NGS, and look at the costs, not just in terms of money, but also in life years lost or gained by patients who were identified and actually went on to get highly effective treatments and live longer because they were discovered to have these targetable alterations.
So, again, this is not a real study of real patients; these are hypothetical patients. So we went through and looked at all of the possible newly diagnosed stage IV lung cancer patients who should be tested for alterations and then looked at different permutations of what percentage of the patients would get either single-gene testing or NGS. We inputted all the costs of reimbursement for testing based on the Centers for Medicare & Medicaid Services costs.
And we estimated that people who were identified and went on a treatment such as osimertinib for EGFR mutations, for example, would live 2 years longer than patients who were never identified and went on to just get chemotherapy and eventually die of their cancer. And we estimated that it would cost $10,000 a year of expensive targeted treatments if they were identified. So trying to make this as realistic as possible, and then alter the assumptions.
So what we basically showed was if you were to substitute at the current, say approximately 80% of patients being tested for EGFR and ALK, if you were to substitute an NGS panel that would test for all seven markers instead, you could save more than 21,000 life years at a reduced cost per life year gained of almost $600 by doing NGS, rather than single-gene testing.
In fact, from the current state of less than 20%, for every 10% increase in testing for NGS, instead of using single-gene tests, we were able to show more than 2,600 life years gained with incrementally decreasing costs for each life year you’re gained by doing NGS. Because it simply costs less to look for all seven than it does to pay for each individual seven single genes.
And then finally, I thought was very interesting, if we were to test every eligible patient in the United States in 2020, for all seven targets using NGS and then treat every patient who is identified at $10,000 a year, it would only cost a little over $16,000 per life year gained to actually do that.
So the bottom line here is, it makes sense to do NGS testing of every patient. It’s not more expensive than we can afford as a society. It actually costs less incrementally than doing the more expensive single-gene tests for all of these. And an enormous number of life years on the one hand can be gained by doing this, but the flip side is tens of thousands of lives are being lost by patients going on and living their lives and then dying without ever being identified as having these very effective treatments available to them.