Biomarker Testing Can Direct Care, but Only If Clinicians Perform the Right Tests

https://doi.org/10.37765/ajmc.2020.42555The promise of precision medicine calls for directing the right therapy to the right patient based on genomic profiling to identify mutations that will predict responses and guide the course of care. But these opportunities are lost if clinicians do not perform biomarker testing or if the testing is insufficient.

How often does this happen? More often than cancer specialists may realize, and payers may be part of the reason, according to Stuart Goldberg, MD, a hematologist/oncologist and chief of the Division of Outcomes and Value Research at the John Theurer Cancer Center at Hackensack University Medical Center in New Jersey.

Previously, Goldberg was involved with COTA Healthcare, a company conceived and built by cancer physicians who wanted to tap electronic health records for insights that would guide better care. As reported last year by Evidence-Based Oncology™ (EBO), COTA has assembled a database of records from both academic centers and community practices, representing the breadth of care offered in the United States.¹

Goldberg and several coauthors used the COTA database for a study whose results showed that patients receiving cancer treatment are not being tested for all the relevant mutations in evidence-based guidelines. He recently spoke with EBO about the findings, which appeared in JCO Precision Oncology,² published by the American Society of Clinical Oncology (ASCO).

The following has been edited slightly for clarity.

EBO: In your study, you note the promise of genomic profiling for mutations that can predict outcomes or response to treatment. But you found that biomarker testing rates are suboptimal. Can you briefly describe the results?

Goldberg: We’ve known for many years that biomarker testing for genomic mutations is an important part of treating colon cancer. This is an important part of the ASCO guidelines; they’ve been part of the [National Comprehensive Cancer Center] guidelines and the [College of American Pathologists] guidelines for many, many years. And we know that, for example, [patients with an EGFR mutation] may not respond to certain monoclonal antibodies. So for the main part, the testing makes the patient not eligible for certain therapies; we don’t, therefore, give them expensive treatments that aren’t going to work.

We wanted to find out [whether] these biomarker tests [are being done] in the real-world community. So we went to the COTA database, which is a large database taken from the electronic health records of patients throughout the country.¹ We looked at multiple states, in both academic centers as well as community [oncology] centers. We pulled the electronic health records for patients with newly diagnosed colon cancer, metastatic colon cancer, and we looked to see [whether] patients had the genomic testing that was done for the year in which [they received their diagnosis] because over the past decade, we’ve added new markers and new genomic [markers]. So we wanted to make sure that the patient in that year—whatever year they [received their diagnosis]—got all the [biomarker tests] recommended under the guidelines. One hypothesis was that…as doctors got more familiar with genomics, the rates would go up. What we found, however, when we looked at over a thousand patients, was that the rates did not go up, and a little bit went down. Yes, doctors were more familiar with genomics and were ordering the tests more often, but because more markers [and] more genomic mutations were now required, they weren’t keeping up and getting all the right testing. So only about 40% of patients were actually tested for all the genomics recommended for that particular year in this retrospective chart review.

EBO: What are some of the barriers to testing, and what are some of the possible solutions? In other words, what can be done to help educate these doctors about the newer tests now available?

Goldberg: In our study—we were doing a retrospective study—we were looking at the charts, so we really couldn’t say exactly why the tests weren’t being done. But we have some hypotheses from other work we’ve done in the past. Part of it is education. Physicians do know that genomics are part of this, and they were ordering tests, but new tests keep coming to them. For markers such as BRAF and HER2, which are some of the newer ones being added to the guidelines, physicians aren’t catching up. They’re ordering the older tests but maybe not catching the newer tests. So [continuing medical education], paying attention to what’s changing, and watching the guidelines are important for [physicians] to stay up-to-date on what needs to be done.

The second piece is educating the insurance industry. The insurance industry wants to pay for as little as possible. And as you know, if you have to order individual tests and the insurance industry says, “Well, we only want to pay for tests A and B, but we’re not going to pay for these large panels,” well, they’re going to end up missing things. So we have to really explain to the insurance industry that payment for these tests in the long run is good for the patient and may be cost-effective. What we saw in our study is that if every single patient had undergone the testing, it probably would have been cost-effective. It actually would been cheaper because many of the patients who weren’t tested ended up getting the monoclonal antibodies that don’t work. So you’re paying for an expensive treatment that doesn’t work and giving the patient no benefit. In the long run, that money could have been used to get the right tests for everybody in the whole group. And it wasn’t that [expensive] because the tests are relatively inexpensive compared with the cost of the therapy.

We also need to work on coordinating care among the different physicians. Often the [gastroenterologist] does the biopsy. That biopsy goes to the pathologist; the hematologist-oncologist doesn’t have access to it. It’s done in a different hospital that can’t get the tests. Now we’re stuck with, “Oh, do we get a liquid biopsy, [or] do we try to track down the tissue?” So there’s a lot of coordination that needs to be worked out in not only colon cancer, we found, [but also] in lung cancer. Coordination of care is still fragmented in our society, especially when patients cross different specialties and different hospitals.

EBO: Does the process become even more complicated if a patient has more than 1 type of cancer that needs to be profiled?

Goldberg: Fortunately, many patients don’t have more than 1 cancer. But the more complex the disease is, the more complex everything becomes. It’s really not so much a question of different diagnoses, but crossing medical systems. [If patients are] treated in one hospital system and then go to another hospital system, getting that biopsy [and] getting the coordination of care often become very difficult. Because, as you know, our electronic health records don’t talk to each other, and our insurance [carriers] and our doctors don’t talk to one another. So coordination of care becomes a big issue in trying to do simple things like getting genomic profiling.

EBO: What are the implications for patient care if insufficient genotyping occurs?

Goldberg: If we want to move to the world of precision medicine, which is where we want to be, [we need to give] the right therapy to the right patient, and I also argue in my other hat that we should be giving it at the right value. But if we want to give the right therapy to the right patient, we need to be able to know what that [patient’s] genomics are. So we really need to be thinking about getting all the genomics and explaining to the patients the importance of this, and trying to make sure we get all the right tests. And that’s another area where we are lacking, in explaining the importance of these tests to the patients.

EBO: You just mentioned your other hat that you wear in your work. Would you say that precision medicine is perhaps another form of value-based care? And can you describe the work that you’re doing now, as you’ve been in a new position for about a year?

Goldberg: At our hospital, I run our new Division of Outcomes and Value Research. We realized that as medicine moves to a value-based world, we really need to have somebody at our center who focuses on the outcomes, the value. [Over are] the days of doing old-fashioned chart reducing and once a year writing my [American Society of Hematology (ASH)] paper or writing my ASCO abstract and sending my residents or fellows or students [to the meetings]. That’s gone. We have electronic records, so we can mine the data in real time for all our patients and learn what we’re doing. And we hope that by doing that, we will actually see the better outcomes—see what things are working, what things aren’t working—and then start moving toward better outcomes, better value.

This is important, not just for academics but also for our pocketbook. Our practices participate in the Oncology Care Model [OCM],³ and we’re going to get paid based on whether we are practicing value-based medicine. Are we being the most effective and most efficient in what we’re doing? So we felt at our center that we needed a physician who’s going to be looking at the outcomes, who’s going to be able to talk to the other physicians in our practice and say, “Hey, you’re doing this, but you know, this biosimilar might be cheaper and give you the same outcomes,” or, “Spend a little more money on a genomic test because in the long run, yes, it’s going to be cheaper for the patient.”

We actually proved that at our center several years ago when we did a study on Oncotype DX.⁴ [It costs] several thousand dollars [to do each] Oncotype DX test. We would say, well, we’re going to add several thousand dollars to every single case of breast cancer. Well, it turns out that Oncotype DX often will move patients from getting expensive chemotherapy to getting less expensive and equally as effective hormonal therapy. So [by performing] the test on every single patient, we ended up moving a lot of patients from chemotherapy to hormones. The patients benefited, and in the long run, it was cheaper. Therefore, we were much more value based, and it helped our practice.

Now, you could argue that in a patient who is 9 years old, you’re not going to [give] chemotherapy, so there’s no reason to do an Oncotype DX. Don’t spend on the genomics. Likewise, if you have a stage I early tumor, very small, you probably don’t need to do the Oncotype [DX test] because you’re not going to give that patient chemotherapy. But in stage II disease, which is what we found in our analysis, where the patient was in the middle, where the doctors—we actually looked at what our doctors were doing— were giving a lot of those patients chemotherapy, and the Oncotype DX test moved enough of them over that they saved money that it paid for the genomic testing.

EBO: You mentioned the OCM. As you know, next year and for the next several years, your practice and many others will be looking at Oncology Care First (OCF). How will the use of Oncotype DX and genomic profiling play a role in savings for your practices?

Goldberg: The [current] OCM…is a 6-month model, which is a little dangerous for some physicians when we talk about genomics because if you’re going to spend $3000 or $4000 for the test now, and you’re going to be graded, what happens in the next 6 months, you’ve got to recoup that money pretty quickly, to make it balance.… We actually found that we saved enough money in those first 6 months to make it worthwhile. So [not only do] some of these genomic tests, although they’re expensive, help the patient, but we actually could [achieve savings] that fit within the short models. Our hope is that as value-based medicine becomes more common, we’ll go away from a 6-month model to a longer model where you can start doing things and hope it’s going to help patients down the road. But right now, we’re stuck with these short models. That’s one of the problems with it, with some of the models, but we’re still learning in the value-based world. So I think we’ll see that.

EBO: Have you seen a big difference regarding the uptake of biosimilars when it comes to genomic profiling and these other tests?

Goldberg: Biosimilars are another important area in value-based medicine, meaning that when we look at the OCM, [we ask], “What are the things that are driving up the costs?” Drugs certainly are at the top. And if we can shift some of those drug costs down by using biosimilars, that [is] going to help everyone. Now the [issue] is, when we change from one drug to a different drug, we don’t want to do it based on just cost—we want to make sure we’re continuing to see the same outcomes.

That’s why looking at value is important. To the medical economist, value is a very specific term—it’s outcomes divided by cost. You must have a similar outcome or a better outcome. I don’t mind paying a little more for a brand-name white bread because it tastes a lot better than store-brand bread. I feel that I’m getting my value. I’m getting my money’s worth. But I don’t want to be paying extra for something [that is not worth my money]. Value is not just what’s cheapest. Unfortunately, I will tell you the OCM model is often based a lot more on cost—trying to reduce costs—than looking at outcomes. In fact, I think the biggest problem with the current OCM is that outcomes are not even considered.

EBO: Do you think the OCF is an improvement?

Goldberg: OCM is the first step. And, hopefully, the OCF model, will be another iteration. We learned a lot from the OCM. January 2020 is a critical point because now we’re going into 2-sided risk, and our practice decided to go to 2-sided risk.⁵ This is the first time that the OCM is going to hit our pocketbooks—until now, the first 3 years were only 1-sided risk. There was no downside to it. So [by] participating, you got your [Monthly Enhanced Oncology Services] payment, and you got your $1000 extra per patient. And that was supposed to help you transform your practice to do better things. Now we’re going to see whether all that practice transformation translates to [practices learning to] reduce costs as we go to 2-sided risk, now that we really have to put our money where our mouth is. We’ll see. If that’s successful, we might see other practices join the next iteration of the OCM, which should be coming in the next 2 years.

EBO: Going back to the study on metastatic colon cancer, did any other findings surprise you?

Goldberg: I was very surprised by that study. I knew from our work in lung cancer that we weren’t going to see all the patients being genotyped properly. We knew that already. We know that education is lacking.… Our hypothesis in doing the study was that yes, doctors were familiar [with the fact] that there are certain genes you have to test for in colon cancer and that over time, [doctors] would learn, and they would get better and better. What I didn’t anticipate was that over time, we got new genes that needed to be added, and physicians weren’t keeping up. That means doctors actually ordered more genomic testing at the end of this study than they did at the beginning. But did they get all the tests they were supposed to get for that particular year? The answer is this actually went down. So yes, we’re familiar with genomics. No, we’re not keeping up with which genomics to order. And I think that was the big surprise to me. I expected [physicians] were going to get better on both sides, and they didn’t.

EBO: Whose responsibility is it to alert physicians, to educate them about the new tests coming out?

Goldberg: I think that’s the doctor’s job. … And if you don’t feel comfortable and know what the latest changes are, you have to go to continuing medical education. You must pay attention; you’ve got to do your work. The field is changing. And it is very, very difficult for oncologists and hematologists today to keep up, especially as things are changing. [It’s hard] to just get back from ASH, to just get back from ASCO, and [see] whole things change overnight. And that’s part of the challenge of our field, which is actually what makes it exciting for many of us. But at the same time, you have to do the work; you have to keep up. And you have to be looking at the guidelines, and you have to be saying, “OK, I know something I knew 2 years ago, but what has changed when I see the patient in front of me?” For each patient you see, you have to go back and start all over again. And that’s what I think we learned from the study, that doctors now [know] they’re going to do genomics—but then we’re using what they learned 2 years ago or 5 years ago and not saying, “OK, genomics is now part of it, but what are the tests I have to order today?” And they’re always lagging behind, and that’s not good for the patients.

EBO: Any final thoughts on educating payers on this issue?

Goldberg: Do I think payers have been the people who have been lagging behind and have been an impediment? Frankly, there’s no question. They often will tell us that they want to pay for an individual tissue marker—they will pay for just the 3 tissue tests and won’t pay for the big panel. But we know that there are many drugs in the pipeline. There are big basket trials being run by ASH and by ASCO, by commercial pharmaceutical [companies], that say, “Look, do a whole gene panel. And depending on which genes mutated, you might be eligible for this particular trial, that particular trial, with these new drugs.” That is world we’re moving to. The approval of immunotherapy for [microsatellite instability-high or mismatch repair-deficient solid tumors], where it’s approved across all cancers,⁵ just based on the genetics has really changed the field. We’re now we’re starting to think of cancer not just as breast cancer, colon cancer, or lung cancer but as a genetic genomic disease, that based on the mutation, we’re going to take different drugs.

That’s a culture shift that is happening in the oncology world. We’re going to have to explain that and teach that to our insurance carriers. Otherwise, they’re going to just clamp down, and we’re never going to get paid. But I think we’re starting to see that sea change even in the insurance industry, but it’s something we’re really going to need to work on over the next several years.

EBO: Is there anything else in oncology that you’re excited to see in 2020?

Goldberg: As a hematologist, I’m extremely excited that the [chimeric antigen receptor (CAR)] T-cell treatments are coming out. And the whole world of immunotherapy—in the solid tumors, yes, with the immunotherapy drugs in hematology, this is another quiver to our group—we have surgery, radiation, we had chemotherapy when I was training, then we added transplantation. Then we added targeted therapies. And now we’re seeing the development of immunotherapies. And very shortly we’re going to see the genomic diagnostic being revolutionized. Where I think we’re going next over the next several years is the data, the mining of the electronic health records, and the use of all these real-world data—which we’re just starting to see—this will shape things so that you won’t need to wait 50 years to get a randomized trial, if you start to see the data and start seeing where things are pointing. The data revolution that’s coming will then direct all these other areas of interest.

For an oncologist, this has been extremely fun 5, 6, 7 years, as we’ve added targeted [therapies] and now immunotherapy. And we see some new things happening with transplants and with CAR T. And I think the data revolution is going to really change everything. We’ll have to see whether it’s cost-effective, whether there’s value, because that could be the thing that really kicks us up. Or it could be the thing that really clamps us back down. And that’s going to be the healthcare debate going on in the country. References

1. Caffrey M. Moving toward the “tipping point” for transformation in cancer care. Am J Manag Care. 2019;25(SP2):SP45.

2. Gutierrez ME, Price KS, Lanman RB, et al. Genomic profiling for KRAS, NRAS, BRAF, microsatellite instability, and mismatch repair deficiency among patients with metastatic colon cancer [published online December 6, 2019]. JCO Precis Oncol. doi: 10.1200/PO.19.00274.

3. Oncology care model. CMS Innovation Center website. innovation.cms.gov/initiatives/oncology-care/. Updated January 14, 2020. Accessed January 26, 2020.

4. Waintraub SE, Isaacs C, Norden AD, et al. Confirmation of the TAILORx 21-gene expression trial using a real world observational database. Cancer Res. 2019;7(suppl 4; abstr P4-14-05). doi: 10.1158/1538-7445.SABCS18-P4-14-05.

5. Dangi-Garimella S. A first: FDA approves pembrolizumab for tissue-agnostic indication. The American Journal of Managed Care^® website. ajmc.com/newsroom/a-first-fda-approves-pembrolizumab-for-tissue-agnostic-indication. Published May 23, 2017. Accessed January 26, 2020.

Download PDF: Biomarker Testing Can Direct Care, but Only If Clinicians Perform the Right Tests