Best Practice in Immuno-Oncology Biomarker Testing – OCTOBER 2018 | Part 3 of a 4-Part Series
Biomarker-Guided Targeted and Checkpoint Inhibitor Therapy in Melanoma: Best Practices from the City of Hope Comprehensive Cancer Center
Treatment of advanced melanoma was revolutionized with the US Food & Drug Administration (FDA) approval in 2011 of an anti-cytotoxic T-lymphocyte–associated protein-4 checkpoint inhibitor and a selective BRAF kinase inhibitor, both of which have been shown to significantly improve overall survival in phase 3 randomized clinical trials.1-3 Since then, treatment with anti–programmed death-1 (PD-1) and anti–programmed death ligand-1 (PD-L1) checkpoint inhibitors, as well as other targeted therapies, have also resulted in improved outcomes for patients with advanced melanoma, prompting the FDA approval in 2014 of 2 anti–PD-1 agents for the treatment of melanoma.1,4
These novel immune-based therapies allow for unique approaches to treatment, either as monotherapy, in combination with one another, or in combination with other treatment modalities, such as targeted therapy, electrochemotherapy, or surgery.1 The increasing number of treatment options now available in advanced melanoma has prompted a need to identify biomarkers that can be helpful in selecting patients for specific treatments. However, several challenges are associated with incorporating biomarker testing into the management of patients with metastatic melanoma and other cancers, and guidance in “what works best” is of interest to many oncology and pathology practices, both at academic centers and in community practice settings.
To this end, development of best practices for incorporating molecular biomarker testing in the delivery of precision therapy to patients with metastatic melanoma was the focus of a conversation between Morganna Freeman, DO, Assistant Clinical Professor, Department of Medical Oncology, City of Hope Comprehensive Cancer Center, Duarte, CA, and Vishwas Parekh, MD, Assistant Clinical Professor, Department of Pathology, City of Hope Comprehensive Cancer Center. These 2 experts shared their perspectives on how they work together as part of the cancer care team to deliver individualized therapy to their patients. Because City of Hope is a mix of an academic center and a community hospital, many of the insights from Dr Freeman and Dr Parekh can be applied to both academic and community-based practices striving to meet the challenge of effectively incorporating biomarker testing into the precision treatment of patients with cancer.
Impact of Biomarker Testing on Cancer Therapy
Dr Freeman and Dr Parekh were asked about the impact of biomarker testing on their practice as a medical oncologist and pathologist, respectively, as it relates to treating patients with metastatic melanoma. Dr Freeman said, “From a melanoma perspective, molecular testing for the BRAF mutation has definitely had a huge therapeutic impact. We know that BRAF‑mutant melanomas behave differently than BRAF wild‑type. Obviously, it is an important marker for therapeutic selection as well, because patients who have the BRAF mutation can also benefit from oral TKI [tyrosine kinase inhibitor] therapy. The PD‑L1 question comes up frequently, and there are evolving viewpoints on whether PD‑L1 is appropriate to test for in melanoma, because based on some of our larger trials using checkpoint inhibitors, it appears to not have a major impact in terms of treatment selection and matching patients to therapy.”
When Dr Freeman was asked whether she begins therapy on a patient with metastatic melanoma before biomarker testing results are in, she replied, “It depends on the clinical urgency of the patient’s case. If the patient has metastatic disease, but they have good performance status, and they are not physically or clinically compromised by their metastasis, then I can probably take some time if I need to and go back and see if the appropriate molecular testing has been done. The same can be said for patients with melanoma in the adjuvant setting now that we have TKIs approved for that indication. I think what is sometimes frustrating for the patient is when they come to see me, there is obviously a sense of urgency to that visit and they want to know how soon they can start treatment. If I do not have all of that information, I have to tell the patient to come back, and that can sometimes lead to a delay in therapy by as much as 2 weeks. If I feel that the need to start treatment is imminent, then of course I will start it, and wait for the information to come back and then see whether biomarker testing results change my treatment plan.”
From a pathologist’s perspective, Dr Parekh said, “We now have predictive biomarkers that essentially predict the response to certain therapies; that is relatively new so it definitely has added to the complexity of our work. In addition, we have to consider whether tissue will be needed in the future for new and emerging predictive markers. We try to be more of a steward of the tissue so that we do not order an unnecessary test that would waste it, because eventually, the tissue may be needed for some of these newer predictive biomarkers. We also try to keep up with the literature, because biomarker testing is a very fast-moving field compared with the diagnostic side. With biomarkers, the filter changes every few weeks or months. New literature is published and then the surgeon comes to us and says, ‘Hey, can you order these?’ If we did not know what it was, that would be difficult. Conversely, ordering unnecessary biomarkers can delay patient care. Biomarker testing has definitely added a lot of complexity to a pathologist’s practice, but at the same time, it has made our work more exciting.”
Multiple potential biomarkers for predicting response to immunotherapy may be needed given the complexity of the tumor microenvironment and its interaction with the immune system. To be of clinical value, biomarkers need to be accurate, reproducible, minimally invasive (depending on the clinical situation), and able to direct selection and sequencing of cost-effective therapy. Current biomarker candidates include PD-L1 expression, CD8+ tumor-infiltrating lymphocytes, tumor mutational burden (TMB), neoantigen burden, and gene-expression-profile (Table).1 Although these biomarkers have been shown to provide benefit in selecting optimal therapy, each is associated with significant challenges, including variability in assays, validation of scoring cutoff points, need for relatively large biopsy tissue samples, and slow turnaround time. These concerns are in addition to the challenges faced by medical oncologists and pathologists, as described by Dr Freeman and Dr Parekh.
Best Practices in Communication Within the Oncologist-Pathologist Team
Communication within the cancer care team is critical for successfully implementing molecular biomarker testing in patient care. But where and how does this communication take place? Dr Freeman noted, “A lot of the discussion happens at tumor boards. In addition, communication between Dr Parekh and I occurs by e-mail on a case‑by‑case basis. If I see a patient in the clinic who I consider to be a candidate for adjuvant therapy but he or she has not had BRAF testing, I will e-mail Dr Parekh to ask him to do a BRAF test on the tissue block. Or I will ask him if we have some unseen slides that we can send for additional testing. A lot of that type of communication happens through e-mail. City of Hope has a precision medicine oversight committee that facilitates communication among members of the cancer care team. Many cancer centers have molecular tumor boards tasked with a review of cases where, based on a mutation analysis, appropriate selection of patient therapy is determined.”
Dr Parekh concurred. “We have 2 channels of communication. The first is tumor boards, where we have more extensive discussions about a patient case from all angles; this is our main method of communication regarding newer tests or whether a specific test is appropriate for a particular patient. The second is emails, which are generally used to communicate about other issues. Do we have enough tissue? What is the turnaround time? Should we do a specific test? Are we thinking of offering this biomarker test in the future? We also have a molecular pathology tumor board where surgeons or medical oncologists can discuss the molecular aspects and technical details of patient cases in much greater detail. In addition, if a new mutation has been identified in one of the in‑house assays and they are not sure whether that is actionable, a discussion is held to determine whether there is enough evidence to go after that mutation. The surgical and medical oncologists can bring the case to the molecular pathology tumor board so they can have a more extensive, in-depth discussion with the molecular pathologists.”
Who “drives the bus” in terms of making decisions regarding biomarker testing (ie, which tests to include in a routine, reflexive biomarker panel, and which tests to perform on a specific patient)? Dr Freeman suggested, “I think by and large, it is the medical oncologist because, ultimately, it is information that we need for decision making. There are different ways in which the oncologist takes the lead. It is a dynamic relationship, so there is no hierarchy in the sense that we are all participating in the patient’s care. Most of the time, if it comes down to a decision about whether biomarker testing should be done, that request is going to be coming from the medical oncologist as opposed to the surgical or radiation oncologist, or even the pathologist.”
Dr Parekh noted, “By their very nature, predictive biomarkers should be driven by oncologists. Essentially, that is going to tell them whether a treatment is going to be effective. I think that is a subject that should be driven by oncologists, and that is generally how it is done. Pathologists offer insights about the technical aspects—the feasibility, the turnaround time, whether sufficient tissue is available, whether there is too much necrosis for the test to be done. However, diagnostic biomarkers are a different issue. Pathologists need a biomarker to make a diagnosis. For example, if I cannot make the diagnosis of melanoma, and I want to send tissue for fluorescence in situ hybridization analysis, or comparative genomic hybridization, that is something that the pathologist would take the lead on. Pathologists are the ones that are ultimately going to diagnose it. That is the difference between predictive biomarkers and diagnostic biomarkers; different people take the lead in terms of when and how to order specific tests.”
The Process of Biomarker Testing: Practical Issues
Although targeted therapy is the standard of care for patients with melanoma with actionable molecular target(s), and immuno-oncology therapy may be the standard for some subsets of patients, there remain practical challenges related to successfully implementing biomarker-driven therapy in clinical practice. One issue is the variety of biomarker testing modalities available (Figure).5 Several biomarkers derived from the tumor microenvironment or peripheral blood have been proposed as biomarkers predictive of responses to immune checkpoint inhibitor therapy. Although innovative advancements in assay technology have made it possible to comprehensively profile the tumor microenvironment and peripheral blood, it is difficult for a single biomarker to predict clinical responses and/or serve as a patient selection criterion, necessitating the measurement of multiple biomarkers, and with different technologies (Figure).5
How do the experts at City of Hope cope with these issues? Dr Parekh reviewed their process. “We do not have reflexive testing. If this is something that is not required for diagnosis, we would have to depend on the oncologists to come to us. Once the oncologists decide they need something done, it is typically a very quick process. We have it pretty much streamlined. They can order add‑on testing on our EPIC mixer system. As soon as they order any test, the pathology laboratory would receive the requisition. The clinician would bring the block to us as pathologists, and ask us to decide which particular block or slide to choose. We immediately order the recut or the unstained slides as necessary. If it is in‑house testing, it starts right away. If it is a send‑out test, then it generally goes out the same day. Once the reference laboratory has issued the result, it is faxed to us immediately. We get it scanned by our pathology staff and an addendum is issued. I think the main time-related issues that come up are when there is a gap between deciding to perform a test and then doing it, or when there is an insufficient amount of tissue. If Dr Freeman orders BRAF or PD‑L1 testing, it takes about 10 days for the actual testing to occur. In fact, the entire NGS [next-generation sequencing] panel takes about 10 days.”
Sufficiency of sample is also often a challenge. As Dr Freeman noted, “That is always a tough situation to face. I think lung cancer is a great example in which a tissue can be exhausted because of the sequential molecular testing that can take place. You have what you need for diagnosis. Then you do EGFR, ALK, ROS1, and PD‑L1 testing. When it comes down to getting additional testing, by the time you get to ROS1, you may be out of tissue. Regarding tissue stewardship, I think a lot of us are just going straight to NGS because we do not run out and we have all the information that we need. If the patient is agreeable to repeat biopsy, then we will obviously try to convince them of it, especially if we think that a therapeutic decision is pending based on the results from additional testing. Regarding melanoma, we are pretty fortunate. Our surgical oncologists are the ones who are most frequently contributing tissue. There is often a surplus of it. With other tumor types or fine‑needle aspirations, or maybe a single core, it can definitely be problematic when it comes to having enough tissue available for testing.”
Neither Dr Freeman nor Dr Parekh uses liquid biopsies for molecular biomarker testing in melanoma. As Dr Parekh indicated, “In most cases, we have the big lymph nodes that have the metastatic deposits. Sometimes, if there is a lung metastasis, we will have similar limitations as for lung cancer, but that is not very often. In most patients with melanoma, we have sufficient tissue to do all of the things that we want to do.”
But what about going beyond BRAF and PD-L1, to include TMB or microsatellite instability (MSI)? Dr Freeman suggested, “Obviously, BRAF is not the only important biomarker that we use in melanoma. We also look at NRAS and, in the case of mucosal and acral melanoma, c‑Kit. NRAS actually has prognostic implications and may also have therapeutic implications. Often, that information can be obtained by doing NGS as opposed to a single molecular test. Another biomarker of interest is the ARID1A mutation, particularly as it relates to patients who would benefit from immunotherapy. This mutation is not specific to melanoma and can only be picked up by doing NGS. I think that a lot of us are going to end up moving towards NGS anyway as our therapeutics are becoming almost tissue‑agnostic. The other thing that I will comment on regarding melanoma is TMB. At this year’s American Society of Clinical Oncology and American Association of Cancer Research meetings, there were some presentations on how TMB may be more influential in terms of response to immunotherapy compared with PD‑L1 expression. Tumors that have very high TMB, such as melanoma or cutaneous squamous cell carcinoma, have very dramatic responses to immunotherapy. I think that it ends up influencing our decision‑making. There is a lot of hope and promise assigned to immunotherapy. Ultimately, across trials, we know that only 30% to 40% of patients are going to have some type of response to immunotherapy. We have to appropriately match patients to treatment.”
Cost and Value in Biomarker Testing
Payment and reimbursement for biomarker testing is an important issue in the current environment of cost-effective medicine. For example, a FoundationOne (Foundation Medicine) NGS panel costs nearly $6000.6 Who pays for that if the cost of the test is not covered by the payer?
As Dr Freeman said, “We have clinical pathways that we have to adhere to when it comes to treatment selection, but the pushback from payers has definitely happened more with diagnostics than it has with molecular or biomarker testing. The Centers for Medicare & Medicaid Services’ 14‑day rule is a certain impediment to a lot of this, and this is what has resulted in some of those delays in initiation of therapy, because it impacts the ability for us to get the testing done in a timely manner if the specimen was obtained while the patient was hospitalized. This creates barriers to care.”
Dr Freeman went on to say, “The question of financial toxicity is definitely entering our conversations more and more. It tends to center more around drugs than it does around diagnostics. In some cases, the patient would call me and say, ‘I got this $5,000 bill for this test, and I do not remember you telling me that you were going to order it.’ It would come down oftentimes to just asking the company to waive the cost of testing, because what ultimately is happening is that the payers are saying, ‘This is not a supported method of testing, so we are not going to pay for it.’ They passed the bill on to the patients and, of course, no one can afford to get that done. I think as NGS is used more frequently, it should get cheaper. It is ultimately going to come down to engaging with payers and deciding on an appropriate reimbursement structure for this type of testing. A lot of cancer centers will develop their own biomarker testing platforms because of the benefit it offers their patients. They will have to find ways to absorb those costs until there is a method of getting them approved and reimbursed by payers.”
Dr Freeman also noted, “Obviously, part of the relationship with an outside testing laboratory is understanding their billing process, and asking them, ‘What’s your means of getting reimbursement for this test? What are some of the workarounds if it is not covered?’ These companies, such as Foundation Medical, obviously want the physician’s business, and they want that relationship to stay strong. Therefore, if it comes down to coding a test appropriately or indicating your reason for testing so that you can improve the likelihood of having it covered by the payer, they are going to do that. A company actually drafted a letter of medical necessity as a template that I could use and include with my requisition so that I could do my best to make sure that that testing was covered. These companies have standard operating procedures in place. They want to help oncologists help their patients. It is really a matter of working with the company and asking them, ‘How are you ensuring that this test gets covered, and if it is not covered, how do we ensure that the patient is not saddled with this large bill?’ Relying on the company that you work with is a key component of that.”
Meeting the Challenges of Incorporating Biomarker Testing into Clinical Practice
When the experts were queried about what they considered the greatest challenges in incorporating biomarker testing into their clinical practice and how they meet these challenges, Dr Parekh replied, “It is deciding whether there is enough evidence on an emerging biomarker to indicate that our patient is actually going to benefit from testing for it. The field is moving very fast and there are many biomarkers being discovered; this results in competing interests for a little bit of tissue. First of all, the pathologist needs to make sure that an accurate diagnosis is made. We have to use the tissue to come up with the right diagnosis and the useful prognostic markers. After that, we must decide on the best biomarker to test for with whatever tissue remains. That becomes the biggest challenge.”
Dr Freeman concurred. “I agree completely. I think that competing interests for tissue is definitely one of the biggest challenges that we have, followed by prioritization of the biomarkers. We go from holding up PD‑L1 as the be‑all and end‑all to a couple months later saying, ‘It is MSI,’ to a couple months after that saying, ‘No, it is TMB.’ For us, it is intellectually stimulating, but that is a frustrating proposition for people in the community who have many more patients to process. They are in a high‑volume, high‑pressure situation. At the end of the day, I think they want to say, ‘Look, just tell me what I should test for. What is it that we are all going to agree on is the right test to select?’ I think that uncertainty spills over to the patients as well. Patients want to know what is going to be the right test for them. As an oncology community, we do not quite have the answer for that yet. Our dialogue continues to evolve. I think that poses a lot of challenges, both positive and negative, for how everyone ultimately will interact and decide on what is going to be the right therapy for the patient sitting in front of you.”
When asked about best practices in meeting these challenges, Dr Freeman noted, “One of the best practices I would share is if you are going to select an NGS partner, find a company that you trust and work with them consistently, whether that is external testing or an internal NGS platform. I think that ultimately everyone is going to end up moving to just that. There are a number of tests that are continuing to evolve from a prognostic standpoint. Doing NGS gives you not only the molecular characteristics but also the MSI and the TMB. I think one of the best practices is to find a platform that you like, that you trust, that is interpretable, and use that consistently.
“The other best practice is maintaining an open dialogue with your pathologists. I could not do my job without them. I think that we learn from each other every day. They are a really important partner when it comes to doing the right thing for the patient.”
Dr Parekh agreed. “Selecting the right platform is very important. We are dealing with a small amount of tissue.
One platform that delivers all of the tests that you may need now and in the future would be the best approach. Also, open channels of communication among the pathologist, radiologist, medical oncologist, and surgeon are critical. We also have interest groups—a melanoma interest group, an immuno‑oncology interest group—where we can focus more on the scientific component that sometimes we do not have time to focus on during the tumor boards. If we have this interest group meeting, we can actually get into the mechanistic aspects of different options that are available. We can try to decide what has more evidence of working for our patients. That is something that any institute can think of to improve their testing. Just as important, working with clinicians to come up with what matters the most to them, coming up with a reflexive testing algorithm so that when the time comes, you can focus on the most important tests first and then go from there.”
Conclusions: Applying Best Practices in Biomarker Testing
Molecular biomarker testing in oncology has come of age. Numerous candidate biomarkers are currently the focus of research, most notably, TMB, tumor or immune cell expressions of PD-L1, and CD8+ T-cell infiltration into the tumor. In the future, bioinformatics technology may allow for the combination of several biomarkers to produce a predictive scoring system, depending on the type and stage of the cancer, the individual treatments, and the timing of the treatment intervention.5 Although there are challenges associated with incorporating biomarker testing into clinical practice (selecting the optimal testing platform and managing costs and reimbursement of the tests), biomarker-driven immunotherapy and targeted therapy have dramatically altered the therapeutic landscape of several cancers.
Dr Freeman summed up her perspectives: “I would discourage never considering biomarker testing or molecular testing for a patient. I will use a specific example of a young man that I saw about a year ago. He had stage IV melanoma and had been treated with immunotherapy because his initial biopsy showed BRAF wild‑type by immunohistochemical analysis. Unfortunately, he experienced disease progression while receiving immunotherapy, so he was sent to me for clinical trial consultation. Most young patients have a BRAF mutation—the prevalence in younger patients is probably about 70%—so I was surprised that he was BRAF wild‑type. I said, ‘Let me go ahead and send this off for NGS because I think that we can sometimes miss BRAF status by immunohistochemical analysis.’ Sure enough, the NGS showed that he had a BRAF mutation, and so rather than put him into a clinical trial, he received oral TKI therapy, and he had a complete response. We never would have seen that clinical result had we not thought to do the additional testing. I encourage clinicians to really think about biomarker testing for their patients, because it can make enormous impacts on treatment decisions, and sometimes patient outcomes.”
Dr Parekh concluded, “These are very exciting times for patients with advanced melanoma and other cancers. Historically, these patients only had a few months to live and there was no hope. Biomarker testing has totally transformed the field of biologics. Pathologists have a vital role to play going beyond the diagnostic and prognostic measures, to utilize newer predictive biomarkers that have transformed cancer therapy. Also, from a pathologist’s perspective, it is important to be able to do further research on potential biomarkers, and to continue working with researchers in helping advance the field of predictive biomarkers and biologics.”
Many of the best practices discussed in this newsletter are tumor-agnostic, so that they can be applied to biomarker testing (including NGS and other comprehensive genetic profiling) across multiple tumor types, including lung cancer, metastatic melanoma, breast cancer, and other solid tumors, as well as hematologic malignancies. Because molecular biomarker testing is increasingly being used to inform targeted therapy and immunotherapy in patients with cancer, these best practices are relevant for the entire cancer care team, including medical and surgical oncologists, pathologists, pulmonologists, interventional radiologists, oncology nurses, and other healthcare professionals involved in the care of these patients.
- Ascierto PA, Puzanov I, Agarwala SS, et al. Perspectives in melanoma: meeting report from the Melanoma Bridge (30 November–2 December, 2017, Naples, Italy). J Transl Med. 2018;16:207. doi:10.1186/s12967-018-1568-6.
- Hodi FS, O’Day SJ, McDermott DF, et al. Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010;363:711-723.
- Chapman PB, Hauschild A, Robert C, et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med. 2011;364:2507-2516.
- Buder-Bakhaya K, Hassel JC. Biomarkers for clinical benefit of immune checkpoint inhibitor treatment–a review from the melanoma perspective and beyond. Front Immunol. 2018;9:1474. doi:10.3389/fimmu.2018.01474.
- Kitano S, Nakayama T, Yamashita M. Biomarkers for immune checkpoint inhibitors in melanoma. Front Oncol. 2018;8:270. doi:10.3389/fonc.2018.00270.
- Goldstein DA, Shaib WL, Flowers CR. Costs and effectiveness of genomic testing in the management of colorectal cancer. Oncology (Williston Park). 2015;29:175-183.
Although many quality measures exist in oncology, few efforts have been undertaken to prioritize, measure, and report quality and costs for an entire region. A recent multiyear, multistakeholder effort to characterize quality of care and costs for Washington State oncology practices revealed that increased quality may be associated with a reduced cost of care in oncology.
Non–small-cell lung cancer (NSCLC)—including squamous-cell carcinoma, adenocarcinoma, and large-cell carcinoma—accounts for 80% to 85% of all lung cancers. Of patients diagnosed with lung cancer (all types), 17% have localized disease (confined to the primary site), 22% have regional disease (spread to regional lymph nodes), 57% have metastatic disease, and 4% have an unknown stage.