Lipson Sheds Light on the Evolving Role of ctDNA in the Realm of Melanoma

Evan J. Lipson, MD

The use of circulating tumor DNA (ctDNA) continues to evolve in the realm of melanoma, with opportunity for use as a potential blood-based marker in which positivity may be indicative of disease recurrence, according to Evan J. Lipson, MD, who added that it may also serve as a tool to clarify pseudoprogression or for long-term monitoring.

One study is focused on examining immunotherapy in patients with stage II melanoma who have ctDNA positivity after disease resection, according to Lipson. In many cases, patients can achieve cure through the surgical procedure alone. With this trial, investigators are seeking to utilize ctDNA positivity as a marker to determine who should receive adjuvant immunotherapy.

“The trial is looking at patients with stage II melanoma who have undergone resection of their disease and who are ctDNA positive after surgery. That seems like a good candidate population [who] might benefit from adjuvant therapy,” Lipson explained. “It also potentially spares the patients whose disease may have been surgically removed completely, and therefore, do not need any adjuvant therapy. Those are the research trials that I am most excited about. We will see if the sensitivity of the assays allows for delineation between those patients who need adjuvant therapy and those who do not.”

In an interview with OncLive^®, Lipson, associate professor of oncology, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins Medicine, discussed the role of ctDNA in the neoadjuvant, adjuvant, and metastatic settings in the realm of melanoma, remaining questions with this marker, and future directions to optimize its use in this disease.

OncLive^®: How does ctDNA fit into the melanoma treatment paradigm?

Lipson: [Currently], ctDNA has an emerging role in the melanoma space, [but] there is a lot of room for development. We use it in a couple of scenarios. One common scenario is when we have a patient who presents with high disease burden and we need to know within a few days whether that patient has a BRAF mutation as a tumor driver, and whether BRAF/MEK inhibitor therapy might be appropriate. We can typically get an answer about whether BRAF is driving that tumor's growth with a ctDNA assay more quickly than we can by assessment of the tumor tissue itself. We do that not infrequently in our practice.

The other place that ctDNA is emerging, and that we sometimes use [it for] here, in our clinic, is in patients who appear to have had a complete response to medical therapy in the locally advanced unresectable or metastatic setting. We oftentimes will check ctDNA levels on a somewhat serial basis when there is any doubt about findings on an imaging study.

We are using ctDNA measurements in that scenario as a blood-based tumor marker to inform our interpretation of imaging studies. [If we see] it on imaging and we have ctDNA positivity, then we are leaning more toward assuming that what we are seeing on imaging does, in fact, represent recurrence of disease.

How does the use of ctDNA vary between the neoadjuvant, adjuvant, and metastatic settings?

First, [I’ll discuss] the adjuvant setting and how ctDNA could be used in [patients with] melanoma. The first question that is most pressing is, who needs therapy? In other words, after a surgical resection of melanoma, could we identify a portion of patients who are not surgically cured and, therefore, need adjuvant therapy? In fact, some ongoing prospective studies are trying to address that question.

Similarly, in patients who have undergone resection of high-risk disease, [such as] stage IIB through IIIC or IIID disease, could ctDNA allow for more frequent, less invasive disease monitoring among that group [regardless of] whether they have received adjuvant therapy? The same would be true for surveillance in patients who do not ordinarily undergo surveillance imaging. For example, patients with resected stage IIA melanoma do not typically get screening or surveillance CT scans, so therefore, could we use ctDNA as a marker of disease progression for those patients?

In the adjuvant setting, sometimes we will come up on a situation where someone who is being surveilled or watched for disease progression, winds up with a small lung lesion, or a small liver lesion, or something that is too small to biopsy or too risky to biopsy. Despite seeing something on imaging, we are left with this question about whether it presents true melanoma growth. [In that case], a positive ctDNA result would support the idea that what we are seeing on imaging does represent disease recurrence, [or] evidence of disease activity.

The neoadjuvant setting is a little bit more challenging, [but] we potentially could use ctDNA as an early marker of the efficacy of whatever neoadjuvant therapy is being administered. In some cases, where we were initially on the fence about the surgical procedure, in a setting where ctDNA drops to undetectable quickly, 1 interesting question might be whether surgery is, in fact, necessary.

We came upon this scenario recently, not in melanoma but in another cutaneous malignancy, where a patient who was slated for neoadjuvant therapy followed by resection, ended up deciding not to undergo surgery when his response to immunotherapy was profound. With tumor regression, he considered his options and decided not to proceed with what would have been a quite large surgical procedure. Perhaps a ctDNA marker could be used as support [in favor of] or against that strategy.

Lastly, in the metastatic setting, several possibilities exist. First, ctDNA could provide early evidence of disease activity that can predict eventual disease progression. We have shown that in a study that we published a few years ago; we demonstrated that ctDNA can predict when the disease will recur. Sure enough, if you look forward on imaging in the following months, it will eventually show up.

ctDNA data can form interpretation of imaging and real-time imaging results where the result is equivocal. [In a situation in which a patient might have] disease progression, a positive ctDNA result clinches that diagnosis. Further into the future, a ctDNA result could facilitate clinical trial enrollment for patients whose tumors don't meet target lesion criteria per RECIST. Occasionally, patients will have, for example, bone metastases only; those lesions are not measurable by RECIST criteria, and they oftentimes do not qualify for studies.

It is the same thing with patients who, for example, have a lot of 8-mm lung metastases; they only have non-target disease, and oftentimes, they do not qualify for studies. Perhaps it could be that a sufficient level of ctDNA might then allow a patient to qualify for a study.

In the same vein, ctDNA could be an alternative measure of drug efficacy. For example, we look for a 30% decrease in the sum of longest diameters per RECIST as a marker of a partial response. It could be that in the future a ctDNA level that drops by half could serve as the marker of efficacy of a drug. You would not need to use measurements by imaging, but a molecular marker could be [used] in its place.

What are some of the clinical scenarios in which ctDNA would serve as a tool, such as pseudoprogression or long-term immunotherapy?

In several scenarios, ctDNA, at least in development as a tool, can clarify pseudoprogression. In some cases, as you mentioned, a few publications mostly from the group in Australia led by Georgina Long AO, BSc, PhD, MBBS, FRACP, FAHMS, of Melanoma Institute Australia, have demonstrated that you can use ctDNA to clarify the patients who are experiencing a desired response to therapy when their imaging is perhaps equivocal.

There is the scenario I mentioned before, in [which] a patient has a complete response, and you can use ctDNA as a long-term monitoring tool. In that scenario, in our experience, and in the study that we published a couple of years ago, when ctDNA is positive, the specificity is excellent. We were never fooled by a positive ctDNA. If it is there, then it is, in fact, real and at some point, the disease will progress—at least in the patients we included in our study.

The same cannot be said of the sensitivity of most of these assays in development. With a small amount of tumor burden, you do not always see evidence in the blood, and that is being worked on. It would be ideal for us to be able to detect ctDNA even when the disease appears to have been surgically removed for the purposes of helping to figure out who might benefit from adjuvant therapy, for example.

Are any ongoing efforts being made that you feel may impact views on how ctDNA can be utilized going forward?

The most exciting studies are the ones that are doing this prospectively. There is a prospective study that is being ramped up and is looking at immunotherapy in the treatment of patients with stage II melanoma who are ctDNA positive following resection of disease. If the sensitivity of the assay is good enough, that is a great population [in whom] to study this.

In many cases, patients with resected, stage II melanoma, are cured of their disease simply [through] surgery alone, so administering adjuvant therapy to many of those patients is only providing them with toxicity and no benefit. In patients who are ctDNA positive, at least there is some evidence that perhaps there's melanoma still hanging around. Those are the patients on this trial who will receive the adjuvant immunotherapy.

What challenges still exist with ctDNA, and what efforts are bein[g made to overcome them?

The greatest current unmet need is developing the sensitivity of the approach. Beyond that, is developing a raison d'être for the technology. Once the sensitivity reaches a critical point where we can be relatively certain that a negative is a negative, and a positive is a positive, then understanding how we are [going to optimally use] it in clinical practice becomes the challenge. We have [many] potential scenarios in which ctDNA would be helpful, and prospective trials, some of which are under development, are [seeking] to address those questions.

What does the future look like for ctDNA in this disease?

The future looks quite bright in terms of the way ctDNA will be applied in the melanoma space. Much of this revolves around meeting some of the challenges that I have mentioned. The primary one is sensitivity of the assays. If [we] could demonstrate that the presence of melanoma was being picked up with reliability, then we could say that we do not have a positive result, and therefore, this patient does not need adjuvant therapy after resection of disease. Or, we could say that this small lung nodule that we see in a patient with high-risk disease does not represent melanoma; it represents something else. Therefore, this patient does not have stage IV disease.

ctDNA is going to be applied more and more as the sensitivity of the assays becomes greater and greater. One important way [in which] that is being addressed is by the way that we detect ctDNA. In the past 5 or 6 years, the strategy has moved [beyond] trying to detect single hotspot mutations. For example, the detection of a BRAF mutation or the detection of an NRAS mutation, [has shifted] to more of a bespoke approach where we analyze multiple mutations in a specific patient's tumor—whether that is 10, or 20, or 50 mutations—and we look for the conglomeration of those mutations in the blood.

That increases the sensitivity of the assay, and as we move further along that road, we are going to find that we are better and better at detecting evidence of disease activity in a specific patient's blood. [That] increases our ability to rely on these data and make [better] clinical decisions.