Misdiagnoses and Sequencing in Melanoma

Jeffrey S. Weber, MD, PhD

Director, Donald A. Adam Comprehensive Melanoma Research Center, Moffitt Cancer Center and Research Institute, Tampa, FL

After decades with few available therapies, a flurry of FDA drug approvals has provided oncologists with a variety of options for treating patients with advanced melanoma. Although it is an exciting time in the field, there are fresh challenges in managing toxicities associated with new therapies and finding the best ways of sequencing these agents to maximize efficacy. Additionally, some struggles continue to affect physicians’ ability to treat patients, such as a high rate of misdiagnosis for primary melanoma in the community.

These topics were among those discussed at the Update for Clinicians on Diagnosis and Treatment of Melanoma and Other Cutaneous Malignancies, a one-day conference that Physicians’ Education Resource (PER) hosted September 28 at the Moffitt Cancer Center in Tampa, Florida.

Jeffrey S. Weber, MD, PhD, senior member and director of the Donald A. Adam Comprehensive Melanoma Research Center at Moffitt, served as one of the course directors of the meeting. A longtime translational researcher, Weber has worked extensively on immunotherapies including ipilimumab and the emerging PD-1 and PD-L1 checkpoint antibodies, as well as targeted therapies, such as the recently approved dabrafenib and trametinib.

In an interview with OncologyLive, Weber discussed current and emerging issuesin the treatment of melanoma.

OncologyLive: One startling statistic regarding melanoma is that as many as 30% of cases are not diagnosed properly in the community. Where is that number coming from?

Weber: That is the estimate by our surgical colleagues who see the primary melanomas, and they estimate that probably 30% of them are not diagnosed correctly. That means you could have a Spitz nevus that’s really a melanoma or a melanoma that’s really a Spitz nevus. You could have an early-stage melanoma that’s really more of a stage I diagnosis instead of a stage II diag - nosis. You could have someone who has ulceration that wasn’t noted. It’s not missed diagnosis—it is misdiagnosis.

Often, the outside reads do not contain all of the information that our specialists [at an academic medical center] would require as the minimum information needed to accurately stage a primary melanoma.

So, if you add up missed diagnoses—which thank goodness we don’t have that often—misdiagnoses, and improperly stated diagnoses, our surgeons think it is probably about 30%, and that’s a big number.

Will the increased understanding of melanoma genetics improve the accuracy of initial diagnoses?

Those data and that technology are not part of the diagnosis of a primary melanoma. Even today, melanoma is diagnosed with a “mark one eyeball,” with a dermatopathologist—or at least we hope a dermatopathologist—looking at slides. The problems arise when you don’t have a trained dermatopathologist. You have a regular anatomic pathologist who is making the diagnosis, who is not experienced with the skin, which in small communities can happen.

Oftentimes, the diagnosis of primary melanoma could be a little difficult. Many dermatopathology clinics have two pairs of eyes that look at it and sometimes you may not have those two pairs available easily. The genetic tests are not used to make the diagnosis of primary disease. They are only used in disease that’s at stage III or more to determine eligibility for therapy.

It is a tough business diagnosing melanoma as a dermatopathologist. When I first looked at slides that I was shown many years ago, I had no clue how anybody could possibly diagnose the difference between a nevus and a melanoma. It was incredibly obscure. With metastatic disease, that’s easy. The cells look ugly. But for primary diagnosis, it is tough, and I absolutely believe that a 30% misdiagnosis rate is believable.

What actions are being taken to curb the rate of misdiagnosis?

I think there has been more of a tendency in the last 10 to 20 years to centralize the reading or rereading of melanoma slides in big referral centers so that pathologists who don’t do a lot of dermatopathology will tend to let them be read by the specialists, who can identify some of the things that people would not have been conscious of in the past, such as angiolymphatic invasion, the number of mitoses per high-power field, and ulceration. Centralization and specialization has improved things, but it still leaves a lot to be desired.

In June, both dabrafenib and trametinib received FDA approval as single agents. What is the appeal of these drugs, and how revolutionary are these approvals in terms of their effect on the treatment of advanced melanoma?

The BRAF inhibitor dabrafenib is a good drug. It is probably similar in its response profile to vemurafenib. It certainly has a different toxicity profile that may lead to more common use in the future because it has a lower likelihood of inducing some of the skin toxicity.

By the beginning of 2014, the combination of dabrafenib and trametinib might be registered. There has been a filing for the combination by GlaxoSmith- Kline and hopefully that combination will get approved. I think that’s extraordinarily important. I think it is going to moot the point of whether we use dabrafenib or vemurafenib. Everybody is going to want to use the combination because not only does it have less toxicity than either drug alone, it seems to me that the data will show that it is significantly more effective than either drug alone.

Does that mean the combination has to be given upfront to have any effect?

Absolutely, I agree. In my view, the time to give the combination is right upfront. I would not deliver a single agent, wait for resistance, and then hope to overcome it. If you want to deal with BRAF inhibitor resistance, the time to deal with it is right upfront. That’s why I think starting in 2014, BRAF-mutated patients are all going to be getting treated with the combination upfront.

When do you test patients for the BRAF mutation?

There are two adjuvant stage III trials going on now. One is from Genentech¹ and one is from GlaxoSmithKline.² We absolutely test every stage III patient who has a positive lymph node who would be eligible for those trials. (At Moffitt, we are only participating in the GlaxoSmith- Kline trial).

FDA-Approved Therapies for Melanoma

^aPrior FDA approvals in other indications.

^bThe FDA has said ipilimumab is the first therapy to “clearly demonstrate” a survival benefit for patients with metastatic melanoma.

I would say at the academic centers, it is becoming a much more common practice to test patients. In the community, it usually varies. Sometimes, if physicians know they can refer people for a stage III trial with no evidence of disease, they will test them at the point of stage III nodal disease. Otherwise, they generally will test patients at the point where they develop metastatic disease. I have noticed that many doctors in the community have been using single-agent vemurafenib or dabrafenib off-protocol upfront for patients who are BRAF-mutated.

However, this raises the issue of how to properly sequence all of these new therapies.

Sequencing is a huge issue. There are some slightly scary data out there suggesting that patients who fail BRAF/MEK will not respond to ipilimumab, but if you fail ipilimumab, your BRAF/MEK response rate is the same as if you have been previously untreated. It is worrisome. It would suggest that those who have indolent disease should always go the immunotherapy route first, but that’s unproven. There are no data.

Michael B. Atkins, MD, [of Georgetown University] is planning an Intergroup trial that would attempt to fix the situation of the “data-free zone” as we call it, and come up with some real data that would drive you to choose one pathway or the other. Frankly, a lot of doctors in the community, even if they have patients who are BRAF-mutated with indolent minimal disease, are giving them a single-agent BRAF inhibitor first. They generally respond, but the question is: Is that going to compromise their chance to respond to a subsequent immunotherapy? I don’t know, and it is a very scary problem.

How would toxicities affect decisions on sequencing therapies, even if patients respond with the associated toxicities?

The toxicity of ipilimumab is not trivial. At the approved dosage of 3 mg/kg, I tell patients that about 10% of them are going to have some significant toxicity requiring treatment, although hospitalization is pretty rare. To go to the 10 mg/kg dosage, you probably almost double the toxicity rate, so it is a potential problem.

However, the toxicity management is on a learning curve, and I would say a lot of doctors in the community are now more comfortable treating patients with ipilimumab. It is just like learning how to use any chemotherapeutic drug, doing a stem cell transplant, or administering interleukin 2. Each drug has its own toxicity profile, and ipilimumab is different from what we are accustomed to, but it is not impossible to deal with. It is, in my view, fairly well tolerated.

What are your thoughts on PD-1 and PD-L1 immunotherapies in melanoma?

I think that the PD-1 data look very good. PD1 plus ipilimumab looks even better, either simultaneous or sequential. I think that it is extremely likely that one of the current registration trials will be positive. The duration of response is probably two years, as Mario Sznol showed at ASCO.³

I think that PD-1 with ipilimumab in some way, shape, or form will probably become the routine first line of treatment for the BRAF wild-type in anyone who can tolerate it. Even in the BRAFmutated population, I think those with indolent disease are going to get the combination, either simultaneously or sequential. I think there will even be the chance to retreat those patients if they get a response and then fail.

What do we know about PD-1 and PD-L1 as targets that have resulted in these very positive responses in early trials?

PD-1 is a molecule that is a brake on the immune system, but it tends to be more expressed on the cells that are the important antitumor factors—CD8. It tends to be highly expressed on antigen-experienced cells that are “exhausted.” It is a very important influence, a driver of immune exhaustion in the setting of chronic antigen exposure.

It is important in the oncologic field because if you can get rid of it or abrogate its activity, you are going to free up immune cells. You sort of de-exhaust them or invigorate them, so to speak. The word “exhaustion” is an immunologic term that describes cells that are not dying, they are not dead, they are not apoptotic, but they are like wet noodles. They are unable to mount an effective cytokine response or a lytic response to their antigenic targets. It literally exhausts the cells.

What other emerging targets or therapies are being evaluated in melanoma?

There are plenty of other checkpoint antibodies. LAG-3 looks very promising, There also are TIM-3, BTLA, and VISTA. I would say the field looks very promising up the road.

Are there any other issues that community oncologists should be aware of with regard to melanoma?

Most patients who have metastatic disease are still going to die of their disease. There is still a major unmet need here. We have some good drugs, but most people who get the BRAF/MEK combination will eventually relapse. Most people who are treated with PD-1 don’t respond—it is only about a 30%-35% response rate. If you do the combination, you may get up to a 40% response rate, but the implication is that most patients will eventually die of their disease.

Be assured that we haven’t cured melanoma yet. We are far from it.

References

1. BRIM8: a study of vemurafenib adjuvant therapy in patients with resected cutaneous BRAF mutant melanoma. NIH Clinical Trials Registry, www.ClinicalTrials.gov website. ID NCT01667419. Updated October 8, 2013. Accessed October 13, 2013.
2. A study of the BRAF inhibitor dabrafenib in combination with the MEK inhibitor trametinib in the adjuvant treatment of highrisk BRAF V600 mutation-positive melanoma after surgical resection (COMBI-AD). NIH Clinical Trials Registry, www.ClinicalTrials. gov website. ID NCT01682083. Updated October 4, 2013. Accessed October 13, 2013.
3. Sznol M, Kluger HM, Hodi FS. Survival and long-term follow-up of safety and response in patients (pts) with advanced melanoma (MEL) in a phase I trial of nivolumab (anti-PD-1; BMS-936558; ONO-4538). J Clin Oncol. 2013;31 (suppl; abstr CRA9006).