Q & A with Dr. Paul Nghiem, MD, PhD


En Español

Paul NghiemMark Teich, executive editor of The Skin Cancer Foundation, spoke with Paul Nghiem, MD, PhD, about recent landmark progress in the detection, monitoring and treatment of patients with Merkel cell carcinoma (MCC), a rare but dangerous skin cancer. Dr. Nghiem (pronounced NEE-em) is the George F. Odland Endowed Chair and head of Dermatology at the University of Washington in Seattle. He is one of the foremost experts on MCC in the U.S.  

Q: Melanoma kills more people than any other form of skin cancer. But while Merkel cell carcinoma (MCC) is much rarer than melanoma, it is more dangerous case by case. What makes it more dangerous?      

A: Early MCCs are typically far less conspicuous than early melanomas. They don’t look concerning, so they often end up being detected at a more advanced stage, when they have metastasized (spread) to other parts of the body and are far harder to cure. In contrast, many melanomas are actually found at the in situ stage (stage 0), when they are contained in the very top layer of the skin (epidermis) and have not invaded into the deeper layers. These in situ melanomas are easy to treat and have virtually no chance of metastasizing. But out of more than 1,000 cases of MCC we are following, we have had only two in situ lesions.  It is very rare to detect them at such an early time.  

When you first find a melanoma or MCC tumor, the initial concern is whether tumor cells may have metastasized from the lesion (the primary tumor) to the nearest lymph nodes. The standard way of checking this today is the sentinel node biopsy (SLNB).  In this test, the physician verifies that no cancer cells have spread to the lymph nodes by removing and microscopically examining the first few nearby nodes that receive lymph fluid drainage from the primary tumor. With melanoma, in most cases, the tumor is small and at an early enough stage that you don’t need to do an SLNB since it is unlikely to have metastasized. However, even the very smallest MCC, as little as four or five millimeters in diameter, has a 15 percent chance of having advanced to the lymph nodes or beyond, so we consider doing an SLNB even if an MCC is very small. 

Q: What can people do to discover more of these lesions when they’re readily curable? How can they recognize potential MCCs earlier?

A: The classic ABCDE signs of melanoma don’t work for Merkel cell carcinoma. Instead, we have the acronym, AEIOU, which can help with early recognition: Asymptomatic/lack of tenderness; Expanding rapidly; Immune suppression; Older than 50 years; and Ultraviolet-exposed/fair skin. About 63 percent of MCC patients have rapidly expanding lesions, and close to 90 percent have three or more of the AEIOU features. It is important to note, however, that many lesions with several of the AEIOU features do not prove to be MCC—some are benign, and others may be a different skin cancer.

merkel cell carcinoma on arm
Merkel cell carcinoma on arm.
merkel cell carcinoma on eyelid
Merkel cell carcinoma on eyelid.

Q: Once an MCC has metastasized to the lymph nodes or beyond, what are the patient’s chances of survival?

A: That’s hard to say right now, because thanks to new treatments, survival is changing quickly for the better. I used to feel hopeless if somebody developed metastatic MCC, but now it is clearly not hopeless. There have been major changes in the field, with a real possibility of cure or near-cure for advanced MCC. I now feel that at least half of the patients with metastatic MCC will be effectively treated.

Q: What treatments have made the difference?

A: In the past, we treated advanced Merkel cell carcinoma with chemotherapy. This approach would shrink tumors in over half of cases, but the benefit would last only three months on average. When the cancer returned, it was resistant to many drugs, and the immune system had been suppressed by the chemotherapy, compounding problems for what is a very immune-sensitive cancer. However, checkpoint inhibitors are an exciting new class of immunotherapies (treatments that boost the ability of the immune system to fight a disease), and are rapidly becoming the standard of care for advanced MCC. By blocking protein receptors that normally keep the immune system in check, these drugs stimulate T cells to fight the tumor.

In 2014, a checkpoint inhibitor drug called pembrolizumab (Keytruda®), which blocks an immune-inhibiting protein receptor called PD-1 (programmed death-1), was originally approved to fight advanced melanoma, sending many patients’ cancers into long-term remission. It is now successfully being used off-label (for another purpose than for what it was approved) with certain other cancers, and has been listed in the National Comprehensive Cancer Network Guidelines as indicated for MCC, which means that insurance companies typically will cover it for MCC. It is currently in clinical trials on MCC, and is being evaluated for FDA approval in MCC. 

In addition to pembrolizumab, a new checkpoint inhibitor immunotherapy moved through the approval process much faster in the past couple of years, receiving fast track, “orphan drug” status (a pharmaceutical that remains commercially undeveloped owing to limited potential for profitability) and “breakthrough therapy” designations to speed the way. In March 2017, the US Food and Drug Administration (FDA) approved the drug, avelumab (Bavencio), for the treatment of adults and pediatric patients 12 years and older with metastatic MCC. FDA approval, the first ever for advanced MCC, was based on data from a clinical trial in which 33 percent of the patients experienced complete or partial shrinkage of their tumors.

Injected intravenously, avelumab stimulates T cells by blocking a molecule called PD-L1 (programmed death-ligand 1) that links up with PD-1 to inhibit these critical immune cells.  

“I used to feel hopeless for patients with metastatic MCC. Now, it is clearly not hopeless.”

Q: Is there any preferred order in which treatments should be given now? Do you have to administer chemotherapy or radiation first, before using one of the checkpoint blockade therapies? 

A: Until recently, chemotherapy was the frontline therapy. But if you treat advanced MCC with a checkpoint inhibitor before doing chemotherapy, about 60 percent of patients respond. If you wait until the patient has failed one round of chemotherapy, it appears that about 40 percent of patients respond. And after two or more rounds of chemotherapy, about 20 percent of patients respond. So, usually the right approach now seems to be to use the checkpoint blockade agent first, then decide what to do next based on the response. Some patients will have a complete response, with total remission of the disease. Some patients’ tumors will partially respond, often decreasing by 70 percent in tumor volume, which is still a major response. Many of these patients end up having a low level of disease that does not affect their health at all. One woman I treated recently had a partial response for a year, then it changed to a complete response. But usually, if the drug is going to work, the response comes fast, and most responses last well over a year. However, we are uncertain what will happen when the drug is stopped.

Since avelumab is now FDA-approved for MCC that has or has not been previously treated with chemotherapy, it is now being used preferentially as a first-line therapy for metastatic MCC, while chemotherapy is not favored in most cases. Current data suggest that immune therapy works better if given prior to chemotherapy, which tends to suppress the immune system and make tumors more resistant to a variety of treatments. 

Although chemotherapy is no longer the usual frontline treatment, radiation may in some cases be used to good advantage early in the game. It synergizes nicely with immunotherapy. The immune system does better at eliminating small tumors than large tumors, and radiation can effectively shrink tumors. So, you may want to irradiate the largest lesions first to shrink them and decrease the cancer burden before using checkpoint blockade therapy. The immune system may then do better at eliminating the smaller tumor.

Q: What about the other FDA-approved checkpoint blockade therapies for melanoma, nivolumab (Opdivo®) and combination ipilimumab (Yervoy®)-nivolumab? Are they being tried for MCC as well?

A: There are no published data on these agents in MCC, though a new trial of the combination of ipilimumab and nivolumab is now gathering patients. While formal data are lacking, we have seen some patients who have had wonderful results with these agents as well.

Q: What happens next if none of the initial checkpoint therapies work?

A: This is a huge area of interest and research at this time. Combining two different types of immune-stimulating drugs is a major area to be explored. It is also possible that local injections of immune-stimulating therapies to eliminate a specific target tumor might be combined with a systemic agent such as an immunotherapy. There is also interest in combining chemotherapy or radiation with checkpoint blockade therapies, but these have not been studied yet and we don’t know how well they would work.

Q: You said you expect about half the patients to do well on the checkpoint blockade therapies. That suggests that about half won’t. Since these are expensive treatments with some potentially serious side effects, is there any way of knowing in advance which patients will benefit, so that you can start the predicted non-responders with a more beneficial treatment?

A: Although more than half of patients respond, for about 40 percent of patients, the cancer continues progressing as if you were giving no treatment at all. Unfortunately, we can’t currently tell in advance who will respond and who will not. We and many others are working very hard to solve that problem, but to date, none of the dozen tests we’ve undertaken are much more accurate than flipping a coin.

Q: How about after a course of treatment has been administered? Is there a way of knowing early on whether the cancer is starting to come back? 

A: We have made exciting strides in this area. Our team developed a blood test, which looks for certain Merkel cell polyomavirus antibodies in the blood, so that we can detect any recurrences early, when the tumor will be more effectively treatable (see www.merkelcell.org/sero).

In 2008, Drs. Patrick Moore and Yuan Chang found that most Merkel cell carcinomas are in part caused by a virus. (Sunlight also often plays a role.) We find this Merkel cell polyomavirus in about 80 percent of MCC tumors. When this virus infects cells, it produces oncoproteins (cancer-promoting proteins) that may cause the cells to grow out of control. We do the blood test at different intervals, starting at the time of diagnosis, and if the number of antibodies to these polyomavirus proteins ever increases, it strongly indicates that the patient is having a recurrence.

We are now using this test routinely, because it is very helpful in managing patients over time. About half of patients do not produce any of these antibodies, and they are actually at about 40 percent higher risk of having their MCC recur than those that make the antibodies, so we know we need to follow them closely with imaging scans. For patients who make the antibodies it’s relatively simple: If the cancer does not come back, the antibodies fall quickly, and in most cases they are undetectable by one year after treatment. If MCC comes back, the antibodies go up. It gives us a great chance of discovering recurrences early, when they are extremely treatable: Treating a grape instead of a grapefruit (in terms of cancer size) makes a big difference. And with a patient whose antibodies keep falling, the blood test gives us great confidence that the cancer is not recurring, so we can skip or reduce the use of costly imaging scans.

About two dozen institutions around the country are using the test now, and a group in Europe is developing a similar test. We published the first description of our test in 2010.  Using support from an NIH grant, we have now followed more than 200 newly diagnosed MCC patients over time to see how useful this test is. We and our patients have in fact found it to be very helpful, and the large second validation study appeared in the journal Cancer in December 2016.

Q: After you have treated MCC patients, how do you decide how often to see them and how much risk they still have of the cancer coming back?

A: For several years, we have had good data on how long patients survive after being diagnosed with various stages of MCC, but we only recently have obtained data on whether patients have had a recurrence. To my surprise, this has proven incredibly useful. For example, when patients are first diagnosed, you can tell them that a stage 3 patient may have a 70 percent chance of recurrence, whereas an early stage 1 patient may have only a 20 percent chance of ever recurring. When patients are seeing you for their one-year follow-up visit, it is now possible to tell them how much ‘residual’ risk of disease they have, by using the data on “recurrence-free survival” (how long patients go without a recurrence). If a lesion has not recurred after three years, a huge amount of the risk is gone.

Furthermore, it’s important to know that even when lesions recur, the stage at which the original tumor was diagnosed plays an important role in how bad the recurrence will probably be. While stage 1 patients do have recurrences, they are usually still near the primary or possibly in the nodes and can be treated successfully. In contrast, when stage 3 patients have recurrences, they are often distant metastases and much more dangerous. 

All of this data really wasn’t available until the past year, and it is proving very helpful in knowing just how closely to keep following patients. We probably used to follow them too aggressively, and for too many years.

In general, we now believe that we don’t have to monitor patients so closely for as many years as we used to do. MCC doesn’t seem to be like melanoma, which often recurs 5 or 10 years after diagnosis. With MCC, if it comes back, it is usually within two to three years. So, we now think that intensive follow-up for two or three years is usually appropriate, with a markedly less aggressive schedule up to about year five. While these recurrence-free data are not yet published, they are available at https://merkelcell.org/prognosis/disease-recurrence.

Q: Are any other new forms of treatment making a difference?

A: The checkpoint blockade immunotherapies have been the game-changer, and no other therapies being tested have made that kind of difference yet. However, in combination with checkpoint therapy, we are now using patients’ own virus-specific T cells to fight the cancer (a technique called adoptive or autologous T-cell transfer therapy). In polyomavirus-driven MCC, you can find these T cells fast, remove them and grow them with growth factors that make them live longer and kill better, then reinject them into the patient. This can give a boost to the immune system, and combining this therapy with checkpoint blockade therapy may be more effective than either approach alone. In a recent small study at the Fred Hutchison University in Seattle, physicians combined avelumab with the T-cell transfer therapy and radiation or another immunotherapy called interferon. Three out of four metastatic MCC patients treated with this experimental combination are in complete remission following the treatment, with no sign of the cancer remaining. 

We are also looking into combining some intralesional therapies with the checkpoint blockade therapies. These are immunotherapies such as interferons, interleukin-12 or toll-like receptor agonists that are injected directly into patients’ tumors. In small studies in MCC, they have often shrunk or eliminated the injected tumors and also reduced some other tumors throughout the body. Studies are ongoing, but it seems likely that some such ‘local’ immune stimulators given along with immune checkpoint agents will help produce higher response rates than the checkpoint agents given by themselves.

Published on November 7, 2016.  Updated June 12, 2017.