Checkpoint inhibitors flourish

Dr. TsaoKey publications from the past year have enhanced our understanding of immunomodulatory drugs for advanced and metastatic skin cancers. Others could portend significant practice changes in areas such as sentinel lymph node (SLN) biopsy, says Hensin Tsao, M.D., Ph.D.

Investigating ipilimumab

Well-balanced in terms of patient numbers and disease substages, the pivotal trial of ipilimumab (Yervoy, Bristol-Myers Squibb) as adjuvant treatment for stage 3 melanoma showed a 24% reduction in recurrence risk at seven years,1 says Dr.  Tsao, who reviewed data during the recent MauiDerm 2017 meeting. He is professor of dermatology at Harvard Medical School and director of the Melanoma and Pigmented Lesion Center/Department of Dermatology at Massachusetts General Hospital.

“Similarly, there’s a 28% reduction in risk of death at five years.” Regardless of treatment, he adds, “Overall survival at five years, even in some stage 3 melanoma settings, can be 50% to 70%.”

Like interferon, Dr. Tsao says, ipilimumab can sometimes be extremely difficult to take due to side effects.

“The most important thing oncologists worry about is colitis,” which impacted 15.5% of study patients versus 1.5% for placebo in the trial. Colitis moreover caused five deaths in ipilimumab-treated patients. “That is significant  -  if one had stage 4 melanoma, and there was some risk of death from the drug, but then the natural outcome is bleak, one might take the risk and work through the side effects.”

But because ipilimumab treatment alone carries a risk of death, he says, a patient with stage 3 melanoma and no evidence of metastases anywhere else might think twice about using it.

Acquired resistance to immune checkpoint inhibitors including ipilimumab and other drugs that target programmed death 1 (PD1) and the ligand PD-L1 has been associated with JAK mutations.2

“Usually we think about acquired resistance with molecular therapies,” Dr. Tsao says. In histology from a patient whose tumor recurred around 300 days after treatment, “Most of the immune markers were still there. But for some reason - perhaps immunoreactivity or a mutation - recurrence occurred.”

In two of four patients who relapsed, deep sequencing uncovered JAK1 mutations.

“JAK mediates signaling from all the interferons. JAK1 mediates all interferons, while JAK2 specifically mediates interferon γ,” Dr. Tsao explains

The investigators from UCLA also showed that the patient-derived cells with the JAK mutations were no longer responsive to the suppressive effects of interferon, he says.

Overall, “We’re beginning to get a sense for what’s going on with the checkpoint inhibitors. This is critical because the ideal patient, dosing schedule and combination of checkpoint/molecular therapies are still not clarified.”

U.S. Food and Drug Administration approval for checkpoint inhibition has moved beyond stage 4 melanoma to approval for stage 3, Dr. Tsao says.

“It’s also moving beyond melanoma - many cancers now are eligible for checkpoint inhibition. The long-term cure rate for ipilimumab is now 22%.

We don’t know for the anti-PD1 drugs yet. And the reasons why some tumors are innately insensitive to anti-PD1/PD-L1 treatments are also not known,” he says.

The most common cutaneous side effect of anti-PD1 treatment is pruritus, he says. In a pooled analysis of 576 patients treated with nivolumab (Opdivo, Bristol-Myers Squibb), 17.2% experienced pruritus.3 As for timing, “The skin complications come up very quickly, some time between five to 10 weeks (median). Gastrointestinal symptoms start appearing a little before 10 weeks also, but the incidence is lower overall.” Generally, he adds, the percentages of patients afflicted with these side effects persisted in the study up to 40 weeks.

However, he says, adverse reactions could be a favorable sign.

“It appears that patients with any adverse reaction have a higher objective response rate (ORR), because these reactions are autoimmune, suggesting; therefore, that the patient can mount an effective immune response.”

Nivolumab-treated patients with no reactions had an ORR of 17.8%, versus 48.6% for patients who experienced any adverse reaction. And the more adverse reactions a patient had, the higher the response rate (although grade 3 or 4 reactions did not confer higher response rates).

Dr. Tsao adds, “The response rates do not appear to worsen with systemic immune modulating agents like steroids. That’s one of the greatest challenges when giving some kind of steroid to these patients. Whether it’s for pruritus or rash, you must work with the oncologist to see if he or she is comfortable with the use of any steroids,” because any anti-inflammatory treatment is theoretically immunosuppressive.

Next: Evidence in MCC

Evidence in MCC

For patients with metastatic Merkel cell carcinoma (MCC), he says, “The outcome is probably even worse than melanoma overall, with all the melanoma treatments available at this point.”

In a small phase 2 trial, pembrolizumab (Keytruda, Merck) proved effective for patients with and without the MCC polyomavirus.4

“The partial response rate is 40%; complete response rate is 16%. And about 36% of patients progressed - estimated progression-free survival at six months was 67%,” Dr. Tsao says. Unlike in melanoma, he adds, PD-L1 expression did not correlate with outcome, though virus-positive tumors had more PD-L1.

In a phase 2 study in stage IV metastatic MCC, the anti-PD-L1 agent avelumab proved effective in previously treated and untreated MCCs.5 Partial and complete response rates were 23% and 9%, respectively, and again, 36% of patients progressed. Median progression-free survival was 2.7 months.

“Checkpoint inhibitors are clearly effective in a subset of metastatic Merkel cell cancers,” Dr. Tsao says. “We wish we knew which ones to give. All of them respond to treatment - but not all of them, as far as we can predict, will eventually show the immune benefits to the extent that we want. This is probably the best option we have right now for metastatic Merkel cell carcinoma.” Avelumab (anti-PD-L1, Bavencio, Pfizer) was approved by the FDA for metastatic MCC in March.

With anti-PD1/PD-L1 inhibitors generally, he adds, “The real question is, to whom should we give the drugs? Should we avoid the side effects, or give them to people in whom we can theoretically predict some element of response?”

The genetic basis of vitiligo also been better characterized, he notes. In the largest genome-wide association study of vitiligo to date, which included 5,000 European cases and 40,000 controls, investigators identified 23 new loci associated with vitiligo.6 The most critical ones are autoimmune and pigment-related, he says. These findings corroborate smaller vitiligo studies, he adds.

Individual vitiligo-related single nucleotide polymorphisms (SNPs) identified by researchers are associated with many immune diseases, such as celiac disease, systemic lupus erythematosus and inflammatory bowel disease, Dr. Tsao says. “So vitiligo is certainly a disease of autoimmunity.”

Other genes involved in vitiligo, such as those governing tyrosinase and the melanocortin 1 receptor gene (MC1R), are also implicated in melanoma, he says. “What’s interesting is the demonstration, at a genetic level, that vitiligo risk and melanoma risk critically interact in some biological way. One would like to surmise that the risk for vitiligo would protect against melanoma, but more functional validation is required for that supposition to be formally proven.”

The involvement of melanocyte signaling genes makes it clear that vitiligo is also a pigmentary disease, Dr. Tsao says.

“The way we see it now, vitiligo is truly a disease of immunity and pigment cells, which at some level was always the assumption among dermatologists.”

Open question

A critical skin cancer question that remains open is whether completion lymph node dissection after a positive SLN biopsy improves survival, Dr. Tsao says. In a German study that included 1,269 patients with positive SLN, researchers randomized patients to observation (241 patients) or dissection (242 patients).

Completion lymph node dissection carries heavy morbidity, including edema, seromas and infection, Dr. Tsao says. “And if we look at the survival, there’s no difference between recurrence-free or overall survival between the two groups.7 About 80% are alive at three years whether you complete their dissection or not.”

Not completing SLN dissection in cases with positive SLN biopsies would represent a monumental practice change, he says. The Multicenter Selective Lymphadenectomy II Trial (MSLT II) is also directly addressing this issue, and accrual is nearly complete, with final results expected sometime around 2022. If this study corroborates the German findings, “There’s a good chance we might see a shift away from completion after positive SLN biopsy.” Because the German trial was underpowered to definitively settle the issue, Dr. Tsao says, “We need the results of MSLT II to change practice. And I believe that over time, we will, especially given the emergence of other surveillance technologies such as nodal ultrasound.”

Next: The potential of Machine learning

The potential of Machine learning

Not to be overlooked, Dr. Tsao says, is the potential impact of machine learning in dermatology. With deep neural network computing, he says, “If you feed the computer a large training set of images and various terms like ‘bridge’ and ‘river,’ then give it an unknown picture, in theory, the computer can take apart its visual elements and correctly assign the unknown to one of the training classes.”

In a Stanford study, such a system proved as effective as 21 dermatologists in distinguishing melanomas from squamous cell and basal cell carcinomas, seborrheic keratoses and benign nevi.8

“Machine learning is huge in image recognition,” Dr. Tsao says. Driverless cars depend on it. “Now it’s coming into dermatology. It can be adapted to mobile devices. It uses advanced neural networks to predict large classes of images. When it comes to these vanguard technologies, we have to focus on the potential benefits for patients and not dwell on the threat to our medical egos.”

The main risk of point-of-care technologies is false negatives, Dr. Tsao says. Nevertheless, he sees true opportunities in the form of patients from the community with suspicious lesions that otherwise would never present to dermatologists.

“If you go to the drug store, there’s often an automated blood-pressure cuff. If someone puts their arm into it and finds they have hypertension 200/100 and then goes to an emergency room, you’ve probably saved the person from a stroke.” The goal of image recognition technology in dermatology is to deploy it with sufficient accuracy that “people coming into your clinic have suspicious lesions of high enough probability that you spend more time accessing melanomas that are being ignored currently in the general population,” he says. DT

Disclosures: Dr. Tsao is a consultant and medical advisory board member for Epiphany Dermatology. He is also on the scientific advisory board for Lubax and has received funding from the National Institutes of Health, the Melanoma Research Alliance and the U.S. Department of Defense, and serves as an editorial advisor for several dermatologic journals.

References:

1. Eggermont AM, Chiarion-Sileni V, Grob JJ, et al. Prolonged survival in stage III melanoma with ipilimumab adjuvant therapy. N Engl J Med. 2016;375:1845-1855

2. Zaretsky JM, Garcia-Diaz A, Shin DS, et al. Mutations associated with acquired resistance to PD-1 blockade in melanoma. N Engl J Med. 2016;375(9):819-29.

3. Weber JS, Hodi FS, Wolchok JD, et al. Safety profile of nivolumab monotherapy: a pooled analysis of patients with advanced melanoma. J Clin Oncol. 2017;35(7):785-792.

4. Nghiem PT, Bhatia S, Lipson EJ, et al. PD-1 blockade with pembrolizumab in advanced Merkel-cell carcinoma. N Engl J Med. 2016;374(26):2542-52

5. Kaufman HL, Russell J, Hamid O, et al. Avelumab in patients with chemotherapy-refractory metastatic Merkel cell carcinoma: a multicentre, single-group, open-label, phase 2 trial. Lancet Oncol. M016;17(10):1374-1385.

6. Jin Y, Andersen G, Yorgov D, et al. Genome-wide association studies of autoimmune vitiligo identify 23 new risk loci and highlight key pathways and regulatory variants. Nat Genet. 2016;48:1418-1424.

7. Leiter U, Stadler R, Mauch C, et al. Complete lymph node dissection versus no dissection in patients with sentinel lymph node biopsy positive melanoma (DeCOG-SLT): a multicentre, randomised, phase 3 trial. Lancet Oncol. 2016;17:757-67.

8.         Esteva A, Kuprel B, Novoa RA, et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature. 2017;