Evidence Helps Refine Melanoma Management

SEATTLE — Evidence from recent and ongoing trials is helping to clarify the best strategies for managing cutaneous melanoma.

A hurdle to better melanoma management has been the high variability of the disease, exemplified in part by its wide-ranging presentations, said Dr. William Dzwierzynski, professor of plastic and reconstructive surgery at the Medical College of Wisconsin in Milwaukee. In fact, accumulating evidence suggests that melanoma may encompass several different diseases with differing biology.

When initially evaluating a suspicious skin lesion, the type of biopsy is critical. "Excisional biopsy is probably the most key thing. You really try not to do an incisional or a shave biopsy," he said, unless the latter is deep and removes the whole lesion. Reassuringly, though, the type of biopsy does not affect survival (Am. J. Surg. 2005;190:913–7)."But we'll never know the depth of the lesion" with an incisional or shave biopsy, he pointed out, "so we'll never have the right prognosis."

Accurate diagnosis of melanoma requires permanent sections. "Melanoma is not accurately diagnosed on frozen sections. Don't do frozen sections on melanoma—you get a lot of false-negatives and a lot of false-positives," Dr. Dzwierzynski said at the annual meeting of the American Society of Plastic Surgeons.

Positron emission tomography (PET) imaging is unreliable for staging in patients with melanoma, yielding a false-negative rate of 79% when used preoperatively to identify occult nodal metastases (Cancer 2005;104:570–9). "There is not any conclusive data that PET scan is any more accurate than a chest x-ray or lab tests," he added. On the flip side, patients should not be assumed to have metastases solely based on a positive PET scan.

"I send everybody who has a melanoma that is 1 mm or greater to an oncologist," he added. "I tell them that the oncologist probably won't have anything to offer you, and that's a good thing. But they are the ones who are going to know if there are any investigational studies or treatment trials."

Whenever possible, patients with advanced disease should be referred for clinical trials. "Investigational therapies—I think this is where the promise is," Dr. Dzwierzynski commented.

When it comes to resecting the tumor, contemporary margins are 1–3 cm for most invasive melanomas. Prospective studies have found no difference in survival between margins of 1–2 cm and larger margins of 3–5 cm, but methodologic limitations leave the issue unresolved, he said.

Mohs surgery for invasive melanoma remains controversial. "There is a lot of distortion when you do Mohs," he noted. "It's really easy to get false-negatives and false-positives." To date, controlled survival data and randomized trials are lacking.

Sentinel node biopsy (SNB) is recommended for patients whose tumors have a Breslow thickness of greater than 1 mm and for those whose tumors are thinner but have adverse features, such as ulceration or a Clark level of IV or V. Currently, it is done to obtain prognostic information and identify the roughly 20% of patients who may benefit from a complete lymph node dissection, Dr. Dzwierzynski noted.

The results of the first Multicenter Selective Lymphadenectomy Trial (MSLT-1) raised the possibility that SNB also may be curing disease in some patients and improving survival (N. Engl. J. Med. 2006;355;1307–17). An ongoing follow-up trial, MSLT-2, is looking more closely at the issue and the possibility that patients with only microscopic disease in the sentinel node may be spared further surgery.

Importantly, there is a learning curve to the SNB procedure. In MSLT-1, the false-negative rate was 10% in a physician's first 25 cases, but fell to 5% thereafter (Ann. Surg. 2005;242:302–13). "So right now, the recommendation is that it does take probably 30 cases for that learning curve," he said.

For pathologic evaluation of sentinel nodes, the combination of step sectioning (at less than 1-mm intervals) and hematoxylin and eosin staining with HMB-45 and S-100 immunohistochemistry has sensitivity approaching 98%, according to Dr. Dzwierzynski. A triple stain used at his institution—the MCW melanoma cocktail (Melan-A, MART-1, and tyrosinase)—has high accuracy (BMC Cancer 2003;3:15), albeit at a fairly high cost. In contrast, polymerase chain reaction has proven to be of limited use because it can be falsely positive in patients with subcapsular nevi.

The National Comprehensive Cancer Network recommends complete lymph node dissection for patients with a positive SNB, but a recent analysis of national data found that only half of such patients underwent the procedure (Ann. Surg. Oncol. 2008;15:1566–76).

"Complete lymph node dissection is a curative procedure," he commented. As such, it is extensive, more so than the lymph node sampling done for, say, breast cancer. "In most of my axillary dissections, I will remove 35–40 lymph nodes," Dr. Dzwierzynski said. "For a superficial inguinal dissection, you should have at least 10 nodes, and for a deep dissection, you have at least 5 nodes."