4 Surgical Treatment of Cutaneous Melanoma
Melanoma is currently the fifth most common cancer in the United States, and its incidence is continuing to increase. In 2017, over 87,000 new cases of invasive melanoma will be diagnosed, accounting for about 4.5% of new cancer cases in the United States. Although the 5-year survival rate continues to improve, and is now estimated at 92% for all stages at presentation, about 10,000 deaths from this disease now occur in the United States annually. 1 Because of its relatively early age of onset, melanoma is one of the most devastating malignant diseases in terms of the years of potential life lost per death. 2
Both environmental and genetic factors increase an individual′s risk for developing melanoma ( Table 4.1). Caucasian race and fair complexion predispose an individual to developing melanoma, as do mutations in several known—and likely multiple as yet unknown—melanoma susceptibility genes. 3 , 4 Melanoma affects a broad age range. According to recent cancer statistics, the peak incidence is between the ages of 55 and 74 years, with a median age at time of diagnosis of 64 years. 5 Three-quarters of all cases will occur before age 70, and approximately 20% occur in individuals under 40 years of age. Environmental exposure in the form of ultraviolent radiation is the most well-studied and modifiable etiologic factor in the development of melanoma. Intermittent, high-intensity exposure (e.g., blistering sunburns) contributes most dramatically to the development of melanoma. Dark-skinned individuals may be less likely than fair-skinned individuals to develop melanoma due to the protective effect of eumelanin. Even in individuals with a genetic predisposition, exposure to ultraviolet radiation is the primary factor determining whether an at-risk person actually develops melanoma. 6 , 7
4.2 Biopsy Techniques and Orientation of Incision
Any patient who presents with a new or abnormal pigmented skin lesion, with increasing size, change in color, or developing irregular borders, should be considered for a biopsy for definitive diagnosis. The approach to biopsy for a suspicious cutaneous lesion often varies depending on the location of the lesion and the preference and training of the physician performing the biopsy. The gold standard for tissue diagnosis when melanoma is suspected is excisional biopsy with a 1 mm to 2 mm margin of normal skin, including some subcutaneous fat (see Chapter 2). This may not be feasible in all cases due to the location of the lesion, and often is more than the minimum necessary for the diagnosis of a benign lesion. 8 Several other biopsy approaches are reasonable to achieve adequate tissue to allow the pathologist to make a diagnosis and assess all relevant histologic features. These techniques include superficial shave biopsy, deep scallop shave biopsy, punch biopsy, incisional biopsy, and the aforementioned excisional biopsy. When any form of partial biopsy is performed, the biopsy should focus on obtaining a sample of the most elevated portion of the lesion to provide the best estimate of tumor thickness. It is ideal to also include a portion of normal skin in the biopsy; however, the main goal should be to obtain the most suspicious and elevated portion of the lesion in the specimen. For lesions that are large or in cosmetically sensitive areas, a punch biopsy that obtains a full-thickness sample is often the best initial diagnostic approach. If a partial biopsy is performed and a portion of the suspicious lesion remains, indicating this on the pathology request helps make the pathologist aware that the lesion was not removed in its entirety, aiding in the diagnostic process.
If a lesion is suspected to be melanoma, cryosurgery and electrodessication should not be performed given they can distort normal tissue planes and do not allow for an adequate histologic evaluation. Partial-thickness shave biopsy of pigmented lesions remains controversial. One of the biggest criticisms of shave biopsy is that partial sampling does not assess the true thickness of a melanoma, if present. Thus, inaccurate microstaging and inappropriate treatment decisions may result. 9 , 10 A multi-institution, retrospective study evaluated over 600 patients diagnosed via shave biopsy with a melanoma that had a Breslow thickness ≤ 2 mm to determine the accuracy of microstaging in these patients. The results demonstrated that the treatment recommendations for excision margin width or sentinel lymph node biopsy changed based on more extensive residual melanoma in the re-excision specimen in only 2% and 1.3% of patients, respectively. 11 This study supports the use of shave biopsy as a diagnostic tool for patients with pigmented lesions, especially when they are not considered highly suspicious for melanoma. In essence, it is better to have a partial biopsy of a melanoma than no biopsy at all.
The orientation of the biopsy incision is also of importance, in order to facilitate subsequent surgical treatment in the event of a melanoma diagnosis. A punch or shave biopsy will not impact the approach to definitive surgery of a melanoma, but incisional and excisional biopsies should be planned with the potential need for wide excision and/or sentinel node biopsy in mind. Utilizing the principles of Langer′s lines may achieve a better cosmetic result for the biopsy scar, but Langer′s lines often run in a transverse direction on the extremity, whereas biopsy incisions there should be vertical in orientation. Doing so allows for subsequent wide excision to be done with the greatest chance for primary closure while minimizing the disruption of lymphatics that could impact the accuracy of sentinel node biopsy, if required.
4.3 Margins of Excision
The goal of the initial biopsy is to obtain a diagnosis, but once a diagnosis of melanoma is made, the emphasis shifts to determining the appropriate margins of excision around the biopsy site or residual pigmented lesion. Adequate margins may reduce the risk of local recurrence, while insufficient margins may predispose the patient to a recurrence of disease. A histologically negative margin is always the goal of surgery on the primary melanoma site. The chosen margin should be measured from the edges of the previous biopsy site or any residual pigmentation. Of note, these are “radial” margins, meaning measured in all directions on the skin except downward. The deep margin is the investing fascia of the muscle, which may or may not be included en bloc at the surgeon′s discretion. The recommendation for how much of a margin of normal tissue is taken beyond the biopsy site is based on the Breslow thickness of the melanoma as well as the anatomic location of the tumor. Different margin widths may be recommended for melanomas that are in situ, thin (≤ 1 mm), intermediate (1.1–2 mm or 2.1–4 mm), or thick (> 4 mm).
4.3.1 Melanoma in situ
Melanoma in situ is a lesion with a capacity to progress to invasive melanoma and, like invasive melanoma, should be resected with a measured margin of normal skin and with the goal of achieving histologically negative margins. Recommendations from the National Comprehensive Cancer Network (NCCN) are for a 0.5 cm to 1 cm margin; however, no randomized, prospective study has been conducted to definitively answer what margin is truly necessary. 12 In the absence of prospective data, we take into account the location of the lesion and the presence of any residual pigmentation (which might signal the presence of invasive melanoma) in deciding on an appropriate margin of excision for melanoma in situ. In areas with greater tissue laxity, a 1 cm margin may be more easily obtained and is therefore favored, especially when residual pigmentation is present. Conversely, in cosmetically sensitive areas such as the face, neck, hands, and feet, or in locations where the skin is under greater tension, a margin of 0.5 cm is more likely to be used.
4.3.2 Invasive Melanoma: Thin and Intermediate Thickness
The recommended margins of excision for invasive melanoma range from 1 to 2 cm based on the Breslow thickness of the primary tumor. Current NCCN recommendations call for a 1-cm margin for thin melanoma (T1), 1 cm to 2 cm for intermediate melanoma (T2–T3), and 2 cm for thick melanoma (T4). 12 The World Health Organization (WHO) trial, two Swedish Melanoma Study Group Trials, and the Intergroup Melanoma Trial have provided prospective randomized trial data that serve as the basis for current margin width recommendations ( Table 4.2). In the WHO trial, over 600 patients with melanomas ≤ 2mm in thickness were randomized to either a 1 cm or 3 cm margin. There was no statistical difference between the study arms in local recurrence, disease-free survival, or overall survival. 13 This trial concluded 3 cm margins were not necessary for melanomas 2 mm or less in thickness. In particular, the risk of any form of recurrence was very low for melanomas 1-mm thick or thinner even with a 1 cm margin. Hence, a 1 cm margin has become the worldwide standard for excision of invasive melanomas up to 1 mm in thickness, regardless of location of the tumor.
The first Swedish Melanoma Study Group Trial evaluated 989 patients with melanomas 0.8 mm to 2 mm in thickness that were resected with either a 2 cm or 5 cm margin. The study evaluated overall and recurrence-free survival and found no statistically significant difference between the arms. 14 The authors concluded that 5 cm margins were not necessary for a melanoma 2 mm or less in thickness.
The Intergroup Melanoma Trial evaluated outcomes when margins of 2 cm versus 4 cm were used for patients with a melanoma between 1 mm and 4 mm in thickness. The study evaluated local recurrence as the site of first relapse and at any time, and found no significant difference in outcomes between the two arms. The local recurrence rates as first relapse and at any time for a 2 cm margin (0.4 and 2.1%, respectively) were not significantly different when compared to a resection margin of 4 cm (0.9 and 2.6%, respectively). A trend toward improved overall survival at 10 years was seen in the 4-cm group, 70 versus 77%, but this was not statistically significant. 15
In the second Swedish Melanoma Study Group Trial, 936 patients with a melanoma greater than 2-mm thick were randomized to margins of either 2 cm or 4 cm. The authors reported no statistically significant differences in local recurrence or overall survival between the two arms. 16
Haigh et al conducted a meta-analysis and systematic review of three of these trials (WHO, Intergroup Trial, and the first Swedish Melanoma Study Group Trial). When the data were combined, local recurrence, disease-free survival, and mortality were not statistically different between patients with primary cutaneous melanoma up to 4 mm in thickness undergoing excision with 1 cm to 2 cm margins compared to those randomized to 3 cm to 5 cm margins. 17
Taken together, the data from all of these studies support the recommendation of 1 cm margins for a thin melanoma and 1 cm to 2 cm margins for an intermediate thickness melanoma. Although no randomized trials have directly compared 1 cm versus 2 cm margins for an intermediate thickness melanoma, we use a 2 cm margin for most melanomas that are 2 mm in thickness or greater except in anatomically constrained areas like the face, ears or distal extremities, where only a 1 cm margin can routinely be obtained. For a melanoma between 1.1 mm and 2 mm in thickness, a 1 cm margin is acceptable when the narrower margin allows for primary closure but a 2 cm margin would require skin grafting or significant tissue undermining to achieve closure. We are particularly likely to use a 1 cm margin for melanomas in this thickness range that are nonulcerated, with a relatively low mitotic rate and no evidence of angiolymphatic invasion. We have shown that using a 1 cm margin in this setting was associated with an acceptably low rate of local recurrence. 18
4.3.3 Thick Melanoma
The NCCN recommends that thick melanomas, defined as tumors greater than 4 in thickness, should be resected with a 2 cm margin. 12 A large, multi-institution retrospective (non-randomized) study of over 270 patients reviewed outcomes in patients with a thick melanoma who had a resection margin of either 2 cm or greater than 2 cm. Margins wider than 2 cm did not improve outcomes, given that local recurrence, disease-free survival, and overall survival were not significantly different between the two treatment arms after a median follow-up of 27 months. 19
Thomas et al randomized 900 patients with tumor thickness of 2 mm or greater to excision with either a 1 cm or 3 cm margin. After a median follow-up of 60 months, there was increased locoregional recurrence in the 1 cm arm (p = 0.05), but no significant difference in overall survival between the arms. This study prohibited any form of nodal staging, be it elective lymph node dissection or sentinel lymph node biopsy. The absence of nodal evaluation raises an important point when considering the impact of the results of this trial. If lymph node evaluation had been conducted, there is a possibility that there would have been an imbalance in regional nodal disease between the two arms, which may have attributed more to the locoregional recurrence rate than the actual margin of excision. When only local and in-transit recurrences were considered, there were 37 recurrences in the 453 patients on the 1-cm arm (8.2%) versus 25 in the 447 patients on the 3-cm arm (5.6%), a difference that was not statistically significant. But there were 20 more nodal basin recurrences in the 1 cm arm than in the 3-cm arm, reinforcing the possibility that any observed outcome differences were at least partially related to imbalances in the number of stage III patients in each arm rather than an effect of the surgical treatment. 20 While there are no data directly comparing a 2 cm and a 3 cm margin for thick melanoma, whenever a 2 cm margin has been prospectively or retrospectively compared to a wider (4 cm or 5 cm) margin, there has been no evidence of an advantage for the wider margin. Hence, we do not recommend either a 1 cm or a 3 cm margin for a thick melanoma, but rather routinely use a 2 cm margin whenever anatomically feasible, even if it requires a reconstruction to achieve this margin.
4.4 Sentinel Lymph Node Biopsy
While the margin of resection is important for control of local disease, melanoma frequently metastasizes via the lymphatic system, and the regional nodal basins are the most common site of melanoma metastasis. Evaluation of the nodal basin may be indicated at the time of initial wide excision. The sentinel lymph node represents the first lymph node in the lymph node basin directly draining the primary tumor. If the melanoma has metastasized via lymphatic channels, the sentinel lymph node would be the first lymph node to contain tumor deposits. The status of the regional nodes as determined by sentinel lymph node biopsy to determine the presence or absence of metastatic disease has been shown in a prospective randomized trial to be the most important prognostic factor in patients with clinically localized, intermediate-thickness melanoma. 21
Sentinel lymph node biopsy should be performed in appropriately selected patients with clinically negative regional nodes at the time of wide excision of the primary tumor. A dual localization technique utilizing both injection of radiocolloid and injection of vital blue dye, either isosulfan blue or methylene blue, is most commonly used. The blue dye allows for direct visualization of the lymphatic channel and the sentinel lymph node itself. The use of blue dye alone has allowed the sentinel node to be identified in approximately 85 to 90% of cases, but the combination of both radiocolloid and blue dye increases the success rate to over 99%. 22 We utilize isosulfan blue whenever a vital blue dye is employed, and do not use methylene blue for sentinel node localization at all due to higher reported local complication rates. 23 One-half to 1 mCi of radiocolloid labeled with 99mtechnetium is injected intradermally, generally 1 to 4 hours before surgery, but if necessary the injection may be performed up to 24 hours prior to surgery. 22 A variety of different radiocolloids are used around the world (e.g., sulfur colloid, tilmanocept, albumin nanocolloid and antimony trisulfide), with little or nothing in the way of head-to-head comparison data to determine which might be superior. 24 , 25 Whichever and whenever the radiocolloid is injected, planar or computed tomographic lymphoscintigraphy is performed during the hour immediately thereafter in order to delineate the number and general location of the sentinel nodes. After the induction of anesthesia in the operating room, the vital blue dye is injected intradermally around the biopsy site or the primary tumor. A total volume of 0.5 to 1.5 mL is generally sufficient to identify the sentinel lymph nodes. Intraoperatively, a handheld gamma probe helps to direct the trajectory of dissection and guide the surgeon down to the “hot” and usually blue sentinel node(s).
Despite the high rate of accuracy with a dual localization technique, some surgeons do not routinely use blue dye. The choice to use radiocolloid alone may be due to limited availability of vital blue dye, concern that the use of blue dye will stimulate an allergic reaction, or because the patient is pregnant (we do not use blue dye during pregnancy because a maternal allergic reaction could be catastrophic for the fetus). 23 , 26 , 27 , 28 , 29 In the Multicenter Selective Lymphadenectomy Trial-I (MSLT-1), the incidence of allergic reaction after isosulfan dye injection was only 0.2%, 30 and in our experience the incidence of allergic reactions is even lower than that. Hence, we routinely utilize isosulfan blue dye in nonpregnant patients. One other issue relating to the use of vital blue dye is the presence of a tattoo on the skin in the drainage area of the nodal basin(s) undergoing sentinel node biopsy. Tattoo ink gradually leaches out of the skin site and accumulates in the regional nodes, coloring them enough to be easily seen. If the tattoo contained blue, green, or black ink, pigments within the node can be confused with vital blue dye or even metastatic melanoma. 31 For this reason, we sometimes do not use vital blue dye in cases where a large tattoo is in the same draining basin as the primary melanoma, and focus our attention only on the radioactivity within the nodes.
The morbidity associated with sentinel lymph node biopsy is low, but potential complications include seroma, lymphocele, wound infection, and, in very rare cases, lymphedema. 30 , 32 The morbidity of sentinel lymph node biopsy has been evaluated in prospective clinical trials. In MSLT-1, the rate of complications for patients undergoing sentinel lymph node biopsy was 10.1%. 30 The Sunbelt Melanoma Trial evaluated complication rates for sentinel lymph node biopsy alone and, for patients with positive sentinel nodes, sentinel lymph node biopsy and completion lymph node dissection. The complication rates were higher in patients undergoing lymph node dissection, while the risk of complications was low for sentinel lymph node biopsy alone (wound infection 1.1%, lymphedema 0.7%, and seroma/hematoma 2.3% for sentinel node biopsy versus 7, 11.7, and 5.9% for sentinel node biopsy followed by lymph node dissection). 33 , 34
While the utility of sentinel lymph node biopsy in determining the spread of cancer and providing a prognostic benefit is clear, the role it plays in providing a survival benefit is not. One prospective, randomized trial has addressed this question. In MSLT-1, patients were randomly assigned to undergo either wide excision of the primary tumor with observation of the nodal basin or wide excision with sentinel lymph node biopsy and subsequent completion lymph node dissection if there was a positive sentinel lymph node. Final analysis of the 1,347 patients with intermediate thickness melanoma who were evaluated in the study demonstrated no difference in melanoma-specific survival at 10 years for patients who underwent sentinel lymph node biopsy versus observation. There was, however, an improvement in disease-free survival for patients undergoing sentinel lymph node biopsy. Patients with a positive sentinel lymph node biopsy who then had completion lymph node dissection were found to have an improved distant disease-free survival compared to patients in the observation group who failed in the nodal basin with clinically obvious disease (54.8 versus 35.6%, p = 0.02). Melanoma-specific survival at 10 years was also improved in the sentinel lymph node biopsy positive group compared to the observation group of patients who failed in the nodal basin (62.1 versus 41.5%, p =0.006). 30
The NCCN currently recommends that sentinel lymph node biopsy be discussed and offered to patients with intermediate-thickness melanoma (1–4 mm). 12 Controversy still exists regarding the routine recommendation for sentinel lymph node biopsy in patients with thin melanoma (< 1 mm). The risk of having a positive sentinel lymph node biopsy in a patient with a thin melanoma has been reported in the literature to be around 5%, but there are clearly subsets of patients within the broader group of T1 melanoma cases with a higher or lower risk of a positive sentinel node. Since we and many others use a 5% sentinel node positivity threshold as the level at which to consider the procedure for an otherwise healthy patient, it is important to identify which subsets of patients with T1 melanoma are or are not good candidates for sentinel node biopsy. Studies have demonstrated that the percentage of cases with a positive sentinel node increases when the melanoma is ≥ 0.76 mm in thickness, ulcerated, or mitotically active (mitotic rate ≥ 1/mm2). 35 , 36 , 37 The NCCN, American Society of Clinical Oncology (ASCO), and Society of Surgical Oncology (SSO) all recommend that sentinel lymph node biopsy be considered in patients with melanomas that are 0.76 to 1 mm if the melanoma is ulcerated or the mitotic rate is ≥ 1/mm2. Sentinel lymph node biopsy, however, should still be considered in appropriately selected patients with thin melanomas in this thickness range (0.76–1 mm) that are not ulcerated or mitotically active. 12 , 38
Controversy also exists regarding sentinel lymph node biopsy for patients with clinically node-negative thick melanomas (≥ 4 mm). The final analysis of MSLT-1 evaluated 290 patients with thick melanomas, defined as greater than 3.5 mm. Fifty-seven (33%) out of the 173 patients in the sentinel lymph node biopsy arm had sentinel lymph node metastasis, while 116 (67%) had no evidence of nodal metastasis. Twelve of the 116 sentinel node-negative patients eventually developed nodal disease, bringing the overall incidence of nodal metastasis in the sentinel lymph node biopsy arm to 39.9% (the rate of nodal metastasis in the nodal observation arm was 37.6%). The 10-year disease-free survival rates for the sentinel lymph node biopsy and observation arms were 50.7 and 40.5%, respectively (p = 0.03). In those patients with thick melanoma found to have sentinel lymph node metastasis, the 10-year melanoma-specific survival was significantly worse (48%) compared to those without sentinel lymph node metastasis (64.6%, p = 0.03). 39 , 40
More recently, Yamamoto et al published a single institution, retrospective review of patients seen over a 10-year period with melanoma ≥ 4 mm. The endpoints evaluated included recurrence-free survival and disease-specific survival. Sentinel lymph node biopsy was performed in 412 patients and of those, 161 (39%) had a positive sentinel lymph node. Forty-six other patients presented with clinically palpable disease. Of the 251 patients who had a negative sentinel lymph node biopsy, 14 (6%) recurred in the mapped nodal basin, for a calculated false-negative rate of 12.3%. The median disease-specific survival of 82 months for patients with a negative sentinel lymph node was significantly better than for those with a positive sentinel lymph node (41 months) or for patients with clinically node-positive disease (27 months; p < 0.0001). The authors also reported a significant difference in median recurrence-free survival between patients with negative sentinel lymph nodes compared to patients with positive sentinel lymph nodes or clinically palpable disease (32 versus 14 and 7 months, respectively; p < 0.0001). The authors concluded sentinel lymph node biopsy should be considered in those with clinically node-negative thick melanoma due to the prognostic information obtained and the acceptably low false-negative rate. 41
Some surgeons have advocated repeat sentinel lymph node biopsy in the event of local recurrence or development of resectable in-transit disease, assuming a prior sentinel lymph node biopsy was negative. 42 , 43 , 44 The technical feasibility of repeat sentinel lymph node biopsy has been evaluated in two separate studies, and the ability to detect a sentinel lymph node using either 99mtechnetium sulfur colloid alone or in combination with blue dye was reported as 12 of 12 (100%) cases in one study and 32 of 33 (97%) cases in another. 42 , 45 However, the prognostic significance and staging reliability of finding a “sentinel” lymph node in the context of recurrence is not clear. At present, we do not perform repeat sentinel node biopsy for patients with local or in-transit recurrence, preferring instead to evaluate and follow the nodal basin with serial ultrasonography. If, however, a patient presents with local recurrence without ever having undergone a sentinel node biopsy, we will consider sentinel node biopsy in conjunction with wide excision of the local recurrence site in selected patients.