Historical Perspective and the Emergence of a Contemporary Paradigm

At the beginning of the twentieth century, melanoma was rarely recognized early and was usually diagnosed when it was locally advanced. Based purely on anecdotal experience, the use of 5 cm radial margins was adopted as the standard approach. The origin of this recommendation is somewhat ambiguous, although it is usually (but probably incorrectly) attributed to a report in 1907 by Handley (1907), who described a wide resection margin that included 1 inch (i.e., 2.5 cm not 5 cm) of surrounding skin, but with an additional en bloc excision of 2 inches of subcutaneous tissue and underlying fascia. Possibly supporting the use of 5 cm skin excision margins were the pathologic descriptions by Olsen and Wong who reported that atypical melanocytes may extend for up to 5 cm beyond the primary lesion (Olsen 1966; Wong 1970). Regardless of the origin of the recommendation, the 5 cm excision approach persisted until the 1970s, when a better understanding of the natural history of melanoma and the prognostic factors that influenced outcome (tumor thickness, in particular) were established (Breslow and Macht 1977; Balch et al. 1979). Surgeons then began to use narrower margins for lower-risk (thinner) lesions and reported excellent results (Kelly et al. 1984; Cosimi 1985; Day Jr. et al. 1982).

Although several published reports documented a low incidence of local recurrence when narrow margins were used for thin melanomas, several retrospective studies suggested that local recurrence was relatively frequent when such margins were used to treat thicker lesions. This indicated that local recurrence was a function of both inherent biology and the extent of the margin (Milton et al. 1985). A paradigm of wider margins being necessary for higher-risk (thicker) lesions emerged and was subsequently tested in several prospective, randomized surgical trials. If such a paradigm was true – that outcome (the frequency of local and regional recurrence and overall survival ) is affected by the extent of excision margins – the following assumptions could be made: (1) microscopic satellite disease is more common and exists at a greater distance from the periphery of the primary lesion in association with thicker melanomas; (2) these microsatellites are a source of subsequent clinically-apparent local, regional, and distant relapses; and (3) wider margins remove microscopic disease that would be left behind if narrower margins were used, resulting in an improved outcome.

In this chapter, evidence relating to the role of excision margins in patients with primary melanomas is summarized. All these trials and retrospective studies were designed to test the hypothesis that wider margins would be superior to narrower margins in terms of locoregional control and survival. A concise description of trial randomization schemes and the number of patients included in each is presented in Table 1.

Table 1 Prospective randomized trials addressing surgical excision margins for primary cutaneous melanoma
Full size table

Wide Excision of Primary Melanomas: Fundamental Concepts

Wide surgical resection of the diagnostic excision biopsy site or residual primary lesion, including a surrounding margin of normal skin en bloc with the underlying subcutaneous tissue, usually down to the deep fascia, has been adopted as the surgical standard for initial management of primary melanomas. Whereas the goal of an appropriately performed diagnostic biopsy is to establish pathologic microstaging (primarily tumor thickness and ulceration status), the goal of wide local excision is to achieve durable local disease control and to cure those patients who do not already harbor disseminated micrometastases in regional lymph nodes or at distant sites.

The optimal width of an excision margin has been a matter of controversy for decades. The vast majority of “narrow” excision procedures (with 1 cm surgical margins) allow primary closure of the resultant surgical defect and can be accomplished using local anesthesia, at minimal cost and with little or no morbidity (Bono et al. 1997). On the other hand, “wide” excision margins (>2 cm) are often performed with the patient under general anesthesia and, depending on the anatomic site, may require skin grafting or reconstruction with sometimes complex full-thickness skin advancement or rotation flaps . However, narrow excision margins for higher-risk lesions are often thought to result in a higher rate of locoregional recurrence, which may impact on survival.

Because local recurrence is generally associated with a poor prognosis and often requires complicated and potentially morbid treatment, prevention of such events has been a compelling reason to recommend wide excision margins. At the same time, minimizing treatment-related morbidity is another important management goal. Use of excessive margins with no improvement in outcome can lead to unnecessary morbidity, additional cost, more complications, and permanent disfigurement. Therefore, establishing rational standards for the extent of surgical excisions has been a focus of extensive clinical investigation.

When considering excision margins, the fundamental difference between surgical margins and histologic margins must be borne in mind. All current treatment recommendations refer to surgical margins, i.e. the in vivo margins determined before wide excision and before histologic examination of the specimen. It must also be borne in mind that there is a predictable degree of tissue shrinkage after the specimen has been removed and fixed in formalin. A detailed study by Friedman et al. (2019) showed that there was usually a 14% shrinkage rate, so that the macroscopic margins measured by the pathologist at the time of specimen cut-up will be less than the margins marked out by the surgeon prior to the wide excision.

T0: Melanoma In Situ/Lentigo Maligna

Melanoma in situ is the earliest recognizable stage of melanoma. It is noninvasive, but may be a precursor to invasive melanoma, which has the potential for metastasis. Lentigo maligna is a subtype of melanoma in situ associated with chronic exposure to ultraviolet radiation. These lesions are insidiously extensive microscopically and, despite the removal of normal-appearing skin surrounding the pigmented lesion, histologically positive margins may be encountered. Lentigo maligna lesions can require sequential operative excisions to achieve adequate, pathologically clear margins, and occasionally the use of a split-thickness skin graft or flap closure is warranted.

There have been no randomized trials and only a limited number of case series to guide excision margins for melanoma in situ (Tzellos et al. 2014; Cancer Council Australia Melanoma Guidelines Working Party 2018) and treatment therefore is based on consensus from clinical experience. Until recently, most guidelines suggested that 5 mm excision margins were adequate for melanoma in situ. However two case series (Bosbous et al. 2009; Akhtar et al. 2014) demonstrated that 5 mm may be inadequate, leading to disease recurrence in some cases. The risk of recurrence of melanoma in situ was particularly great for the lentigo maligna subtype and for head and neck sites. In many cases, adequate clearance margins for melanoma in situ can be accurately determined preoperatively by careful examination and confirmed by pathology. However, defining the extent of the lesion can be difficult in lentigo maligna because it is characterized by the subtle presence of scattered atypical melanocytes, meaning positive margins can remain after excision, leading to possible recurrence at excision sites (Akhtar et al. 2014; Bosbous et al. 2009). A recent study reported that the use of in vivo confocal microscopy may improve lesion boundary identification and allow for more effective excision (Guitera et al. 2013). However, further studies are required to validate this technique.

Current Excision Margin Recommendations for T0 Melanomas

European guidelines suggest 5 mm excision margins for melanoma in situ (Dummer et al. 2012; Garbe et al. 2016; NICE 2015) although a British Best Practice statement (BMJ Best Practice 2016) includes the caveat that 5 mm is inadequate in up to 50% of cases of melanoma in situ, particularly lentigo maligna, and 10 mm margins, staged excision or Mohs surgery are options. Dermatology guidelines from the USA recommend 5–10 mm margins and state that wider margins may be necessary for the lentigo maligna subtype of melanoma in situ (Bichakjian et al. 2011). Table 2 details the recommendations from these and various other national guidelines. Based on all the available evidence, it would seem most appropriate to widely excise melanoma in situ with a minimum margin of 5 mm, but with a margin of 10 mm when this is readily possible.

Table 2 National guidelines for excision margins for T0 (in situ) cutaneous melanoma
Full size table

T1: Invasive Melanomas ≤1 mm in Thickness

Randomized Trials of Excision Margins for T1 Melanomas

No randomized trials have exclusively included only patients with primary tumors ≤1 mm thick (often referred to as “thin” melanomas). Three trials included 159/326 (49%), 356/612 (58%), and 244/989 (25%) patients with primary tumors 5 to 10 mm in Breslow thickness (Khayat et al. 2003; Veronesi and Cascinelli 1991; Ringborg et al. 1996). Two trials compared 2 cm with 5 cm margins (Khayat et al. 2003; Ringborg et al. 1996), while the WHO trial compared 1 cm with 3 cm margins (Veronesi et al. 1988). Individually, the trials reported nonsignificant differences in the risk of death, local recurrence, nodal recurrence, or distant metastasis between the narrow (1 or 2 cm) and wider (3, 4, or 5 cm) excision margins (Table 1). In combining these trials, the pooled estimate for risk of death associated with narrower versus wider margins was 0.95 (95%CI 0.76–1.17), risk of local recurrence 0.92 (95%CI 0.72–3.37), risk of nodal recurrence 1.15 (95%CI 0.88–1.49), and risk of distant metastasis 1.00 (95%CI 0.66–1.50) (Fig. 1a–c). Thus, all estimates showed no significant differences when patients were treated with 1 or 2 cm margins compared to 3, 4, or 5 cm margins. Two of these trials (Cascinelli 1998; Khayat et al. 2003) reported local recurrence within the group of patients with primary tumors ≤1 mm, and when combined gave a risk ratio of 1.30 (95%CI 0.18–9.70) but with only five events, the risk estimate ranged from 82% less likely to almost 10 times more likely that a local recurrence would occur in a patient treated with a narrow (1–2 cm) compared to a wide (3–5 cm) margin (Fig. 1e). These trials do not report outcomes of death, nodal metastasis or distant metastasis within the group of patients with primary tumors ≤1 mm thick; therefore, it is not possible to determine the applicability of these findings to this subgroup.

Fig. 1
figure 1

(a) Risk of death from all causes in trials of patients with primary tumors ≤2 mm thick randomized to narrow (1–2 cm) margins or wide (3–5 cm) margins. (b) Risk of local recurrence in trials of patients with primary tumors ≤2 mm thick randomized to narrow (1–2 cm) margins or wide (3–5 cm) margins. (c) Risk of nodal recurrence in trials of patients with primary tumors ≤2 mm thick randomized to narrow (1–2 cm) margins or wide (3–5 cm) margins. (d) Risk of distant metastases in trials of patients with primary tumors ≤2 mm thick randomized to narrow (1–2 cm) margins or wide (3–5 cm) margins. (e) Risk of local recurrence in the subgroup of patients with primary tumors ≤1 mm thick randomized to narrow (1 or 2 cm) margins or wide (3 or 5 cm) margins

Full size image

Nonrandomized Studies of Excision Margins for T1 Melanomas

In a case control study, 174 patients with primary melanomas ≤1 mm thick who experienced a recurrence of disease were matched with patients who did not have a recurrence; these two groups were selected from a patient cohort of 10,505 and were analyzed for associations with recurrence. (MacKenzie Ross et al. 2016). Excision margin was associated with time to recurrence with a hazard ratio of 0.95 (95%CI 0.92–0.98), meaning that for every 1 mm decrease in excision margin, the patient had a 5% reduction in time to recurrence. Categorizing the margin into <8 mm histological margin (<1 cm surgical margin) and ≥8 mm (≥1 cm surgical margin) showed a significant difference in time to recurrence, with the <8 mm margin group having earlier recurrence than the ≥8 mm group. Interestingly, however, multivariate analysis did not show an association with presence or absence of local recurrence .

A registry study of New Zealand melanoma patients included 620 patients with primary tumors ≤1 mm thick (Ng et al. 2001). This analysis showed that the optimum margin for zero local recurrence in lesions ≤1 mm was 1 cm; however, their dataset included only 16 recurrences and must therefore be interpreted with caution.

A Swedish study of 585 patients with primary melanomas ≤0.8 mm in thickness reported a median time to recurrence of 50 months in 26 patients (Månsson-Brahme et al. 1994). No difference in the recurrence rate in patients treated with 1–2 cm excision margins compared to 5 cm margins was reported, suggesting that a 1 cm margin is adequate in this group but providing no data for margins <1 cm.

Current Excision Margin Recommendations for T1 Melanomas

Excision with a 1 cm margin of clinically normal skin and underlying subcutaneous tissue is the consistent recommendation for melanomas ≤1 mm thick in guidelines across the world (see Table 3). Except for some lesions arising on the distal extremities, scalp, or face, essentially all 1 cm margin excision defects can be closed primarily.

Table 3 National guidelines for excision margins for T1 cutaneous melanoma (≤1 mm thick)
Full size table

T2: Invasive Melanomas >1–2 mm Thick

Randomized Trials of Excision Margins for T2 Melanomas

As was the case for T1 melanomas, trials have not exclusively randomized patients with primary tumors 1–2 mm thick, but four trials included them: 245/612 (40%), 745/989 (75%), 166/326 (51%), and 272/486 (56%). Two of these trials compared 2 cm with 5 cm margins (Khayat et al. 2003; Ringborg et al. 1996), one trial compared 1 cm and 3 cm margins (Veronesi et al. 1988), and one trial compared 2 cm with 4 cm margins (Balch et al. 1993). The Intergroup trial (Balch et al. 1993), which included primary tumors 1–4 mm in thickness, reported separately the local, nodal, and distant metastasis outcome data for the groups <2 mm and >2 mm, and the data from the <2 mm group are reported and discussed here. Unfortunately, deaths within this group were not reported. Meta-analysis of all four trials gave estimates of local recurrence (risk ratio 1.19, 95%CI 0.35–4.02), nodal recurrence (risk ratio 1.15, 95%CI 0.89–1.48), and distant metastasis (risk ratio 1.10, 95%CI 0.82–1.46) that showed no significant differences in risk between the narrow margin group (1 or 2 cm) and the wider margin groups (3, 4, or 5 cm) (see Fig. 2a–c). Risk of death in the three trials reporting this outcome gave a pooled estimate of 0.95 (95%CI 0.80–1.13) (Fig. 1). Findings were quite consistent for all outcomes except local recurrence, which showed large variability, probably reflecting the low event rates (15/1088; 1.4% and 12/1126; 1.2%) and chance effect. Three trials reported the number of patients with local recurrence in the subgroup of patients with primary tumors 1–2 mm thick, and combining their data gave a risk ratio of 1.75 (95%CI 0.37–8.40) (Fig. 2d). However, although the narrow margin group may have been almost twice as likely to experience a local recurrence than the wide margin group, this estimate ranged from 60% less likely to up to eight times more likely, as events were rare (only 14 among the 707 patients). None of the four trials reported the numbers of skin grafts, wound infections, or surgical complications, so no information about morbidity is available.

Fig. 2
figure 2

(a) Risk of local recurrence in four trials of excision margins 1–2 cm versus 3–5 cm in trials of patients with primary tumors <2 mm thick. (b) Risk of nodal recurrence in four trials of excision margins 1–2 cm versus 3–5 cm in trials of patients with primary tumors <2 mm thick. (c) Risk of distant metastases in four trials of excision margins 1–2 cm versus 3–5 cm in trials of patients with primary tumors <2 mm thick. (d) Risk of local recurrence in the subgroup of patients with primary tumors 1–2 mm thick in three trials of excision margins 1–2 cm versus 3–5 cm

Full size image

To date, one randomized trial, coordinated by the Australian and New Zealand Melanoma Trials Group, has compared 1 cm and 2 cm excision margins in patients with primary tumors >1 mm in thickness, of whom 223/377 (59%) had primary tumors 1–2 mm thick. To date, only 12-month data have been reported, with outcomes of reconstruction, wound necrosis, wound dehiscence, and infection, along with quality of life. Patients undergoing a 1 cm excision were much less likely to require reconstruction with a flap or skin graft (risk ratio 0.39, 95%CI 0.26–0.59) and somewhat less likely to experience wound necrosis (risk ratio 0.15, 95%CI 0.02–1.20) than those treated with a 2 cm excision margin. When first reported in 2018, the trial had insufficient follow-up time for survival and recurrence outcomes to be assessed (Moncrieff et al. 2018).

A second study coordinated by the University of Kansas is comparing 1 cm and 2 cm margins in patients with T2 melanomas (1–2 mm thick) (NCT 03034395), but no results are yet available.

Nonrandomized Studies of Excision Margins for T2 Melanomas

A retrospective analysis of 2131 Australian patients with primary melanomas between 1 and 2 mm thick who were treated between 1992 and 2012 was undertaken to assess whether excision margins influenced prognosis and survival, and to identify a minimum safe excision margin (Haydu et al. 2016). A total of 326 patients were treated with a histologic excision margin of <8 mm (equivalent to a <1 cm surgical margin), 1219 with an 8–16 mm excision margin (equivalent to a 1–2 cm surgical margin), 544 with a 16–24 mm (equivalent to a 2–3 cm surgical margin), and 42 with a ≥24 mm excision margin (equivalent to a ≥3 cm surgical margin). For the whole cohort, local or in-transit recurrence occurred in 11.1% and any recurrence in 26.4%. A higher recurrence rate (38.5%) was seen in those with a <8 mm pathologic margin (<1 cm surgical margin) compared to those with a margin ≥8 mm (24.8%). While the <8 mm excision margin patients had more in-transit, local, nodal, and distant metastases, this did not translate to significantly reduced survival rates. In an earlier study that had some patient overlap with this one, 2681 patients with primary melanomas 0.1 to 2.0 mm thick, local recurrence occurred in 55 patients (McKinnon et al. 2005). Patients with local recurrence had a narrower mean and median excision margin (13.5 mm/16 mm) compared to the whole cohort (17 mm/19.9 mm), and there was a significant difference in local recurrence in patients with margins <8 mm compared to those treated with margins ≥8 mm.

The finding that margins less than 1 cm may be too narrow and increase the risk of local recurrence is consistent with the findings of a meta-analysis published in 2003 (Haigh et al. 2003). However, for the outcome of melanoma-specific survival , analysis did not show an association between margin width and survival. This lack of association between margin status and overall or melanoma-specific survival was confirmed in two subsequent meta-analyses (Lens et al. 2007; Sladden et al. 2009).

A retrospective analysis of 576 patients with primary melanomas 1–2 mm thick who were treated with excision margins of 1 cm (n = 224) or 2 cm (n = 352) also showed an increased rate of local recurrence in the 1 cm margin group (8/224; 3.6%) compared to patients treated with a 2 cm margin (3/352; 0.9%) (Hudson et al. 2013). This study found no difference in the rates of regional/in-transit or distant recurrences in the 1 cm and 2 cm margin groups. As in the Australian study, there was no impact on overall survival .

The Intergroup trial, mentioned above (Balch et al. 2001), included a group of 272 patients with head and neck and distal extremity melanomas who were not randomized and received a 2 cm excision margin; 17 had a local recurrence within 10 years, giving an event rate of 6.3%. This is considerably higher than the local recurrence rates in the randomized groups (2.1% and 2.6%) suggesting that patients with head, neck, and distal extremity melanomas may be more susceptible to recurrence. The authors performed a multifactorial stepwise regression analysis based on the Cox model for the entire group of 742 patients and separately for the randomized and nonrandomized groups (Karakousis et al. 1996). The results showed that poorer survival was associated with the presence of tumor ulceration and head and neck site but was not influenced by the width of the excision margin (Table 4). The presence of tumor ulceration and head and neck site was also significantly associated with more frequent local recurrence . Separate analyses within the randomized and nonrandomized groups were consistent, again showing that tumor ulceration was strongly associated with local recurrence, anatomic site much less so (Table 4).

Table 4 Multivariate analysis (Cox model) or risk for local recurrence (at Any Time) among 740 patients in the Intergroup trial and Intergroup nonrandomized subgroup
Full size table

Current Excision Margin Recommendations for T2 Melanomas

As outlined above, randomized trial data evaluating different excision margins is limited for the specific group of patients with primary tumors 1–2 mm thick, and interpreting retrospective data is problematic because of the inability to adjust for unknown confounding factors. However, available data suggest that a minimum margin of 1 cm does not reduce survival rates in these patients compared with a greater margin, although there may be a slightly increased risk of local recurrence. Most national guidelines recommend either a 1 cm or 1–2 cm excision margin for primary tumors 1–2 mm in thickness (see Table 5) .

Table 5 National guidelines for excision margins (in cm) for T2 cutaneous melanoma (1–2 mm thick)
Full size table

T3: Invasive Melanomas >2–4 mm in Thickness

Randomized Trials of Excision Margins for T3 Melanomas

As with previously discussed tumor thickness trials, no trials exclusively include patients with primary tumors 2–4 mm thick, and outcomes within this tumor thickness subgroup are not reported separately. However, three trials have included 664/936 (71%), 656/900 (73%), and 214/486 (44%) patients with primary melanomas between 2 and 4 mm thick, randomizing them to 2 cm or 4 cm excisions (two trials; (Balch et al. 1993; Gillgren et al. 2011)) and 1 cm or 3 cm excision margins (one trial; (Thomas et al. 2004)). Ten-year data have been reported for these trials (Gillgren et al. 2011). For all three trials, the rates of local, regional, nodal, and distant metastasis were reported, and deaths in two of them were reported, while one – the Intergroup trial – reported deaths only for the complete group with tumor thicknesses of between 1 and 4 mm (Balch et al. 2001). Pooled data from these trials are shown in Fig. 3a–e. The risk of death from any cause in the two trials reporting this outcome was not different between the 1–2 cm excision margin group and the 3–4 cm margin group (risk ratio 1.04, 95% CI 0.94–1.14), and this estimate was identical for the individual trials (Fig. 3a). The risks of local recurrence (risk ratio 1.17, 95%CI 0.49–2.78), regional recurrence (risk ratio 1.30, 95%CI 0.82–2.07), nodal recurrence (risk ratio 0.98, 95%CI 0.79–1.20), and distant metastatic recurrence (0.98 95% CI 0.70–1.36) were all not different for the narrow (1–2 cm) and wide (3–4 cm) excision margin groups across the three trials (Fig. 3b–e). The Intergroup trial reported local recurrence within subgroups of primary tumor thickness (Karakousis et al. 1996); counterintuitively, the risk ratio for 2 cm margins versus 4 cm margins in the group of patients with primary tumors 2–4 mm thick was 0.33 (95%CI 0.07–1.59), implying that the risk of local recurrence was 67% less in the 2 cm margin group compared to the 4 cm margin group, although it could range from 93% less likely to 1.6 times more likely due to the infrequent events, only 8 in 177 patients. Pooled analysis of the local recurrence data for all patients with primary tumors >2 mm thick showed inconsistent findings, with two trials very similar and one quite different. This might be because the deviant trial (the Intergroup study (Balch et al. 1993)) included a majority of thin tumors (<2 mm), whereas the other two (The Swedish Melanoma Study Group (SMSG) II trial (Gillgren et al. 2011) and the UK Melanoma Study Group (UKMSG) trial (Thomas et al. 2004)) had a majority of thicker tumors, or that the definition of local recurrence differed between the studies. Importantly, patients in the UKMSG trial (Thomas et al. 2004) did not have regional node staging by sentinel node biopsy or elective node dissection, only 81 patients (9%) of those in the SMSGII trial (Gillgren et al. 2011) had sentinel node biopsy (51 positive in the 2 cm groups and 31 positive in the 4 cm group) and the third trial (the Intergroup study (Balch et al. 1993)) included a subset of patients who had an elective node dissection (130 in the 4 cm margin group and 133 in the 2 cm margin group), but no details of the number who were positive were reported. Because nodal disease was not determined for all patients in these trials, it is possible there was a chance imbalance in this feature between the treatment groups which may have contributed to variability in the findings.

Fig. 3
figure 3

(a) Risk of death from any cause in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins for excision. (b) Risk of local recurrence in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins for excision. (c) Risk of regional recurrence in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins for excision. (d) Risk of nodal recurrence in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins for excision. (e) Risk of distant metastases in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins for excision

Full size image

Analyses of predictors of recurrence and survival within these trials identified increased tumor thickness, presence of ulceration, site, and male gender as being associated with an increased likelihood of recurrence and death (Hayes et al. 2016; Karakousis et al. 1996; Gillgren et al. 2011; Thomas et al. 2004). Thus, features of the primary tumor, i.e., thickness, ulceration, and site, appear to be better indicators of the likelihood of spread of disease and death than excision margin.

Wider excision margins were associated with increased use of skin grafts in the SMSG II trial (Gillgren et al. 2011) (77/465 in the 2 cm margin group and 250/471 in the 4 cm group), giving an almost 70% lower chance of a skin graft in the 2 cm group compared with the 4 cm group (risk ratio 0.31, 95%CI 0.25–0.39). The UKMSG trial (Thomas et al. 2004) reported many more inpatient procedures with general anesthesia in the 3 cm margin group (297/447) compared to the 1 cm margin group (145/453), giving a risk ratio of 0.48 (95%CI 0.41–0.56), meaning that the 1 cm margin group was about half as likely as the 3 cm group to have an inpatient procedure with general anesthesia (Thomas et al. 2004). In the UKMSG trial, the risk of wound dehiscence was not significantly different between the 1 cm and 3 cm margin groups (risk ratio 0.70, 95%CI 0.29–2.04), but surgical complications were approximately half as likely in the 1 cm margin group (risk ratio 0.53, 95% CI 0.36–0.78) (Hayes et al. 2016).

Add-on studies to these trials have assessed the effects on quality of life and provided psychosocial measurements for the different excision margin groups (Newton-Bishop et al. 2004; Bergenmar et al. 2010). Both studies showed that the impact on quality of life of different excision margins was limited and that differences between the groups resolved over time.

Nonrandomized Studies of Excision Margins for T3 Melanomas (>2–4 mm)

One retrospective study included 324 patients with tumors >2 mm, but with no further description of the number within subgroups of tumor thickness. In this study, 228 patients had 1 cm excision margins, while 97 had 2 cm excisions. There were no significant differences in the rates of local, locoregional, or distant recurrences or melanoma deaths between the two groups (Hunger et al. 2015) although all event rates were slightly higher in the 2 cm margin group.

Another retrospective analysis of 1587 patients with melanomas 2–4 mm thick, of whom 855 had outcome data, showed that a surgical margin of <1 cm (histopathologic margin of <8 mm) was associated with more local recurrence and shorter survival than excisions ≥1 cm (Lamboo et al. 2014). These data suggest that a margin of <1 cm is suboptimal for these patients.

Current Excision Margin Recommendations for T3 Melanomas

Randomized trial data and retrospective studies are consistent in showing no convincing evidence of a benefit for excision margins >2 cm for primary tumors 2–4 mm thick. National guidelines from around the world are variable for this group; those from Australia suggest 1–2 cm excision margins, while Canada, the USA, the Netherlands, France, and Germany suggest 2 cm and the UK 2–3 cm (Table 6). A caveat in the Australian guidelines is a recognition that in certain areas of the body (e.g., face) and in frail or elderly patients, excision margins greater than 1 cm may not be appropriate.

Table 6 National guidelines for excision margins (in cm) for T3 cutaneous melanoma (2–4 mm thick)
Full size table

T4: Melanomas >4 mm in Thickness

Randomized Trials of Excision Margins for T4 Melanomas (>4 mm)

Two trials have included patients with primary melanomas >4 mm in thickness (often referred to as “thick” melanomas). In these trials, 270/936 (29%) and 242/900 (27%) patients were randomized to 1 cm versus 3 cm excisions (UKMSG (Thomas et al. 2004)) or 2 cm versus 4 cm excisions (SMSGII (Gillgren et al. 2011)). Neither trial reported outcomes separately for the group of patients with primary tumors >4 mm. Pooled results from these two trials show no significant differences in the risk of death for the narrower margins (risk ratio 1.04, 95%CI 0.94–1.14) (Fig. 3), local recurrence (risk ratio 1.58, 95%CI 0.81–3.08), in-transit or regional recurrence (risk ratio 1.32, 95%CI 0.77–2.28), nodal recurrence (risk ratio 1.01, 95%CI 0.80–1.27), or distant metastasis (risk ratio 0.95, 95%CI 0.54–1.61) (see Fig. 4a–d).

Fig. 4
figure 4

(a) Risk of local recurrence in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins of excision. (b) Risk of in-transit or regional recurrence in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins of excision. (c) Risk of nodal recurrence in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins of excision. (d) Risk of distant metastases in trials of patients with primary melanoma tumors of >2 mm thickness and randomized to narrow (1–2 cm) or wide (3–4 cm) margins of excision

Full size image

Nonrandomized Studies of Excision Margins for T4 Melanomas

A retrospective analysis of 632 patients with primary melanomas >4 mm in thickness treated in Australia between 1992 and 2009 with an average excision margin of 1.5 cm showed a 9% lower risk of local recurrence for every additional 1 mm of excision margin (Pasquali et al. 2013). There was no impact of excision margin on survival, but the authors acknowledged the lack of power for this analysis.

An earlier study of 278 patients with melanomas >4 mm thick showed that neither local recurrence nor survival was associated with excision margins of <2 cm or >2 cm (Heaton et al. 1998). A similar study of 108 patients with primary tumors >4 mm also showed no reduction in local recurrence or difference in survival with excision margins of ≥2 mm (Ruskin et al. 2016).

The inference, albeit limited by the retrospective nature of these two 10-year studies, is that 2 cm margins can safely be used for melanomas >4 mm thick, thereby avoiding the potential morbidity, cost, and cosmetic disfigurement associated with margins greater than 2 cm.

Current Excision Margin Recommendations for T4 Melanomas

Since randomized trial data and retrospective analyses show no benefit for excision with wide margins of 3, 4, or 5 cm compared with 2 cm for patients with melanomas >4 mm thick, most national guidelines suggest an excision margin of 2 cm for melanomas >4 mm thick (Table 7) .

Table 7 National guidelines for excision margins (in cm) for T4 cutaneous melanoma (>4 mm thick)
Full size table

Excision Margins Summary

The results of randomized and nonrandomized studies detailed above provide the basis for the current recommendations about the width of surgical margins, based on primary melanoma thickness. Simply stated, for melanomas <2 mm and >2 mm, 1 cm and 2 cm margins, respectively, appear to be adequate, with still some lingering uncertainty about the appropriate margin in the 1–2 mm subgroup. No data exist that margins wider than 2 cm (3 cm, 4 cm, or 5 cm) result in any superior disease-specific outcome, but these wider margins are associated with increased surgical morbidity. However, excision margins <1 cm may be too narrow, particularly for the T2 and T3 subgroups, and may contribute to unfavorable outcomes in the form of increased local and regional recurrences. The surgical trial outcomes remind us that we should not be cavalier about excision margins, especially since the defects created by the recommended 1 cm or 2 cm margins, except in a few anatomically challenging areas, can be closed primarily and result in little early or late surgical morbidity (see Table 8).

Table 8 Margins of excision – summary of recommendations according to primary melanoma thickness range
Full size table

Techniques for Routine Wound Closure

Most patients with early-stage cutaneous melanoma can be effectively treated with a simple elliptical excision of the primary lesion or biopsy site followed by primary closure. Procedures that involve excision margins of 2 cm or less can usually be undertaken on an outpatient basis with local anesthesia alone or in conjunction with intravenous sedation. From a technical point of view, the long axis of the elliptical excision should be oriented to make maximal use of the available surrounding skin to achieve a dependable, full-thickness primary closure. Orientation toward the appropriate draining node basin is considered desirable by some surgeons, but there is no convincing evidence that this is of value, and orienting the long axis of the excision in the line of the natural skin creases will produce a better cosmetic result and reduce the risk of wound complications . A variety of options for designing the ellipse can be used to facilitate closure (Fig. 5). The planned excision is influenced by the anatomic site and the surgeon’s assessment of the elasticity and mobility of the local soft tissues. The elliptical excision may not necessarily be directly linear but may be in a “lazy S”-type pattern.

Fig. 5
figure 5

Examples of excision margins. (a) A 1 cm excision oriented in a linear fashion was planned for a proximal extremity lesion. (b) Various options for excision of a 2 cm lesion on the back to allow primary closure. (c) The depth of excision extends to, but does not include, the muscular fascia

Full size image

The width of the oncologically appropriate excision is first circumferentially measured with a ruler from the edge of an intact tumor or previous biopsy site. An elliptical excision, with the long axis of the excision three or four times the width, is usually the technique of choice to facilitate mobilization of full-thickness flaps of skin and subcutaneous tissue for primary closure (Fig. 6). The excision should include removal of the skin and underlying subcutaneous tissue, usually down to the muscular fascia (Fig. 5c); there is no evidence that removal of the muscular fascia provides any improvement or impairment of local control or survival (Olsen 1964; Kenady et al. 1982; Hunger et al. 2014). Indeed, one study suggested harm through increased local recurrence when the deep fascia is removed (Grotz et al. 2013). In some body sites, e.g., the buttock or the thigh or abdominal wall of an overweight patient, the underlying muscular fascia is many centimeters deep to the skin surface. In these circumstances, the excision depth can reasonably be limited to a depth corresponding to the lateral excision margin (e.g., 1 cm depth when a 1 cm lateral excision margin is performed, 2 cm depth for a 2 cm lateral excision margin). The specimen should be marked for orientation, to allow accurate histologic margin assessment, before being removed from the patient.

Fig. 6
figure 6

Schematic showing the surgical margin 2 cm from the biopsy site or the lateral margin of the intact melanoma. The long axis of the incision should be three to four times its width. After the melanoma is excised, skin flaps are raised in a plane above the deep fascia for a sufficient distance to enable the closure of the skin edges without undue tension. The most extensive area of mobilization is near the center of the flaps, and it is often necessary to mobilize the skin flaps for a distance twice that of the excised margin. The flaps are usually joined in a two-step procedure; the subcutaneous tissues and deep dermis are joined first, and then the skin is approximated by either a standard suture technique or a subcuticular closure

Full size image

Closure of a wide excision defect usually requires the creation of simple advancement flaps. This is done by detaching the skin and subcutaneous tissue from the underlying fascial connections in a plane just deep to the fatty layer of the skin. The flaps should be tailored so that the most central part of the incision is mobilized to the greatest degree because this is the area where the tension will be greatest. The opposing flaps can then be stretched over the defect and the tissues approximated using absorbable interrupted or continuous deep sutures that include the deep dermis (Fig. 7). The skin can then be closed with a running subcuticular closure, with a stapling device, or with interrupted or continuous nylon sutures depending on the closure tension, the anticipated range of motion, and the wound stress predicted for specific anatomic locations. Depending on the size and location of the soft tissue defect, a closed suction drain may be considered necessary to minimize the risk of a seroma (Fig. 7b).

Fig. 7
figure 7

Completed full-thickness primary closure (a) without and (b) with a suction drain

Full size image

The surgical morbidity , cosmetic disfigurement, and financial cost associated with skin grafts are greater than for primary wound closure. Thus, primary wound closure with local advancement flaps is almost always preferable. However, in some anatomic locations, primary closure may not be possible, such as on the distal extremities or scalp. In such circumstance, closure of the defect is often accomplished using either a split-thickness or full-thickness skin graft or a rotation flap closure. In esthetically sensitive locations, such as on the face, rotation flaps offer a cosmetically more satisfactory alternative (Fig. 8a–e). Patient dissatisfaction with the appearance of the healed surgical wound is most often associated with the degree of soft tissue indentation or depression rather than the length of the incision (Cassileth et al. 1983). Avoiding skin grafts when possible, providing the patient with accurate information about the expected postoperative physical appearance, and a candid but empathetic discussion of the potential risks of inadequately treated melanoma all give the patient an accurate and appropriate perspective to alleviate the negative psychological impact of surgical treatment of the primary lesion by wide excision.

Fig. 8
figure 8

(a) Patient with a relatively large lentigo maligna melanoma on the face. (b) Planned excision and rotation flap reconstruction. (c–e) Resultant surgical defect and raised flap, closure of the defect, and long-term cosmetic result, respectively

Full size image

Excisions for Melanomas in Unusual or Restrictive Locations

The surgical techniques described above can be employed for melanomas in most anatomic sites, such as the trunk, the proximal extremities, and a large percentage of melanomas on the distal extremities. However, surgical approaches may need to be individualized and modified at other anatomic sites, such as in the head and neck region and on the distal extremities (see chapter “Reconstructive Options Following Surgery of Primary Melanoma”).