‘Primum non nocere’: how harmless is routine wide local excision for AJCC stage IA melanoma?

Abstract

Background

Management of primary cutaneous malignant melanoma is with initial excision biopsy followed by a wide local excision to achieve locoregional control. For low-risk thin melanomas, the added survival benefit from the wide local excision is minimal. In this study, we investigated the morbidities of wide local excision and evaluated the current clinical practice in managing stage IA malignant melanoma.

Methods

Patients with confirmed stage IA malignant melanoma who had undergone a wide local excision in the 2013/14 period were identified using a specialist cancer multidisciplinary team-held database. Primary pathology, surgical data and follow-up documentation were analysed.

Results

A total of 231 cases were identified; 95% of patients (n = 220) had malignant melanoma excised completely at first excision biopsy, mean margin 2.8mm (range 0.5–8.0mm). Postoperative morbidities occurred in 25% of patients (n = 57), including 6.6% wound problems, 10.9% scarring problems, 10.0% psychological stress and 0.4% cosmetic concern. Wide local excision reconstructions were performed with primary closure in 82% of patients, split skin grafts in 4%, full-thickness skin grafts in 3% and flaps in 10%. Of the total, 44% of patients (n = 101) had further excisions and 17 received new low-risk melanoma diagnoses.

Conclusions

We demonstrated that 1cm wide local excision is associated with significant morbidity, which can affect patients’ physical, psychological and social wellbeing. Since wide local excision does not give a survival advantage, and its margin is already frequently reduced in cosmetically sensitive areas, the need for a second full 1cm wide local excision procedure for thin melanoma should be re-evaluated, especially when 95% of our study cohort had their malignant melanoma completely excised by the initial biopsy alone.

Introduction

There has been a global increase in the incidence of cutaneous melanoma, with a fourfold increase in the UK over the past 20 years.¹ Despite this, mortality rates have been observed to be levelling off or even decreasing in European countries.² This has been attributed to both earlier diagnosis,³ as indicated by studies which showed a decrease in Breslow thickness of cutaneous melanoma at the time of initial diagnosis,³ and perhaps overcautious diagnoses.⁴

In our supraregional unit, malignant melanomas with Breslow thickness 1mm or less comprise over 75% of our caseload with, before 2018, 60% of these being stage IA,⁵ following the American Joint Committee on Cancer (AJCC) 2009 criteria.⁶ The current National Institute for Health and Care Excellence (NICE) guidelines advise that initial biopsy of a suspected lesion should be with a clinical margin of 2mm of normal skin.⁷ Upon histopathological review, confirmed malignant melanoma with a Breslow thickness less than 1mm are to be treated with wide local excision to complete a 10mm margin.⁷ As a result, patients who are confirmed as having the lowest risk AJCC stage IA melanoma have to undergo two procedures: the first biopsy to identify but almost always to completely excise the suspected lesion, and the second to remove a further margin of clinically uninvolved skin. It is accepted that wide local excision does not significantly improve survival but is considered necessary to minimise the unpleasant and distressing effects of local recurrence, where that might otherwise have occurred.

When considering stage IA malignant melanoma, it is accepted that this has the least metastatic potential, with even pre-AJCC 2017 studies of Breslow thickness suggesting local recurrence rates of less than 1.5%.⁸ This means that there are increasing numbers of patients who would not have metastasised, even without wide local excision, who are automatically receiving this treatment. Despite continuing research,⁹ there are still no better prognostic markers to differentiate the higher risk stage IA melanomas from the rest of the stage IA cohort. Thus, wide local excision continues to be the standard treatment in stage IA melanoma because it is perceived as a safe and minimally harmful procedure.¹⁰

However, the senior author (SR-E) became increasingly concerned, from discussion with patients, that wide local excision does in fact carry frequent postoperative morbidity, such as wound problems including infection, poor healing and scarring, as well as the psychological stress and anxiety caused not only by the ‘cancer’ diagnosis, but also the surgical sequelae. Furthermore, the continuing global application of this potentially unnecessary procedure in so many has an implication for resource allocation, as well as creating an increasing ethical dilemma.¹¹

In this study, we investigated the morbidities associated with wide local excision in patients with stage IA melanomas and discuss whether the morbidities are justified by the anticipated benefits of the procedure.

Materials and methods

Design and settings

A prospective database of referrals to the melanoma screening clinic in the 2013/14 period was retrospectively reviewed. All patients with a primary histopathological diagnosis of AJCC (2009) stage IA malignant melanoma were identified. We note that, at the time of submission, AJCC has been updated to 2017. From the same database, every fifth patient treated chronologically who had a negative result for melanoma from their excision biopsy was selected as a control group.

Cases were reviewed using histopathology reports, discharge summaries and specialist skin cancer multidisciplinary team letters. Patient demographics, tumour-related data including site, size and the minimum margin on excision were collected. Outcome measures were any documented wound problems such as infection, healing issues, scarring and anxiety. Numbers of further excisions not managed at the first appointment were also recorded, including subsequent confirmed malignant melanoma during the follow-up period.

Patients who did not receive wide local excision as treatment were excluded. All data were recorded on a Microsoft Excel spreadsheet and data analysis was performed with MS Excel.

Follow-up and end points

All patients were followed up for at least 12 months, as per NICE guidelines for management for stage IA malignant melanoma. All data in the 12-month period were included.

Postoperative morbidities

Patient-reported morbidities were categorised into four groups: wound problems such as infections and haematoma; scarring problems such as dog-ear or hypertrophic scars; psychological stress and anxiety (symptoms according to General Anxiety Disorder Assessment Scale); and cosmetic concern. When patients reported more than one postoperative morbidity, all complaints were recorded.

Results

Patient and tumour characteristics

A total of 230 patients with 232 new primary diagnoses of stage IA melanoma were identified. One patient was excluded from the study because he was on palliative care for another malignancy and did not receive wide local excision.

The median age of our study group was 57 years (n = 229, range 18–92 years, standard deviation, SD, 16.5 years), male to female ratio was 1:1.3. Of the 231 cases analysed, 220 (95%) had the primary lesion completely removed at the first biopsy. The mean histological margin of excision in these was 2.8mm (range 0.5–8.0mm). Some 18 patients had a 5mm or greater histological margin. All cases received a wide local excision following the NICE guidelines, performed by consultant surgeons or specialist registrars. Primary closure was performed in 189 patients (82%), split skin grafts in 10 (4%), full-thickness skin grafts in 8 (3%) and flaps in 24 (10%). All patients were followed up for at least 12 months according to NICE guidelines. Eight patients were followed for longer, between 15 and 24 months, but the postoperative morbidities reported outside the standard 12-month follow-up period were not included in this study.

The control group consisted of 255 patients. Median age was 46 years (range 15–91 years, SD 18.71 years), male to female ratio was 1:1.5. The control group was significantly younger than the study group (p < 0.001).

Postoperative morbidities

In the melanoma positive cohort, 57 (25%) patients had 64 morbidities recorded post-excision and wide local excision (Table 2), with seven patients (3%) reporting more than one morbidity. In this subgroup, the male to female ratio was 1 : 3 and the median age was 47 years (range 18–80 years, SD 14.3 years). There were no complications reported in any of the non-melanoma control group.

Table 2.

Melanoma diagnoses in further excisions

New diagnoses	Patients (n)
	(Study group)	(Control group)
One pTx	1
One pTis	2
One pT1a	7	2
One pT1a + one pTx	1
One pT1a + one pTis	1
Two pT1a	2	1
Three pT1a	1
Six pT1a	1
Two pT1a + one pT1b	1

There were 15 (6.6%) reports of wound problems, 25 (10.9%) reports of scarring problems, 23 (10.0%) related to anxiety and 1 (0.4%) reported concern regarding wound cosmesis (Table 1).

Table 1.

Postoperative morbidities reported by study cohort to hospital clinicians

Morbidity	Patients
	(n)	(%)
Wound problems:	15	6.6
Infections	4
Extruded or retained sutures	5
Wound breakdown	2
Haematoma	2
Nerve damage	2
Scarring	25	10.9
Hypertrophic scar	13
Stretched scar	6
Dog ear	6
Psychological stress	23	10.0
Due to new lesion	18
Due to initial diagnosis	4
Due to scar	1
Cosmesis	1	0.4

Six patients (2.6% of the total study group) needed further interventions to manage their wide local excision-associated morbidities. Three patients required drainage of a haematoma or seroma developing from their wide local excision. In patients with scarring problems, one needed an additional scar revision and one needed liposuction. One patient who was experiencing anxiety due to their initial diagnosis was referred to clinical psychology for further counselling.

Further excisions

In the following 12 months, further excisions were performed in 101 (44%) of the melanoma positive group. 17 (7.4% of the total study cohort) had a further positive diagnosis of malignant melanoma. These patients received 30 new malignant melanoma diagnoses, with seven having more than one malignant melanoma diagnosed (Table 2). In the control group, 16 (6%) patients had further excisions and three patients (1.2%) received new melanoma diagnoses. Of these 30, there were two ‘fully regressed’ and one stage IB melanoma, with the other 27 being either stage IA or in situ lesions.

Discussion

Completion wide local excision to 1cm Breslow thickness is often considered to be a procedure with low morbidity, and thus an acceptable treatment of increasingly large numbers of patients to benefit the few. Our study found that 25% of patients were troubled enough to have reported various degree of morbidity ranging from physical discomfort to psychological distress to their plastic surgeon. However, we believe that 25% is likely to be an underestimate because only retrospective clinic letters were analysed, and there was an unexplained preponderance of younger females reporting compared with the study cohort average. We expect the rate would have been higher if we had considered the full hospital record including dressings clinic and clinical nurse specialist records, covering perioperative complications as well as minor wound concerns redirected to primary care. It would probably have been higher still if primary care records were available.

Physical morbidities and scarring

In most cases, a simple linear closure along radioactive seed localisations can be achieved without significant cosmetic or functional impairment. However, 18% of patients required reconstruction with skin grafts or flaps. This compares with 13.6% in the MelMarT group.¹²

Even with the majority being closed directly, 21% of our patients reported surgically related adverse events. With reference to the classification of surgical complications by Dindo et al MelMarT 84% were classed as grade I, 8% grade II and 8% grade III.¹³

Many patients complained of itch and discomfort around the excision area, which usually resolves without complications, so were not formally calculated here. Nonetheless, 17.5% of patients reported other physical morbidities related to wound or scars.

Most patients wholly appreciate that any surgery leads to scarring, but scarring problematic enough to be reported, was the most common complaint in our study cohort (10.9%, n = 25). Because scarring is the normal physiological endpoint of wound repair and considered a necessary outcome by clinicians, its effect on patients’ quality of life is often underplayed especially in the cancer context. However, plastic and dermatological surgeons should be more attuned to the impact that scarring has on the five domains of patients’ quality of life: emotional wellbeing, confidence in the nature and management of scars, social functioning, acceptability to self and others, and physical discomfort.¹⁴ Moreover, melanomas commonly occur in the head and neck regions, thus unnecessary scarring by wide local excision in critical anatomical structures on the face can cause more functional and cosmetic deficit.¹⁵ As a result, routine practice follows suggestions that excision margins should be reduced for melanomas on the head and neck area.¹⁶

Psychology of receiving treatment: distress or peace of mind?

Studies suggested that at least 20% of patients with pigmented skin lesions had clinically high levels of anxiety.^17,18 Although not as well studied, as with other patients ‘living with a diagnosis of cancer’ there is presumably a degree of anxiety associated simply with a confirmed diagnosis of malignant melanoma.¹⁹ In our study, 10% of patients reported an increase in anxiety level post-wide local excision in the follow-up period. This is a higher than expected incidence considering they have already received treatment and had the primary melanoma excised.

Previous studies have shown that patients given a ‘cancer’ diagnosis are often more inclined to opt for active treatments for minimal survival benefits, even when the treatments carry significant risks and morbidities.²⁰ It is suggested that patients do this to provide themselves with peace of mind that they are doing something to treat their disease.²¹ Patients who are more likely to overestimate the perceived survival benefits of treatments with significant adverse effects tend to be of a younger age group.²² It is this younger age group that is more commonly affected in melanoma. Considering these findings, it is not surprising that patients are willing to undergo wide local excision when the associated morbidities were already perceived as being insignificant to start with. However, it is important that we as their clinicians have more balanced perceptions.

Balancing risks and benefits

Ever since the 1990s, melanoma clinicians have sought to reduce the unnecessary harm being done to patients from the previously much wider excisions, popularised by Handley in 1907.²³ There has always been concern that ‘the excess morbidity caused by larger excision defects, including increased hospital stay, complications, and need for reconstructive surgery, may not be necessary, particularly because [as] previous randomised controlled trials (RCTs) have shown, local recurrence rates are low, ranging from 1.3% for intermediate-risk primaries to 3.3–4.3% for high-risk primaries’.¹² Yet, as the Cochrane review of surgical margins stated, even though the ‘size and depth of the excision should therefore err on the side of safety first, [the] quality of life after surgery is an important consideration and unnecessary disfigurement should be avoided’.¹⁵

This led to the various trials at the turn of the century,^24,25 which reduced the recommended margins quite dramatically, but which not unreasonably focused on the intermediate and higher risk primaries. Yet since then, there have been only a few published or planned randomised controlled trials looking at even narrower margins, but again these focus on primaries greater than 1mm.^12,15

Where just Breslow thickness has been considered, the local recurrence rate has been estimated at less than 1.5%.^26,27 However, these studies have not differentiated between stage IB lesions with ulceration and stage IA lesions without, let alone the difference between those with mitoses or not.

The evidence indicates the question is whether any disease recurrences were due to incomplete excision or true lymphatic spread. If in melanoma in situ, a 5mm margin is considered sufficient, perhaps a 5mm wide local excision may also be sufficient in these very-low-risk lesions. Indeed, the consensus is still that there has been no statistically proven survival benefit from wide local excision across all the stages.¹⁵ Furthermore, the Cochrane review also accepted that for less than 1mm primaries, ‘there is insufficient RCT data on which to base clinical recommendations’, and that it would be sensible ‘to look for evidence of equivalence or non-inferiority of a narrower excision margin compared with a wider one. In this situation, that is whether ‘a narrower margin [is] not [significantly] worse than a wider one?’.¹⁵ If these official margins could be reduced safely, it could enable the initial excision biopsy, albeit with a 5mm margin, to be sufficient to avoid further unnecessary surgery and reduce the burden on skin oncology services as a whole.

It is interesting to note that the Cochrane reviewers also felt that the benefit from wide local excision, and presumably for all levels of risk, was limited enough to suggest that, ‘a prospective trial for facial melanomas, perhaps comparing 0.5cm and 1.0cm excision margins, would be clinically very useful’.¹⁵

Perhaps, if one were proposing a new treatment today, one would need to balance the published evidence for proposed benefits, and the ‘numbers needed to treat’ against that of any significant risks and potential morbidities. We therefore wonder whether, when it comes to completion wide local excision for the lowest-risk malignant melanoma, that has truly been done. The current evidence seems to suggest that up to 49, and perhaps more, patients are being treated to treat the one who may otherwise have developed local recurrence, which would anyway have been unlikely to have improved their overall survival.²⁸ And the debate concerning overtreatment of the lowest risk cancers has already begun.¹¹

Resource implications

We found that patients with a previous melanoma diagnosis were more likely to receive further excision biopsies than patients with a previous negative diagnosis (44% vs 6%, p < 0.01, one-end t-test). This is not surprising considering that patients with a history of melanoma tend to be more vigilant of new lesions and clinicians often have a lower diagnostic threshold for these patients. Although the rate of positive melanoma diagnoses within these further excisions was in fact, found to be higher in patients with a previous negative diagnosis (25% vs 17%), this difference was not statistically significant. One reason for this apparent difference could be that patients who had previously had a negative biopsy may have been appropriately reassured about identifying new lesions and thus had a higher threshold for seeking medical advice, while the melanoma group were more cautious.

The cost of associated postoperative morbidities in melanoma patients and their implication on scarce resources such as operating time and clinic appointments is currently under study in intermediate-risk melanoma,¹² but would be even more relevant in the more prevalent stage IA cohort. And since the consensus is that there has been no statistically proven survival benefit from wide local excision ‘across all stages’,¹⁵ let alone in Breslow thickness, it could be significantly more cost effective to redirect resources towards identifying and treating patients with higher risk melanoma.

Limitations

The reported morbidities could be associated with factors such as the operating surgeons’ technique, level of postoperative care and patients’ preoperative status. In particular, no standard clinical questionnaires, such as the FACT-M, were used in follow-up clinics, so it was challenging to quantify different anxiety and depression symptoms in the patient complaint. Nonetheless, all psychiatric complaints documented were classed as psychological distress to preserve accuracy.

Conclusion

In our study, we have demonstrated that completion 1cm wide local excision does have a significant rate of morbidity that affect patients’ physical, psychological and social wellbeing. This could also have implications on the associated resource allocations at a national level. While we agree that locoregional control of malignant melanoma should always be the primary objective, it is critical that patients are better informed of the potential adverse outcomes as well as a more accurate presentation of the benefits, in order for them to make decisions on their management options. We suggest that with further control studies comparing the standard margin of 10mm to lesser margins, stage IA malignant melanoma that had been further risk-stratified by dermoscopic assessment could be managed with a single, initially wider primary biopsy and greater reassurance. Thus, allowing selected patients to avoid a second wide local excision surgery while preserving patient outcomes.