Local Recurrence and Survival in Patients With Melanoma In Situ

This cohort study evaluated the rate of local recurrence, metastasis, and melanoma-specific death in patients with melanoma in situ treated with excisional biopsy or wide excision alone.

Key Points

Question

What is the frequency of local recurrence in non-lentigo maligna /non-acral lentiginous melanoma in situ (non-LM/non-ALM MIS)?

Findings

In this cohort study of 401 patients with 403 non-LM/non-ALM MIS and follow-up extending up to 26 years (median, 5.2 years), no local recurrences, metastasis, or melanoma-specific death were observed among patients with histologically clear margins treated with initial excisional biopsy alone or via wide excision.

Meaning

For non-LM/non-ALM MIS, clear histopathological margins either at excisional biopsy or after wide excision may be sufficient, although further studies comparing this approach with current standard management are needed.

Abstract

Importance

The overdiagnosis of melanoma in situ (MIS) is well documented. There is limited evidence on the rate of local recurrence of the non-lentigo maligna (non-LM)/non-acral lentiginous melanoma (non-ALM) subtypes.

Objective

To investigate local recurrence and prognosis in non-LM/non-ALM MIS, the histopathological clearance of the excisional biopsy margins, and the association with the size of wide excision margins.

Design, Setting, and Participants

This retrospective cohort study included patients with non-LM/non-ALM MIS diagnosed from 1991 to 2023 who were followed up for at least 1 year at the Skin Cancer and Melanoma Unit of Andreas Sygros University Hospital in Athens, Greece. Patients with a history of invasive melanoma or a histopathological diagnosis of LM or ALM in situ were excluded. Median (IQR) follow-up was 5.2 (2.9-7.9) years. Deidentified data on patient demographics and clinical characteristics, including the histopathological clearance of margins of the initial excisional biopsy and the size of margins of the wide excision, were obtained from medical records.

Main Outcomes and Measures

The primary outcomes were local recurrence, metastasis, and melanoma-specific survival.

Results

A total of 401 patients (214 [53.4%] women) with 403 non-LM/non-ALM MIS and a median (IQR) age of 52 (40-62) years were included in the analysis. MIS was frequently located on the trunk (201 lesions [49.9%]), followed by the lower extremity (99 [24.6%]), the upper extremity (71 [17.6%]), and the head and neck (32 [7.9%]). All lesions were initially treated with excisional biopsy, followed by wide excision for 372 (92.3%). There was only 1 local recurrence in a patient with involved margins at the excisional biopsy who did not undergo wide excision and developed an invasive melanoma 14 months later. Thirty lesions in 30 patients had clear excisional biopsy margins with no wide excision and had no recurrence at a median (IQR) follow-up of 8.1 (4.1-12.9) years. In 23 patients with 23 lesions that had wide excision with narrower than the standard 0.5-cm margins (mean [SD] margin size, 0.36 [0.07] cm), no recurrences were found at a median (IQR) follow-up of 4.3 (2.7-6.2) years. During follow-up, 6 patients (1.5%) developed a lesion suspicious for recurrence near the excision scar, which was excised and showed nevus or solar lentigo on histopathology. No patients had metastasis or melanoma-specific death.

Conclusions and Relevance

This cohort study found that diagnostic excisional biopsies with clear margins may be sufficient for treating MIS; however, larger studies are necessary.

Introduction

Cutaneous melanoma in situ (MIS) (stage 0) is currently included in European and US melanoma treatment guidelines, which recommend a wide excision following the initial excisional or shave biopsy regardless of the initial histopathological margin clearance achieved.^1,2 MIS incidence is increasing worldwide.^3,4 Consequently, there is a substantial number of patients expected to be diagnosed with MIS who will be treated with wide excision as part of standard care.

Nevertheless, patients diagnosed with MIS have been found to live longer than the general population.^5,6 The phenomenon of increasing diagnoses of indolent in situ melanomas that do not affect patient mortality is termed overdiagnosis.⁷ In Australia, the overdiagnosis of MIS is estimated to cost their health care system approximately A$18 million annually and is considered the primary contributor to melanoma overdiagnosis.⁸

Local recurrence of MIS is a prognostic event of importance for patients; a diagnosis of MIS carries a similar fear of recurrence as a diagnosis of invasive melanoma.^9,10 Because evidence remains limited about the rates of local recurrence in MIS of the non-lentigo maligna (non-LM) and non-acral lentiginous (non-ALM) subtypes after wide excision,¹¹ we conducted a retrospective study of patients with non-LM/non-ALM MIS with extended follow-up, aiming to evaluate local recurrence, metastasis, and melanoma-specific survival and to assess the histopathological margin clearance of the excisional biopsy and association with wide excision margin size.

Methods

We performed a retrospective cohort study of adult patients with a first diagnosis of MIS from 1991 to 2023 at the Skin Cancer and Melanoma Unit of Andreas Sygros University Hospital, in Athens, Greece. Initially, an excisional biopsy was performed with a 2-mm clinical safety margin, followed by a wide excision, according to guidelines.¹² Patients with MIS were followed up at least annually with total body skin examination for local recurrence and new primary melanomas. Patients with a history of invasive melanoma were excluded, as it would be difficult to attribute the origin of a possible metastasis. Also, patients with a histopathological diagnosis of LM or ALM in situ were excluded. The histopathologic diagnosis was made by expert dermatopathologists at the study site. In cases of diagnostic uncertainty, a second or third external expert review was conducted. The study was approved by the Institutional Review Board at Andreas Sygros Hospital, which waived the patient informed consent requirement due to the noninterventional retrospective nature of the study and the use of deidentified data. The study is reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.¹³

Variables of Interest

The medical records of eligible patients were reviewed, and deidentified data were collected on patients’ demographic and clinical characteristics, including age, sex, medical history and medication history. The primary variables of interest were local recurrence, metastasis, and death (from any cause or melanoma). Local recurrence was defined as recurrence in the excisional scar.² To avoid missing a possible local satellite or in-transit recurrence, we scrutinized the medical records for any report of a new lesion near the site of a previously excised MIS. If new lesions were suspected near the scar, the clinical dermoscopic photos and histology reports of the lesion’s excision were examined. Histopathological variables included the maximum diameter of the MIS, histopathological margin status at excisional biopsy (clear or involved), the margin size of the wide excision (eFigure in Supplement 1), and the histopathological margin status of the wide excision.

Statistical Analysis

Continuous variables are reported as mean (SD) or median (IQR), while categorical variables are presented as frequencies and percentages. The Mann-Whitney test was used for the comparison of medians. Follow-up time was calculated from the date of excisional biopsy until the last follow-up. Patients and lesions with at least 1 year of follow-up (n = 403) were considered eligible for the main analysis. A subgroup analysis was performed for patientswith less than 1 year of follow-up (n = 44) to explore possible early recurrences. Statistical tests were 2-sided, and a P < .05 was considered statistically significant. Stata, version 12 (StataCorp) was used to analyze the data.

Results

The sample included a total of 401 patients (214 [53.4%] women; 187 [46.6%] men) diagnosed with 403 non-LM/non-ALM MIS who had a minimum follow-up of 1 year. Two patients presented with 2 synchronous MIS lesions each. The median (IQR) patient age was 52 (40-62) years. Most patients (395 [98.5%]) were immunocompetent. The median (IQR) maximum diameter of MIS on the histopathological report was 0.7 (0.5-1.1) cm. MIS were frequently located on the trunk in (201 lesions [49.9%]), followed by the lower extremity (99 [24.6%]), the upper extremity (71 [17.6%]), and the head and neck (32 [7.9%]). All lesions had an initial excisional biopsy, with clear histopathological margins in 381 (97.9%). Median (IQR) follow-up was 5.2 (2.9-7.9) years), and the range of follow-up was 1.01-25.8 years. (Table)

Table. Characteristics of 403 MIS Lesions in 401 Patients.

Characteristic	Total	Without wide excision	Wide excision with <0.5-cm margins
Patients
No.	401	31	23
Age, median (IQR), y	52 (40-62)	43 (35-61)	50 (40-61)
Sex, No. (%)
Female	214 (53.4)	20 (64.5)	14 (60.9)
Male	187 (46.6)	11 (35.5)	9 (39.1)
Immunosuppression, No. (%)
Yes	6 (1.5)	0	1
No	395 (98.5)	31	22
MIS Lesions
No.	403	31	23
Body location
Head/neck	32 (7.9)	5 (16)	6 (26.1)
Trunk	201 (49.9)	9 (29)	4 (17.4)
Upper extremity	71 (17.6)	4 (13)	7 (30.4)
Lower extremity	99 (24.6)	13 (42)	6 (26.1)
Initial excisional biopsy histopathological margin, No. (%)a
Clear	381 (97.9)	30	23
Involved with MIS	6 (1.6)b	1	0
Involved with nevus	2 (0.5)	0	0
Wide excision performed, No. (%)
Yes	372 (92.3)	0	23
No	31 (7.7)	31	0
Local recurrence, No.	1	1	0
Metastasis, No.	0	0	0
Melanoma-specific death, No.	0	0	0
Follow-up, y
Median (IQR)	5.2 (2.9-7.9)	8.1 (4.1-12.9)c	4.3 (2.7-6.2)
Range	1.0-25.8	NA	NA

Abbreviations: MIS, melanoma in situ; NA, not applicable.

^a Data on initial histopathological margin clearance were missing for 14 lesions.

^b Of the 6 lesions with involved histopathological margins with MIS, one also had a clinically evident (macroscopic) residual lesion.

^c Follow-up for the 30 patients who did not undergo wide excision and did not have recurrence.

Wide excision was performed in 372 lesions (92.3%). The remaining 31 lesions did not have wide excision after the excisional biopsy of MIS, with clear margins (n = 26), missing margin status (n = 4), or involved histopathological margins (n = 1). The patient who had involved histopathological margins was diagnosed with MIS on the lower leg, and 1 peripheral histopathological margin was involved with MIS in the excisional biopsy. The patient did not undergo wide excision and had a local recurrence diagnosed as invasive melanoma (Breslow 0.8 mm) at the first follow-up visit at 14 months. The remaining 30 of 31 lesions (96.8%) that did not have wide excision had no recurrence at a median (IQR) follow-up of 8.1 (4.1-12.9) years. During follow-up of all 401 patients, 6 patients (1.5%; 5 had wide excision and 1 had only the initial excisional biopsy with clear margins) developed a suspicious lesion near the scar that was excised and histopathology showed either a nevus, bleeding, or solar lentigo.

Twenty-three MIS lesions (5.7%) in 23 patients were treated with wide excision with margins narrower than 0.5 cm (mean [SD] margin size, 0.36 [0.07] cm); histopathological margins were clear. There were no observed recurrences after a median (IQR) follow-up of 4.3 (2.7-6.2) years. The median (IQR) diameter of these 23 MIS lesions was significantly smaller compared with MIS treated with wide excision with at least 0.5-cm margins (0.6 [0.4-0.8] cm vs 0.8 [0.5-1.1] cm, P = .03); narrower margins were used for smaller MIS lesions. The MIS lesions that were excised with narrower wide excisional margins had a relatively balanced distribution in terms of location: head and neck (n = 6), trunk (n = 4), upper extremity (n = 7), or lower extremity (n = 6) (Table).

There were no cases of in-transit, nodal, or distant metastasis in our study. At the last follow-up, 399 patients were alive; there were no melanoma-specific deaths, and 2 patients died from unrelated causes. A subgroup analysis of the 44 patients with less than 1 year of follow-up, who were not included in the main analysis, showed no early recurrence.

Discussion

Herein, we aim to build on the limited existing literature on the recurrence of non-LM/non-ALM MIS, with 3 main findings. First, no recurrences were observed in MIS lesions after wide excision with clear histopathological margins. Sun et al¹⁴ reported on 351 MIS (31% LM) treated with wide excision and a 5-mm safety margin after excisional or shave biopsy. Three local recurrences (0.87%) were noted after a median 7-year follow-up. It is possible that lentigo maligna was the primary recurrent MIS subtype in this study; however, this information was not available. In a meta-analysis of previous studies of MIS of the head and neck, the local recurrence rate was 3.3% in 141 cases treated with wide excision, 1.3% with staged excision, and 0.7% with Mohs micrographic surgery (MMS).¹⁵ However, 8 of the 11 studies on wide excision included fewer than 10 cases of MIS each, while the studies on MMS or staged excision either included mostly LM, or the MIS subtypes were not reported. Another meta-analysis on MIS treated with MMS reported a local recurrence rate of 0.3%, but the MIS subtype was not described.¹⁶

Second, we found no recurrence in 53 patients who either did not have wide excision after the initial excisional biopsy with clear histopathological margins or had wide excision with narrower than standard-size margins. There was 1 recurrence observed in a patient who had involved histopathological margins and declined wide excision, highlighting the importance of achieving clear histopathological margins to avoid missing a possible invasive melanoma component.¹⁷ We are aware of only 1 other study by Smith et al¹⁸ that reported on follow-up in a small number of non-lentiginous MIS with clear histopathological margins, some of which did not have subsequent wide excision, and had no recurrence. Smith et al¹⁸ studied 234 non-lentiginous MIS with clear histopathological margins after excisional biopsy, of which 46 (19.7%) did not have wide excision. However, a relatively short mean follow-up of 29.7 months was documented in 56.6% of the total 234 cases and it was not reported whether those that did not have wide excision were included in the follow-up.¹⁸

Third, we focused on non-LM/non-ALM MIS to provide a risk-stratified histologic subtype-specific approach. Most previous studies reported recurrence in lentigo maligna or for a combination of MIS subtypes but there is limited evidence for non-LM/non-ALM MIS.^{11,14,19,20,21} Considering the subtypes of MIS separately may contribute to a more refined patient risk stratification, eg, based on histology (non-lentiginous MIS vs LM or vs ALM in situ).^5,22 In addition, non-LM MIS is an entity posing diagnostic challenges to differentiate from dysplastic nevi in histopathology. Misclassification may arise from upgrading the histopathological diagnosis of dysplastic nevus to MIS.⁷ This diagnostic drift may arise from pressure of medical liability and the variability and challenges in histopathological diagnosis of dysplastic nevus vs MIS.^23,24

The recent Melanocytic Pathology Assessment Tool and Hierarchy for Diagnosis classification categorizes severely dysplastic nevi together with MIS as high-grade lesions with a low risk of progression to invasive melanoma, suggesting their histopathological and biological equivalence.²⁵ We believe it would be unwise to dismiss the accumulating evidence that MIS are indolent lesions and not obligate precursors of invasive melanoma.²⁶ In the US Nurses’ Health Study, invasive melanomas were diagnosed at a younger age compared with MIS, and this finding was replicated in studies using Surveillance, Epidemiology, and End Results 9 as well as data from cancer registries in Australia and Scotland.^3,26 This finding is in contrast to what would be expected for a progression model from MIS to invasive melanoma later in time. Melanocytic nevi and melanomas in situ have been described as precursors in this stepwise progression model of invasive melanoma development^{27,28,29,30,31,32,33}; however, this stepwise progression model seems to be a rare process.^26,34,35,36

Strengths and Limitations

Strengths of our study are the rigorous ascertainment of local recurrence, metastasis, and survival, the length of follow-up, and reporting results for 3 treatment groups, including wide excision with standard size margins, wide excision with narrower than standard margins, and not having wide excision after the initial excisional biopsy. Study limitations include the single-center and retrospective design. Additionally, the wide excision safety margins were not measured clinically but were inferred based on the size of the wide excision sample from the macroscopic description of the histopathology report. Tumor shrinkage of up to 15% has been reported after formalin tissue fixation,^37,38 which could result in a small underestimation of up to 0.75 mm for a margin size of 5 mm. A small number of lesions did not have standard wide excision. In addition, there was a nonstandardized histopathological diagnosis across the 32 years of data collection of our study.

Conclusions

In this retrospective cohort study, we found no local recurrences, metastases, or melanoma-specific deaths in 401 patients with non-LM/non-ALM MIS. These findings provide support that non-LM/non-ALM MIS excised with clear histopathological margins in the excisional biopsy may not require additional treatment. However, our findings should be confirmed with prospective, randomized trials and large multicenter retrospective studies that compare standard management with initial excisional biopsy with clear margins only.

Supplement 1.

eFigure. Safety Margins for the Wide Excision of Melanoma In Situ

Supplement 2.

Data Sharing Statement