Funding sources The clerical support needed for this study was providedby the Australian and New Zealand Melanoma Trials Group.
Radiotherapy for lentigo maligna: a literature review and recommendations for treatment
Version of Record online: 13 JAN 2014
© 2013 British Association of Dermatologists
British Journal of Dermatology
Volume 170, Issue 1, pages 52–58, January 2014
How to Cite
Fogarty, G.B., Hong, A., Scolyer, R.A., Lin, E., Haydu, L., Guitera, P. and Thompson, J. (2014), Radiotherapy for lentigo maligna: a literature review and recommendations for treatment. British Journal of Dermatology, 170: 52–58. doi: 10.1111/bjd.12611
Conflicts of interest None declared.
- Issue online: 13 JAN 2014
- Version of Record online: 13 JAN 2014
- Accepted manuscript online: 6 SEP 2013 02:14AM EST
- Manuscript Accepted: 28 AUG 2013
Lentigo maligna (LM) incidence is increasing. LM frequently involves the face near critical anatomical structures and as a consequence clinical management is challenging. Nonsurgical therapies, including radiotherapy (RT), are increasingly used. Evidenced-based treatment guidelines are lacking. We conducted a review of previously published data analysing RT treatment of LM. A search of PubMed, Embase and Medline databases to June 2012 identified nine clinical studies that examined the use of RT for LM treatment in at least five patients. Nine studies described 537 patients with LM treated with definitive primary RT, between 1941 and 2009, with a median reported follow-up time of 3 years. Eight articles could be reviewed for oncological outcome data. There were 18 recurrences documented in a total of 349 assessable patients (5%). Salvage was successful in the majority of recurrent LM cases by using further RT, surgery or other therapies. Progression to LM melanoma (LMM) occurred in five patients (five out of 349, 1·4%) who all had poor outcomes. There were five marginal recurrences documented out of 123 assessable patients (4%). There were eight in-field recurrences documented with either LM (five) or LMM (three) out of 171 assessable patients (5%). A series of recommendations were then developed for RT parameters for treatment of LM. These parameters include treatment volume, dose, dose per fraction and outcome measures. These may be of use in prospective data collection.
Lentigo maligna (LM), a form of in situ melanoma,[1-3] is characterized in its earliest stages by scattered single atypical melanocytes involving the dermoepidermal junction (Fig. 1a) and extending down skin appendageal structures. It usually occurs within severely sun-damaged skin. As the tumour progresses, melanocyte density and degree of cytological atypia increase (Fig. 1b). Advanced cases of LM often show confluent growth of melanocytes along the basal epidermis, nest formation and pagetoid epidermal invasion, with florid adnexal involvement (Fig. 1c). The transition at the peripheral margin of the lesion from neoplastic to non-neoplastic melanocytes can be poorly defined.[5, 6]
When dermal invasive melanoma develops in an LM, it is termed LM melanoma (LMM). LMM is the third most common of the four main invasive melanoma subtypes and has metastatic potential. The lifetime risk of developing LMM within an LM has been reported to be as low as 5% and as high as 50%, with the time to transformation varying between a few months to more than 30 years. Physician bias is to treat LM to avoid the risk of missing an LMM.[9, 10]
Current melanoma treatment guidelines suggest simple excision of LM with a 5-mm margin beyond the visible lesion, but some consider this to be inadequate.[12-14] Nonsurgical therapies for LM have included among others radiotherapy (RT), topical therapies such as imiquimod, photodynamic therapy and laser. RT is superior to surgery in conserving normal tissue within the treatment field. RT can be used in LM as definitive treatment or as adjuvant treatment on finding positive margins following excision.
Superficial RT gives a full dose to the skin surface; the dose then falls off exponentially. Grenz rays are very superficial RT (typically 10–30 kV), which only penetrate to the epidermal–dermal interface. A much higher dose can be given in a small number of fractions as there is less tissue penetration (Fig. 2). Brachytherapy moulds can also be used. Penetration in tissue from moulds depends on the distance of the source to the skin and this can be varied to achieve the depth required.
RT is fractionated, i.e. the total dose is divided into parts and given over time, often as daily treatments. The main reason for fractionation is to take advantage of the difference in repair capacity between normal cells and tumour cells. Normal cells within the radiation treatment volume are able to repair a small amount of radiation damage but tumour cells cannot, and die. Fractionation selectively kills tumour cells and preserves normal tissue. Dose per fraction protocols for skin malignancy in most RT departments are usually a balance between achieving disease control, adequate cosmesis, staffing of machines and patient convenience. A greater dose per fraction means that fewer visits to the RT department are needed. However, it also means that there is less chance for complete normal cell recovery, and the damaged normal cells are replaced over time by fibrous tissue. This may lead to less satisfactory functional and cosmetic outcomes. One reason why RT may not be offered is because of the perceived poorer functional and cosmetic outcomes due to these RT late effects. In skin, late effects include hypopigmentation, tissue thinning and telangiectasia. Late effects are related to fraction size.[18, 19]
Depth of treatment is important for prescribing the correct radiation modality. A study that investigated the depth of extension of skin appendages in 20 patients with LM from our institution showed that hair follicles extended to a median depth of 1·5 mm with a range of 0·5–4·5 mm (R.A. Scolyer, unpublished data). To ensure that LM is adequately treated, RT treatment should extend to a depth of 5 mm.
Materials and methods
The purpose of this study was to review the published experience of RT in the treatment of LM, focusing on technical aspects in order to develop the best evidenced-based RT protocol. Within the articles, published information on when recurrence typically occurs after RT, and how this occurs (i.e. locally, regionally or distantly), was also investigated. Searches were conducted in June 2012 using search terms ‘radiation therapy’ and ‘lentigo maligna’ and ‘Hutchinson's melanotic freckle’.
A systematic review or meta-analysis was not attempted as this would have entailed retrieving individual patient records, which was beyond the scope of the study. This study is therefore a collation of the final results of the published articles.
With PubMed, the search terms ‘radiation therapy’ and ‘lentigo maligna’ identified 89 publications. The search terms ‘radiation therapy’ and ‘Hutchinson's melanotic freckle’ identified 40 publications, all of which had previously been identified by the ‘radiation therapy’ and ‘lentigo maligna’ search. The Embase.com database search identified 101 publications. Only one new clinical study that examined the use of RT for LM treatment was identified. The remaining publications were reviews, did not specifically address RT and LM, or had been identified in the PubMed searches. The search of the Medline (OvidSP 1950 to present) database did not identify further publications. The search strategy is outlined in Figure 3.
Nine articles satisfied the search criteria and are listed in Table 1.[20-28] All the identified published articles were retrospective studies from single institutions. Harwood and colleagues coauthored five articles documenting the experience from the same institution.[21, 29-32] As there appeared to be significant overlap of the patient cohorts in these studies (with the treatment being analysed from different perspectives), the last article published, which involved the highest number of patients, was utilized for the purpose of this analysis. The article by Tsang et al. is from the same institution as Harwood and there is overlap in the years of accrual (1968–1988 vs. 1 July 1958–1982, respectively). Tsang et al. say that their findings confirm those of Harwood, which is then cited, so we have presumed that there is no overlap between the two cohorts. The article by Petratos et al. was from the same institution as one by Kopf et al. and for similar reasons only the latter was included in this analysis. The other six articles reported the experience of six other institutions.
|Reference||Year of publication||Number of patients with LM treated definitively with RT||Patient age of whole cohorts (years), median (range)||Accrual years of whole cohorts||Follow-up of whole cohorts (months), median (range)||All recurrences reported, n|
|Kopf et al.||1976||12||69 (54–83)||1964–73||50 (6–132)||5|
|Harwood||1983||23||72 (56–92)||1958–82||26 (5–96)||2|
|Campolmi et al.||1994||7||(52–82)||1987–92||Unable to determine for just LM||0|
|Tsang et al.||1994||36||71 (50–92)||1968–88||72 (12–144)||4|
|Christie and Tiver||1996||7||80 (76–89)||1979–95||20 (8–37)||0|
|Panizzon||1999||129||Not stated||1941–98||114 (6–228)||2|
|Schmid-Wendtner et al.||2000||42||76 (54–87)||1987–98||15 (1–96)||0|
|Farshad et al.||2002||93||71 (35–95)||1950–2000||96 (24 to >96)||5|
|Hedblad and Mallbris||2011||188||72 (30–90 + )||1990–2009||Unable to determine for just LM||Unable to report for LM alone|
|Total where relevant||N/A||537||30–95||1941–2009||1–228||18|
Table 1 summarizes the studies published to date reporting the use of RT to treat LM. In total, 537 patients with LM were treated with definitive primary RT between 1941 and 2009 with a median of reported follow-up times of 3 years.
Only eight articles could be reviewed for oncological outcome data. A large series by Hedblad and Mallbris could not be included as the outcomes of LM and LMM were combined. There were 18 recurrences documented in a total of 349 assessable patients (5%) after a median of reported follow-up times of 3 years. Salvage was attempted successfully in the majority of recurrent LM by using further RT or surgery or other therapies. Recurrence was simply observed if the patients were considered too frail for further treatment. Progression to LMM occurred in five of 349 patients: three of 12 patients in Kopf et al. and two of 36 patients in Tsang et al., all of whom had poor outcomes involving significant salvage surgery.
Table 2 summarizes the effect of RT field size in the three studies where this was documented. The protocol in all nine studies was to determine the treatment margins by expanding the radiation field for a fixed distance from the visible lesion. There were five marginal recurrences out of 123 assessable patients. The data suggest to us that the larger the treatment field margin, the lower the recurrence rate. Details of whether skin collimation was used are not noted in the studies. This is important for assessing how much of the field was penumbra.
|Reference||Margin from visible pigment to field edge (mm)||Number of patients with LM||Number of LM marginal recurrences||Crude recurrence rate (%)|
|Campolmi et al.||10–20||7||0||0|
|Farshad et al.||7–10||93||4||4·3|
Table 3 summarizes the effect of depth of treatment on in-field recurrence. This was specified in four studies. There were a total of eight in-field recurrences with either LM (five) or LMM (three) out of 171 assessable patients. The recurrence in the Harwood study is probably influenced by inadequate dose rather than inadequate depth. In the other studies, the results suggest to us that the more penetrating beams give better in-field control.
|Reference||Number of patients||Number of in-field LM recurrences||Crude recurrence rate (%)||Prescribed radiation depth dose||Modality|
|Harwood||23||1||4·3||50% at 6 mm||SXRT|
|Tsang et al.||36||1||2·7||80% dose to 5 mm||SXRT|
|Campolmi et al.||7||0||0||50% at 3–12·5 mm||SXRT|
|Farshad et al.||93||1||1·1||50% at 1 mm||Grenz|
|Kopf et al.||12||3||25·0||50% at 1·3 mm||Grenz|
Table 4 outlines the specialty of the first author, the radiation modality used and the cosmetic outcomes and time to pigment resolution relating to total dose and dose per fraction. All the studies could be included in these nononcological endpoints. Generally, first author radiation oncologists used superficial RT whereas dermatologists used Grenz rays. Overall, pigment resolution was not related to any radiation criteria. Cosmesis could not be related to fraction size. Anecdotally, in terms of functional outcome, one patient in the Kopf et al. series treated with Grenz rays with high dose per fraction developed an ectropion, presumably from in-field fibrosis.
|Reference||Clinician specialty||Modality||Radiation ranges per week, dose (Gy)/#/#||Time to pigment resolution (months),median (range)||Cosmesis|
|Tsang et al.||RO||SXRT||35/5/5–50/15/5||6 (3–12)||Poor in 11%|
|Christie and Tiver||RO||SXRT||44/11–57/23/5||3||Acceptable|
|Campolmi et al.||Derm||SXRT||42/7/2– 50/20/3||4 (2–8)||Acceptable|
|Farshad et al.||Derm||Grenz||120/10/2||Not given||Acceptable|
|Hedblad and Mallbris||Derm||Grenz||100–160/10–12/2||Not given||Excellent|
|Kopf et al.||Derm||Grenz||100/5/2||Not given||Fair–poor|
|Schmid-Wendtner et al.||Derm||Grenz||100/10/5||(2–8)||Good–excellent|
Other issues of an anecdotal nature arose out of the review. Kopf et al. had a quality assurance problem. The department's Grenz ray machine was recalibrated during the time period between the publications of their two papers. The beam penetration increased from depositing the reference dose at 0·7 mm in their 1972 paper to 1·3 mm in their final paper published in 1976; i.e. the penetration almost doubled in 4 years. There were no other quality assurance issues mentioned in any other study.
All the studies have been from single institutions and are all of a retrospective nature. No study was complete in providing details of all the RT parameters specified. Selection bias was found – if surgery and RT were reported together in the one study (e.g. Tsang et al.), those treated with RT tended to be the patients with the least favourable characteristics, those with lesions that were not suitable for surgery because of factors such as lesion size and patient comorbidity.
This study has reviewed the published experience of RT in treating LM. It has paid special attention to technical aspects so that the most appropriate, evidenced-based RT protocol can be recommended in this disease. The review was not very helpful in defining the optimal radiation parameters that should be used to treat LM. However, they have been taken into account in the following recommendations for the treatment of LM.
Radiotherapy techniques recommended for the treatment of lentigo maligna
These recommendations have been based on the preceding review and supplemented by our initial experience from our new multidisciplinary LM clinic at the Melanoma Institute Australia, the largest melanoma unit in the world. The recommendations coming from the review would be at best level 3, those from our experience at best level 4.
The RT field treated for LM has traditionally been determined by treating the visible lesion with a margin. This practice assumes that the nonvisible extensions of LM around the visible lesions extend equally in all directions. The review suggested to us that in the three studies (Table 2) with sufficient data, the greater is the margin between the visible lesion and the treatment field edge, the fewer are the out-of-field recurrences. It is now possible to define the treatment volume more accurately by using in vivo reflectance confocal microscopy (RCM).[34, 35]
For LM treatment in our new clinic, the area marked as involved by RCM is further expanded by 1 cm all around, if anatomically possible, to define the treated area. These areas are photographed and recorded. Future events, e.g. recurrence, can then be recorded as in-field or out-of-field. In RT planning terminology, the RCM-defined field will be the gross tumour volume (GTV), and the clinical target volume will be a 1-cm expansion on GTV.
Depth of treatment should be determined by the depth of extension of LM. LM can migrate down skin appendages that are in direct continuity with the overlying epidermis (where LM arises). Kopf et al. reports in case 12 recurrent LM in a sweat gland, which may have been physically below the effective dose given by Grenz rays. There is little documentation of the actual depth of skin appendageal involvement in the LM literature to date. The review suggested to us that in five studies with sufficient data, the more penetrating radiation was associated with less in-field recurrence (Table 3). Based on a small study at our institution, RT treatment should extend to a depth of 5 mm to treat LM adequately (R.A. Scolyer, unpublished data). The depth actually treated will be determined by the RT modality used. Beams that attenuate too superficially may result in geographical miss at depth.
Dose parameters include total dose, dose per fraction and number of fractions given per week. It is not possible to give a dose response from the review except to say that 20 Gy in one fraction is insufficient based on one case. Doses as low as 35 Gy in five fractions were associated with long-term control. Our practice has been to prescribe 50 Gy for adjuvant treatment, with 54 Gy for definitive treatment; and no more than 60 Gy in 2 Gy fractions or equivalent in greater fraction sizes.
Dose per fraction
In the review, dose per fraction varied from 20 Gy in one fraction with superficial RT, to 160 Gy in six fractions at two treatments per week using Grenz rays. A conclusion on this issue could not be made from the review. There was one ectropion possibly caused by fibrosis from high dose per fraction. We use a dose per fraction of 2 Gy to decrease potential late effects but a higher dose per fraction of up to 4 Gy is reasonable if patient factors require this.
Outcome measures reported in lentigo maligna radiotherapy treatment
Outcomes include progression to LMM, recurrence or persistence of LM, resolution of pigmentation, and cosmetic and functional results.
Progression to lentigo maligna melanoma
One of the reasons why LM is treated is to stop progression to LMM (which has metastatic potential and can cause death), so this is clearly an important outcome to measure. In the review overall, progression to LMM occurred in five of 349 patients (1·4%) at a median reported follow-up time of 3 years. Three of 12 patients treated with Grenz rays in the study of Kopf et al. progressed with LMM and may have been due to dosimetric miss at depth of unrecognized microinvasive disease at presentation. Two of 36 patients in the study of Tsang et al. treated with superficial RT also developed LMM. It is unclear whether these two were in-field lesions. All five LMM recurrences were salvaged by using surgery but some then progressed with distant metastatic disease.
Recurrence of lentigo maligna after radiotherapy
In the review, eight articles that could be reviewed for oncological outcome data reported 18 recurrences in a total of 349 assessable patients (5%) at a median reported follow-up time of 3 years. RCM can be used to check for recurrence or persistence of LM over time. In our practice, RCM imaging is done no earlier than 6 months post-treatment to ensure RT changes have resolved and will not act as false positives.
Resolution of pigmentation
When LM cells are destroyed, the melanin that was within the cells is left in the extracellular space and eventually phagocytosed by homeostatic mechanisms. Resolution of pigmentation in the review studies occurred after 2–24 months, with an average of 6 months. Bigger, darker lesions were treated with higher doses. Pigment resolution was not related to RT parameters per se. RCM can distinguish melanophages from melanocytes.
Cosmetic and functional outcomes
RT late effects consist mainly of fibrosis and are expressed clinically as in-field hypopigmentation, telangiectasia, alopecia and decreased skin elasticity. These are determined by total dose, fraction size, time since treatment and, perhaps, patient factors. Cosmetic and functional outcomes can be monitored with quality of life, physician review and serial photography.
This study aimed to review the literature on how LM has been treated with RT in order to define evidenced-based recommendations for the irradiation of LM. All studies found were single institution retrospective series. There was overall a crude 5% LM recurrence rate in 349 irradiated patients, for whom the median study follow-up was 3 years. Unfortunately, there was a lack of standardization of the studies and overall they were not a great help in defining the optimum RT parameters to be used. However, a series of recommendations were developed for field, dose and outcome measures from the review and from our initial experience with a new multidisciplinary LM clinic. We will use these recommendations to collect data in a prospective manner. Based on the measured depth of extension of skin appendageal structures on histopathology in a series of our patients with LM, we recommend that the depth of effective treatment reach 5 mm.
- 2Pathology and Genetics of Skin Tumours. Lyon, France: IARC Press, 2006., , et al.
- 3Weedon's Skin Pathology, 3rd edn. London: Elsevier Publications, 2010..
- 11Australian Cancer Network Melanoma Guidelines Revision Working Party. Clinical Practice Guidelines for the Management of Melanoma in Australia and New Zealand. Cancer Council Australia and Australian Cancer Network, Sydney and New Zealand Guidelines Group: Wellington, 2008. Available at: http://www.nhmrc.gov.au/guidelines/publications/cp111 (last accessed 17 October 2013).
- 15Recent Australian advances in the radiotherapy of skin cancer. Int J Bioautomation 2012; 16:43–52..
- 22Radiotherapy of lentigo maligna and lentigo maligna melanoma. Skin Cancer 1994; 9:23–6., , et al.
- 25Radiotherapy of lentigo maligna and lentigo maligna melanoma. Skin Cancer 1999; 14:203–7..
- 36International Commission on Radiation Units and Measurements. Prescribing, recording, and reporting photon beam therapy. ICRU Report 50. Bethesda, MD: ICRU, 1993.