- Email: [email protected]
Melanoma versus dysplastic naevi: microtopographic skin study with noninvasive method
Journal of Plastic, Reconstructive & Aesthetic Surgery (2006) 59, 700–705
Melanoma versus dysplastic naevi: microtopographic skin study with noninvasive method Vittorio Mazzarello, Daniela Soggiu, Daniela Rita Masia*, Pasquale Ena, Corrado Rubino Department of Plastic Surgery, University of Sassari, San Pietro, Sassari, Italy Received 6 July 2005; accepted 27 October 2005
KEYWORDS Malignant melanoma; Atypical naevi; Scanning electron microscopy (SEM); Silicone replica; Polygonal tumoral areas (PTA)
Summary Malignant melanoma (MM) is a cutaneous tumour that originates from the epidermal melanocytes, and whose prognosis is strongly determined by tumour vertical thickness. ABCDE system is currently used for the clinical diagnosis of MM: lesion asymmetry, border irregularity, change or variegation in colour, diameterO6 mm and rapid evolution. Different methods have been used to improve the diagnosis of MM, but the results have not been convincing. Our work investigates the differences between MM and atypical naevi through skin surface analysis, using scanning electron microscopy (SEM): in fact ABCDE classical criteria are often insufficient for differential diagnosis of melanoma. In our study, we analysed 15 malignant melanomas surfaces from 15 patients and of 15 atypical naevi from 10 patients. On histological examination five superficial spreading melanoma (SSM) and 10 nodular melanoma (NMM) were found. Inclusion criteria used for selection were clinical diagnosis of SSM or clinical doubt between MM and atypical naevi (dysplastic naevi). Superficial skin texture was analysed using the silicone replica technique. The replica material is a polyvinylsiloxane derivative for dental application. Our results show that the melanoma has a very irregular surface, with a marked reduction or disappearance of cutaneous furrows, and formation of new lines with varying direction. In conclusion, the method above based on skin surface analysis, may be of help in diagnosis of early stage MM and in differentiating from atypical naevi. q 2006 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. E-mail address: [email protected] (D.R. Masia).
There is a world wide concern about the increasing incidence of malignant melanoma (MM), a skin cancer that develops from epidermal melanocytes.
S0007-1226/$ - see front matter q 2006 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2005.10.018
Melanoma versus dysplastic naevi MM is more common in Caucasians than in Asians or Hispanics, and rare in African people. Several research studies confirm that UV-radiation skin damage can cause MM and that it occurs mostly in skin areas that have not been exposed to excessive sunlight.1,2 Prognosis is related to the depth of invasion and to nodal status at diagnosis. Early stage MM is curable, but, once the MM has metastasised, prognosis is poor. Clinical factors that identify higher risk patients are the presence of multiple atypical naevi, large congenital naevi, and a family history of numerous atypical naevi or MM. This can help the clinician for early detection and diagnosis with accurate skin examination or noninvasive diagnostic techniques like ELM-D (digital videodermoscopy).3 Most MM occur ‘de novo’ in the Caucasians, in absence of an obvious precursor lesion. However, 20–50% of MM arises from a pigmented precursor, usually an acquired naevus. Caucasians generally have 20 or more melanocytic naevi. Therefore, a correct noninvasive diagnostic evaluation of every skin pigmented lesions is essential and should be followed by excision of all atypical naevi. Effective early diagnosis remain a goal, we do not have objective clinical markers that identify MM. Conventional diagnosis is made almost entirely on the basis of clinical inspection. Clinical characteristic include the ABCDE: lesion asymmetry, border irregularity, change or variegation in colour, diameter over 6 mm, and rapid evolution. Most of these can only been estimated semi quantitatively.4 Therefore, several methods have been investigated for diagnosis improvement such as measurement of vascularisation, 32P uptake, thermal abnormalities, sonography, NMR-tomography. However, results are not convincing.4 Clinically, there have been various reports on surface irregularities in MM in comparison to nevocellular naevi and normal skin. In other reports, however, identification of MM comes about by quantitative assessment of surface parameter by means of laser profilometry. Observations presented in these studies are controversial.5–7 Our studies investigates differences to be found in MM’s and atypical nevocellular naevi by means of skin surface parameters using scanning electron microscopy (SEM), in order to add new diagnostic parameter for differentiation between benign melanocytic lesion like dysplastic naevi (Clark) and MM’s. Conventional ABCDE criteria are not always sufficient for the differential diagnosis.
701
Materials and methods In our study, we analysed 15 malignant melanomas surfaces from 15 patients and of 15 atypical naevi from 10 patients. All clinical diagnoses were histologically confirmed. Among melanomas, histological finding showed five superficial spreading melanoma (SSM) and 10 nodular melanoma (NMM). Inclusion criteria were clinical diagnosis of SSM or clinical doubt between MM and atypical naevi (dysplastic naevi). The age of the patients—15 males and 10 females—ranged from 31 to 77 years. Superficial skin texture was analysed using the silicone replica technique. The replica material is a polyvinylsiloxane derivative Coltenew President Light body) for dental application. Before taking the mould, the skin was washed to ensure perfect hygiene and dried. Sebum was removed by means of swabbing with ether or acetone. After silicone solidification, a three-dimensional positive contra replica was obtained using an epoxid substance (araldite). Each positive cast was then covered with palladium gold and analysed by SEM Zeiss 962 A. For every cast, monochrome digital photographs were filed on hard disk during SEM observation. All photographs included the surrounding skin. The photographs included approximately 10–20 enlargements, 20 kV, 10 mm working distance, gamma 3. Morphology of the processed replica samples was filed on a PC hard disk and then classified using the Sibirani and Gammal guidelines. All images were processed using software based on combination algorithm, binary system, thinning, panning and linearisation (Takayashi, 1994).8 Computer analysis facilitated a geometric description of the superficial skin surface based on the different skin surface topography parameters: roughness, anisotropy, primary and secondary skin furrow number, widths of primary furrows (Table 1). Objective numerical data were statistically analysed: skin texture variation was obtained by means of Student data analysis.
Results Malignant melanoma Skin-morphology MM’s observed with scanning electron microscopy (Fig. 1), show anarchic morphology with both large and low magnification (10! to 100!). One can easily see its unusual aspect with variations between different zones in each MM. It is possible
702 Table 1
V. Mazzarello et al. Skin surface topography parameters
Parameters
Description
Roughness Anisotropy
Standard deviation of the grey level of each pixel in the graphic image Variation coefficient of the black dots number in each 13!13 network that makes up the binarized graphic image Straight lines number in the graphic image after computer linearisation Point relationship between binary and thin image (cutaneous sulci mean width)
Furrows number Furrows width
to describe common skin texture morphology for different MM’s. Two different areas in each MM, a central anarchic area and a peripheral anarchic area, can be defined. The former is characterised by a complete loss of primary and secondary furrows and by development of new tumoral lines (TL) which may follow the internal neoplastic growth. These lines have lost sharpness and show changeable directions within the anarchic area. Few features, curvilinear and semi-circle or winding linescan be recognised. TL can be flat or wavy. The latter have a variable waviness, compared to the neighbouring skin, seeming to be deeper and can delimitate polygonal tumoral areas (PTA) with smooth surface or with parallel thin lines.
PTA show odd circular lines. Typical of PTA are different and variable dimension within the same tumoral lesion. The wavy TL are peculiar to superficial spreading melanomas, appearing under the form of several thin parallel lines, intercepting at different angles so giving a rough surface similar to tissue paper. Peripheral anarchic areas show a surface similar to normal skin with few differences: reduced secondary furrows, presence of polygonal areas delineated by primary furrows with variable depths and direction. Image analysis Table 2 shows the results of computer analysis obtained on MM and surrounding normal skin. MM show, statistically significant differences for all the analysed parameter except for primary furrow widths when compared to normal surrounding skin.
Dysplastic naevi Skin-morphology Atypical naevi skin texture seen with high and low magnification (10! to 200!) show a regular texture (Fig. 2). The decreased number of secondary furrows with raised polygonal area surface is typical. A few primary furrows with greatly reduced depth or even disappearance seem to create coalition of polymorphic polygonal shapes. This slightly irregular repeated shape in the lesional area presents the same shape we find in normal skin. This is common both in flat than in elevated or nodular naevi. Elevated naevi also present deeper residual primary furrows following the normal directions perpendicular to surrounding normal skin. These furrows may be caused by new tension lines created by nevocytic skin columns of papillary dermis or dermoepidermal junction.
Figure 1 Skin texture of three MM (left, from top to bottom) and surrounding normal skin (right, from top to bottom) (SEM 20!).
Image analysis Table 3 shows the results of computerised analysis obtained on atypical naevi and surrounding normal skin.
Melanoma versus dysplastic naevi Table 2
Normal Melanoma P
703
MeanGSD of the cutaneous parameters in normal skin and in melanoma
Roughness
Anisotropy
Primary furrows number
Secondary furrows number
Width primary furrows
5.5G0.43 7.8G1.01 0.01
0.5G0.06 0.3G0.15 0.001
45.1G6.71 19G4.3 0.001
26.4G3.04 41.4G7.04 0.001
2.8G0.62 3.1G0.45 0.134
Number of furrows and anisotropy results to be statistically significant for atypical naevi. Anisotropy is increased because of the number of furrows reduction with increased regularity in primary lines direction with a minimal dispersion. Furrows are reduced in number (especially secondary) because of epidermal or dermal proliferation of naevic cells. Thus, we also have a roughness reduction.
Figure 2
Width increases in atypical naevi, is highly significant in elevated naevi.
Discussion Diagnosis of melanoma can be easy when the ABCDE’s clinical characteristics are all satisfied. It is often difficult to differentiate early melanoma
Skin texture of six dysplastic naevi (SEM 20!).
704 Table 3
V. Mazzarello et al. MeanGSD values of the cutaneous parameters in the normal skin and in the atypical naevi
Normal Atypical naevi P
Roughness
Anisotropy
Primary furrows number
Secondary furrows number
Width primary furrows
4.53G1.01 3.9G0.90 0.277
0.44G0.11 0.51G0.06 0.123
57G11.70 46G5.1 0.045
39G10.2 18G7.1 0.001
2.9G0.31 3.9G0.16 0.075
from benign pigmented lesions of similar clinical appearance, or from amelanotic lesions, because surface details cannot be determined quantitatively by visual inspection.5,6 Skin surface parameters are objective well defined means that can be used for melanoma diagnosis.8 In the last decade efforts have been made to improve diagnosis using several techniques, means for example ELM-D. Many authors report an improved accuracy of preoperative diagnosis using ELM-D but some technological items still require improvements.9,10 Kreusch in 1997 studied with laser profilometry superficial skin texture in melanomas and naevi.11 This technique showed a statistically significant difference between MM and naevi, and this data can be used to improve diagnosis. However, if used for roughness parameters, this method is more expensive and needs an expert to manipulate replicas. It is also a time-consuming procedure.11 Moreover, it is a difficult procedure that may lead to imprecise skin value results. Our work is the first that applies silicone replica technique and scanning electron microscopy (SEM) to melanoma and dysplastic naevi. SEM is an instrument available in many hospitals. With SEM
we obtained several digital images (10!, 20!) of skin lesions and surrounding normal skin. We compared normal surrounding skin texture to the lesion skin texture, because skin morphology and texture change in various regions and with age. Melanoma skin texture shows anarchic morphology with complete loss of primary and secondary furrows, smooth areas, and development of new tumoral lines with changeable directions. These furrows can be seen in all MM surface. All analysed parameters, except one, show a statistically significative difference between melanoma and surrounding normal skin. Thus, we have a demonstration of a high difference in surface skin parameters (Tables 1 and 2—Graphic 1). MM compared to surrounding normal skin showed statistically significant differences in all the parameters examined, with exception of furrow width: (1) increased superficial roughness due to the anarchic skin texture and to superficial skin gathering; (2) reduced anisotropy showing the abnormality of skin texture with changeable furrow direction and new tumoral lines which intercepted several times, at different angles with a strong directional dispersion. This value is not statistically significant in nodular MM’s due to the presence of smooth tumoral areas; (3) a remarkable decreased
Melanoma versus dysplastic naevi number of primary furrows that appears deeper than in surrounding normal skin. In nodular MM the increased furrow width are statistically significant due to the reduced number of furrows with an increased depth; (4) a raised number of secondary furrows, in consequence of well represented tumoral lines especially in nodular MM. In particular, a greatly reduced number of primary furrows (45–19 for mm2), together with increased secondary furrows, due to the presence of new tumoral lines (TL), have been observed. Lines with changeable direction, variable waviness, anarchic morphology, in pathological areas with anisotropy close to zero and leading to increased superficial roughness, were found. The latter, is an objective parameter which indicates superficial irregularity caused by new lines growth and anisotropy. On the contrary, atypical naevi skin texture seen with SEM shows a decreased number of secondary furrows with raised polygonal areas surface and no anarchy. The same shape found in normal skin was also evidenced. All analysed parameters are not statistically significant if compared to surrounding skin, if not the secondary furrows reduction. Scanning electron microscopy surface melanoma analysis shows clear objective and subjective differences compared to surrounding normal skin, not found in atypical naevi. In fact, in the latter, a regular texture bearing little differences compared to normal skin, was noticed. This technique shows that surface parameter can be a useful objective means for melanoma diagnosis, especially in case of doubts with dysplastic naevi. Well defined differences comparing melanomas skin surface parameters to peripheral normal skin,
705 as shown in this study, confirm what Kreusch obtained with different methods. This is an appropriate input for an effort to research for typical morphological patterns (texture formZF factor) and could enable clinician to make more highly sensitive and specific diagnosis of early curable melanoma.
References 1. Elwood JM, Koh HK. Etiology, epidemiology, risk factors and public health issues of melanoma. Curr Opin Oncol 1994;6: 179. 2. Ernstoff MS. Melanoma: screening and education. Clin Plast Surg 2000;27(3):317. 3. Friedman RJ, Rigel DS, et al. Early detection of malignant melanoma: the role of physician examination and selfexamination of the skin. CA Cancer J Clin 1985;35:130. 4. Hughes BR, Black D, et al. Comparison of techniques for the non-invasive assessment of skin tumors. Clin Exp Dermatol 1987;12:108. 5. Grin CM, Kopf AW, et al. Accuracy in the clinical diagnoses of malignant melanoma. Arch Dermatol 1990; 126:763. 6. Kopf AW, Mintzis M, Bart RS. Diagnostic accuracy in malignant melanoma. Arch Dermatol 1975;111:1291. 7. Mihm MC, Clark WH, Reed RJ. The clinical diagnosis of malignant melanoma. Semin Oncol 1975;2:105. 8. Takahashi M. Image analysis of skin surface contour. Acta Derm Venereol (Stockh) 1994;185:9. 9. Bondi EE, Elder DE, et al. Skin markings in malignant melanoma. J Am Med Assoc 1983;250:503. 10. Lederman JS, Fitzpatrick TB, Sober AJ. Skin markings in the diagnosis and prognosis of cutaneous melanoma. Arch Dermatol 1984;120:1449. 11. Kreusch JF, Busche H, Conneman BJ, et al. Differentiation between nevocellular naevi and malignant melanoma by skin surface parameters. In: Wilhelm KP, Elsner P, Berardesca E et al., editors. Bioengineering of the skin: skin surface imaging and analysis. Boca Raton: CRC Press; 1997. p. 289–99.
Melanoma versus dysplastic naevi: microtopographic skin study with noninvasive method Vittorio Mazzarello, Daniela Soggiu, Daniela Rita Masia*, Pasquale Ena, Corrado Rubino Department of Plastic Surgery, University of Sassari, San Pietro, Sassari, Italy Received 6 July 2005; accepted 27 October 2005
KEYWORDS Malignant melanoma; Atypical naevi; Scanning electron microscopy (SEM); Silicone replica; Polygonal tumoral areas (PTA)
Summary Malignant melanoma (MM) is a cutaneous tumour that originates from the epidermal melanocytes, and whose prognosis is strongly determined by tumour vertical thickness. ABCDE system is currently used for the clinical diagnosis of MM: lesion asymmetry, border irregularity, change or variegation in colour, diameterO6 mm and rapid evolution. Different methods have been used to improve the diagnosis of MM, but the results have not been convincing. Our work investigates the differences between MM and atypical naevi through skin surface analysis, using scanning electron microscopy (SEM): in fact ABCDE classical criteria are often insufficient for differential diagnosis of melanoma. In our study, we analysed 15 malignant melanomas surfaces from 15 patients and of 15 atypical naevi from 10 patients. On histological examination five superficial spreading melanoma (SSM) and 10 nodular melanoma (NMM) were found. Inclusion criteria used for selection were clinical diagnosis of SSM or clinical doubt between MM and atypical naevi (dysplastic naevi). Superficial skin texture was analysed using the silicone replica technique. The replica material is a polyvinylsiloxane derivative for dental application. Our results show that the melanoma has a very irregular surface, with a marked reduction or disappearance of cutaneous furrows, and formation of new lines with varying direction. In conclusion, the method above based on skin surface analysis, may be of help in diagnosis of early stage MM and in differentiating from atypical naevi. q 2006 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. E-mail address: [email protected] (D.R. Masia).
There is a world wide concern about the increasing incidence of malignant melanoma (MM), a skin cancer that develops from epidermal melanocytes.
S0007-1226/$ - see front matter q 2006 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2005.10.018
Melanoma versus dysplastic naevi MM is more common in Caucasians than in Asians or Hispanics, and rare in African people. Several research studies confirm that UV-radiation skin damage can cause MM and that it occurs mostly in skin areas that have not been exposed to excessive sunlight.1,2 Prognosis is related to the depth of invasion and to nodal status at diagnosis. Early stage MM is curable, but, once the MM has metastasised, prognosis is poor. Clinical factors that identify higher risk patients are the presence of multiple atypical naevi, large congenital naevi, and a family history of numerous atypical naevi or MM. This can help the clinician for early detection and diagnosis with accurate skin examination or noninvasive diagnostic techniques like ELM-D (digital videodermoscopy).3 Most MM occur ‘de novo’ in the Caucasians, in absence of an obvious precursor lesion. However, 20–50% of MM arises from a pigmented precursor, usually an acquired naevus. Caucasians generally have 20 or more melanocytic naevi. Therefore, a correct noninvasive diagnostic evaluation of every skin pigmented lesions is essential and should be followed by excision of all atypical naevi. Effective early diagnosis remain a goal, we do not have objective clinical markers that identify MM. Conventional diagnosis is made almost entirely on the basis of clinical inspection. Clinical characteristic include the ABCDE: lesion asymmetry, border irregularity, change or variegation in colour, diameter over 6 mm, and rapid evolution. Most of these can only been estimated semi quantitatively.4 Therefore, several methods have been investigated for diagnosis improvement such as measurement of vascularisation, 32P uptake, thermal abnormalities, sonography, NMR-tomography. However, results are not convincing.4 Clinically, there have been various reports on surface irregularities in MM in comparison to nevocellular naevi and normal skin. In other reports, however, identification of MM comes about by quantitative assessment of surface parameter by means of laser profilometry. Observations presented in these studies are controversial.5–7 Our studies investigates differences to be found in MM’s and atypical nevocellular naevi by means of skin surface parameters using scanning electron microscopy (SEM), in order to add new diagnostic parameter for differentiation between benign melanocytic lesion like dysplastic naevi (Clark) and MM’s. Conventional ABCDE criteria are not always sufficient for the differential diagnosis.
701
Materials and methods In our study, we analysed 15 malignant melanomas surfaces from 15 patients and of 15 atypical naevi from 10 patients. All clinical diagnoses were histologically confirmed. Among melanomas, histological finding showed five superficial spreading melanoma (SSM) and 10 nodular melanoma (NMM). Inclusion criteria were clinical diagnosis of SSM or clinical doubt between MM and atypical naevi (dysplastic naevi). The age of the patients—15 males and 10 females—ranged from 31 to 77 years. Superficial skin texture was analysed using the silicone replica technique. The replica material is a polyvinylsiloxane derivative Coltenew President Light body) for dental application. Before taking the mould, the skin was washed to ensure perfect hygiene and dried. Sebum was removed by means of swabbing with ether or acetone. After silicone solidification, a three-dimensional positive contra replica was obtained using an epoxid substance (araldite). Each positive cast was then covered with palladium gold and analysed by SEM Zeiss 962 A. For every cast, monochrome digital photographs were filed on hard disk during SEM observation. All photographs included the surrounding skin. The photographs included approximately 10–20 enlargements, 20 kV, 10 mm working distance, gamma 3. Morphology of the processed replica samples was filed on a PC hard disk and then classified using the Sibirani and Gammal guidelines. All images were processed using software based on combination algorithm, binary system, thinning, panning and linearisation (Takayashi, 1994).8 Computer analysis facilitated a geometric description of the superficial skin surface based on the different skin surface topography parameters: roughness, anisotropy, primary and secondary skin furrow number, widths of primary furrows (Table 1). Objective numerical data were statistically analysed: skin texture variation was obtained by means of Student data analysis.
Results Malignant melanoma Skin-morphology MM’s observed with scanning electron microscopy (Fig. 1), show anarchic morphology with both large and low magnification (10! to 100!). One can easily see its unusual aspect with variations between different zones in each MM. It is possible
702 Table 1
V. Mazzarello et al. Skin surface topography parameters
Parameters
Description
Roughness Anisotropy
Standard deviation of the grey level of each pixel in the graphic image Variation coefficient of the black dots number in each 13!13 network that makes up the binarized graphic image Straight lines number in the graphic image after computer linearisation Point relationship between binary and thin image (cutaneous sulci mean width)
Furrows number Furrows width
to describe common skin texture morphology for different MM’s. Two different areas in each MM, a central anarchic area and a peripheral anarchic area, can be defined. The former is characterised by a complete loss of primary and secondary furrows and by development of new tumoral lines (TL) which may follow the internal neoplastic growth. These lines have lost sharpness and show changeable directions within the anarchic area. Few features, curvilinear and semi-circle or winding linescan be recognised. TL can be flat or wavy. The latter have a variable waviness, compared to the neighbouring skin, seeming to be deeper and can delimitate polygonal tumoral areas (PTA) with smooth surface or with parallel thin lines.
PTA show odd circular lines. Typical of PTA are different and variable dimension within the same tumoral lesion. The wavy TL are peculiar to superficial spreading melanomas, appearing under the form of several thin parallel lines, intercepting at different angles so giving a rough surface similar to tissue paper. Peripheral anarchic areas show a surface similar to normal skin with few differences: reduced secondary furrows, presence of polygonal areas delineated by primary furrows with variable depths and direction. Image analysis Table 2 shows the results of computer analysis obtained on MM and surrounding normal skin. MM show, statistically significant differences for all the analysed parameter except for primary furrow widths when compared to normal surrounding skin.
Dysplastic naevi Skin-morphology Atypical naevi skin texture seen with high and low magnification (10! to 200!) show a regular texture (Fig. 2). The decreased number of secondary furrows with raised polygonal area surface is typical. A few primary furrows with greatly reduced depth or even disappearance seem to create coalition of polymorphic polygonal shapes. This slightly irregular repeated shape in the lesional area presents the same shape we find in normal skin. This is common both in flat than in elevated or nodular naevi. Elevated naevi also present deeper residual primary furrows following the normal directions perpendicular to surrounding normal skin. These furrows may be caused by new tension lines created by nevocytic skin columns of papillary dermis or dermoepidermal junction.
Figure 1 Skin texture of three MM (left, from top to bottom) and surrounding normal skin (right, from top to bottom) (SEM 20!).
Image analysis Table 3 shows the results of computerised analysis obtained on atypical naevi and surrounding normal skin.
Melanoma versus dysplastic naevi Table 2
Normal Melanoma P
703
MeanGSD of the cutaneous parameters in normal skin and in melanoma
Roughness
Anisotropy
Primary furrows number
Secondary furrows number
Width primary furrows
5.5G0.43 7.8G1.01 0.01
0.5G0.06 0.3G0.15 0.001
45.1G6.71 19G4.3 0.001
26.4G3.04 41.4G7.04 0.001
2.8G0.62 3.1G0.45 0.134
Number of furrows and anisotropy results to be statistically significant for atypical naevi. Anisotropy is increased because of the number of furrows reduction with increased regularity in primary lines direction with a minimal dispersion. Furrows are reduced in number (especially secondary) because of epidermal or dermal proliferation of naevic cells. Thus, we also have a roughness reduction.
Figure 2
Width increases in atypical naevi, is highly significant in elevated naevi.
Discussion Diagnosis of melanoma can be easy when the ABCDE’s clinical characteristics are all satisfied. It is often difficult to differentiate early melanoma
Skin texture of six dysplastic naevi (SEM 20!).
704 Table 3
V. Mazzarello et al. MeanGSD values of the cutaneous parameters in the normal skin and in the atypical naevi
Normal Atypical naevi P
Roughness
Anisotropy
Primary furrows number
Secondary furrows number
Width primary furrows
4.53G1.01 3.9G0.90 0.277
0.44G0.11 0.51G0.06 0.123
57G11.70 46G5.1 0.045
39G10.2 18G7.1 0.001
2.9G0.31 3.9G0.16 0.075
from benign pigmented lesions of similar clinical appearance, or from amelanotic lesions, because surface details cannot be determined quantitatively by visual inspection.5,6 Skin surface parameters are objective well defined means that can be used for melanoma diagnosis.8 In the last decade efforts have been made to improve diagnosis using several techniques, means for example ELM-D. Many authors report an improved accuracy of preoperative diagnosis using ELM-D but some technological items still require improvements.9,10 Kreusch in 1997 studied with laser profilometry superficial skin texture in melanomas and naevi.11 This technique showed a statistically significant difference between MM and naevi, and this data can be used to improve diagnosis. However, if used for roughness parameters, this method is more expensive and needs an expert to manipulate replicas. It is also a time-consuming procedure.11 Moreover, it is a difficult procedure that may lead to imprecise skin value results. Our work is the first that applies silicone replica technique and scanning electron microscopy (SEM) to melanoma and dysplastic naevi. SEM is an instrument available in many hospitals. With SEM
we obtained several digital images (10!, 20!) of skin lesions and surrounding normal skin. We compared normal surrounding skin texture to the lesion skin texture, because skin morphology and texture change in various regions and with age. Melanoma skin texture shows anarchic morphology with complete loss of primary and secondary furrows, smooth areas, and development of new tumoral lines with changeable directions. These furrows can be seen in all MM surface. All analysed parameters, except one, show a statistically significative difference between melanoma and surrounding normal skin. Thus, we have a demonstration of a high difference in surface skin parameters (Tables 1 and 2—Graphic 1). MM compared to surrounding normal skin showed statistically significant differences in all the parameters examined, with exception of furrow width: (1) increased superficial roughness due to the anarchic skin texture and to superficial skin gathering; (2) reduced anisotropy showing the abnormality of skin texture with changeable furrow direction and new tumoral lines which intercepted several times, at different angles with a strong directional dispersion. This value is not statistically significant in nodular MM’s due to the presence of smooth tumoral areas; (3) a remarkable decreased
Melanoma versus dysplastic naevi number of primary furrows that appears deeper than in surrounding normal skin. In nodular MM the increased furrow width are statistically significant due to the reduced number of furrows with an increased depth; (4) a raised number of secondary furrows, in consequence of well represented tumoral lines especially in nodular MM. In particular, a greatly reduced number of primary furrows (45–19 for mm2), together with increased secondary furrows, due to the presence of new tumoral lines (TL), have been observed. Lines with changeable direction, variable waviness, anarchic morphology, in pathological areas with anisotropy close to zero and leading to increased superficial roughness, were found. The latter, is an objective parameter which indicates superficial irregularity caused by new lines growth and anisotropy. On the contrary, atypical naevi skin texture seen with SEM shows a decreased number of secondary furrows with raised polygonal areas surface and no anarchy. The same shape found in normal skin was also evidenced. All analysed parameters are not statistically significant if compared to surrounding skin, if not the secondary furrows reduction. Scanning electron microscopy surface melanoma analysis shows clear objective and subjective differences compared to surrounding normal skin, not found in atypical naevi. In fact, in the latter, a regular texture bearing little differences compared to normal skin, was noticed. This technique shows that surface parameter can be a useful objective means for melanoma diagnosis, especially in case of doubts with dysplastic naevi. Well defined differences comparing melanomas skin surface parameters to peripheral normal skin,
705 as shown in this study, confirm what Kreusch obtained with different methods. This is an appropriate input for an effort to research for typical morphological patterns (texture formZF factor) and could enable clinician to make more highly sensitive and specific diagnosis of early curable melanoma.
References 1. Elwood JM, Koh HK. Etiology, epidemiology, risk factors and public health issues of melanoma. Curr Opin Oncol 1994;6: 179. 2. Ernstoff MS. Melanoma: screening and education. Clin Plast Surg 2000;27(3):317. 3. Friedman RJ, Rigel DS, et al. Early detection of malignant melanoma: the role of physician examination and selfexamination of the skin. CA Cancer J Clin 1985;35:130. 4. Hughes BR, Black D, et al. Comparison of techniques for the non-invasive assessment of skin tumors. Clin Exp Dermatol 1987;12:108. 5. Grin CM, Kopf AW, et al. Accuracy in the clinical diagnoses of malignant melanoma. Arch Dermatol 1990; 126:763. 6. Kopf AW, Mintzis M, Bart RS. Diagnostic accuracy in malignant melanoma. Arch Dermatol 1975;111:1291. 7. Mihm MC, Clark WH, Reed RJ. The clinical diagnosis of malignant melanoma. Semin Oncol 1975;2:105. 8. Takahashi M. Image analysis of skin surface contour. Acta Derm Venereol (Stockh) 1994;185:9. 9. Bondi EE, Elder DE, et al. Skin markings in malignant melanoma. J Am Med Assoc 1983;250:503. 10. Lederman JS, Fitzpatrick TB, Sober AJ. Skin markings in the diagnosis and prognosis of cutaneous melanoma. Arch Dermatol 1984;120:1449. 11. Kreusch JF, Busche H, Conneman BJ, et al. Differentiation between nevocellular naevi and malignant melanoma by skin surface parameters. In: Wilhelm KP, Elsner P, Berardesca E et al., editors. Bioengineering of the skin: skin surface imaging and analysis. Boca Raton: CRC Press; 1997. p. 289–99.