The role of spectrophotometry in the diagnosis of melanoma
- 8.6k Downloads
Spectrophotometry (SPT) could represent a promising technique for the diagnosis of cutaneous melanoma (CM) at earlier stages of the disease. Starting from our experience, we further assessed the role of SPT in CM early detection.
During a health campaign for malignant melanoma at National Cancer Institute of Naples, we identified a subset of 54 lesions to be addressed to surgical excision and histological examination. Before surgery, all patients were investigated by clinical and epiluminescence microscopy (ELM) screenings; selected lesions underwent spectrophotometer analysis. For SPT, we used a video spectrophotometer imaging system (Spectroshade® MHT S.p.A., Verona, Italy).
Among the 54 patients harbouring cutaneous pigmented lesions, we performed comparison between results from the SPT screening and the histological diagnoses as well as evaluation of both sensitivity and specificity in detecting CM using either SPT or conventional approaches. For all pigmented lesions, agreement between histology and SPT classification was 57.4%. The sensitivity and specificity of SPT in detecting melanoma were 66.6% and 76.2%, respectively.
Although SPT is still considered as a valuable diagnostic tool for CM, its low accuracy, sensitivity, and specificity represent the main hamper for the introduction of such a methodology in clinical practice. Dermoscopy remains the best diagnostic tool for the preoperative diagnosis of pigmented skin lesions.
The incidence of melanoma in Europe has been steadily increasing, more rapidly than for any other cancer, during recent decades. Incidence of melanoma deeply varies across Europe, with the highest incidence in Scandinavian countries.
Lifetime risk of developing melanoma within the entire caucasian population is estimated to be 1:535 individuals. Incidence is higher in individuals with fair skin who have been exposed to high levels of UV-B radiation. Mortality due to metastatic melanoma has increased rapidly in both males and females; such a disease accounts for only 4% of skin cancers, but for 79% of skin-cancer related deaths. In particular, mortality rates are higher in males than in females; mortality has doubled in males over the past 25 years. Again, death rates are higher in individuals with fair skin. From 2002 to 2006, the median age at death due to metastatic melanoma of the skin was 68 years. Both incidence and mortality rates are highest in Australia and New Zealand. The 5-year survival rate for patients with advanced melanoma is less than 10% [1, 2, 3, 4, 5, 6].
In melanoma patients, survival is strongly related to tumor thickness; earlier diagnosis and complete excision of lesions may thus lead to a decline in mortality for such a disease. The most effective tool in fighting malignant melanoma is detecting the disease with a depth of ≤ 1 mm (Breslow thickness) and without ulceration, which is associated with an excellent prognosis (95% 5-year survival rate), in comparison to detection of a malignant melanoma with a depth of > 4 mm (40% 5-year survival rate), as reported by the American Joint Committee on Cancer .
Diagnostic accuracy for pigmented skin lesions using the naked eye has been estimated to be about 60% . To improve the accuracy of melanoma diagnosis, a variety of diagnostic instruments has been developed: dermoscopy, multispectral imaging, confocal laser microscopy, ultrasonography, optical coherence tomography, or magnetic resonance imaging .
From its introduction in 1990s, dermoscopy or epiluminescence microscopy (ELM), a non-invasive tool for cutaneous screenings, has become the most popular technique for early diagnosis of melanoma [10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33], also reducing the excess of cases addressed to excision biopsy . However, ELM is an useful tool in expert hands. As stated in a systematic review by Kittler and colleagues, dermoscopy does improve the diagnostic accuracy for melanoma in comparison with eye-based inspection, when used by experienced examiners only . Recentely, a meta-analysis evaluating diagnostic accuracy for melanoma confirmed that sensitivity was much higher for dermoscopy (0.90; 95% confidence interval [CI], 0.80-0.95) than for naked eye examination alone (0.71; 95% CI, 0.59-0.82), with an estimated difference of about 0.18 (95% CI, 0.09-0.27; P = 0.002) . Conversely, there was no statistical evidence of a significant difference in specificity [dermoscopy 0.90 (95% CI, 0.57-0.98) versus naked eye examination 0.81 (95% CI 0.48-0.95); difference 0.09 (95% CI, 0.06-0.23, P = 0.18)] . In addition, a meta-analysis of 30 studies using dermoscopy showed a sensitivity of 0.88 (95% CI, 0.87-0.89) and a specificity of 0.86 (95% CI, 0.85-0.86) .
Since sensitivity and specificity of dermoscopy significantly vary among different studies without reaching the 100% of validity, additional non-invasive techniques, such as spectrophotometric intracutaneous analysis (SIA) and spectrophotometry (SPT), have been introduced to improve early diagnosis of melanoma.
The SIA technique (SIAscopy, as named by Moncrieff et al. ), which was developed using the model of light transport [39, 40], is aimed at differentiating melanomas from non-melanoma skin cancers, benign melanocytic naevi and other pigmented lesions such as seborrhoeic keratosis .
The spectrophotometry is a method that evaluate the color of a lesion by measuring its reflectance as a function of the wavelength. A pioneer in this field was Marshall  who described a comprehensive analysis of pigmented lesions under both ultraviolet and infrared radiations. In such a study, he stated that the infrared photographs tends to discriminate the different types of lesions, with melanoma showing a relatively high degree of correlation with low infrared reflectance.
The SPT (Spectroshade, MHT Verona, Italy) system allows to extract information regarding the skin microarchitecture, like the distribution, position, and quantity of blood, collagen and melanin throughout the epidermidis and papillary dermis, converting data from reflected near infrared radiation into in vivo histological images of the lesion . The SPT system provides information including a series of 15 multispectral images into the near infrared bandwidth. On this regard, permanence of the image into the infrared area of the spectrum has been demonstrated to represent a significant indication of malignancy of the pigmented lesion under examination. Seven parameters (mean reflectans, MR; variegation, V; area, A; dark area ratio, DAR; dark island reflectance, DA; dark distribution factor, DDF) are processed automatically by software, describing the pigmented lesion features. Three spectral areas play a major role in quantification of all these parameters: 584 nm, where the highest contrast between lesion and the surrounding skin is determined; 650-750 nm, where a pigmented lesion can be better discriminated since the light adsorption due to melanine is much higher than that due to blood; 750-950 nm, where the lesion color is quantified in the near infrared, a spectral area invisible to the clinician's eye. After exposing the skin to visible light and infrared radiation, the spectrophotometric system converts the reflected radiations into in vivo histological images of the lesion by means of a sophisticated computer algorithm. Such an image allow to extract information regarding the microarchitecture of the cutaneous alteration under examination.
Starting from our experience in the use of both ELM and SPT, we intend to define the role of spectophotometric diagnosis in the early detection of cutaneous melanoma.
In recent years, physicians of the Melanoma Cooperative Group at the National Cancer Institute of Naples have performed a considerable amount of visits for cutaneous screening (about 10.000 visits per year). For the evaluation of clinically-relevant pigmented lesions, epiluminescence microscopy (ELM) was carried out using a Molemax II (Derma Instruments, Vienna) video-dermatoscope.
ELM-based classification of cutaneous lesions and corresponding histology
angiokeratoma, verrucous nevus, pigmented basal cell carcinoma, seborrhoeic keratosis, angioma, keratoacantoma, solar keratosis, etc.
(without pigment network)
compound nevus, intradermal nevus, papillomatous compound nevus, Spitz nevus, blue nevus.
(with pigment network)
junctional nevus, lentigo simplex, pigmented spindle cell nevus (Reed nevus), nevus spilus, cockarde nevus, atipycal nevus, melanoma.
squamous carcinoma, trichoepithelioma, Becker's nevus, Ota's nevus, Ito's nevus, Halo nevus.
ELM-based risk classification for cutaneous pigmented lesions
(very high risk)
Lesion with a pigment network and any of the classical ELM features specific for melanoma (pseudopods, radial streaming, blue-gray veil, atypical vessel, etc.)
Lesion with a pigment network and subtle new ELM features that may suggest melanoma but often are also seen in atypical nevi.
Lesion with a pigment network carrying subtle perturbations that can be detected in atypical naevus as well as in melanocytic hyperplasia.
Lesion with a benign appearing network.
(very low risk)
Lesion with a benign appearing network and with a globular pattern or another benign ELM pattern.
Surgical excision was advised for high and very high risk lesions. In type A of medium, low, and very low risk lesions, surgery may be justified on the basis of cosmetic and/or functional reasons (in absence of contraindications for surgical excisions). Anyway, patients with medium risk lesions entered a close follow-up program. Surgery was always recommended for all type B lesions . In case of advise for surgery of a melanocytic lesion, patients were invited to participate in our study; a written informed consent was obtained before patient's enrollment into the study.
SPT diagnostic classes and corresponding ELM risk classes
PROBABILITY OF MELANOMA
CORRESPONDING ELM RISK CLASS
Non Melanocytic and Type 5
doubtful melanoma (< 25%)
suspected melanoma (< 75%)
Types 3 and 2
probable melanoma (> 75%)
The study was reviewed and approved by the Ethycal Committee at the National Cancer Institute of Naples (Italy).
SPT sensitivity and specificity as well as positive and negative predictive values were determined according to a computer-assisted statistical analysis using the SPSS software for Windows, 8.0 version (SPSS Inc., Chicago, USA). Taking histology as standard, a lesion was considered true positive (TP) or true negative (TN) if results from SPT or ELM and histology were consistent (in other words, a concordance was achieved using such two diagnostic approaches). Conversely, a lesion was considered as false positive (FP) or false negative (FN) if histology did not confirmed the classification of SPT/ELM (in other words, SPT/ELM was positive but histology was negative or SPT/ELM was negative but histology was positive). Sensitivity was calculated as TP/TP+FN; specificity was calculated as TN/TN+FP; positive predictive value was calculated as TP/TP+FP; negative predictive value was calculated as TN/TN+FN.
In a period of one year, total-body skin examinations for detection of early malignant melanoma were carried out at National Cancer Institute of Naples; they allowed to identify cutaneous pigmented lesions to be addressed to surgical excision and histological examination in 54 patients [35 females and 19 males; median age, 41 (range, 19-73 years)]. Before surgery, lesions were evaluated by both ELM and SPT approaches.
Agreement between ELM diagnosis and histology
Type of lesion
Histological Diagnosis Positive/Negative
Very low risk
Very high risk
Before surgery, the same 54 selected pigmented lesions were evaluated by SpectroShade software, which automatically classified them in four different classes: group 1, non melanoma; group 2, doubtful melanoma (< 25%); group 3, suspected melanoma (< 75%); group 4, probable melanoma (> 75%). Since criteria for classification into such groups appeared to be not particularly accurate, we decided to modify such a classification and correlate it to the ELM-based risk assessment of each specific cutaneous lesion. Therefore, we defined the following classes: group 1, non melanocytic and very low risk lesions; group 2, low risk lesions; group 3, medium and high risk lesions; group 4, very high risk lesions (Table 3).
Agreement between SPT diagnosis and histology
Histological Diagnosis Positive/Negative
Agreement between SPT and ELM classifications
ELM Diagnosis Positive/Negative
Sensitivity and specificity for SPT in group 4 lesions
Sensitivity = 66.6
Specificity = 76.2
Breslow thickness of melanomas diagnosed in the series of 54 cutaneous lesions
BRESLOW THICKNESS CATEGORIES
N. OF CASES
MEDIAN BRESLOW THICKNESS
< 1 mm
1.01 to 2 mm
2.01 to 4.0 mm
> 4.01 mm
Classification of pigmented lesions is actually based on both clinical evaluation, by naked-eye, and dermoscopy of the skin features; however, diagnosis strongly depends on the experience of physicians. At present, the ELM methodology, even in the hands of experienced examiners, does not have 100% sensitivity and specificity (overall, 88% sensitivity and 86% specificity ). To improve the early detection of melanoma, new diagnostic tools providing an automated classification of pigmented skin lesions (in other words, new devices which could be also used by non-expert physicians) have been proposed. Among others, the SPT is considered as the most promising method for such a purpose.
Although the series of analyzed patients in the present study was quite small, our findings seem to indicate that the SPT may have a limited role into the early diagnosis of melanoma (66.6% sensitivity; 76.2% specificity; 44.4% positive predicitive value; 88.9% negative predictive value). About sensitivity, one melanoma (0.8 mm breslow thickness, with a large dermal regression) was classified in group 1 (very low risk) by SPT; the other three melanomas (median breslow thickness, 0,4 mm; range 0,3-0,7 mm) were classified in group 2 by SPT. Similarly, specificity was quite low since only 8 out of 18 lesions, which have been classified in group 4 by SPT, were really malignant after histological diagnosis (consequently, the agreement with histology was also low).
Another important issue is the difficulty to perform the SPT evaluation on palpable lesions such as papillomatous naevi. In fact, 6 of these latter lesions and 3 dermal naevi were diagnosed in group 3 by SPT; therefore, this contributed to lower the agreement with the histology also in such a group.
Summary of the most important studies on sensitivity and specificity of SPT
Tomatis S et al., 1998 
Farina B et al., 2000 
Wallace VP et al., 2000 
Elbaum M et al., 2001 
Moncrieff M et al., 2002 
Haniffa MA et al., 2007 
Govindan K et al., 2007 
Glud M et al., 2009 
The present study
Consistently with this latter indication, Haniffa et al. (2007)  reported high values of sensitivity (87%) and specificity (91%) for SPT among 860 lesions examined using the SIAscope spectrophotometer; when the same lesions were however evaluated by a dermatologist with a dermatoscope and at least 3 years of experience, sensitivity and specificity resulted to be even higher (94% and 91%, respectively). Since diagnostic accuracy of dermoscopy and SPT was similar, again authors suggested that a spectrophotometer does not improve the diagnostic ability of experienced dermatologists . Finally, the recent study by Glud et al. (2009)  revealed a 100% sensitivity but a 59% specificity for SPT (as in Farina et al., 2000 ), further indicating that such an approach may overestimate the proportion of possible malignant lesions (high amount of false positives).
Although the SPT is still considered as a valuable diagnostic tool, the low specificity of such a methodology (probably, due to the interference of seborrhoeic keratoses, which are not always recognized as benign lesions) represents the main hamper for the introduction of such a diagnostic tool into the clinical practice. No seborrhoeic keratosis was instead present in our series; therefore, we may speculate that the difference in diagnostic accuracy between ELM and SPT was due to additional factors in our study. Probably, the automated software of the SPT might use some parameters that do not discriminate the features of melanoma or overestimate the role of others skin components (blood, melanin, and collagen in the infrared band) as markers of malignancy. Finally, the agreement between the two screening methods was unsatisfactory (68.5%), and the high number of false positives in group 3 and 4 makes the ELM evaluation as the best technique to help clinicians in early diagnosis of cutaneous melanoma. Consistently, SPT should be utilized in clinical trials only.
The spectrophotometry (SPT) allows to accurately define the skin microarchitecture features. However, the low accuracy of SPT for either thin melanocytic lesions (particularly, those with regression) or palpable lesions (such as papillomatous and dermal naevi) clearly indicates that dermoscopy remains the best diagnostic tool for the preoperative diagnosis of pigmented skin lesions.
Authors would like to thank MHT (Medical High Technologies S.p.A., Verona, Italy) for providing technical support in using the Spectroshade. A special thank to Alessandra Trocino, for providing excellent bibliography service and assistance, and Ilenia Visconti, for data management. Work was supported by Italian Ministry of Health "Progetto Ricerca Finalizzata" and Associazione UMANA Onlus.
- 3.Ferlay J, Bray F, Pisani P, Parkin DM: GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. In IARC CancerBase No. 5. IARCPress, Lyon; 2004. version 2.0Google Scholar
- 4.National Cancer Institute Surveillance Epidemiology and End Results (SEER) Fact sheet[http://seer.cancer.gov/statfacts/html/melan.html]
- 7.Balch CM, Gershenwald JE, Soong SJ, Thompson JF, Atkins MB, Byrd DR, Buzaid AC, Cochran AJ, Coit DG, Ding S, Eggermont AM, Flaherty KT, Gimotty PA, Kirkwood JM, McMasters KM, Mihm MC Jr, Morton DL, Ross MI, Sober AJ, Sondak VK: Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol 2009, 27: 6199–206. 10.1200/JCO.2009.23.4799CrossRefPubMedPubMedCentralGoogle Scholar
- 11.Cascinelli N, Ferrario M, Bufalino R, Zurrida S, Galimberti V, Mascheroni L, Bartoli C, Clemente C: Results obtained by using a computerized image analysis system designed as an aid to diagnosis of cutaneous melanoma. Melanoma Res 1992, 2: 163–70. 10.1097/00008390-199209000-00004CrossRefPubMedGoogle Scholar
- 12.Nachbar F, Stolz W, Merkle T, Cognetta AB, Vogt T, Landthaler M, Bilek P, Braun-Falco O, Plewig G: The ABCD rule of dermatoscopy. High prospective value in the diagnosis of doubtful melanocytic skin lesions. J Am Acad Dermatol 1994,30(55):1–9.Google Scholar
- 17.Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Sammarco E, Delfino M: Epiluminescence microscopy for the diagnosis of doubtful melanocytic skin lesions. Comparison of the ABCD rule of dermatoscopy and a new 7-point checklist based on pattern analysis. Arch Dermatol 1998, 134: 1563–70. 10.1001/archderm.134.12.1563CrossRefPubMedGoogle Scholar
- 25.Bauer P, Cristofolini P, Boi S, Burroni M, Dell'Eva G, Micciolo R, Cristofolini M: Digital epiluminescence microscopy: usefulness in the differential diagnosis of cutaneous pigmentary lesions. A statistical comparison between visual and computer inspection. Melanoma Res 2000, 10: 345–9. 10.1097/00008390-200008000-00005CrossRefPubMedGoogle Scholar
- 26.Ascierto PA, Palmieri G, Celentano E, Parasole R, Caracò C, Daponte A, Chiofalo MG, Melucci MT, Mozzillo N, Satriano RA, Castello G: Sensitivity and specificity of epiluminescence microscopy: evaluation on a sample of 2731 excised cutaneous pigmented lesions. The Melanoma Cooperative Study. Br J Dermatol 2000, 142: 893–8. 10.1046/j.1365-2133.2000.03468.xCrossRefPubMedGoogle Scholar
- 27.Piccolo D, Ferrari A, Peris K, Diadone R, Ruggeri B, Chimenti S: Dermoscopic diagnosis by a trained clinician vs. a clinician with minimal dermoscopy training vs. computer-aided diagnosis of 341 pigmented skin lesions: a comparative study. Br J Dermatol 2002,147(48):1–6.Google Scholar
- 29.Rubegni P, Burroni M, Cevenini G, Perotti R, Dell'Eva G, Barbini P, Fimiani M, Andreassi L: Digital dermoscopy analysis and artificial neural network for the differentiation of clinically atypical pigmented skin lesions: a retrospective study. J Invest Dermatol 2002, 119: 471–4. 10.1046/j.1523-1747.2002.01835.xCrossRefPubMedGoogle Scholar
- 31.Soyer HP, Argenziano G, Zalaudek I, Corona R, Sera F, Talamini R, Barbato F, Baroni A, Cicale L, Di Stefani A, Farro P, Rossiello L, Ruocco E, Chimenti S: Three-point checklist of dermoscopy. A new screening method for early detection of melanoma. Dermatology 2004, 208: 27–31. 10.1159/000075042CrossRefPubMedGoogle Scholar
- 33.Menzies SW, Bischof L, Talbot H, Gutenev A, Avramidis M, Wong L, Lo SK, Mackellar G, Skladnev V, McCarthy W, Kelly J, Cranney B, Lye P, Rabinovitz H, Oliviero M, Blum A, Varol A, De'Ambrosis B, McCleod R, Koga H, Grin C, Braun R, Johr R: The performance of SolarScan: an automated dermoscopy image analysis instrument for the diagnosis of primary melanoma. Arch Dermatol 2005, 141: 1388–96. [Erratum in Arch Dermatol 2006;142:558] 10.1001/archderm.141.11.1388CrossRefPubMedGoogle Scholar
- 42.Ascierto PA, Palmieri G, Botti G, Satriano RA, Stanganelli I, Bono R, Testori A, Bosco L, Daponte A, Caraco C, Chiofalo MG, Melucci MT, Calignano R, Tatangelo F, Cochran AJ, Castello G, Melanoma Cooperative Group: Early diagnosis of malignant melanoma: Proposal of a working formulation for the management of cutaneous pigmented lesions from the Melanoma Cooperative Group. Int J Oncol 2003, 22: 1209–15.PubMedGoogle Scholar
- 44.Tomatis S, Bartoli C, Bono A, Cascinelli N, Clemente C, Marchesini R: Spectrophotometric imaging of cutaneous pigmented lesions: discriminant analysis, optical properties and histological characteristics. J Photochem Photobiol B 1998, 42: 32–9. 10.1016/S1011-1344(97)00113-9CrossRefPubMedGoogle Scholar
- 47.Elbaum M, Kopf AW, Rabinovitz HS, Langley RG, Kamino H, Mihm MC Jr, Sober AJ, Peck GL, Bogdan A, Gutkowicz-Krusin D, Greenebaum M, Keem S, Oliviero M, Wang S: Automatic differentiation of melanoma from melanocytic nevi with multispectral digital dermoscopy: a feasibility study. J Am Acad Dermatol 2001, 44: 207–18. 10.1067/mjd.2001.110395CrossRefPubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-5945/10/5/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.