Characteristic Epiluminescent Microscopic Features of Early Malignant Melanoma on Glabrous Skin a Videomicroscopic Analysis

STUDY Characteristic Epiluminescent Microscopic Features of Early Malignant Melanoma on Glabrous Skin A Videomicroscopic Analysis

Shinji Oguchi, MD; Toshiaki Saida, MD, PhD; Yoko Koganehira, MD; Sachiko Ohkubo, MD; Yasushi Ishihara, MD; Shigeo Kawachi, MD, PhD

Objective: To investigate the characteristic epilumi- Results: On epiluminescent microscopy, malignant mela- nescent microscopic features of early lesions of malig- noma in situ and the macular portions of invasive malignant melanoma affecting glabrous skin, which is the most nant melanoma showed accentuated pigmentation on the prevalent site of the neoplasm in nonwhite populations. ridges of the skin markings, which are arranged in parallel patterns on glabrous skin. This “parallel ridge pattern” Design: The epiluminescent microscopic features of vari- was found in 5 (83%) of 6 lesions of malignant melanoma ous kinds of melanocytic lesions affecting glabrous skin in situ and in 15 (94%) of 16 lesions of malignant melanoma. were investigated using a videomicroscope. All the di- The parallel ridge pattern was rarely found in the lesions agnoses were determined clinically and histopathologi- of benign melanocytic nevus. Most benign melanocytic nevi cally using the standard criteria. showed 1 of the following 3 typical epiluminescent patterns: (1) a parallel furrow pattern exhibiting pigmentation on the Setting: A dermatology clinic at a university hospital. parallel sulci of the skin markings (54%), (2) a latticelike pattern (21%), and (3) a fibrillar pattern showing filamen- Patients: The following 130 melanocytic lesions con- tous or meshlike pigmentation (15%). The remaining 11 secutively diagnosed at our department were examined: benign nevi (10%) showed a nontypical pattern. 16 lesions of acral lentiginous melanoma, 6 lesions of malignant melanoma in situ, and 108 lesions of benign me- Conclusion: Because epiluminescent microscopic fea- lanocytic nevus (acquired or congenital). tures of early malignant melanoma on glabrous skin are characteristic, we can effectively detect early lesions us- Main Outcome Measure: The incidence of each char- ing this noninvasive method. acteristic epiluminescent feature was compared among disease categories. Arch Dermatol. 1998;134:563-568

PILUMINESCENT micros- lignant melanoma on glabrous skin, copy, also called dermatos- comparing them with those of benign me- copy or surface micros- lanocytic nevi on these sites. copy, is now widely used for the clinical evaluation of RESULTS pigmentedE skin lesions.1-11 The most important use of this technique is in the ac- MALIGNANT MELANOMA IN SITU curate diagnosis of malignant melanoma, particularly in its early stages. Although The clinical data of 6 patients with malig- many investigators have reported valu- nant melanoma in situ are shown in able findings on the subject, most au- Table 2. The main epiluminescent mi- thors investigated pigmented lesions of croscopic features of malignant mela- nonglabrous skin. In a previous study of noma in situ are summarized in Table 3. the use of a videomicroscope,12 the macu- All the lesions showed an abrupt edge, From the Department of lar and plaque portions of acral lentigi- namely, a sharply cut-off margin of pig- Dermatology, Shinshu nous melanoma of glabrous skin exhib- mentation, at least in a portion. Five of the University School of Medicine, ited accentuated pigmentation of the ridges 6 lesions showed accentuated pigmenta- Matsumoto (Drs Oguchi, Saida, Koganehira, Ohkubo, and of the skin markings that run mostly in tion on the ridges of the skin markings that Kawachi), and the Division of parallel on these anatomical sites. In this run mostly in a parallel pattern on the ana- Dermatology, Hokushin study, also with the use of a videomicro- tomical sites (Figure 1). In contrast, the General Hospital, Nakano scope, we investigated epiluminescent mi- furrows of the skin markings were com- (Dr Ishihara), Nagano, Japan. croscopic features of early lesions of ma- paratively devoid of pigmentation. This

ARCH DERMATOL / VOL 134, MAY 1998 563

©1998 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 Table 1. Melanocytic Lesions on Glabrous Skin MATERIALS AND METHODS Investigated With Videomicroscopy

Diagnosis No. of Lesions A total of 130 pigmented lesions on glabrous skin, Benign melanocytic nevus 108* including 6 lesions of malignant melanoma in situ, Junctional type 80† 16 of acral lentiginous melanoma, and 108 of ac- Compound type 28 quired or congenital melanocytic nevus, were inves- Intradermal type 0 tigated using a videomicroscope. All the lesions were Acral lentiginous melanoma 16 seen in Japanese patients at the Department of Der- Malignant melanoma in situ 6 matology, Shinshu University Hospital, Matsu- Total 130 moto, Japan, from November 5, 1990, to June 30, 1996 (Table 1). Informed consent was obtained from all *Ten lesions were diagnosed as congenital nevus. patients. The diagnoses were all determined clini- †Seven lesions of lentigo simplex were included. cally and histopathologically based on the standard criteria. Lentigo simplex was classified as acquired melanocytic nevus. Most of the melanocytic nevi were Table 2. Clinical Data of Patients from the foot, as follows: 72 lesions on the sole or With Malignant Melanoma In Situ of the Sole volar aspect of the toes and 36 lesions on the palm or volar aspect of the fingers. All the lesions of ma- Patient No./Sex/Age, y Size of Lesion, mm lignant melanoma and malignant melanoma in situ were from the sole of the foot. 1/F/75 16.0 ϫ 4.0 As described in detail in a previous report,12 each 2/F/27 15.0 ϫ 4.0 3/F/46 11.5 3.5 lesional site was immersed with mineral oil and cov- ϫ 4/M/54 6.0 ϫ 5.0 ered with a glass slide. With the camera probe ap- 5/M/67 16.0 ϫ 11.0 plied to the slide, the magnified features, ϫ20, ϫ50, 6/F/73 22.0 ϫ 11.0 and ϫ100, were viewed on a color monitor of the videomicroscope and recorded on a video floppy disk.

Table 3. Incidence of Videomicroscopic Features pigmentation pattern was observed in large areas of all of Melanocytic Lesions on Glabrous Skin (N = 130) cases. We designated this feature as a “parallel ridge pattern.” Corresponding to this epiluminescent finding, his- Melanocytic Lesions, No. (%) topathological examination revealed that a proliferation Acral Malignant Benign of atypical melanocytes with melanin granules was promi- Lentiginous Melanoma Melanocytic nent on the crista profunda intermedia, ie, epidermal rete Melanoma* In Situ Nevus ridges beneath the surface ridges and passed through by Features (n = 16) (n=6) (n = 108) an intraepidermal eccrine duct. Diffuse multicompo- Parallel ridge pattern 15 (94) 5 (83) 0 nent pigmentation composed of pigmented blotches of Abrupt edge 12 (75) 6 (100) 0 Peripheral dots 10 (62) 4 (67) 0 variegated shades of brown was observed in portions of Diffuse multicomponent 3 lesions. Peripheral black dots of various sizes were de- pigmentation 12 (75) 3 (50) 3 (3) tected in 4 lesions, which corresponded histopathologi- Fibrillar pattern 7 (44)† 1 (17) 16 (15) cally to the aggregated melanin granules shed into the Parallel furrow or cornified layer of the epidermis. One lesion of mela- latticelike pattern 4 (25)† 0 81 (75) noma in situ (patient 2 in Table 2) did not show the par- *Findings detected in the macular portions. allel ridge pattern but exhibited the fibrillar pattern of †Detected only in small areas within the lesions. subtype A (Figure 2), one observed in some lesions of melanocytic nevus as described later. In this lesion, de- pigmentation was observed on the periphery, resulting Table 3. The epiluminescent findings of the macular por- in an irregular, asymmetric configuration of the lesion. tions were essentially identical to those observed in the Pseudopods were rarely found in the lesions of mela- malignant melanoma in situ lesions. The parallel ridge noma in situ on the sole. Furthermore, the parallel fur- pattern was a predominant feature of 15 (94%) of the 16 row and latticelike patterns, which were the most com- lesions. The diffuse multicomponent pigmentation was mon patterns in benign nevi, were not detected in any found in 12 lesions of malignant melanoma (75%) at least lesions of melanoma in situ. in some portions within the lesions. The abrupt edge (75%) and peripheral dots (62%) were also frequently MALIGNANT MELANOMA detected. In the invasive portions of malignant melanoma, blotches of varying shades of brown, randomly dis- The 16 invasive primary lesions of acral lentiginous mela- tributed black dots or brown globules, the blue-gray veil noma ranged from 15 to 75 mm in maximum diameter (ie, diffuse bluish-gray pigmentation with a somewhat and from 1.0 to 8.0 mm in Breslow tumor thickness whitish hue), and depigmented areas were frequently rec- (Table 4). The epiluminescent microscopic features of ognized. The fibrillar pattern was detected in some small the macular portions of the lesions are summarized in areas of 7 (44%) of the 16 lesions. The parallel furrow

ARCH DERMATOL / VOL 134, MAY 1998 564

B B

C C Figure 1. Clinical (A), videomicroscopic (B, magnification ϫ100), and Figure 2. Clinical (A), videomicroscopic (B, magnification ϫ50), and histopathological (C, magnification ϫ75) features of malignant melanoma in histopathological features (C, magnification ϫ150) of malignant melanoma situ of the sole of the foot. The characteristic parallel ridge pattern exhibiting in situ. The fibrillar pattern of subtype A is clearly observed with selective pigmentation of the ridges of the skin markings is clearly videomicroscopy. demonstrated with videomicroscopy.

and the latticelike patterns were rare in the lesions of ma- the lesions showed solitary arranged melanocytes in the lignant melanoma, detected only in small portions in 4 lower epidermis. of the 16 lesions. Epiluminescent microscopic features of 97 (90%) of the 108 benign nevi were classified into 1 of the fol- BENIGN MELANOCYTIC NEVUS lowing 3 typical patterns: (1) a parallel furrow pattern exhibiting pigmentation on the parallel sulci of the skin The results of epiluminescent microscopy of 108 le- markings, which was observed in 58 (54%) of 108 lesions of acquired or congenital melanocytic nevus on gla- sions (Figure 3). In 19 (33%) of the 58 lesions show- brous skin are summarized in Table 5. Ten lesions were ing the parallel furrow pattern, small brown globules were judged clinically, histopathologically, or both, to be con- arranged in regular manner on the ridges of the skin mark- genital melanocytic nevi. The maximum diameter of the ings, usually sparing the eccrine pores. (2) A latticelike lesions of acquired melanocytic nevus ranged from 1 to pattern, considered a variant of the parallel furrow pat- 12 mm and that of congenital nevus from 6 to 15.5 mm. tern, showed pigmentation on the furrows and on the lines All these acquired and congenital lesions were judged his- crossing the furrows (Figure 4). This pattern was found tologically benign because of their overall symmetrical in 23 nevi (21%). (3) A fibrillar pattern, in which fila- structure and because the proliferating melanocytes (ne- mentous or meshlike pigmentation slanted the skin mark- vus cells) showed no nuclear atypia, although some of ings, was found in 16 lesions (15%).

ARCH DERMATOL / VOL 134, MAY 1998 565

©1998 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 Table 4. Clinical and Histopathological Data of Patients With Acral Lentiginous Melanoma on the Sole

Patient No./ Size of Tumor Clark Sex/Age, y Lesion, mm Thickness, mm Level 1/M/55 30 ϫ 23 2.9 IV 2/F/83 33 ϫ 20 3.5 IV 3/F/81 22 ϫ 12 2.5 IV 4/M/67 75 ϫ 60 4.5 V 5/M/55 34 ϫ 25 1.0 III 6/M/64 26 ϫ 16 4.0 IV 7/F/71 23 ϫ 20 3.8 IV A 8/M/46 35 ϫ 30 2.2 IV 9/F/56 15 ϫ 9 1.5 III 10/M/63 33 ϫ 25 5.0 V 11/M/65 40 ϫ 30 2.0 IV 12/M/37 42 ϫ 30 2.9 IV 13/M/55 21 ϫ 11 2.0 IV 14/M/72 35 ϫ 25 1.7 IV 15/M/78 30 ϫ 25 3.5 IV 16/M/69 50 ϫ 30 8.0 V

Table 5. Correlation Between Epiluminescent Patterns and Histopathological Subtypes of Melanocytic Nevi (N = 108)

B Histopathological Subtype Epiluminescent No. (%) of Pattern* Lesions Junctional Compound Intradermal Figure 3. Videomicroscopic (A, magnification ϫ50) and histopathological (B, magnification ϫ75) features of a melanocytic nevus on the sole showing Parallel furrow 58 (54) 44 14 0 the parallel furrow pattern. The furrows of the skin markings, shown in a Latticelike 23 (21) 19 4 0 parallel arrangement, are selectively pigmented, sparing the ridges of the Fibrillar 16 (15) 11 5 0 marking. Corresponding to the videomicroscopic pattern, findings from the Nontypical 11 (10) 6 5 0 histopathological examination revealed a preferential proliferation of nevus Total 108 (100) 80 28 0 cells (melanocytes) in the crista profunda limitans, ie, the rete ridges beneath the furrow. *See the “Results” section for a description of the patterns.

The following 2 subtypes of the fibrillar pattern were recognized: delicate filamentous pigmentation arranged in parallel marks (subtype A, 8 lesions) (Figure 5) and fibrillar pigmentation densely arranged in a meshlike pattern (subtype B, 8 lesions) (Figure 6). In benign melanocytic nevi more than 7 mm in maximum diameter, the incidence of the fibrillar pattern increased to 50% (10 of 20 lesions). In most nevi showing the 3 typical patterns, the entire lesion was composed of only 1 pattern, whereas 3 lesions exhibited the parallel furrow and the fibrillar patterns. The remaining 11 lesions (10%) of benign melanocytic nevi showed epiluminescent features that did Figure 4. Videomicroscopic features (magnification ϫ50) of a melanocytic not belong to any of the 3 typical patterns. These were nevus on the sole showing the latticelike pattern. classified as a nontypical pattern. Some lesions with the nontypical pattern were accompanied by diffuse pigmentation, and some of them were congenital nevi. The par- were more likely to have the nontypical pattern than junc- allel ridge pattern characteristic of malignant mela- tional type nevi. In the parallel furrow and the lattice- noma and melanoma in situ was not detected in any like patterns, nests of melanocytes containing melanin benign melanocytic nevi. In addition, the margin of granules were mainly found in the crista profunda limi- the benign melanocytic nevi mostly faded out, and the tans, the rete ridge corresponding to the furrow of the abrupt edge was rare. Peripheral dots were also seldom skin marking (Figure 3). In lesions showing the fibrillar detected. pattern, a proliferation of melanocytes, arranged as The relationship between the epiluminescent pat- solitary cells and in small nests, was observed in the terns and the histopathological subtypes of the nevi is lower epidermis not restricted to any particular rete summarized in Table 5. All 7 lesions of lentigo simplex ridges (Figures 5 and 6). In the nontypical pattern, the showed the parallel furrow pattern. Compound-type nevi proliferation of melanocytes was prominent in the

ARCH DERMATOL / VOL 134, MAY 1998 566

B B

C Figure 5. Clinical (A), videomicroscopic (B, magnification ϫ50), and Figure 6. Clinical (A), videomicroscopic (B, magnification ϫ50), and histopathological (C, magnification ϫ75) features of a melanocytic nevus histopathological (C, magnification ϫ75) features of a melanocytic nevus on the sole showing the fibrillar pattern, subtype A. Delicate filamentous on the sole showing the fibrillar pattern, subtype B. Fibrillar pigmentation is pigmentation slanting the skin markings is clearly observed. arranged in a meshlike manner, producing a characteristic feature.

lower epidermis, arranged as solitary cells and in nests, As in the previous study, some lesions of benign nevus and the nests occasionally bridged the neighboring rete were found whose magnified features did not belong to ridges. Three lesions showed the histopathological fea- any typical pattern, and these were grouped as a “non- tures of so-called dysplastic nevi. Including these 3 le- typical pattern.” Similar findings of melanocytic nevi on sions, all the lesions with the nontypical pattern were glabrous skin have been reported by other investigators judged benign because of their overall symmetry and the using a dermatoscope or a videomicroscope.13 limitation of proliferation of the melanocytes to the lower The most important finding in this study is that all epidermis and because the proliferating melanocytes had but 1 lesion of malignant melanoma in situ exhibited a no distinct nuclear atypia. unique epiluminescent feature, namely the parallel ridge pattern showing accentuated pigmentation on the ridges COMMENT of the skin markings. This feature was not detected in any lesions of benign melanocytic nevus, and thus, the We focused the current epiluminescent microscopic study specificity of the finding was 100%. The sensitivity of the on melanocytic lesions affecting glabrous skin. These ac- finding was also high, being 83% in malignant mela- ral areas are the most prevalent sites of malignant melanoma in situ and 94% in the macular portion of acral len- noma in nonwhite populations. The findings in benign tiginous melanoma. In a previous study,12 a serrated pat- melanocytic nevi elicited in this study are essentially the tern in the macular portions of malignant melanoma of same as those seen in a previous study,12 although the the sole was reported. The serrated pattern probably cor- term used to describe the parallel pattern in the previ- responds to the radial streaming reported in melanoma ous report was changed to the “parallel furrow pattern.” lesions of nonglabrous skin. The serrated pattern is con-

ARCH DERMATOL / VOL 134, MAY 1998 567

©1998 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 sidered to be extensions of the parallel ridge pattern at Accepted for publication November 17, 1997. the margin of the lesions. The histological background This study was supported in part by the Grant-in-Aid of the parallel ridge pattern was a prominent prolifera- for Cancer Research 9-23 from the Ministry of Health and tion of atypical melanocytes in the crista profunda Welfare, Tokyo, the Government of Japan. intermedia. It is still unclear why the melanocytes of ma- Presented in part at the 19th World Congress of lignant melanoma, including the in situ lesions, pre- Dermatology and the Fourth World Conference on Mela- dominantly affect these particular rete ridges of the epi- noma, Sydney, Australia, June 12 and 16, 1997. dermis. The abrupt edge and peripheral dots may also Reprints: Toshiaki Saida, MD, PhD, Department of Der- be important in diagnosing malignant melanoma on gla- matology, Shinshu University School of Medicine, 3-1-1 brous skin, as they were almost exclusively found in le- Asahi, Matsumoto 390-8621, Japan (e-mail: tosaida@gipac sions of malignant melanoma and melanoma in situ. The .shinshu-u.ac.jp). diagnostic significance of these findings has already been 1-11 reported in malignant melanoma of nonglabrous skin. REFERENCES Based on the results of this study, melanocytic lesions on glabrous skin showing predominantly the par- 1. Pehamberger H, Steiner A, Wolff K. In vivo epiluminescence microscopy of pig- allel furrow pattern or the latticelike pattern may be judged mented skin lesions, I: pattern analysis of pigmented skin lesions. J Am Acad to be benign. One lesion of melanoma in situ on the sole Dermatol. 1987;17:571-583. 2. Steiner A, Pehamberger H, Wolff K. In vivo epiluminescence microscopy of pig- showed almost entirely the fibrillar pattern. The fibril- mented skin lesions, II: diagnosis of small pigmented skin lesions and early de- lar pattern observed in this lesion corresponded to sub- tection of malignant melanoma. J Am Acad Dermatol. 1987;17:584-591. type A and could not be distinguished from that found 3. Soyer HP, Smolle J, Hodl S, Pachernegg H, Kerl H. Surface microscopy: a new approach to the diagnosis of cutaneous pigmented tumors. Am J Dermatopa- in the benign nevi. The fibrillar pattern may reflect the thol. 1989;11:1-10. proliferation of melanocytes arranged as solitary cells or 4. Bahmer FA, Fritsch P, Kreusch J, et al. Terminology in surface microscopy. J Am Acad Dermatol. 1990;23:1159-1162. in small nests in the lower epidermis. Further study is 5. Kenet RO, Kang S, Kenet BJ, Fitzpatrick TB, Sober AJ, Barnhill RL. Clinical di- necessary to clarify how to manage the lesions showing agnosis of pigmented lesions using digital epiluminescence microscopy: grad- the fibrillar pattern. Another point to be investigated is ing protocol and atlas. Arch Dermatol. 1993;129:157-174. 6. Schindewolf T, Stolz W, Albert R, Abmayr W, Harms H. Comparison of classifi- the differentiation of benign nevi showing the nontypi- cation rates for conventional and dermatoscopic images of malignant and be- cal pattern from malignant melanoma. This may be pos- nign melanocytic lesions using computerized colour image analysis. Eur J Der- sible because benign nevi showing the nontypical pat- matol. 1993;3:299-303. 7. Pehamberger H, Binder M, Steiner A, Wolff K. In vivo epiluminescence micros- tern are usually devoid of the characteristic parallel ridge copy: improvement of early diagnosis of melanoma. J Invest Dermatol. 1993; pattern, which is found in almost all lesions of malig- 100(suppl 2):356S-362S. 8. Yadav S, Vossaert KAA, Kopf AW, Silverman M, Grin-Jorgensen C. Histopatho- nant melanoma including in situ lesions. We cannot com- logic correlates of structures seen on dermoscopy (epiluminescence micros- pletely exclude the possibility, however, that early le- copy). Am J Dermatopathol. 1993;15:297-305. sions of melanoma in situ that cannot be histologically 9. Stolz W, Braun-Falco O, Bilek P, Landthaler M, Cognetta AB. Color Atlas of Der- matoscopy. Oxford, England: Blackwell Science Publishers; 1994. diagnosed at present show the nontypical pattern. In ad- 10. Stanganelli I, Burroni M, Rafanelli S, Bucchi L. Intraobserver agreement in in- dition, interobserver and intraobserver variability in the terpretation of digital epiluminescence microscopy. J Am Acad Dermatol. 1995; recognition of the characteristic patterns revealed in this 33:584-589. 11. Menzies SW, Ingvar C, McCarthy H. A sensitivity and specificity analysis of the study must be further evaluated. surface microscopy features of invasive melanoma. Melanoma Res. 1996;6: The findings of the present study in which we used 55-62. 12. Saida T, Oguchi S, Ishihara Y. In vivo observation of magnified features of a videomicroscope support the usefulness of epilumi- pigmented lesions on volar skin using video macroscope: usefulness of epilu- nescent microscopy in detecting and diagnosing early ma- minescence techniques in clinical diagnosis. Arch Dermatol. 1995;131: lignant melanoma on glabrous skin. To confirm the va- 298-304. 13. Akasu R, Sugiyama H, Araki M, Ohtake N, Furue M, Tamaki K. Dermatoscopic lidity of this conclusion, however, many more melanocytic and videomicroscopic features of melanocytic plantar nevi. Am J Dermatopa- lesions must be investigated with this method. thol. 1996;18:10-18.

ARCH DERMATOL / VOL 134, MAY 1998 568