Conflicts of interest: none declared.
The limited specificity of histological examination in the diagnosis of dysplastic nevi
Version of Record online: 16 OCT 2008
© Blackwell Munksgaard 2008
Journal of Cutaneous Pathology
Special Issue: Festschrift in Honor of N. Scott McNutt, MD
Volume 35, Issue Supplement s2, pages 20–23, November 2008
How to Cite
Rabkin, M. S. (2008), The limited specificity of histological examination in the diagnosis of dysplastic nevi. Journal of Cutaneous Pathology, 35: 20–23. doi: 10.1111/j.1600-0560.2008.01131.x
- Issue online: 16 OCT 2008
- Version of Record online: 16 OCT 2008
- Accepted for publication July 2, 2008
Dysplastic nevi were originally described as a distinct entity with specific clinical and histological features of importance as direct precursors of malignant melanoma and as markers of patients at increased risk of developing melanoma in the setting of familial melanoma. Nevi with the clinical and histological features described first as ‘B-K moles’ and later as ‘dysplastic nevi’ clearly do exist and do sometimes represent melanoma precursors or melanoma risk markers, but it is now recognized that most dysplastic nevi never progress to melanoma, that the histological features originally described in nevi in familial melanoma patients are poorly correlated with dysplastic nevi as they are defined clinically, and that overlapping or identical histological features are found in a variety of other melanocytic lesions including small (< 5 mm) melanocytic nevi, lentiginous nevi, atypical (dysplastic) lentiginous nevi, lentiginous melanoma, lentigo maligna and nevi in an ever-growing number of ‘special sites’. This article will briefly review the evolution of our understanding of the histological range of nevi and the histological differential diagnosis of dysplastic nevi.
Clark et al.1 reported in 1978 that some members of families with an unusually high incidence of familial malignant melanoma exhibited an association between the development of melanoma and the presence of melanocytic nevi that these investigators believed to have unique clinical and histological features. Clark et al. reported that these nevi could serve as both a marker for patients at increased risk of developing melanoma and a direct melanoma precursor. The relationship between these unusual nevi and melanoma was reported in two familial melanoma kindreds, and the nevi were initially named after the initials of the index families under the designation ‘B-K mole’. In 1980, Greene et al.2 suggested that the lesions be termed ‘dysplastic nevi’ because these nevi were believed to potentially represent an intermediate step in the development of melanoma and to have clinical, architectural and cytological features that deviated from those of normal nevi. The significance of these nevi and appropriate nomenclature have been and continue to be very controversial.3 The nevi described by Clark et al. in their 1978 report and subsequent reports will be referred to as dysplastic nevi for purposes of this article.
Individual dysplastic nevi have been reported to have some or all the clinical features of malignant melanoma, but with atypical features often less pronounced than in melanoma.4 Individuals in the melanoma-prone kindreds were classified as clinically affected with dysplastic nevi if they had multiple lesions at least 5 mm in diameter with a flat component and at least two of the following three characteristics: variable pigmentation, irregular asymmetric outline and indistinct borders.5 There is imperfect agreement between observers in the clinical assessment of dysplastic nevi,6 and even with the most advanced techniques of clinical examination currently available, dysplastic nevi cannot be distinguished with certainty from melanoma or histologically unremarkable nevi.7 Reported histological features of dysplastic nevi include specific architectural and cytological features and a characteristic inflammatory and fibrotic host response.1,5,8,41 Refy1 Dysplastic nevi are either entirely junctional or they contain a junctional component, commonly present as a ‘shoulder’ adjacent to a compound papular ‘head’. The junctional component displays an increased number of single and nested melanocytes with nests predominating, melanocyte nests are commonly found near the tips and sides of elongated rete, and often some nests bridge adjacent rete. The rete ridges tend to be uniformly and irregularly elongated. In contrast to the predominance of moderately to severely cytologically atypical melanocytes usually seen in melanoma, in dysplastic nevi, a minority of cells exhibits cytological atypia (‘random’ cytological atypia).
The histology of dysplastic nevi was originally thought to be specific, with histological examination used to confirm a clinical diagnosis of dysplastic nevus not just in melanoma-prone families with numerous atypical nevi but also in patients with only one or a few nevi.9 Clark et al. did not study biopsies of control populations, and they did not study the histological features of common nevi with a diameter smaller than 5 mm. Other investigators subsequently reported that many nevi from control populations showed histological features similar to those described by Clark et al. in dysplastic nevi.10,11 It has become clear since the original work of Clark et al. that many small melanocytic nevi (nevi < 5 mm in diameter) show architectural features resembling those of dysplastic nevi, including nevi from members of the general population. Cytological atypia in these small nevi is often minimal or mild, but a significant proportion of small nevi has been reported to show architectural and cytological features indistinguishable from those of classic dysplastic nevi (Braun-Falco et al. 2003).12,13,43
Currently available data suggest that acquired melanocytic nevi of every size may exhibit a continuum of clinical and histological atypia, in contrast with the dichotomy between ‘dysplastic’ and ‘benign’ nevi originally proposed by Clark et al.42 There does appear to be some correlation between increasing degrees of clinical and histological atypia, but this correlation is poor, especially in small nevi.12,13 Histological study of small nevi and nevi from control populations have led to the suggestion that a histological diagnosis of dysplastic nevus should only be applied to relatively large nevi (> 5–6 mm).8,14,15 This suggestion is consistent with clinical studies showing that clinically atypical nevi > 5 mm in diameter are markers of increased risk of melanoma.4,16 Limiting the diagnosis of dysplastic nevus to large nevi undoubtedly will increase the specificity of the diagnosis, but with a corresponding loss of sensitivity. In addition to the correlation between large atypical nevi and melanoma risk, there is an independent correlation between the melanoma risk and the number of small diameter nevi.17,18 Small clinically atypical nevi are commonly found in familial melanoma patients. When they first appear, nevi that will eventually become clinically dysplastic often are clinically indistinguishable from nevi that will become ordinary compound or dermal nevi.5 Several recent studies have suggested that nevi showing moderate or high-grade cytological atypia may be somewhat over-represented among patients at increased risk for malignant melanoma.19,20 However, advanced cytological atypia is relatively rare even in clinically atypical nevi in melanoma patients and melanoma-prone populations. Small diameter nevi and nevi with minimal or mild histological atypia are much more common (if less specific) in these populations.
Thanks to sophisticated techniques of clinical evaluation such as whole body photography, epiluminescence microscopy and digital dermoscopy, and close clinical follow up and frequent biopsies of high-risk populations such as patients with familial melanoma and dysplastic nevus syndrome, it is now recognized that small clinical lesions may show histologically unequivocal in situ and invasive melanoma, and indeed, small diameter melanomas have metastasized.21,22 Biopsy of many small melanocytic lesions to accomplish the diagnosis and surgical extirpation of melanoma at an early, low-risk stage has inevitably led to biopsy of many small nevi, as small nevi are very common in the general population, and in melanoma patients, and nevi cannot be separated from melanoma with certainty on the basis of clinical examination.7 As noted previously, small nevi from control populations may show considerable histological overlap with classic dysplastic nevi,10,11 and (except for size) small nevi may be histologically indistinguishable from classic large dysplastic nevi (Braun-Falco et al. 2003).12,13 Many small nevi have prominent lentiginous features including elongated rete, increased numbers of non-nested junctional melanocytes, a non-nested peripheral or shoulder component, poorly formed junctional melanocyte nests and dermal fibrosis and inflammation. Small nevi with these features appear to be very common in the general population. They have been reported under a variety of names including lentiginous nevus, incipient nevus and jentigo.23,24
Atypical (dysplastic) lentiginous nevi25,26 are relatively newly described lesions that show overlap with both lentiginous nevi and dysplastic nevi. Atypical lentiginous nevi may be seen as isolated lesions, they may merge with lesions indistinguishable from lentigo maligna and malignant melanoma in situ and they may be associated with invasive nevoid malignant melanoma. Atypical lentiginous nevi commonly occur in heavily sun-damaged skin of the elderly. In contrast to lentigo maligna, which is classically seen primarily on the head and neck, atypical lentiginous nevi usually present on the trunk and limbs. In contrast to lentiginous nevi, atypical lentiginous nevi show an irregular rete ridge system, nests and single melanocytes irregularly distributed along rete, focal confluent growth of melanocytes in the suprapapillary plates and focal cytological atypia. Adding further difficulty to the difficulty of specifically diagnosing dysplastic nevi in elderly individuals and in sun-damaged skin, Farrahi et al.27 recently reported that intralesional heterogeneity and histological overlap with dysplastic nevi are substantial in many cases of lentigo maligna, with many cases of clinically classic lentigo maligna showing focal or widespread dysplastic nevus-like features such as elongated rete and bridging of rete by nests of melanocytes.
Cases reported as lentiginous melanoma28,29 and de novo intra-epidermal epithelioid melanocytic dysplasia24,30 are newly described lesions that have many features in common with atypical lentiginous nevi. Lesions classified in the literature by these three diagnoses may comprise identical or overlapping biological processes.29 Lentiginous melanoma shows tremendous histological overlap with common lentiginous nevi. Pagetoid spread in lentiginous melanoma is very focal, and cytological atypia usually is not severe. Partial biopsies of the three index cases of lentiginous melanoma reported by King et al. were originally interpreted as lentiginous nevi and therefore not excised immediately following the initial biopsy. Persistence and progressive enlargement of the index cases eventually led to complete excision, and only in the excisional specimens did the malignant nature of the process became more clear. Lentiginous melanoma is currently defined histologically by confluent growth of atypical melanocytes across the entirety of a large (> 1 cm.) lesion, so only large or excisional specimens allow for confident distinction between lentiginous melanoma, common lentiginous nevi and atypical melanocytic hyperplasia in heavily sun-damaged skin. Two of the 16 excised specimens reported by King et al. showed focal level II dermal invasion. One of the five cases of lentiginous melanoma reported by Davis and Zembowicz was initially biopsied and interpreted clinically and histologically as a lentigo, and the lesion was followed clinically and biopsied on two other occasions before being totally excised 16 years after initial presentation. The excisional specimen of this lesion showed unequivocal focal invasive malignant melanoma.
The 82 biopsies reported as de novo intra-epidermal epithelioid melanocytic dysplasia by Sachdeva et al. have histological similarities to the cases of King et al. and Davis and Zembowicz. Severe cytological atypia and a high density of pagetoid melanocytes are not seen in either entity. Melanocyte density may be lower and nesting less prominent in the cases reported by Sachdeva et al. This entity seems to have a significant association with dysplastic nevus syndrome and melanoma risk, as over half of the patients reported by Sachdeva et al. had clinical and in most cases histological evidence of dysplastic nevus syndrome, 5 of 16 patients (33%) had a family history of melanoma, and 20 of 75 (26.6%) had a personal history of melanoma (including three patients having more than one melanoma). Eight of their 82 specimens showed fully evolved malignant melanoma contiguous to the melanocytic dysplasia, and this dysplasia was thought to represent a precursor to the melanoma.
Congenital nevi31 and nevi in an ever-growing list of ‘special sites’ including genital,24,32 acral,33,34 embryonic milk line/flexural (axillary, inguinal and umbilical),24,35 breast,36 scalp of adolescents37,38 and external ear39,40 commonly show atypical histological features that have similarities to melanoma and dysplastic nevi and differ from those of stereotypical ‘benign’ nevi. Atypical features may include asymmetry, origin of melanocyte nests at the sides of rete, melanocyte nests that vary in size and shape and are sometimes quite large, poor circumscription, a nested and lentiginous pattern, upward melanocyte spread, widespread or focal melanocyte cytological atypia and an inflammatory and fibrotic host response. Differential diagnosis may be quite challenging, as melanoma and bona fide dysplastic nevi may occur in these locations. While these lesions often lack the lentiginous pattern of regular rete ridge elongation classically seen in dysplastic nevi, there is enough variability in the appearance of some congenital nevi, special site nevi and dysplastic nevi that they often cannot be distinguished with confidence, especially on partial biopsies (Elder 2006).
In summary, while the histology of the large, clinically atypical nevi seen in familial melanoma patients was originally believed to be specific, it is now clear that there is considerable histological overlap between dysplastic nevi in this setting and many nevi that do not appear to be associated with increased melanoma risk or with a familial or known genetic predisposition to develop melanoma. Many small nevi show architectural features of dysplastic nevi with minimal cytological atypia, but small nevi also may show complete histological overlap with classic dysplastic nevi, many clinically atypical nevi and nevi in melanoma-prone patients show minimal cytological atypia, and it may be difficult or impossible to histologically distinguish dysplastic nevi from a variety of melanocytic lesions including lentigo maligna, atypical lentiginous nevi, lentiginous melanoma and nevi of special sites, especially on partial biopsies.
- 23The lives of lesions. Chronology in dermatopathology. New York: Masson Publishing, 1984., .
- 24The melanocytic proliferations. A comprehensive textbook of pigmented lesions. New York: Wiley-Liss, 2001., , .