Glatz, Katharina MD*; Hartmann, Christoph MD*; Antic, Milos MD*; Kutzner, Heinz MD†
The presence and distribution of mitotic figures is mentioned as an important discriminatory parameter in the assessment of melanocytic lesions in most textbooks of dermatopathology. However, no references are indicated to corroborate statements concerning proliferative activity in banal nevi which is supposed to be low or absent in most of these lesions. It follows, that those statements are mainly based on the personal experience of the book authors but not on systematic studies. For example, McKee and Calonje state in their Diagnostic Atlas of Melanocytic Pathology that “mitotic figures are rare in banal nevi although if sufficient sections and levels are examined they will almost always be identified.”1 There are only 2 recent publications on mitoses in benign melanocytic nevi with only 157 and 48 samples, respectively.2,3 In past publications on the subject, proliferative activity in melanocytic nevi has mostly been determined by immunohistochemical evaluation of proliferation markers, for example, Ki-67.4-7 We have recently shown that the application of the commercially available mitotic markers Phospho-Histone H3 Ser28 (PHH3) and MPM2 improves the efficiency and reproducibility of mitotic counting in various tumor types.8 Both markers can be assessed immunohistochemically on routinely processed formalin-fixed, paraffin-embedded tissue specimens. For this study, PHH3 and MPM2 were used in addition to traditional hematoxylin and eosin (H&E) staining to evaluate the occurrence and distribution of mitotic figures in a random sample of 353 benign melanocytic nevi.
MATERIALS AND METHODS
Paraffin-embedded blocks of 353 consecutively excised benign melanocytic nevi of 206 female and 147 male patients were retrieved from the routine files at Dermatopathologie, Friedrichshafen. The mean age of the patients was 34.3 years (range 2-78 years). In 8 cases, the referring clinician suspected a melanoma. Histologically, these clinically suspicious lesions were undoubtedly benign (1 Spitz nevus, 2 cellular blue nevi, and 5 common nevi). All except for 12 lesions had been excised between October 1 and March 31. Follow-up times for the 353 patients ranged from 16 to 65 months (mean 42.2 ± 23.1). No recurrences were reported during the follow-up period.
The 353 excisional specimens included junctional common nevi (3), compound common nevi (147), dermal common nevi (85), junctional dysplastic nevi (3), compound dysplastic nevi (68), compound Sutton nevi (7), junctional Spitz nevi (4), compound Spitz nevi (2), dermal Spitz nevi (3), atypical compound Spitz nevi (2), junctional Reed nevus (1), compound Reed nevi (3), blue nevi (13), cellular blue nevi (5), combined compound nevi (4), and combined dermal nevi (3).
In all cases, the first and the senior author of this study could establish a consensus diagnosis using well-established histopathologic criteria. All dysplastic nevi had mild to moderate cytologic and architectural atypia based on the criteria published by Barnhill et al.9
The nevi had a mean diameter of 5.5 mm (range 2-18 mm) and a mean dermal surface area (compound and dermal nevi) of 11.5 mm2 (range 0.05-150 mm2).
Solar elastosis was graded semiquantitatively as being absent (274), mild (64) or moderate to severe (15). Common nevi with severe destructive intra- and peritumoral inflammation were designated Sutton nevi (21). Mild peritumoral inflammation was present in 74 samples and absent in 258. Traumatization was considered being present if there was a serocrust, erosion or ulceration of the overlying epidermis accompanied by an inflammatory infiltrate and fibrin deposits in the edematous papillary dermis. Trauma was present in 11/353 lesions.
Congenital features10 were present in 127 and a polypoid or verruccous exophytic architecture in 85 of the 300 common and dysplastic nevi with a dermal component.
Counting of Mitotic Figures
Mitotic figures in the H&E and the immunostained slides were assessed as previously described in detail.8 Briefly, mitotic figures from the beginning of metaphase to telophase were counted at high power in 3 consecutive step sections stained with hematoxylin and eosin, anti-PHH3, and anti-MPM2, respectively. The location of mitotic nevus cells was recorded as being either in the epidermis, the upper half or the lower half of the dermis. In addition, dermal mitotic activity was determined per mm2 hotspot (5 consecutive high power fields with a Zeiss Axioskop microscope) in the area of highest mitotic activity. Mitotic counts were performed without knowledge of the results of the other stainings or of clinical information.
Determination of Dermal Surface Area
The dermal surface area of compound and dermal nevi was determined by using the contour tool from the measurement menu of Carl Zeiss MicroImaging GmbH Axio Vision software version 188.8.131.52.
Tissues were fixed in 4% buffered formalin immediately after excision to prevent an impact of delayed fixation on mitotic counts. After fixation, the specimens were routinely processed, and embedded in paraffin. Four-micron thick sections were stained with H&E. In addition, a 4-micron consecutive serial section was immunostained with the commercially available antibodies polyclonal rabbit Anti-phospho-Histone H3 (Ser28) (PHH3, 1:2000, Upstate Biotechnology, Lake Placid, NY; manufacturer No. 07-145) and monoclonal mouse Anti-MPM2 (MPM2, 1:5000, DCS Innovative Diagnostik-Systeme, Hamburg, Germany; manufacturer No. MI792C002). Antigen retrieval was performed by incubating the tissue in citrate buffer at pH 6.1 in an open water bath without pronase. Staining was done by standard procedure using the labeled streptavidin biotin method on a Dako TechMate 500.
PHH3- and MPM2-positive mitotic figures were counted in the same way as in H&E-stained sections and indicated as number of labeled mitoses per lesion and per square millimeter. Positively stained interphase nuclei in the absence of an obvious prophase chromosomal organization were disregarded. MPM2 positive cytoplasm was only accepted as mitotic figure if condensed chromosomes were recognizable in the same cell.
Statistical analyses were performed using the χ2 test. Calculations were done with JMP software. Results were considered significant at a P value of P < 0.05.
In all 3 stainings nevi with at least one mitosis or more than one mitotic figure per lesion were significantly more frequent in the youngest age group (0-20 years) than in patients older than 50 years (Fig. 1). The difference was most marked in the PHH3 staining with 26.2% mitotically active nevi in the youngest age group versus 2.9% in the oldest age group (P < 0.0001).
Staining Technique and Nevus Subtype
At least one mitotic figure was present in 19.5%, 31.3%, and 42.8% of H&E-, PHH3-, and MPM2-stained lesions. Mitotic activity was particularly frequent in Spitz nevi (Fig. 2), compound nevi, and cellular blue nevi. Mitotic figures were present in 19.7 (H&E) to 44.2% (MPM2) of all banal compound nevi. The percentage of mitotically active nevi per subtype and staining is indicated in Table 1. The 5 nevi with the highest absolute number of mitotic figures comprised 4 Spitz nevi and a polypoid compound nevus (Fig. 3). Most mitotic figures were found in a compound Spitz nevus on the volar lower arm of a 2-year-old boy (H&E: 11; PHH3: 29; MPM2: 26) and in a traumatized polypoid compound nevus from the breast of a 40-year-old man (H&E: 11; PHH3: 26; MPM2: 24).
Mitotic Activity in Association With Selected Morphologic Findings
The only morphologic findings that were associated with a higher frequency of dermal mitoses were signs of traumatization (Table 2) and an exophytic polypoid or verruccous architecture. Of 11 traumatized nevi 6 showed mitotic activity in the H&E staining and 9 in the MPM2 staining.
The percentage of nevi with more than one dermal mitosis was higher for exophytic lesions in all 3 stainings but this difference was only significant in the MPM2 staining (15.3% nonexophytic vs. 27% exophytic nevi with more than one dermal mitosis; P = 0.019).
Severity of solar elastosis, and the presence or absence of congenital morphology had no impact on the presence or absence of mitotic figures. Nevi with severe inflammation more often showed mitotic figures than those without inflammation in the immunohistochemical stainings but not in the H&E staining. However, the difference was not significant and some of the mitotically active cells in the immunohistochemical stainings probably represented inflammatory cells which were difficult to discriminate morphologically from nevus cells within a dense inflammatory infiltrate. Mitotic figures in inflammatory cells were present in 42.2% and 35.4% of the lesions with inflammation in the PHH3 and the MPM2 staining, respectively. The highest absolute number of mitotically active inflammatory cells (n = 50) was counted in the PHH3 staining of the polypoid compound nevus shown in Figure 3.
Junctional and Dermal Mitoses
Compound nevi more often showed at least one mitotic figure than nevi of the junctional or dermal type. Mitoses were more frequently present in the dermal than in the junctional portion (Table 3). If the junctional nevomelanocytic nests were small, discrimination of mitotic figures in nevus cells from mitotic figures in keratinocytes was difficult (Fig. 4).
In the dermis, mitotic activity was roughly 3 times as frequent in the upper than in the lower half of the dermis (Fig. 5, Table 3). Clusters of mitotic figures were absent in all of the 45 nevi with more than one dermal mitosis. The highest mean number of dermal mitotic figures per mm2 surface area was observed in Spitz nevi (H&E: 0.175/mm2) and the lowest in blue nevi (H&E: 0.006/mm2). Dermal mitotic figures were more abundant in common compound nevi (H&E: 0.024/mm2) than in dysplastic compound nevi (H&E: 0.018/mm2) and in common dermal nevi (H&E: 0.009/mm2), respectively. The mean number of dermal mitoses per mm2 of the whole dermal surface area in different nevus subtypes is shown in Table 4.
The presence of mitotic figures depended on the size of the dermal component which was 37.5% larger in mitotically active nevi as compared with mitotically inactive nevi.
Nevi with mitotic figures in the lower half of the dermal component could be detected by all three staining methods (H&E 11/342; PHH3 26/342; MPM2 36/342). The highest number of deep dermal mitoses (n = 11; 2/mm2) was found in the MPM2 staining of a cellular blue nevus from the buttocks of an 8-year-old girl. The mitotic figures were evenly distributed within the dermis and other worrisome signs as cellular atypia or necrosis were absent.
Using immunohistochemical markers for enhanced detection of mitotic figures our study revealed that brisk and deep dermal mitotic activity in melanocytic lesions is not restricted to malignant melanomas but can be observed in a considerable number of otherwise banal melanocytic nevi. Mitotic rate has been extensively studied in malignant melanomas11,12 because of its paramount prognostic relevance. Attis and Vollmer indicated a mean mitotic rate of 3.2/mm2 (range 0-92) in 1268 primary cutaneous melanomas.11 In our study, the mean dermal mitotic rate per mm2 in different nevus subtypes (H&E staining) was considerably lower (from 0.006/mm2 in blue nevi to 0.024/mm2 in common compound nevi to 0.175/mm2 in Spitz nevi). As the range of mitotic activity is high, individual cases of both benign and malignant melanocytic lesions may considerably deviate from this mean. In a study on the mitotic rate in primary cutaneous melanomas, Azzola et al12 found a complete absence of dermal mitotic figures in 22% of the 3661 melanomas analyzed. It follows that this important diagnostic criterion shows considerable overlap in benign and malignant melanocytic lesions.
The percentage of 19.5% banal melanocytic nevi with mitotic figures in the conventional H&E staining detected by us was considerably higher than in previously published studies on the subject.2,3,13,14 The percentage of nevi harboring at least one mitosis was even higher in the more sensitive PHH3 (31.3%) and MPM2 (42.8%) immunohistochemical stainings. In 2 recent studies, Jensen et al2 found mitotic figures in 4% of 157 shaved or punch biopsies of conventional nevi, whereas Nasr et al3 detected none in 40 compound and dysplastic nevi stained by PHH3. The lower rate of mitotic activity compared with our findings may in part be due to the fact that Jensen et al. had assessed shave or punch biopsies and not excisional biopsies comprising the whole lesion. Nasr et al had only counted 10 high-power fields while we had examined the whole lesion measuring 57.5 high-power fields on average. In both studies, the size of the examined lesions is not mentioned. As the number of mitotic figures is higher in larger lesions, the rather high mean dermal surface area (11.5 mm2) of our study samples may have contributed to the high percentage of mitotically active nevi. In addition, we had considered all mitotic phases from beginning metaphase to telophase while a precise description of what was considered a mitotic figure is lacking in the 2 other studies.
We found mitotically active nevi to be most prevalent in the youngest age group. The percentage of nevi showing any mitotic activity was nearly constant between the age of 21 and 50 and then decreased in patients 50 years or older. With the H&E staining nevi with more than one mitosis could be detected in only 1.4% of the lesions of patients aged more than 50 years but in 10.0% of patients younger than 20 years. As the natural history of nevi is one of greatest growth by the third decade, and gradual regression of lesions with increasing age, this result is not unexpected.1 Polypoid or verruccous configuration of the nevus and signs of traumatization were associated with higher mitotic activity. Mechanical irritation of exophytic polypoid nevi may stimulate proliferative activity. The association of trauma and mitotic activity has been reported before.15,16
Massi and Leboit state in their textbook that in heavily inflamed Sutton nevi the presence of mitoses in numbers, mitoses clustered together or in the deep part of the lesion were an important clue in favor of melanoma.10 We found mitotic figures in half of our 21 Sutton nevi. In some lesions, they were very numerous (up to 32 in the MPM2 staining) but never in clusters. Mitoses in the deeper half of the dermis were present in 2 of the H&E, 4 of the PHH3, and 6 of the MPM2-stained innocent looking lesions. Altogether, mitoses in Sutton nevi are not uncommon even in deeper parts of the lesion but occurrence of clusters should raise suspicion. Massi and Leboit10 also warn that few lymphocytes can commonly be seen in mitosis in a dense infiltrate.10 We had separately analyzed mitotic figures in nevomelanocytes and inflammatory cells. Mitotic figures in inflammatory cells were present in 35.4% of the PHH3 and 42.2% of the MPM2-stained sections, respectively. Mitotic figures were more frequent in heavily inflamed lesions in which proliferatively active nevomelanocytes and inflammatory cells were intermixed and could hardly be distinguished from each other. Under these circumstances, the number of mitotic figures in nevomelanocytes may be overestimated by the application of immunohistochemistry. In H&E-stained sections, mitoses in inflammatory cells were much less obvious.
Recent ultraviolet irradiation has been shown to influence the proliferative rate of nevi.17 We cannot exclude an impact of recent ultraviolet irradiation on the number of mitotic figures in our cohort. However, as all except for 12 nevi had been excised during the winter half year this impact is probably minimal.
Not only mitotic counts but also distribution of mitotic figures are important in the differential diagnosis of benign and malignant melanocytic tumors.18 In malignant melanomas, the distribution of proliferating cells is often highly heterogeneous.19 In our subset of nevi with high mitotic activity, the mitotic figures were evenly distributed never forming clusters.
Mitotic activity was more frequently present in the dermis than in the junctional component. Within the dermal compartment mitotic figures were roughly 3 times as frequent in the upper half than in the lower half of the dermis. Occasional deep dermal mitoses can be expected in banal benign common nevi particularly in those with a polypous architecture. However, if more than one deep dermal mitosis is present in conventional H&E-staining a diagnosis of a benign common nevus must be carefully reconsidered. Common nevi of the compound type featured more dermal mitoses per mm2 than common dermal nevi supporting the concept of an evolutionary process of moles from junctional to compound and eventually to purely intradermal lesions.
Immunohistochemical detection of mitotic figures is usually not needed for the diagnosis of banal melanocytic nevi. However, mitotic markers may be a very helpful tool for the evaluation of suspected nevoid melanomas, spitzoid lesions, and cellular blue nevi as they give a rapid overall impression of the spacial distribution and the approximate number of mitotic figures at a low magnification. In heavily pigmented lesions, mitotic figures can easily be overlooked in the H&E staining but are well detectable by immunohistochemistry.
Due to the higher sensitivity of immunohistochemistry for mitotic figures, higher thresholds for PHH3 and especially MPM2 have to be applied.
In our study we could show that mitotic activity in obviously benign melanocytic nevi is not rare and occasional mitotic figures can even be detected in the deep dermal part.
More than 2 mitotic figures per lesion are rare and can usually be explained either by the nevus subtype (eg, Spitz nevus), mechanical irritation (exophytic architecture), traumatization, or inflammation. Many anecdotal statements were confirmed by our study such as the high mitotic activity in Spitz nevi in comparison to other subtypes, higher numbers of mitoses in larger nevi, nevi with inflammation or signs of trauma, and nevi of younger patients. Atypical mitoses or clusters of mitotic figures were not found in any of our 353 lesions and should therefore be considered an ominous sign. More than one mitotic figure in a melanocytic tumor of an elderly patient or in skin with marked solar elastosis also points to malignant melanoma as brisk mitotic activity is very rare in this setting.
The detection of evenly distributed mitotic figures in a melanocytic tumor even in a deep dermal location per se is not sufficient for the diagnosis of malignant melanoma. However, in combination with other signs of malignancy, especially nuclear pleomorphism or atypia, mitoses are an important additional clue of malignancy.10,15
1. McKee PH, Calonje E. Diagnostic Atlas of Melanocytic Pathology. Oxford: Elsevier; 2009.
3. Nasr MR, El-Zammar O. Comparison of pHH3, Ki-67, and surviving immunoreactivity in benign and malignant melanocytic lesions. Am J Dermatopathol. 2008;30:117-122.
2. Jensen SL, Radfar A, Bhawan J. Mitoses in conventional melanocytic nevi. J Cutan Pathol. 2007;34:713-715.
4. Bergman R, Malkin L, Sabo E, et al. MIB-1 monoclonal antibody to determine proliferative activity of Ki-67 antigen as an adjunct to the histopathologic differential diagnosis of Spitz nevi. J Am Acad Dermatol. 2001;44:500-504.
5. Kanter L, Blegen H, Wejde J, et al. Utility of a proliferation marker in distinguishing between benign naevocellular naevi and naevocellular naevus-like lesions with malignant properties. Melanoma Res. 1995;5:345-350.
6. Smolle J, Soyer HP, Kerl H. Proliferative activity of cutaneous melanocytic tumors defined by Ki-67 monoclonal antibody. A quantitative immunohistochemical study. Am J Dermatopathol. 1989;11:301-307.
7. Stefanaki C, Stefanaki K, Antoniou C, et al. G1 cell cycle regulators in congenital melanocytic nevi. Comparison with acquired nevi and melanomas. J Cutan Pathol. 2008;35:799-808.
8. Tapia C, Kutzner H, Mentzel T, et al. Two mitosis-specific antibodies, MPM-2 and phospho-histone H3 (Ser28), allow rapid and precise determination of mitotic activity. Am J Surg Pathol. 2006;30:83-89.
9. Barnhill RL, Piepkorn M, Busam KJ. Pathology of Melanocytic Nevi and Malignant Melanoma. New York: Springer Verlag; 2004.
10. Massi G, LeBoit PE. Histological Diagnosis of Nevi and Melanoma. Darmstadt: Steinkopff Verlag; 2004.
11. Attis MG, Vollmer RT. Mitotic rate in melanoma: a reexamination. Am J Clin Pathol. 2007;127:380-384.
12. Azzola MF, Shaw HM, Thompson JF, et al. Tumor mitotic rate is a more powerful prognostic indicator than ulceration in patients with primary cutaneous melanoma: an analysis of 3661 patients from a single center. Cancer. 2003;97:1488-1498.
13. Lund HZ, Stobbe GD. The natural history of the pigmented nevus; factors of age and anatomic location. Am J Pathol. 1949;25:1117-1155, incl 1114 pl.
14. Stegmaier OC Jr, Montgomery H. Histopathologic studies of pigmented nevi in children. J Invest Dermatol. 1953;20:51-64.
15. Culpepper KS, Granter SR, McKee PH. My approach to atypical melanocytic lesions. J Clin Pathol. 2004;57:1121-1131.
16. Strungs I. Common and uncommon variants of melanocytic naevi. Pathology. 2004;36:396-403.
17. Tronnier M, Rudolph P, Koser T, et al. One single erythemagenic UV irradiation is more effective in increasing the proliferative activity of melanocytes in melanocytic naevi compared with fractionally applied high doses. Br J Dermatol. 1997;137:534-539.
18. Okun MR, Edelstein LM, Kasznica J, et al. What criteria reliably distinguish melanoma from benign melanocytic lesions? Histopathology. 2000;37:464-472.
19. Rudolph P, Lappe T, Schubert C, et al. Diagnostic assessment of two novel proliferation-specific antigens in benign and malignant melanocytic lesions. Am J Pathol. 1995;147:1615-1625.
© 2010 Lippincott Williams & Wilkins, Inc.