Case 1 -
Melanoma with regression
Knoxville Dermatopathology Laboratory
Virtual Slides as well as Still Images are displayed below.
For the fastest viewing of virtual slides, click:
under each thumbnail image below. You must have Aperio ImageScope installed on your PC.
Or, click on slide thumbnail images to view each slide
If you do not already have Aperio ImageScope, Windows users with administrator privileges may download and install a free version in order to view USCAP Virtual Slides. Click the icon on the right to get your free copy: ||
in a Web-based slide viewer, which is somewhat slower.
If you have any difficulties viewing these slides, email or call George Clay at +1.724.449.1137.
Case 1. 63 Year old male presented to his dermatologist with a pigmented lesion on
his right posterior shoulder. The patient was unaware of the duration of the lesion and has a history of
multiple benign moles removed from his back. His wife noticed this lesion and is not sure if this was a
site of previous biopsy.
Case 2. A 42 year old female presented with a pigmented lesion on her left back/flank area. The
patient stated to her dermatologist that she had noticed the lesion changing. the patient had a history
of multiple benign moles removed from her trunk and abdomen.
Pathological/Microscopic Findings and any Immunohistochemical or Other Studies:
Additional biopsy demonstrates a broad atypical melanocytic proliferation spanning the entire
dermoepidermal junction (Figure A). In one half of the biopsy there is a markedly atypical confluent
proliferation of melanocytes with associated epidermal hyperplasia and no evidence of dermal scar (Figure
B). In the opposite half of the biopsy, there is effacement of the retiform epidermis with proliferation
of an atypical compound melanocytic proliferation with associated dermal fibrosis and lymphocytic
infiltrate (Figure C and D).
Case 1 - Slide 1
Figure 1. Histologic patterns of recurrent nevus phenomenon. In the histologic patterns types 1and 2, there is effacement of the retiform epidermis with junctional (a) and compound melanocytic hyperplasia (b), respectively. Dermal scar accompanies both biopsies. This is in contrast to types 3 and 4, where there is retiform epidermal hyperplasia, associated with junctional (c) and compound melanocytic hyperplasia (d), and dermal scar.
Figure 2. (a) In 20 cases, prominent pagetoid upward migration of melanocytes was present. This may be confused with melanoma in situ, if taken out of context. (b) Cytologically, the dominant melanocyte present had an epithelioid configuration with round nucleus and even chromatin pattern. The dermal melanocytes retained the morphology of the junctional component, and transition to types B and C melanocytes was not evident (b).
Figure 3. (a and b). In those cases with a retiform epidermis and accompanying atypical features, such as confluent growth pattern, pagetoid spread, and cytologic atypia, there was histologic overlap with primary melanoma (a and b).
Figure 4. Histologic overlap of recurrent nevus and melanoma with regression. Recurrent nevus (a-c) and melanoma with regression (d-f). In partial biopsies, there may be considerable overlap between these two lesions. This was especially true for cases of MM with late regression with loss of tumor, effacement of the retiform epidermis, and fibrosis.
1) Melanoma with regression
2) Recurrent nevus phenomenon
3) Atypical nevus with regression
Melanoma with regression
This case is intended to highlight the difficulty in distinguishing between
recurrent nevus phenomenon and melanoma with regression. In the submitted slide, the biopsy shows
effacement of the retiform epidermis with an atypical junctional melanocytic proliferation and associated
dermal fibrosis/scar. These histologic findings may be seen in both recurrent nevus as well as melanoma
with regression. The term "pseudomelanoma" has been used to describe recurrent nevus simulating
melanoma, further indicating the histologic overlap that may occur between these two entities. The
following discussion highlights the clinical findings and histologic spectrum of recurrent nevus
phenomenon and compares this to melanoma with regression.
Review of the Literature/Treatment Options (if applicable):
We recently analysed 357 cases of recurrent nevus and only those cases in which the clinical history,
original melanocytic lesion, and recurrent lesion were available for review, were included in the study
The original melanocytic lesion and recurrent nevus hematoxylin and eosin stained slides were reviewed
and the following criteria were used to assess recurrent nevi: symmetry, presence/absence of effacement
of the retiform epidermis, junctional and/or dermal melanocytic hyperplasia, growth pattern, relationship
of the melanocytic proliferation to the dermal scar, cytologic atypia of melanocytes, and presence /
absence of residual nevus.
For comparison, 34 cases of unequivocal melanoma with regression were obtained from the KDL melanoma
database, and using criteria generated Kang et al  , regression was defined as:
Early: dense lymphoid infiltrates replacing nests of melanocytes;
Intermediate: absence/loss of tumor with replacement by mix of lymphocytes and melanophages and early
Late: tumor absence with extensive fibrosis and telangiectasia, macrophages and epidermal effacement
Original Nevus Histological Data
The primary melanocytic neoplasms covered a broad spectrum of lesions: the most common type was
characterized as ordinary nevi (64.29%), this included junctional (12 cases, 5.3%), dermal (35 cases,
15.5%), and compound (178 cases, 79%). Dysplastic nevi accounted for 27.43% of all cases, congenital
nevi 6.29%, and specialized nevi (which included blue nevi and spindle cell nevi) accounted for 1.14% of
cases. Three cases of malignant melanoma where also identified (0.86%) (Table 1).
Recurrent nevus Clinical Data
325 patients fulfilled our criteria and were included in this study. There was a female predominance
with females accounting for 71.69%, and males 28.31%. Patients ranged in age from 7 to 92 years; with
the average age being 32 years and the median age 30 years (Table 2). The vast majority of recurrences
(73.54%) occurred in patients under the age of 40 years. By far the most common site of recurrence was
the back, followed by the abdomen and extremities (Table 3). The time to recur ranged from 1 to 63
months, with the average time to recur 8.04 months and the median 5 months (table 4). Most of the nevi
(64%) recurred within six months (Table 4). In addition, 23 patients had more than one recurrent nevus
occurring at separate locations and 6 patients had multiple recurrences of the same lesion, resulting in
a total of 357 biopsies. There were 88 cases (25.14%) in which the primary lesion appeared to be
Recurrent Nevus Histological Data
Repeatable histologic patterns were identified and these were primarily based on A) epidermal changes
i.e. retention or effacement of retiform epidermis; and B) junctional versus junctional and dermal
melanocytic hyperplasia. Based on these criteria, we were able to broadly classify our lesions into four
categories and calculate their relative frequencies:
Type 1: junctional melanocytic hyperplasia with effacement of the retiform epidermis- 49.02%
Type 2: compound melanocytic hyperplasia with effacement of the retiform epidermis-36.85%
Type 3: junctional melanocytic hyperplasia with retention of the retiform epidermis- 5.04%
Type 4: compound melanocytic hyperplasia with retention of the retiform epidermis- 10.08%
All these changes had to be associated with a dermal scar.
| ||Junctional melanocytic proliferation ||Compound melanocytic proliferation|
|Epidermis: effacement |
|Epidermis: retiform |
The melanocytes in the majority of our cases were arranged in nests in 265 cases (74.23%) with the
remainder present as single cells. As noted above, almost half of our cases had a dermal component.
Extension of melanocytes down adnexal structures was present in 5.6% of cases and pagetoid upward
migration was also observed in 5.6% of cases. A confluent growth pattern was observed in 7 cases
(1.68%), and a prominent dermal inflammatory reaction was noted in 21 cases (6.16%). Residual nevus was
present in 117 cases (32.77%). Cytologically, the dominant melanocyte present had an epithelioid
configuration with round nucleus and even chromatin pattern. Of note, concerning the cases with a dermal
component, the dermal melanocytes retained the morphology of the junctional component, and transition to
type B and C melanocytes was not evident. Melanocytic atypia was defined as melanocytes with
hyperchromasia and enlarged nuclei (greater than the nucleus of a keratinocyte in the stratum
spinosum). Utilizing these criteria, 26.05% of biopsies demonstrated melanocytic atypia (Table 5).
MM with regression histologic data
In all our cases, invasive MM was present adjacent to areas of regression. In our 34 cases, there
were 6 early, 14 intermediate, and 14 with late regression. Early regression was most easy to recognize
as melanoma, as dermal nests of melanoma cells were present intermingled with the lymphoid infiltrate and
loss of melanocytes. However, in intermediate and late regression, with increasing fibrosis, loss of
melanoma cells, diminishing inflammatory response, and effacement of the retiform epidermis, unequivocal
diagnosis of MM in these areas was challenging, were it not for the adjacent changes of MM.
In our series of cases, the majority of recurrent nevi occurred in females (72%) under the age of
forty, with the back being the most common site and accounting for over half (57%) the patients. The
time to recurrence was less than 6 months for over 64% of cases, with only 4% recurring after 24 months
following the primary biopsy/excision. These findings are similar to those reported by Park et al  ,
with the exception of fewer head and neck cases seen in our series (28% versus 6%).
With regards to the patients' original nevi, almost two thirds were ordinary nevi (64%). Of these,
junctional accounted for 5.3% (12 cases), dermal 15.5% (35 cases) and compound 79% (178 cases).
Dysplastic nevi accounted for 27% (96 cases) of the primary biopsies and this phenomenon has not been
described in previous large studies of recurrent nevi, because of the fact that these studies predate the
dysplastic nevus concept. In a more recent smaller study of 15 cases of recurrent nevi, Hoang et al 
classified 3 of their 15 original biopsies as Clark (dysplastic) nevi. Although their series was small,
it did account for 20% of their cases, similar to our findings. Given the relationship of dysplastic
nevus to melanoma, a recurrent melanocytic lesion occurring at the site of a previously biopsied
dysplastic nevus may be cause for increased alarm and suspicion for malignancy. However, our series
demonstrate that this phenomenon does occur with relative frequency, and proper correlation with the
original biopsy will aid in correctly diagnosing this process.
We were able to identify repeatable histologic patterns occurring in our cases, based on a combination
of epidermal changes and location of recurrent nevus cells. The vast majority of RN were readily
identified histologically by the presence of the entire melanocytic proliferation (either junctional or
compound) occurring within the confines of the scar. However, in partial biopsies where the scar
extended to the edges of the biopsy differentiating RN from an atypical melanocytic
proliferation/melanoma, proved problematic. This was particularly the case in which there was retention
of the retiform epidermis (usually not observed overlying scars) and in which there was confluent
proliferation of atypical melanocytes, histologically overlapping with primary melanoma. Previous
studies have not emphasized retiform epidermal changes and which we observed in 15% of our RN cases.
The majority of our cases (85%) demonstrated epidermal effacement with either a junctional or compound
melanocytic proliferation, a pattern most commonly illustrated in prior publications
often showing pagetoid upward migration of melanocytes. In our series, it was this histologic pattern
that demonstrated pagetoid spread in 13 % of cases (as opposed to 1% of cases of RN with retiform
epidermis). In addition, an inflammatory response was also noted in 13% of these cases. When we
compared these cases with intermediate and late stage regression in MM, there were considerable
overlapping histological changes that without knowledge of the patients' prior biopsy, RN could be
interpreted as MM with intermediate/late stage regression. This was especially true in cases where
pagetoid and adnexal spread of melanocytes was present together with a dermal inflammatory response.
Similarly, in our known cases of MM, all had an invasive component to ensure that the cases we were
dealing with were bone fide cases of regression. In analyzing the areas of regression, there were cases
of late regression which were indistinguishable from cases of RN with effacement of the retiform
epidermis and pagetoid and adnexal spread of cytologically atypical melanocytes. This may be a
diagnostic pitfall with potentially serious consequences, and we have observed 2 cases in which the
biopsy was thought to represent RN and the re-excised specimen demonstrated unequivocal invasive MM
(unpublished observations). In one case, we had no record of a prior nevus removal and the lesion
extended to edge margins, leading us to recommend complete removal of the lesion. The re-excised
specimen demonstrated MM, Clark level II, with regression. In the second case, we received in
consultation, an excision from the lateral second toe demonstrating MM, acral lentiginous type, Clark
level II, with associated dermal scar. The prior biopsy, performed 1 year previously at another
institution, was signed out as "consistent with recurrent nevus". The lesion did extend to edge
margins. No mention was made of the initial biopsy which was interpreted at a third institution as an
acral compound nevus, 3 years prior to that biopsy. In analyzing the entire case together, we
interpreted the initial biopsy as MM, Clark level II, the second biopsy as MM, Clark level II with
regression and scar, and the third biopsy as MM, Clark level II, with regression. Both these cases
highlight the overlapping histologic features of MM with regression and RN, and further emphasize the
importance of the clinical findings and review of previous melanocytic biopsies at the site of recurrent
nevi, especially in partial biopsies. This may become an increasingly important issue in view of the
fragmentation of the health care system and the shipping of specimens to laboratories around the country.
Various theories have been proposed regarding the origin of RN, and these include melanocytes
repopulating from the adjacent epidermis
repopulation from adnexal structures
regrowth from residual dermal nevus  . In our series, in 88 cases the primary lesion appeared to be
completely excised which would suggest that regrowth from residual nevus unlikely. Repopulation from
adjacent epidermis would also seem less likely, and this is supported by the fact that in the completely
re-excised RN, the melanocytic proliferation was confined to the scar, and the adjacent unaltered
epidermis did not demonstrate a similar proliferation. We believe that repopulation of melanocytes occur
from remaining adnexal structures and migrate into the regenerating epithelium. This is supported by the
elegant study of Nishimura et al  , who in a mouse model demonstrated that melanocyte stem cells can
be found in the lower permanent portion of hair follicles. These stem cells were shown to be able to
migrate to "niches" lacking melanocytes and repopulate them. They further suggest that these stem cells
could potentially be the source for both hair follicle and epidermal melanocytes.
Dermal nests of melanocytes occurred in a significant number of cases (151 biopsies, 42%), similar to
that seen in the study by Park et al  . In their study, they observed what they term "dropping off" of
epidermal nevus cells into the dermis in 36% of their cases. Interestingly, in their study, 9 of their
RN cases were purely intradermal, a finding not seen in our study. We believe that the dermal nests do
originate from the junctional component, as all our cases were associated with a junctional component and
the cytology of the melanocytes in the dermis was identical to those in the junctional component. These
findings suggest these lesions were attempting to recapitulate the organization of ordinary nevi. On a
practical note, awareness that dermal nests do occur with some frequency in RN and that they are
cytologically similar to the junctional melanocytes may help in correctly interpreting these
In summary, RN tend to occur in young females with the back being the most common site and recurrence
occurring relatively soon after the primary biopsy (8 months). Dysplastic nevi can recur and in our
series did so frequently. Histologically, the vast majority of RN are readily identifiable, however, in
partial biopsies or in cases without prior knowledge of the original biopsy may lead to misdiagnosis.
This is especially true in RN and regression in MM, where these two lesions share overlapping histologic
features. Correlation with the clinical findings and prior biopsy will avoid these pitfalls.
RN tend to occur in young females with the back being the most common site and
recurrence occurring relatively soon after the primary biopsy (8 months). Dysplastic nevi can recur and
in our series did so frequently. Histologically, the vast majority of RN are readily identifiable,
however, in partial biopsies or in cases without prior knowledge of the original biopsy may lead to
misdiagnosis. This is especially true in RN and regression in MM, where these two lesions share
overlapping histologic features. Correlation with the clinical findings and prior biopsy will avoid
- Schoenfeld RJ, Pinkus H. The recurrence of nevi after incomplete removal. AMA Arch Derm. 1958; 78(1): 30.
- Cox AJ, Walton RG. The Induction of Junctional Changes in Pigmented Nevi. Arch Pathol. 1965; 79: 428.
- Kornberg R, Ackerman AB. Pseudomelanoma: recurrent melanocytic nevus following partial surgical removal. Arch Dermatol. 1975; 111(12): 1588.
- Hiss Y, Shafir R. "Pseudomelanoma" in a keloid. J Dermatol Surg Oncol. 1978; 4(12): 938.
- Kang S, Barnhill RL, Mihm MC, Jr., Sober AJ. Histologic regression in malignant melanoma: an interobserver concordance study. J Cutan Pathol. 1993; 20(2): 126.
- Park HK, Leonard DD, Arrington JH, 3rd, Lund HZ. Recurrent melanocytic nevi: clinical and histologic review of 175 cases. J Am Acad Dermatol. 1987; 17(2 Pt 1): 285.
- Hoang MP, Prieto VG, Burchette JL, Shea CR. Recurrent melanocytic nevus: a histologic and immunohistochemical evaluation. J Cutan Pathol. 2001; 28(8): 400.
- Cox AJ, Walton RG. Pigmented nevi. Induced changes in the junctional component. Calif Med. 1966; 104(1): 32.
- Imagawa I, Endo M, Morishima T. Mechanism of recurrence of pigmented nevi following dermabrasion. Acta Derm Venereol. 1976; 56 (5): 353.
- Walton RG, Cox AJ. Electrodesiccation of pigmented nevi. Arch Dermatol. 1963; 87: 342.
- Walton RG, Sage RD, Farber EM. Electrodesiccation of pigmented nevi; biopsy studies: a preliminary report. AMA Arch Derm. 1957; 76(2): 193.
- Schreus HT. [Nevus and melanoma. I. Contribution to the histogenesis of nevus cell nevi.]. Hautarzt. 1960; 11: 440.
- Nishimura EK, Jordan SA, Oshima H, et al. Dominant role of the niche in melanocyte stem-cell fate determination. Nature. 2002; 416 (6883): 854.