Benign Lichenoid Keratosis

Transcription

1 Benign Lichenoid Keratosis ALAN F. FRIGY, M.D. AND PHILIP H. COOPER, M.D. The microscopic spectrum of benign lichenoid keratosis (BLK) was studied by examination of 30 examples. BLK consists of a segment of hyperplastic epidermis accompanied by a lymphoid infiltrate in the papillary dermis. Although termed "lichen planus-like," saw-tooth acanthosis predominated in only seven lesions, whereas irregular acanthosis was widespread in 23. The infiltrate had both a band-like (lichenoid) and perivascular arrangement, but a lichenoid appearance predominated in only half the specimens. Liquefaction of the basal layer was widespread, and hydropic change often involved the midepidermis. Mild cytologic atypicality was present in each case, but other features of actinic keratosis were lacking. Changes of senile lentigo were observed adjacent to 7 lesions. Partial involution was present in several cases. BLK often is confused clinically with a variety of other cutaneous tumors. Nevertheless, the diagnosis usually can be made on histologic grounds alone when the history indicates a solitary lesion. (Key words: Lichenoid keratosis; Lichen planus-like keratosis; Solitary lichen planus; Benign keratosis; Lichenoid inflammation; Senile lentigo; Involution) Am J Clin Pathol 985; 83: BENIGN LICHENOID KERATOSIS (BLK), first defined nearly two decades ago, 5 ' 8 is a relatively common cutaneous lesion, frequently excised for therapeutic and diagnostic purposes. Despite several publications describing BLK, 2 " 8 all in the dermatologic literature, a recent report 2 indicates that the lesion usually is misdiagnosed, both clinically,and pathologically. Although there may be some basis for the clinical confusion, the pathologic features of BLK are sufficiently characteristic that the diagnosis can be suggested strongly on histologic grounds alone and usually can be made with confidence when the clinical impression indicates a solitary lesion. There are no descriptions of BLK in the pathology literature, however, and textbooks contain only brief sections of parenthetic comments about the lesion. Moreover, most descriptions in the dermatologic literature are of a summary nature, and only a few touch on most of the salient features. 2 ' 8 We studied a series of BLK in order to better document its microscopic spectrum, to clarify certain of its histologic features that remain unsettled, and to provide a description sufficiently thorough that pathologists could become skillful at researching a secure morphologic diagnosis. Received May 30, 984; accepted for publication June 26, 984. Address reprint requests to Dr. Cooper: Department of Pathology, Box 24, University of Virginia Medical Center, Charlottesville, Virginia Departments of Pathology and Dermatology, University of Virginia Medical Center, Charlottesville, Virginia Materials and Methods Included for study were 30 BLK evaluated by the Department of Pathology, University of Virginia Medical Center, between January 98 and June 983. Twentytwo of the lesions were received from dermatologists for processing and diagnosis in our laboratory. Slides of eight others were submitted for consultation by dermatologists or pathologists. We obtained a clinical description, history, and followup information for each case. In order to document the histologic range of BLK, we recorded the presence and extent of a variety of specific microscopic features during joint examination of the lesions. Clinical Data Results There were 29 patients (one patient had two lesions). They ranged in age from 28 to 8 years (mean, 52; median, 54). All were white, and 2 (72%) were women. Two-thirds of the lesions were located on the chest, upper extremities, or face (Table ). Their duration was from two weeks to two years, and 60% were present less than one year. The lesions were described as red or brown papules or plaques with shiny, waxy, or slightly scaly surfaces. They varied in size from 3 to 20 mm. The majority were apparently asymptomatic, but three were pruritic. The examining dermatologists entertained a variety of diagnoses (Table 2). BLK was considered in nine cases, but in only two instances was it the only diagnosis offered. Twenty-six of the lesions were treated with either shave or excisional biopsy. There were no recurrences during periods of follow-up that ranged from to 2 months (mean, 5.8; median, ). Microscopic Observations Shave and excisional biopsy specimens of BLK had a characteristic low-power appearance (Figs. -3). The lesions were composed of a segment of thickened epidermis and a coinciding mononuclear cell infiltrate in Downloaded from by guest on 25 September

2 440 FRIGY AND COOPER A.J.C.P. April 985 Table. Anatomic Location of Benign Lichenoid Keratosis Location Number of Cases Upper anterior chest 0 Arm 8 Leg 4 Shoulder 3 Temple 2 Back Hand Thigh Total 30 the superficial dermis. The epidermal surfaces were gently curved or slightly undulating or angular. There were often accentuated dells, however, where follicular infundibula (Fig. 3) or acrosyringeal units with little or no hyperplasia connected with hyperplastic epidermis. The margins of the lesions were rounded, sloped, or more gradually tapered as they joined adjacent normal epidermis. They were obvious in ellipse excisions and often discernable in shave biopsy specimens. Mild to moderate hyperkeratosis, usually diffuse, was present in all BLK, but a basket-weave pattern was retained partially in each. The keratin layer was often loosely adherent or partially detached. Two-thirds of the lesions displayed focal or segmental parakeratosis. Hypergranulosis was mild to moderate in all specimens. The degree of epidermal hyperplasia ranged from slight to marked, and there were several patterns, often present in combination. Irregular elongation and widening of rete ridges was the predominant pattern in 23 lesions (Fig. ). A saw-tooth appearance was present in 20 cases but predominated in only seven (Fig. 2). Areas of more uniform, plaque-like acanthosis, with effacement of the rete ridges, were observed in 4 specimens (Fig. 3). Three lesions had foci of epidermal thinning; in one, this was extensive. The dermis contained a lymphoid infiltrate, most marked in the papillary dermis but also present about blood vessels of the superficial reticular dermis. The infiltrate was arranged in two basic patterns: either as Table 2. Clinical Diagnoses Considered in Differential Diagnosis of Benign Lichenoid Keratosis bands that closely hugged the dermal-epidermal junction (Figs. 2 and 3) or as more discrete aggregates, separate from the epidermis and centered about vessels in the papillary dermis (Fig. ). The band-like arrangement predominated in half the specimens but was the sole pattern in only six lesions. Combinations of both patterns were observed in 2 cases. The band-like infiltrate often was associated with regions of saw-tooth acanthosis, whereas the perivascular pattern was most common beneath segments of irregular epidermal hyperplasia. Foci of mononuclear cell exocytosis extending to the midepidermis were observed in nearly every lesion. In addition to lymphoid cells, a few eosinophils were observed in six cases, and rare plasma cells or neutrophils were noted in four and three lesions, respectively. Five cases had small extravasations of red blood cells. Melanin incontinence, mild to focally marked, was present in 27 biopsies. A minority of specimens had fields or larger segments, either within the lesions or at its margin, where the infiltrate was scattered haphazardly and markedly diminished (Fig. 4). The papillary dermis in such areas was thickened and slightly sclerotic and often contained melanin. Liquefaction of the basal cell layer, segmental or diffuse, was present in each specimen (Fig. 5). The change was particularly marked in areas with a bandlike infiltrate where there were numerous apoptotic (colloid, Civatte) bodies, small dermal-epidermal clefts, and vacuolar change in the midepidermis. Liquefaction involved follicular infundibula and acrosyringia in some instances. It was minimal or absent in epidermis, overlying thickened, sclerotic papillary dermis (Fig. 4). Epidermal cells within the lesions often were mildly enlarged and often had increased eosinophilia in comparison with adjacent, unaffected epidermis (Figs. and 5). In addition, some degree of cytologic atypicality affected the basal and suprabasal cells in nearly every lesion (Fig. 6). There was some loss of normal polarity, and nuclei were enlarged, vesicular, and had prominent nucleoli. Solar elastosis was usually mild, and in only a few lesions was it of the degree typically associated with actinic keratosis. In 7 cases, the adjacent epidermis had changes consistent with senile lentigo (Fig. ). In two, small fields suggested a preexisting seborrheic keratosis (Fig. 7). Downloaded from by guest on 25 September 208 Diagnosis Basal cell carcinoma Bowen's disease Seborrheic keratosis Benign lichenoid keratosis Actinic keratosis Keratosis, not otherwise specified Nevus Wart Lichenoid dermatitis Epidermoid carcinoma Prurigo nodularis Number of Cases Discussion The diagnosis of BLK is often apparent upon lowpower microscopic inspection of excised lesions. 2 The two essential findings are a segment of thickened epidermis and a mononuclear cell infiltrate in the papillary dermis that, in combination, create a well-defined papule or plaque. Microscopic recognition of the borders of the lesion is diagnostically useful, in our experience, as this allows the pathologist to appreciate its circumscribed

3 Vol. 83 No. 4 LICHENOID KERATOSIS 44 FIG. (upper). Low-power view of BLK. The epidermal hyperplasia is irregular, and the infiltrate is primarily perivascular. Paler staining epidermis within the lesion contrasts with darkly staining epidermis at the margin. Lentiginous epidermal change is present adjacent to the lesion at far left. Hematoxylin and eosin (X40). FIG. 2 (center). Low-power view of BLK. The saw-tooth pattern of epidermal hyperplasia is obscured partially by a dense hugging dermal inflammatory infiltrate. Note the loosely adherent keratotic scale and the contrast between the pale epidermis of the lesion and darker staining epidermis laterally (left). Hematoxylin and eosin (X65). Downloaded from by guest on 25 September 208 FIG. 3 (lower). Segments of uniform, plaque-like epidermal thickening were present in some BLK. There often were accentuated dells at adnexal orifices. The infiltrate in this lesion is distinctly band-like. Hematoxylin and eosin (X60). nature. The transition to normal skin is often abrupt and unmistakable, but in BLK with tapered edges the transition is more gradual. The surfaces of BLK are slightly curved or gently undulating or angular. In addition, we noted small but distinct dells at appendigeal orifices. When numerous, they created more complex surface configurations, but well-developed papillomatosis is uncharacteristic of BLK. 8 Mild to moderate hyperkeratosis is an invariable feature, but the basket-weave pattern almost always is retained to some dgree, and dense, compact hyperkeratosis is rare. Parakeratosis often is mentioned in descrip-

4 442 FRIGY AND COOPER A.J.C.P. -April 985 tions of BLK, but in two series 2,6 it was noted specifically in substantially less than half the cases. We observed scattered foci and occasional broader segments of parakeratosis in two-thirds of the lesions. Epidermal hyperplasia and a cellular infiltrate in the papillary dermis, the two most characteristic microscopic findings in BLK, have been described only in general terms. An early report, noting microscopic similarities between BLK and lichen planus, emphasized a sawtooth pattern of epidermal hyperplasia, 5 but others described the acanthosis as irregular 2,7 or uniform 3 or supplied no descriptive details. Irregular elongation and widening of rete ridges was the predominant pattern in 23 of the 30 specimens in this series. A saw-tooth configuration was predominant in only seven lesions, and ten lacked this finding. Approximately one-half the specimens had segments of uniform epidermal thickening. Combinations of these patterns were common, and the degree of epidermal hyperplasia varied considerably within individual specimens. Regarding the inflammatory infiltrate, some observers emphasized a band-like distribution, 2 " 5 but others, noted, in addition, that the infiltrate could be more discrete and perivascular. ' 7,8 We found that combinations of the two arrangements were common and that a distinctly lichenoid, band-like pattern was dominant in only onehalf the lesions. Moreover, the band-like pattern was associated typically with saw-tooth acanthosis, whereas the perivascular pattern, with relative sparing of the dermal-epidermal junction, was usually present in zones with irregular acanthosis. Liquefactive change is an essential feature of BLK, but some of its features have not been emphasized. The change is best developed in areas with a band-like infiltrate where it is by no means limited to basal cells. The midepidermis can show prominent vacuolar change, accompanied by striking mononuclear cell exocytosis. < FIG. 4 {upper). Portions of some BLK (left) contained only sparse mononuclear cell infiltrate and showed sclerosis of the papillary dermis. Such fields were interpreted as evidence of partial involution. Hematoxylin and eosin (X75). FIG. 5 (upper, center). Liquefaction of basal and suprabasal cells often was marked, with clustering of apoptotic (colloid, Civatte) bodies and prominent mononuclear cell exocytosis. Hematoxylin and eosin (X20). Downloaded from by guest on 25 September 208 FIG. 6 (lower, center). A mild degree of epidermal disorder and cytologic atypicality was present in each BLK. Hematoxylin and eosin (X250). FIG. 7 (lower). BLK were observed to merge with seborrheic keratosis in two instances. Note that the inflammatory component of the BLK is contiguous with similar changes affecting the seborrheic keratosis. Hematoxylin and eosin (X50).

5 Vol. 83 No. 4 LICHENOID KERATOSIS 443 In contrast, a substantial minority of specimens in this study had fields, occasionally large, that suggested partial involution of the lesion. The epidermis was hyperplastic but lacked liquefaction, and the papillary dermis was thickened, slightly sclerotic, and contained sparse, scattered inflammatory cells and variable amounts of melanin. Clinical observations indicate that BLK can involute,' but the histologic correlate of this has not been recorded. As dermatologists continue to confuse BLK with a variety of other lesions, 2 ' 4 pathologists need to recognize its clinical aspects. Published reports vary in detail, and the description and clinical photographs of Laur and associates 4 are recommended. Patients with BLK. are, with few exceptions, white. Most are in the fifth to seventh decade of life, and women outnumber men two to one. The lesions are nearly always solitary and have a predilection for the forearm and dorsum of the hand (30%), face (25%), and upper trunk (20%). The onset is often abrupt, and the clinical duration is frequently just a few weeks or months. Longer histories are not unusual, however. The lesions consist of round, polygonal, or irregular papules or plaques with distinct borders. Most are between 5 and 5 mm in diameter. The color can be pink, red, violaceous, tan, or brown. The surfaces are smooth, velvety, waxy, slightly scaly, or occasionally more ketatotic. Some patients complain of mild pruritis, stinging, or burning. As a variety of other more common cutaneous lesions manifest one or more of these clinical features, it is not surprising that BLK often are misinterpreted clinically. In a recent series, the proper clinical diagnosis was considered in only 3 of 43 instances. 2 BLK most commonly is confused with basal cell carcinoma, actinic keratosis, unspecified keratoses, Bowen's disease, and melanocytic nevi. 4 Inflammatory dermatoses rarely are considered. Pathologists, suspecting a BLK, usually can make the diagnosis with confidence when the specimen is accompanied by a clinical impression, albeit incorrect, indicating that the lesion is likely to be isolated or solitary. The nature of BLK is uncertain. Lumpkin and Helwig, 5 in an initial report, speculated that it might be a solitary variant of lichen planus. Shapiro and Ackerman, 8 in the same year, concluded that BLK was a distinct entity, an opinion shared by subsequent observers. BLK only infrequently resembles lesions of lichen planus, clinically, and the latter has not developed in patients with BLK. The possibility that BLK is actinically induced has been considered by some. Hirsch and Marmelzat 3 described five lesions with epidermal atypia and a lichenoid reaction that they termed lichenoid actinic keratosis. These lesions now are accepted as BLK. 4,7 Nevertheless, subsequent reports either failed to comment on dysplasia, 4 ' 8 noted its occasional presence, 2 or, in one instance, emphasized the finding. Cellular and nuclear enlargement, prominent nucleoli, and loss of cellular polarity were invariably present in our series. Features typical of actinic keratosis, however, such as disproportionate nuclear-cytoplasmic ratios, nuclear hyperchromatism, alternating areas of epidermal atrophy and hyperplasia, columns of parakeratosis, and suprabasal acantholysis were uniformly absent. Other factors that mitigate against an actinic etiology for BLK include their occasional appearance on non-sun-exposed portions of the body or in dark-skinned individuals, the absence of actinically related lesions in most patients, and the frequent presence of only mild solar elastosis in the underlying dermis. The cytologic atypicality in BLK probably represents an epidermal reaction to inflammatory injury, perhaps enhanced, to some degree, by the cytologic changes of regeneration. There were scattered mitotic figures in the majority of lesions. A dense mononuclear cell infiltrate, at times lichenoid, is a well-recognized histologic sign of host response or regression in a variety of benign and malignant cutaneous tumors. That BLK may represent a host reaction to a preexisting epidermal lesion is supported by our findings and those of others. Laur and associates 4 provided clinical evidence of preexisting senile lentigines at the sites of 29 of 59 cases, and Mehregan 6 claimed that there was histologic evidence of such lesions in all BLK he examined. We identified lentiginous changes adjacent to 57% of BLK, and two lesions contained recognizable foci of seborrheic keratosis. A host response to a preexisting lentigo or, rarely, a flat seborrheic keratosis would account for the abrupt onset of BLK, many of its clinical and microscopic features, and the fact that it can involute over a period of months. References. Berman A, Herszenson S, Winkelman RK: The involuting lichenoid plaque. Arch Dermatol 982; 8: Goette DK: Benign lichenoid keratosis. Arch Dermatol 980; 6: Hirsch P, Marmelzat WL: Lichenoid actinic keratosis. Dermatol Internat 967;6: Laur WE, Posey RE, Waller JD: Lichen planus-like keratosis, a clinicohistopathologic correlation. J Am Acad Dermatol 98; 4: Lumpkin LR, Helwig EB: Solitary lichen planus. Arch Dermatol 966; 93: Mehregan AH: Lentigo senilis and its evolutions. J Invest Dermatol 975; 65: Scott MA, Johnson WC: Lichenoid benign keratosis. J Cutan Pathol 976; 3: Shapiro L, Ackerman AB: Solitary lichen planus-like keratosis. Dermatologica 966; 32: Downloaded from by guest on 25 September 208