The “neurothekeoma”: Immunohistochemical analysis distinguishes the true nerve sheath myxoma from its mimics

The “neurothekeoma”: Immunohistochemical analysis distinguishes the true nerve sheath myxoma from its mimics

The “Neurothekeoma”: Immunohistochemical Analysis Distinguishes the True Nerve Sheath Myxoma From Its Mimics WILLIAM B. LASKIN, MD, JOHN F. FETSCH, MD...

1MB Sizes 0 Downloads 37 Views

The “Neurothekeoma”: Immunohistochemical Analysis Distinguishes the True Nerve Sheath Myxoma From Its Mimics WILLIAM B. LASKIN, MD, JOHN F. FETSCH, MD, AND MARKKU MIETTINEN, MD In contrast with the myxoid variant of neurothekeoma (nerve sheath myxoma), evidence of neurosustentacular (NS) differentiation in the so-called cellular and mixed (intermediate) variants of neurothekeoma remains controversial. In this study, we selected 22 tumors coded as neurothekeoma or nerve sheath myxoma from the Soft Tissue Registry of the AFIP. Each tumor was histologically subtyped as either a myxoid/hypocellular neurothekeoma (MN) (N ⴝ 11) or as a “cellular” or “mixed” (intermediate) neurothekeoma variant (C&MV) (n ⴝ 11) and analyzed immunohistochemically. The MNs were composed of small, cytologically bland cells arranged in a loose cellular network or in files within highly myxomatous nodules delineated by dense collagen. The tumors showed clear-cut evidence of NS differentiation by exhibiting consistent immunoreactivity for S-100 protein (11 of 11 cases) and low-affinity nerve growth factor receptor, p75NGFR, (NGFR) (10 of 10), and variable reactivity for glial fibrillary acidic protein (GFAP) (10 of 11) and CD57 (Leu-7) (5 of 9). They also showed pericellular collagen type IV (CIV) expression (9 of 9), scattered intralesional CD34-positive spindled cells (10 of 10), epithelial membrane antigen (EMA)-positive spindled cells located within the adjacent dense collagen (8 of 11), and immunoreactivity for alpha-smooth muscle actin (SMA) (3 of 10) and calponin (4 of 9). In 4 cases, scattered intralesional neuraxons were detected by the Bodian histochemical method or immunohistochemically with anti-neurofilament protein. The tumors had a male-to-female ratio of

6:5, a peak incidence in the 4th decade of life, and an anatomic distribution that included the upper and lower limbs and back. The C&MVs included 9 “mixed” and 2 “cellular” variants. C&MVs differed histologically from MNs by their higher cellularity and presence of larger spindled or epithelioid cells with vesicular nuclei. Immunohistochemically, the tumor cells expressed CIV (9 of 10), calponin (7 of 9), SMA (5 of 10), Leu-7 (1 of 7), S-100 protein (1 of 11), but not NGFR, GFAP, or CD34. EMA-positive spindled cells surrounded tumor fascicles in 1 case. Intralesional neuraxons were not identified. Clinically, these tumors differed from the MNs by exhibiting a male-tofemale ratio of 4:7, a peak incidence in the 2nd decade, and an upper body distribution. Our results indicate that the MN shows NS differentiation and is the bona fide nerve sheath tumor, whereas the C&MVs fail to show convincing evidence of NS differentiation and probably warrant a separate classification. HUM PATHOL 31:1230-1241. This is a US Government work. There are no restrictions on its use. Key words: nerve sheath myxoma, neurothekeoma, neurosustentacular, cellular neurothekeoma, immunohistochemistry. Abbreviations: NTK, neurothekeoma; NSM, nerve sheath myxoma; NS, neurosustentacular; AFIP, Armed Forces Institute of Pathology; MN, myxoid, hypocellular variant of neurothekeoma; C&MV, mixed or cellular neurothekeoma variant; NGFR, nerve growth factor receptor; GFAP, glial fibrillary acidic protein; EMA, epithelial membrane antigen; CIV, collagen type IV.

Neurothekeoma (NTK) and nerve sheath myxoma (NSM) are the currently favored designations for a benign, myxomatous, chiefly cutaneous neoplasm of purported peripheral nerve sheath or neurosustentacular (NS) origin. The NSM was first described as a specific pathologic entity by Harkin and Reed1 in 1969. In 1980, Gallagher and Helwig2 reported their experience with 53 examples of a superficially located tumor that

they termed “neurothekeoma” and mentioned that some of their cases bore close histologic resemblance to the previously described NSM. A subsequent series of 70 cases of NSM published by Pulitzer and Reed3 in 1985 reiterated the histologic findings documented in the former 2 studies and reported clinical data that paralleled those previously detailed by Gallagher and Helwig. The NTK/NSM was initially considered a peripheral nerve sheath tumor based on its light microscopic resemblance to non-neoplastic neural structures,3,4 association with small peripheral nerves,1,2,3,5 ultrastructural evidence of schwannian2,6-10 or perineurial3,11,12 differentiation, and expression of S-100 protein2,5,8,10,12-16 or myelin-basic protein6,14,16 by the tumor cells. However, the nerve sheath derivation of NTK came under scrutiny when some investigators reported potential examples that did not express S-100 protein.17-22 Complicating matters further, Rosati et al23 in 1986 described 3 tumors under the designation of “cellular neurothekeoma” that featured increased cellularity, a prominent epithelioid cell component, and absence of S-100 protein expression. Subsequent studies of this variant documented the presence of cytologic atypia and conspicuous mitotic activity in some exam-

From the Department of Pathology, Northwestern University Medical School, Chicago, IL; and the Department of Soft Tissue Pathology, Armed Forces Institute of Pathology, Washington, DC. Accepted for publication June 29, 2000. Presented in part at the meeting of the United States and Canadian Academy of Pathology; New Orleans, Louisiana; March 27, 2000. The opinions and assertions contained herein are the expressed views of the authors and are not to be construed as official or reflecting the views of the Departments of the Navy, Army, or Defense. Address correspondence and reprints requests to William B. Laskin, MD, Department of Pathology, Northwestern Memorial Hospital, Feinberg Pavilion, 7-325, 251 East Huron St, Chicago, IL 606113053. This is a US Government work. There are no restrictions on its use. 0046-8177/00/3110-0010$00.00/0 doi:10.1053/hupa.2000.18474

1230

THE “NEUROTHEKEOMA” (Laskin et al)

TABLE 1. Immunoreagents Used in the Study of Neurothekeoma Immunoreagent

Clone

Pretreatment

Antibody Dilution

Source

S100 protein Glial fibrillary acidic protein Nerve growth factor receptor (p75ngfr) CD34 Collagen type IV ␣-Smooth muscle actin Epithelial membrane antigen CD57 (Leu-7) Neurofilament protein Calponin Factor XIIIa

Polyclonal Polyclonal

None* None

1:2560 1:2560

Dakopatts, Carpinteria, CA Dakopatts

QBEnd/10 CIV 1A4 Mc5 HNK-1 2F11 CALP Polyclonal

Citra antigen retrieval, 20 min Protease, 5 min Microwave, 15 min Ventana Protease 2, 4 min None Ventana Protease 1, 8 min Sigma pepsin, 20 min Ventana Protease 1, 8 min

1:40 1:100 1:160 Prediluted 1:20 1:800 1:100 1:800

Beckton-Dickinson, Mountainview, CA Dakopatts Dakopatts Ventana, Tucson, AZ Beckton-Dickinson Dakopatts Dakopatts CalBioChem Corp, San Diego, CA

* None, no pretreatment.

ples,10,15,24-26 and a lack of immunohistochemical24-29 and ultrastructural10,15,30 evidence of NS differentiation. In addition to the myxoid NTK and the so-called cellular variant, a third purported subtype, termed “mixed”25 or “intermediate”16 NTK, was introduced as a tumor with hybrid immunomorphologic features. Some authors support the notion that the 3 tumors under discussion form a morphologic spectrum,16,25,26 whereas others contend that they represent clinicopathologically distinct neoplasms.15,29,30 Our study was designed to address the dilemma regarding NS differentiation in tumors falling within the currently accepted histomorphologic spectrum of NTK. We selected a group of 22 superficially located, soft tissue tumors coded as nerve sheath myxoma, neurothekeoma, or variant thereof, from the Soft Tissue Registry of the Armed Forces Institute of Pathology (AFIP) and performed a clinicopathologic and immunohistochemical analysis. Our study provides evidence that the highly myxomatous, relatively hypocellular NTKs are of true nerve sheath derivation, whereas the so-called mixed and cellular variants of NTK lack convincing evidence of NS differentiation and probably have a different histogenetic derivation. MATERIALS AND METHODS Twenty-two archival cases accessioned between 1971 and 1999 and coded as “neurothekeoma,” “nerve sheath myxoma,” or a variant thereof, were selected from Soft Tissue Registry of the AFIP and formed the basis of this study. Hematoxylin and eosin–stained sections were used to histomorphologically subtype each lesion as either a myxoid, hypocellular variant of neurothekeoma (MN), or a so-called mixed or cellular neurothekeoma variant (C&MV). The latter tumors differed from the former by being generally less myxoid, exhibiting areas of increased cellularity, and possessing a component of larger spindled or epithelioid-appearing cells. Bodian histochemical method for detecting neuraxons was performed on 16 tumors (MN, n ⫽ 7; C&MV, n ⫽ 9). Immunohistochemical stains using the avidin-biotin complex immunoperoxidase technique31 were obtained in cases in which formalin-fixed, paraffin-embedded material or unstained slides were available. The antibodies used for im-

munohistochemical study and their sources are listed in Table 1. All immunohistochemical stains were analyzed and semiquantitatively graded on a scale of 0 to 4 in the following manner: 0 points correlated with no detectable immunoreactivity; 1 point was assigned if 10% or less of the tumor cells were reactive; 2 points were given if more than 10% but 25% or less of the tumor cells were reactive; 3 points were alloted if greater than 25% but 50% or less of the tumor cells were reactive; and 4 points were given if greater than 50% of the tumor cells were reactive. Mitotic activity was evaluated by examining 50 randomly selected high-power fields (⫻400) from the most cellular areas of the tumor. The mitotic counts were performed with an Olympus BH2 microscope with a high-power field area of 0.237 mm2.

CLINICAL FINDINGS MN (n ⫽ 11) The clinical data for patients with MN are presented in Table 2. The patients ranged in age from 8 to 69 (median, 28; mean, 32) years and included 6 males and 5 females. Ten lesions arose in the limbs, including the foot (n ⫽ 2), knee (n ⫽ 2), ankle, shin, toe, forearm, thumb, and hip (1 each), and the remaining tumor was located on the back. A preoperative clinical impression was available for 5 of the patients. Two lesions were thought to represent lipomas. The tumor situated over the patella was considered a bursal cyst. A ganglion cyst was the clinical impression for the lesion located on TABLE 2. Summary of the Clinical Features of Patients With Myxoid Neurothekeoma and the “Cellular” and “Mixed” Variants of Neurothekeoma

Age range (yr) Median age Mean age Males:females Location Head and neck Upper extremity Lower extremity Trunk

1231

MN

C&MV

8-69 28 32 6:5

4-29 16 15 4:7

— 2 8 1

4 7 — —

HUMAN PATHOLOGY

Volume 31, No. 10 (October 2000)

FIGURE 1. Microscopic features of myxoid neurothekeoma (MN). (A) MN is characterized by the presence of small, spindled, stellate, and multinucleated cells loosely arranged within variably sized, highly myxomatous nodules partitioned by dense collagen. (Hematoxylin and eosin, original magnification ⫻59.) (B) A linear arrangement of tumor cells resulting from end-to-end attachments was a growth pattern observed in many of the MNs. (H&E, original magnification ⫻296.) (C) Cytologically, the cells were small with welldefined cell borders and possessed oval to round, cytologically bland nuclei. (H&E, original magnification ⫻370.) Multinucleated cells with similar cytologic characteristics were a consistent feature in almost all cases.

the dorsum of the foot. One patient, a 28-year-old, developed a mass over a prior injection site where a polio vaccine was administered as a child. This lesion was thought to represent myositis ossificans.

C&MV (n ⫽ 11) The clinical data for patients with C&MV are presented in Table 2. The patients ranged in age from 4 to 29 (median,

1232

THE “NEUROTHEKEOMA” (Laskin et al)

FIGURE 1. (Continued ) (D) Structures resembling Verocay bodies were observed in 3 examples of MN. (H&E, original magnification ⫻150.) (E) A cellular focus (right) in an otherwise classic example of MN is composed of elongated cells with enlarged nuclei arranged in short, intersecting fascicles. (H&E, original magnification ⫻59.) Inset: Note the resemblance of the cellular focus to Antoni A component of conventional schwannoma. (H&E, original magnification ⫻172.)

16; mean, 15) years, and included 7 females and 4 males. The tumors involved the upper portion of the body, including the forearm (n ⫽ 3), shoulder, upper arm, hand, finger, chin, face, neck, and scalp (1 each). A preoperative impression was available for 6 patients and included a cyst (n ⫽ 4), a calcifying epithelioma of Malherbe (n ⫽ 1), and a lipoma (n ⫽ 1).

GROSS EXAMINATION The MNs (n ⫽ 8) measured 0.8 to 3.5 cm in maximum dimension with a median size of 1.8 cm. The C&MVs (n ⫽ 11) ranged from 0.5 to 1.8 cm in maximum dimension, with a median size of 1.0 cm. Three MNs and 2 C&MVs were local-

1233

HUMAN PATHOLOGY

Volume 31, No. 10 (October 2000)

FIGURE 2. Microscopic features of the so-called cellular and mixed neurothekeoma variant (C&MV). (A) This example of a “mixed” neurothekeoma variant features ill-defined nests composed of loosely aggregated and tightly clustered plump spindled cells. (H&E, original magnification ⫻15.) (B) A “cellular” neurothekeoma variant has minimal stromal mucin and consists of infiltrating, solid nests of epithelioid tumor cells. (H&E, original magnification ⫻370.)

TABLE 3. Immunohistochemical Results for Myxoid Neurothekeoma and the “Cellular” and “Mixed” Neurothekeoma Variants

ized to the subcutis on macroscopic examination by the pathologist or surgeon. The tumors were described as nodular or oval masses with a rubbery to hard consistency. The surface of the lesions ranged in color from gray-white to yellow-tan. A mucoid cut surface was described for 2 examples of MN and for one of the C&MVs. Cystic change was noted in 2 examples of MN.

MICROSCOPIC EXAMINATION MN (n ⫽ 11) All tumors were centered in the dermis, but 5 lesions extended into the subcutaneous fat. At low-power magnification, the tumors had a multinodular or lobulated architecture and consisted of infiltrating, varying-sized nodules that were partially or completely invested by a dense band of collagen (Fig 1A). The tumor nodules showed low to moderate cellularity and were composed of small cells that characteristically grew in a loose cellular network (Fig 1A) or formed end-to-end files that were arranged in lamellar and whorled patterns (Fig 1B). Cytologically, the neoplastic cells had a spindled, oval, or

Number of (⫹) Cases/ Number Tested

Antigen/Antibody

Myxoid NTK

Cellular & Mixed NTK

Nerve growth factor receptor (p75ngfr) S-100 protein Glial fibrillary acidic protein CD57 (Leu-7) CD34 (QBEND-10) Collagen type IV Epithelial membrane antigen Neurofilament protein Smooth-muscle actin Calponin Factor XIIIa

10 /10 11 /11 10 /11 5 /9 10 /10 9 /9 8 /11† 1 /10‡ 3 /10 4 /9 5 /9§

0 /10 1 /11* 0 /10 1 /7 0 /9 9 /10 1 /10*† 1 /10‡ 5 /10 7 /9 7 /9§

1234

* Rare reactive cell. † Spindled cells in surrounding collagen. ‡ Neuraxons reactive. § Non-neoplastic dendritic cells reactive.

THE “NEUROTHEKEOMA” (Laskin et al)

FIGURE 2. (Continued) (C) Cytologically, the tumor cells possessed ample amounts of eosinophilic, finely vacuolated cytoplasm and large, vesicular nuclei often with conspicuous nucleoli. (H&E, original magnification ⫻480.) (D) This neoplasm shows a close histologic resemblance to MN. (H&E, original magnification ⫻148.) Immunohistochemical analysis showed no evidence of neurosustentacular differentiation.

stellate-shape and well-marginated, pale eosinophilic cytoplasm (Fig 1C). Most cells possessed a small, round to oval, cytologically bland nucleus. Small intranuclear inclusions were frequently encountered. In all cases, multinucleated cells with sharp cytologic borders and nuclei identical to the mononuclear cells were readily identified. Mitotic activity in the lesional cells ranged from 0 to 2 (mean, 0.6) mitoses per 50 high-power fields examined. In 4 separate cases, tumor cells focally grew in patterns resembling nonneoplastic neural structures. In 3 cases, tightly aggregated spindled tumor cells with clustered or palisaded nuclei formed structures resembling Verocay bodies (Fig 1D). In a single microscopic focus in an otherwise conventional MN, elongated spindled cells with enlarged, wavy, and “buckled” nuclei proliferated in short, intersecting fascicles reminiscent of the Antoni A growth pattern characteristic of conventional schwannoma (Fig 1E). The stroma within the tumor nests consisted of a sparse number of delicate collagen fibers, an occasional capillarysized vessel, scattered mast cells, and a copious amount of mucin that formed pools and clefts.

FIGURE 3. Immunohistochemical results for the myxoid neurothekeoma % reactive cells: 䡺, 0; , 1-10; , 11-25; , 26-50; f, ⬎50. NGFR, nerve growth factor receptor (p75ngfr); GFAP, glial fibrillary acidic protein, CD57, Leu-7; CD34, QBEND-10; CIV, collagen type IV; EMA; epithelial membrane antigen (rare spindled cells in collagenous septa); SMA, alpha–smooth-muscle actin; CALP, calponin; F13, Factor 13a (intralesional dendritic cells).

1235

HUMAN PATHOLOGY

Volume 31, No. 10 (October 2000) FIGURE 4. Immunohistochemical evidence of neurosustentacular differentiation in MN. Strong reactivity for (A) nerve growth factor receptor (p75ngfr), (B) S-100 protein, and (C) glial fibrillary acidic protein in an example of MN. (Original magnification ⫻99.) (D) Anti-CD34 (QBEND-10) marks scattered spindled cells with dendritic cytoplasmic processes in a nest of tumor. (Original magnification ⫻244.) (E) Anti-epithelial membrane antigen highlights scattered spindled elements, presumably perineurial cells, in the collagen surrounding a nodule of tumor. (Original magnification ⫻240.)

Rare foci of perivascular lymphocytes were identified in the collagenous septa in 8 tumors. Segments of peripheral nerve were noted adjacent to tumor nests in seven lesions and, in 2 examples, they showed myxoid change. The Bodian method identified a few, scattered neuraxons not associated with intact peripheral nerve located principally at the edge of a tumor nodule in 3 of the 7 tumors examined.

C&MV (n ⫽ 11) These tumors were located primarily in the dermis, with 7 of the lesions extending into subcutaneous fat. Although some examples had a discrete, multinodular appearance, most of the lesions were composed of infiltrating, variably sized, ill-defined nests of tumor cells that frequently coa-

1236

THE “NEUROTHEKEOMA” (Laskin et al)

axons not associated with intact peripheral nerve within tumor nests in any of the 9 cases evaluated.

IMMUNOHISTOCHEMICAL RESULTS MN (n ⫽ 11)

FIGURE 5. Immunohistochemical results for the cellular and mixed neurothekeoma variants % reactive cells; 䡺, 0; , 1-10; , 11-25; , 26-50; f, ⬎50. NGFR, nerve growth factor receptor (p75ngfr); GFAP, glial fibrillary acidic protein, CD57, Leu-7; CD34, QBEND-10; CIV, collagen type IV; EMA; epithelial membrane antigen (rare spindled cell in collagenous septa); SMA, alpha–smooth-muscle actin; CALP, calponin; F13, Factor 13a (intralesional dendritic cells).

lesced, resulting in lobulated (Fig 2A) or a vaguely plexiform architecture. The cellularity of the tumors varied from moderate to high in any given case. In most examples of the “mixed” NTK variant, epithelioid-appearing cells formed small, tightly clustered whorls near the center of the nest with loosely arranged spindled and stellate-shaped cells at the periphery (Fig 2A). The 2 examples of “cellular” NTK variant were composed of plump, spindled, and epithelioid cells arranged in solid, infiltrating nests and fascicles (Fig 2B). Cytologically, the neoplastic cells were relatively large and contained lightly eosinophilic, finely vacuolated cytoplasm. The nuclei were round to oval with vesicular chromatin and often possessed conspicuous nucleoli (Fig 2C). Multinucleated giant cells, including osteoclastlike giant cells, were encountered in 9 tumors but were not numerous. A mild to moderate degree of nuclear atypia manifested by slight anisonucleosis and mild hyperchromatism was appreciated in 4 of the cases. The mitotic activity ranged from 0 to 9 (mean, 2.5) mitoses per 50 highpower fields examined, with no atypical mitotic figures identified. Two tumors showed close histologic resemblance to MN, except for focal, mild increased cellularity and the occasional presence of slightly larger cells (Fig 2D). A few examples of the C&MV showed small foci vaguely reminiscent of MN, but increased cellularity or cellular enlargement in other areas. The stroma consisted of wispy collagen fibers that tended to condense in the more epithelioid areas, scattered capillarysized vessels, and stromal mucosubstance. The amount of stromal mucin was considerably reduced in comparison with that seen in MN, especially in the 2 examples of the so-called cellular NTK. Deposits of perivascular lymphocytes in the surrounding tissue were identified in almost every case, and intratumoral infiltration by lymphocytes was observed in 7 tumors. Xanthomatous histiocytes were admixed with tumor cells in 1 example. Scattered mast cells were identified in a few examples. Peripheral nerve twigs were identified in the surrounding tissue in 6 examples, and in 3 cases, segments of peripheral nerve were entrapped by infiltrating tumor. However, none of the nerve segments showed histologic features similar to the adjacent tumors. The Bodian method failed to identify neur-

The immunohistochemical results for the MNs are given in Table 3 and illustrated in Figure 3. The tumor cells in all 10 examples tested for nerve growth factor receptor (NGFR) showed strong and diffuse cytoplasmic and membranous reactivity (Fig 4A). In almost all cases, focal NGFR expression in spindled cells located in the dense collagen surrounding tumor nodules was observed. Moderate to strong nuclear and cytoplasmic expression of S-100 protein by tumor cells was observed all 11 lesions tested (Fig 4B). Variable glial fibrillary acidic protein (GFAP) (Fig 4C) immunoreactivity was detected in 10 of 11 neoplasms tested. CD57 (Leu-7) immunoreactivity was observed in 5 of 9 tumors evaluated, but was generally weak. Anti-CD34 stained scattered spindled cells with dendritic cytoplasmic processes within tumor nodules in all 10 neoplasms tested (Fig 4D). Anti-epithelial membrane antigen (EMA) failed to stain tumor cells but did decorate an occasional spindled element in the dense collagen surrounding tumor nodules in 8 of the 11 neoplasms evaluated (Fig 4E). Strong pericellular expression of collagen type IV (CIV) was observed in all 9 tumors tested. Fragments of neuraxons were identified within a single tumor nodule in only 1 of 10 tumors tested with anti-neurofilament protein (NFP). These neuraxons were not observed in sections prepared from the same paraffin block and stained by the Bodian method. The 2 myoid antigens, calponin and alpha-smooth muscle actin, were expressed by scattered lesional cells in 4 of 9 and 3 of 10 cases, respectively. Rare, non-neoplastic cells with dendritic cytoplasmic processes were identified by the anti-factor XIIIa antibody within the tumor nodules in 5 of 9 tumors tested. C&MV (n ⫽ 11) The immunohistochemical results for this group are given in Table 3 and illustrated in Figure 5. Tumor cells in 3 of the neoplasms showed limited reactivity to the NS immunomarkers used in this study. In 1 of the “mixed” NTK variants, rare S-100 protein–positive cells were identified in the cellular area of the tumor, whereas in another example, tumor cells focally expressed Leu-7. In a “cellular” NTK variant, rare, scattered spindled cells surrounding infiltrating tumor nests expressed EMA. The remaining tumors tested for expression of S-100 protein (n ⫽ 10), NGFR (n ⫽ 10), GFAP (n ⫽ 10), CD57 (n ⫽ 6), and CD34 (n ⫽ 9) were negative. Pericellular staining of tumor cells with anticollagen type IV was identified in 9 of 10 neoplasms tested (Fig 6A), but the degree of reactivity was more variable than in the MNs. Reactivity with anti-calponin (7 of 9 tumors) (Fig 6B) and anti-smooth muscle actin (5 of 10 tumors) (Fig 6C) was variable, but generally stronger than in the MNs. Scattered intralesional factor

1237

HUMAN PATHOLOGY

Volume 31, No. 10 (October 2000)

FIGURE 6. Immunohistochemical findings in C&MV. (A) Pericellular expression of collagen type IV was observed in 90% of cases, but was more variable in strength than in the MNs. (Original magnification ⫻244.) Myoid immunomarkers, (B) calponin (Original magnification ⫻488).

XIIIa–immunoreactive, non-neoplastic dendritic cells were observed in 7 of the 9 neoplasms tested, whereas numerous factor XIIIa–positive dendritic cells were identified in the connective tissue surrounding nests of tumor in several examples. Fragmented neuraxons were identified with the anti-NFP antibody within a tumor nodule in a single case (n ⫽ 10), but lesional cells within this nodule failed to express any of the NS markers used in our study. DISCUSSION Although there is ample immunohistochemical and ultrastructural evidence showing that the myxoid, hypocellular variant of NTK is of nerve sheath origin, studies evaluating cytodifferentiation in the “mixed” and “cellular” NTK variants have not convincingly shown NS differentiation. In this study, we provide immunohistochemical data validating the peripheral nerve sheath origin of the MN. Furthermore, the ab-

sence of convincing NS differentiation in the C&MV group disputes the notion that these tumors are histogenetically related to, and form a morphologic continuum with, the MN. Although the number of cases analyzed was relatively small, we did note several clinical differences between the 2 groups, which are summarized in Table 2. The age difference between the 2 groups in our study and the tendency of the C&MVs to involve the head and neck region and upper extremity parallel the clinical data previously reported for NSM,30 NTK,2,30 and the so-called cellular NTK variant.15,24 We found that the MN can be distinguished from C&MV on light microscopic examination in almost all cases. Compared with the plumper, more epithelioid-appearing cells with vesicular nuclei that comprise the C&MV, the mononuclear and multinucleated cells of the MN are smaller and have more uniform-appearing, cytologically bland nuclei. Architecturally, the cells of MN commonly formed orderly

1238

THE “NEUROTHEKEOMA” (Laskin et al)

FIGURE 6. (Continued) (C) alpha-smooth muscle actin (Original magnification ⫻389) were reactive in 78% and 50% of cases, respectively. Reactivity with these latter 2 immunoreagents and factor XIIIa (not shown) was generally stronger in examples ofC&MV than in the MNs.

files and were disposed in well-delineated nests generally surrounded by a band of dense collagen resembling perineurium. In comparison, the tumor cells in the C&MV typically grew in tight whorled aggregates or in a fascicular arrangement within compartments that frequently coalesced resulting in a lobular or, rarely, a plexiform architecture. A lymphocytic infiltrate around vessels within the adjacent connective tissue was a feature more commonly observed in examples of C&MV. Zelger et al29 previously noted that a peripheral lymphocytic infiltrate is a characteristic finding in the so-called cellular NTK variant. Table 4 compares the salient pathologic differences between the 2 tumor types. Relying on S-100 protein, GFAP, and NGFR as the principal immunomarkers for eliciting evidence of NS differentiation, we found that all 11 examples of MN showed moderate to strong immunoexpression of at least 2 of these antigens, whereas the C&MVs were virtually nonreactive (Table 4). This latter group also included 2 tumors that showed significant histomorphologic overlap with MN. Furthermore, we observed microscopic foci vaguely resembling MN in other examples of C&MV that, if present alone in a limited biopsy, could potentially lead to an erroneous diagnosis of MN. Our study shows that this panel of immunomarkers identified the true NSM and is an important diagnostic adjunct for separating tumors with histologic overlap, especially on a limited biopsy. S-100 protein immunoreactivity has been reported in approximately 93% of tumors classified as NTK or NSM in large series.3,5,10,15,25 In a study subtyping NTK into 3 histologic variants, Husain et al16 reported that tumors toward the myxoid end of the histologic spectrum of NTK more often expressed S-100 protein. Kao and Penneys14 in abstract form reported “spotty” S-100 protein immunoreactivity in

TABLE 4. Comparison of Pathologic Features of Myxoid Neurothekeoma with the “Cellular” and “Mixed” Variants of Neurothekeoma

Histopathology Architecture Cellularity (relative to native dermis) Cell type(s) Cell arrangement Nuclear characteristics Perivascular lymphocytic infiltrate Immunohistochemistry NGFR/S100/GFAP CD34 EMA† SMA/Calponin

Myxoid NTK

Cellular & Mixed NTK

Well defined nodules; less often lobulated Hypocellular to minimally hypercellular Small spindled, stellate, and multinucleated cells

Rare

Lobulated; occasionally, plexiform Moderately to markedly hypercellular Moderate to large spindled and epithelioid cells; rare multinucleated cell Loosely arranged and tightly clustered cells; fascicles and whorls Variable sized, round to oval; vesicular chromatin, occasional conspicuous nucleolus Common

⫹⫹-⫹⫹⫹ ⫹ 0⫺⫹ 0⫺⫹

0* 0 0‡ 0⫺⫹⫹⫹

Loose cellular network and interconnecting files; occasionally, Verocay-like bodies Small, oval to round; rarely wavy and buckled

* Rare S-100 (⫹) tumor cell in 1 case. † Spindled cells in surrounding collagen. ‡ Rare cell (⫹) in 1 case.

1239

HUMAN PATHOLOGY

Volume 31, No. 10 (October 2000)

the 34 examples of NTK tested. Unfortunately, this study, which represents the largest immunohistochemical examination of NTK to date, did not take histomorphologic differences into account when interpreting the immunohistochemical results. Conversely, only 3 of 68 tumors reported as examples of “cellular” NTK10,15,23-29,32 exhibited S-100 protein immunoexpression.10,32 GFAP is an intermediate filament found in non– myelin-forming schwann cells,33 and its immunoexpression has been evaluated in NTK.9,10,14,15,25,29,34,35 We are aware, however, of only 2 cases of conventional NTK described in the literature that were positive.9,35 We are the first to report immunoexpression of low-affinity NGFR in tumor cells of NTK. This 75-kd surface receptor glycoprotein is found on non–myelinforming schwann cells, embryonic schwann cells, perineurial cells, and nerve fibers.33,36 Low-affinity NGFR immunoreactivity, although not specific for neural tissue,37 has been reported in a variety of benign peripheral nerve sheath tumors, including neurofibroma,37-40 schwannoma,36,37,39,40 intraneural perineurioma,41 and granular cell tumor.37,39 Anti-NGFR also marked spindled cells located in the dense collagen surrounding tumor nodules. Although it is sometimes difficult to separate these spindled cells from tumor cells and from NGFR-positive perivascular cells,37 we believe that this immunomarker is highlighting, in part, perineurial cells. In comparison, anti-EMA, which also detects a subset of perineurial cells,21,42 identified fewer spindled elements in the surrounding collagen. Our findings support the observation that perineurial cells are commonly present in the collagen band surrounding tumor nests of MN.25 Anti-CD57 (Leu-7) is believed to detect a carbohydrate epitope of a myelin-associated glycoprotein present on schwann cells.43 Our literature review found that all but a few reported examples of conventional NTK were CD57 (Leu-7) negative,10,15,25 including 34 examples of NTK studied by Kao and Penneys.14 Variable expression of another structural component of myelin, myelin-basic protein, has been reported in 2 series of NTK,14,16 whereas in other large studies, myelin-basic protein was not detected.15,23-25 The low frequency of myelin sheath–related antigen expression in tumors reported as examples of MN may be attributable to the principal involvement of a GFAP-/NGFR-positive, non–myelin-forming schwann cell in the process, the inability of the antibody to recognize a myelin sheath–related epitope, or the inclusion of so-called cellular or mixed NTK variants. Our immunomorphologic analysis of 11 examples of MN indicates that this tumor is composed primarily of schwann cells based on the consistent coexpression of S-100 protein with NGFR, common immunoexpression of GFAP,33 the identification of Verocay-like structures and Antoni A growth patterns, and the invariable presence of pericellular CIV.43 In addition, the relatively few number of CD34-positive fibroblastlike cells44 and near-total absence of intralesional neuraxons iden-

tified by Bodian method and by anti-NFP immunoreagent further suggest that MN may be more closely related to the benign schwannoma than neurofibroma. However, we are uncommitted as to whether the MN is a variant of one of these aforementioned tumors and would prefer to simply regard the MN as a unique entity amongst the benign peripheral nerve sheath neoplasms. Our inability to show immunohistochemical evidence of NS differentiation in the examples of C&MV raises questions regarding the histogenesis of this group of neoplasms. We cannot completely exclude the possibility that some examples of C&MV are of peripheral nerve sheath derivation. Hypothetically, a peripheral nerve sheath element that lacks NS differentiation, such as the embryonic perineurial cell, which has fibroblastic and myofibroblastic features,45,46 could potentially serve as the cell of origin of the C&MV. In addition, some investigators have reported expression of the putative neural/neuroendocrine immunomarkers, NK1/C3 (CD57),26,27,29 and PGP9.547 in examples of “cellular” NTK. Nevertheless, we contend that the virtual absence of expression of conventional NS markers, which are commonly found in all well-characterized benign peripheral nerve sheath tumors, seriously militates against this possibility. The C&MVs in our study showed a relatively greater degree of smooth muscle actin and calponin immunoexpression than did the MNs. Furthermore, intralesional and peritumoral factor XIIIa–immunoreactive dendritic cells and perivascular lymphocytes were more prevalent in cases of C&MV than MN. These findings, along with the cytologic appearance of the cells comprising the C&MVs, are features shared with “fibrohistiocytic” tumors. In a study of “cellular” NTK, Zelger et al29 noted additional similarities between the “cellular” NTK and fibrous histiocytoma, including a preoperative suspicion of a dermatofibroma or fibrous histiocytoma in cases of “cellular” NTK, acanthosis of the overlying epidermis and accentuation of adjacent collagen bundles in both lesions, and immunohistochemical expression of similar antigens. The plexiform growth pattern noted in some examples of C&MV is also shared by the plexiform fibrohistiocytic tumor. Indeed, some investigators contend that the “cellular” NTK may be related to the above-mentioned entity.48 We also have noted overlapping morphologic features between the plexiform fibrohistiocytic tumor and some of the more cellular and epithelioid examples of the C&MV in our study and believe that they are related entities. In summary, our analysis shows that the hypocellular, myxoid NTK is a unique form of benign peripheral nerve sheath tumor and deserves the appellation of “nerve sheath myxoma.” All 11 examples of MN reacted convincingly, with at least 2 of the 3 principal immunomarkers (S-100 protein, GFAP, and NGFR) used in our study to detect NS differentiation, and this immunohistochemical panel also aided in the classification of 2 C&MVs that histologically mimicked the MN. Our data

1240

THE “NEUROTHEKEOMA” (Laskin et al)

further indicate that the so-called “cellular” and “mixed” NTK variants lack NS differentiation and may be potentially related to tumors included in the “fibrohistiocytic” category. REFERENCES 1. Harkin JC, Reed RJ: Solitary benign nerve sheath tumors, in Harkin JC, Reed RJ (eds): Tumors of the Peripheral Nervous System. Atlas of Tumor Pathology, Second series, Fascicle 3. Washington, DC, Armed Forces Institute of Pathology 1969, pp 29-65 2. Gallager RL, Helwig EB: Neurothekeoma: A benign cutaneous tumor of neural origin. Am J Clin Pathol 74:759-764, 1980 3. Pulitzer DR, Reed RJ: Nerve sheath myxoma (perineurial myxoma). Am J Dermatopathol 7:409-421, 1985 4. Macdonald DM, Wilson-Jones E: Pacinian neurofibroma. Histopathology 1:247-255, 1977 5. Fletcher CDM, Chan JK-C, McKee PH: Dermal nerve sheath myxoma: A study of three cases. Histopathology 10:135-145, 1986 6. Aronson OJ, Fretzin DF, Potter BS: Neurothekeoma of Gallagher and Helwig (dermal nerve sheath myxoma variant): Report of a case with electron microscopic and immunohistochemical studies. J Cutan Pathol 12:506-519, 1985 7. Goldstein J, Lifshitz T: Myxoma of the nerve sheath: Report of three cases, observations by light and electron microscopy and histochemical analysis. Am J Dermatopathol 7:423-429, 1985 8. Blum AK, Kay S, Adelaar RS: Nerve sheath myxoma of digital nerve. Cancer 63:1215-1218, 1989 9. Paulus W, Jellinger K, Perneczky G: Intraspinal neurothekeoma (nerve sheath myxoma): A report of two cases. Am J Clin Pathol 95:511-516, 1991 10. Argenyi ZB, Kutzner H, Seaba MM: Ultrastructural spectrum of cutaneous nerve sheath myxoma/cellular neurothekeoma. J Cutan Pathol 22:137-145, 1995 11. Webb JN: The histogenesis of nerve sheath myxoma: Report of a case with electron microscopy. J Pathol 127:35-37, 1979 12. Angervall L, Kindblom LG, Haglid K: Dermal nerve sheath myxoma: A light and electron microscopic, histochemical and immunohistochemical study. Cancer 53:1752-1759, 1984 13. Nogita T, Someya T, Nakagawa H, et al: Nerve sheath myxoma: An immunohistochemical study of a case. Dermatologica 181: 317-319, 1990 14. Kao GF, Penneys NS: Immunohistochemical findings of 34 neurothekeomas (benign peripheral nerve sheath tumor). J Cutan Pathol 17:304, 1990 (abstr) 15. Barnhill RL, Dickersin GR, Nickeleit V, et al: Studies on the cellular origin of neurothekeoma: Clinical, light microscopic, immunohistochemical, and ultrastructural observations. J Am Acad Dermatol 25:80-88, 1991 16. Husain S, Silvers DN, Halperin AJ, et al: Histologic spectrum of neurothekeoma and the value of immunoperoxidase staining for S-100 protein in distinguishing it from melanoma. Am J Dermatopathol 16:496-503, 1994 17. Enzinger FM, Weiss SW: Benign tumors of peripheral nerves, in Enzinger FM, Weiss SW (eds): Soft Tissue Tumors (ed 1). St Louis, MO, Mosby, 1983, pp 580-624 18. Epstein J, Urmacher C: A cutaneous neoplasm with neural differentiation. Am J Dermatopathol 7:591-593, 1984 19. Henmi A, Sato H, Wataya T, et al: Neurothekeoma: Report of a case with immunohistochemical and ultrastructural studies. Acta Pathol Jpn 36:1911-1916, 1986 20. Isoda M, Katayama M: Neurothekeoma. Cutis 41:255-256, 1988 21. Ariza A, Bilbao JM, Rosai J: Immunohistochemical detection of epithelial membrane antigen in normal perineurial cells and perineurioma. Am J Surg Pathol 12:678-683, 1988 22. Tuthill RJ: Nerve-sheath myxoma: A case report with immunohistologic evidence of its perineurial cell origin. J Cutan Pathol 15:348, 1988 (abstr) 23. Rosati LA, Fratamico FCM, Eusebi V: Cellular neurothekeoma. Appl Pathol 4:186-191, 1986 24. Barnhill RL, Mihm MC: Cellular neurothekeoma: A distinctive variant of neurothekeoma mimicking nevomelanocytic tumors. Am J Surg Pathol 14:113-120, 1990

25. Argenyi ZB, LeBoit PE, Santa Cruz D, et al: Nerve sheath myxoma (neurothekeoma) of the skin: Light microscopic and immunohistochemical reappraisal of the cellular variant. J Cutan Pathol 20:294-303, 1993 26. Busam KJ, Mentzel T, Colpaert C, et al: Atypical or worrisome features in cellular neurothekeoma: A study of 10 cases. Am J Surg Pathol 22:1067-1072, 1998 27. Calonje E, Wilson-Jones E, Smith NP, et al: Cellular neurothekeoma: an epithelioid variant of pilar leiomyoma? Morphological and immunohistochemical analysis of a series. Histopathology 20:397-404, 1992 28. Tomasini C, Aloi F, Pippione M: Cellular neurothekeoma. Dermatology 192:160-163, 1996 29. Zelger BG, Steiner H, Kutzner H, et al: Cellular “neurothekeoma”: An epithelioid variant of dermatofibroma? Histopathology 32:414-422, 1998 30. Scheithauer BW, Woodruff JM, Erlandson RA: Miscellaneous benign neurogenic tumors, in Scheithauer BW, Woodruff JM, Erlandson RA (eds): Atlas of Tumor Pathology, Tumors of the Peripheral Nervous System. Third series. Fascicle 24, Washington, DC, Armed Forces Institute of Pathology 1999, pp 219-282 31. Hsu SM, Raine L, Fanger H: Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:577-580, 1981 32. Chow LT, Ma TK, Chow WH: Cellular neurothekeoma of the hypopharynx. Histopathology 30:192-194, 1997 33. Jessen KR, Mirsky R: Schwann cell precursors and their development. Glia 4:185-194, 1991 34. Paulus W, Warmuth-Metz M, Sorensen N: Intracranial neurothekeoma (nerve-sheath myxoma): Case report. J Neurosurg 79: 280-282, 1993 35. Katsourakis M, Kapranos N, Papanicolaou SI, et al: Nervesheath myxoma (neurothekeoma) of the oral cavity: A case report and review of the literature. J Oral Maxillofacial Surg 54:904-906, 1996 36. Garin Chesa P, Rettig WJ, Thomson TM, et al: Immunohistochemical analysis of nerve growth factor receptor expression in normal and malignant human tissues. J Histochem Cytochem 36:383-389, 1988 37. Perosio PM, Brooks JJ: Expression of nerve growth factor receptor in paraffin-embedded soft tissue tumors. Am J Pathol 132: 152-160, 1988 38. Ross AH, Grob P, Bothwell M, et al: Characterization of nerve growth factor receptor in neural crest tumors using monoclonal antibodies. Proc Nat Acad Sci 81:6681-6685, 1984 39. Hoshi N, Hiraki H, Yamaki T, et al: Frequent expression of 75kDa nerve growth factor receptor and phosphotyrosine in human peripheral nerve tumours: An immunohistochemical study on paraffin-embedded tissues. Virchows Arch 424:563-568, 1994 40. Yasuda T, Sobue G, Mitsuma T, et al: Nerve growth factor receptor immunoreactivity in human benign peripheral nerve sheath tumor. Acta Neuropathol 77:591-598, 1989 41. Sciacco M, Scarpini E, Baron PL, et al: Sural nerve immunoreactivity for nerve growth factor receptor in a case of localized hypertrophic neuropathy. Acta Neuropathol 83:547-553, 1992 42. Perentes E, Nakagawa Y, Ross GW, et al: Expression of epithelial membrane antigen in perineurial cells and their derivatives: An immunohistochemical study with multiple markers. Acta Neuropathol 75:160-165, 1987 43. Ortiz-Hidalgo C, Weller RO: Peripheral nervous system, in Sternberg SS (ed): Histology for Pathologists (ed 2). Philadelphia, PA, Lippincott-Raven Publishers, 1997, pp 285-311 44. Weiss SW, Nickoloff BJ: CD-34 is expressed by a distinctive cell population in peripheral nerve, nerve sheath tumors, and related lesions. Am J Surg Pathol 17:1039-1045, 1993 45. Bunge MB, Wood PM, Tynan LB, et al: Perineurium originates from fibroblasts: Demonstration in vitro with a retroviral marker. Science 243:229-231, 1989 46. DuPlessis DG, Mouton YM, Muller CJ, et al: An ultrastructural study of the development of the chicken perineurial sheath. J Anat 189:631-641, 1996 47. Wang AR, May D, Bourne P, et al: PGP9.5. A marker for cellular neurothekeoma. Am J Surg Pathol 23:1401-1407, 1999 48. Requena L, Sangueza OP: Benign neoplasms with neural differentiation: A review. Am J Dermatopathol 17:75-96, 1995

1241