Abstract
To verify the pathologic features, anaplastic lymphoma kinase (ALK) expression and biologic behavior of inflammatory myofibroblastic tumors (IMTs) of the central nervous system (CNS), we analyzed 10 cases of IMTs-CNS (8 cranial, 1 spinal, and 1 orbital). Our series of IMTs of the CNS showed a male predominance (male:female = 6:4) and a wide age range (10-60 years; mean age, 46.7 years). Lesion location also varied, but they were basically dura-based. Radiologically, they showed two patterns: isolated mass forming (n = 6) and an en plaque-like pattern (n = 4). Histopathologically, plasma cell granuloma (PCG)-like (n = 5) or fibrohistiocytic (FHC) variant (n = 5) was present. No correlation was found between the radiologie and histopathologic patterns. Spindle-shaped mesenchymal cells of IMTs expressed smooth muscle actin (SMA) in all cases. ALK expression was not found in our IMTs of the CNS. Late recurrence was found in 2 cases in different sites (20%). Pathologically, IMT-CNS could be subclassified into PCG-like and FHC. Immunostaining for SMA was found to helpfully discriminate myofibroblastic cells and to make a differential diagnosis. Although our cases did not show ALK immunoreactivity, some IMTs-CNS can recur, which suggests the neoplastic potential of these tumors. The rearrangement of the ALK gene in IMTs-CNS should be verified by an examination of more cases.
Key Words: Anaplastic lymphoma kinase, Central nervous system, Inflammatory myofibroblastic tumor, Inflammatory pseudotumor. Plasma cell granuloma.
INTRODUCTION
Inflammatory myofibroblastic tumor (IMT) is a lesion characterized by the proliferation of myofibroblastic spindle cells with mixed inflammatory infiltrates of plasma cells, lymphocytes, eosinophils, and histiocytes. The constituent inflammatory cells are mature and polyclonal, and occur in virtually every organ system, including the liver, mesentery, gastrointestinal tract, retroperitoneum, urinary bladder, upper respiratory tract, and mediastinum (1, 2), as well as in the lung, the originally described and most common site. Because of the wide spectrum of its histologic and clinical features, this tumor has been referred to variously as “inflammatory pseudotumor,””plasma cell granuloma,””pseudosarcomatous myofibroblastic proliferation,””inflammatory myofibrohistiocytic proliferation,” etc. In the 2002 World Health Organization classification of soft tissue tumors, these lesions were renamed “inflammatory myofibroblastic tumors” and allocated to the soft tissue tumor category (3). Recently, chromosomal translocations involving the ALK gene were reported in about 35% to 50% of IMTs, and resultant ALK protein overexpression in myofibroblastic cells was found in 35% to 60% of IMT cases, suggesting neoplasm of these tumors rather than a reactive or reparative process (4-6). However, the concept of the same lesions in the central nervous system (CNS) was not changed and was still referred to as “plasma cell granuloma” or “inflammatory pseudotumor” but rarely as “inflammatory myofibroblastic pseudotumor” when myofibroblasts were rich in the mass. These different nosologies for the same kind of tumor inevitably cause confusion during the diagnosis and management of these lesions. We consider that they are the same lesion that occurs in the CNS and suggest that they be referred to in unison as IMT. The CNS, including the spinal cord, may be one of the rare sites affected by IMTs, and if we included the lesions described using the various terminologies above, some 100 sporadic cases have been reported in the literature (7, 8). To verify the clinicopathology, ALK expression, and biologic features of IMT-CNS, we analyzed clinicopathologic features and performed ALK immunohistochemistry on 10 cases of IMTs-CNS.
MATERIALS AND METHODS
Ten cases of IMTs-CNS (8 intracranial, 1 spinal, and 1 intracranially extended orbital case) were selected from the archives of three referral hospitals: Seoul National University Hospital, Samsung Medical Center, and Inje University Hospital. The patients presented between 1996 and 2003. Clinical histories and radiologic findings were obtained from medical records. As a control for the ALK immunohistochemical study, 10 cases of extra-CNS and extra-orbital IMTs were included. The IMTs-CNS group was composed of 6 males and 4 females, and patients’ ages ranged from 10 to 65 years with a mean age 46.7 years. Ten cases of extracranial IMTs were used as controls for immunohistochemical study. Extracranial IMTs were located in the lung (n = 3), urinary bladder (n = 2), liver (n = 2), and one in each of colon, chest wall, and retroperitoneum; patients’ ages ranged from 5 to 72 years with a mean of 31.1 years. Three cases of cerebral and spinal idiopathic hypertrophie pachymeningitis were included to compare smooth muscle actin (SMA) immunoreactivity with that of IMTs-CNS.
Two pathologists independently reviewed the hematoxylin and eosin slides and agreed to a diagnosis of IMT. Immunohistochemical staining was performed using a conventional labeled streptavidin- biotin-peroxidase method (LSAB Kit, DAKO, Glostrup, Denmark) according to the manufacturer’s protocol. Briefly, 4-m tissue sections were placed on silanecoated slides, deparaffinized, and rehydrated. After antigen retrieval by microwaves blocking endogenous peroxidase activity with hydrogen peroxide, goat serum (DAKO) to prevent nonspecific reaction and primary antibodies were applied sequentially. The primary antibodies used were as follows: vimentin, SMA, epithelial membrane antigen (EMA), Ki-67, p53, ALK, CD3, CD5, CD10, CD20, and kappa-, and lambda-light chains, Tdt, S- 100 protein, c-kit, CD21 (all from DAKO), cyclin D1 (Novocastra, Newcastle-upon-Tyne, UK and p16 (Pharmingen, San Diego, CA), The slides were incubated in biotinylated goat anti-mouse/rabbit immunoglobulin and then in a solution of streptavidin-biotin complex. Immunoreactivity was visualized by using 3,3- diaminobenzidine. Tonsil, intestine, and ALK-positive anaplastic large cell lymphoma tissues were used as appropriate positive control and as an antibody control primary antibodies were omitted.
RESULTS
Clinical and Radiologic Features
The clinical and radiologic features of the 10 IMTs-CNS patients are summarized in Table 1. Tumor locations varied and included the supratentorial convexity, including the falx (n = 5), infratentorial convexity (n = 2), anterior cavernous sinus (n = lumbar spine (n = 1), and the orbit (n = 1). All intracranial tumors, except for one cavernous sinus mass (Patient 5), occurred as dura-based meningeal lesions with or without brain involvement (4 cases with cerebral parenchymal invasion and 2 cases with cerebellar parenchymal invasion), and no example of an isolated intraaxial tumor was presented. Radiologically, IMTs-CNS showed two different types of growth pattern, an isolated mass-forming type and en plaque-like type, which were accompanied by leptomeningeal thickening with or without infiltration into the brain parenchyma or perilesional brain edema (Fig. 1). In enhancing studies, all lesions were well enhanced except Patient 7, which showed multifocal enhancement. A case of orbital IMT presented as a mass-forming lesion with intracranial extension and the fibrohistiocytic (FHC) variant (Fig. 1). Four cases had synchronously multiple occurrences in intracranial sites, and the remaining 6 cases had a single lesion (Fig. 1). Combined orbital and mastoid involvements were found in 20% of our cases. All patients initially presented with neurologic manifestations depending on the location of tumor but were without constitutional symptoms. Laboratory data were normal in all except for an elevated erythrocyte sedimentation rate in Patient 2. Patient 3 had a temporal lesion and PCR for herpes simplex virus in cerebrospinal fluid was negative. Two cases (Patients 1 and 7) experienced intracranial recurrences in sites other than the original location; 1 recurred twice in the right mastoid and in the right cerebellar tentorium, with cerebellar parenchymal involvement 12 and 15 years after initial complete resection of the right orbital IMT and despite radiotherapy. The other case recurred in the infratentorial convexity 7 years after resection of a mastoid tumor in the same side. The histologies of the recurred tumors were identical to those of the primary tumors.
Pathologic Features
Like IMTs arising in other organs, IMTs-CNS showed a wide spectrum of histologic features. Depending on the predominant histology, IMTs-CNS could be classified into two types: the so- called plasma cell granuloma (PCG)-like type or the FHC variant (Tables 1 and 2; Fig. 2). The PCG-like type (Patients 1, 3, 4, 6, and 9) was characterized by predominant lymphoplasma cell infiltrations in vascular stroma with a minor component of myofibroblastic cells. Within the PCG-like type, the compositions of inflammatory cells varied. Plasma cells occasionally had two nuclei and an immature appearance. However, the CD79a-positive plasma cells were polytypic populations and expressed both κ and λ light chains. Patients 1, 3, 6, and 9 exhibited extensive lymphocytic infiltrations into the brain parenchyma through Virchow- Robin spacesas well as the leptomeninges. These 4 cases with both leptomeningeal and brain involvement were difficult to distinguish from lymphoproliferative disorders or infectious conditions. In Patient 3, the lymphoid cells were mainly small B cells with no atypism and low mitotic activity. They were heterogeneously positive for lymphoid markers, CD3 or CD20, but negative for CD10, Tdt, and CD5. The spinal IMT case (Patient 4) showed infiltrates of various inflammatory cells and vague granulomas.
FHC variant (Patients 2, 5, 7, 8, and 10) was characterized by distinct myofibroblastic proliferation with a minor population of mixed inflammatory cells. The myofibroblastic cells had oval nuclei with no pleomorphism and small indistinct or distinct nucleoli (Fig. 2). The hyalinization of stroma was variable from case to case.
As summarized in Table 2, SMA was positive in the most fibroblastic cells of all CNS and extra-CNS IMTs (control cases), but SMA positivity was absent or very rarely found in fibroblastic cells of the cerebral and spinal pachymeningitis (Fig. 2). ALK- positive cells were not found in any IMTs-CNS; however, 20% of the extra-CNS IMTs revealed ALK expression (2 of 10 cases) (Fig. 2). EMA, CD21, and c-kit were all negative in CNS- and extra-CNS IMTs; these were undertaken to rule out meningioma, follicular dendritic cell tumor, and stromal or embryonal tumor, respectively. No aberrant p53 expression was found in any IMTs-CNS. The Ki67 labeling index in infiltrating lymphocytes was moderately high (about 5%) but in myofibroblasts was low (
DISCUSSION
IMTs are enigmatic tumors that may involve virtually every organ but rarely originate in the CNS. IMT-CNS is more commonly called an “inflammatory pseudotumor,” which represents a simple reactive or benign inflammatory process. However, the biologic, pathologic, and clinical features of IMT-CNS have not been well established. Molecular genetic and immunohistochemical studies for ALK are extremely rare in IMT-CNS (7, 9).
TABLE 1. Clinical, Radiologic, and Histologic Features of IMTs- CNS
According to a review of 57 cases of inflammatory pseudotumors of the CNS by Hausler et al, they arise predominantly from dural/ leptomeningeal structures (60%) (7). They commonly manifest as intracranially and extra-axially with or without brain parenchymal involvement, or rarely as pure intraaxial (12%) or intraventricular lesions (12%). They also may extend from intracerebral to extracerebral sites (9%). Based on a review of 38 intracranial and spinal IMTs by Buccoliero et al, the maleifemale ratio was 7:3 and median age was 32 years, ranging from 5 to 76 years. Of these 38 cases, 82% presented as a single intracranial lesion, 3% as a solitary spinal lesion, and 16% as multiple synchronous or metachronous lesions (10). One patient with synchronous multiple IMTs showed involvement of the cerebral, cerebellar, brainstem, and spinal leptomeninges (11), and another patient with dual IMTs located in the spinal meninges and the cerebral falx (12). Even coexisting IMTs in the lung and CNS have been reported (13). The recurrence rate has been reported to be as high as 40% within 2 years, in general, especially following incomplete tumor resection (7, 14).
FIGURE 1. The multiple en plaque type (Patient 1) showing leptomeningeal thickening with enhancement and edema in the surrounding brain. (A) Contrast-enhanced, T1-weighted axial MR image. (B) T2-weighted axial MR image. (C) T2-weighted axial MR image of orbit. (D) Mass forming type; CT (Patient 5) exhibiting a wellenhanced, round mass (arrow) in the right cavernous sinus.
The present study shows that IMTs-CNS can arise everywhere in the CNS including the spinal cord but that they occur as dura-based, mass-forming lesions (60%) or en plaquelike lesions (40%). It also shows a male dominance (male: female = 6:4) and a mean age in the mid-fifth decade. All cases appeared as homogeneously well-enhanced lesions on MRI except 1 case that showed heterogeneous enhancement and an intratumoral hemorrhage. Associated orbital or mastoid involvements by IMTs-CNS were not uncommon; each was found in 20% of cases in the present study, which were synchronously or metachronously presented. Brain parenchymal infiltration was higher in our series (60% of cases) than that of Hausler et al review (16% of cases) (7).
Histopathologically, IMTs were classified in several ways by Coffin et al ( 1 ). They described three major histologic patterns, which depended on location. In nonpulmonary IMTs, they were of a myxoid/vascular pattern, a compact spindle cell pattern, and a hypocellular fibrous pattern; whereas in pulmonary IMTs, they were of an organizing pneumonia pattern with central hyalinization, a fibrous histiocytoma-like pattern, and a lymphoplasmacytic pattern. We modified the Coffin et al classification of pulmonary IMTs into a two-tiered classification according to predominant features, i.e. PCG-like type and FHC variant. The latter is characterized by distinct myofibroblastic proliferation with a minor inflammatory cell component, and the former as the reverse, i.e. by a distinct inflammatory cell component with slight myofibroblastic proliferation. The schema of nonpulmonary IMTs by Coffin et al is not applicable to IMT-CNS because myxoid/vascular or hypocellular fibrotic patterns were not found in our series, although it should be added that our study included a limited number of cases. Even though our series could be divided into two subtypes, a mixed pattern composed of relatively equal proportion of PCG-like and FHC component exists.
The histopathologic diagnosis of IMTs-CNS is not always easy. Important differential diagnoses might be intracranial plasmacytoma, lymphoma or lymphoproliferative disorder, idiopathic hypertrophic pachymeningitis (15), and a lymphoplasma cell-rich variant of meningioma. Myofibroblasts are spindle mesenchymal cells that have ultrastructural features in common with smooth muscle cells and fibroblasts. The confirmation of the presence of SMA-positive myofibroblasts is potentially important for the differential diagnoses from idiopathic hypertrophie pachymeningitis or meningioma because we also confirmed the lack of or very rare SMApositive myofibroblasts in 3 cases of pachymeningitis (Fig. 2) and meningiomas (data not shown). In addition, idiopathic hypertrophic pachymeningitis usually shows unique abortive granulomas with central basophilic necrosis (or smudging) and contains multinucleated but small giant cells and some neutrophilic infiltration, which were not found in our IMTsCNS (15). Unlike the lymphoplasma cell-rich variant of meningioma, there are neither meningothelial proliferation with whorls nor EMA expression in IMTs- CNS. The infiltration of polyclonal plasma cells and other mixed inflammatory cells is a helpful feature when differentiating IMT- CNS and lymphoma or plasmacytoma. However, in cases having IMTs-CNS with parenchymal involvement, it was difficult to establish a correct diagnosis, and in such cases, molecular studies may verify the polyclonality of infiltrating lymphoid cells. We also had diagnostic difficulties, and 1 case was initially misdiagnosed as plasmacytoma.
FIGURE 2. Plasma cell granulomalike type (Patient 9). (A) Predominant lymphocytic infiltrations with scant myofibroblastic cells (hematoxylin and eosin, original magnification = 100). (B) Underlying brain parenchymal involvement (hematoxylin and eosin; left, original magnification = 100; right, original magnification = 200). (C) Fibrohistiocytic variant (Patient 2) (hematoxylin and eosin; original magnification = 200). (D) Abundant SMA-positive myofibroblastic cells in fibrohistiocytic variant (left panel, Patient 2) (immunoperoxidase stain; original magnification = 200). Only scattered SMA (+) cells in the plasma cell granuloma-like type (right panel, Patient 3) (immunoperoxidase stain; original magnification = 400). (E) The spindle cells of the idiopathic hypertrophic pachymeningitis do not express SMA (immunoperoxidase stain; original magnification = 200). (F) ALK is not expressed in the cranial IMT (left panel, Patient 7) (immunoperoxidase stain; original magnification = 200). ALK is positive in the cytoplasm of myofibroblastic cells in the IMT of the urinary bladder (right panel, control case) (immunoperoxidase stain; original magnification = 200).
TABLE 2. Immunohistochemical Findings
Despite an apparently benign morphologic pattern, evolving evidence strongly suggests the neoplastic nature of IMTs, including the translocation and clonal gene rearrangement of anaplastic lymphoma kinase (ALK) gene on the chromosome 2p23, and biologic behaviors such as local recurrences, and rare distant metastases, even malignant transformation (16-20). In cases of IMTs-CNS, the recurrence rate was reported to be 12.5% to 40% (8).
The tropomyosin genes TPM3 and TPM4 (20), the clathrin heavy chain gene CLTC (21 ), the cysteinyl-tRNA synthetase gene CARS (22), and the Ran-binding protein 2 gene RANBP2 (23) have been identified as fusion partners of the ALK gene. There have been only a few sporadic case reports of ALK overexpression in IMTs-CNS; one was a frontal tumor (9) and the other was a spinal tumor (7). Interestingly, both suffered from multiple local recurrences. However, in our study, which is the largest series to study the immunohistochemical expression of ALK in IMTs-CNS, no ALK-positive cell was found in IMTs-CNS. However, of the control cases of the extracranial and extra-orbital IMTs, 20% expressed ALK. Coffin et al reported that the overexpression of ALK was most frequent in abdominal and pulmonary IMTs in the first decade of life and that this is associated with a higher frequency of recurrence (5). A single case of IMT-CNS that underwent malignant transformation has been reported (7). In our series, the histology of recurred tumors was the same as that of the original tumors. \Our study indicates that ALK-expression in IMT-CNS is rare, and it is possible that some other discrete biologic subset identical to neoplastic IMT is included in the category “IMT-CNS,” including “inflammatory pseudotumors” or “plasma cell granulomas of the CNS.” Close clinical follow-up is recommended, especially if surgical excision is incomplete, because recurrence may occur even 10 years later. The rearrangement of the ALK gene in CNS and orbital IMTs should be verified in more cases.
REFERENCES
1. Coffin CM, Watterson J, Priest JR, et al. Extrapulmonary inflammatory myofibroblastic tumor (inflammatory pseudotumor): A clinicopathologic and immunohistochemical study of 84 cases. Am J Surg Pathol 1995; 19:859-72
2. Coffin CM, Humphrey PA, Dehner LP. Extrapulmonary inflammatory myofibroblastic tumor: A clinical and pathologic survey. Semin Diagn Pathol 1998;15:85-101
3. Fletcher CDM, Unni K, Mertens F. Pathology and Genetics, Tumors of Soft Tissue and Bone. World Health Organization Classification of Tumors. Lyon, France: IARC Press, 2002:91-93
4. Cook JR, Dehner LP, Collins MH, et al. Anaplastic lymphoma kinase (ALK) expression in the inflammatory myofibroblastic tumor: A comparative immunohistochemical study. Am J Surg Pathol 2001;25:1364- 71
5. Coffin CM, Patel A, Peking S, et al. ALK1 and p80 expression and chromosomal rearrangements involving 2p23 in inflammatory myofibroblastic tumor. Mod Pathol 2001;14:569-76
6. Cessna MH, Zhou H, Sanger WG, et al. Expression of ALK1 and p80 in inflammatory myofibroblastic tumor and its mesenchymal mimics: A study of 135 cases. Mod Pathol 2002;15:931-38
7. Hausler M, Schaade L, Ramaekers VT, et al. Inflammatory pseudotumors of the central nervous system: Report of 3 cases and a literature review. Hum Pathol 2003;34:253-62
8. Tresser N, Rolf C, Cohen M. Plasma cell granulomas of the brain: Pediatric case presentation and review of the literature. Childs Nerv Syst 1996;12:52-57
9. Lacoste-Collin L, Roux FE, Gomez-Brouchet A, et al. Inflammatory myofibroblastic tumor: A spinal case with aggressive clinical course and ALK expression. J Neurosurg 2003;98:218-21
10. Buccoliero AM, Caldarella A, Santucci M, et al. Plasma cell granulomaan enigmatic lesion: Description of an extensive intracranial case and review of the literature. Arch Pathol Lab Med 2003;127:e220-23
11. Kilinc M, Erturk IO, Uysal H, et al. Multiple plasma cell granuloma of the central nervous system: A unique case with brain and spinal cord involvement: case report and review of literature. Spinal Cord 2002;40:203-6
12. Hsiang J, Moorhouse D, Barba D. Multiple plasma cell granulomas of the central nervous system: case report. Neurosurgery 1994:35:744-47
13. Le Marc’hadour F, Lavieille JP, Guilcher C, et al. Coexistence of plasma cell granulomas of lung and central nervous system. Pathol Res Pract 1995;191:1038-45
14. Brandsma D, Jansen GH, Spliet W, et al. The diagnostic difficulties of meningeal and intracerebral plasma cell granulomas: Presentation of three cases. J Neurol 2003;250:1302-6
15. Park SH, Whang CJ, Sohn M, et al. Idiopathic hypertrophic spinal pachymeningitis: A case report. J Korean Med Sci 2001;16:683- 88
16. Sciot R, Dal Cin P, Fletcher CD, et al. Inflammatory myofibroblastic tumor of bone: Report of two cases with evidence of clonal chromosomal changes. Am J Surg Pathol 1997;1:1166-72
17. Su LD, Atayde-Perez A, Sheldon S. et al. Inflammatory myofibroblastic tumor: Cytogenetic evidence supporting clonal origin. Mod Pathol 1998;11:364-68
18. Griffin CA, Hawkins AL, Dvorak C, et al. Recurrent involvement of 2p23 in inflammatory myofibroblastic tumors. Cancer Res 1999;59:2776-80
19. Biselli R, Boldrini R, Ferlini C, et al. Myofibroblastic tumors: Neoplasias with divergent behavior. Ultrastructural and flow cytometric analysis. Pathol Res Pract 1999;195:619-32
20. Lawrence B, Perez-Atayde A, Hibbard MK, et al. TPM3-ALK and TPM4-ALK oncogenes in inflammatory myofibroblastic tumors. Am J Pathol 2000;157:377-84
21. Bridge JA, Kanamori M, Ma Z, et al. Fusion of the ALK gene to the clathrin heavy chain gene, CLTC, in inflammatory myofibroblastic tumor. Am J Pathol 2001;159:411-15
22. Cools J, Wlodarska I, Somers R, et al. Identification of novel fusion partners of ALK, the anaplastic lymphoma kinase, in anaplastic large cell lymphoma and inflammatory myofibroblastic tumor. Genes Chromosomes Cancer 2002;34:354-62
23. Ma Z, Hill DA, Collins MH, et al. Fusion of ALK to the Ran- binding protein 2 (RANBP2) gene in inflammatory myofibroblastic tumor. Genes Chromosomes Cancer 2003;37:98-105
Yoon Kyung Jeon, MD, PhD, Kee-Hyun Chang, MD, PhD, Yeon-Lim Suh, MD, PhD, Hee Won Jung, MD, PhD, and Sung-Hye Park, MD, PhD
From the Departments of Pathology (YKJ, SHP), Radiology (KHC), and Neurosurgery (HWJ), Seoul National University College of Medicine, Seoul, Korea; and Department of Pathology (YLS), Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea.
Send correspondence and reprint requests to: Sung-Hye Park, MD, PhD, Department of Pathology, Seoul National University College of Medicine, 28 Yongon-dong, Chongno-gu, Seoul, 110-799, Republic of Korea; E-mail: [email protected].
Supported by a grant from the Seoul National University College of Medicine.
Copyright American Association of Neuropathologists, Inc. Mar 2005
Comments