Rudimentary Meningocele: Remnant of a Neural Tube Defect?
Total Page:16
File Type:pdf, Size:1020Kb
STUDY Rudimentary Meningocele: Remnant of a Neural Tube Defect? Laila El Shabrawi-Caelen, MD; Wain L. White, MD; Hans Peter Soyer, MD; Bang-Soon Kim, MD; Ilona J. Frieden, MD; Timothy H. McCalmont, MD Background: Rudimentary meningocele, a malforma- munohistochemistry studies were performed for each case. tion in which meningothelial elements are present in the A panel of immunoperoxidase reagents (EMA, CD31, skin and subcutaneous tissue, has been described in the CD34, CD57, S-100, and CAM 5.2) was used to assess past under a variety of different terms and has also been lineage and to confirm the meningothelial nature of these referred to as cutaneous meningioma. There has been de- lesions. bate as to whether rudimentary meningocele is an atretic form of meningocele or results from growth of menin- Results: Recent evidence indicating a multisite clo- geal cells displaced along cutaneous nerves sure of the neural tube in humans suggests that classic meningocele and rudimentary meningocele are on a con- Objective: We reviewed the clinical, histological, and tinuous spectrum. immunohistochemical characteristics of rudimentary me- ningocele in an attempt to assess the most likely patho- Conclusion: Rudimentary meningocele seems to be a logic mechanism for it. remnant of a neural tube defect in which abnormal at- tachment of the developing neural tube to skin (compa- Design: Retrospective study. rable to that in classic meningocele) could explain the presence of ectopic meningeal tissue. In the majority of Setting: University hospitals. cases, no underlying bony defect or communication to the meninges could be detected. However, in light of the Patients: Thirteen children with rudimentary men- probable pathogenesis, imaging studies to exclude any ingocele. communication to the central nervous system should precede any invasive evaluation or intervention. Main Outcome Measures: Medical records were re- viewed and histopathologic examination as well as im- Arch Dermatol. 2001;137:45-50 UDIMENTARY meningocele is similar to that observed in experimental an uncommon develop- animals.10,11 Multiple distinct closure sites mental anomaly in which of the neural tube explain the most fre- meningothelial elements are quent locations of neural tube defects. Fail- displaced into the skin and ure to achieve complete fusion results in subcutaneous tissue. In the past it has been various forms of dysraphia. From the Department of R Dermatology, University Graz, described under a variety of different terms, We present 13 cases of rudimentary Graz, Austria (Drs El such as cutaneous meningioma, hamar- meningocele that show a strikingly simi- Shabrawi-Caelen and Soyer); toma (of the scalp) with ectopic meningo- lar anatomical distribution to the neural the Department of Pathology, thelial elements, sequestrated meningo- tube fusion sites observed experimentally. North Carolina Baptist cele, acoelic meningeal hamartoma, and The multisite closure model explains the Hospital–Bowman Gray School cutaneous heterotopic meningeal nod- congenital nature and distribution of these of Medicine, Wake Forest ules. 1-8 Controversy exists as to whether lesions and also coincides with the distri- University, Winston-Salem, this entity represents a form of meningo- bution of classic meningoceles. NC (Dr White); SNU Clinic cele in which the underlying connection to of Dermatology Seoul, the meningeal space is obliterated, or South Korea (Dr Kim); RESULTS and the Departments of whether lesions develop because of prolif- Dermatology (Drs Frieden eration of meningeal cells along the routes HISTOPATHOLOGIC FINDINGS 9 and McCalmont) and Pathology of cutaneous nerves. (Dr McCalmont), University of There is recent evidence for a multi- The microscopic findings are summa- California, San Francisco. site closure of the neural tube in humans rized in Table 2. In general, skin append- (REPRINTED) ARCH DERMATOL / VOL 137, JAN 2001 WWW.ARCHDERMATOL.COM 45 ©2001 American Medical Association. All rights reserved. Downloaded From: http://archderm.jamanetwork.com/ by a UQ Library User on 10/13/2015 PATIENTS AND METHODS The 2 patients with lesions along the spine showed more complex underlying abnormalities. Both had bifid PATIENTS laminae and tethered spinal cords. In one of these pa- tients, there was associated diastematomyelia and syrin- The clinical data are summarized in Table 1. All 13 pa- gomyelia. The patient also had attachment of the penis to tients were children from 0 to 6 years of age (median, 19 the scrotum (penis palma) with vesicoureteral reflux. In months). Eight were boys and 5 were girls. All lesions, vari- the second patient, a fibrovascular bundle pierced the dura ably described as patches, papules, nodules, or exophytic and extended into the subdural space (Figure 1C). masses, were congenital and ranged in size from 0.5 to 8 All defects were locally excised. No recurrence was ob- cm (Figure 1A). Some lesions had increased in size with served, and there was no evidence of postsurgical menin- the growth of the child. All were midline lesions and all gitis or cerebrospinal fluid leakage after a median fol- but 2 occurred on the scalp, preferentially involving the low-up of 28 months. occiput and vertex. The remaining 2 lesions occurred over- lying the cervical spine (C2-C3) and lumbosacral spine METHODS (Figure 1B). The “hair collar” sign, a ring of coarse hair sur- rounding the malformation and thought to be character- Formalin-fixed and paraffin-embedded tissue from 13 skin istic of cranial dysraphism, was found in 3 infants. Eleven lesions was examined in conventional and immunoperoxi- patients had solitary defects. In each of the remaining 2 cases, dase sections via a light microscope. Immunoperoxidase stains 2 lesions were apparent. Other clinical abnormalities in- were prepared with an avidin-biotin method using monoclo- cluded port wine stains at a separate location in 2 patients nal antibodies directed against epithelial membrane antigen and a congenital melanocytic nevus in patient 1. The clini- (EMA)(Dako,Capenteria,Calif),CD31(Biogenex,SanRamon, cal diagnoses included aplasia cutis, epidermal inclusion Calif), CD34 (Novocastra, Newcastle-upon-Tyne, England), cyst, congenital melanocytic nevus, hemangioma, skin CD57 (Becton-Dickinson, San Jose, Calif), and CAM 5.2 (Bec- tag, dermal sinus tract, meningocele, and encephalocele. ton-Dickinson),aswellaspolyclonalantibodiesdirectedagainst Small bony defects were identifiable in 2 patients and a S-100 protein (Dako). Two cases were evaluated further with connecting fibrous tract to the dura was identifiable in one anti–glial fibrillary acidic protein (GFAP) (Boehringer- based on radiologic imaging results or direct surgical vi- Mannheim; Indianapolis, Ind) and vimentin (Monosan; Uden, sualization. the Netherlands). ages, vessels, nerves, and melanocytes were quantified, lesion but showed strong expression in the deep portion. and remnants of brain tissue were sought. Psammoma Similarly, in specimen 5, which was transected at the level and collagen bodies and calcification were also sought. of the midreticular dermis, the loose network of pseudo- The overall architecture varied in that 8 lesions vascular spaces did not express EMA, but vimentin nicely showed replacement of both the dermis and subcutis by outlined these spaces. CD31 and CD34, known to be vas- the malformation, while in 4 lesions only 1 of the com- cular markers, did not react with these anastomosing chan- partments was involved. In case 5 the biopsy was too su- nels, confirming their nonvascular nature. perficial to exclude deep involvement. In some speci- Antibodies directed against S-100 protein, CD57, and mens a loose network of pseudovascular spaces in the CAM 5.2 failed to label meningothelial cells. S-100 was superficial dermis (Figure 2A) contrasted with dense used to quantify melanocytes, which were increased at collagenous tissue containing clusters of meningocytes the dermo-epidermal junction in 1 specimen and within in the deeper parts. Cystic structures lined by meningo- the rudimentary meningocele in 2 patients. Adjacent thelial cells were present in 5 cases (Figure 2B). The me- nerves were also identified via S-100 immunostaining in ningocytes, recognizable by their ovoid to spindle shape 4 patients. Anti-GFAP detected small nests of apparent with scant eosinophilic cytoplasm and small nuclei with glial tissue in patient 7 and also stained a few cells in pa- finely stippled chromatin and indistinct cell borders, tient 12. However, results of analysis for neurons with formed cords, strands, and small nests and had the ten- neuron-specific enolase and for axons with neurofila- dency to encompass collagen fibers and adnexal struc- ment in these 2 cases were negative, unlike what has been tures. We observed decreases as well as increases in ec- reported in most cases of heterotopic glial tissue. Me- crine glands and hair follicles. There were also distorted ningothelial elements like those observed in both cases follicles and an increase in apocrine glands and smooth are not a feature of heterotopic glial tissue, but rather fa- muscle bundles. In one case, a follicular cyst was evi- vor a diagnosis of rudimentary meningocele. dent. Prominent vessels were noted in 8 cases. Necrotic glial tissue was absent. Calcification was noted either in COMMENT the form of psammoma bodies (4 patients) or in an un- usual reticulate pattern (5 patients) (Figure 2C). The pathogenesis of human neural tube defects is com- plex and poorly understood.