Original Paper Fetal Diagn Ther 2010;28:28–33 DOI: 10.1159/000315299
Received: February 26, 2010 Accepted after revision: May 18, 2010 Published online: July 21, 2010
Etiopathology of Arnold-Chiari Malformation: A Fetal Rat Model of Dysraphism G. Corti T. Manzur C. Nagle M. Martinez-Ferro CEMIC, Centro de Investigaciones en Reproducción Humana y Experimental, Hospital Universitario Saavedra, Buenos Aires, Argentina
Key Words Arnold-Chiari malformation ⴢ Cerebrospinal fluid ⴢ Experimental fetal surgery
Abstract Objectives: We report an experimental fetal rat model with the aim of comparing two surgical methods used to check Arnold-Chiari Malformation (ACM) by dysraphism. We also wanted to (1) determine which type(s) of ACM akin to human anatomical findings were generated with the model and (2) study whether a cerebrospinal fluid pressure gradient could be responsible for ACM’s etiopathology. Materials and Methods: At E20, a mean of two fetuses per pregnant rat underwent an incision at the 2–3 lumbar level, deep into the medulla oblongata central canal, by two different surgical methods. Cesarian section was performed at E22. Dysraphic fetuses were examined clinically. Those born alive and controls without lesions were anatomically and histologically studied. Results: Method 2 was better than method 1 at reproducing the model. 100% of operated fetuses showed no spontaneous motility or sensibility to pressure on the posterior limbs in addition to anatomopathological evidence of type II ACM. Conclusions: A high rate of ACM could be checked by dysraphism with both methods. The opening of the central canal was demonstrated to generate a cerebro-
© 2010 S. Karger AG, Basel 1015–3837/10/0281–0028$26.00/0 Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/fdt
spinal fluid pressure gradient responsible for the herniation of encephalic structures comparable with human ACM. We believe this model may be useful for evaluating further strategies for prenatal treatment. Copyright © 2010 S. Karger AG, Basel
Introduction
Arnold-Chiari malformation (ACM) is a pathology which is difficult to diagnose due to its association with numerous and diverse symptoms, including dizziness, muscle weakness, numbness and headache, as well as vision, balance and coordination problems. According to magnetic resonance studies, it consists of malformations that produce a downward displacement of the cerebellum, and occasionally the brainstem, towards the spinal cord at least 3–5 mm below the foramen magnum. It is classified into four types. Type I consists of the herniation of cerebellar tonsils and is generally asymptomatic during childhood. Type II is usually accompanied by myelomeningocele (MMC) and can be diagnosed prenatally [1]. It consists of the downward displacement of the cerebellar tonsils, vermis and even the 4th ventricle into the foramen magnum. Hydrocephalus and syringomyelia are frequently present [2]. Type III and IV are inMarcelo Martinez-Ferro, MD Crámer 4601 (C1429AKK) Pediatric Surgery Department, Fundación Hospitalaria Ciudad Autónoma de Buenos Aires (Argentina) Tel. +54 11 4702 3116, Fax +54 11 4704 6006, E-Mail martinezferro @ fibertel.com.ar
a
b
c
Fig. 1. Creation of dysraphism at the 2–3 lumbar level, deep into the medulla oblongata central canal. a Exteriorization of the bicornuate uterus and of the inferior dorsal side of the fetus. b Incision with a microsurgical scissor with c observed loss of cerebrospinal fluid.
compatible with life, the first characterized by a complete descensus of the cerebellum, associated with encephalocele, and the latter by a lack of cerebellar development. Surgery is the only treatment available to correct functional disturbances or to halt the progression of damage to the central nervous system. More than one operation may be needed to treat the condition. Its etiopathology is not yet clear. It was originally considered a pure congenital syndrome that occurred during early embryogenesis. Numerous theories, however, describe a secondary origin [3–8]. Surgical [9–13] and non-surgical [14–16] animal models have been developed for the experimental study of MMC and ACM. The main aim of this study was to create an experimental fetal rat model in order to compare two surgical methods used to check ACM by dysraphism. The secondary objectives were to (1) determine which type(s) of ACM akin to human anatomical findings are generated and (2) study whether a cerebrospinal fluid pressure gradient could be responsible for ACM’s etiopathology.
Materials and Methods Fourteen virgin female Sprague-Dawley rats (age: 12 8 4 weeks) were cohabited with two male Sprague-Dawley rats per female. Day 0 of gestation was considered when sperm was found present in the vagina (vaginal smear: day 0, term: 22 days). Pregnant rats (n = 12) underwent laparotomy at E20 to expose the gravid bicornuate uterus. A mean of two fetuses per pregnant female were lesioned at the 2–3 lumbar level, deep into the medulla oblongata central canal, by two different surgical methods. Fetuses were classified into two groups (fig. 1). In group 1 (G1) dysraphism was created by a transuterine puncture with a 25-G nee-
Etiopathology of Arnold-Chiari Malformation
Ventral
Central canal Dorsal
Fig. 2. 63% of G1 (n = 5) showed histomorphological disorganiza-
tion of the medulla oblongata and thickening of the dura and dorsal pia mater. The white dotted line denotes the possible needle tract.
dle (method 1; n = 11). In group 2 (G2) dysraphism was created by an incision with microsurgical scissors (method 2; n = 15). Loss of cerebrospinal fluid was observed in the majority of the cases. At E22, the rats underwent cesarean sections. Dysraphic fetuses (n = 26) and controls (fetal rats without a dysraphic lesion; n = 26) were removed and examined clinically. The presence of vital signs and dysraphic signs, such as absence of spontaneous motility and sensibility to pressure on the posterior limbs, were registered. A pincer was used to evaluate the latter sign. All the animals were sacrificed with carbon dioxide. G1, G2 and controls born alive were fixed in Bouin liquid for 24 h. Dead fetuses were not considered since they showed signs of necrosis.
Fetal Diagn Ther 2010;28:28–33
29
Table 1. Clinical examination results
from dysraphic fetuses and controls
Variable
Vital signs Dysraphic signs
Table 2. Histological and
anatomopathological study of ACM in dysraphic fetuses born alive and controls
% of fetuses group 1
group 2
controls
73 (n = 8) 73 (n = 8)
80 (n = 12) 100 (n = 15)
100 (n = 26) 100 (n = 26)
Variable
% of fetuses
Opening of central canal Herniation of encephalic structures Cerebellomedullary cistern occlusion Modification of the brain ventricle’s spaces Silvio’s aqueduct stenosis Type II ACM Severe ACM Hydrocephalus Syringomyelia
Transversal cuts at the level of the lesion were evaluated for histological evidence of neural disorganization under a microscope magnification of !5 and !10. Sagital cuts of the craniocephalic region were compared for morphological evidence of ACM under a magnification glass of !10. Fetal rat ACM was defined as the downward displacement of the cerebellum towards the spinal canal (the cerebellum adopts the shape of a stake, thereby compressing the brainstem) and of the brainstem below the foramen magnum. The results were analyzed using descriptive statistics and Yates’ 2 test.
Results
73% of G1 (n = 8) and 80% of G2 (n = 12) were born alive. There were no dead controls. 27% of G1 (n = 3) and 100% of controls showed spontaneous motility and sensibility to pressure on the posterior limbs. None of G2 showed these signs (table 1). 63% of G1 (n = 5) showed histomorphological disorganization of the medulla oblongata and thickening of the dura and dorsal pia mater (fig. 2). In the remaining 27% (n = 3), there was a dorsoventral hemorrhagic area and an intact central canal (fig. 3). 100% of G2 (n = 12) showed complete disorganization of the medulla oblongata, with inflammatory infiltrates of white and grey matter (fig. 4). 30
Fetal Diagn Ther 2010;28:28–33
group 1
group 2
controls
63 (n = 5) 63 (n = 5) 63 (n = 5) 63 (n = 5) 63 (n = 5) 63 (n = 5) 25 (n = 2) 0 0
100 (n = 12) 100 (n = 12) 100 (n = 12) 100 (n = 12) 100 (n = 12) 100 (n = 12) 17 (n = 2) 0 0
0 0 0 0 0 0 0 0 0
The proportion of dysraphic fetuses born alive with central canal disruption was statistically higher in G2 than in G1 (p ! 0.05; table 2). The anatomopathological findings were cerebellomedullary cistern occlusion, modification of the brain ventricle’s spaces and Silvio’s aqueduct stenosis (fig. 5). Severe ACM was observed in 15% (n = 4) of the cases associated to brain anteroposterior detachment from the crane, thus leaving a space behind the bulbus olfactorius. The entire cerebellum, the brainstem and even the 4th ventricle were prolapsed below the foramen magnum. The medulla oblongata was seen to adopt a ‘Z’ shape (fig. 6), too. ACM was observed in 63% of G1 (n = 5) and in 100% of G2 (n = 12). No control was diagnosed with fetal rat ACM (table 2). The presence of ACM was associated with the presence of dysraphic signs (p ! 0.05), but was not associated with the absence of vital signs (p 1 0.05). The absence of vital signs was unrelated to the surgical method (p 1 0.05). The presence of ACM was associated with the anatomopathological findings in both G1 and G2 (p ! 0.05). The severe cases of ACM were not associated with the surgical methods (p 1 0.05). From the present investigation, neither hydrocephalus nor syringomyelia was found in the experimental rat fetuses or controls (table 2). Corti/Manzur/Nagle/Martinez-Ferro
Hemorrhagic area Lesion
Hemorrhagic area
Central canal
Fig. 3. In 27% of G1 (n = 3), there was a
dorsoventral hemorrhagic area and an intact central canal.
Discussion
Dysraphism is a defect in neural tube development, with MMC being one of its forms and ACM one of its associated pathologies, together with hydrocephalia and syringomyelia. Various experimental models have been developed in mice [17], rats [18], rabbits [12] and sheep [9] to study MMC physiopathology. Postulated mechanisms include trauma to the herniated neural elements or progressive injury from amniotic fluid exposure as gestation proceeds. A component of amniotic fluid that may contribute to neural injury is meconium. Experimental animal models have also been developed to evaluate the relevance of ACM malformations, clinical parameters and grade of histological lesions [10]. Based on earlier published experience, we generated a model of dysraphism that led to MMC at birth that closely resembled human MMC phenotypes in terms of morphology and functional deficit. Method 2 was better than method 1 at reproducing this model as 100% of operated fetuses showed no spontaneous motility and sensibility to pressure on the posterior limbs. Also, 100% of dysraphic fetuses born alive and operated on by method 2 showed histological evidence of central canal opening. The model enabled us to see a high rate of ACM by dysraphism (100% with method 2) and one of its possible causes as well. We believe the association between the Etiopathology of Arnold-Chiari Malformation
Site of lesion: medulla oblongata, skin and soft tissue destruction Fig. 4. 100% of G2 (n = 12) showed complete disorganization of
the medulla oblongata.
presence of ACM and dysraphic signs resulted from the irreversible damage of the medulla oblongata at the 2–3 lumbar level. As in MMC, damage of the medulla oblongata produced loss of nerve-related function below the defect. In those cases where the central canal could be Fetal Diagn Ther 2010;28:28–33
31
Fig. 5. Photographic images of craniocephalic sagital cuts of control (left) and dysraphic (right) fetuses. BO = Bulbus olfactorius; CC = cerebellomedullary cistern; CE = cerebellum; FM = foramen magnum; CR = brain; TE = brainstem; ME = medulla oblongata. The arrows indicate cerebrospinal fluid flow. Note, in the dysraphic fetus, the cerebellum adopts the shape of a stake thereby compressing the brainstem and occluding the cerebellomedullary cistern.
reached, a pressure gradient was generated as a consequence of cerebrospinal fluid loss. The downward displacement of encephalic structures provoked cerebellomedullary cistern occlusion, modification of the brain ventricle’s spaces and Silvio’s aqueduct stenosis. Cerebrospinal fluid flow was most probably blocked due to the prolapse of the encephalic structures. Severe ACM could result from a more aggressive incision and cerebrospinal fluid loss. The morphologic investigation revealed an ACM akin to human type II ACM because of the downward displacement of (1) the cerebellum towards the central canal and (2) the brainstem below the foramen magnum in addition to the association of the fetal ACM with the surgically created MMC. We postulate no hydrocephalia or syringomyelia will be found in the experimental rat fetuses since the animal model’s gestational period was not long enough to enable these pathologies to develop. In contrast to Hung [13], who also did not observe hydrocephalia in his fetal rat model of ACM with ethylenethiourea, we do not reject the hydrodynamic theory that human type II ACM can be induced by increasing intracranial pressure in hydrocephalus. Our experiment proved that the generation of a cerebrospinal pressure gradient can cause herniation of encephalic structures. In contrast to Hung, we observed Silvio’s aqueduct stenosis to be associated with every case of ACM, too. In conclusion, the described fetal rat model of dysraphism anatomically reproduced human type II ACM, thus facilitating investigation of ACM etiopathology. The opening of the central canal was demonstrated to generate a cerebrospinal fluid pressure gradient responsible for the herniation of encephalic structures. ACM could be 32
Fetal Diagn Ther 2010;28:28–33
Fig. 6. Photographic image of the craniocephalic sagital cut of a dysraphic fetus born alive with severe ACM. BO = Bulbus olfactorius; CE = cerebellum; B = brain; BS = brainstem; MO = medulla oblongata. Note the brain anteroposterior detachment from the crane, thus leaving a space behind the BO.
obtained by both a puncture and an incision deep into the medulla oblongata central canal, though the last method proved to be more efficient. We believe this model could be useful for evaluating further strategies for prenatal treatment. Even though the expectations for improved neurological outcome have not been fulfilled, preliminary findings suggest an improvement in hindbrain herniation [19]. Further research is still needed. Corti/Manzur/Nagle/Martinez-Ferro
References 1 Iruretagoyena JI, Trampe B, Shah D: Prenatal diagnosis of Chiari malformation with syringomyelia in the second trimester. J Matern Fetal Neonatal Med 2010;23:184–186. 2 Zderkiewics E, Kaczmarczyk R: Diagnostics of syringomyelia coexisting with ArnoldChiari syndrome (in Polish). Neurol Neurochir Pol 2008;42:37–42. 3 Tubbs RS, Shoja MM, Ardalan MR, et al: Hindbrain herniation: a review of embryological theories. Ital J Anat Embryol 2008; 113:37–46. 4 Roth M: Cranio-cervical growth collision: another explanation of the Arnold-Chiari malformation and of basilar impression. Neuroradiology 1986;28:187–194. 5 Barry A, Patten BM, Stewart BH: Possible factors in the development of the ArnoldChiari malformation. J Neurosurg 1957; 14: 285–301. 6 Marín-Padilla M, Marín-Padilla TM: Morphogenesis of experimentally induced Arnold-Chiari malformation. J Neurol Sci 1981;50:29–55. 7 Sathi S, Stieg PE: ‘Acquired’ Chiari I malformation after multiple lumbar punctures: case report. Neurosurgery 1993;32:306–309, discussion 309.
Etiopathology of Arnold-Chiari Malformation
8 Boyles AL, Enterline DS, Hammock PH, et al: Phenotypic definition of Chiari type I malformation coupled with high-density SNP genome screen shows significant evidence for linkage to regions on chromosomes 9 and 15. Am J Med Genet A. 2006; 140:2776–2785. 9 Heffez DC, Aryanpur J, Cuello-Rotellini NA, et al: Intrauterine repair of experimental surgically created dysraphism. Neurosurgery 1993;32:1005–1010. 10 Weber Guimaraes Barreto M, Martinez Ferro M, Guimaraes Bittencourt D, et al: Arnold-Chiari in a fetal rat model of dysraphism. Fetal Diagn Ther 2005;20:437–441. 11 Sutton S: Fetal surgery for neural tube defects. Best Pract Res Clin Obstet Gynaecol 2008;22:175–188. 12 Housley HT, Graf JL, Lipshultz GS, et al: Creation of myelomeningocele in the fetal rabbit. Fetal Diagn Ther 2000;15:275–279. 13 Hung CF: The relationship between hydrocephalus and Chiari type II malformation in the experimental rat fetuses with ArnoldChiari malformation. Proc Nati Sci Repub China B 1986;10:118–126.
14 Danzer E, Schwartz U, Wehrli S, et al: Retinoic acid induced myelomeningocele in fetal rats: characterization by histopathological analysis and magnetic resonance imaging. Exp Neurol 2005;194:467–475. 15 Briner W, Lieske R: Arnold-Chiari-like malformation associated with a valproate model of spina bifida in the rat. Teratology 1995;52: 306–311. 16 Diez-Pardo JA, Mariño JM, Baoquan Q, et al: Neural tube defects: an experimental model in the foetal rat. Eur J Pediatr Surg 1995; 5: 198–202. 17 Reis JL, Correia-Pinto J, Monteiro MP, et al: In utero topographic analysis of astrocytes and neuronal cells in the spinal cord of mutant mice with myelomeningocele. J Neurosurg 2007;106(Suppl 6):472–479. 18 Correia-Pinto J, Reis JL, Hutchins GM, et al: In utero meconium exposure increases spinal cord necrosis in a rat model of myelomeningocele. J Pediatr Surg 2002;37:488–492. 19 Fichter MA, Dornseifer U, Henke J, et al: Fetal spina bifida repair – current trends and prospects of intrauterine neurosurgery. Fetal Diagn Ther 2008;23:271–286.
Fetal Diagn Ther 2010;28:28–33
33
