Pseudomeningocele after in utero repair of myelomeningocele as an early sign of hypertensive hydrocephalus – a case report and review of the literature

Available at: http://www.archpedneurosurg.com.br/ Background: Myelomenigocele (MMC) is the most common congenital defect of the spine. The Management of Myelomeningocele study (MOMS trial) demonstrated that the prenatal repair decreased shunt implant, reversal of hindbrain herniation and better neurologic function compared to postnatal repair. Several ultrasound findings can predict the risk of postnatal hydrocephalus after intrauterine MMC repair. This report shows a prenatal pseudomenigocele after intrauterine correction of MMC as an early sign of hydrocephalus.


INTRODUCTION
Myelomeningocele (MMC) is the most common congenital defect of the spine and spinal cord [1]. MMC is characterized by sensorimotor deficits, hydrocephalus, neurogenic bladder, and cognitive impairment [2][3][4][5]. Currently, the benefits of treating MMC in utero are well established especially regarding the prevention of hydrocephalus [6]. The regression of Chiari malformation (CM) type II after in utero repair of MMC may be associated with changes in the dynamics of cerebrospinal fluid (CSF) flow and improvement in brainstem function [7,8]. Studies have evaluated the cephalic and intracranial parameters after in utero repair of MMC to determine the prognosis of hydrocephalus after birth [9]. However, the dynamics of CSF flow involves the entire neuraxis; and changes in the spinal cord canal prenatally may impact the risk of progression to hydrocephalus [10]. In this article, we present a case report of a child who underwent in utero repair of MMC, who developed bulging of the repair site (pseudomeningocele) prenatally despite the partial regression of CM type II. We discuss the possible pathophysiology of hydrocephalus and the value of pseudomeningocele as a prognostic factor in this setting.

CASE REPORT
A 34-year-old female, gesta 2 para 1, with a 5-year-old healthy child, was referred to our Fetal Center because of congenital spinal defect and ventriculomegaly in the fetus. The fetus had a prenatal diagnosis of MMC with anatomical level L4 and ventricular enlargement (posterior horn 14 and 15 mm) and met the eligibility criteria for intrauterine MMC correction [11]. The MMC surgery was performed at 24 weeks of gestation. The surgical technique used in the procedure was open surgery by mini-hysterotomy (4 centimeter). After fetal exposure, the neurosurgery team performed MMC closure under microscope visualization using the conventional 3-layer technique (dura mater, muscle fascia, and skin); no relief incisions were needed. The total procedure time was 4 hours, and there were no additional complications.
Follow up ultrasound examinations were performed every 2 weeks after in utero repair. Progressively, the ultrasound examination showed regression of the lemon sign and partial regression of CM type II (confirmed on prenatal MRI at 30 weeks of gestation), but progression of ventricular enlargement, especially after 28 weeks of gestation. (Figure 1).  Table 1 presents the ultrasound findings before and after in utero repair of MMC for this patient. Unexpectedly, ultrasound examination showed a progressive increase in the wound with local bulging and signs of externally thick tissue, suggesting integrity and concomitant expansion of the skin. This ultrasound finding was initially identified at 28 weeks of gestation ( Figure 2).  A 2.995 grams, male newborn was delivered by cesarean section at 37 weeks of gestation, with APGAR 9/10/10 and head circumference 34.5 cm). The newborn was transferred to neonatal intensive care unit, without needing respiratory support. Physical examination confirmed a large local bulging with distended skin and slight scarification at the bottom of the bulging sac, but without signs of CSF leakage. The newborn was referred for surgical correction of the lesion. Intra-operatively, we observed a distension of the skin with CSF content and, after reopening the skin, a small opening in the muscle fascia without exposure of the placode or dura mater ( Figure 3).
The muscle fascia was resutured and the skin was closed using layered anatomic closure. The patient remained under clinical observation without evidence of fontanelle bulging and/or worsening of ventricular enlargement. After 10 days, there was a CSF leakage through the incision of the surgical site repair associated with worsening of the ventricular enlargement. The clinical picture of hypertensive hydrocephalus was clear, VPS was indicated and the CSF leakage regressed after the procedure.

DISCUSSION
The pathophysiology of hydrocephalus in MMC remains unclear and still merits further studies. One of the most accepted theories for the development of hydrocephalus in these children is based on the unified McLone theory [12]. According to this theory, a continuous fetal CSF leakage during pregnancy induces the formation of a small and overcrowded posterior fossa with consequent obliteration of the outflow of the fourth ventricle, causing CM type II and aqueductal changes with consequent ventricular dilation. However, other theories have been proposed, such as venous hypertension of the posterior fossa structures caused by bone structural changes during mesenchymal development [13]. According to this theory the main factor causing CSF flow disturbance would not be obstructive leading to a communicating hydrocepahalus. Changes in the tectum of mesencephalon (beaking of the tectum) could also contribute to the genesis of ventricular enlargement due to aqueduct stenosis, which would explain the obstructive nature of hydrocephalus in these patients [14]. Contrast studies have shown the obstruction and possible stenosis at the level of the aqueduct of Sylvius. However, one-third of the studied patients have permeable aqueduct [15]. These theories reinforce the peculiar nature of hydrocephalus in MMC combining communicating and obstructive components on its genesis.
The MOMs trial showed clearly benefits of in utero repair MMC by demonstrating significant reduction of the frequency of postnatal hydrocephalus in newborns who underwent prenatal repair compared to those who underwent postnatal repair [11]. However, this randomized controlled trial showed that 40% of the newborns who underwent prenatal repair still required ventriculoperitoneal shunt (VPS). Indications for VPS are a combination of increased head circumference, bulging of the fontanelle, increased ventricular index or a cephalic percentile above 97.5 [11]. Other isolated indications include symptoms of CM type II, syringomyelia and bulging or CSF leakage in the surgical site repair. In a review of the MOMs trial cases, the non-regression of CM type II was shown to be the worst prognostic factor for hydrocephalus. In addition, ventricular enlargement > 15 mm was associated with a 79% need for VPS in the postnatal period [16]. The analysis of postoperative prenatal ultrasonography after MMC treatment could predict the postnatal prognosis of hydrocephalus and provide information on which subgroups of patients would have a higher risk of VPS after birth.
Studies have already shown the importance of CSF flow during the embryonic phase for adequate cortical formation and structuring of cerebral ventricles [17]. Morais et al. showed that ventricular enlargement is present in up to 94% Pseudomeningocele after in utero repair of myelomeningocele as an early sign of hypertensive hydrocephalus -a case report and review of the literature of the MMC patients in a magnetic resonance study of MMC postnatal related brain abnormalities [18]. This shows that most indications of VPS during the postnatal period are related to the presence of intracranial hypertension and ventricular enlargement. However, in a discussion about the failure of endoscopic treatment of hydrocephalus in patients with MMC, Marlin [19] emphasizes the importance of the communicating role in hydrocephalus. The theory is that although there is blockage in the outflow of the fourth ventricle in the subaracnoid spaces by Luscha and Magendie foramen obstruction, there is still permeability and CSF flow through the obex and central canal of the spinal cord, which explains the presence of syringomyelia in these patients.
In a meta-analysis conducted by Kabagambe et al. [6] the indications for VPS in children who have undergone MMC correction during the prenatal period ranged from 32% to 45%. In patients who underwent fetoscopy MMC fetal surgery, Graf et al. [20] described a need for wound re-suture due to CSF leakage in up to 35% after birth. Pedreira et al. [21] described syringomyelia in 30% of operated patients. Belfort et al. [22] described a case of an extremely premature (26 weeks) newborn who underwent in utero MMC treatment by fetoscopy with pseudomeningocele with CSF leakage requiring VPS in the postnatal period. However, there was no report of the appearance of pseudomeningocele on prenatal ultrasound. The lack of detail in the description of VPS indications and the high rate of complications of the surgical site repair of the MMC in most of the studies in the literature indicate that some cases described as technical closure failures could instead be a consequence of hypertensive hydrocephalus, especially if there is a combination of pseudomeningocele with ventriculomegaly [23].
In our patient, a progressive bulging of the surgical MMC site was observed prenatally after the fetal surgery, despite the partial regression of CM type II. Initially, the bulging of the surgical site repair observed on prenatal ultrasound was considered as a technical failure of MMC closure, but the possibility of disturbance of the CSF flow was not ruled out. Tamburrini et al. [24] reported that most neurosurgeons consider CSF leakage and surgical site repair bulging as signs of hypertensive hydrocephalus. However, the simultaneous treatment of the site repair and the performance of VPS could be associated with a higher rate of infectious complications [25]. For this reason and the absence of bulging of the fontanelle, it was decided to initially treat the site repair and monitor the enlargement of the ventricle.
The restructuring of the outlet of the fourth ventricle was not the only cause of hydrocephalus in our patient; and the changes in the subarachnoid space were responsible for communicating hydrocephalus. This hypothesis is reinforced by the fact that the CSF leakage occurred ten days after the correction of the pseudomeningocele. The bulging of the surgical MMC site repair in cases of fetal MMC correction is considered an absolute indication of VPS in the postnatal period, according to the MOMs trial criteria. However, the appearance of pseudomeningocele in the postoperative prenatal period has not yet been described, which indicates that some cases reported in other studies may have been described as CSF leakage or suture dehiscence. The pseudomeningocele combined with ventricular enlargement in this case was an early sign of hydrocephalus decompensation, and immediate treatment with VPS should be considered.
In conclusion, the presence of pseudomeningocele associated with progressive prenatal ventriculomegaly after in utero MMC repair may represent an early sign of hypertensive hydrocephalus present even before birth. Correction of pseudomeningocele alone may not be effective and treatment of hydrocephalus should be performed.