Subgaleoatrial shunt: Further progress in the management of iatrogenic cranial pseudomeningoceles
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/neuroindia.NI_881_16
Source of Support: None, Conflict of Interest: None
A postoperative or iatrogenic pseudomeningocele is a troublesome entity commonly encountered by the practicing neurosurgeon. The incidence of cranial pseudomeningoceles has been reported to be as high as 33% following posterior fossa surgery. A small proportion of pseudomeningoceles become persistent or recurrent in nature and pose the risk of wound dehiscence, cerebrospinal fluid (CSF) fistula formation, aseptic meningitis, intracranial hypotension, and rarely, death. Occasionally, they may cause significant pain due to their tense nature or produce diffuse or focal neurological deficits., Lastly, one must not overlook the psychosocial impact of the disfiguration. It is, therefore, imperative that such pseudomeningoceles be dealt with in a more aggressive manner.
A host of treatment options have been described in the literature. An initial conservative management includes elevation of the head, wrapping of the head with a compression bandage (either primarily or after tapping the pseudomeningocele under sterile precautions), and instituting a lumbar subarachnoid drain. When these methods fail, a direct repair of the dural defect or a lumboperitoneal (LP) shunt are commonly employed. In the rare occasion of pseudomeningoceles associated with hydrocephalus, a ventriculoperitoneal (VP) shunt usually results in resolution of both. A recent case series by Kiran et al., proposed the use of a subgaleoperitoneal (SP) shunt as a simpler, safer, and equally effective alternative to an LP shunt in the management of cranial pseudomeningoceles.
In the current study, we add to the array of treatment options for persistent or recurrent cranial pseudomeningoceles by describing the first successful use of a subgaleoatrial (SA) shunt in the treatment of a persistent cranial pseudomeningocele for a patient in whom CSF diversion procedures to the peritoneum were not an option.
A 22-year old male patient was brought to our emergency department with an exteriorized VP shunt draining CSF into a plastic bag. He had severe headache and multiple episodes of vomiting. On examination, he had a tense left temporal pseudomeningocele [Figure 1] and signs of meningitis.
At the age of 7 years, a cystoperitoneal shunt had been done for a left temporal arachnoid cyst elsewhere. A year before he presented to us, he was diagnosed to have a Chiari Type I malformation, for which a foramen magnum decompression and C1 posterior arch excision had been performed at another hospital. Following the operation, he developed a posterior fossa pseudomeningocele complicated by bacterial meningitis. The meningitis resolved after treatment with intravenous antibiotics. The pseudomeningocele subsided only after the placement of an LP shunt. Thereafter, he intermittently complained of headache, vomiting, and diplopia, and had a recurrence of the pseudomeningocele, a month later.
The LP shunt was revised at the same centre, after which his symptoms improved and the occipital meningocele subsided. In July 2015, he presented to a local hospital with altered sensorium and an episode of extensor posturing. He underwent a programmable VP shunt, as there was radiological evidence of hydrocephalus. A week after the surgery, he developed peritonitis leading to severe sepsis. An emergency laparotomy and peritoneal lavage was done. The previously inserted LP and cystoperitoneal shunts were removed and the VP shunt was exteriorized, as mentioned earlier. It was at this point that a tense left temporal meningocele developed. Cultures from the peritoneal fluid, VP shunt tip, and CSF grew Klebsiella spp. He was started on parenteral cefepime in accordance with the antimicrobial sensitivity report and was referred to our centre for further management.
The left temporal pseudomeningocele remained tense and painful despite free flow of CSF from the abdominal end of the VP shunt but completely subsided following insertion of a lumbar subarachnoid drain. We, therefore, removed the VP shunt completely. After completion of the course of cefepime, a repeat CSF culture once again grew Klebsiella spp., this time being carbapenem-resistant. Based on the sensitivity pattern, he was started on a 6-week course of tigecycline, colistin, and meropenem. During this period, it was determined that he was dependant on the lumbar drain because every attempt to clamp it resulted in headache and vomiting, and more importantly, in the enlargement of the pseudomeningocele. However, there was no recurrence of the earlier posterior fossa pseudomeningocele or development of ventriculomegaly.
As the option of CSF diversion to the peritoneum was ruled out, we performed a left temporal craniotomy, fenestration of the arachnoid cyst into the basal cisterns, and direct repair of the dural defect with fascia lata. Although the operation initially led to resolution of the temporal pseudomeningocele, it recollected within a week. It was then that we proceeded with a SA shunt. However, the pseudomeningocele recurred on the second postoperative day. We revised the shunt the same day and found that the cause of shunt failure was a tight silk tie around the facial vein. Following the shunt revision, the pseudomeningocele completely subsided [Figure 2] and he remained asymptomatic at follow-up after 9 months.
The low-pressure Chhabra shunt system (SH403, G. Surgiwear Limited, India) was used for the SA shunt. The cranial end was inserted into the pseudomeningocele in the subgaleal plane and tunnelled subcutaneously into the neck. In the neck, an oblique skin incision was made along the anterior border of the sternocleidomastoid muscle. It was deepened in layers till the carotid sheath. The internal jugular vein was dissected off the carotid artery and the facial vein was identified. The distal end of the shunt without the reservoir was passed via the facial vein into the right atrium of the heart and its position confirmed with fluoroscopy at the T5-T6 level. The cranial and atrial ends of the SA shunt were anchored to the periosteum of the skull and the facial vein, respectively, using 4-0 silk.
A pseudomeningocele is an epidural collection of CSF at the operative site following either cranial or spinal surgery. They can appear immediately after surgery in some patients and weeks-to-months later in others. The pathogenesis is yet to be clearly understood although various iatrogenic, traumatic, and congenital factors have been implicated. One of the simplistic mechanisms proposed to account for their formation is the inadequate closure of the dura. A meticulous closure of the dura is recommended, and when that is not possible, a dural graft should be used. Another mechanism described is the "ball-valve" phenomenon, wherein a tissue flap that allows only a one-way passage of CSF into the pseudomeningocele is formed. Lastly, there is a possibility that the pseudomeningocele is formed when CSF fills up a potential space created during surgery and thereafter enlarges when manoeuvres such as coughing, sneezing, or postural changes occur, which raise the intracranial pressure.
Quite often, pseudomeningoceles follow an episode of aseptic meningitis (a well described entity in posterior fossa surgery), implying an inflammatory etiology. It has been hypothesized that blood from the operative cavity causes inflammation of the meninges, ultimately resulting in the scarring of the subarachnoid space. This can affect the CSF drainage and give rise to hydrocephalus if the outlets of the fourth ventricle are blocked by the scarring. Our patient had bacterial meningitis which probably had a significant role in the formation of the pseudomeningocele. The increased pressure in the subarachnoid space may have caused the pseudomeningocele to gradually grow. However, a combination of all these factors are likely to be responsible for development of pseudomeningocele.
In our patient, we noticed that, despite adequate drainage of CSF via a VP shunt, there was no decrease in the size of the pseudomeningocele. However, once a lumbar subarachnoid drain was inserted, it rapidly resolved. The implication of this was that the pseudomeningocele was in communication with the subarachnoid space but not the ventricular system. Even after fenestration of the temporal arachnoid cyst into the basal cisterns, the pseudomeningocele failed to subside. This demonstrated the altered CSF dynamics in pseudomeningoceles and reaffirms the fact that a one-dimensional approach towards their treatment often results in failure.
Various treatment modalities have been described for pseudomeningoceles. The most basic include head-end elevation and pressure dressings. The technique of percutaneous tapping of the pseudomeningocele under sterile precautions is quite popular, albeit more prone to infection. In patients in whom these conservative measures fail, direct surgical repair of the dural defect with layer-by-layer closure is an option. Temporary measures such as therapeutic lumbar punctures or lumbar subarachnoid drains are occasionally sufficient. However, a majority of such patients finally require permanent diversion of CSF. Although it was established that our patient would benefit from CSF diversion from the subarachnoid space, the commonly performed LP shunt was not advisable in his case due to recent peritonitis. The recently described SP shunt prompted us consider attempting a SA shunt.
Ventriculoatrial (VA) shunts revolutionized the treatment of hydrocephalus in the 1950s. It was subsequently replaced by the VP shunt due to its better safety profile and ease of placement even though both were equally effective. Shunt infection and obstruction are the most common complications attributed to any shunt procedure, with the overall incidence of shunt dysfunction reaching 15% at 6 years in one series of adult patients. However, the worrying complications specific to VA shunts include pulmonary embolism, pulmonary hypertension, and shunt nephritis. Pulmonary embolism and pulmonary hypertension were seen to occur in 0.4% and 0.3% of patients with VA shunts, respectively. It is believed that pulmonary thrombi occur as a result of brain thromboplastin in the CSF mixing with venous blood. The cause of pulmonary hypertension is not fully understood but is possibly due to repeated endothelial damage by the thrombotic components.
It is reasonable to assume that these complications may occur in a SA shunt as well, implying that our patient will need to be kept on a close follow-up. On the other hand, the SA shunt, like the SP shunt, has the advantage of avoiding the ventricular system and subarachnoid space entirely. It, therefore, is free of drainage-related complications of LP shunts such as the intracranial hypotension syndrome, subdural hematomas, subdural hygromas, and tonsillar herniation.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]