Some of the smaller breeds of dog are prone to hydrocephalus, which can be associated with various health problems; William Thomas describes how to best diagnose the condition and the preferred options to treat and manage affected patients.
Pediatric hydrocephalus most commonly affects young, small breed dogs.
Diagnosis is based on skull conformation, neurologic deficits and brain imaging.
Ultrasound through a persistent fontanelle is a practical way to image the lateral ventricles.
Definitive treatment requires surgical placement of a ventriculoperitoneal shunt.
Hydrocephalus is defined as active distension of the ventricular system of the brain, caused by obstruction of flow of cerebrospinal fluid (CSF) (1). Produced at a constant rate by the choroid plexuses, the ependymal lining of the ventricular system, and blood vessels in the subarachnoid space, CSF circulates through the ventricular system into the subarachnoid space, where it is absorbed by arachnoid villi. Obstruction anywhere along this pathway causes active distension of the ventricular system, i.e., hydrocephalus. Several conditions, such as infarction, necrosis, and atrophy can result in decreased volume of brain parenchyma, in which the loss of brain tissue leaves a vacant space that fills passively with CSF; this was previously called hydrocephalus ex vacuo, but since there is no active distension of the ventricles it is not true hydrocephalus (1).
Hydrocephalus can be caused either by developmental abnormalities or acquired lesions, such as tumors or inflammatory disease. The site of obstruction influences the portion of the ventricular system that enlarges, with the distension typically occurring proximal to the obstruction – for example, obstruction of the third ventricle causes dilatation of both lateral ventricles but not the fourth ventricle. Hydrocephalus initially damages the ependymal lining of the ventricles, allowing water and larger molecules to leak into the adjacent white matter, causing periventricular edema. Further enlargement of the ventricles compresses the white matter, causing demyelination and axonal degeneration. The septum pellucidum separating the lateral ventricles can become fenestrated or completely destroyed, giving rise to one single large ventricle (Figure 1). The grey matter of the cerebral cortex is initially preserved, and at this point surgical placement of a shunt can result in re-expansion of the white matter and regeneration of remaining axons. In more advanced cases, the cortex becomes thin, with neuronal vacuolation and loss of neurons; once this happens the neuronal damage may persist even after shunting (2). With an acute obstruction, the volume of CSF can increase so fast that it causes increased intracranial pressure, which impairs blood flow to the brain and causes further brain damage.
Based on the age of onset, hydrocephalus can be classified as pediatric or acquired. Pediatric hydrocephalus is usually caused by developmental abnormalities, and clinical signs are often noticed by several months of age. Toy and brachycephalic breeds are at increased risk, including the Maltese Terrier, Yorkshire Terrier, English bulldog, Chihuahua, Lhasa Apso, Pomeranian, Toy Poodle, Cairn Terrier, Boston Terrier, Pug, and Pekingese (3). In most cases, an obvious site of obstruction is not apparent; if the lesion is at the level of the subarachnoid space or arachnoid villi it is difficult to detect. Another possibility is obstruction during a critical stage of development in which the obstructive lesion later resolves, leaving only the ventricular enlargement. Pediatric hydrocephalus may also be associated with other malformations, such as meningomyelocele, Chiari malformation, Dandy-Walker syndrome, and cerebellar hypoplasia.
Clinical signs of pediatric hydrocephalus include an enlarged, dome-shaped head with persistent fontanelles and open cranial sutures. However, not all patients with a persistent fontanelle have hydrocephalus and not all patients with pediatric hydrocephalus have a persistent fontanelle. Enlargement of the calvarium can be subjectively estimated by assessing if the most lateral aspect of the parietal bone extends laterally beyond the level of the zygomatic arch. There may be ventral or ventrolateral strabismus due to either malformation of the orbit or brainstem dysfunction (Figure 2).
Puppies with hydrocephalus are often unthrifty and smaller than normal. Common neurologic deficits include abnormal behavior and cognitive dysfunction, such as an inability to become house-trained. Visual deficits include unilateral or bilateral blindness with normal pupillary function (although it should be noted that the menace response may not develop until at least 4 weeks of age in normal puppies). Ataxia, seizures, circling, vestibular dysfunction and head/neck pain are also possible. The clinical course is variable and difficult to predict. Neurologic deficits can progress over time, remain static, or even improve after 1 to 2 years of age (4). Patients with very large lateral ventricles and a thin cerebral cortex are at risk of intracranial hemorrhage caused by minor head trauma that can tear bridging veins in the skull. This can result in chronic, subclinical hematomas or sudden neurologic deterioration due to intracranial hemorrhage (5).
Acquired hydrocephalus can develop at any age due to neoplasia, head trauma or meningoencephalitis. Neurologic deficits are similar to those in young animals, but if hydrocephalus develops after the cranial sutures have closed, malformation of the skull does not develop.
The diagnosis of hydrocephalus is based on clinical signs and brain imaging. Magnetic resonance imaging (MRI) is the best modality to assess ventricular size and identify obstructive lesions. Masses such as tumors, granulomas, and cysts may be identified, especially on post-contrast images. MRI is more sensitive than computed tomography (CT) in demonstrating small focal lesions, especially those in the caudal fossa. CT is usually sufficient for follow-up of previously diagnosed patients and those with an existing shunt. Ventricular size is usually assessed subjectively, but a ventricular: brain index provides a more objective measurement. Based on dorsal plane image, the maximum distance between the borders of the lateral ventricles is divided by the maximum width of the brain at the same level; a ratio > 0.6 increases the risk of clinically significant hydrocephalus (6) (Figure 3).
In dogs with a persistent fontanelle, ultrasound can determine obviously enlarged ventricles. Normal-sized ventricles appear as paired, slit-like anechoic structures, just ventral to the longitudinal fissure, on either side of the midline. Enlarged ventricles are easily seen as paired hypoechoic regions. With marked ventricular enlargement, the septum pellucidum that normally separates the lateral ventricles is absent and the ventricles appear as a single, large anechoic structure (Figure 4).
Periventricular edema is best appreciated on heavily T2-weighted fluid-attenuated inversion recovery (FLAIR) sequences, where the CSF is dark and white matter edema is bright (Figure 5). Periventricular edema is usually associated with acute hydrocephalus and increased intraventricular pressure, rather than chronic, relatively compensated hydrocephalus with normal intra ventricular pressure (7).
It is important to differentiate between hydrocephalus and ventriculomegaly secondary to brain atrophy. Atrophy is characterized as widening of the cerebral sulci and suarachnoid space (Figure 6). On the other hand, effacement of sulci, periventricular edema and rounding of the frontal portion of the lateral ventricles and ventral displacement of the third ventricle suggest hydrocephalus with increased intraventricular pressure.
If meningoencephalitis is suspected, an increased white blood cell count and protein content may be found on analysis of CSF. However, CT or MRI should be performed before sampling to identify any shifting of brain tissue, such as caudal cerebellar herniation, or other abnormalities that may increase the risk of CSF collection from the cerebellomedullary cistern. In some cases, it may be safer to collect CSF from an enlarged lateral ventricle through a persistent fontanelle. In patients with a fontanelle, an enlarged lateral ventricle can be punctured with a 25-gauge needle inserted at the lateral aspect of the fontanelle, avoiding the sagittal sinus on the midline. Ultrasound is helpful in determining the depth of the center of the ventricle. Approximately 2 mL of CSF can be safely removed in most patients.
Medical therapy is used when surgery is not an option or not indicated, and for short-term management of acute deterioration prior to surgical intervention. Several medications have been used to decrease CSF production and may provide temporary relief of clinical signs. Acetazolamide (10 mg/kg PO q8H) is a carbonic anhydrase inhibitor that decreases CSF production. Furosemide (1 mg/kg PO q24H) inhibits CSF formation to a lesser degree by partial inhibition of carbonic anhydrase. Omeprazole (0.5 mg/kg PO q24H) has been advocated to reduce CSF production, although studies in normal dogs have provided conflicting results (8, 9). Glucocorticoids are also commonly used to treat hydrocephalus in veterinary patients; one protocol is to administer prednisone at 0.25-0.5 mg/kg q12H until signs improve, then reduce the dose at weekly intervals until 0.1 mg/kg every other day. In patients with severe or rapidly progressing signs, removal of CSF is sometimes used as a temporary measure to decrease intraventricular pressure and to help predict which patients will benefit from surgical shunting.
Definitive treatment for the condition requires placement of a ventriculoperitoneal shunt to divert CSF from the ventricles to the peritoneal space. The presence of a fontanelle or enlarged ventricles with no clinical signs does not indicate the need for surgery. A young patient with neurologic deficits, ventriculomegaly and evidence of increased intraventricular pressure is a clear indication for shunting. Progressive ventriculomegaly over time is also an indication, unless it is secondary to cortical atrophy. Older patients with stable clinical signs and stable ventriculomegaly are generally not considered for treatment (10). Placement of a shunt may provide palliation of signs due to acquired hydrocephalus and should be considered in the case of obstructed CSF flow (e.g., tumor, inflammation).
A variety of shunt systems designed for human patients are available and have been used in both dogs and cats. Pediatric or low-profile versions designed for infants work well in small dogs. A shunt consists of three components; a ventricular catheter that is inserted into the ventricle, a valve, and a peritoneal catheter placed in the abdomen (Figure 7). A differential pressure valve is most commonly employed, which opens when the pressure difference across the valve exceeds a predetermined threshold. Most manufacturers provide fixed pressure valves in 3 or 4 categories, e.g., very low (< 1 cm H20), low (1-4 cm H20), medium (4-8 cm H20) and high (> 8 cm H20). Diaphragm valves are the most commonly produced and involve deflection of a silicone membrane in response to pressure. Some shunts employ a slit valve; these consist of one or more slits in the tubing (usually at the distal end) that open and close based on the thickness and stiffness of the tube material. In human patients distal slits are associated with a greater incidence of shunt obstruction due to omentum or proteinaceous debris (11). Externally adjustable (“programmable”) valves are also available that allow the clinician to percutaneously adjust the opening pressure as the patient’s clinical course changes.
No data are available to determine the ideal opening pressure in dogs. In human pediatric patients, low-pressure valves are more likely to fail than medium- or high-pressure valves, usually due to obstruction of the ventricular catheter as the ventricles decrease in size (12). For most cases, the surgeon should become familiar with a specific system and employ that product consistently.
Surgical technique is similar for all shunts, although details vary slightly based on the specific system. Aseptic technique and meticulous hemostasis are critical to minimize the chance of shunt infection and obstruction. The site of the incision over the skull is determined using preoperative brain imaging, so that the catheter tip will be placed in the center of the occipital horn or frontal horn, avoiding the choroid plexus. The abdominal incision is located 2-3 cm caudal to the last rib, about halfway between the lumbar spine and the ventral aspect of the abdomen. The patient is clipped and prepared from the head to the abdominal incision site. The ventricular catheter is placed into the ventricle through a burr hole in the skull and secured with sutures placed via one or two small holes in the bone. Securely anchoring the catheter to the skull is important to prevent dislodgement (Figure 8). A second incision is created to place the distal end of the shunt into the peritoneal cavity. A subcutaneous tunnel is created connecting the two incisions, and the shunt is pulled from the cranial incision through the subcutaneous tissues to the abdominal incision (Figure 9). The shunt tubing is secured to the abdominal wall (using non-absorbable suture secured to an anchoring clip, or tied in a finger-trap fashion to the tubing) and the abdominal muscle apposed with absorbable suture. The subcutaneous and skin incisions are closed routinely.
Preoperative antibiotics are occasionally continued for 7-10 days after surgery, but prolonged antibiotic therapy is not indicated in uncomplicated patients (13). Pain control is provided by injectable analgesics that are transitioned to oral medication, and any preoperative anti-seizure medications should be continued as needed. Two-view radiographs of the entire shunt from skull to abdomen should be obtained to serve as a reference for any future complications (Figure 10). Preoperative neurologic deficits usually resolve quickly, and patients are re-assessed within the first 2 or 3 months with ultrasound, CT or MRI to measure ventricular size and serve as a baseline for subsequent follow-up (Figure 11).
Potential complications include obstruction, over-drainage, infection and disconnection. Obstruction can occur at any point along the shunt system, but most commonly occurs at the ventricular catheter. Obstruction of the valve is less common and usually occurs soon after shunt insertion, presumably from blood or cellular debris. Kinking of the shunt system can also cause obstruction. In all cases, obstruction may cause recurrence of the original neurologic signs (13, 14, 15).
Over-drainage can result in collapse of the ventricle and cerebral cortex with accumulation of extra-axial blood or fluid; this is most common in patients with very large ventricles and a thin cerebral cortex. Subdural fluid accumulation is often asymptomatic, but a large or rapidly expanding hematoma can cause progressive neurologic deficits (16). Over-drainage probably also increases the risk of obstruction, because as the ventricle collapses the catheter can adhere to the ventricular wall or become embedded in the choroid plexus (17). In human patients, over-drainage can lead to very small ventricles and episodes of increased intracranial pressure and headache, known as slit ventricle syndrome or non-compliant ventricle syndrome. Very small ventricles allow the brain to almost completely fill the intracranial space, which decreases the ability to compensate for transient increases in intracranial volume. Episodes of pain can occur after shunting in dogs and may be similar to slit ventricle syndrome in human patients (14).
Shunt infections present as shunt obstruction, meningitis, or non-specific signs such as fever and lethargy (15, 16). Diagnosis is based on cytology and culture of CSF collected from the shunt system. Infection may resolve with 4 weeks of antibiotic therapy, chosen based on culture and sensitivity, (15) but persistent infection requires replacement of the shunt.
Occasionally shunt components can become disconnected, or the ventricular catheter can slide out of the ventricle, or the distal catheter can move out of the abdomen (13, 15, 16). These complications usually occur soon after placement and are detectable on plain radiographs. Approximately 72-85% of dogs treated with shunting have long-term improvement, whilst 15-20% of patients require shunt revision, usually due to shunt obstruction, fracture or migration (13, 14, 15).
Optimal management of hydrocephalus in small breed dogs is based on accurate diagnosis, which in turn is based on history, examination findings and brain imaging. Patients with mild signs can be managed with medical therapy, although the prognosis is difficult to predict. Definitive treatment is placement of a ventriculoperitoneal shunt, which is associated with a good outcome in the majority of patients.
- Rekate HL. A contemporary definition and classification of hydrocephalus. Semin Pediatr Neurol 2009;16:9-15.
- Yamada H, Yokota A, Furuta A, et al. Reconstitution of shunted mantle in experimental hydrocephalus. J Neurosurg 1992;76:856-862.
- Selby LA, Hayes HM, Jr, Becker SV. Epizootiologic features of canine hydrocephalus. Am J Vet Res 1979;40:411-413.
- Simpson ST. Hydrocephalus. In: Kirk RW, Bonagura, JD. (eds.) Current Veterinary Therapy X Small Animal Practice. Philadelphia: WB Saunders; 1989;842-845.
- Nykamp S, Scrivani P, DeLahunta A, et al. Chronic subdural hematomas and hydrocephalus in a dog. Vet Radiol Ultrasound 2001;42:511-514.
- Laubner S, Ondreka N, Failing K, et al. Magnetic resonance imaging signs of high intraventricular pressure – comparison of findings in dogs with clinically relevant internal hydrocephalus and asymptomatic dogs with ventriculomegaly. BMC Vet Res 2015;11:181.
- Naidich TP, Schott LH, Baron RL. Computed tomography in evaluation of hydrocephalus. Radiol Clin North Am 1982;20:143-167.
- Javaheri S, Corbett WS, Simbartl LA, et al. Different effects of omeprazole and Sch 28080 on canine cerebrospinal fluid production. Brain Res 1997;754:321-324.
- Girod M, Allerton F, Gommeren K, et al. Evaluation of the effect of oral omeprazole on canine cerebrospinal fluid production: a pilot study. Vet J 2016;209:119-124.
- Iantosca MR, Drake JM. Cerebrospinal fluid shunts. In: Albright AL, Pollack IF, Adelson PD, (eds.) Operative Techniques in Pediatric Neurosurgery. New York: Thieme; 2001;3-14.
- Cozzens JW, Chandler JP. Increased risk of distal ventriculoperitoneal shunt obstruction associated with slit valves or distal slits in the peritoneal catheter. J Neurosurg 1997;87:682-686.
- Robinson S, Kaufman BA, Park TS. Outcome analysis of initial neonatal shunts: does the valve make a difference? Pediatr Neurosurg 2002;37:287-294.
- Biel M, Kramer M, Forterre F, et al. Outcome of ventriculoperitoneal shunt implantation for treatment of congenital internal hydrocephalus in dogs and cats: 36 cases (2001-2009). J Am Vet Med Assoc 2013;242:948-958.
- Shihab N, Davies E, Kenny PJ, et al. Treatment of hydrocephalus with ventriculoperitoneal shunting in twelve dogs. Vet Surg 2011;40:477-484.
- de Stefani A, de Risio L, Platt SR, et al. Surgical technique, postoperative complications and outcome in 14 dogs treated for hydrocephalus by ventriculoperitoneal shunting. Vet Surg 2011;40:183-191.
- Kitagawa M, Kanayama K, Sakai T. Subdural accumulation of fluid in a dog after the insertion of a ventriculoperitoneal shunt. Vet Rec 2005;156:206-208.
- Sainte-Rose C, Piatt JH, Renier D, et al. Mechanical complications in shunts. Pediatr Neurosurg 1991;17:2-9.