External Hydrocephalus

in Two Cats
External hydrocephalus describes an accumulation of cerebrospinal fluid (CSF) between the
cerebral hemispheres and the overlying arachnoid membrane, rather than within the lateral ven-
tricles. Two young cats with encephalopathic signs were diagnosed with external hydrocephalus,
one via magnetic resonance imaging and one via computed tomography. Both cats had abnor-
mally large, broad heads, with no evidence of open fontanelles. A surgical shunt was placed in
each cat to divert the accumulated CSF within the cranial cavity to the peritoneal space. Both
cats improved dramatically soon after surgical shunting was performed, and they continue to do
well clinically, approximately 42 months and 8 months postoperatively, respectively.
J Am Anim Hosp Assoc 2003;39:567-572.

Curtis W. Dewey, DVM, MS, Introduction

Diplomate ACVIM (Neurology), Hydrocephalus is an abnormal accumulation of cerebrospinal fluid (CSF)
Diplomate ACVS
within the ventricular system of the brain.1-3 There are numerous classi-
Joan R. Coates, DVM, MS, fication schemes for hydrocephalus in dogs and cats, the majority of
Diplomate ACVIM (Neurology) which are based either on etiology (e.g., congenital versus acquired) or
suspected location of CSF flow interruption (e.g., obstructive versus
Julie M. Ducoté, DVM, nonobstructive).1-3 The location of CSF accumulation with respect to the
Diplomate ACVIM (Neurology)
lateral ventricles is typically not specified. Internal hydrocephalus refers
Joseph D. Stefanacci, VMD, to excessive CSF within the lateral ventricles of the brain; with the excep-
Diplomate ACVR tion of one dog,4 all reported cases of canine and feline hydrocephalus
have fallen into this category. External hydrocephalus is a condition in
Michael A. Walker, DVM, which the accumulated CSF is primarily in an extra-axial location, rather
Diplomate ACVR than within the lateral ventricles.4-7
(Radiology, Radiation Oncology)
This report describes the clinical aspects and imaging findings of two
Dominic J. Marino, DVM, young cats with external hydrocephalus. An analogous disorder of human
Diplomate ACVS infants, usually found in combination with abnormally large cranium size
(i.e., macrocephaly), is also discussed.

Case Reports

Case No. 1
This case was briefly mentioned in a previously published textbook
From Long Island Veterinary Specialists chapter.8 An 8-month-old, male castrated, Flame Point Siamese cat pre-
(Dewey, Stefanacci, Marino), sented to the Texas A&M University Veterinary Medical Teaching
163 South Service Road,
Hospital (TAMU-VMTH) with an 18-week history of intermittent
Plainview, New York 11803
and the Departments of Small Animal lethargy, abnormal behavior, pelvic limb weakness, and apparent head
Medicine and Surgery (Coates, Ducoté) pain. The owners described what they suspected was a short seizure
and Large Animal Medicine episode at the onset of the cat’s difficulties. At 6 months of age, the cat
and Surgery (Walker), had experienced a very prolonged anesthetic recovery following castra-
College of Veterinary Medicine,
Texas A&M University, tion and declaw procedures. During the period of the patient’s illness
College Station, Texas 77843-4474. prior to referral to the TAMU-VMTH, he had been treated on several
occasions with varying dose regimens of corticosteroids and antibiotics,
Doctor Ducoté’s current address is with little to no apparent effect. Approximately 1 week before referral,
Dallas Veterinary Surgical Center,
4444 Trinity Mills Road, Suite 203,
the cat had been started on oral prednisone (0.5 mg/kg body weight, q
Dallas, Texas 75287. 48 hours) by the referring veterinarian.

JOURNAL of the American Animal Hospital Association 567

568 JOURNAL of the American Animal Hospital Association November/December 2003, Vol. 39

Upon presentation to the TAMU-VMTH, the cat was

alert and responsive. The only notable physical abnormality
was a large, broad skull, which seemed disproportionate to
the patient’s body size [Figure 1]. No open fontanelles were
palpable. Neurological abnormalities included intention
tremors of the head, decreased menace responses bilaterally
with associated visual deficits, bilateral positional ventrolat-
eral strabismus, a hypermetric gait, decreased propriocep-
tive placing responses in the right pelvic limb, and cervical
spinal hyperesthesia evident on palpation of the neck. The
neuroanatomical localization was multifocal/diffuse
encephalopathy, primarily involving the forebrain (i.e.,
cerebrum/diencephalon) and cerebellum. A complete blood
count (CBC) and serum biochemical panel were within ref-
erence ranges. Serology was negative for feline leukemia
virus (FeLV), feline immunodeficiency virus (FIV), and
feline infectious peritonitis virus (FIPV). The patient was
anesthetized, and a computed tomography (CT) scan of the
brain was performed. The cerebral hemispheres were shift-
ed axially by extra-axial fluid accumulation, and the cere-
bellum was displaced caudally by accumulated fluid [Figure
2A]. The lateral ventricles appeared moderately enlarged.
On some of the CT images, it was apparent that a section of
the left occipital cerebral region was absent, at which point
the CSF of the left lateral ventricle was confluent with that
of the overlying subarachnoid space [Figure 2B]. The CT
findings supported a diagnosis of external hydrocephalus. A
CSF sample obtained from the cerebellomedullary cistern
was visibly xanthochromic, consistent with prior hemor-
rhage. Cytopathological evaluation of the CSF was consis-
tent with chronic hemorrhage with a suspected secondary
mixed-cell inflammatory response. The red blood cell
(RBC) count was 16 × 103 cells/µL. The white blood cell
(WBC) count was 256 cells/µL (reference range, 0 to 5
cells/µL), with 60% macrophages, 30% nondegenerate neu-
trophils, and 10% lymphocytes. The CSF protein level was
1,487 mg/dL (reference range, <30 mg/dL). A CSF corona-
virus antibody titer was negative (<1:25). Since the previ-

Figures 2A, 2B—Computed tomography (CT) scan of the

cat from Figure 1, demonstrating external hydrocephalus.
On the transverse (2A) and dorsal (2B) images, both cere-
bral hemispheres are laterally compressed, and a defect in
the left occipital lobe is apparent.

ously prescribed prednisone had not resulted in any obvious

clinical improvement, it was discontinued.
During the ensuing 3-week period, the cat experienced
several more episodes of lethargy and apparent head pain. At
this point, it was decided to surgically place a shunting
devicea to divert excessive CSF fluid from the cranial cavity
into the peritoneal space. The rostral end of the shunting
device was placed in the expanded subarachnoid space later-
al to the left cerebral hemisphere; the caudal end was placed
Figure 1—Photograph of an 8-month-old, Flame Point in the peritoneal cavity [Figure 3]. At the time of surgery, a
Siamese cat (case no. 1) with external hydrocephalus; CSF sample was obtained from the dilated subarachnoid
note the domed-shaped calvarium and macrocephaly. space in the cranial vault and submitted for evaluation. This
November/December 2003, Vol. 39 External Hydrocephalus 569

Figure 3—Right (R) lateral, postoperative radiograph of

shunt placement in the cat from Figures 1 and 2. The cra-
nial end (arrow) of the shunt is in the left side of the calvari-
um, and the caudal end extends into the peritoneal cavity.

CSF sample was also xanthochromic. The RBC count for this
second CSF sample was 1.29 × 103 cells/µL. The WBC count
was 6 cells/µL, with a distribution of 65% macrophages, 19%
Figure 4—Computed tomographic scan 6 weeks after
neutrophils, 13% lymphocytes, and 3% eosinophils. The CSF shunt placement in the cat from Figures 1 through 3. Note
protein level was 1,065 mg/dL. The cat recovered unevent- that the fluid accumulation is still present. The shunt is the
fully from surgery. Forty-eight hours after surgery, the radio-opaque object in the left subarachnoid space.
patient’s rectal temperature was elevated (104.2˚F), although
he was active, had a normal appetite, and his surgical inci-
sions were healing well. A CBC performed at this time was examination was normal. However, he had been experienc-
also within reference ranges. Oral amoxicillin/clavulanic acid ing generalized seizures with increasing frequency during
(22 mg/kg body weight, q 12 hours for 14 days) was institut- the 4 months prior to readmission. At the time of readmis-
ed. By day 5 following surgery, the cat’s rectal temperature sion, the cat was having approximately two seizures a day.
was near normal limits (102.7˚F), and he was sent home. The serum phenobarbital level at this time was 12.1 µg/mL.
At a recheck examination approximately 6 weeks after The cat’s phenobarbital dose was increased to 3.2 mg/kg
surgery, the owner reported that the cat had been acting nor- body weight, q 12 hours. A serum phenobarbital level 2
mally since discharge, with no further episodes of lethargy weeks later was 35.3 µg/mL. No further seizure activity was
or apparent head pain. Also, according to the owner, the cat observed following the increase in phenobarbital dose.
had been very playful and had an increased appetite. The cat
The owner was contacted via telephone approximately
had gained 0.73 kg since hospital discharge. The only neu-
42 months after initial presentation. The cat was still doing
rological abnormality at this time was decreased proprio-
well at this time and had not experienced any further seizure
ceptive placing reactions in both pelvic limbs. The patient
activity.
was anesthetized, and a repeat CT scan of the brain was per-
formed. Correct positioning of the rostral end of the shunt- Case No. 2
ing device was confirmed; however, there was no obvious
difference in the extent of subarachnoid CSF accumulation A 14-month-old, female spayed, domestic shorthaired cat
[Figure 4]. presented to Long Island Veterinary Specialists (LIVS) on an
Approximately 19 months later, the cat was readmitted to emergency basis after having several short focal seizures
the TAMU-VMTH for having experienced three generalized (i.e., facial muscle twitching) and one prolonged (approxi-
seizures during the 2 weeks preceding readmission. Other mately 30 minutes) generalized seizure. The patient was
than the seizure episodes, the patient was still doing well actively seizuring at the time of admission; the seizure was
neurologically. Oral phenobarbital was instituted at a dose halted after bolus administration of diazepam (0.5 mg/kg
regimen of 1.6 mg/kg body weight, q 12 hours. A serum body weight, intravenously). Upon physical examination, it
phenobarbital level measured 2 weeks later was 8.2 µg/mL was noted that the cat had a large, broad cranium, which
(therapeutic range, 15 to 45 µg/mL). Despite the low serum appeared disproportionately large in relation to her body
phenobarbital level, the cat had not experienced any further size. No open fontanelles were palpable. Abnormal findings
seizure activity. The dose regimen was not changed at this on neurological examination included obtunded mental sta-
time. tus, absent menace responses bilaterally, and apparent hyper-
The patient returned for evaluation of worsening gener- esthesia upon manipulation of the calvarium. Some of the
alized seizure activity approximately 7 months later, 29 patient’s neurological abnormalities were suspected to be
months after initial presentation. The cat’s neurological either postictal signs or related to recent diazepam adminis-
570 JOURNAL of the American Animal Hospital Association November/December 2003, Vol. 39

tration. The only abnormal blood-work finding was a blood The antibiotics to which both organisms were sensitive
glucose (BG) level of 318 mg/dL (reference range, 70 to 150 included gentamicin, amikacin, carbenicillin, piperacillin,
mg/dL). Because this BG result was based on blood pulled ceftazidime, cefotaxime, ciprofloxacin, orbifloxacin, and
during a seizure episode, a repeat BG was measured 6 hours imipenem.
later; at that time the BG level was 70 mg/dL. The cat’s vac- The patient recovered well from surgery and was eating
cination status was current, and serological samples submit- voluntarily and walking around her cage approximately 36
ted previously by the referring veterinarian for FeLV and hours after the shunting procedure. The cat was discharged
FIV were negative. Approximately 8 hours following the from the hospital 72 hours after surgery. The clindamycin
cat’s seizure episode, a repeat neurological examination was was prescribed for a 3-week period. After the culture/sensi-
performed. Neurological abnormalities identified at this time tivity results from the second CSF sample were available,
included obtunded mental status, bilateral horizontal nystag- oral enrofloxacin (2.9 mg/kg body weight, q 12 hours for 14
mus, positional ventrolateral strabismus on the left side, and days) was added to the clindamycin therapy. The cat was
circling to the left side. The neuroanatomical localization presented again to LIVS, 72 hours after discharge, after
was multifocal encephalopathy, involving both the forebrain experiencing four short episodes that were believed to be
(i.e., cerebrum/diencephalon) and the cerebellomedullary focal seizures. The episodes consisted of transient ataxia
region of the hindbrain. and hypersalivation. The cat was hospitalized, and the oral
The cat was anesthetized, and a magnetic resonance phenobarbital dose was increased (2.5 mg/kg body weight,
image (MRI) of her brain was obtained. A large amount of q 12 hours). Over the next 24 hours, the cat experienced
extra-axial fluid was evident, displacing the cerebrum axial- four, short (<30 seconds) generalized seizures.
ly, and the lateral ventricles were moderately enlarged Approximately 24 hours later, the patient was released from
[Figures 5A, 5B]. Meningeal enhancement was evident on the hospital.
the T1-weighted images with contrast (i.e., gadolinium) Ten days following hospital discharge, the cat was
[Figure 5C]. An apparent fluid line was evident on trans- rechecked at LIVS. The owners reported that the cat was
verse images, suggesting a difference in signal intensity doing well, with no further seizure activity. The only neuro-
between dorsal and ventral regions of the accumulated fluid logical deficit identified was absent menace responses bilat-
[Figure 5C]. The imaging diagnosis was external hydro- erally. The cat appeared to have some visual ability, however.
cephalus. A CSF sample was obtained from the cerebel- The surgical incisions were healed at this time.
lomedullary cistern for both cytopathological analysis and Six weeks later, the patient was readmitted to the hospi-
bacterial culture. The CSF analysis was supportive of a tal after having two short (<1 minute) seizures the evening
mild, mixed-cell pleocytosis. The WBC count was 6 prior to presentation. Other than an absent menace response
cells/µL (range, 0 to 5 cells/µL), with a distribution of 47% on the right eye, the cat’s neurological examination was
monocytoid cells, 29% neutrophils, and 24% lymphocytes. within normal limits. A serum phenobarbital level was
The RBC count was normal (2 cells/µL). The CSF protein measured and found to be 23.9 µg/mL (therapeutic range,
level was within reference ranges. The CSF culture failed to 15 to 45 µg/mL). The phenobarbital dose was increased to
yield any growth after 5 days. 3.2 mg/kg, q 12 hours. The cat had no more seizures over
Pending CSF culture results, the cat was started on oral the next 24 hours and was sent home.
clindamycin (14 mg/kg body weight, q 12 hours) and oral Approximately 8 months following initial presentation,
phenobarbital (1.8 mg/kg body weight, q 12 hours). The the owners were contacted by telephone. The cat was still
patient was alert when handled at this time, but was gener- doing well neurologically. She had experienced two short
ally somnolent, preferring to lay in lateral recumbency. The (<20 seconds) generalized seizures approximately 4 and 6
cat had to be hand-fed. Approximately 72 hours after admis- months after initial presentation.
sion to LIVS, the cat was anesthetized again, and a shunting
devicea was surgically placed to divert excess CSF from the Discussion
cranial vault to the peritoneal cavity. At the time of surgery, Hydrocephalus is not commonly encountered in cats. All
a second CSF sample and a meningeal biopsy were reported cases of feline hydrocephalus have been internal
obtained. The CSF sample was red-tinged and opaque, sug- hydrocephalus, in which the CSF accumulation is primarily
gestive of recent hemorrhage. The RBC count of this CSF within the lateral ventricles.1-3,8 The two cases of this report
sample was 4.83 × 106/µL. The WBC count was 4.6 × represent an unusual manifestation of feline hydrocephalus.
103/µL, with a distribution similar to that of peripheral External hydrocephalus is occasionally reported in
blood. Macrophages engulfing erythrocytes were also humans, most commonly in infants. In these cases, external
appreciated on CSF cytopathology. The CSF protein level hydrocephalus is usually associated with an abnormally
was 8,000 mg/dL (reference range, <48 mg/dL). The inter- large cranium. This latter malformation is called macro-
pretation of these CSF findings was subacute to chronic cephaly and is often the sole indication of a neurological
intra-arachnoid hemorrhage. The meningeal biopsy was abnormality.5-7,9-11 This combination of macrocephaly and
interpreted as moderate, chronic-active, fibrosing meningi- external hydrocephalus is typically mild clinically and usu-
tis. Two organisms, Escherichia coli and Pseudomonas ally resolves spontaneously by 2 to 3 years of age. In a small
aeruginosa, were cultured from the second CSF sample. percentage of human cases, however, the condition pro-
November/December 2003, Vol. 39 External Hydrocephalus 571

Figures 5A-5C—Magnetic resonance image (MRI) appear-

ance of external hydrocephalus in a 14-month-old domestic
shorthaired cat (case no. 2). The transverse T1-weighted
MRI scan (5A) demonstrated hypointense cerebrospinal
fluid (CSF). A transverse T2-weighted MRI (5B) depicts an
increase in signal intensity indicative of fluid, most likely
CSF. A contrast-enhanced, transverse T1-weighted MRI
scan (5C) revealed meningeal enhancement (arrow) and a
difference in signal intensity between dorsal and ventral
regions of CSF accumulation.

gresses to the point at which surgical shunting is required.5- An alternate theory suggests that external hydrocephalus is a
7,10,11 Macrocephaly/external hydrocephalus is believed to sequela to severe internal hydrocephalus; in this theory, CSF
be a heritable congenital syndrome.10 External hydro- accumulation within the lateral ventricle eventually leads to
cephalus has also been reported in adults, following sub- rupture of a region of the surrounding cerebral parenchyma.
arachnoid hemorrhage.6,12 The pathogenesis of external The fluid subsequently surrounds the cerebrum, shifting it
hydrocephalus is unknown. Most theories propose either a axially. There exists a rat model of hydrocephalus in which
congenital (e.g., developmental abnormality) or acquired this second scenario has been documented. In these rats,
(e.g., hemorrhage, inflammation) deficiency of the arach- severe internal hydrocephalus leads to rupture of the caudal
noid villi in their ability to absorb CSF as the reason for occipital pole of the cerebrum; external hydrocephalus
intracranial CSF accumulation. A lower resistance to CSF develops following this rupture.14 An area of possible cere-
flow within the extra-axial subarachnoid space in compari- bral parenchymal rupture in the caudal occipital region was
son with the ventricular system is the suspected reason for identified on the CT images of case no. 1.
preferential subarachnoid CSF accumulation. One view of The pathogenesis of external hydrocephalus in the two
the pathogenesis of external hydrocephalus is that it repre- cats of this report is unknown. The abnormally large head
sents an early, clinically mild stage of hydrocephalus that size of both cats would suggest they were affected by a
may eventually advance to become internal hydrocephalus.13 congenital syndrome similar to macrocephaly/external
572 JOURNAL of the American Animal Hospital Association November/December 2003, Vol. 39

hydrocephalus of humans. However, there are a number of Addendum

features of the feline disorder that are very dissimilar to the Shortly after submission of this manuscript, case no. 2 re-
human syndrome. In neither case could the disorder be con- presented to LIVS on an emergency basis. The cat had expe-
sidered mild or self-limiting, as is the rule for human macro- rienced two, short, generalized seizures within a 24-hour
cephaly/external hydrocephalus. It is possible that both cats period. Although the cat recovered from the seizures, the
of this report suffered from a severe form of macro- owner elected for euthanasia because of frustration over
cephaly/external hydrocephalus. If the feline condition is continued seizure activity and financial limitations. A
similar to that in humans, it is likely that the majority of necropsy was not permitted.
cases would not be clinically apparent. Pet owners and vet-
erinarians are unlikely to seek CT or MRI of a cat’s brains
based solely on the presence of large cranium size. The CSF a Accura shunt systems; Phoenix Biomedical Corp., Valley Forge, PA
analysis was abnormal in both cats, suggestive of subacute
to chronic hemorrhage. Cerebrospinal fluid is typically nor-
mal in cases of human macrocephaly/external hydro- References
cephalus.10 The subarachnoid hemorrhage in these cats 11. Harrington ML, Bagley RS, Moore MP. Hydrocephalus. Vet Clin
could represent either a cause of hydrocephalus (i.e., North Am (Sm Anim Pract) 1996;26:843-856.
12. Summers BA, Cummings JF, deLahunta A. Malformations of the
obstruction of arachnoid villi) or a consequence of rapidly
central nervous system. In: Summers BA, Cummings JF, deLahunta
developing hydrocephalus (i.e., tearing of stretched A, eds. Veterinary neuropathology. St. Louis: Mosby, 1995:68-94.
meningeal and cerebral blood vessels). In one cat (case no. 13. Dewey CW. Encephalopathies: disorders of the brain. In: Dewey
2), repeat CSF culture results were positive for Escherichia CW, ed. A practical guide to canine and feline neurology. Ames:
coli and Pseudomonas aeruginosa. Since the initial CSF Iowa State Press, 2003:99-178.
14. Dewey CW. External hydrocephalus in a dog with suspected bacteri-
culture was negative, and the cat did not exhibit “classic”
al meningoencephalitis. J Am Anim Hosp Assoc 2002;38:563-567.
signs (e.g., pyrexia, neck pain) of bacterial meningoen- 15. Boaz JC, Edwards-Brown MK. Hydrocephalus in children: neurosur-
cephalitis, the possibility exists that these organisms repre- gical and neuroimaging concerns. Neuroim Clin North Am
sent contaminants. However, in a recent study of bacterial 1999;9:73-91.
meningoencephalomyelitis of dogs, it was found that “clas- 16. Maytal J, Alvarez LA, Elkin CM, Shinnar S. External hydrocephalus:
radiologic spectrum and differentiation from cerebral atrophy. Am J
sic” signs of bacterial central nervous system infection are
Radiol 1987;148:1223-1230.
often absent.15 It is possible that there truly was an infec- 17. Andersson H, Elfverson J, Svendson P. External hydrocephalus in
tious component to the cat’s condition. Whether this sus- infants. Child’s Brain 1984;11:398-402.
pected infection was causative or contributory is unknown. 18. Coates JR, Sullivan SA. Congenital cranial and intracranial malfor-
A puppy with external hydrocephalus was recently mations. In: August JR, ed. Consultations in feline internal medicine.
4th ed. Philadelphia: WB Saunders, 2001:413-423.
described, in which a bacterial meningoencephalitis was
19. Modic MT, Kaufman B, Bonstelle CT, et al. Megalocephaly and
suspected as the etiology for the fluid accumulation.4 hypodense extracerebral fluid collections. Neuroradiol 1981;141:93-
Unfortunately, CSF culture was not pursued for case no. 1. 100.
The neurological status of both cats improved dramati- 10. Alvarez LA, Maytal J, Shinnar S. Idiopathic external hydrocephalus:
cally shortly after surgical placement of a shunting device. natural history and relationship to benign familial macrocephaly.
Pediatrics 1986;77:901-907.
A 6-week recheck CT scan of case no. 1 verified correct
11. Sandler AD, Knudsen MW, Brown TT, Christian RM.
shunt placement within the cranial vault, although there was Neurodevelopmental dysfunction among nonreferred children with
no appreciable difference in the amount of extra-axial fluid idiopathic megalencephaly. J Pediatr 1997;131:320-324.
accumulation. It is possible that expecting an obvious dif- 12. Cardoso ER, Schubert R. External hydrocephalus in adults: report of
ference in appearance after 6 weeks was premature. Another three cases. J Neurosurg 1996;85:1143-1147.
13. Robertson WC, Gomez MR. External hydrocephalus: early findings
possibility is that providing a continuous conduit for the
in congenital communicating hydrocephalus. Arch Neurol
excessive CSF allowed for functional improvement that was 1978;35:541-544.
not likely to be structurally apparent on CT or MRI scans. 14. Sasaki S, Goto H, Nagano H, et al. Congenital hydrocephalus
Long-term follow-up CT or MRI in these patients would be revealed in the inbred rat, LEW/Jms. Neurosurg 1983;13:548-554.
valuable in determining whether or not obvious structural 15. Radaelli ST, Platt SR. Bacterial meningoencephalomyelitis in dogs:
a retrospective study of 23 cases (1990-1999). J Vet Intern Med
changes are to be expected following surgical shunting.
2002;16:159-163.
The persistence of seizure activity in both cats following
surgery could be due to several factors. Since clinical signs
of neurological dysfunction other than seizures dramatical-
ly improved or resolved in both cats, shunt malfunction as a
cause of continued seizure activity is not likely. Permanent
brain damage from the CSF accumulation, leading to a
“seizure focus,” is a more likely explanation. The seizures
were eventually well controlled in both cats with phenobar-
bital therapy.