ORIGINAL ARTICLE EAJNS 1(2):20-26

The effect of cerebrospinal fluid’s biochemical and

cellular properties on ventriculoperitoneal shunt
survival
Ben K Mutiso1, Michael Magoha1, Julius Kiboi1, John Boore3, Angela Amayo2
1 University of Nairobi, Department of Surgery
2 University of Nairobi, Department of Clinical chemistry
3 Kenyatta National Hospital, Department of Neurosurgery

Corresponding address: Dr Ben Mutiso, email: [email protected]

Abstract
Background: The burden of hydrocephalus in our region is huge. Ventriculoperitoneal (VP)
shunting is the main mode of treatment of hydrocephalus. These shunts frequently
malfunction, with shunt blockage being the commonest cause of shunt failure. The
biochemical and cellular properties of CSF are hypothesized to contribute to shunt blockage.
This study aimed to demonstrate the impact of these CSF properties on shunt survival.
Materials and methods: This was a prospective cohort study. Patients were recruited into
the study following VP shunt surgery for hydrocephalus. Based on their baseline CSF
biochemical and cellular properties, the patients were divided into two groups; those with
normal and those with abnormal CSF biochemistry and cellularity. Patients were then followed
up in the neurosurgical clinics for a period of 3months during which shunt function was
assessed clinically. This data was then entered to the statistical package for social sciences
(IBM SPSS statistics 25.0) for data analysis. Results: A total of 82 patients met the inclusion
criteria and were recruited into the study. Forty-six (56.1%) were male while 36(43.9%) of
them were female. The mean age was 15.5 months (SD 23.4 months). Majority (82.9%) of the
patients had congenital hydrocephalus with Dandy walker malformation being the most
common congenital anomaly seen. Most patients (52.44%) had normal CSF proteins while
the remainder had elevated CSF proteins. Majority of the patients had reduced levels of CSF
glucose at 65.9%. The CSF cell count was normal for most of the patients at 86.6%. Elevated
CSF protein concentration was associated with an increased likelihood of ventriculoperitoneal
shunt failure by 8.7 times compared to the patients with normal CSF protein concentration.
Reduced CSF glucose concentration was associated was also found to increase the likelihood
of shunt failure in this study. There was no correlation between the CSF cell count and the
likelihood of shunt failure.

Key words: Hydrocephalus, CSF biochemistry and cellularity, ventriculoperitoneal shunt

survival

Introduction proximal ventricular end can all block shunts.

Ventriculoperitoneal (VP) shunts routinely Kinking or disconnection of the shunt tubing
fail, with new research revealing that up to 30- can also cause catheters to become
40% of VP shunts fail within the first year of obstructed along the tract. Proteinaceous
placement (1). Shunt obstruction is the most debris might clog the catheters' lumens as
common cause of shunt failure (2). Our well. The omentum, adhesions, or
setting has one of the highest rates of shunt pseudocysts are the most common causes of
blockage (3,4). The reasons for our abdominal end blockage.
comparatively high failure rates in
comparison to the Western world have yet to The typical cerebrospinal fluid (CSF) protein
be determined. The ependymal lining, concentration is 0.2 to 0.4g/L. 10 High CSF
choroid plexus, or gliotic brain tissue in the protein concentrations are widely thought to

20 The East African Journal of Neurological Sciences www.theeajns.org Volume 1 Issue 2 - 2022
MUTISO ET AL. CSF AND VENTRICULOPERITONEAL SHUNT SURVIVAL
impede shunt function via a variety of ways signs of shunt failure to confirm the diagnosis.
(2). Despite the lack of strong data to support The duration of shunt survival, defined as the
this practice, clinical practice avoids placing number of days from the insertion of the
VP shunts in patients with increased CSF ventriculoperitoneal shunt to the diagnosis of
proteins in favor of alternative CSF diversion shunt failure, and the etiology of shunt failure,
methods (5). These include the use of defined as shunt blockage or shunt infection,
external ventricular drains or frequent were the primary outcomes of interest during
ventricular taps, both of which are associated this follow-up period. Mechanical causes of
with a significant risk of infection and the shunt failure such as shunt component
associated morbidity and mortality. The aim breakage, kinking, disconnection, or
of this study was to evaluate the biochemical migration were excluded from the study.
and cellular properties of CSF in
Results
hydrocephalic patients undergoing
ventriculoperitoneal shunting, as well as the Demographic Data
impact of these CSF parameters on outcome.
A total of 82 patients met the inclusion criteria
Materials and methods and were enrolled into the study. There were
46 (56.1%) males and 36 (43.9%) females.
This was a prospective cohort study
The patients' ages ranged from 3 weeks to 10
conducted at Kenyatta National Hospital in
years, with a mean of 15.5 months. The
Kenya. Patients were enrolled in the research
after undergoing VP shunt surgery for majority of the patients (78.04%) were under
the age of one year (Figure 1).
hydrocephalus. Data obtained upon
recruiting included study identification, age,
gender, residence, and the etiology of
hydrocephalus. In addition, data on CSF
biochemistry, including proteins and glucose,
as well as cellularity, was extracted from
patient files. It is routine practice in our setting
to collect CSF during venticuloperitoneal
shunt placement for biochemical and
microscopic analysis. This is common
practice in the hospital for people with
hydrocephalus. The patients were split into
two groups based on their baseline CSF
biochemical and cellular properties: normal
and abnormal.
Patients were subsequently followed up on in
neurosurgical clinics for three months, at Figure 1: A bar chart illustrating the patient
which time shunt function was evaluated. The distribution by age
data collection tool was utilized to screen for
shunt failure, which included clinical signs Etiology of Hydrocephalus
and symptoms as well as imaging findings. The etiology of hydrocephalus was
Clinical signs of shunt obstruction included a determined by clinical history, laboratory
bulging fontanelle, fluid collecting along the tests and radiologic tests performed. Most of
shunt tract, a reduced level of consciousness, the patients had congenital hydrocephalus
agitation, abdominal pain, nausea, and (82.9%). The commonest congenital
vomiting. CSF pleocytosis in the presence of intracranial malformation leading to
fever was utilized to establish a diagnosis of hydrocephalus was Dandy walker
shunt infection. A CT scan imaging of the malformation accounting for about 36.76% of
head was performed on patients with clinical all patients with congenital hydrocephalus

21 The East African Journal of Neurological Sciences www.theeajns.org Volume 1 Issue 2 - 2022
MUTISO ET AL. CSF AND VENTRICULOPERITONEAL SHUNT SURVIVAL
(Figure 2). Eighteen (21.9%) of the patients patients in this study who had increased CSF
had spina bifida as well. The rest (78.1%) had proteins or low CSF glucose. Shunt failure
hydrocephalus with no associated spinal occurred on average 60 days after shunt
neural tube defects placement in patients with increased
proteins. At three months after insertion, 20%
Type of surgery
of the patients with increased CSF proteins
The majority of patients, 76(92.7 %), had shunt failure, while only 2.3 percent of the
underwent primary shunt placement, patients with normal CSF proteins had shunt
whereas 6(7.3 %) underwent shunt revision. failure. According to the findings of this study,
The majority of shunt surgeries (96.3 %) were increased CSF proteins increase the risk of
conducted by residents, the majority of whom shunt failure by 8.7 times in the first three
were in their sixth year of residency (62.2 %). months after insertion.It was also observed
The rest were inserted by consultant that the protein concentration in individuals
neurosurgeon (3.6 %), fifth year residents who experienced shunt failure was
(7.32%), and fourth year residents 1250mmol/L, compared to a mean protein
(30.49%).In all the patients in the study concentration of 545mmol/L in patients
population, a medium pressure Chhabra whose shunts were functioning at 3 months.
shunt was used with the distal shunt catheter
The outcomes of the patients post recurrence
placed in the peritoneum.
were evaluated using the Eastern
Laboratory Findings Cooperative Oncology Group (ECOG) /
WHO/ Zubrod score. An outcome of 0-1 was
The majority of patients (52.44%) had normal defined as a good outcome, 2 as moderate
CSF proteins, while the rest had increased outcome, 3 as severe outcome, 4 as
CSF proteins. The reference range chosen vegetative state and 5 as death. The most
was 150-400mMol/L. The mean CSF protein frequent outcome was moderate at 38.89%
content was 588.3mMol/L (range from 5 to (Figure 3).
2045mMol/L). Most of patients (65.9%) had
low CSF glucose levels, with the mean CSF Two patients died before intervention at
glucose content being 2.42mMol/L. The CSF recurrence. One patient presented with an
cell count was normal in 86.6% of the infratentorial papillary adenocarcinoma
patients. A cell count of more than 5 cells per whereas the other patient had a pineal gland
high power field was regarded as abnormal. tumor. The outcome of death was equal in
both males and females and was significantly
Shunt failure more if the tumor recurred within the first 6
Eight (9.8%) of the patients had their shunts months. Surgical intervention post recurrence
fail in the 3 month follow up period The resulted in patients having better outcomes
average shunt survival for the shunts that relative to other patients who underwent
failed was 60 days. Failure of all these shunts other therapies.
was attributed to shunt blockage. Analysis of data
Factors associated with Shunt Failure There was significant difference (p = 0.047)
Elevated CSF protein levels and decreased between age and outcome as assessed by
CSF glucose levels were significantly Fisher-Freeman-Halton exact test. The
associated with shunt failure at 3 months significance tests carried out between the
(Table 1). Other variables investigated were other variables were not significant. The
not shown to be significantly related to shunt results of the exact test are displayed in Table
failure. Shunt failures were seen in all 1.

22 The East African Journal of Neurological Sciences www.theeajns.org Volume 1 Issue 2 - 2022
MUTISO ET AL. CSF AND VENTRICULOPERITONEAL SHUNT SURVIVAL
Table 1: Factors associated with shunt failure

Shunt Failed p-value

No Yes
Gender Female 30 6 0.062
Male 44 2
Etiology Acquired 13 1 0.717
Congenital 61 7
Spina bifida No 57 5 0.284
Yes 15 3
Type of Shunt Surgery First 68 8 0.403
Revision 6 0
Resident year of training Year 4 24 1 0.472
Year 5 5 1
Year 6 45 6
CSF proteins Elevated 31 8 0.002
Normal 43 0
CSF Glucose Abnormal 46 8 0.046
Normal 28 0
CSF Cellularity Abnormal 9 2 0.311
Normal 65 6

Figure 2: Frequency of various causes of congenital hydrocephalus in the study population

Discussion incidence is higher than that of sub-Saharan

Africa which is estimated at 0.14 per 100 live
Hydrocephalus remains a prevalent births. From a global perspective, this is the
neurosurgical disorder in Kenya with more
highest incident rate reported with Latin
than 80 children treated in KNH in 2 months.
America and North America reporting rates of
Crude estimates of the prevalence of 0.32 and 0.07 per 100 live births respectively
hydrocephalus in the East African region are
(6).
between 0.47 to 0.60 per 100 live births with
an estimated incidence of between 4900 and There is an observed male predominance in
8200 new cases per year in Kenya. This the patients included in this study, with

23 The East African Journal of Neurological Sciences www.theeajns.org Volume 1 Issue 2 - 2022
MUTISO ET AL. CSF AND VENTRICULOPERITONEAL SHUNT SURVIVAL
46(56.1%) being male while 36(43.9%) of populations in each of the studies with the
them were female. Kitunguu et al in 2017 later study having included patients of all age
reported a similar pattern of male groups while the former only studied pediatric
predominance with 55.8% of the patients in hydrocephalus. Additionally, an earlier study
his study being male. Previous studies in in 1989 by Gichuhi et al on pediatric
KNH by Noorani et al in 2003 and Omulo et hydrocephalus in KNH found congenital
al in 1993 have had similar findings as well hydrocephalus to be the commonest entity
(7). Worldwide, a similar observation has (9).
been made in most parts of the world except
In the current study, 18(21.9%) of the patients
in Mozambique where the male to female
had spina bifida. Of the these, 10 (55.5%) had
ratio is 1:1.
elevated CSF protein while 8 (44.5%) had
As regards the radiologic investigation of normal CSF proteins. There was no
children presenting with features of correlation between elevated CSF protein
hydrocephalus, the majority (91.46%) of and the presence of spina bifida in this study.
patients had cranial CT scan as the This variable could not be found in existing
diagnostic imaging work up performed with literature both locally and internationally. The
only 4.8% of the patients having magnetic presence of spina bifida did not affect shunt
resonance imaging performed in the initial survival at 3 months in the present study.
work up for hydrocephalus. Imaging tools
For the large majority of patients in this
available to the clinician in the initial
study,76(92.7%) shunt surgery performed
assessment of hydrocephalic children include
during the study period was the first-ever
ultrasound, CT scans, and magnetic
shunt insertion. This finding is in keeping with
resonance imaging. The modality of choice in
the findings of previous studies in the same
imaging pediatric hydrocephalus is MRI, as it
institutions where primary shunt insertions
provides detailed anatomical and functional
account for 75-95% of children undergoing
information important in managing these
shunt surgery (7,9,11,12).
patients (8). The reason most patients in this
study underwent CT scan, which is not the Most (96.3%) shunt surgeries in this study
gold standard imaging modality, is likely due were performed by neurosurgery residents.
to its wide availability, its relatively lower cost, D. Cochrane et al in 2005 reported that there
and ease of performing in children without is a relationship between the surgeon’s
requiring sedation, compared to MRI. operative experience and the shunt surgery
outcomes with higher shunt failure rates at 6
Congenital hydrocephalus was the most
months after surgery in less experienced
common entity, diagnosed in 89.2% of the
surgeons (13). Anderson et al in 2021, in a
patients in this study with Dandy walker
prospective study, reported that the presence
spectrum of malformations being the most
of a pediatric neurosurgeon during shunt
identifiable congenital anomalies in 36.76%
surgery had better 30-day outcomes than
of those with congenital hydrocephalus. A
surgeries performed by residents alone.39. In
similar finding was recorded by Omulo et al in
the year 2000 in a study conducted in KNH the current study, there was no statistically
significant difference in 90-day outcomes
with 74.4% of the patients he had studied
between shunt outcomes performed by
having congenital hydrocephalus (9).
residents in different stages of their training.
Kitunguu et al in 2017 however had a different
This finding is in keeping with that of a
observation in a study conducted in the same
previous study conducted in the same
institution, with a majority (57.1%) of the
institution in 2010 by the principal investigator
patients in his study having post-infectious
in this study (11).
hydrocephalus (10). This difference in
etiology of hydrocephalus among patients With regards to the CSF protein content,
managed in the same institution can be 52.4% of the patients had normal CSF
attributed partly to the different patient proteins while 47.6% had elevated CSF

24 The East African Journal of Neurological Sciences www.theeajns.org Volume 1 Issue 2 - 2022
MUTISO ET AL. CSF AND VENTRICULOPERITONEAL SHUNT SURVIVAL
proteins. This variable could not be found in 3 months. A similar observation was made in
literature both locally and internationally for neonates undergoing VP shunting after
comparison. This is an important finding that intraventricular hemorrhage and also in
highlights that nearly half of the patients patients with tuberculous meningitis
undergoing shunt insertion for non-tumor undergoing VP shunting (14).
hydrocephalus in our hospital have elevated
There was a positive correlation between
CSF protein. It was further noted that
reduced CSF glucose and the risk of shunt
elevated CSF proteins increase the likelihood
failure at 3 months. studies in neonates with
of shunt failure in the first 3 months after
post hemorrhagic hydrocephalus (1) and with
insertion by 8.7 times.This study thus shows
post tuberculous hydrocephalus (14) have
the impact that elevated CSF proteins have
had a contrary finding with both concluding
on shunt survival. Elevated CSF proteins has
that there was no link between CSF glucose
been reported to increase the likelihood of
level and shunt failure. It is also important to
shunt failure due to shunt blockage in
note that all patients with shunt failure at 3
patients with tuberculous meningitis (5, 14).
months and low CSF glucose at shunt
In an in vitro study established that the
insertion also had elevated CSF proteins.
average incubation period to shunt blockage
The author here hypothesizes that partial
attributed to high CSF proteins was 46days
treatment of preceding meningitis could
(2). In this current study, the average
explain this finding of reduced CSF glucose
incubation period was 60days. A study by
with elevated proteins. Further studies to
Kitunguu et al (10) in the same institution
validate this finding are therefore warranted.
showed that among the shunts that failed, the
average incubation period was 67 days. Conclusion
The overall shunt failure rate in this study was Shunt blockage remains the leading cause of
9.8% at 3 months of follow-up. All shunt early shunt failure in KNH. About half of the
failures were attributed to shunt blockage. A pediatric patients undergoing
previous prospective study in the same ventriculoperitoneal shunting in KNH have
institution reported a shunt blockage rate of elevated CSF proteins. Elevated CSF
10% (7). A retrospective study by Mwachaka proteins or reduced CSF glucose at the time
et al in 2010 in the same institution reported of surgery for shunt insertion were shown to
that shunt blockage accounted for a majority increase the likelihood of shunt failure by 3
(53.8%) of the shunt complications seen in months of follow-up. We recommend closer
KNH. follow up of patients with elevated CSF
proteins of reduced CSF glucose at
Nurhayat et al in a retrospective multicenter
ventriculoperitoneal shunt insertion so as to
study noted that a persistently elevated CSF
screen for features of shunt failure. Further
protein level above 100mg/dl was associated
studies on the chemical interaction between
with shunt infection (11). No case of shunt
shunt tubings and CSF proteins that leads to
infection was noted in this current study
shunt blockage are also recommended. The
among the patients with elevated CSF
proteins. Fulkerson et al, in his 10year findings of such a study can later inform the
materials used to manufacture shunts that
retrospective cohort study on the risk factors
will have less interactions with CSF proteins.
for shunt failure in low-birth-weight preterm
infants, observed a similar finding that there Limitations and delimitations
was no correlation between CSF protein
levels and the risk of shunt infection (1). This study provides the first ever description
of common biochemical parameters in
With regards to CSF cellularity and shunt children with hydrocephalus, not only in
blockage, this study finds no correlation Kenya but in the East African region. A longer
between elevated CSF cell count at shunt duration of follow-up would have been better
insertion and the likelihood of shunt failure at to screen for the effects of the CSF

25 The East African Journal of Neurological Sciences www.theeajns.org Volume 1 Issue 2 - 2022
MUTISO ET AL. CSF AND VENTRICULOPERITONEAL SHUNT SURVIVAL
biochemical and cellular characteristics over data about this important neurosurgical
many years of follow-up. A larger patient disorder.
group would have also yielded higher-quality
References
1. Fulkerson, D. H., Vachhrajani, S., Bohnstedt, B. N. 8. Mwang'ombe NJ, Omulo T. Ventriculoperitoneal
et al. Analysis of the risk of shunt failure or infection shunt surgery and shunt infections in children with
related to cerebrospinal fluid cell count, protein non-tumour hydrocephalus at the Kenyatta
level, and glucose levels in low-birth-weight National Hospital, Nairobi. East Afr Med J. 2000
premature infants with post-hemorrhagic Jul;77(7):386-90. doi: 10.4314/eamj.v77i7.46684.
hydrocephalus, Journal of Neurosurgery: PMID: 12862159.
Pediatrics PED 2011;7(2), 147-151. 9. Gichuhi K.M. Complications of ventriculo-
2. Joseph T. Cheatle, Alexis N. Bowder, Jonathan L. peritoneal shunts in management of congenital and
Tefft, et al Effect of Protein Concentration on the acquired hydrocephalus as seen at Kenyatta
Flow of Cerebrospinal Fluid Through Shunt Tubing, National Hospital. University of Nairobi e-
Neurosurgery, Volume 77, Issue 6, December repository, 1989.
2015, Pages 972–978. 10. Kitunguu P.K ventriculoperitoneal shunt survival at
3. Brydon HL, Keir G, Thompson EJ. Protein the Kenyatta national hospital; a registry based
adsorption to hydrocephalus shunt catheters: CSF assessment; 2017. University of Nairobi repository.
protein adsorption Journal of Neurology, 11. Nurhayat Yakut, Ahmet Soysal, Eda Kepenekli
Neurosurgery & Psychiatry1998;64:643-647. Kadayifci et al. Persistant High Cerebrospinal
4. Zwimpfer, T.J., Zwimpfer, M.J. Assessment of Fluid Protein May Be a Risk Factor for
ventriculoperitoneal (VP) shunt malfunction in an in Ventriculoperitoneal Shunt Reinfection in Children,
vitro model of artificial CSF flow: influence of CSF Open Forum Infectious Diseases, Volume 3, Issue
protein concentration, CSF contamination and time suppl_1, December 2016, 1195.
of shunt incubation. Fluids Barriers CNS 12, P60 12. Warf BC. Hydrocephalus in Uganda: the
(2015) doi:10.1186/2045-8118-12-S1-P60 predominance of infectious origin and primary
5. Kamat, A S. “CSF Protein Concentration management with endoscopic third
Associated with Ventriculoperitoneal Shunt ventriculostomy. J Neurosurg. 2005 Jan;102(1
Obstruction in Tuberculous Meningitis.” The Suppl):1-15. doi: 10.3171/ped.2005.102.1.0001.
International Journal of Tuberculosis and Lung PMID: 16206728.
Disease : the Official Journal of the International 13. Anderson, I. A., Saukila, L. F., Robins, J. M. W. et
Union against Tuberculosis and Lung Disease, al (2018). Factors associated with 30-day
U.S. National Library of Medicine, 2018. 31. 6-8. ventriculoperitoneal shunt failure in pediatric and
6. Kassebaum NJ, Arora M, Barber RM, Bhutta ZA, adult patients, Journal of Neurosurgery
Brown J et al Global, regional, and national JNS, 130(1), 145-153
disability-adjusted life-years (DALYs) for 315 14. Ambekar, Sudheer, Dwarakanath et al. Does CSF
diseases and injuries and healthy life expectancy composition predict shunt malfunction in
(HALE), 1990–2015: a systematic analysis for the tuberculous meningitis?. The Indian journal of
Global Burden of Disease Study tuberculosis, 2011. 58. 77-81.
2015. Lancet 388:1603–1658, 2016
7. Pan P. Outcome analysis of ventriculoperitoneal
shunt surgery in pediatric hydrocephalus. J Pediatr
Neurosci 2018;13:176-81

26 The East African Journal of Neurological Sciences www.theeajns.org Volume 1 Issue 2 - 2022