Prospects of CSF shunt independence among chronically shunted patients

doi:10.21203/rs.3.rs-4125850/v1

Background and objectives

CSF shunt placement for hydrocephalus and other etiologies has arguably been the most life-saving intervention in pediatric neurosurgery in the past 6 decades. Yet, chronic shunting remains a source of morbidity for patients of all ages. Neuroendoscopic surgery has made shunt independence possible for newly diagnosed hydrocephalic patients. In this study, we examine the prospects of shunt independence with or without endoscopic third ventriculostomy (ETV) in chronically shunted patients.

Methods

After IRB approval, a retrospective analysis was completed on patients whose shunt was ligated or removed to achieve shunt independence, with or without ETV. Clinical and imaging data were collected.

Results

Eighty-eight patients with CSF shunts had their shunt either ligated or removed, 57 of whom had a concomitant ETV. Original reasons for shunting included: congenital hydrocephalus 20 (23%), post-hemorrhagic hydrocephalus (PHH) of prematurity 14 (16%), aqueductal stenosis 10 (11%), intracranial cyst 8 (9%), tumor 8 (9%), infantile subdural hematomas 8 (9%), myelomeningocele 7 (8%), post-traumatic hydrocephalus 7 (8%) and post-infectious hydrocephalus 6 (7%). The decision to perform a simultaneous ETV was made based on etiology. Forty-nine (56%) patients became shunt independent. The success rate was 46% in the ETV group and 73% in the no ETV group. Using multivariate analysis and Cox Proportional Hazards models, age > 4 months at shunt placement (p = 0.032), no shunt revisions (p = 0.01), select etiologies (p = 0.043), and ETVSS > 70 (in the ETV group) (p = 0.017), were protective factors for shunt independence.

Conclusion

Considering the long-term complications of shunting, achieving shunt independence may provide hope for improved quality of life. While this study is underpowered, it provides pilot data identifying factors that predict shunt independence in chronically shunted patients, namely age, absence of prior shunt revision, etiology, and in the ETV group, the ETVSS.

CSF shunts have saved or improved the lives of countless patients with hydrocephalus[1, 2]. Yet, long-term complications of shunting remain a challenge to neurosurgeons, patients, and families[3–7], and shunt dependence is generally expected to be lifelong. The advent of endoscopic surgery, namely endoscopic third ventriculostomy (ETV), but also endoscopic cyst fenestration, tumor resection, more recently ETV with choroid plexus coagulation (CPC), and others, has made shunt avoidance an achievable, if not expected goal depending on etiology, age, and other factors[8]. Achieving shunt independence in those already shunted would eliminate the long-term complications of shunting and potentially improve quality of life. We conducted a retrospective analysis on a series of children and adults whose shunts were removed or ligated, with or without concomitant endoscopic surgery. Indications for surgery included symptoms of shunt malfunction in some, or patient/family interest in achieving shunt independence in others.

After approval by the Institutional Review Board, a retrospective analysis was completed on patients who underwent surgery to either remove or ligate their CSF shunt, with or without ETV. The decision to ligate the shunt or remove it depended on patient presentation and etiology of hydrocephalus, and in some cases on patient desire for shunt independence. The decision to perform a concomitant ETV was based on expected success rate, which was primarily related to etiology of hydrocephalus. Since most of these decisions preceded the publication of the ETV Success Score (ETVSS)[9], ETVSS played no role in surgical decision in most patients. Instead, ETVSS was computed retrospectively. Data collected included reason for and age at shunt removal/ligation surgery, type of surgery (ligation or removal of shunt, with or without ETV), reason for and age at original shunt placement, type of shunt, number of shunt revisions and infections, imaging findings, intraoperative findings, results of intervention, and follow-up period. Patients with childhood hydrocephalus and infantile subdural hematoma/hygroma were included. Patients with normal pressure hydrocephalus, pseudotumor cerebri and cyst-peritoneal shunts were excluded because of low sample size (2 or less).

Statistical Analysis

Time-to-event analysis was used to model the success of shunt independence. An event was defined as surgery to re-establish shunt function after failure to achieve shunt independence. The follow-up interval was defined as the length of time from surgery for removal/ligation of the shunt to 1) last clinic visit in those who maintained shunt independence, or 2) surgery to re-establish shunt function in those who failed to maintain shunt independence. Median time-to-event was defined as the period of time after shunt ligation/removal when half the population had their shunt reactivated or reimplanted, and the other half remained shunt independent.

A Kaplan-Meier survival curve for the entire cohort estimated the fraction of shunt-independent patients among those at risk for shunt reactivation/reimplantation at each time point following attempted removal/ligation. For the categorical variable etiology, we compared survival curves between etiologies and used a log rank test to measure overall differences in survival between the categories[10].

A Cox proportional hazards (CPH) model was used to evaluate continuous predictive variables for successful shunt-independence. These included: age at shunt placement, age at shunt removal/ligation, shunt duration, number of shunt revisions, number of shunt infections, and the ETVSS. Models using ETVSS were used only for the subset of patients who received an ETV.

A multivariate hazards model was used on continuous variables that appeared predictive in the univariate models. For these variables, we estimated the optimal cutpoint at which the variable would be most predictive of a successful outcome. To do this, we used the maxstat package[11] in R, which selects the maximum log-rank statistic for all cuts of the data, and accounts for the multiple testing done when selecting the maximum. All statistical analyses were done in R[12].

Patient demographics

Eighty-eight patients had their shunts removed or ligated, fifty-seven of which (65%) underwent simultaneous ETV. There are 49 males with an average age of 20 years at shunt removal/ligation, and an average duration of shunting prior to removal/ligation of 11.6 years. The reasons for original shunting include: congenital hydrocephalus in 20 (23%), post-hemorrhagic hydrocephalus (PHH) of prematurity in 14 (16%), intracranial cyst in 8 (9%), aqueductal stenosis in 10 (11%), brain tumor in 8 (9%), infantile subdural hematomas in 8 (9%), myelomeningocele in 7 (8%), post-traumatic hydrocephalus in 7 (8%) and post-infectious hydrocephalus in 6 (7%). 44 (50%) patients had no history of shunt revision surgery (Table 1). Clinical and imaging presentations that led to a decision to attempt shunt independence include symptomatic shunt overdrainage in 24 (27%); proximal obstruction in 16 (18%); shunt infection in 14 (16%%); distal obstruction, disconnection, or painful shunt track in 16 (18%); aqueductal stenosis in 10 (11%); and clinical suspicion that the shunt is no longer required, namely subdural shunts in 8 (9%). Indications for ETV was based on etiology of hydrocephalus. We did not think ETV was clinically indicated in patients who had successful surgical treatment of colloid cyst (2), complex intraventricular cyst (3), choroid plexus papilloma (2), posterior fossa cyst (2), or Chiari I malformation (2); and in one case, a patient whose shunt was placed late in life for questionable high ICP. Retrospectively, the median ETVSS in the ETV group was 70 (average 72). Most of the predictor variables of interest were continuous measures: age, shunt duration, number of revisions or infection, and ETVSS. Cox proportional hazards estimates for univariate continuous predictors are shown in Table 2. Indications for surgery to re-activate or re-implant a shunt included symptoms of elevated ICP, CSF leak, high ICPs measured through an external ventricular drain (EVD), shunt tap, and ICP monitoring.

Table 1

**Patient demographics.** AS: aqueductal stenosis; ETV: Endoscopic Third Ventriculostomy; ETVSS: Endoscopic Third Ventriculostomy Success Score; HC: hydrocephalus; MM: Myelomeningocele; PHH: Post-Hemorrhagic Hydrocephalus.
Demographics	No ETV = 32	ETV = 56	Combined = 88
Demographics	n (%)	n (%)	n (%)
Age at shunt placement (years)
0–1	18 (56.2%)	31 (55.4%)	49 (55.7%)
1 + − 10	7 (21.9%)	8 (14.3%)	15 17%)
>10	7 (21.9%)	17 (30.4%)	24 (27.3%)
Age at shunt ligation or removal (years)
0–1	1 (3.1%)	3 (5.4%)	4 (4.5%)
1–10	14 (43.8%)	9 (16.1%)	23 26.1%)
>10	17 (53.1%)	44 (78.6%)	61 (69.3%)
Time to shunt ligation or removal (years)
0–5	14 (43.8%)	18 (32.1%)	32 (36.4%)
5 + − 10	8 (25%)	6 (10.7%)	14 (15.9%)
10 + − 20	6 (18.8%)	19 (33.9%)	25 (28.4%)
> 20	4 (12.5%)	13 (23.2%)	17 (19.3%)
Number of shunt revisions
0	19 (59.4%)	26 (46.4%)	45 (51.1%)
1	7 (21.9%)	7 (12.5%)	14 (15.9%)
2	2 (6.2%)	9 (16.1%)	11 (12.5%)
3–7	4 (12.5%)	14 (25%)	18 (20.5%)
Number of shunt infections
0	28 (87.5%)	45 (80.4%)	73 (83%)
1	4 (12.5%)	9 (16.1%)	13 (14.8%)
2	0 (0%)	2 (3.6%)	2 (2.3%)
Etiology
Congenital HC	7 (21.9%)	13 (23.2%)	20 (22.7%)
Post-traumatic	4 (12.5%)	3 (5.4%)	7 (8%)
AS	0 (0%)	10 (17.9%)	10 (11.4%)
Cyst	4 (12.5%)	4 (7.1%)	8 (9.1%)
Subdural	8 (25%)	0 (0%)	8 (9.1%)
Post-infectious	3 (9.4%)	3 (5.4%)	6 (6.8%)
Tumor	2 (6.2%)	6 (10.7%)	8 (9.1%)
PHH	3 (9.4%)	11 (19.6%)	14 (15.9%)
MM	1 (3.1%)	6 (10.7%)	7 (8%)
ETVSS
30–60	0 (0%)	7 (12.5%)	7 (8%)
70	0 (0%)	20 (35.7%)	20 (22.7%)
80	0 (0%)	29 (51.8%)	29 (33%)
NA	32 (100%)	0 (0%)	32 (36.4%)

Table 2

**Cox Proportional Hazard** **univariate** **predictor model.** Factors that predict shunt independence include (1) age at shunt placement (p = 0.032), (2) number of shunt revision (p = 0.01), and (3) ETVSS (p = 0.017). ETV: Endoscopic Third Ventriculostomy; ETVSS: Endoscopic Third Ventriculostomy Success Score; HR: hazards ratio; LCL: Lower Confidence Limit; UCL: Upper Confidence Limit; Pr: Probability
	HR	LCL	UCL	Pr(>\|z\|)
Age at shunt placement (years)	0.969	0.941	1	0.032
Age at shunt ligation/removal (years)	0.991	0.971	1.01	0.37
Time to shunt ligation/removal	1.023	0.997	1.05	0.086
Number of shunt revisions	1.21	1.05	1.41	0.01
Number of shunt infections	1.360	0.763	2.42	0.3
ETVSS of ETV patients, n = 58	0.973	0.951	1	0.017

Success of shunt removal/ligation

Median time-to-event, i.e., when half of patients failed shunt ligation/removal and had their shunts activated/reimplanted, is 4 years (Fig. 1, Table 3). The 6-month and 4-year shunt-independent survival periods were estimated from the Kaplan-Meier curve. 56.8% and 50.1% of patients were shunt independent 6 months and 4 years after shunt removal/ligation, respectively (Table 4). The median shunt-independent duration, i.e., period of time when half of patients at risk had their shunt reactivated/reimplanted, is 4 years (95% CI= [11, NA]). Shunt-independent survival in patients whose shunt was placed before 4 months of age was 66.7% and 64.3% at 6 months and 4 years respectively, and in those whose shunt was placed at or after 4 months of age, it was 38.7% and 25.4% respectively. Shunt-independent survival in patients with zero shunt revisions was 64.4% at 6 months as well as 4 years, and in those with 1 or more shunt revisions, it was 48.8% and 34.7%, respectively (Table 3). Six-month shunt independence was achieved in 47% of the ETV group, and 73% of the No ETV group (Fig. 2). In patients with ETVSS < 70, the shunt-independent survival at 6 months and 4 years was 28.6% and 14.3%, respectively. In patients whose ETVSS was 70, the shunt-independent survival at 6 months and 4 years was 45% and 26.2% respectively. And in patients with an ETVSS 80, the shunt-independent survival at 6 months and 4 years was 58.6% and 54.7%, respectively (Table 3).

Table 3. 6-month and 4-year Percent Survival, [95% CI], Categorized Predictors. Following shunt removal/ligation, 56.8% and 50.1% of patients remained shunt independent at 6 months and 4 years, respectively. Median shunt independence (the time period in which half the patients at risk had a shunt reactivated/reimplanted) was approximately 4 years (95% CI= [11, NA]). Predictors of shunt independence are listed by category. AS: Aqueductal stenosis; CI: Confidence Interval; HC: hydrocephalus; MM: Myelomeningocele; PHH: Post-Hemorrhagic Hydrocephalus.; Number of failures: number of patients who failed to achieve shunt independence throughout the duration of the study.

		n	Number of failures (n)	6-month survival [95% CI]	4-year survival [95% CI]	Median survival (years) [95% CI]	Logrank p
Etiology	All	88	47	56.8 [47.4,68.2]	50.1 [40.5,62]	4.06 [0.11,NA]	0.0043
	Congenital HC	20	12	55 [37,81.8]	48.9 [30.9,77.4]	1.86 [0.06,NA]
	Post-traumatic	7	4	42.9 [18.2,100]	42.9 [18.2,100]	0.22 [0.08,NA]
	AS	10	3	70 [46.7,100]	70 [46.7,100]	NA [0.11,NA]
	Cyst	8	4	62.5 [36.5,100]	50 [25,100]	0.63 [0.03,NA]
	Subdural	8	0	100 [100,100]	100 [100,100]	NA [NA,NA]
	Post-infectious	6	3	50 [22.5,100]	50 [22.5,100]	0.04 [0.02,NA]
	Tumor	8	3	87.5 [67.3,100]	70 [42,100]	9.35 [3.03,NA]
	PHH	14	12	28.6 [12.5,65.4]	21.4 [7.9,58.4]	0.06 [0.02,NA]
	MM	7	6	28.6 [8.9,92.2]	14.3 [2.3,87.7]	0.06 [0.01,NA]
Age at Shunt Placement (months)	0-4	31	24	38.7 [24.9,60.3]	25.4 [13.8,46.7]	0.07 [0.03,2.98]	0.0003
Age at Shunt Placement (months)	>4	57	23	66.7 [55.5,80.1]	64.3 [52.8,78.3]	NA [8.9,NA]	0.0003
Number of Shunt Revisions	0	45	18	64.4 [51.9,80.1]	64.4 [51.9,80.1]	NA [8.9,NA]	0.024
Number of Shunt Revisions	1-7	43	29	48.8 [36,66.3]	34.7 [22.5,53.5]	0.3 [0.07,NA]	0.024
ETVSS of ETV Patients, n=56	30-60	7	6	28.6 [8.9,92.2]	14.3 [2.3,87.7]	0.04 [0.02,NA]	0.015
	70	20	16	45 [27.7,73.1]	26.2 [11.9,57.8]	0.06 [0.02,NA]
	80	29	14	58.6 [43.2,79.6]	54.7 [39.2,76.4]	12.22 [0.11,NA]

Using Cox proportional hazards estimate, we identified factors that correlate with shunt independence after removal/ligation. These include age at shunt placement, number of shunt revisions, etiology, and ETVSS. Factors that did not independently correlate with success of shunt removal/ligation are duration of shunting prior to removal/ligation, number of shunt infections, and age at shunt removal/ligation (Table 2). A cox proportional hazards multiple regression analysis was completed on the continuous predictive variables identified in the univariate model. The number of shunt revisions is protective of shunt independence (P = 0.039) (Table 4).

Table 4

**Cox Proportional Hazard *multiple* regression model.** This analysis was completed on the continuous predictive variables identified in the univariate model in Table 2. HR: hazards ratio; LCL: Lower Confidence Limit; UCL: Upper Confidence Limit; Pr: Probability
	HR	LCL	UCL	Pr(>\|z\|)
Age at shunt placement (years)	0.974	0.946	1.00	0.073
Number of shunt revisions	1.172	1.008	1.36	0.039

Table 5. Late ETV failures. Eight patients presented with late failure (beyond 6 months) after shunt ligation or removal, with an average of 6.8 years. Five of these were from the ETV group. Cyst-P: Cyst-Peritoneal shunt; ETV: ETV: Endoscopic Third Ventriculostomy; ETVSS: Endoscopic Third Ventriculostomy Success Score; ICP: intracranial pressure; VP: Ventricular-Peritoneal shunt; IVH: Intra-ventricular hemorrhage.

ID	Etiology	Shunt type	Age at shunt Independence (years)	Indication for shunt ligation/ removal	Surgery	Time to shunt reactivation/reimplantation (years)	Indication for shunt reactivation or reimplantation
1	Cyst	Cyst-P	48.8	Overdrainage	Shunt ligation	6.9	Headache recurrence requiring repetitive high-volume reservoir taps
2	Congenital	VP	26	Disconnection	Shunt removal	8.8	Headaches and progressive deterioration in mental status
3	Congenital	VP	17	Proximal Obstruction	ETV without shunt ligation/removal, as shunt was considered obstructed (dry shunt tap)	1.8	Shunt tap showing normal pressure and flow in the shunt.
4	Tumor	VP	14	Disconnection	ETV + shunt removal	9.3	Headaches requiring repetitive high-volume lumbar punctures
5	Tumor	VP	30.4	Proximal Obstruction	ETV + shunt removal	3	Headaches
6	IVH	VP	13.9	Proximal Obstruction	ETV + shunt removal	9	Headaches, memory issues requiring high-volume lumbar punctures
7	Congenital	VPleural	25.5	Shunt infection	Shunt removal	12.1	Headaches and presumed high ICP (shunted at a different institution)
	IVH	VP	20.5	Shunt infection	ETV + shunt removal	4	Progressive headaches and ventricular enlargement

Age at shunt placement

The average ages at initial shunt placement are 10.2 years (median 2) and 5.6 years (median 0.25 years) in the groups that succeeded or failed to achieve shunt independence, respectively (p = 0.032). Maximal Log Rank estimated an optimal cut-point of 0.34 years for age at shunt placement (p = 0.0086), below which the success rate is very low (18.75%) (Fig. 3, Table 2).

Number of shunt revisions

While less shunt revisions correlates with success of shunt removal/ligation (HR = 1.2, p = 0.01), a history of zero shunt revisions is a statistically significant predictor of success (p = 0.048) (Fig. 4, Table 2).

ETVSS:

The average ETVSS is 72 (median 70) and 69 (median 70) in the groups that succeeded or failed to achieve shunt independence, respectively (Fig. 5). Higher ETVSS scores (cut-point of 70) are protective (HR = 0.98, p = 0.0103, Table 2). ETVSS was calculated retroactively in 47% of patients who underwent surgery before the ETVSS was first described in the literature.[8] ETV success is similar before (55%) and after (52%) ETVSS publication.

Original etiology/reason for shunt placement

A subdural shunt placed for chronic subdural hematoma in infancy was successfully removed/ligated in all 8 patients (100%). The etiology for shunt placement in chronic subdural collections was post traumatic (6), post-meningitis (1) and secondary to leukemia (1). Success of VP shunt removal/ligation at 6 months and 4 years, respectively, correlates with etiology of hydrocephalus, as follows: tumor (88% and 70%), aqueductal stenosis (70% and 70%), cyst (63% and 50%), infection (50% and 50%), trauma (43% and 43%), and PHH (29% and 21%) myelomeningocele (29% and 14%) (Table 3). Survival curves differ by etiology (p = 0.004 and Fig. 6).

Indication for shunt re-implantation

Failure of shunt ligation/removal was recognized via elevated ICP captured in 26 patients (67%) via a postoperative external ventricular drain EVD in 18, Codman parenchymal ICP monitoring in 2, shunt tap in 5 and lumbar puncture in 1; 4 failures (10%) were diagnosed after an incisional CSF leak, and the remaining 9 (23%) developed clinical symptoms such as headaches, balance problems, macrocephaly, papilledema, and behavioral problems.

Late failures

The average time to shunt reactivation/reimplantation was 18.4 days (0 days to 3 months). Twenty-five patients (64%) had their shunt reactivated/reimplanted in the same hospital admission, while the remaining patients were readmitted for surgery. Eight patients (9%) presented with late failures (> 6 months), with an average time to failure of 6.8 years (range 1.8 years to 12 years) in that group. Five of the 8 were in the ETV group (Table 5).

Complications

Operative complications occurred in 4 patients and include 2 intraventricular hemorrhages requiring temporary external drainage and 2 superficial surgical site infections treated with oral antibiotics. CSF leaks (reported above) occurred in the setting of failure of shunt ligation/removal.

Until recently, the consensus among pediatric neurosurgeons was that most shunted children are fated for a lifetime of shunt dependence[13], following the adage “once a shunt, always a shunt” [14]. Neuroendoscopy has proven the contrary, revealing that patients with hydrocephalus can indeed be shunt independent if the causative obstruction is removed or bypassed. Still, many neurosurgeons who have grown weary of repetitive shunt surgery shy away from attempting shunt removal for fear of complications leading to additional surgery, following the principle of “if it ain’t broke, don’t fix it.” But one may consider the alternative logic, that the best way to minimize long-term shunt complications is to no longer need the shunt.

The introduction of ETV motivated more surgeons to attempt shunt removal[15–17]. In the 1990s, ETV for shunt removal was used for the treatment of slit ventricle syndrome (SVS)[18–20] and shunt infections[21–23]. A recent study by the Hydrocephalus Clinical Research Network (HCRN) showed a post-shunt ETV with or without CPC success rate of 41%[16]. The study did not clarify the surgeons’ indications for ETV vs. ETV/CPC. Another multicenter, retrospective study reviewed 80 patients who had ETV when patients presented with shunt malfunction or infection, with a success rate of 64%[24]. In addition to a high complication rate of 34%, a significant limitation of the paper is that most shunts were not removed or ligated after ETV, which makes it impossible to determine whether the patients indeed achieved shunt independence.

Factors associated with shunt independence

In the present study, approximately 56% of patients were rendered shunt-independent at 6 months, and the majority (50% of the entire cohort) maintained shunt independence at 4 years. The factors that increase the likelihood of shunt independence are 1) shunt placement age > 4 months, 2) a history of no shunt revisions, 3) in those who had ETV, an ETVSS of 70 or above; and 4) etiology of the hydrocephalus. Specifically, the likelihood of shunt independence is highest in patients with hydrocephalus related to aqueductal stenosis (70%) and tumor (70%) and lowest in patients with history of myelomeningocele repair (14.4%) and prematurity-related PHH (21.4%) (table 4). Notably, 2 of our PHH patients remained shunt independent after ETV long-term (9 and 7 years), and 2 others required shunt reactivation several years later, implying that the reason for ETV failure is more likely closure of the third ventriculostomy site rather than lack of CSF absorption. A special note should be made about patients treated in infancy with subdural shunts, as they all responded to shunt removal or ligation. Similarly, Erwood, et al. reported a success rate of 96% after removal of subdural shunts[25]. While a valid reason to remove subdural shunts is to spare a growing child from the complications of subcutaneous calcification and tissue scarring along shunt tracts, some argue against removal of subdural shunts in asymptomatic patients, as there is no published evidence of long-term neurological complications from chronic subdural shunting[26–29].

Surgical complications and study limitations

Our data indicate that surgery to ligate/remove a CSF shunt with or without ETV is safe. Moreover, we argue that, as the primary care-givers of this patient population with proven significant long-term shunt morbidity and mortality, pediatric neurosurgeons bear the responsibility to not only ensure adequate shunt function, but to also inform patients of all options at their disposal, including ETV and shunt removal, along with each procedure’s expected success and complication rates. Inevitably, the data obtained from this and the other published studies are underpowered by virtue of their retrospective nature and small sample size, which limits the quality of evidence needed to inform patients on risk and benefit. Thus, pending large registries and prospective clinical trials, surgeons should supplement the ‘weaker’ evidence available to them with fastidious informed consent and careful clinical judgment.

Improving patient quality of life is the predominant goal of caring for patients with CSF shunts. The optimal approach to avoid shunt complications is achieving shunt independence. This study shows that ligating/removing shunts (with ETV in select cases) carries a reasonable success rate in line with the success rates reported in ETV+/-CPC studies. In addition, similar to ETV studies, factors that correlate with surgical success include age, etiology, and number of shunt revisions. In particular, patients with shunts placed for chronic subdural hematoma, aqueductal stenosis and tumor have a high likelihood of achieving shunt independence, whereas those related to PHH of prematurity seldom succeed. ETVSS is a good predictor of shunt independence in those who undergo simultaneous ETV, but our results also show that many patients can achieve shunt independence without the additional potential morbidity of ETV. As neurosurgeons, we carry the responsibility of guiding patients and families in caring for their shunt. We join others in proposing that this should include helping them understand the difference in risk profiles between chronic shunting and additional surgery for shunt removal. While our underpowered data do not provide definitive recommendations for care, they add to other studies in guiding important family discussions.

Funding:

Theodore W. Batterman Family Foundation.

Author Contribution

J.K.: acquisition, analysis and interpretation of dataA.B.: statistical analysis interpretation of dataJ.K. and B.I.: wrote the main manuscript

Acknowledgements:

none

Beuriat PA et al (2017) Hydrocephalus treatment in children: long-term outcome in 975 consecutive patients. J Neurosurg Pediatr 20(1):10–18
Drake JM, Kulkarni AV, Kestle J (2009) Endoscopic third ventriculostomy versus ventriculoperitoneal shunt in pediatric patients: a decision analysis. Childs Nerv Syst 25(4):467–472
Browd SR et al (2006) Failure of cerebrospinal fluid shunts: part II: overdrainage, loculation, and abdominal complications. Pediatr Neurol 34(3):171–176
Browd SR et al (2006) Failure of cerebrospinal fluid shunts: part I: Obstruction and mechanical failure. Pediatr Neurol 34(2):83–92
Drake JM et al (1998) Randomized trial of cerebrospinal fluid shunt valve design in pediatric hydrocephalus. Neurosurgery 43(2):294–303 discussion 303-5
Kraemer MR et al (2018) Overdrainage-related ependymal bands: a postulated cause of proximal shunt obstruction. J Neurosurg Pediatr 22(5):567–577
Sainte-Rose C et al (1991) Mechanical complications in shunts. Pediatr Neurosurg 17(1):2–9
Kulkarni AV et al (2009) Endoscopic third ventriculostomy in the treatment of childhood hydrocephalus. J Pediatr 155(2):254–9e1
Kulkarni AV et al (2010) Predicting who will benefit from endoscopic third ventriculostomy compared with shunt insertion in childhood hydrocephalus using the ETV Success Score. J Neurosurg Pediatr 6(4):310–315
Harrington P, David FRT (1982) A Class of Rank Test Procedures for Censored Survival Data. Biometrika 69(3):553–566
Hothorn TaLB (2003) On the exact distribution of maximally selected rank statistics. Comput Stat Data Anal 43(2):121–137
Team RC (2020) A language and environment for statistical computing. R Foundation for Statistical Computing ; Available from: https://www.R-project.org
Epstein F (1985) How to keep shunts functioning, or the impossible dream. Clin Neurosurg 32:608–631
Hemmer R, Bohm B Once a shunt, always a shunt? Dev Med Child Neurol Suppl, 1976(37): p. 69–73
Maldonado IL, Valery CA, Boch AL (2010) Shunt dependence: myths and facts. Acta Neurochir (Wien) 152(9):1449–1454
Rocque BG et al (2022) Endoscopic third ventriculostomy in previously shunt-treated patients. J Neurosurg Pediatr, : p. 1–9
Guida L et al (2023) Predicting endoscopic third ventriculostomy success in pediatric shunt dysfunction: a monocentric retrospective case series of 70 consecutive children, systematic review, and meta-analysis. J Neurosurg Pediatr, : p. 1–11
Baskin JJ, Manwaring KH, Rekate HL (1998) Ventricular shunt removal: the ultimate treatment of the slit ventricle syndrome. J Neurosurg 88(3):478–484
Goumnerova LC, Frim DM (1997) Treatment of hydrocephalus with third ventriculocisternostomy: outcome and CSF flow patterns. Pediatr Neurosurg 27(3):149–152
Reddy K et al (1988) Slit ventricle syndrome with aqueduct stenosis: third ventriculostomy as definitive treatment. Neurosurgery 23(6):756–759
Cinalli G et al (1998) The role of endoscopic third ventriculostomy in the management of shunt malfunction. Neurosurgery, 43(6): p. 1323-7; discussion 1327-9
Jones RF et al (1993) Third ventriculostomy for shunt infections in children. Neurosurgery, 32(5): p. 855-9; discussion 860
Kelly PJ (1991) Stereotactic third ventriculostomy in patients with nontumoral adolescent/adult onset aqueductal stenosis and symptomatic hydrocephalus. J Neurosurg 75(6):865–873
Hersh DS et al (2020) Converting Pediatric Patients and Young Adults From a Shunt to a Third Ventriculostomy: A Multicenter Evaluation. Neurosurgery 87(2):285–293
Erwood A et al (2022) Case Series on Removal of Subdural to Peritoneal Shunt After Resolution of Subdural Collection in the First 2 Years of Life. Oper Neurosurg (Hagerstown) 23(1):8–13
Kombogiorgas D, Sgouros S (2005) Removal of subdural-peritoneal shunts in infants. Childs Nerv Syst 21(6):458–460
Litofsky NS, Raffel C, McComb JG (1992) Management of symptomatic chronic extra-axial fluid collections in pediatric patients. Neurosurgery 31(3):445–450
Morota N et al (1995) Infantile subdural fluid collection: diagnosis and postoperative course. Childs Nerv Syst 11(8):459–466
Vinchon M et al (2001) Subduroperitoneal drainage for subdural hematomas in infants: results in 244 cases. J Neurosurg 95(2):249–255

No competing interests reported.

Prospects of CSF shunt independence among chronically shunted patients

Status:

Journal Publication

Version 1

Abstract

Background and objectives

Methods

Results

Conclusion

Figures

Introduction

Methods

Statistical Analysis

Results

Patient demographics

Success of shunt removal/ligation

Age at shunt placement

Number of shunt revisions

ETVSS:

Original etiology/reason for shunt placement

Indication for shunt re-implantation

Late failures

Complications

Discussion

Factors associated with shunt independence

Surgical complications and study limitations

Conclusion

Declarations

Funding:

Author Contribution

Acknowledgements:

References

Additional Declarations

Status:

Journal Publication

Version 1