Subtemporal Decompression in Resistant Slit Ventricle Syndrome in Children: An Observational Study and Survival Analysis

Introduction: Slit ventricle syndrome (SVS) remains a challenging problem in the early-shunted paediatric population. Various surgical and non-surgical treatments have been devised for this condition. However, there is currently no gold standard for its optimal management. Among various treatment modalities, subtemporal decompression (STD) is often performed as a last resort. We present our experience of STD in paediatric patients with SVS in whom initial treatment with programmable valves and anti-syphon device were not successful. Methods: This is a single-centre retrospective observational study and survival analysis. Patients who underwent STD for SVS were included. Pre- and post-operative imaging data and clinical outcomes were collected. Results: There were 20 patients (12 M, 8 F) with a mean age of 9 years (SD: 4) at first STD. 90% (n = 18) of patients had multiple shunt revisions pre-STD. At first STD, 70% (n = 14) and 30% (n = 6) of patients had unilateral or bilateral STD, respectively. STD led to a reduction in the frequency of shunt revisions in 60% (n = 12) of patients. The median time required before further STD, shunt surgery, or cranial vault surgery was 14 months. The median time before a further STD was required (either revision or contralateral side) was 89 months. At a median follow-up of 66.5 months (range: 1–159), 65% (n = 13) of patients had improvement in symptoms. Conclusions: A large proportion of patients with persistent SVS symptoms, refractory to multiple shunt revisions, benefitted from STD in combination with shunt optimization. It was also safe and well-tolerated. Therefore, in patients who have multiple failed shunts, STD may reduce the morbidity associated with further shunt revisions and can significantly improve symptomatology.


Introduction
Slit Ventricle Syndrome Slit ventricle syndrome (SVS) is a condition in shunted patients with hydrocephalus (HCP) characterized by small ventricular volumes on radiological findings, associated with intermittent headaches, vomiting, nausea, and drowsiness [1].Various causes have been outlined relating to disturbances in CSF dynamics, including noncompliant ventricles and intermittent proximal catheter obstructions [2,3].These patients display a unique slitlike appearance of the ventricles, and often have a slow refill of the shunt valve mechanism.What is common between them, however, is that they often present with a complex history over several years burdened by numerous shunt revisions and blockages.Typically, these patients are shunted in their early years of life, following which they remain symptom-free for a period [2][3][4][5].Eventually, the symptoms and slit-like ventricles warrant shunt revisions as well as implementation of anti-syphon or programmable valves.

Subtemporal Decompression
In a proportion of patients, shunt revisions and adjustments do not relieve the patient's symptoms of SVS.At this point, subtemporal decompressions (STDs) may be considered as an alternative treatment option.The multitude of treatment alternatives presented in the literature include shunt revisions, programmable valve and anti-syphon device implementation, endoscopic third ventriculostomy and shunt removal, STD, and cranial vault expansion.Given the paucity of evidence reporting outcomes after these interventions, the choice of treatment remains predominately based on local preferences and anecdotal experience.

Aim
In this study, we report our single-centre experience of STD in patients with resistant symptoms of SVS after multiple shunt revisions, with the aim of determining safety, clinical and radiology outcomes, and the likelihood for further surgeries.

Study Design and Participants
This was a retrospective, observational survival analysis.We reviewed patients who had STDs for SVS between November 1995 and December 2020 at our institution.All data were retrieved from a prospective hospital operative database.All patients included a diagnosis of SVS, with slit-like ventricles confirmed on CT imaging, with corresponding headache symptoms.Patients without follow-up after surgery were excluded from the study.This service evaluation project is registered with the Great Ormond Street Hospital Quality Committee (Registration number 3467).

Variable Baseline Data
For identified patients, demographic baseline data were collected from the hospital electronic records, including age, sex, HCP aetiology prior to the initial shunt, and number of pre-STD shunt revisions.

Radiological Outcomes
In addition, assessing clinical outcomes of SVS, we also assessed changes in the temporal horns and peri-cerebral spaces.We defined the changes on imaging as follows: temporal horn change: (1) no improvement; no change from preoperative scan.Less than <1 mm (collapsed) or (2) satisfactory: slit-like but patent and >1 mm.Sylvian fissure change (1) no improvement: no change from preoperative scan and width (transaxial imaging) narrow <2 mm or collapsed or (2) satisfactory: >2 mm in diameter on coronal slices.Cortical sulci (1) no improvement with a tight convexity or (2) satisfactory: with improved sulci definition.Those with imaging unavailable were defined as "unavailable."

Surgical Technique
Under general anaesthesia, the patient is positioned in a supine position with a shoulder roll.Following a full aseptic preparation, a temporalis linear incision is made to allow temporal access.The temporalis muscle is then split with a temporal incision towards the zygomatic arch.A 4 × 4 cm temporal craniectomy is then performed to the level of the zygoma.Dura is opened in a cruciate fashion and covered with a large sheet of haemostatic oxidised cellulose.Layered closure is performed afterwards.

Statistical Analysis
Baseline demographic and descriptive statistics were performed.The unilateral and bilateral STD post-operative outcomes (improved, stable, or deteriorated) were compared using a χ 2 test.Number of subsequent surgeries required before versus after STD surgery was compared with a paired Student's t test (normality of data was determined using a QQ plot confirming "number of subsequent surgeries" data to be parametric).Complications were reported descriptively.For survival analysis, the defined endpoint criterion was the point at which a further surgery was required, including shunt surgery, STD revision, or a cranial vault operation.The time at which 50% of patients required a form of further surgery was assessed using Kaplan-Meier curves.Comparison between survival between primary unilateral and bilateral STDs was performed with a Logrank (Cox-Mantel) test.All statistical tests were performed on GraphPad Prism 8.4.2c.A p value of less than 0.05 was considered significant.

Baseline Demographics
A total of 20 patients with SVS after shunting had undergone an STD (12 M, 8 F), with a mean age at primary STD being 9 years (range: 2-15 years) and a mean follow-up of 73.6 months (range: 1-159 months).All 20 patients had undergone VP shunt insertion within the first year of life, and 18 patients had these shunts within the first 8 weeks of life.The aetiology for the initially treated HCP is shown in Table 1.A total of 11 (55%) had been shunted for post-haemorrhagic HCP, 6 (30%) for congenital, 2 (10%) for IIH, and 1 (5%) for HCP secondary to a tumour.

Symptom Outcomes
Overall, at last clinical follow-up, 13 (65%) patients improved (of which 6 had total symptom resolution), 5 (25%) remained stable, and 2 (10%) deteriorated symptomatically (Table 1).Interestingly, the figures suggest that those with an initial congenital HCP were more likely to benefit from a rescue STD, when compared to the other aetiologies; however, due to low numbers, this was not statistically significant.
There was a significant difference in the symptom outcome between patients with unilateral or bilateral STD (p = 0.014) (Fig. 1).Notably, the unilateral STD had better outcomes, with 11 (100%) patients improving and none deteriorating (Fig. 1).

Radiological Outcomes
Regarding radiological outcomes at the last clinical follow-up, satisfactory improvements across all three parameters were found in 50-55% (10-11/20) of patients (Table 1; Fig. 3). Figure 3a shows an example CT scan of bilateral STD at 4 years post-surgery, with ventricles that were no longer slit-like in appearance (Fig. 3b), and an element of bony regrowth on both sides (Fig. 3c, d).Radiological outcomes postintervention were unavailable for 3 patients.
Further Surgery Post STD Overall Following initial STD, 35% (n = 7) of patients required shunt revisions alongside repeat STD, 35% (n = 7) of patients required multiple shunt revisions alone, and 30% (n = 6) of patients required no further interventions.No patients required repeat STD alone.

Shunt Revisions
Prior to the primary STD, 18 of the 20 patients required multiple shunt revisions, with a total of 149 shunt revisions being performed.Of the shunts revised, the reason for shunt failure included catheter blockages in 7 patients and suboptimal symptom control in 9. Post STD, 14 patients required shunt revisions, with an overall 80 shunt revision procedures.Of the shunts revised, the reason for shunt failure included catheter blockages in 7 patients and suboptimal symptom control in another 7. STD surgery significantly reduced the mean number of shunt revisions required from 7.45 to 4.0 procedures (p = 0.02) (Fig. 4).

STD Revisions
A total of 7 (35%) patients required a further STD, with a total of 15 subsequent STD procedures.Of those procedures, 2 (13.3%) were to create a contralateral STD and 13 (86.7%)were due to bony overgrowth of the original craniectomy.Cranial Vault Expansions A total of 4 (20%) patients underwent cranial vault expansion subsequently.Four of these patients had HCP of post-haemorrhagic aetiology, and 1 patient had HCP of congenital aetiology.One patient presented with severe headaches and nausea at 4 months of clinical follow-up, however, these symptoms resolved following a vault expansion.

Complications
There was one acute post-operative complication following STD, which was extra-axial CSF collection.This was aspirated and a head-bandage was applied, following which the patient's symptoms fully resolved.Chronic complications included bone regrowth (n = 7, including 1 patient where the site became fully ossified) and overdrainage symptoms with a sunken wound site (n = 1).

Survival Analysis
The survival analysis is summarized in Figure 5.The median time before further STD, shunt surgery, or cranial vault surgery required was 14 months (Fig. 5a).The median time before a further STD was required (either a revision or surgery to the contralateral side) was 89 months (Fig. 5b).There was no significant difference in the time to revisions between the unilateral STD (median survival of 17.5 months) compared to bilateral decompressions (median survival of 14.0 months) (p = 0.65) (Fig. 5c).

Key Findings
Resistant symptoms of SVS remain a challenging problem in the early-shunted paediatric population.Various surgical and non-surgical treatments have been devised for this condition.Among various treatment modalities, STD is often performed as a last resort.We reviewed our experience of STD in paediatric patients with resistant symptoms of SVS in whom treatment with programmable valves and anti-syphon devices had not been successful.Our six main findings are highlighted below.1.All the patients had shunts inserted within the first year of life, and majority had undergone multiple revisions of these shunts.2.Over half of all patients who had undergone STD had symptomatic improvement, with over a quarter showing symptom resolution.Interestingly, those with unilateral decompressions (ipsilateral to the shunt) had a significantly greater improvement compared to patients with a bilateral decompression.3. We found that around 30% of patients who underwent primary STDs did not require further surgical intervention and had symptom improvement.Therefore, in this subset of patients, it was an effective first-line treatment.However, for the majority of patients (70%), a further procedure is likely.4. STD significantly reduced the number of subsequent shunt revisions required.5. Revisions of the STD craniectomy were less common and were performed in patients whose craniectomy had ossified, a common phenomenon in young patients due to raised levels of osteoblastic activities at this age.
6. Finally, STD as a procedure was found to be safe, with no significant cases of morbidity or mortality.
Only one acute complication of pseudomeningocele is reported, which was managed effectively.

SVS Aetiology
Slit ventricles may develop in up to 53% of paediatric patients with shunted HCP [1,3,6].However, only up to 5% of the paediatric shunted population displays

Subtemporal Decompression in Resistant
Slit Ventricle Syndrome in Children concurrent symptoms of SVS [3,4,7].SVS has been described in patients with (1) normal ICP (where the ventricle is unable to expand due to subependymal stiffness), (2) low ICP (due to overdrainage, which can in turn lead to raised ICP if the proximal catheter obstructs), and (3) raised ICP (for example, due to early suture fusion and cranial vault incapacity) [8].
For those in the last group, poor ventricular compliance is a consistent feature, likely owing to sub-optimally treated chronically high ICP.Once established, the logical solution of adding an anti-siphon, or adjusting a valve for shunt underdrainage (with the aim to expand ventricular volume) unfortunately often does not relieve symptoms.
All the patients in this cohort had their shunts in infancy, and most during the first 8 weeks of life.Rekate noted that SVS can occur in patients shunted in infancy, owing to their differing CSF dynamics [9].Rekate reported that infant HCP often develops when there is a pressure difference from the ventricles and a space exterior to the brain.When the fontanelles close (post-shunting), the ventricles can collapse due to changes in these dynamics.Raised pressure can remain, despite collapsed ventricles [9,10].In this group of patients, decompression may be an effective adjunct to manage their symptoms.STD for SVS STD was initially described by Cushing as a treatment for raised ICP secondary to tumours [11].STD for the treatment of slit ventricles was first reported by Epstein et al. [12] in 1974, as an alternative to shunt surgery.Epstein et al. [12] (1988) suggested subtemporal craniectomies in patients with recurring symptomatology, normal ventricular volume, and shunt function if shunt revision surgery is not successful [3].
STDs are usually been performed unilaterally in the same side as the shunt, to allow space for parenchyma, hence allowing the ipsilateral collapsed ventricles to dilate [3,[13][14][15].Similarly, we found significantly better outcomes in patients undergoing unilateral STD, possibly owing to the same concept of local "relaxation" of the brain.
In our study, we found that patients with STDs had improved symptomatology and reduced shunt revisions.These results concur with Buxton et al. [16], who reported decreased shunt revisions and hospital attendances following STD in 15 patients.However, in contrast to our results, these authors found the rate of shunt revisions to be 3 times higher during the first 3 years post-STD when compared to the 3 years pre-STD.It is unclear why this finding occurred, but there have been reports of paradoxically smaller ventricles following STD [13].
Roth et al. [17] report symptom resolution in 73.4% of patients with a modified bilateral STD technique, in which the dura and arachnoid mater were opened.They proposed a treatment algorithm comprising of conservative medical management with lumbar punctures, STD, and reserving cranial expansions for patients with severe cranio-cerebral disproportions.Rekate (1993) and Olson (2004) propose a similar treatment algorithm reserving such calvarial expansions for the latter group [1,18].

Our Current Protocol
A similar treatment algorithm to that used by the aforementioned groups was employed for patients in this cohort, with conservative measures trialled first.Vault expansions and STDs were reserved for resistant SVS in patients who already had programmable valves and antisyphon devices and in those where attempts to repositioning the ventricular catheter had failed.

Reasons for STD Failure
There are many possible reasons for failure of STD including insufficient decompression, acute temporalis muscle oedema, temporal lobe herniation (not found in our cohort), and bone re-ossification.An insufficient decompression could be due to a too small decompression, or due residual compression from the dura.An expansion duroplasty can provide further decompression volume.In our cohort, a sheet of absorbable haemostatic oxidised cellulose was laid on an unclosed dura mater.
One acute complication is temporalis oedema which can cause localized further mass effect.One option is to resect the temporalis muscle [19,20].It is important however to be aware of and counsel patients regarding the cosmetic and functional implications of temporalis resection.
A common complication observed in our cohort was bone re-ossification, a phenomenon also observed in another series of STD procedures for resistant SVS [17].Paediatric patients are known to possess strong potential for calvarial regrowth, as evidenced in the craniosynostosis literature [21].It is possible that the "opened" cranial vault could lead to dampening of the arterial pulse and result in local ossification around the STD site.An alternative hypothesis is that the developing temporalis muscle in infants can exert forces and influence osteoblast activity and subsequent regeneration of the temporal pericranium [22].Progenitor cells in the pericranium are also known to play a significant role in spontaneous ossification at craniectomy sites [23][24][25].Cytokines and prostaglandin E1 enable an environment suitable for bone reformation following trauma [26,27].Cyclooxygenase inhibitors delivered locally could in turn slow regrowth of bone in these patients [26].
Alternative Treatments for SVS Anti-migraine therapy reduced headache symptoms in 7 patients treated by Obana et al. [28], avoiding further shunt revisions and cranial expansion-related procedures.None-the-less, many children have symptoms requiring more than medical management.First-line treatments for SVS include shunt-related procedures, including addition of programmable valves and anti-syphon devices.These aim to prevent overdrainage and to avoid proximal catheter obstruction due to coaptation of the ventricular wall around the proximal catheter.
Shunt removal followed by endoscopic third ventriculostomies have also been shown to have significant improvement to symptoms in 73% [29].A treatment algorithm proposed by Rekate (2008) suggests ETV in patients with ventricular dilatation associated with raised ICP 24 h after the shunts are externalized and ligated off [30].However, the aetiology of the SVS is important when selecting an algorithm [30].When all of these alternatives fail, calvarial expansion-related procedures in the form of STD and cranial vault expansions have been reported, with varying degrees of success.

Future Work
The optimal solution for SVS would be to prevent the development of the condition in the first place.This could be achieved by avoidance of shunting, optimizing shunt settings, avoiding states of overdrainage or chronically raised ICP (e.g., avoiding the choroid plexuses with catheters, or early treatment of high-pressure states), or avoiding excessive fluctuations between over-and underdrainage.Therefore, protocols addressing SVS should include a period of ICP monitoring.However, effective prevention is yet to be achieved, and many patients requiring shunts will still go on to develop SVS.Therefore, ongoing research into prevention, and management of SVS is required.

Limitations and Strengths
STD is a relatively rare procedure, reserved for only the most complex of patients with failed alternative procedures.Like prior studies of STD, the number of patients was limited.Furthermore, this study is reviewing STD as a secondary rescue procedure, after multiple failed shunts, and not as a primary procedure.None-the-less, we show a clear reduction in shunt revisions, suggesting a role for STD in this group of complex patients.

Conclusions
A large proportion of patients with persistent SVS symptoms, refractory to multiple shunt revisions, benefitted from STD in combination with shunt optimization.It was also safe and well-tolerated.Therefore, in patients who have multiple failed shunts, STD may reduce the morbidity associated with further shunt revisions and can significantly improve symptomatology.

Statement of Ethics
This research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki.No patient-identifiable information or images were used.The service evaluation project is registered with the Great Ormond Street Hospital (GOSH) Quality Committee (Registration number 3467).The study has been approved as being within the NHS Health Research Authority criteria for service evaluation.Ethical approval and informed consent are not required for this study in accordance with UCL Ethics Committee guidelines.

Fig. 2 .
Fig. 2. ICP data pre-and post-STD.a ICP before STD.b ICP after STD.c PA before STD.d PA after STD.Example of ICP recordings and pulse amplitude values before and after STD.ICP, intracranial pressure; PA, pulse amplitude.

Fig. 3 .
Fig. 3. CT scan 4 years post-STD surgery showing bilateral decompression (a), satisfactory lateral ventricle dimensions (b), right STD (c), and left STD (d), both showing some reduction in diameter due to an element of bony regrowth.

Fig. 4 .Fig. 5 .
Fig. 4. Number of shunt revisions pre-and post-STD.Graphs showing the number of shunt revisions as well as the mean difference in the number of shunt revisions before and after STD procedure for all individual patients.STD, subtemporal decompression.

Table 1 .
Summary of the primary aetiology (for initial VP shunt) and outcomes (clinical and radiological) following STD