Surgical treatment of long-standing overt ventriculomegaly in adults (LOVA)

The clinical onset of LOVA can occur at any stage in adulthood, with a range between 22 and 81 years (median 54) in our case series, which has been mirrored by others [4, 7, 8]. No clear-cut differences in patient sex have been identified thus far. Some series have identified a trend towards sub-normal IQ and cognitive abilities in LOVA patients, although this feature does not appear to be universal across all cases of LOVA [4, 7, 8]. Given the progressive nature of LOVA, early identification of patients with potential symptoms and signs of the condition is an attractive notion, as it may help to maximise the impact of any surgical intervention to implement it earlier in the disease process. However, the subtle myriad of early symptoms and the non-specific nature of large head circumference make early detection of LOVA a clinical challenge.

The key features of LOVA are aqueductal stenosis, with arrest of hydrocephalus before the onset of gross macrocephaly and the symptoms of raised ICP. During the arrested hydrocephalic period CSF flow is likely maintained via a combination of rencannalisation of the aqueduct, utilisation of alternative CSF flow routes, and modification of CSF production or absorption [1]. This asymptomatic interval is followed later in life by a failure of the compensatory processes, leading to the symptomatic phase of LOVA, which is typically progressive and long term. Patients with an identifiable insult to alter CSF flow are excluded from the definition of true LOVA, and the mechanisms that alter CSF dynamics to end the asymptomatic phase of the condition remain unclear.

It is thought that the symptoms of LOVA develop as a consequence of (1) failure of adequate CSF flow, (2) skull base changes as a consequence of chronic and progressive rise in ICP and (3) chronic pressure effects upon brain parenchyma [1, 9, 10]. As such, early intervention has the potential to modify the disease process before irreversible changes occur as a result of chronically raised ICP [4].

Reports of persistent symptoms in patients with LOVA after both ETV and shunt procedures have led to a theory that, in some cases, the pathological process in adulthood leading to CSF volume imbalance may actually take place more distally than the aqueduct; it may be a phenomenon of failure of CSF re-absorption, thus unnameable to traditional CSF diversion procedures. One case reported by Rekate [13] illustrated a LOVA patient whose symptoms remained refractory to CSF shunt diversion and ETV, which in itself was complicated by shunt infection. A venogram revealed bilateral transverse sinus stenosis, which was successfully treated utilising a neuro-radiological venous-stenting procedure. The patient’s ICP subsequently normalised. It is possible that back-pressure from blockade of dural venous sinuses could instigate a chronic, potentially reversible, aqueductal stenosis.

Al-Jumaily et al. identified the most common features in LOVA were headaches and imbalance [4], which were also the most common presenting symptoms in our case series (Table 3).

The same group also highlighted that patients with LOVA may demonstrate a myriad of cognitive and psychological problems such as decline in memory, attention and language skills, along with depression, anxiety and disinhibition [4, 7, 8]. Some case series have retrospectively identified that a proportion of LOVA patients demonstrated sub-normal IQs or evidence of early dementia before the onset of more overt symptoms of raised ICP, such as headaches and gait disturbance [4, 7, 8]. However other patients demonstrate normal-to-high IQs and report no evidence of hydrocephalic symptoms at childhood, with a large head circumference the only pre-symptomatic indicator of a potential CSF flow problem [3, 4]. Whether any pre-symptomatic decline in cognitive function identified in a proportion of LOVA patients is a consequence of the arrested hydrocephalus at childhood or whether it represents the earliest phase of the decompensation of CSF flow in adulthood remains to be explored.

These symptoms may be difficult to identify clinically, as the chronic nature of LOVA may permit some patients’ brains undergo sufficient neuroplasticity to permit functional reorganisation, minimising any clinical neuropsychological deficits until late in the disease process [4, 8]. Nevertheless, when cognitive decline, inattention and mood disturbances do occur in LOVA, they can have significant impacts upon quality of life [4]. These symptoms were marked in one case from the cohort at our institution. In this case there was significant improvement in the social functioning, memory and mood after ETV (Table 3). Whilst there are other reports of improvement in these faculties after surgical treatment of LOVA [8], it is thought that these symptoms are less likely to improve if treatment is implemented late in the disease process [4]. Neuropsychological symptoms should therefore be assessed and considered early in the therapeutic decision-making process for LOVA patients who may appear to function highly even after the point of clinical decompensation ensues [4, 6‐8].

All reviewed studies advocated surgery for symptomatic cases of LOVA. However, the optimal CSF diversion procedure for patients with LOVA has been debated for over a decade [1, 2, 9‐12]. This debate has centred on a debate between neuroendoscopic procedures versus CSF shunt diversion. Oi et al. [1] reported outcomes for 18 patients with LOVA, 9 undergoing neuroendoscopic ETV (8) or aqueduct-plasty (1) and 9 receiving ventriculoperitoneal (VP) shunts [7 differential pressure valve (DPV) shunts, 2 pressure programmable valve (PPV) shunts]. All seven patients receiving DPV shunts developed subdural haematomas, most likely as a consequence of over-drainage. All of these patients required shunt revision to a PPV shunt and/or ETV. Six of the nine patients undergoing neuroendoscopic procedures developed radiological and clinical arrest of hydrocephalus. Two patients in this study suffered intra-axial haemorrhages after ETV. However, subsequent series have demonstrated the relative safety of ETV, with an overall complication risk of approximately 6 % [4, 13]. In one case series, a single ETV provided long-term symptomatic relief for 88 % (21/24) of patients with LOVA with no long-term complications port-operatively [6]. This correlates with our series, where there was only one complication post ETV: a self-limiting generalised seizure with no long-term sequelae for the patient. Conversely 100 % (n = 14) of patients in our cohort reported improvement or absence of progression in their symptoms at their first follow-up appointment after ETV (3 months).

There are no case reports of significant change in ventricular diameter after ETV or shunt insertion. The explanation touted for this is that only a small volume of CSF needs to be drained to achieve near-physiological ICP in adult hydrocephalus [11]. Improvements in radiological CSF flow and ICP after ETV for LOVA have been identified [1, 13]. However, these do not appear to correlate with clinical improvement, and recurrence of symptoms does not reliably relate to failure of flow through the ETV stoma [13]. It seems that after initial drainage of CSF through the stoma, a more complicated reorganisation of CSF dynamics takes place. Unpacking this mechanism may then be the key to understanding how to achieve sustained relief from the symptoms of LOVA.

The majority of studies, our series included, demonstrate that a majority of LOVA patients can achieve post-operative clinical improvement in headaches, balance, motor skills and neuropsychological function after ETV [3, 4, 7, 8]. One large retrospective study of ETV in adult patients (n = 163) recently identified that the majority of patients undergoing primary ETV (87 %; n = 130) reported subjective improvement in symptoms 3 months post-operatively [3]. Only ten patients (6 %) suffered post-operative complications. It should be noted that the authors present data relating to their cohort as a whole, which encompassed adult patients with normal pressure hydrocephalus, partially treated childhood hydrocephalus (i.e. failed VP shunt), hydrocephalus secondary to other intracranial pathologies and LOVA. Separate outcome data for these specific causes of adult hydrocephalus were not presented; however, 97 patients in the whole cohort (60 %) did meet the diagnostic criteria for LOVA, and their outcomes for ETV success and complications align with data from other case series specifically reviewing cases of LOVA (Table 2).

Despite these impressive outcomes after ETV, other works report less promising results. In one series, six patients with LOVA underwent ETV [10]. All six patients required a secondary procedure for persistent symptoms: shunt insertion, repeat ETV or venous stenting. This small series highlighted the need for continued follow-up whatever intervention is implemented, as 50 % of patients (n = 3) suffered recurrence of symptoms at 18–36 months after their primary ETV, despite an initial improvement or arrest in progression of symptoms. Other studies subsequently identified varying long-term success rates for ETVs carried out for secondary hydrocephalus and LOVA, with recurrence of symptoms occurring a number of months or years post-operatively in a small proportion of patients in some series [1, 4, 7‐10]. Furthermore, studies have presented patients in whom ETV did not facilitate objective improvement in cognitive or neuropsychological tests, which may reflect the progressive nature of the condition and inter-individual differences in neural compliance after a period of chronic hydrocephalus [4]. In patients with recurrent symptoms repeat ETV or shunt insertion can sometimes deliver symptomatic improvement, but this is again unpredictable. Thus, decisions to re-operate need to be guided by individual patient circumstances. Many cases of LOVA treated with ETV, including those in our series, have demonstrated a benefit in the short-to-medium term. However long-term follow-up, over a period of years, should be considered, given the reports of recurrence and progression of symptoms in a minority of LOVA patients up to 2–3 years post-operatively [4, 10, 13].

Two major advantages of ETV over traditional shunting procedures are the high infection (and thus revision) rates for shunt devices and the risk of shunt over-drainage and its sequale in the latter. After the dramatic complications following PPV insertion in the initial series by Oi et al., focus shifted onto use of gravitational shunts for LOVA patients to minimise the risk of over-drainage [1, 2, 11, 12]. In an important study of shunt procedures for LOVA patients, 23 patients elected gravitational shunt insertion after being offered the choice of ETV or shunt insertion [11, 12]. Two patients (9 %) developed small subdural haematomas, one of which required drainage. Eighty-two percent (19) of the cohort demonstrated clinical improvement after shunt insertion (follow-up 6–75 months). The authors conclude that gravitational shunts are a viable alternative to ETV in the management of LOVA. However, it is of note that 15 % (4) of patients developed symptoms of overdrainge post-operatively, two of which necessitated shunt revision surgery. Median follow-up was 29 months, which may not have been extensive enough to capture cases of later shunt failure or infection in the cohort. Furthermore, the work was limited by the non-randomised methodology for the selection of patients for shunt insertion and the institution’s relatively large, rigid endoscopes leading to strict criteria for excluding ETV as a therapeutic option (foramen of Munro width <6.5 mm and third ventricle floor thickness >2.5 mm).

When discussing treatment options for LOVA, the significant infection and revision rates associated with shunt procedures, and the risks of under- or over-drainage, must be considered [4]. Conversely, the ventricular anatomy may not always be amenable to endoscopic CSF diversion procedures. A comparative study between ETV and shunt implantation for LOVA is yet to be completed; such work could provide useful data to aid therapeutic decision-making for these patients.