Posthemorrhagic hydrocephalus management strategies

Case history: Infant A

A mother in her first pregnancy had an uncomplicated pregnancy and received the first dose of corticosteroids 4 hours prior to an emergency Caesarean section at a gestation of 30 weeks and 5 days, because of suspected intra-uterine infection with fetal tachycardia and decelerations on the CTG. A female infant was born weighing 1400 g at delivery. Her Apgar scores were 3, 6, and 8 at 1, 5, and 10 minutes, respectively. Her respiration was supported with nasal continuous positive airway pressure and she received surfactant using the minimally invasive surfactant therapy (MIST) procedure. Her CRP was 118 mg/L on the second day after birth. Listeria monocytogenes was cultured from her blood culture. There was no pleocytosis (<7 10 ⁶ /L white cells) in the CSF. Amoxicillin was given intravenously for 3 weeks. Her initial cranial ultrasound scan was unremarkable ( Fig. 6.1 A) but a repeat scan on day 4 showed bilateral IVH grade II–III ( Fig. 6.1 B). She was then scanned twice a week. Ventricular dimensions progressively enlarged and LPs were performed from day 10 onward (see Fig. 6.1 C). These were successful and 10 mL/kg of port wine–colored cerebrospinal fluid (CSF) could be obtained each time.

As there was a need for repeated tapping of CSF and repeated LPs were becoming impractical, an Ommaya reservoir (ventricular access device) was inserted frontally in the right ventricle with the patient under general anesthesia. The reservoir was tapped daily, at 10 mL/kg/day for 4 days. Thereafter, the reservoir was tapped 7.5 mL/kg twice a day as the ventricular dimensions did not decrease sufficiently. Tapping was increased to 15 mL/kg twice a day for 3 days and finally to 20 mL/kg twice a day. This increase in CSF removal was based on ventricular measurements on daily cranial ultrasound scans.

CSF protein was initially 1.7 g/L. Tapping the reservoir continued to be necessary for 7 weeks, and it was then decided to insert a ventriculoperitoneal low-pressure shunt when Infant A reached full-term equivalent age. Postoperatively, there was no pulmonary problem, CSF leak, or infection. Magnetic resonance imaging (MRI) at term showed very mild ventricular dilation.

Question 1: What measurements of ventricular size are used in diagnosis of PHVD?

The likelihood of progressive ventricular dilation increases with the amount of blood visible in the ventricles. With a small grade II intraventricular hemorrhage, measurement of ventricular size once a week for 4 weeks and then at discharge is appropriate; with a large grade III or IV IVH, twice-weekly ultrasonography is needed as a minimum because dilatation is likely and may be rapid. Although large, balloon-shaped ventricles are obvious without formal measurements, quantitative documentation is essential for decision making. Reference ranges for measurement of the width (midline to lateral border) of the lateral ventricles at the midcoronal level were first published in 1981. Since 1984, an “action line,” defined as width 4 mm higher than the 97th centile width for age, has been used as a definition of serious PHVD in therapeutic trials ^, and as a secondary outcome in randomized trials of neonatal intensive care interventions ( Fig. 6.2 ). This measurement has the advantage that it is highly reproducible among observers because it is relatively unaffected by anterior or posterior angulation of the scan head as the lateral wall of the ventricle in this orientation runs fairly parallel to the midline. However, ventricular enlargement is not always sideways, and sometimes the most marked change is posterior enlargement or a change from thin slit to round balloon. With this in mind, Davies and colleagues published reference ranges for anterior horn width (to capture the change in shape to balloon) (95th centile approximately 3 mm), thalamo-occipital width (to capture posterior enlargement) (95th centile approximately 25 mm), and third ventricle width (95th centile approximately 2 mm). New graphs have become available including infants below 26 weeks’ gestation. An electronic spreadsheet is available to be downloaded, stored in secure servers, and used for individual patients in two formats: one is for postmenstrual age 24–42 weeks ( https://tinyurl.com/PHVD-Measures-1 ) and another for postmenstrual age 24–29 weeks ( https://tinyurl.com/PHVD-Measures-2 ). The fronto-temporal or fronto-occipital horn ratio (FTHR or FOHR; widest distance of the frontal horns plus temporal or occipital horns, respectively, divided by twice the largest bi-parietal distance) are also used. ^, A recent study showed that the intra-observer and inter-observer reliability are best for the VI and AHW and that the AHW best predicted subsequent development of PHVD requiring neurosurgical intervention. A combination of ventricular width over 97th centile and anterior horn width over 6 mm was used as eligibility for the ELVIS trial.

Question 2: How can ventricular dilation driven by cerebrospinal fluid under pressure be distinguished from ventricular dilation due to loss of periventricular white matter?

To distinguish between CSF under pressure and loss of periventricular white matter as the cause of ventricular dilation is important because removing fluid that has accumulated as a replacement for dead brain (ex vacuo dilatation) is unlikely to improve outcome.

CSF-driven ventricular enlargement can be slow or rapid, it is characterized by balloon-shaped lateral ventricles, and if CSF pressure is measured, it is found to be raised or near the upper limit of normal, mean 3 mm Hg, upper limit 6 mm Hg. Furthermore, head circumference growth over time is accelerated, although it may lag behind ventricular enlargement by 1 to 2 weeks. In contrast, ventricular enlargement from atrophy is always slow, it is more irregular in outline rather than balloon shaped, and if CSF pressure is measured, it is found not to be raised. Head circumference velocity is either normal or slow but is not accelerated. There may also be notable accumulation of extra-axial fluid such as fluid between the hemispheres and between the cortical surface and the skull. Nonprogressive mild ventricular dilation at term is common and a marker of white matter injury.

Question 3: How is excessive head enlargement defined?

Head circumference normally enlarges by approximately 1 mm per day between 26 weeks of gestation and 32 weeks, and about 0.7 mm per day between 32 and 40 weeks. We regard a persistent increase of 2 mm per day as excessive. Measuring head circumference accurately, although “low-tech,” is not as easy as it sounds. The relevant measurement is the maximum fronto-occipital circumference. Detecting a difference of 1 mm from day to day is difficult, and we do not react to a difference of 2 mm from one day to the next unless there is other evidence of raised intracranial pressure. However, an increase of 4 mm over 2 days is more likely to be real, and an increase of 14 mm over 7 days or less is definitely excessive. This is, however, a late sign of PHVD and poor correlation between head circumference and cranial ultrasound findings in premature infants with intraventricular hemorrhage has been reported. It is therefore not recommended to wait for rapid increase in head-circumference for a decision to start with intervention.

Question 4: How is raised intracranial pressure recognized?

It is possible to detect a change in palpation of the fontanelle from concave to bulging and to document excessive head enlargement. The preterm skull is very compliant and can easily accommodate an increase in CSF by expanding with separation of the sutures. When CSF pressure was measured with an electronic transducer in infants in whom ventricles were expanding after IVH, the mean CSF pressure was approximately 9 mm Hg, three times the mean in normal infants. There was a considerable range, with ventricle and head expansion in some infants at a pressure of 5 to 6 mm Hg, and in a small number with CSF pressure around 15 mm Hg. A CSF pressure of 9 mm Hg does not necessarily produce clinical signs but may be associated with an increase in apnea or vomiting, hypotonia, hypertonia, or decreased alertness. A structured neurological examination such as that published by Dubowitz is recommended.

Obtaining serial calculations of the Doppler flow-velocity resistance index (RI) on the anterior cerebral artery is a useful and practical way of detecting impairment of cerebral perfusion by raised intracranial pressure and can easily be done during ultrasound imaging. The resistance index is calculated as follows: (systolic velocity − diastolic velocity)/systolic velocity. This measurement is independent of the angle of insonation. If intracranial pressure rises to a level exceeding the infant’s compensation, end-diastolic velocity tends to decrease, eventually becoming zero (RI is then 1.0) ( Fig. 6.3 ). Serial increases in RI above 0.85 while the ventricles are rapidly expanding would be evidence that pressure is rising. This statement assumes that the infant does not have a significant left-to-right shunt at the ductal level and that P co ₂ has not decreased recently, because both of these physiologic changes could increase RI. Severe intracranial hypertension may cause reversed end-diastolic velocities. The sensitivity of resistance index can be increased by applying pressure to the fontanelle during the examination. An infant who is close to the limit of cranial compliance responds with a large decrease in end-diastolic velocities—that is, an increase in RI. Somato-sensory evoked potentials and amplitude-integrated electroencephalography (EEG) may show a deterioration, with electroencephalographic activity becoming less frequent as dilation increases and improving with effective CSF drainage. Several groups have shown that repeated near-infrared spectroscopy (NIRS) demonstrated deteriorating cerebral oxygen saturation with increasing ventricular size and improvement after decompression. We regard all these signs of pressure as late in the development of PHVD and do recommend to start intervention before these signs have occurred.

Question 5: What is infant A’s prognosis?

The prognosis at diagnosis of PHVD using the preceding criteria is influenced by the presence of identifiable parenchymal lesions. The Ventriculomegaly and PHVD Drug Trials used the 4 mm + 97th centile definition of PHVD and had standardized follow-up. Of children in whom ultrasonographic examination shows no persistent periventricular echodensities or echolucencies (cysts), approximately 40% had cerebral palsy. ^, In the DRIFT trial, using the same treatment criteria, follow-up at 10 years of age in infants having similar management to A showed mean developmental quotient without hemorrhagic infarction to be 71 (+/–36) with 37% having cerebral palsy, 46% having extra educational resources and 23% at special schools. In the more recent ELVIS trial where intervention started before or just after crossing the 4 mm + 97th centile, less than 10% of the infants with a grade III hemorrhage developed CP. In the infants enrolled in the Preterm Erythropoietin Neuroprotection Trial, who were less mature, 19% of those with grade III developed CP. Magnetic resonance imaging (MRI) at term can reveal parenchymal injury which cannot be easily detected with ultrasound, such as noncystic white matter injury, grey matter abnormality, and cerebellar hemorrhage and/or atrophy.

Question 6: What is the mechanism of PHVD?

Following a large IVH, multiple blood clots can obstruct the ventricular system or channels of reabsorption, initially leading to a phase of CSF accumulation. Although tissue plasminogen activator can be demonstrated in posthemorrhagic CSF, fibrinolysis is very inefficient in the CSF, which has low levels of plasminogen and high levels of plasminogen activator inhibitor. ^, This potentially reversible obstruction by thrombi may lead to a chronic obliterative, fibrosing arachnoiditis, and subependymal gliosis involving deposition of extracellular matrix proteins such as laminin in the foramina of the fourth ventricle and the subarachnoid space ( Fig. 6.4 A and B).

Transforming growth factor-β (TGF-β) is involved in the initiation of wound healing and fibrosis and is likely to be one mediator of this process. TGF-β elevates the expression of genes encoding fibronectin, various types of collagen, ^, and other extracellular matrix components, and is involved in a number of serious diseases in which there is excessive deposition of collagen, including diabetic nephropathy and cirrhosis. TGF-β1 is stored in platelets and thus provides a store of TGF-β1 for many weeks in the CSF after IVH. TGF-β is elevated in the CSF of adults with hydrocephalus after subarachnoid hemorrhage, and intrathecal administration of TGF-β to mice resulted in hydrocephalus. ^, TGF-β1 and TGF-β2 concentrations in CSF from infants with posthemorrhagic ventricular dilation are 10 to 20 times those in nonhemorrhagic CSF and the concentration of TGF-β in CSF is higher in those shunted later. A product of TGF-β, aminoterminal propeptide of type 1 collagen is elevated in CSF from PHVD. Chow and associates have demonstrated elevation of TGF-β2 and nitrated chondroitin sulfate proteoglycans (an extracellular matrix protein) in CSF from preterm infants with posthemorrhagic hydrocephalus.

A rat pup model of PHVD has also provided evidence of the involvement of TGF-β and its downstream products, fibronectin, laminin, and vitronectin. ^, Transgenic mice that overexpress TGF-β1 in the central nervous system are born with hydrocephalus. Thus there is a strong possibility of a role for the TGF-βs in the development and/or maintenance of hydrocephalus after ventricular hemorrhage. However, two drugs that inhibit production of TGF-β, pirfenidone and losartan, did not reduce ventricular size or improve neuromotor performance in a rat pup model of PHVD.

Amyloid precursor protein (APP) and L1 cell adhesion molecule (L1CAM) were also noted to be increased in infants with PHVD and were significantly related to fractional anisotropy in the corpus callosum.

Question 7: How can PHVD injure white matter?

Damage to periventricular white matter is probably exacerbated by ischemia due to raised intracranial pressure and parenchymal compression, by oxidative stress due to the generation of free radicals, and by the actions of inflammatory cytokines.