Due to the vulnerability of preterm infants, VP shunt placement is not conducted as an initial treatment option as an intervention for PHH. VP shunt is inevitably chosen when temporary management to drain CSF is unsuccessful. It is important to note that there are some cases in which PHH becomes arrested hydrocephalus and no additional treatment is necessary. In rare cases, ventricular dilatation might improve with spontaneous resolution.

As for when exactly to convert to VP shunt placement, the size of the ventricles and degree of head circumference (HC) enlargement are indicators, not to mention if the clinical signs do not improve with temporary CSF management. And, as will be discussed later, the infant’s physical growth and the degree of clearance in hemorrhagic CSF are also important determinants. Full fontanelle, splaying sutures, irritability, apneas, bradycardia and sunset phenomenon are typical clinical signs of progressive hydrocephalus [20,37]. We must recognize that VP shunt surgery after the appearance of these clinical signs is somewhat too late to make a decision.

Quantitative measurement of the ventricles by transcranial ultrasonography is routinely performed. The most frequently used index is ventricular enlargement of 4 mm above the 97th percentile on the ventricular index. Another indicator is progressive HC enlargement. HC 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. A persistent increase of 2 mm per day is regarded as excessive [40]. An increase of 6 mm over 3 days is more likely to be real and an increase of 14 mm over 1 week is definitely excessive [40]. Before chronic intracranial hypertension is left untreated and various clinical symptoms appear, shunt placement should be considered based on objective judgment using measurements such as ventricular size and head circumference as indicators.

There is insufficient evidence to recommend a specific body weight or CSF parameters for the timing of shunt creation in premature infants with PHH, but the most important parameter is the body weight [31]. Even if PHH is progressive, VP shunt is done when infants have gained sufficient weight. Very preterm infants have an immature immune system, inadequate peritoneal absorption capacity, and fragile and thin skin, which leads to an extremely high incidence of shunt-related complications, including shunt infection [5,6,9,24,29,37,41]. Some reports recommend to until the infant weighs 2.5 kg [31,40], while others recommend surgery once the infant weighs 2 kg [29]. In addition, it has been reported that good neurologic functional outcome was achieved if surgery was performed when the weight reached 1.5 kg [4,32]. On the other hand, performing VP shunt placement at a weight of less than 2.0 kg is likely to result in shunt infection and ultimately worsen the neurological function. If a VP shunt is recklessly performed for a very small infant, ulceration of the skin at the valve site would be inevitable.

In a recent single-center retrospective study, permanent shunts were placed at an average gestational age of 43 weeks and an average weight of 2.9 kg [15]. Each children’s hospital has its own criteria for when to place a VP shunt, but it is common practice to wait until the infant has grown to a weight of 2.5 to 3.0 kg.

If CSF contains large amounts of blood and protein, it is very likely that the shunt system will not drain properly and will require early revision surgery. Lower CSF protein levels have been advocated. It is generally recommended that a shunt be inserted once CSF protein has decreased to less than 150–200 mg/dL, although some have suggested that less than 1000 mg/dL is acceptable [16,18]. Of course, the absence of signs of infection in the CSF is also an essential requirement.

A wide variety of shunt systems are now available to clinicians, but their impact on shunt failure rates is still unknown [24]. A comparative study of newly developed valves has not identified any differences in reoperation rates or incidence of over drainage between valve types [3].

Evidence-based guidelines also point to insufficient evidence to recommend the use of programmable versus nonprogrammable valves. Both programmable and non-programmable valves are possible treatments for PHH [1], but in actual clinical practice, small and low-profile shunt valves tend to be preferred because of the fragile scalp of infants.

Previous reports have shown that placement of a VP shunt in premature PHH infants is more likely to result in shunt revision, slit ventricle syndrome, loculated hydrocephalus, and shunt infection than in infants with hydrocephalus of other causes [30]. Based on reports from various centers, shunt revision rates ranged from 45.2% to 84.1%, and 1-year and 5-year survival rates ranged from 40% to 73.6% and 13.8% to 44.1%, respectively (Table 1) [5,6,9,24,29,37,41]. It is clear that children with PHH are more likely to undergo shunting earlier than other categories of patients in terms of weeks of gestation rather than weeks after birth [24]. It is widely known that very young age and low body weight at the time of initial shunt insertion are associated with higher rates of shunt failure [23,35].

Thus, younger age and lower weight at the time of initial shunt placement undoubtedly have a significant impact on shunt survival in PHH, which is attributed to the difficulty of the surgical technique and the fragility of the infant’s brain, intestinal tract, and skin. Fragile infants are vulnerable to any invasion.

In PHH infants undergoing shunt surgery, the condition of being post-IVH is also significantly associated with shunt survival. Excessive CSF protein levels contribute to shunt dysfunction and catheter occlusion [29]. This is also true for the treatment of PHH in adults.

Another troublesome problem associated with shunt creation in preterm infants is the complication of infection. The high incidence of shunt infections is likely due to an immature immune system and inherently weak antimicrobial ability [30]. Shunt infections have been reported in 7.1–22.4% of preterm infants (Table 1) [5,6,29,37,41].

For these susceptible infants, antibiotic-impregnated catheter is highly recommended [27,32], and every effort should be made to reduce shunt infections to near zero using strict infection prevention protocols [17]. New Hydrocephalus Clinical Research Network protocol advocated by Kestle et al. [17] in 2016 and the protocol of my department are presented in Table 2.

An isolated TFV is a peculiar complication of VP shunt placement for PHH and is caused by over drainage of the supraventricular shunt [25]. PHH occurring in preterm infants is known to be a complex combination of both communicating and non-communicating hydrocephalus factors, but the Sylvian aqueduct and foramen Luschka and Magendi may have membranous obstruction due to inflammatory scarring after IVH [28].

When the supratentorial CSF shunt is overactive, the occlusive mechanism of the Sylvian aqueduct becomes even more pronounced and combined with the obstruction of the fourth ventricular outlet, the flow path of CSF produced by the choroid plexus in the fourth ventricle is blocked and the fourth ventricle is isolated and enlarged.

Although TFV may result from other etiology, the frequency of TFV is higher in PPH in preterm infants. Pomeraniec et al. [28] reported that the frequency of TFV following VP shunt for PHH in preterm infants was 15.4%.

Treatment options include fourth ventriculostomy with posterior fossa craniotomy, placement of an additional shunt in the fourth ventricle, and endoscopic procedures (aqueductoplasty and/or stenting, cystoventricular stenting) [10]. Note, however, that even if the fourth ventricle is isolated and enlarged on computed tomography or magnetic resonance imaging imaging, it is often asymptomatic, and furthermore, because cerebellar atrophy is not rarely seen in preterm infants with PHH, the indication for treatment of TFV should be carefully considered.

In recent years, there has been a great deal of interest in how to treat PHH without CSF diversion surgery, and attention has focused on endoscopic ventricular lavage [33] and fibrinolytic therapy based on ventricular drainage management [26]. Given the non-negligible probability of troublesome problems in VP shunt surgery and subsequent management of PHH in preterm infants, the treatment concept of VP shunt avoidance is an important issue for us pediatric neurosurgeons to address. However, since not all children’s hospitals are able to implement the two aggressive treatment strategies described above, it is essential that VP shunt surgery and subsequent long-term management be carried out without trouble after appropriate temporary CSF management.

A female infant was born at 24 weeks 6 days of gestational and weighted 401 g. Her Apgar score was 1 point (1 minute) and 1 point (5 minutes). A grade IV IVH occurred on day 1 (Fig. 1A). Ventricular dilatation progressed (Fig. 1B) and ventricular drainage with peripherally inserted catheter underwent on 28 days after birth. VP shunt with a programmable valve was performed at the age of 4 months (Fig. 1C) because of the enlargement of her HC. Although shunt infection or revision are not required for 10 years, she has moderate disability. If surgical intervention such as endoscopic ventricular lavage [32] or fibrinolytic therapy [25] had been performed earlier, the prognosis might be better.