First Treatment in Infants With Hydrocephalus

For infant hydrocephalus, shunting is the current mainstay of treatment.1 The decades-long experience with shunts and their complications has, however, led us to look, sometimes desperately, for alternatives. Endoscopic third ventriculostomy (ETV), with or without choroid plexus cauterization (CPC), is the only current alternative for the treatment of infant hydrocephalus.2,3 Despite the fact that ETV success tends to be low in infants,3 there remains very strong interest in pursuing this alternative to shunt. If, however, we examine the issues critically, is this interest in ETV justified for infant hydrocephalus? I was tasked with defending the case for shunt in infant hydrocephalus. To do this, I will outline here the issues I believe are relevant and critically describe areas in which shunt might be objectively superior to ETV. I will also ask us to honestly examine some of the decision-making biases we as neurosurgeons bring to this debate. It is important to recognize that, given the debate format of this article, the facts presented intentionally favor shunt, leaving unmentioned the counterarguments that would favor ETV. In this manner, this article avoids some of the nuance that would be necessary to fully appreciate all aspects of this issue, and it must be understood in that context. WHAT IS INFANT HYDROCEPHALUS? To begin to defend the case for shunt for infant hydrocephalus, we first need to understand what infant hydrocephalus is in current practice. I will limit this to the North American practice scenario, using data from the Hydrocephalus Clinical Research Network (HCRN), recognizing that the scenario differs, often dramatically, in other parts of the world.4,5 The HCRN currently comprises 9 pediatric neurosurgery centers across North America that, as part of a core data registry, collect very detailed prospective data on all hydrocephalus surgeries performed at each member institution.6 The practice pattern within the HCRN is broadly representative of North American pediatric hydrocephalus practice. Therefore, data from the HCRN give us an accurate picture of what infant hydrocephalus actually is. Figure 1 provides data on the cases of infant hydrocephalus (age < 24 months) treated within the HCRN since 2008. The largest groups by far are infants < 1 month of age with myelomeningocele and those < 6 months of age (uncorrected) with intraventricular hemorrhage of prematurity. Traditionally, these are groups that have been found to do poorly with ETV, as can readily be seen from the ETV Success Score (ETVSS).3 The ETVSS was developed to predict the probability of ETV success at 6 months based on the child’s age, cause of hydrocephalus, and presence of previous shunt and has been validated in multiple settings around the world.7-13 The ETVSS reveals that young age is the strongest single predictor of poor ETV success. Therefore, for the largest group of infants, that is, those < 6 months old, the ETVSS predicts ETV success in no more than 50%. Thus, ETV does not appear inherently appealing. In fact, as Figure 2 shows, the vast majority of HCRN surgeons understand this: Despite the fact that many HCRN surgeons are quite experienced, even aggressive, neuroendoscopists, the majority of infants they treat are still shunted. This is also despite the fact that ETV-CPC (discussed later in more detail) is currently being aggressively carried out within the HCRN, resulting in ETV being slightly overrepresented in this infant cohort since about 2012.FIGURE 1: Age of patients and cause of infant hydrocephalus within the Hydrocephalus Clinical Research Network. IVH, intraventricular hemorrhage.FIGURE 2: Treatment choice for infant hydrocephalus within the Hydrocephalus Clinical Research Network. ETV, endoscopic third ventriculostomy.IS THERE EVIDENCE TO SUGGEST THAT ETV IS BETTER THAN SHUNT? Regardless of the fact that current practice in North America for infant hydrocephalus greatly favors shunt, it is fair to ask if there is evidence to suggest the superiority of ETV. This is a complex question, requiring the consideration of different types of outcome. These outcomes include, for example, the rates of surgical failure requiring repeat surgery, long-term neurocognitive outcome, and, perhaps most important, long-term quality of life (QoL). With respect to failure of treatment requiring repeat surgery, the literature suggests a likely benefit for shunt over ETV for most infants, at least in the short term. It must be noted, however, that direct comparisons of these treatments in infants are quite lacking. Using propensity scoring adjustment, our group has previously shown that, once major prognostic factors are accounted for, there is on average a higher failure rate for ETV compared with shunt, although this reverses about 3 months after treatment.14 These analyses, however, considered the full range of the hydrocephalus population and did not speak specifically to infants. Another analysis showed that, in patients with the lowest ETVSS (≤ 40% chance of ETV success, which is the group to which most infants belong), shunt appears to have lower failure than ETV initially.8 However, as with the previous study, the risk pattern reverses a few months after surgery, and ultimately, both treatments do quite poorly, with a 3-year success of < 40% for both. Probably the best available data to address this issue will come from the International Infant Hydrocephalus Study (IIHS, NCT00652470), a prospective comparison of ETV and shunt for aqueductal stenosis in infants < 24 months of age.15 The IIHS stopped recruiting in December 2013, and initial results on early surgical failure will be published shortly while we await results of the 5-year long-term outcome assessment. A potential, albeit unproven, benefit of shunting on long-term neurocognitive outcome might be the more consistent reduction in ventricular size seen with shunt compared with ETV. Previous work has established that even successful ETVs often lead to stable, asymptomatic ventriculomegaly.16 Although these children look well on the surface, it is not known whether this persistent ventriculomegaly causes subtle harm to long-term cognitive function. A small initial study of 11 patients with hydrocephalus from aqueductal obstruction showed no difference in long-term neurocognitive function between those with ETV and those with shunt.17 More recently, we examined a slightly larger group of children with pure obstructive hydrocephalus, excluding patients with other brain anomalies or other major neurosurgical interventions other than for treatment of hydrocephalus.18 In examining this group of 23 patients (17 who had ETV and 6 who had shunt), we saw a range in ventricular size, from nearly slit-like to massive. This study did not find a noticeable association between larger ventricles and either cognitive function or measurable white matter injury on the basis of diffusion tensor imaging. The main limitations of this study, however, were the small sample size and the lack of baseline, pretreatment control data. Therefore, these results should not be viewed as definitive, and the long-term effects of persistent ventriculomegaly on neurocognitive outcome remain unknown. In children with hydrocephalus, long-term QoL is perhaps the most important, overriding outcome.19-22 Although there is no standard, universally accepted definition of QoL, the World Health Organization’s 1948 definition acknowledged that the related concept of health was multidimensional and involved aspects of physical, mental and social wellness.23 A key hallmark of QoL, however, is that it should ideally measure how people feel about what they can do rather than just measuring what they can do. This is often challenging in pediatric populations because natural cognitive immaturity or developmental delay can hinder attempts to directly access how children feel about their functioning.19 Understanding these limitations, previous studies have suggested that, not surprisingly, it is the cognitive aspects of QoL that are the most affected in children with hydrocephalus.22,24,25 Studies have also shown that the main determinants, or at least associations, with poorer QoL include certain causes of hydrocephalus, the presence of epilepsy, and socioeconomic factors.24 As it relates to shunt vs ETV, it would be tempting (and to some intuitive) to assume that ETV patients have better QoL than shunt patients. The data, however, are not so clear. It is true that large, retrospective studies have shown that increased shunt complications do seem to be associated with poorer QoL.22,24,25 However, this might not be as clear an indictment of shunt as it seems because it does not speak to whether shunt or ETV was the first treatment in these children. That is to say, some children who suffer multiple shunt revisions and a poorer QoL might have been those who failed an initial ETV attempt. Similarly, many children undergo shunt insertion, suffer few, if any, shunt complications, and have a good QoL. To more directly compare QoL in ETV and shunt, we previously examined a small, select group of 47 children with aqueductal obstruction and no other brain abnormalities, roughly half of whom were initially shunted and half initially treated with ETV.17 No convincing difference in QoL was seen between these 2 groups. In a second study, we examined QoL in a more heterogeneous population of 58 children with ETV and 545 with shunt. Multivariate regression analysis was used to help adjust for confounders, including differences in ages and causes of hydrocephalus.26 Once again, we found no convincing evidence of a difference in QoL outcome between ETV and shunt after adjusting for potential confounders. Taken together, the current data provide no compelling evidence that shunting itself leads to worse QoL compared with ETV. IS CPC A GAME CHANGER? The cumulative experience with ETV alone for infant hydrocephalus has been somewhat disappointing, as highlighted in the previous sections. In 2005, Warf2 published his pioneering work on the combined use of ETV and CPC for the treatment of infant hydrocephalus in Uganda. His initial and subsequent publications have shown great promise for CPC in improving the chances of ETV success in these infants.4,27-30 The results for infants in North America, however, are more limited. To date, the largest series has come from Warf’s own experience in Boston.31 Of 91 North American infants treated with ETV-CPC, the success rate was 57%. Some of the initial failures responded to repeat ETV, bringing the overall shunt-free rate to 65%. An important predictor of failed ETV-CPC, however, was age at treatment < 6 months, which, as mentioned previously, accounts for the majority of the infant hydrocephalus population. In this group of infants < 6 months old, the 1-year success rate was < 50%. The HCRN has also investigated the use of ETV-CPC in the North American population in a smaller retrospective review of 36 infants.32 Although there were no major perioperative morbidities or mortalities, the procedure was successful in only 52% at 1 year. This compares with the 65% 1-year success rate of shunts in infants treated contemporaneously within the HCRN. Therefore, although ETV-CPC might hold promise for a subset of infants with hydrocephalus, it is difficult to say at this time that a majority of infants will benefit from this procedure, especially the youngest infants. ARE NEUROSURGEONS UNFAIRLY BIASED AGAINST SHUNT? A final consideration in the ETV vs shunt debate is to acknowledge some inherent biases that we as neurosurgeons might carry. First, the technical aspects of ETV generally provide a greater challenge and are of greater interest to most neurosurgeons; all else being equal, most neurosurgeons would rather spend time in the operating room performing an ETV than placing a shunt. Second, the perception of shunts is often sullied by personal memories of having to deal with difficult shunt cases that required multiple repeat surgeries for recurrent complications, whereas the same is not true for ETV. We must recognize, however, that almost all patients with failed ETV eventually get shunted, so there really is no opportunity for ETVs to fail repeatedly; after the first or, at most, second failure, these patients become shunt patients. Following on this practice, the ETV patients we see in our long-term follow-up clinics are therefore successful cases exclusively. Think for a moment how different our perception of ETV might be if all ETV failures were treated only by having to perform another ETV, without being able to resort to shunt. Our perception of ETV might well suffer as a result. Although these biases almost certainly do not directly affect our decision making for an individual case, they might, at the margins, sway us slightly more to favoring ETV in a way that might not be supported by evidence. CONCLUSION In describing the case for shunt for first treatment of infant hydrocephalus, this article has presented fairly consistent data showing lower failure rates associated with shunt compared with ETV and, in some cases, ETV with CPC. The data also do not show convincing differences in long-term neurocognitive outcome or QoL between shunt and ETV. Future work will likely help us better identify infants who might truly benefit from ETV, with or without CPC, but for now and the foreseeable future, shunt remains the standard treatment for infant hydrocephalus. Disclosure The author has no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Acknowledgment The HCRN data were kindly provided by Drs John Kestle, Jay Riva-Cambrin, and Richard Holubkov on behalf of the Hydrocephalus Clinical Research Network.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.