Ghobrial, George M. MD; Lang, Michael J. MD
Department of Neurosurgery, Thomas Jefferson University Hospital, Philadelphia, Pennsylvania
Correspondence: George M. Ghobrial, MD, Department of Neurosurgery, Thomas Jefferson University, 909 Walnut Street, 3rd Floor, Philadelphia, PA 19107. E-mail: Georgeghobrial@gmail.com
Received June 05, 2013
Accepted July 19, 2013
I. SIGNIFICANCE/CONTEXT AND IMPORTANCE OF THE STUDY
Idiopathic normal pressure hydrocephalus (INPH) was first defined by Adams et al1 in 1965 and its symptoms later classified by the clinical triad of gait dysfunction, urinary incontinence, and dementia. The exact pathophysiology of this disease is not well understood.2 Surgical options for the treatment of INPH are ventriculoperitoneal shunt (VPS) placement (most commonly with a programmable valve), and endoscopic third ventriculostomy (ETV). VPS is by far the most common method used to treat INPH worldwide. Debate exists as to the superiority of the 2 management options. Historically, VPS placement with a programmable valve has led to improved outcomes in INPH.3 More recent use of ETV has been reported in the form of retrospective data, demonstrating neurological improvement in up to 69% of patients.4 However, a cited limitation of this study is the less stringent diagnostic criteria that fail to discriminate secondary normal pressure hydrocephalus (NPH) from INPH. This is important because of the higher success rates of treatment in secondary NPH.2 This study by Pinto et al should be commended for its attempt to compare ETV with VPS with a nonprogrammable valve for patients with the diagnosis of INPH prospectively. Given that the natural history of VPS carries a significant rate of shunt revision, there have been no previous attempts to provide level I evidence demonstrating equivalence or superiority of ETV over VPS placement.
II. ORIGINALITY OF THE WORK
Literature pertaining to the surgical management of INPH exists largely in the form of retrospective cohort studies. Previous retrospective studies have been previously reported showing that VPS placement is beneficial in 75% of 132 patients at 18 months follow-up with INPH.2 Likewise, there are no level I data pertaining to the use of ETV for the surgical treatment of INPH. There are, however, level I data (randomized, not controlled) comparing various settings of fixed-pressure valves in the setting of INPH. This study is the first randomized clinical trial to compare therapeutic measures for INPH.
III. APPROPRIATENESS OF THE STUDY DESIGN OR EXPERIMENTAL APPROACH
The authors compare the treatment of INPH with a VPS using a nonprogrammable valve (PS Medical, Medtronic, Minneapolis, Minnesota) with an ETV in a randomized, parallel, open-label trial with enrollment in a 1:1 ratio. The selection of a nonprogrammable valve is due to cost and availability. Part of the inclusion criteria involves the tap test (TT) in which a lumbar puncture is performed and 40 mL of cerebrospinal fluid (CSF) is drained. They chose the TT given its widely known validation and ease of use, despite its low sensitivity (<30%). More sensitive methods such as the prolonged lumbar drain (up to 100%) require greater technical expertise. The authors improve on previous methodology of retrospective studies that only use the classic symptoms in better defining INPH for trial inclusion by the addition of radiographic evidence, the TT, and excluding patients with any history of primary dementia, intracranial pathology, or medical comorbidities that may introduce bias into the diagnostic process. These stringent diagnostic criteria aim to limit the variability of results seen in the retrospective literature for surgical treatment of INPH.
The trial, however, is likely severely underpowered to detect a significance between the study populations. The authors calculated the study population size using Altman's nomogram and BERG score outcomes, for which they calculated 22 patients in each trial arm (α = .05) and a power of 80%. The BERG scores from ETV and VPS retrospective trials used for this calculation are not explicitly cited. Recalculating the sample size using binary response rate (the primary outcome of this study being improvement of at least 2 points of the NPH score) yields significantly different results. If the response rates of the cited retrospective studies are used, with a 70% and 80% response rate after ETV and VPS, respectively, a standard difference of 0.23 is calculated, resulting in study arms of 350 patients required to detect significant difference at α = .05 and a power of 80%. This failure to appropriately power the study may have been even greater had the authors explicitly defined the hypothesis. Based on the stated rationale, it can be presumed that this trial is intended to be a noninferiority trial, given the relatively similar response rates between ETV and VPS patients reported in the literature. Using the above-stated response rates, with δ = 0.1, α = .05, and a power of 80%, each study arm would require 878 patients to detect a significant difference.
Finally, failure of the trial to demonstrate equivalence in primary outcome can be anticipated from the primary cited study justifying ETV. A response rate of 69.1% was reported in the 2008 Italian study from Gangemi et al.4 Although not reported directly in that report, post hoc analysis of the results demonstrates a mean improvement in NPH score of 1.34 points with a standard deviation of 1.4. As a result, only roughly one third of the patients in that study would have had a positive response as defined by the author's primary outcome.
IV. ADEQUACY OF EXPERIMENTAL TECHNIQUES
Although many early studies reviewed the efficacy of CSF diversion for INPH, their inclusion criteria are usually limited to ventriculomegaly, dementia, and ataxia. More modern study criteria include clinical and radiographic criteria (Evans index >0.30), as well as lumbar CSF drainage testing in attempts to limit many of the confounding factors that cloud the diagnosis of INPH.
V. SOUNDNESS OF CONCLUSIONS AND INTERPRETATION
The authors conclude that neurological improvement is superior in the VPS group, specifically with gait findings, at 12 months. It is difficult to reconcile that neurological improvement is superior in the VPS group, and it should be stressed that for this population of patients with INPH, conclusions can be made when they correlate with the specific diagnostic criteria in this study.
As previously mentioned, INPH is a disease with multiple confounding factors. Across a number of studies, inclusion and exclusion criteria differ. One of the largest series reported by Vanneste et al5 of 131 patients found only a 31% improvement in symptoms on follow-up, in contrast to the 75% found by McGirt et al.2 These dramatically different results should illustrate to the clinician interpreting this study that the variance in methods across studies matter. For the conclusions of this study to translate over to a particular practice, ideally the diagnostic criteria for INPH should correlate between this trial and neurosurgical practice.
The authors also state that it would be difficult to recommend ETV based on their study. This statement is hard to reconcile given the design of the study. The small patient population, the lack of a control population, and more than half of the patients diagnosed with INPH excluded (n = 48) by their criteria highlight the challenge of this study, along with the trial design limitations discussed earlier. In addition to the large number of patients excluded before randomization, 5 of the 21 patients in the ETV arm were excluded due to anatomy that would add procedural risk. Four additional patients treated with ETV did not improve and underwent VPS placement, whereupon they were removed from the final analysis. The number of patients excluded from this study amounted to almost two thirds of all patients diagnosed with INPH (57/90). Therefore, the authors rightly conclude that future multicenter studies with larger patient populations are needed.
VI. RELEVANCE OF DISCUSSION
The authors begin the discussion with its limitations. They emphasize that diagnostic criteria are important in defining INPH, a disease with symptoms that cannot be completely controlled for. They draw on all of the previous literature and provide a more comprehensive diagnostic criteria for the study. As a result, they exclude a large number of patients (45/90). The authors use the TT as an inclusionary criterion for the study, despite its reportedly low sensitivity (26%-61%). They also mention the alternatives in diagnostic criteria. Another key point raised is the type of VPS used, which is a fixed-pressure valve in this study. Programmable valves are commonly used in the United States, often in conjunction with antisiphon devices that prevent overdrainage. Minor adjustments in these valves have been shown in retrospective studies to improve overall outcomes in NPH.3 The rationale for this practice in the United States is illustrated by the 20% rate of subdural hematoma, all of which required a surgical evacuation.
VII. CLARITY OF WRITING, STRENGTH, AND ORGANIZATION OF THE PAPER
The strength of this paper is derived from this novel attempt to compare 2 established treatments for INPH. It is clearly written and easily understandable. The discussion is well organized, beginning with the limitations of this study, which comprises roughly half of the overall discussion. The discussion then follows a logical progression of generalizability of the study and then interpretation.
VIII. ECONOMY OF WORDS
This paper is concise and well organized. Given the exceedingly low number of clinical series on the surgical treatment of INPH, a table listing the previous contributions would be helpful to the reader, broken down into 2 sections, ETV and VPS.
IX. RELEVANCE, ACCURACY, AND COMPLETENESS OF BIBLIOGRAPHY
This paper is well referenced and includes all of the clinical cohort studies on surgical treatment of INPH. The first few references also include references to the original works by Hakim and Adams that discussed the early clinical symptoms of INPH.
X. NUMBER AND QUALITY OF FIGURES, TABLES, AND ILLUSTRATIONS
Figure 1 illustrates nicely the flow of enrollment in this study, illustrating the high number of patients excluded. Table 1 summarizes the score at various time intervals for all of the clinical outcomes measured. The scoring systems for the clinical outcomes are listed separately in the methodology. The scales are not organized alike in that higher scores for all groups do not interpret as improvement or decline. It would be easier to read if the table had a legend that correlated increased numerical score with clinical improvement or decline. Figure 2 shows a typical patient who underwent an ETV, pre- and post-intervention. This does not add much to the understanding of the study. It may be more useful to show sagittal magnetic resonance imaging of the patients who were excluded to give the author an idea of what was deemed too unsafe to perform an ETV. Excluding nearly 20% of potential candidates based on anatomic variance alone is a high number and would be of interest to the reader. Likewise, for a typical VPS patient, a pre- and postoperative computed tomography of the head was included to demonstrate what the authors considered an adequate cathether tip position. This does not add much clarification to the methodology and probably can be removed from the study.
XI. FUTURE/NEXT STEPS THE PAPER LOGICALLY LEADS TO
Despite being clinically characterized in 1965, INPH, to date, does not have a clear pathophysiology nor has there been a demonstration in the literature of a superior form of surgical management. The authors of this study lay the groundwork for comparing surgical treatments for INPH. Future resources should be devoted to multicenter, randomized, controlled trials (RCTs) with larger patient populations. Given the considerable variation that is encountered between studies regarding diagnostic criteria for INPH, future studies should aim at standardizing this diagnostic process to limit confounding variables due to disease overlap. All steps must be taken to maximize the relevance of future RCTs, which would especially be the use of programmable valves with siphon guards. Follow-up times should exceed 1 year to gain a better sense of the need for reoperation rate, especially in the ETV group.
1. Adams RD, Fisher CM, Hakim S, Ojemann RG, Sweet WH. Symptomatic occult hydrocephalus with “Normal” cerebrospinal-fluid pressure. A treatable syndrome. N Engl J Med. 1965;273:117–126.
2. McGirt MJ, Woodworth G, Coon AL, Thomas G, Williams MA, Rigamonti D. Diagnosis, treatment, and analysis of long-term outcomes in idiopathic normal-pressure hydrocephalus. Neurosurgery. 2005;57(4):699–705; discussion 699-705.
3. Zemack G, Romner B. Adjustable valves in normal-pressure hydrocephalus: a retrospective study of 218 patients. Neurosurgery. 2002;51(6):1392–1400; discussion 1400-1392.
4. Gangemi M, Maiuri F, Naddeo M, et al.. Endoscopic third ventriculostomy in idiopathic normal pressure hydrocephalus: an Italian multicenter study. Neurosurgery. 2008;63(1):62–67; discussion 67-69.
5. Vanneste J, Augustijn P, Dirven C, Tan WF, Goedhart ZD. Shunting normal-pressure hydrocephalus: do the benefits outweigh the risks? A multicenter study and literature review. Neurology. 1992;42(1):54–59.