Introduction

The treatment of posthemorrhagic hydrocephalus (PH) in premature neonates remains a challenge in many aspects. Both the intraventricular hemorrhage (IVH) and a subsequently developing hydrocephalus are likely to harm the normal neurological development of the child. The early treatment of PH is complicated due to the low body weight of the infant and the bloody ventricular cerebrospinal fluid (CSF) with high protein concentration. Despite ongoing global progress in medicine, a significant rate of neonates continues to suffer from mortality and disability after IVH. Aside from the primary brain damage caused by the hemorrhage itself, secondary brain damage is caused by CSF circulation disturbances, elevation of intracranial pressure (ICP), and inflammatory reactions due to resolution of blood degradation products [1,2,3]. According to a recent 10-year American health care database study, 9.3% of neonates with IVH of all grades will develop hydrocephalus. Of those, 38.3% will receive a ventricular peritoneal shunt, while 81.7% of those who receive an initial temporizing neurosurgical treatment for PH will develop shunt dependency [4]. Possible complications with CSF shunts include a significant reoperation rate during life-long follow-up [5,6,7], with neonates with PH being in the group with the highest rate of shunt complications [8,9,10,11,12,13]. The goal remains to enhance the neurodevelopmental outcome by reducing the complications during treatment, to prevent further progression of hydrocephalus, and to reduce the rate of shunt dependency, as well as the rate of surgical revisions linked to the CSF diverting implant [14,15,16].

With the presented collaborative effort, we defined the need for a comparative prospective synchronized data collection for different types of initial neurosurgical treatment for neonates with posthemorrhagic hydrocephalus. Our aim with the TROPHY study is to collect relevant data on safety, efficacy, and long-term neurodevelopmental outcome of four different initial treatment methods, included in a frame of a multicenter international registry.

Methods

International collaboration

Under the umbrella of the International Federation of Neuroendoscopy (previously the International Study Group for Neuroendoscopy), pediatric neurosurgical centers throughout the world have already conducted a number of collaborative studies, creating an established network of collaborating clinicians. The first study dealt with the merits of repeat-endoscopic third ventriculostomy in 20 patients from 4 centers [17]. The second study dealt with endoscopic third ventriculostomy (ETV) following infection and/or hemorrhage in 101 patients from 7 centers [18]. The third study was conducted on safety and accuracy of neuroendoscopic biopsies and included 293 patients in 13 centers [19]. The fourth study was the International Infant Hydrocephalus Study (IIHS), which included 158 patients from 27 centers and compared ETV and shunt for infant hydrocephalus [20, 21]. Uniquely, the IIHS was prospective, which has set the stage for the current proposed prospective registry. We have thus proved our commitment and ability to collaborate and collect relevant data internationally. Such multicenter studies allow an accelerated rate of data collection for rare diseases or surgical procedures. Not only that, but the data provided are more generalizable and universally applicable.

Rationale and concept of the TROPHY registry

The heterogeneity of approaches to surgically treat posthemorrhagic hydrocephalus in neonates and the lack of data for safety and efficacy of the different treatment options warrant a prospective data acquisition on this topic. Different established surgical approaches like external ventricular drainage (EVD), ventricular access device (VAD), and ventriculo-subgaleal shunt (VSGS) have never been compared in a comparative prospective investigation for complication rate and neurological outcome parameters. A meta-analysis investigation from historical cohort studies showed a higher mortality (19.1%) and revision rate (47.3%) for EVDs, however with a diminished shunt rate (68.2%) [22]. A new technique of neuroendoscopic lavage (NEL) appears to be a safe and effective treatment modality to diminish the intraventricular blood and achieve CSF relief as shown in a retrospective single-center study, but requiring further evaluation on a multicenter basis [23]. A more recent bi-center report combined the experience from two centers reporting a mortality rate of 5.4%, a re-bleeding rate with surgical indication of 3.6%, a shunt rate of 57%, and a shunt revision rate of 1.7 ± 1.9 during the first 12 months in a cohort of 56 neonates [24].

Recommendations from the hydrocephalus guidelines are rather limited [15]. This clearly highlights the necessity for a multicenter, prospective registry in order to provide standardized data collection on different treatment modalities for posthemorrhagic hydrocephalus in neonates.

We decided after intense discussion against a randomized study approach due to the following reasons: (1) The IIHS study showed very limited compliance of parents and caretakers to agree on randomization, with parental preference strongly dictating treatment. (2) Some centers with standard protocols are unable to switch their treatment regimen away from a standard procedure, so institutional compliance would also be limited. Thus, it was decided to perform a comprehensive online registry for multicenter prospective data collection in order to analyze immediate and long-term outcome of neonates with IVH and progressive hydrocephalus and to compare NEL with other established treatment options among the participating centers. Outcome parameters include surgical complications, mortality, shunt dependency, ventricular size, and neurological outcome. A synchronized collection of patient data from participating centers will establish an international comparative database.

Results

The TROPHY registry is a password-protected online database, with individual access for the respective individuals of the participating centers. Possible initial neurosurgical techniques are EVD, VAD, VSGS, and NEL, which will be subsequently compared in the analysis of the study data (Fig. 1). The registry is structured to include patient characteristics, preoperative status, imaging, operative techniques, complications, and reoperations, as well as 6, 12, 24, 36, and 60 months follow-up measuring shunt dependency, shunt revision rate, and neurological outcome. The data will be pseudonymized and are entered into the database by investigators of each participating hospital. The centers’ investigators with access to the database are enabled to review only their own data. The principle investigators (administrator) located at the Charité Universitätsmedizin, Berlin, and Universitätsmedizin Göttingen, Germany, will manage the entire dataset and oversee the analyses at the predefined time points for later statistical analysis.

Fig. 1
figure 1

Flowchart of TROPHY registry design

Full size image

Administration

Data recording and collection will be accomplished by a study-database application based on Filemaker® software allowing individualized username- and password-protected access only for registered participants and administrative staff. The central data server is physically located and expertly hosted at the Charité, Berlin, Germany, to provide online access to the study database-application via Filemaker® Server software. Data transfer between the clinical server and the study server is encrypted (SSL coding) to assure data privacy. Access to the online registry application is provided over the study homepage: http://trophy-registry.org (Fig. 2).

Fig. 2
figure 2

Representative screenshots of the online registry (https://trophy-registry.org) referring to the patient list overview and the basic data entry for perioperative clinical condition and surgical technique used

Full size image

All clinical data of affiliated patients are assigned to an automatically generated Case-ID (pseudonym). Clinical data is directly entered online into the study database application by the investigating physicians. Registered study participants can create multiple datasets and complement or update the corresponding clinical data content prospectively at any time by their individual log-in. All documents concerning patients’ treatment-related data are linked to the Case-ID and are recorded in folders of the Electronic Case Report Forms (eCRF).

As soon as a follow-up form has to be filled in, the participating department will be informed with an e-mail and a reminder will be sent after 2 weeks. Issues with data entry compliance and integrity could lead to expulsion of a center from the registry by the steering committee. The online Web-based registry for recruitment and data collection will enable contemporaneous data entry and improved completeness, as the user is obliged to complete required fields before proceeding to the subsequent steps. In addition, it will allow the administrators to have a real-time view of recruited patients; their eligibility and appropriate guidance for the participating centers may be provided through the study at any time.

Inclusion and exclusion characteristics

Surgical interventions eligible for patient inclusion in the TROPHY registry are NEL, EVD, VAD, and VSGS, performed according to standard local procedure of the participating center. The participating centers who offer NEL as a surgical intervention have neuroendoscopic and pediatric neurosurgical expertise with more than 10 neurosurgical procedures in neonates per year per surgeon. Completion of a training workshop on NEL is highly recommended before starting this technique and at least one NEL surgery in neonates per surgeon per year is required. A strong commitment to the study design, consent, follow-up, and smooth data supply and the ability to follow-up for at least 5 years are required for every medical center.

Inclusion in the TROPHY registry study will be achieved by written and signed consent by parents or caretakers to this project.

The clinical condition of the patient must meet the following criteria for inclusion:

  • No previous surgical treatment for hydrocephalus

  • Gestational age ≤ 41 weeks at first neurosurgical intervention

  • IVH documented by ultrasonography

  • Hydrocephalus with progressing ventricular size (US ventricular indices above 97th percentile; according to Levene) diagnosed by preoperative ultrasonography

  • Surgically treated with any of the following: NEL, EVD, VAD, VSGS

  • Ability to follow-up 5 years after initial surgical treatment

The following criteria for exclusion must be absent:

  • Comorbidity-related instability forbidding surgical intervention

  • Diagnosed impaired hemostasis in the most recent laboratory results within up to 5 days before surgery (platelet counts < 50,000/ml; prothrombin time (Quick) > 20 s; PTT > 50 s)

  • Proven CNS tumor or vascular malformation

Outcome parameters

For each patient registered into the TROPHY registry, we are combining a primary factor together with a weighted set of secondary factors to reach meaningful measures of the treatment quality and patient outcome.

The primary outcome for each study participant is based on the neurobehavioral outcome when the patient reaches 5 years of age.

The secondary outcomes include the following factors: number of hydrocephalus-related operations; surgical morbidity; initial intensive care hospitalization time; need for second surgery at follow-up; radiological measures of hydrocephalus; mortality; complications such as CNS infection, secondary hemorrhage, seizures requiring medications.

Neurodevelopmental aspects

The long-term neurodevelopmental effects of the primary damage of hemorrhage and comorbidities, as well as of any surgical intervention, are a major concern for the group of neonates. Comprehensive, standardized evaluation of the participants’ development and abilities over the follow-up course is an essential element of this registry, since the primary criteria for each intervention are based on the neurobehavioral outcome when the patient reaches 5 years of age.

For this reason, a schedule of careful assessments of each child’s development is included in this registry. We have chosen a set of comprehensive, objective evaluation tests that are standardized, with consistent numeric scoring scales, to enable objective, statistically meaningful comparisons and analyses of the results, for a large patient body that is drawn from a broad range of places over an extended period of time. Post-operative, ongoing evaluations are planned to track the patient’s development and progress over time on a regular basis. Testing is scheduled at regular intervals after 24, 36, and 60 months of follow-up.

Developmental assessments are based on the following:

  • At the time of 24 and 36 months of follow-up, the Bayley Scale of Infant Development (BSID) and the Gross Motor Function Classification System (GMFCS) will be used. Due to the extended efforts required of the BSID, which might not be available at all centers, this item will be strongly recommended, but is not defined as mandatory.

  • After 60 months of follow-up, the Gross Motor Function Classification System (GMFCS) will be used, in addition to the Health Utility Index (HUI®) and the Hydrocephalus Outcome Questionnaire (HOQ) for evaluation of the neurological outcome and quality of life.

Statistical considerations

As a prospective registry, data collection will be ongoing, with no prespecified endpoint for data collection. Because it is unclear how many patients with VAD, VSGS, and EVD will be recruited, we are not prespecifying a strictly defined comparative analysis. According to the present information, the following analyses shall be made:

Analysis of complication incidence

Assuming the incidence of each possible complication will not be greater than 0.1 and aiming for an estimate precision of ± 0.05 (95% confidence interval), this analysis will be conducted after 139 patients following NEL have reached 6 months follow-up. An estimate for each of the following complications will be calculated and compared to the other surgical techniques: infection, reoperation, subcutaneous CSF collection, CSF fistula, re-hemorrhage, mortality.

Analysis of shunt dependency

An estimate of the incidence of patients requiring shunt after NEL will be conducted and compared to the other surgical techniques. Assuming a shunt dependency incidence of approximately 0.6 and aiming for an estimate precision of ± 0.1 (95% confidence interval), this analysis will be conducted after 100 patients following NEL have reached 12 months follow-up.

Analysis of long-term developmental outcome

Developmental outcome will be analyzed as follows: Bayley Scale of Infant Development, 3rd Edition, Scale Scorers, at 24 months and 36 months; GMFCS at 24, 36, and 60 months; Health Utilities Index-Mark 3 at 60 months; Hydrocephalus Outcome Questionnaire at 60 months. The aforementioned outcome parameters will be analyzed once at least 100 patients have been assessed for the respective outcome.

Discussion

Neurosurgical treatment of posthemorrhagic hydrocephalus in neonates remains challenging in the context of high comorbidities of these patients with very low body weight and high CSF protein polluted CSF. Thus, initial temporizing neurosurgical treatment is mostly applied before patients are eligible for possible shunting procedures. These procedures include EVD, VAD, and VSGS and are associated with relevant perioperative complications [22]. Until today, no relevant guidelines or any consensus is given in the literature for this rather complex clinical condition [11, 25,26,27,28]. A meta-analysis was able to make the following very limited recommendations. First, clinical judgment is necessary for the decision regarding which temporizing measure is applied among the possibilities of EVD, VAD, VSGS, or even lumbar puncture (LP). The latter one, however, is not considered as a neurosurgical intervention and is not recommended as a serial treatment option especially if non-communicating hydrocephalus is present. Second, VSGS reduces the amount of daily CSF aspirations compared to VAD, whereas VAD reduces the morbidity and mortality compared to EVD. However, all recommendations are linked with only a moderate degree of clinical certainty. Therefore, the need for better evidence in these complex cases remains essential [15].

Recently, some promising results were reported by applying neuroendoscopic lavage in neonates with posthemorrhagic hydrocephalus as the initial neurosurgical treatment option [23]. The data available for this technique is certainly limited since it was collected in only a few centers, but the potential advantages include a possibly reduced rate of shunt dependency, infection, and mortality, as well as a possible reduction in subsequent complications for shunted patients. However, in order to draw a clear conclusion on this technique, a multicenter approach and further clinical evaluation is warranted.

By introducing the multicenter prospective TROPHY registry, we aim to achieve statistically meaningful results for issues that to date have had no objective basis for solid interpretation. The registry is the only comprehensive database to date that focuses on long-term developmental outcomes for neonates with posthemorrhagic hydrocephalus in relation to the applied intervention. As such, the TROPHY registry will serve as an essential tool, providing any means for better parental counseling and for defining standards on the treatment of this challenging patient group. As was already established in previous studies, the registry will further strengthen international collaboration in the field of a rare condition and can settle important debates in a scientific manner. The international framework that we construct with the TROPHY registry will as well serve as a platform for future prospective studies.