Giant Intracranial Aneurysm Registry

Definition of aims, eligibility of patients and data flow

In August 2009 a steering committee was appointed and its first meeting took place in Berlin. Here the aims of the registry were defined as follows: 1) To generate detailed insight into which therapies are being conducted. 2) To document the natural history of patients and follow it over years after diagnosis. 3) To follow imaging data of giant aneurysms over years after diagnosis, whether treated conservatively or not. 4) To expand the register into several disciplines (by inviting neuroradiologic and neurologic centers to join) and internationally.

Inclusion criteria are a) the diagnosis of a giant intracranial aneurysm by CT, MRI or regular angiography, b) age 18 years and older and c) informed consent. There are no exclusion criteria. If a patient is not able to give informed consent due to his or her clinical situation, the legal representative of the patient (e.g. family member) will be informed and asked to sign the consent form on behalf of the patient.

Patients are recruited consecutively in responsibility of each center. In addition to the prospective part there is a retrospective part in which patients are included that were admitted after January 2006. Data are gathered at admission, discharge and 1/3/5 years after discharge. Starting from February 2010 a 9-month pilot phase was conducted in order to identify and correct possible problems of data acquisition.

Development of a study protocol

Relevant variables were discussed on the basis of existing literature on giant intracranial aneurysms. All variables were chosen so that they can be determined in a routine clinical context even in small centers with minimal technical equipment. Data collection was decided to be carried out using a modular structure, consisting of a basic module and additional modules for specific research questions.

The steering committee decided on the content of a mandatory basic data collection for all centers. This basic module was designed to keep the amount of variables at a minimum so that data collection can be conducted as safely and quickly as possible by physicians of all disciplines. It focuses on the clinical course between initial presentation and discharge but also includes data on patient history, such as risk factors and possible prior aneurysm therapy. A neurological examination of cranial nerve functions, motor capability and speech is conducted initially, at discharge and during 3 mandatory follow-up examinations. At each examination the patients' physical condition is recorded using internationally accepted and reliable grading systems. The investigator is asked to not only record a clinical diagnosis, such as epileptic seizure or incidental finding, but also use imaging data to produce a radiological diagnosis and characterize the aneurysm regarding its location, size, shape and structure. The mode of therapy is documented choosing between conservative, surgical and endovascular options and combinations thereof. Possible complications are recorded as well as radiologic results of intervention.

Each center is given the option to take part in additional extra modules, which serve as a basis for spin-off projects resulting from the discussion within the study group. This comprises for example an imaging module to lead to a better understanding of how giant intracranial aneurysms develop over time. It will monitor changes in aneurysm configuration and flow dynamics using 3D-imaging data. A cost module will integrate data on treatment expenses making it possible to compare the costs of each mode of therapy.

An external review of the study protocol was conducted to ensure high methodological quality and wide acceptance of the results. For this purpose a preliminary working report was forwarded to 4 independent experts in the field of neurovascular research. The reviewers were asked to comment on the appropriateness of the methods, the practicability of our variables and the coverage of relevant aspects of giant intracranial aneurysm care. The comments of the external reviewers were gathered and discussed within the steering committee. Based on the recommendations of the external reviewers, a final revision of the study protocol took place.

Primary hypothesis, study outcome and statistical analyses

The primary hypothesis of the study is that conservative, endovascular and surgical therapy result in equal aneurysm rupture rates. Therefore primary outcome will be aneurysm rupture rate at 5 years of follow up.

The sample size calculation is based on the estimated 5-year cumulative rupture rates for patients with unruptured giant aneurysm given in the ISUIA-Trial. According to their results we assume a 5-year rupture rate of 0.30 irrespective of which specific treatment is applied. Because of a lack of data on special rupture rates regarding patients treated with the three considered methods, sample size calculation was done for one proportion based on precision analysis. Consequently, a two-sided 95% confidence interval with a width equal to 0.10 is achieved by a sample size of 320 patients. For the outcomes death or rupture the corresponding survival functions with 95% confidence intervals will be estimated by the Kaplan-Meier method. The null-hypotheses of equal survival curves for two treatment groups will be tested by the Log-Rank test. With documented dates of rupture events the sample size of 320 calculated above is sufficient to detect moderate effects of possible predictors for rupture by Cox proportional hazards regression within a secondary analysis. Furthermore we can model the logit of rupture after 5 years by binary logistic regression detecting possible moderate to considerable effects of predictors. Therapy (conventional, surgery or endovascular) will be the main factor of interest for both multivariable analyses.

Patient recruitment / Data storage / Quality assurance

In 2008 patient recruitment began at the Charité Berlin, one year later data from other centers were included as well, initially only in Germany, later on throughout Europe and Japan. Patient data are stored using pseudonyms. Centers can either use a fax-based data inclusion form or a web-based entry form. All data are stored at the Charité - Berlin. For quality assurance a monitoring visit to the 7 most-including centers was conducted in 2013.