The New York Islands Arteriovenous Malformation Study

Procedure Manual

1 Introduction

For too long there has been uncertainty about the best treatment of cerebral arteriovenous malformations (AVMs). Until recently there has been insufficient data to estimate the natural course risk of AVM hemorrhage and recent work has shown that the risk of invasive AVM treatment is largely unknown as well. Therefore, for many patients with AVM it remains unclear whether conservative management or invasive intervention is the best approach. Disregarding the issue of an unclear risk/benefit balance of invasive AVM management, the high costs of serious disability caused by intracranial hemorrhage are reason enough for interest in AVM therapy.

In proposing this study, we take the position that insufficient data exist as to the incidence, prevalence, and neurologic sequelae of brain AVMs. All current evidence is burdened by retrospective data acquisition and the restriction to - often small and highly selected - samples from tertiary referral centers. No prospective population-based study of AVM has so far been undertaken, raising concerns about the applicability of results from tertiary centers to the total population of persons carrying an AVM.

1.2 Aim

This investigator-initiated, population-based study is designed to assess the incidence, prevalence, and neurologic sequelae of cerebral AVMs in the geographical area of the New York Islands (i.e. Manhattan island, Staten Island, and Long Island including the New York City boroughs Queens and Brooklyn, as well as Suffolk and Nassau County).

The study will be divided into two phases:
• Phase 1 will (1) assess the total number of cerebral AVMs occurring in 1998 and thereafter, using neuroradiological criteria for mode of presentation (e.g., hemorrhage, infarction, no signs of hemorrhage but signs of AVM). The purpose of phase 1 is to make a count of cases and (2) to standardize criteria for the neuroradiologic detection of AVMs.
• Phase 2 will occur after phase 1 has been completed. In addition to continuing the enrollment of AVMs based on neuroradiologic criteria, it will expand the focus of the project to (1) assess the type of initial AVM presentation (i.e. hemorrhage, seizure, migrainous and non-specific headaches, focal neurologic deficits, other types of presentation, and asymptomatic AVM), (2) assess the neurologic deficit, impairment, disability, and death rate, and (3) investigate the outcome during natural course follow-up and for any type of treatment.

1.3 Research Plan

Patients with the clinical diagnosis of cerebral AVM - supported by brain imaging and/or cerebral angiography - will be eligible to be included into the study. Phase 1 will establish the detection network in the geographical area of the New York islands, in the first year of the study (1998) only baseline neuroradiologic diagnostic data of patients with AVM will be recorded. Phase 2 will occur in the following years (1999 and thereafter) in which patients will be followed to detect primary endpoints (spontaneous intracranial hemorrhage, seizure, focal neurologic deficit, and treatment outcome).

Those eligible for Phase 1 and Phase 2:
• are age 15 or older
• have undergone brain imaging or cerebral angiography prior to inclusion showing an AVM rated "definite" or "possible" (the latter will be subject to an expert panel adjudication).

Those ineligible have:
• A cerebral AVM diagnosed before January 1st, 1998
• No evidence of a cerebral AVM
• A dural AVM
• A vein of Galen malformation
• An extracerebral (including spinal) AVM

Those enrolled in phase 2 who suffer neurological events - either during the natural course or treatment related - will prompt notification of the Data Management Center (DMC), clinical evaluation; and collection of relevant data. For phase 2, an adjudication panel will judge the occurrence of any relevant clinical end point, including stroke severity, death, treatment complications, CT and/or MRI findings. A Performance Board will meet at intervals and receive status reports from the investigators and the DMC.

Study Timetable

1998 1999 2000 2001 2002

1998 |--------------| 1999; Geographical area AVM incidence detection system organized

1999 |----------------------------------------| 2002; Incidence detection and follow-up of patients

2 Organization

2.1 Coordinating Center

The Coordinating Center (CC) for The New York Islands Arteriovenous Malformation Study is located at the Stroke Service of The Neurological Institute of the Columbia-Presbyterian Medical Center. The CC is responsible for all aspects of the study. Questions concerning patient enrollment, diagnosis, termination should be directed to the CC.

2.2 Data Management Center (DMC)

The DMC is also located at the Stroke Service. The DMC staff includes Quality Control Coordinator, Data Manager, Site Monitor, and Computer Maintenance Consultant, who are responsible for data management and analysis during the trial. Questions concerning use of computers, data entry, and completing case report forms should be directed to the DMC. In keeping with an efficient design of this study, the DMC is configured to provide timely and efficient handling and quality control of data at a minimum of cost.

Operations of the DMC will be supported by appropriate computers for the daily project administration and quality control.

2.3 Current Consortium of Participating Centers

Site Name	Local Investigator
Beth Israel Medical Center	A Berenstein, MD
Columbia-Presbyterian Medical Center	JP Mohr, MD
Lenox Hill Hospital	LM Rothman, MD
Long Island Jewish Medical Center	R Libman, MD
Mount Sinai School of Medicine	BP Drayer, MD
New York University - New York VA Medical Center	K Siller, MD
North Shore University Hospital	J Halperin, MD
NYU Medical Center	P Kim Nelson, MD
St. Johns Episcopal Hospital, South Shore	D Dorfman, MD
St. Mary's Hospital of Brooklyn	F Carillo, MD
The Brooklyn Hospital Center	BS Leitman, MD

2.4 Operations Committee

The Operations Committee meets weekly to review issues arising from daily implementation of the project. These include the status of data entry, frequency of deviations from the protocol, and completeness of the data set from the individual institutions.

2.5 Adjudication Committee

For phase 1, a panel of senior neuroradiologists will adjudicate patient enrollment conformation with the study protocol and judge on cases rated "possible AVM" by the local centers.

For phase 2, a panel of senior neuroradiologists, neurosurgeons, and neurologists will adjudicate adjudicate the endpoint for each of the patients suffering an event. The review will utilize the information available from the initial and subsequent clinical and laboratory data.

3 Enrollment

Each patient arriving at a study center with a cerebral AVM (diagnosed 1998 or later) will be enrolled. Patients with the diagnosis "possible" AVM will be evaluated by the adjudication committee and enrolled if adjudication reaches a "definite" AVM decision. The clinical and neuroradiologic evaluation must be sufficient to establish the diagnosis of AVM. Clinical information documenting the medical and neurological history, hospital course, neurologic examination, general physical examination, results of diagnostic evaluations including cut film cerebral angiography, extracranial and transcranial Doppler, CT or MRI, and EEG will be used to determine the clinical presentation, morphology, and hemodynamic features. The diagnostic information for the presence of an AVM which qualifies the patient for enrollment must be available before enrollment.

The number of AVM patients who are treated outside the hospital system is most likely small (the gold standard diagnostic procedure of cerebral angiography and most invasive treatment procedures require hospital admission). Nevertheless, a catchment system for ambulatory AVM cases will be developed drawing from experienced of the Northern Manhattan Stroke Study.

4 Data Transfer and Transmission

Data will be recorded and transfered in two ways, (1) on hard-copy report forms faxed to the DMC after completion or (2) via Internet into an interactive website. All data will be stored in hardcopy version and on stand-alone personal computers to insure data safety and confidentiality. No personal patient identifyers will be transmitted to the DMC.

5 Consent and Compliance

All patients enrolled in 1999 and later will be asked for written permission to be followed within the study. Some may refuse to participate in follow-up procedures, as is their right. For such patients, only the status at the point of initial evaluation will be recorded.

6 Follow-Up

The DMC will be in contact with the patient by telephone as required to verify the status of each individual participant. As of 1999, in phase 2, the patient will make regular personal contact with the local clinical investigator for clinical evaluation. At this visit, questions will be asked concerning neurologic events and the neurologic status will be assessed by physical examination. At the end of five consecutive years of observation, each patient will be discontinued from study participation.