OBJECTIVEThe gold standard for documentation of surgical cure of a brain arteriovenous malformation (AVM) is a postoperative angiogram. Intraoperative angiography also has been used for assessing surgical obliteration of AVMs. The objective of this work is to determine the incidence of unexpected residual AVM in patients undergoing intraoperative angiography after brain AVM surgery, the incidence of false-negative intraoperative angiography, and whether there are any identifiable factors that would predict such an occurrence.METHODSPatient age and sex, AVM location and size, clinical presentation of the AVM, day of surgery after hemorrhage, whether embolization was performed preoperatively, presence of intraoperative brain swelling or substantial bleeding, and postoperative course were recorded prospectively on 34 consecutive patients who underwent surgery for brain AVMs. Intraoperative angiography was performed after the surgeon thought that the AVM was completely obliterated. The incidence of unexpected residual AVM and false-negative intraoperative angiography was determined. Factors predicting these findings were identified by multivariate analysis.RESULTSTwenty-five of 34 patients underwent intraoperative angiography to assess the extent of resection, and two patients underwent the examination to localize the AVM. Postoperative angiograms were obtained for 26 patients. Intraoperative angiography showed unexpected residual AVM in 2 (8%) of 25 patients. In two patients, intraoperative angiography was useful to locate a small AVM in the wall of a hematoma cavity. Three patients (18%) whose intraoperative angiograms had not shown AVM had postoperative angiograms that showed residual or recurrent AVM. One (11%) of nine patients who had only postoperative angiography had an unexpected residual nidus; the patient underwent a reoperation and successful resection. There were no significant clinical or radiological features that predicted the intraoperative angiographic finding of residual AVM or of false-negative intraoperative angiogram.CONCLUSIONIntraoperative angiography is useful to demonstrate residual AVM in about 8% of patients undergoing AVM resection. It can be used to localize small AVMs, but other methods for localization may be as useful and may avoid the risks and cost of additional angiography. Intraoperative angiography does not replace postoperative angiography to confirm AVM removal because of false-negative findings, which occurred in 18% of patients in this series.