Functional MRI (fMRI) as a Replacement for the Wada Test in Evaluating Memory Function?

Abstracts & Commentary

Synopsis: Although further technical improvements and prospective clinical validation are required, these results suggest that mesial temporal memory activation, detected by fMRI during complex visual scene encoding, correlates with post-surgical memory outcome, and supports the notion that this approach will ultimately contribute to patient management.

Source: Rabin ML, et al. Functional MRI Predicts Post- Surgical Memory Following Temporal Lobectomy. Brain. 2004;127:2286-2298.

The intracarotid amobarital (or Wada) test has been used for over half a century to pre-operatively determine language laterality and to assess the risk of amnesia in pharmacologically refractory epilepsy patients being evaluated for temporal lobectomy (TLX). Ideally, injection of the ipsilateral carotid artery with amobarabital produces no memory deficit, while contralateral injection induces severe (but temporary) amnesia. Such a highly lateralizing Wada indicates that memory is subserved solely by the non-epileptogenic temporal lobe, and the patient is not at significant risk for developing a global memory deficit. Memory function is also modality-specific for either verbal or non-verbal memory. The neuropsychological testing battery administered during the Wada test uses different tasks to subdivide memory functions to predict whether removal of mesial temporal structures may be contraindicated because of significant verbal or non-verbal amnesia complicating TLX.

Rabin and colleagues provide important data to suggest that fMRI testing of visual memory correlates well with the results of Wada testing. Moreover, fMRI predicted post-operative visual memory deficit in patients going on to TLX (the true gold standard for both fMRI and Wada). The study enrolled 35 patients undergoing pre-surgical evaluation for epilepsy and 30 normals. All subjects underwent both fMRI and Wada during a complex scene encoding task. Encoding performance was assessed by follow-up recognition testing. Twenty-three patients who subsequently underwent TLX, completed the same task outside the scanner an average of 6.9 months post-operatively.

A region of interest (ROI) analysis was performed to quantify hippocampal (H) activation alone, with a separate ROI including hippocampus, parahippocampus, and fusiform gyrus (HPF). Asymmetry ratios (AR) were calculated for activation for each ROI as ipsilateral minus contralateral divided by ipsilateral plus contralateral. Normal subjects showed symmetrical AR for fMRI (mean = 0.02 for hippocampus and < 0.007 for HPF). For patients, AR correlated with Wada for HPF but not the smaller H ROI. In addition, ipsilateral HPF AR showed a significant inverse correlation with good post-TLX memory outcome (ie, the lower ipsilateral fMRI absolute activation, the better the memory outcome). Surprisingly, a statistically significant correlation did not exist between contralateral fMRI AR and post-operative memory outcome. Finally, AR differences in either H or HPF ROI activation did not show the same correlation between seizure outcome and Wada memory results as previously reported (Neurology. 1994;44:2322-2324 and Epilepsia. 1995;36:851-856).

Commentary

Neurology Alert has previously (October 2002, p. 15 and August 2003, p. 92) commented upon the use of fMRI as a replacement for the Wada test and deficiencies related to inadequate memory testing paradigms as the reason why fMRI was not ready for prime time in this application. Gaillard and colleagues (Neurology. 2002;59:256-265) addressed the validity of fMRI for lateralizing language. Sabsevitz and colleagues (Neurology. 2003;60:1788-1792) looked at fMRI in conjunction with the Boston naming test to show that fMRI compares favorably with Wada in predicting verbal naming deficits following TLX. The current study by a collaborative group from the Thomas Jefferson University and the University of Pennsylvania convincingly demonstrates that fMRI can also predict post-TLX visual memory deficit in medically refractory epilepsy patients.

Over just a few years, it appears that fMRI has advanced almost to the point of a clinically relevant tool. Two key questions remain unanswered. First, can all 3 neuropsychological paradigms provide the same information in a single study session, as each has reported separately? There is no a prior reason to think otherwise. Second, can a combined verbal and nonverbal memory fMRI yield the same positive predictive information about seizure control following TLX as a highly lateralized Wada test can? The only data thus far available to answer this question are from Rabin et al, and these are unfortunately negative. —Andy Dean

Andy Dean, MD Assistant Professor of Neurology and Neuroscience; Director of the Epilepsy Monitoring Unit, Department of Neurology, New York Presbyterian Hospital Cornell Campus is Assistant Editor of Neurology Alert.