Setting the Gold Standard

Sallie A Baxendale, Neuropsychologist,

Other Contributors:

November 11, 2013

Dear Sirs,

We read the above study with interest and appreciated the thorough analysis. Whilst the authors acknowledge in their discussion that the Wada test is a 'silver standard', they nevertheless conclude that the Wada test is warranted when fMRI fails to show clear left lateralisation. We would make the following points:

1. The true "gold standard" for language dominance tests is prediction of outcome. The authors acknowledge this in their discussion.

2. Recent studies suggest that high-quality fMRI is somewhat more accurate than the Wada test in predicting postoperative language outcomes. Sabsevitz et al showed better prediction of naming outcome by fMRI than by Wada in an unselected sample of left anterior temporal lobectomy patients. Janecek et al demonstrated that in patients with discordant fMRI and Wada, fMRI is generally more accurate at predicting language outcomes than the Wada test. Thus it is erroneous to consider fMRI results as inaccurate simply because the technique picks up more activation in the right hemisphere. It turns out that this "non-essential activation" probably helps patients recover.

3. For these reasons, it makes little sense to use fMRI as a screening tool and Wada as the standard in patients with atypical language representation.

4. The conclusion that the Wada test is warranted cannot be justified using this methodology employing a suboptimal clinical standard.

References:

Sabsevitz DS, Swanson SJ, Hammeke TA, Spanaki MV, Possing ET, Morris GL, Mueller WM, Binder JR. Use of preoperative functional neuroimaging to predict language deficits from epilepsy surgery. Neurology, 2003; 60: 1788 -1792.

Janecek JK, Swanson SJ, Sabsevitz DS, Hammeke TA, Raghavan M, Mueller W, Binder JR. Naming outcome prediction in patients with discordant Wada and fMRI language lateralization. Epilepsy & Behavior, 2013; 27: 399- 403.

Conflict of Interest:

None declared

Conflict of Interest

None declared