The role of the Wada test in the surgical treatment of temporal lobe epilepsy: An international survey


The Wada test, (intraarterial amobarbital procedure, IAP) has been used to determine language dominance and assess the risk of a postoperative amnesic syndrome in temporal lobe epilepsy (TLE) surgery candidates since its inception in the mid twentieth century (Wada J, 1949). In 1993, over 95% of epilepsy surgery centers worldwide reported using the IAP, with 85% conducting the procedure on all surgical candidates, primarily to assess language dominance and memory capacity (Rausch et al., 1993). Scores from the IAP were subsequently used as an adjunct in lateralizing the seizure focus (Davies et al., 1999; Kanemoto et al., 1999; Swearer et al., 1999; Alpherts et al., 2000; Akanuma et al., 2003) and to predict changes in postoperative memory function (Bell et al., 2000; Chiaravalloti & Glosser, 2001; Sabsevitz et al., 2001; Stroup et al., 2003; Lacruz et al., 2004; Lee et al., 2005; Andelman et al., 2006).

However, the IAP is an invasive procedure with small, but serious risks of significant morbidity (Rausch et al., 1993; Loddenkemper et al., 2002). Morbidity risks are increased in selective procedures. Procedural failures, including somnolence can invalidate test scores and the accurate assessment of memory function can be difficult following perfusion of the language dominant hemisphere or perfusion of anterior regions leading to perseverative responses and/or behavioral disinhibition (Baxendale et al., 1996; Hamer et al., 2000; Masia et al., 2000; Bengner et al., 2003; de Paola et al., 2004). Topiramate, an antiepileptic drug, may also interfere with the validity of the procedure (McCabe & Eslinger, 2000; Kipervasser et al., 2004). Adverse reactions may be more common in procedures employing newer compounds such as etomidate, propofol, or brevital rather than amobarbital (Grote & Meador, 2005; Mikuni et al., 2005).

High-definition structural MRIs enable the detailed examination of ipsilateral and contralateral structures in presurgical candidates (Duncan, 2007). Some centers now use data from these images, combined with detailed baseline neuropsychological assessments to assess the structure and the functional capacity of the contralateral mesial temporal lobe in surgical candidates, in order to assess the risk of a postoperative amnesic syndrome and to predict postoperative changes in memory function (Stroup et al., 2003; Baxendale et al., 2006). Functional imaging paradigms are also beginning to show promise in this regard (Richardson et al., 2004; Janszky et al., 2005). The potential of functional MRI (fMRI) to lateralize language functions prior to epilepsy surgery is becoming established. Large series of patients have undergone language evaluation with both IAP and fMRI and the latter is rapidly making the transition from an experimental technique to a clinical tool (Baxendale, 2002; Powell & Duncan, 2005).

In the light of these developments and increasing divergence in the literature regarding the future role of the IAP and its counterparts in the presurgical evaluation of TLE candidates (van Emde, 1999; Valton & Mascott, 2004; Grote & Meador, 2005; Jones-Gotman et al., 2005), we initiated an international survey of current practice to determine the prevalence and rationale behind the employment of the IAP in the presurgical evaluation of TLE patients across the world at the present time.


Centers offering elective epilepsy surgery were identified via a search of the literature, January 2000 to June 2007, using the key words “temporal lobe epilepsy surgery.” A total of 1,139 papers were found, originating from 40 countries. Potential centers associated with epilepsy surgery were identified from the corresponding authors' affiliations. Since a single epilepsy surgery service is often linked to multiple academic and clinical institutions, potential authors were cross-referenced to try to ensure that only one clinician was invited to complete the survey for each epilepsy surgery centre.

Clinicians from 207 centers were invited via email to take part in the Web-based survey. Invitations to participate in the survey were returned from 18 centers due to invalid email addresses. An email contact could not be traced for four institutions. The survey link remained open for 1 month after the initial invitation and an email reminder was sent in the week prior to the close of the survey. Since named individuals working within each centre were identified from the literature and emailed individually the chances of “spoof” responses contaminating the survey are extremely low.

Survey Design

The survey was designed to be as brief as possible to maximize the response rate. In the context of the survey the phrase “Wada test” was used to refer to all intraarterial procedures, involving standard carotid and more selective injections with amobarbital, propofol, brevital, or etomidate in order to assess cognitive functions as part of a presurgical evaluation. After completing details about their epilepsy surgery centre, clinicians were asked to rate their confidence in allowing patients with TLE to proceed to surgery without data from the IAP with respect to four functions (see Table 2).

Table 2.  Survey responses
inline image
  • 1Lateralizing the seizure focus
  • 2Screening for amnesic risk
  • 3Predicting postoperative memory decline
  • 4Determining language lateralization


We received 92 survey responses, from clinicians in 31 countries, giving an overall response rate of 40%. Response rates varied by region; see Fig. 1. Three responses were incomplete, accounting for the variations in the total number of responses for each question.

Figure 1.

Responses to “How often do you perform a Wada procedure in the presurgical evaluation of temporal lobe epilepsy patients?” by geographical region.

Prevalence of the Wada test in the presurgical evaluation of temporal lobe epilepsy patients (See Table 1)

Table 1.  Prevalence of the IAP in the surgical evaluation of TLE patients
How often do you perform a Wada procedure in the presurgical evaluation of temporal lobe epilepsy patients?
inline image

The frequency of employment of the Wada test in the presurgical evaluation of temporal lobe epilepsy patients was not related to the type of service offered (pediatric vs. adult vs. both) or the annual number of operations conducted by the centre (less than 10 vs. 11–30 vs. more than 30).

However, there were significant regional variations in the employment of the Wada test (χ2= 28.7; df = 5; p < 0.001) with the Wada test being employed more frequently in more centers from North America, than in Europe; see Fig. 1. A total of 48% of the respondents in North America reported that 75% or more of their patients underwent a Wada test as part of their preoperative evaluation. None had ceased to use the procedure on all patients. In Europe, 48% of the respondents reported that they never or very rarely used the Wada test in the presurgical evaluation of TLE patients. Whilst the numbers of respondents in the other regions were too small for statistical analysis, the descriptive data also suggest a move away from the Wada procedure in Australia and Central and South America.

Reliance on the Wada test for lateralization of the seizure focus

The majority of clinicians indicated that they do not rely on the Wada test to lateralize the seizure focus, with 89% of respondents reporting that they would feel confident localizing the seizure focus without Wada data in all or most cases (Table 2). Although there was a tendency for centers doing less than 10 operations annually to report more reliance on the Wada test for lateralizing seizure focus, this trend was not statistically significant (χ2= 0.2; df = 2; p > 0.05). Fewer exclusively pediatric services relied on the Wada for seizure lateralization, than adult and mixed services, but again this difference was not statistically significant (χ2= 3.5; df = 2; p > 0.05).

One clinician wrote:

“My understanding is that the Wada provides some localizing value, but is not tended to be a strong localizer. We use Wada data as a warning, if there is a strong asymmetry in the wrong direction, prompting perhaps more extensive investigations or a fresh look at the existing findings.”

Reliance on the Wada test in the assessment of amnesic risk

Just over half of the respondents (53%) indicated that they would feel confident in assessing the risk of a postoperative amnesic syndrome without Wada data in most or all of their cases (Table 2). Confidence in predicting amnesic risk without Wada test data was not related to the annual number of cases seen or the patient group (pediatric vs. adult vs. both).

Reliance on the Wada test for prediction of postoperative memory change

More than half of the respondents (56%) indicated that they would feel confident in predicting postoperative memory change without Wada data in most or all of their cases (Table 2). Confidence in postoperative memory change without Wada test data was not related to the annual number of cases seen or the patient group (pediatric vs. adult vs. both).

Language lateralization

In the majority of centers (86%), resections on the language dominant side are less extensive than those on the nondominant side. A total of 14% of the centers reported that the extent of resection was standardized regardless of language dominance. Most of these respondents indicated that they typically conducted a selective amygdalohippocampectomy, rather than a standard anterior temporal lobe resection.

Two-thirds of the respondents (66%) indicated that they would feel confident in proceeding to surgery without Wada data to lateralize language function in most or all of their cases (Table 2). Not surprisingly, confidence in proceeding to surgery with Wada test data to lateralize language function was higher in the centers who did not tailor the extent of the resection on the basis of language dominance (Mann–Whitney U= 346.0, p < 0.05). As with the previous roles, confidence ratings in proceeding to surgery without Wada test data to lateralize language function was not related to the annual number of cases seen or the patient group (pediatric vs. adult vs. both). One clinician reported that they performed a selective amygdalohippocampectomy without a Wada test in most patients, but in MRI-negative cases they performed a Wada-Test in order to guide the extent of resection. fMRI is not available as a clinical tool for language lateralization in all epilepsy surgery centers and confidence in the procedure is not universal. One clinician reported:

“We have very poor fMRI capacities at our facility at present. If these were clearly established here and reliable, I obviously would feel more confident going with these alone (although I have never been convinced that the Wada and functional imaging are measuring the same things, as one is an inactivation study and the other is not.”

Additional Information

A total of 45 of the respondents returned additional comments on their experience with the Wada test. Of those who conducted a Wada test on all TLE cases, one commented: “This is very important issue to discuss and further developments are needed. You did not ask whether surgery is refused on the basis of Wada; we still have 1 or 2 of these cases a year and I think this is the real measure of our reliance on the Wada.” Another commented:

“Despite our widespread use of the Wada test, and my belief that if the patient passes the risk of severe post-operative amnesia is low, it is unclear to our group how to use the results to counsel patients as to possible subtle but functionally meaningful memory problems.”

Practical, clinical and theoretical reasons for not employing the Wada test were given. A number of centers in Europe and South America reported difficulties in obtaining amobarbital. Others were concerned with the risk/benefit ratio of the procedure.

“We performed this procedure until 2003 (more than 150 patients). But the last patient presented a vascular accident due to catheterization. We then reanalyzed all previous decisions and it appeared that the Wada test had never influenced our decision.”

Other centers reported a gradual move toward letting clear-cut cases through to surgery without the Wada test. One clinician commented:

“Many of these responses reflect relatively recent changes in practice (within the past year); e.g., we previously performed Wada's on all surgical candidates but now typically do not if patient has unilateral MTS on MRI and all data are consistent with that.”

In centers who did not conduct Wada tests on everyone preoperatively, Wada data was felt to be valuable for some patients. One clinician reported:

“Rarely we will perform a Wada for language lateralization to assist in the interpretation of the preoperative neuropsychology in left-handed patients with nonconcordant preoperative data and if fMRI has been inconclusive.” And another that “There is a different situation in patients with bilateral temporal lesions and/or bilateral EEG spikes more than 80:20% and/or bilateral memory difficulties according to neuropsychological testing. In These cases we always perform Wada and – as far as possible – fMRI.”

Many of the clinicians reported that fMRI paradigms had replaced the Wada test in some or all of their cases, particularly in the assessment of language lateralization. One wrote

“We can use fMRI for language but it has not been as useful for memory assessment. However, if you know that the person is left language dominant, is being considered for a right ATL, then one would not expect memory decline or amnesia after a nondominant resection and you might forego the Wada.”

While some did not have access to fMRI techniques, others relied on other available investigations to fulfil the role of the Wada, including PET studies and structural MRI in combination with baseline neuropsychological testing. On pediatric neurologist wrote:

“We have not done Wada for >20 years at our pediatric institution. We have good MRI and neuropsych to be able to estimate risks to memory and risk of amnesic syndrome. We use fMRI to lateralize language.”

Some of those who had abandoned the procedure in all patients, highlighted the methodological confounds of the Wada and the advantages of newer techniques. A neurologist wrote:

“The WADA test is an invasive procedure. It is difficult, patients often fall asleep during the test making interpretation at best difficult. Failure on a WADA test has prevented effective epilepsy surgery in many patients, and modern imaging techniques, such as fMRI, are better than the WADA test in my opinion. It is strange that this test is still around in an age of evidence-based medicine. This is a test for the history books of medicine.”


The results of this survey suggest that there has been a sea change in the employment of the Wada procedure in the presurgical investigation of TLE surgery candidates over the past 15 years. In 1993, Rausch et al. concluded from the second Palm Desert Epilepsy Surgery study that “it is extremely rare that an epilepsy surgery centre does not perform an IAP” (Rausch et al., 1993). A total of 85% of the centers, which took part in the 1993 study, reported that all their prospective TLE surgery patients underwent a Wada test prior to surgery. In our survey, only 12% of the centers indicated that every TLE surgery candidate underwent a Wada test prior to surgery whilst over one-third of the centers, who responded indicated that they never or very rarely (in less than 5% of cases) employed a Wada procedure in these patients.

Whilst the Wada procedure may provide some useful adjunctive data in terms of seizure lateralization, 89% of centers would be happy to proceed without Wada data in this respect for most or all of their cases. It appears from the responses given that the primary rationale for continuing to use the Wada test in the presurgical evaluation of TLE patients remains the assessment of language and memory function.

Two-thirds of the respondents indicated that they would be happy for most or all of their surgical candidates to proceed to surgery without language lateralization determined by the Wada test. For centers offering a selective amygdalo-hippocampectomy and some offering a standardized anterior temporal lobe resection, language dominance was not a critical determinant in the resection strategy. Others were using fMRI paradigms. Others indicated that they used a combination of baseline neuropsychological measures and ictal and postictal tests of language function to assess language dominance. Over half of the respondents indicated that they would feel confident assessing the risk of a postoperative amnesic syndrome without the Wada test. Only 6% of the respondents felt that Wada test was necessary in all cases to assess this risk. A similar pattern was seen in predicting postoperative memory decline.

This survey was limited to the largest group of epilepsy surgery cases (those with TLE), and the current prevalence and rationale behind the use of the Wada test in other epilepsy surgery candidates is unknown. Whilst the response rates for North America and Europe were similar, centers in Asia were underrepresented, possibly due to language barriers.

Nevertheless these findings have important ethical and medico-legal ramifications for prospective epilepsy surgery candidates. It could be argued that it is inappropriate to conduct an invasive test to assess risk, if the risks associated with the test outweigh those it is designed to assess. For example, the risks associated with a Wada procedure, though small, may be greater than the risk of a postoperative amnesic syndrome in a right-handed patient with clear cut right hippocampal sclerosis, intact contralateral structures, concordant EEG and baseline neuropsychological measures and an ictal semiology consistent with a nondominant seizure focus.

These findings indicate that the majority of epilepsy surgery centers no longer advocate the Wada procedure for all TLE surgical candidates. This suggests that the rationale for conducting this test should be carefully determined on an individual, case by case basis; examining the efficacy of all the alternative and available non invasive methods to address the outstanding concerns for each patient, in order to ensure best clinical practice (Abou-Khalil, 2007; Pelletier et al., 2007). Further, if fMRI techniques mature into standard clinical practice, the role of the Wada test is likely to be further reduced in the coming years.


With thanks to the all the clinicians in Argentina, Australia, Austria, Belgium, Brazil, Canada, Colombia, Czech Republic, Denmark, Finland, France, Georgia, Germany, Hong Kong, Hungary, Ireland, Israel, Italy, Japan, Mexico, The Netherlands, Norway, Saudi Arabia, Spain, South Korea, Sweden, Switzerland, Taiwan, Turkey, U.K. and the U.S.A. who took the time to participate in the survey. And with special thanks to all those in the U.S.A. who accepted with humor and good grace, the indignity of their country classified as a “minor outlying island” due to a glitch in the survey software.

Conflict of interest: We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.