Studies have summarized the high degree of concordance of language lateralization of functional magnetic resonance imaging (fMRI) and either the Wada test or direct electrical stimulation. (1) In this summary, fMRI was concordant with the Wada test in 78 of 83 (94%) cases and with direct electrical stimulation in 23 of 26 (88%) cases. In 2003, Sabsevitz and colleagues reported on a series of 24 consecutive patients who underwent both fMRI and Wada testing before left anterior temporal lobectomy for seizure disorders. While both tests were predictive of language changes, in this study fMRI had a sensitivity of 100% and specificity of 57%, while results for the Wada test were 100% and 43%, respectively. (2)
In 2008, Bizzi and colleagues assessed the sensitivity and specificity of fMRI for mapping language and motor functions using intraoperative intracortical mapping as the reference standard. (3) Thirty-four consecutive patients with a focal mass adjacent to eloquent cortex were included in the study. A site-by-site comparison between fMRI and intracortical mapping was performed with verb generation or finger tapping of the contralateral hand. A total of 251 sites were tested; 141 in patients evaluated with verb generation and 110 in patients evaluated with finger tapping. For hand motor function alone, sensitivity and specificity were 88% and 87%, respectively. For language, sensitivity and specificity were 80% and 78%, respectively. The fMRI for Broca’s area showed 100% sensitivity and 68% specificity, while the fMRI for Wernicke’s area showed 64% sensitivity and 85% specificity. Sensitivity of fMRI decreased from 93% for World Health Organization grade II gliomas to 65% for grade IV gliomas.
Another 2008 study assessed the language laterality index (LI) across different statistical thresholds in 13 patients with brain tumor and 7 controls; results were not compared with the Wada test. (4) In both groups, the language LI varied as a result of statistical thresholding, presence of tumor, prior surgery, and language task. Three patients demonstrated a shift in the LI between hemispheres as a function of statistical threshold. The study found no optimal threshold for reliably determining the LI. In another 2008 report, Wellmer et al. assessed whether currently recommended thresholds for the fMRI-lateralization index (LI) allowed identification of atypical dominant patients (i.e., not left-dominant) with sufficient safety for presurgical settings. (5) Of 65 patients who had presurgical workup for epilepsy surgery, 22 were identified as atypical dominant by the Wada test. Lateralization indices were calculated for 3 functionally determined regions of interest comprising Broca’s area, a prefrontal area outside Broca’s, and temporoparietal cortex overlapping with the Wernicke area. In patients in whom the Wada test results were compatible with typical left dominance, the fMRI-LI ranged from 1 to -0.61. Among patients with atypical language dominance according to the Wada test, fMRI-LI of the frontal and temporoparietal regions of interest ranged from 1 to -1. Depending on the chosen LI threshold for unilateral language dominance, between 2 and 5 patients (9% and 23%) of this sample of atypical dominant patients would have been misclassified as typical dominant. Another study compared presurgical mapping by fMRI with either verb generation or semantic decision/tone decision and the Wada test in 28 patients with epilepsy. (6) The study found moderate correlation between the two tasks (r: 0.495) and between the language tasks and the Wada test (r: 0.652 and r: 0.735). It was estimated that the language tasks explained approximately 58% of the variability of the Wada test with moderate convergent validity. With a LI threshold of + 0.25, 8 of the 28 patients (29%) may have been misclassified based on fMRI alone.
Presurgical Mapping of Eloquent Cortex
Medina and colleagues evaluated 60 consecutive patients prior to surgery in a 2005 report. (1) Language mapping was performed in 53 patients; motor mapping was done in 33, and visual mapping was conducted in 7. The fMRI study revealed change in anatomic location or lateralization of language-receptive (Wernicke) in 28% of patients and in language-expressive (Broca) in 21%. In 38 (63%) patients, fMRI helped to avoid further studies, including Wada test. In 31 (52%) and 25 (42%) of the patients, intraoperative mapping and surgical plans were altered because of fMRI results.
In 2006, Petrella and colleagues reported on the impact of fMRI preoperatively on 39 consecutive patients with brain tumors. (7) In 4 patients, additional tests, e.g., the Wada test, were not ordered because of the fMRI result. Treatment plans differed in 19 patients after fMRI, with a more aggressive approach recommended after imaging in 18 patients. However, the impact of the altered treatment plans on patient outcome was not assessed. The fMRI resulted in reduced surgical time for 22 patients; it also led to decisions to perform craniotomy in 13 patients in whom less invasive approaches had been initially planned. Other authors have reported that successful pre-operative fMRI decreased intracortical mapping time from about 50 minutes to 30 minutes and total operating time from an average 8.5 hours to about 7 hours. (8)
Research published in 2007 and 2008 appears to focus on improving and establishing standardized protocols for pre-surgical evaluation of the eloquent cortex. One report from 2007 described a routine preoperative fMRI protocol in 81 consecutive patients (70 with tumors on the left side and 11 with tumors on the right side and language deficits). (9) Patients were trained to recall simple sentences (picture cues) or to generate words in a category (word cues). Although 11 patients were not able to complete the more cognitively demanding word generation task, the combination of tasks allowed localization of both the Broca and Wernicke areas and determination of hemispheric language dominance in 79 (98%) patients. Surgical plans were modified in 9 (11%) patients based on the fMRI findings (7 patients underwent radiation therapy instead of surgery, and 2 patients had partial resection of large malignant gliomas). Results of the surgeries were not described. The authors noted that, although fMRI is capable of localizing the center of a functional area, resection borders cannot be reliably determined by this technique.
Use of preoperative fMRI in combination with intra-operative MRI (ioMRI) was reported in 2009 to allow more complete resection of tumors without affecting eloquent neurologic function. (10) In this case series of 29 patients, preoperative fMRI was performed to identify and coregister areas of brain activation for motor, speech, and short-term memory prior to brain tumor resection. Areas of brain activation that were identified preoperatively were superimposed on 1.5-T or 3-T scanners during the operative procedure, allowing the surgeon to avoid brain areas where damage would result in a postoperative neurologic deficit. Post-operative neurologic morbidity was reported to be low in the 27 patients in whom an fMRI-guided tumor resection was possible; 7 patients (26%) had transient neurologic impairments consisting of left hemiparesis, speech apraxia, motor apraxias, speech and motor apraxia, or temporary word-finding difficulty. No permanent neurologic impairment was observed in the 27 patients.
In a 2011 report, Wengenroth et al. compared localization of eloquent tumor-adjacent brain areas by fMRI versus structural MRI imaging in 77 consecutive patients with brain tumors of the central region. (11) During fMRI, the patients performed self-paced tongue up and down movements with closed lips, complex finger tapping with sequential finger-to-thumb opposition, as well as repetitive toe flexion-extension of the side contralateral to the respective lesion. The motor hand area was localized in 76/77 patients (99%) by fMRI and in 66/77 patients (86%) by structural MRI. Motor areas of the foot and tongue were investigated in 70 patients and could be identified by fMRI in 96% (tongue representation) and 97% (foot representation) of patients. Morphological landmarks for the motor hand area were found to be reliable in the unaffected hemisphere (97% success rate) but not in the tumor-affected hemisphere (86% success rate). In 14% of patients it was not possible to identify the motor hand area at all according to anatomical criteria. There are no reliable morphological landmarks for motor foot and tongue areas, and these representations could only be located by fMRI. After consideration of the clinical condition, tumor etiology, and fMRI results, the decision for neurosurgical operation was made in 52 patients (67.5%). In 16 patients, the decision against surgery was based mainly on fMRI results, which provided evidence that major neurological impairments would be expected after surgery. fMRI-based risk assessment before surgery had a high correlation with the clinical outcome and corresponded in 46 of 52 operative patients (88%) who had only minimal deficits or functional improvement postoperatively.
Localization of Seizure Focus with EEG-fMRI
In a 2007 report, the preoperative localization of epileptic focus was assessed in 29 complex cases (unclear focus and/or multifocality) that had been rejected for epilepsy surgery. (12) Patients were included in the study if they had no contraindications for MRI, had more than 10 interictal discharges in 40 minutes of a previously recorded electroencephalogram (EEG), and if the reason for rejection was the inability to localize a single source with EEG. The results of the fMRI were considered robust if a consensus-defined interictal electrical discharge was associated with a significant positive blood oxygen level-dependent (BOLD) response. In 8 (28%) patients, a robust fMRI response was considered to be topographically related to interictal electrical discharges. The EEG-fMRI findings improved localization in 4 of 6 unclear foci and advocated 1 of multiple foci in another patient; in 4 other patients multiple foci were confirmed. As a result of the testing, 4 patients (14%) were considered to be surgical candidates and 1 of the 4 had undergone surgery at the time of the publication. The authors of this European-based study describe this as the first report of the clinical use of EEG-fMRI.
Practice Guidelines and Position Statements
The American College of Radiology (ACR) and the American Society of Neuroradiology (ASNR) jointly published a 2007 guideline stating that fMRI using BOLD is a proven and useful tool for the evaluation of eloquent cortex in relation to a focal brain lesion, such as neoplasm or vascular malformation. (13) The guideline’s primary indications for fMRI include presurgical planning, surgical planning, and therapeutic follow-up for the assessment of intracranial tumoral disease and assessment of language functions for epilepsy surgery.
Overall, the literature indicates that fMRI is complementary to the Wada test and direct electrical stimulation in localizing certain eloquent functions. Evidence suggests that although bilateral activation patterns in fMRI cannot be conclusively interpreted, fMRI in patients who are to undergo neurosurgery for seizures or brain tumors may help to define eloquent areas, reduce surgical time, and alter treatment decisions. Therefore, fMRI may be considered medically necessary in the preoperative evaluation for patients being considered for neurosurgery, when the lesion is in close proximity to an eloquent area of the brain (e.g., controlling verbal or motor function) and testing is expected to have an important role in assessing the spatial relationship between the lesion and eloquent brain area.
Although promising, the use of EEG-fMRI to localize epileptic foci requires additional study. Use of EEG-fMRI to identify seizure focus is considered experimental, investigational and unproven.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.