Are somatosensory evoked potentials of the tibial nerve the most sensitive test in diagnosing multiple sclerosis?
Background : Multiple sclerosis (MS) is mostly diagnosed clinically, but the diagnosis has significantly improved through the use of brain magnetic resonance imaging (MRI), testing of cerebrospinal fluid, and multimodal evoked potentials (MEPs). Even though MRI is the superior method in diagnosing this illness, MEPs remain important because they can detect clinically silent lesions in the sensory and motor pathways of the central nervous system (CNS). Aim : The aim of the study is to test the diagnostic sensitivity of MEPs and MRI and the ratio of their sensitivity in patients with MS. Materials and Methods : The study subjects included 293 patients with MS with disease duration of two to six years: 249 patients with relapsing-remitting (RR) MS and 44 with primary-progressive (PP) MS. All patients were subjected to an MRI brain scan, visual evoked potentials (VEPs), median somatosensory evoked potentials (SEPs), tibial somatosensory evoked potentials (SEPs), and auditory evoked potentials (AEPs). Abnormal Findings Included : changed wave morphology, interside difference in wave amplitude, absolute and interwave latency increased by 2.5 SD as compared with the control group. The control group comprised of 35 healthy subjects. Results : In this study the most abnormal findings were tibial SEPs, median SEPs, and VEPs. Our results suggest different sensitivity of MEPs in patients suffering from different forms of MS. In RR-MS the sensitivity of tibial SEPs was statically significant (Fischer's exact probability test) as compared to other evoked potential modalities. Similarly VEPs were more sensitive as compared to AEPs. In the PP-MS, median SEPs have been found to be more sensitive than VEPs, while tibial SEPs have been found to be more sensitive than AEPs. There was no significant difference in the sensitivity of MRI and MEPs both the forms of MS. Conclusion : Tibial SEPs produce the most abnormal results and the highest sensitivity in the RR-MS. We propose that this test as useful criterion for the diagnosis of MS.
Keywords: Evoked potentials, magnetic resonance imaging, multiple sclerosis
Multiple sclerosis (MS) is a chronic inflammatory disease with involvement of sensory, visual, auditory, and motor pathways both clinically and subclinically. The diagnosis is based on clinical features, magnetic resonance imaging (MRI) findings, cerebrospinal fluid, (CSF) findings, and multimodal evoked potential (MEP) studies. Evoked potentials test the function of visual, sensory, auditory, and motor pathways of the central nervous systems (CNS) and can provide evidence for topographic dissemination of demyelinating lesions, both clinical and subclinical. The clinical use of MEPs in the diagnosis of MS is somewhat limited because of the low specificity. Now a days MEPs are rarely done with the availability of MRI, which has a better diagnostic sensitivity in detecting the pathoanatomical and subclinical lesions.  For these reasons, MRI features have been given importantance in the McDonald criteria., , Of the MEPs, only visual evoked potential (VEP) findings have been considered as a useful criterion.  A number of previous studies have shown the importance of MEPs in the diagnosis of MS and also thier usefulness in monitoring the disease evolution and therapeutic effects. ,, More recent research has shown that MEPs are a good predictor of the outcome and gravity of the disease, and that they could be a good marker in MS.  The aim of this prospective study is to test the diagnostic sensitivity of MEPs and MRI in the patients with MS.
The study subjects included 293 patients with clinically diagnosed MS and monitored for five years at the Department of Clinical Neurophysiology of the Clinic of Neurology, Clinical Centre Nis, Serbia. The duration of the disease was between two and six years. The subjects included 121 men and 172 women, aged 18-45, with mean age of 32 years. All the patients had evoked potentials and MRI brain scan studies.
The evoked potentials tests included: visual evoked potentials (VEPs), median nerve somatosensitive evoked potentials (median SEPs), tibial nerve somatosensitive evoked potentials (tibial SEPs) and auditory evoked potentials (AEPs). Saphire II 4E (Medelec, UK) machine was used to record the MEPS. Control group included 35 healthy participants. Recording of various evoked potentials was by the well described conventional techniques.
Absolute P100 latency >113.4 ms (mean + 2.5 SD of control) and interside latency difference >8 ms (mean + 2.5 SD) were considered as abnormal. Interside difference in the latency of more than 50% was also considered abnormal.
Absolute N22 latency >25.4 ms (mean + 2.5 SD) and interside latency difference longer than the control group by more than 2.5 SD were taken as abnormal. Interside difference in the amplitude of N20 more than 50% was also considered abnormal
Absolute P40 latency >43.5 ms (mean + 2.5 SD) and interside latency those in difference longer than the control group by 2.5 SD were taken as abnormal. Interside difference in the amplitude of P40 more than 50%, was also considered as abnormal.
Absolute latencies longer than the controls (mean + 2.5 SD) and interside differencies in latencies longer than the controls (mean + 2.5 SD) (I wave>2.02; III wave > 4.18; V wave > 6.12; I-III > 2.51; III-V > 2.23; I-V > 4.43) were taken as abnormal. The changed morphology of I, III and V waves and ratio of amplitudes of I and V waves was also considered as abnormal.
Magnetic resonance imaging
The MRI scans were obtained with the Siemens Avanto 1.5T and the protocol included: T2 FAST Localizer, T1 SE Saggital, T2 TSE Transversal, FLAIR Transversal, T1 SE Transversal, T2 TSE Coronal, and contrast (gadolinium), T1 SE - Triplanar.
To compare the sensitivity of the methods, we used Fischer's exact test. Sensitivity was calculated as the percentage of abnormal findings with regard to the number of tested patients with the clinical diagnosis of MS, according to the McDonald criteria.
Of the 293 patients with MS studied during five years, 249 patients were with relapse-recurrent MS (RR-MS) and 44 patients with primary-progressive MS (PP-MS).
[Figure 1] and [Figure 2] and [Table 1] show the MEP results. In the RR-MS group the abnormal findings were: tibial SEPs (80.3% of 173 patients tested), VEPs (66.5% of 194 patients tested), median SEPs (59.5% of 163 patients tested) [Figure 1], and AEPs (43.1% out of 123 patients tested). In the PP-MS group [Figure 2], the abnormal findings were: median SEPs (90.3% of 31 patients tested), tibial SEPs (88.6% of 35 patients tested), VEPs (70.3% of 37 patients tested) and AEPs (58.6% of 29 patients tested). The frequency of abnormal findings on MRI were: RR-MS (92.2% of 180 patients tested) and PP-MS (94.7% of 38 patients tested).
There was no statistical difference between the sensitivity of the MRI and EP modalities in the RR-MS group [Table 2]. When the modalities of evoked potentials were compared, highly significant difference was found between tibial SEPs and median SEPs, where the tibial SEPs have shown higher sensitivity. The sensitivity of tibial SEPs was stastically significant than VEPs. The sensitivity of VEPs was statistically significant than median SEPs and AEPs.
In the PP-MS, there has been no significant difference in the sensitivity of the MRI and the modalities of evoked potentials [Table 3]. Moreover, there was no difference in the sensitivity between AEPs and VEPs and AEPs and median SEPs. There was, however, a significant difference in the sensitivity of median SEPs and VEPs, and also of median SEPs and tibial SEPs, where median SEPs were more sensitive. Tibial SEPs were significantly more sensitive than AEPs.
In MS, MEPs represent an electrophysiological correlation of demyelination and axonal degeneration of sensory, visual, auditory and motor pathways. Our study of MEPs in patients with MS showed significant abnormalities of MEPs. Strong correlation between abnormalities of MEPs and clinical symptoms has been shown in a number of studies. , The predictive value of MEPs in MS has also been shown.  MEPs are less sensitive in detecting demyelinating lesions of the CNS as compared to MRI, because they detect functional changes is some of the pathways, and not whole of the white matter.  However, they are significant for the early detection of clinically suspected lesions, especially in the acute phase of optic neuritis. , As the optic nerve is a predilective site for demyelination, and often signals the onset of MS, the percentage of abnormal results is very high. Our results show a high percentage of abnormal VEP findings in both the RR-MS (66.5%) and the PP-MS (70.3%). Similar were the observations in the earlier studies. ,,,, Older studies showed a significantly higher percentage of abnormal VEP findings in the definite or probable form of MS. ,, In our study VEPs were more sensitive than median SEPs and AEPs in the RR-MS while they were more sensitive than AEPs in the PP-MS.
Median and tibial SEPs evaluate the function of somatosensory pathways  and detect the lesions in these pathways.  Our results suggest a higher percentage of abnormal SEPs in both the RR-MS and PP-MS. Comparing the MEPs in the RR-MS, we found a highly significant difference in the sensitivity of tibial and median SEPs, where tibial SEPs were more sensitive than median SEPs and VEPs, while VEPs were more sensitive than median SEPs and AEPs. These observations can be explained by the anatomical length of the sensory pathways and also by the high frequency of demyelinating lesions in the spinal cord. Recent MRI studies in MS have shown demyelinating lesions in the spinal cord in as many as 90% of the patients.  The commonest lesion is found in the cervical cord, in the form of focal lesions and atrophy of the spinal cord. , In our study in the PP-MS, the difference in the sensitivity between median SEPs and VEPs was significant, and also between median SEPs and tibial SEPs, where median SEPs have shown more sensitivity in relation to both modalities. Tibial SEPs were more sensitive than AEPs. However, one should be cautious in interpreting these results, as there were considerably fewer patients with the PP-MS. In the milder clinical forms of both PP- and RR-MS, the only abnormal findings were tibial SEP. In contrast, in more severe forms, in addition to an abnormal tibial SEP findings, abnormal median SEP, VEP, and AEP findings were also recorded.
The McDonald criteria for the diagnosis of MS are based on MRI and VEPs (the only useful evoked-potential test in diagnosing MS). Our study has shown higher sensitivity of tibial SEPs and median SEPs, as compared with VEPs and AEPs in a large group of patients. We fell that subsequent revision of diagnostic criteria should consider including tibial and median SEPs as useful tests in the diagnosis of MS. VEPs only detect demyelinating pathology of anterior visual pathway and testing of retrochiasmal pathway requires time consuming half-field stimulation. Tibial and median SEPs are a better, less time-consuming methods for detecting demyelinating lesions, not only in the brain, but also in the spinal cord. Complementary use of VEPs and SEPs would significantly improve the diagnostic yield especially in developing countries, where MRI is not widely available and is also expensive.
AEPs detect demyelinating lesions in the brainstem. Earlier studies, pre-MRI period showed a high percentage of abnormal AEP results in MS, up to 77%. , In some other studies the percentage of abnormal AEP findings is much less.  Our study also showed low percentage abnormal AEP findings in both forms of MS forms and the lowest sensitivity as compared with other evoked potential modalities.
Even though the diagnostic value of MEPs in MS has decreased after the introduction of MRI, there are studies suggesting a significant correlation between MEPs and MRI.  There is general agreement in many studies that MEPs are still relevant to the diagnosis of MS, because they show good correlation with the clinical symptoms of the disease. ,, Our results have shown that there is no statistical difference in the sensitivity of MEPs and MRI. The results must be taken with some limitations, however, since only few patients had normal MRI. This limiting factor in statistical analysis of the data.
Our study shows different sensitivity of MEPs in the RR and PP forms of MS. Tibial SEPs produce the most abnormal results and the highest sensitivity in the RR-MS, so, in addition to VEPs, we propose that this test as a useful criterion for the diagnosis of MS. There is no significant difference in the sensitivity of MRI and MEPs in the diagnosis of MS, thus warranting the use of MEPs in diagnosis of MS.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]