Monday, May 2, 2016
My wife was diagnosed with myasthenia quite a few times based on severe clinical symptoms. None of the neurologists that saw her had any doubt, even though some of her symptoms were atypical. Despite that, her SFEMG was always normal, even at times of profound weakness, and this made the diagnosis doubtful repeatedly. As a scientist, with a good understanding of medical sciences, I know that there is no test that is 100% accurate. So my question is- What is the exact sensitivity, specificity, positive and negative predictive values of this test? What studies is this information based on? Are there any studies that addressed the accuracy of this test in seronegative as opposed to seropositive patients?
SFEMG is extremely sensitive in picking up MG in a clinically weak muscle. If the muscle being studied is definitely weak (let's say generating 50% of the normal force -- a "4" out of "5" on the strength grading system) and the SFEMG is normal, then the cause of that muscle's weakness is not MG. The results would be abnormal in both sero-positive and sero-negative MG.
The science is as follows: 1) Weakness due to MG occurs because synaptic transmission fails to generate an end-plate potential that is sufficiently large to generate an action potential on the post-synaptic muscle membrane. 2) Blocking is seen on SFEMG when a muscle fiber membrane in a "pair" of fibers fails to fire. Blocking should therefore be seen in muscles currently weakened by MG. 3) Increased jitter is seen before muscle fibers exhibit blocking. If blocking were missed for some reason, the jitter in a weakened muscle should be picked up. So, if the jitter is normal and no blocking occurs (normal SFEMG results), the weakness can not be due to a defect in synaptic function (the primary defect in MG).
Several caveats are in order: 1) To be dogmatic about the results, the muscle being studied must be clinically weak. (In MG, abnormal SFEMG results may be found in a muscle that is quite strong. This is because the method is so sensitive.) But it is possible that SFEMG results can be normal in a strong muscle in a patient that does have MG. 2) The EMGer must be extremely skilled. The technique is technically quite demanding and there are quite a few ways of rendering false negatives or false positives if the practitioner is not completely and very well trained, very experienced and active enough to maintain proficiency. (I personally have never achieved what I would consider proficiency.) 3) There are quite a few conditions other than MG than can give abnormal SFEMG results (false positives).
I will not be able to give you an adequate answer about the overall accuracy -- that would really be a dissertation of sorts. It seems however that the issue regarding your wife's diagnosis is one of SFEMG's test sensitivity. It is extremely sensitive provided one pays attention to the cautions noted above. Hope this helps.
John G Quinlan, MD
Professor of Neurology
College of Medicine
University of Cincinnati