Journal of Neurology, Neurosurgery, and Psychiatry, 1980, 43, 622-633 Clinical electrophysiology in myasthenia gravis ERIK STALBERG From the Department of Clinical Neurophysiology, University Hospital, Uppsala, Sweden SUMMARY Effective diagnostic methods are of great importance in order to recognise myasthenic patients among those with muscle fatigability. Intracellular recordings are useful for research work within the field and for detailed description of the motor end-plate's physiology in the individual case. The method is not used for the routine diagnosis of myasthenia gravis. The decrement of the electrical muscle response with nerve stimulation is the most commonly used method. The diagnostic yield is higher in proximal muscles, in warmed muscles, after exercise, and after ischaemia. A significant number of patients may be undiagnosed with this technique. The mechanical response with nerve stimulation shows the same type of decrement but also an abnormal response to long stimulation. The diagnostic value of this is under dispute. Single fibre EMG needs more patient cooperation than do these tests. The diagnostic yield is significantly higher. Some patients considered to have myasthenia gravis do not show any abnormalities with this technique, particularly those with the pure ocular form. Conventional EMG is not useful for the diagnosis of myasthenia, but may be indicated in these patients when concurrent nerve or muscle disease is in question. Tests for eye movement fatigue have not proved useful. Stapedius reflex fatigability is demonstrated in about the same proportion of patients as have positive SFEMG findings. The technique is not uncomfort- able for the patient and requires minimal cooperation. The general usefulness must be assessed by further routine use. Even with the advent of immunological tests, neurophysiological investigations are indispensable in helping establish the diagnosis of myasthenia gravis. Discrepancies between the results comparing electrophysiological and immunological tests may indicate that myasthenia gravis is a heterogenous entity within which subgroups may be identified. Neurophysiological methods have been used in the investigation of myasthenic patients for two main reasons. The diagnostic value of different methods have been demonstrated since the early reports of changes in muscle response to repetitive nerve stimulation. Over the decades these methods have also been used to help understand the pathophysio- logical mechanisms of myasthenia gravis (MG) and, in spite of the recent recognition of the immunological defect in MG, neurophysiological methods still have to be used to demonstrate the functional effects of the morphological or molecular changes demonstrated with other techniques. This paper will deal with some clinical neurophysiological methods for the diagnosis of disturbed neuromuscular transmission, both those commonly used and others at present used only in certain laboratories. MG has attracted considerable attention out of proportion to the actual number of patients with the Address for reprint requests: Dr Erik St5lberg, Department of Clinical Neurophysiology, University Hospital, S-750 14 Uppsala 14, Sweden. disease. This is mainly due to the early report of positive therapeutic effects of cholinesterase inhibitors, thymectomy, lymph drainage, cortico- steroids, 4-aminopyridine, and other drugs as well as exciting new knowledge about the pathogenesis, including the development of a useful animal model. The disease may also provide a model for the further understanding of other immunologically induced diseases. Efficient diagnostic methods are of great importance in order to recognise myasthenic patients among those with muscle fatiguability and to follow the effect of different therapeutic measures. Over the years Ian Simpson has contributed much to the knowledge of MG and has inspired enthusiasm among those working with these problems. As a clinical scientist he acknowledged and used neurophysiological methods in the battery of tests of neuromusclar transmission. He was the first to draw attention to the correlation of MG with autoimmune disorders. As editor of the Journal of Neurology, Neurosurgery and Psychiatry he has always been receptive to papers concerning neuromuscular 622 by copyright. on 2 April 2019 by guest. Protected http://jnnp.bmj.com/ J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Downloaded from
Transcript
Journal of Neurology, Neurosurgery, and Psychiatry, 1980, 43, 622-633
Clinical electrophysiology in myasthenia gravisERIK STALBERG
From the Department of Clinical Neurophysiology, University Hospital, Uppsala, Sweden
SUMMARY Effective diagnostic methods are of great importance in order to recognise myasthenicpatients among those with muscle fatigability. Intracellular recordings are useful for research workwithin the field and for detailed description of the motor end-plate's physiology in the individualcase. The method is not used for the routine diagnosis of myasthenia gravis. The decrement of theelectrical muscle response with nerve stimulation is the most commonly used method. The diagnosticyield is higher in proximal muscles, in warmed muscles, after exercise, and after ischaemia. A significantnumber of patients may be undiagnosed with this technique. The mechanical response with nervestimulation shows the same type of decrement but also an abnormal response to long stimulation.The diagnostic value of this is under dispute. Single fibre EMG needs more patient cooperationthan do these tests. The diagnostic yield is significantly higher. Some patients considered to havemyasthenia gravis do not show any abnormalities with this technique, particularly those with thepure ocular form. Conventional EMG is not useful for the diagnosis of myasthenia, but may beindicated in these patients when concurrent nerve or muscle disease is in question. Tests for eyemovement fatigue have not proved useful. Stapedius reflex fatigability is demonstrated in aboutthe same proportion of patients as have positive SFEMG findings. The technique is not uncomfort-able for the patient and requires minimal cooperation. The general usefulness must be assessed byfurther routine use. Even with the advent of immunological tests, neurophysiological investigationsare indispensable in helping establish the diagnosis of myasthenia gravis. Discrepancies between theresults comparing electrophysiological and immunological tests may indicate that myasthenia gravisis a heterogenous entity within which subgroups may be identified.
Neurophysiological methods have been used in theinvestigation of myasthenic patients for two mainreasons. The diagnostic value of different methodshave been demonstrated since the early reports ofchanges in muscle response to repetitive nervestimulation. Over the decades these methods havealso been used to help understand the pathophysio-logical mechanisms of myasthenia gravis (MG) and,in spite of the recent recognition of the immunologicaldefect in MG, neurophysiological methods still haveto be used to demonstrate the functional effects of themorphological or molecular changes demonstratedwith other techniques. This paper will deal with someclinical neurophysiological methods for the diagnosisof disturbed neuromuscular transmission, both thosecommonly used and others at present used only incertain laboratories.MG has attracted considerable attention out of
proportion to the actual number of patients with the
Address for reprint requests: Dr Erik St5lberg, Department ofClinical Neurophysiology, University Hospital, S-750 14 Uppsala 14,Sweden.
disease. This is mainly due to the early report ofpositive therapeutic effects of cholinesteraseinhibitors, thymectomy, lymph drainage, cortico-steroids, 4-aminopyridine, and other drugs as wellas exciting new knowledge about the pathogenesis,including the development of a useful animalmodel. The disease may also provide a model forthe further understanding of other immunologicallyinduced diseases.
Efficient diagnostic methods are of greatimportance in order to recognise myasthenicpatients among those with muscle fatiguability andto follow the effect of different therapeutic measures.Over the years Ian Simpson has contributed muchto the knowledge ofMG and has inspired enthusiasmamong those working with these problems. As aclinical scientist he acknowledged and usedneurophysiological methods in the battery of tests ofneuromusclar transmission. He was the first to drawattention to the correlation ofMG with autoimmunedisorders. As editor of the Journal of Neurology,Neurosurgery and Psychiatry he has always beenreceptive to papers concerning neuromuscular
622
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
transmission and so promoted the reporting of newmethods and new findings in this field.The first neurophysiological investigation of MG
was published in 1895 by Jolly' who used submaximalfaradic stimulation of the nerve and voluntaryactivation to study fatigability of muscle. Heobserved a continuous decrease in muscle responseduring faradisation with recovery after rest. Aftervoluntary exercise the muscle gave less response tosubsequent stimulation and conversely, gave lessvoluntary force after faradisation. A similartechnique is used today when we test the muscleresponse with repetitive nerve stimulation and whichis sometimes called the Jolly test, although we useanother stimulation pattern and usually asupramaximal stimulation strength.
In 1941 Harvey and Masland2 reported that theEMG potential recorded with concentric needleelectrodes varied in amplitude with consecutivedischarges during voluntary activation in myastheniagravis, a phenomenon which still is used as anindication of disturbed neuromuscular transmission.We now know that this depends on intermittentdisappearance or increased temporal variability ofthe individual spike components constituting the socalled motor unit potential. Their other finding wasthe decrementing muscle response with repetitivenerve stimulation.3 This method has been modifiedbut is probably the single most widely used methodfor the diagnosis of myasthenia today. Attemptshave been made to use the method of repetitivenerve stimulation to differentiate presynaptic andpostsynaptic defects, particularly together withpharmacological tests. It was used by Grob et a14-6 intheir studies of pathophysiological mechanism inMG.The similarity between the decremental response
in MG and that of the curarised normal muscle, theworsening in MG after intra-arterial injection ofacetylcholine (ACh), and the abnormal response toother depolarising agents was taken as evidence thatMG was due to a postsynaptic defect. Desmedtet al7-9 using the same techniques found greatersimilarity between MG and the effects ofhemicholinium which inhibits the uptake of cholineproducing a presynaptic block due to reduced AChsynthesis, and therefore considered MG to be apresynaptic disorder.
Since the decrement method does not allowlocalisation of the defect in greater detail a moredirect method was necessary. This came when thetechnique of intracellular recordings from humanintercostal muscles was introduced for the study ofmyasthenic motor end-plates.10 11
Intracellular recordings are used in somelaboratories for diagnosis of MG but the method is
623
complicated and has its main use in research. It hasbeen the most important method for understandingthe pathophysiology of MG and has helped inexplaining the results obtained with many otherneurophysiological methods.With the tip of the recording glass capillary elec-
trode inside the muscle fibre at the end-plateregion small sub-threshold depolarisations are re-corded which are too small to elicit a muscle fibreaction potential. These are called miniature end-plate potentials (mepps) and have amplitudes ofabout 1 mV. They represent the quantal releaseof ACh from the nerve terminal, and occur spon-taneously in the resting state. After depolarisationof the nerve terminal a large number of quantaare released and their mepps summate to producean end-plate potential (EPP) with an amplitudewhich would reach 70-80 mV in the normalmuscle if it was developed undisturbed. Thisexceeds the threshold for firing a muscle actionpotential and the EPP can therefore normallynot be examined since it is transformed into alarge propagated action potential. To study theEPP its amplitude must be reduced to subthresholdlevels. This is obtained by decreasing the pro-duction of release of ACh (usually by magnesium)or the postsynaptic sensitivity (usually by curare).In the latter case the mepp amplitude is dimin-ished to the same extent.From mepp frequency, mepp amplitude, and
EPP amplitude, information is gained aboutstorage and release of ACh and about postsynapticconditions. In the curarised preparation of a nor-mal motor end-plate the amplitude of EPPs de-crease during the first few of a train of stimuli.This is seen even at 2 Hz stimulation rate but ismore pronounced at higher frequencies. Whentested directly after a period of high frequencystimulation (50-200 Hz) the decrement with lowfrequency stimulation is initially less than before,and the amplitude is higher, a change known asposttetanic facilitation. After a few minutes thereis an even more pronounced posttetanic exhaus-tion. In the motor end-plate treated with mag-nesium (which causes a presynaptic block), theamplitude of successive EPPs shows a greatvariability at stimulation rates below 10 Hz sincethey are composed of a reduced but variable num-ber of normal sized quanta. At higher stimulationrates the EPP amplitude increases and shows lessvariability owing to an increased quantal contentof the EPP.The myasthenic motor end-plate was reported to
have reduced EPP amplitudes.12 Amplitudes weresometimes too low to reach the threshold for
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
firing a muscle fibre action potential and so gaveimpulse blocking. The same authors also foundreduced amplitudes of the mepps but normalnumber of mepps in each EPP, ie normal quantalcontent. The reduced mepp amplitude could eitherbe due to a presynaptic or a postsynaptic defect.Normal sensitivity to application of ACh analogueswas found and the reduced mepp amplitude wasinterpreted as indicating a presynaptic defect.probably defective synthesis of ACh since releasewas normal, tested with increased potassium inthe bath. Later investigations by others have con-firmed the EPP and mepp changes but they showeda reduced sensitivity to applied ACh in the experi-mentally induced MG13 probably due to reducedor physiologically blocked cholinergic receptors inaccordance with a postsynaptic defect, also shownwith other techniques in human myasthenicmuscles14 "', and now considered to be the principalsite of lesion in MG.These are the basic phenomena behind the
typical findings in the routine repetitive nervestimulation tests in disorders of neuromusculartransmission.
Repetitive nerve stimulationGeneral findings The change in muscle response torepetitive nerve stimulation has become the mostcommonly used test for the diagnosis of MG. It hasshown to be a useful diagnostic technique providedit is used correctly and to its full capability. It hasalso given additional knowledge about MG.A minimal programme for an investigation with
this method could be as follows. Surface electrodesare used, one over the belly of the muscle, the otherat a position remote from the muscle. Low frequencystimulation is given by a surface electrode on thecorresponding nerve at a rate of 2-3 Hz. Movementsinduced by the muscle contiaction have to beprevented by convenient fixation. The negativeamplitude of the first response is measured togetherwith the relative difference between the fourth (orfifth) and first response (fig 1). In our departmentwe use a computer for the analysis and have alsoincluded measurements of the area between thesignal envelope and base-line, also used by others.9 22
The change in this value is usually close to theamplitude change. If not, recording artefacts havefirst to be suspected. With these excluded thedifference may then be due to so called pseudo-facilitation, an increase in the amplitude of thesummated action potential due to "synchronisation."Here the area is a better indication of the changes.In the normal muscle the amplitude and area
measures exceed certain minimal values, varyingwith different muscles and different ages. The
Erik Stdlberg
amplitude and area normally change less than 500with repetitive nerve stimulation but in MG theydecrease more than that. As mentioned earlier thereis a normal reduction in end-plate potential (EPP)amplitudes during the first stimuli after a period ofrest but due to a high safety factor in the normalmuscle this does not lead to impulse blocking and istherefore not detected with a recording of the gross
muscle response. In MG or any condition with re-
duced safety factor, that is, having low EPPs (Eaton-Lambert'6 or other myasthenic syndromes, botulism,curare), this physiological decrease in EPP amplitudeis unmasked and causes blocking in successivelyincreasing number of motor end-plates, giving a
decremental response.
143
4 8
- 48
-'t 4.3
D. 55023 Date: 04 Jan80 Time 12:48:43:Muscle: 5 deltoidesDiagnosis: 22 myasthenia gravis
F
2-0 45Time (ms)
70 E 15 120 145 170 195
Fig 1 Computer printout of two decrement investigationsfrom the deltoid muscle before (upper) and after (lower)20 s of maximal voluntary activation. Black bars indicatethe negative amplitude of the first four responses. Thefirst (dotted) and fourth (full line) responses are displayedin detail.A =amplitude of the first response.
DA=decrement in amplitude of the fourth relative tothe first.DS=decrement in surface of the fourth relative tothe first.Note the facilitation with increased amplitude andreduced decrement after activation.
~~~~~A=10-4DA= -40~DS =
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
After a period of maximal activation or highfrequency repetitive stimulation (the latter is notused in our department since it is uncomfortable forthe patient although it is considered to be a bettercontrolled activation) the amplitude and area aremainly unchanged in the normal muscle. In themyasthenic muscle (not overtreated withcholinesterase inhibitors) the decrement becomesless pronounced or disappears and an eventuallyreduced resting amplitude is increased towardsnormal values, an effect seen only during the first20-30 seconds after the activation.
This is again the normal phenomenon of facilita-tion of ACh release in the motor end-plates. It isunmasked in situations where the increase in AChmay contribute positively to restoration of trans-mission. In conditions where disturbed neuro-muscular transmission is mainly caused by impairedrelease, typically in Eaton-Lambert syndrome, theeffect of this facilitation is particularly pronounced.On the other hand in conditions where an over-exposure to ACh is causing the dysfunction (such asovertreatment with cholinesterase inhibitors,17intoxication with organophosphorus compounds,'8)there is no increase of the recorded response butrather a reduced amplitude directly after maximalvoluntary contraction or tetanic stimulation. Theamplitude of the response at the height of thefacilitation gives an indication of the total number o fmotor end-plates that can be activated, that is, notirreversibly blocked or destroyed and the measureis thus of practical value.
After this short period of facilitation in themyasthenic muscle the amplitude decreases and thedecrement becomes more pronounced. This is calledthe postactivation exhaustion and was first describedby Desmedt.19 This parameter seems physiologicallycloser to the fairly reversible myasthenic fatiqueafter exercise than the decrement. The cause of thepostactivation exhaustion is not yet clear. Thereceptors are known to show "desensitisation"experimentally after even a short exposure to ACh20and this may play a role but further investigationsare necessary.
In all tests it should be remembered that facilita-tion and exhaustion are two separate phenomenaacting at the same time but with different time scales,and that the obseived response is determined by thesummated effect of the two.9 This affects the resultsof the routine electrophysiological testing. In apatient with pronounced MG a period with maximalvoluntary activity of 20 seconds may be too long toshow facilitation since more pronounced exhaustionis already dominating after this time. Here a shorterperiod may shown facilitation. In a patient withmyasthenia of moderate degree a 20 second period
625
may be enough to give facilitation but too short toshow any significant exhaustion. One has to beaware of these effects of activity when testingpatients. To compare results from different occasionsthe same basic conditions must be present. We restthe patient for 30 minutes before the test. In longterm studies it is important to have the patientcomfortably positioned and completely relaxed. Theeffect of activity on base-line conditions isparticularly pronounced in severely affected muscles.Neuromuscular transmission can be quantified by
means of a similar technique using different stimula-tion patterns. By using double pulse stimulation andtesting the relative amplitude of the second response arecovery curve of the neuromuscular transmission isobtained.2 9 21 In the MG there is a depression of thesecond response not seen in the normal muscle atintervals of 0-1 to 10 seconds, correlated to theseverity of the disease.By using repetitive nerve stimulation (1-8 Hz) for
10 to 40 minutes other aspects of the dynamics ofneuromuscular transmission can be tested.9 22-24 Thearea of the action potential reaches a plateau levelafter the initial decrement and a transient facilitation.Bergmans et al.22-24 assumed that this level representsa steady state of emptying and refilling thetransmitter store and used the information from theprolonged stimulation experiments to study thekinetics of the transmitter. New knowledge about thepostsynaptic defects in MG may to some extentchange the interpretation of the above mentionedspecial tests which therefore are not described indetail here.
Difference between mutscles In the clinical examina-tion MG usually shows a proximal distribution.This must also be noted when making electro-physiological tests. The decrement is as a rule lowerin hand muscles (properly warmed) than in proximalmuscles in the same patient. The diagnostic yield ofthe method is significantly lower for distal than forproximal muscles. This was shown in a study of 80MG patients.25 A decremental pattern was obtainedin 82% in deltoid muscle, 50% in abductor digitiminimi (ADM) muscle, 62-5% in orbicularis oculimuscle and 52% in the wrist flexors. When all thesemuscles were considered together 95 % of the patientsshowed abnormalities. In our cases (table) we havenot seen a single case where the hand muscles haveshown a decrement and the proximal muscles havebeen normal at the same time. This may notnecessarily indicate that the immunological attackto the motor end-plates is more pronouncedproximally but may indicate different safety factorsin different muscle groups or reflect other differences.From Single Fibre EMG (SFEMG) studies it has
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
Table A summary of test results of the whole MG material of 164 cases at the time when the patient was firstexamined at our laboratory. No steroid treatment. Some patients were receiving cholinesterase inhibitors. The casesincluded ocular (23), mild-mod?rate generalised (139) and severe MG (6) patients. ADM=abductor digiti minimi muscle,EDC=extensor digitorum communis muscle, antibodies= cholinergic receptor antibodies
Decrement: ADM pos 23/74 (31 %) Deltoid pos in 48% of ADM neg casesDeltoid pos 52/80 (65%) ADM pos in 0% of Deltoid neg cases
been reported26 that inADM muscles with minimal orno decrement some individual motor end-platesshowed impulse blockings and some of themfacilitation even at low stimulation rates (2 Hz)which more or less effectively compensated for thedrop-out in the others and explained the small neteffect measured as the decremental response withsurface electrodes.
Effect of temperature The effect of temperature onmyasthenic muscle has been discussed by Simpson27and others.28 In our dec-ement studies deteriorationwith increasing temperature is a constant finding.With a change from 26 to 350 intramuscularly, dec-rement may increase from 0% to 29% (fig 2B). This isalso the case in patients that do not report any subjec-tive worsening with increasing temperature. The effectof temperature on neuromuscular transmission hasbeen studied in animals. 29 30 The cause for theimpaired neuromuscular transmission at highertemperatures is not known in detail. Intracellular
40 A
36322824201612840 IFi
40
36
3228
24
20
16
128
4
0
B
*-S*'I
280 310 340 370 40)0Temperature (IC)
Fig 2 Effect oftemperature onthe decrement(first/fourthresponse) in twopatients withmyastheniagravis. A fromdeltoid muscle,Bfrom ADMmuscle. Thetemperature isslowly changedwith a heatinglamp andicebags andis measuredwith anintramuscularthermocouple.
recordings show a dramatic increase in miniatureend-plate potentials (mepp) frequency. In themagnesium blocked motor end-plate the EPPamplitude decreases with increasing temperatureprobably due to decreased quantum content(presynaptic effect); in curarised muscle the EPPamplitude is seen to increase but is dramaticallyshortened, most likely due to postsynaptic membraneeffects, and increased cholinesterase activity.
Because of a high safety factor the normal muscledoes not show any neuromuscular block within thephysiological temperature range. In cases withdisturbed neuromuscular transmission this factormay have considerable influence. In order tocompare the results of one investigation withanother in the same patient it is therefore of greatestimportance to make the test at a standardisedtemperature. A thermostatically controlled lamp isrecommended. Intramuscular temperature in thehand decreases normally to below 30°C (260 is notuncommonly seen) within 30 minutes of rest in awarm laboratory room. Proximal muscles show lesschange and a temperature reduction of 2°C (from36 to 340) is more commonly seen. This parametermust always be checked in long term investigations,for example when following the effect of intra-venously injected or per orally administratedcholinesterase inhibitors or other drugs over morethan 30 minutes.
Provocative tests The easiest way of increasing themyasthenic defect temporarily is to fatigue themuscle by activity, either by voluntary exercise or bya more standardised electrical stimulation, forexample 3 Hz for five minutes. The effect can bestronger if the muscle temperature is slightlyincreased.The neuromuscular transmission is sensitive to
ischaemia. Therefore, a combination of exercise andischaemia is reported9 to be effective. The test hasbeen made in two steps, first 3 Hz stimulation forfive minutes with free blood circulation, then thesame activation under ischaemia. With this adecrement can be seen in cases of mild MG that donot show any significant abnormality with standardtesting before provocation.
crw0)
Erik Sttalberg626
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
Cliniic al electrophy7s,iology inmyasthenia gravIis
Based on the increased sensitivity of patients withMG to curare, this drug has been used to reveal aneuromuscular distuitbance. In order to reduce therisk the test is made after regional intravenousadministration.31 A diagnostic yield of Ip to 960 ingeneralised MG is reported. 12 A curare test isconsidered potentially hazardous and its use hastherefore been discouraged in most laboratories.
Staircase phenlomnenonl Decrement of the electricalresponse as described above is usually the commonlyused test of the MG defect but it can be measuredfrom the mechanical response as well. In thissituation not only is neuromuscular transmissiontested but also the contractile characteristics of themuscle fibres. With low frequency stimulation (2 Hz)normal muscle shows a slow increase in twitchamplitude during the first minutes of stimulation.This is assumed to be an effect of intensifiedexcitation-contraction process. In myasthenia this socalled staircase phenomenon was reported to beabnormally wveak or by some authlors absent33whereas others9 reported that the staircasephenomenon was usually within normal limits. If iti.s abnormal, after the amplitude reduction in theelectrical response caused by neuromuscular trans-mission is considered, it should suggest defects inMG other than those located at the motor end-plate.
627
This should be tested further. Today the value oftesting the staircase phenomenon for the diagnosis ofmyasthenia gravis is still uncertain.
Sinigle fibre EAIGWith the development of a technique called SingleFibre EMG (SFEMG)1' :13 it has become possible tostudy the microphysiology of the motor unit inhumans including the functional status of individualmotor end-plates in sitia. A needle electrode with arecording surface of 25 ,um in diameter in a side port ofthe cannula is positioned in the voluntarily activatedmuscle to record activity from a few adjacent musclefibres. For the study of neuromuscular transmissionan electrode position is soughlt where activity fromtwo muscle fibres belonging to the same motor unitis recorded as a potential pair. There is usually atime interval between the two action potentialsdepending on the difference in propagation timefrom the common brancling point along the nervetwig to the recording electrode. With consecutivedischarges there is a variability in the interpotentialintervals called the jitter. When expressed as themean value of consecutive differences of these timeintervals (MCD), this jitter is of the order of 5-50 psin the normal muscle. The jitter increases in situationswith disturbed neuromiiuscular transmission such asafter a small dose of curare; when the disturbance is
I
Fig 3 SFE,GAG jitterl Ir'(''0o illg.s' fi'Oml EDC nst.scle of' I patient wit/i un17a*sthlieniaglcllis (an1td allylignan1l1cy. Sivniptoins were o01/11 presenit (durnit7g perliOds of fevcr.Th7e oscilloscope sweep i.s ti iggg1iere hi, tile first aCtiOn7 pOten7tiCal, delaycd byI,1-2 m,is. Thle jitter iS SeeCn aIS (I a'6l)i blc pOSitiOn7 of the secon7d Ind i/l C clsNO aIthilrd aCtiOn1 pOt1itial. L'ppe tln CCiclgs: 20 sUoperin1iposed sweeps, lower tracings:sweeps Moved dw10nWardS ithliIm thlc soic Nscle 01 cami7 see miorlal77jittCer ( ),imicr-eased jitter- hbit l1o h/lo Akiligs (B) anct inbcreasedjitter and7l imlterIolittenltblockings (sec('01d potemiticil iti C). C d en701nStrates dlilferent degree of ahnormnalitva11n10omg mllotOr en17d-plaltes iii the Isamn7c m11otor- ull7it. Jitter A - 29 ps, B 65 p.s,C-81 ps aind 49 ps.
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
more pronounced partial or total impulse blocking isseen. The jitter expresses a varying neuromusculardelay time which may be due to a variable rise timeof consecutive EPPs, which has not been seen withintracellular recordings in the normal muscle. It ismore likely due to a slight variation of the thresholdat which the EPP initiates a muscle action potential.The EPP will therefore reach the threshold afterdifferent times from initiation. When the EPP ischanged in shape, for example after curare reachingthe threshold under a less steep slope, the effect of thethreshold fluctuations will be more pronounced andgive rise to a larger jitter. If the EPP is too low toreach the threshold, impulse blocking occurs.The jitter value is different for motor end-plates
even within the same motor unit. Different muscleshave a different mean jitter value within the samesubject and the mean value for one particular musclediffers between subjects. After a small standarddose of curare the jitter value increases less in thosemotor end-plates with an initially low value thanin those with an initially high jitter value. It seemsthat the jitter value in a motor end-plate is anindicator of its safety factor expressed as sensitivityto curare, that is, low jitter suggests less sensitivity.In myasthenia gravis the jitter is typically increased.The degree of abnormality in one muscle shows arange from normal motor end-plates to those withpronounced changes including partial or totalblocking (fig 3). This range of abnormality is alsoseen within the same motor unit (fig 3C). Whenusing the SFEMG method it is therefore necessaryto make recordings from at least 20 differentelectrode positions. The jitter mean value varies fordifferent muscles in the myasthenic patient. Usuallywe start recording from the extensor digitorumcommunis (EDC) muscle. It is easy to activate andshows abnormalities even in slight MG, clinically aswell as neurophysiologically. If this muscle is normal,recordings are made from other muscles dependingupon the symptoms, usually deltoid, frontalis ororbicularis oculi muscles. The findings are quantifiedas jitter value and degree of blocking for eachindividual potential pair, often summarised for thewhole recording as a percentage of recordings withnormal potential pairs, those with increased jitter,and those showing partial impulse blockings(fig 4). The mean jitter value for all recordings isgiven as well. The criterion for abnormality inindividual recordings is a jitter value exceeding anupper limit determined from the normal material(about 50ps but different for different muscles). Theinvestigation in one muscle of at least 20 recordingsis abnormal if more than two individual recordingsshow abnormal jitter or if the mean jitter exceedsa certain value obtained from the normal material.
Fig 4 Computer printout showing distribution ofjittervalues (jts) in EDC muscle from the same patient asdescribed in fig 3. Vertical line indicates upper normallimit for the jitter in this muscle. Filled symbols indicaterecordings with partial impulse blockings, oval symbols=double potentials, square symbols=more complexpotentials. Some statistics are given. 9-30-61 means that9% of the recordings show intermittent blockings, 30%show increased jitter but not blockings and 61% shownormal jitter. (In this patient repetitive stimulationshowed a decrement of 3% in ADM and 4% in thedeltoid muscle at rest and 120% during postactivationexhaustion.)
The abnormal jitter and degree of blocking usuallydecrease after the injection of Tensilon in patientswith untreated myasthenia gravis. Normal motorend-plates in myasthenia do not change after theinjection of Tensilon. In patients receiving cholin-esterase inhibitors the effect of Tensilon may varyfor different motor end-plates.36 Some of them mayshow improvement, others in the same muscle maynot change their abnormality, while others at thesame occasion may show even increase in jitter anddegree of blocking. These findings indicate thevarying degree of myasthenic defect and sensitivityto treatment. Some motor end-plates are under-treated, others are optimally or even overtreated.This is certainly true not only for individual motorend-plates but for muscle groups. In the clinicalsituation some muscles may be undertreated whileothers may be overtreated. In these situations theoptimum response has to be judged from the clinicalsituation with particular regard to vital musclegroups, such as those concerned with respiration orswallowing. The effect of temperature is principallythe same as indicated by the findings in repetitivenerve stimulation. With increasing temperature theabnormality in a motor end-plate increases. Duringcontinuous and steady activity the jitter may increaseduring the first few seconds but will then usuallyremain relatively constant. With decreasing meanfiring frequency the jitter may decrease and atincreasing rate the jitter increases.
628
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
In Eaton-Lambert syndrome the jitter is increased.It has the same characteristics as jitter in MGexcept for the reaction to increased firing rate. InMG the jitter and degree of blocking increases, inEaton-Lambert syndrome they decrease when theinnervation rate is increased.
In some clinically diagnosed patients with MG aproportion of the abnormal motor end-plates show,in contrast to the larger part of the motor end-plates, a slight improvement during continuousvoluntary activation as described for the Eaton-Lambert syndrome. Thus patients with clinicalmyasthenia may have a mixture of motor end-platedefects, some typical for MG, some similar to thosefound in the myasthenic syndrome. This inter-mediate form is sometimes seen also with decrementtesting in these patients.37SFEMG has been performed in 164 of our
patients with MG diagnosed on clinical grounde.The jitter values have been abnormal in 94% of thecases. The degree of neuromuscular disturbancsexpressed in terms of jitter and impulse blocking isgrossly correlated to the degree of clinical involve-ment. When the patients are subgrouped into thoseocular, moderate generalised and severe generalisedMG the degree of abnormality is successively morepronounced for each group when the extensordigitorum communis muscle is tested. The SFEMGabnormalities in two groups have been compared,38one with and the other without clinical involvementof the tested muscle, the extensor digitorum com-munis muscle. The jitter was significantly moreabnormal in the group where clinical symptomswere present, 114+8-5 jis than in the nonsympto-matic group, 47 ±25-is. In the symptomatic groupabnormalities were found in 63 of 64 and in thenonsymptomatic group in 56 of 74 patients.
In a few cases we have made SFEMG investiga-tions before and after complete remission, that isdisappearance of clinical signs and symptoms andon no medication. The average jitter decreased butsome motor end-plates still showed abnormalities.In none of these cases did SFEMG normalisecompletely (fig 5). In another study38 7 out of 9patients with clinical remission after corticosteroids-showed abnormal SFEMG values.A comparison between the diagnostic yield of
repetitive nerve stimulation (temperature-con-trolled proximal muscle), antibody titre againstacetylcholine receptors, and SFEMG results aresummarised in the table. It is seen that the SFEMGinvestigations have the highest proportion ofpositive tests. In no case was a decrement presentwhere SFEMG was negative. The titre of receptorantibodies was abnormal in all seven tested cases outof the 10 SFEMG negative cases. On the other hand
A.,_730110 (Thymect. Jun 1970)l 00-81-19,/Moderate gen.symptoms
TH--:* Cholinesterase inhib.0 00
O 0 000 0 0T - - -
77070500-30-701.Complete clin.remission
0. No treatment0 00 0 0
0 20 410 60 80 100 120 140 160 180s200 jitterJitter in 2 cases of MG in complete clinical remission
Fig 5 Distribution ofjitter values in two patients beforeand after complete clinical remission. Vertical lines uppernormal value. Filled symbols indicate recordings withpartial impulse blockings. The three hyphenated valuesexpressed in per cent (arrows) indicate respectively% recordings with increased jitter and blockings,increased jitter and normal jitter. The improvement isclearly seen but in both cases neurophysiologicalabnormalities are still present.
SFEMG was abnormal in all 17 cases with normalantibody titre. In patients with ocular myasthenia,22all three tests in general had the lowest percentageof positive findings, 28% for repetitive nervestimulation, 67% for SFEMG in EDC and 86%when also a facial muscle is investigated, and 67%for antibodies.
It should be stressed that increased jitter andpartial impulse blockings are not pathognomonic forMG, but indicate disturbed neuromuscular trans-mission or sometimes an abnormal impulse propaga-tion in the terminal nerves. Increased jitter andblocking are also seen during early stages of re-innervation,35 for example, posttraumatic, lowermotor neurone disorders, polyneuropathies, also inpronounced electrolyte disturbances, and to someextent in myopathies, particularly in polymyositis.Other SFEMG parameters such as fibre density andduration usually typify these disorders. They arerarely a problem in the differential diagnosis fromMG clinically electrophysiologically.Fibre Density The arrangement of muscle fibreswithin the motor unit can also be determined bySFEMG.The average number of muscle fibres belonging to
the same motor unit is determined from 20 different
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
electrode positions and the local fibre density (FD) ismeasured. The FD differs for different normalmuscles but is of the order of 1P3 to 1P5. From otherstudies we know that the FD is increased inconditions with reinnervation and in situations ofverified fibre type grouping.39 40
fn our total material or patients with MG the FDis increased on average. Statistical analysis has notrevealed any correlation with severity or durationof the disease but there is a statistically significantdifference between patients treated and not treatedwith cholinesterase inhibitors.
In the patients where the treatment had been usedfor more than one month the FD was increasedfrom the normal mean by 2-2 SD (p<0-01) in thebiceps brachi muscle compared to 1-2 SD in theuntreated patients. The difference was most pro-nounced in older treated patients. In the cases wherebiopsy had been taken the degree of fibre typegrouping was correlated to the FD. The increasedfibre density most likely reflects denervation causedby the cholinesterase inhibitors. Signs of denervationcan also be induced in animals.41 42 The slightincrease in average fibre density in the nontreatedgroup as compared to a normal value is statisticallysignificant (p < 0-01). This may be due to themorphological changes of the myasthenic motor end-plate43 44 which sometimes may be pronouncedprobably leading to denervation. Reinnervation byaxonal sprouts from adjacent motor units willproduce remodelling indicated by abnormal fibredensity values for the motor unit. Morphologicalabnormalities of the motor end-plate have beendiscussed by Simpson earlier.46
Conventional electromyographyConventional electromyography (EMG) isperformed in MG mainly for two reasons: one is tofind signs of this particular disease, the other andmore impoitant is to exclude or detect otherdiseases which clinically may have similar symptoms.Some of these may occur concurrently with MG suchas thyroid myopathy and polymyositis; others maybe induced by the steroid treatment.The typical finding in conventional needle EMG
is varying shape of the motor unit potential. Harveyand Masland2 reported amplitude variation as acharacteristic sign in MG. This is due to increasedvariability in the time dispersion between individualspike components constituting the motor unitpotential, that is increased jitter. Owing to drop outof individual fibres in more pronounced MG thenumber of muscle fibres in the active motor unit maybe reduced which give motor unit potentials ofdecreased amplitude and short duration, the same asis seen in primary myopathies. This is particularly
Erik Stailbergseen during activity. After a period of rest the motorunit potential may regain a normal shape. This EMGsign should not be misinterpreted as an indication of"myopathic" changes in MG. This phenomenon hasto be taken into consideration when an EMGinvestigation is made to evaluate the presence of amyopathy in MG, eg induced steroid myopathy.The conventional needle EMG has thus not turned
out to be very useful in the diagnosis of MG but isof value for the study of concomitant musculardisorders.
Other testsMuscle fatigability in myasthenia gravis can be seenin most, perhaps all, skeletal muscles in patientswith definite symptoms, more frequently in severecases. Other tests than repetitive nerve stimulationcan be used to quantify this.One such test concerns the extraocular muscles.
The patient is asked to follow with his eyes a dotmoving on a screen in front of him.46 The eyemovements are recorded with surface electrodeslateral to the eye similarly to the technique used innystagmography. In cases ofMG the eye movementwill progressively become slower and the amplitudeof the excursion smaller. The test has been used to alimited extent for special studies. The value of thistest for the diagnosis of MG as compared to othershas not been evaluated so far but the yield isprobably less than with repetitive nerve stimulation.Another measure of muscular fatigue is obtained
by measuring fatigue of the stapedius reflex. Thestapedius muscle controls the tension of the ear drumand reacts reflexly to sound stimuli. The contractiondeveloped by the muscle can be inferred bymeasuring the acoustic impedance of the tympanicmembrane. It is used in many audiology departmentsin routine neuroaudiological investigations. Whena continuous tone is given to the ear the musclecontracts and remains so until the sound is removed.The measured impedance is increased to a constantvalue during the stimulation. In MG there is acontinuous decrease of the acoustic impedance dueto a fatigue of the stapedius muscle.47-49 Owing tobase-line shifts it is sometimes difficult to measurethe impedance over many seconds of stimulation.The technique was modified49 to a stimulation with apulsating 500 Hz tone where the duration of thepulses were 200 or 500 ms (sound pulse stimulationSPS 200 and SPS 500) separated by an equalduration "off line". In this way the muscle isrepetitively stimulated in a similar way used innerve stimulation tests. When the change inimpedance is measured between the responsesobtained during the first and last 10 seconds of astimulation period of 300 seconds the normal subject
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
shows a slight increment for the 500 ms pulses anda slight decrement for the 200 ms pulses. In MGthere is a decrementing response to this stimulationthat improves after cholinesterase inhibitors. WithSPS 500 stimulation, 830% of MG patients showedabnormal responses and with SPS 200 77% wereabnormal. At least one of the stimulation patternswas abnormal in 96% of the cases tested. Thetechnique is easy to apply in places where theaudiology department is equipped for stapediusreflex testing. The test is painless for the patient andthe diagnostic yield seems to be very high, com-parable to SFEMG. Experience so far is limited andit is recommended that control values be obtainedbefore using this test for the routine diagnosis of MG.
Comparison of techniquesThe diagnosis of MG is mainly based on clinicalsymptoms and signs. Laboratory investigations arehowever important to confirm the diagnosis or tomake it unlikely. These tests are also indispensableto describe the status of the neuromuscular trans-mission in atypical cases. Many of the myasthenicsyndromes simulate MG clinically in many, thoughnot all, respects. New subgroups of myasthenia willcertainly be described by means of combinedelectrophysiological and immunological testing.Usually it suffices to use only one of the electro-physiological tests when it reveals a neuromusculardefect. When the first used method fails to demon-strate an abnormality other tests must be utilised.Their accessibility in the individual situation, thediagnostic yield of the method, and discomfort forthe patient are factors of importance for the choice.Decrement studies are commonly available in
most EMG laboratories; they are technicallysimple although technical pitfalls can easily ruin theresults. Studies of the hand muscles are toleratedby most patients, but even with warming the handbefore testing and with measurements after exercisethe diagnostic yield is relatively low. Ischaemia andhigh frequency long term nerve stimulation mayprovoke abnormalities but are more painful for thepatient and may require local anaesthesia. Decre-ment studies in proximal muscles, eg biceps,deltoid or facial muscles, are somewhat more painfulthan hand muscles but usually tolerated by thepatients, including children. The tests in thesemuscles are more prone to artefactual changes;however the neuromuscular abnormalities are muchmore demonstrable. A decrement test is incompleteif only hand muscles are studied and the resultsobtained are negative. SFEMG measurements arestill only available in a few laboratories. It requiresminimal changes in routine EMG equipment withsome extra training and experience by the electro-
631
myographer. The patients' cooperation is neededand a complete test can usually not be performed inpatients below the age of seven years. For the patientthe test gives the same discomfort as an ordinaryEMG investigation; some patients find this testless uncomfortable than repetitive nerve stimulation.The diagnostic yield is very high and is particularlyvaluable in cases with mild forms of myastheniagravis and in cases with the ocular form. Usuallyonly one muscle has to be tested. If SFEMG in EDCmuscle is normal in a patient with ocular symptomsthe frontalis or orbicularis oculi muscles should betested. If a complete investigation gives normalresults in a clinically weak muscle the diagnosis ofmyasthenia is very unlikely.
Tests of stapedius reflex and eye movements arepainless and require minimal patient cooperation.The test can be performed in any well equippedaudiological and ENT department respectively buttheir clinical applicability for the diagnosis ofmyasthenia remains to be assessed by furtherroutine use.
In this paper the value of measuring the antibodytitre against acetylcholine receptor has not beendiscussed. This test demonstrates the immunologicalabnormality and has obvious advantages. So far thetest is not correlated to the clinical severity of thedisease and has unfortunately, like the other tests,the lowest yield in ocular myasthenia.The discrepencies in the results when using the
different electrophysiological and immunologicaltests may indicate that myasthenia gravis is aheterogenous entity within which subgroups may beidentified. It is a challenge for further studies ofmyasthenia gravis and the myasthenic syndromes.
The project is supported by the Swedish ResearchCouncil (Grant 135). The collaboration of Dr P 0Osterman who examined most of the patients in theSFEMG study and Dr Ann Kari Lefvert whomeasured the receptor antibodies is acknowledged.
References
I Jolly F. Uber Myasthenia gravis pseudoparalytica.Berl klin Wschr 1895; 32:1-7.
5 Grob D, Johns RJ. NM transmission in myastheniagravis with particular reference to the Ach insensitiveblock. In: Viets, ed. Myasthenia gravis. Springfield:Thomas, 1961: 127-49.
G
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
15 Engel AG, Lambert EH, Howard FM. Immunecomplexes (IgG and C3) at the motor end-plate inmyasthenia gravis. Ultrastructural and light micro-scopic localisation and electrophysiologic correla-tion. Mayo Clin Proc 1977; 52:267-80.
16 Eaton LM, Lambert EH. Electromyography andelectric stimulation of nerves in diseases of motorunit: observations on the myasthenic syndromeassociated with malignant tumours. JAMA 1957;163: 1117-24.
17 Roberts DV, Wilson A. Electromyography indiagnosis and treatment. In: Greene R, ed. Myas-thenia gravis. London: Heinemann, 1969: 29-42.
18 Jager KW, Roberts DV, Wilson A. Neuromuscularfunction in pesticide workers.BJIndMed 1970;27:273.
19 Desmedt JE. Nature of the defect of neuromusculartransmission in myasthenic patients. 'Post-tetanicexhaustion'. Nature 1957; 179:156-7.
20 Magazank LG, Vyskocil F. Desensitisation at theneuromuscular junction. In: Thesleff S, ed. Motorinnervation ofmuscle. Academic Press. 1976: 151-76.
21 Johns RJ, Grob D, Harvey AM. Effects of nervestimulation in normal subjects and in patients withmyasthenia gravis. BullJohns Hopk 1956; 99:125-35.
22 Bergmans J, Rosselle N, Verheyen G, Schellens L.The kinetics of transmitter release in myastheniagravis. I. An electrophysiological analysis of thestorage of transmitter. Electromyogr clin Neuro-physiol 1972; 12:443-88.
Erik Stdlberg
23 Bergmans J, Rosselle N, Verheyen G, Schellens L.The kinetics of transmitter release in myastheniagravis. II. An electrophysiological analysis of thereplenishment of transmitter stores. Electromyogrclin Neurophysiol 1973; 13: 3-58.
24 Bergmans J, Rosselle N, Verheyen G, Schellens L.The kinetics of transmitter release in myastheniagravis. III. An electrophysiological analysis of therelease of transmitter. Electromyogr clin Neuro-physiol 1973; 13:145-73.
25 Ozdemir C, Young R. The results to be expectedfrom electrical testing in the diagnosis of myastheniagravis. In: Ann NY Acad Sci 1976; 274:203-22.
26 Schwartz MS, Stalberg E. Single fibre electromyo-graphic studies in myasthenia gravis with repetitivenerve stimulation. J Neurol Neurosurg Psychiat 1975;38:678-82.
27 Simpson JA. Myasthenia gravis: a new hypothesis.Scott Med J 1960; 5:419-36.
28 Borenstein S, Desmedt JE. New diagnostic pro-cedures in myasthenia gravis. In: Desmedt JE, ed.New Dev in Electromyogr clin Neurophysiol. Basel:Karger, 1973; 1:350-74.
29 Hubbard J, Jones SF, Landau EM. The effect oftemperature change upon transmitter release,facilitation and post-tetanic potentiation. J Physiol1971; 216:591-609.
30 Barrett EF, Barrett JN, Botz D, Chang DB,Mahaffey D. Temperature-sensitive aspects ofevoked and spontaneous transmitter release at thefrog neuromuscular junction. J Physiol 1978; 279:253-73.
31 Foldes F, Klonymus F, Maisel W, Osserman K. Anew curare test for the diagnosis of myastheniagravis. J Am M A 1968; 203:649-53.
32 Horowitz SH, Sivak M. The regional curare test andelectrophysiologic diagnosis of myasthenia gravis:Further studies. Muscle & Nerve 1978; 1:432-4.
33 Slomic A, Rosenfalck A, Buchthal F. Electrical andmechanical responses of normal and myasthenicmuscle, with particular reference to the staircasephenomenon, Brain Res 1968; 10:1-75.
34 Ekstedt J, Stalberg E. Myasthenia gravis. Diagnosticaspects by a new electrophysiological method.Opuscula medica 1967; 12:73-6.
35 Stalberg E, Trontelj J. Single fibre electromyography.Old Woking, Surrey: The Mirvalle Press Ltd 1979;1-244.
36 Stalberg E, Trontelj J, Schwartz MS. Single musclefibre recording of the jitter phenomenon in patientswith myasthenia gravis and in members of theirfamilies. Ann NY Acad Sci 1976; 274:189-262.
37 Schwartz MS, Stalberg E. Myasthenia gravis withfamilies. Ann NY Acad Sci 1976; 274:189-262.tion with electrophysiologic methods including singlefibre electromyography. Neurology 1975; 25:1,80-84.
38 Sanders DB, Howard JF, Johns TR. Single fiberelectromyography in myasthenia gravis. Neurology1979; 29:1, 68-76.
39 Hakelius L, Stalberg E. Electromyographical studiesof free autogenous muscle transplants in man. ScandJPlast Reconstr Surg 1974; 8:211-9.
by copyright. on 2 A
pril 2019 by guest. Protected
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.43.7.622 on 1 July 1980. Dow
40 Henriksson KG, Stalberg E. The terminal innerva-tion pattern in polymyositis: A histochemical andSFEMG study. Muscle & Nerve 1978; 1:3-13.
41 Roberts DV, Thesleff S. Acethylcholine release frommotor-nerve endings in rats treated with neostigmine.Eur J Clin Pharmacol 1969; 6:281-5.
42 Engel AG, Lambert EH, Santa T. Study of long-term anticholinesterase theraphy. Effects on neuro-
muscular transmission and on motor end-plate finestructure. Neurology 1973; 23:1273-81.
43 Coers C, Woolf AL. The innervation of muscle. Abiopsy study. Oxford: Blackwell, 1951.
44 Engel AG, Santa T. Histometric analysis of theultrastructure of the neuromuscular junction inmyasthenia gravis and in the myasthenic syndrome.Ann NY Acad Sci 1971; 183:46-63.
45 Simpson JA. A morphological explanation of thetransmission defect in myasthenia gravis. Ann NYAcad Sci 1971; 183:241-7.
46 Baloh RW, Keesey JC. Saccade fatigue and responseto edrophonium for the diagnosis of myastheniagravis. Ann NY Acad Sci 1976; 274:631-41.
47 Blom S, Zakrisson JE. The stapedius reflex in thediagnosis of myasthenia gravis. J Neurol Sci 1974;21:71-6.
