The diagnosis of multiple sclerosis

Journal of the Neurological Sciences, 1986, 72:273-285

273

Elsevier JNS 2616

The Diagnosis of Multiple Sclerosis Contribution of Non-Clinical Tests E . A . C . M . Sanders ~, J . P . H . Hogenhuis 4

Reulen 2, E . A . Van der Velde 3 and L . A . H .

Departments of mNeurology and 3Medical Statistics, State University, Leiden; 2Department of Medical Physics, Free University. Amsterdam; and 4Department of Neurology, Hospital "'De Goddelijke Voorzienigheid", Sittard (The Netherlands)

(Received 6 June, 1985) (Revised, received24 September, 1985) (Accepted 25 September, 1985)

SUMMARY A group of 89 patients in whom multiple sclerosis (MS) has been clinically diagnosed with varying degrees of certainty, and 25 patients with optic neuritis (ON), were subjected to the following electrophysiological tests: visual evoked response (VER), auditory brainstem-evoked response (ABER), somatosensory-evoked response (S S ER), blink reflex and electronystagmography (ENG). All these patients also underwent computerized tomography (CT scan) and analysis of cerebrospinal fluid (CSF). A new diagnostic procedure is proposed, combining optimum detection of definite MS with optimally economical use of the above-mentioned non-clinical tests. The results for the MS patients show that definite MS can be diagnosed much more frequently (72~) if abnormal results in the above-mentioned tests are accepted as evidence of a (subclinical) CNS lesion. Application of the clinical diagnostic criteria of McAlpine yielded "definite MS" only in 27~o of our patient material. Our diagnostic criteria showed evidence for MS in 36~o of the patients clinically diagnosed as having ON. The test results were inconclusive as regards the possibility of the remaining ON patients developing MS in the future.

Key words: Diagnosis - Electrophysiological tests - Multiple sclerosis - Optic neuritis

Reprint requests to: J. P. H. Reulen, Ph.D., Department of Medical Physics,Free University,Van der Boechorststraat7, 1081BT Amsterdam,The Netherlands. 0022-510X/86/$03.50 © 1986 Elsevier Science Publishers B.V. (BiomedicalDivision)

274 INTRODUCTION A definite diagnosis of multiple sclerosis (MS) may be difficult to achieve because of the wide range of possible signs and symptoms and a great variability in clinical course of the disease. While there may be disagreement about details, the six clinical criteria put forward by Schumacher et al. (1965) may be taken as representative for a clinically definite diagnosis. Poser (1965) has pointed out the essential uniformity of correct - and incorrect - diagnoses of MS among neurologists of varying experience training and nationality. Nevertheless, an appreciable degree of uncertainty seems to be built into the diagnosis of this complaint. In various diagnostic schemas proposed for MS (McAlpine et al. 1965; Rose et al. 1976; McDonald and Halliday 1977), a distinction is always drawn between definite, probable and possible MS (or some similar categories). This uncertainty may be explained as reflecting two separate conflicts. On the one hand, that between the clinician's confidence in his intuitive ability to recognize the picture of MS (as pointed out e.g. by Poser et al. 1983) and the lack, in a number of cases, of objective evidence satisfying all the criteria of Schumacher et al. (1965). The lack of the clinical evidence of involvement of two or more separate parts of the central nervous system (CNS) may be mentioned in particular here. Moreover, clinical signs and symptoms in MS may not always reflect the actual damage done to the CNS, as has been emphasized in reports on pathological (Georgi et al. 1961; McAlpine 1961; MacKay and Hirano 1967; Fog and Linnemann 1970; Castaigne etal. 1981; Gilbert and Sadler 1983; Padke and Best 1983), electrophysiological and CT-scan investigations (Namerow and Etemadi 1970; Halliday et al. 1972, 1973; Asselman et al. 1975; Kimura 1975; Gyldensted 1976; Lowitzsch et al. 1976; Hennericci et al. 1977; Solingen et al. 1977; Cala et al. 1978; Small et al. 1978; Clifford and Jones 1979; Lacquaniti et al. 1979; Mastaglia et al. 1979; Matthews and Esiri 1979; Matthews and Small 1979; Trojaborg and Petersen 1979; Chiappa 1980; Mikol et al. 1980; Prasher and Gibson 1980; Robinson and Rudge 1980; Tackmann et al. 1980; Khoshbin and Hallett 1981; Green and Walcoff 1982; Yamada et al. 1982; Kjear 1983; Reulen et al. 1983; Sanders et al. 1984). The fact that the underlying cause of MS is not known, so that detection of the causative agent cannot be used as a basis for diagnosis, may also be seen as contributing to this conflict. On the other hand, we have the conflict between the desire to diagnose MS as early as possible and the reluctance to make a wrong diagnosis. If MS can be diagnosed early, there is a good hope of slowing down or halting the progression of the disease as future therapies are in development. Appreciable advances have been made on the therapeutic front of recent years; we may mention in this connection the recent report (Mertin et al. 1982) of certain immunosuppressive drugs, which appear to give good results in the treatment of young and slightly disabled MS patients. Methods promising an early definite diagnosis of MS could also help to provide a fn'mer basis for the experimental design of clinical trials of new therapies for MS. Both the above-mentioned conflicts can be largely resolved, and the chances of an early definite diagnosis of MS increased, if means can be found to detect CNS lesions reliably at an early, subclinical stage.

275 A variety of electrophysiological tests, and computerized tomography (CT scan), have been used to this end by various authors (Namerow and Etemadi 1970; Halliday et al. 1972, 1973; Asselman et al. 1975; Kimura 1975; Gyldensted 1976; Lowitzsch et al. 1976; Hennericci et al. 1977; Solingen et al. 1977; Cala et al. 1978; Small et al. 1978; Clifford and J ones 1979; Lacquaniti et al. 1979; M astaglia et al. 1979; Matthew s and Esifi 1979; Matthews and Small 1979; Trojaborg and Petersen 1979; Chiappa 1980; Mikol etal. 1980; Prasher and Gibson 1980; Robinson and Rudge 1980; Tackmann et al. 1980; Khoshbin and Hallett 1981; Green and Walcoff 1982; Yamada et al. 1982; Kjear 1983; Reulen et al. 1983; Sanders et al. 1984). Most of these authors tried out a combination of tests, though Mastaglia et al. (1977) were the only ones to use a battery of electrophysiological tests together with CT scan on a patient group with suspected MS. It has furthermore been suggested (McAlpine et al. 1965) that the presence of elevated immunoglobulin-G (IgG) levels in CSF should be added to the list of diagnostic criteria for MS. Trojaborg and Saxtrup (1979) and Bartel et al. (1983) combined evoked potentials with cerebrospinal fluid (CSF) analysis. Application of a large battery of non-clinical tests as a routine matter would be too time-consuming and expensive; and it is not yet clear which combination(s) of tests would give optimum diagnostic results under various circumstances, or which tests give most information about the presence of lesions at various levels in the CNS. Furthermore, none of these tests is specific for MS. Finally, exact data are not available on the frequency of false positive results in the various tests. In an attempt to answer these questions, we applied a battery of electrophysiological tests, VER, ABER, SSER, ENG, blink reflex, CT scan and CSF analysis uniformly to a group of 114 patients clinically diagnosed as having definite or suspected MS, and subjected the results to statistical analysis. PATIENTS AND METHODS Patients

The 114 patients studied comprised 89 MS patients (53 females and 36 males) and 25 ON patients (17 females and 8 males). The entire population of patients in these two categories were examined at the University Hospital, Leiden, the Netherlands, during the period from 1983 to 1984. The MS patients were classified according to the clinical criteria of McAlpine et al. (1965) into definite (n = 31, mean age A = 48 years, mean duration of disease D = 14.5 years, mean Kurtzke (1961) disability K = 6.7); probable (n = 31, A = 41 years, D -- 7.6 years, K = 3.7) and possible (n = 27, A = 37 years, D = 5.1 years, K = 1.6). Most of the MS (n = 68) and all ON patients were seen in the out-patient's clinic, while 21 MS patients were investigated during a period of hospitalization lasting several weeks. The tests were performed shortly after clinical examination. The ON patients were investigated within the first two months after the initial complaints of visual loss.

276 Methods VER, ABER, SSER, blink reflex and ENG tests, high-dose contrast enhanced (telebrix), CT scan and CSF analysis (Mancini and isoelectric focussing) were performed on all 114 patients. CT scans were judged abnormal when they showed areas of hypo- or hyperdense contrast enhancement indicating the presence of CNS lesions. Signs of cerebral atrophy or ventricular enlargement in CT scans were not regarded as indicative of MS. Further details of the above-mentioned methods were given in previous reports (Reulen et al. 1983; Sanders et al. 1984). Three CSF parameters were investigated in the present patient group since previous studies suggested that an abnormally high IgG index (> 0.85), an abnormally high IgG concentration (> 0.036 g/l) or the presence of oligoclonal bands (Poloni et al. 1979; Hosein and Johnson 1981) are the most common CSF abnormalities found in MS patients. Statistical analysis was performed with help of the SPSS software package. The following tests were used for statistical evaluation: Chi-square, Wilcoxon, Kendall, Fisher's exact test and the Mann-Witney U-test. The statistical associations between specific parameters were examined with the aid of contingency tables. RESULTS

Clinical Seventy-five MS patients showed clinical signs and symptoms of brainstem, cerebellar or spinal-cord involvement at the time of this study. The overlap of clinical signs attributable to these three CNS levels is demonstrated in Fig. 1. The associations between clinical brainstem and cerebellar abnormalities, and between clinical brainstem and spinal-cord abnormalities were found to be statistically highly significant (both P < 0.005). Clinical cerebellar and spinal-cord abnormalities often occurred simultaneously too, although the association was not significant. INCIDENCE

Clinical function investigated 25

Brainstem

(n=37)

Cerebellum

(n = 4 t )

Spinal cord

(n =68)

50

Patients 75

100% J

Fig. 1. Distribution of the brainstem, cerebellar and spinal cord lesions detected in the 114 patients on clinical examination, showing the overlap between the three structural types of lesions. Each disorder in one patient is represented by a black rectangle; each row represents the incidence of a particular disorder in the over-all patient group, while each column represents the various disorders detected in a given patient.

277 Electrophysiological, CT scan testing and CSF results Except for the VER results in the ON group, the electrophysiological tests revealed the highest amount of abnormalities in definite MS patients. Detailed results of the eleetrophysiological and CT scan are presented in Table 1, while Table 2 shows the results of CSF examination in the 114 patients. The electrophysiological signs of brainstem, cerebellar and spinal-cord involvement also show considerable overlap, as shown in Fig. 2. The VER results were significantly associated with clinical optic-nerve involvement (P < 0.005). A similar association was found between clinical involvement of the posterior columns of the spinal cord and abnormally prolonged SSER Nt4-N2o latencies (P = 0.001). TABLE 1 F R E Q U E N C I E S O F T H E E L E C T R O P H Y S I O L O G I C A L A N D CT SCAN A B N O R M A L I T I E S IN MS A N D ON PATIENTS Tests

Definite a (n = 31)

Probable ~ (n = 31)

Possible a (n = 27)

Total (n = 89)

Optic neuritis (n : 25)

n

n

n

%

n

%

n

%

%

%

VER ABER SSERNla-N2o Blink reflex ENG CT-scan

21 26 30 25 26 18

68 84 97 81 84 58

21 23 21 20 25 20

68 74 68 65 80 65

12 16 11 15 17 10

45 59 41 56 63 32

52 65 62 60 68 48

61 73 70 67 76 54

25 8 5 3 9 5

100 32 20 12 36 20

Total

31

100

31

100

25

83

87

98

25

100

a MS categories according to the criteria of McAlpine et al. (1965). TABLE 2 F R E Q U E N C I E S O F T H E CFS A B N O R M A L I T I E S IN MS A N D O N P A T I E N T S CSF parameter

IgG concentration ( > 0 . 0 3 6 g/l)

Definite a

Probable a

Possible a

Total (n = 89)

Optic neuritis (n = 25)

Overall total (n = !14)

(n = 31)

(n = 31)

(n = 27)

n

%

n

%

n

%

n

%

n

%

n

~o

26

84

18

58

21

78

65

73

3

12

68

71

IgG index ( > 0.85)

18

58

10

32

8

30

36

40

-

-

36

32

Oligoclonal

27

87

24

77

25

93

76

85

9

36

85

75

Total

28

90

27

87

25

93

82

92

9

36

91

80

a MS categories according to the criteria of McAlpine et al. (1965).

278 INCIDENCE Electrophyslological

Patients

function

investigated

2j5

ABER

(n = 7 3 )

SSER

(n=67)

Blinkreflex

(n = 5 9 )

ENG

(n= 77)

50

75

100%

mB mm

Fig. 2. Incidence of ABER, S S ER N j4-N2o , blink reflex and E N G abnormalities in patient material. (For further details see caption to Fig. 1).

The other three electrophysiological tests (ABER, ENG, blink reflex) were all significantly associated with signs of brainstem and/or cerebellar involvement (0.005 < P < 0.05). These findings are reported in greater detail in our previous report in which we advise the use of ENG and ABER tests in the search for a silent brainstem lesion in patients suspected of having MS (Sanders et al. 1985). Hypo- or hyperdense areas in the CT scan showed no association with any of the clinical signs and symptoms. Of all the electrophysiological tests, the ABER revealed the highest percentage of asymptomatic lesions (53~, Table 3). The other electrophysiological tests and CT scan demonstrated asymptomatic lesions in 26-50~o of the patients (Table 3).

Age, disease duration and disability The association between the elctrophysiological and CT scan results on the one hand and (a) patients' age, (b) disease duration and (c) Kurtzke disability on the other, were subjected to statistical analysis. TABLE 3 I N C I D E N C E OF T H E CLINICAL A N D S U B C L I N I C A L D I S O R D E R S D E T E C T E D BY T H E V A R I O U S TESTS IN T H E P A T I E N T M A T E R I A L Test

VER ABER SSER N14-N2o Blink reflex ENG CT-scan

Symptomatic lesions

Asymptomatic lesions

n

%

n

%

46 43 56 41 51 0

88 70 78 64 84 0

31 28 11 22 26 53

50 53 26 41 49 46.5

279 Abnormal VER or ENG recordings showed no statistical relationship with these three disease characteristics. The incidence of hypo- or hyperdense areas in CT scan was higher (P < 0.05) among patients with severe Kurtzke disability. The incidence of disorders indicated by ABER, SSER and blink reflex testing increased with age, duration of the disease and Kurtzke disability (0.005 < P < 0.05). DISCUSSION

Comparison with previous findings The incidence ofVER, ABER, SSER N 1 4 - N 2 0 , blink reflex, ENG and CT scan abnormalities found in the present patient group agreed well with the findings of other authors (Namerow and Etemadi 1970; Halliday et al. 1972, 1973; Asselman et al. 1975; Kimura 1975; Gyldensted 1976; Lowitzsch et al. 1976; Hennericci et al. 1977; Solingen et al. 1977; Cala et al. 1978; Small et al. 1978; Clifford and Jones 1979; Lacquaniti et al. 1979; Mastaglia et al. 1979; Matthews and Esiri 1979; Matthews and Small 1979; Trojaborg and Petersen 1979; Chiappa 1980; Mikol et al. 1980; Prasher and Gibson 1980; Robinson and Rudge 1980; Tackmann et al. 1980; Khoshbin and Hallett 1981; Green and Walcoff 1982; Yamada et al. 1982; Kjear 1983; Reulen et al. 1983; Sanders et al. 1984). Surprisingly, our observations on the frequencies of abnormally raised IgGconcentrations, IgG-indices and the presence of oligoclonal bands detected with isoelectric focusing, show a discrepancy with previous f'mdings (Poloni et al. 1979; Hosein and Johnson 1981). These previous f'mdings indicate a significantly lower frequency of immunological reactions in the CSF of probable and possible MS patients as compared with that for definite MS patients. In our MS patient group (N = 89) on the other hand the incidence of abnormally high IgG indices or concentrations and the presence of oligoclonal bands were distributed practically uniformly over the three clinical groups (Table 2). At present we do not have a good explanation for this fact, but it seems that CSF investigation has a high sensitivity as diagnostic test for MS in this study. The frequency of immunological reactions in the CSF of our ON patients (36 ~ ) agrees with that reported by other authors (35-40 ~o) (Sandberg-Wolheim 1975; Hutchinson 1976). The contribution of non-clinical tests to definite MS diagnosis As mentioned in the Introduction, a major objective of this study was to obtain a clear picture of the extent to which non-clinical tests can supplement clinical examination and hence reduce the uncertainty of the diagnosis of MS. In order to investigate this point, we propose the diagnostic classification shown in Table 4. As already suggested by Poser et al. (1983), this classification is based on the clinical signs and symptoms and the information gained with the non-clinical tests. Group I of this classification contains all patients considered, on the basis of clinical examination and/or the above-mentioned non-clinical tests, to have two or more strictly separate demyelinating lesions of the CNS. The great deal of overlap between the different signs and symptoms and/or the elctrophysiological test results concerning

280 TABLE 4 D I A G N O S T I C CLASSIFICATION FOR MULTIPLE SCLEROSIS USED AS A BASIS FOR THE SURVEY OF THE RESULTS SHOWN IN TABLE 5

Diagnostic group

No. of lesions (symptomatic or asymptomatic) detected

Intrathecal immunoglobulin synthesis

I II IIl IV

2 or more 2 or more one one

yes no yes no

the brainstem (see Figs. 1 and 2) does not allow to distinguish separated lesions in such a small region. These lesions are said to be "symptomatic" if they are discovered by clinical examination, and "asymptomatic" if they are indicated by the non-clinical tests. In addition, patients in group I are found on CSF analysis to have signs of intrathecal immunoglobulin synthesis - also in line with the suggestion by Poser et al. (1983) - that all definite MS patients show such CSF signs. Group I may thus be considered to contain the patients most likely to have definite MS according to the definition we are proposing here. Group II comprises all patients with two separate CNS lesions as in group I, but without the CSF signs. Group III contains the patients with one symptomatic or asymptomatic demyelinating CNS lesion and CSF signs of intrathecal immunoglobulin synthesis; these patients may be regarded as those with "incipient MS". Finally, group IV comprises those patients with one CNS lesion but no CSF signs; these patients are characterized as (provisionally) "non-MS". The distribution of our patient material between these four classes is shown in Table 5. Seventy-two percent of our patients fell into group I, only 4',50 into group I1, 11 ~o into group III and 13~/o into group IV. These results provide very fair confirmation of the assertion by Poser et al. (1983) that all defmite MS patients show intrathecal immunoglobulin synthesis: only 2 patients were found to possess all other signs of definite MS, but not to manifest intrathecal immunoglobulin synthesis (group II). Further, the data of Table 5 show how well founded the clinician's recognition of the picture of MS usually is. Only 31 of our MS patients met all the clinical criteria for MS of McAlpine et al.; all other 83 may be regarded as patients considered by the clinician on the basis of his experience probably to have MS, but where insufficient objective clinical evidence was present to support this diagnosis fully. Our results show that non-clinical tests reveal evidence of the "missing" CNS lesions in 51 of these 83 "doubtful" cases.

281 TABLE 5 OVER-ALL SURVEY OF DATA ON OUR PATIENT MATERIAL, ON THE BASIS OF THE CLASSIFICATION OF TABLE 4 For further details see section "The contribution of non-clinical tests to definite MS diagnosis (p. 279). Our diagnostic group

I

II III IV

Definitea (n = 31)

Probablea (n = 31)

Possible" (n = 27)

Optic neuritis (n = 25)

Total (n = 114)

n

%

n

%

n

%

n

%

n

%

28 1

90 3

28 0

90 0

17 1

63 4

9 3

36 12

82 5

72 4

2 0

7 0

3 0

9 2

7 2

26 7

0 13

0 52

12 15

11 13

" MS categories according to the criteria of McAlpine et al. (1965). It must be noted that one shortcoming of our results, considered as a basis for a proposed diagnostic procedure, is that we have no estimate of the number of false positives yielded by our non-clinical tests (i.e. the number of cases in which abnormal results in a given test do not correspond to the presence of a demyelinating CNS lesion); nor do we have any means for taking this possibility of false positives into account in the diagnostic procedure. The only reliable way for gaining an estimate of the incidence of false positives is a prospective study in which patients who have shown abnormal results in the various non-clinical tests are followed until clinical evidence can be obtained that their abnormal test results are indeed (or are not) followed by demonstrable clinical sings of a demyelinating CNS lesion; autopsy data on the presence or absence of the predicted lesion would provide the final certainty. Finally, a few remarks about the implications of our results for the relationship between O N and MS. The possibility that O N can be regarded as a sign of incipient MS has been discussed in the literature for a considerable time. The concrete estimates of the extent to which O N is actually followed by MS vary widely, from 8 ~ to 87 ~o (Cohen et al. 1979). This very wide spread doubtless depends largely on variation in the definition of MS used as basis for the comparison. In our study, using more sensitive diagnostic tools, we actually find that 36~o of our O N patients already have def'mite MS according to our definition (which does not take into acount the number of bouts). O f the remaining O N patients, 1 2 ~ fell into group II (signs of 2 CNS lesions, but no intrathecal immunoglobulin synthesis), and 5 2 ~ into group IV (signs of one CNS lesion without intrathecal immunoglobulin synthesis). It is striking that no O N patients were found in group III of our classification: this suggests that C S F analysis might provide a sensitive means of distinguishing between O N patients who already have (subclinical) MS, and those who do not (Moulin et al. 1983; Stendahl-Brodin and Luik 1983). However, our results do not provide a basis for predictions about how many O N patients are likely to develop MS in the future.

282

A diagnostic procedure for MS On the basis of our results, we propose the following scheme for the diagnosis of MS, using a combination of clinical examination and non-clinical tests to achieve maximum certainty while keeping the examination as short and simple as possible in order to save time and money and to keep the already considerable load on the patient as light as possible. The basis of this new diagnostic procedure is summarized in a flow diagram of the procedure in Fig. 3. The first step in this procedure is conventional clinical examination using e.g. the diagnostic criteria of McAlpine et al. (1965) or Schumacher et al. (1965). This will yield a certain number of definite MS patients, and an appreciable number in whom only one demyelinating CNS lesion is found on clinical grounds. Now our results have shown

I

Diagnostic clinical criteria of McAlpine et al.

I

In which CNS structures have clinical lesions been detected? I

.s

I : Apply ist non-clinical test

: Apply 2nd non-clinlcal test

~o~ not_ h a v e y

Fig. 3. A diagnostic procedure for MS.

283 that the majority of these monosymptomatic patients may be expected to have at least a second, asymptomatic CNS lesion detectable by one of the non-clinical tests discussed in the paper. We can look for a second CNS lesion with one of these non-clinical tests; Figure 3 presents guidelines for the choice of the most suitable test(s). Since the ABER, ENG, SSER and blink reflex tests yielded interdependent data where the brainstem function was concerned, we previously concluded that the blink reflex and the SSER tests gave hardly supplementary information (Sanders et al. 1985). If the diagnosis of definite MS (where symptomatic and asymptomatic CNS lesions are considered as equally valid as basis for this diagnosis) is already clinched during the fh'st round of these tests, there will be no need to apply the second round. This is particularly useful in connection with CT scan, which we include in Fig. 3 as a second-round test: while this examination can yield a great deal of information, it does subject the patient to X-ray load which must in principle be avoided if not absolutely necessary. Since our results showed only 2 MS patients with two or more CNS lesions, obtained either clinically and/or paraclinically, without CSF signs, as compared with 73 with two or more CNS lesions and CSF signs - indicating a "false positive" rate of less than 3 ~o - we have to consider that routine use of this examination adds an important biochemical and immunological confirmation of the diagnosis in more than 95 ~o of the cases. Bearing in mind the above-mentioned provision about the possibility of false positives in the non-clinical tests, the procedure described above would seem to provide a relatively simple means of increasing the certainty with which MS can be diagnosed in patients where the clinical examination reveals a fairly well based suspicion of the presence of this condition. In the near future N M R scanning probably replaces CT scanning in MS diagnosis (Yoang et al. 1981; Luker et al. 1983). The flow chart presented in Fig. 3 does not yet include results obtained by a Nuclear Magnetic Resonance (NMR) technique. So one could hesitate about the real life span of the diaguostical scheme. However, the definite value of the N M R scan for MS diagnosis can only be determined by a comparative study of this technique and all here used paraclinical tests including CSF analysis. ACKNOWLEDGEMENT The authors gratefully thank Dr. R.H. Bathgate (Eindhoven) for critically reviewing this paper. REFERENCES Asselman, P., D.W. Chadwick and C.D. Marsden (1975) Visual evoked responses in the diagnosis and management of patients suspected of multiple sclerosis, Brain, 98: 261-282. Bartel, D.R., O.N. Markand and O.J. Kolar (1983) The diagnosis and classificationof multiple sclerosis -- Evoked responses and spinal fluid electrophoresis,Neurology, 23:611-617. Cala, L.A., F.L. Mastaglia and J.L. Black (1978) Computerizedtomographyof brain and optic nerve in multiple sclerosis-- Observations in 100 patients including several studies in 16, J. Neurol. Sci., 36 (3): 41 i-426.

284 Castaigne, P., F. Lhermitte, R. Escourolle, J.J. Hauw, F. Gray and O. Lyon-Caen (1981) Les sclrroses en plaques asymptomatiques, Rev. Neurol. (Paris), 12: 729-739. Chiappa, K.H. (1980) Pattern shift visual, brainstem auditory and short latency somatosensory evoked potentials in multiple sclerosis, Neurology (Minneap.), 30 (2): 110-123. Clifford-Jones, R. E., G.P. Clarke and P. Mayles (1979) Crossed acoustic responses combined with visual and somatosensory evoked responses in the diagnosis of multiple sclerosis, J. Neurol. Neurosurg. Psychiat., 42 (8): 749-752. Cohen, M.M., S. Lessell and P.A. Wole (1979) A prospective study of the risk of developing multiple sclerosis in uncomplicated optic neuritis, Neurology (Minneap.), 29: 208-213. Fog, T. and F. Linnemann (1970) The course of multiple sclerosis, Acta NeuroL Scand., 47: Suppl. 1. Georgi, F., P. Hall and H.R. Muller (1961) Zur Problematik der Multiplen Sklerose, Basle. Gilbert, J.J. and M. Sadler (1983) Unsuspected multiple sclerosis, Arch. Neurol. (Chic.), 40: 533-536. Green, J.B. and M.R. Walcoff (1982) Evoked potentials in multiple sclerosis, Arch. Neurol. (Chic.), 39: 696-697. Gyldensted, C. (1976) Computer tomography of the brain in multiple sclerosis - - A radiological study of 110 patients with special reference to the determination of cerebral plaques, A cta Neurol. Scand., 53: 386-389. Halliday, A.M., W.J. McDonald and J. Mushin (1972) Delayed visual evoked response in optic neuritis, Lancet, i: 982-985. Halliday, A.M., W.J. McDonald and J. Mushin (1973) Visual evoked response in diagnosis of multiple sclerosis, Brit. Med. J., IV: 661-664. Hennericci, M., D. Wenzel and H.J. Freud (1977) The comparison of small-size rectangle and checkerboard stimulation for the evaluation of delayed evoked responses in patients suspected of multiple sclerosis, Brain, 100:119-136. Hosein, Z.Z. and K.P. Johnson (1981) Iso-electric focusing of cerebrospinal fluid proteins in the diagnosis of multiple sclerosis, Neurology (Minneap.), 31: 70-76. Hutchinson, W. M. (1976) Acute optic neuritis and the prognosis for multiple sclerosis, J. Neurol. Neurosurg. Psychiat., 39: 283-289. Khoshbin, S. and M. Hallett (1981) Multi-modality evoked potentials and blink reflex in multiple sclerosis, Neurology (Minneap.), 31 : 138-144. Kimura, J. (1975) Electrically elicited blink reflex in the diagnosis of multiple sclerosis - - A review of 260 patients over a seven year period, Brain, 98 (3): 583-598. Kjear, M. (1983) Evoked potentials - - With special reference to the diagnostic value in multiple sclerosis, Acta Neurol. Scand., 67: 67-89. Kurtzke, J. F. ( 1961) On the evaluation of disability in multiple sclerosis, Neurology (Minneap.), 11 : 686-694. Lacquaniti, F., P. Benna, M. Gilli, W. Troni and B. Bergamasco (1979) Brainstem auditory evoked potentials and blink reflex in quiescent multiple sclerosis, Electroenceph. Clin. NeurophysioL, 47: 607-610. Lowitzsch, U., U. Kuhnt, Ch. Sakmann, K. Maurer, H.C. Hope, O. Schott and K. Tahter (1976) Visual pattern evoked responses and blink reflexes in assessment of M S diagnosis - - A clinical study in 135 multiple sclerosis patients, J. Neurol., 213: 17-32. Luker, S.A., L. E. Crooks, M.J. Aminoff, L. Kaafman, S. Panitch, C. Mills and D. Norman (1983) Nuclear magnetic resonance imaging in multiple sclerosis, Ann. Neurol., 13: 592-601. McAlpine, D. ( 1961) The benign form of multiple sclerosis - - A study based on 241 cases seen within three years of onset and follow up until the tenth year of the disease, Brain, 84: 186-203. McAlpine, D., C.E. Lumsden and E.D. Ackeson (1965)Multiple Sclerosis - - A Reappraisal, Livingstone, Edinburgh, London. McDonald, W.I. and A.M. Halliday (1977) Diagnosis and classification of multiple sclerosis, Brit. Med. Bull., 33(1): 4-8. MacKay, R. and A. Hirano (1967) A form of benign multiple sclerosis, Arch. NeuroL (Chic.), 17: 588-600. Mastaglia, F.L., J.L. Black, L.A. Cala and D.W.K. Collins (1977) Evoked potentials, saccadic velocities and computerised tomography in diagnosis of multiple sclerosis, Brit. Med. J., I: 1315-1317. Mastaglia, F.L., J.L. Black and D. W.K. Collins (1979) Quantitative studies of saccadic and pursuit eye movements in multiple sclerosis, Brain, 102: 817-834. M atthews, W. B. and M. Esiri (1979) Multiple sclerosis plaques related to abnormal somatosensory evoked potentials, J. Neurol. Neurosurg. Psychiat., 42: 940-942. M atthews, W.B. and D.G. Small (1979) Serial recording of visual and somatosensory evoked potentials in multiple sclerosis, J. Neurol. Sci., 40:11-21. Mertin, J., M. Kremer, S. C. Knight, J. R. Batchelor, A. M. Halliday, P. Rudge, M.J.R. Healey, A. Compston.

285 E.J. Thompson, M. Denman and P.B. Medawar (1982). Double-blind controlled trial of immunosuppression in the treatment of multiple sclerosis - - Final report, Lancet, ii, 351-354. Mikol, F., A. Bouchareine, M.L. Aubin and J. Vignaud (1980) La tomodensitomttrie dans la scltrose en plaques, Rev. Neurol. (Paris), 136 (6-7): 481-490. Moulin, D., D.W. Paty and G.C. Ebers (1983) The predictive value of cerebrospinal fluid electrophoresis in "possible" multiple sclerosis, Brain, 106: 809-816. Namerow, N.S. and A. Etemadi (1970) The orbicularis oculi reflex in multiple sclerosis, Neurology (Minneap.), 20: 1200-1203. Padke, J.G. and P.I. Best (1983) A typical and clinically silent multiple sclerosis - - A report of 12 cases discovered unexpectedly at necropsy, J. Neurol. Neurosurg. Psychiat., 46: 414-420. Poloni, M., B. Rochelli, R. Seelst and R. Pinelli (1979) Intrathecal IgG synthesis in multiple sclerosis and other neurological diseases - - A comparative evaluation by IgG index and iso-electric focusing, J. Neurol., 221: 245-255. Poser, C.M. (1965) Clinical diagnostic criteria in epidemiological studies of multiple sclerosis, Ann. N.Y. Acad. Sci., 122: 506-519. Poser, C.M., D.W. Paty, L. Schernberg, W.J. McDonald, F.A. Davis, G.C. Ebers, K.P. Johnson, W.A. Sibley, D.H. Silberberg and W.W. TourteUotte (1983) New diagnostic criteria for multiple sclerosis - - Guidelines for research protocols, Ann. Neurol., 13 (3): 227-231. Prasher, D.K. and W. P.R. Gibson (1980) Brainstem auditory evoked potentials - - A comparative study of monaural versus binaural stimulation in the detection of multiple sclerosis, Electroenceph. Clin. Neurophysiol., 50: 247-253. Reulen, J. P.H., E.A.C.M. Sanders and L.A.H. Hogenhuis (1983) Eye-movement disorders in multiple sclerosis and optic neuritis, Brain, 106: 121-140. Robinson, K. and P. Rudge (1980) The use of the auditory evoked potentials in the diagnosis of multiple sclerosis, J. Neurol. Sci., 45 (2-3): 235-244. Rose, A. S., G.W. Ellison, L.W. Myers, and W.W. Tourtellotte (1976) Criteria for the clinical diagnosis of multiple sclerosis. In: United States - Japan Conference on Multiple Sclerosis, Neurology (Minneap.), No. 6, Part 2: 20-22. Sandberg-Wolheim, M. (1975) Optic neuritis - - Studies on the cerebrospinal fluid in relation to clinical course in 61 patients, Acta Neurol. Scand., 52: 167-178. Sanders, E.A.C.M., J. P. H. Reulen and L. A. H. Hogenhuis (1984) Central nervous system involvement in optic neuritis, J. Neurol. Neurosurg. Psychiat., 47: 241-249. Sanders, E.A.C.M., J.P.H. Reulen, L.A.H. Hogenhuis and E.A. Van der Velde (1985) Brainstem involvement in multiple sclerosis, Acta Neurol. Scand., 71: 54-61. Schumacher, G.A., G. Beebe, R.F. Kibler, L.T. Kurland, J.F. Kurtzke, F. McDowell, B. Nagler, W.A. Sibley, W.W. Tourtellotte and T.L. Willmon (1965) Problems of experimental trials on therapy in multiple sclerosis - - Report by the panel on the evaluation of experimental trials of therapy in multiple sclerosis, Ann. N.Y. Acad. Sci., 122: 522-586. Small, D. G., W.B. Matthews and M. Small (1978) The cervical somatosensory evoked potentials (SSEP) in the diagnosis of multiple sclerosis, J. Neurol. Sci., 35:211-224. Solingen, L. D., R.W. Baloh, L. Myers and G. Ellison (1977) Subclinical eye movement disorders in patients with multiple sclerosis, Neurology (Minneap.), 27: 614-619. Stendahl-Brodin, L. and H. Luik (1983) Optic neuritis - - Oligo-clonal bands increase the risk of multiple sclerosis, Acta Neurol. Scand., 67: 301-304. Tackmann, W., H. Strenge, R. Barth and A. Sojka-Raytscheff (1980) Evaluation of various brain structures in multiple sclerosis with multi-modality evoked potentials, blink reflex and nystagmography, J. Neurol. 224: 33-46. Trojaborg, W. and E. Petersen (1979) Visual and somatosensory evoked cortical potentials in multiple sclerosis, J. NeuroL Neurosurg. Psychiat., 42 (4): 323-330. Trojaborg, W. and O. Saxtrup (1979) Evoked potentials and immunoglobulin abnormalities in multiple sclerosis, Neurology (Minneap.), 31: 866-871. Yamada, T., E. Shivapour, T. Wilkinson and J. Kimura (1982) Short and long latency somatosensory evoked potentials in multiple sclerosis. Arch. Neurol. (Chic.), 39: 88-94. Yoang, J.R.S. Hall and C.A. Pallis (1981) Nuclear magnetic resonance imaging of the brain in multiple sclerosis, Lancet, ii: 1063-1066.