Betahistine in Meniere's disease By T. J. WILMOT (Omagh, N. Ireland) and G. N. MENON (Weesp, The Netherlands) Summary
THE effects of betahistine hydrochloride (Sere) on the clinical features of Meniere's disease were assessed in two double-blind, placebo-controlled, cross-over, clinical studies. The diagnosis was based on a peripheral, fluctuating, recruiting, cochlear (sensorineural) deafness in one or both ears, tinnitus (usually of low tone) and paroxysmal attacks of rotational vertigo. Appropriate auditory and vestibular analyses confirmed the diagnosis. Twenty-four patients were admitted to the studies after careful screening over two-and-a-half years. Twenty-two patients completed the studies, ten of whom received betahistine and placebo for eight weeks each whereas the remaining twelve were given betahistine and placebo for twelve weeks each. The dose of betahistine was the same (16 mg. t.i.d.) in both studies. Daily symptom score cards kept by all patients throughout the studies showed a statistically significant preference for betahistine over placebo with regard to vertigo (p= 0-025), tinnitus (p = o-oio) and fullness of the ear (p = 0-036). Symptom scores of deafness and vomiting indicated trends in favour of betahistine but these did not attain statistical significance. Objective measurements of deafness (mean db. loss), however, showed a highly significant improvement in favour of betahistine when compared with placebo (p 10 (= Sere markedly better)
: 9
Number of patients (P-S)^—10 (= Placebo markedly better)
: 0
Number of patients—io< (P-S)< 10 8 (= no important change)
P- value, one-sided Sign test of marked changes
837
0-004
T. J. Wilmot and G. N. Menon TABLE I I I B . MEAN AND STANDARD DEVIATIONS OF HEARING LOSS IN PATIENTS WITH BILATERAL HEARING LOSS
Hearing loss (dB) (n = 5) Standard deviation
Mean
Sere (S)
Number of patients (P-S) > 10 in both ears : 4 (= Sere markedly better)
—left —right
13-2 12-6
io-6 7-1
Placebo (P)—left —right
30-2 27-4
21 -i "•3
Number of patients (P—S)< —10 in both ears : 0 (-)- Placebo markedly better) Number of patients—io< (P-S)< 10 : 1 in both ears (no important change)
P-value, one-sided Sign test of marked changes 0 • 06
Sign Test (p = o • 004) and a significant trend in favour of betahistine in patients with bilateral deafness (p = 0-06). 4. Objective evaluation of vestibular function While clinical balance tests, positional, optokinetic, caloric and rotation tests were performed on all patients before the studies (pretreatment), at the cross-over of the treatments and at the end of the studies, the only test results which were studied in depth were the rotation tests. No clinical alteration in the other tests were noted and the indelicacy of the stimuli applied made detailed observation unnecessary. The results of the rotation tests are expressed in Table IV TABLE IV. OBJECTIVE EVALUATION OF VESTIBULAR FUNCTION EXPRESSED AS AVERAGE EYE SHIFT PER SECOND
Rotation test results (n = 22) Pre-treatment (Pre) Mean
Sere (S)
SD
Mean
SD
Placebo (P) SD Mean
°/sec*—left
3-2
2-9
2-8
2 -2
33
2-5
3 "/sec—left
8-3
5-1 6-2
7'3
4-0
7-2
13-7 24-0
7-8 13-4
12 -2 22-9
3"5 7-3 n-7
3-6 5-o
2-4 7-1
2-9
8-9
13-7 30-6
1-6 2-7 8-6
I
b "/sec—left
9 0 "/sec*—left 1 "/sec—right*
3 °/sec2—right
6 °/secJ—right 90 °/sec*—right
I
5'5
32-5 3-6 8-9 i4-b 36-0
15-3
15-5
* One missing value; n = 21.
838
6-4 13-3 28-5
i-8 3-1
7-5
15-5
P-value, one-sided (S>P) Wilcoxon signed rank test 0-84 0-44 O-22 0-44 o-8i 0-28 0-15 0-18
Betahistine in Meniere's disease in terms of mean values, standard deviations and p-values as calculated by using the Wilcoxon Signed Rank Test. To summarize, the patients' symptoms as assessed by themselves and by the investigator were improved by betahistine hydrochloride. Objective assessment of hearing was also in favour of betahistine while the attempted objective assesment of vestibular function showed no important improvement with either betahistine or placebo. Discussion The assessment of a multi-factorial problem such as Meniere's disease is fraught with difficulty. In the first place the severity of the disease is extremely variable and the degree of involvement of the cochlea and vestibule is frequently not parallel. As examples, there may be unilateral cochlear involvement with evidence of bilateral loss of vestibular function, or bilateral cochlear involvement with retention of useful vestibular function, or slight cochlear involvement and severe vestibular hypofunction. Many other combinations may occur. Secondly, study of the natural history of the disease shows that natural remissions occur and that these may be entirely independent of treatment. It is therefore against this physical complexity that any drug must first be judged. Thirdly, the psychosomatic background of many cases of Meniere's disease suggests that, if not actually involved in the causation of the disorder (which we believe to be the case), it is of considerable importance in initiating recurrences. For the above reasons, it was important to provide in our two studies a constant background of medical support for each individual over a prolonged period of time—16 weeks for one group of cases and 24 weeks for the other—and to ensure that this support was not conditioned by any knowledge on the part of the investigator as to whether the patient was on drug or placebo at any particular moment. This, and the performance of routine time-consuming medical and technical examinations, posed a considerable burden on the investigator and his chief technician, who were individually responsible for these aspects and who had to arrange their holidays and schedules to ensure no disruption over a period of z\ years. It was felt that a study of small numbers of cases in depth under personal supervision of a single investigator might be more informative than that of a larger number by multiple observers. Lastly, the ethical problems involved in a long-term evaluation of this nature are not insignificant and were carefully weighed before the studies commenced. The results of the studies suggest that during the period of 2\ years when patients were being treated on a randomized basis, the majority 839
T. J. Wilmot and G. N. Menon of patients were symptomatically better and had improved hearing during the time they were on betahistine hydrochloride. On the other hand the carefully repeated battery of vestibular tests done under constant conditions, failed to show any convincing evidence of objective improvement in vestibular function. In particular, the detailed and extensive rotational testing programme which was designed expressly for this purpose also failed in this respect. While we had not expected caloric testing to be delicate enough to show changes in function we had hoped that the same would not apply to more delicate stimuli such as rotational accelerations. It would appear therefore that it is more difficult to reverse the vestibular damage which occurs in this disease than the cochlear damage or alternatively that betahistine hydrochloride is less effective in restoring the micro-circulatory changes in the vestibule than in the cochlea. At no time during the 2J years of these studies were any side effects of treatment with either placebo or betahistine observed although we have occasionally encountered gastric discomfort since we started using this drug in 1969. Acknowledgements
We wish to thank Mr. R. H. Allen, Chief Audiology Technician, Tyrone County Hospital and Mr. A. Heyting, Head of the Department of Statistics, Philips Duphar, without whose expert help these studies could not have been completed and the results analyzed. REFERENCES CAWTHORNE, T. (1964) In Neurological Aspects of Auditory and Vestibular Disorders.
Edited by Fields, W. S., and Alford, B. R. Published by C. C. Thomas, Springfield, Illinois, p. 271. ELIA, J. C. (1966) Journal of the American Medical Association, 196, 187. HICKS, J. J., HICKS, J. N., and COOLEY, H. N. (1967) Archives of Otolaryngology
{Chicago), 86, 610. HINCHCLIFFE, R. (i972) Ada Oto-laryngologica, Supplement 305, 10. MARTINEZ, D. M. (1972) Ada Oto-laryngologica, Supplement 305, 29. SCHAYER, R. W. (1964) Annals of the New York Academy of Sciences, 116, 891. SEYMOUR, J. C. (1960a) Journal of Laryngology and Otology, 74, 133. (19606) Journal of Laryngology and Otology, 74, 599. WILMOT, T. J. (1970) Journal of Laryngology and Otology, 84, 1033. (1971) Journal of Laryngology and Otology, 85, 369. ( I 974) Proceedings of the Royal Society of Medicine, 67, No. 5, 331.
840
THE effects of betahistine hydrochloride (Sere) on the clinical features of Meniere's disease were assessed in two double-blind, placebo-controlled, cross-over, clinical studies. The diagnosis was based on a peripheral, fluctuating, recruiting, cochlear (sensorineural) deafness in one or both ears, tinnitus (usually of low tone) and paroxysmal attacks of rotational vertigo. Appropriate auditory and vestibular analyses confirmed the diagnosis. Twenty-four patients were admitted to the studies after careful screening over two-and-a-half years. Twenty-two patients completed the studies, ten of whom received betahistine and placebo for eight weeks each whereas the remaining twelve were given betahistine and placebo for twelve weeks each. The dose of betahistine was the same (16 mg. t.i.d.) in both studies. Daily symptom score cards kept by all patients throughout the studies showed a statistically significant preference for betahistine over placebo with regard to vertigo (p= 0-025), tinnitus (p = o-oio) and fullness of the ear (p = 0-036). Symptom scores of deafness and vomiting indicated trends in favour of betahistine but these did not attain statistical significance. Objective measurements of deafness (mean db. loss), however, showed a highly significant improvement in favour of betahistine when compared with placebo (p 10 (= Sere markedly better)
: 9
Number of patients (P-S)^—10 (= Placebo markedly better)
: 0
Number of patients—io< (P-S)< 10 8 (= no important change)
P- value, one-sided Sign test of marked changes
837
0-004
T. J. Wilmot and G. N. Menon TABLE I I I B . MEAN AND STANDARD DEVIATIONS OF HEARING LOSS IN PATIENTS WITH BILATERAL HEARING LOSS
Hearing loss (dB) (n = 5) Standard deviation
Mean
Sere (S)
Number of patients (P-S) > 10 in both ears : 4 (= Sere markedly better)
—left —right
13-2 12-6
io-6 7-1
Placebo (P)—left —right
30-2 27-4
21 -i "•3
Number of patients (P—S)< —10 in both ears : 0 (-)- Placebo markedly better) Number of patients—io< (P-S)< 10 : 1 in both ears (no important change)
P-value, one-sided Sign test of marked changes 0 • 06
Sign Test (p = o • 004) and a significant trend in favour of betahistine in patients with bilateral deafness (p = 0-06). 4. Objective evaluation of vestibular function While clinical balance tests, positional, optokinetic, caloric and rotation tests were performed on all patients before the studies (pretreatment), at the cross-over of the treatments and at the end of the studies, the only test results which were studied in depth were the rotation tests. No clinical alteration in the other tests were noted and the indelicacy of the stimuli applied made detailed observation unnecessary. The results of the rotation tests are expressed in Table IV TABLE IV. OBJECTIVE EVALUATION OF VESTIBULAR FUNCTION EXPRESSED AS AVERAGE EYE SHIFT PER SECOND
Rotation test results (n = 22) Pre-treatment (Pre) Mean
Sere (S)
SD
Mean
SD
Placebo (P) SD Mean
°/sec*—left
3-2
2-9
2-8
2 -2
33
2-5
3 "/sec—left
8-3
5-1 6-2
7'3
4-0
7-2
13-7 24-0
7-8 13-4
12 -2 22-9
3"5 7-3 n-7
3-6 5-o
2-4 7-1
2-9
8-9
13-7 30-6
1-6 2-7 8-6
I
b "/sec—left
9 0 "/sec*—left 1 "/sec—right*
3 °/sec2—right
6 °/secJ—right 90 °/sec*—right
I
5'5
32-5 3-6 8-9 i4-b 36-0
15-3
15-5
* One missing value; n = 21.
838
6-4 13-3 28-5
i-8 3-1
7-5
15-5
P-value, one-sided (S>P) Wilcoxon signed rank test 0-84 0-44 O-22 0-44 o-8i 0-28 0-15 0-18
Betahistine in Meniere's disease in terms of mean values, standard deviations and p-values as calculated by using the Wilcoxon Signed Rank Test. To summarize, the patients' symptoms as assessed by themselves and by the investigator were improved by betahistine hydrochloride. Objective assessment of hearing was also in favour of betahistine while the attempted objective assesment of vestibular function showed no important improvement with either betahistine or placebo. Discussion The assessment of a multi-factorial problem such as Meniere's disease is fraught with difficulty. In the first place the severity of the disease is extremely variable and the degree of involvement of the cochlea and vestibule is frequently not parallel. As examples, there may be unilateral cochlear involvement with evidence of bilateral loss of vestibular function, or bilateral cochlear involvement with retention of useful vestibular function, or slight cochlear involvement and severe vestibular hypofunction. Many other combinations may occur. Secondly, study of the natural history of the disease shows that natural remissions occur and that these may be entirely independent of treatment. It is therefore against this physical complexity that any drug must first be judged. Thirdly, the psychosomatic background of many cases of Meniere's disease suggests that, if not actually involved in the causation of the disorder (which we believe to be the case), it is of considerable importance in initiating recurrences. For the above reasons, it was important to provide in our two studies a constant background of medical support for each individual over a prolonged period of time—16 weeks for one group of cases and 24 weeks for the other—and to ensure that this support was not conditioned by any knowledge on the part of the investigator as to whether the patient was on drug or placebo at any particular moment. This, and the performance of routine time-consuming medical and technical examinations, posed a considerable burden on the investigator and his chief technician, who were individually responsible for these aspects and who had to arrange their holidays and schedules to ensure no disruption over a period of z\ years. It was felt that a study of small numbers of cases in depth under personal supervision of a single investigator might be more informative than that of a larger number by multiple observers. Lastly, the ethical problems involved in a long-term evaluation of this nature are not insignificant and were carefully weighed before the studies commenced. The results of the studies suggest that during the period of 2\ years when patients were being treated on a randomized basis, the majority 839
T. J. Wilmot and G. N. Menon of patients were symptomatically better and had improved hearing during the time they were on betahistine hydrochloride. On the other hand the carefully repeated battery of vestibular tests done under constant conditions, failed to show any convincing evidence of objective improvement in vestibular function. In particular, the detailed and extensive rotational testing programme which was designed expressly for this purpose also failed in this respect. While we had not expected caloric testing to be delicate enough to show changes in function we had hoped that the same would not apply to more delicate stimuli such as rotational accelerations. It would appear therefore that it is more difficult to reverse the vestibular damage which occurs in this disease than the cochlear damage or alternatively that betahistine hydrochloride is less effective in restoring the micro-circulatory changes in the vestibule than in the cochlea. At no time during the 2J years of these studies were any side effects of treatment with either placebo or betahistine observed although we have occasionally encountered gastric discomfort since we started using this drug in 1969. Acknowledgements
We wish to thank Mr. R. H. Allen, Chief Audiology Technician, Tyrone County Hospital and Mr. A. Heyting, Head of the Department of Statistics, Philips Duphar, without whose expert help these studies could not have been completed and the results analyzed. REFERENCES CAWTHORNE, T. (1964) In Neurological Aspects of Auditory and Vestibular Disorders.
Edited by Fields, W. S., and Alford, B. R. Published by C. C. Thomas, Springfield, Illinois, p. 271. ELIA, J. C. (1966) Journal of the American Medical Association, 196, 187. HICKS, J. J., HICKS, J. N., and COOLEY, H. N. (1967) Archives of Otolaryngology
{Chicago), 86, 610. HINCHCLIFFE, R. (i972) Ada Oto-laryngologica, Supplement 305, 10. MARTINEZ, D. M. (1972) Ada Oto-laryngologica, Supplement 305, 29. SCHAYER, R. W. (1964) Annals of the New York Academy of Sciences, 116, 891. SEYMOUR, J. C. (1960a) Journal of Laryngology and Otology, 74, 133. (19606) Journal of Laryngology and Otology, 74, 599. WILMOT, T. J. (1970) Journal of Laryngology and Otology, 84, 1033. (1971) Journal of Laryngology and Otology, 85, 369. ( I 974) Proceedings of the Royal Society of Medicine, 67, No. 5, 331.
840
