A Multicenter Randomized Double-Blind Study of Ginkgo Biloba Extract Versus Placebo in the Treatment of Tinnitus
Service d’Oto-Rhino Laryngologie, Hospital Saint-Antoine, F 75012 Paris
This important multicenter study of 103 tinnitus out-patients during a 13-month treatment period was carried out by ten E.N.T. specialists, using the double blind, drug versus placebo method. The results were conclusive as regards the effectiveness of Ginkgo Biloba Extract and made it possible to determine the prognostic value of different parameters. Of special importance among these parameters were site and periodicity of the disease. However, the Ginkgo Biloba Extract treatment improved the condition of all the tinnitus patients, irrespective of the prognostic factor.
Tinnitus, a very frequent symptom, troublesome both for the patient and for the physician who has difficulty-when he is able to do so-in being explicit about the physiopathological mechanism and lacks certainty as to what therapeutic posture to adopt.
The fact that tinnitus is relatively benign when present alone explains why it is encountered most frequently in everyday practice. In order to study tinnitus cases of recent onset, that is, less than a year old, it was necessary to carry out a private study, despite the complexity of organization.
This study had the purpose of qualifying the efficacy of treatment of recent tinnitus with Ginkgo Biloba Extract (EGb 761, Rokan, Intersan), a product widely prescribed in this indication because of the etiopathogenic hypothesis considered. Ginkgo Biloba Extract (EGb) has been found active in ischemic processes, in membral exchanges, and on certain neurotransmitters. This study also sought to demonstrate prognostic factors making it possible to determine profiles for patients responding to treatment.
*Rokan (Ginkgo Biloba). Recent results in Pharmacology and Clinic Edited by E.W. Funfgeld Springer-Verlag Berlin Heidelberg New York 1998, pp 245-250
Presse Med 1986 (in French)
Sept 25:15(31) 1562-4
Ten otorhinolaryngology specialists in private practice participate in the study and contributed 103 case histories in 12 months. After randomization, the patients were treated in a double-blind for three months either with EGb or with a placebo strictly identical in appearance. The dosage was 4 mL/day in two doses. The duration of treatment could not be less than one month in order to draw conclusions as to a possible therapeutic failure.
The patients included had recent tinnitus, that is, appearing less than a year earlier. Excluded from the study were patients suitable for surgical or anti-infectious treatment, those suffering with acute inner or middle ear infections (otitis, Eustachian salpingitis), or associated pathological disorders capable of skewing analysis of results.
Also excluded were patients already under treatment and for whom the evolution was favorable, as well as those who required medical treatment which might interfere with evaluation of the results.
The criteria for establishment of the diagnosis and for monitoring were essentially clinical and comprised the definition of the patients and description of the tinnitus: timbre, schedule, rhythm, impairment, intensity, overall evolution, and time before disappearance; other associated ORL symptoms such as vertigo and hypoacusis were routinely investigated.
The paraclinical examinations comprised essentially tonal audiometry performed at the time the diagnosis was given and possible repeated on the basis of the clinical condition. The performance of other tests (impedancemetry, electronystagmography) were left to the discretion of the physician.
Statistical analysis was by means of the X test for the qualitative variables and the Student t-test for the quantitative variables. In order to evaluate the influence of the various parameters on the evolution, a step-by-step decreasing logistic regression was done.2
Comparability of the groups at the time of admission.
After breaking of the double-blind, both treatment groups were analyzed in order to very comparability at the time of admission to the study; 161 variables were analyzed.
The mean value of severity, graded from 1 to 4, was identical in both groups, but distribution by class presented a slight difference. This corresponds to a broader distribution within the placebo group, and more centered around the values of 2 and 3 in the EGb group. Therefore one group was not more severely affected than the other.
The intensity of the tinnitus, graded from 0 to 3, was comparable in the two groups. Impairment, coded from 0 to 3, was also comparable in both groups (Table I).
Comparison of the groups at the time of admission
Ginkgo Biloba Extract
3 degrees of freedom
3 degrees of freedom
|X2=0.5782 degrees of freedom|
|Both groups were also comparable for length of time the syndrome had been present, age and sex: -time syndrome had been present: EGb, 127 days; placebo, 143 days; t=0.67; -age: EGb, 50.97 years; placebo, 49.76 years; t=0.39; -sex: EGb, 29 men, 29 women; placebo, 25 men, 20 women (X2=0.31-1 degree of freedom).|
Ginkgo Biloba Extract
|1-2=exacerbation; 3=no change; 4=slightly improved; 5=improved; 6=much improved; 7=suspension because of intolerance.|
Percent of Good Evolution
Ginkgo Biloba Extract
|Class 1, old, bilateral, intermittentClass 2, recent, unilateral, permanentClass 3, recent, bilateral, permanentClass 4, old, unilateral, permanent|
Class 5, old, bilateral, permanent
There was also no difference with respect to the consumption of tobacco or alcohol, the ORL history (effect on the middle ear, barotraumas, phonotrauma, ototoxic involvement), and the associated syndromes (vertigo, hypoacusis).
Analysis of the tonal audiometry performed at the time of admission to the study also showed no difference between the two groups.
Analysis of therapeutic efficacy
Overall evolution: comparison of the two groups showed a significant difference in favor of the EGb group (X2=4.44-2 degrees of freedom, P=0.05)(Table 2).
Time before disappearance, or distinct improvement: there is a statistically significant difference in favor of the group treated with EGb, the evolution of which was much faster (X2=3.90-1 degree of freedom, unilateral test P=0.03).
Thus, the median, that is to say the time before disappearance or distinct improvement in 50% of the tinnitus cases, was 70 days in EGb group and 119 days in the placebo group.
Evolution of intensity of the tinnitus between the first and last consultations: this appeared to be statistically better in the EGb group. Difference in intensity: EGb=1.00; placebo=0.67; Student t-test=1.79; unilateral P test=0.03.
Decrease in impairment between the first and the last consultations: this was greater for the treated group (0.84) than for the control group (0.59); this difference is at the limit of significance (Student t-test=1.38; unilateral P test=0.08).
Study of the prognostic variables
Statistical analysis showed that three variables have prognostic value: the history of the disease (more than 30 days, less than 30 days), the site (unilateral, bilateral), and periodicity (permanent, intermittent). Study of the links among these three variables made it possible to demonstrate five different prognostic classes (Table 3).
This multicenter study, performed by ORL specialists in private practice, with a rigorous methodology, despite the difficulties which this includes in a field in which the subjective component is significant, as it confirms the results of treatment with placebo, brought together 103 case histories of recent tinnitus.
It was thus possible to meet the two objectives established:
- to verify the efficacy of treatment with EGb;1
- to analyze the prognostic value of the various parameters by demonstrating those whose influence was determinant (history, site and periodicity), as well as to specify the modalities of their interaction.
Tinnitus characterized as recent, unilateral and intermittent, has a good prognosis. Treatment with EGb improves the evolution of the tinnitus regardless of its prognostic factors.
- 1 – Meyer B: Etude multicentrique des acouphenes, epidemiologie et therapeutique. Annal. ORL 1986, 103, 185-188.
- 2 – Nakache J.P., Guegen A., Pierart H.: Utilisation du modele logistique dans l’etude de l’influence des variables initiales et du traitement sur l’evolution de l’acouphene. Revue de Statistique Appliquee (sous presse).