Multicenter, Open-Label, Prospective Study Shows Safety and Therapeutic Benefits of a Defined Ginkgo Biloba Extract for Adults with Major Neurocognitive Disorder

Major neurocognitive disorder, also known as dementia, is a devastating neurological disease. Affected patients undergo a progressive cognitive decline that affects a broad range of their abilities such as memory, executive function, language, attention, or spatial vision. This decline greatly impacts their daily activities and social life [1]. Major neurocognitive disorder is differentiated based on its causes [2]. Pure Alzheimer’s disease (AD) or pure vascular disease are rarely seen in neuropathologic studies, whereas mixed forms are often observed [3].

Since aging is the most important risk factor, the prevalence of major neurocognitive disorder increases as the worldwide population is aging [4, 5]. In India, it is estimated that people over 60 years old will represent almost 20% of the population by 2050, accounting for 316 million individuals. The rise in life expectancy among the Indian population will probably lead to a dramatic increase in the prevalence of major neurocognitive disorder as well as to significant public health concerns. The “Dementia in India 2020 report” calculated that 5.3 million Indians over 60 years of age had this disease in 2020, and this number is projected to rise above 14 million by 2050 [6]. The growing prevalence of major neurocognitive disorder entails an increasing demand for treatments, rehabilitation, and support services for patients and caregivers [5]. Caregivers often need to abandon the life they have been used to, for instance, to help patients with their daily activities and to liaison with social services and the healthcare system [7]. Thus, the burden of caregiving also encompasses physical, emotional, and financial issues for families or relatives [1].

The proprietary extract EGb 761^® is the dry extract manufactured from the leaves of the tree Ginkgo biloba L. The production process is defined in the European Pharmacopoeia. The final preparation must contain 22.0–27.0% ginkgo flavonoids and 5.4–6.6% terpene lactones consisting of 2.8–3.4% ginkgolides A, B, C, and 2.6–3.2% bilobalide. The content of biologically active proanthocyanidins was found to be approximately 7% [8]. The rigorously controlled production process ensures the very high batch-to-batch consistency of the herbal drug and its reliable, reproducible pharmacokinetic and pharmacodynamic properties which are the basis for its clinical efficacy and safety [9]. In contrast, unregulated products such as food supplements are not subject to strict quality control procedures. They were shown to contain only low amounts of flavone glycosides or terpene lactones, and a significant number of products were adulterated [10].

EGb 761^® has been proven effective in the symptomatic treatment of major neurocognitive disorder, as several meta-analyses [11‒14] have shown. In randomized, placebo-controlled trials, it improved cognitive performance, neuropsychiatric symptoms, activities of daily living, and independent clinicians’ global rating of change in overall condition [15]. EGb 761^® treatment also had beneficial effects on caregiver distress related to the patients’ neuropsychiatric symptoms [14, 15]. The extract targets various pathophysiological mechanisms related to AD and cerebrovascular disease, including antioxidant, neuroprotective, and perfusion-enhancing effects; improvement of mitochondrial function; and stimulation of neurogenesis and synaptogenesis [16, 17].

Many older adults also experience dizziness and/or tinnitus. In studies of EGb 761 treatment in major neurocognitive disorder, prevalence rates between 13% and 77% for dizziness and between 13% and 52% for tinnitus were found. As shown in a meta-analysis of these placebo-controlled studies, both symptoms were significantly improved by EGb 761^® treatment [18]. Another meta-analysis showed that the incidence of both symptoms during the treatment period was lower in the EGb 761-treated patients than in the placebo group; the odds ratios were 0.50 for dizziness and 0.38 for tinnitus [12].

So far, no clinical studies with EGb 761^® have been conducted in Indian settings. This study assessed the safety, tolerability, and treatment effects in the Indian population.

This open-label, multicenter, single-arm phase IV trial enrolled patients with major neurocognitive disorder at 12 clinical sites in India. After obtaining written informed consent, we screened 157 patients for eligibility according to inclusion and exclusion criteria on visit 1 and enrolled 150 patients. Participants took one film-coated tablet containing 120 mg EGb 761^® (Tebokan^® forte; Dr. Willmar Schwabe GmbH & Co. KG, Karlsruhe, Germany) twice daily for 18 weeks. The treatment was initiated on trial visit 2 (baseline, day 0). The safety and treatment effects were subsequently evaluated at visits 3 (week 4, day 28), 4 (week 8, day 56), 5 (week 12, day 84), and 6 (week 18, day 126, end-of-trial visit). A caregiver accompanied the participants during the trial as part of the study prerequisites for the patients’ safety and care. Those who discontinued the trial early or were withdrawn from the trial after allocation were not replaced. An overview of the patient distribution is provided in Figure 1.

This study followed the protocols of the Declaration of Helsinki and the International Council for Harmonization Consolidated Guideline E6 for Good Clinical Practice (CPMP/ICH/135/95) as well as applicable laws. Before enrollment, the patients were provided with adequate verbal and written information about the nature, purpose, and possible risks and benefits of the trial. Written informed consent was obtained from all the study participants or their legally authorized representatives. Ethics Committee approvals were obtained for all study sites. The trial was prospectively registered in the Clinical Trials Registry of India (trial No. CTRI/2017/02/007764).

Participants were eligible if they were at least 50 years old; had a diagnosis of major neurocognitive disorder (due to AD, major vascular neurocognitive disorder, or a mixed form of both) according to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) criteria; and had a Mini-Mental State Examination (MMSE) [19] score of ≥12 and ≤24 indicating mild-to-moderate intensity. The diagnosis was based on clinical assessment, cognitive testing, and magnetic resonance imaging. The latter provided essential information about the etiology of the disease. The patients or their legal representatives had to give their informed consent and agree to follow all trial procedures; they also had to consent to the possibility of using the support of their caregivers.

Patients with any of the following ongoing conditions were excluded from the study: diagnosis of a different type of major neurocognitive disorder, major depression or other psychiatric or neurological disorders; active malignant disease; infection with the human immunodeficiency virus; any other serious infection; a severe or insufficiently controlled cardiovascular, renal, or hepatic disorder; diabetes; anemia; thyroid dysfunction; and a history of any bleeding abnormalities (other than in the context of antiplatelet or anticoagulant drugs). Further criteria for exclusion were the use of Gingko biloba-based medications within 8 weeks prior to baseline; use of cognitive enhancers; antidepressants; cholinergic, anticholinergic, or hemorheology influencing drugs; use of cholinesterase inhibitors within 6 months prior to baseline; hospitalization for major depression within 1 year prior to baseline; or participation in any other clinical trial in the previous 6 months.

For the statistical analysis, we defined two different datasets. The safety analysis set included all participants (n = 150) who took at least one dose of EGb 761^®. The effectiveness analysis set included all patients (n = 146) who took at least one dose of EGb 761^® and had at least measurement of one therapeutic effect parameter during the active treatment phase. The latter was chosen since it captures the benefits observed in real-world practice. This is more relevant for the practitioners than the population without significant protocol deviations.

The primary outcomes were the number of adverse events (AEs) and the proportion of patients who experienced AEs, including presumed adverse drug reactions and serious adverse events. The safety assessments included evaluating changes in vital signs as well as physical and laboratory examinations between baseline and week 18.

The secondary outcomes included the change from baseline to week 18 on various cognitive scales. The instruments used were the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) constructional praxis (CP) [20] and recall of CP [21] which is a picture-learning test assessing attention and memory; the Trail-Making Test (TMT) A and B evaluating speed and executive functioning; the number and percentage of patients with improvement in cognitive functioning per TMT or CERAD or both; the Behavioral Pathology in Alzheimer’s Disease (BEHAVE-AD) rating scale [22] total score and individual scores for seven items (paranoid and delusional ideation, hallucinations, activity disturbances, aggressiveness, diurnal rhythm disturbances, affective disturbances, and anxieties and phobias), which represents an assessment of the patient’s behavior; the Clinical Global Impressions (CGI) Scale [23] describing the global improvement; and the number and percentage of patients who improved or recovered from concomitant vertigo and/or tinnitus according to 11-point box scales [24‒26].

All AEs were recorded during the period between signing the informed consent and 2 days after the last dose of study medication. This took place regardless of whether the investigator observed an AE or the patient reported it and whether the AE was attributed to the study drug. Any event was reported and managed as a potential adverse drug reaction which occurred during the treatment period or up to 2 days after the last dose and which the investigators assessed as possibly related to the study drug.

A physician performed a physical examination of the patient which included a thorough review of all body systems. Data were collected and documented at screening and end-of-trial visits. The vital signs, including pulse, body temperature (oral), respiration rate, and blood pressure, were collected and documented at each trial visit. Safety-related laboratory tests included blood count; coagulation tests (partial thromboplastin time and international normalized ratio); renal and liver function tests; and serum chemistry. Serological tests were performed for human chorionic gonadotropin and human immunodeficiency virus at the screening visit. The evaluation of therapeutic effects included the assessment of the CERAD (CP and recall of CP), TMT (A and B), CGI (severity at baseline, global improvement from week four onward), and BEHAVE-AD at all trial visits. Furthermore, 11-point box scales for vertigo and tinnitus were used to assess patients who had reported such symptoms at baseline.

A sample size of 150 patients was large enough to analyze differences between the baseline and the end of the trial in vital signs, laboratory values, and treatment effects. For this sample size, the probability of observing AEs that occur with a frequency ≥2% is higher than 90%. When single items or total scores of effect measures were missing at baseline, they were replaced by the median value of all patients without missing values. When data were missing during the treatment period, these were replaced by the last observation carried forward method. An exception was made for missing values for the 11-point box scales and the TMT. If a patient was not able to count loudly from 1 to 25, the TMT was not applied at all. If a patient was not able to say the first 25 letters of the alphabet, he/she was asked to do only the TMT A. In every other case, both TMT A and B were applied. As this rule justified most missing values for the TMT, no replacement strategy was applied.

All statistical analyses were performed using SAS^® software (version 9.2; SAS Institute Inc., Cary, USA) on a MS Windows platform. Continuously distributed data were analyzed using the following descriptive statistics: number of patients (n), mean, standard deviation (SD), first quartile (Q₁), median, third quartile (Q₃), minimum, maximum, and two-sided 95% confidence intervals (CI). All p values of the paired t tests were two-sided at the 5% significance level for a change from baseline. For categorical data, the number and percentage of patients were calculated.

All 150 patients were Indians: 85 (56.7%) participants were men, and 65 (43.3%) were women. Their mean age was 62.8 (SD 8.34) years. The mean overall treatment compliance, calculated from pill counts, was 99.2% (95% CI, 99.0–99.4%).

In accordance with clinical criteria and from the findings of brain magnetic resonance imaging scans, major neurocognitive disorder was due to AD in 21.3% of the patients, cerebrovascular disease in 48.0%, and mixed pathology in 30.7%. The MMSE scores ranged from 12 to 24 at enrollment; the mean (SD) and median scores were 19.3 (2.8) and 20.0, respectively.

The occurrence of AEs was assessed in the safety set (n = 150). During the treatment period (from baseline to week 18), 45 AEs occurred in 33 (22.0%) patients, including 43 nonserious AEs of mild-to-moderate severity and 2 SAEs. The most frequently reported AEs (at least 2 cases) were headache (8 AEs in 8 [5.3%] patients), diarrhea (5 AEs in 4 [2.7%] patients), pyrexia and vomiting (4 AEs in 4 [2.7%] patients each), decreased blood pressure, dyslipidemia, and arthralgia (2 AEs in 2 [1.3%] patients each). Among the AEs, ten presumed ADRs were reported in 9 (6.0%) patients including 3 cases of headache, 2 cases of diarrhea, 2 cases of dyslipidemia, 1 of blurred vision, 1 of pain, and 1 of pruritic rash. However, the causal relationship of EGb 761^® with those ADRs was deemed unlikely. The two nonfatal SAEs were an exacerbation of preexisting chronic obstructive pulmonary disease (1 patient, 0.7%) and angina pectoris (1 patient, 0.7%). However, these were determined to be unrelated to the trial drug. Three AEs of mild intensity (increased blood glucose, increased blood creatinine, decreased hemoglobin) were reported for 1 patient each (0.7%) during the posttreatment exposure phase. No significant difference in laboratory assessments and physical examinations was found between baseline and week 18.

The treatment effects were assessed in the effectiveness set and reported as a change from baseline to the end of study treatment at week 18 (n = 146).

Table 1 and Figure 2 show that the mean CERAD CP scores and the mean CERAD recall of CP scores significantly increased from baseline to week 18 by 1.2 and 1.8, respectively. The increase in CERAD CP and CERAD recall of CP scores was statistically significant at all visits compared to baseline.

At week 18, the mean time to complete part A of the TMT showed a significant reduction of 21.6 s from baseline. Compared to the baseline, the mean time to complete part A of the TMT significantly decreased across all visits. The mean time to complete part B of the TMT decreased by 27.8 s from baseline to week 18. These results are displayed in Table 1 and Figure 3.

One hundred forty-one (96.6%) patients showed improved cognitive functioning as evaluated by TMT or CERAD, whereas improvements in both CERAD and TMT scores were observed in 99 (67.8%) patients.

The mean BEHAVE-AD symptomatology total score gradually decreased from baseline, reaching a reduction of 5.5 points at week 18 as displayed in Figure 4. Compared to baseline, the mean BEHAVE-AD symptomatology total score improvements were significant at all visits. The mean BEHAVE-AD symptomatology scores for individual items showed gradual improvements for all of them as shown in Figure 4. These improvements were significant at all trial visits for the individual items of the BEHAVE-AD symptomatology scores of paranoid delusional ideation, activity disturbances, hallucinations, aggressiveness, and affective disturbances. For the domain of diurnal rhythm disturbances, the difference was significant at all visits except for visit three on day 28.

In the effectiveness set, the mean scores on the CGI scale decreased from 3.2 at week 4 to 2.4 at week 18 (Table 1). At week 4, 1 (0.7%) patient showed a “very much improved” rating, and 14 (9.6%) patients were rated as “much improved.” The number of patients reported as “very much improved” and “much improved” increased to 42 (28.8%) and 34 (23.3%) by week 18, respectively, as displayed in Figure 5.

Two patients reported tinnitus at the screening with a mean severity score at baseline of 3.5 on an 11-point box scale. At the end of the trial, the score reported by only 1 of the patients was 3.0 (Table 1). The mean score on an 11-point box scale for vertigo was 4.4 at baseline for the 15 participants who reported vertigo and decreased to 3.7 at week 18 for all 14 patients with vertigo still available for analysis (Table 1).

This phase IV study assessed the safety and therapeutic effects of EGb 761^® at a daily dose of 240 mg (120 mg twice daily) in Indian patients with major neurocognitive disorder. 22.7% of the patients reported mild-to-moderate AEs during the 18-week active treatment period and the subsequent 2-day washout phase. Two SAEs were assessed as unrelated to the study drug, especially since both were exacerbations of preexisting somatic diseases. The AEs reported most frequently (headache, diarrhea) were regarded to be possible side effects of EGb 761^®.

Two meta-analyses were conducted of randomized, placebo-controlled trials involving treatment with Ginkgo biloba: one involved 9 trials in 2,561 patients with major neurocognitive disorder or mild cognitive impairment, and one reviewed 44 trials in 6,415 patients with a variety of diseases. In the absence of a placebo control in our study, it is noteworthy that these two meta-analyses suggest that the rates of headache and diarrhea we observed were not significantly higher in patients treated with EGb 761^® than in those taking placebo [13, 27]. No AEs suggestive of previously unknown drug risks were documented in our study. Thus, the results confirm the safety profile presented in the patient information leaflet of Tebokan^® forte registered in India.

Following 18 weeks of treatment with EGb 761^®, significant improvement in cognitive functioning and behavioral changes were observed. These changes were consistent with the clinicians’ global assessments. Results for subgroups defined by age and gender showed no apparent differences.

A limitation of the study is that the design did not include a placebo group. In general, studies without control groups have inherent biases; thus, the treatment effect is difficult to assess. However, the findings from the present study are clearly in line with those of previous studies. Several trials in patients with major neurocognitive disorder found significant improvements under treatment with EGb 761^® in, e.g., Short Cognitive Performance Test (Syndrom-Kurztest [SKT]), Alzheimer’s Disease Assessment Scale – cognitive subscale (ADAS-cog), Verbal Fluency Test, Clock-Drawing Test. Some of the tests include assessments of memory (e.g., SKT, ADAS-cog) and some include aspects of constructional praxis (ADAS-cog, Clock-Drawing Test) [28‒31]. The SKT Short Cognitive Performance Test, which had been applied in most clinical trials of EGb 761^® in dementia, has been validated in many countries, cultures, and languages, but so far not in India. The ADAS-cog, which was widely used in anti-dementia drug trials and in a few trials of EGb 761^®, uses word lists for the testing of memory. In our study, participants from various parts of India, with different primary languages, were included. Therefore, the CERAD subtests Constructional Praxis and recall of Constructional Praxis were chosen to avoid the risk of increased variability of results in memory tests due to different linguistic backgrounds. These are nonverbal tests with components of memory and attention, as well as visuospatial abilities and constructional praxis. CP and recall of CP of the CERAD cognitive battery belong to the standard test batteries for clinical trials in major neurocognitive disorder. They were complemented by the Trail-Making Test Forms A and B tapping the domains information processing speed and attention, as well as executive function.

Unfortunately, only few studies with drugs for the treatment of major neurocognitive disorder have been conducted in India so far, and even fewer have been conducted using plant-based medicines. To our knowledge, this is the first study to investigate EGb 761^®’s safety and therapeutic effects in India. Further limitations of this study include the relatively small sample size and the fact that not all geographical regions of India were represented. These limitations should be taken into consideration when interpreting the results. However, the findings of this study are in line with the comprehensive clinical database of several 1,000 patients. They do not reveal unexpected findings requiring further research.

The average improvement in the TMT A (speed of processing) compares well to that found in a 24-week study of the same daily dose of EGb 761^® conducted by Gavrilova et al. in patients with mild neurocognitive disorder [32]. The improvement in the more demanding TMT B (executive functioning) was, however, less pronounced than in the previous study.

The cognitive testing with the same test methods was repeated every 4 weeks. In theory, this might have contributed to a practice effect in patients. In the absence of a placebo control, the role of learning effects to the changes in cognitive test scores cannot be ruled out with certainty. For the TMT, Gavrilova et al. found practice effects in patients with MCI, but Duff et al. did not [33]. In cognitively healthy people who later progressed to MCI or major neurocognitive disorder, no practice effects were seen in various tests for non-memory domains [34]. In earlier trials of EGb 761^® in patients with major neurocognitive disorder, no learning effects were detected in any cognitive test. It is therefore unlikely that training contributed to a relevant extent to the results of the cognitive tests applied in our study.

In the present study, the proportion of patients rated as “very much improved” or “much improved” on the CGI scale at the end of treatment (week 18) was 52.1%. This is higher than the proportions of patients with major neurocognitive disorder considered to be “much improved” or “very much improved” upon EGb 761^® treatment in the trials by Kanowski et al. [35] or Herrschaft et al. [31], which were slightly above 30%. Major clear-cut improvements in clinicians’ global assessments support the clinical relevance of changes in cognitive performance, as assessed by neuropsychological tests or test batteries.

The presence of clinically significant behavioral symptoms was acceptable, but not a prerequisite for enrollment. Consequently, the average severity of these symptoms was rather mild. Nevertheless, the patients’ behavioral symptoms (as reviewed by the BEHAVE-AD) improved significantly under treatment with EGb 761. This corresponds to observations from other trials in patients with major neurocognitive disorder in which symptom severity and caregiver distress improved considerably as assessed by the Neuropsychiatric Inventory.

Tinnitus and vertigo are frequently diagnosed in older adults, especially those with major neurocognitive disorder. EGb 761^® had shown beneficial effects on these symptoms in a meta-analysis [18]. Therefore, we applied 11-point box scales to assess tinnitus severity, vertigo, and treatment-related changes. The mean decrease from baseline for vertigo of 0.8 on the 11-point box scale was the same as the average improvement in former trials in patients with major neurocognitive disorder which applied the same scale. For tinnitus, posttreatment data of only 1 patient were available.

To summarize, in the present study, EGb 761^® improved cognitive function and behavioral symptoms to similar degrees as in earlier studies in which the extract was used at the same daily dose to treat mild cognitive impairment or major neurocognitive disorder. This occurred even though the measures used to evaluate outcomes were different in previous studies than they were in the current study. With its effects on vascular and neurodegenerative pathologies, EGb 761^® is a feasible therapeutic modality for the treatment of most patients with major neurocognitive disorder seen in everyday practice. Its low rate of potentially drug-related AEs, most of which were mild in nature, means the drug can be safely administered to Indian patients.

Our trial confirmed the favorable safety profile and the therapeutic benefits of EGb 761^® in Indian patients diagnosed with major neurocognitive disorder after 18 weeks of treatment. The findings align with previous controlled studies and real-world data from other countries.

The authors thank the patients who participated in the clinical trial. Medical writing support was provided by Patricia Buchholz, Angela Garcia-Perez, and Małgorzata Biernikiewicz of Valid Insight, London, UK, following Good Publication Practice (GPP3) guidelines. The trial, medical writing services, and publication were funded by Dr. Willmar Schwabe GmbH & Co. KG, Karlsruhe, Germany.

This study protocol was reviewed and approved by the Institutional Ethics Committee of Maulana Azad Medical College and Associated Lok Nayak Hospital (Reg. No. ECR/329/Inst/DL/2013/RR-16), Approval No. F.1/IEC/MAMC/(55/4/2016/No. 50 dated 21/3/17 for DC. Each of the 12 participating study centers received approval from their responsible Institutional Ethics Committee. Written informed consent was obtained from all participants before enrolling in the study.

The trial was conducted in India, and D.C. received grants for the conduct of the trial and honoraria from Dr. Willmar Schwabe GmbH & Co. KG, Karlsruhe, Germany. R.H. is an employee of Dr. Willmar Schwabe GmbH & Co. KG, Karlsruhe, Germany. A.K.R., V.R.R., Y.K.G., and P.N. have no conflicts of interest to declare.

Dr. Willmar Schwabe GmbH & Co. KG, Karlsruhe, Germany, sponsored the research and funded medical writing support and publication. The authors declare that this study received funding from Dr. Willmar Schwabe GmbH & Co. KG, Karlsruhe, Germany. The funder was involved in the study design, analysis, and preparation of the manuscript. Editorial/medical writing support was provided by Valid Insight, London, UK, and was funded by Dr. Willmar Schwabe GmbH & Co. KG. The final decision to submit the manuscript for publication remained with the authors.

D.C. contributed to conceptualizing and designing the study, collecting and interpreting data, and critically reviewing the manuscript. A.K.R. contributed to the collection of data and critically reviewed the manuscript. V.R.R., Y.K.G., and P.N. contributed to data collection and critically reviewed the manuscript. R.H. contributed to conceptualizing and designing the study, interpreting data, and drafting the manuscript. All authors read and approved the final manuscript.

Raw data cannot be shared both due to ethical reasons and to data protection laws. To the extent permitted by law, the trial data required for validation purposes have already been disclosed in result reports about corresponding databases. Reasonable requests to access the datasets should be directed to the corresponding author.