Ginkgo biloba Extract EGb 761 in the Treatment of Patients with Mild Neurocognitive Impairment: A Systematic Review

Introduction

The Ginkgo biloba extract EGb 761 has been assessed in more than 40 clinical trials involving patients with a variety of clinical conditions. Data from older trials in cognitive disorders may be difficult to interpret, because diagnostic terms for cognitive impairment and diagnostic criteria are different from those commonly used today. In this review, we summarize the evidence for clinical efficacy of EGb 761 in mild neurocognitive disorder (mild NCD) while taking into account most recent guideline developments and diagnostic criteria.

Cognitive impairment is a condition frequently seen in the elderly. Detailed diagnostic criteria for dementia syndromes and Alzheimer’s disease (AD) have been available and commonly used since 1980s (Diagnostic and Statistical Manual of Mental Disorders, third edition, DSM-III)¹ (National Institute of Neurological and Communicative Disorders/Alzheimer’s Disease and Related Disorders Association, NINCDS-ADRDA).² On the other hand, until the end of 1990s, a variety of terms were used for milder forms of aging-associated or early cognitive impairment related to a specific pathology (eg amyloid-related early stages of AD). Scientists coined terms such as impairment of cerebral functions, cerebrovascular insufficiency or organic brain syndrome, and they introduced other terms representing more or less different concepts, such as benign senescent forgetfulness,³ age-associated memory impairment (AAMI),⁴ aging-associated cognitive decline (AACD)⁵ and cognitive impairment – no dementia (CIND).⁶ The term “mild cognitive impairment” (MCI), although previously used to denote pre-dementia conditions of cognitive impairment,⁷ became prominent when defined by Petersen et al in the late 1990s.^8,9 The concept of the MCI syndrome and the diagnostic criteria were developed further by Petersen and an international working group^10,11 and later integrated into the modern systematology of the Alzheimer’s disease continuum.¹² By contrast, an International Working Group (IWG) led by B. Dubois referred to the MCI patient population using the label of prodromal AD.^13,14 With the taxonomy of the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) of 2013, the concept of mild neurocognitive disorder (mild NCD) was introduced, with additional criteria for sub-classification by etiology.¹⁵

Cholinesterase inhibitors, which are first-line treatments in Alzheimer’s disease dementia, have shown minor symptomatic effects in patients with MCI,¹⁶ whereas in ApoE4 carriers 3-year treatment with donepezil reduced the risk of progression to dementia.¹⁷ Studies of anti-amyloid substances have yielded more or less inconsistent results.^18,19 Concerns have been expressed with regard to safety and tolerability of such treatments in patients with mild symptoms.^16,19 Patients whose ailments are still mild and not too distressing are likely to be more compliant, if a treatment with proven efficacy is safe and well tolerable.

In neuropathology studies of MCI patients, mixed pathology was observed most frequently, predominantly including features of Alzheimer’s disease and cerebro-vascular lesions.²⁰ This ties in with findings from clinically healthy older adults followed longitudinally in the Harvard Aging Brain Study. Patients with both amyloid pathology on positron emission tomography (PET) scans and a high vascular risk score showed faster cognitive decline than those with a single pathology.²¹ In summary, the view that considers mainly mono-etiological pathologies is outdated. In addition to cerebrovascular pathologies, other neurodegenerative comorbidities (eg synucleinopathies, tauopathies, Lewy bodies, etc.) are also common.²⁰

Ginkgo biloba extract EGb 761 (a proprietary extract of Dr. Willmar Schwabe GmbH & Co. KG) has shown vaso-regulatory, vaso-protective and blood flow-enhancing properties in clinical trials.^22–25 In animal models, it inhibited the formation of toxic amyloid-oligomers, attenuated amyloid toxicity and improved neurogenesis and synaptogenesis.^26–28 It also showed antioxidant activity and protection of mitochondrial function against various toxicities,^23,29 a process that is strongly linked to Alzheimer’s disease pathophysiology.³⁰

Clinical trials assessing the effects of EGb 761 on cognitive functions were conducted long before internationally or even globally recognized terms, concepts and diagnostic criteria for the condition now called MCI or mild NCD existed. Most of these studies demonstrated improvements in cognitive performance that were comparable to later studies, which enrolled patients with criteria-based diagnoses of MCI or dementia. Based on these results, the Herbal Medicinal Products Committee (HMPC) of the European Medicines Agency (EMA) recommended the use of EGb 761 for “the improvement of (age-associated) cognitive impairment and of quality of life in mild dementia.”³¹ Expert consensus papers and practice guidelines of several countries also recommend EGb 761 for the treatment of MCI.^32,33 The beneficial effects of EGb 761 in mild-to-moderate dementia have been summarized by several systematic reviews and meta-analyses.^34,35 However, in older studies of patients with MCI, the variability of previous inclusion criteria and the lack of precisely defined and widely accepted diagnostic procedures were difficult to interpret. With the implementation of the DSM-5 and the retrospective application of its diagnostic criteria,³⁶ a systematic review of the older studies together with those conducted in the early 2000s became possible.

Both the DSM-5³⁷ and the core diagnostic criteria of the National Institute on Aging – Alzheimer’s Association workgroups (NIA-AA)¹² for MCI require the independence in functional abilities to be preserved, although subtle difficulties with complex functional tasks are acceptable. Acknowledging that the clinical characteristics of patients with MCI (due to AD) may overlap with those at the milder end of the (AD) dementia spectrum and that these two stages of the disease are not clearly demarcated, the “Guideline on the clinical investigation of medicines for the treatment of Alzheimer’s disease” issued recently by the European Medicines Agency permits subjects with MCI and mild dementia to be studied together.³⁸ This has, in fact, already been done in trials of anti-amyloid agents.^39,40

Objectives

The objective of this paper is to review the data on efficacy and safety of EGb 761 in the treatment of mild NCD, also encompassing studies dating back to 1980s which enrolled only patients with mild NCD as well as those which admitted patients with mild NCD and mild stages of major NCD. Outcomes of interest are cognitive performance, neuropsychiatric symptoms, geriatric assessments, and global impression of change.

Methods

The Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement^41,42 was taken into account.

Eligibility Criteria

We included randomized, placebo-controlled, double-blind clinical trials in patients diagnosed with mild NCD who were treated with the Ginkgo biloba extract EGb 761 as the only drug treatment for cognitive deficits for at least 8 weeks. Patients had to meet the mild NCD criteria of the DSM-5.³⁷ Studies that enrolled mainly patients with mild NCD but did not strictly exclude mild dementia (ie, mild stage of major NCD) were also included. Studies were only considered eligible if the outcome measures were appropriate to assess therapeutic efficacy in patients with mild NCD, ie, standardized neuropsychological tests, established scales to assess neuropsychiatric symptoms, geriatric rating scales (eg, Sandoz Clinical Assessment Geriatric, SCAG) and/or Clinical Global Impression (CGI) scales.

The following were deemed ineligible: cohort studies, studies assessing primary preventive effects, studies of Ginkgo biloba products not containing EGb 761 as active substance, studies of patients with diagnoses not compatible with mild NCD (eg, aging-associated memory impairment as defined by Crook et al,⁴ subjective cognitive decline), studies of EGb 761 as add-on treatment for cognitive deficits (eg, EGb 761 and cholinesterase inhibitors) or studies in which the outcome measures were not appropriate to assess therapeutic efficacy in patients with mild NCD. We also excluded studies published in languages other than English, French, German, Czech, Spanish, Italian or Portuguese.

Information Sources and Search Strategy

We searched the electronic databases Embase (between 1947 and 12 May 2022), Medline (between 1946 and 06 May 2022) and PubMed (from the beginning until 06 May 2022) for eligible studies. Briefly, the search strategy included the terms “ginkgo” (as well as common abbreviations and brand names for ginkgo extracts), “maidenhair tree”, “mild neurocognitive disorder”/“mild NCD”, “mild cognitive impairment”/“MCI”, “preclinical AD”, “randomized controlled trial” and common variations of these terms. A full search strategy is presented in Online Supplementary Content.

We adopted two further approaches to identify relevant studies: 1) screening the references of the systematic reviews retrieved through the searches above, and 2) screening the references of a former publication,³⁶ in which the diagnoses for inclusion in randomized clinical trials of EGb 761 were retrospectively classified according to the DSM-5.

Selection Process

Two reviewers independently screened the retrieved records but used a common approach: Firstly, they eliminated duplicates by manual screening; secondly, they screened the titles of all articles for relevance, including systematic reviews. If titles did not reveal relevance or irrelevance, abstracts were then screened; if the abstracts did not provide enough information to ascertain eligibility, the reviewers screened the full text. Finally, the reviewers’ ratings of the articles were compared, then disagreements were discussed to reach a consensus.

Data Collection Process, Data Items Collected and Effect Measures

Data were extracted from published study reports as presented. The following outcomes were considered eligible: cognitive function, behavioral and psychological symptoms of dementia (BPSD), global assessments (CGI, geriatric rating scales, dementia rating scales), functional measures (activities of daily living, ADL), quality of life, adverse events (AEs) and serious adverse events (SAEs). We considered changes from baseline (for each group or difference between groups), response rates or both, as reported in the published study reports.

In addition to the outcome measurements, the following data were also collected: author and year of publication, diagnosis according to the DSM-5 and original diagnosis reported in the publication, main inclusion criteria, treatment dose and duration, number of patients included in the analyses.

Due to the heterogeneity of the study designs and outcome measures, the methodological criteria to perform meta-analyses were not met.

Study Risk of Bias Assessment

We used the Jadad score⁴³ to assess the risk of bias in the included studies. This instrument addresses bias in three domains: randomization, blinding and withdrawals/dropouts. Scores can range between 0 (high risk of bias) to 5 (low risk of bias).

Results

Study Selection

Of 395 records identified from databases, 106 duplicates were removed, and 289 records were screened for eligibility. Of these, 269 could be excluded due to the information in the title or abstract, for 20 records the published reports were retrieved. Based on the information provided by the reports, two studies were identified as meeting our eligibility criteria and 18 had to be excluded. See PRISMA flow diagram (Figure 1) for details. The reference lists of the 12 systematic reviews identified by database searches and of the former publication focusing on DSM-5 criteria³⁶ yielded 148 further records, of which 72 duplicates were removed and 67 could be excluded due to the information given in the reviews and in the titles. The published reports were retrieved for nine records. Seven studies meeting our eligibility criteria were identified using this approach (Figure 1). Altogether, nine studies involving 946 patients could be included in this review.

Figure 1 PRISMA flow diagram of the literature search of this systematic review. Adapted from Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. Creative Commons.41

Study Characteristics and Outcome Measures

Nine studies that met the selection criteria had been published between 1987 and 2019. Detailed study characteristics are provided in Table 1. Daily doses of EGb 761 ranged between 120 mg and 240 mg, and treatment periods were between 8 weeks and 52 weeks. Daily doses of 240 mg were applied in 3 studies (2 of which were conducted after the year 2000), whereas daily doses of 120 mg or 160 mg were used in 6 studies (5 of which were conducted before the year 2000). Seven studies followed a parallel group design with two groups, one treated with EGb 761, the other one receiving placebo. In one study,⁴⁴ patients were randomly assigned to one of four parallel groups, two of which were treated with EGb 761, while the other two received placebo. One of the EGb 761 groups and one of the placebo groups underwent additional standardized memory training. Another study⁴⁵ was run with three parallel groups, receiving a traditional Chinese medicine (TCM) treatment, EGb 761 or placebo.

Table 1 Randomized, Placebo-Controlled, Double-Blind Clinical Trials of EGb 761 in Patients with Mild NCD – Description of Included Trials

The trials were heterogeneous with respect to the outcome measures. Neuropsychological tests were administered in all trials. In two trials, tests or test batteries were used for which summary scores were reported.^45,46 Seven trials used specific tests for single cognitive domains.^44,47–52 Geriatric assessment scales were administered in two trials;^51,52 one used the clinician’s global impression of change (CGI-C) rating.⁴⁷ In one trial that included MCI patients with clinically significant neuropsychiatric symptoms (NPS), a rating scale for NPS (Neuropsychiatric Inventory, NPI) was administered;⁴⁷ two trials focused on depressive symptoms in patients with mild NCD using the Hamilton Depression Scale (HAMD).⁵¹ In one trial, a computerized test battery consisting of six validated tests of information processing^49,53 was administered and scores were combined for the domains speed and accuracy.⁵² One trial focused on memory, using an elaborate “memory battery” to assess four different aspects of memory.⁴⁴ In a one-year trial, the progression from MCI to dementia was analyzed.⁴⁵ One study included patients with very mild (borderline to slightly abnormal cognitive function) and mild (clearly abnormal cognitive function) cognitive impairment. Ceiling effects were therefore expected (and found) in patients with very mild impairment, due to the limited sensitivity of available tests. Therefore, the conduct and reporting of subgroup analyses, which included patients with more pronounced cognitive impairment (ie clearly abnormal cognitive function) were specified in advance by the protocol.⁴⁸

Risk of Bias in Studies

Jadad scoring supports the assumption of a low risk of bias for most of the studies included. Six studies achieved a score of 5, one study a score of 4 and two studies a score of 3; recent studies tended to score higher (Table 1). The Jadad rating does not take into account concealment of treatment allocation. However, with identical appearance of drug and placebo tablets, identical packaging and indistinguishable labelling, effective concealment was assured in the seven trials for which the double-blinding procedures were described appropriately, and it was likely for the two trials, which were merely described as double-blinded.

Efficacy Results of Individual Studies

The published results of the nine studies are shown in detail in Table 2. Superiority of EGb 761 compared to placebo in cognitive tests was reported in all studies, except in one that mainly focused on depression scales.⁴⁹ Treatment effects were reported for composite neuropsychological test scores (Alzheimer’s Disease Assessment Scale – cognitive subscale, ADAS-cog; Mini-Mental Status Examination, MMSE; SKT short cognitive performance test), tests of working memory, short- and long-term memory (short-term retention capacity,⁵⁴ Appointments Test, Wechsler Memory Scale (WMS-III) Faces I and II; Israël’s Memory Battery), neuropsychological tests of attention (Trail-Making Test – Form A, TMT-A, Vienna Test System – Attention Performance Series, WTS-ALS) and executive functioning (TMT-B). In the three studies which enrolled patients with mild NCD and symptoms of depression or other significant neuropsychiatric symptoms at baseline, changes in specific rating scales (NPI, HAMD) indicated significant benefits of EGb 761 treatment. A geriatric rating scale (SCAG), an observer global assessment (CGIC), and a rating of treatment satisfaction also found superiority of EGb 761 over placebo. A rating scale for dementia severity (Clinical Dementia Rating – Sum of Boxes, CDR-SB) used in patients with mild NCD showed a trend in favor of EGb 761, whereas an ADL scale for patients with MCI (Alzheimer’s Disease Cooperative Study ADL Scale for MCI, ADCS-ADL-MCI-24) showed a significant treatment effect of EGb 761. In a trial with a 52-week treatment period, 1 of 104 patients (0.96%) treated with EGb 761, but 7 of 70 patients (10%) of the placebo group progressed to dementia (p = 0.0076).

Table 2 Randomized, Placebo-Controlled, Double-Blind Clinical Trials of EGb 761 in Patients with Mild NCD – Results

Safety Results

Numbers and percentages of patients who experienced AEs or SAEs were reported only for the most recent trials.^45,47,48 In all three trials, there was no difference between EGb 761 and placebo groups regarding the proportions of patients with AEs. During 52-week treatment, 41% and 43% of patients taking 160 mg per day of EGb 761 or placebo, respectively, reported AEs.⁴⁵ In studies of EGb 761 at 240 mg per day, in 46% of those randomized to active drug and in 45% of those on placebo, AEs were observed during a 24-week treatment period,⁴⁷ whereas 15% of patients of each treatment group experienced AEs during a 12-week period.⁴⁸ In two trials, no SAEs were observed;^47,48 only SAEs from active treatment groups were reported from the third trial,⁴⁵ but none of these SAEs was considered to be causally related to EGb 761. Most frequently reported AEs were headache, gastrointestinal symptoms (eg dyspepsia, diarrhea and constipation), respiratory tract infections and urinary tract infections. AE information from all older trials is limited, because events were reported only for active treatment groups, or only if they were suspected to be side effects or if they led to study discontinuation. In more recent studies, numbers for specific AEs were reported at different levels of precision^47,48 or not at all,⁴⁵ so that a representation across studies is not possible.

Discussion

Overall, nine randomized, placebo-controlled, double-blind trials involving 946 patients demonstrated significant clinical effects of the Ginkgo biloba extract EGb 761 in patients with mild NCD. Significant improvements in cognitive function, neuropsychiatric symptoms, overall geriatric assessments, and global ratings of change were reported. Consistent positive effects were found for more specified neuropsychological tests in several cognitive domains including memory, attention, processing speed and executive functioning. Improvements in neuropsychiatric symptoms including depression and anxiety were also reported.

The limitations of this review are the retrospective application of the DSM-5 diagnostic criteria and the heterogeneity of the trials with respect to design and outcome measures. No widely recognized scientific recommendations or guidelines were available when the early trials were performed. This is due to the lack of clear diagnostic guidelines or diagnostic criteria at that time. The US Food and Drug Administration (FDA) published a draft guideline for the development of drugs for the treatment of early-stage Alzheimer’s disease⁵⁵ only in 2013, and the European Guideline on the clinical investigation of medicines for the treatment of Alzheimer’s disease first included a short section on studies in patients with MCI in 2018.³⁸ This, to some extent, explains the heterogeneity of study designs, phrasing of inclusion diagnoses and choice of endpoints in the EGb 761 trials.

The review was restricted to studies in patients who met the criteria for a defined diagnostic entity, ie mild NCD. This may, however, weaken the power of the data presented, because some studies included similar patients and their findings could corroborate the favorable effects of EGb 761 on factors such as neuropsychological performance.^56–58 However, these studies were excluded, since it could not be ascertained that the patients who were enrolled fulfilled the diagnostic criteria for mild NCD.

When the early trials of EGb 761 were conducted, most of the tests to evaluate cognitive functioning were designed to detect, characterize and assess the severity of cognitive deficits. Thus, many of these tests had a rather low sensitivity as well as a statistical ceiling effect regarding treatment-related changes over time or measurement of smaller differences between treatment groups. Measurements of minor cognitive or neuropsychological changes in patients with very mild symptoms in the early stages of the disease are still subject to further discussion, eg, in the context of current disease-modifying antibody therapies for the treatment of Alzheimer’s disease.^40,59 Here, too, many authors have discussed the pronounced ceiling effect of eg the CDR-SB or the MMSE as the primary endpoint.

While cognitive assessments for clinical drug trials in dementia due to Alzheimer’s disease were available in the early 1980s (eg ADAS-cog;⁶⁰ Brief Cognitive Rating Scale, BCRS),⁶¹ the development of neuropsychological tests focused on treatment trials in MCI began only in the late 1990s. At this time, Mohs et al extended the ADAS-cog by adding two items, which addressed cognitive abilities that are typically impaired at the MCI stage of dementia disorders.⁶² Even later, with the advent of potentially disease-modifying therapies, further instruments were designed for the assessment of neuropsychological changes in early stages of Alzheimer’s disease (eg Alzheimer’s Disease Cooperative Study – Preclinical Alzheimer Cognitive Composite, ADCS-PACC;⁶³ Alzheimer’s Disease Composite Score, ADCOMS).⁶⁴ The availability of these new assessments may lead to more homogeneous trials with better comparability of results in the future but are still significantly less established than the well-known tests, such as the MMSE or the CDR.

In terms of neuropsychological domains, the most consistent effects of EGb 761 treatment were seen for memory (eg Appointments Test, WMS-III Faces I/II and Israël’s Memory Battery); clearcut effects were also found for attention (WTS-ALS), processing speed (TMT-A) and executive functioning (TMT-B). This is in line with findings reported by Kaschel⁶⁵ in a review of clinical trials of Ginkgo extracts in patients with less well-specified cognitive deficits. Interestingly, memory impairment is characteristic for MCI and dementia due to Alzheimer’s disease,¹² whereas impaired processing speed, executive functioning and attention are considered typical for vascular NCD.⁶⁶ In fact, composite scores of neuropsychiatric tests (ADAS-cog, SKT) showed superiority of EGb 761 over placebo in patients with Alzheimer’s disease and vascular dementia.^35,67 The specific cognitive domains mainly covered by these tests are attention, memory, speed of processing and executive functioning.

Some domain-specific neuropsychological tests apparently were not sensitive enough to clearly distinguish between treatments in patients with borderline or very mildly impaired cognitive functioning. One reason may be that at the very early stages of neurocognitive disorders, many patients have deficits only in one or two domains, so the numbers of patients with deficits in specific domains and therefore the statistical power may be too low. This renders the demonstration of between-group differences in tests for individual domains difficult. Composite scores from test batteries and rating scales, which assess cognitive abilities together with performance in daily life and neuropsychiatric symptoms (eg SCAG) may therefore be more useful in studies of patients with mild NCD.

In mild NCD, cognitive deficits have no major impact on activities of daily living (ADL) but may render instrumental activities more difficult to perform. This is in line with the lack of effect of EGb 761 on the number of activities still possible or still practiced,⁵² whereas a significant improvement was found in an ADL scale specifically designed for use in mild NCD.⁴⁵ The observed improvement in neuropsychiatric symptoms concurs with the results of trials in dementia, as shown in a meta-analysis by Savaskan.⁶⁸ This is noteworthy, because patients with MCI have a higher risk of progressing to dementia and a lower chance for reverting to normal, if they have neuropsychiatric symptoms.⁶⁹

As expected, EGb 761 was well tolerated. The rates of adverse events and suspected adverse drug reactions were similar in the placebo groups, the rates of serious adverse events were very low, and clinical examinations, ECG and lab tests did not reveal significant changes.

Conclusion

Overall,^36,45 the randomized, placebo-controlled trials included in this systematic review provide evidence of efficacy of Ginkgo biloba extract EGb 761 in the treatment of patients with mild neurocognitive disorder. Improvements in cognitive performance, neuropsychiatric symptoms, geriatric rating scales and clinical global impression were demonstrated. The drug was safe and well tolerated.