Attempts to Develop Vaccines Against Alzheimer’s Disease: A Systematic Review of Ongoing and Completed Vaccination Trials in Humans

In this systematic review, we evaluate the safety, tolerability, and immunogenicity of vaccination efforts against Alzheimer’s disease (AD) in human subjects from both ongoing and completed vaccination trials. Databases like PubMed, Embase, and Scopus were used to identify relevant articles on completed vaccination trials whereas the clinicaltrials.gov database was used for identifying ongoing clinical trials for vaccination against AD in humans until January 2022. Only interventional randomized or non-randomized clinical trials which reported on the safety and immunogenicity of vaccine against AD in humans were included. Cochrane risk of bias tool-2 (RoB-2) or risk of bias in non-randomized studies- of intervention (ROBINS-I) was used for risk of bias assessment as appropriate. A narrative descriptive synthesis of the findings was done. Sixteen randomized/non-randomized clinical trials (phase I: six and phase II: 10) for seven different types of vaccines against AD were identified comprising a total of 2080 participants. Apart from the development of meningoencephalitis in 6% of patients receiving AN1792 in an interrupted phase II trial, the rest of the trial reported promising results on the safety and immunogenicity of vaccines. While only a subset of reported adverse events was treatment related, none of the fatalities reported during the trial were considered related to vaccine administration. The serological response rate ranged from 100% (4/16 trials) to 19.7% in an interrupted trial. Although current trials show promising results, adequately powered phase III studies are needed to conclusively establish the safety, immunogenicity and therapeutic efficacy of vaccines.


Introduction And Background
Alzheimer's disease (AD), the leading cause of dementia in the elderly, is a neurodegenerative disorder believed to be due to the deposition of amyloid plaques and abnormal tau protein in the brain [1]. Over 50 million people are estimated to be living with dementia worldwide in 2020 with more than half (60%) of the patient population belonging to low-and middle-income countries. In addition, the number of patients living with dementia is projected to double every 20 years, reaching a staggering number of 152 million by 2050 [2]. Available treatment options for AD include cholinesterase inhibitors (like donepezil, rivastigmine, and galantamine), N-methyl-D-aspartate (NMDA) receptor antagonists (like memantine), and monoclonal antibodies against amyloid beta-like aducanumab [3,4]. Although currently, available treatment options for AD have a modest therapeutic effect on neuropsychiatric and functional outcomes, any conclusive evidence on stopping the progression of AD or even decreasing the amyloid burden to achieve symptomatic recovery is lacking [5][6][7][8].
On the other hand, efforts to shift the paradigm of management of AD from tertiary prevention to primary preventive approaches could help reduce the burden of AD in the future. Novel active and passive immunotherapeutic approaches are, therefore, being actively explored as potential preventive measures against AD [9][10][11]. Active immunotherapeutic approaches targeting Aß or tau-related pathologies in transgenic mice have shown promising outcomes in terms of reduction in the extent and progression of ADlike pathologies, reducing inflammation, improving cognitive performance, and reducing memory loss [12][13][14][15][16]. This has led to the progression of preclinical studies in transgenic mouse models to clinical trials in human subjects to evaluate the safety, tolerability, and immunogenicity of active immunotherapy. However, a systematic literature review and qualitative analysis of safety, tolerability, and immunogenicity of earlyphase clinical trials, both ongoing and completed, for active immunotherapeutic efforts against AD is lacking. This study, therefore, is done to systematically review the safety, tolerability, and immunogenicity of vaccination efforts against AD in human subjects from randomized or non-randomized clinical trials.

Protocol and Registration
We conducted this review following preferred reporting items for systematic review and meta-analysis (PRISMA) guidelines [17]. PRISMA checklist is available in S1 File. The protocol was drafted and registered before conducting the systematic review in the international prospective register for systematic review (PROSPERO) with registration number: CRD42022299172 which is available in S2 File.

Inclusion Criteria
We included peer-reviewed published full-text articles which assessed the safety and immunogenicity of vaccines against AD in human beings. Only interventional randomized or non-randomized clinical trials published in the English language were included.

Exclusion Criteria
Clinical trials for vaccines against AD among non-human or animal models were not included. In addition, clinical trials evaluating the safety and efficacy of passive immunotherapy like monoclonal antibodies, drug therapies, interventions like deep brain intervention techniques, etc. against AD in humans or animal models were excluded. Prospective or retrospective cohort studies, case-control studies, review articles, editorials, case reports, and case series were also excluded.

Search Strategy
We conducted a systematic search for relevant articles in databases like PubMed, Embase, and Scopus. We also searched the clinicaltrials.gov database for identifying relevant ongoing clinical trials for vaccination against AD in humans. A free search was conducted without any predetermined timeframe for published studies. The search strategy included keywords like "Alzheimer*" and "vaccine" combined with the Boolean operator "AND". Only full-text articles published in the English language were included. The last search was conducted in January 2022.

Data Screening and Extraction
AT, SB, and SP independently screened and retrieved the articles using the systematic search strategy. Studies were reviewed for eligibility by a screening of titles and abstracts followed by full-text screening. AT, SB, and SP cross-examined the search results to check for their decisions. Upon disagreement, a careful review of the inclusion and exclusion criteria was the basis of the final decision. The principal investigator (AT) made the final decision after thorough reviewing when an agreement could not be reached. Zotero, a research tool to collect, organize, and manage research publications was used to keep a record, and remove duplicates.
Studies retrieved from a systematic search were imported to Zotero. Screened studies were placed into appropriate subfolders created in Zotero based on the decision to include or exclude them. Final data were extracted from the included studies. The following information was extracted from included studies (if available): type of study, vaccine characteristics, route of administration, a dosing schedule of the vaccine, inclusion, and exclusion characteristics for participants in vaccine trials, number of participants both in vaccination and control arms, duration of follow up, reported adverse events, tolerability and immunogenicity of the vaccine, etc. All records were entered into an Excel spreadsheet. AT, SB, and SP independently extracted the data and both authors cross-checked the extracted data in an alternate fashion.

Risk of Bias
AT, SB, and SP independently assessed the risk of bias in included studies, and disagreements between the authors were resolved by further discussion.
The risk of bias was assessed using the revised Cochrane risk of bias tool-2 (RoB-2) for RCTs [18]. RoB-2 consists of five key domains which assess bias based on aspects of trial design, conduct, and reporting. Domain-level risk of bias and subsequent overall risk of bias judgment was generated by an algorithm based on answers to the signaling questions on each domain. Risk of bias judgment was categorized as "low risk of bias," "high risk of bias" or risk of bias expressing some concerns." The risk of bias for non-randomized interventional clinical trials was evaluated using the Risk of Bias in Non-randomized Studies-of Intervention (ROBINS-I) assessment tool [19]. It evaluated the risk of bias under seven domains namely bias due to confounding, bias due to selection of participants into the study, bias in classification of interventions, bias due to deviations from intended interventions, bias due to missing data, bias in measurement of outcomes, and bias in selection of reported results. The judgment on risk of bias was made based on the answers to the signaling questions. Both domain level and overall risk of bias judgment were categorized as low/ moderate/serious/critical risk of bias and no information.

Data Analysis and Data Synthesis
A narrative descriptive synthesis of the findings was done. The results of this review were primarily focused on summarizing and reporting the safety, tolerability, and immunogenicity of vaccines against AD in various clinical trials. Safety of the vaccines was reported in terms of the reported number of adverse events along with adverse events considered related to study treatment and a number of reported fatalities. Tolerability to the vaccines was reported in terms of a number of participants who dropped out of the trial due to adverse events. Lastly, immunogenicity was reported in terms of serological response rate and/or antibody titers against the antigen component of the vaccine.

Study Selection
We identified a total of 4,443 potentially eligible studies through a combination of database search and free hand search for related articles. After excluding 1,593 duplicate records, 2,850 articles were screened for inclusion via title and abstract screening. This yielded 73 articles for full-text screening after excluding 2,777 articles that failed to address the review question. Finally, 16 studies were included in the systematic review after full-text screening of the articles. The remaining 57 articles were excluded from the review due to the reasons mentioned in the PRISMA flow diagram in Figure 1.

Characteristics of Included Studies
Out of the 16 included studies [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35], six were phase I clinical trials, ten were phase II clinical trials, and comprised a total of 2080 participants. The earliest phase I trial with human aggregated Aß42 (AN1792) by Bayer et al. was conducted in 2000 in the UK across four centers to evaluate the safety, tolerability, and immunogenicity of the vaccine with 80 trial participants [20]. Several clinical trials with different AD vaccine designs were conducted in subsequent years. Among the 16 included studies in this review, four are randomized, double-blind placebo-controlled phase I clinical trials, two are open-label single arm phase I interventional clinical trials, nine are randomized double blind/third party unblinded placebo-controlled phase II clinical trials and one is open label single arm phase II clinical trial. A total of five out of 16 included studies evaluated the safety, tolerability and other exploratory endpoints among patients receiving Vanutide Cridificar (ACC-001) in phase II clinical trials. Potential safety and tolerability of CAD106 as vaccine against AD was explored in one phase I and two phase II clinical trials. Likewise, vaccines like AN1792, AADvac1, UB-311, ABvac40 and AD02 were the candidates for vaccine against AD in rest of the studies. A summary of the characteristics of the included studies is presented in Table 1.

Risk of Bias in Included Studies
For randomized controlled trials (13 out of 16 studies), revised Cochrane risk of bias tool (ROB 2) evaluated risk of bias for primary outcomes in five key domains, i.e., bias arising from randomization process, bias due to deviation from intended interventions, bias due to missing outcome data, bias in measurement of the outcome and bias in selection of the reported result. All of the studies except the studies by Arai et al. [27] and Schneeberger et al. [29] were found to have low overall risk of bias. Arai et al. and Schneeberger et al., on the other hand, reported no information about the method of randomization and whether or not the allocation sequence was concealed until participants were enrolled and assigned to interventions. This led to risk of bias judgment in domain one and overall risk of bias as "some concern". All of the randomized controlled trials ensured that participants, carers and people delivering the interventions were unaware of the intervention groups during the trial and an appropriate analysis was used to estimate the effect of assignment to intervention leading to low risk of bias judgment in domain two. In addition, outcome data were available for all, or nearly all, randomized participants resulting in low risk of bias in domain three. Likewise, in domain four, the method of measurement of outcome was not inappropriate and the measurement or ascertainment of the outcome did not differ between intervention groups. Furthermore, the outcome assessors were also unaware of the intervention received by the study participants resulting in low risk of bias. Finally, risk of bias in selection of reported result was low in all studies because they analyzed the data in accordance to pre-specified plan and reported results for the outcome domain corresponding to all intended outcome measurements and analyses. Traffic light plots of the domain level judgments for each individual studies and weighted bar plots of the distribution of risk of bias judgments within each bias domain are presented in Figures 2, 3, respectively.

Overview of Completed Vaccination Trials
First in human phase I clinical trial for AD vaccine, AN1792 was conducted by Bayer et al. which demonstrated a lack of significant safety concerns for the vaccine [20]. The study, however, reported four cases of nonfatal serious adverse events considered related to the study treatment namely rash, confusion and syncope, encephalitis (T-lymphocyte meningoencephalitis detected on autopsy) and worsening of dementia. In addition, treatment related adverse events like hostility (n = 3) and hallucination (n = 1) lead to discontinuation of trial in four patients, three in vaccine arm and one in placebo arm. The subsequent phase IIa trial for the vaccine was interrupted because of reported cases of meningoencephalitis in 6% of immunized patients, although no indication of its association with the vaccine was reported in prior studies. An induction of T-cell inflammatory response was hypothesized to be the cause behind incidence of meningoencephalitis rather than due to the development of antibodies against human aggregated Aß42 [26].
Safety, tolerability and immunogenicity of novel active Aß immunotherapy, CAD106 was assessed in phase I clinical trial by Winblad et al. [21] which showed a favorable safety profile and an antibody response. Although cases of meningoencephalitis were not reported, two patients had pericerebral meningeal MRI changes attributed due to lumbar puncture or minor head trauma. None of the serious adverse events were considered related to the study treatment. Two phase IIa randomized placebo-controlled trials and their subsequent open label extension studies for CAD 106 conducted by Farlow et al. [28] supported the safety and tolerability profile of the vaccine. A single case of intracerebral hemorrhage (ICH) as a serious adverse event considered related to treatment was seen in core study. In addition, four patients, two of them in cores studies and two in extension studies, presented with amyloid related imaging abnormalities corresponding to micro hemorrhages (ARIA-H). No cases of amyloid related imaging abnormalities corresponding to vasogenic edema (ARIA-E) or meningoencephalitis were reported. Similarly, Vandenberghe et al. [34] in a phase IIb placebo-controlled trial for CAD106 reported a balance between antibody response and tolerability. Three serious adverse events namely allergic dermatitis, atrial fibrillation and acute psychosis seen in vaccine arm were considered related to treatment. MRI findings compatible with ARIA (ARIA-H, n= 5; ARIA-E, n=1) were seen in six patients in vaccine arm; however, no signs or symptoms of CNS inflammation were reported.
Tau vaccine AADvac1 for AD was investigated for its safety, tolerability and immunogenicity in phase I and phase II clinical trials by Novak et al. [22,25,35] In placebo-controlled phase I trial, apart from mild injection site local reactions, viral infection and epileptic seizure were the only two serious adverse events that were considered related to treatment. Furthermore, besides a case of microhemorrhage in one patient with preexisting microhemorrhage, no cases of meningoencephalitis or amyloid related imaging abnormalities (ARIA-H or ARIA-E) were reported [22]. A 72-week interventional phase I follow-up study of parent trial also found AADvac1 to have a benign safety profile [25]. Clinically silent new micro-hemorrhages were observed in one ApoE4 homozygote, and superficial hemosiderin was found in one ApoE4 heterozygote, consistent with background incidence of such lesions. The subsequent phase II trial also found the vaccine to be safe and well tolerated [35].
Vanutide Cridificar (ACC-001) and its safety, tolerability and immunogenicity in patients with AD was reported in five studies, four of them being placebo controlled randomized phase II trial and one interventional single arm long term extension study [27,[30][31][32][33]. van Dyck et al. [32] reported that treatment with ACC-001 was not associated with a higher rate of serious adverse events compared with placebo. The study also reported no incidences of immune-mediated adverse events along with any amyloid related imaging abnormalities. Similar conclusions were drawn from two phase II studies by Arai et al. [27]. However, Ketter et al. [30] in a phase II trial reported five cases of ARIA-E in vaccine arm leading to early discontinuation of the study medication; the events were however, considered mild to moderate in severity and their frequency too small to draw any conclusions. Similarly, Pasquier et al. [31] in two phase IIa multiple ascending dose studies reported two cases of ARIA-E, one symptomatic and other asymptomatic, in ACC-001 group. Symptomatic ARIA-E was resolved after treatment with steroids for three days, and asymptomatic ARIA-E resolved without treatment. Hull et al. [33] in single arm interventional extension studies observed similar safety profile as that in parent trials and suggested that side effects did not pose a principal limitation for anti-amyloid immunotherapy. An overview of completed vaccination trials in human is summarized in Table 2.

Overview of Ongoing Vaccination Trials
A systematic search on the website clinicaltrials.gov retrieved a single ongoing randomized, double blind, placebo controlled 24 months phase II study to investigate the safety, tolerability and immune response of repeated subcutaneous injections of ABvac40, an active vaccine against C-terminal end of Aß40 [36]. The trial is designed with an estimated enrollment of 120 patients, 55-80 years old with amnestic mild cognitive impairment or very mild AD. It was started in February 2018 and the estimated completion date was December 2022. Upon successful completion, the study aims to confirm the safety and tolerability data obtained from phase I clinical trial of ABvac40 by Lacosta et al. [24] Dosing schedule of ABvac40/placebo is six subcutaneous injections, the first five administered every four weeks and the sixth at week 42. Safety, tolerability and immune response, i.e., levels of anti-abeta40 antibodies in plasma are the primary outcome measures.

Discussion
While the early phase clinical trials for active immunotherapeutic approaches for AD have revealed promising results in terms of safety and immunogenicity, the therapeutic efficacy of the vaccines against AD is yet to be conclusively explored in adequately powered phase III studies. However, development of T-cell meningoencephalitis among 6% of patients who received AN-1792 in phase IIa clinical trial has led to the development of Aß targeting second generation vaccines designed to stimulate B-cell activation and antibody production, avoiding an Aß specific T-cell response [21,23,26,37]. In addition, even though the trial was interrupted with none of the participants receiving complete doses of the vaccine, long-term follow up of vaccinated patients, who were labelled as responders, maintained low but sustained, detectable antibody titers with significantly reduced functional decline [38]. These findings help emphasize the prospects of Aß immunotherapy as a potential means of reducing disease morbidity. Lastly, since the study population in all of the clinical trials consisted of elderly patients with AD, only a subset of reported adverse events were considered treatment related whereas majority of them fell into the spectrum of age-related comorbidities [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35].
The main limitation of this review is the characteristics of included studies. Assessments of safety and efficacy of the vaccines were made from phase I and phase II clinical trials with limited sample size and follow up duration. Since, none of the vaccines completed phase III trials, safety and efficacy assessments could not be confirmed with results of phase III trials with larger study population. With the inclusion of only interventional randomized or non-randomized clinical trials and exclusion of observational studies, comprehensive review of literature may have been impeded. Due to the differences in trial design along with formulation, dosage and administration of study treatments, comparative assessments of study outcomes could not be made. Furthermore, phase I clinical trials for three out of seven vaccines for AD (UB-311, AD02, ABvac40) were not followed up with subsequent phase II trials [23,24,29].
Despite the limitations of the study, this review serves to bridge the research gap by providing a systematic literature review and qualitative analysis of safety, tolerability and immunogenicity of early phase clinical trials, both ongoing and completed, for active immunotherapeutic efforts against AD. It is important to assert that continuous advancements in vaccine development with promising vaccine trials could not only help pioneer a safe and effective vaccine against AD, but it could also open doors to explore potential immunotherapeutic options for the wide spectrum of neurodegenerative disorders. However, attempts to develop vaccines against AD should always strive towards achieving highest therapeutic efficacy coupled with highest safety standards and cost-effectiveness. Large scale adequately powered clinical trials should, therefore, be conducted to detect rare adverse events associated with vaccination, explore changes in any potential disease biomarkers in vaccinated individuals along with the effect of host immune system competence to mount an immune response to vaccination. Furthermore, the role of anti-amyloid beta antibodies in non-AD tauopathies could also be explored. The authors are hopeful that clinical trials will pick up the pace in near future to further explore the role of vaccines against AD.

Conclusions
This systematic review summarized the safety, tolerability, and immunogenicity of vaccines against AD where the majority of clinical trials reported promising results on all of the three outcome measures. This study also attempted to highlight the considerations which need to be heeded as one moves toward the development of active immunotherapeutic options for AD. Further high-quality phase III studies are needed to conclusively explore the clinical implications of active immunotherapies for AD patients.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.