Intravenous Immunoglobulin for Patients With Alzheimer’s Disease: A Systematic Review and Meta-Analysis

Abstract

Aim:

To assess the efficacy and safety of intravenous immunoglobulin (IVIg) for patients with Alzheimer’s disease (AD).

Materials and Methods:

We searched electronic databases and other sources for randomized controlled trials comparing IVIg with placebo or other treatment for adults with AD. Primary outcome was change from baseline in Alzheimer’s Disease Assessment Scale–Cognitive subscale (ADAS-Cog).

Results:

Five placebo-controlled trials were included in the meta-analysis. Compared to placebo, IVIg 0.2 and 0.4 g/kg once every two weeks did not change ADAS-Cog score (weighted mean difference: 0.37, 95% confidence interval: −1.46 to 2.20 and 0.77, −1.34 to 2.88, respectively). Furthermore, except for an increase in the incidence of rash, IVIg did not affect the incidence of other adverse events.

Conclusion:

IVIg, albeit safe, is inefficacious for treatment of patients with AD. Future trials targeting earlier stages of disease or applying different dosing regimens may be warranted to clarify its therapeutic potential.

Introduction

The ever-growing prevalence of Alzheimer’s disease (AD) and its detrimental effect on patients’ quality of life has triggered the scientific community to thoroughly investigate for effective treatments. Nevertheless, to date, the only approved treatment modalities for AD are acetylcholinesterase inhibitors ¹ and memantine, ² which serve merely as symptomatic cognitive-enhancing drugs and cannot alter the course of the disease. The quest for disease-modifying therapies has largely been informed by the dominant scientific theory to explain AD, the so-called amyloid hypothesis, which posits that the pathogenic cascade begins with extracellular amyloid-beta (Aβ) accumulation, which exerts toxic effects on proximal neurons leading to hyperphosphorylation of microtubule-associated protein tau, ³ its accumulation into neurofibrillary tangles intracellularly, and eventually neurodegeneration with synaptic and neuronal loss. ⁴ This has led to the development of a line of therapeutics aiming at active and passive immunization against Aβ, including a new class of drugs, the anti-Aβ monoclonal antibodies. However, results of clinical trials have so far been either disappointing ^{5 -7} or inconclusive. ^8,9 On the other hand, the active immunization of patients with synthetic Aβ peptide (AN1792) was marked by the development of a substantial number of meningoencephalitis cases and was, thus, deemed unsafe. ¹⁰ These disappointments have led many researchers to question whether the amyloid hypothesis can serve as a guide for therapeutic development in AD. ¹¹ What is more, most of the potential anti-tau therapies have been discontinued mainly because of concerns about toxicity and lack of efficacy. ¹²

Intravenous immunoglobulin (IVIg) was suggested as an alternative therapeutic approach, given that it is a source of naturally occurring human antibodies against both Aβ peptides and tau protein. ^{13 -15} These antibodies can induce several immune-mediated responses that can lead to the clearance of accumulated Aβ and the decrease in tau aggregation. ^16,17 Additionally, as opposed to anti-Aβ monoclonal antibodies, they are polyclonal, meaning that they target a number of Aβ species rather than a single one conforming to the theory that the whole spectrum of aggregation-prone Aβ species needs to be targeted. ¹⁸ Moreover, in vitro evidence suggested that IVIg may protect neurons against Aβ toxicity by attenuating cell death pathways. ¹⁹ Finally, the rationale for also targeting tau with IVIg has been strengthened by recently emerging evidence for its transneuronal spread and its presence in the extracellular environment. ²⁰ In light of this evidence, a number of clinical trials of IVIg in AD have been conducted.

The aim of this systematic review and meta-analysis is to summarize and critically appraise all existing evidence on the efficacy and safety of IVIg for patients with AD.

Methods

This systematic review and meta-analysis was based on a publicly available protocol registered in the PROSPERO database (CDR42018107423) and is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. ²¹

Data Sources

We searched Medical Literature Analysis and Retrieval System Online (MEDLINE), Excerpta Medica Database (EMBASE), Psychological Information Database (PsycINFO), Cochrane Central Register of Controlled Trials (CENTRAL), and Cumulative Index to Nursing and Allied Health Literature (CINAHL) from inception to May 30, 2018. We used relevant free-text and controlled vocabulary terms without imposing any restrictions regarding language or publication status. ²² The detailed search strategy is available in Supplementary Table S1. We also searched clinical trial registries, including ClinicalTrials.gov, International Standard Randomised Controlled Trials Number (ISRCTN) registry, University Medical Information Network (UMIN) Clinical Trials Registry (Japan’s trial register), and the World Health Organization (WHO) International Clinical Trials Registry Platform. Finally, we sought for additional studies from handsearching the proceedings of the Alzheimer’s Association International Conference since 2010 and the references of pertinent reviews and retrieved articles.

Study Selection

We included randomized controlled trials (RCTs) comparing any IVIg preparation with placebo or any other treatment agent in adults with either dementia or mild cognitive impairment (MCI) due to AD, irrespective of treatment duration and dosage regimen. Records retrieved from electronic databases were imported into a reference management software (EndNote X7; Clarivate Analytics, Massachusetts, USA), and after deduplication, the remaining records were exported to an online software (Covidence; Veritas Health Innovation Ltd, Melbourne, Australia) for screening. Two reviewers independently screened all records at title and abstract level and subsequently assessed the full text of potentially eligible studies. Any disagreements were resolved by consensus between the two reviewers. Grey literature sources were searched by a single reviewer, and the results were juxtaposed against the published records.

Data Extraction

Two reviewers, working independently, extracted data from eligible studies using a predesigned extraction form, and any discrepancies were arbitrated by a senior reviewer. When multiple reports of the same study were retrieved, we collated all data, and in case of contradictory data, we extracted those from the published report. We preferably extracted data for the doses of 0.2 g/kg and 0.4 g/kg administered once every 2 weeks (q2wk), since preliminary studies have shown promising results with these regimens. ^13,23,24 If a trial did not assess these doses, we extracted data for the closest dosing schemes to these regimens. In order to detect a possible disease-modifying effect, we extracted outcome data for the longest treatment duration reported in every individual trial. Finally, in case data for the outcomes of interest were not reported, we contacted the corresponding authors for further information. Our primary outcome was change from baseline in cognitive performance as measured by Alzheimer’s Disease Assessment Scale–Cognitive subscale (ADAS-Cog). Secondary efficacy outcomes included change from baseline in Mini-Mental State Examination (MMSE), Alzheimer’s Disease Cooperative Study–Activities of Daily Living Inventory (ADCS-ADL), and Clinical Dementia Rating–Sum of Boxes (CDR-SOB) scores. We also extracted data on the incidence of cerebral microhemorrhage and vasogenic edema, which are both well-known adverse effects of anti-amyloid immunotherapy. ^25,26 Other safety outcomes included incidence of rash, pyrexia, headache, nausea, vomiting, which have been associated with administration of IVIg, ²⁴ and death.

Risk of Bias Assessment

Quality assessment of the included studies was performed using the revised Cochrane Collaboration’s Risk of Bias (RoB 2.0) tool. ²⁷ Two reviewers independently assessed the risk of bias for the primary outcome, and any disagreements were resolved by consensus. Overall risk of bias for each study was considered low if all individual domains were at low risk of bias, high if at least one domain was at high risk of bias, and at some concerns in any other case. Finally, we also planned to explore the presence of small-study effects for the primary outcome both visually by a funnel plot and formally with the Egger's test. ²⁸

Data Synthesis and Analysis

We handled ordinal data (scale measurements) as continuous for which we calculated weighted mean differences (WMDs) with 95% confidence intervals (CIs) applying an inverse-variance weighted random-effects model with the DerSimonian and Laird method. ²⁹ When no measure of dispersion was reported, we imputed standard deviations using appropriate methodology. ³⁰ For dichotomous outcomes, we calculated odds ratios (ORs) and 95% CIs using the fixed-effect Mantel-Haenszel method. ³¹ In case of studies reporting zero events, we applied a continuity correction proportional to the reciprocal of the opposite treatment arm size. ³² Statistical heterogeneity across studies was assessed with the I² statistic, with values more than 50% indicating substantial heterogeneity. ³³

We conducted two separate sets of analyses, one for each eligible dosage regimen (0.2 and 0.4 g/kg q2wk). We also planned to perform subgroup analyses based on the stage of AD (MCI or dementia) and a sensitivity analysis for the primary outcome only synthesizing data from studies at low overall risk of bias. Finally, since we included both completed and truncated trials, we conducted a post hoc sensitivity analysis excluding trials that were prematurely terminated. ³⁴ All analyses were performed using RevMan 5.3 (Nordic Cochrane Centre, Copenhagen, Denmark) and Stata 13.1 (StataCorp, College Station, Texas, USA).

Results

Search Results and Study Characteristics

The study selection process is depicted in Figure 1. Twelve records describing six trials (788 patients) were included in the systematic review. ^{35 -40} Results for one trial were available solely from conference abstracts and were not reported in a way that could be used in a meta-analysis. ³⁸ Data were requested from the corresponding author, ⁴¹ but were not made available; hence, this study was reviewed but could not be incorporated in the meta-analysis.

Figure 1.

Flow diagram of study selection process.

The characteristics of the included studies and participants’ baseline characteristics are presented in Table 1. All trials had a parallel group design comparing different commercial IVIg preparations of 10% concentration with placebo. Five trials recruited patients with a diagnosis of probable mild-to-moderate dementia due to AD established according to the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) criteria. ^{35,36,38 -40,42} The remaining trial recruited patients with single or multidomain amnestic MCI due to AD. ³⁷ IVIg was administered for a time period of 8 to 70 weeks, while the mean number of infusions per participant ranged from 5 to 29.9. Notably, one trial did not reach its planned duration of 18 months but was terminated prematurely at 9 months due to the inefficacy of IVIg demonstrated by a concurrently run trial. ⁴⁰ Three trials assessed both regimens of 0.2 and 0.4 g/kg q2wk. ^35,39,40 One trial utilized only the regimen of 0.4 g/kg q2wk, ³⁷ while the two remaining studies assessed multiple dosing schemes. ^36,38 Of note, for one of these trials apart from the 0.4 g/kg q2wk treatment arm, we also extracted data for the dosing regimen of 0.25 g/kg q2wk, since this was the closest to our predetermined regimen of 0.2 g/kg q2wk. ³⁶ Participants’ mean age ranged from 68.3 to 73.8 years, while the mean baseline cognitive performance ranged from 10.29 to 19.4 and 19.0 to 26.75 as assessed by the ADAS-Cog and MMSE, respectively.

Table 1.

Study and Patient Baseline Characteristics.

Study ID	Duration of Treatment	AD Severity	Study Arms Included in the Meta-Analysis	No of Patients Randomized, n	Age, Mean (SD), years	Male, n (%)	Baseline Cognitive/Functional Scale, Mean (SD)	APOE ∊4 Carrier, n (%)
Arai et al 35	12 weeks	Mild to moderate	IVIg 0.2 g/kg q2wk	6	72.0 (10.2)	2 (25.0)	MMSE: 19.0 (2.3)	NR
			IVIg 0.4 g/kg q2wk	6	73.8 (9.1)	1 (16.7)	MMSE: 20.2 (3.1)	NR
			Placebo	4	71.5 (11.3)	0 (0.0)	MMSE: 21.3 (4.4)	NR
Dodel et al, 36 NCT00812565	24 weeks	Mild to moderate	IVIg 0.25 g/kg q2wk	7	68.3 (4.2)	4 (57.0)	ADAS-Cog: 19.4 (8.2) MMSE: 21.4 (2.4) ADCS-ADL: 65.3 (14.0) CDR-SOB: 4.3 (1.8)	5 (71.0)
			IVIg 0.4 g/kg q2wk	7	72.9 (5.0)	4 (57.0)	ADAS-Cog: 17.6 (6.5) MMSE: 22.4 (3.2) ADCS-ADL: 67.1 (6.8) CDR-SOB: 4.3 (1.1)	5 (71.0)
			Placebo	7	72.6 (9.2)	3 (43.0)	ADAS-Cog: 17.7 (4.2) MMSE: 21.6 (1.7) ADCS-ADL: 62.1 (10.4) CDR-SOB: 4.8 (2.6)	6 (86.0)
Kile et al, 37 NCT01300728	8 weeks	Single or multidomain MCI	IVIg 0.4 g/kg q2wk	24	72.26 (7.91)	10 (41.7)	ADAS-Cog: 10.63 (4.23) MMSE: 26.75 ( 2.15) CDR-SOB: 1.96 (0.95)	15 (62.5)
			Placebo	25	72.4 (7.36)	11 (44.0)	ADAS-Cog: 10.29 (5.68) MMSE: 26.44 (2.60) CDR-SOB: 1.58 (0.90)	14 (56.0)
Relkin et al, 38 NCT00299988a	6 months	Mild to moderate	–	24	NR	NR	NR	NR
Relkin et al, 39 NCT00818662	70 weeks	Mild to moderate	IVIg 0.2 g/kg q2wk	138	70.1 (8.3)	61 (44.2)	MMSE: 21.5 (3.1)	94 (68.1)
			IVIg 0.4 g/kg q2wk	129	70.6 (9.7)	59 (45.7)	MMSE: 21.3 (3.2)	87 (67.4)
			Placebo	123	70.2 (9.9)	57 (46.3)	MMSE: 21.1 (3.2)	85 (69.1)
NCT01524887 40	9 monthsb	Mild to moderate	IVIg 0.2 g/kg q2wk	85	69.9 (8.59)	48 (56.5)	NR	NR
			IVIg 0.4 g/kg q2wk	83	72.0 (8.36)	40 (48.2)	NR	NR
			Placebo	83	70.6 (9.98)	31 (37.3)	NR	NR

Abbreviations: AD, Alzheimer’s disease; ADAS-Cog, Alzheimer’s Disease Assessment Scale–Cognitive subscale; ADCS-ADL, Alzheimer’s Disease Cooperative Study–Activities of Daily Living Inventory; APOE, apolipoprotein E; CDR-SOB, Clinical Dementia Rating–Sum of Boxes; IVIg, intravenous immunoglobulin; MCI, mild cognitive impairment; MMSE, Mini-Mental State Examination; NCT number, ClinicalTrials.gov identifier; NR, not reported; q2wk, every 2 weeks.

^aNot included in the meta-analysis.

^bThis study was scheduled to have a duration of 18 months; however, due to early termination, 9-month analyses were performed in the subset of participants who completed at least 9 months of treatment.

Risk of Bias Assessment

Risk of bias assessment for the primary outcome is presented in Supplementary Table S2. We did not evaluate risk of bias for one study because it did not assess the primary outcome. ³⁵ Overall, two studies were at low risk of bias. ^36,37 One study was deemed at high risk of bias due to early termination, ⁴⁰ while the remaining one was deemed as having some concerns due to missing outcome data. ³⁹ Small-study effects could not be assessed due to the limited number of included studies. ⁴³

Primary Efficacy Outcome

Our primary outcome was change from baseline in ADAS-Cog. Compared to placebo, the administration of IVIg 0.2 and 0.4 g/kg did not result in a change in ADAS-Cog score (WMD, 0.37, 95% CI, −1.46 to 2.20; I² = 0%, 3 studies and 0.77, 95% CI, −1.34 to 2.88; I² = 28%, 4 studies, respectively; Figure 2). Similarly, no treatment effect was observed in a sensitivity analysis excluding the prematurely terminated trial (0.01, 95% CI, −2.14 to 2.16; I² = 0% and 1.13, 95% CI, −2.22 to 4.48; I² = 51% for IVIg 0.2 and 0.4 g/kg, respectively). ⁴⁰ We could not perform a subgroup analysis based on the stage of AD, as only one trial recruited participants with MCI. ³⁷ However, in a post hoc sensitivity analysis excluding this trial results remained unchanged (0.78, 95% CI, −1.87 to 3.42; I² = 49% for IVIg 0.4 g/kg). Finally, the sensitivity analysis restricted to trials at low risk of bias did not reveal any treatment effect of IVIg (3.19, 95% CI, −0.41 to 6.80; I² = 0% for IVIg 0.4 g/kg).

Figure 2.

Forest plot showing change in Alzheimer’s Disease Assessment Scale–Cognitive subscale (ADAS-Cog) for intravenous immunoglobulin compared to placebo. CI indicates confidence interval; IV, inverse variance; IVIg, intravenous immunoglobulin; q2wk, every 2 weeks; SD, standard deviation.

Secondary Efficacy Outcomes

As secondary efficacy outcomes, changes from baseline in MMSE, ADCS-ADL, and CDR-SOB scores were assessed. Regarding the cognitive performance assessed by MMSE, no significant difference was observed with the use of IVIg compared to placebo (1.95, 95% CI, −0.54 to 4.44; I² = 0% and −0.45, 95% CI, −2.12 to 1.23; I² = 0%, 2 studies with IVIg 0.2 g/kg and 3 studies with 0.4 g/kg, respectively; Supplementary Figure S1). Results did not differ in a sensitivity analysis excluding the trial recruiting patients with MCI. ³⁷ Similarly, the functional performance as evaluated by ADCS-ADL was not significantly improved with either dose of IVIg compared to placebo (−1.93, 95% CI, −4.54 to 0.67; I² = 3% and −3.08, 95% CI, −7.39 to 1.23; I² = 58%, 3 studies with IVIg 0.2 g/kg and 3 studies with 0.4 g/kg, respectively; Supplementary Figure S2). These findings were consistent in a sensitivity analysis excluding the truncated trial. ⁴⁰ Finally, no significant change was observed in the score of CDR-SOB (1.06, 95% CI, −0.16 to 2.28; I² = 0%, 2 studies with IVIg 0.4 g/kg; Supplementary Figure S3).

Safety Outcomes

Table 2 presents a summary of findings of meta-analyses on all safety outcomes. Treatment with IVIg did not result in an increased incidence of cerebral microhemorrhage in comparison to placebo (OR, 0.55, 95% CI, 0.13 to 2.37; I² = 0% and 1.71, 95% CI, 0.58 to 5.02; I² = 0%, 3 studies with IVIg 0.2 g/kg and 4 studies with IVIg 0.4 g/kg, respectively). Similarly, no differences were observed on the incidence of vasogenic edema, pyrexia, headache, nausea, vomiting, and death. However, both doses of IVIg were associated with an increased incidence of rash compared to placebo (3.73, 95% CI, 1.62 to 8.60; I² = 0% and 3.56, 95% CI, 1.64 to 7.72; I² = 0%, 3 studies with 0.2 g/kg and 4 studies with IVIg 0.4 g/kg, respectively).

Table 2.

Findings of Fixed-Effect Meta-Analyses Comparing IVIg With Placebo on Safety Outcomes.

Outcome	Comparison	Studies Contributing Data, n	Participants Analyzed, n		OR (95% CI), I2
			IVIg	Placebo
Cerebral microhemorrhage	IVIg 0.2 g/kg q2wk vs. placebo	3	148	132	0.55 (0.13-2.37), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	4	164	157	1.71 (0.58-5.02), 0%
Vasogenic edema	IVIg 0.2 g/kg q2wk vs. placebo	3	148	132	1.00 (0.10-10.36), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	4	164	157	1.52 (0.24-9.56), 0%
Rash	IVIg 0.2 g/kg q2wk vs. placebo	3	226	208	3.73 (1.62-8.60), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	4	240	233	3.56 (1.64-7.72), 0%
Pyrexia	IVIg 0.2 g/kg q2wk vs. placebo	2	142	128	1.54 (0.47-5.08), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	3	158	153	1.78 (0.58-5.43), 0%
Headache	IVIg 0.2 g/kg q2wk vs. placebo	3	227	211	1.65 (0.98-2.78), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	4	241	236	1.46 (0.86-2.45), 0%
Nausea	IVIg 0.2 g/kg q2wk vs. placebo	3	227	211	1.14 (0.54-2.39), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	4	241	236	1.74 (0.88-3.47), 0%
Vomiting	IVIg 0.2 g/kg q2wk vs. placebo	3	148	132	1.20 (0.40-3.60), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	4	164	157	1.95 (0.79-4.86), 0%
Death	IVIg 0.2 g/kg q2wk vs. placebo	4	233	215	1.21 (0.29-5.03), 0%
	IVIg 0.4 g/kg q2wk vs. placebo	5	247	240	0.73 (0.16-3.33), 0%

Abbreviations: CI, confidence interval; IVIg, intravenous immunoglobulin; OR, odds ratio; q2wk, every 2 weeks.

Discussion

In this systematic review and meta-analysis, we explored the efficacy and safety of IVIg for patients with AD. Overall, IVIg was safe conferring no additional risk in comparison to placebo for the occurrence of cerebral microhemorrhage, vasogenic edema, pyrexia, headache, nausea, vomiting, and death. Nevertheless, we observed an increase in the incidence of rash with both dosing regimens of 0.2 and 0.4 g/kg q2wk of IVIg. On the other hand, compared to placebo, IVIg had no effect on the cognitive and functional performance of patients as reflected in the change of ADAS-Cog, MMSE, ADCS-ADL, and CDR-SOB scores.

In line with our findings, a recent meta-analysis showed that the administration of IVIg, albeit well tolerated, did not improve the cognitive status of participants. ⁴⁴ However, the validity of its conclusions was undermined by the lack of a prespecified publicly available protocol, the absence of a full publication, and the use of the suboptimal Jadad scale for the quality assessment of the included studies. ⁴⁵ In addition, the researchers focused on a similar but rather vague research question not reporting the dosing regimens and the duration of IVIg administration they evaluated. Finally, they pooled most of the adverse events and classified them as serious or any others. By contrast, we adhered to a publicly available protocol with minimal deviations and report our systematic review and meta-analysis according to recommended guidelines. ²¹ Furthermore, we performed an extensive literature search including grey literature sources and evaluated the methodological quality of included trials using a valid and robust tool. ²⁷ Moreover, we assessed the occurrence of several, diverse, adverse effects from different system organ classes that have been related to the use of IVIg. Finally, we conducted separate meta-analyses for each of the two most commonly used dosing regimens of IVIg to explore for potential dose-dependent effects and performed pertinent sensitivity analyses.

Nevertheless, some limitations of our study should also be acknowledged. Most trials recruited a small number of participants; hence, our effect estimates were mostly driven by the results of the phase III trial conducted by Relkin et al. ³⁹ Furthermore, included studies provided inadequate data with regard to the differential effects anti-amyloid immunotherapy can have on the clinically significant subgroups of apolipoprotein (APOE) ∊4 carriers and noncarriers. ⁴⁶ Finally, we could not assess the impact the stage of AD has on response to treatment by performing the prespecified subgroup analyses, since there was only one trial recruiting patients with MCI.

Based on our synthesis of the current existing evidence from RCTs, IVIg should not be considered as a therapeutic option for AD. However, our meta-analysis provides some important insights into the safety profile of IVIg that need to be highlighted. Previous trials examining the effects of anti-amyloid immunotherapies resulted in several cases of amyloid-related imaging abnormalities (ARIA), either cerebral microhemorrhages (ARIA-H) or vasogenic brain edema (ARIA-E). ²⁶ However, we showed that passive immunization with IVIg is not related to either of these two adverse events. On the other hand, we verified previous reports on the increased incidence of rash. ²⁴ Of note, the reoccurrence of rash, most commonly maculopapular, was prevented in one trial with the prophylactic use of antihistamines prior to an IVIg infusion. ⁴⁷ The safety profile of IVIg has always been an issue of special interest, since it is administered in a variety of immunological and neurological conditions for over three decades. Some of the reported events, such as flushing and fatigue, are considered mild; however, some others, including thrombosis and renal impairment, require increased attention. ⁴⁸ Of note, a recent meta-analysis showed that IVIg is not related to an increased risk of thromboembolic events. ⁴⁹ Two ongoing clinical trials that examine the use of IVIg in post-polio syndrome and influenza, respectively, are going to provide further insight into the tolerability of IVIg. ^50,51

Although so far clinical trials have demonstrated non-promising results regarding efficacy, there are some matters that need to be taken into consideration before abandoning any use of IVIg in patients with AD. Emerging evidence suggests that APOE ∊4 genotype has a significant effect on the efficacy and safety of immunotherapy in AD. ⁴⁶ Indeed, this was also evident in the trial by Relkin et al, in which subgroup analyses showed that APOE ∊4 carriers had more favorable outcomes compared to non-carriers. ³⁹ Hence, future studies might be warranted in APOE ∊4 carrier patients with AD. In addition, in most trials, IVIg was administered at a late stage of AD. However, the pathophysiologic process of Aβ accumulation begins decades before the onset of first symptoms of cognitive decline, ⁵² while in later stages of AD, several other mechanisms, independent of Aβ pathology, seem to contribute to the neuronal and synaptic loss. ⁵³ Therefore, researchers should ideally assess the effects of IVIg at an earlier stage of AD before dementia and even MCI occur. ⁵⁴ This has been the case in one trial so far, which albeit failing to meet its primary end point, showed a significant reduction in brain atrophy. ³⁷ On this ground, an ongoing trial (NCT03319810) is evaluating the effect of IVIg on cerebral and retinal amyloid levels in patients with MCI. ⁵⁵ Finally, the duration of the trials might have been too short for a disease-modifying effect to become apparent, while also the doses of IVIg administered might have been too low to exhibit any anti-inflammatory or immunomodulatory effect. Given the short supply and relatively high cost of IVIg, innovative modes of production of IVIg preparations may be explored to overcome these limitations. ²³

In conclusion, evidence from available RCTs support that IVIg is a safe, nevertheless not efficacious treatment option for patients with AD. Further research with studies of longer duration targeting earlier stages of disease in patients with specific genetic markers and at potentially higher doses of IVIg may be warranted to clarify its therapeutic potentials.