Development of AFFITOPE vaccines for Alzheimer's disease (AD) — From concept to clinical testing

Abstract

Based on the notion that cerebral accumulation of certain Aβ species is central to AD pathogenesis and endowed with the knowledge that emerged during clinical testing of the first human Alzheimer vaccine, AN1792, we designed a new generation of Alzheimer vaccines. Rather than relying on full-length Aβ itself or fragments thereof, AFFITOPE vaccines use short peptides, mimicking parts of the native Aβ sequence, as their antigenic component. The technology created to identify these peptides, termed AFFITOPE-technology, at the same time provides the basis for the multi-component safety concept realized in AFFITOPE vaccines. First, as they are non-self, AFFITOPES don't need to break tolerance typically established against self proteins. This allows us to use aluminium hydroxide, the agent first approved as immunological adjuvant for human use and, thus, exhibiting an excellent safety profile. Second, AFFITOPES employed in Alzheimer vaccines are only 6 amino acids in length, which precludes the activation of Aβ-specific autoreactive T cells. Third, and above all, the AFFITOPE technology allows for controlling the specificity of the vaccine-induced antibody response focusing it exclusively on Aβ and preventing crossreactivity with APP. In a program based on two AFFITOPES allowing neoepitope targeting of Aβ (free N-terminus), this approach was taken all the way from concept to clinical application. Early clinical data support the safety concept inherent to AFFITOPE Alzheimer vaccines. Further clinical testing will focus on the identification of the optimal vaccine dose and immunization schedule. Together, result of these trials will provide a solid basis for clinical POC studies.

Abbreviations

Aa

amino acid

Aβ

amyloid β

Ab

antibody

AD

Alzheimer's disease

AD

adverse event

APP

Amyloid Precursor Protein

BBB

blood brain barrier

CMI

cell mediated immunity

FDA

Federal Drug Agency

IFN

Interferon

ME

meningoencephalitis

NINCDS/ADRDA

National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association criteria

POC

Proof of concept

SUSAR

suspected unexpected serious adverse reaction

Introduction

The concept of active Aß immunotherapy of AD was introduced in 1999 (1). Results published by Schenk and colleagues showed that active immunization reduces cerebral Ab load of APP transgenic mice, ameliorates their Aß-induced neuropathology, and, as a result, improves their cognitive function. The effects appeared to be mediated by vaccination-induced antibodies (Abs). Importantly, these results indicating disease-modifying potential of Aß immunization, were consequently corroborated by other investigators (2, 3, 4, 5).

As a next step, ELAN took the approach to the clinic. The vaccine used initially was comparable to the one of the mouse studies. It consisted of the whole Aß42 peptide adjuvanted with QS21 and was termed AN1792. During the clinical development program, the vaccine’s formulation was changed to include the stabilizer PS-80, also known to shift the quality of the immune response induced towards a type-1 phenotype (IFN-γ high, CMI – high, IgG1 high). This change in the vaccine’s formulation was associated with the occurrence of meningoencephalitis (ME) in 6% of vaccine recipients (6). Research into the pathogenesis of the ME cases, carried out by ELAN/Wyeth in collaboration with the FDA, identified Aß-specific, TH1-type CD4+ Τ lymphocytes as the causative agent. This result is supported by many other studies showing that the induction of a cellular immune response tackling cerebral proteins such as myelin basic protein eventually leads to inflammatory processes manifesting as encephalitis (7). Thus, active immunization with the 42 aa Aß peptide under TH1-promoting conditions had activated Τ cells that mediated an autoimmune reaction of clinical significance. However, this does not explain every aspect of the ME complex. In particular, all vaccinees that mounted an Ab response must have had such activated Aß-reactive, TH1-type CD4+ Τ cells (as Τ cell help is needed for an Ab response). A humoral immune response was seen in 22% of vaccine recipients. This number exceeds by far the ME rate of 6% thus indicating the existence of (a) cofactor(s). In analogy to the animal vaccine models, where encephalitis is only seen upon coadministration of Pertussis toxin, an agent known to affect the integrity of the blood brain barrier (BBB), one could incriminate substances disrupting the BBB such as cytokines released during inflammatory processes.

Of note, active immunization of AD patients with AN1792 did not only produce side-effects. Despite the fact that, on average, only 2 rather than the projected 6 vaccinations were applied, the study generated evidence for the proposed activity, that is reduction of the cerebral Aß load as well as a clinical benefit (8, 9, 10, 11, 12). More recently, by relating neuropathological findings with the clinical course in a small cohort of AD patients having received AN1792, Nicoll and colleagues questioned the clinical benefit of sole cerebral Aß reduction (13).

Implications for the design of second generation AD vaccines

The AN1792 example stresses the importance of considering autoimmunity when immunizing with self-proteins, a phenomenon well-known from cancer vaccines (14). As, in the case of AD, the therapeutic principle is based on the Ab response, the task is to build vaccines eliciting Aß-specific Abs without activating Aß-reactive Τ cells. A possible solution to this problem is deduced from Τ cell biology. Τ cells recognize peptides only in association with MHC molecules: CD4+ Τ cells interact with peptides of 14-20 aa length that are bound to MHC class II molecules, while CD8+ Τ cells are activated by 8-9 aa long peptides on MHC class I moieties (15). Thus, the use of short antigenic peptides (i.e., <8aa) is an as effective as simple measure to preclude the activation of specific Τ cells. This general strategy is adopted by most vaccine-developing companies (ELAN/Wyeth, Novartis/Cytos, AFFiRiS).

AFFITOPE AD vaccines integrate all lessons learned from the first human AD vaccine

AFFiRiS takes the AN1792 autoimmunity lesson a step further. Rather than using short fragments of Aß itself, AFFiRiS uses so called AFFITOPES, i.e. short (6 aa) peptides mimicking parts of the native A 42 molecule. By definition, they differ from Aß. As the selected ones do also not exhibit sequence identity with other human proteins, they are “foreign” to the human immune system. This offers a conceptual advantage. It is easier to elicit an immune response towards foreign- as compared to self-proteins as the organism protects “self” from being destroyed by an immune attack via “central” and “peripheral” tolerance mechanisms (16). This allows AFFiRiS to use aluminium hydroxide, the agent first approved as immunological adjuvant for human use and, thus, exhibiting an excellent safety profile. The process of AFFITOPE selection, also termed AFFITOPE-technology, delivers another safety feature to AFFITOPE vaccines. It directs the Ab response towards parts of the Aß molecule that are highly specific, i.e. not present in its precursor APP or other human proteins. This prevents crossreactivity of the Aß-induced Ab response, thus humoral autoimmunity. For example, the first two AD vaccine candidates of AFFiRiS, AFFITOPE AD01 and AD02, target the N-terminus of Aß in a way that the peptide sequence is only recognized if its N-terminus is free. As predicted, sera from mice immunized with AFFITOPE AD01 or AD02 react with Aß but fail to do so with its precursor APP as within APP Aß is flanked at both ends. Consistent with the fact that AFFITOPE AD01 and AD02 do not show sequence identity with human proteins, tissue microarray studies did not reveal any relevant reaction profile of AFFITOPE AD01- and AD02-induced sera. How this is accomplished is illustrated in Fig. 1. Central to the technology is the selection antibody. Its features are imprinted into the AFFITOPE candidates selected with its help from the pool of all possible 6- or 7-mer peptides. Upon immunization, the imprinted features are passed on to the ensuing Abs. Candidates become true AFFITOPES if they are immunogenic, transfer the features of the selecting Ab (AD01 and AD02 - recognition of the Aß peptide sequence only if its N-terminus is free) and are successful in POC studies in APP transgenic mice.

Figure 1.

The AFFITOPE technology

The case of AFFITOPES AD01 and AD02

Using Abs specifically recognizing the free N-terminus of Aß, we identified a series of “N-terminus AFFITOPES”. The set identified and patented most likely encompasses all peptides sharing this feature. Most of them could be manufactured and were found to be immunogenic in mice. Two of them were developed through POC studies. Both exhibited disease modifying potential in that their application to APP transgenic mice (TG2576, TASD41) was found to reduce cerebral Aß load, improved their AD-like neuropathology as well as their cognitive function (M. Mandler, manuscript in preparation). Toxicity studies in rats and guinea pigs (their Aß sequence is identical to the human one) revealed an excellent safety profile for both, AD01 and AD02.

The strategy behind clinical testing of AFFITOPE vaccines

According to current thinking, an agent with Aß-lowering activity possesses disease-modifying potential. Thus, the general strategy has to consider relevant clinical parameters and biomarkers over a sufficient time period using appropriate statistical tools.

As a first step, safety and toxicity of AFFITOPE AD01 and AD02 had to be addressed. The rationale behind the respective trials (AFFiRiS 001 investigates AFFITOPE AD01, AFFiRiS 002 AFFITOPE AD02) was based on the following facts. AFFITOPES AD01 and AD02 were designed to preclude cellular as well as humoral autoimmunity and proved to be safe in animal experiments. True toxicity testing of peptide-based vaccines has to consider immunotoxicity. This requires (i) vaccine administration in a way that allows for inducing an immune response and (ii) application of the vaccine to patients, as only they are expressing the targeted structure, i.e. Aß, in sufficient amounts.

Based on the thinking outlined above, both studies were designed as two-arm, parallel group studies in which participants receive the AFFITOPE/carrier conjugate either on itself or adjuvanted with aluminium hydroxide. Both studies are directed to patients with mild to moderate AD (MMSE 16-26) as defined by the NINCDS/ADRDA criteria. Having passed the screening visit, a given participant is planned to receive 4 vaccinations at 4-week intervals (Figure 2). Each vaccination visit is followed by two structured telephone interviews (1 and 3 weeks later) and an in-person safety visit (2 weeks later). The final visit is performed 8 weeks after the last immunization. Primary objective of either trial is the evaluation of the vaccines’ safety and tolerability. Secondary objective is the assessment of the vaccines’ immunological as well as clinical activity, which – given the number of study participants (n=24/trial) – will be looked at in an explorative manner only.

Figure 2.

Study plan

Safety measures built into the trial include the availability of a caregiver, stepwise enrolment, emergency plans, frequent patient contacts and regular review of the data by an external data safety monitoring board.

Early clinical results support the safety concept realized in AFFITOPE vaccines

AFFiRiS 001, the study assessing AFFITOPE AD01 started late summer 2007. It will be completed November 2008. As of September 2008, all vaccinations have been applied. As of the writing of this report (October 2008), none of the participants had developed signs reminiscent of meningoencephalitis or another SUSAR (suspected unexpected serious adverse reaction). All AEs were of mild to moderate degree, two thirds were found to be local reactions. AFFiRiS 002, evaluating AFFITOPE AD02, started in February 2008. By October 2008, two thirds of the projected study participants had been recruited and had received a total of 49 vaccinations. The safety and toxicity profile was as favourable as the one of AFFITOPE AD01 (A. Schneeberger et al., unpublished results).

Clinical development of AFFITOPE vaccines – the next steps

Each participant of AFFiRiS 001 and 002 is offered follow-up over a period of 1 year to assess long-term effects with respect to the vaccines’ safety and tolerability. Again, in an explorative manner, we will assess the kinetics of the specific immune response and the clinical effects of vaccination with AFFITOPES AD01 and AD02, respectively. Parallel to the long-term follow-up studies, we will optimize the vaccination protocol with regard to the dose applied and the vaccination schedule used. Employing the optimal vaccination protocol, a phase II study will be initiated. This will focus on one of the two N-terminal Aß AFFITOPE vaccines and will be designed to analyze the vaccine’s effect on the pathology underlying AD by assessing the effects on the clinical symptoms as well as currently available AD biomarkers (e.g., MRI volumetry).

Conclusion

There is increasing evidence for the notion that specific immunotherapy can reduce cerebral Aß levels. This enables us to address one of the, if not the most, thrilling question in AD research: Is the reduction of cerebral Aß concentrations in and of itself sufficient to modify the course of the disease? We now have various immunotherapeutic strategies at hand to face that question. Particularly interesting in that regard is the AFFITOPE approach. Specific safety features of the AFFITOPE vaccines and the fact that the AFFITOPE technology delivers multiple AFFITOPES that can be combined encompass the major advantages of this strategy. Moreover, the technology can readily be used to safely tackle other target structures in AD or other indications.

Acknowledgements: This work was supported by grants from the Austrian Research Promotion Agency (FFG; grants 807.619, 809.649 and 811.169) and the Center of Innovation and Technology (ZIT), City of Vienna, Call Life Sciences Vienna 2006.