determine how fast the drug is cleared from the blood stream in the presence and absence of ADAs. It can be assumed that significant amounts of ADAs could strongly reduce circulation time and serum concentration of the monoclonal antibody. An ELISA based on murine TNF-α, the antigen to the monoclonal antibody, was performed. The recovery of muAb in the formulations prior to injection was determined. In all formulation, except for the formulation containing only insoluble aggregates after light exposure, substantial amounts of muAb were quantified with the ELISA (data not shown). Figure 8-7 provides an overview of the results obtained for muAb pharmacokinetics.
The mice that received the placebo formulation or the formulations containing the insoluble aggregates showed no quantifiable amounts of muAb in the serum samples, neither in C57BL/6 nor in BALB/c mice (data not shown). In C57BL/6 mice both, the native muAb formulation and the native muAb plus adjuvant formulation, show serum levels that do not point to the generation of ADAs. The maximum serum levels of muAb were measured on day 6, after three booster injections on days 1, 2 and 4. Later on, when only administered weekly, the levels decreased slowly reaching the basic level after 6 weeks washing out phase. Since monoclonal antibodies are characterized by half-lifes of several days to weeks the results obtained for those groups are reflecting a regular, undisturbed pharmacokinetic profile.
However, the results for the native muAb formulation and the adjuvant formulation were considerably different in BALB/c mice, were muAb was only detected on day 6. This is similar in mice of groups 2 and 3 (see Figure 8-7, A2 and B2). Neither after administration of native muAb formulation nor after administration of native muAb plus adjuvant circulating muAb molecules were detectable at any other bleeding time point than day 6.
A B
160
The repetitive administration of the formulation containing soluble aggregates resulted in no detectable free circulating muAb molecules in C57BL/6, whereas levels of muAb were detectable in BALB/c. The absorbance ratios in group 4 of BALB/c were lower compared to those on day 6 in animals of groups 2 or 3 and had a maximum on day 13 that was constantly decreasing until on day 27 the basic value was achieved again.
C57BL/6 mice Group BALB/c mice
(2) Native muAb
(3) Native muAb
+ adjuvant
(4) Light Solubles
Figure 8-7 – Absorbance unit ratios of PK ELISA in C57BL/6J and BALB/c mice.
A) After receipt of the native muAb formulation. B) After receipt of the formulation consisting of native muAb + adjuvant. C) After receipt of the soluble aggregates after 48 h of light exposure.
A1
B1
C1
A2
C2 B2
161 8.4 DISCUSSION
The study for the first time investigates the immune response to protein aggregates in two different wild-type mouse strains. The C57BL/6 animals were selected, since the monoclonal antibody investigated in this study was generated in this strain. Additionally BALB/c mice were selected, because of their well-known strong Th2 immune response, resulting in strong antibody elicitation [Watanabe et al., 2004]. The subcutaneous route of administration was chosen as this is commonly used in patient’s therapy using monoclonal antibodies besides the intravenous administration [Berger et al., 2002]. Also, the subcutaneous administration is, in general, thought to be more immunogenic than other administration routes [Wierda et al., 2001].
The antibody elicitation after administration of five different aggregate formulations of muAb was investigated. Those formulations were prepared under different conditions and consisted of aggregates differing in size and structural properties. According to the immunon model hypothesized by Dintzis et al., aggregates of large sizes and repeating epitopes on their surface are capable to induce an immune response [Dintzis et al., 1976]. The mechanisms behind immune responses to protein aggregates have not yet been clarified. Currently, administrative organizations and industrial research focus on anti-drug antibodies that potentially can have an impact on the efficacy of the drug, altering pharmacokinetics and thus the success of the therapy. Thus, it is supposed, that aggregates with structural properties close to the native protein are foremost crucial in the formation of ADAs. However, it is entirely conceivable that antibodies towards aggregate-specific structures can be generated as well, which probably do not have a large impact on the pharmacokinetics of the therapeutic protein, but might lead to adverse effects such as allergic reactions.
The structure and size of the administered species was investigated beforehand. The secondary structure of the native and the aggregated muAb samples was investigated by FTIR after the completion of the stress procedure, but not of the final injection samples, since the final concentration of 25 µg/mL of total protein is too low for FTIR analysis. Of the five formulations containing aggregates, three contained protein species with a strongly modified secondary structure (Groups 4, 5 and 8). Exposure to light (Groups 4 and 5) as well as storage at elevated temperatures (Group 8) resulted in strong changes of the protein as indicated by unfolding, whereas after shaking (Group 7) and stirring (Group 6) the protein mostly maintained its secondary structure.
Unfolding and structural modifications might imply the formation of antibodies that are not capable of binding to muAb. In this case, the ADAs will not be detected in this assay.
Therefore, in theory the subcutaneously injected formulations of aggregated antibody in this study should still be able to bind to the antigen, indicating a native-like structure and ensuring the generation of antibodies directed towards the original native structure of the drug molecule.
162
For the duration of the study, the animals were maintained in SPF conditions. Despite the repetitive administration of the native muAb to the host strain, the meticulous cleaning and preparation of all instruments and samples was suggested to keep constantly low background levels of immunoglobulins. The results detected in C57BL/6 mice indicated the absence of pathogens; however these mice are known to elicit fewer antibodies when compared to other mouse strain [Mills et al., 2000]. The study only determined the generation of antibodies directed against the native muAb, but not towards the aggregates administered to the mice.
Therefore, the results do not reflect the total levels of IgG isotype in the serum samples, but the levels specifically directed towards the muAb molecule.