Activation of the complement cascade and pseudo-allergic reactions, by Pr. Dr Jacques Descotes

The term ‘drug-induced hypersensitivity’ is increasingly preferred to ‘drug allergy’ as it is often ambiguous, if not misleading. Drug-induced hypersensitivity reactions are either immune- or non-immune mediated. Immune-mediated – formerly called immuno-allergic – reactions are always unexpected and so far, unpredictable based on nonclinical findings, because they are dose-independent and host-specific. In sharp contrast, non-immune-mediated or ‘pseudo-allergic’ reactions develop rather consistently, are dose/concentration-dependent, and thus, can be predicted using in vitro assays or animal studies, at least to some extent. Either type has been estimated to account for roughly 50% of clinical reactions. It can prove essential to substantiate whether one given reaction is immune-mediated or not. Changing the route of administration (e.g. IV bolus vs. slow infusion), the pharmaceutical formulation, or the dose to be administered to human subjects can, indeed, help eliminate the risk for a drug to induce non-immune-mediated reactions, but definitely not the risk of immune-mediated reactions.

Direct activation of the complement cascade is one among several mechanisms of drug-induced pseudo-allergic reactions. The complement system serving as a first-line host defense against microbial infections is an essential component of innate immunity. It consists of over 30 proteins, the majority of which are inactive zymogens in the plasma that can be sequentially cleaved, hence the notion of complement cascade. Complement activation can occur through the classical, alternate or lectin pathway: the classical pathway is initiated by antigen-associated antibodies, the lectin pathway after binding to microbial pathogen surface involving carbohydrate moieties, and the alternate pathway by immune-mediated, but more often non-immune-mediated triggers. Finally, neutrophil and macrophage proteases as well as plasma factors, such as kallikrein and kinins, can activate the complement cascade. Potentially damaging consequences of activation are continuously checked and prevented by a complex set of regulatory proteins.

The cleavage of inactive complement components leads to a large active fragment that can cleave a target downstream inactive component and a small fragment with variable biological effects. The anaphylatoxins (C3a, C4a, and C5a) are examples of such small fragments. If produced in large amounts, they will cause acute reactions clinically mimicking an anaphylactic shock (hence the term ‘pseudo-allergic reaction’). The anaphylatoxins are very potent triggers of smooth muscle contraction in vessels, bronchi and the gut, vascular permeability, and activation of mast cells resulting in the release of mediators including histamine or TNF-α. Typical clinical manifestations include cough, abdominal pain, flushing and often moderate changes in both heart rate and blood pressure. Breathing difficulties and cardiovascular disorders may also be prominent and sometimes life-threatening. The acronym CARPA for ‘complement activation related pseudo-allergic reaction’ has been coined to refer to these potentially severe reactions.

Because direct activation of the complement is dose/concentration-dependent, these reactions are essentially recorded after intravenous administration. By far, the leading pharmaceutical cause is the solvent Cremophor El° (polyoxyethylated castor oil) involved in reactions associated with the general anesthetic Althesin® (withdrawn from the market in 1984), or IV formulations of the immunosuppressive agent cyclosporine and the anticancer drug paclitaxel. The clinical risk associated with nanomedicines (or nanoparticles) is still poorly established. For instance, huge variability has been reported among liposomes, from large liposomes (>200nm), which are potent activators, to functionalized liposomes (e.g. mPEG-liposomes), which are not. To date, assessing the potential for nanomedicine candidates to activate the complement system using in vitro assays is recommended as a first screen and to meet regulatory expectations. Therapeutic proteins, such as monoclonal antibodies can induce acute infusion reactions. Among the various underlying mechanisms, activation of the complement has been shown to be involved, for instance with the anticancer agent rituximab.

The potential to induce direct complement activation can be assessed in vitro from human blood samples incubated with graded concentrations of the test item. In this context, a CH50 assay has limited, if any value. Laboratory measurements (ELISAs) should preferably focus on fragments, such as C3a, iC3b, C4d, C5a, Bb, SC5b-9. Because any of the 3 activation pathways may be involved, it is recommended to combine several measurements (for instance, C5a, Bb and SC5b-9). Importantly, in vitro results using human blood samples have long been shown to be consistent with clinical findings.