Paclitaxel Hypersensitivity: It's the Vehicle That Gets You There


Patients receiving paclitaxel frequently develop signs and symptoms of hypersensitivity including dyspnea, flushing, rash, chest pain, tachycardia, hypotension, and generalized urticaria. Consequently, paclitaxel is usually administered slowly over three or 24 hours following premedication with glucocorticoids and histamine antagonists. Despite these efforts at prevention, more than 40% of patients still experience some minor reactions such as flushing or rash. The mechanism of these side effects is unclear. Some have hypothesized that the reaction is IgE-mediated classical immediate hypersensitivity, and others have suggested that direct damage to mast cells or basophils results in histamine release independent of IgE. In addition, the stimulus to the reaction is unclear. Is it paclitaxel or the vehicle, cremaphor EL, that is used to solubilize it?

The suggestion that IgE is responsible for the hypersensitivity response to paclitaxel is made somewhat less plausible by the fact that in most patients, the response is seen with the initial exposure. Of course, it may be possible that paclitaxel looks a lot like an antigen that these people have seen before and that has elicited an IgE response. However, this explanation is also made less plausible by the high incidence of reactions. IgE-mediated hypersensitivity is just not this common. The reaction to paclitaxel bears some resemblance to the reactions seen as a consequence of complement activation as occurs in some patients at the start of extracorporeal circulation in hemodialysis, plasmapheresis, and cardiopulmonary bypass. Activated complement components, particularly C3a, can act as anaphylatoxins and directly release histamine from mast cells and basophils. This similarity between paclitaxel reactions and syndromes related to complement activation prompted Szebeni and colleagues to examine whether paclitaxel and/or the cremaphor EL/ethanol vehicle in which it is dissolved could activate complement in vitro.

Szebeni et al examined the capacity of several compounds to activate serum complement: paclitaxel (Taxol) dissolved in cremaphor EL and 50% ethanol, cremaphor EL and alcohol, cremaphor EL alone, alcohol alone, docetaxel (Taxotere) in polysorbate 80 (Tween 80), and cyclosporine in cremaphor EL. Complement activation was measured by quantitating two complement components: a soluble protein S-bound form of the terminal complex called SC5b-9, which measures C5a formation, a consequence of both the classical and the alternative pathways, and Bb, an activated form of factor B that is generated as a consequence of only the alternate pathway. Any agent (i.e., cyclosporine or paclitaxel) dissolved in cremaphor EL and 50% alcohol significantly elevated both SC5b-9 and Bb levels in serum. The alcohol was insufficient to activate complement alone, but it boosted the complement-activating capacity of cremaphor EL. Both classical and alternative pathways were activated. Docetaxel only activated complement when it was used together with 50% ethanol.

A soluble form of the complement receptor (sCR1) was able to inhibit the activation of complement by paclitaxel and by the vehicle in a dose-dependent fashion. Clinical studies are underway to see if administration of sCR1 to patients receiving paclitaxel prevents the hypersensitivity reactions. (Szebeni J, et al. J Natl Cancer Inst 1998;90:300-306.)


Cremaphor EL is made by reacting castor oil with ethylene oxide under conditions of increased heat and pressure. The final product is comprised by unmodified castor oil and a variety of polyethylene glycols, polyethoxylated glycerols, polyethoxylated fatty acids, and glycerol esters with ricinoleic acid. A major problem in the clinical development of drugs that are natural products is formulating them into drugs that can be solubilized and administered intravenously. Cremaphor EL has been useful as a vehicle. However, based on these data, it appears to be responsible for complement activation and the anaphylotoxin-mediated release of histamine to a degree sufficient to induce systemic vascular reactions. At least this is a hypothesis that can be tested.

These results suggest that sCR1, a truncated soluble form of the complement receptor (CD35), might prevent the effect. sCR1 inhibits complement activation by accelerating the decay of C3 and C5 convertases and by promoting the proteolysis of C3b and C4b by factor I. It is active at in vitro concentrations that should be readily achievable in vivo.

If it does become possible to block the hypersensitivity reactions that have occurred from paclitaxel infusion, it may be necessary to go back and evaluate bolus injection in phase I and II studies.