Cetuximab-Associated Infusion Reactions: Pathology and Management
Cetuximab-Associated Infusion Reactions: Pathology and Management
ABSTRACT: Cetuximab (Erbitux), a chimeric antiepidermal growth factor receptor monoclonal antibody currently used to treat metastatic colorectal cancer, is in clinical development for several other solid tumors. Although cutaneous manifestations are the most common toxicities associated with cetuximab, they are rarely life-threatening. Cetuximab-related infusion reactions are less common, but they may become severe and cause fatal outcomes if not managed appropriately. Little about the specific etiology of these events is known; however, an overview of infusion reactions observed with other compounds may shed some light and help characterize cetuximab-related reactions. For physicians administering cetuximab, familiarity with acute reaction treatment protocols and preparedness to identify and manage symptoms promptly and effectively are most important to minimize potential risks.
The epidermal growth factor receptor (EGFR) is an important therapeutic target in treating cancer. Overexpression of EGFR, an indicator of poor prognosis,[1,2] occurs in several malignancies, including colorectal, pancreatic, and head and neck cancers, as well as non-small-cell lung cancer (NSCLC).[3,4] Recently, agents that include small-molecule tyrosine kinase inhibitors and monoclonal antibodies have been developed that target EGFR. Randomized trials with cetuximab (Erbitux), a chimeric anti-EGFR monoclonal antibody composed of the variable regions of a murine anti-EGFR antibody with human immunoglobulin (Ig) G1 constant regions, have demonstrated important antitumor activity in metastatic colorectal cancer and head and neck cancer.
In irinotecan (Camptosar)-refractory metastatic colorectal cancer, disease control (defined as response or stable disease) was achieved in 32% of patients on cetuximab monotherapy and 56% of those given a combination of cetuximab and irinotecan. In head and neck cancer, cetuximab therapy plus radiation prolonged survival when compared with radiotherapy alone (median survival: 54 vs 28 months; P = .02).
Although its overall safety profile is favorable, cetuximab therapy has been complicated by acute infusion-related reactions. This review will describe the types of acute infusion-related reactions reported with cetuximab, compare them with infusion-related reactions seen with other antineoplastic agents, and suggest recommendations for managing these reactions in the clinic.
Cetuximab Safety Overview
The overall safety profile of cetuximab is favorable; however, adverse events (Table 1) have been reported in clinical trials of cetuximab given alone for advanced colorectal cancer. The majority of toxicities were of grade 1 or 2 severity, as defined by the Cancer Therapy Evaluation Program Common Toxicity Criteria for Adverse Events (CTEP-CTCAE), version 2.0. Current criteria for evaluating these toxicities are presented in Table 2.
The most common toxicity related to cetuximab therapy is an acneiform rash noted among 72% to 90% of all advanced colorectal cancer patients treated with cetuximab in clinical trials. The majority of these rashes were mild or moderate in nature, although 6% to 18% of patients developed a grade 3 or 4 rash.[5,9-12] This dermatologic manifestation probably results from inhibition of EGFR in the skin; the histology of rash lesions usually involves neutrophilic infiltration in dermal tissues (particularly in the infundibular region of the hair follicle) and loss of differentiation in the stratumcorneum of the epidermis, without presence of comedones. This unique histology indicates this rash is a pathologic entity different from acne or hypersensitivity-related manifestations.
Infusion reactions of all severities occurred in approximately 20% of colorectal cancer patients participating in trials using cetuximab alone or with irinotecan. Approximately 3% of patients developed severe reactions (grade 3 or 4), and a small number of these events (< 0.1% of all patients) were fatal.[5,7,9-12]
Preliminary reports of randomized trials with cetuximab in advanced head and neck cancer and NSCLC indicate that the safety profile of cetuximab is very similar in these settings; further, the monoclonal antibody has not worsened toxicity associated with conventional cytotoxic or radiation therapies. Among these trials, grade 3/4 reactions described as infusion reactions or hypersensitivity reactions occurred in 2% to 7% of patients.[6,14,15]
Reactions to Monoclonal Antibodies
Monoclonal antibodies can be divided into four categories based on the origin of their different domains: animal, chimeric, humanized, or human (Figure 1).
Monoclonal antibodies derived from animal sources, most commonly mice, elicit the formation of human antimouse antibodies (HAMAs) and are substantially immunogenic, which limits multiple dosing schedules of these agents.[18-20] Subsequent investigations have focused on engineering antibodies to limit or eliminate the murine portions in the molecule. Three general methods currently are used to achieve this goal: (1) combining the full murine variable region of an antibody with the human constant region to create a chimeric antibody, (2) adding only the complementarity-determining regions (or hypervariable regions) of the murine variable regions to a human IgG framework to create a humanized antibody, and (3) using phage-display technology to screen for human-derived hypervariable regions to create a fully human monoclonal antibody.[21-24] The formation of HAMAs is reduced with these newer molecules-HAMAs developed in approximately 75% of patients during early trials of murine monoclonal antibodies,[18-20] whereas reported rates of HAMA development using modern molecules fell below 15%.[7,25,26]
Although these newer therapeutic monoclonal antibodies feature lower immunogenicity, they can still induce various types of acute infusion-related reactions with unclear underlying mechanisms.[5,27-29] Many of these reactions are mild-to-moderate in severity; however, severe reactions occur with an incidence of < 5%, generally.[7,30,31] Rarely, patients have died from these reactions.[7,27,31]
Infusion Reactions: Clinical Experience Across Therapeutic Agents
Infusion reactions occur with a wide variety of therapeutic agents, including cytotoxic agents and biologic agents such as monoclonal antibodies.[27,32-34] The characteristics of these reactions vary in symptoms, time course, and severity. A small, yet significant, percentage of patients experience acute, severe, infusion-related, potentially life-threatening adverse reactions.
As physicians integrate cetuximab into their cancer treatment strategies, they must become familiar with the identification and management of cetuximab-associated reactions. We will review clinical experience involving similar agents. Observations related to agents that have been available longer than cetuximab contribute to our current understanding of the potential mechanisms behind these reactions and may help us to manage cetuximab-related infusion reactions more effectively.
Depending on their underlying immunologic basis, severe infusion-related reactions may be categorized as either anaphylactic or anaphylactoid in nature. Fundamentally, anaphyl-actic reactions are IgE-mediated, whereas anaphylactoid reactions are not.
Anaphylactic reactions are caused by rapid release of immune mediators from tissue mast cells and peripheral blood basophils triggered by IgE. Anaphylactoid reactions, also referred to as pseudoallergic reactions, mimic the symptoms of anaphylactic reactions. They are caused by the direct release of mediators from mast cells and basophils in the absence of specific IgE antibodies or antigen-antibody bridging on the mast or basophil cell membrane.
Clinically, "anaphylaxis" typically describes a rapid systemic reaction that involves a constellation of symptoms including wheezing/bronchospasm, angioedema (particularly laryngeal edema), hypotension, pruritus, urticaria, gastrointestinal sequelae (nausea, vomiting, pain, diarrhea), uterine contractions, and/or direct cardiac effects such as arrhythmias. Symptoms typically appear within moments of exposure, although, less commonly, the reaction may be delayed for 30 to 60 minutes or longer. Because these symptoms have a rapid onset and life-threatening potential, the clinician must be familiar with them and their time course to identify severe cases as quickly as possible.
Careful examination of the similarities and differences between reactions associated with other agents may yield valuable insights on the etiology and management of cetuximab-related reactions. Perhaps the best known type of reaction occurs with the use of paclitaxel,[32,36] a small-molecule taxane that blocks cell division by inhibiting tubulin depolymerization and that currently is used across tumor types, from those involving the breast to NSCLC. Paclitaxel reactions may be caused by either direct mast cell degranulation by the drug's excipient, Cremophor EL (a polyethoxylated castor oil) or the involvement of sensory neuropeptides. These reactions differ from those observed with monoclonal antibodies in time course and symptomatology.
The most data on infusion reactions exists for the monoclonal antibody trastuzumab (Herceptin). Trastuzumab inhibits HER2, a growth factor receptor from the same family as EGFR, and currently is used against HER2-overexpressing breast tumors.
Table 3 compares the characteristic infusion-related reactions associated with each of these compound classes.[25,27,32,33,39-42] Key timing features distinguish the onset of these reactions. Trastuzumab reactions typically are associated with the first infusion and generally occur within the first 2 hours after administration. Paclitaxel-related reactions also tend to occur shortly (10 minutes) after administration of the first infusion.[27,39,40] Reactions to oxaliplatin (Eloxatin), however, usually do not appear at first infusion; instead, they occur after repeated administration.[27,41]
Some characteristics of acute infusion-related reactions associated with monoclonal antibodies suggest that they may not be IgE-mediated. First, these reactions typically occur with the first infusion; the lack of prior sensitization argues against development of specific IgE antibodies against these agents.[27,43] Second, rechallenges have resulted in repeated reactions in few patients.[27,43] In retrospective studies of trastuzumab, 39 patients who suffered serious infusion reactions were rechallenged, and 33 (85%) of these successfully continued therapy. Further, in a single-center report using infliximab (Remicade), an antitumor necrosis factor chimeric antibody, 21 patients who experienced mild or moderate infusion-related reactions were able to receive further therapy if given prophylactic medications. Of three patients who were retreated following severe reactions, two had no further reactions, and one had a repeated severe acute reaction. Third, the lack of IgE-mediated anaphylaxis was further corroborated by findings that serum tryptase levels were normal in 11 patients who experienced 14 acute infusion reactions after receiving infliximab.
Different mechanisms may occur in different patients; severe reactions may be anaphylactic, whereas less severe reactions would be anaphylactoid. Data concerning infusion reactions linked to radiographic contrast material suggested this variable etiology—a retrospective study found that specific IgE is present in < 4% of patients who have reactions. However, in a separate study, specific IgE was found in all patients suffering severe reactions, and histamine and tryptase release correlated with the severity of the reaction.
Cetuximab-Associated Infusion Reactions
Clinical trials testing cetuximab used alone or with irinotecan in advanced colorectal cancer provide the most complete characterization of infusion-related reactions associated with the monoclonal antibody. Grade 1 or 2 infusion reactions occurred in 16% of patients receiving cetuximab plus irinotecan and in 19% of those receiving cetuximab alone; these included chills, fever, and dyspnea. The incidence of grade 3 or 4 reactions associated with cetuximab in clinical trials ranged from 2.3% to 5%.[5,9,10]
Severe reactions typically were characterized by rapid onset of airway obstruction (bronchospasm, stridor, hoarseness), urticaria, and/or hypotension. Approximately 90% of severe reactions were associated with the first infusion; however, some were associated with later infusions. These reactions occurred despite use of prophylactic antihistamines and were not reliably predicted by the use of a test dose (20 mg of cetuximab was administered IV over 10 minutes prior to the initial loading dose in all patients enrolled in clinical trials). Fatalities, although rare (occurring in < 0.1% of patients), did occur.
Other limited data are available from clinical studies of cetuximab in various tumor types and a cetuximab safety review.[5,9,10,14,15,45-47] In these reports, infusion reactions were generally described as being hypersensitivity, allergic, anaphylactic, or anaphylactoid reactions that were characterized by such clinical symptoms as shortness of breath and chest tightness.
A few trials have provided information on the time course, management, and/or outcome of patients experiencing these reactions. The safety review reported that allergic reactions occurred in 31 of 419 patients (7.4%) treated with cetuximab. Of these reactions, 11 (2.6%) were of grade 3 severity and 6 (1.4%) were of grade 4 severity and were described as being anaphylactic reactions. The majority of all reactions occurred with the first dose, but 8 of the 25 patients (33%) who developed grade ≥ 3 reactions did so with later infusions. Conversely, all of the grade 4 reactions occurred within minutes of the first infusion.
The true etiology of cetuximab infusion-related reactions remains unclear. As with reports of such reactions related to the use of other monoclonal antibodies, the fact that the majority of reactions occur with the first dose would suggest that these reactions are not IgE-mediated. One possible explanation for this discrepancy is that patients who develop reactions may have been exposed to a murine antigen (11% to 27% of children in inner cities in the United States are sensitized to mouse allergen). Furthermore, different mechanisms may be involved in grade 4 reactions than in grade 1-3 reactions, since the time course appears to differ between the two. Data on rechallenged patients who experienced mild-to-moderate reactions supported a lack of IgE-mediated effect; however, rechallenge data for severe reactions are sparse.