Carboplatin skin testing: a skin-testing protocol for predicting hypersensitivity to carboplatin chemotherapy

A high incidence of moderate to severe hypersensitivity reactions (HRs) is noted in patients who have been treated with multiple courses of carboplatin. Presently, there is no reliable way to predict which patients may be at risk for this potentially severe adverse reaction. We developed a skin-test protocol to identify patients at high risk for HR to carboplatin chemotherapy.

Patients undergoing more than seven courses of carboplatin received a 0.02-mL intradermal injection of an undiluted aliquot of their planned carboplatin infusion 1 hour before each course of the agent. A positive skin test was prospectively defined as that resulting in a wheel of at least 5 mm with a surrounding flare. We recently reported a 27% incidence of HRs in patients receiving more than seven courses of carboplatin. These patients served as historical controls for the current study.

Forty-seven patients with recurrent ovarian or primary peritoneal carcinoma receiving carboplatin were skin tested. Thirteen of 47 patients (28%) manifested a positive skin test at a median of nine total courses of carboplatin (range, eight to 17 courses). This rate of skin-test positivity was not significantly different from the incidence of documented HR reported in a historical control group (P =.89), suggesting comparable populations. A negative skin test accurately predicted the absence of HR in 166 of 168 courses of chemotherapy. Only two of 47 patients (4%) experienced a HR after a negative skin test. Thus, administering carboplatin only to patients with a negative skin test may result in a significant reduction in HRs relative to historical controls (P =.002).

An easily performed skin test appears to predict patients in whom carboplatin may be safely administered. Treatment modifications based on the results of skin testing may reduce the incidence of HRs in patients receiving repeated courses of carboplatin.