Natural oil-based emulsion containing allantoin versus aqueous cream for managing radiation-induced skin reactions in patients with cancer: a phase 3, double-blind, randomized, controlled trial.

PURPOSE To investigate the effects of a natural oil-based emulsion containing allantoin versus aqueous cream for preventing and managing radiation-induced skin reactions. METHODS AND MATERIALS A total of 174 patients were randomized and participated in the study. Patients received either cream 1 (the natural oil-based emulsion containing allantoin) or cream 2 (aqueous cream). Skin toxicity, pain, itching, and skin-related quality of life scores were collected for up to 4 weeks after radiation treatment. RESULTS Patients who received cream 1 had a significantly lower average level of Common Terminology Criteria for Adverse Events at week 3 (P<.05) but had statistically higher average levels of skin toxicity at weeks 7, 8, and 9 (all P<.001). Similar results were observed when skin toxicity was analyzed by grades. With regards to pain, patients in the cream 2 group had a significantly higher average level of worst pain (P<.05) and itching (P=.046) compared with the cream 1 group at week 3; however, these differences were not observed at other weeks. In addition, there was a strong trend for cream 2 to reduce the incidence of grade 2 or more skin toxicity in comparison with cream 1 (P=.056). Overall, more participants in the cream 1 group were required to use another topical treatment at weeks 8 (P=.049) and 9 (P=.01). CONCLUSION The natural oil-based emulsion containing allantoin seems to have similar effects for managing skin toxicity compared with aqueous cream up to week 5; however, it becomes significantly less effective at later weeks into the radiation treatment and beyond treatment completion (week 6 and beyond). There were no major differences in pain, itching, and skin-related quality of life. In light of these results, clinicians and patients can base their decision on costs and preferences. Overall, aqueous cream seems to be a more preferred option.