Development of an Illustrated Scale for Acute Radiation Dermatitis in Breast Cancer Patients

Scales for rating acute radiation dermatitis (ARD) have not been validated despite decades of clinical use, and little is known regarding the relationship between toxicity scores and patient-reported symptoms. Skin tone also complicates assessment of ARD, and as such we sought to design an illustrated scale to consistently describe ARD across several skin tone types in breast cancer patients undergoing radiation (RT).

Patients undergoing RT for breast cancer were enrolled on a prospective study with photographs obtained at 2-week intervals. Photographs were clustered according to the apparent severity of acute radiation dermatitis and a descriptive photonumeric scale was developed. Four clinically experienced raters used both the illustrated photonumeric scale and the Common Terminology Criteria for Adverse Events to independently score the collection of photographs in 2 independent sessions.

Among 80 unique patients with 192 photographs, 47 patients (59%) completed questionnaires about their symptoms during RT. Physicians completed toxicity forms at the point-of-care for 52 patients (65%). Photonumeric ratings compared against patient reports of dry and moist desquamation demonstrated high specificity (95% and 93%, respectively) and negative predictive value (84% and 92%), indicating correct identification of patients who did not report dry or moist desquamation. The sensitivity and positive predictive value for separate measures of dry and moist desquamation were considerably lower. A combined measure of any desquamation (dry or moist) portrayed higher diagnostic accuracy, resulting in 72% sensitivity, 93% specificity, 75% positive predictive value, and 92% negative predictive value. Photonumeric ratings of dry or moist desquamation were significantly associated with patient reports of itching, burning or stinging, hurting, and swelling.

The Michigan scale for acute radiation dermatitis is a simple grading rubric that is distinguished by characterization of its intra- and interrater reliability and diagnostic accuracy, correlation with patient-reported symptoms of bother and pain, and applicability across the spectrum of skin pigmentation.