Risk of Radiation Dermatitis in Patients with Skin of Color Who Undergo Radiation to the Breast or Chest Wall Irradiation and Regional Nodes

Abstract

<h3>Purpose/Objective(s)</h3> Radiation dermatitis (RD) is common after RT for breast cancer with data indicating potentially worse RD in African American (AA) patients (pts). Current measures of RD, such as the CTCAE, do not include hyperpigmentation, which may disproportionately affect how RD is classified and treated in pts with skin of color (SOC). We aim to characterize RD in SOC and identify factors, including baseline skin pigmentation (BSP) that predict RD. <h3>Materials/Methods</h3> Pts treated with whole breast (WB) or chest wall (CW) with regional nodal RT or high tangents with 50 Gy in 25 fractions from 2015-2018 were identified. Three dermatologists independently classified BSP using photographs from CT simulation based on the Fitzpatrick scale ([FS], range=I-VI; I=light/pale white to VI=black/ very dark brown). SOC was defined as FS IV-VI. Pt characteristics were investigated for association with interventions to treat RD, clinician-graded acute RD, and late skin toxicity (NCI CTCAE scale) with Chi-squared and logistic regression analyses. <h3>Results</h3> 325 pts met eligibility criteria (58 African American [AA], 42 Asian, 151 Caucasian, 77 other). 40% (n=129) had SOC, 60% underwent CW RT, 40% WB RT and 82% had systemic therapy. Pts with SOC were more likely to be Hispanic (14% vs 8% p=0.007), AA (43% vs 1%, p<0.001) and have greater mean BMI (28.0 vs 26.5, p=0.02). Acute grade 2/3 RD was lower in SOC (FS I 60%, FS II 63%, FS III 52%, FS IV 64%, FS V 40%, FS VI 41%; p=0.049). Increased BSP (OR 0.83; p=0.01) and AA pts (OR: 0.22; p<0.001) had lower odds of acute grade 2/3 RD, whereas bolus and dosimetric parameters such as increased PTV volume had increased odds. On multivariable analysis (MVA), AA pts and bolus remained significant (OR: 0.14, p=0.01; OR: 6.63 p<0.001, respectively). Topical steroid use to treat RD was less frequent and oral analgesic use was more frequent in SOC (43% vs 63%, p<0.001; 50% vs 38%, p=0.05, respectively). Pts with increased BSP (OR 0.73, p<0.001), AA race (OR 0.19, p<0.001) and greater BMI had lower use of topical interventions whereas any boost phase, bolus, IMN RT and increased PTV volume had greater use. On MVA, AA pts (OR 0.27, p=0.04), boost (OR 2.04, p=0.033), IMN RT (OR 2.73, p=0.003) and PTV V105% (OR=1.002, p=0.03) retained significance. Late grade 2/3 hyperpigmentation was greater in SOC (16% vs 3%, p=0.01). Increased BSP (OR 2.14, p=0.001), AA pts (OR 8.18, p=0.02), bolus and CW boost had greater odds of grade 2/3 hyperpigmentation. On MVA, increased BSP (OR: 3.76, p=0.03) and bolus (OR: 14.1, p=0.01) retained significance. <h3>Conclusion</h3> We found less clinician-graded acute RD in SOC and AA pts, less frequent use of topical interventions but more oral analgesic use. We also found higher rates of late pigmentation change with increased BSP independent of race. These findings suggest that RD may be under-diagnosed in SOC. This study confirms the necessity for objective measures of RD that account for variability in BSP to accurately classify the severity of radiation skin toxicity in SOC and treat accordingly.