Parotid-Sparing Versus Conventional Whole Brain Radiation Therapy: A Multi-Institutional Randomized Clinical Trial.

PURPOSE: Xerostomia (dry mouth) is common after conventional whole brain radiation therapy (WBRT) for brain metastases and is associated with parotid dose. No studies have reported whether parotid-sparing WBRT reduces this toxicity. We assessed whether parotid-sparing WBRT reduces xerostomia compared with conventional WBRT.

METHODS AND MATERIALS: Patients receiving WBRT (30-35 Gy in 10-15 fractions) using 3-dimensional fields were enrolled at 3 academic centers from 2018 to 2021. Patients were randomized to conventional fields (covering C1-C2 vertebrae inferiorly without prospective delineation of parotids) versus parotid-sparing fields (omission of C-spine coverage; parotids delineated and V20Gy limited to 25%). Patients completed the University of Michigan Xerostomia Questionnaire (scored 0-100, higher = worse xerostomia) at baseline, within 5 days of WBRT completion (EndRT), 2 weeks, 1, 3, and 6 months. The primary endpoint was xerostomia score at 1 month (1M), analyzed as proportion with ≥15-point increase and as model-estimated mean xerostomia score. The secondary endpoint was marginal tumor relapse (≤1 cm outside WBRT fields).

RESULTS

A total of 56 patients were randomized and 48 (24 in each arm) were eligible for analysis. Mean parotid dose was 17 versus 10 Gy with conventional versus parotid-sparing WBRT. At 1M, mean xerostomia score was 17 versus 8 (P = .06). The primary endpoint at 1M was not met; the proportion with ≥15-point increase in xerostomia score did not differ significantly between conventional versus parotid-sparing fields (12% vs 16%; P = .56). However, proportion with ≥15-point increase in xerostomia score at EndRT was higher with conventional versus parotid-sparing treatment (52% vs 22%; P = .05). Parotid V5Gy (P = .04) and V10Gy (P = .05) were associated with xerostomia at EndRT. On multivariable analysis, parotid-sparing WBRT was associated with improved xerostomia at EndRT (P = .04) but not at 1M (P = .06). There were no marginal relapses in either arm.

CONCLUSIONS

Parotid sparing without coverage of upper cervical vertebra appears safe and potentially reduces acute xerostomia in patients with limited life expectancy receiving conventional WBRT.