Volumes Changes in Adrenal Metastases during Delivery of 5-Fraction Stereotactic Magnetic Resonance-Guided Adaptive Radiotherapy

Stereotactic MR-guided adaptive radiotherapy (SMART) improves adrenal target coverage and reduces organ at risk doses. Changes in adrenal tumor volumes were observed during contouring for daily-adaptive planning. We hypothesized that such changes in GTV may correlate with long-term outcomes.An Ethics approved institutional database was accessed for details of 5-fraction adrenal SABR treatments on a 0.35T MR-Linac between 2016-2023. Volumetric MR-scans were acquired at baseline, and prior to each of 5 fractions. Manually contoured GTVs were accessed. Volume changes between simulation and first fraction (ΔPF1), as well as between first and subsequent fractions (ΔF1F2 to ΔF1F5) were computed. Daily relative adrenal volume change (DRVC) and the absolute volume change (DAVC) before treatment, were calculated by dividing ΔPF1 by the number of days between simulation and F1. Paired Wilcoxon tests were used to compare volumes at different timepoints. Univariate linear regression (LinR) models were created to study per-treatment volume changes. A logistic regression (LogR) analysis explored associations between baseline characteristics and volumetric changes with radiological complete response (CR) at the date of last news.Adrenal SMART in 5 fractions of either 8 or 10Gy was delivered to 70 patients. The median interval between simulation and F1 was 13 days (range, 2-35), and 13 days between F1-F5 (range, 9-31). Solitary metastasis comprised 44%; metastases were mainly from non-small cell lung (64%) or renal (9%) primaries. Concurrent systemic therapy (± 3 months) was delivered in 41% of patients. Mean baseline GTV at simulation and F1 were respectively 38.5 and 41.4cc (p<0.001). Mean ΔPF1 was +9.1% (-10.2% to +50.1%) or +2.9cc (-4.4 to +27.4cc). Mean DRVC was +0.8% per day (-1.0% to +8.4%), and did not correlate with histology, GTV or concurrent systemic treatment. Mean DAVC was +0.26cc per day (-0.31 to +2.4cc), and was associated with GTV size (p = 0.04). In total, 47% of adrenal GTVs decreased in volume at F5 compared to F1. A ≥10% increase in GTV was observed in 39% of patients during SMART. GTV variations of ≥20% occurred in 36% patients at some point during SMART. At a median follow-up of 20.3 months, radiological CR was seen in 23% of 64 evaluable patients. Radiological CR was significantly associated with baseline GTV (p = 0.03) and ΔF1F5 (p = 0.03), but not DRVC (p = 0.6) or ΔF1F3 (p = 0.3). Local relapses were seen in 6% of patients.Changes in adrenal volumes during SMART were common, with variations ≥20% occurring in 36% patients. The probability of a radiological CR correlated with both baseline GTV size and median relative volume decline between fractions 1 and 5. These changes support use of daily SABR plan adaptation.