3408 - Sacral Plexus Dose Constraints in the SBRT Era: A Retrospective Analysis of Radiation-Induced Lumbosacral Plexopathy

Purpose/Objective(s):

Radiation-induced lumbosacral plexopathy (RILSP) is a rare but potentially debilitating late toxicity of pelvic radiotherapy. Existing sacral plexus dose constraints are largely extrapolated from conventional fractionation data and surrogate neural structures, with minimal prospective validation. We hypothesize that the risk of RILSP following pelvic SBRT is low and associated with high-dose sacral plexus subvolume exposure.

Materials/Methods:

We retrospectively reviewed 64 patients treated with SBRT involving the lumbosacral plexus region (lower lumbar spine or pelvis) between 2023 and 2025. Patients received either 3-fraction (f) SBRT (n=21) or 5f SBRT (n=43). The sacral plexus was contoured using the following standardized contouring guidelines. Extracted dosimetric parameters included D5cc, D0.5cc, and D0.1cc. The primary endpoint was clinically and radiographically defined RILSP. Continuous variables were compared between patients with and without RILSP by the Mann-Whitney test.

Results:

Median age was 73.2 years. Median follow-up was 10.5 months. Median prescription dose was 30.0 Gy (range 21.0 – 36.0 Gy) in 3f and 37.5 Gy (range 25.0 – 50.0 Gy) in 5f. In the 3f cohort, median (interquartile range) D5cc and D0.1cc sacral plexus doses were 18.1 Gy (14.7, 20.3) and 23.4 Gy (22.3, 24.0) in 3f. No cases of RILSP were observed in the 3f cohort. Two cases of RILSP occurred (overall incidence 3.1%), both in the 5f cohort; neither patient received concurrent systemic treatment. Within the 5f cohort, sacral plexus dose metrics were numerically and statistically higher among patients who developed RILSP (n=2) compared with those who did not (n=41). Median (IQR) D5cc was 33.7 Gy (30.9–36.5) versus 26.6 Gy (22.3–29.0), respectively (p<0.01). Corresponding D0.5cc values were 41.8 Gy (31.9–51.7) versus 29.7 Gy (27.5–31.2) (p<0.05), and D0.1cc values were 42.6 Gy (32.2–52.9) versus 31.6 Gy (29.6–32.0) (p<0.05). After conversion to EQD2 (a/ß = 2.5), differences remained pronounced. Among patients without RILSP (n=62), median (IQR) EQD2 D5cc, D0.5cc, and D0.1cc were 57.3 Gy (45.7–61.7), 63.3 Gy (58.6–68.1), and 66.2 Gy (60.9–70.6), respectively. In contrast, patients who developed RILSP (n=2) had substantially higher EQD2 values of 71.1 Gy (70.2–72.0), 98.4 Gy (74.4–122.4), and 101.0 Gy (75.2–126.8), respectively (all p<0.01).

Conclusion:

In this contemporary SBRT cohort, RILSP was uncommon and occurred only in patients with elevated sacral plexus near-maximum doses in the 5f setting. In our experience, maintaining sacral plexus D0.1cc =32 Gy and D0.5cc =30 Gy in 5f was associated with a <5% observed risk of plexopathy. These findings support the safety of commonly used SBRT constraints while underscoring the need for longer follow-up and prospective validation.