SGRT reduced pelvic pitch and longitudinal setup errors in vulvar radiotherapy

In 20 vulvar cancer patients, SGRT reduced pitch error from 1.21° to 0.07° and longitudinal error from 4.9 to 0.4 mm.

KEY POINTS

  • This retrospective study included 20 patients receiving definitive or adjuvant radiotherapy for vulvar squamous cell carcinoma in the frog-leg position. Ten patients were positioned using conventional skin marks and lasers, and ten received additional surface-guided radiotherapy.
  • Setup accuracy was evaluated from 500 daily helical fan-beam CT datasets, using the first 25 fractions from each patient. Registration was performed in two stages: initial alignment to pelvic bones and femora, followed by manual soft-tissue adjustment around the vulvar and nodal targets.
  • Surface guidance produced its largest improvement in pelvic pitch. Mean pitch error fell from 1.214° ± 0.481° with conventional positioning to 0.069° ± 0.325° with surface guidance, with p < 0.001 and a very large effect size of Cohen’s d = 2.79.
  • Mean longitudinal displacement decreased from 0.487 ± 0.256 cm to 0.036 ± 0.241 cm, corresponding to a reduction from approximately 4.9 mm to 0.4 mm. This difference was significant at p < 0.001, with Cohen’s d = 1.82.
  • Vertical error was also reduced, from 0.285 ± 0.462 cm to 0.010 ± 0.150 cm (p < 0.001; d = 0.80). No significant differences were observed for lateral translation, roll, or yaw.
  • In the conventional group, body mass index correlated strongly with pitch and yaw errors, with Spearman ρ = −0.81 and p = 0.005 for both measurements. No statistically significant association between body mass index and setup error remained in the surface-guided group.
  • The study did not directly measure target dose, organs-at-risk dose, toxicity, or clinical outcomes. Its proposed reduction of planning target volume margins from 5–7 mm to 3–5 mm therefore remains hypothetical and requires validation based on soft-tissue targeting and dosimetric consequences.

CLINICAL TAKEAWAY

Surface guidance appears useful for detecting pelvic tilt and longitudinal sliding that may be missed by tattoos in the biomechanically unstable frog-leg position. The magnitude of geometric improvement is notable, but the cohort included only 20 patients, and the data do not yet justify routine margin reduction without institution-specific validation.

SOURCE

Journal of Applied Clinical Medical Physics