KEY POINTS
- The retrospective analysis included 34 patients with anal squamous cell carcinoma undergoing definitive chemoradiotherapy and 212 CBCT scans, with a median of six scans per patient.
- All patients were treated with IMRT/VMAT, with mean RT dose 51.1 Gy and an approximately 7-mm isotropic CTV-to-PTV expansion. CBCT registration used bony anatomy, while treatment CTVs were retrospectively recontoured on selected scans.
- Target volume progressively decreased during treatment. Mean CBCT-derived CTV volume was 6.9% smaller than the planning CTV, corresponding to a mean absolute reduction of 71.3 cm³.
- Shrinkage became more pronounced over time: median CTV deviation progressed from approximately −1.5% at fraction 1 to −12.7% at fraction 30, with increasing interpatient variability later in treatment.
- Despite this overall regression, substantial spatial excursions still occurred. 78 of 212 CBCTs (36.8%) contained at least one CTV extension >10 mm beyond the planning PTV.
- Most large deviations involved the primary region: 66.3% occurred in the anal canal and 33.7% in the inguinal region. Anterior displacement was most frequent, accounting for 30.9% of >10-mm deviations, followed by posterior displacement at 25.5%.
- The accumulated CTV envelope remained mostly inside the planned PTV: the mean volume extending beyond the PTV was only 1.7%, and 85% of patients had <3% outside. However, one patient reached 13.9%, illustrating marked patient-specific variability.
- Bladder filling showed large day-to-day variation, including individual volumes above 500 cm³, but bladder volume was not correlated with CTV deviation, suggesting that target variability cannot be explained by bladder preparation alone.
- No dose recalculation was performed. A geometric excursion outside the PTV therefore cannot be equated with target underdosage or worse local control, and the 10-mm threshold was an exploratory geometric definition rather than a validated clinical cutoff.
CLINICAL TAKEAWAY
Anal cancer targets may shrink substantially over six weeks while still moving beyond the original PTV on individual fractions. These data argue against indiscriminate margin reduction and support robust image guidance, with selective adaptive strategies worth investigating for patients showing repeated or pronounced geometric deviations.