KEY POINTS
- This retrospective single-centre planning study included 20 patients with locally advanced esophageal cancer, 19 of whom had squamous cell carcinoma. Each patient underwent separate CT-only and FDG-PET/CT-guided target delineation and treatment planning.
- PET/CT-guided GTV definition used metabolically active disease with anatomical correlation, while CTV margins remained unchanged. Both alternative plans prescribed 50.4 Gy in 28 fractions, allowing the effect of target definition itself to be compared.
- Mean PTV decreased substantially from 602.3 ± 267.4 cm³ with CT alone to 453.2 ± 237.4 cm³ with PET/CT (p<0.001), a mean reduction of approximately 149 cm³. Most patients had smaller PET-guided volumes, although some increased because additional PET-positive disease was incorporated.
- Importantly, PET/CT did not simply shrink targets. In 4/20 patients (20%), more than 100 cm³ of the PET-defined PTV lay outside the original CT-based PTV, and additional metabolically active lymph nodes that would not have been covered by the CT-based fields were identified.
- Geometric agreement between CT and PET/CT targets was only moderate, with mean Dice coefficient 0.74 and HD95 22.1 mm. Eight patients (40%) had HD95 >25 mm, demonstrating that PET integration frequently changed not just target size but its spatial configuration.
- Smaller and more individualized PET-guided targets translated into lower lung exposure: mean lung V20 fell from 10.39% to 7.85% (p=0.001) and V30 from 4.41% to 2.46% (p=0.003). Heart V30, heart D1cc and LAD mean dose were numerically lower but not significantly different.
- Interobserver agreement was broadly similar, although PET/CT showed slightly better geometric consistency: mean Dice was 0.86 vs 0.85 and HD95 13.3 vs 15.8 mm, with the greatest apparent improvement in the less experienced observer.
- The study included only 20 patients and assessed planning endpoints rather than local control, survival or clinical toxicity. It therefore supports PET/CT as a contouring aid but cannot establish that the observed geometric changes improve patient outcomes.
CLINICAL TAKEAWAY
PET/CT may matter in esophageal RT planning for two opposite reasons: it can reduce unnecessary elective volume while simultaneously revealing metabolically active disease outside CT-defined targets. The 20% rate of major PET-defined extension is clinically provocative, but this remains a small retrospective planning study without outcome validation.