FDG-PET-CT reduces the interobserver variability in rectal tumor delineation

Area of Science:

• Oncology
• Radiotherapy
• Medical Imaging

Background:

• Previous studies demonstrated a correlation between pathology and automatically generated PET contours in rectal cancer.
• Interobserver variation in gross tumor volume (GTV) definition is a challenge in rectal cancer radiotherapy planning.

Purpose of the Study:

• To analyze the effect of PET-CT on interobserver variation in GTV definition for rectal cancer.
• To evaluate the influence of PET-CT on clinical target volume (CTV) and treatment volumes in rectal cancer.

Main Methods:

• Forty-two rectal cancer patients underwent FDG-PET-CT for radiotherapy planning.
• GTVs were delineated by 5 observers using CT, MRI, and PET (with and without auto-contours).
• Concordance indices (CI) and GTV volumes were compared; CTV and treatment volumes were analyzed for PET influence.

Main Results:

• PET-based GTV volumes were significantly smaller, and CIs increased significantly with PET use.
• The best interobserver agreement was achieved using PET auto-contours.
• PET-CT altered CTV definition in up to 29% of patients and allowed tailoring of the caudal treatment border, especially in low-seated tumors.

Conclusions:

• PET-CT enhances interobserver agreement in GTV definition for rectal cancer.
• PET-CT aids in avoiding geographical misses and allows precise tailoring of treatment volumes.
• The use of PET-CT, including auto-contours, improves radiotherapy planning accuracy for rectal cancer.