Radiotherapy Planning of Carcinoma Esophagus - Role of PET Fusion with CT Scan

Abstract

Introduction: Oesophageal cancers are treated with a multimodality approach. The concept of organ preservation has led to concurrent chemoradiation becoming the standard of care in cervical as well as mid-thoracic oesophageal tumors. Radiotherapy can be delivered by various conventionall and with conformal techniques. Defining target volumes adequately is important in radiotherapy planning and delivery of treatment as esophageal and gastroesophageal cancers have a high propensity of loco-regional recurrence. The present study aims to evaluate and compare the dosimetry parameters in patients with cancer esophagus planned on CT-based contours and PET-CT fusion-based contours by conformal technique. Material and Methods: Fifty biopsy-proven cases of cancer esophagus and gastroesophageal junction were selected in our institute between February 2021 to July 2022. After immobilization, spiral CT for simulation was performed. Thereafter, PET was performed in the same treatment position as in CT imaging protocol. The (DICOM) images were transferred to the Eclipse treatment planning system (TPS) and registered. Gross tumor volume was contoured on CT and PET-CT scans, followed by CTV contouring. Volumetric margin given depending on the institutional protocol to account for microscopic tumor extension and mean motion of the lesion to generate planning target volume (PTV). Two PTV volumes were finally contoured that is PTV - PTV-CT and PTV - PET-CT. The organs at risks (bilateral lungs, heart, spinal cord) were generated in accordance with the Radiation Therapy Oncology Group (RTOG) protocol and dose constraints given as per QUANTEC. The dose prescribed to PTV-CT and PTV-PETCT in the range of 45Gy - 59.4Gy, depending upon the site in 25- 33 fractions. Two plans (groups 1 and 2) were generated for comparison and were optimized to maximize the dose to the PTV and limit the dose to normal tissue. PTV dosimetric parameters evaluated were V95, D2, D50, D95, D98, Dmax, Dmean, conformity index and homogeneity index. Dosimetric parameters evaluated for OARs were both lungs combined (D-mean, D-max, V-5, and V-20), heart (D-mean, D-max, and V-40) and spinal cord (D-max). Statistical analysis was done using Paired T-test. A comparison of mean value of dosimetric parameters and P-Value was done. Results: The mean age is 61 years, with male to female ratio 0.7:1. The most common subsite is the mid-thoracic esophagus (46%). Histopathology seen was squamous cell carcinoma (88%) and adenocarcinoma (12%) with a majority having moderately differentiated grade (80%). Median standardized uptake value is 14.9 and the mean is 14.8 (range 0–37.4). The variation of gross tumor size and gross tumor volume on PET-CT scan as compared to CT scan ranged from -1.7 (-44.7%) cm to 2.7 cm (55.1%) and from -11.9 cm3 (-28.6%) to 13.2 cm3 (36.8%) which was not statistically significant (p = 0.38 and p = 0.41 respectively). There was a difference of 18% in the detection of nodes by PET CT scan as compared to CT scan which was statistically significant (p = 0.0001). Increased PTV was seen in 26% of patients (p=0.0001). No difference in dosimetric parameters of PTV was found in terms of V9, D2, D50, D95, D98, Dmax, and Dmean. Similarly, no statistical difference was found in the CI and HI of both plans. Dosimetric parameters of both lungs show statistically significant differences in D mean (14.73 Gy vs 16.13 Gy; p = 0.0005) and V5 (89.16 Gy vs. 28.4 Gy; p = 0.0056). Dosimetric parameters in the heart did not show a statistically significant difference. Dmax for the spinal cord in both groups was within dose constraints (38.7 Gy vs 39.1 Gy) with no statistically significant difference (p = 0.22). Conclusion: Delineation of primary gross tumor is better with PET fusion than as compared to CT alone. Further, PET-CT scans detect more lymph nodes than CT alone. Therefore, by incorporating PET–CT scans in radiotherapy planning of carcinoma esophagus, more accurate and precise treatment planning can be done, which will lead to fewer chances of geographical miss and less chances of loco-regional failure.