Dynamic contrast enhanced CT aiding gross tumor volume delineation of liver tumors: an interobserver variability study

Nikolaj K G Jensen; Danielle Mulder; Michael Lock; Barbara Fisher; Rebecca Zener; Ben Beech; Roman Kozak; Jeff Chen; Ting-Yim Lee; Eugene Wong

doi:10.1016/j.radonc.2014.01.026

Dynamic contrast enhanced CT aiding gross tumor volume delineation of liver tumors: an interobserver variability study

Radiother Oncol. 2014 Apr;111(1):153-7. doi: 10.1016/j.radonc.2014.01.026. Epub 2014 Mar 13.

Authors

Affiliations

¹ Physics & Engineering, London Regional Cancer Program, Canada.
² Radiation Oncology, London Regional Cancer Program, Canada; Department of Oncology, University of Western Ontario, London, Canada.
³ Radiology, London Hospitals, Canada.
⁴ Radiology, St. Joseph's Health Care, London, Canada.
⁵ Physics & Engineering, London Regional Cancer Program, Canada; Department of Oncology, University of Western Ontario, London, Canada; Department of Medical Biophysics, University of Western Ontario, London, Canada.
⁶ Department of Oncology, University of Western Ontario, London, Canada; Radiology, St. Joseph's Health Care, London, Canada; Imaging Research Lab, Robarts Research Institute, London, Canada; Department of Medical Biophysics, University of Western Ontario, London, Canada; Imaging Program, Lawson Health Research Institute, London, Canada.
⁷ Physics & Engineering, London Regional Cancer Program, Canada; Department of Oncology, University of Western Ontario, London, Canada; Department of Medical Biophysics, University of Western Ontario, London, Canada; Department of Physics & Astronomy, University of Western Ontario, London, Canada. Electronic address: ewong4@uwo.ca.

PMID: 24631143
DOI: 10.1016/j.radonc.2014.01.026

Abstract

Purpose: To evaluate the application of perfusion CT for gross tumor volume (GTV) delineation for radiotherapy of intrahepatic tumors.

Materials and methods: 15 radiotherapy patients with confirmed liver tumors underwent contrast enhanced 4D-CT (Philips Brilliance Big-bore) as well as dynamic contrast enhanced (DCE) CT (GE 750HD). Perfusion maps were generated with CT perfusion v5 from GE. Five observers delineated GTVs of all intrahepatic foci on the 4D-CT, time-averaged DCE-CT and perfusion CT for every patient. STAPLE consensus contours were generated. Dice's coefficients were compared between GTVs generated by observers on each image set and the corresponding consensus GTVs. Comparisons were also performed with patients stratified by hepatocellular carcinoma (HCC) metastatic tumors, and by tumor volume.

Results: Overall, mean Dice's coefficients were 0.81±0.14, 0.84±0.10, and 0.81±0.14 for 4D-CT, DCECT and perfusion. DCE-CT performed significantly better than 4D-CT and perfusion (p=0.005 and p=0.01 respectively). For patients with HCC, DCE-CT reduced interobserver variability significantly compared to 4D-CT (Dice's coefficients 0.87 vs. 0.84, p<0.05). For patients with metastatic disease time-averaged DCE-CT images decreased variability compared to 4D-CT (Dice's coefficient 0.81 vs. 0.76, p<0.05), especially true for tumors<100cc. The smaller tumors results are important to be included here.

Conclusions: DCE-CT imaging of liver perfusion reduced interobserver variability in GTV delineation for both HCC and metastatic liver tumors.

Keywords: 4D-CT; CT perfusion; Contouring; Dynamic contrast enhanced CT; GTV delineation; Hepatocellular carcinoma.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Adult
Aged
Aged, 80 and over
Carcinoma, Hepatocellular / pathology*
Carcinoma, Hepatocellular / radiotherapy*
Contrast Media
Female
Four-Dimensional Computed Tomography / methods
Humans
Liver Neoplasms / pathology*
Liver Neoplasms / radiotherapy*
Male
Middle Aged
Observer Variation
Radiotherapy Planning, Computer-Assisted / methods*
Tomography, Spiral Computed / methods
Tumor Burden

Substances

Contrast Media