MO-G-17A-09: Quantitative Autoradiography of Biopsy Specimens Extracted Under PET/CT Guidance

Authors


Abstract

Purpose:

To develop a procedure for accurate determination of PET tracer co ncentration with high spatial accuracy in situ by performing Quantitative Autoradiography of Biopsy Specimens (QABS) extracted under PET/CT guidance.

Methods:

Autoradiography (ARG) standards were produced from a gel loaded with a known co ncentration of FDG biopsied with 18G and 20G biopsy needles. Specimens obtained with these needles are generally cylindrical: up to 18 mm in length and about 0.8 and 0.6 mm in diameter respectively. These standards, with similar shape and density as biopsy specimens were used to generate ARG calibration curves.Quantitative ARG was performed to measure the activity co ncentration in biopsy specimens extracted from ten patients. The biopsy sites were determined acco rding to PET/CT's obtained in the operating room. Additional CT scans were acquired with the needles in place to co nfirm co rrect needle placements. The ARG images were aligned with the needle tip in the PET/CT images using the open source CERR software. The mean SUV calculated from the specimen activities (SUVarg) were co mpared to that from PET (SUVpet) at the needle locations.

Results:

Calibration curves show that the relation between ARG signal and activity co ncentration in those standards is linear for the investigated range (up to 150 kBq/ml). The co rrelation co efficient of SUVarg with SUVpet is 0.74. Discrepancies between SUVarg and SUVpet can be attributed to the small size of the biopsy specimens compared to PET resolution.

Conclusion:

The calibration procedure using surrogate biopsy specimens provided a method for quantifying the activity within the biopsy cores obtained under FDG-PET guidance. QABS allows mapping the activity concentration in such biopsy specimens with a resolution of about 1mm. QABS is a promising tool for verification of biopsy adequacy by comparing specimen activity to that expected from the PET image.

A portion of this research was funded by a research grant from Biospace Lab, 13 rue Georges Auric 75019 Paris, FRANCE.

Ancillary