Dissertation Abstract

Multi-modality imaging for improved staging of prostate cancer

The treatment of patients with prostate cancer depends heavily on determining disease extent, but such information is difficult to obtain with conventional empirical staging methods or structural imaging. ¹¹¹ In-ProstaScint ® is a nuclear medicine agent that binds to the prostate-specific membrane antigen (PSMA), enabling it to identify prostate cancer cells anywhere in the body. Despite this useful biological property, ¹¹¹ In-ProstaScint ® suffers limitations inherent to nuclear medicine studies: poor photon statistics, attenuation, collimator blurring, and non-specific uptake. To overcome these limitations, we developed a method of ¹¹¹ In-ProstaScint ® imaging based on a combined CT/SPECT system. This system acquires CT and SPECT images of the patient sequentially and registers the images together using fiducial markers. The CT provides anatomic localization, which can discriminate potential tumor sites from non-specific uptake, and is also used to generate a patient-specific attenuation map, which models the dual-energy emissions from ¹¹¹ In. This attenuation map and a model of the distance-dependent collimator resolution are incorporated into an iterative SPECT reconstruction that increases the quality and quantitative accuracy of the SPECT image. Components of the reconstruction are also used to apply template-projection reconstruction techniques for measuring the activity concentration in CT-defined regions of interest.

We tested the quantitative accuracy of the CT/SPECT system by imaging phantoms that simulated ¹¹¹ In-ProstaScint ® patients. In these phantoms, we were able to measure the activity concentration of simulated tumors to within 10% accuracy for large (3.2 cm) objects and 30% to 40% accuracy for smaller (1.8 cm and 1.2 cm) objects. We also performed a pilot clinical study of combined CT/SPECT ¹¹¹ In-ProstaScint ® imaging on patients undergoing conventional ¹¹¹ In-ProstaScint ® SPECT studies. The iterative reconstruction (with corrections for photon attenuation and collimator blurring) significantly improved the quality of the nuclear medicine images. Furthermore, in three of thirteen patients, the ability to register CT and SPECT images together clarified SPECT images that might otherwise have produced equivocal results. Quantitative measurements of ¹¹¹ In-ProstaScint ® uptake did not correlate well with empirical staging algorithms predicting lymph node metastases; however, future studies with a larger patient population could help illuminate the relationship between these two different measures of disease.

This dissertation citation and abstract are published with permission of ProQuest Information and Learning. Further reproduction is prohibited without permission.