A67-16 Characterizing Pulmonary Embolism From Non-Contrast Computed Tomography Imaging in an Animal Model

Abstract Rationale Outcomes for patients with acute Pulmonary embolism (PE) is dependent on a timely diagnosis. While computed tomography angiography (CTA) is the primary PE diagnosis modality, patients who are allergic to iodinated contrast agents can have significantly longer time between hospital administration and CTA acquisition. CT-derived perfusion imaging (CTP) is an image processing-based method for quantifying pulmonary perfusion from non-contrast CT imaging. We hypothesize that CTP imaging can serve as a contrast-free alternative to CTA for patients with iodinated contrast allergies. We conducted a pre-clinical study to test our hypothesis. Methods CTP algorithms identify mass changes between temporal phases in thoracic 4DCT images as a surrogate for perfusion. Using deformable image registration, we generate a vector map between successive 4DCT phases. After accounting for volume changes, the difference in Hounsfield units (HU) represents the mass change between two image phases. We acquired micro-4DCT images in a cohort of three female C57BL/6 mice. In each animal, we identified peak inhale and exhale CT phases. CTP images were constructed using these inhale/exhale pairs, representing healthy perfusion. Gadolinium alginate hydrogels (Gd-Alg) were electrosprayed to produce 1-mm radiopaque beads and surgically injected into the external jugular vein to serve as synthetic emboli, one per animal. Following surgery, the mice were imaged with the same protocol to compute CTP images of PE. Results One mouse was excluded from the analysis as fluid was identified in the right middle lobe prior to surgery. Of the remaining mice (n = 2), one mouse had an embolism in the left lung (Mouse A), and the other in the right (Mouse B). In mouse A, we saw a 7% CTP drop in the PE lung, and a 12% increase in the healthy lung. Mouse B had an 8% CTP drop in the PE lung and a 12 percent increase in the healthy lung. Local CTP pre- and post-embolism can be visualized in Figure 1. Conclusions Our results indicate that CTP can be an effective alternative to CTA for diagnosing PE. Interestingly, the unaffected lung in both cases had increased CTP, indicating the healthy lung may be compensating for functional decreases in the PE lung. We observed overall decrease in CTP in the PE lungs, primarily distal to the embolism. Near the embolisms, CTP drastically increased. We hypothesize this stems from an inflammatory response or increased blood pressure resulting from blockage to a major artery, which will be investigated in future studies. This abstract is funded by: National Cancer Institute through MD Anderson’s Cancer Center Support GrantP30CA016672