Can Dynamic Contrast-Enhanced CT Quantify Perfusion in a Stimulated Muscle of Limited Size? A Rat Model.

BACKGROUND Muscle injury may result in damage to the vasculature, rendering it unable to meet the metabolic demands of muscle regeneration and healing. Therefore, therapies frequently aim to maintain, restore, or improve blood supply to the injured muscle. Although there are several options to assess the vascular outcomes of these therapies, few are capable of spatially assessing perfusion in large volumes of tissue. QUESTIONS/PURPOSES Can dynamic contrast-enhanced CT (DCE-CT) imaging acquired with a clinical CT scanner be used in a rat model to quantify perfusion in the anterior tibialis muscle at spatially relevant volumes, as assessed by (1) the blood flow rate and tissue blood volume in the muscle after three levels of muscle stimulation (low, medium, and maximum) relative to baseline as determined by the non-stimulated contralateral leg; and (2) how do these measurements compare with those obtained by the more standard approach of microsphere perfusion? METHODS The right anterior tibialis muscles of adult male Sprague Dawley rats were randomized to low- (n = 10), medium- (n = 6), or maximum- (n = 3) level (duty cycles of 2.5%, 5.0%, and 20%, respectively) nerve electrode coupled muscle stimulation directly followed by DCE-CT imaging. Tissue blood flow and blood volume maps were created using commercial software and volumetrically measured using NIH software. Although differences in blood flow were detectable across the studied levels of muscle stimulation, a review of the evidence suggested the absolute blood flow quantified was underestimated. Therefore, at a later date, a separate set of adult male Sprague Dawley rats were randomized for microsphere perfusion (n = 7) to define blood flow in the animal model with an accepted standard. With this technique, intra-arterial particles sized to freely flow in blood but large enough to lodge in tissue capillaries were injected. Simultaneously, blood sampling at a fixed flow rate was simultaneously performed to provide a fixed blood flow rate sample. The tissues of interest were then explanted and assessed for the total number of particles per tissue volume. Tissue blood flow rate was then calculated based on the particle count ratio within the reference sample. Note that a tissue's blood volume cannot be calculated with this method. Comparison analysis to the non-stimulated baseline leg was performed using two-tailed paired student t-test. An ANOVA was used to compare difference between stimulation groups. RESULTS DCE-CT measured (mean ± SD) increasing tissue blood flow differences in stimulated anterior tibialis muscle at 2.5% duty cycle (32 ± 5 cc/100 cc/min), 5.0% duty cycle (46 ± 13 cc/100 cc/min), and 20% duty cycle (73 ± 3 cc/100 cc/min) compared with the paired contralateral non-stimulated anterior tibialis muscle (10 ± 2 cc/100 cc/min, mean difference 21 cc/100 cc/min [95% CI 17.08 to 25.69]; 9 ± 1 cc/100 cc/min, mean difference 37 cc/100 cc/min [95% CI 23.06 to 50.11]; and 11 ± 2 cc/100 cc/min, mean difference 62 cc/100 cc/min [95% CI 53.67 to 70.03]; all p < 0.001). Similarly, DCE-CT showed increasing differences in tissue blood volumes within the stimulated anterior tibialis muscle at 2.5% duty cycle (23.2 ± 4.2 cc/100 cc), 5.0% duty cycle (39.2 ± 7.2 cc/100 cc), and 20% duty cycle (52.5 ± 13.1 cc/100 cc) compared with the paired contralateral non-stimulated anterior tibialis muscle (3.4 ± 0.7 cc/100 cc, mean difference 19.8 cc/100 cc [95% CI 16.46 to 23.20]; p < 0.001; 3.5 ± 0.4 cc/100 cc, mean difference 35.7 cc/100 cc [95% CI 28.44 to 43.00]; p < 0.001; and 4.2 ± 1.3 cc/100 cc, mean difference 48.3 cc/100 cc [95% CI 17.86 to 78.77]; p = 0.010). Microsphere perfusion measurements also showed an increasing difference in tissue blood flow in the stimulated anterior tibialis muscle at 2.5% duty cycle (62 ± 43 cc/100 cc/min), 5.0% duty cycle (89 ± 52 cc/100 cc/min), and 20% duty cycle (313 ± 269 cc/100 cc/min) compared with the paired contralateral non-stimulated anterior tibialis muscle (8 ± 4 cc/100 cc/min, mean difference 55 cc/100 cc/min [95% CI 15.49 to 94.24]; p = 0.007; 9 ± 9 cc/100 cc/min, mean difference 79 cc/100 cc/min [95% CI 33.83 to 125.09]; p = 0.003; and 18 ± 18 cc/100 cc/min, mean difference 295 cc/100 cc/min [95% CI 8.45 to 580.87]; p = 0.023). Qualitative comparison between the methods suggests that DCE-CT values underestimate tissue blood flow with a post-hoc ANOVA showing DCE-CT blood flow values within the 2.5% duty cycle group (32 ± 5 cc/100 cc/min) to be less than the microsphere perfusion value (62 ± 43 cc/100 cc/min) with a mean difference of 31 cc/100 cc/min (95% CI 2.46 to 60.23; p = 0.035). CONCLUSIONS DCE-CT using a clinical scanner is a feasible modality to measure incremental changes of blood flow and tissue blood volume within a spatially challenged small animal model. Care should be taken in studies where true blood flow values are needed, as this particular small-volume muscle model suggests true blood flow is underestimated using the specific adaptions of DCE-CT acquisition and image processing chosen. CLINICAL RELEVANCE CT perfusion is a clinically available modality allowing for translation of science from bench to bedside. Adapting the modality to fit small animal models that are relevant to muscle healing may hasten time to clinical utility.