Imagify™ Clinical Program Overview
The overall objective of the Imagify™ (perflubutane polymer microspheres) injectible suspension clinical program was to demonstrate that Imagify with cardiac stress ultrasound provides equivalent efficacy to nuclear stress testing for detection of CAD. Eleven clinical studies have been conducted to date, including two pivotal studies.
Summary of Pivotal Results
The pivotal studies, RAMP-1 and RAMP-2, were designed to demonstrate that Imagify with ultrasound (“AI-700 ECHO”) has equivalent efficacy to nuclear stress testing (“SPECT”). RAMP-1 and -2 enrolled patients at intermediate risk of CAD, the intended patient population, with disease prevalence of approximately 50% in each trial. Acusphere evaluated three efficacy endpoints using a hierarchical statistical analysis: 1) the overall primary endpoint was accuracy, i.e., all the patients in trial, whether or not they had disease. If the accuracy endpoint was met, then additional analyses were performed: 2) patients with disease (i.e. sensitivity) and 3) patients without disease (i.e. specificity). There were three ultrasound blinded readers in each trial and one nuclear blinded reader in RAMP-1 and 3 nuclear blinded readers in RAMP-2. Blinded readers had access to images only (e.g., no access to patient history or other information). The truth standard in both trials for most patients was angiography.
Primary Efficacy Endpoint: Accuracy
The primary efficacy endpoint was to demonstrate that ultrasound with Imagify had non-inferior accuracy to nuclear stress testing. The criteria for success were that two out of the three ultrasound readers be non-inferior to the nuclear reader. Imagify exceeded the criteria of success in both trials because all three of the ultrasound readers were non-inferior to the nuclear reader as demonstrated in the following chart where each bar represents the accuracy of each reader in the trial:
Accuracy of All Six Imagify Readers Was Non-Inferior to SPECT
Analysis of Sensitivity and Specificity
Since the first endpoint (accuracy) in the hierarchical analysis had been met, the sensitivity and specificity of AI-700 were also evaluated. In a non-inferiority analysis of sensitivity and specificity, AI-700 ECHO did not meet the criteria for success as only one AI-700 ECHO reader in each trial was non-inferior to the SPECT reader. However, when the non-inferiority endpoint for one parameter was missed, the other parameter in the same trial was shown to be superior to SPECT. This is a typical result caused by reader bias, where conservative ultrasound readers were matched with aggressive nuclear readers and vice versa.
A number of other analyses clearly demonstrate the equivalence of AI-700 sensitivity and specificity to SPECT.
Receiver Operating Characteristic (ROC) Curve
The chart below shows all of the data from the RAMP-1 and 2 trials in an ROC curve, a standard technique used to compare testing techniques while removing the biases of individual readers. Sensitivity and specificity are inversely related and therefore readers routinely trade-off sensitivity and specificity based on their own judgment and biases. Aggressive readers read with higher sensitivity than specificity and conservative readers read with higher specificity than sensitivity. ROC curves are used to display the data to remove these individual reader biases. In the chart below, one curve represents the performance of AI-700 ECHO readers and the other curve represents the performance of the SPECT (e.g. nuclear) readers. The shape of the curves, as well as the area under the curve (AUC), is the same, demonstrating that ultrasound with Imagify has equivalent performance to nuclear stress.
AI-700 ECHO Has the Same Sensitivity and Specificity as SPECT
Pooled Analysis
In a pooled analysis of all the ECHO and SPECT readers from both trials using weighted mean averages, the sensitivity and specificity of AI-700 ECHO was the same as that of SPECT as shown in the chart below:
AI-700 ECHO Has the Same Sensitivity and Specificity as SPECT
Safety Results
Imagify™ has been administered to over 1,000 patients with suspected CAD and was well tolerated. The majority of adverse events (AEs) were mild or moderate in intensity, transient, and resolved without treatment and without residual effects. Eleven patients had non-life threatening serious AE’s that were treated with standard techniques routinely available in a stress echo lab. Because of the time of onset, type and severity, most adverse events were associated with the use of a pharmacological stressor, dipyridamole. A safety signal was observed in a small minority of patients who experienced cardiovascular AEs (primarily hypotension and dyspnea) sometimes leading to discontinuation. These were consistent with complement activation, the body’s natural pathway for removing particulates from the blood. These were generally self-limiting, and resolved spontaneously or with normal treatment. Similar AEs have been reported for approved ultrasound contrast agents.
The overall AE profile of the AI-700 pharmacologic perfusion stress echo was similar to current pharmacologic stress imaging techniques in common use (SPECT or dobutamine echo).