2128

There is a need for an early, effective, robust and patient-friendly method to monitor the PDE of anticancer agents in patients (PATS), especially of AA. DCE-MRI has been used for such measurements (e.g. AACR #5399, 2003), and we believe that this method, when used properly, is ready for wider applicability in both drug development and serial patient monitoring. Administration of an MRI contrast agent (COA) such as Gd-DTPA, allows the visualization of the vasculature into which that solute distributes, and using a fast acquisition method, one monitors the rate of influx of blood into tumor and other organs, as well as the retention and the rate of efflux. We believe that high temporal resolution (e.g. 1 second) is critical to capture the early influx phase, which appears to be most critically associated with PAT response. Data acquisition is done as follows: upon COA administration as a bolus (15 ml Gd-DTPA at 2 ml/sec), MR images are acquired using a volume sequence with a temporal resolution of 1 second for the first 2 minutes, followed by data acquisition every 15-30 seconds for another 20 minutes. Data analysis is the next critical step. The time course of proton enhancement (PRE) can be evaluated in any desired Region-of-Interest (ROI) using programs such as ImageJ, an image analysis software developed at the NIH (http://rsb.info.nih.gov/ij), and the data (e.g., mean vs time) allow the estimate of the influx rate (ICAR, Initial Contrast Accumulation Rate) and of the efflux rate (DCAR, Delayed Contrast Accumulation Rate). Statistical programs have been incorporated into ImageJ using a plugin implemented by one of us (JAP). Alternatively, commercially available software may be used for estimating these rates. Full details of this methodology and examples of results obtained will be presented. Another important consideration is that the value of the PRE depends on the local concentration of the COA and on the relaxivity of that particular tissue and its effect on the T1 relaxation process. Because it is still difficult to obtain accurate T1 values, our approach relies on the changes in the ICAR and the DCAR in a given PAT during the course of that PATs treatment. In a cooperative study we and others had validated that in the absence of treatment DCE-MRI measurements are reproducible to better than 20% (ASCO 20:#303,2001), and hence, that a change >25%is statistically significant. An alternate method for data analysis uses a compartmental model approach (Tofts, JMRI 10, 223, 1999). Because a single model may not be equally suitable for the analysis of all ROI’s, we believe that at this time the determination in the changes of the slopes (ICAR and DCAR) provide an effective, robust, reliable and patient friendly method that can and should be used in both drug development studies as well as in PAT monitoring.

[Proc Amer Assoc Cancer Res, Volume 45, 2004]