A MRI Spatial Mapping Technique for Microvascular Permeability and Tissue Blood Volume Based on Macromolecular Contrast Agent Distribution Franci Demsar, Timothy P. L. Roberts, Heidi C. Schwickert, David M. Shames, Cornelis F. van Dijke, Jeffry S. Mann, Maythem Saeed. Robert C. Brasch A rapid and automated method for two-dimensional spatial depiction (mapping) of quantitative physiological tissue char- acteristics derived from contrast enhanced MR imaging was developed and tested in disease models of cancer, inflamma- tion, and myocardial reperfusion injury. Specifically, an estab- lished two-compartment kinetic model of unidirectional mass transport was implemented on a pixel-by-pixel basis to gen- erate maps of tissue permeability surface area product (PS) and fractional blood volume (BV) based on dynamic MRI in- tensity data after administration of albumin-(Gd-DTPA),, a prototype macromolecular contrast medium (MMCM) de- signed for blood pool enhancement. Maps of PS and BV in disease models of adenocarcinoma. intramuscular abscess inflammation, and myocardial reperfusion injury clearly de- picted zones of increased permeability (up to approximately 500 pllcch-compared to <25 pl/cc/h in normal tissues). As revealed on PS maps, the rank ordering of studied permeabil- ity abnormalities was reperfusion injury > inflammation > tumors. A rapid, automated mapping technique derived from dynamic contrast-enhanced MRI data can be used to facilitate the identification and characterization of pathophysiologic abnormalities, specifically relative increases in blood volume and/or microvascular permeability. Key words: blood pool contrast agent; MRI mapping; perme- ability. INTRODUCTION The use of contrast-enhancing agents in conjunction with MRI provides an opportunity to extract physiological information, in addition to the superb anatomical data offered by unenhanced images. For example, small mo- lecular weight contrast media, represented by Gd-DTPA dimeglumine, have provided clinically useful for MRI depiction of abnormalities in the blood brain barrier (1, 2). Other classes of contrast agents, now under develop- ment, are designed to define the physiology in various tissues, for example, blood pool agents, for microvascular characteristics (3-6), macrophage-directed particles for reticuloendothelial function (7-9), bile-extracted com- pounds for hepatocyte function (lo), paramagnetic cho- lesterol derivatives for adrenal function (111, and iron MAM 37236-242 (1997) From the Department of Radiology, University of California, San Francisco, California.. Visiting from the Jozef Stefan Institute, University of Ljubljana, Ljubljana, Slovenia. Address correspondence to: Franci Demsar, Ph.D., Jozef Stefan Institute, Jamova 39, 1 .OOO Ljubljana, Slovenia.. Received August 1,1995; revised December 7,1995; accepted August 13, 1996. Copyright 0 1997 by Williams & Wilkins All rights of reproduction in any form reserved. 0740-3194197 $3.00 oxide particulates for lymph node function (12, 13). For each of these various types of contrast media and phys- iologic characteristics, optimized MRI techniques and display formats must be developed. Variations in tissue physiology as defined by contrast agent kinetics should be displayed on an anatomic template, in a form easily interpreted by the diagnostic physician. Complex and time-consuming postprocessing techniques may limit clinical utilization (14). In this study a novel, rapid, and automated method for generating anatomical maps of physiological tissue char- acteristics was tested in animals with a range of pathol- ogies. A rodent model of R3230 mammary adenocarci- noma represented neoplasm, an intramuscular sterile abscess represented inflammation, and a myocardium reperfusion injury model represented ischemic disease. All of the above disease models have been previously shown to increase transendothelial microvascular per- meability (15-22). This new mapping technique was im- plemented on dynamic MRI intensity data enhanced with a prototype blood-pool contrast agent, albumin-(Gd- DTPA),,, a macromolecule that has been shown to be useful for quantification of fractional blood volume and endothelial permeability in selected regions (21-23). MATERIALS AND METHODS MR Imaging MRI was performed on an Omega CSI-I1 system (Bruker Instruments, Fremont, CA), operating at 2T. The system is equipped with Acustar S-150 self-shielded gradient coils (20 G/cm, 15-cm inner diameter) and “birdcage” radiofrequency coils. For imaging rat tumor and the in- flammation model, TI-weighted three-dimensional (3D)- pooled gradient-refocused acquisition in a steady state (SPGR) images were obtained with the following param- eter settings: TR 50 ms, TE 3 ms, 1 NEX, flip angle go”, matrix 128 x 128 X 16, section thickness 3 mm, FOV 50 X 50 X 48 mm, acquisition time 102 s. For rat myo- cardial imaging, ECG-gated, T,-weighted single-slice spin echo images were obtained with the following pa- rameter settings: TR = 250 ms, TE 6.8 ms, 2 NEX, matrix 256 X 128, section thickness 2 mm, FOV 70 x 70 mm, acquisition time 50 s. A nonenhanced image set was obtained before contrast medium administration. Subsequently, serial MRI was performed at 2-min intervals for 60 min to generate 30 postcontrast image sets for computer analysis. 236