Note: In this Q&A we adopt the convention that substances "bright" on T1-weighed images have short T1 values. This is true for most (but not all) pulse sequences. For a detailed explanation, click here.
The deposition of cholesterol in the walls of vessels has been a topic of extensive scientific investigation over the last several decades. As such, a wealth of MRI and MRS data concerning cholesterol derives directly from research on atherosclerotic plaques. This body of knowledge reveals that cholesterol exists in several forms and states, each with different imaging characteristics.
At body temperatures, atherosclerotic plaque lipids are simultaneously present in an isotropic liquid state (cholesterol esters and triglycerides), liquid crystalline state (phospholipid/cholesterol bilayers), and solid state (cholesterol crystals). The narrow-line MRS signals from plaque arise predominantly from the liquid lipid phase (mostly cholesterol esters). Cholesterol in all its states has a short T1 value and range of chemical shifts similar to other lipids. The T2 values of cholesterol, however, differ significantly according to state.
Liquid cholesterol esters have fairly short T2 values (~5-10 msec), but these times are long enough to be detectable/imageable by routine MRI. By contrast, the solid and crystalline forms of cholesterol associated with cell membranes have extremely short T2 times (measured in microseconds); their signals decay too quickly for detection using routine MR techniques.
The earliest MR investigations into the imaging of atherosclerotic plaque were focused on detecting these short T1 cholesterol and lipid components. It was soon recognized, however, that non-hemorrhagic plaques were not particularly bright on T1-weighted images. The reason for this is that water content in the atheromatous core is 10x higher than lipid, thus dominating the MR signal. Only in recent years have advanced techniques (including diffusion and water suppression methods) allowed some headway to be made into detecting and imaging these lipid components.
Craniopharygiomas frequently contain high T1 signal cystic areas, and it is common to hear or read that cholesterol is at least partially responsible for this. Indeed, these cysts often do contain large amounts of cholesterol which surgeons can see as golden flecks floating on a sea of "motor-oil" colored fluid. In 1992 Ahmadi et al compared the MR appearance to cyst content in 10 cases and showed that the T1 signal was unrelated to cholesterol or lipid content. Instead, the short T1 seen in craniopharyngiomas is likely due to high protein content, blood products (methemoglobin), or both.
Cholesterol granulomas are uncommon lesions that typically occur at the apex of the petrous temporal bone. These lesions are characteristically bright on T1-weighted images and do contain very high concentrations of cholesterol. However, the cholesterol they contain is not a liquid ester, but is crystalline (with T2 values so short as to render their signals undetectable). Further proof that lipids are not responsible for the MR appearance is that these lesions do not have a surrounding chemical shift artifact nor suppress with fat saturation pulses. The high T1 signal is instead due to the paramagnetic effects of blood products characteristically associated with these lesions.
Cholesterol gallstones serve as a final example, which are well known to have low signal on both T1 and T2 weighted images. This is because the cholesterol in gallstones is in a solid state, with very short T2 values. A bright gallstone on T1-weighted images is therefore most likely a bilirubin stone, whose T1 is shortened by paramagnetic blood products.
Advanced Discussion (show/hide)»
A few random additional comments. . .
The structure and characteristics of solid crystalline forms of cholesterol have been investigated, but using ex vivo techniques of solid state NMR including magic angle spinning (MAS).
Cholesteatomas, in spite of their name, do not contain cholesterol.
Colloid cysts are often T1-bright and some do contain cholesterol. However, they also frequently contain mucin, blood products, and other lipids, so the role of cholesterol remains in doubt.
Other cholesterol-containing tumors include hepatocellular carcinoma and pancreatic acinar cell carcinoma. Neither is bright on T1-weighted images.
Pearlman JD, Zajicek J, Merickel MB, et al. High-resolution ¹H NMR spectral signature from human atheroma. Magn Reson Med 1988;7:262-279.
Toussaint J-F, LaMuraglia GM, Southern JF, et al. Magnetic resonance images lipid, fibrous, calcified, hemorrhagic, and thrombotic components of human atherosclerosis in vivo (full text link). Circulation 1996; 94:932-938.
Choudhury RP, Fuster V, Badimon JJ, et al. MRI and characterization of atherosclerotic plaque. Emerging applications and molecular imaging. Arterioscler Thromb Vasc Biol 2002;22:1065-1074.
Tsai H-M, Lin X-Z, Chen C-Y et al. MRI of gallstones with different compositions. AJR 2004;182:1513-1519.
Ahmadi J, Destian S, Apuzzo MLJ, et al. Cystic fluid in craniopharygiomas: MR imaging and quantitative analysis. Radiology 1992; 182:783-785
What is T1 relaxation?
Can you explain a little more about the dipole-dipole interaction? I still don't quite understand.