Renal Cysts May Artificially Enhance Due to Surrounding Parenchymal Enhancement

ABSTRACT & COMMENTARY

Synopsis: Lack of enhancement of renal lesions of homogeneous low (water) attenuation is a key criterion in the CT diagnosis of a renal cyst. However, following the administration of intravenous radiographic iodinated contrast material, some true cysts may falsely appear to enhance as the surrounding renal parenchyma increases in attenuation.

Source: Maki DD, et al. Renal cyst pseudoenhancement: Beam-hardening effects on CT numbers. Radiology 1999;213:468-472.

Ct performed in a dedicated manner to identify the maximum number of renal lesions and assess their imaging characteristics is generally obtained during the maximum degree of renal parenchymal enhancement. This serves two purposes: 1) to maximize the attenuation difference (and hence conspicuity) between normally enhancing parenchyma and the lesser enhancement of hypovascular or avascular lesions; and 2) to produce the optimal conditions for demonstrating subtle enhancement in low-vascularity but neoplastic lesions. The absence of this latter enhancement (taken as enhancement < 10 HU, though some have argued < 15 HU) is accepted as evidence of avascularity (i.e., evidence that the lesion is a renal cyst).

When a lesion fails to enhance, that is accepted as one criterion for the diagnosis of a renal cyst. However, is the converse true? Do renal cysts never enhance? Many radiologists have experienced the apparent enhancement of a lesion that is shown by imaging or surgery to be a renal cyst, and have difficulty explaining the phenomenon. Is it due to technique? To "beam-hardening?" To slice misregistration? To partial volume effects?

To investigate these questions, Maki and colleagues developed a phantom to test the phenomenon. They created an abdominally shaped polyethylene phantom containing a central, hollow, kidney-sized cylinder that could be filled with various concentrations of iodine solutions to simulate varying degrees of renal parenchymal enhancement. Suspended within the "kidney" cylinder were water-filled straws or balloons of various sizes. Some of these water-filled cysts were spherical (to allow for testing of partial-volume effects) and some were cylindrical (so CT slices through them would be free of partial-volume effects). Helical imaging was performed through the cylindrical cysts and both axial and helical CT imaging were performed through the spherical cysts. Attenuation levels were measured.

For the cylindrical cysts there was a progressive increase in attenuation as the background attenuation increased, and the effect was greater (though not uniformly so) in smaller-diameter cylinders than in larger-diameter cylinders. For example, as the average background "renal" attenuation rose by 90 HU, the attenuation of 5- or 10-mm-diameter cysts increased by 11-17 HU compared to an increase of 7-9 HU in cysts of 15- to 30-mm diameter. For a background attenuation increase of 180 HU, which can be commonly achieved during helical CT at peak renal enhancement, cyst attenuation rose by 18-28 HU in the smaller cysts and by 10-15 HU in the larger cysts. Similar findings occurred with the spherical cysts, and there was an added effect from partial-volume effects when a small cyst (10 mm) was scanned with relatively wide (7 mm) collimation. In almost all experiments, there was no substantial difference in the attenuation measurements between axial and helical CT scans. Maki et al also performed a computer simulation to show that the beam-hardening effect explained little of the observed effects. Their conclusions include a need for better CT reconstruction algorithms to cure the pseudoenhancement phenomenon but, until that occurs, a modification of the 10-HU threshold for designating the presence of lesion enhancement to 15-20 HU and greater consideration for MRI to characterize small intrarenal cysts.

Comment by James H. Ellis, MD

Classic teaching about the interpretation of abdominal CT scans states that a renal cyst can be diagnosed if a renal lesion is homogeneous in attenuation, has attenuation near that of water, has smooth margins without internal nodules, has an imperceptible wall, and does not enhance following administration of intravenous radiographic iodinated contrast material. Not all of these criteria can always be assessed in every examination. For example, many patients are scanned without preliminary noncontrast imaging, making the assessment of lesion enhancement problematic. Some lesions are entirely surrounded by parenchyma and, thus, the thickness of the lesion wall cannot be determined; for these intrarenal lesions, the assessment of contrast enhancement takes on greater importance. For difficult-to-assess lesions, dedicated thin section pre- and post-contrast CT imaging of the kidneys is often recommended, which permits an assessment of enhancement along with the other criteria. Helical CT (with its rapid scanning ability) is considered better for this purpose than standard axial CT because of expected reduction in patient motion (especially respiratory) artifact, the ability to flood the kidneys with a high level of contrast material administered in a short high-flow intravenous bolus, and the ability to reconstruct slices at any table index to ensure comparable pre- and post-contrast images through the lesion center.

This article by Maki et al (along with presentations about in vitro and in vivo studies by other groups at national meetings)1,2 forces us to re-evaluate our understanding of the enhancement (or rather, lack of enhancement) criterion for the diagnosis of a renal cyst. While lack of enhancement still remains a useful diagnostic element for a renal cyst, the presence of enhancement greater than the commonly accepted standard of 10 HU should not be blindly taken as evidence of a more serious renal lesion.

Not often does bench research translate so immediately to the clinical situation as in this study, though it is not without its limitations. Maki et al tested only one CT scanner, so the magnitude (lesser or greater) of the effect on CT scanners of other models or other manufacturers is not known. For a more complete understanding of the effects of renal lesion pseudoenhancement, testing should include alterations in slice scan time, mA, and kV. A more anatomic phantom would have had the "kidney" off-center, paired, and kidney-shaped (rather than cylindrical), and would have had representations of the spine, fat, and other organs. Lesions could be tested at the edge of the phantom kidney and not just in the center. However, in my opinion, additional testing is more likely to expand the applicability of Maki et al’s work than to contradict it.

This work by Maki et al begs the question of whether scanning at peak renal enhancement may in fact be the wisest course. Yuh and Cohan discussed the similar conspicuity of all but the smallest (and presumably clinically insignificant) renal lesions in the excretory phase of renal enhancement (i.e., with less parenchymal enhancement) than in the nephrographic phase, with its greater parenchymal enhancement.3 Now we learn from Maki et al that there may be less pseudoenhancement of renal lesions if we scan when the parenchyma is less enhanced (perhaps by changing the timing of the scan or by changing the duration or volume of the contrast material bolus). The question of optimal technique for dedicated renal CT remains to be determined, but the present work clearly points to a re-evaluation of the criteria for renal cyst enhancement using the current technique.

Maki et al speculate that renal lesion pseudoenhancement results from an inadequate correction for beam hardening in the CT reconstruction algorithm. Were this to be mended, both the criterion for nonenhancement of a renal cyst and the optimal technique for dedicated renal CT might have to be redetermined. That is in the future. For now, it suffices to be aware that cysts may falsely appear to enhance and that rigid adherence to cyst nonenhancement criteria may lead to error. Confirmation with alternate tests such as ultrasound or MRI, or follow-up over time, may help in management of selected patients who have renal lesions in which enhancement may be more apparent than real.

References

1. Coulam CH, et al. Comprehensive evaluation of pseudoenhancement of renal cysts during contrast-enhanced CT. Am J Roentgenol 1999;172(S):31.

2. Bae KT, et al. CT assessment of renal cysts: Are attenuation values artifactually increased on contrast-enhanced images? Radiology 1999;213(P):175.

3. Yuh BI, Cohan RH. Different phases of renal enhancement: Role in detecting and characterizing renal masses during helical CT. Am J Roentgenol 1999;173: 747-755.