Applied Radiology

Paul M. Silverman, MD
Professor of Radiology; Gerald D. Dodd, Jr. Distinguished Chair; Chief, Section of Body Imaging; Department of Radiology; M.D. Anderson Cancer Center; Houston

Advances in multislice CT technology have resulted in a 4- to 6-fold increase in imaging speed and have enabled both the acquisition of thinner slices and an increase in z-axis coverage. Together, these improvements are prompting reconsideration of conventional approaches to the use of contrast media. Now, along with setting appropriate parameters for pitch, slice thickness, and the number of passes to make during various phases of contrast enhancement, radiologists tailor the delivery of iodinated contrast by adjusting volume, concentration, rate, and the timing between injection and scanning.

The organ whose imaging is most affected by advances in multislice CT is the liver. Identification of pathology within the liver is related directly to differences in normal hepatic parenchyma and abnormal tissue. In the vast majority of cases, metastatic disease presents as lower-density, hypovascular lesions against a bright background, made so by substantially greater contrast enhancement. ¹ The goal in hepatic imaging is to increase the difference in enhancement between the two tissues, so that the conspicuity of lesions is maximized. The best way to do this is by giving adequate amounts of contrast material, as gauged by the total grams of iodine. The rate of contrast administration determines the timing of peak contrast enhancement.

The enhancement curve of the liver is characterized by a fairly steep uprise. It then reaches a plateau, followed by a slower downward curve. It is at the peak that the maximal difference in the enhancement of the liver and hypovascular liver lesions becomes apparent. It is, therefore, imperative that scanning be performed during the maximal phase of contrast enhancement. ²

Single-slice helical scanners require a significant amount of time to scan the liver, and therefore must include some of the upward phase of the contrast enhancement curve, as well as the plateau and the downward phase (Figure 1). Scanning during the early upward phase of the enhancement curve is inefficient, as images are obtained when the concentrations of contrast are still low, even though the patient has received a large contrast load.

Scanning during the downward phase of the enhancement curve is worse. As contrast reaches equilibrium, the aortic enhancement curve and the liver enhancement curve begin to run parallel, and hypovascular liver lesions begin to fill in from the periphery. During this equilibrium phase, liver lesions become much less conspicuous, they appear smaller, and, finally, they disappear. This results in a less accurate representation of the extent of disease and could, for example, give the false impression that lesions were improving with chemotherapy or radiation therapy when, in fact, their change in size was simply a hemodynamic artifact caused by encroachment of contrast into the lesion.

A major advantage of multislice CT over standard helical CT is that it takes significantly less time to scan a target organ. Consequently, less of the upward and downward portions of the enhancement curve need be included. Instead, most of the scanning closely surrounds the area of peak enhancement of the liver (Figure 2).

Ideally, imaging would capture just the peak of liver enhancement, with none of the up- or downslope of the enhancement curve. To do that, image acquisition in the entire volume of the liver would need to be completed within only seconds. Although we have not yet achieved that goal, the development of 4-, 8-, and even 16-slice CT scanners has brought us much closer.

Computer Automation

As 8- and 16-slice CT scanners become available and, in the future, volumetric CT is introduced, it will become even more critical to time scanning precisely to correspond to the peak contrast enhancement of the liver. We can do this by using either a test bolus of contrast material or, even better, computer automated scanning technology (CAST). SmartPrep (GE Medical Systems, Milwaukee, WI) is one example of this technology.

Using a test bolus has several disadvantages. It requires giving additional contrast material, at additional expense. In addition, it is done separately from routine scanning and, therefore, adds to the length of the examination.

By comparison, CAST enables contrast to be given at an acceptable rate (2.5 to 6 mL/sec) and a series of low-radiation-dose scans taken at short intervals until a specific enhancement threshold over baseline is achieved in the liver. The computerized software is now sophisticated enough to automatically track with cursor placement in aortic and liver enhancement curves.

Our research has shown that an enhancement threshold of 50 HU results in excellent lesion detection in the liver. ^2-5 When the contrast reaches this threshold, the technologist simply performs the diagnostic scan. The time it takes to switch from enhancement detection scanning to the actual diagnostic study is acceptable for parenchymal organs, such as the liver, though it may be slightly longer than is ideal in some applications, such as CT angiography.

The development of CAST was prompted by the observation that contrast circulation times vary from person to person. The average time from injection to portal venous phase scanning of the liver is 65 to 70 seconds. While this time frame would likely be appropriate for a very fit person, a patient with cardiovascular disease would benefit from CAST, which automatically adapts to the patient's decreased cardiac output and delays the liver scan until enhancement is optimal.

Computer automation may also reduce the volume of contrast used. Research in this area has demonstrated that the degree of liver enhancement achieved with a 125-mL bolus of contrast using CAST is equivalent to that achieved with 150 mL of contrast using a conventional time delay between injection and imaging. ^3-5

Another technique used to achieve better contrast enhancement is the "saline-push." By chasing the contrast with saline, a greater degree of contrast enhancement can be achieved. This approach requires additional time, a more experienced technologist, and may compromise contrast sterility if not well managed. In the appropriate clinical environment, however, it can be a useful method to improve contrast enhancement.

Contrast Concentration

For most CT examinations, it is typical to use approximately 150 mL of contrast with a concentration of 300 to 320 mgI/mL. As multislice CT technology advances, the ability to deliver higher concentrations of contrast in smaller volumes will become important. With faster scanners, including those capable of volumetric scans of an entire organ in one acquisition, it is absolutely critical to deliver contrast rapidly. Volume, on the other hand, becomes less important. With increasingly fast scanners, it is no longer necessary to use volume to prolong the examination, as has been done in the past. Higher concentrations (for example, 370 to 400 mgI/mL) and lower volumes likely represent the future of contrast administration as scanners become faster and faster.

Some caution is warranted, however. Rates of contrast administration have increased significantly in the past decade. Two surveys conducted by the Society of Computed Body Tomography/Magnetic Resonance, in 1995 and 1998, ⁶ indicated contrast delivery rates for routine body examinations have increased from a range of 1 to 1.5 mL/sec to 2.5 to 3 mL/sec, and in the case of evaluation for hypervascular lesions, to 5 to 6 mL/sec. To remain comfortable in delivering contrast at such high rates, we must be sure that there is an extremely low incidence of contrast extravasation.

In our practice, we are committed to universal use of low-osmolality contrast media, and have been able to achieve exceptional efficiency with an extremely low incidence of adverse reactions. Careful attention to safety will be even more critical as we venture into intravenous use of higher-concentration iodinated contrast material in CT.

Another potential problem may be the use of relatively high concentrations of iodinated contrast when scanning such areas as the neck and upper thorax, where streak artifacts emanating from undiluted contrast material can be detrimental. Despite this, higher concentrations of contrast, assuming they can remain safely within the low-osmolar range, will likely become widely used.

Additional Efficiencies

The use of prefilled syringes is now becoming more popular, as scan times are reduced and workplace efficiency critically examined. At the University of Texas M.D. Anderson Cancer Center, we perform nearly 300 CT procedures each day. If the cost of using prefilled syringes is comparable to that of manually drawing up contrast for each procedure, the added advantages of sterility and time savings for nurses and technologists are likely to be deciding factors that propel the use of prefilled syringes.

Another cost-saving approach might be to buy contrast in bulk, perhaps in 1000-mL containers, rather than the 500-mL containers now more generally in use. We could then spike these containers once and, using a series of one-way valves to protect against bacterial and viral contamination, fill syringes for multiple patients. This approach is actually being used now in many hospitals, as the contents of today's 200- to 500-mL contrast bottles are used for multiple examinations in multiple patients.

Conclusion

Current and future multislice CT scanners will generate interest in using higher concentrations of contrast, lower volumes, and techniques to clearly define contrast circulation and peak enhancement, such as computer automated scanning. Prefilled syringes will probably be more widely used as pressure to increase efficiency continues. Purchasing contrast in bulk containers may be another effective means of reducing costs. *