Adopting low-dose protocols for the two sets of applications created a natural laboratory for Dr. Jeremy Collins and colleagues at Northwestern University to test their impact on patient dose and image quality. Collins, as assistant professor of radiology, described the results in a SIR session on Sunday.

Dose reduction has become a clinical priority, with a sixfold increase since 1980 in the per capita exposure to medical radiation in the U.S., and diagnostic CT accounting for 45% of the increase, Collins said in a news conference. The long-term implications of this growth are unclear, but the associated risk from exposure to ionizing radiation from medical imaging is considered linear and additive, with no minimum threshold for a safe dose, he said.

Additional CT utilization may result from positive findings from the National Lung Screening Trial (NLST), which indicated that CT screening of an at-risk population can reduce mortality by 20%. This could lead to more serial CT screenings and CT-guided biopsies of pulmonary nodules detected by screening, Collins noted.

"Anything we can do to lower dose to patients is appropriate," he said.

Northwestern's interventional radiologists anticipated the trend in August 2010 by adopting two low-dose protocols for CT guidance: one for lung biopsies and another for extrapulmonary biopsies and drainages.

The CT-guidance protocol for lung biopsies recommends a helical planning acquisition at 120 kV and 60 mAs. The guidance acquisition is sequential, at 100 kV and 50 mAs. Collimation is set at 1.5 mm for planning and guidance.

The department previously followed the CT vendor's recommended protocol, which also called for a planning acquisition at 120 kV at 60 mAs settings; however, the guidance protocol was 120 kV at 50 mAs. Again, 1.5-mm collimation was used.

Fifty consecutive cases were evaluated retrospectively before and after the new CT-guidance protocol for lung biopsies was adopted. Collins and colleagues found that low-dose protocols resulted in an average effective dose reduction of 66%, to a dose of 3.5 mSv (1.0-8.0 mSv), compared with 10.3 mSv (1.7-52.4 mSv) for the standard protocol (p < 0.0001).

Procedures performed by an attending interventional radiologist typically exposed patients to slightly less radiation than those performed by trainees (mean dose of 2.8 mSv versus 3.8 mSv). For the standard protocol, attending physicians applied slightly more dose on average (10.6 mSv) than the trainees (10.2 mSv). No significance difference in the incidence of pneumothorax was observed.

Nodule size did not vary notably (2.0 cm versus 2.1 cm diameter) for the two approaches. An average of 15 scans were performed for the low-dose approach, compared with a 17-scan average for the standard approach.

Representative low-dose images were noisier in large patients than views generated with the standard protocol, but the slight degradation was not enough to affect the procedures, Collins said.

Low-dose CT for extrapulmonary biopsies

The low-dose protocol for guiding extrapulmonary biopsies and drainages retained the standard approach's recommendation for a 120-kV helical scan for planning. The new planning recommendation for tube current limited settings to 50 mAs to 80 mAs (compared with 60 mAs to 80 mAs for the standard protocol), depending on tissue density and thickness.

For guidance scans, the low-dose protocol reduced the recommended tube voltage to 100 kV, compared with 120 kV for the standard protocol. For tube current, 50 mAs to 80 mAs was once again used.

Collimation settings of 1.5 mm were recommended for both low-dose and standard protocols. As with lung biopsies, the interventional radiologist was free to modify the settings when necessary. Technologists were urged to use sequential acquisition when possible.

Collins and colleagues retrospectively reviewed consecutive cases six months before and after the protocol was modified to measure its effect. The low-dose protocol was applied to 54 biopsies performed on 52 patients. Twenty-eight biopsies aimed at targets in musculoskeletal anatomy, while there were 16 retroperitoneal biopsies, three pancreatic biopsies, one renal biopsy, and one adrenal biopsy. Four drainages were examined.

Fifty-five procedures on 50 patients were performed with the standard protocol. The procedural mix did not differ substantially from that of the low-dose protocol, though only one drainage was performed.

Overall, the low-dose protocol reduced the average effective dose by 46%, decreasing from 10.2 mSv for the standard protocol to 4.5 mSv for the low-dose protocol (p = 0.03) among experienced interventional radiologists. For trainees, the average effective dose fell from 11.1 mSv for the standard protocol to 6.7 mSv for the low-dose approach, but the results weren't statistically significant (p = 0.75).

Both studies were limited by the retrospective view, an institutional preference for ultrasound-guided abdominal biopsies and drainages, and minor protocol changes implemented independently by the CT technologists. Collimator thickness was reduced to 0.75 mm, for example, to increase the reconstruction speed of helical acquisitions

"The modification to the scanner's technique is simple and can be applied to any existing CT scanner system," Collins said, "but interventional radiologists will still need to evaluate each person on a case-by-case basis, especially smaller people or those who have anatomy that is more difficult to image."