Low-Dose CT Protocol Performs Well in Lung Nodule Assessment

Caroline Helwick

September 07, 2011

September 7, 2011 — For monitoring lung nodules with computed tomography (CT), radiation dose can be reduced to approximately 3% of the current standard dose without compromising efficacy, according to a recent study by investigators from Stanford University in California and the University of Bern in Switzerland.

The study, which appears in the September issue of the American Journal of Roentgenology, found that tube current-time settings could be set as low as 10 milliampere-seconds (mAs), compared with the standard 100 to 300 mAs, without affecting diagnostic image quality. The lower tube current-time setting translates into a substantial decrease in radiation dose.

"The risk of radiation-induced cancer death for one standard chest CT is estimated at roughly one in 3000, and could be reduced hypothetically to one in 120,000 with this low-dose protocol," said lead author Andreas Christe, MD, from Stanford University and the University of Bern in a press release.

On August 24, 2011, the US Joint Commission issued a Sentinel Event Alert on the radiation risks of diagnostic imaging, noting that during the past 2 decades, the US population's total exposure to ionizing radiation has nearly doubled. Although experts disagree on the extent of the risks for cancer from diagnostic imaging, there is agreement that steps should be taken to eliminate avoidable exposure to radiation.

Study Evaluated Multiple Dose Levels

The study involved CT images from 50 patients referred for cancer staging, lung nodule screening, or follow-up. A total of 125 cuboidal regions of interest showing a single nodule (≤8 mm) and 27 normal cuboids were selected. Image sets were reconstructed with software from raw data simulating different dose levels: 300 (original dose), 220, 180, 140, 100, 80, 60, 50, 40, 30, 20, 10, and 5 reference mAs. Three blinded radiologists reviewed the reconstructed images.

"The tube current-times level of 10 mAs did not exhibit a significant difference in individual reader sensitivity compared with the standard 300 mAs," said Dr. Christe.

At the lowest tube current level of 5 mAs, only 1 reader had a significant decrease in sensitivity, from 82% to 77% (P = .0035). Radiologists identified an average of 88% of the lung nodules at 10 mAs compared with 91% at 300 mAs. The specificity at the standard dose level was 92% and was no lower at lower dose levels.

The authors write: "The readers had one thing in common: sensitivity for all 125 nodules was not significantly influenced by dose level."

Dr. Christe noted that in his clinical practice, he and his colleagues will monitor lung nodes at 40 mAs until they learn the results of a follow-up clinical study. "We are using the slightly higher dose because the 10 mAs resulted from a study design with dose simulations and smaller field of view than is routine in clinical practice," he said.

Human vs Machine

The study also reviewed the use of a computer to measure nodule size. Size is important because volume-doubling time is a predictor of malignancy, and there can be considerable interreader and intrareader variability in assessments.

The study found that the use of a computer to measure nodule size decreased the variability that is observed with manual measurements. The mean nodule volume measurement error between 5 and 300 mAs was 2.2% with the computer, which is much lower than the interobserver volume measurement error rate of 38%.

At each dose level, the readers measured a significantly larger volume than was achieved with the computer-aided lesion measurement solution. The readers probably tended to include large parts of the small ground-glass transition around the nodule in the measurement of the diameter, the authors suggest.

"[O]ur study showed that volume-doubling time estimation would be more accurate if an automated measurement tool were used for both measurements than if individual manual measurements were made by two readers," the report states.

Agreement on the Need to Lower Radiation Dose

Other investigators have evaluated outcomes with lower doses of radiation, but the current study takes this approach a step farther, Kavita Garg, MD, professor of radiology at the University of Colorado, Denver explained to Medscape Medical News. "What is unique is that they used a wide range of dosages, as low as 5 mAs. The other unique feature is that they addressed small nodules," she said.

Dr. Garg said that as an investigator in a number of lung cancer screening trials, she too has evaluated screening with low-dose CT protocols. "I have always felt we can use much lower dosages than we routinely use, especially in the follow-up of small nodules. This article validates that," she said.

"I also agree that computerized methods would be better for screening. They would be more consistent and reproducible, with less inter- and intrareading variability that you get with 2 readers looking at 2 scans at different time points, which relies a lot on the reader's experience."

The study was supported by the Swiss National Science Foundation, the Swiss Gottfried und Julia Bangerter-Rhyner Foundation, and the Swiss Huggenberg-Bischoff Foundation. Dr. Christe and Dr. Garg have disclosed no relevant financial relationships.

Am J Roentgenol. 2011;197:1-8. Abstract