Lung Nodule Enhancement at CT
Abstract & Commentary
Synopsis: The absence of enhancement of a solitary pulmonary nodule greater than 15 HU on CT is predictive of benignity.
Source: Swensen SJ, et al. Lung nodule enhancement at CT: Multicenter study. Radiology 2000;214:73-80.
The characterization of the solitary pulmonary nodule (SPN) is one of the most common diagnostic problems encountered by radiologists. As approximately 50% of surgically resected SPNs are found to be benign, a noninvasive means of accurately distinguishing benign from malignant lesions would have a significant effect on the morbidity and cost of surgical resection for benign lesions. It has been shown that malignant lesions are relatively hypervascular as compared to benign lesions and, therefore, enhance to a greater degree on angiography, contrast-enhanced conventional tomography, positron-emission tomography using radioactive glucose analog fluorodeoxyglucose (FDG), gadolinium-enhanced magnetic resonance imaging, and contrast-enhanced computed tomography (CT). In this prospective multicenter study, Swensen and colleagues build on their previously published experience with the technique of CT nodule enhancement to distinguish benign from malignant SPNs.
The study enrolled 550 patients at six medical centers over a 22-month period, of which 356 patients had a definitive clinical or pathologic diagnosis and a technically adequate lung nodule enhancement study. The nodules had to meet the following entry criteria: between 5- and 40-mm diameter (calculated as the mean of the long and short axis diameters as measured on lung windows); relatively spherical, homogeneous attenuation on precontrast examination; absence of CT artifacts due to motion or beam hardening from adjacent bone; and absence of fat, calcium, or cavitation on thin-section CT scans. Patients allergic to iodinated contrast, creatinine greater than 1.5 mg/dL, and those unable to consistently breath-hold were likewise excluded. The technique of CT nodule enhancement consisted of preliminary unenhanced helical 3-mm collimated scans obtained in a single breath-hold using a pitch of 1:1 through the entire nodule, followed by serial 5-second, 3-mm collimated helical acquisitions performed at one, two, three, and four minutes after intravenous administration of iodinated contrast (420 mg/kg of 300 mg/mL) injected at a rate of 2 mL/s. The mean attenuation values of the SPN at baseline and at one, two, three, and four minutes post-injection were manually calculated by choosing the scan closest to the nodule equator and drawing a round or oval region of interest that approximated the nodule’s shape and encompassed approximately 70% of the nodule’s short- and long-axis diameter. The value of nodule enhancement was defined as the difference between the maximum mean attenuation value measurement in the four minutes after contrast administration and the pre-enhancement measurement as expressed in Hounsfield units (HU).
The findings showed that there was a significantly greater median enhancement of the malignant neoplasms than for benign neoplasms or granulomas (38 vs 10 HU; P < 0.01). More important, by setting an enhancement value greater than 15 HU as a marker for malignancy, the sensitivity and specificity of contrast-enhanced CT were 98% and 58%, respectively. Of 171 malignant neoplasms, only four failed to enhance more than 15 HU. Swensen et al conclude that the absence of significant nodule enhancement (£ 15 HU) was a strong predictor of benignity.
They point out that since all four false-negative malignant SPNs enhanced 14-15 HU, retrospectively, lowering the threshold of a positive study to 10 HU would increase the sensitivity of the technique to 100%.
Comment by Jeffrey S. Klein, MD
This study confirms the previous work of Swensen et al and several Japanese investigators by demonstrating a high sensitivity of CT nodule enhancement in the detection of malignant SPNs. This technique is much more practical than other methods of assessing nodular vascularity such as MRI and PET, and, as Swensen et al point out, CT nodule enhancement can be performed immediately following routine staging CT of chest and abdomen for suspected lung cancer without the need for additional contrast adminstration and with little additional scan time (< 5 minutes) and radiation dose.
The problems with this study relate to the large number of cases excluded due to insufficient follow-up or technically inadequate CT studies (35% in this series), and the lack of pathologic diagnosis of nonenhancing probably benign nodules, as these tend to be followed rather than biopsied or resected.
Caution should be exercised in using this technique to evaluate larger (> 2 cm) nodules or those with necrosis, as these have a greater tendency to be false negative on CT nodule enhancement studies.
The relative role of CT nodule enhancement in the evaluation of SPNs remains to be defined. I believe that the technique will prove most useful in situations where pathologic diagnosis via transthoracic needle biopsy is either unavailable, cannot be performed, or proves nondiagnostic.