Can a CT-based quantification method evaluate chronic liver disease?
Quantifying liver surface nodularity (LSN) from CT scans can accurately and quickly identify clinically significant portal hypertension in cirrhosis patients, according to a Radiology study.
Currently, the reference standard for assessing portal hypertension is the hepatic venous pressure gradient (HVPG). While many noninvasive methods are available to evaluate HVPG, such as transient elastography, ultrasound-based shear wave elastography and MR elastography, those methods are plagued by failure rates or have not been validated for defining clinically significant portal hypertension (CSPH).
Lead author Riccardo Sartoris, with University Hospitals Paris Nord Val de Seine in Beaujon Clichy, France, and colleagues set out to evaluate the diagnostic performance of the quantification of LSN score from CT scans to diagnose CSPH in patients with cirrhosis.
The retrospective study included 189 patients, 102 of whom had CSPH with 78 separate patients in an external validation group. All underwent abdominal CT and HVPG measurement, which was used as reference standard, between 2010 and 2016.
Overall, the liver surface nodularity score accurately classified 84 percent of patients with CSPH, verified using the HVPG measurements.
Additionally, the researchers found “in patients with compensated cirrhosis from two centers considered for surgery, the rule-in and rule-out cutoff values of LSN could correctly classify about six of 10 patients as either having or not having CSPH,” they wrote.
The method had a risk of false-negative or false-positive results below 10 percent, Sartoris et al. noted.
Andrew D. Smith, with the University of Alabama’s department of radiology declared “the era of quantitative liver CT is upon us,” in an accompanying editorial.
Smith noted one of the strengths of the technique demonstrated by Sartoris et al. was its ability to use existing liver CT images—saving patients from additional expenses and ionizing radiation.
Overall, Smith found the semiautomated technique promising, and noted the technology continues to grow.
“Quantitative liver CT image-processing algorithms are on the rise, are widely applicable, and can be applied retrospectively to routine liver CT images,” he concluded.
Matt joined Chicago’s TriMed team in 2018 covering all areas of health imaging after two years reporting on the hospital field. He holds a bachelor’s in English from UIC, and enjoys a good cup of coffee and an interesting documentary.