Quantitative imaging predicts pancreatic fatty infiltration on routine CT examination

The pancreas undergoes marked changes throughout life due to physiological and pathological processes. These changes may be macroscopic with modification in the volume or lobulated contours of the pancreas as well as microscopic, affecting tissue composition and architecture [1,2]. Fatty infiltration and fibrotic deposits are considered to be two of the most clinically relevant changes that may affect the pancreas. Fatty infiltration of the pancreas is associated with a higher body mass index (BMI), visceral fat, precancerous pancreatic lesions [3], and an increased risk of post-pancreatectomy pancreatic fistula [4], [5], [6], [7]. Fatty infiltration can be classified as extralobular (adipocytes located between the lobules of the pancreas) and intralobular fatty infiltration (adipocytes located within the lobules) [3,8]. Pancreatic fibrosis is associated with a lower risk of post-pancreatectomy pancreatic fistula [4], [5], [6].

Despite their substantial clinical value, these modifications are rarely assessed because a biopsy of the pancreatic parenchyma is associated with a non negligible risk of morbidity [9,10]. A few studies have suggested that imaging could be a non-invasive option to assess the histological changes in the pancreas, using the qualitative visual analysis of ultrasound, computed tomography (CT), or magnetic resonance imaging (MRI) [11,12].

Other studies have used a variety of quantitative MRI methods to evaluate the fatty pancreas. In this regard, Fukui et al. reported a significant association between histological pancreatic fatty infiltration and proton density fat fraction (PDFF) [13]. Schawkat et al. reported a significantly greater PDFF in patients with advanced fat deposition than in patients without [14]. Pancreatic fibrosis has been effectively evaluated with dedicated MRI techniques [15], [16], [17].

Despite promising results, these techniques are not routinely used because they have not been externally validated, they often require specific acquisitions or substantial post-processing. Only a few studies have evaluated CT, mainly showing lower pancreatic attenuation (PA) on unenhanced images [18], [19], [20], [21], larger areas of visceral fat [3], and more recently a greater fat volume fraction on dual-energy CT [22] to assess pancreatic fatty infiltration. A recent study by Sartoris et al. assessed the feasibility and reproducibility of pancreatic surface lobularity (PSL) quantification on CT images in a population of adult patients free of pancreatic disease [23]. The goal of PSL quantification is to provide a quantitative index of the degree of contour lobularity of the pancreas. This study showed that PSL correlated with the BMI and age, suggesting that PSL could reflect histological changes in the pancreas, mainly fatty infiltration [3,24,25]. We hypothesized that the amount of fat could influence the macroscopic appearance of the surface of the pancreas, increasing lobularity.

The purpose of this study was to evaluate the performance of quantitative CT-based analysis, including PSL quantification, to predict the presence of fatty infiltration of the pancreas in a series of patients who underwent surgery for a pancreatic tumor, using pathology as the standard of reference.

留言 (0)

沒有登入
gif