In this study, we assessed the effect of steroid therapy on FDG uptake of the liver and DA by comparing PET images taken before and during steroid therapy. We assessed both the manual and automated methods of liver uptake analysis because several methods to set the VOI in the liver were reported. Wahl et al. proposed a manual method to set VOI on the normal inferior right lobe (RL) which is a widely used method in accordance with the Positron Emission Tomography Response Criteria in Solid Tumors (PERCIST) criteria [7]. Hirata et al. proposed an automated method to set the VOI in the liver, which was not always in the inferior RL, with very high inter-operator reproducibility [9]. Similar results were obtained by both methods.
The SUVs obtained from the liver during steroid therapy were significantly higher than those of the pre-therapy scan, whereas the SUVs from DA were not altered by steroid therapy.
CS is increasingly recognized as a cause of heart failure and arrhythmias. FDG PET is a promising tool to assess the activity of CS. The volume-based assessment of FDG uptake is a more precise predictor of cardiac events than SUVmax [1, 8]. An appropriate SUV threshold is important for the identification of the precise CMV. This study is the first to assess the effects of steroid therapy on the FDG thresholds used to estimate the metabolic volume.
CMV is defined as the volume within a given boundary determined using the FDG uptake threshold, such as the liver uptake and the blood pool [8, 10]. However, as our study suggested, the liver uptake in CS patients was significantly increased from the time point before to that during steroid therapy. Glucocorticoids promote changes in body composition that correlate with insulin resistance, hyperinsulinemia, and eventual onset of hyperglycemia [15,16,17]. Steroid-induced diabetes mellitus (SIDM) has been recognized as a complication of steroid use. The effect of glucocorticoids on glucose metabolism is likely the result of impairment of multiple pathways including sensitivity to glucose and the ability to release insulin due to the beta cell dysfunction and insulin resistance in other tissues [17]. One of the etiologies of SIDM is based on the effect of glyceroneogenesis in the liver and adipose tissue [18]. In the adipose tissue, glyceroneogenesis controls the rate of free fatty acid (FFA) release in the blood. On the other hand, glyceroneogenesis is responsible for the synthesis of triacylglycerol (TAG) from FFA and glycerol 3-phosphate (G-3-P) in the liver [18]. The regulation of this process in both the liver and adipose tissue occurs via the enzyme phosphoenolpyruvate carboxykinase (PEPCK). In patients using glucocorticosteroids, PEPCK gene expression in adipose tissue is suppressed, inhibiting glyceroneogenesis [19].
Because FDG uptake increases in the presence of inflammation, the increase of liver uptake may be due to an inflammatory process caused by steroid therapy [20, 21]. In the liver, PEPCK upregulates the synthesis of TAG from FFA and G-3-P, and liver fat increases. Although the HUs on CT did not show a significant difference, the accumulation of triglycerides in hepatocytes was assumed to have increased. While most fatty liver diffusely involves the whole liver, focal or multi-focal fat deposition in the liver is occasionally encountered and causes a diagnostic challenge. Some regions of the liver are well known as common sites of focal fat deposition [22, 23]. A 30-mm-diameter spherical VOI cannot assess the whole liver fat content and the CT HUs can be affected by focal fat deposition. [22].
Insulin resistance might be another potential reason for the increased liver uptake during steroid administration. In the liver, PEPCK stimulates glycerol production and FFA concentrations increase in the blood. In the end, the amount of FFAs released into the blood increases and the increased FFA level interferes with glucose utilization and results in insulin resistance, especially in skeletal muscle [24]. Though the specific effect of insulin on hepatic glucose uptake remains unclear, insulin stimulates glucose uptake in the liver of both insulin-sensitive and insulin-resistant subjects [25]. Iozzo et al., employing the euglycemic hyperinsulinemic clamp, confirmed that insulin increases hepatic phosphorylation of FDG to FDG-6-phosphatase (FDG-6-P) [25]. There is a strong evidence to suggest that hepatic steatosis and insulin resistances are driven by obesity-induced adipokines, and the association between insulin resistance and hepatic steatosis has been established [26, 27].
Although serum ALT levels are often used as a surrogate marker for liver inflammation, ALT is typically elevated in only 50% of non-alcoholic fatty liver disease (NAFLD) cases [20, 28]. In our patients, neither AST nor γ-GTP levels showed a statistically significant difference between time points before and during steroid therapy. Our results thus suggested that ALT alone was a poor marker for the presence of hepatic steatosis. Previous studies reported that patients with advanced fibrosis had significantly lower ALT levels than those with no/mild fibrosis and that ALT had no role in identifying patients with advanced disease [29, 30].
Our results showed that oral steroid therapy and individual patient factors showed significant effect on the blood pool and liver uptakes and UFH administration before FDG injection did not affect the liver or blood pool SUV [14]. UFH increases plasma FFA levels due to activation of lipoprotein and hepatic lipases [4]. As mentioned above, increased FFA interferes with glucose utilization and results in insulin resistance [24]. However, preadministration of UFH is just before FDG injection and the assessment of effect on liver uptake is difficult.
Oral steroid therapy showed no effect on the SUV obtained from the blood pool. Our group previously reported that individual FDG uptake thresholds from the DA were preferable to those from the liver, due to the high inter-operator reliability and non-dependence on dietary relations [13]. Our present findings show further evidence that the threshold for the evaluation of a therapy response should be determined by DA rather than liver values.
Limitations
This study had some methodological limitations. First, the study was retrospective from a single center and the sample size was relatively small, partly because the study focus was patients with active cardiac sarcoidosis. Second, while all patients in our study underwent PET imaging, other tests, such as Holter monitoring and cardiac magnetic resonance imaging (MRI), were not routinely performed. Third, four patients with T2DM, who might have higher liver fat content compared with healthy subjects, were included in this study [26]. Fourth, we measured CT HUs using non-diagnostic low-dose CT as a tool of attenuation correction in PET/CT imaging. Compared to unenhanced diagnostic CT, a low-dose CT scan can be reliably used to exclude NAFLD, if neither liver attenuation of < 40 HU nor a liver-to-spleen ratio < 1.1 is present [31]. However, chemical shift images on MRI are desirable for the diagnosis of diffuse hepatic steatosis because they can demonstrate suppression of the signal from mixtures of microscopic lipids and water at the cellular level [20, 22]. Finally, we assessed the liver and the blood pool FDG uptake in patients using medium-dose steroids (29.7 ± 1.6 mg/day of prednisolone) but did not investigate whether the dose of the steroid affected the results. Oral steroid treatment for CS is generally initiated with 30 mg and is tapered down to 15–25 mg in most patients by the 3-month follow-up visit [1]. Further assessment of patients with tapered-down steroid use is warranted.