In this study, we proposed a predictive model of adenomatous gallbladder polyps and found that low viscosity less than 7.5 sec-1 with an OR of 22.539, low bile cholesterol less than 414.5 mg/dl with an OR of 10.004, and age older than 55 years with an OR of 23.550 were predisposing factors for adenomatous polyps. The predictive scoring model using these factors showed high adherence with a sensitivity of 90.9% and specificity of 80.2% and an area under the curve of 0.845. Our results seemed superior to the previously reported quantitative contrast-enhanced ultrasound sensitivity and specificity parameters of 76.5% and 75%, respectively, reported by Bae et al. [10], or 68.1% and 70.2%, respectively, reported for differentiation based on polyp size over 1 cm in a study by Wennmacker et al. [11].
Bile acid is synthesized in the liver and its conjugation and transportation occur through the enterohepatic circulation. After that, the transformed bile acid is secreted into the bile and stored in the gallbladder. If there is abnormal bile secretion or functional impairment of the gallbladder, the bile contents stored in the gallbladder will change accordingly [12]. Several studies [13, 14] have reported that dyslipidemia, such as high low-density lipoprotein, would decrease gallbladder sensitivity to cholecystokinin and inhibit cholecystokinin secretion, resulting in a reduction of gallbladder motility and alteration of the mucosal esterification of free sterols from bile. This physiological change in the hepatic cholesterol mechanism could then promote cholesterol polyp formation. Zhao et al. [6] reported that the accumulation of cholesterol esters in epithelial cells of the submucosa of the gallbladder is thought to be a key factor in this association. The accumulated substances in the gallbladder submucosa are phagocytized by macrophages and generate foam cells. Then, finally, the surface of the gallbladder mucosa is converted into a polyp-like appearance with swollen villi. Our results showed that the level of bile cholesterol was a very useful marker for the differentiation of polyps.
Interestingly, in this study, a significant low viscosity was seen in the adenomas. Bile viscosity can be useful for examining the characteristics of bile, but there have been few reports on human bile viscosity. During storage in the gallbladder, mucus, a highly viscous substance, is secreted from goblet cells in the epithelium of the gallbladder. Mucus contains mucins, a family of large glycosylated proteins, and bile viscosity is drastically increased by mucus secretion. Previously, Yoo et al. [4] reported a relationship between cholesterol-related gallbladder disease and expression of the gallbladder mucin (MUC) gene in gallbladder epithelial cells. The expression of MUC3 and MUC5B genes was higher in cholesterol polyps than normal tissue of gallbladder and the upregulated expression of MUC genes contributed to the hypersecretion of mucin and the increased mucin in bile consecutively, led to increases in bile viscosity. Although we did not measure mucin levels, we believe that the high elevation in bile viscosity due to this characteristic mucin secretion process in cholesterol polyps can be useful in distinguishing it effectively from adenomatous polyps.
Additionally, increased age was also a significant predisposing factor for adenomatous polyps in the current study. Aging is closely related to changes in lipid profiles. Cha et al. [15] reported that ages of 65 years or older were correlated with neoplastic or malignant GBPs and our results showed that 56 years or older was a potent parameter in the predictive scoring model. The cutoff value of 55-year-old is noteworthy, which is relatively younger than the cutoff value commonly mentioned in previous studies. Given that our study consisted of a small sample size, the results need to be confirmed by a study with a larger number of samples.
This novel scoring system for distinguishing adenomas and cholesterol polyps could have clinical significance. Adenomas have malignant potential that generally follows the adenoma-carcinoma sequence, similar to that of colorectal cancer [16]. Gallbladder malignancy has been regarded as aggressive cancer with a dismal prognosis of 5 – 12% 5-year survival, but its survival could be prolonged to 80% if it could be detected with in situ disease [17]. These characteristic differences in the survival rate of gallbladder cancer patients according to the time of diagnosis emphasizes the importance of early detection of adenomatous polyps to prevent the spread the tumor cells or invade adjacent organs. In contrast, cholesterol polyps, which comprise most of the GBPs, have benign characteristics until they become symptomatic. Thus, the accurate discrimination of cholesterol polyps from adenomas can help optimize the indications for cholecystectomy for GBP and avoid unnecessary surgery, thus reducing the burden of medical costs and avert risks related to surgery, such as bile duct injury or bleeding.
To date, the most widely accepted indication for surgery for GBP is the diameter of polypoid lesions exceeding 10 mm based on ultrasonographic findings [5, 18]. Although we agree that these criteria are comparatively valid, it seems unreasonable to simply determine the malignant potential of GBP using this criterion only and use it to decide whether to perform surgery. Sonograms may vary in accuracy depending on the sonographer’s proficiency and there might be a discrepancy between the actual polyp size and the size measured by sonographic imaging. In addition, there are other reports that the 1 cm surgical threshold for GBP has insufficient diagnostic accuracy [11]. To overcome these limitations, several diagnostic modalities have been reported, but computed tomography or magnetic resonance imaging have also been reported to be unreliable in differentiating benign from premalignant lesions [19]. Therefore, our proposed discrimination method that reflects age and the biochemical characteristics of bile can reflect the nature of polyps more than size-based discrimination methods that merely reflect morphological aspects. It will help facilitate precise therapeutic decision-making and establish precise guidelines for the management of GBP.
Despite interesting outcomes, this study had some limitations that require attention to interpret. As described in the methodology section, only patients with GBPs with a size of 1 cm or more who underwent cholecystectomies were enrolled and those with GBPs less than 1 cm who underwent regular follow-up surveillance without surgery were not included. Therefore, this study did not represent all GBP patients and limits the application of the prediction model to the whole population of patients with GBPs. However, there is no general disagreement regarding surgery for GBPs of 1 cm or more and it is still the most common indication for surgical intervention. Our results, therefore, can provide practical guidance that is useful for those who are considering clinical intervention. Finally, our data had selection bias and some clinicopathologic data were incomplete because it was a retrospectively designed study. However, the selection bias could be alleviated because we prospectively performed sampling and collected data for all patients operated on due to GBP. A prospectively designed study involving a large number of adenomatous and cholesterol polyps is required in the next study. Also, we need to register and analyze all cases of GBP having sonographic evidence, regardless of the size of the GBP, to establish guidelines for the entire GBP population.