There are two main causes of neonatal cholestasis jaundice including BA and non-BA cholestasis diseases. Non-BA cholestasis diseases include infection, metabolic disorders, genetic abnormality, extrahepatic obstruction from common duct gallstones or choledochal cyst, panhypopituitarism, Alagille syndrome, parenteral nutrition associated liver disease and other rare disorders. It is difficult to distinguish BA from non-BA cholestasis diseases. BA is the most common cause of cholestatic jaundice in infants. Early evaluation and timely treatment play an important role in improving the prognosis of BA. Early diagnosis is the key to the treatment of BA. However, the early clinical symptoms of BA are atypical, and misdiagnosis of BA can lead to unnecessary exploratory surgery. At present, there are no non-invasive methods to diagnose BA definitely, and how to diagnose BA early is still a difficult problem for doctors[15].In recent years, many BA diagnosis and prediction models have been applied, but these models have many indexes, complicated methods and high cost, so they were difficult to be applied in grass-root level[16, 17]. When BA was suspected, ultrasound was undoubtedly the first choice imaging method. As a convenient, simple and economical examination method[18], radiologists in almost all medical institutions will measure the length and width of gallbladder.
It has been reported that when gallbladder LTWR was 5.2 as the cut-off value, it had a high diagnostic value to distinguish BA from other cholestasis diseases, and the AUC could reach to 0.844[14]. Therefore, gallbladder LTWR could be used as another reliable index for diagnosing BA. It was not only easy to identify and measure its size, but also has strong applicability, objective and reliable results. Because other ultrasonic signs (such as triangular cord sign, abnormal gallbladder) have strong subjectivity[19]. Panjai Choochuen et al. found that when gallbladder LTWR > 4.1, its sensitivity was 71.7% and specificity was 58.3%[13]. In this study, it was found that the sensitivity and specificity of diagnosing BA were 78.9% and 66.7% when gallbladder LTWR > 3.47 regardless of age, which was comparable to that reported by Panjai Choochuen et al[13]. However, the cut-off value of gallbladder LTWR in this study was significantly lower than 4.1. A review of relevant data found that the gallbladder width of the BA group in our study was longer than theirs. Analysis of the most possible reason for this situation was previous study[13] measurement method of the maximum width of gallbladder different from ours.Our study measured the maximum width from external wall to external wall,while previous study measured the maximum width from internal wall to internal wall. As the same time, we also found that the gallbladders wall of most BA infants were thickness, which was also one of the reasons for the small gallbladder LTWR.
In our study, the patients were divided into five age groups with 30 days as the boundary, and the gallbladder LTWR in different age groups were compared between the two groups There were only 5 cases of BA patients aged ≤30 days, and 11 cases of non-BA group. Although the sample size was small, it also reflected certain diagnostic value. The gallbladder LTWR of patients aged ≤30 days had a relatively high sensitivity, up to 80%. Why were fewer patients aged ≤30 days with BA? It might be due to early jaundice in children, parents did not pay attention to, mistaken for physiological jaundice,or in the internal medicine treatment time was too long, into the surgery of older age and other circumstances. When patients’ jaundice were less than 30 days, we should be alert to the possibility of BA if the gallbladder LTWR > 3.16 as measured by ultrasound.
In this study, the sensitivity, specificity, positive predictive value, negative predictive value and AUC of different age groups and undifferentiated age groups were significantly different. The gallbladder LTWR had different AUC values in different ages. Group III (61-90 days) had the highest AUC of 0.831, while group V (≥121 days) had the lowest AUC of 0.548. Among the patients aged 31 to 60 days(gallbladder LTWR>3.37), the sensitivity was 68.4%, the specificity was 85%, the positive predictive value was 81.2%, the negative predictive value was 73.9%, and the accuracy was 76.9%. It was also clear that the diagnostic accuracy was higher, second only to group III (61-90 days). Some medical centers have applied the gallbladder LTWR to the diagnosis of BA[13, 14]. In our study, the gallbladder LTWR was helpful to the diagnosis of BA aged 61-90 days. Once found the gallbladder LTWR > 3.6, it was most likely BA and can be used as a signal of surgical exploration. Our study found that gallbladder LTWR of children in group IV (91-120 days) and group V (≥121 days) had low diagnostic accuracy and sensitivity, but high specificity with 100%, which could be used as a criterion to exclude BA. Meanwhile, it was also found that there were fewer non-BA patients in these two groups, which might be due to the fact that most parents of children were admitted to the hospital in time when jaundice was found. More non-BA patients in the group IV (91-120 days) and group V (≥121 days) should be included in the later stage and the diagnostic value of LTWR should be further discussed. Based on our results, it was reliable to diagnose BA in combination with age, which not only improved the accuracy of diagnosis, but also avoided the missed diagnosis or misdiagnosis of younger children, and provided great help for later treatment.
Additionally, other ultrasound parameters have been described as useful indicators to diagnosis BA. Takamizawa et al. [20] reported that the gallbladder length combined with triangular cord sign and gallbladder contractility was helpful to diagnosis BA. El Guindi et al. [21] found that some parameters such as abnormal gallbladder, gallbladder length, positive triangular cord sign, hepatic artery diameter and hepatic subcapsular flow were predictors of BA significantly.As the same time, the parameters of the gallbladder with the relatively high sensitivity and specificity in the diagnosis of BA were abnormal gallbladder (93.3% and53.3%) and gallbladder length (76.7% and 80%), respectively[21]. In a meta-analysis, Zhou et al.[22]reported that abnormal gallbladder was one of the most widely accepted ultrasound criteria to distinguish BA from other hepatic cholestasis. Lee et al. [23] reported that hepatic subcapsular flow had sensitivity of 100% and specificity of 86% in the diagnosis of BA. It can be seen that ultrasound is widely used in the diagnosis of BA. The diagnosis of BA by gallbladder LTWR is a comparison of single parameters in our study,which is easy to operate, low technical access, and more conducive to rapid promotion. In the next step, the LTWR of gallbladder and other parameters such as triangular cord sign, hepatic artery diameter, hepatic artery resistance index, and hepatic subcapsular flow, etc. will be combined to diagnose BA, so as to improve the accuracy of non-invasive examination in diagnosing BA.
Ultrasound enabled doctors at all levels to evaluate the possibility of BA quickly and accurately, improve the diagnostic accuracy, reduce the probability of surgical exploration for children with non-BA, and avoid unnecessary trauma and radiation injury as much as possible[10]. The main limitation of this study is that the number of cases in group I (≤30 days), IV (91-120 days) and V (≥121 days) are somehow relatively limited. Besides, this is a retrospective single-center study and the data of other ultrasound parameters are incomplete. In the future, more patients will need to be included in multi-center, prospective studies, to further explore the diagnostic value of gallbladder length combined with other parameters for BA.