Semiquantitative parameters on hepatobiliary scintigraphy to differentiate biliary atresia from neonatal hepatitis in patients with no identifiable biliary to bowel transit

Abstract

Introduction

The current gold standard for imaging to rule out extra-hepatic biliary atresia (EHBA) is hepatobiliary scintigraphy (HBS), which involves visually assessing the tracer appearance in the bowel. However, in cases where hepatic function is impaired, biliary-to-bowel transit may not be observable, even in neonatal hepatitis (NH) cases. This study aims to assess the utility of semi-quantitative parameters on HBS to distinguish biliary atresia (BA) from NH when there is a lack of biliary-to-bowel transit for up to 24 h.

Materials and methods

The study involved retrospective analysis of patients with the diagnosis of neonatal cholestasis where HBS failed to differentiate BA and NH. Histopathological examination was taken as the gold standard. Semiquantitative parameters calculated include: Liver: blood pool (LBR) and liver: kidney ratios (LKR) at 5 min, 30 min, 1 h and 24 h. Mean values for the two groups were calculated. Student’s t-test was employed to assess the statistical significance of difference of mean between the two groups. Receiver operating characteristic (ROC) curve was also drawn to determine a cut-off of these ratios to differentiate between the two groups using SPSS v26.0. P-value < 0.05 was considered statistically significant.

Results

The study included 53 patients (37 males) with a median age of 3 months (range: 24 days to 10 months). Of these, 32 patients had a histopathological diagnosis of BA, while 21 had NH. Mean LBR and LKR at 24 h were statistically different in the two groups (p-value < 0.05). Receiver operating characteristic (ROC) curve analyses showed highest AUC for LBR at 24 h 0.683 (CI:0.532–0.834, p-value 0.017) and LKR at 24 h − 0.669 (CI: 0.511–0.827, p-value:0.036). For diagnosis of BA a cut-off value of ≤ 4.18 for LBR at 24 h (sensitivity and specificity of 62.5% and 61.9% respectively) and ≤ 4.64 for LKR at 24 h (sensitivity and specificity of 68.8% and 66.7% respectively) were found to be pertinent.

Conclusion

HBS serves as non-invasive imaging of choice to rule out EHBA. Semi-quantitative indices LBR and LKR at the 24-hour can differentiate between EHBA and NH even in cases with compromised hepatic function where traditional visual interpretation of tracer transit proves inadequate.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13139-025-00909-1.

Introduction

Neonatal cholestasis lasting more than 2 weeks affects one child in 2500 live births [1]. The two most prevalent causes of neonatal cholestasis are biliary atresia (BA) and neonatal hepatitis (NH). These account for approximately 70–80% of all cases of neonatal cholestasis [2]. Though both conditions have a similar clinical presentation, these have distinct etiologies, management, and prognoses. BA is a surgical emergency, with Kasai portoenterostomy as the treatment of choice, ideally performed within the first 60 days of life. If the surgery is not performed within 90 days, it shows a grim prognosis. The disease progresses to liver cirrhosis and hepatic failure. While conservative treatment is preferred for the management of NH [3]. Thus, underscoring the importance of accurately differentiating BA from medical causes, mainly NH.

There is no preoperative diagnostic algorithm that has proven 100% diagnostic accuracy for BA. The current imaging of choice to rule out extra-hepatic biliary atresia (EHBA) is hepatobiliary scintigraphy (HBS) using 99mTc-mebrofenin. In a normal HBS study, the tracer gets rapidly cleared from the blood pool with uniform hepatic uptake within 5–10 min. The tracer is visualized in the bowel by 30–40 min [2]. Appearance of tracer in the bowel rules out the diagnosis of BA, while no biliary-to-bowel transit favors the diagnosis of BA [4]. However, in cases where hepatic function is impaired, biliary-to-bowel transit may not be observable, even in NH cases [1]. Thus making an accurate pre-operative diagnosis and further management difficult. The authors conducted this study to identify the diagnostic semi-quantitative parameters on HBS to distinguish between BA and NH. The study also evaluated the diagnostic efficacy of the semi-quantitative parameters on HBS to distinguish BA and NH when there is a lack of biliary-to-bowel transit for up to 24 h.

Materials and methods

This was a retrospective study, performed after approval from Institutional Ethics Committee. All the patients referred to the Department of Nuclear Medicine with Neonatal Cholestasis for HBS were evaluated.

Data collection

Clinical details of the patients including birth history, serial biochemical parameters, family history and histopathology results were obtained from the electronic hospital record system. Furthermore, raw imaging data was obtained from Picture Archiving and Communication System (PACS) and evaluated.

Patients adequately prepared for the study with no observable biliary-to-bowel transit were included in the study. Histopathology reports were taken as the gold standard for diagnosis. When histopathology reports were not available, improvement on follow-up with conservative management was considered as NH.

Patients who were either not adequately prepared or showed biliary-to-bowel transit in the HBS study, ruling out BA were excluded from the study. Also, the patients whose final diagnosis could not be confirmed due to the absence of histopathology report or follow-up were excluded from the study.

Image acquisition and analysis

All the studies were acquired after patient preparation with either phenobarbital (5 mg/kg/day) or ursodeoxycholic acid (15 mg/kg/day) for 5 days. The patients were fasting on the day of study for 2–4 h before injection. Patients were injected with 0.5–1 mCi 99mTc-mebrofenin intravenously [5]. Images were acquired on a large field of view gamma camera with low energy, high-resolution parallel hole collimator (GE Discovery NM/CT 670). Planar images were acquired in anterior and posterior views at 5 min, 30 min, 1 h, 2 h, 4 h, 6 h and 24 h. The images were visually assessed by expert Nuclear Medicine physician to look for biliary-to-bowel transit. If no biliary-to-bowel transit was observed, regions of interest (ROI) were drawn over blood pool (heart), liver and kidney at 5 min, 30 min, 1-hour and 24-hour images. The ratio of average counts per pixel in liver: blood pool and liver: kidney ROIs were calculated for the respective time frames.

Statistical analysis

Data analysis was performed using SPSS v 26.0. Based on the histopathology reports, patients were divided into BA and NH groups. All the semi-quantitative parameters in the two groups were presented as Mean ± Standard Deviation (SD). Independent samples student’s t-test was employed to assess the difference of mean between the two groups. Receiver operating characteristic (ROC) curve was also drawn to determine a cut-off of these ratios to differentiate between the two groups. Sub-group analysis in patients < 3 months of age was also done to assess the difference of mean in the BA and NH groups. P-value < 0.05 was considered statistically significant.

Results

The study included 53 patients (37 males) with a median age of 3 months (range: 24 days to 10 months). Of these, 32 patients had a diagnosis of BA while 21 had NH. Median age in BA sub-group was 3 months (range: 24 days to 10 months), while median age in NH sub-group was 2 months (range: 1 month to 8 months). Mean semi-quantitative parameters in the two groups are given in Table 1. Mean Liver: Blood pool ratio (LBR) and Liver: Kidney ratio (LKR) at 24 h were statistically different in the two groups (p-value < 0.05) (Figs. 1 and 2).

Table 1.

Mean semiquantitative parameters in BA and NH. BA: biliary atresia; NH: neonatal hepatitis; SD: standard deviation

Semi-quantitative index	Mean ± SD in BA (n = 24)	Mean ± SD in NH (n = 15)	Significance (p-value)
LBR 5 min	1.27 ± 0.76	1.42 ± 0.46	0.387
LKR 5 min	2.56 ± 1.35	3.16 ± 1.15	0.09
LBR 30 min	2.34 ± 1.35	2.57 ± 0.88	0.454
LKR 30 min	4.24 ± 2.39	4.51 ± 2.15	0.663
LBR 1 h	2.64 ± 1.43	3.17 ± 1.81	0.268
LKR 1 h	4.29 ± 2.17	5.39 ± 3.11	0.167
LBR 24 h	3.61 ± 1.64	5.08 ± 2.41	0.02*
LKR 24 h	3.81 ± 1.47	5.2 ± 2.48	0.029*

Fig. 1.

2 months old child, suspected case of EHBA. 99mTc-mebrofenin HBS performed to rule out EHBA revealed no biliary to intestinal transit of tracer, giving the final differential of EHBA vs. Neonatal hepatitis. On semiquantitative analysis, Mean LBR and LKR at 24 h were 3.3 and 4.4, thus favouring EHBA which was confirmed on final histopathology

Fig. 2.

1 month-old-child, suspected case of EHBA. 99mTc-mebrofenin HBS performed to rule out EHBA revealed no biliary to intestinal transit of tracer, giving the final differential of EHBA vs. Neonatal hepatitis. On semiquantitative analysis, Mean LBR and LKR at 24 h were 7.2 and 5.8, thus favouring Neonatal Hepatitis which was confirmed on final histopathology

Receiver operating characteristic (ROC) curve analyses showed highest AUC for LBR at 24 h 0.683 (CI:0.532–0.834, p-value 0.017) and LKR at 24 h − 0.669 (CI: 0.511–0.827, p-value:0.036). For diagnosis of BA a cut-off value of ≤ 4.18 for LBR at 24 h (sensitivity and specificity of 62.5% and 61.9% respectively) and ≤ 4.64 for LKR at 24 h (sensitivity and specificity of 68.8% and 66.7% respectively) were found to be pertinent (Fig. 3).

Fig. 3.

Receiver operating characteristic (ROC) curve analyses LBR and LKR at 24 h for diagnosing BA

In the sub-group analysis of patients less than 3 months of age, 39 patients (28 males) were included. 24/39 patients had a final diagnosis of BA while 15/39 had NH. Mean semi-quantitative parameters in the two groups are given in Table 2. Mean LBR and LKR at 24 h were statistically different in the two groups (p-value < 0.05).

Table 2.

Mean semiquantitative parameters in BA and NH in subgroup analysis of patients less than 3 months of age. BA: biliary atresia; NH: neonatal hepatitis; SD: standard deviation

Semi-quantitative index	Mean ± SD in BA (n = 24)	Mean ± SD in NH (n = 15)	Significance (p-value)
LBR 5 min	1.39 ± 0.75	1.57 ± 0.40	0.337
LKR 5 min	2.82 ± 1.32	3.44 ± 1.13	0.124
LBR 30 min	2.65 ± 1.34	2.87 ± 0.78	0.536
LKR 30 min	4.77 ± 2.36	4.95 ± 2.19	0.808
LBR 1 h	2.92 ± 1.47	3.56 ± 1.92	0.274
LKR 1 h	4.72 ± 2.17	6.09 ± 3.26	0.164
LBR 24 h	3.97 ± 1.6	5.56 ± 2.4	0.034*
LKR 24 h	4.10 ± 1.41	5.2 ± 2.29	0.041*

Receiver operating characteristic (ROC) curve analyses of patients less than 3 months of age showed highest AUC for LBR at 24 h 0.711 (CI:0.538–0.885, p-value 0.017) (Table 3). For diagnosis of BA a cut-off value of ≤ 4.8 for LBR at 24 h (sensitivity and specificity of 66.7% and 66.7% respectively) was found to be pertinent (Fig. 4).

Table 3.

ROC curve analysis results: AUC, cut-off values, respective sensitivities and specificities of imaging semi-quantitative parameters for diagnosing BA. AUC: area under the curve; ROC: receiver operating characteristic curve; BA: biliary atresia

Semi-quantitative index	Cut off value	Sensitivity	Specificity	AUC	95% CI	p-value
LBR at 5 min	≤ 1.39	66.7	73.3	0.692	0.523–0.860	0.026
LKR at 5 min	≤ 2.99	70.8	66.7	0.681	0.509–0.852	0.039
LBR at 24 h	≤ 4.8	66.7	66.7	0.711	0.538–0.885	0.017
LKR at 24 h	≤ 4.64	62.5	73.3	0.660	0.475–0.845	0.049

Fig. 4.

Receiver operating characteristic (ROC) curve analyses LBR and LKR at 24 h in patients less than 3 months of age for diagnosing BA

Discussion

Neonatal cholestasis refers to the accumulation of bilirubin in the blood, liver and extrahepatic tissues, either due to excessive formation or due to an obstruction leading to reduced excretion of bile. Neonatal cholestatic jaundice, characterized by elevated levels of conjugated bilirubin, is diagnosed when the concentration exceeds 1 mg/dl when the total serum bilirubin is either 5 mg/dl or less, or when total bilirubin surpasses 5 mg/dl with more than 20% being conjugated bilirubin [6].

Among the causes of neonatal cholestasis, idiopathic neonatal hepatitis is the most prevalent, followed by BA [7]. In the initial 3–4 months of life, some degree of physiologic cholestasis can be seen due to inefficient uptake of bile acids and other organic anions by hepatocytes, along with the presence of immature pathways for bile acid conjugation and biliary secretion [6]. The primary concern in this context is to distinguish between pathologic cholestasis and the typically benign physiologic manifestations of this condition.

Biliary atresia is described as a condition involving the complete or partial blockage of the extrahepatic biliary tree. This disorder is marked by the gradual inflammation and fibrosis of the intrahepatic or extrahepatic bile ducts [8]. Hepatobiliary scintigraphy (HBS) is generally regarded as a safe, accurate and non-invasive diagnostic method for this condition [9, 10]. Nevertheless, some studies have reported variable diagnostic accuracy for biliary atresia using hepatobiliary scintigraphy.

Chatziioannou et al. reported sensitivity and specificity of 100% and 64.3%, with an accuracy of 85.6% for hepatobiliary scintigraphy in diagnosing BA [11]. In contrast, Norman et al. found sensitivity, specificity, and accuracy to be 100%, 49.3%, and 60.5%, respectively [12], while Liu X et al. reported values of 91.4%, 83.8%, and 85.9%, respectively [13]. These discrepancies in results may be attributed to factors such as immature hepatocytes and compromised hepatic reserve.

Consequently, few researchers have explored additional semiquantitative parameters to enhance the differentiation between BA and other conditions. For instance, Liu X et al. observed no significant difference in the liver-to-kidney ratio (LKR) between biliary atresia and neonatal hepatitis groups in their study of 128 patients [13]. On the other hand, Mohamadien et al. found that median values of kidney-to-liver ratio (KLR), intestinal-to-liver ratio (ILR), and background-to-liver ratio (BLR) at 30 min were significantly higher in patients with biliary atresia, aligning with the findings of the present study [2]. The authors highlighted the kidney-to-liver ratio (KLR) as the most reliable among the semiquantitative parameters for predicting biliary atresia, with an area under the curve (AUC) of 0.707 (CI: 0.573–0.840). In the current study, amongst all the semi-quantitative parameters at different time frames, LBR at 24 h was found to be most reliable for diagnosing BA, with an AUC of 0.683 (CI: 0.532–0.834).

EHBA is generally believed to have better hepatic extraction function which deteriorates as the fibrosis and hepatocyte damage sets in compared to NH where hepatocyte function is impaired early in the disease presentation. However, the current study revealed worse hepatic extraction function in EHBA with lower LBR and LKR values compared to NH. Mohamadien et al. also found similar results with higher background activity in patients with BA. The authors believed it to be due to higher mean age of their study population [2]. Progressive hepatic dysfunction is seen in EHBA with increasing age while hepatic function shows improvement in neonatal hepatitis. Thus, in the present study, we also performed subgroup analysis for patient with age < 3 months to alleviate the effect of increasing age. The subgroup analysis revealed similar findings i.e. lower LBR and LKR in BA. Also, LBR at 24 h was found to be most reliable semi-quantitative parameter for diagnosing BA, with an AUC of 0.711. This depicts other factors at play rather than simple hepatocytic damage. The responsible factors could be blockage of small biliary canaliculi by inflammatory infiltrates and giant cells in NH as well as component of biliary tree damage from inflammatory infiltrates [14, 15]. It is important to note that the limitations of current study, including a relatively small sample size, single-centric study and its retrospective nature. Thus, further larger studies are needed to establish the applicability of these semi-quantitative parameters in clinical practice.

Conclusion

HBS serves as non-invasive gold standard to rule out EHBA. Semi-quantitative indices LBR and LKR at the 24-hour can accurately differentiate between EHBA and NH even in cases with compromised hepatic function where traditional visual interpretation of tracer transit proves inadequate. Further research is warranted to establish the broader applicability of these semi-quantitative parameters in clinical practice.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Author contributions

Dr Tejasvini and Dr Parneet participated in drafting of the manuscript, and data analysis. Dr Girish participated in the conceptualization, study design, and approval of the final content of the manuscript. Dr Dr Ralph and Ashique participated in data acquisition. Dr Aditi, Dr Kanhaiya and Dr P Sai participated in the manuscript revision. All authors read and approved the final manuscript.

Data availability

The data that support the findings of this study are available from the corresponding author, [Dr Girish], upon reasonable request.

Declarations

Ethics approval

The present study was approved by the Institute Ethical Committee of our institution (IEC No. T/IM-NF/Nucl-Med/23–54) and with the 2013 Helsinki declaration and its later amendments or comparable ethical standards.

Consent to participate

The institutional review board at our institution approved this retrospective study, and the requirement to obtain informed consent was waived.

Consent for publication

As this is a retrospective study, patient consent was not needed. However, patient identity has not been revealed at any stage of the study.

Conflict of interest

The authors (Tejasvini Singhal, Parneet Singh, Girish Kumar Parida, Ralph Emerson, Ashique Rehman, Aditi, Kanhaiyalal Agrawal, and P Sai Sradha Patro) declare that they have no conflict of interest.