Serum Matrix Metalloproteinase 7 as a Diagnostic and Prognostic Biomarker for Extrahepatic Biliary Atresia.

Background: Differentiation of neonatal cholestasis into neonatal hepatitis (NH) and extrahepatic biliary atresia (EHBA) is essential to formulate the treatment plan; promptness is indispensable for optimal outcomes. The clinical and nonoperative algorithms lack precision; the gold standard investigations (liver biopsy or per-operative cholangiogram) are invasive. There is a need for a noninvasive test which is both, sensitive and specific and has a high likelihood ratio. Aim: To study the (diagnostic) role of matrix metalloproteinase 7 (MMP-7) as a serum biomarker to differentiate between EHBA and NH and evaluate the prognostic significance in EHBA based on its correlation with liver histopathology and serological predictors of liver fibrosis - Aspartate-to-Platelet Ratio Index (APRI) and Fibrosis-4 (FIB-4). Materials and
Methods: This was a prospective study conducted upon patients of neonatal cholestasis presenting with acholic stools (n = 46) with equal number of controls (n = 45) with no liver pathology. Observational parametric included disease-specific workup and serum MMP-7 levels (all participants); liver biopsyl and APRI-FIB-4 (EHBA).
Results: (Diagnostic) Serum MMP-7 levels were significantly elevated in EHBA (n = 25; 28 ng/mL) as compared to those in NH (n = 21; 1.88 ng/mL) and normal infants (n = 45; 1.2 ng/mL) (P < 0.001 for both). Serum cutoff at 4.99 ng/mL differentiated EHBA-NH with a high sensitivity (96%), specificity (90.5%), and a negative predictive value (95%), with the number needed to misdiagnose being 23. (Prognostic) Inflammatory activity and fibrosis-stage on liver histopathology (METAVIR-and-Ishak scores) correlated with MMP-7 levels. APRI and FIB-4 scores also depicted a strong correlation with each other, age of the patient, and liver fibrosis. Conclusions: MMP-7 has a diagnostic value in differentiating EHBA from NH and may also be used as a prognostic biomarker in the follow-up of these patients. MMP-7 levels in controls may be used as a baseline for future studies. Copyright:

INTRODUCTION

The prompt differentiation of “neonatal cholestasis” into neonatal hepatitis (NH) and extrahepatic biliary atresia (EHBA) is essential for optimal patient outcomes.[1] The breastmilk jaundice in the 2nd week of life masquerades the presence of associated cholestasis, and the subsequent amelioration of indirect hyperbilirubinemia is falsely assuring.[2] Besides, age is an important determinant of the outcomes following surgery for EHBA.[3] Therefore, any delay in the diagnosis of EHBA leads to an undesirable sequelae of pathological developments inside the liver and leaves the caregivers with a limited time for diagnostic evaluation and surgery. The situation is further complicated by the presence of closely related clinical, biochemical, and imaging profiles of the two entities.[12] Although a liver biopsy can diagnose EHBA with high sensitivity and specificity, it is an invasive procedure and limited by the absence of classic differentiating histopathological features early in the course of the disease.[4] A per-operative cholangiogram (POC) is considered the “gold standard” for differentiating medical jaundice (NH) from surgical pathology (EHBA); yet it is a surgical procedure in itself and necessitates a mini-laparotomy.[5]

The persistent uncertainty in the diagnostic algorithm and need to shorten the time to surgery has been a strong motivation to the research minds, and multiple experimental markers are under various stages of evaluation such as the liver elastography,[67] serum nitric oxide levels,[5] serum matrix metalloproteinase 7 (MMP-7),[891011] serum Interleukin 33,[12] and gene panels identified by the new gene sequencing technologies.[13] The tissue remodeling and liver fibrosis associated with biliary atresia are associated with degradation and deposition of protein components of the extracellular matrix and basement membrane, and the serum levels of MMP-7 are elevated manifold.[111415] Preliminary studies have suggested that the serum levels of MMP-7 can be used to discriminate between EHBA and NH.[8] It has also been suggested in literature that the serum MMP-7 levels rise with the age of the patient, which may be ascribable to progression of obliterative cholangiopathy.[9] Based on these findings, the need for evaluating MMP-7 as a prognostic biomarker based on its correlation with direct and surrogate markers of liver injury becomes pertinent. Despite a myriad of studies, there is an impending need for a biomarker which can differentiate between EHBA and NH, simplify the diagnostic algorithm, decrease the outpatient department-to-operating room time lag for patients with EHBA and save a diagnostic procedure (POC) of operative nature in those with medical jaundice (NH).

The current study was designed to study the role of MMP-7 as a serum biomarker to differentiate between EHBA and NH. The study had also been configured to evaluate the functional and prognostic significance of serum MMP-7 levels in EHBA by looking into its correlation with liver inflammation and fibrosis. The relevance of APRI and FIB-4 scores in quantifying liver injury vis-à-vis serum MMP-7 and liver histopathology has also been studied in patients with EHBA.

MATERIALS AND METHODS

This prospective, single-center, cross-sectional study was conducted in the Department of Paediatric Surgery at the parent institute after seeking approval of the Institute's Ethics Committee (IECPG-235/22.04.2019) over a period of 14 months (May 2019–June 2020). Consecutive cases of neonatal cholestasis presenting with acholic stools (n = 46) were enrolled in the study. An equal number (n = 45) of age-and sex-matched infants presenting for unrelated noninfectious conditions not involving the liver and not known to affect the serum MMP-7 levels were included as controls.

Predesigned templates were used to record demographic parameters, clinical findings, complete blood count, liver function tests, and findings upon ultrasound evaluation and postpriming hepatobiliary scintigraphy.

At the time of routine blood sampling, an additional 2.5-mL sample was collected in a plain vial for serum MMP-7 estimation. The sample was transported to the laboratory in ice, centrifuged at 2500 rpm for 3 min, and stored in aliquots at − 80°C to avoid loss of bioactive MMP-7. Commercially available human MMP-7 enzyme-linked immunosorbent assay (ELISA) kit (RayBio Human MMP-7 ELISA kit, Parkway Lane, GA, USA), was used to quantify serum MMP-7 and the values were recorded in ng/mL. All measurements were performed in triplicate to enhance the reliability of the results.

Participants who demonstrated gut activity on postpriming hepatobiliary scintigraphy were labeled as NH, whereas others were labeled as suspect EHBA and further evaluated with a POC.[5]

A wedge liver biopsy was taken at the time of surgery for EHBA and processed for standard histopathology. Staging of fibrosis and activity assessment were done using the METAVIR scoring system[16] and the Ishak's modification of Knodell's Histological Activity Index score.[16] The METAVIR scores were as follows: Inflammatory activity (A0 = none; A1 = mild; A2 = moderate; and A3 = severe inflammation) and fibrosis stage (F0 = no fibrosis; F1 = portal fibrosis without septa; F2 = portal fibrosis with few septa; F3 = portal fibrosis with many septa; and F4 = cirrhosis). The Ishak scores for necrosis and inflammatory activity were as follows: histological grading scores ranging 0–18; re-stratified as mild 0–6, moderate 7–12 and marked 13–18 and fibrosis stage were: F0 = no fibrosis; F1 = fibrous expansion of some portal areas, F2 = fibrous expansion of most portal areas; F3 = fibrous expansion of most portal areas with occasional bridging, F4 = fibrous expansion of most portal areas with marked bridging; F5 = marked bridging with occasional nodules/incomplete cirrhosis; F6 = probable or definite cirrhosis. Patients were re-grouped into early (METAVIR stage 1, 2; Ishak's Stage 1–3) and advanced fibrosis (METAVIR stage 3, 4; Ishak's stage 4–6).

Aspartate-to-Platelet Ratio Index (APRI) and Fibrosis-4 (FIB-4) scores were calculated from the recorded information.[17]

Statistical analysis was carried out using Stata v15.0 (College Station, Texas, USA). Serum MMP-7 levels, APRI, and FIB-4 were assessed for normal distribution using Shapiro–Wilk test. MMP-7 levels were then compared among the three groups using Kruskal–Wallis test, followed by pair-wise comparison using Wilcoxon rank-sum test. The P value was adjusted using Bonferroni correction. Receiver-operating characteristic (ROC) curves were drawn to assess the diagnostic performance of MMP-7 in EHBA versus NH and NH versus controls groups. The results were presented as cutoff, sensitivity, specificity, and likelihood ratio. Area under ROC was calculated. The association of MMP-7 and APRI and FIB-4 scores with METAVIR and Ishak activity grades and fibrosis staging was assessed using Jonkheere–Terpstra test and Wilcoxon rank sum test, respectively. P <0.05 was considered significant.

RESULTS

The participants were substratified into EHBA (n = 25) and NH (n = 21) based on postpriming hepatobiliary scintigraphy and POC. The participants in the three groups (EHBA, NH, and controls) were comparable for age (median): EHBA = 87 days (interquartile range [IQR]: 74–121 days), NH = 88 days (IQR: 70–120 days), and controls = 86 days (IQR: 59–103 days) [Table 1].

Table 1

Outline of the study participants

	Study participants (sample size)			*P
	EHBA (n=25)	NH (n=21)	Controls (n=45)
Median age (days) (IQR, days)	87 (74-121)	88 (70-120)	86 (59-103)	0.417
Gender profile, n (%)
Male	19 (76.0)	15 (71.4)	32 (71.1)	0.901
Female	6 (24.0)	6 (28.6)	13 (28.9)
Serum MMP-7, median level (ng/mL) median (IQR, ng/mL)	28.1 (19.1-30.7)	1.88 (1.37-2.09)	1.2 (1.08-1.32)	<0.001

*P<0.05 was considered significant. IQR: Interquartile range, MMP: Matrix metalloproteinase, EHBA: Extra-hepatic biliary atresia

Levels of serum matrix metalloproteinase 7 in normal infants

The median serum MMP-7 level in normal infants (n = 45) was 1.2 ng/mL (IQR: 1.08–1.32 ng/mL). The levels of MMP-7 were relatively constant across the age range of the participants in this group (Spearman's correlation coefficient ρ = −0.010, P = 0.96). MMP-7 levels were also comparable between male and female participants (M: F =32:13; males = 1.21 ng/mL, IQR 1.12–1.32; females = 1.14 ng/mL, and IQR 0.97–1.22; P = 0.139).

Levels of serum matrix metalloproteinase 7 in neonatal cholestasis and its role in differentiating extrahepatic biliary atresia from neonatal hepatitis

The median serum MMP-7 level was 28.1 ng/mL in EHBA (n = 25; IQR: 19.1–30.7) which was 15-fold higher as compared to the values observed in NH (n = 21; median: 1.88 ng/mL, IQR: 1.37–2.09). As compared to the control patients, the levels of MMP-7 were 23-fold higher in patients with EHBA. The levels in NH were 1.5 fold higher than those in normal infants.

The difference between the three groups was statistically significant (P < 0.001) [Figure 1].

Figure 1

Box-and-Whisker plot depicting the distribution of matrix metalloproteinase 7 (ng/mL) in patients with extrahepatic biliary atresia and neonatal hepatitis and in control participants. The middle horizontal line represents the median matrix metalloproteinase 7 (ng/mL) and the upper and lower bounds of the box represent the 75th and the 25th centile of matrix metalloproteinase 7 (ng/mL), respectively

Levels of MMP-7 were not different between male and female patients of EHBA (M: F = 19:6; males = 28.1 ng/mL, IQR 21.2–30.7; females = 24.3 ng/mL, IQR 13.9–47.5; P = 0.975) or NH (M: F = 15:6; males = 1.89 ng/mL IQR 1.63–2.08; females = 1.44 ng/mL, IQR 1.2–2.84; P = 0.392). However, there was a rise in the serum levels of MMP-7 with age (positive correlation), which was significant (ρ =0.450, P = 0.039) for NH but not for EHBA (ρ =0.180, P = 0.081). The serum MMP-7 levels were not different for EHBA patients who were <90 days of age and those who were >90 days.

Correlation of matrix metalloproteinase 7 levels with other facets of extrahepatic biliary atresia

No correlation of MMP-7 levels was detected with other facets of EHBA such as total bilirubin (ρ =0, P = 0.986), direct bilirubin (ρ =0.03, P = 0.883), aspartate transaminase (ρ =0, P = 0.988), alanine transaminase (ρ =0.11, P = 0.595), alkaline phosphatase (ρ =0.11, P = 0.585), total serum proteins (ρ =0.16, P = 0.442), and serum albumin (ρ =0.16, P = 0.399).

Role of serum matrix metalloproteinase 7 levels as a diagnostic biomarker for extrahepatic biliary atresia

The area under the ROC curve (AUROC) for MMP-7 (ng/mL) predicting EHBA vis-a-vis NH was 0.99 (95% confidence interval [CI]: 0.973–1), thus demonstrating excellent diagnostic performance. At a serum cutoff of 4.99 ng/mL, 96% of the patients (24 of 25) were correctly classified as EHBA, approaching a sensitivity of 96% (95% CI: 80–100) and a specificity of 90.4% (95% CI: 76–100) [Figure 2 and Table 2]. The negative predictive value of MMP-7 for differentiating EHBA from NH was calculated at 95% (95% CI: 75–100). The number needed to misdiagnose (NNM) a case was calculated as 23.

Figure 2

Receiver-operator characteristic curve analysis showing diagnostic performance of matrix metalloproteinase 7 (ng/mL) in predicting extrahepatic biliary atresia (n = 25) versus neonatal hepatitis (n = 21)

Table 2

Receiver-operator characteristic analysis between patients with extrahepatic biliary atresia, patients with neonatal hepatitis, and controls (normal infants)

MMP-7 (ng/mL)	Cutoff for serum MMP-7 levels	Sensitivity (%)	Specificity (%)	Positive likelihood ratio	Negative likelihood ratio	AUC (95% CI)
EHBA versus NH	>/=4.99	96	90.4	20.16	0.04	0.99 (0.97-1.00)
NH versus controls	>/=1.56	76	71	16.07	0.3	0.85 (0.74-0.96)

MMP-7: Matrix metalloproteinase 7, EHBA: Extrahepatic biliary atresia, NH: Neonatal hepatitis, AUC: Area under curve, CI: Confidence interval

The AUROC for MMP-7 (ng/mL) predicting NH vis-a-vis controls was 0.851 (95% CI: 0.74–0.962), depicting good diagnostic performance. At a cutoff of MMP-7 (ng/mL) ≥1.56, it predicted NH with a sensitivity of 76% and a specificity of 71%, correctly classifying 87.8% of patients [Figure 3 and Table 2].

Figure 3

Receiver-operator characteristic curve analysis showing diagnostic performance of matrix metalloproteinase 7 (ng/mL) in predicting neonatal hepatitis (n = 21) versus controls (n = 45)

Functional and prognostic significance of matrix metalloproteinase 7: Correlation with liver histopathology

The grading of EHBA patients based on METAVIR and ISHAK is shown in Table 3. There is a very high concordance of the grades of liver injury between the ISHAK and METAVIR scores. The inflammatory activity was graded as mild, moderate, and severe in 16% (n = 4), 60% (n = 15), and 24% (n = 6) of the patients, respectively, upon either scale with respective median MMP-7 levels as 4.99 ng/ml (IQR: 4.1–17.1 ng/mL), 27.5 (IQR: 21.1–29.7 ng/mL), and 47.9 (IQR: 30.7–63.3 ng/mL). Higher grades of inflammatory activity were associated with higher MMP-7 levels, which was significant statistically (P= 0.003) [Table 4].

Table 3

Quantification of fibrosis and inflammation in liver biopsy specimens in extrahepatic biliary atresia patients with the METAVIR and Ishak’s (modified Knodell’s histological activity index) scoring systems

Histopathology type	Grade/stage	n (%)
METAVIR activity grade	Mild (A1)	4 (16)
	Moderate (A2)	15 (60)
	Marked (A3)	6 (24)
METAVIR fibrosis stage	F1	1 (4)
	F2	2 (8)
	F3	8 (32)
	F4	14 (56)
METAVIR fibrosis category**	Early fibrosis (F1-F2)	3 (12)
	Advanced fibrosis (F3-F4)	22 (88)
Ishak activity grade*	Mild (0-6)	4 (16)
	Moderate (7-12)	15 (60)
	Marked (13-18)	6 (24)
Ishak fibrosis stage	F1	0
	F2	1 (4)
	F3	2 (8)
	F4	0 (0)
	F5	8 (32)
	F6	14 (56)
Ishak fibrosis category**	Early fibrosis (F1-F3)	3 (12)
	Advanced fibrosis (F4-F6)	22 (88)

*Patients were re-grouped into mild (Ishak activity grade 0-6), Moderate (Ishak activity grade 7-12) and marked (Ishak activity grade 13-8) categories for the purpose of statistical analysis, **Patients were re-grouped into early and advanced fibrosis categories (METAVIR stage 1, 2 versus 3, 4 and Ishak’s stage 1-3 versus 4-6, respectively) for the purpose of statistical analysis

Table 4

Correlation between inflammation and fibrosis upon liver histopathology (METAVIR/Ishak scoring) and serum matrix metalloproteinase-7 Levels

Histopathologic observations upon liver biopsy	Number of patients (n)	Serum MMP-7 value (ng/mL), median (IQR)	*P
METAVIR/Ishak activity grading			0.003
Mild	4	4.99 (4.1-17.1)
Moderate	15	27.5 (21.1-29.7)
Marked	6	47.9 (30.7-63.3)
METAVIR fibrosis grading			0.013
Early	3	4.99 (3.3-12.5)
Advanced	22	28.8 (21.1-32.3)

*P<0.05 were considered significant. MMP-7: Matrix metalloproteinase-7, IQR: Inter-quartile range

This study has also depicted that the serum MMP-7 levels bear a strong correlation with the inflammatory activity in the liver (Ishak scale: 0–18) [ρ =0.578, P = 0.003, Figure 4].

Figure 4

Correlation between serum matrix metalloproteinase 7 levels (ng/mL) and inflammatory activity in the liver (Ishak scale 0–18)

Similarly, patients of EHBA were substratified into early fibrosis (n = 3; 12%) and advanced fibrosis (n = 22; 88%) based on either score with corresponding MMP-7 levels as 4.99 ng/mL (3.3–12.5 ng/mL) and 28.8 ng/mL (21.1–32.3 ng/mL). Higher stages of fibrosis were associated with higher MMP-7 levels (P = 0.013) [Table 4].

Predictive accuracy of Aspartate-to-Platelet Ratio Index and fibrosis-4 markers in quantifying liver injury vis-à-vis serum matrix metalloproteinase 7 and liver histopathology

The FIB-4 and APRI scores were calculated for all the patients of EHBA. The median value of FIB-4 was 0.0169 (IQR: 0.00612–0.02902; n = 25) and the median APRI score was 1.69 (IQR: 0.95–3.54; n = 25) for patients with EHBA. Both APRI and FIB-4 depicted a strong correlation with the age of the patient: APRI vs. age: ρ =0.503, P = 0.010 and FIB-4 vs. age: ρ =0.634, P = 0.001. APRI and FIB-4 also correlated with each other strongly (ρ =0.832, P < 0.001).

Furthermore, the APRI scores were calculated for patient subgroups based on METAVIR/Ishak activity and fibrosis categories [Table 5]. There was no significant association between the FIB-4 or APRI values and the grades of inflammatory activity on either of the two histopathology scores. However, the METAVIR/Ishak fibrosis stage depicted a positive association with the FIB-4 scores (P = 0.04). Higher stage of fibrosis on the METAVIR/Ishak systems was also observed in patients with high APRI scores, but the association failed to reach significance (P = 0.054).

Table 5

Aspartate-to-Platelet Ratio Index and fibrosis-4 scores in extrahepatic biliary atresia patients and their correlation with inflammation and fibrosis upon liver histopathology (METAVIR/Ishak scoring)

Histopathologic observations upon liver biopsy	Number of patients (n)	FIB-4, median (IQR)	APRI, median (IQR)
METAVIR/Ishak activity grading
Mild	4	0.00566 (0.00528-0.02840)	1.34 (0.97-2.92)
Moderate	15	0.02184 (0.00993-0.03552)	2.01 (0.95-3.54)
Marked	6	0.01309 (0.00710-0.01690)	1.5 (0.94-6.66)
P		METAVIR/Ishak activity versus FIB-4: 0.861	METAVIR/Ishak activity versus APRI: 0.573
METAVIR fibrosis grading
Early	3	0.00563 (0.00535-0.00597)	0.68 (0.63-1.43)
Advanced	22	0.01988 (0.00993-0.03552)	1.91 (1.25-4.06)
P		METAVIR/Ishak fibrosis versus FIB-4: 0.040	METAVIR/Ishak fibrosis versus APRI: 0.054

FIB-4: Fibrosis-4, IQR: Interquartile range, APRI: Aspartate-to-platelet ratio index

Higher values of MMP-7 were observed in EHBA patients with higher FIB-4 and APRI scores, but the association was not significant.

DISCUSSION

EHBA is a rare, though devastating disease with an obscure, probably multifactorial etiology.[25] The pathophysiology involves obstructive cholangiopathy of the extrahepatic bile duct, which culminates into liver fibrosis, cirrhosis, and eventual liver failure. The outcomes are suboptimal despite the best possible treatment; age at surgery is one of the crucial factors contributing to the final outcomes.[3] The clinical presentation overlaps with that of neonatal cholestasis with alternate causes, and prompt differentiation from the same is necessary to offer the patient the best chances of survival. The proposed clinical algorithms for differentiation between the two entities are imprecise and lead to an inadvertent delay in diagnosis and treatment.[11819]

MMP-7, also been referred to matrilysin (lysis of matrix), uterine metalloproteinase, or PUMP-1 protease, is a protease enzyme which is a member of the MMP family consisting of structural related zinc-dependent endopeptidases.[2021] MMP-7 is constitutively expressed by the normal cholangiocytes. The primary role of activated MMP-7 is to breakdown the extracellular matrix during normal physiological processes including tissue remodeling by degrading macromolecules including casein, gelatins (Types I, II, IV, and V), fibronectin, and proteoglycan.[2223] The normal liver has relatively minor extracellular matrix; however, the cascade of events leading to liver fibrosis involve extensive tissue remodeling and a gain in the extracellular matrix. MMP-7 has been implicated in tissue remodeling during biliary atresia associated liver fibrosis.[2425] The expression of MMP-7 is weak in normal liver and is highly upgraded in response to progressive biliary atresia.[142627]

The median age of presentation of EHBA in the current study is 87 days (IQR 74–121 days). Considering that the study has been conducted at a tertiary care center of the country and the assumption that the data is a true representation of the Indian statistics, it is a matter of significant concern that the age at presentation compares unfavorably with corresponding studies from other countries – Yang et al.[8] have reported a mean of 54 days (0–6 months; n = 75), Wu et al.[9] have reported a mean age of 42.4 ± 3.6 days (n = 36), while Jiang et al.[10] reported a median age of 59 days (IQR 48–69 days; n = 187). Because the levels of MMP-7 are dependent on the extent of liver fibrosis, the levels are likely to rise with age and progression of disease pathology. Moreover, a delayed treatment is likely to culminate in poorer outcomes.

The study has established the serum levels of MMP-7 (median value 1.2 ng/mL, IQR 1.08–1.32 ng/mL; n = 45) in normal infants across the age range of 59 days to 103 days, and there were no age-related differences in this bracket. These data may serve as a baseline reference for further studies. The reported value of MMP-7 in normal infants from China[10] is 2.4-fold higher (2.86 ng/mL, IQR: 1.32–5.32 ng/mL, n = 54); the authors have also tested the MMP-7 values in older healthy cohort of 6 months–2 years (n = 8) and adults (n = 10) and found no difference across the entire spectrum.

The study has witnessed that the serum levels of MMP-7 levels are 23-fold higher in EHBA (median MMP-7: 28.1 ng/mL, IQR: 19.1–30.7) as compared to normal babies and 15-fold higher as compared to those with NH. The results are in concurrence with the previous reports.[8910] However, there is a wide variation in median values between different studies, with the highest values being reported by Yang et al. from China with a rise of 42-fold when compared to age-matched controls. This could be ascribable to various factors such as the age of the patient, ethnicity, differences in sample collection, technique of measurement (including dilution technique and kit used), sample storage, and severity of liver injury. However, two studies from China itself have reported a difference in MMP-7 values[810] although the age group of the patients under study is not grossly different (54 days in Yang et al. and 59 days in Jiang et al. for EHBA). Moreover, the two studies have estimated the MMP-7 values by the same technique (ELISA) and used same kits. Prolonged storage of the samples will lead to degradation and false readings. The extent of liver injury on the other hand will increase with age, with the disease being progressive. Wu et al. have reported upon the youngest age group and the serum MMP-7 levels are the lowest of all. However, with India reporting in older patients, the values should have been higher than any of these reports.[8910] Whether there are any ethnic reasons for the diverse MMP-7 values across various studies, or the differences are due to varying laboratory protocols, remains elusive.

The serum MMP-7 levels were comparable in samples obtained from infants with EHBA at >90 (n = 11) days of age compared with <90 (n = 14) days of age. The arbitrary value of 90 days was chosen based on the prognostic cutoff for surgery.[28] Subgroup stratification at a cutoff of 30 days and 60 days was not possible as the study cohort included 1 and 4 patients in the respective subgroups. Wu et al.[9] have, however, shown that the serum MMP-7 levels were significantly higher in EHBA patients beyond 30 days of age. The difference may be explained by the difference in the progression of the disease pathology with age and establishment of liver fibrosis.

The data from the current study have supported the diagnostic role of serum MMP-7 for biliary atresia and established a serum cutoff level of 4.99 ng/mL to stratify neonatal cholangiopathy into EHBA and NH with sensitivity, specificity, and negative predictive values of 96%, 90.4%, and 95%, respectively [Figure 2]. This compares favorably with the data reported in literature.[8910] Statistical analysis of the data in the index study has shown that the NNM a case is 23, which is very close to the value of 25 reported in literature.[8] However, the cutoff levels suggested by different studies are widely separated including 1.43 ng/mL and 52.85 ng/mL by Wu et al. and Yang et al., respectively. The cutoff value is a function of baseline and pathologic MMP-7 levels; similar factors may be responsible for this observation.

It is known that liver histopathology presents with features which are instrumental in differentiating biliary atresia from other causes of neonatal cholestasis. The METAVIR scoring system and the Ishak's modification of Knodell's Histological Activity Index were used in this study to quantify the extent of liver fibrosis in EHBA patients. Both these scoring systems are accepted widely for the assessment of fibrosis and necroinflammation.[29] A “strong” or “almost perfect” agreement was observed for staging of fibrosis using both these scoring systems both in diagnostic and surgical liver biopsies.[16] The METAVIR and Ishak scores were used individually to quantify the processes of “inflammatory activity” and “fibrosis.” Both these parameters were found to correlate strongly with the serum MMP-7 levels. These observations have further suggested the potential role of MMP-7 as a noninvasive prognostic biomarker or a surrogate indicator of the severity of liver injury. These observations also suggest indirectly that follow-up MMP-7 levels after Kasai's portoenterostomy may be used to monitor the progress of the disease pathology. Wu et al.[9] reported a positive correlation between serum MMP-7 levels and collagen content in the liver. Jiang et al.[10] had documented an association with liver fibrosis but not with inflammatory activity over the METAVIR scale, which was contrary to our observations. The discrepancy in observations may partly be explained by the age difference of the patients in the two studies. Our patients were older at the time of presentation and treatment and the disease pathology is expected to be more advanced.

The APRI and FIB-4 index are two of the most popular scoring systems to evaluate liver fibrosis. High degree of accuracy in cases with fibrosis and cirrhosis due to chronic hepatitis C has been demonstrated previously.[3031] The utility of APRI in EHBA as a predictor of liver fibrosis and cirrhosis, native liver survival, and prognosis has been disputed between studies.[32333435363738] Reports pertaining to the utility of FIB-4 in biliary atresia as a serological indicator of liver fibrosis are also conflicting.[73940] However, Hwang et al. documented a positive correlation of both APRI and FIB-4 with progressing clinical categories in EHBA.[41] The current study also failed to identify any association of the APRI and FIB-4 scores in EHBA with the inflammatory activity upon METAVIR and ISHAK scores upon histopathology. An association of liver fibrosis on either scale was documented with FIB-4 but not with APRI. The serum MMP-7 levels are a reflection of the regenerative activity inside the liver, while the serological markers under study (APRI and FIB-4) are indicative of liver fibrosis. Although higher levels of MMP-7 were observed in EHBA patients with higher APRI and FIB-4 scores, the study could not establish an correlation between these parameters. A similar comparison could not be found in literature. These findings may be explained by the observations from the meta-analysis by Lin et al. wherein a pooled analysis of the data from forty different studies suggested that the score is more valid for cases with advanced pathology.[30] The same holds true for FIB-4 as well.[42] It has been concluded often that the use of multiple indices in combination and an algorithmic approach may yield higher diagnostic accuracy.[43]

The study has few limitations. It was conducted at a single center. However, the chosen sample size was sufficient for the disease prevalence, and all efforts were made to collate the findings of this study, with all studies reporting upon similar variables. The study design is suited to determine the diagnostic value of MMP-7 levels for EHBA and differentiating it from NH. However, it would require a longitudinal study to establish the prognostic value of MMP-7 as a biomarker of liver injury and postoperative recovery.

CONCLUSIONS

The diagnostic and prognostic value of MMP-7 and its role as noninvasive biomarker in the diagnosis and management of EHBA has been strongly suggested by the study and is in concordance with prior reports from other countries. The current study has also established the baseline values of serum MMP-7 in normal infants <6 months of age which may be used as a baseline reference. The high concentrations of MMP-7 in EHBA may be used to sub-stratify the patients of neonatal cholangiopathy into EHBA and NH with high sensitivity, specificity, and negative predictive values. The NNM has been calculated at 23. The strong correlation of MMP-7 with histopathological injury has also suggested that MMP-7 may be used as a prognostic biomarker of liver fibrosis and a follow-up marker in EHBA. Last but not the least, the role of APRI and FIB-4 as markers of liver fibrosis in the management of EHBA could not be authenticated.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.