Blood-Based Non-Invasive Tests of Hepatic Fibrosis in Autoimmune Hepatitis: Application among Selected Patients Leads to Higher Accuracy
1
Center for the Study and Treatment of Autoimmune Diseases of the Liver and Biliary System, Policlinico di Sant’Orsola, Istituto di Ricovero e Cura a Carattere, Scientifico Azienda Ospedaliero-Universitaria di Bologna, University of Bologna, 40138 Bologna, Italy
2
Department for the Science of the Quality of Life (QUVI), University of Bologna, 40138 Bologna, Italy
*
Author to whom correspondence should be addressed.
Gastroenterol. Insights 2022, 13(3), 286-295; https://doi.org/10.3390/gastroent13030029
Submission received: 31 July 2022
/
Revised: 4 September 2022
/
Accepted: 7 September 2022
/
Published: 9 September 2022
(This article belongs to the Special Issue Feature Papers in Liver Research)
Abstract
:Background. Assessment of liver fibrosis is essential to guide treatment in autoimmune hepatitis (AIH), but non-invasive tests (NITs) showed poor accuracy. Our study aims to evaluate the performance of NITs among different AIH presentations. Methods. Monocentric retrospective study among 122 AIH patients. NITs were compared to histological grading of liver fibrosis. We performed an accuracy analysis among acute (jaundice and/or transaminases > 10 times upper limit of normal) and non-acute patients. Results. A significant difference in the distribution of NIT values for each Ishak stage was found for spleen-diameter-to-platelet-count ratio (SD/PC) (p < 0.001), fibrosis-4-score (FIB-4) (p = 0.002), AST-to-ALT ratio (AAR) (p = 0.002), red-blood-cell-width-distribution-to-platelet-count ratio (RDW/PC) (p = 0.008) and AST-to-platelet-count ratio (APRI) (p = 0.029). The AUC for advanced fibrosis of SD/PC, FIB-4, RDW/PC, APRI and AAR were, respectively, 0.814, 0.770, 0.768, 0.708 and 0.694. The AUC of SD/PC, FIB-4 and APRI in non-acute subgroup were 0.902, 0.834 and 0.758, while in acute patients they were 0.754, 0.724 and 0.716. RDW/PC and AAR weren’t different among the two subgroups. Conclusions. For SD/PC, FIB-4 and APRI, diagnostic accuracy is higher in patients with non-acute presentation. In this context, SD/PC and FIB-4 showed an overall performance that could be of interest in clinical practice alongside other non-invasive techniques.
1. Introduction
Autoimmune hepatitis (AIH) is defined as an immune-mediated inflammation of the liver, characterized by the presence of hypergammaglobulinemia, autoantibodies and interface hepatitis on biopsy [1]. AIH has a wide spectrum of clinical manifestations, ranging from asymptomatic forms and forms with non-hepatospecific symptoms, to forms of acute hepatitis (occasionally leading to liver failure) and presentations with symptoms of advanced liver disease or portal hypertension [2].
Assessment of liver fibrosis stage is essential to determine prognosis, guide treatment strategies, evaluate their effectiveness, and to suggest the need for regular screening for complications of cirrhosis [1]. Liver biopsy is considered the gold standard for the evaluation of liver fibrosis stage, but it is limited by sampling variability and its intrusiveness, with a reported risk of death of 0.2% and of severe bleeding of 0.6% [3]. Moreover, it has high costs, and it is unacceptable to patients to undergo periodic biopsies to monitor liver fibrosis stage [4]. Liver stiffness measurement with vibration controlled or shear wave-based elastography methods is widely accepted as the most accurate non-invasive method for detection of fibrosis in AIH [5,6], but it still has a relatively high cost and 5–10% failure rate [4], and it seems to be affected by inflammation activity of acute forms of AIH [5,7].
Several non-invasive tests, including laboratory and radiological tests, have been proposed, but unlike studies in chronic viral hepatitis or NAFLD, they showed contrasting data about their diagnostic accuracy among AIH patients. Moreover, at the day there are no studies taking into account the different performances of these tests in acute and non-acute AIH clinical subsets.
2. Materials and Methods
This is a retrospective study among 122 AIH patients at the time of diagnosis, from the Center of Autoimmune Diseases of the Liver and Biliary System, Policlinico di Sant’Orsola, University of Bologna. AIH diagnosis was defined in accordance with guidelines of the International Autoimmune Hepatitis Group. Patients with primary biliary cholangitis, primary sclerosing cholangitis, overlap syndromes or other forms of liver disease were excluded. Patients with inadequate liver biopsy were also excluded. Demographic and clinical data, blood chemistry, abdominal ultrasound and histological data were collected after reviewing the outpatient records.
Clinically, the presentation of AIH was arbitrarily defined as “acute” when there was evidence of acute hepatitis with jaundice and, under the laboratory profile, when transaminases 10 times the normal limit and/or a value of total bilirubin greater than 5 mg/dl were found. The presentation was defined as “non-acute” when non-specific symptoms were present (such as asthenia, weight loss, amenorrhea, fever, nausea, loss of appetite) or in case of asymptomatic presentation, without “acute” laboratory features.
Biochemical and hematological parameters collected included white blood cells (WBC), neutrophils, lymphocytes, red blood cells (RBC), hemoglobin (Hb), mean corpuscular volume (MCV), red blood cell volume distribution width (RDW), platelet count (PC), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) and their respective increase from the upper limit of norm (ULN), alkaline phosphatase, gamma-glutamyltransferase and albumin. From abdominal ultrasound reports we collected spleen diameter (SD). All patients underwent liver biopsy, and the samples obtained were analyzed by expert pathologists of the Polyclinic. Liver fibrosis stage was assessed according to Ishak. Advanced fibrosis (AF) was defined as Ishak stage equal or greater than 4. Since there was only one patient with Ishak = 6, he was grouped to Ishak 5.
All the evaluations (abdominal ultrasound, laboratory analyses) were performed on average 23 days before the liver biopsy; all patients were treatment-naive at the time of the investigations. Non-invasive tests (NIT) calculated are fibrosis 4 score (FIB-4), AST to PC ratio (APRI), AST to ALT ratio (AAR), RDW to PC (RDW/PC), SD to PC ratio (SD/PC).
Data are presented as medians with interquartile ranges (IQRs) and n (%) unless otherwise stated. Comparisons between two groups were made using Mann–Whitney U-test and chi-square test as appropriate. The correlation between NIT values and fibrosis stages was estimated using ANOVA Kruskal–Wallis test. Diagnostic value of each NIT was assessed by the area under the receiver operating characteristic curve (AUC). Comparison AUCs and 95% confidence intervals (CIs) were performed using the nonparametric Delong test. A value of p < 0.05 was considered significant. Youden’s index was used to identify the optimum cut-off point of the AUCs. Positive and negative predictive values (PPV and NPV) and likelihood ratios (LRs) for the appropriate cut-offs were calculated. Data were analyzed using JASP statistics software version 0.11.1.0 and easyROC, a web-tool for ROC curve analysis (ver. 1.3.1). Diagnostic performance is considered acceptable for AUC greater than 0.7 [8] and of value in clinical practice for AUC greater than 0.8 [9].
3. Results
Mean age was 49 years (33.25–58.00) and female sex accounted for 73.70% of patients involved. Histological Ishak stage was 1 in 27, 2 in 41, 3 in 27, 4 in 17 and 5 in 10 patients. Advanced fibrosis (Ishak stage ≥ 4) was observed in 27 patients (22.1%). 63 patients (51.63%) presented with an acute form of hepatitis. Comparison between laboratory features and NITs among absent/minimal fibrosis and advanced fibrosis groups is presented in Table 1.
Distribution of NITs for each Ishak stage is shown on Figure 1. Using Kruskal–Wallis test, we found a significant difference among the different stages of fibrosis when comparing the values for SD/PC (p < 0.001), FIB-4 (p = 0.002), AAR (p = 0.002), RDW/PC (p = 0.008) and APRI (0.029).
According to the results of receiver operating characteristics curves analysis (Table 2), the best performance for the diagnosis of AF was seen for SD/PC (AUC = 0.814; 95% CI 0.719–0.909), followed by FIB-4 (AUC = 0.770; 95% CI 0.670–0.870), RDW/PC (AUC = 0.768; 95% CI 0.640–0.896), APRI (AUC = 0.708, 95% CI 0.592–0.823) and AAR (AUC = 0.694; 95% CI 0.568–0.819) (ROC curves analysis in Table 3 and Figure 2). Statistical comparison of the curves with DeLong test did not show statistical superiority of one of these tests (not shown in tables), due to the limitation of the sample.
For each NIT we identified the best cut-off by Youden’s index and its performance (Table 3). For SD/PC, optimal cut-off was calculated at 0.054 (cm/unit × 109/L), where it gives high sensitivity and NPV, with a good specificity (88.0%, 94.0% and 68.9%), keeping a PPV of more than 50%. FIB-4 and APRI (optimal cut-off calculated at the maximum of Youden’s index 2.00 and 6.04 respectively) showed great ability to rule out AF with good sensitivity and NPV (96.0% and 97.5%, 70.8% and 88.7% and respectively), but quite low specificity and PPV (48.7% and 36.9% and 68.8% and 40.5% respectively). RDW/PC and AAR (optimal cut-off calculated at 0.087 and 0.93 respectively) showed high NPV (both close to 90%), keeping a good specificity (85.9% and 87.7% respectively).
The performance of NITs in identifying AF in patients with non-acute presentation and patients with acute hepatitis was compared using ROC curves analysis (Table 2, Figure 2). The AUCs of SD/PC (non-acute AUC = 0.902, CI 95% 0.814–0.990; acute AUC = 0.754, CI 95% 0.606–0.902), FIB-4 (non-acute AUC = 0.834, CI 95% 0.679–0.990; acute AUC = 0.724, CI 95% 0.567–0.881) and APRI (non-acute 0.758, 95% CI 0.579–0.938; acute AUC = 0.716, 95% CI 0.553–0.879) are higher in non-acute patients. Comparison with DeLong test did not show statistical significance (p = 0.12, p = 0.37, and p = 0.73 respectively). AUCs of AAR and RDW/PC did not show relevant differences in the two subsets of patients. Best cut-off for FIB-4 in the non-acute group is 3.3, where it keeps high sensitivity and NPV, giving higher specificity and PPV compared to the general population set. For SD/PC, in non-acute patients the best cut-off was 0.06, where it has a NPV of 96.7% (Table 3).
4. Discussion
The central driver of hepatic fibrosis in AIH is adaptive immune response, which triggers the extrinsic pathway of hepatocyte apoptosis, inducing myofibroblast transformation of Kupffer and stellate cells, which in turn expands the extracellular matrix [10]. This process is now recognized as dynamic, since there is a high potential for fibrosis regression in AIH if liver inflammation is effectively controlled [1,10]. There is agreement that assessment of hepatic fibrosis in addition to complete biochemical remission may help to assure a lack of disease progression over the long term, and may reduce the need for follow-up liver biopsy [11,12]. In this context, our work aims to evaluate the performance of blood-based NITs that, to date, have given contrasting results in literature. Moreover, we wanted to clarify whether they are affected by different forms of presentation of AIH. In Table 4 we summarize the literature for blood-based NITs among AIH and other chronic liver diseases (CLD) [13,14,15,16,17,18,19,20].
The main result of our study is that the diagnostic performance of SD/PC, FIB-4, RDW/PC and APRI can be considered acceptable (AUC greater than 0.7), but only SD/PC and FIB-4 in the non-acute subset have accuracies that could be of value in clinical practice (AUC greater than 0.80) [9].
In more detail, the best performance was seen for SD/PC, which showed a good correlation with different fibrosis stages and a good accuracy (AUC = 0.814) that is higher among non-acute AIH, reaching an AUC of 0.902. For FIB-4 we found an AUC of 0.770, which is higher in non-acute patients (AUC = 0.834). For RDW/PC we obtained an AUC of 0.768, which was not affected by acute onset, while for APRI we found an AUC of 0.708, and a mildly better performance among non-acute patients (AUC = 0.758). For AAR we identified an AUC of 0.694, which is too low for the application in clinical practice. As shown by our results, the main application of these tests is mainly in the exclusion of AF, since at the best cut-off calculated they all show great sensitivity and NPV. Overall, specificity and PPV were quite low for all the NITs at the optimal cut-off, which means that up to 60% of patients will be misdiagnosed as having an advanced form of liver disease.
Interestingly, our study is the first that evaluate the performance of NITs among different presentation of AIH. Regarding the identification of acute-onset AIH, there is no globally accepted definition nor unique serological parameters that could help in the differentiation. In our study, we apply the “acute” definition provided by a multicenter Italian study, in which our cohort was partially included [2]. Although increasingly recognized, in the literature there are no exclusion criteria for acute forms of AIH, and treatment status is not often clear, two biases that inevitably affect the results. Our study suggests that at least for SD/PC, FIB-4 and APRI, application in a well-stratified cohort may show better performance. Performance presented in the literature for AIH is lower that what we found among non-acute patients (Table 4). Probably, large metanalyses fail to demonstrate a benefit in using these tests given the great heterogeneity of AIH manifestations. Our results cannot be ascribed to a greater prevalence of AF in the non-acute patients, since Ishak stage ≥ 4 was comparable between the two subgroups (p = 0.886) (Supplementary material, Table S1). The obvious explanation of the different performances is that tests directly proportional to AST level (such as FIB-4 and APRI) will fail in acute patients. Speculating on the performance of SD/PC, we found that SD, PC and their ratio are not different between acute and non-acute patients; this means that in some acute patients, reactive splenomegaly unrelated to portal hypertension can lead the test to underperform.
Comparing our results to the literature (Table 4), we can see that performance of NITs for the diagnosis of AF in AIH (especially if applied in selected patients) seems in line to what has been published in literature for other CLDs where these tests are routinely used. Moreover, we must acknowledge that studies among chronic viral hepatitis have a higher prevalence of advanced fibrosis (as high as 33.5% for CHB and 44% for CHC) that affects diagnostic accuracy rising pre-test probability [9,10,11]. In contrast, in our study the prevalence of AF was 22.1%, which is more similar to that reported for metabolic forms (28%) [11]. Finally, cut-offs identified by our study are comparable to those reported in literature, except for APRI, whose reported cut-off (APRI = 0.5) is not suitable for our cohort, since the prevalence of acute forms of AIH leads to higher AST levels and an inevitably higher ratio.
There is a broad consensus that in AIH the most adequate method for staging hepatic fibrosis is hepatic elastography by VCTE or 2DSWE, which in all studies proved to be superior to blood-based NITs. As in other CLDs, this method offers the chance to obtain a real-time estimate of the degree of hepatic fibrosis, and, from the available data, it does not seem significantly affected by the activity of the disease [5,6,21]. Nevertheless, blood-based NITs do not incur additional expense, involve routinely available data, and are highly reproducible, so they could be widely implemented with relative ease. Given these specifications, they could represent an adjunctive tool in the hand of the physician, aware of their limitations and of their field of application.
Limitations of this study are that it is monocentric, retrospective and that performance is evaluated at the time of diagnosis, when biopsy is necessarily obtained. Regarding RDW/PC, it may be useful to know folate, vitamin B12 and iron levels to contextualize results and to minimize false positives, but we can indirectly state that none of the 122 patients included were under supplementation at the time of diagnosis. Future study should apply these NITs in a longitudinal manner to evaluate reliability for the diagnosis of fibrosis stage during follow-up.
5. Conclusions
Our study shows that, in a wide cohort of AIH patients, blood-based NITs have an overall acceptable performance. Higher accuracy was seen for FIB-4, SD/PC and RDW/PC, while for SD/PC, FIB-4 and APRI diagnostic performance is higher when applied among patients with non-acute presentation. In this context, SD/PC and FIB-4 showed an overall performance that is consistent (AUC > 0.8) and could be of interest in clinical practice alongside other non-invasive techniques. Comparing these results with available literature, their diagnostic accuracy is comparable to what was reported about other forms of CLDs. Cut-offs identified in the literature are suitable among AIH patients except for APRI, and generally they lead to high NPV that permits excluding advanced fibrosis. Future studies should take into account the great heterogeneity of AIH presentations and should exclude patients with an acute hepatitis. Longitudinal application of these scores might be helpful to non-invasively monitor progression of liver fibrosis alongside the complete biochemical response to exclude the evolution of the disease despite treatment.
Supplementary Materials
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/gastroent13030029/s1, Table S1: Comparison of collected data and derived non-invasive tests of fibrosis among acute and non-acute AIH patients.
Author Contributions
Conceptualization, M.F.; Data curation, M.F. and P.M.; Formal analysis, L.M. and P.M.; Investigation, M.F.; Methodology, M.L. and L.M.; Supervision, M.L., L.M. and P.M.; Validation, M.L. and L.M.; Writing—original draft, M.F. All authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy.
Conflicts of Interest
The authors declare no conflict of interest.
References
- European Association for the Study of the Liver. EASL Clinical Practice Guidelines: Autoimmune hepatitis. J. Hepatol. 2015, 63, 971–1004.
- Muratori, P.; Carbone, M.; Stangos, G.; Perini, L.; Lalanne, C.; Ronca, V.; Cazzagon, N.; Bianchi, G.; Lenzi, M.; Floreani, A.; et al. Clinical and prognostic implications of acute onset of Autoimmune Hepatitis: An Italian multicentre study. Dig. Liver Dis. 2018, 50, 698–702. [Google Scholar] [CrossRef] [PubMed]
- West, J.; Card, T.R. Reduced mortality rates following elective percutaneous liver biopsies. Gastroenterology 2010, 139, 1230–1237. [Google Scholar] [CrossRef] [PubMed]
- Cai, Y.; Liu, D.; Cui, J.; Sha, Y.; Zhou, H.; Tang, N.; Wang, N.; Huang, A.; Xia, J. Diagnostic accuracy of red blood cell distribution width to platelet ratio for predicting staging liver fibrosis in chronic liver disease patients: A systematic review and meta-analysis. Medicine 2019, 98, e15096. [Google Scholar] [CrossRef] [PubMed]
- Wu, S.; Yang, Z.; Zhou, J.; Zeng, N.; He, Z.; Zhan, S.; Jia, J.; You, H. Systematic review: Diagnostic accuracy of non-invasive tests for staging liver fibrosis in autoimmune hepatitis. Hepatol. Int. 2019, 13, 91–101. [Google Scholar] [CrossRef] [PubMed]
- Xu, Q.; Sheng, L.; Bao, H.; Chen, X.; Guo, C.; Li, H.; Ma, X.; Qiu, D.; Hua, J. Evaluation of transient elastography in assessing liver fibrosis in patients with autoimmune hepatitis. J. Gastroenterol. Hepatol. 2017, 32, 639–644. [Google Scholar] [CrossRef] [PubMed]
- Yada, N.; Sakurai, T.; Minami, T.; Arizumi, T.; Takita, M.; Hagiwara, S.; Ida, H.; Ueshima, K.; Nishida, N.; Kudo, M. Influence of Liver Inflammation on Liver Stiffness Measurement in Patients with Autoimmune Hepatitis Evaluation by Combinational Elastography. Oncology 2017, 92 (Suppl. S1), 10–15. [Google Scholar] [CrossRef] [PubMed]
- Swets, J.A. Measuring the accuracy of diagnostic systems. Science 1988, 240, 1285–1293. [Google Scholar] [CrossRef] [PubMed]
- European Association for the Study of the Liver; List of panel members (alphabetical order); Berzigotti, A.; Boursier, J.; Castera, L.; Cazzagon, N.; Friedrich-Rust, M.; Petta, S.; Thiele, M.; Tsochatzis, E. Easl Clinical Practice Guidelines (Cpgs) On Non-Invasive Tests For Evaluation Of Liver Disease Severity And Prognosis-2020 Update. J. Hepatol. 2021; 3, 659–689. [Google Scholar]
- Montano-Loza, A.J.; Thandassery, R.B.; Czaja, A.J. Targeting Hepatic Fibrosis in Autoimmune Hepatitis. Dig. Dis. Sci. 2016, 61, 3118–3139. [Google Scholar] [PubMed]
- Hartl, J.; Ehlken, H.; Sebode, M.; Peiseler, M.; Krech, T.; Zenouzi, R.; von Felden, J.; Weiler-Normann, C.; Schramm, C.; Lohse, A.W. Usefulness of biochemical remission and transient elastography in monitoring disease course in autoimmune hepatitis. J. Hepatol. 2018, 68, 754–763. [Google Scholar] [CrossRef] [PubMed]
- Hartl, J.; Denzer, U.; Ehlken, H.; Zenouzi, R.; Peiseler, M.; Sebode, M.; Hübener, S.; Pannicke, N.; Weiler-Normann, C.; Quaas, A.; et al. Transient elastography in autoimmune hepatitis: Timing determines the impact of inflammation and fibrosis. J. Hepatol. 2016, 65, 769–775. [Google Scholar] [CrossRef] [PubMed]
- Xiao, G.; Yang, J.; Yan, L. Comparison of diagnostic accuracy of aspartate aminotransferase to platelet ratio index and fibrosis-4 index for detecting liver fibrosis in adult patients with chronic hepatitis B virus infection: A systemic review and meta-analysis. Hepatology 2015, 61, 292–302. [Google Scholar] [CrossRef] [PubMed]
- Vilar-Gomez, E.; Chalasani, N. Non-invasive assessment of non-alcoholic fatty liver disease: Clinical prediction rules and blood-based biomarkers. J. Hepatol. 2018, 68, 305–315. [Google Scholar] [CrossRef] [PubMed]
- Siddiqui, M.S.; Yamada, G.; Vuppalanchi, R.; Van Natta, M.; Loomba, R.; Guy, C.; Brandman, D.; Tonascia, J.; Chalasani, N.; Neuschwander-Tetri, B.; et al. NASH Clinical Research Network. Diagnostic Accuracy of Noninvasive Fibrosis Models to Detect Change in Fibrosis Stage. Clin. Gastroenterol. Hepatol. 2019, 17, 1877–1885.e5. [Google Scholar] [CrossRef] [PubMed]
- de Oliveira, A.C.; El-Bacha, I.; Vianna, M.V.; Parise, E.R. Utility and limitations of APRI and FIB4 to predict staging in a cohort of nonselected outpatients with hepatitis C. Ann. Hepatol. 2016, 15, 326–332. [Google Scholar] [CrossRef] [PubMed]
- Guo, L.; Zheng, L.; Hu, L.; Zhou, H.; Yu, L.; Liang, W. Transient Elastography (FibroScan) Performs Better Than Non-Invasive Markers in Assessing Liver Fibrosis and Cirrhosis in Autoimmune Hepatitis Patients. Med. Sci. Monit. 2017, 23, 5106–5112. [Google Scholar] [CrossRef] [PubMed]
- Sheptulina, A.; Shirokova, E.; Nekrasova, T.; Blum, H.; Ivashkin, V. Platelet count to spleen diameter ratio non-invasively identifies severe fibrosis and cirrhosis in patients with autoimmune hepatitis. J. Gastroenterol. Hepatol. 2016, 31, 1956–1962. [Google Scholar] [CrossRef] [PubMed]
- Liu, L.; Cao, J.; Zhong, Z.; Guo, Z.; Jiang, Y.; Bai, Y.; Xu, J. Noninvasive indicators predict advanced liver fibrosis in autoimmune hepatitis patients. J. Clin. Lab. Anal. 2019, 33, e22922. [Google Scholar] [CrossRef] [PubMed]
- Kruger, F.C.; Daniels, C.R.; Kidd, M.; Swart, G.; Brundyn, K.; van Rensburg, C.; Kotze, M. APRI: A simple bedside marker for advanced fibrosis that can avoid liver biopsy in patients with NAFLD/NASH. South Afr. Med. J. 2011, 101, 477–480. [Google Scholar]
- Boursier, J.; de Ledinghen, V.; Zarski, J.P.; Rousselet, M.C.; Sturm, N.; Foucher, J.; Leroy, V.; Fouchard-Hubert, I.; Bertrais, S.; Gallois, Y.; et al. A new combination of blood test and fibroscan for accurate non-invasive diagnosis of liver fibrosis stages in chronic hepatitis C. Am. J. Gastroenterol. 2011, 106, 1255–1263. [Google Scholar] [CrossRef] [PubMed] [Green Version]
Figure 1.
Correlation between non-invasive tests and fibrosis stages. Plot of (A) AAR, (B) APRI, (C) FIB-4, (D) RDW/PC, (E) SD/PC score values for fibrosis stages. The point through the middle of each line represents the median. The length of the line represents the interquartile range.
Figure 1.
Correlation between non-invasive tests and fibrosis stages. Plot of (A) AAR, (B) APRI, (C) FIB-4, (D) RDW/PC, (E) SD/PC score values for fibrosis stages. The point through the middle of each line represents the median. The length of the line represents the interquartile range.
Figure 2.
Receiver operator characteristics curves of non-invasive tests for the diagnosis of advanced fibrosis. AUC for (A) AAR: entire population = 0.694, acute patients = 0.674, non-acute patients = 0.697, (B) APRI: entire population = 0.708, acute patients = 0.716, non-acute patients = 0.758, (C) FIB-4: entire population = 0.770, acute patients = 0.724, non-acute patients = 0.834, (D) RDW/PC: entire population = 0.768, acute patients = 0.773, non-acute patients = 0.733, (E) SD/PC: entire population = 0.814, acute patients = 0.754, non-acute patients = 0.902. Black line = entire population, red line = acute patients, green line = non-acute patients.
Figure 2.
Receiver operator characteristics curves of non-invasive tests for the diagnosis of advanced fibrosis. AUC for (A) AAR: entire population = 0.694, acute patients = 0.674, non-acute patients = 0.697, (B) APRI: entire population = 0.708, acute patients = 0.716, non-acute patients = 0.758, (C) FIB-4: entire population = 0.770, acute patients = 0.724, non-acute patients = 0.834, (D) RDW/PC: entire population = 0.768, acute patients = 0.773, non-acute patients = 0.733, (E) SD/PC: entire population = 0.814, acute patients = 0.754, non-acute patients = 0.902. Black line = entire population, red line = acute patients, green line = non-acute patients.
Table 1.
Comparison of collected data and derived non-invasive tests of fibrosis among Ishak 0–3 and Ishak ≥ 4 patients.
Table 1.
Comparison of collected data and derived non-invasive tests of fibrosis among Ishak 0–3 and Ishak ≥ 4 patients.
| Missing | Ishak 0–3 | Ishak ≥ 4 | p-Value | |
|---|---|---|---|---|
| (n = 95) | (n = 27) | |||
| Age | 0 | 47.0 (31.0–57.5) | 57.0 (42.0–60.5) | 0.099 |
| WBC | 16 | 5.90 (5.05–7.60) | 5.46 (4.37–6.78) | 0.127 |
| Neutrophil | 20 | 3.14 (2.59–4.05) | 2.63 (2.05–3.82) | 0.095 |
| Lymphocyte | 20 | 1.95 (1.50–2.54) | 2.05 (1.49–2.39) | 0.684 |
| RBC | 18 | 4.60 (4.23–4.87) | 4.50 (4.24–4.87) | 0.846 |
| Hb | 16 | 13.40 (12.60–14.50) | 12.90 (12.10–14.20) | 0.279 |
| MCV | 21 | 89.00 (84.78–91.83) | 88.10 (85.80–91.00) | 0.962 |
| PC | 17 | 245.50 (187.00–290.00) | 177.00 (130.00–222.00) | <0.001 |
| RDW | 40 | 13.90 (13.40–15.00) | 14.10 (13.90–16.75) | 0.027 |
| MPV | 46 | 9.20 (8.20–11.00) | 9.60 (8.85–10.70) | 0.579 |
| AST | 2 | 254.00 (95.00–739.00) | 488.00 (279.00–982.50) | 0.044 |
| AST/ULN | 2 | 7.00 (2.70–20.10) | 13.00 (7.50–23.70) | 0.077 |
| ALT | 2 | 434.00 (124.00–962.00) | 650.00 (317.50–986.00) | 0.341 |
| ALT/ULN | 2 | 11.22 (3.00–27.00) | 14.00 (8.15–23.90) | 0.704 |
| GGT | 9 | 79.50 (29.75–137.00) | 151.00 (68.00–190.00) | 0.014 |
| ALP | 14 | 132.50 (81.25–219.25) | 172.00 (127.25–238.25) | 0.165 |
| Albumin | 14 | 40.00 (36.85–42.00) | 36.00 (34.53–39.00) | 0.003 |
| SD | 21 | 11.00 (10.00–12.50) | 13.80 (12.70–14.53) | <0.001 |
| FIB–4 | 17 | 2.21 (1.21–4.53) | 6.17 (3.32–11.03) | <0.001 |
| APRI | 17 | 0.04 (0.01–0.11) | 0.11 (0.05–0.12) | 0.002 |
| AAR | 17 | 0.70 (0.59–0.89) | 0.97 (0.71–1.24) | 0.002 |
| RDW/PC | 20 | 0.06 (0.05–0.08) | 0.09 (0.06–0.13) | <0.001 |
| SD/PC | 29 | 0.04 (0.04–0.06) | 0.08 (0.06–0.10) | <0.001 |
AAR = AST to ALT ratio; ALP = alkaline phosphatase; ALT = alanine aminotransferase; APRI = AST to platelet count ratio; AST = aspartate aminotransferase; FIB-4 = fibrosis 4 score; GGT = gamma-glutamyl transpeptidase; Hb = hemoglobin; MCV = mean corpuscular volume; MPV = mean platelet volume; PC = platelets count; RBC = red blood cells; RDW = red blood cell volume distribution width; RDW/PC = red blood cell volume distribution width to platelet count ratio; SD = spleen diameter (SD); SD/PC = spleen diameter to platelet count ratio; ULN = upper limit of norm; WBC = white blood cells.
Table 2.
Areas under the receiver operating characteristic curves of non-invasive tests for the prediction of advanced fibrosis among all patients, acute and non-acute AIH.
Table 2.
Areas under the receiver operating characteristic curves of non-invasive tests for the prediction of advanced fibrosis among all patients, acute and non-acute AIH.
| Test | Population | AUC | 95% CI | Z | p-Value |
|---|---|---|---|---|---|
| FIB-4 | All patients (n = 105) | 0.770 | (0.670–0.870) | 5298.42 | <0.001 |
| Acute (n = 56) | 0.724 | (0.567–0.881) | 2792.685 | 0.005 | |
| Non-acute (n = 49) | 0.834 | (0.679–0.990) | 4206.383 | <0.001 | |
| APRI | All patients (n = 105) | 0.708 | (0.592–0.823) | 3.522 | <0.001 |
| Acute (n = 56) | 0.716 | (0.553–0.879) | 2.605 | 0.009 | |
| Non-acute (n = 49) | 0.758 | (0.579–0.938) | 2.816 | 0.005 | |
| AAR | All patients (n = 120) | 0.694 | (0.568–0.819) | 3022.145 | 0.003 |
| Acute (n = 62) | 0.674 | (0.500–0.848) | 1958.879 | 0.050 | |
| Non-acute (n = 58) | 0.697 | (0.500–0.894) | 1969.192 | 0.049 | |
| RDW/PC | All patients (n = 82) | 0.768 | (0.640–0.896) | 4114.128 | <0.001 |
| Acute (n = 41) | 0.773 | (0.614–0.933) | 3362.831 | 0.001 | |
| Non-acute (n = 41) | 0.733 | (0.506–0.959) | 201.469 | 0.044 | |
| SD/PC | All patients (n = 93) | 0.814 | (0.719–0.909) | 6478.971 | <0.001 |
| Acute (n = 49) | 0.754 | (0.606–0.902) | 3371.503 | 0.001 | |
| Non-acute (n = 44) | 0.902 | (0.814–0.990) | 8988.844 | <0.001 |
AAR = AST to ALT ratio; APRI = AST to platelet count ratio; AUC = area under the receiver operating characteristic curve; CI = confidence intervals; FIB-4 = fibrosis 4 score; RDW/PC = red blood cell volume distribution width to platelet count ratio; SD/PC = spleen diameter to platelet count ratio.
Table 3.
Cut-off and diagnostic performance of non-invasive tests for determination of advanced fibrosis among all patients, acute and non-acute.
Table 3.
Cut-off and diagnostic performance of non-invasive tests for determination of advanced fibrosis among all patients, acute and non-acute.
| Test | Population | Cut-Off | Sensitivity | Specificity | PPV | NPV | LR pos | LR neg |
|---|---|---|---|---|---|---|---|---|
| FIB-4 | All patients | 2.000 | 0.960 | 0.487 | 0.369 | 0.975 | 1.873 | 0.082 |
| Non-acute | 3.300 | 0.818 | 0.872 | 0.643 | 0.944 | 6.382 | 0.209 | |
| Acute | 6.200 | 0.643 | 0.756 | 0.474 | 0.861 | 2.636 | 0.472 | |
| APRI | All patients | 6.044 | 0.708 | 0.688 | 0.405 | 0.887 | 2.267 | 0.424 |
| Non-acute | 3.422 | 0.636 | 0.842 | 0.538 | 0.889 | 4.03 | 0.432 | |
| Acute | 7.232 | 0.846 | 0.548 | 0.367 | 0.92 | 1.87 | 0.281 | |
| AAR | All patients | 0.930 | 0.593 | 0.787 | 0.444 | 0.871 | 2.785 | 0.518 |
| Non-acute | 0.934 | 0.583 | 0.848 | 0.5 | 0.886 | 3.833 | 0.491 | |
| Acute | 0.825 | 0.733 | 0.66 | 0.407 | 0.886 | 2.154 | 0.404 | |
| RDW/PC | All patients | 0.087 | 0.632 | 0.859 | 0.571 | 0.887 | 4.491 | 0.429 |
| Non-acute | 0.061 | 0.889 | 0.594 | 0.381 | 0.95 | 2.188 | 0.187 | |
| Acute | 0.087 | 0.700 | 0.800 | 0.538 | 0.889 | 3.5 | 0.375 | |
| SD/PC | All patients | 0.054 | 0.880 | 0.691 | 0.512 | 0.94 | 2.85 | 0.174 |
| Non-acute | 0.062 | 0.909 | 0.853 | 0.667 | 0.967 | 6.182 | 0.107 | |
| Acute | 0.045 | 0.929 | 0.588 | 0.481 | 0.952 | 2.255 | 0.121 |
AAR = AST to ALT ratio; APRI = AST to platelet count ratio; FIB-4 = fibrosis 4 score; LR neg = likelihood ratio negative; LR pos = likelihood ratio positive; NPV = negative predictive values; PPV = positive predictive values; RDW/PC = red blood cell volume distribution width to platelet count ratio; SD/PC = spleen diameter to platelet count ratio.
Table 4.
Literature review of non-invasive tests’ performance among different chronic liver diseases and AIH.
Table 4.
Literature review of non-invasive tests’ performance among different chronic liver diseases and AIH.
| NIT | Article Type | CLD | N | AUC | Cut-Off | Diagnostic Accuracy |
|---|---|---|---|---|---|---|
| FIB-4 [13] | Metanalysis | CHB | 3139 | 0.80 (0.74–0.91) | 0.5 | SE 73%, SP 55% |
| FIB-4 [14] | Review | NAFLD | 686 | 0.80–0.86 | 1.3 | NPV 90–95% |
| 3.25 | PPV 75% | |||||
| FIB-4 [15] | Retrospective multicenter | NAFLD | 1904 | NR | 1.37 | SE 75%, SP 71% |
| FIB-4 [16] | Retrospective monocentric | CHC | 798 | 0.84 | NR | NR |
| FIB-4 [17] | Retrospective monocentric | AIH | 108 | 0.64 (0.53–0.74) | 3.21 | SE 24%, SP 78% |
| FIB-4 [18] | Retrospective monocentric | AIH | 76 | 0.74 (0.62–0.86) | 2.37 | SE 74%, SP 71% |
| FIB-4 [5] | Metanalysis | AIH | 421 | 0.76 (0.72–0.79) | NR | SE 60%, SP 76% |
| FIB-4 [19] | Retrospective monocentric | AIH | 45 | 0.76 (0.61–0.90) | 2.26 | SE 77%, SP 74% |
| RDW/PC [4] | Metanalysis | Various | 1489 | 0.83 (±0.03) | 0.07 | SE 78%, SP 70% |
| RDW/PC [19] | Retrospective monocentric | AIH | 45 | 0.79 (0.65–0.92) | 0.24 | SE 86%, SP 61% |
| AAR [15] | Retrospective multicentre | NAFLD | 1904 | 0.68 (0.66–0.71) | 0.85 | SE 54%, SP 73% |
| AAR [19] | Retrospective monocentric | AIH | 45 | 0.71 (0.56–0.86) | 0.77 | SE 82%, SP 57% |
| AAR [18] | Retrospective monocentric | AIH | 76 | 0.73 (0.61–0.85) | 1.0 | SE 59%, SP 71% |
| AAR [14] | Review | NAFLD | 174 | 0.83–0.90 | 0.8 | NPV 93%, PPV 44% |
| AAR [5] | Metanalysis | AIH | 252 | 0.73 (0.69–0.76) | NR | SE 84%, SP 64% |
| APRI [16] | Retrospective monocentric | CHC | 798 | 0.82 | NR | NR |
| APRI [13] | Metanalysis | CHB | 3139 | 0.76 (0.68–0.87) | NR | NR |
| APRI [20] | Retrospective monocentric | NAFLD | 111 | 0.85 | 0.98 | SE 75%, SP 86% |
| APRI [14] | Review | NAFLD | 175 | 0.56–0.67 | 1 | NPV 84%, PPV 37% |
| APRI [15] | Retrospective multicentre | NAFLD | 1904 | NR | 0.84 | SE 75%, SP 65% |
| APRI [5] | Metanalysis | AIH | 506 | 0.74 (0.70–0.78) | NR | SE 72%, SP 64% |
| APRI [17] | Retrospective monocentric | AIH | 108 | 0.65 (0.54–0.75) | 2.13 | SE 43%, SP 89% |
| APRI [18] | Retrospective monocentric | AIH | 76 | 0.71 (0.58–0.83) | 0.84 | SE 83%, SP 63% |
| SD/PC [18] | Retrospective monocentric | AIH | 76 | 0.88 (0.80–0.97) | 0.001 | SE 88%, SP 85% |
AAR = AST to ALT ratio; AIH = autoimmune hepatitis; APRI = AST to platelet count ratio; AUC = area under the receiver operating characteristic curve; CHB = chronic hepatitis B; CHC = chronic hepatitis C; CLD = chronic liver disease; FIB-4 = fibrosis 4 score; NAFLD = non-alcoholic fatty liver disease; NIT = non-invasive tests; NPV = negative predictive values; NR = not reported; PPV = positive predictive values; RDW/PC = red blood cell volume distribution width to platelet count ratio; SE = sensitivity; SD/PC = spleen diameter to platelet count ratio; SP = specificity.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
MDPI and ACS Style
Ferronato, M.; Lenzi, M.; Muratori, P.; Muratori, L. Blood-Based Non-Invasive Tests of Hepatic Fibrosis in Autoimmune Hepatitis: Application among Selected Patients Leads to Higher Accuracy. Gastroenterol. Insights 2022, 13, 286-295. https://doi.org/10.3390/gastroent13030029
AMA Style
Ferronato M, Lenzi M, Muratori P, Muratori L. Blood-Based Non-Invasive Tests of Hepatic Fibrosis in Autoimmune Hepatitis: Application among Selected Patients Leads to Higher Accuracy. Gastroenterology Insights. 2022; 13(3):286-295. https://doi.org/10.3390/gastroent13030029
Chicago/Turabian StyleFerronato, Marco, Marco Lenzi, Paolo Muratori, and Luigi Muratori. 2022. "Blood-Based Non-Invasive Tests of Hepatic Fibrosis in Autoimmune Hepatitis: Application among Selected Patients Leads to Higher Accuracy" Gastroenterology Insights 13, no. 3: 286-295. https://doi.org/10.3390/gastroent13030029
