An evaluation of the AST:ALT ratio in identifying patients in primary care for appropriate referral to gastroenterology

The need to differentiate patients with advanced liver disease from those with earlier stage, or more benign diseases for optimal management and allocation of resources is an ever present challenge. In this article we discuss our experiences of using the aspartate transaminase (AST) : alanine transaminase (ALT) ratio as part of a pathway to screen patients for referral to secondary care.

by Dr Raphael Buttigieg and Dr Sara Jenks

Introduction
Deaths from liver disease in Scotland are on the increase [1]. More often than not patients are picked up at a late stage of their disease with significant fibrosis and/or cirrhosis already present. As a result there is a need to try to identify patients with progressive disease earlier on in the course of their illness.

Abnormal liver function tests (LFTs) are frequently picked up on general screening blood samples done in primary care. The degree of abnormality correlates poorly with the extent of liver disease. The gold standard test for liver disease diagnosis and staging is considered to be a liver biopsy; however, there are many other considerations to this invasive procedure including clinical risk, technical ability of person doing the biopsy, inter-pathologist variation in scoring and others. These limitations have led to the development of non-invasive methods for the assessment of liver fibrosis. Although there have been suggestions by different groups regarding the appropriate use of non-invasive fibrosis scoring systems, no one guideline is currently in use.

Non-invasive methods rely on two different approaches [2]:
(a) A biomarker-based approach using serum samples. Advantages are their high applicability (>95%) and good inter-laboratory reproducibility.
(b) A physical approach based on the measurement of liver stiffness. Liver stiffness corresponds to an intrinsic physical property of liver parenchyma. Physical approaches include transient elastography such as FibroScan^® and MR elastography.

Because of noted variations in care, as well as to ensure appropriate referrals, NHS Lothian made a guideline for GPs in 2013. At the time, based on the best available clinical evidence, the aspartate transaminase (AST) : alanine transaminase (ALT) ratio was chosen as a scoring system to guide referrals, which was developed in recognition that as liver fibrosis develops, the normal ratio tends to reverse. An abnormal AST:ALT ratio can, thus, be used to pick up patients who should be referred to secondary care for further investigation, as well as closer monitoring and treatment.

However, other biomarker-based fibrosis risk scores have also been developed [2], which have been used for this purpose including the Fibrosis-4 (FIB-4) [3], NAFLD (non-alcoholic fatty liver disease) fibrosis score, and APRI (AST to platelet ratio index), which may have a better performance than the AST:ALT ratio [4]. Each of these has been validated for different liver diseases – and in many cases different cut-off points are recommended for diagnosis of advancing fibrosis based on the likely primary pathology involved in the individual patient. For example, alcohol use in itself will raise the AST and, thus, the same AST:ALT ratio is likely to indicate more advanced fibrosis in someone with HCV-related liver disease than in alcoholic liver disease with ongoing ethanol excess. This adds to the complexity of using any one score in a guideline to ensure the right balance between sensitivity and specificity.

A final consideration to note is that specifically for NAFLD/non-alcoholic steatohepatitis (NASH), the continued development of pharmaceuticals for the prevention of disease progression means that, once again, the threshold for diagnosis may need to change as therapies to target earlier stages become available [5].

This article will discuss our guideline (Fig. 1) and conclusions drawn from an audit of its use.

Method
A list of all the requests for a AST:ALT ratio in a 6-month period in NHS Lothian was obtained from laboratory records – in terms of date of request, patient name and CHI number (unique patient identifier) (n=874). These records were encoded into a spreadsheet and a plan for analysis made.
Following this, various data were audited retrospectively from the patient electronic record. Individual notes and files were not used because of the logistical difficulty in analysing large numbers of case notes.
Of the total number of ratios (n=874) requested in the 6-month period, 49 were elevated at >1.0 and 295 were normal at ≤1.0; 530 ratio requests were cancelled due to ALT being within the reference range.

Results
The various aspects of the referral process from primary to secondary care were audited with the following aims.

1. To identify all the abnormal ratios in a 6-month period (n=49) (Table 1).

2. To identify all the ratios in the same 6-month period that were in the range 0.8–1.0 (n=53) (Table 2).
This was carried out to assess whether there should be concern about the ratio producing false negative results, and how useful it was to actually exclude liver disease. We thus audited all patients with a borderline ratio of 0.8–1.0.
Additionally, we asked if the FIB-4 score or APRI score was used, would this have affected referral?

3. To identify the first 50 individuals in a 6-month period tested with an ALT level of 40–49 on whom the AST:ALT ratio had been cancelled (n=50) (Table 3).
Although an upper limit of 50 is taken for the normal range of ALT, there is evidence that even at levels below this a certain amount of liver inflammation is present, and, thus, different health boards use other values – such as an upper limit of normal of 40.

This last part of the project set out to identify people with a borderline abnormal ALT of 40–49, and assess whether using different scores – such as the FIB-4 or APRI scores would potentially label these individuals as having liver disease and needing to be referred

Limitations of our study
Most of these patients had a very short-term follow-up, which in many cases did not allow proper determination of their disease severity, as well as assessment of long-term mortality/morbidity risk using different scoring systems.

Secondly, we were unable to compare scores to a gold standard as in many cases a liver biopsy had not been carried out. Transient elastography and hyaluronic acid testing had been carried out in a selection of patients which allowed further characterization of fibrosis staging; however, it is appreciated that neither of these are the gold standard.

Conclusions and considerations
The AST:ALT ratio is a good test for assessing whether people should be referred to secondary care or not. This conclusion is based on the fact that many patients who were referred with a positive ratio were seen in secondary care and kept under review. However, better tests are needed to further assess their stage of disease, ideally non-invasively.

The FIB-4 (possibly in association with further tests below) may be a more sensitive/specific score to be used in diagnosing patients; however, cut-off points would need to be determined to guide the most effective use of available resources in primary and secondary care. As can be seen in the second group FIB-4 and the APRI were raised in patients which would not have been picked up by the AST:ALT ratio, which thus increases pick-up. Another consideration is, as previously mentioned, that the AST:ALT ratio tends to be raised in patients drinking excessive ethanol, even if their disease is not very advanced. Since in our cohort alcohol use was very prevalent, other scores may possibly be better suited.
The plan from now is to adopt a pathway of cascading lab tests based on patients’ alcohol consumption, BMI/metabolic syndrome markers, LFT results and automatic scoring with interpretation will be issued to GPs. Also possible is further testing – either in the community or in secondary care to further guide patients in different scoring groups – including either transient elastography (FibroScan), or further biochemical testing. NHS Lothian currently uses hyaluronic acid, and this may be a way of further classifying/evaluating people in ‘intermediate’ categories. The elastography (FibroScan) test could be another option and this is the current recommendation in the current NICE guidelines.

For any further information please feel free to contact the authors:
Raphael Buttigieg: ST3 Chemical Pathology and Metabolic Medicine, NHS Greater Glasgow and Clyde, UK; raphael.buttigieg@nhs.net.
Sara Jenks: Consultant in Chemical Pathology and Metabolic Medicine, NHS Lothian, Edinburgh, UK; sjenks@nhs.net.

References
1. Gray L, Leyland AH. Alcohol. The Scottish Health Survey 2014: Volume 1: Main report (http://www.gov.scot/Publications/2015/09/6648/318753)
2. European Association for Study of Liver. EASL-ALEH Clinical Practice Guidelines: Non-invasive tests for evaluation of liver disease severity and prognosis. J Hepatol 2015; 63(1): 237.
3. McPherson S, Anstee QM, Henderson E, Day CP, Burt AD. Are simple noninvasive scoring systems for fibrosis reliable in patients with NAFLD and normal ALT levels? Eur J Gastroenterol Hepatol 2013; 25(6): 652–658.
4. Parkes J, Guha IN, Harris S, Rosenberg WM, Roderick PJ. Systematic review of the diagnostic performance of serum markers of liver fibrosis in alcoholic liver disease. Comp Hepatol 2012; 11(1): 5.
5. Dyson JK, Anstee QM, McPherson S. Non-alcoholic fatty liver disease: a practical approach to diagnosis and staging. Frontline gastroenterology 2014; 5(3):211–218.

The authors
Raphael Buttigieg*¹ Sara Jenks^2
1Department of Clinical biochemistry, Glasgow Royal Infirmary, NHS Greater Glasgow and Clyde, UK
²Department of Clinical biochemistry, NHS Lothian, Royal Infirmary of Edinburgh, Edinburgh, UK

*Corresponding author
E-mail: raphael.buttigieg@nhs.net