Balance of benefits and harms
The Guidelines Development Group recommended the use of NITs to assist in the assessment of stage of liver disease and diagnosis of cirrhosis, to help prioritize those at greatest risk of morbidity and mortality for antiviral therapy. This avoids the use of liver biopsy, which is an expensive and invasive procedure associated with patient discomfort, carries a small risk of serious bleeding and requires specialist histological interpretation for accurate staging. Based on evidence from the systematic review, the Guidelines Development Group considered that transient elastography (FibroScan) (where resources permit) and APRI were the most useful tests for the assessment of cirrhosis in LMICs. However, the recommendation was conditional because the PPV for detection of cirrhosis was low for all NITs, and in particular for APRI (detecting only one third of persons with cirrhosis), and there has been very limited evaluation of their use in sub-Saharan Africa. FIB-4 was not considered or recommended because it has been developed and validated for the detection of fibrosis stages ≥F3 and not cirrhosis. FibroTest is a commercial assay and less accurate than transient elastography (FibroScan) for diagnosing cirrhosis. Standard ultrasound was also not considered as it only detects advanced cirrhosis, and therefore its use would result in an unacceptably high number of false-negative results.
Potential harms from the use of NITs include treatment decisions based on either false-positive or false-negative APRI test results. A false-positive test result may lead to a patient being treated unnecessarily or prematurely, which would expose them to the inconvenience of long-term treatment, potential drug resistance as well as a small risk of drug toxicities. Conversely, a false-negative result means that a person with cirrhosis would not be identified by NITs, and may therefore not receive prompt antiviral therapy, which might prevent progression to decompensation or decrease the risk of developing HCC.
APRI is based on two indirect markers of fibrosis (AST and platelet count), which are readily available in resource-limited settings. An approach that combined a high and a low cut-off value of APRI would be optimal (a high cut-off with high specificity [i.e. fewer false-positive results] and a low cut-off with high sensitivity [i.e. fewer false-negative results]). However, the Guidelines Development Group recommended the use of a single high cut-off >2 for identifying adults with cirrhosis (F4) and in need of antiviral therapy, and those ≤2 without cirrhosis for several reasons.
Although in adults an APRI score of >2 would detect only one third of persons with cirrhosis, this high cut-off of >2 was used, because the low cut-off would result in an unacceptibly high number of false-positive test results (approximately one quarter of those tested).
It is also likely that adults with cirrhosis not detected using an APRI score >2 would be identified as being in need of antiviral therapy because of other eligibility criteria (such as persistently abnormal ALT levels
a as well as evidence of ongoing HBV replication (HBV DNA >20 000 IU/mL) (
see also
Chapter 5: Who to treat and not to treat).
It is also simpler and more feasible to use a single cut-off in resource-limited settings.
Clinical evidence of cirrhosis or an APRI score >2 are recommended in these guidelines as key criteria for prioritizing initiation of antiviral therapy among adults in resource-limited settings. Conversely, treatment can be deferred in those without clinical features of cirrhosis (or based on APRI score ≤2), who also have persistently normal ALT concentrations and low levels of HBV replication (HBV DNA <2000 IU/mL), and who can be re-evaluated at subsequent visits. For those with an APRI score ≤2, a proportion will fulfil other criteria for treatment such as persistently abnormal ALT or raised HBV DNA levels. Adults with indeterminate APRI scores (i.e. between 1 and 2 based on the combined APRI cut-off) in particular would need retesting and evaluation every one or two years.
Caveats in the use of NITs: Overall, the Guidelines Development Group considered that the benefits of using NITs outweighed these potential harms. The benefits included the potential increase in treatment availability resulting from access to non-invasive monitoring, and reduced risk of adverse events from liver biopsy.
However, a number of very important caveats were noted in the use of NITs. Overall, the PPV of all NITs for the diagnosis of cirrhosis was low, especially for APRI, and many cases of cirrhosis will be missed using NITs alone. It is therefore important that NITs are used alongside clinical criteria and other laboratory criteria (ALT and HBV DNA levels) to identify those in need of treatment. Second, the results of NITs may be impacted by intercurrent diseases that may falsely increase or decrease the scores. For example, heavy alcohol intake (due to AST elevation from alcoholic hepatitis), and conditions such as malaria or HIV (due to a decrease in platelet count), or use of drugs and traditional herbal medicines may also cause falsely high APRI scores. Hepatitis flares or acute hepatitis, congestive heart failure or a recent meal may also cause a high liver stiffness measurement on elastography (81). The impact of different comorbidities on the diagnostic accuracy of the APRI score has not been fully evaluated and, in particular, there has been no evaluation of NITs, particularly APRI in people from sub-Saharan Africa or among children.
Limitations with transient elastography (FibroScan) include the following: it uses a single cut-off and therefore reported sensitivities and specificities of FibroScan may be overestimated across fibrosis stages; there are no uniformly established and validated cut-offs for specific stages of fibrosis; accuracy is diminished in obese persons, in the presence of moderate/severe necroinflammation, right-sided heart failure, and food intake. Examination is not feasible in the presence of ascites and is contraindicated in pregnant women. Data on its use in children are limited, and special probes are required.
Although the data on HBV/HIV coinfection were limited, the performance of NITs in such persons is unlikely to be significantly different from that in HBV-monoinfected persons.
Values and preferences
From a patient’s perspective, the Guidelines Development Group felt that the APRI test would be acceptable, as it requires only phlebotomy, is routinely available and can be undertaken by untrained staff. Similarly, transient elastography (FibroScan) is non-invasive, takes less than 10 minutes to perform, can be undertaken in outpatient or community settings, and health-care staff can be easily trained in its use. Factors that limit the use of transient elastography include the high cost of the equipment, the need for preventive and corrective maintenance, regular service/recalibration, trained operators, and the lack of extensively validated cut-off values for specific stages of fibrosis.
Resource use
The lower cost of the blood-based NITs compared to transient elastography was a key factor in the recommendation for the use of APRI as the preferred NIT. The blood tests that are needed to calculate APRI score are routinely available at most health-care facilities, even in LMICs, and are inexpensive (less than a few dollars each). The results of APRI using a high cut-off of >2 in adults to diagnose cirrhosis are also relatively straightforward to interpret. Cost considerations were a concern with the FibroTest. It is a patented test that is costly (approximately US$ 73/test) and requires a certified laboratory or the processing of specimens at a centralized laboratory in France.
In contrast to APRI, the cost of acquiring, running and maintaining (requires regular service/recalibration) a transient elastography machine such as the FibroScan is high; the machine costs US$ 50 000 (or US$ 34 000 for the portable machine), and yearly maintenance is US$ 8500/year. However, consumable costs are minimal for FibroScan, and the cost per test could be less than US$ 10 in some settings. FibroScan also requires a trained operator, and the interpretation of the results needs an understanding of the indications and limitations of the method, especially given the lack of well-validated cut-off values for specific stages of fibrosis. However, the training process is relatively straightforward and the inter- and intra-observer variability of the test is low (81). FibroScan in children requires a specially designed probe and a different specific probe for those with a body mass index (BMI) >30 kg/m2. For these reasons, the use of transient elastography and FibroTest was considered to be less feasible in most LMICs.
