Keywords
To achieve the World Health Organization (WHO)’s global strategy of eliminating hepatitis B virus (HBV) infection by 2030, it is vital to scale up screening, clinical staging and treatment services in low- and middle-income countries (LMICs) which account for the largest proportion of people living with chronic HBV infection worldwide.
[1]
Although both screening tests for hepatitis B surface antigen (HBsAg) and generic antiviral medicines are now available at affordable prices in LMICs, access to the conventional diagnostic tests to examine treatment eligibility (liver biopsy/FibroScan®/HBV DNA) is still limited and non-affordable.[2]
Consequently, the WHO guidelines provided simplified criteria specific for LMICs where HBV DNA quantification is inaccessible: cirrhosis (clinical diagnosis or aspartate aminotransferase-to-platelet ration index (APRI) >2.0), or persistently elevated alanine aminotransferase (ALT), defined by 3 ALT determinations over a period of 6–12 months.[3]
Recently, using the data from the PROLIFICA (Prevention of Liver Fibrosis and Cancer in Africa) Study in The Gambia, West Africa, we developed TREAT-B (Treatment Eligibility in Africa for the Hepatitis B Virus), a novel simple score free from HBV DNA and based on solely ALT level and hepatitis B e antigen (HBeAg) sero-status. The total TREAT-B score is obtained by adding HBeAg score (negative (0 point) or positive (1 point)), and ALT score (<20 IU/L (0 point), 20–39 (1 point), 40–79 (2 points) or ≥80 (3 points)). TREAT-B ranged from 0 (HBeAg-negative and ALT <20 IU/L) to 4 (HBeAg-positive and ALT ≥80 IU/L). TREAT-B had high diagnostic accuracy for the selection of treatment eligible patients (area under the receiver operating characteristic curve [AUROC] 0.85; sensitivity 84.5%; specificity 77.3%), based on the reference tests used in the European Association for the Study of the Liver (EASL) guidelines, performed in the validation dataset from other African populations.[4]
Since the publication of the original work, further validation of TREAT-B and the simplified WHO criteria was performed, using the EASL guidelines as a reference. In the UK and Germany, among predominantly migrant populations from Africa or Asia, the AUROC of TREAT-B (0.84; 95% CI 0.80–0.88) was significantly higher than the WHO criteria (0.67; 95% CI 0.63–0.70), with sensitivity and specificity of 82.9% and 73.7% for TREAT-B and 92.4% and 41.0% for WHO.
[5]
In Ethiopia, Johannessen and colleagues showed that the AUROC of TREAT-B (0.73; 95% CI 0.68–0.78) was again significantly better than the WHO criteria (0.62; 95% CI 0.56–0.68), with sensitivity and specificity of 53.0% and 83.4% for TREAT-B and 56.7% and 67.4% for the WHO criteria.[6]
Here, we report the performance of TREAT-B and simplified WHO criteria evaluated in an independent dataset from Burkina Faso. Of 1,313 HBsAg-positive patients evaluated at the Yalgado Ouédraogo University Hospital, Ouagadougou, in 2006–2012, there were 255 treatment-naïve patients who had ≥3 ALT measurements, in addition to the comprehensive baseline assessment including HBeAg (Elecsys®, Roche Diagnostics), HBV DNA PCR (Cobas® Taqman) and fibrosis assessment (liver biopsy [n = 78], FibroScan [n = 8], and FibroTest [n = 169]). For FibroScan, significant fibrosis (≥F2) and cirrhosis (F4) were defined as ≥7.9 kPa and ≥9.5 kPa, respectively.
[7]
After excluding patients coinfected with HCV/HDV/HIV, decompensated cirrhosis, or hepatocellular carcinoma, 222 were finally included in the analysis.Table 1 shows their baseline characteristics. Median age was 34 years (interquartile range: 28–40), and 150 (67.6%) were men. HBeAg was positive in 36 (16.2%). Treatment was indicated in 63 (28.4%), 91 (41.0%), and 73 (32.9%) patients, based on the reference tests used in the EASL 2017 criteria,
[8]
TREAT-B, and simplified WHO criteria, respectively. Table 2 presents the performance of TREAT-B and WHO criteria to diagnose treatment eligibility, as determined by the EASL guidelines. The AUROC of TREAT-B (0.76; 95% CI 0.69–0.83) was significantly higher than that of the WHO criteria (0.68; 95% CI 0.61–0.75), using the Wald statistic (p = 0.03).[9]
The sensitivity and specificity were 69.8% and 70.4% for TREAT-B with a cut-off of ≥2, and 58.7% and 77.4% for the simplified WHO criteria.Table 1Characteristics of the study participants (n = 222).
| All (n = 222) | Treatment eligible by EASL (n = 63) | Treatment ineligible by EASL (n = 159) | p value | |
|---|---|---|---|---|
| Median age (IQR), years | 34 (28–40) | 32 (26–40) | 34 (28–40) | 0.8 |
| Male, n (%) | 150 (67.6) | 51 (81.0) | 99 (62.3) | 0.01 |
| Median HBV DNA (IQR), log 10 IU/ml | 3.0 (2.1–4.3) | 5.8 (3.9–7.2) | 2.4 (1.8–3.3) | <0.0001 |
| Positive HBeAg, n (%) | 36 (16.2) | 22 (34.9) | 14 (8.8) | <0.001 |
| Median ALT (IQR), IU/L | ||||
| 1st measurement | 30 (20–53) | 51 (28–77) | 25 (18–42) | <0.0001 |
| 2nd measurement | 29 (18–46) | 43 (25–87) | 26 (16–38) | <0.0001 |
| 3rd measurement | 28 (19–47) | 44 (27–77) | 24 (17–37) | <0.0001 |
| Liver fibrosis, n (%) | ||||
| No or mild (F0–1) | 103 (46.4) | 4 (6.4) | 99 (62.3) | <0.001 |
| Significant (F2) | 82 (36.9) | 32 (50.8) | 50 (31.5) | |
| Severe (F3) | 24 (10.8) | 15 (23.8) | 9 (5.7) | |
| Cirrhosis (F4) | 13 (5.9) | 12 (19.1) | 1 (0.6) | |
| Treatment eligible by TREAT–B (score ≥2), n (%) | 91 (41.0) | 44 (69.8) | 47 (29.6) | <0.001 |
| Treatment eligible by the simplified WHO criteria, n (%) | 73 (32.9) | 37 (58.7) | 36 (22.6) | <0.001 |
ALT, alanine aminotransferase; EASL, the European Association for the Study of the Liver; HBeAg, HBV e antigen; HBV, hepatitis B virus; WHO, World Health Organization.
* p value was obtained using Wilcoxon rank-sum test for continuous variables, and Chi-square test or Fisher’s exact test for categorical variables.
Table 2Performance of TREAT-B and the simplified WHO criteria (n = 222).
| TREAT-B | Simplified WHO criteria | p value | |
|---|---|---|---|
| AUROC (95% CI) | 0.76 (0.69–0.83) | 0.68 (0.61–0.75) | 0.03 |
| Sensitivity (%) | 69.8% | 58.7% | |
| Specificity (%) | 70.4% | 77.4% | |
| Positive predictive value (%) | 48.4% | 50.7% | |
| Negative predictive value (%) | 85.5% | 82.6% |
AUROC, area under the receiver operating characteristic curve; WHO, World Health Organization.
* p value was obtained using the Wald test.
Despite substantial variation in sensitivity and specificity, all the previous external validation studies, including the one reported here, constantly showed that TREAT-B is better than the simplified WHO criteria at discriminating between patients who are eligible or ineligible for antiviral therapy. The sensitivity of TREAT-B observed in Ethiopia (53.0%) and in Burkina Faso (69.8%) may be suboptimal; nevertheless, from a public health perspective, reduced cost and improved operational characteristics can compensate for the low sensitivity of a test, which can lead to better access to diagnostic services. A modeling study of tuberculosis has shown that the diagnostic sensitivity was just one of multiple parameters in resource-limited countries; a point-of-care test with 40% sensitivity had a similar impact on population-level mortality to a molecular assay with 70% sensitivity, because the former improved the linkage to care.
[10]
The advantage of TREAT-B is that both ALT and HBeAg can be potentially measured using a point-of-care test, although the detection limit of HBeAg rapid diagnostic tests needs to be improved.[11]
A health-economic study using modeling is underway by our group to assess the effectiveness and cost-effectiveness of the simplified HBV treatment algorithms with varying diagnostic sensitivity and specificity in resource-limited settings.Different approaches exist to simplify HBV treatment criteria. TREAT-B was derived through a purely statistical method: out of clinical, haematological, biochemical, and serological markers widely available in LMICs, an automated stepwise selection procedure identified a set of variables (ALT and HBeAg) that gave a good prediction of the treatment eligibility. Another approach is to replace each of the conventional reference tests used to define treatment eligibility with an affordable and available alternative. For example, liver fibrosis can be evaluated through APRI or gamma-glutamyltransferase to platelet ratio (GPR).
[7]
Instead of real-time PCR, high HBV viral replication in treatment-naïve patients can be diagnosed using a simpler molecular assay (e.g., loop-mediated isothermal amplification (LAMP))[12]
or inexpensive serological test (e.g., hepatitis B core-related antigen (HBcrAg)).[13]
Further evaluation of simple, affordable, and reliable diagnostic algorithms, including TREAT-B, is crucial to reach the global elimination goal.- Shimakawa Y.
- Ndow G.
- Njie R.
- Njai H.F.
- Takahashi K.
- Akbar S.M.F.
- et al.
Hepatitis B core-related antigen: an alternative to hepatitis B virus DNA to assess treatment eligibility in Africa.
Clin Infect Dis. 2019; https://doi.org/10.1093/cid/ciz412
Financial support
The work was partly supported by the NéoVac (Neonatal Vaccination against Hepatitis B in Africa) project, funded by the Total Foundation.
Conflict of interest
YS has served as consultant of Gilead Sciences. The other authors declare no conflict of interest.
Please refer to the accompanying ICMJE disclosure forms for further details.
Supplementary data
The following are the Supplementary data to this article:
- Supplementary Data 1
- Supplementary Data 2
References
Author names in bold designate shared co-first authorship
WHO. Global hepatitis report, 2017. Geneva, Switzerland: 2017.
- Prevalence of chronic hepatitis B virus infection and infrastructure for its diagnosis in Madagascar: Implication for the WHO’s elimination strategy.BMC Public Health. 2017; 17: 636https://doi.org/10.1186/s12889-017-4630-z
WHO. Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection. Geneva, Switzerland: 2015.
- Development of a simple score based on HBeAg and ALT for selecting patients for HBV treatment in Africa.J Hepatol. 2018; 69: 776-784https://doi.org/10.1016/j.jhep.2018.05.024
- Clinical utility of TREAT-B score in African and non-African HBV-infected patients living in Europe.J Hepatol. 2019; 70: 1295-1297https://doi.org/10.1016/j.jhep.2019.03.008
- A novel score to select patients for treatment in chronic hepatitis B: results from a large Ethiopian cohort.J Hepatol. 2019; 71: 840-841
- The gamma-glutamyl transpeptidase to platelet ratio (GPR) predicts significant liver fibrosis and cirrhosis in patients with chronic HBV infection in West Africa.Gut. 2016; 65: 1369-1376https://doi.org/10.1136/gutjnl-2015-309260
- EASL 2017 Clinical Practice Guidelines on the management of hepatitis B virus infection.J Hepatol. 2017; 67: 370-398https://doi.org/10.1016/j.jhep.2017.03.021
- Estimation and comparison of receiver operating characteristic curves.Stata J. 2009; 9: 1
- Modeling the impact of alternative strategies for rapid molecular diagnosis of tuberculosis in Southeast Asia.Am J Epidemiol. 2013; 178: 1740-1749https://doi.org/10.1093/aje/kwt210
- Poor sensitivity of commercial rapid diagnostic tests for hepatitis B e antigen in senegal, West Africa.Am J Trop Med Hyg. 2018; 99: 428-434https://doi.org/10.4269/ajtmh.18-0116
- Pan-genotypic loop-mediated isothermal amplification assay for HBV: a simple, rapid and affordable point-of-care test to semi-quantify HBV DNA.J Hepatol. 2018; 68: S483-S484https://doi.org/10.1016/S0168-8278(18)31217-0
- Hepatitis B core-related antigen: an alternative to hepatitis B virus DNA to assess treatment eligibility in Africa.Clin Infect Dis. 2019; https://doi.org/10.1093/cid/ciz412
Article info
Publication history
Published online: July 19, 2019
Accepted:
June 26,
2019
Received in revised form:
June 24,
2019
Received:
May 13,
2019
Identification
Copyright
© 2019 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
