ABSTRACT.
There is a lack of real-world data on hepatitis B (HBV) treatment in Africa. We conducted a single-center 18-month prospective cohort study in Ethiopia to understand clinical, laboratory, and demographic variables associated with HBV treatment. One hundred fifty HBV-positive patients were included: 51 on treatment, 99 with no treatment. Median age was similar between groups. Those on treatment were more likely to be male (86%), report higher coffee intake (90% versus 70%, P < 0.05), lower khat intake (0% versus 9%, P = 0.08), lower alcohol consumption (0% versus 5%, P = 0.1), and had attained higher levels of education (56% versus 42%, P = 0.19). Individuals on treatment had higher median aspartate aminotransferase (AST), alanine aminotransferase (ALT), HBV DNA, and median Aminotransferase-to-Platelet Ratio Index and Fibrosis-4 scores. At 6 and 12 months, those on treatment showed a decrease in median AST, ALT, and fibrosis scores and had less hepatocellular carcinoma development at 6 months (2% versus 4%). Our study highlights potential demographic disparities in HBV treatment as well as benefits in a real-life setting in Africa.
Hepatitis B virus (HBV) infection represents a significant worldwide burden.1–3 In Ethiopia, prevalence of HBV has been reported as 7.4%, and there has not been a robust real-world description of HBV treatment in the region.4 The progression of HBV-related liver disease can depend on HBV genotype and can be hastened with increased viral load, coinfection with other viruses, and exposures such as alcohol, khat (Catha edulis, a widely used recreational drug), and tobacco use.5–7
Antiviral medications can prevent liver fibrosis by suppressing HBV replication.8 However, most studies showing benefit are performed in resource-rich settings and do not highlight the challenges associated with antiviral treatment in resource-limited settings.8–10 Africa has one of the highest prevalence of HBV in the world with few studies addressing real-world variables.11–13 Although some studies have described characteristics of patients undergoing treatment of chronic hepatitis B in Africa, to our knowledge, no studies have evaluated the impact of socioeconomic background and exposures on HBV treatment in Ethiopia.14,15
In this study, we aimed to identify characteristics of individuals who were on treatment versus on no treatment of HBV in Ethiopia to assess whether baseline demographic differences and the cumulative effect of environmental exposures may have affected outcomes such as laboratory findings, ultrasound data, and the development of hepatocellular carcinoma (HCC).
This was a prospective cohort study conducted at Tikur Anbessa Hospital in Addis Ababa, Ethiopia. Patients with positive hepatitis B surface antigen (HBsAg) were identified and recruited in the gastroenterology referral clinic starting in January 2020. Inclusion criteria included positive HbsAg regardless of presence of cirrhosis or advanced fibrosis, age ≥18 years, and having documented follow-up in the hospital (at least two prior clinic visits). Demographic, clinical, laboratory, and ultrasound data were obtained during clinic visits at baseline and at 6, 12, and 18 months when available. Data were extracted from patient charts at 6-month intervals by participating investigators and recorded in REDCap (Research Electronic Data Capture). Data collected in REDCap was integrated into Excel, which was used to calculate the aspartate aminotransferase-to-platelet ratio index (APRI) and Fibrosis-4 (FIB4) scores using their respective formulas. Specific HCC scores such as modified PAGE-B or REAL-B scores could not be calculated because few patients had alpha fetoprotein or albumin values. Patients were divided into on-treatment versus no-treatment groups for the analyses. Analyses were performed using the χ2 test and Wilcoxon signed-rank test via IBM SPSS (Version 24) with P < 0.05 considered significant.
Patients were initiated on treatment according to standard of care in the facility, and by using the WHO HBV guidelines (APRI > 2, or age > 30 years with abnormal ALT and HBV DNA > 20,000 IU/mL).8 The diagnosis of HCC was based on either advanced imaging (computed tomography, magnetic resonance imaging) or biopsy. The study was approved by the Department of Medicine at Addis Ababa University, Ethiopia.
During the study period, 150 HBsAg-positive patients were included, of whom 51 were receiving treatment for HBV (on treatment) and 99 were not (no treatment). Median age was similar between groups (35, interquartile range [IQR] 29–46, and 34 years, IQR 28–41 years) with the on-treatment group more likely to be male (86%, N = 44) compared with the no-treatment group (76%, N = 75, P = 0.1). Tenofovir was the most common antiviral used (98%, N = 50), with only one patient treated with entecavir and one treated with tenofovir-lamivudine. Interestingly, the on-treatment group reported a higher level of education (56% versus 42%) with more patients reporting a college or professional degree (P = 0.19) (Table 1).
Table 1.
Characteristics of patients in the on-treatment and no-treatment groups for hepatitis B
| Baseline characteristics | On treatment (n = 51) | No treatment (n = 99) | P value |
|---|---|---|---|
| Median age, years | 35 (IQR 29–46) | 34 (IQR 28–41) | |
| Female sex (%, n) | 14% (7) | 25% (25) | 0.10 |
| Level of education (n) | 30 | 64 | 0.19 |
| Elementary (%, n) | 17% (5) | 27% (17) | |
| High school (%, n) | 27% (8) | 31% (20) | |
| College (%, n) | 36% (11) | 23% (15) | |
| Professional degree (%, n) | 20% (6) | 19% (12) | |
| Current smoking | 0% (0) | 0% (0) | |
| Current alcohol use | 0% (0) | 5% (5) | 0.1 |
| Khat use (n) | 30 | 65 | 0.09 |
| Yes (%, n) | 0% (0) | 9% (6) | |
| No (%, n) | 100% (30) | 91% (59) | |
| Coffee use (n) | 31 | 64 | 0.03 |
| Yes (%, n) | 90% (28) | 70% (45) | |
| No (%, n) | 10% (3) | 30% (19) | |
| Frequency of coffee use (n) | 28 | 45 | 0.74 |
| > 2 times per day (%, n) | 39% (11) | 51% (23) | |
| Daily (%, n) | 50% (14) | 36% (16) | |
| 3 times per week (%, n) | 11% (3) | 7% (3) | |
| Weekly (%, n) | 0% (0) | 7% (3) | |
| Ascites (%, n) | 43% (22) | 5% (5) | <0.01 |
| Ultrasound, abnormal (n) | 48 | 90 | <0.01 |
| Yes (%, n) | 60% (29) | 20% (18) | |
| No (%, n) | 40% (19) | 80% (72) | |
| Type of abnormality (n) | 29 | 18 | |
| Mass (%, n) | 17% (5) | 0% (0) | |
| Gallstones (%, n) | 0% (0) | 7% (2) | |
| Ascites (%, n) | 62% (18) | 14% (4) | |
| Fatty liver (%, n) | 7% (2) | 35% (10) | |
| Nodularity (%, n) | 0% (0) | 7% (2) | |
| Other (%, n) | 14% (4) | 0% (0) | |
| Median AST (IU/L) | 36 (IQR 27–56), N = 48 | 27 (IQR 22–36), N = 97 | |
| Median ALT (IU/L) | 32 (IQR 21–39), N = 48 | 26 (IQR 20–38), N = 96 | |
| Median total bilirubin (mg/dL) | 0.67 (IQR 0.51–0.83), N = 39 | 0.63 (IQR 0.5–0.81), N = 72 | |
| Median Log HBV DNA (IU/mL) | 25,921 (IQR 3,162–178,430), N = 23 | 389 (IQR 102–1,816), N = 55 | |
| HBeAg positive (%, n) | 6% (1), N = 18 | 6% (3), N = 54 | |
| Median APRI | 0.54 (IQR 0.37–1.51), N = 44 | 0.31 (IQR 0.21–0.46), N = 74 | |
| Median FIB-4 | 1.79 (IQR 0.86–2.65), N = 44 | 0.85 (0.51–1.28), N = 73 | |
| 6-month characteristics | |||
| Treatment (%, n) | 39% (20) | 40% (39) N = 98 | 0.95 |
| Ultrasound, abnormal (%, n) | 40% (20) N = 50 | 41% (39) N = 96 | 0.94 |
| Type of abnormalities (n) | 20 | 39 | |
| Mass (%, n) | 5% (1) | 10% (2) | |
| Gallstones (%, n) | 10% (2) | 5% (1) | |
| Ascites (%, n) | 30% (6) | 75% (15) | |
| Fatty liver (%, n) | 25% (5) | 35% (7) | |
| Nodularity (%, n) | 10% (2) | 20% (4) | |
| Other (%, n) | 20% (4) | 50% (10) | |
| HCC (%, n) | 2% (1) N = 50 | 4% (4) N = 95 | 0.49 |
| Median APRI | 0.48 (IQR, 0.28–0.75), N = 38 | 0.36 (IQR 0.24–0.56), N = 84 | |
| Median FIB 4 | 1.0 (IQR 0.6–1.76), N = 36 | 1.0 (IQR 0.62–1.55), N = 83 | |
| Median AST (IU/L) | 29 (IQR 22–41), N = 49 | 30 (IQR 24–43), N = 97 | |
| Median ALT (IU/L) | 29 (IQR 21–38), N = 47 | 28 (IQR 20–39), N = 97 | |
| Median Log HBV DNA (IU/mL) | 488 (IQR 129–3,517), N = 20 | 392 (IQR 68–1,103), N = 32 | |
| HbeAg positive (%, n) | 0% (0), N = 36 | 0% (0), N = 53 | |
| 12-month characteristics | |||
| Treatment | 39% (9), N = 23 | 44% (24), N = 55 | |
| Ultrasound, abnormal | 42% (8), N = 19 | 35% (15), N = 43 | |
| Type of abnormalities | N = 8 | N = 15 | |
| Mass (%, n) | 0% (0) | 0% (0) | |
| Gallstones (%, n) | 25% (2) | 7% (1) | |
| Ascites (%, n) | 13% (1) | 13% (2) | |
| Fatty liver (%, n) | 25% (2) | 7% (1) | |
| Nodularity (%, n) | 0% (0) | 7% (1) | |
| Other (%, n) | 38% (3) | 67% (10) | |
| HCC (%, n) | 0, N = 22 | 0, N = 53 | |
| Median APRI | 0.51 (IQR 0.44–0.75), N = 22 | 0.6 (IQR 0.45–0.79), N = 51 | |
| Median FIB 4 | 0.96 (IQR 0.71–1.74), N = 22 | 1.0 (IQR 0.74–1.6), N = 52 | |
| Median AST (IU/L) | 26 (IQR 24–33), N = 22 | 29 (IQR 21–39), N = 48 | |
| Median ALT (IU/L) | 25 (IQR 20–38), N = 22 | 27 (IQR 20–34), N = 47 | |
| Median Log HBV DNA (IU/mL) | 102 (IQR 5–425), N = 6 | 337 (IQR 17–2,153), N = 10 | |
| HbeAg positive (%, n) | 0% (0), N = 18 | 0% (0), N = 34 | |
| 18-month characteristics | |||
| Treatment (%, n) | 100% (1), N = 1 | 44% (10), N = 23 | |
| Ultrasound, abnormal (%, n) | No data | No data | |
| Type of abnormalities (n) | No data | No data |
ALT = alanine aminotransferase; APRI = Aminotransferase-to-Platelet Ratio Index; AST = aspartate aminotransferase; FIB4 = Fibrosis-4; HbeAg = hepatitis B e antigen; HBV = hepatitis B virus; HCC = hepatocellular carcinoma; IQR = interquartile range. Bold indicates statistically significant difference.
No tobacco use was reported in either group, and alcohol consumption was low (5% in the no-treatment versus 0% in the on-treatment group, P = 0.1). Among those with available data, 0% of on-treatment individuals reported using khat compared with 9% (N = 6) of the no-treatment group (P = 0.08), and 90% (N = 28) of the on-treatment group used coffee versus 70% (N = 45) of the no-treatment group (P=0.03).
On-treatment individuals had higher median aspartate aminotransferase (AST; 36 versus 27 IU/L), ALT (32 versus 26 IU/L), and HBV DNA (25,921 versus 389 IU/mL). The on-treatment group also had higher median APRI (0.54 versus 0.31) and median FIB4 scores (1.79 versus 0.85). Higher median ALT, HBV DNA, and APRI is consistent with treatment decision per WHO guidelines. Females in the on-treatment group had higher median APRI (1.29 versus 1.52) and FIB4 scores (2.22 versus 1.73) compared with males.
Among 138 individuals for whom liver ultrasound data was available, on-treatment individuals (N = 48) had more abnormal findings (P < 0.01), specifically ascites (62%, N = 18) and liver mass (17%, N = 5). Of the individuals on-treatment with a liver mass on baseline ultrasound, four of five patients had HCC confirmed via imaging or biopsy.
Our cohort showed strong 6-month follow-up, with at least 94% of patients in both groups returning to clinic (Figure 2). Those on treatment at baseline showed a decrease in median AST (36 to 29 IU/L; P = 0.09), median ALT (32 to 29 IU/L; P = 0.3), and median FIB4 (1.79 to 1.0; P = 0.005) and APRI (0.54 to 0.48; P = 0.1) scores (Figure 1A and B). The no-treatment group showed a mild increase in median AST (27 to 30 IU/L), median ALT (26 to 28 IU/L), and FIB4 (0.85 to 1.0) and APRI (0.31 to 0.36) scores. One case of HCC was reported in the on-treatment group, whereas HCC developed in 4% (N = 4) of the no-treatment group during the same time. At 6-month follow-up, 40% (N = 39) of the no-treatment individuals had started on treatment.
Figure 2.
Bar graph showing percentage of patients at follow up at baseline, 6, 12, and 18 months in the on-treatment and no-treatment groups.
Figure 1.
Line graph showing (A) Aminotransferase-to-Platelet Ratio Index (APRI) and (B) Fibrosis-4 (FIB4) scores in patients in the on-treatment and no-treatment groups over 12 months.
At 12 months, when patients who had initiated treatment at 6 months were added to the on-treatment group, the follow-up was more similar between groups with (53% versus 54%) (Figure 2). In the on-treatment group, from baseline to 12 months, there was a decrease in median AST (36 to 26 IU/L), ALT (32 to 25 IU/L) and APRI (0.54 to 0.51), and FIB4 (1.79 to 0.96) scores (Figure 1A and B). The no-treatment group showed an increase from baseline in median AST (27 to 29 IU/L), median ALT (26 to 27 IU/L), FIB4 (0.85 to 1.0), and APRI (0.31 to 0.6) scores. No new HCC was reported in either group at 12 months, and two additional patients initiated treatment.
At 18 months, there was significant loss to follow-up. When all patients who had initiated treatment at 6 and 12 months were added to the on-treatment group, 19% of the on-treatment group followed up versus 23% of the no-treatment group. There was not enough data obtained to compare median laboratory values, and no ultrasound data were available for either group.
We found that patients on HBV treatment were more likely to be male and have higher baseline liver-specific laboratory findings, suggesting males are more likely to have immune-active hepatitis B or are more likely to seek and advocate for treatment because not all patients who were treated had > 20,000 IU/mL HBV DNA. Females initiated on treatment tended to have higher APRI and FIB4 scores, suggesting a higher fibrosis need to be considered for treatment. Decision for initiation of treatment was likely affected by factors including cost, availability of medications in the area, and ability to afford laboratory tests. Guidelines followed by practitioners were not the focus of this study, so it is possible that treatment decisions were made based on internal guidelines or practice and may be an area for further study.16
There was a trend toward a higher level of education in the on-treatment group when compared with the no-treatment group (although not statistically significant), suggesting that those with more education may be more likely to seek treatment. Those not on treatment showed more liver-unhealthy habits, such as alcohol and khat consumption, and less coffee intake, which has been suggested to reduce the risk of liver fibrosis.17,18 The exact mechanism of caffeine reducing liver fibrosis has not been elucidated; however, one hypothesis is caffeine decreasing transforming growth factor beta, which is a mediator of fibrosis, and without this, hepatic stellate cells are unable to differentiate into myofibroblasts.19,20 Interestingly, we found that those not on treatment tended to develop more HCC than those on treatment in short-term follow-up. This finding is of importance but should be taken with caution because it is specific to a very limited period of time.
At 6 and 12 months, those in the on-treatment group showed a decrease in their liver enzymes and FIB4, and APRI scores, aligning with effective treatment. Nearly half of those in the no-treatment group at baseline transitioned to treatment within a year of follow-up. A major concern in areas of resource-limited settings is that follow-up is challenging, and when patients with HBV are lost to follow-up, the risk of HCC development increases. At 18 months, only ∼1 in 5 of the individuals in either group followed up, highlighting some of the difficulties in considering treatment discontinuation and longitudinal clinical benefit assessment. Loss of follow-up is likely due to a minimalist approach from patients due to costs of care, including medications, laboratory fees, and transportation. However, future approaches to increase patient engagement and participation in surveillance programs should be considered.
The limitations of our study include being a single-center study with a relatively small number of patients, which may not be generalizable to the entire country, let alone continent. In addition, the exact time of treatment initiation could not be obtained because of inconsistent documentation and may have impacted the study end points. However, the study does show significant changes in the liver enzymes and noninvasive parameters, adding to the knowledge regarding benefits of therapy in a short period of time. We relied on patient-reported habits, and given our inclusion criteria requiring two prior visits, our data are subject to sampling bias selecting motivated participants with a propensity to follow up. In addition, our data lost strength at 18 months. Strengths of our study include the solid two time-point follow-up of the study patients and the description of real-world pattern of the treatment of HBV in this region. Other studies have followed a cohort with strong support from study organizers on follow-up, and no implications of cost; our study shows the challenges and outcomes when these strict measures are not in place.15 In addition, we were able to obtain lifestyle habits associated with chronic liver disease as part of this study.
In conclusion, those who were not treated for HBV in Ethiopia tended to be female and have more liver-unhealthy habits, with a trend toward lower levels of education. Further studies are needed to assess reasons for loss of follow-up to ensure treatment of HBV and prevention of its complications.
REFERENCES
- 1. Roth GA. et al. , 2018. Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 392: 1736–1788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. WHO , 2017. Global Hepatitis Report, 2017. Available at: http://www.who.int/hepatitis/publications/global-hepatitis-report2017/en/. Accessed January 17, 2023.
- 3. World Health Organization WHO Hepatitis B Fact Sheet. Available at: http://www.who.int/mediacentre/factsheets/fs204/en/. Accessed January 2, 2023.
- 4. Belyhun Y, Maier M, Mulu A, Diro E, Liebert UG, 2016. Hepatitis viruses in Ethiopia: a systematic review and meta-analysis. BMC Infect Dis 16: 761. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Anugwom CM, Allaire M, Akbar SM, Sultan A, Bollipo S, Mattos AZ, Debes JD, 2021. Hepatitis B–related hepatocellular carcinoma: surveillance strategy directed by immune-epidemiology. Hepatoma Res 7: 23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kuper H, Tzonou A, Kaklamani E, Hsieh CC, Lagiou P, Adami HO, Trichopoulos D, Stuver SO, 2000. Tobacco smoking, alcohol consumption and their interaction in the causation of hepatocellular carcinoma. Int J Cancer 85: 498–502. [PubMed] [Google Scholar]
- 7. Orlien SMS, Berhe NB, Morgan MY, Johannessen A, 2019. Khat-related liver disease in sub-Saharan Africa: neglected, yet important. Lancet Glob Health 7: e310. [DOI] [PubMed] [Google Scholar]
- 8.Leathers JS, Pisano MB, Re V, Van Oord G, Sultan A, Boonstra A, Debes JD, 2019. Evaluation of Rapid Diagnostic Tests for Assessment of Hepatitis B in Resource-Limited Settings. Ann Glob Health 85: 98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Marcellin P. et al. , 2013. Regression of cirrhosis during treatment with tenofovir disoproxil fumarate for chronic hepatitis B: a 5-year open-label follow-up study. Lancet 381: 468–475. [DOI] [PubMed] [Google Scholar]
- 10. Kim WR. et al. , 2015. Impact of long-term tenofovir disoproxil fumarate on incidence of hepatocellular carcinoma in patients with chronic hepatitis B. Cancer 121: 3631–3638. [DOI] [PubMed] [Google Scholar]
- 11. Yang JD. et al. , 2017. Characteristics, management, and outcomes of patients with hepatocellular carcinoma in Africa: a multicountry observational study from the Africa Liver Cancer Consortium. Lancet Gastroenterol Hepatol 2: 103–111. [DOI] [PubMed] [Google Scholar]
- 12. Wong DKH. et al. , 2020. Among patients with undetectable hepatitis B surface antigen and hepatocellular carcinoma, a high proportion has integration of HBV DNA into hepatocyte DNA and no cirrhosis. Clin Gastroenterol Hepatol 18: 449–456. [DOI] [PubMed] [Google Scholar]
- 13. Zampino R, Boemio A, Sagnelli C, Alessio L, Adinolfi LE, Sagnelli E, Coppola N, 2015. Hepatitis B virus burden in developing countries. World J Gastroenterol 21: 11941–11953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Aberra H, Desalegn H, Berhe N, Medhin G, Stene-Johansen K, Gundersen SG, Johannessen A, 2017. Early experiences from one of the first treatment programs for chronic hepatitis B in sub-Saharan Africa. BMC Infect Dis 17: 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Desalegn H, Aberra H, Berhe N, Mekasha B, Stene-Johansen K, Krarup H, Pereira AP, Gundersen SG, Johannessen A, 2018. Treatment of chronic hepatitis B in sub-Saharan Africa: 1-year results of a pilot program in Ethiopia. BMC Med 16: 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Johannessen A, Tadesse Gebremedhin L, Desalegn H, 2022. Comment new WHO hepatitis B treatment guidelines: look to Ethiopia. Lancet Glob Health 10: e1711–e1712. [DOI] [PubMed] [Google Scholar]
- 17. Setiawan VW, Wilkens LR, Lu SC, Hernandez BY, Le Marchand L, Henderson BE, 2015. Association of coffee intake with reduced incidence of liver cancer and death from chronic liver disease in the US multiethnic cohort. Gastroenterology 148: 118–125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kennedy OJ, Roderick P, Buchanan R, Fallowfield JA, Hayes PC, Parkes J, 2016. Systematic review with meta-analysis: coffee consumption and the risk of cirrhosis. Aliment Pharmacol Ther 43: 562–574. [DOI] [PubMed] [Google Scholar]
- 19. Heath RD, Brahmbhatt M, Tahan AC, Ibdah JA, Tahan V, 2017. Coffee: the magical bean for liver diseases. World J Hepatol 9: 689–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Gressner OA, Lahme B, Rehbein K, Siluschek M, Weiskirchen R, Gressner AM, 2008. Pharmacological application of caffeine inhibits TGF-β-stimulated connective tissue growth factor expression in hepatocytes via PPARγ and SMAD2/3-dependent pathways. J Hepatol 49: 758–767. [DOI] [PubMed] [Google Scholar]