Journal of Clinical Virology 61 (2014) 305–308
Contents lists available at ScienceDirect
Journal of Clinical Virology journal homepage: www.elsevier.com/locate/jcv
Letter to the Editor Calibration of qualitative HBsAg assay results for quantitative HBsAg monitoring
a r t i c l e
i n f o
Keywords: Hepatitis B Surface antigen Qualitative Quantitative HBV HBsAg
a b s t r a c t Evidence is accumulating that quantitative hepatitis B surface antigen monitoring may be useful in managing patients with chronic HBV infection on certain treatment regimens. Based on these results with the Abbott Architect qualitative and quantitative HBsAg assays, it seems feasible to convert qualitative to quantitative HBsAg values for this purpose. © 2014 Elsevier B.V. All rights reserved.
Evidence is accumulating for the value of quantitative hepatitis B surface antigen (qHBsAg) testing and monitoring for understanding and predicting the natural clearance of HBsAg in chronic hepatitis B-infected (CHB) and treatment-naïve patients [1–4], and assessing the likely responses to interferon- and nucleos(t)ide analoguebased therapies [4–6]. The use of qHBsAg monitoring has been recently recommended in national guidelines for the management of chronic hepatitis B (HBV) infection [7]. The Abbott Architect (Abbott Diagnostics, Abbott Park, IL, USA) is a popular platform for high throughput qualitative HBsAg (HBsAg Qualitative II kit, B2G220) testing. A quantitative version of this kit (HBsAg, B6C3G0) offers qHBsAg testing. During the evaluation and the validation of the Architect qHBsAg assay, for introduction into our routine diagnostic service, we noticed a remarkable linearity of the Architect qualitative HBsAg (qlHBsAg) assay. Previous studies have calibrated qlHBsAg assays using qHBsAg assays for specific internal and research use [8,9], so we decided to investigate this further to determine if these qlHBsAg results could be accurately and consistently converted to qHBsAg results. Our evaluation of the qlHBsAg vs qHBsAg Abbott Architect assays took place in two stages. First, from our archived samples, we selected high HBsAg concentration serum samples from each of the HBV A, B and C genotypes (the most common HBV genotypes in this Alberta population). These were diluted to give a wide linear range and tested in parallel on the qlHBsAg and qHBsAg assays (Fig. 1). Note that the linear regression curves are relatively similar across the three different HBV genotypes, so for the rest of this analysis we are assuming that the different HBV genotypes give similar results.
http://dx.doi.org/10.1016/j.jcv.2014.07.017 1386-6532/© 2014 Elsevier B.V. All rights reserved.
The second stage of this exercise took place once the qHBsAg assay was formally introduced into routine service. Each week, for 5 consecutive weeks, samples received routinely (about 5–10 samples/week) for qHBsAg (for which an automatic or manual 1:500 dilution is required as per manufacturer’s instructions) were also run at a 1:500 dilution on the qlHBsAg assay. The resulting qHBsAg (IU/mL) (x-axis) and qlHBsAg S/CO (signal to cut-off) ratio (y-axis) values were plotted and a linear regression curve was fitted. For each of the subsequent weeks after the first week, the 1:500 diluted qlHBsAg S/CO value was inserted into the previous week’s linear regression equation to obtain a ‘calculated’ qHBsAg result for each sample. Note that this 1:500 dilution step (which is not routinely required) for the qlHBsAg assay is important to obtain this calculated qHBsAg value in this way. Examples of these results, including the linear regression curve obtained from the full 5 weeks of qlHBsAg/qHBsAg data (Fig. 2A) are shown in Fig. 2B–F. Although there is some variation in the calculated vs measured qHBsAg values derived from the 5-week linear regression curve, the overall comparison over this 5-week period is reasonably close (all R2 values > 0.93). Note that the weekly calculated vs measured qHBsAg values cover variable ranges, depending on the clinical samples received for qHBsAg testing during each week. Based on these results, we would like to suggest that a mean of these linear regression curves may be useful for converting qlHBsAg values to qHBsAg values – especially in laboratories that are already running the qlHBsAg assay, where the introduction of the relatively low volume qHBsAg assay may not be deemed particularly cost-effective.
306
Letter to the Editor / Journal of Clinical Virology 61 (2014) 305–308
A Qualitative (S/CO)
1,000,000.00 100,000.00
Genotype A:
10,000.00
y = 317573x0.8829
1,000.00
Genotype B:
100.00
y = 613035x0.9685
Genotype C: 0.9747
y = 149254x
10.00 1.00 0.10 1
0.1
0.01
0.001
0.0001
0.00001
0.000001
Dilution Genotype A
Genotype B
Genotype C
Trendline Genotype A
Trendline Genotype B
Trendline Genotype C
B Quantitative (IU/mL)
100,000.00 10,000.00
Genotype A:
1,000.00
y = 18753x0.9542
100.00
Genotype B: 0.9643
10.00
y = 19039x
Genotype C:
1.00
0.9513
y = 3011.7x
0.10 0.01 1
0.1
0.01
0.001
0.0001
0.000001
0.00001
Dilution Genotype A
Genotype B
Genotype C
Trendline Genotype A
Trendline Genotype B
Trendline Genotype C
C Qualitative (S/CO)
1,000,000.00 Genotype A:
100,000.00
0.9232
y = 35.572x
10,000.00
Genotype B: 1.0035
y = 30.923x
1,000.00 Genotype C:
100.00
1.0246
y = 40.681x
10.00 1.00 100,000.00
10,000.00
1,000.00
100.00
10.00
1.00
0.10
0.01
Quantitative (IU/mL) Genotype A
Genotype B
Genotype C
Trendline Genotype A
Trendline Genotype B
Trendline Genotype C
Fig. 1. Plots showing the linearity of the qualitative (A) and quantitative HBsAg (B) assay performance with serial dilutions of serum samples of HBV genotypes A, B and C. Combined plots relating the qualitative (qlHBsAg) to the quantitative (qHBsAg) assays for each HBV genotype A, B and C (C): y = 35.572x0.9232 , y = 30.923x1.0035 and y = 40.681x1.0246 , respectively.
Letter to the Editor / Journal of Clinical Virology 61 (2014) 305–308
A
307
Week 1
B Calculated (IU/ml)
10,000
qlHBsAg (S/CO)
1,000 100 y = 0.071x0.9842 R2 = 0.9921
10
150,000 100,000 50,000
y = 0.7711x + 3823.7 R2 = 0.9851
0 1 100
50,000
0 1,000
10,000
100,000
1,000,000
100,000
150,000
Measured (IU/ml)
qHBsAg (IU/ml)
Week 3
Week 2
C
Calculated (IU/ml)
Calculated (IU/ml)
D 50,000 40,000 30,000 20,000
y = 0.8361x + 218.6
10,000
2
R = 0.9993
2,500 2,000 1,500 y = 1.1814x - 33.482
1,000
R2 = 0.9329
500 0
0 0
10,000
20,000
30,000
40,000
0
50,000
500
Measured (IU/ml)
1,000
1,500
2,000
2,500
Measured (IU/ml)
Week 4
Week 5
Calculated (IU/ml)
40,000 30,000 20,000
y = 1.0743x + 345.68
10,000
R2 = 0.9946
0
Calculated (IU/ml)
F
E
20,000 15,000 10,000
y = 1.1155x - 323.55
5,000
R2 = 0.973
0
0
10,000
20,000
30,000
40,000
0
Measured (IU/ml)
5,000
10,000
15,000
20,000
Measured (IU/ml)
Fig. 2. (A) Plot (with linear regression line) of 5 consecutive weeks of samples received for routine qHBsAg, also tested by qlHBsAg in parallel. (B–F) Examples of how the calculated qHBsAg result (by applying the linear regression curve obtained from A) compare with the measured qHBsAg values over each of the 5 consecutive weeks.
Funding None. Competing interests None declared. Ethical approval Not required. Acknowledgments This work was supported in part by Abbott Laboratories Ltd., Ontario, Canada, in the form of some free validation/evaluation kits and supplemental material. However, Abbott Laboratories Ltd. played no role in the final writing of this paper.
References [1] Nguyen T, Thompson AJV, Bowden S, Croagh C, Bell S, Desmond PV, et al. Hepatitis B surface antigen levels during the natural history of chronic hepatitis B: a perspective on Asia. J Hepatol 2010;52:508–13. [2] Jaroszewicz J, Serrano BC, Wursthorn K, Deterding K, Schlue J, Raupach R, et al. Hepatitis B surface antigen (HBsAg) levels in the natural history of hepatitis B virus (HBV)-infection: a European perspective. J Hepatol 2010;52: 514–22. [3] Seto W-K, Wong DK-H, Fung J, Hung IF-N, Fong DY-T, Yuen JC-H, et al. A large case–control study on the predictability of hepatitis B surface antigen levels three years before hepatitis B surface antigen seroclearance. Hepatology 2012;56:812–9. [4] Hadziyannis E. Quantification of HBsAg in serum: characteristics of the assays. OA Hepatology 2013;1:1–6. [5] Zoutendijk R, Hansen BE, van Vuuren AJ, Boucher CAB, Janssen HLA. Serum HBsAg decline during long-term potent nucleos(t)ide analogue therapy for chronic hepatitis B and prediction of HBsAg loss. J Infect Dis 2011;204:415–8. [6] Martinot-Peignoux M, Lapalus M, Asselah T, Marcellin P. The role of HBsAg quantification for monitoring natural history and treatment outcome. Liver Int 2013;33:125–32. [7] National Institute for Health and Care Excellence (UK). Hepatitis B (chronic): diagnosis and management of chronic hepatitis B in children, young people and adults. NICE clinical guideline 165: guidance.nice.org.uk/cg165. Issued: June 2013.
308
Letter to the Editor / Journal of Clinical Virology 61 (2014) 305–308
[8] Tuaillon E, Mondain A-M, Nagot N, Ottomani L, Kania D, Nogue E, et al. Comparison of serum HBsAg quantitation by four immunoassays, and relationships of HBsAg level with HBV replication and HBV genotypes. PLOS ONE 2012;7:e32143. [9] Congly SE, Wong P, Al-Busafi SA, Doucette K, Fung SK, Ghali P, et al. Characterization of hepatitis B virus genotypes and quantitative hepatitis B surface antigen titres in North America tertiary referral liver centres. Liver Int 2013;33:1363–9.
Hans Gunning Dena Adachi Alberta Provincial Laboratory for Public Health, Edmonton, Alberta, Canada a
Julian W. Tang a,b,∗ Alberta Provincial Laboratory for Public Health, Edmonton, Alberta, Canada b Department of Medical Microbiology and Immunology, University of Alberta, Edmonton, Alberta, Canada
∗ Corresponding author at: Alberta Provincial Laboratory for Public Health, University of Alberta Hospital, 8440 – 112 Street, 2B1.03, Edmonton, Alberta T6G 2J2, Canada. Tel.: +1 780 407 3068; fax: +1 780 407 8961. E-mail addresses:
[email protected],
[email protected],
[email protected] (J.W. Tang).
27 June 2014