Rapidly decreased HBV RNA predicts responses of pegylated interferons in HBeAg-positive patients: a longitudinal cohort study

Background As an important anti-HBV drug, pegylated interferon α (PegIFNα) offers promising clinical efficacy, but biomarkers that accurately forecast treatment responses are yet to be elucidated. Here, we evaluated whether HBV RNA could act as an early monitor of pegylated interferon responses. Methods We analyzed a phase 3, multicenter, randomized cohort of 727 HBeAg-positive non-cirrhotic patients receiving a 48-week treatment of PegIFNα-2a or PegIFNα-2b and a 24-week treatment-free follow-up. Serum levels of HBV RNA, HBV DNA, HBeAg, and HBsAg were measured at weeks 0, 12, 24, 48, and 72. Results HBeAg seroconversion and HBsAg loss at week 72 were observed in 217 (29.8%) and 21 (2.9%) patients, respectively. During the 48-week treatment, HBV RNA decreased more rapidly than HBV DNA and HBsAg, but HBV RNA and HBeAg shared similar dynamics with positive correlations. Multivariate regression analyses consistently revealed the significance of HBV RNA at weeks 0, 12, 24, and 48 to monitor HBeAg seroconversion but not HBsAg loss. Although baseline HBV RNA only showed a modest AUC performance, HBV RNA with a significant increase of AUC at week 12 outperformed other HBV biomarkers to forecast HBeAg seroconversion (p value < 0.05). HBV RNA ≤ 1000 copies/mL was an optimized cutoff at week 12 that offered better prediction than other HBV biomarkers. This optimized cutoff plus patient age, HBV genotype B, and HBeAg offered a strong estimation of HBeAg seroconversion (accuracy 95.2%, true negative rate 99.8%). Conclusion HBV RNA at week 12 is an effective monitor of HBeAg seroconversion in HBeAg-positive patients treated with pegylated interferons. Electronic supplementary material The online version of this article (10.1007/s12072-020-10015-3) contains supplementary material, which is available to authorized users.


Supplementary Methods
In this section, we provide details about HBV RNA extraction and quantification.

Primers, probes, and internal controls
In order to measure serum HBV RNA, primers of 3'RACE, HBV RNA-forward, HBV RNA-reverse, and HBV RNA-probes were obtained in literature [12]. Noncompetitive internal controls were added to monitor the extraction and amplification of viral loads [13]. Information on primers, probes, and internal controls is summarized below.  Here, we collected HBV genomic sequences to prove that primer sequences above were located within the conserved regions of the HBV genome. First, fifteen HBV genomic sequences were extracted from the NCBI nucleotide database  and ambiguous nucleotides were taken into account (e.g. "W" represents "A" or "T"; "Y" represents "C" or "T"). MEGA X was used for data visualization.

HBV RNA extraction and amplification
HBV RNA extraction, PCR reaction, and amplification were performed on the Natch S system (Sansure Biotech, Inc, China) using protocols, reagents, and software provided by the manufacturer [14]. HBV RNA was extracted from patient serum samples (200 μL) using a nucleic acid extraction kit which was developed based on the magnetic bead technology. According to the manufacturer's protocol, HBV RNA extraction could be briefly described by seven-fold processes.
(1) Prepare an appropriate number of centrifugation tubes (1.5 mL). Mark negative control, positive control, and serum sample names on tubes accordingly. Add 600 μL of RNA Extraction Solution 1 into each tube. Note that RNA Extraction Solution 1 to 4 are available in the nucleic acid extraction kit.
(2) Add serum sample (200 μL), negative control, and positive internal control into each centrifugation tube. Close the lip, vortex it for 10 seconds, then centrifuge them instantaneously.
(3) Add 100 μL of RNA Extraction Solution 2 into each tube (mix it thoroughly before pipetting), vortex it for 10 seconds, then keep it at the temperature of 60℃ for 10 minutes, followed by 4℃ for 10 minutes.
(4) Centrifuge them instantaneously and then put centrifugation tubes in a separator.
After 3 minutes, pipette the liquids out slowly (do not touch the brown substance on the tube wall). Notably, all centrifugation tubes, pipettes, PCR tubes, and related equipment used above should be RNase free to prevent the interference of RNase.
Eluted HBV RNA (30 μL) was obtained from HBV RNA extraction above and 10 μL was used for reverse transcription. Real-time fluorescence quantitative PCR technology was applied to quantify HBV RNA using fluorescence probes and specific primers that target conserved HBV regions (see Table S

Internal control
To provide effective internal control, armored RNA technology was applied to encode predefined RNA sequences (see internal control in

HBV RNA standard
Stable standards are required for quantifying HBV RNA levels in patient serums.
We prepared HBV RNA standards similar to the approach proposed in the literature [18]. Here, we briefly describe the preparation of HBV RNA standards in threefold procedures. First, a plasmid pGS-1 ligated with a replication sequence was constructed:

Limit of detection (LOD)
In order to identify the limit of detection of HBV RNA assay, we prepared three batches to test HBV RNA levels based on 7 different concentrations ranged from 100 copies/mL to 1000 copies/mL. Each batch with 40 samples was tested independently to measure the number of successful detections at different concentrations. New results are summarized in Table S 7. Given any serum sample, we did conduct another test if weak signals of HBV RNA could not be observed.
We performed statistical analyses to show the robustness of the HBV RNA method (Table S 7). Based on experimental results, HBV RNA ≥ 250 copies/mL maintained 100% (120/120) of detection rates over three batches (Figure S 5). With 100% of the detection rate, 250 copies/mL of HBV RNA were potentially considered as the limit of detection, though the focus of our study was not to prove the best detection limit for the HBV RNA assay.   RNA assays were evaluated using HBV RNA specimens at the concentration of 5×10 2 and 5×10 3 , whose results were illustrated in red and blue, respectively. Tests upon each specimen were repeated for 20 times. Coefficients of variability for HBV RNA at concentrations of 5×10 2 and 5×10 3 were 2.09% and 1.56%, respectively.  For HBV RNA at 250 copies/mL, we also demonstrated the amplification curves and characterized the coefficient of variation, defined by the ratio of the standard deviation to the mean. Given 120 individual tests, the CV value of specimens harboring HBV RNA at concentrations of 250 copies/mL was 4.84% (see Figure S 7, PCR raw data is available in the supplementary file). Moreover, the CV values of HBV RNA at 5×10 2 and 5×10 3 were 2.09% and 1.56%, respectively (Figure S 6).
In addition to the analysis of HBV RNA at 250 copies/mL, we demonstrated the amplification curves of HBV RNA at the concentrations from 5×10 2 to 5×10 9 copies/mL (Figure S 8). The linear regression curve mapped the standard curve along with the HBV RNA concentrated from 5×10 2 copies/mL (log10 value: 2.699) to 5×10 9 copies/mL (log10 value: 9.699). The correlation coefficient was -3.479002 with R 2 =0.999. Taken together, we observed a good linear relation in standard curves of HBV RNA based on diluted plasmids from 5×10 2 to 5×10 9 copies/mL.