Short Communication: Serum-Based Assay Accurately Detects Single Nucleotide Polymorphisms of IL28B and SOCS3 in HIV/Hepatitis C Virus-Coinfected Subjects

Introduction

Recent advances have explored the value of using single nucleotide polymorphisms (SNP) of the human genome as an invaluable resource for prediction susceptibility and outcome of therapy. ¹ It is widely expected that such efforts will be more commonly employed and applied to most diseases. However, there exist significant restrictions in collecting DNA samples from patients, and thus, performing SNP detection using noncellular DNA would greatly enhance our ability to individualize treatment [Genetics in medicine ACMG statement]. In this regard, SNPs have become important in predicting treatment response in both hepatitis C virus (HCV) monoinfected ² and HIV/HCV coinfected patients. ³ Recent studies demonstrate that a polymorphism of the IL28B gene (SNP rs12979860) is a reliable, clinically useful predictor of cure or sustained virologic response (SVR) among patients treated with peg-interferon and ribavirin. ⁴ In addition, pretreatment expression levels and polymorphisms of another gene, SOCS3, an inhibitor of the interferon (IFN)-α-induced JAK/STAT pathway associated with obesity and insulin resistance, have also been associated with interferon-based treatment outcomes. ^{2,5 –8} While SVR is associated with reduced morbidity and improved survival, ⁵ predicting which patients will respond to therapy is challenging.

Due to lack of availability of DNA samples collected in clinical trials, a reliable method for extracting host DNA from serum for genotyping assays would present a practical alternative for clinicians and investigators seeking to perform SNP analyses, particularly in resource-limited settings. In this regard, there is at least one recent report that has used serum samples to perform IL28B SNP typing with dried blood spots, serum, and epithelial cells. ⁹ However, this technique has not been extended to other SNPs. In this study, we compare the reliability and reproducibility of genotyping analysis for IL28B SNP rs12979860 (hereafter “IL28B SNP”) and SOCS3 SNP rs4969170 (hereafter “SOCS3 SNP”) from both human peripheral blood mononuclear cells (PBMCs) and serum.

Materials and Methods

Patients coinfected with HIV and HCV on stable antiretroviral therapy were enrolled in three National Institute of Allergy and Infectious Disease (NIAID) IRB approved trials and treated with pegylated IFN-α2a (180 μg weekly), pegylated IFN-α2b (1.5 μg/kg weekly), or albinterferon-α2b (900 μg every other week), all in combination with RBV (1–1.2 g/day) for 48 weeks. Eligibility criteria have been described previously. ^{10 –12} All patients signed an informed consent for participation that was approved by the NIAID Institutional Review Board. Serum and whole blood samples were collected at regular protocol-defined intervals, processed, and stored at −80°C. The time of storage of samples ranged from 4 to 9 years (mean 6.9 years). All subjects signed informed consent that included genotyping in the original treatment protocol.

Results

Fifty-one HIV/HCV coinfected patients with available PBMCs and serum were included. Table 1 shows a summary of the patients' demographic data. The study subjects were predominantly male (88%), African American (63%), HCV genotype 1 (80%), and had relatively higher levels of CD4 T cells. Overall, there was a 37% SVR rate among this cohort. Results using plasma are quite similar and serum was chosen due to sample availability (data not shown).

A comparison of the results of the genotype analyses performed using DNA isolated from PBMCs or cell-free serum showed a 100% agreement between the IL28B genotyping results from the serum and PBMC isolates and 98% agreement for SOCS3 SNP (Table 2). One subject could not be genotyped for SOCS3 SNP in one serum sample on multiple attempts and therefore was considered undetermined.

Discussion

In this study we describe a novel method to isolate DNA fragments from the serum of HIV/HCV-coinfected subjects in order to accurately determine each subject's genotype for each of two SNPs: IL28B (rs12979860) and SOCS3 (rs4969170). When compared to the standard assay using DNA isolated from PBMCs, we were able to obtain 100% and 98% accuracy for IL28B and SOCS3, respectively, using the serum-based assay. This assay could be immediately valuable for detecting clinically relevant SNPs from serum in cases in which PBMCs are not available.

Since the description of polymorphisms in the IL28B SNP as an important predictor of HCV treatment outcomes, ^4,13 genotyping for this SNP has become a routine clinical assay used by physicians in the evaluation of HCV-infected patients prior to treatment. Additionally, in a recent multicenter study that treated 500 HCV genotype 3-infected subjects, baseline steatosis was an independent predictor of SVR ¹⁴ ; however, a correlation among the SOCS3 SNP in particular, hepatic steatosis, and treatment outcomes was not performed because PBMCs were not collected. Our assay provides a practical approach to determine the SOCS3 and IL28B genotypes in patients with stored serum to determine its role in clinical liver outcomes.

Moving forward, we believe this technique will have wider applications in various diseases. As described previously, the collection of cellular DNA material in larger multicenter trials is complex and limited. Hence, this technique offers a unique opportunity to perform targeted SNP analysis using serum stored over many years. As treatment for HCV is rapidly evolving into all oral directly acting agents, our assay, which provides accurate determination of the IL28B SNP and SOCS3 SNP genotypes from DNA extracted from serum, will allow clinicians and researchers to test all patients undergoing novel treatment for HCV more easily. In addition, this serum-based assay technique has the potential to be adapted to other novel SNPs, to identify and confirm novel host genetic mediators of disease susceptibility and treatment outcome. Although our assay can be useful in determining SNPs from stored serum samples, it raises the important issue about genetic testing of stored samples. In many cases, samples stored over time may not have linked informed consents from study subjects that permit genetic testing. In such cases, a special ethics committee review of proposal weighing in on the advantages of such testing needs to be conducted prior to use.

Conclusions

In conclusion, we describe the results of a serum-based genotyping assay for the determination of two SNPs that are predictive of HCV treatment response and are associated with clinical outcomes. Although we tested only two SNPs, this method could likely be adapted to detect additional SNPs, including those that are commercially available and custom designed, and may have wider applications in medical research in which SNP determination is relevant. This serum-based assay provides clinicians and researchers with a simpler alternative to tedious PBMC collection and processing and may be of particular use in resource-limited settings.