PBMC Not A Reservoir for HCV for Patients Who Clear HCV RNA, Study Reports
Clearance of hepatitis C virus RNA from the peripheral blood mononuclear cells of blood donors who spontaneously or therapeutically control their plasma viremia
Hepatology March 2008 Advance Publication
Flavien Bernardin 1 2, Leslie Tobler 1, Irina Walsh 1, Joan Dunn Williams 3, Mike Busch 1 2, Eric Delwart 1 2 *
1Blood Systems Research Institute, San Francisco, CA
2University of California, San Francisco, CA
3Blood Systems, Tempe, AZ
"On the basis of our results, the clearance of HCV from PBMC therefore appears complete in both spontaneously aviremic and successfully treated seropositive blood donors......These results indicate that PBMC-associated HCV is unlikely to be maintained as a viral reservoir with the potential to rekindle plasma viremia in aviremic subjects as determined by plasma TMA assays. This conclusion is supported by a recent analysis that similarly failed to detect PBMC-associated HCV RNA in 9 spontaneous and 2 treatment-induced aviremic patients. A greater than 90% rate of clearance of HCV RNA in the liver of sustained virological response also indicates that a long-lived hepatic reservoir is unlikely to exist"
We determined whether hepatitis C virus (HCV) RNA could be detected associated with peripheral blood mononuclear cells (PBMC) of seropositive blood donors who had spontaneously or therapeutically cleared their plasma viremia. Blood donor plasma viremia status was first determined with a highly sensitive transcription-mediated amplification (TMA) test performed in duplicate assays. PBMC from 69 aviremic and 56 viremic blood donors were then analyzed for the presence of HCV RNA with TMA adapted to detect viral RNA in PBMC and with a reverse transcription-nested polymerase chain reaction assay. PBMC-associated HCV RNA was detected in none of the 69 aviremic donors, including all 6 subjects with a sustained viral response following antiviral therapy. PBMC-associated HCV RNA was detected in 43 of the 56 viremic donors. The 13 viremic donors with no detectable PBMC-associated HCV RNA all had very low viral loads (6 positive only in 1 of 2 duplicate plasma TMA assays, 6 with viral loads below 100 HCV RNA copies/mL, and 1 with a viremia of 2700 HCV RNA copies/mL). The absence of detectable PBMC HCV RNA detection in all 69 aviremic donors reported here contrasts with prior studies, possibly as a result of the higher sensitivity of the TMA assay used to test for plasma viremia.
Conclusion: Our results indicate that PBMC are unlikely to serve as a long-lived reservoir of HCV in aviremic subjects.
Clearance of hepatitis C virus (HCV) from plasma occurs spontaneously within the first year of infection in 15%-45% of infections, depending on the route of infection and demographics of the subjects.[1-6] Combination antiviral therapies can also achieve sustained virological response, defined as persistent clearance of plasma viremia, in a growing fraction of treated patients. Numerous publications have analyzed extrahepatic compartments of viral replication that could potentially contribute to plasma viremia, most frequently peripheral blood mononuclear cells (PBMC).[7-14] The contribution of PBMC-based HCV replication to total viremia remains unclear, however, as the level of negative-strand HCV RNA in PBMC is very low with respect to that in liver.[15-18]
The continued presence of viral RNA in the PBMC of subjects who had either spontaneously cleared their plasma viremia or cleared viremia following antiviral therapy has recently been reported, raising concerns that PBMC may serve as a long-lived HCV reservoir capable of rekindling systemic infection.[19-22] In this study, plasma and corresponding PBMC were prepared from recalled confirmed HCV-seropositive blood donors who had tested either HCV RNA-positive or HCV RNA-negative in their original index donation. The presence of plasma viremia was tested for using a highly sensitive transcription-mediated amplification (TMA) assay.[23-25] After determining that this TMA assay could also be used to sensitively test for HCV RNA in PBMC and confirming our results using reverse transcription-nested polymerase chain reaction (RT-nPCR), we investigated whether clearance of plasma viremia was associated with clearance of viral RNA from PBMC. We have found that clearance of plasma viremia is tightly associated with clearance of HCV RNA from PBMC.
Using CA-TMA and RT-nPCR, two sensitive tests for the detection of viral RNA in PBMC, we found no evidence for cell-associated HCV RNA in aviremic seropositive subjects. These results contrast with those from studies that reported the detection of HCV RNA in the PBMC of spontaneously resolved and treatment-induced aviremic subjects.[20-22] These authors speculated that persistent HCV RNA in PBMC may reflect the presence of a long-lived HCV reservoir[20-22] (reviewed by Feld and Liang). The difference with our results may be due to the more stringent definition of plasma aviremia used here. To detect viremia, we used a TMA assay with a 95% detection limit of 30 RNA copies/mL performed in duplicate tests on a large plasma volume (2 X 0.5 mL). Prior studies used either a commercial HCV RNA test (Roche Amplicor 2.0) or in-house PCR assays with cDNA input derived from smaller plasma volumes.[20-22] TMA and the Amplicor 2.0 assay have been directly compared with respective sensitivities of 6 and 35 HCV IU/mL, respectively. Prior studies have also reported TMA-based detection of plasma HCV RNA in Amplicor-negative samples.[33-35] By using a more sensitive assay in duplicate tests, it is therefore likely that we detected low-level viremia that may have gone undetected by less sensitive assays.
Increased levels of cell-associated HCV RNA have also been reported following nonspecific in vitro stimulation and culture of PBMC. The PBMC tested here were purified directly from blood without external stimulation. It therefore remains formally possible that nonspecific PBMC stimulation could have increased residual PBMC-associated HCV RNA to detectable levels.
A strong association was found between the detection of plasma virus, determined with duplicate TMA tests, and the detection of PBMC-associated HCV RNA. PBMC HCV RNA was detected in 43 of 56 viremic donors but remained undetected in all 69 aviremic donors tested (P < 0.000001, Fisher's exact test). The 13 of 56 viremic donors with undetectable PBMC HCV RNA all had low-level plasma viremia [stochastically positive in only 1 of their 2 duplicate TMA tests (n = 6), viral loads below 100 (n = 6), or viral load of 2700 HCV RNA copies/mL (n = 1)], possibly reflecting a threshold level of plasma viremia before detectable association with PBMC. All 6 donors with stochastically positive plasma TMA tests were negative for PBMC HCV RNA, whereas 43 of the 50 duplicate TMA-positive donors had detectable levels of PBMC HCV RNA (P = 0.00005, Fisher's exact test). Absence of detectable PBMC HCV RNA was therefore limited to donors with no or low-level plasma viremia.
These results indicate that PBMC-associated HCV is unlikely to be maintained as a viral reservoir with the potential to rekindle plasma viremia in aviremic subjects as determined by plasma TMA assays. This conclusion is supported by a recent analysis that similarly failed to detect PBMC-associated HCV RNA in 9 spontaneous and 2 treatment-induced aviremic patients. A greater than 90% rate of clearance of HCV RNA in the liver of sustained virological response also indicates that a long-lived hepatic reservoir is unlikely to exist,[37-39] although another report detected hepatic HCV RNA in 50% of spontaneously aviremic seropositive subjects. The slow decrease in anti-HCV antibody titers in subjects with spontaneously cleared viremia[40-44] as well as the complete seroreversion detected in 7% of transfusion-transmitted infections may also reflect an absence of ongoing antigenic stimulation, indirectly supporting clearance of infection in persons who test HCV RNA-negative in plasma. Finally, our findings are supported by the lack of HCV transmission following 11 of 12 fresh whole blood transfusions (containing approximately 109 PBMC) from aviremic donors (using duplicate TMA). The single large blood volume transfusion that was infectious was most likely caused by HCV present below the detection limit of even duplicate TMA tests. On the basis of our results, the clearance of HCV from PBMC therefore appears complete in both spontaneously aviremic and successfully treated seropositive blood donors.
Assays for HCV RNA in PBMC.
Two different assays were used to test for the presence of HCV RNA in PBMC, TMA and RT-nPCR. The HCV RNA TMA test used (Procleix HIV-1/HCV assay, Gen-Probe) has a 95% detection sensitivity of 30 HCV RNA copies/mL, using an input of 0.5 mL of plasma, and detects all 6 HCV genotypes.[23-25]
The sensitivity of the TMA assay for PBMC-associated HCV RNA was first measured with total cellular RNA extracted from the PBMC of 3 viremic blood donors. HCV RNA levels in these PBMC were first measured with a commercial qPCR assay and then serially diluted to 1000 to 0.1 copies of HCV RNA (see the Patients and Methods section). These serial dilutions were then tested by replicate TMA assays. The results showed a CA-TMA sensitivity of 5 to 10 HCV RNA copies (Table 1A).
The sensitivity of the CA-TMA assay was then measured in the presence of a constant amount of cellular nucleic acids. Serial dilutions of 2 plasma samples of known viral loads (A and B) were added to a constant number of PBMC (5 X 106). The infected plasma/PBMC mixtures were then tested by CA-TMA (see the Patients and Methods section). The sensitivity of CA-TMA in the presence of a constant amount of PBMC was estimated at 2 to 50 HCV RNA copies in 5 X 106 PBMC (Table 1B).
Serial dilutions of infected plasmas added to a constant number of PBMC prior to total RNA extraction were also used to estimate the sensitivity of an RT-nPCR protocol (see the Patients and Methods section). An RT-nPCR sensitivity of 1 to 10 HCV RNA was measured (Table 1C). Considering that only a fraction of the extracted RNA (10/30 uL) and cDNA (5/25 uL) was used, the assay sensitivity was 15 to 150 HCV RNA molecules in 5 X 106 PBMC.
Aviremic Blood Donors.
Sixty-seven aviremic seropositive donors were selected on the basis of negative HCV RNA TMA minipool results followed by duplicate negative HCV TMA results on plasma from their individual (not pooled) index blood donation. Sixty of these donors remained plasma HCV RNA-negative on duplicate TMA assays performed on recall plasma collected at an average of 2.5 years following the index donations (Fig. 1A, dotted-line box). Of the 7of 67 initially aviremic seropositive donors who became TMA-positive at the later date (Fig. 1A, bold-line box), 5 were positive in both duplicate TMAs, whereas 2 were positive in only 1 of 2 TMA assays. The viral loads in the 5 consistently TMA-positive samples (that is, in 2 of 2 duplicate TMA assays) were <100 RNA copies/mL for 4 samples and 1.6 X 105 RNA copies/mL for 1 sample. The viral load in the stochastically TMA-positive samples (that is, in 1 of 2 duplicate TMA assays) was assumed to be near the limit of detection of the TMA assay (30 HCV RNA copies/mL). Stochastically positive TMA results have been previously reported for samples with very low HCV and human immunodeficiency virus viral loads.[28-30] Sixty of 67 initially aviremic seropositive donors therefore remained aviremic over an average of 2.5 years, and of the 7 donors that became viremic, only 1 donor had a viral load > 100 RNA copies/mL. These 7 seropositive subjects may reflect low-level plasma viral load fluctuations near the limit of detection or, in the case of the high viral load follow-up sample, possible reinfection with another HCV strain.
Viremic Blood Donors.
Fifty-eight viremic subjects (based on TMA testing of both minipool and individual samples from the first seropositive donations) were selected as controls for this study. Forty-nine of these donors were still viremic at the later sampling date (Fig. 1B, bold-line box), including 45 cases in which HCV RNA was detected in both duplicate TMA assays and 4 cases in which only 1 of the duplicate TMA tests was positive, indicating very low viral loads (2 of these 4 donors reported receiving antiviral therapy subsequent to their index donations). The remaining 9 initially viremic donors tested HCV RNA-negative in both duplicate TMA assays on the recall samples (Fig. 1B, dotted-line box); 6 of these 9 donors reported undergoing antiviral treatment for HCV following their index donation. The other 3 now aviremic donors may represent either spontaneous clearance of acute infections in the interval between their index and later donations or fluctuation of low-level viremia around the detection limit of TMA.
Detection of HCV RNA in PBMC.
PBMC samples were first grouped on the basis of plasma viremia detection at the time of the donors' initial seropositive donations (Fig. 1). Duplicate testing of 0.5 mL aliquots of plasma from their follow-up donations using HCV TMA classified their later recall samples into 69 aviremic and 56 viremic donations (Fig. 1, dotted-line and bold-lines boxes, respectively). Viremic donors were further categorized on the basis of whether recall plasmas were reactive either in both TMA tests (n = 50) or in only 1 of 2 replicate TMA tests (n = 6; Fig. 1 grey boxes with bold lines). Five million cryopreserved PBMC from both recall aviremic and viremic donations were then thawed and extensively rinsed. Half of the resulting PBMC was then analyzed for the presence of HCV RNA with CA-TMA, and the other half used for cellular RNA extraction and RT-nPCR (see the Patients and Methods section). When tested for the presence of HCV RNA, none of the 69 PBMC from aviremic donors were positive by either CA-TMA or RT-nPCR. This included the 6 donors who were viremic at the time of their index donations but became aviremic following reported anti-HCV therapies (Fig. 1B). PBMC from 43 of the 56 viremic donors were positive for HCV RNA (41 with CA-TMA and the same 41 plus another 2 cases with RT-nPCR). We next determined if the absence of detectable PBMC-associated HCV RNA in 13 otherwise viremic donors was associated with a low plasma viral load. Six of 13 PBMC HCV RNA-negative donors contained very low plasma viral loads as reflected by the stochastic nature of their plasma TMA results, with only 1 of 2 replicate plasma TMA tests positive. The other 7 plasma viremic donors were TMA plasma-positive in both duplicate TMA tests and were therefore likely to contain higher plasma viral loads. Plasma viral loads were measured for these 7 samples and found to be <100 RNA copies/mL for 6 samples and 2700 RNA copies/mL for 1 sample. The HCV RNA-negative PBMC were therefore all obtained from people with either no detectable plasma viremia (n = 69) or very low plasma viremia (n = 13). Each of the remaining 43 samples with detectable PBMC-associated HCV RNA was positive in both of their duplicate plasma TMA assays.
Demographics and Liver Functions of Viremic and Aviremic Donors.
No significant differences were detected in the demographics (gender, percentage of Caucasians, and age) of viremic (n = 56) and aviremic (n = 69) donors (Table 2). A comparison of the rate of abnormal liver biochemical results between the 2 groups at recall revealed significant differences for elevated alanine aminotransferase (ALT; P < 0.0001) and aspartate aminotransferase (AST; P < 0.0001) as expected from chronic versus resolved infections. No significant difference was observed for elevated total or direct bilirubin and rates of hypoalbuminemia. When the plasma viremic donors with (n = 43) and without (n = 13) PBMC HCV RNA were compared, significantly elevated ALT (P = 0.0008) and increased AST (P = 0.1) were detected in the HCV RNA positive PBMC group, but the demographics of the groups were similar (data not shown). This increase in liver dysfunction may be related to the higher HCV plasma viral loads seen in the PBMC HCV RNA-positive donors (that is, the proportion of repeatedly versus stochastically positive plasma HCV TMA results). When the 60 originally aviremic donors who remained aviremic at recall were compared to those 7 originally aviremic donors who later became viremic, no demographic or liver function differences were observed (data not shown), and this likely reflected the very low viral loads seen in these 7 subjects.
Patients and Methods
Subjects were identified from lists of donors confirmed to be HCV-seropositive detected following voluntary blood donations at blood centers in the southern and western United States operated by Blood Systems [HCV Enzyme Immunoassay (EIA) 3.0 and Recombinant Immunoblot Assay (RIBA) 3.0, Ortho Diagnostics, Raritan NJ]. Matched groups of HCV RNA-positive and RNA-negative donors (based on minipool TMA testing; discussed later) were recalled an average of 2.5 years following their index seropositive donations. Following informed consent, the donors filled out a questionnaire related to HCV risk factors, symptoms, and treatment and had blood drawn and processed. The protocol was approved by the Committee for Human Research at the University of California at San Francisco.
The TMA detection procedure is initiated by a sample lysis step of 0.5 mL of plasma with a detergent and denaturant buffer followed by the specific capture of HCV RNA with oligonucleotide-coated magnetic beads. The beads are rinsed, and the HCV RNA is amplified by isothermal TMA.[23-25] Blood donations that were HCV EIA 3.0-seropositive (specificity confirmed with RIBA 3.0) were originally tested for HCV RNA with TMA of plasma minipools of 16 or 24 donations, as is customary for all blood donations. Seropositive index donations that were TMA minipool-positive were resolved to a single donor by TMA testing of the individual donations using the residual plasma, and donors were notified of their infection status. Frozen plasma aliquots from seropositive but minipool HCV RNA-negative index donations were subsequently tested in duplicate by individual donation TMA. Both HCV RNA-positive and RNA-negative donors were recalled an average of 2.5 years following their index donations for a blood draw from which both plasma and PBMC were prepared.
HCV RNA plasma viral loads for samples A and B [used to measure the sensitivity of cell-associated transcription-mediated amplification (CA-TMA) and RT-nPCR] and for a subset of 7 blood donor plasma samples were quantified using a commercial polymerase chain reaction (PCR) assay (National Genetic Institute, Los Angeles, CA).
HCV RNA in PBMC Tests.
PBMC were prepared with standard Ficoll Hypaque centrifugation protocols, which included 2 phosphate buffer saline (PBS) rinsing stages prior to cryopreservation with 10% dimethyl sulfoxide. Five million cryopreserved PBMC from HCV-seropositive donors were later thawed at 37ºC and again rinsed twice with PBS. Two and a half million PBMC were then resuspended in 0.5 mL of PBS before the addition of 0.4 mL of lysis buffer containing detergent and denaturing agents used for TMA assay (GenProbe, Inc., San Diego, CA). HCV RNA molecules in the lysate were then captured and purified with a magnetic bead-bound HCV-specific capture probe and amplified by TMA prior to specific signal detection (Procleix HIV-1/HCV assay, Gen-Probe). CA-TMA signal/cutoff results were then scored with the same cutoff used for the TMA performed using plasma as input (TMA-positive for signal/cutoff > 1.0).
An RT-nPCR assay was also used to test for the presence of HCV RNA in PBMC from seropositive donors. Total RNA from 2.5 million cells was extracted with the RNeasy minikit (Qiagen, Valencia, CA) following the manufacturer's instructions. The homogenization step was performed with QiaShredders (Qiagen). Purified RNA was collected in 30 uL of elution buffer plus 1 uL of Protector RNase inhibitor (Roche). First-strand complementary DNA (cDNA) synthesis was initiated with 10 uL of RNA, 0.5 ug of AS1 primer (ATAGARAARGAGCAACCRGG, HCV H77 position 849-868), 0.5 mM deoxyribonucleotide triphosphate (each), and 200 U of murine leukemia virus reverse transcriptase (Promega, Madison, WI) in a final volume of 25 uL. Amplification by nested polymerase chain reaction (nPCR) was performed with primers S1 (TTGTGGTACTGCCTGATAGGG, HCV H77 position 281-301) and AS1 for the first round and with primers S2a (GTAGACCGTGCAYCATGAGC, HCV H77 position 328-347) and AS2a (ADCCGCAYGTDAGGGTATC, HCV H77 position 711-729) for the second round with 5 uL of cDNA, 25 pmol of each primer, 0.2 mM deoxyribonucleotide triphosphate (each), and 1.5 U of GoTaq DNA polymerase in a 1X colorless reaction buffer (Promega). PCR cycles, identical for both rounds, were as follows: 2.5 minutes at 94ºC, 35 cycles of 20 s at 94ºC, 20 s at 57ºC, and 30 s at 72ºC, and a final extension step of 5 minutes at 72ºC. Nested primers were designed to anneal to highly conserved regions of the HCV core gene and has allowed the amplification of HCV RNA belonging to subtypes 1a, 1b, 2a, 2b, 3a, 4a, and 6o (data not shown). RT-nPCR sensitivity tests were performed following serial dilutions of 2 plasmas of known viral loads (virus A and virus B in Table 1C) and their addition to 5 million PBMC prior to total RNA extraction and RT-nPCR. A final input of 1.55 to 62.5 RNA molecules into the reverse transcription step or, assuming 100% efficiency of HCV cDNA synthesis, 0.31 to 12.5 copies of HCV cDNA in the first-round PCR (only 5/25 uL of total cDNA was used to initiate nPCR) was used to measure sensitivity.
For determination of CA-TMA sensitivity, total RNA was extracted from PBMC with and without added plasma using the RNeasy minikit (Qiagen) as described previously. Sensitivity testing for CA-TMA was performed with serial dilutions of PBMC nucleic acids whose HCV viral loads of 2.4 X 103 to 3.9 X 104 HCV RNA/106 PBMC were measured with quantitative polymerase chain reaction (qPCR). qPCR was done on a COBAS TaqMan 48 analyzer (Roche Diagnostics, Indianapolis, IN) with real-time PCR and TaqMan HCV analyte-specific reagents (Roche Molecular System, Inc., Branchburg, NJ).
Analysis of 2 binomial variables was performed with the SAS computer program (version 9.1) to determine if there was a statistically significant difference between groups. If the normal approximation to the binomial distribution was valid, then Pearson's chi-square test for differences in proportion was used. If the normal approximation to the binomial distribution was not valid, then Fisher's exact test was used.