The purpose of the current study was to better understand how rapidly hepatitis C virus (HCV) clears the bloodstream. As the only ethical way to infect a human liver in a human patient is by transplanting an HCV-uninfected liver into an HCV-infected recipient, we calculated the rates of viral clearance from the bloodstream during liver transplant as a surrogate for viral entry.
Research Article
Rate of hepatitis C viral clearance by human livers in human patients: Liver transplantation modeling primary infection and implications for studying entry inhibition
Michael G. Hughes Jr. , William W. Tucker, Sreelatha Reddy, Michael E. Brier, David Koch, Craig J. McClain, Colleen B. Jonsson, Nobuyuki Matoba, Donghoon Chung
Published: July 21, 2017 https://doi.org/10.1371/journal.pone.0180719
Abstract
To better understand the dynamics of early hepatitis C virus (HCV) infection, we determined how rapidly non-cirrhotic HCV-uninfected liver allografts clear HCV from the circulation of cirrhotic HCV-infected patients at the time of transplantation but before administration of immunosuppression. Specifically, we characterized serum HCV kinetics during the first 90 min of reperfusion for 19 chronically HCV-infected patients transplanted with an HCV-uninfected liver by measuring serum viral load immediately prior to reperfusion (t = 0) and then every 15 min for a total of 90 min (t = 90). Immunosuppression was withheld until all samples were taken to better model primary infection. During this period, rates of viral clearance varied more than 20-fold with a median rate constant of 0.0357 1/min, range 0.0089–0.2169; half-life (minutes) median 19.4, range 3.2–77.8. The majority of viral clearance occurred within the first 60 min. The amount of blood transfused during this 90-min period (a potential confounding variable of this human liver transplant model of primary infection) accounted for 53% and 59% of k (r = 0.53, p = 0.05) and half-life (r = 0.59, p = 0.03) variability, respectively. No other clinical variables tested (age, allograft weight, and degree of reperfusion injury as assessed by peak postoperative ALT or AST) accounted for the remaining variability (p>0.05).
Conclusion: In a human liver transplant model of primary infection, HCV rapidly clears the bloodstream. With approximately 90% of clearance occurring in the first 90 minutes of reperfusion, studies of HCV entry inhibition could utilize rate of clearance during this early period as an outcome measure.
Conclusion: In a human liver transplant model of primary infection, HCV rapidly clears the bloodstream. With approximately 90% of clearance occurring in the first 90 minutes of reperfusion, studies of HCV entry inhibition could utilize rate of clearance during this early period as an outcome measure.
Discussion Only
View Full Text Article
The purpose of the current study was to better understand how rapidly hepatitis C virus (HCV) clears the bloodstream. As the only ethical way to infect a human liver in a human patient is by transplanting an HCV-uninfected liver into an HCV-infected recipient, we calculated the rates of viral clearance from the bloodstream during liver transplant as a surrogate for viral entry. To more accurately model primary infection, immunosuppression was withheld during the study period. This is a key difference from other published studies on HCV kinetics during liver transplantation.
We hypothesized that the rate of viral clearance (and likely entry) varies widely during initial infection and focused our investigation on the first 90 min of reperfusion to ensure that the dynamics of initial infection were captured. Decreasing serum viral levels over the 90 min of initial reperfusion were considered to represent viral clearance by allografts, as it has been previously demonstrated that serum viral levels dramatically decrease with transition from the anhepatic phase to the reperfusion phase of liver transplant [7, 12]. The current study significantly adds to existing knowledge regarding this transition by 1) evaluating earlier time points and 2) excluding immunosuppression as a confounding variable.
We elected to use an in-house HCV qRT-PCR assay rather than utilize a commercial assay to minimize variability. The internal controls for the commercial assays were qualitative rather than quantitative and therefore could not be used for calibration. For our study, we anticipated that we would need to detect potentially very small changes in viral load as we were going to sample every 15 minutes. We therefore used a quantitative internal control (Armored HIV RNA quant) that could be used to calibrate our HCV levels and reduce variability. Compared with commercially available assays, our variability was lower [13], though we did sacrifice some sensitivity. For this study, we prioritized minimizing variability over maximizing sensitivity.
In our current study of 19 patients who were viremic at the time of reperfusion, the rate of viral clearance by allografts varied more than 20-fold but demonstrated first order kinetics for all patients but one (patient 5). The linearity of the process throughout the 90 min of the study suggests that virus may be cleared at a rate greater than production. Most viral clearance occurred in the first 60 min of reperfusion. Of the tested clinical variables, only blood transfusion could account for any of the observed variability. Patient 5 deviated from log-linear towards the end of the 90-minute period. Though this could be due the newly infected liver supporting rapid viral replication and release back into the bloodstream, this plateau could rather represent the intermediate phase of a biphasic decline consistent with what has been described in other studies [7, 14]. If the minimal intra-cellular delay time is truly 6 hours [14], then this more likely represents a second, extra-hepatic replication compartment that generated a more significant contribution to HCV levels for patient 5 than for the other patients [14]. The duration of our study was too short to calculate the contribution of a second compartment for these patients.
Bleeding and resulting blood transfusion is a likely a confounding variable for this model of primary infection. Virus is lost during bleeding with a reduction in circulating blood volume. Restoring blood volume with transfusion of HCV-uninfected blood and fluids then dilutes out circulating viral levels. However, blood transfusion could not fully account for the observed differences in this study. The amount of blood transfused during this 90-min period accounted for 53% and 59% of k (r = 0.53, p = 0.05) and half-life (r = 0.59, p = 0.03) variability, respectively. It may be that blood transfusion is an imperfect marker for blood loss and volume replacement. Patients transfused just 1 or 2 units of blood were observed to clear virus twice as quickly as those that did not. It seems unlikely that a <500 cc/unit blood transfusion could dilute viral levels to this degree when circulating blood volumes are typically 3–5 liters. Measuring actual blood loss in real-time during reperfusion was not possible as this process is too dynamic. We therefore propose that the intrinsic rate viral clearance would be best estimated by the median rate of clearance for those patients not transfused (0.0287 1/min). However, this still likely overestimates the true rate of viral clearance as it cannot separate out the contribution of degradation or other non-hepatic clearance mechanisms. If we had determined the rate of clearance during the anhepatic phase, then we could have used these data to more accurately determine hepatic clearance. This represents a limitation of our study design.
The study was designed to carefully and prospectively analyze a very discrete period during the transplant event and follow up on the findings of Garcia-Retortillo et al. [7] and Powers et al. [12]. In a study of 20 patients undergoing liver transplantation, Garcia-Retortillo et al. showed that virus decreased with reperfusion despite having a longer half-life (3.4 hours compared with 2.2 hours during the anhepatic phase). This discrepancy is likely because they did not measure any circulating viral levels during the first 4 hours of reperfusion. By evaluating the first 90 min of reperfusion in 15-min intervals, the current study shows that clearance occurs much earlier and faster (t1/2 19.4 min). The half-life of HCV during reperfusion (as calculated by Garcia-Retortillo et al.) was likely longer than the anhepatic phase because viral production by the newly infected allograft was starting to increase relative to clearance.
In a smaller study of 6 patients, Powers et al. (9) sampled serum during reperfusion earlier than Garcia-Retortillo et al. (10). However, the Powers’ study still did not capture the first hours of reperfusion, missing most clearance, and calculated viral clearance by lumping the anhepatic phase within the first 4 hours of reperfusion. The purpose and design of their study was not to evaluate viral clearance, but rather to characterize viral resurgence following initial reperfusion. They calculated a similar half-life during reperfusion (3.4 hours) as Garcia-Retortillo et al., likely because they also started 4 hours after reperfusion. Their graphical depiction of viral levels does show, like Garcia-Retortillo et al., that the largest decrease in viral levels occurred between the last anhepatic and the first reperfusion samples drawn. Therefore, viral clearance during initial reperfusion (as described in the current study) links viral decay during the anhepatic phase and viral resurgence during later reperfusion (as described by these two prior studies). We have included the viral kinetics following the 90-minute reperfusion phase from 4 patients to show that our findings are similar to the other cited studies when measured over the same time period.
It appears that the clearance mechanism was not saturated in any of the patients studied. This was determined by the observation that the concentrations fall in a linear fashion (on the straight line of natural log of HCV concentration over time). It is therefore unlikely that cell surface receptors for the virus (such as CD81 [15], SR-BI [16], Claudin-1 [17] and Occludin [18]) were present at low enough levels to become saturated. With such high titer inoculum relative to other primary infections (e.g. needle stick transmission), it is unlikely other modes of transmission would result in receptor saturation either. During the first 90 min of reperfusion, new viral production was minimal relative to clearance due to the HCV concentration over time not departing from linearity. Therefore, studies of viral evolution during this time period are unlikely to be significantly confounded by production of new mutations [19, 20].
The large variation in rates of viral clearance not fully accounted for by the tested clinical variables must be explained by other factors. As clearance is determined by blood flow and extraction efficiency, variable perfusion pressure during reperfusion will contribute to the observed differences. This is a likely contributor as hemodynamic instability occurs frequently during reperfusion. Furthermore, differences in extraction efficiency could be due to differences in allograft or viral factors. Allografts may differ in HCV receptor density and/or turnover [20, 21] and therefore may vary in susceptibility to infection. In a study of liver transplant recipients in the first year following transplantation, Mensa et al. found that higher concentrations of SR-BI in the allograft correlated with more rapid clearance of HCV in the early reperfusion phase (first 24 hours), whereas higher levels of occludin and claudin-1 correlated with faster rate of viral production in the following week [22]. Alternatively, a study targeting SR-BI showed that an SR-BI antagonist, ITX5061, had no impact on rates of viral clearance in the first 24 hours, but may have limited subsequent viral rebound [23]. Additionally, viral populations have been shown to differ between patients, leading to variable levels of infectivity [19, 20, 24]. The same study of ITX5061 showed that SR-BI blockade limited viral quasispecies evolution at hypervariable region 1 of E2 [23], further supporting that SR-BI may bind HCV at HVR1 [25]. Patients therefore may differ in viral inoculum fitness.
We believe that the data presented herein can provide background to design clinical studies measuring the efficacy of HCV entry inhibitors. Entry inhibitors can play a role during liver transplant by preventing transmission of recipient bloodstream derived HCV into the newly transplanted uninfected liver without damaging vulnerable hepatocytes [8, 10]. Given that secondary non-hepatic compartments for viral production may exist, such inhibitors may need to address these compartments through anti-viral activity, either through an intrinsic mechanism or when given in combination with direct acting anti-viral agents [23]. Rates of viral clearance over the first 90 minutes of reperfusion would be relevant outcome measure for entry inhibitors that target either the virus or host-entry factors. Liver transplantation could be used to study the impact of these inhibitors on viral entry prior to testing them in chronically infected patients.
Conclusion
With most HCV clearance occurring in the first 90 minutes of transplantation, we believe that liver transplantation represents a prime opportunity to study the impact of HCV entry inhibitors.
View Full Text Article......
The purpose of the current study was to better understand how rapidly hepatitis C virus (HCV) clears the bloodstream. As the only ethical way to infect a human liver in a human patient is by transplanting an HCV-uninfected liver into an HCV-infected recipient, we calculated the rates of viral clearance from the bloodstream during liver transplant as a surrogate for viral entry. To more accurately model primary infection, immunosuppression was withheld during the study period. This is a key difference from other published studies on HCV kinetics during liver transplantation.
We hypothesized that the rate of viral clearance (and likely entry) varies widely during initial infection and focused our investigation on the first 90 min of reperfusion to ensure that the dynamics of initial infection were captured. Decreasing serum viral levels over the 90 min of initial reperfusion were considered to represent viral clearance by allografts, as it has been previously demonstrated that serum viral levels dramatically decrease with transition from the anhepatic phase to the reperfusion phase of liver transplant [7, 12]. The current study significantly adds to existing knowledge regarding this transition by 1) evaluating earlier time points and 2) excluding immunosuppression as a confounding variable.
We elected to use an in-house HCV qRT-PCR assay rather than utilize a commercial assay to minimize variability. The internal controls for the commercial assays were qualitative rather than quantitative and therefore could not be used for calibration. For our study, we anticipated that we would need to detect potentially very small changes in viral load as we were going to sample every 15 minutes. We therefore used a quantitative internal control (Armored HIV RNA quant) that could be used to calibrate our HCV levels and reduce variability. Compared with commercially available assays, our variability was lower [13], though we did sacrifice some sensitivity. For this study, we prioritized minimizing variability over maximizing sensitivity.
In our current study of 19 patients who were viremic at the time of reperfusion, the rate of viral clearance by allografts varied more than 20-fold but demonstrated first order kinetics for all patients but one (patient 5). The linearity of the process throughout the 90 min of the study suggests that virus may be cleared at a rate greater than production. Most viral clearance occurred in the first 60 min of reperfusion. Of the tested clinical variables, only blood transfusion could account for any of the observed variability. Patient 5 deviated from log-linear towards the end of the 90-minute period. Though this could be due the newly infected liver supporting rapid viral replication and release back into the bloodstream, this plateau could rather represent the intermediate phase of a biphasic decline consistent with what has been described in other studies [7, 14]. If the minimal intra-cellular delay time is truly 6 hours [14], then this more likely represents a second, extra-hepatic replication compartment that generated a more significant contribution to HCV levels for patient 5 than for the other patients [14]. The duration of our study was too short to calculate the contribution of a second compartment for these patients.
Bleeding and resulting blood transfusion is a likely a confounding variable for this model of primary infection. Virus is lost during bleeding with a reduction in circulating blood volume. Restoring blood volume with transfusion of HCV-uninfected blood and fluids then dilutes out circulating viral levels. However, blood transfusion could not fully account for the observed differences in this study. The amount of blood transfused during this 90-min period accounted for 53% and 59% of k (r = 0.53, p = 0.05) and half-life (r = 0.59, p = 0.03) variability, respectively. It may be that blood transfusion is an imperfect marker for blood loss and volume replacement. Patients transfused just 1 or 2 units of blood were observed to clear virus twice as quickly as those that did not. It seems unlikely that a <500 cc/unit blood transfusion could dilute viral levels to this degree when circulating blood volumes are typically 3–5 liters. Measuring actual blood loss in real-time during reperfusion was not possible as this process is too dynamic. We therefore propose that the intrinsic rate viral clearance would be best estimated by the median rate of clearance for those patients not transfused (0.0287 1/min). However, this still likely overestimates the true rate of viral clearance as it cannot separate out the contribution of degradation or other non-hepatic clearance mechanisms. If we had determined the rate of clearance during the anhepatic phase, then we could have used these data to more accurately determine hepatic clearance. This represents a limitation of our study design.
The study was designed to carefully and prospectively analyze a very discrete period during the transplant event and follow up on the findings of Garcia-Retortillo et al. [7] and Powers et al. [12]. In a study of 20 patients undergoing liver transplantation, Garcia-Retortillo et al. showed that virus decreased with reperfusion despite having a longer half-life (3.4 hours compared with 2.2 hours during the anhepatic phase). This discrepancy is likely because they did not measure any circulating viral levels during the first 4 hours of reperfusion. By evaluating the first 90 min of reperfusion in 15-min intervals, the current study shows that clearance occurs much earlier and faster (t1/2 19.4 min). The half-life of HCV during reperfusion (as calculated by Garcia-Retortillo et al.) was likely longer than the anhepatic phase because viral production by the newly infected allograft was starting to increase relative to clearance.
In a smaller study of 6 patients, Powers et al. (9) sampled serum during reperfusion earlier than Garcia-Retortillo et al. (10). However, the Powers’ study still did not capture the first hours of reperfusion, missing most clearance, and calculated viral clearance by lumping the anhepatic phase within the first 4 hours of reperfusion. The purpose and design of their study was not to evaluate viral clearance, but rather to characterize viral resurgence following initial reperfusion. They calculated a similar half-life during reperfusion (3.4 hours) as Garcia-Retortillo et al., likely because they also started 4 hours after reperfusion. Their graphical depiction of viral levels does show, like Garcia-Retortillo et al., that the largest decrease in viral levels occurred between the last anhepatic and the first reperfusion samples drawn. Therefore, viral clearance during initial reperfusion (as described in the current study) links viral decay during the anhepatic phase and viral resurgence during later reperfusion (as described by these two prior studies). We have included the viral kinetics following the 90-minute reperfusion phase from 4 patients to show that our findings are similar to the other cited studies when measured over the same time period.
It appears that the clearance mechanism was not saturated in any of the patients studied. This was determined by the observation that the concentrations fall in a linear fashion (on the straight line of natural log of HCV concentration over time). It is therefore unlikely that cell surface receptors for the virus (such as CD81 [15], SR-BI [16], Claudin-1 [17] and Occludin [18]) were present at low enough levels to become saturated. With such high titer inoculum relative to other primary infections (e.g. needle stick transmission), it is unlikely other modes of transmission would result in receptor saturation either. During the first 90 min of reperfusion, new viral production was minimal relative to clearance due to the HCV concentration over time not departing from linearity. Therefore, studies of viral evolution during this time period are unlikely to be significantly confounded by production of new mutations [19, 20].
The large variation in rates of viral clearance not fully accounted for by the tested clinical variables must be explained by other factors. As clearance is determined by blood flow and extraction efficiency, variable perfusion pressure during reperfusion will contribute to the observed differences. This is a likely contributor as hemodynamic instability occurs frequently during reperfusion. Furthermore, differences in extraction efficiency could be due to differences in allograft or viral factors. Allografts may differ in HCV receptor density and/or turnover [20, 21] and therefore may vary in susceptibility to infection. In a study of liver transplant recipients in the first year following transplantation, Mensa et al. found that higher concentrations of SR-BI in the allograft correlated with more rapid clearance of HCV in the early reperfusion phase (first 24 hours), whereas higher levels of occludin and claudin-1 correlated with faster rate of viral production in the following week [22]. Alternatively, a study targeting SR-BI showed that an SR-BI antagonist, ITX5061, had no impact on rates of viral clearance in the first 24 hours, but may have limited subsequent viral rebound [23]. Additionally, viral populations have been shown to differ between patients, leading to variable levels of infectivity [19, 20, 24]. The same study of ITX5061 showed that SR-BI blockade limited viral quasispecies evolution at hypervariable region 1 of E2 [23], further supporting that SR-BI may bind HCV at HVR1 [25]. Patients therefore may differ in viral inoculum fitness.
We believe that the data presented herein can provide background to design clinical studies measuring the efficacy of HCV entry inhibitors. Entry inhibitors can play a role during liver transplant by preventing transmission of recipient bloodstream derived HCV into the newly transplanted uninfected liver without damaging vulnerable hepatocytes [8, 10]. Given that secondary non-hepatic compartments for viral production may exist, such inhibitors may need to address these compartments through anti-viral activity, either through an intrinsic mechanism or when given in combination with direct acting anti-viral agents [23]. Rates of viral clearance over the first 90 minutes of reperfusion would be relevant outcome measure for entry inhibitors that target either the virus or host-entry factors. Liver transplantation could be used to study the impact of these inhibitors on viral entry prior to testing them in chronically infected patients.
Conclusion
With most HCV clearance occurring in the first 90 minutes of transplantation, we believe that liver transplantation represents a prime opportunity to study the impact of HCV entry inhibitors.
View Full Text Article......
No comments:
Post a Comment