We read the study by Dandri et al.1 with much interest. The authors sought to estimate hepatitis B virus (HBV) half-life based on several baseline serum and intrahepatic parameters measured from 80 untreated patients with chronic HBV, including intrahepatic HBV-DNA–containing capsids and serum HBV DNA levels. A mathematical model2 that couples baseline viremia and intracellular virion productivity was applied by Dandri et al. and showed a strong inverse correlation between baseline viremia and HBV half-life. In addition, they estimated that virion clearance is very fast (median half-life [t1/2] 46 minutes [interquartile range, 4 minutes to 3.7 hours] in patients positive for hepatitis B e antigen (HBeAg) and 2.5 minutes [interquartile range, 24 seconds to 13 minutes] in HBeAg-negative patients). More striking is their finding (shown in their figure 2) that the HBV t1/2 varies more than 6 orders of magnitude among patients, with t1/2 being about 0.01 minutes when HBV DNA levels are 103 copies/mL or lower. For simian immunodeficiency virus, the rate of clearance in the presence and absence of high titer neutralizing antibodies differs by a factor of about four,3 leaving unexplained how a 6 order of magnitude increase in HBV clearance could occur in infected patients with low baseline viral loads. Further, when infection is initiated, HBV DNA levels should be low and clearance with a t1/2 of the order of seconds might preclude infection from taking hold. Estimated HBV half-life versus baseline HBV DNA. Free HBV half-life was estimated via mathematical modeling from the first phase HBV DNA decline during therapy in HBeAg-positive patients (circles or triangles) and HBeAg-negative patients (squares). Values were copied from Lewin et al.,7 Colombatto et al.,8 Mihm et al.,9 and Wolters et al.10 Patient 13 from Lewin et al.,7 with HBV DNA = 5.7 × 10 5 copies/mL and HBV half-life = 91 hours, is not shown but was included in the statistical analysis that shows no significant correlation (P = 0.5) between baseline HBV DNA and virion half-life (SPSS version 15; SPSS Inc., Chicago, IL). A number of previous studies have analyzed the rate of HBV virion clearance by using antiviral drugs to inhibit viral production and have come to decidedly different conclusions. Drug perturbation analysis showed no correlation between virion half-life and baseline viremia (Fig. 1), and estimated significantly slower virion clearance (median t1/2 ≈21 hours and ≈9 hours in treated HBeAg-positive and HBeAg-negative patients, respectively (Fig. 1) with less than one order of magnitude variation among patients; see also table 3 by Sypsa et al.4). The discrepancy between the findings of Dandri et al.1 and drug perturbation analysis (Fig. 1) is striking. To resolve the discrepancy, we suggest other approaches such as (1) infusion of viral particles5 and (2) plasma apheresis.6 For simian immunodeficiency virus the clearance rate was estimated from the disappearance of viral particles from plasma after intravenous bolus injection (or constant rate infusion) into rhesus macaques.5 This approach seems feasible to apply in chimpanzees. For both human immunodeficiency virus (type 1) and hepatitis C virus clearance rates were measured using plasma apheresis in which viral clearance was temporarily increased while viral production was unaffected.6 Interestingly, estimates of the hepatitis C virus half-life of 2-3 hours from plasma apheresis6 is similar to drug pertubation analysis estimates.11 This approach seems feasible to apply in patients with chronic HBV as well. This research was performed under the auspices of the U.S. Department of Energy under contract DE-AC52-06NA25396 and supported by National Institutes of Health (NIH) grants RR06555, AI28433, and AI065256 (to A.S.P.). H.D. is supported by the University of Illinois Gastrointestinal and Liver Disease (GILD) Association. Harel Dahari*, Scott J. Cotler*, Thomas J. Layden*, Alan S. Perelson , * Department of Medicine, University of Illinois at Chicago, Chicago, IL, Theoretical Biology and Biophysics, Los Alamos National Laboratory, Los Alamos, NM.