Pharmacodynamics and Viral Kinetics during Albinterferon Alfa-2b Therapy of Chronic Hepatitis C Genotype 1 Treatment-Naïve Patients

Individual patient pharmacodynamic parameters. Individual patient changes in HCV RNA and corresponding alb-IFN serum concentrations in the 900 µg and 1200 µg treatment groups are shown in Figure 2. Different viral kinetic patterns were observed among patients in both dose groups in the context of similar pharmacokinetic profiles. Nonlinear fitting of the combined pharmacokinetic and viral kinetic data allowed the estimation of individual pharmacodynamic parameters for each patient (Table 3 and Table 4).

Table 3. Individual Viral Kinetic and Pharmacodynamic Parameters for Each Patient

Table 4. Individual Normalized Pharmacokinetic/Pharmacodynamic Parameters for Each Patient

Figure 2A. Nonlinear fitting of viral load and albinterferon alfa-2b (alb-IFN) concentration data by a pharmacodynamic model for individual patients. Nonlinear fitting (solid lines; see Methods, Eq 1–Eq 6) allowed estimation of the viral dynamic and pharmacodynamic parameters loss rate constant d, 90% effective concentration, and second-order sensitivity to changes in alb-IFN levels. PK, pharmacokinetic; VK, viral kinetic. Open red circles, outlier single data points that were excluded by the fitting procedure (Jackknife algorithm).

Figure 2B. Nonlinear fitting of viral load and albinterferon alfa-2b (alb-IFN) concentration data by a pharmacodynamic model for individual patients. Nonlinear fitting (solid lines; see Methods, Eq 1–Eq 6) allowed estimation of the viral dynamic and pharmacodynamic parameters loss rate constant d, 90% effective concentration, and second-order sensitivity to changes in alb-IFN levels. PK, pharmacokinetic; VK, viral kinetic. Open red circles, outlier single data points that were excluded by the fitting procedure (Jackknife algorithm).

Pharmacodynamic parameters were similar in the 900 µg and 1200 µg groups (Table 2). In all patients, the median alb-IFN EC₉₀ was 4.3 log₁₀ pg/mL; median MQ was 2.6, and median IQ was 1.1. Drug levels were > EC₉₀ for a median of 13.6 days (25%, 75% quartiles: 2.8, 13.9), preventing rebound in most patients. The median N_hill coefficient (ie, the second-order sensitivity to changes in alb-IFN levels) was low (1.0), which resulted in sustained levels of viral suppression (> EC₉₀) even in the context of declining alb-IFN levels at the end of Week 2. Thus, at Day 14, the end of the first dosing interval, IQ was > 1 in 54% (14/26) of patients. After the second dose, drug accumulation resulted in an increased MQ at Day 18 and increased IQ at Day 28 in 81% (21/26) and 77% (20/26), respectively, of patients (Table 4). Furthermore, 4 patients maintained blocking effectiveness > EC₉₀ even as drug levels further declined between Day 28 and Day 42.

Predictors of Day-28 virologic response. Individual patient pharmacokinetic and pharmacodynamic parameters as a function of alb-IFN dose and virologic response (= 2-log₁₀ decline at Week 4) are shown in Figure 3. The concentration of alb-IFN at Day 4 (C_max) did not correlate with virologic response. Lower EC₉₀ (P = not significant), higher MQ (P < .001), higher IQ (P < .001), and greater number of days that alb-IFN remained > EC₉₀ (P < .001) were, however, associated with virologic response. In addition, a rapid loss rate of infected cells (d) was significantly associated with virologic response (P = .002) and correlated with the magnitude of viral decline at Day 28 (R = 0.76; P < .001).

Figure 3. Pharmacokinetic and pharmacodynamic parameters for individual patients. These parameters were given as a function of albinterferon alfa-2b (alb-IFN) dose and virologic response (> 2-log₁₀ decline at Week 4). A, no difference in alb-IFN maximum concentration (C_max) was observed between any of the groups. B, 90% effective concentration (EC₉₀) was lower (P = not significant) for patients who achieved virologic response. C, infected cell loss rate d was significantly predictive (P = .002) of virologic response. The combined pharmacokinetic/pharmacodynamic parameters of days of interferon level > EC₉₀ during the first week (D), maximal antiviral quotient (MQ; C_max/EC₉₀; E), and inhibitory quotient (IQ; C[d14]/EC₉₀; F) were significantly predictive (P < .001) of therapeutic response. C, concentration.

The interactions between parameters that were predictive of virologic response and specific predictive values of these parameters were also examined (Figure 4). As expected, there was a strong correlation between MQ and first-phase viral decline (R = 0.6; P < .001; Figure 4A). Virologic response was predicted by MQ > 3 (positive predictive value [PPV] = 100%, P = .007; Figure 4B) and HCV-RNA reduction > 1 log₁₀ IU/mL at Day 4 (PPV = 94.4%, P < .001; Figure 4C). Patients with MQ < 3 did not achieve virologic response unless their loss rate of infected cells was sufficiently rapid, ie, d > 0.1 (negative predictive value = 100%, P < .001; Figure 4B). High blocking effectiveness—e(t) > 90%—at the end of the dosing interval (IQ > 1) was highly predictive of virologic response (PPV = 100%, P < .001; Figure 4C and 4D). High IQ was associated with high initial virologic response (HCV-RNA reduction > 1 log₁₀ IU/mL at Day 4; Figure 4C), high MQ (data not shown), and low N_hill coefficient (Figure 4D).

Figure 4. Interactions between albinterferon alfa-2b (alb-IFN) pharmacodynamic parameters predictive of virologic response. A, maximal antiviral quotient (MQ) drives the first-phase viral decline and, therefore, is naturally correlated with it. B, a high MQ (> 3) was predictive of a virologic response at Day 28 (positive predictive value = 100%), whereas a low MQ (< 3) together with a low loss rate of infected cells (d < 0.1), assumed to be mostly host dependent, negatively predicted virologic response (negative predictive value = 100%). C, greater (> 1 log₁₀) viral decline at Day 4 was predictive of virologic response at Week 4 (positive predictive value = 95%) and was associated with an inhibitory quotient (IQ) > 1 at Day 14, and an IQ > 1 was predictive of virologic response at Day 28. D, a low second-order sensitivity to changes in alb-IFN levels (N_hill coefficient) was associated with an IQ > 1 at Day 14. C, concentration; C_max, maximum concentration; EC₉₀, 90% effective concentration; HCV, hepatitis C virus.