Lack of effect of moderate hepatic impairment on the pharmacokinetics of oral oseltamivir and its metabolite oseltamivir carboxylate


Paul Snell, PhD, Department of Clinical Pharmacology, Roche Products Ltd, 40 Broadwater Road, Welwyn Garden City, Hertfordshire AL73AY, UK.
Tel: + 44 (0)17 0736 5766
Fax: + 44 (0)17 0736 5887



To compare the pharmacokinetics of oseltamivir and oseltamivir carboxylate in hepatically impaired patients and healthy subjects.


Hepatically impaired patients (n = 11) and healthy subjects (n = 11) were individually paired on the basis of gender, age (±10 years) and body weight (±20%) and administered a single dose of oseltamivir (75 mg).


Oseltamivir and oseltamivir carboxylate Cmax were ≤6% and ≤19% lower, and their AUC(0,∞) 33% higher and ≤19% lower, respectively, in hepatically impaired patients compared with healthy subjects. These changes are within the safety limits for the drug.


The metabolism of oseltamivir is not compromised minimally in hepatically impaired patients. No dose adjustment is required in these patients when receiving oseltamivir.


Oseltamivir has been approved for the treatment (75 mg twice daily) and prevention (75 mg once daily) of influenza A and B infection in adolescents and adults, and treatment in children ≥1 year of age (30–75 mg twice daily according to body weight). Oseltamivir is rapidly absorbed from the gastrointestinal tract and undergoes rapid and extensive presystemic conversion by high capacity hepatic carboxylesterases [1, 2]. Hepatic impairment is known to alter drug absorption and disposition. For example, liver cirrhosis can result in changes in hepatic architecture and the formation of intra- or extra-hepatic shunts [3], providing routes for blood and drugs to bypass hepatic metabolism. Based on in vitro models and mass balance studies in animals and humans, the potential for oseltamivir pharmacokinetics to be affected by hepatic disease is low.

This study investigated the effect of hepatic impairment on the pharmacokinetics of oseltamivir and its active carboxyl metabolite in order to determine whether dose adjustments should be recommended in such patients.


Study subjects

This was a multicentre, open-label, parallel-group study, which was performed at the request of regulatory authorities. Male or female subjects aged 18–65 years with (n = 11) or without (n = 11) stable moderate hepatic impairment satisfied the inclusion criteria. They were paired by age (±10 years), gender and body weight (±20%), prior to enrolment. All subjects had a creatinine clearance greater than 60 ml min−1. The degree of hepatic dysfunction was assessed using the Child-Pugh score [4]. All liver function tests and urinalysis results were required to be within typical ranges for that patient group.

This study was performed in full compliance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines. Written informed consent was obtained from each subject prior to enrolment. Approval was obtained from the relevant ethics committees (Canterbury Ethics Committee, Christchurch, New Zealand [Dr Robson]; Investigational Review Board Inc, Florida, USA [Dr Galitz]; Bayerische Landesarztekammer Ethik-Kommision, Munich, Germany [Dr Weil]).

Study procedures

Subjects fasted from midnight before dosing until oseltamivir was administered (on the following morning). For pharmacokinetic analysis, blood samples were collected from each subject at predose and up to 48 h postdose. Urine was collected within 30 min prior to dosing and up to 24 h after dosing. A follow-up medical examination was performed.

Sample analysis

Plasma and urine samples were analyzed by HPLC/MS/MS [5] for oseltamivir and the carboxylate (BAS Analytics, Kenilworth, UK). The limits of quantification for oseltamivir and the carboxylate were 1.0 and 10.0 ng ml−1, respectively, for plasma and 5.0 and 30.0 ng ml−1 for urine.

Data analysis

Pharmacokinetic parameters for oseltamivir and oseltamivir carboxylate were determined by standard noncompartmental methods using WinNonlin® (version 3.1, Pharsight Corporation, Mountain View, CA, USA). AUC(0,∞) (area under the curve) and Cmax (maximum observed plasma concentration) were the primary pharmacokinetic parameters.

No power calculation was performed. A sample size of 11 hepatically impaired patients is considered generally acceptable for this type of study [4]. Analysis of variance (using SAS for Windows, version 8.2) was performed using a mixed model including terms for subject and group (healthy, impaired) and treating pairing as a random effect. Confidence intervals (CI, 90%) on the difference between least squares means (LSM) were calculated and used to test for equivalence [4].


Eleven pairs of hepatically impaired patients and healthy subjects received a single oral dose (75 mg) of oseltamivir. Most subjects were Caucasian (>90%) and male (82%). All the patients had cirrhosis, (seven alcohol induced). The mean BMI (27 [range 21–36]vs 26 [20–35 kg m−2]) and estimated serum creatinine clearance (101 [68–303]vs 95 [68–139 ml min−1]) were evenly matched between groups.

For oseltamivir, the ratio of the LSM (90% CI) for Cmax was 6% lower in the hepatically impaired patients relative to the matched healthy subjects (Table 1). However, the mean AUC(0,∞) of oseltamivir was 33% higher. The mean times to tmax (maximum concentration) and t1/2 (terminal half-life) were comparable between the groups, but the apparent oral clearance of oseltamivir (CL/F) was lower in hepatically impaired patients. The percent of drug excreted in the urine (Ucum (0−24 h)) and renal clearance (CLR (0−24 h)) were similar the between groups.

Table 1.  Mean (SD) values and 90% confidence intervals (CI) of pharmacokinetic parameters of oseltamivir (OSE) and oseltamivir carboxylate (OC) for paired hepatically impaired patients and healthy volunteers
 Hepatically impaired (n = 11)Healthy volunteers(n = 11)Ratio (90% CI)a(%)
  1. NA = Not applicable; AUC = area under the curve. aRatios were calculated from the log transformed least squares means (LSM); bOnly the evaluable subjects with 24 h urine collections were included; cIncludes dose adjustment (OSE dose × ratio of molecular-weights of OC [284] to that of OSE [312]). *90% CI of the ratio of the least squares mean falls within 70–143% CI range for equivalence.

Cmax (ng ml−1)100 (89)260 (83)95 (61)315 (93)94 (64, 140) 81 (61, 107)
AUC(0,∞) (ng ml−1 h)210 (133)3100 (1250)141 (57)3310 (577)133 (94, 187) 88 (73, 106*)
tmax (h)0.8 (0.5)4.5 (1.6)0.6 (0.2)4.0 (1.2)110 (78, 156)111 (84, 146)
t1/2 (h)1.7 (0.4)7.3 (2.5)2.3 (1.8)6.6 (1.4)83 (66, 104)109 (83, 143)
CL/F (ml min−1)8830 (5670)421 (153)c10700 (5530)353 (55)c75 (53, 106)113 (94, 136)
OC : OSE AUC(0,∞) ratio22 (15)c 31 (17)c NA 
Ucum (0−24 h) (%)b5.3 (4.1)53.0 (20)c4.3 (1.0)66.0 (7.8)c79 (39, 159) 75 (58, 96)
n = 10n = 10n = 10n = 10  
CLR (0−24 h) (ml min−1)b372 (262)244 (122)398 (115)255 (38)67 (35, 127) 82 (59, 116)
n = 10n = 10n = 10n = 10  

For oseltamivir carboxylate the ratio of the LSM (90% CI) for Cmax and AUC(0,∞) was ≤ 19% lower in hepatically impaired patients compared with the healthy subjects. The mean tmax, t1/2, oral clearance, CLR(0−24 h) and Ucum (0−24 h) were comparable between groups.

The metabolite : parent ratios were slightly lower in the hepatically impaired patients compared with the healthy subjects (Table 1). The parent moiety did not accumulate in the presence of low metabolite concentrations in the patients, demonstrating that hepatic conversion of oseltamivir remains intact. However, the plasma concentration of oseltamivir in these patients with liver disease tended to be higher and more variable than that in healthy subjects.

Oseltamivir was well tolerated by all patients and healthy subjects. Only six adverse events occurred, affecting three hepatically impaired patients (nausea [n = 3], diarrhoea [n = 1], nasopharyngitis [n = 1], arthralgia [n = 1]) and one healthy volunteer (nausea [n = 1]). All cases of nausea were considered to be either possibly related or unrelated to the study medication. There were no deaths or serious adverse events during the study and no subject was withdrawn because of an adverse event.


The results of the present study confirm previous in vitro data (Roche unpublished data), that hepatic carboxylesterase capacity in patients with moderate hepatic disease is sufficient to metabolize oseltamivir to its active antiviral metabolite. The resultant carboxylate metabolite concentrations were consistent in the presence of highly variable oseltamivir concentrations. Thus, whereas first pass metabolism of the parent compound can vary, the metabolite can still be formed due to rapid hepatic recirculation. This may also explain the second peak in the plasma vs concentration time profile following the major oseltamivir peak observed in hepatic patients (Figure 1). Mean exposure to oseltamivir was marginally increased in hepatically impaired subjects, probably as a result of intra- and/or extra-hepatic shunting and reduced first pass metabolism. The slightly lower metabolite to parent ratios for hepatic patients compared to their matched controls could suggest the possible presence of such shunts in some or all of the patients in our study. Probably as a consequence of these phenomena, the metabolite : parent ratio for CL/F was greater in hepatic patients than healthy subjects, whereas that for Ucum(0−24 h) and CLR(0−24 h) was lower (Table 1).

Figure 1.

Mean oseltamivir and oseltamivir carboxylate plasma concentration-time profiles in paired hepatically impaired patients and healthy volunteers. Inset shows log transformed data with SD. OSE HI (•), OSE HV (○), OC HI (◆), OC HV (◊)

In hepatically impaired patients, plasma concentrations of oseltamivir carboxylate rose rapidly, exceeded the effective antiviral concentration (∼100 ng ml−1) at the same time as in healthy subjects, and remained sufficiently high to provide antiviral activity throughout the 12 h dose interval. This further demonstrates the efficiency of metabolic conversion of oseltamivir to oseltamivir carboxylate in the presence of hepatic impairment. The Cmax and AUC(0,∞) of oseltamivir in isolated patients with hepatic impairment may be higher than expected, but will not cause drug accumulation because of its short half-life.

Although oseltamivir concentrations varied in patients with hepatic impairment, values for oseltamivir and oseltamivir carboxylate Cmax and AUC(0,∞) fell well within the therapeutic margin of safety already documented for this drug [6]. Oseltamivir carboxylate concentrations were less variable and also fell within the  efficacious  range.  These  findings  were  supported by  the  good  tolerability  of  oseltamivir  in  these  high-risk patients with liver disease who had various underlying diseases and were taking a number of other medications.

In conclusion, this study demonstrates that in hepatically impaired patients, oseltamivir (75 µg, od) is adequately metabolised, well tolerated and should result in no loss of clinical benefit. Therefore, no dose adjustment is required in patients with mild and moderate hepatic impairment receiving oseltamivir for either treatment or prevention of influenza.

We would like to thank the staff of the following three study centres for their contribution to this study; Christchurch Clinical Studies Trust, Christchurch, New Zealand; South Florida Bioavailability Clinic, Miami, FL. 331–3405 USA; Apex Research GMbH, D-81241, Munich, Germany.

We would also like to thank the following current and former Roche staff for their valued contribution to this study; Dr Charles Oo (Aventis Pharmaceuticals, Bridgewater, NJ, USA), Dr Baolian Liu and Dr Al Dorr (Hoffmann-La Roche Inc, Nutley, NJ, USA), Dr Miranda Norton (Bio Products Laboratory, Elstree, Herts, UK), Dr Nicole England, Mr Darren Martin, Mrs Tracy Simkins, Dr Ian Small and Mr Rob Doody (all Roche Products Ltd, Welwyn Garden City, Herts, UK).