Importance of Collection Tube during Clinical Studies of Oseltamivir

Abstract

Ex vivo conversion of the anti-influenza drug oseltamivir to its active metabolite can be inhibited by the esterase inhibitor dichlorvos or by using commercial fluoride-oxalate tubes. Oseltamivir and its active metabolite remain intact in plasma samples during a proposed virus heat inactivation step: incubation at 60°C for 45 min.

The threat of a human H5N1 pandemic has focused attention on the control and treatment of influenza (7). Currently, the only reliably effective and orally bioavailable antiviral agent is the neuraminidase inhibitor oseltamivir (11), which has been extensively studied for the prevention and treatment of epidemic seasonal influenza and is being stocked for mass administration. Information on its use in severe influenza remains limited, however (2, 4), and research efforts are under way to optimize dosing by assessing its pharmacokinetic and pharmacodynamic properties.

Oseltamivir (OP) is an ethyl ester prodrug hydrolyzed in vivo by high-capacity hepatic carboxyesterases to the active metabolite oseltamivir carboxylate (OC) (2, 8). We previously showed that human plasma esterases can also cause variable and often extensive ex vivo conversion of OP in blood and plasma samples under conditions likely to be encountered during clinical studies and during assay preparation (5). The esterase inhibitor dichlorvos prevented this conversion, but the use of organophosphate compounds in the laboratory raises safety issues. We therefore investigated whether collection of clinical samples in any commercially available tube would be as effective in preventing ex vivo conversion of OP into OC. We also investigated whether OP and OC would remain intact during a proposed virus inactivation procedure (i.e., heating plasma at 60°C for 30 to 90 min). When handling biological samples, there is always a risk of exposure to virus present in the sample, and viremia with H5N1 infection is higher than in other influenza infections. Most viruses are thermolabile, and avian influenza virus is inactivated by heating at 60°C for 30 min (6, 9, 10; http://www.oie.int).

The extent of ex vivo hydrolysis in various collection tubes was studied in plasma obtained from a single healthy volunteer, the individual with the highest esterase activity in our previous report (5). Heparinized plasma (2 ml; previously stored frozen at −86°C for 6 months) was transferred into duplicate sets, each consisting of an EDTA tube, a fluoride-oxalate tube, an Li-heparin tube (all from Teklab) and a Li-heparin tube containing the esterase inhibitor dichorvos (200 μg/ml of plasma). OP in water (20 μl) was then added to each tube to produce a final concentration of 200 ng/ml of plasma. The samples were briefly vortexed, and one set was placed in the refrigerator (about 8°C), and one set was kept at room temperature (about 25°C). Triplicate 100-μl plasma samples from each tube were removed at 16 h, and the concentrations of OP and OC were determined by liquid chromatography-mass spectrometry and solid-phase extraction (12). Heat inactivation was studied using plasma obtained from human blood drawn into fluoride-oxalate tubes. Triplicate sample sets of plasma containing 3 and 300 ng of OP/ml and 30 and 4,000 ng of OC/ml were prepared. One set was incubated for 45 min at 60°C, one set was incubated for 90 min at 60°C, and one set was kept at ambient temperature as a reference. Three aliquots (100 μl) of each sample were analyzed as described above.

Significant temperature-dependent ex vivo conversion of OP to OC occurred in the Li-heparin and EDTA tubes, reaching 64 and 41%, respectively, at room temperature and 30 and 11%, respectively, at 8°C (Table 1). In the presence of dichlorvos or in fluoride-oxalate tubes, conversion reached only 4 and 3%, respectively, at room temperature with no significant difference observed between storage at 8°C or room temperature for 16 h. In addition, the concentrations of the OC metabolite were less than 5 ng/ml (i.e., less than the lower limit of quantification) in fluoride-oxalate tubes or in the presence of dichlorvos at both temperatures (Table 1). Only a small nonsignificant decrease (<3%) in the concentrations of OP after incubation at 60°C for 90 min was observed (P > 0.05; Table 2). The results for all time points and all levels also remained well inside the stability criteria (i.e., ±15%) suggested by the U.S. Food and Drug Administration (1) for routine bioanalysis. OP and OC can thus be considered stable for at least 90 min at 60°C (Table 2). The characteristics of the plasma samples are altered with time when heated. The samples gradually become more viscous, which could lead to difficulties in handling them (e.g., aspiration of a correct volume with the pipette). However, heating for up to 45 min was possible without any noticeable impact on the characteristics of the matrix.

TABLE 1.

Degradation of OP in different collection tubes

Treatment group (200 ng OP/ml added)a	Degradation results (SD)b
	Refrigerator (about 8°C) for 16 h			Ambient temp (about 25°C) for 16 h
	Mean amt of OP found (ng/ml)	% OP degradation	Mean amt of OC found (ng/ml)	Mean amt of OP found (ng/ml)	% OP degradation	Mean amt of OC found (ng/ml)
Heparin	141 (3)	30	48 (0.5)	72 (1.3)	64	111 (3)
EDTA	178 (4)	11	7* (0.2)	119 (0.6)	41	72 (0.8)
Fluoride-oxalate	190 (4)	5	2* (0.1)	192 (4)	4	5* (0.3)
Dichlorvos	193 (2)	4	-†	194 (6)	3	2* (0.1)

^an = 3 replicates per experiment.

^bThe percent OP degradation was calculated as follows: (added OP − found OP)/added OP. *, Below the lower limit of quantification (10 ng/ml); †, below the limit of detection (2 ng/ml).

TABLE 2.

Stability of OP and OC in fluoride-oxalate plasma during heat inactivation at 60°C^a

Treatmentb	OP			OC
	Amt added (ng/ml)	Mean amt found (ng/ml)	% Deviation	Amt added (ng/ml)	Mean amt found (ng/ml)	% Deviation
No heat inactivation	3	3.02 (0.05)	0.5	30	30.5 (0.5)	1.7
	300	301 (15)	0.3	4,000	4,000 (160)	0
Heat inactivation at about	3	3.17 (0.09)	5.7	30	30.8 (1.9)	2.8
60°C for 45 min	300	297 (3)	−1.1	4,000	3,930 (30)	−1.7
Heat inactivation at about	3	2.97 (0.08)	−1.1	30	30.9 (1.7)	3.1
60°C for 90 min	300	292 (2)	−2.7	4,000	4040 (20)	0.9

^aThe percent OP degradation was calculated as follows: (added OP − found OP)/added OP.

^bn = 3 replicates per experiment.

We conclude that use of fluoride-oxalate collection tubes is as effective as the use of the esterase inhibitor dichlorvos in preventing ex vivo conversion of OP in plasma samples, irrespective of the ambient temperature. The use of either should assure the highest-quality measurements in clinical studies of the pharmacokinetics of OP. However, the use of commercial fluoride-oxalate tubes is logistically simpler and raises no safety concerns, making it the preferred option for such studies in the future. We further conclude that a virus heat inactivation step can be included for routine analysis of samples from avian influenza patients without risk of compromising sample integrity.