In spite of the fact that the thymidine analogues zidovudine and stavudine are never co-administered, Becher et al.  recently found measurable intracellular concentrations of the active triphosphate form of stavudine in 26 out of 31 HIV-infected patients under zidovudine treatment, thus suggesting some degree of intracellular transformation of zidovudine into the triphosphate form of stavudine. No stavudine was detected in the plasma of the same individuals by immunoassay .
In our routine practice of therapeutic drug monitoring (TDM), plasma concentrations of antiretroviral drugs are measured by means of a solid-phase extraction high performance liquid chromatography method , allowing the simultaneous detection of all licensed antiretroviral agents.
Over a 6-month period, 65 patients under antiretroviral regimens all including zidovudine (300 mg twice a day) underwent TDM. Measurable concentrations of zidovudine (limit of quantification 10 ng/ml) were found in 52 patients (80%), whereas no zidovudine was found in 13 individuals. Among the former, zidovudine peak concentrations (Cmax; median time of sampling 1 h, range 0.5–2 h) and trough concentrations (Ctrough; median time of sampling 12 h, range 10–13 h) were determined in 11 (median Cmax 522 ng/ml, range 235.5–857.5) and 49 patients (median Ctrough 52.5 ng/ml, range 10–387.5), respectively. In the same plasma samples of 31 out of 52 patients, stavudine concentrations above the limit of quantification (10 ng/ml) were measured (median 20 ng/ml, range 10–910), with a median stavudine : zidovudine ratio of 0.13 (0.01–30.5) and no correlation between zidovudine and stavudine concentrations (linear regression).
Over the same study period, 38 patients under antiretroviral treatment not including zidovudine or stavudine underwent TDM, and no measurable plasma concentrations of either drug were found in any case, whereas stavudine concentrations were regularly detected in 30 out of 36 stavudine recipients tested in the same period with the same method.
In our zidovudine recipients, no statistically significant association was found between stavudine concentrations and the following variables: sex, age, CD4 cell count (median 429.5 cells/μl), HIV-RNA copies/ml (< 50 in 73% of patients), concomitant intake of non-antiretroviral drugs, nucleoside reverse transcriptase inhibitors (NRTI) other than zidovudine, non-nucleoside reverse transcriptase inhibitors or protease inhibitors, previous stavudine intake, hepatitis B virus or hepatitis C virus co-infection, liver cirrhosis, alanine aminotransferase increase and time of blood sampling (Cmax or Ctrough). The presence of detectable stavudine concentrations was found to be associated with zidovudine administration started less than 12 months before sampling (χ2 test, P = 0.018), and inversely related to the total duration of antiretroviral therapy (Pearson's r = −0.30; P = 0.05).
Our findings provide evidence that this phenomenon spreads out the intracellular compartment. The reason as to why in the study by Becher et al.  stavudine was not found in the plasma might lie in the different method the authors employed for assessing stavudine concentrations in plasma (enzyme immunoassay instead of high performance liquid chromatography). However, the best analytical technique to assess stavudine and zidovudine plasma and intracellular concentrations is still unclear, and further investigations using different highly sensitive methods are warranted.
We observed a high variability in stavudine plasma concentrations and only 59.6% of zidovudine takers showed detectable stavudine concentrations. The wide inter-subject variability is also in accordance with a multimodal distribution of the stavudine : zidovudine ratio. At one extreme of such distribution three subjects were found to have stavudine concentrations even higher than zidovudine concentrations (ratios 1.28, 6.47 and 30.5). The association we found between the presence of detectable stavudine concentrations and a shorter duration of both zidovudine and total antiretroviral therapy suggests that the rate of zidovudine–stavudine transformation might depend upon individual factors that are going to be counterbalanced by some late developing compensatory mechanisms.
It is worth noting that the stavudine plasma concentrations measured here were not negligible. Among the 26 patients who had detectable stavudine Ctrough, 16 (61.5%) had stavudine concentrations above 16 ng/ml, which is the median Ctrough reported in patients taking effective stavudine-containing antiretroviral regimens . In the same period, the median Ctrough we measured in 30 stavudine recipients was 18.5 (range 10–49).
These findings raise some question about the current interpretation of both the reciprocal selection of resistance by thymidine analogues  and the attributable responsibility of these drugs in some NRTI-associated untoward effects . Further studies in this area are needed to clarify whether a suboptimal exposure to stavudine during zidovudine therapy selects for thymidine-associated mutations in spite of adequate zidovudine concentrations, and if stavudine-associated side-effects occur in zidovudine takers whose rate of zidovudine–stavudine transformation is particularly high.
Although the category of NRTI has been the first to appear in this therapeutic setting, there seems still to be much to investigate in order to make the most appropriate clinical use of these invaluable therapeutic weapons.
1. Becher F, Pruvost AG, Schlemmer DD, Creminon CA, Goujard CM, Delfraissy JF, et al
. Significant levels of intracellular stavudine triphosphate are found in HIV-infected zidovudine-treated patients.AIDS
2. Simon VA, Thiam MD, Lipford LC. Determination of serum levels of thirteen human immunodeficiency virus-suppressing drugs by high-performance liquid chromatography.J Chromatogr
3. Legrand M, Goujard C, Mentrè F, Diquet B, Taburet AM, and the Cophart-ANRS102 Study Group. Nucleoside reverse transcriptase inhibitors plasma concentrations in patients with sustained virological response on HAART.
In: 4th International Workshop on Clinical Pharmacology of HIV Therapy
. Cannes, 27–29 March 2003 [Abstract 30:P5.5].
4. Shulman NS, Machekankano RA, Shafer RW, Winters MA, Zolopa AR, Liou SH, et al
. Genotypic correlates of a virologic response to stavudine after zidovudine monotherapy.J Acquired Immune Defic Syndr
5. Cote HC, Brumme ZL, Craib KJ, Alexander CS, Wynhoven B, Ting L, et al
. Changes in mitochondrial DNA as a marker of nucleoside toxicity in HIV-infected patients.N Engl J Med