Concomitant treatment with lamivudine renders cladribine inactive by inhibition of its phosphorylation

Authors


In the management of cancer patients with previous or chronic hepatitis B infection, anti-viral agents, such as lamivudine (2-deoxy-3′-thiacytidine; 3-TC), are increasingly co-administered with antitumoural agents in order to prevent reactivation of hepatitis B that may sometimes take a severe course (Kohrt et al, 2006). Clinically relevant drug interactions may occur when two nucleoside drugs that both require activation by phosphorylation are co-administered, as demonstrated by the following patient and the subsequent in vitro studies.

Case report

A 51-year-old patient with a B-cell chronic lymphocytic leukaemia (B-CLL) was treated with cladribine (2-chloro-2-deoxyadenosine; CdA) because of progressive anaemia and a decreasing platelet count. At the same time he was started on prophylactic 3-TC, 100 mg/d, to prevent reactivation of previous hepatitis B. Unexpectedly, no decrease in the lymphocyte count occurred in the weeks following the first 5-d course of subcutaneous CdA at a dose of 0·1 mg/kg per day (Fig 1).

Figure 1.

 Time course of the lymphocyte count during therapy with CdA and 3-TC. The dashed bars indicate the duration of each chemotherapy cycle with CdA, the stippled bar the duration of 3-TC prophylaxis.

An interaction between 3-TC and CdA was suspected and 3-TC was stopped. Another 5-d cycle of CdA was initiated 2 weeks after stopping 3-TC at the same dose and under the same conditions as the previous one. This time, the lymphocyte count decreased to 25·24 × 109/l after 11 d and reached a nadir of 22·04 × 109/l after 1 month. After a third course of CdA the lymphocyte count dropped further to 10·77 × 109/l after 11 d, reaching a nadir of 9·24 × 109/l 2 months later. During this period, hepatitis B virus-DNA remained undetectable.

In order to investigate the mechanism underlying the suspected interaction we investigated the phosphorylation of the two drugs in vitro. Peripheral blood mononuclear cells (PBMC) were isolated from the blood of a healthy volunteer by centrifugation on a Ficoll gradient. The conditions for the CdA-phosphorylation assay essentially followed published methods (Arner et al, 1992). Isolated PBMCs were lysed by three freezing and thawing cycles and, after centrifugation, the supernatant cell lysate was used as a source of deoxycytidine kinase for the phosphorylation assay. [8-3H]-CdA (specific activity: 259 GBq/mmol, Moravek Biochemicals, Brea, California, USA) was incubated at a final concentration of 25 μmol/l in 50 mmol/l Tris-HCl buffer (pH 7·6) with 5 mmol/l ATP, 10 mmol/l sodium fluoride, 4 mmol/l dithiothreitol and 5 mmol/l MgCl2 with increasing concentrations of 3-TC (Biotrend Chemicals, Zurich, Switzerland, 10 μmol/l to 1 mmol/l) for 30 min at 37°C. The reaction was started by the addition of 10 μl of the cell lysate to the mixture. Aliquots of 10 μl of the incubation mixture were loaded at different time points on Whatman DE-81 filter paper (Whatman, Maidstone, UK), which adsorbs phosphorylated but not unphosphorylated CdA (Ives & Wang, 1978). The filters were washed three times for 5 min in 5 mmol/l ammonium formate to remove unphosphorylated CdA. Phosphorylated CdA retained on the filter was eluted with 250 μl of 0·4 mol/l perchloric acid and the radioactivity was measured by liquid scintillation spectrophotometry. The radioactivity was corrected for unspecific binding and expressed as a percentage of the binding in the absence of inhibitory 3-TC. Mean values (±SD) of three incubations of 30 min are reported.

In control experiments with CdA as the only substrate, a linear increase in the generation of phosphorylated metabolites of CdA (R2 = 0·995) was observed over a 30 min incubation period. When 3-TC was added to the incubations the linearity of the reaction (R2 = 0·999) was preserved, but there was a concentration-dependent inhibition of the phosphorylation of CdA. At 10 μmol/l 3-TC the generation of phosphorylated CdA metabolites decreased by 21% (±1·6 SD). At equimolar concentrations (25 μmol/l) of CdA and 3-TC, CdA phosphorylation was inhibited by 26% (±3·0 SD) and at a ratio of CdA:3-TC of 1:40 (1 mmol/l 3-TC) phosphorylation decreased to 15% (±1·8 SD) of control values (Fig 2).

Figure 2.

 Concentration-dependent inhibition of the formation of phosphorylated CdA metabolites by 3-TC. Mean ± SD of three incubations.

2-Chloro-2-deoxyadenosine, a purine analogue effective in treating lymphoproliferative disorders (Beutler, 1992), is a prodrug that requires activation by intracellular phosphorylation to exert its cytotoxic effects. The initial intracellular phosphorylation of CdA to its monophosphate form (CdAMP) is mainly catalyzed by deoxycytidine kinase (dCK). Phosphorylation allows the intracellular retention of CdAMP which is then further phosphorylated. The cytotoxicity mainly depends on the accumulation of the active triphosphate form (Arner et al, 1994). The initial phosphorylation step is shared by other nucleoside analogues, such as the anti-viral agent 3-TC, a reverse transcriptase inhibitor that is also activated by intracellular phosphorylation (Kewn et al, 2000). As highlighted by the present case, 3-TC may render CdA inactive. The most likely explanation is an inhibition of the phosphorylation of CdA by 3-TC, as demonstrated by the in vitro experiments. Therapeutic plasma concentrations of 3-TC and CdA are in the range of 1–5 μmol/l and 10–100 nmol/l, respectively (Albertioni et al, 1998; Johnson et al, 1999). Thus, the chosen ratios of CdA:3-TC of up to 1:40 leading to a markedly decreased activation of CdA are clinically relevant. As the three courses of CdA were administered under identical conditions and subcutaneous CdA has a reproducible bioavailability (Sonderegger et al, 2000) the plasma concentrations of CdA may be expected to have been similar during the three treatment cycles. A genetically determined resistance to CdA cannot explain the lack of initial response in our patient either as he responded to CdA after stopping 3-TC. An additional interaction between the two compounds at the level of cellular uptake cannot be ruled out. The extent to which CdA affects the metabolism and effectiveness of 3-TC cannot be deduced from the present patient who remained HBV-DNA negative with and without 3-TC prophylaxis.

This report highlights the clinical relevance of the interaction between CdA and 3-TC. The in vitro experiments showed that 3-TC inhibited the production of phosphorylated metabolites of CdA in a concentration-dependent manner. Clinicians should be aware that they may compromise the therapeutic benefit of CdA treatment by prophylactic treatment of hepatitis B with 3-TC.

Conflict-of-interest disclosure

The authors declare no competing financial interests.

Ancillary