Management of drug-to-drug interactions between cyclosporine A and the protease-inhibitor lopinavir/ritonavir in liver-transplanted HIV-infected patients



Highly active antiretroviral therapy (HAART) has improved the life expectancy of HIV-infected patients, allowing orthotopic liver transplantation as a reasonable treatment option for selected patients with terminal liver disease. Both non-nucleoside reverse transcriptase inhibitors and protease inhibitors, key elements of HAART, give rise to substantial drug-to-drug interactions with immunosuppressive drugs such as tacrolimus and cyclosporine A. After studying 12-hour pharmacokinetic profiles in 3 HIV-positive patients after liver transplantation, we describe how dosing of cyclosporine A can be adjusted to maintain effective immunosuppressive drug levels on a daily dosing schedule when ritonavir-boosted indinavir or lopinavir-based antiretroviral therapy is given. To avoid toxic drug levels, we used an orally available cyclosporine A formulation prepared from the commercial available intravenous solution, which enabled dose adjustments in 1-mg increments. Under ritonavir-boosted HAART, cyclosporine A levels showed markedly altered absorption/elimination characteristics with more or less constant blood-levels throughout the dosing interval and prolonged elimination half-lives up to 38 hours. To obtain equivalent areas under the curve of cyclosporine A, daily doses were reduced to 5–20% of the individual standard doses given before initiation of ritonavir-boosted HAART. Because of the flat absorption/elimination profiles under ritonavir-boosted HAART cyclosporine A, dosing could be reliably monitored long term by measuring cyclosporine A trough-levels. (Liver Transpl 2004;10:939–944.)

Highly active antiretroviral therapy (HAART) of HIV infection effectively prevents opportunistic diseases and has greatly enhanced overall life expectancy in HIV-positive patients.1, 2 However, several large cohort studies have shown that morbidity caused by chronic hepatitis B and C has become a leading cause for death in HIV-infected patients with concomitant chronic viral hepatitis.3, 4 Thus, orthotopic liver transplantation may be a rational therapeutic option for selected HIV-infected patients with concomitant liver disease.5–11 Non-nucleoside reverse transcriptase inhibitors and protease inhibitors are the key elements of HAART, but they show substantial drug-to-drug interactions via interaction with cytochrome P450 enzymes and p-glycoprotein, both in the liver and gut. Likewise, cyclosporine A and tacrolimus, the backbone of immunosuppression, have great potential for drug-to-drug interactions via the same metabolic pathways. In this investigation, we studied 12-hour pharmacokinetic profiles in 3 HIV-positive patients after liver transplantation to characterize the interaction between antiretroviral therapy containing ritonavir-boosted indinavir and lopinavir and cyclosporine A. On the basis of these results, we suggest new guidelines for immunosuppression with cyclosporine A in HIV-positive patients whose HIV infection must be controlled by ritonavir-boosted HAART after liver transplantation.


HAART, highly active antiretroviral therapy; b.i.d., twice daily; AUC, area under the curve; OLT, orthotopic liver transplantation; HIV, human immunodeficiency virus.


Two HIV-positive patients transplanted for end-stage chronic hepatitis C (patients 1 and 2) and 1 HIV-positive patient transplanted for fulminant liver failure caused by acute hepatitis B (patient 3) were included in this study. Criteria for liver transplantation were a CD4-cell count greater than 150 cells/μL, the absence of any opportunistic disease, and efficacious control of HIV-infection by HAART. The patients were transplanted at our center in 2002 and 2003; baseline demographic data are summarized in Table 1. All patients provided written informed consent, and the study was conducted in full agreement with the declaration of Helsinki and its subsequent revisions.

Table 1. Clinical Characteristics of the Patients in the Study
PatientGenderAge [y]Duration of HIV Infection at OLTRisk Factor For HIVIndication  for OLTHAARTCD4 CellsHIV-RNA
BeforeAfter (Time of Start)Before1 Year After§Before1 Year After#
  • Abbreviations: Age, age at the time of transplantation; OLT, orthotopic liver transplantation; ESLD, end-stage liver disease; HCV, hepatitis C infection; ALF, acute liver failure; HBV, hepatitis B infection; AZT, zidovudine; ABC, abacavir; EFV, efavirenz; TDF, tenofovir; LPV/r, boosted lopinavir; d4T, stavudine; 3TC, lamivudine; HCC, hepatocellular carcinoma.

  • *

    HIV-infection was newly diagnosed during the routine work-up before transplantation.

  • Switched to TDF, ABC, and boosted indinavir week 12 after OLT.

  • 5 months after OLT.

  • §


  • #


1male4417HemophiliaESLD due to chronic HCVAZT, ABC, EFVTDF, 3TC, LPV/r; (week 3)18731518,265< 50
2male520*Not knownALF due to acute HBVABC, 3TC, LPV/r; (week 6)223534> 500,000< 50
3male5918HemophiliaESLD and HCC due to chronic HCV3TC, d4TTDF, ABC, LPV/r (week 8)16389681< 50

Standard immunosuppression consisting of cyclosporine A and prednisolone for induction was used in all patients. Cyclosporine A was dosed twice daily. Doses were adjusted individually to match target trough levels of 150–200 ng/mL, 100–150 ng/mL, and 75–125 ng/mL at weeks 0–2, weeks 3–11, and after week 11 after liver transplantation, respectively. To enable dose adjustments in 1-mg increments, a cyclosporine A formulation was prepared by mixing 1 mL cyclosporine A solution for intravenous use (1 mL = 50 mg cyclosporine A) with 93 mL 5% (weight/volume) glucose and 6 mL 50% (weight/volume) magnesium sulfate solution. One milliliter of the resulting solution contained 0.5 mg cyclosporine A and 12 mmol magnesium. Pharmaceutical stability up to 36 hours has been demonstrated for this preparation.12

Twelve-hour pharmacokinetic profiles of cyclosporine A whole blood levels were measured in patients 2 and 3 before and in all patients 3–6 weeks after initiation of HAART with ritonavir-boosted HAART. In all patients, lopinavir was given at a dose of 400 mg b.i.d. together with 100 mg ritonavir b.i.d. In patients 1 and 2, measurements of 12-hour pharmacological profiles were repeated 1 year after initiation of ritonavir-boosted HAART.

Because of drug intolerance, the HAART regimen in patient 3 was switched to ritonavir-boosted indinavir, given at a dose of 800 mg b.i.d. A 12-hour pharmacokinetic profile was measured in this patient 6 weeks after changing the HAART regimen. Because of high blood levels, indinavir was reduced to 400 mg b.i.d. 2 weeks after measuring the 12-hour pharmacokinetic profile.

Lopinavir and indinavir trough levels were determined in serum samples using high-performance liquid chromatography.13 Cyclosporine A levels were measured in whole-blood samples via a commercially available immunoassay (Dimension, Dade Behring Inc., Newark, Del).

On the day of the pharmacokinetic studies, patients were allowed to eat a light breakfast, and they received a single oral dose of cyclosporine A at 8:00 AM. Blood samples were taken at baseline and 30, 60, 90, 120, 180, 240, 300, 360, 480, 600, and 720 minutes after taking the cyclosporine A medication. Measured cyclosporine A blood levels were analyzed by the TOPFIT Version 2.0 software14 on a personal computer system to characterize pharmacokinetic profiles. Areas under the curve (AUCs) were determined for the 12-hour observation intervals using the trapezoidal rule. Terminal half-lives were calculated by log-linear regression analysis of the last data points without model assumption (noncompartmental analysis). Additionally, several models, including a 4-compartment, 4-segment absorption model, were used to fit the data.14


All patients are currently alive 22 months (patient 1), 15 months (patient 2), and 8 months (patient 3) after liver transplantation, respectively. None of the patients has experienced any acute or chronic rejection episode during this observation period. Likewise, no hepatotoxicity from HAART or of other cause was observed. Cotrimoxazole 480 mg once daily was added to the standard care, in case CD4-cells were less than 200/μL, to prevent opportunistic infections. Anti-HBs-immunoglobulins, in addition to lamivudine as part of his HAART, successfully prevented reinfection with hepatitis B in patient 2. Hepatitis C reinfection occurred in the 2 patients in whom liver transplantation had to be done because of chronic hepatitis C. Reinfection took an accelerated course of disease in patient 1, requiring early initiation of pegylated interferon ribavirin combination treatment.

Boosted lopinavir serum trough levels were consistently greater than the targeted limit of 1000 ng/mL in all patients. When patient 3 was switched to a ritonavir-boosted indinavir regimen because of lopinavir associated gastrointestinal intolerance, his indinavir trough levels were consistently greater than 100 ng/mL, even after the indinavir dose had been reduced to 400 mg b.i.d.

Before initiation of HAART, daily cyclosporine doses varied from 75–150 mg b.i.d. to keep cyclosporine A blood levels in the desired ranges. Twelve-hour pharmacokinetic profiles performed in patients 2 and 3 before initiation of HAART showed a typical cyclosporine A absorption and elimination profile, with Tmax of approximately 1 hour and terminal half-lives of 4–6 hours, as reported previously for HIV-negative patients (Fig. 1a and b, Table 2).15

Figure 1.

(A) Cyclosporine A whole blood 12-hour concentration profile with and without ritonavir-boosted lopinavir (LPV/r) in patient 2. Cyclosporine A blood levels were determined without HAART (cyclosporine A was dosed 75 mg b.i.d.) and when the patient was taking HAART (cyclosporine A was dosed 15 mg b.i.d.).Values are shown together with the intra-assay variation coefficient as error bars. (B) Cyclosporine A whole blood 12-hour concentration profile with and without ritonavir-boosted indinavir (IDV/r) in patient 3. Cyclosporine A blood levels were determined without HAART (cyclosporine A dosed 150 mg b.i.d.) and when the patient was taking HAART (cyclosporine A dosed 35 mg in the morning and 25 mg in the evening). Values are shown together with the intra-assay variation coefficient as error bars.

Table 2. Pharmacokinetic Parameters of Cyclosporine A With and Without Concomitant Ritonavir-Boosted Lopinavir or Indinavir
PatientAfter OLT [d] Time of PKAfter start HAART [d] Time of PKCYA Dose* [mg]Weight [kg]Protease InhibitorCtrough [ng/mL]Cmax [ng/mL]tmax [h]AUC0–12 [ng∗h/mL]Terminal Half Life [h]R2
  • Abbreviations: PK, pharmacokinetic; OLT, orthotopic liver transplantation; HAART, highly active antiretroviral therapy; CYA, cyclosporine A; AUC, area under the curve; R24 comp. 4 segment, squared correlation coefficient of the model fit using a 4-compartment/4-segment absorption model; LPV/r, lopinavir 400 mg/ritonavir 100 mg bid; IDV/r, indinavir 800 mg/ritonavir 100 mg bid.

  • *

    CYA dose, morning-evening dose, 12-hour dosing interval.

  • Ctrough, concentration of cyclosporine A 12 hours after the last oral dose.

  • Cmax, maximum whole blood concentration of cyclosporine A.

1654210-5 LPV/r1262172212638,40,992
 5935707-7 LPV/r1161561156816,70,995
239–475-75 875601,525854,20,998
 4233807,5-7,5 LPV/r961144124119,40,997
319–35150-150 74393120125,60,998
 1055135-25 IDV/r113168514786,50,998

HAART was started after consistent cyclosporine A trough levels within the target range had been reached and serum aminotranferases were recovering and less than 200 IU/mL (37°C). After the start of HAART, patient 1 continued with his previous cyclosporine A standard dose of 150 mg b.i.d., resulting in toxic trough levels of more than 900 ng/mL. To maintain cyclosporine A blood trough levels within the desired therapeutic range of 75–125 ng/mL, dose reductions to 5% of the original dose were necessary (Fig. 2). Likewise, cyclosporine A doses had to be reduced to 5% and 20% of the original doses in patients 3 and 2, respectively, when ritonavir-boosted lopinavir containing HAART was added. When patient 3 was switched to ritonavir-boosted indinavir, cyclosporine A levels could be maintained in the therapeutic range still using 20% of his previous standard dose without HAART.

Figure 2.

Cyclosporine A whole-blood trough-levels and cyclosporine A doses during the first 1½ years after orthotopic liver transplant in patient 1.

Cyclosporine A 12-hour pharmacokinetic profiles showed a fundamentally altered shape of absorption/elimination kinetic under both ritonavir-boosted lopinavir and indinavir, respectively (Fig. 1a and b). Cyclosporine A blood levels showed a rather flat absorption/elimination kinetic, with little variation over the 12-hour observation period. Time to maximum blood concentration and, in particular, terminal half lives of cyclosporine A elimination were markedly delayed, and maximum blood concentration was conspicuously reduced (Table 2). However, AUCs remained unaltered when cyclosporine A doses were adjusted according to measured cyclosporine A blood trough levels. Targeting cyclosporine A trough levels into the conventional 75–125 ng/mL range enabled safe and efficient long-term immunosuppression for our patients taking ritonavir-boosted HAART (Fig. 2).


The long-term success of liver transplantation in HIV-positive patients critically depends on the antiretroviral efficacy of HAART. Ritonavir-boosted protease inhibitors have demonstrated high efficacy in pretreated and naive patients and are commonly used agents in antiretroviral therapy. Good tolerability, in particular with regard to hepatotoxicity, make ritonavir-boosted protease inhibitors preferable as part of HAART in the orthotopic liver transplantation setting. Ritonavir blocks p-glycoprotein and cytochrome P450 in the gut and liver, both of which are critically involved in the metabolism of cyclosporine A and tacrolimus16, 17 and many other commonly used drugs such as antiarrhythmics, antibiotics, analgesics, sedatives, and others.18 Thus, dosing of these immunosuppressants must be adjusted in patients concomitantly taking ritonavir-boosted regimens. Prolongation of dosing has been proposed for tacrolimus,19–22 but this strategy may be associated with an increased risk of toxicity because higher loading doses are needed to maintain a sufficient trough level throughout the dosing interval. In this article, we describe how an orally available preparation of cyclosporine A can be successfully applied to adjust cyclosporine A doses within the milligram range, enabling stable immunosuppressive therapy by 12-hour dosing intervals. To strengthen our rationale, 12-hour single-dose pharmacokinetic profiles were measured before and after initiation of HAART.

The observed pharmacokinetic profiles revealed that the curve characteristic of cyclosporine A absorption/elimination was fundamentally altered, resulting in unprecedented stable cyclosporine A blood levels under ritonavir-boosted lopinavir and indinavir-containing HAART. Elimination of cyclosporine A was markedly delayed in all of our patients because of the profound inhibition of cytochrome P450 and p-glycoprotein by the protease inhibitors. The prolonged elimination necessitated a profound reduction of the daily dose of cyclosporine A, which could not be dosed appropriately by commercially available tablets or solution. Thus, a new formulation was prepared using the commercially available solution for intravenous application. It is known that cyclosporine A is less well absorbed from our solution than from the newer commercially available microemulsion,23 which may have helped us to adjust the cyclosporine A dosing in the presence of markedly inhibited elimination. However, cyclosporine A absorption also may vary in a temporal discontinuous fashion, and this assumption could at least in part explain the observed fluctuations of cyclosporine A concentrations. Moreover, recirculation of cyclosporine via the enterohepatic pathway may have contributed to these variations.24 When using various pharmacokinetic models to fit the observational data, the highest correlation coefficient was obtained for a 4 compartment/4 absorption segment model (Table 2), suggesting that in the presence of markedly reduced elimination marginal variations in absorption become detectable and influence the pharmacokinetic profile.

Rejection-free survival is correlated to sufficient AUCs of cyclosporine A. Blood levels measured 2 hours after intake of cyclosporine A have been proposed as the best surrogate markers of AUC, which are difficult to measure in clinical routine.15 The observed pharmacokinetic profiles under HAART illustrate that 2-hour blood concentrations as generally recommended for cyclosporine A monitoring should not be applied in patients receiving ritonavir-boosted HAART. This approach is likely to considerably underestimate AUC and may result in significant toxicity. In contrast, cyclosporine A trough levels appear to be more reliable surrogate markers to assess the AUC in this situation. Monitoring of cyclosporine A therapy via drug trough levels enabled safe and rejection-free immunosuppression over prolonged treatment periods at 5 to 20% of the original doses. Conversely, coadministration of cyclosporine A to ritonavir boosted lopinavir or indinavir did not significantly alter lopinavir or indinavir trough levels (data not shown).


We have demonstrated that cyclosporine A–based immunosuppression is feasible in HIV-positive organ recipients receiving ritonavir boosted protease inhibitors as part of their HAART. However, cyclosporine A doses must be adjusted individually, and dose reductions down to 5% of the standard dose may be necessary to maintain cyclosporine A levels within the target range.