American Journal of Transplantation 2003; 3: 1576–1580
`Blackwell Munksgaard
`
`Copyright C(cid:2) Blackwell Munksgaard 2003
`doi: 10.1046/j.1600-6135.2003.00242.x
`
`Influence of Delayed Initiation of Cyclosporine on
`Everolimus Pharmacokinetics in de Novo Renal
`Transplant Patients
`
`John M. Kovarika,∗, Jacques Dantalb, Giovanni
`Civatic, Gaetano Rizzod, Marisel Rouillya, Olga
`Bettoni-Ristica and Christiane Rordorfa
`
`aNovartis Pharmaceuticals, Basel, Switzerland
`bH ˆopital Hotel Dieu, Nantes, France
`cOspedale Niguarda Ca’Granda, Milan, Italy
`dOspedali Riuniti Santa Chiara, Pisa, Italy
`∗Corresponding author: John M. Kovarik,
`john.kovarik@pharma.novartis.com
`
`We quantified the influence of delayed initiation of
`cyclosporine on everolimus pharmacokinetics in or-
`der to provide dosing guidance for kidney transplant
`patients.
`
`In a randomized multicenter study, 56 de novo kidney
`transplant patients received everolimus, basiliximab,
`corticosteroids and either immediate (n == 40) or de-
`layed (n == 16) initiation of cyclosporine based on renal
`function. Everolimus and cyclosporine predose blood
`levels (Cmin) were obtained over the first 3 months
`post-transplant.
`
`Everolimus Cmin averaged 9–11 ng/mL in the imme-
`diate cyclosporine group over the first 3 months. In
`the delayed cyclosporine group, average everolimus
`Cmins were significantly lower by 2.9-fold in the ab-
`sence vs. presence of cyclosporine: 2.9 ±± 2.8 vs.
`8.3 ±± 3.7 ng/mL (p < 0.001). Likewise, the within-
`patient
`ratio of everolimus Cmins in the pres-
`ence/absence of cyclosporine averaged 2.9 (range,
`0.7–5.6).
`
`Both everolimus and cyclosporine blood concentra-
`tions need to be monitored in kidney transplant pa-
`tients with delayed graft function during the period
`when cyclosporine is withheld and shortly after its
`initiation. Dosing of everolimus needs to be adjusted
`to take into account an average threefold increase in
`everolimus exposure when cyclosporine is added to
`the regimen.
`
`Key words:
`everolimus,
`tation
`
`drug interactions,
`Cyclosporine,
`immunosuppression, kidney transplan-
`
`Received 12 May 2003, revised 3 July 2003 and ac-
`cepted for publication 10 July 2003
`
`1576
`
`Introduction
`
`Everolimus is a macrolide immunosuppressant which in-
`hibits T-lymphocyte proliferation signaling in response to
`organ allografting (1). Everolimus is intended for use in
`combination with cyclosporine and has demonstrated sim-
`ilar efficacy to mycophenolate mofetil in phase 3 kidney
`transplant trials (2) and superior efficacy to azathioprine in
`a phase 3 heart transplant trial (3). Both everolimus and
`cyclosporine are extensively metabolized via cytochrome
`CYP3A with elimination of metabolites via the bile. In ad-
`dition, everolimus is a substrate of the countertransporter
`P-glycoprotein, while cyclosporine is both a substrate and
`an inhibitor. Because of these shared pathways, there is the
`potential for drug–drug interactions between the agents.
`
`to the influence of everolimus on cy-
`With respect
`closporine pharmacokinetics, a crossover study in main-
`tenance kidney transplant patients did not identify any sig-
`nificant change in cyclosporine steady-state profiles upon
`the addition of everolimus to the regimen for 1 month
`(4). Over the first year after kidney transplantation, the
`blinded phase 3 trials noted that 10% lower cyclosporine
`doses were used in everolimus-treated patients to achieve
`similar cyclosporine blood concentrations as in mycophe-
`nolate mofetil-treated patients (5). Taken together, these
`data indicate that there is a minor influence of everolimus
`on cyclosporine, which would likely be compensated for
`in the context of routine cyclosporine therapeutic drug
`monitoring.
`
`Uncovering an influence of cyclosporine on everolimus,
`however, has been more difficult inasmuch as everolimus
`was always administered in a combined regimen with cy-
`closporine in drug development trials. The pharmacokinetic
`data from the phase 3 studies demonstrated that as cy-
`closporine blood levels were down-titrated over the first
`year post-transplant, everolimus blood levels remained sta-
`ble (5,6). Only in the case of a crossover healthy sub-
`ject study could the complete absence of cyclosporine on
`everolimus be quantified. That study demonstrated that
`the AUC of everolimus after a single 2-mg dose given
`alone increased 2.7-fold when a single 175-mg dose of cy-
`closporine (Neoral, Novartis) was coadministered (7). To-
`gether these studies suggest that as cyclosporine expo-
`sure is varied within the normal therapeutic blood level
`range, there is no differential
`influence on everolimus
`
`Ex. 1048-0001
`
`

`
`exposure. However if cyclosporine is removed from the
`regimen, a change in everolimus blood levels may occur.
`
`An opportunity to test this hypothesis at steady state in
`kidney transplant patients came in the context of a devel-
`opment study in patients at increased risk of developing
`delayed kidney graft function in whom cyclosporine ther-
`apy was withheld in the early days after transplantation and
`initiated later when renal function recovered. We compared
`everolimus and cyclosporine dosing histories and blood
`concentrations over the first 3 months post-transplant in
`patients with immediate and delayed cyclosporine in order
`to quantify the influence of cyclosporine on everolimus and
`to provide practical guidance on the use of this drug com-
`bination under these clinical conditions.
`
`Methods
`
`Study design
`This was a 1-year, multicenter, randomized, open-label study in de novo
`kidney transplant recipients at increased risk of delayed graft function. The
`study protocol was reviewed at Ethics Committees at each clinical center
`and patients gave written informed consent to participate in the trial. The
`clinical results of the study will be reported separately. This communication
`presents the pharmacokinetic results from the first 3 months of the trial.
`
`Patients received an immunosuppressive regimen consisting of everolimus
`(Certican, Novartis Pharmaceuticals, Basel, Switzerland), basiliximab
`(Simulect, Novartis), corticosteroids, and cyclosporine (Neoral, Novartis).
`Everolimus was administered orally 1.5 mg twice daily. Basiliximab was
`given by intravenous bolus injection of 20 mg before surgery and again on
`day 4 after transplantation. Prednisone or equivalent was dosed according
`to local center practice. Cyclosporine was administered orally twice daily
`simultaneously with everolimus. Cyclosporine doses were individualized to
`achieve predose trough concentrations of 150–350 ng/mL in weeks 1–2,
`150–300 ng/mL in weeks 3–4, 100–250 ng/mL in months 2–6, and 100–
`200 ng/mL in months 7–12.
`
`Patients were assigned to three groups based on post-transplant renal func-
`tion, as follows: Patients with well-established renal function within 24 h of
`transplantation, as measured by urine output >2 L, were not randomized
`but remained in the study in a stand-alone arm and received the same treat-
`ment as patients in the immediate cyclosporine group. Patients with urine
`output <2 L in thefirst 24 h were considered to be at increased risk of devel-
`oping delayed graft function and were randomized on a 1 : 1 basis to receive
`either early or delayed initiation of cyclosporine. In the early cyclosporine
`group, cyclosporine was initiated not later than 48 h after transplantation
`with a half-dose regimen allowed up to day 7 for patients who subsequently
`developed delayed graft function. In the delayed cyclosporine group, cy-
`closporine was initiated when renal function recovered (serum creatinine
`<3.4 mg/dL) or by day 14 at the latest. If by day 14, renal function had not
`adequately recovered, cyclosporine was initiated at half the conventional
`dose and increased when function recovered.
`
`Cyclosporine–Everolimus Drug Interaction
`
`Bioanalytics
`Bioanalytics were performed at a central laboratory. Everolimus blood con-
`centrations were determined with an enzyme-linked immunosorbent as-
`say (ELISA) after extraction with tert-butylmethylether. Assay performance
`was assessed by a five-point quality control concentration range from 2
`to 80 ng/mL. The interassay coefficient of variation ranged from 9.5% to
`38.3% and the bias from –8.5% to 16.6%. The lower limit of quantifica-
`tion was 2 ng/mL. Cyclosporine blood concentrations were determined
`by a radioimmunoassay (RIA) method using the specific reagents of the
`Incstar Cyclo-trac SP-whole blood radioimmunoassay kit (Diasorin, Stillwa-
`ter, MN) according to the manufacturer’s instruction manual. The assay in-
`volves methanol extraction, incubation with radiolabeled cyclosporine and
`a cyclosporine-specific mouse monoclonal antibody, and a double-antibody
`separation technique. The calibration and quality control blood samples
`were prepared by spiking drug-free human blood with cyclosporine instead
`of using the samples supplied with the kit. The four-point quality control
`concentration range was 50–1800 ng/mL. The interassay coefficient of vari-
`ation ranged from 13.1% to 30.4% and the bias from –8.0% to –4.5%. The
`lower limit of quantification was 50 ng/mL.
`
`Statistical analysis
`Data are presented as mean ± standard deviation unless otherwise noted.
`Everolimus blood levels were compared in the absence and presence of
`cyclosporine by an independent t-test for groups with unequal variances. A
`p-value of 0.05 was taken as significant.
`
`Results
`
`Study population and evaluation groups
`Evaluable concentration data were obtained from 56 pa-
`tients at 13 clinical centers. Initially the pharmacokinetic
`data were evaluated based on the three treatment as-
`signments: not randomized, early initiation of cyclosporine,
`and delayed initiation of cyclosporine. The first two groups
`yielded similar results, indicating that the decisive factor
`to take into account was the timing of cyclosporine initi-
`ation. Therefore, for the purposes of this clinical pharma-
`cology evaluation, patients were categorized into two eval-
`uation groups according to the day of cyclosporine initia-
`tion. Those who began cyclosporine therapy within 2 days
`of the start of everolimus were designated the immedi-
`ate cyclosporine group; hence, this group included patients
`in the nonrandomized and early cyclosporine arms of the
`study. Patients who began cyclosporine 3 days or later af-
`ter the start of everolimus were designated the delayed
`cyclosporine group.
`
`There were 40 patients in the immediate cyclosporine
`group, 30 men and 10 women, aged 55 ± 11 years and
`weighing 71 ± 13 kg. The delayed cyclosporine group was
`demographically similar with 16 patients, 11 men and five
`women, aged 54 ± 13 years and weighing 77 ± 12 kg.
`
`Pharmacokinetic assessments
`Venous blood samples for the determination of everolimus and cyclosporine
`whole blood concentrations were obtained before the morning drug doses
`on day 2; at least three times weekly in the first 2 weeks; and at protocol-
`scheduled clinic visits in months 1 and 3. Blood samples were drawn into
`EDTA-containing collection tubes, and the tubes were gently inverted sev-
`eral times, and frozen at –20 ◦C.
`
`Drug dosing
`Cyclosporine was initiated in the immediate cyclosporine
`group by day 2 in all patients. Doses were 330 ±
`154 mg/day in the first 2 weeks and 295 ± 133 mg/day at
`month 1. By month 3 the conventional dose-reduction after
`the de novo period was evident with an average dose of
`
`American Journal of Transplantation 2003; 3: 1576–1580
`
`1577
`
`Ex. 1048-0002
`
`

`
`16
`14
`12
`10
`
`02468
`
`Everolimus Cmin (ng/mL)
`
`0
`
`2
`
`4
`
`6
`Week
`
`8
`
`10
`
`12
`
`Figure 2: Mean everolimus trough blood levels (Cmin) in pa-
`tients in the immediate cyclosporine group ( ✉) and those in
`the delayed cyclosporine group ( ❡). Bars represent 95% confi-
`dence intervals.
`
`Thereafter, with cyclosporine dose reduction, troughs de-
`clined slightly by month 3. In patients receiving delayed
`cyclosporine, the earliest initiation was day 4 with pro-
`gressively more patients in this treatment group on cy-
`closporine at each successive visit. By month 1, the mean
`cyclosporine level was similar to that of the immediate cy-
`closporine group.
`
`Everolimus mean trough trajectories are shown in Figure 2.
`In patients receiving immediate cyclosporine, everolimus
`reached steady state between days 4–6. Thereafter, lev-
`els remained relatively stable with means between 9 and
`11 ng/mL in the first 3 months post-transplant. The av-
`erage everolimus trough progressively rose over the first
`14 days after transplantation in patients receiving delayed
`cyclosporine. During this period, everolimus troughs were
`approximately two- to threefold lower in this group com-
`pared with troughs in the immediate cyclosporine group.
`By month 1 and thereafter, the everolimus troughs were
`similar in the two treatment groups.
`
`Influence of cyclosporine on everolimus
`Three evaluations were performed on the everolimus
`trough data from patients receiving delayed cyclosporine to
`assess the drug interaction of cyclosporine on everolimus:
`(1) a graphical assessment generating a time-shifted plot
`of the serial troughs by patient relative to the visit at which
`cyclosporine was added to the regimen; (2) a population
`comparison comparing all troughs across the population
`before vs. after initiation of cyclosporine; and (3) an intrain-
`dividual comparison based on the individual ratios of expo-
`sure before and after initiation of cyclosporine.
`
`As shown in Figure 3, each patient’s everolimus trough tra-
`jectory was plotted by a relative visit, whereby ‘visit 0’ was
`the visit at which cyclosporine was started, negative visits
`were before cyclosporine initiation with decreasing num-
`ber signifying sequentially earlier visits, and positive visits
`were after cyclosporine initiation with increasing number
`
`Kovarik et al.
`238 ± 112 mg/day. The average day on which cyclosporine
`was initiated in the delayed cyclosporine group was day
`11.1 ± 3.5 with a range from day 4 to day 15. Consequently,
`cyclosporine average doses were lower in this group com-
`pared with the immediate cyclosporine group for all visits
`in the first 2 weeks. For example, average doses were
`28 ± 83 mg/day at day 6, 192 ± 192 mg/day at day 12,
`and 264± 218 mg/day at day 16. Thereafter, cyclosporine
`doses were similar in both groups, averaging 346 ±
`128 mg/day at month 1 and 228 ± 98 mg/day at month 3.
`
`The protocol-specified everolimus dose of 1.5 mg twice
`daily was maintained in the majority of patients. An
`everolimus dose reduction to 1 mg twice daily was made
`in one patient in the immediate cyclosporine group and
`two patients in the delayed cyclosporine group in the first
`3 months post-transplant.
`
`Pharmacokinetics
`There was a total of 374 evaluable everolimus-cyclosporine
`concentration pairs. Concentration pairs from day 2 (n = 35)
`were included in plots of the data but omitted from infer-
`ential evaluation because everolimus was not yet at steady
`state at this visit. The main pharmacokinetic evaluation was
`therefore based on 339 concentration pairs collected from
`day 4 to month 3. There were 6.2 ± 2.0 concentration pairs
`per patient in the immediate cyclosporine group (range, 1–
`9 pairs) and 6.9 ± 1.7 concentration pairs per patient in
`the delayed cyclosporine group (range, 4–9 pairs). Within
`the delayed cyclosporine data set, pairs were evenly di-
`vided between the period when cyclosporine was withheld
`(3.6 ± 1.5 pairs per patient) and the period after cy-
`closporine was initiated (3.3 ± 2.4 pairs per patient).
`
`Figure 1 compares the mean cyclosporine trough trajec-
`tories in patients receiving immediate and delayed cy-
`closporine. In patients receiving immediate cyclosporine,
`mean troughs rose over the first week post-transplant and
`then remained relatively stable from week 2 to month 1.
`
`350
`
`300
`
`250
`
`200
`
`150
`
`100
`
`50
`
`0
`
`Cyclosporine Cmin (ng/mL)
`
`0
`
`2
`
`4
`
`6
`Week
`
`8
`
`10
`
`12
`
`Figure 1: Mean cyclosporine trough blood levels (Cmin) in pa-
`tients in the immediate cyclosporine group ( ✉) and those in
`the delayed cyclosporine group ( ❡). Bars represent 95% confi-
`dence intervals.
`
`1578
`
`American Journal of Transplantation 2003; 3: 1576–1580
`
`Ex. 1048-0003
`
`

`
`Cyclosporine–Everolimus Drug Interaction
`
`available or the precyclosporine levels were unquantifiable
`(<2 ng/mL). Of the 10 evaluable patients, three had essen-
`tially unaltered everolimus exposure when cyclosporine
`was initiated as evidenced by ratios of 0.7, 0.8, and 1.2.
`For the remaining seven patients, ratios were generally
`clustered between 1.7 and 3.8 with two outliers at 4.7 and
`5.6. The overall median ratio for all 10 patients was 2.9.
`
`Discussion
`
`A single-dose study in healthy subjects has demonstrated
`that coadministration of cyclosporine as the microemulsion
`formulation (Neoral) increases everolimus blood levels by
`an average 2.7-fold (7). As phase 3 everolimus drug de-
`velopment trials in kidney and heart transplantation used
`this drug combination from the de novo through mainte-
`nance periods after transplantation, this pharmacokinetic
`influence of cyclosporine on everolimus is incorporated in
`the proposed standard dose regimen and in the range of
`concentrations proposed for everolimus therapeutic drug
`monitoring (8). In the case of renal transplant patients with
`delayed kidney graft function, however, the initial with-
`holding of cyclosporine needs to be taken into account
`in dosing everolimus. We quantified the pharmacokinetic
`influence of cyclosporine on everolimus in this study to pro-
`vide guidance for the use of everolimus under these clinical
`conditions.
`
`This study revealed that when everolimus is used in
`an immunosuppressive regimen without cyclosporine,
`everolimus exposure is significantly lower than in a regi-
`men with cyclosporine. The basis for this difference is a
`drug interaction of cyclosporine on everolimus, likely by
`their shared CYP3A and/or P-glycoprotein disposition path-
`ways. Delayed kidney graft function itself is unlikely to have
`played a role in this interaction inasmuch as elimination of
`everolimus by the kidney is negligible (9).
`
`When cyclosporine was added to an everolimus-based reg-
`imen, everolimus concentrations were increased on aver-
`age by 2.9-fold; however, the increase was highly variable
`between patients ranging in this study from no change to
`a 5.6-fold increase. The average change in everolimus ex-
`posure in the presence of cyclosporine observed in these
`patients is in agreement with the previous healthy subject
`crossover study in which the AUC of everolimus increased
`in all 12 subjects by an average 2.7-fold with an individual
`range of 1.5- to 4.7-fold (7). Similar observations have been
`made for sirolimus. When a single 10-mg dose of sirolimus
`was coadministered simultaneously with 300 mg of cy-
`closporine microemulsion, the sirolimus AUC increased
`3.3-fold (10).
`
`The clinical implications of our study are that the dosing of
`everolimus needs to be adjusted to take into account an av-
`erage threefold increase in blood levels when cyclosporine
`is added to the regimen. Given the wide variability in the
`
`without cyclosporine with cyclosporine
`
`20
`18
`16
`14
`12
`10
`
`02468
`
`Everolimus Cmin (ng/mL)
`
`-8
`
`-6
`
`-4
`
`-2
`
`2
`0
`Relative visit
`
`4
`
`6
`
`8
`
`Figure 3: Individual everolimus trough blood level (Cmin) tra-
`jectories in patients in the delayed cyclosporine group (–).
`Mean Cmin trajectory ( ✉). Data are standardized by relative visit
`whereby visit 0 is the start of cyclosporine therapy, negative vis-
`its are while cyclosporine was withheld, and positive visits are
`those after the initiation of cyclosporine. Cmins less than the as-
`say quantification limit were set to 2 ng/mL for inclusion in the
`plot.
`
`signifying progressively later visits. Patients had different
`numbers of total visits depending on how long cyclosporine
`was withheld and how many evaluable troughs they pro-
`vided. Everolimus troughs were notably lower before the
`initiation of cyclosporine averaging between 2 and 4 ng/mL.
`At the first visit after the start of cyclosporine, everolimus
`troughs rose but did not stabilize until the second visit.
`Thereafter, they remained relatively stable in the group as
`a whole averaging approximately 8 ng/mL. The delay in
`stabilization was likely the result of an everolimus accu-
`mulation to reach a new steady state in the presence of
`cyclosporine. Inasmuch as the visits were separated by
`2 days, it took approximately 4 days to attain a new steady
`state as a result of the drug interaction.
`
`Given the indication that everolimus troughs were not at
`steady state at visit 1 after the initiation of cyclosporine,
`these data were omitted from this analysis, yielding a total
`of 96 everolimus-cyclosporine concentration pairs in pa-
`tients with delayed cyclosporine. The everolimus troughs
`were divided into two groups yielding the following popula-
`tion average troughs: 2.9 ± 2.8 ng/mL (n = 57) without cy-
`closporine vs. 8.3 ± 3.7 ng/mL (n = 39) with cyclosporine
`present (p < 0.001). The ratio of the means was 2.9, in-
`dicating that everolimus troughs increased 2.9-fold when
`cyclosporine was added to these patients’ regimens.
`
`Using the same dataset described earlier, each patient’s
`median trough was determined from the values collected
`before the initiation of cyclosporine and after initiation of
`cyclosporine. The magnitude of the interaction was cal-
`culated as the postcyclosporine/precyclosporine ratio for
`each patient. This could be derived in 10 patients; in the
`other six patients either no postcyclosporine data were
`
`American Journal of Transplantation 2003; 3: 1576–1580
`
`1579
`
`Ex. 1048-0004
`
`

`
`Kovarik et al.
`
`magnitude of this drug interaction, both everolimus and cy-
`closporine blood concentrations need to be carefully mon-
`itored in delayed graft function patients, especially in the
`period when cyclosporine is withheld and shortly after its
`initiation.
`
`Acknowledgment
`
`This study was funded by Novartis Pharma AG, Basel, Switzerland.
`
`References
`
`1. Schuler W, Sedrani R, Cottens S et al. SDZ RAD, a new rapamycin
`derivative. Transplantation 1997; 64: 36–42.
`2. Nashan B. Review of proliferation inhibitor everolimus. Expert
`Opin Invest Drugs 2002; 11: 1845–1857.
`3. Valantine H, Eisen H, Dorent R et al. 12-month results of a mul-
`ticenter study comparing efficacy and safety of everolimus to
`azathioprine in de novo cardiac transplantation. Am J Transplant
`2002; 2: 247.
`4. Kahan BD, Wong RL, Carter C et al. Phase 1 study of a 4-week
`course of SDZ-RAD in quiescent cyclosporine-prednisone-treated
`
`renal transplant recipients. Transplantation 1999; 68: 1100–
`1106.
`5. Kovarik JM, Kaplan B, Tedesco-Silva H et al. Pharmacokinetics
`of an everolimus-cyclosporine immunosuppressive regimen over
`the first 6 months after kidney transplantation. Am J Transplant
`2003; 3: 606–613.
`6. Kovarik JM, Rouilly M, Rordorf C, Somberg K. Influence of cy-
`closporine on the pharmacokinetics of everolimus: searching for
`the interaction threshold in renal transplant patients. Am J Trans-
`plant 2002; 2: 186.
`7. Kovarik JM, Kalbag J, Figueiredo J et al. Differential influence of
`two cyclosporine formulations on everolimus pharmacokinetics:
`a clinically relevant pharmacokinetic interaction. J Clin Pharmacol
`2002; 42: 95–99.
`8. Kovarik JM, Kaplan B, Tesdesco Silva H et al. Exposure-response
`relationships for everolimus in de novo kidney transplantation:
`defining a therapeutic range. Transplantation 2002; 73: 920–
`925.
`9. Kovarik JM, Sabia H, Rouilly M et al. Influence of renal and hep-
`atic impairment on everolimus pharmacokinetics: are dose ad-
`justments necessary? Am J Transplant 2001; 1: 385.
`10. Rapamune (sirolimus). Physician’s Desk Reference. Mont-
`vale, NJ: Medical Economics Company,
`Inc., 2001: 3443–
`3448.
`
`1580
`
`American Journal of Transplantation 2003; 3: 1576–1580
`
`Ex. 1048-0005