Combining cytochrome P-450 3A4 modulators and cyclosporine or everolimus in transplantation is successful

Abstract

AIM: To describe the long term follow-up of kidney allograft recipients receiving ketoconazole with calcineurin inhibitors (CNI) alone or combined with everolimus.

METHODS: This is an open-label, prospective observational clinical trial in low immunologic risk patients who, after signing an Institutional Review Board approved consent form, were included in one of two groups. The first one (n = 59) received everolimus (target blood level, 3-8 ng/mL) and the other (n = 114) azathioprine 2 mg/kg per day or mycophenolate mofetyl (MMF) 2 g/d. Both groups also received tapering steroids, the cytochrome P-450 3A4 (CYP3A4) modulator, ketoconazole 50-100 mg/d, and cyclosporine with C0 targets in the everolimus group of 200-250 ng/mL in 1 mo, 100-125 ng/mL in 2 mo, and 50-65 ng/mL thereafter, and in the azathioprine or MMF group of 250-300 ng/mL in 1 mo, 200-250 ng/mL in 2 mo, 180-200 ng/mL until 3-6 mo, and 100-125 ng/mL thereafter. Clinical visits were performed monthly the first year and quarterly thereafter by treating physicians and all data was extracted by the investigators.

RESULTS: The clinical characteristics of these two cohorts were similar. During the follow up (66 + 31 mo), both groups showed comparable clinical courses, but the biopsy proven acute rejection rate during the full follow-up period seemed to be lower in the everolimus group (20% vs 36%; P = 0.04). The everolimus group did not show a higher surgical complication rate than the other group. By the end of the follow-up period, the everolimus group tended to show a higher glomerular filtration rate. Nevertheless, we found no evidence of a consistent negative slope of the temporal allograft function estimated by the modification of the diet in renal disease formula in any of both groups. At 6 years of follow-up, the uncensored and death-censored graft survivals were 91% and 93%, and 91% and 83% in the everolimus plus cyclosporine, and cyclosporine alone groups, respectively. The addition of ketoconazole saved 80% of cyclosporine and 56% of everolimus doses.

CONCLUSION: Combining CYP3A4 modulators with CNI or mammalian target of rapamycin inhibitor, in low immunological risk kidney transplant recipients is feasible, effective, safe and affordable even in the long term.

Key Words: Kidney transplant; Immunosuppressive; Cyclosporine; Ketoconazole; Everolimus; Cytochrome P-450; Cytochrome P-450 3A4 modulator

Core tip: Several immunosuppressive (IS) drugs, used in clinical transplantation, are metabolized by the hepatic cytochrome P-450 system as many other drugs. The co-prescription of IS and ketoconazole reshapes the IS pharmacokinetics and appears to confer benefit to patients receiving calcineurin inhibitors (CNI) and mammalian target of rapamycin inhibitors. We describe the long term follow-up of kidney allograft recipients receiving ketoconazole with a CNI alone or combined with everolimus and report good graft and patient survivals and low rates of acute rejection episodes. These combinations, in low immunological risk kidney transplant recipients are feasible, effective, safe and affordable even in the long term.

INTRODUCTION

The prognosis of kidney transplantation has improved as new immunosuppressive (IS) drugs have been introduced in clinical practice and as prescribing physicians have learned to combine and prescribe them[1]. Most of the time, IS doses are monitored by measuring patients’ drug blood levels based on the results of clinical trials designed to prove that a specific drug blood level window is associated with maximal IS efficacy to prevent acute rejection episodes and minimal incidence of drug-related adverse events.

Several IS drugs are metabolized by the hepatic cytochrome P-450 system[2]. This elimination pathway is shared by a lot of drugs commonly prescribed both in internal medicine and in clinical transplantation, creating the opportunity for the appearance of drug interactions that could translate to adverse effects. For instance, while rifampin and phenytoin induce activity of the cytochrome, macrolides and azole antifungal agents decrease it, in such a way that certain drug metabolism is secondarily accelerated or retarded, respectively[2].

Intending to prescribe IS with cytochrome P-450 inhibitors simultaneously, particularly on the cytochrome P-450 3A4 isozyme (CYP3A4), is a practice that has been repeatedly reported in transplant literature, beginning with cyclosporine[3-26] and tacrolimus[27-29] and followed by sirolimus and everolimus[30-33]. These combinations have been associated with favorable clinical short and long term outcomes, but occasionally with more adverse events due to drug induced toxicities. At the same time, these drug combinations give health payers the opportunity to save financial resources[32,34-39]. Few authors have already shown that for other clinical conditions than transplantation the proposed combination has no adverse effects and saves money.

Combining IS drugs with a low dose of ketoconazole, a well-known CYP3A4 inhibitor, gives the possibility to modulate the isozyme activity in order to change the drug blood concentration vs time curve shape in such a way that the drug’s maximal concentration (C_max) is reduced alongside its metabolic disposal rate and the area under the time concentration curve (AUC) is reshaped to approximately the pharmacokinetic profile described by a Gamma’s distribution curve, from one with lower to another with higher alpha and beta parameters for that function (Figure 1)[40]. In other words, the addition of a CYP3A4 modulator gives the AUC a more rectangular graphical shape as C_max decreases but maintains the clinically driven C0 target (concentration at the end of the dosing interval and before the next drug intake) and, at the same time, stabilizes AUC, whose magnitude has been related to acute rejection risk in cyclosporine or tacrolimus users.

Figure 1 Gamma distribution curves with varying alpha and beta parameters.

The interaction between ketoconazole and the IS drugs is believed to result from the imidazole’s inhibition of the hepatic microsomal cytochrome P-450 dependent mixed function oxidase system that deactivates drugs. Two mechanisms have been proposed: Competitive inhibition at the substrate binding site and interaction of ketoconazole with the haem moiety of the cytochrome P-450 itself, preventing the binding and activation of oxygen and consequently inhibiting the metabolism of IS drugs[41].

This therapeutically intended reshaping in IS drug exposure has been correlated, in prospective randomized trials, to a decreased incidence and severity in clinical allograft acute rejection rate and to a better graft function in cyclosporine or tacrolimus treated patients[42-47]. Preliminary results with sirolimus and everolimus are also promising[32,33].

The aim of this report is to describe the long term follow-up of two cohorts of kidney allograft recipients whose CYP3A4 was modulated with a low ketoconazole dose and who were receiving an IS treatment consisting in a calcineurin inhibitor (CNI) alone or in combination with another CYP3A4 metabolized drug, such as everolimus.

MATERIALS AND METHODS

Study design

We performed an open-label, observational, nonrandomized, prospective, cohort, comparative clinical trial among low immunologic risk patients, who were defined as adult males or non-pregnant females undergoing primary deceased donor, living-unrelated or human leukocyte antigen-mismatched living-related donor kidney transplantations.

Subjects were required to display a rate of panel reactive antibodies (PRA) < 20%, cold ischemia time of < 30 h and a warm ischemia time lower than 45 min in order to undergo transplantation. All patients signed a written informed consent form approved by the local ethics committee. All participating women consented to use an effective contraceptive method.

Immunosuppressive therapy

After transplantation, all patients received IV methylprednisolone for the first 3 d and then oral prednisone at doses tapered to reach 15 mg/d at 6 mo; 10 mg/d at 12 mo; and 5 mg/d thereafter. From 0 d, all patients received oral modified cyclosporine (Neoral, Novartis Pharma AG, Basel, Switzerland), ketoconazole (100 mg/d) and azathioprine (2.0-2.5 mg/kg per day). After 5 d, a cohort of patients without a significant delayed graft function (defined as a requirement for less than one week of dialysis), were switched from azathioprine to everolimus 0.25 mg twice a day without loading dose. The other group continued receiving mainly azathioprine, but some patients were switched to mycophenolate mofetyl (2 g/d) by the treating physicians. No induction therapy was allowed, but one patient inadvertently received basiliximab.

Immunosuppressant doses were modified according to the following through blood level targets. Everolimus group: Everolimus, 3-8 ng/mL (Innofluor, Seradyn); cyclosporine, 200-250 ng/mL the first month, 100-125 ng/mL the second month and 50-65 ng/mL thereafter (Axym, Abbott). Azathioprine or mycophenolate mofetyl (MMF) group: Cyclosporine 250-300 ng/mL the first month, 200-250 ng/mL the second month, 180-200 ng/mL until the end of the sixth month and 100-125 ng/mL thereafter.

Primary aim

To describe the pharmacological interaction between the CYP3A4 modulator ketoconazole and cyclosporine alone or in combination with everolimus in kidney transplanted patients.

Secondary aim

To describe, in both groups, the incidence of biopsy proven acute rejection episodes, graft survival and kidney graft function by serum creatinine and modification of the diet in renal disease (MDRD) estimated glomerular filtration rate (GFR) at six years of follow-up. To describe, in both groups, the incidence of selected medical complications, such as new-onset diabetes mellitus (NODAT), neoplasia, and post-transplant lymphoproliferative disorder (PTLD) and BK virus nephropathy and cytomegalovirus (CMV) disease.

Statistical analysis

As this was not a randomized trial, we do not have the intention to formally and strictly compare both groups. All analyses were performed on an intention-to-treat basis. Analysis of variance was used for continuous variables and covariance for repeated measurements; χ² and Fisher exact tests for categorical variables. Survival analysis was done with the Kaplan-Meir method and the log-rank test.

RESULTS

Between January 1^st 2005 and December 31^st 2012, 254 transplants were performed. From them, 2 patients abandoned controls and one patient’s clinical registries were lost, leaving 251 patients. The sixty one patients having PRA > 20%, those who suffered from a non-functioning graft (n = 12; 4.8%) and the five patients who died before they were discharged from first hospitalization (2%) were not considered in further analysis, leaving a total of 173 patients for follow up. From these, 59 patients (34%) began everolimus immunosuppressive treatment during the first month and the other 114 patients (66%) continued receiving azathioprine or MMF combined with cyclosporine, ketoconazole and tapering steroids.

The clinical characteristics of these two cohorts are showed in Table 1. Both groups were very similar, but the group receiving azathioprine/MMF either received more kidneys from non-living or hypertensive donors or underwent a longer warm ischemia time and, as expected, they suffered more delayed graft function (DGF).

Table 1 Characteristics of kidney donors and recipients.

	Everolimus (n = 59)	Azathioprine/MMF (n = 114)	P value
Donor
Age (yr)	38.4 ± 13.7	44.1 ± 13.0	< 0.01
Male gender	30 (51%)	65 (57%)	0.44
Living	15 (25%)	14 (12%)	0.03
Non-living	44 (75%)	100 (88%)	0.03
Extended criteria donor	5 (9%)	20 (18%)	0.11
Stroke as donor's cause of death	10 (23%)	28 (28%)	0.51
Hypertension	3 (5%)	22 (19%)	0.01
Type 2 diabetes	0 (0%)	4 (4%)	0.15
Serum creatinine (mg/dL)	0.83 ± 0.26	0.90 ± 0.36	0.19
Cold ischemia time (h)	18.9 ± 5.4	20.1 ± 7.1	0.33
Warm ischemia time (min)	37.3 ± 9.25	41.3 ± 11.2	0.02
Recipient
Age (yr)	43.1 ± 12.5	45.0 ± 12.1	0.35
Male gender	32 (54%)	79 (69%)	0.05
List waiting time (mo)	27.9 ± 22.7	30.4 ± 28.3	0.57
Previous kidney transplant	0 (0%)	0 (0%)
Total time in dialysis (mo)	49.0 ± 26.5	58.4 ± 33.6	0.57
PRA (%)	3.0 ± 4.3	3.8 ± 5.2	0.35
HLA-mismatch	2.9 ± 1.4	2.8 ± 1.2	0.68
Double kidney transplant	1 (2%)	5 (4%)	0.36
Hypertension	42 (71%)	79 (69%)	0.80
Type 2 diabetes	0 (0%)	7 (6%)	0.10
Coronary artery disease	1 (2%)	1 (1%)	0.63
IgG CMV (+)	56 (97%)	98 (88%)	0.06
Immunosuppressive treatment
Induction	0 (0%)	1 (1%)	0.47
Cyclosporine	59 (100%)	114 (100%)
Azathioprine	59 (100%)	111 (97%)	0.21
Mycophenolate mofetyl	0 (0%)	3 (3%)	0.21
Delayed graft function	3 (5%)	65 (57%)	< 0.01

MMF: Mycophenolate mofetyl; PRA: Panel reactive antibodies; CMV: Cytomegalovirus.

During the follow up (66 + 31 mo, median 66.6 mo, range 1-133), both groups showed comparable clinical courses. However, the biopsy proven acute rejection rate during the full follow-up period seemed to be lower in the everolimus group (20% vs 36%; P = 0.04) (Table 2). As expected, those patients who received azathioprine/MMF tended to show more leukopenia, thrombocytopenia or to develop more pneumonias than those receiving everolimus. The everolimus group did not show a higher surgical complication rate.

Table 2 Follow up clinical findings and complications n (%).

	Everolimus (n = 59)	Azathioprine/MMF (n = 114)	P value
Surgical complication	11 (19)	25 (22)	0.61
Vascular complication	2 (3)	10 (9)	0.22
First year acute rejection episode	6 (10)	25 (22)	0.06
Acute rejection episode during entire follow up period	12 (20)	41 (36)	0.04
Cyclosporine toxicity	8 (14)	22 (19)	0.35
New onset diabetes after transplant	3 (5)	8 (7)	0.75
CMV disease	0 (0)	6 (5.3)	0.10
BK virus nephropathy	1 (2)	5 (4)	0.67
New onset neoplasia	2 (3)	6 (5)	0.72
Post-transplant Lymphoproliferative disease	1 (2)	2 (2)	0.98
Hospitalizations/yr	0.50 ± 0.72	0.62 ± 0.78	0.32
Leucopenia	12 (20)	58 (51)	< 0.01
Thrombocytopenia	29 (49)	73 (64)	0.06
Pneumonia	6 (10)	25 (22)	0.06
Urinary tract infection	27 (46)	46 (40)	0.49

MMF: Mycophenolate mofetyl; CMV: Cytomegalovirus.

Other adverse events were not consistently observed. Nevertheless, at the beginning of each immunosuppressive treatment much attention had to be devoted to adjusting drug doses in order to achieve the therapeutic windows without surpassing their upper limits. There were several times that cyclosporine blood levels transiently reached supra-therapeutic concentrations without more adverse events than tremor. Liver functions tests were monitored at each clinical visit and no alterations were observed.

Renal function and grafts survival

The everolimus group had less DGF than the azathioprine/MMF group, but this happened because of the design of the immunosuppressive protocols, as patients suffering of DGF for more than a week could not receive the mammalian target of rapamycin (m-TOR) inhibitor because of concerns of a risk of prolonging the graft dysfunction.

Regardless of the DGF incidence, both groups recovered kidney function in a comparable way. However, by the end of the follow-up period, the everolimus group tended to show a higher glomerular filtration rate. Nevertheless, we found no evidence of a consistent negative slope of the temporal allograft function in any of both groups (Figure 2).

Figure 2 Kidney allograft function estimated by plasma creatinine (A) and glomerular filtration rate estimated by mdrd formula (B). MMF: Mycophenolate mofetyl.

The uncensored and death-censored graft survival at different time periods are shown in Tables 3 and 4 and Kaplan-Meier graphs are shown in Figure 3. Log-rank tests did not show statistical significant differences between both groups.

Table 3 Graft survival uncensored by recipient death with a functioning graft at different periods after kidney transplant.

Time	Everolimus (n = 59)	Azathioprine/MMF (n = 114)
Year 1	98%	97%
Year 2	98%	94%
Year 3	96%	93%
Year 4	94%	88%
Year 5	94%	86%
Year 6	91%	83%

MMF: Mycophenolate mofetyl.

Figure 3 Graft survival un-censored (A) and graft survival censored (B) for patient death with a functioning graft. MMF: Mycophenolate mofetyl.

Table 4 Graft survival censored by recipient death with a functioning graft at different periods after kidney transplant.

Time	Everolimus (n = 59)	Azathioprine/MMF (n = 114)
Year 1	100%	97%
Year 2	100%	94%
Year 3	98%	93%
Year 4	96%	88%
Year 5	96%	88%
Year 6	93%	83%

MMF: Mycophenolate mofetyl.

Figure 4 Cyclosporine daily dose (A), cyclosporine blood concentrations (B) and everolimus daily dose (C) during the first two years of follow-up. MMF: Mycophenolate mofetyl.

CYP3A4 modulator effect

The addition of ketoconazole was associated to a lower dose requirement of both everolimus and cyclosporine in order to achieve the therapeutic blood concentrations. The usual recommended initial cyclosporine and everolimus doses of 8 mg/kg per day and 1.5 mg/d, respectively, were allowed to be decreased, at 30 d post transplantation, to 1.63 + 0.83 mg/kg per day and 0.67 + 0.23 mg/d of cyclosporine and everolimus, respectively. That is to say, the CYP3A4 modulator saved 80% and 56% of drug doses.

In the cyclosporine only group, the same 80% dose reduction necessity was observed. At day 30 post transplantation cyclosporine daily dose was 1.67 + 0.47 mg/kg.

The immunosuppressant daily doses and blood levels during the first year of follow up are shown in Figures 4 and 5. The most relevant findings deployed in those figures are a lesser dispersion of the daily doses of both IS, cyclosporine and everolimus, in order to achieve and maintain the therapeutic blood concentration windows in all time periods of the follow-up. Obviously, the cyclosporine blood levels in both groups are not comparable, because the target ones are different in both schemes.

Figure 5 Everolimus blood concentrations during the first years of follow-up.

There was a slight positive correlation between cyclosporine blood levels and serum creatinine in the everolimus group: r = 0.1637; two-tailed probability: 0.004 (Figure 6), but not in the Azathioprine/MMF group: r = 0.064; two-tailed probability: 0.256 (Figure 6).

Figure 6 Correlation between plasma creatinine and blood cyclosporine concentration in the everolimus group (A) and mycophenolate mofetyl group (B).

DISCUSSION

The addition of a CYP3A4 modulator to kidney transplant recipients who use a cyclosporine or a cyclosporine and everolimus based immunosuppressive regimen allows to consistently and importantly reduce the drug doses without jeopardizing the ability to achieve and maintain therapeutic blood levels of the IS in both regimens. Moreover, the addition of low doses of ketoconazole stabilizes medium and long term of both everolimus and cyclosporine and makes the periodic control clinical visits easier.

The use of ketoconazole has been a controversial issue in clinical transplantation, in spite of prospective randomized trials that do not show worse clinical results in comparison to not using the CYPA34 modulator[42-47]. Moreover, it has been suggested that ketoconazole could behave as an immunomodulator agent, as it reduced the acute rejection rate in heart transplant patients[44]. Our biopsy proven acute rejection (BPAR) rate of both groups was comparable to similar schemes without the CYP3A4 modulator. For example, the everolimus and cyclosporine group showed a first year BPAR of 10% that compares favorably with the three arms containing a calcineurin inhibitor in the Elite-Symphony trial[48] (low-dose tacrolimus 12.3%, standard-dose cyclosporine 25.8% and low-dose cyclosporine 24.0%) and also with another trial with a similar design of everolimus and low exposure of cyclosporine that reported a first year incidence of BPAR of 16.2%[49]. For the cyclosporine only group, the first year BPAR rate was 22% in comparison with 23% in the azathioprine group and 18% of the mycophenolate mofetyl group of the MYSS trial[50] and also alike the cyclosporine and MMF rates in the Elite-Symphony trial[48].

We did not construct formal pharmacokinetic time-curves in any of the study groups. However, in a previous experience, we learned that in order to maintain the blood cyclosporine concentration constant before the next dose (C0) combining cyclosporine with ketoconazole, it is necessary to adjust the CNI dose in such a way that the pharmacokinetic profile changed decreasing both C_max and AUC[51]. That is to say that ketoconazole changed the cyclosporine blood concentration time function in the same way as increasing the alpha and beta parameters of a Gamma type distribution (Figure 1)[40].

The main limitation of using CYP3A4 modulators could be related to the occurrence of adverse events due to a theoretically increased exposure to IS drugs, which could translate to more infective episodes or a higher frequency of hospitalizations. Nevertheless, our data does not show an increase in the incidences of NODAT, CMV or BK virus diseases, new onset neoplasia or PTLD or more hospitalizations as compared with the other trials[48-50]. The key issue to achieve these comparable rates is to actively adjust the IS doses to the usual therapeutic windows reducing everolimus and cyclosporine in almost 60% and 80%, respectively (Figures 3 and 4).

Both graft survival functions, censored and uncensored by recipients death with a functioning graft, were positive. At year six of follow-up, those receiving everolimus show 93% and 91%, respectively, and those receiving azathioprine/MMF 83% and 81%. Both compare favorably with the follow-up of the Elite-Symphony trial that showed uncensored graft survival between 85% and 90% in the four experimental groups after 3 years of follow-up, and they are certainly better than other clinical trials exploring CNI and m-TOR inhibitor combination[49,52,53].

The kidney allograft functions of both immunosuppressive regimens show similar behaviors. In spite of several time points with significant differences in plasma creatinine or MDRD estimated glomerular filtration rate, both show follow-up stability and, interestingly, there appears to be no progressive GFR deterioration in the full cyclosporine exposure scheme in comparison with the other scheme with reduced exposure of the CNI. These findings put in doubt the real importance of CNI exposure and its postulated related nephrotoxicity that was once named as chronic allograft nephropathy and correlated with histological kidney graft interstitial fibrosis and tubular atrophy[54].

Still more important, we found no evidence that the CYP3A4 modulator could predispose to a graft functional progressive deterioration, either because of a deficient immunosuppressive efficacy or chronic CNI associated nephrotoxicity, as both regimens did have different CNI exposures.

This idea of co-administering CYP3A4 modulators enhancing the immunosuppressive efficacy and safety of commonly used drugs in solid organ transplantation has been transferred to a completely different clinical field such as medical oncology. In fact, there is an increasing interest of exploring this particular pharmacological interaction to better preserve the health of cancer patients[55-57]. Nevertheless, it is necessary to be conscious that ketoconazole could be related to adverse events, mainly liver injury, if they are prescribed in higher doses than 200 mg a day[58] and that newer combinations of drugs in internal medicine, solid organ transplantation or oncology can be a better choice that the use of CYP3A4 modulators.

In summary, we have described our long term experience of combining the CYP3A4 modulator ketoconazole with a lone CNI or in combination with an m-TOR inhibitor, in low and medium immunological risk kidney transplant recipients and our main findings were that these combinations are clinically feasible, effective, safe and affordable even in the long term. In spite of that, these strategies have not received much attention and have not been explored in adequately designed, prospective, randomized and long term trials; they deserve all of the transplant community’s attention because they could potentially allow for better global clinical results in kidney, and even other solid organ, transplantation.

COMMENTS

Background

Kidney transplantation is a well-accepted treatment for end stage renal disease as it maximizes patient survival in comparison to remaining in chronic dialysis. Immunosuppressive (IS) treatment is the main therapy used to prevent acute rejection episodes and to avoid premature allograft losses. In spite of improving IS schedules, graft survival is not satisfactory.

Research frontiers

At the beginning of the 1990s, it was reported in biomedical literature that combining IS drugs metabolized by the hepatic cytochrome P-450 system with ketoconazole or diltiazem could slow the disposal metabolic rate of IS, giving the opportunity to save money in disadvantaged countries. Shortly afterwards, it was also postulated that the addition of ketoconazole could, in fact, modulate the cytochrome function allowing some kind of accommodation of the IS regimens that could theoretically improve graft survivals. In fact, this imidazole agent changes the pharmacokinetic curve both of calcineurin and mammalian target of rapamycin (m-TOR) inhibitors.

Innovations and breakthroughs

With the entry of newer IS, like mycophenolate acid derivatives and m-TOR inhibitors, that strategy was abandoned, just remaining in isolated clinical reports. In Hospital del Salvador, in Chile, the modulation of the cyctochrome P-450 system with ketoconazole is part of almost all IS regimens since the early 1990s. In the middle of the last decade, the authors began an experience combining ketoconazole, cyclosporine and everolimus that is yet continuing and in this paper, the authors communicate this experience compared with another similar cohort receiving only cyclosporine and ketoconazole [plus azathioprine or mycophenolate mofetyl (MMF)].

Applications

The obtained results are certainly encouraging as the authors observed similar or even lower acute rejection episode and viral infection rates and similar or better 5 year graft survival compared with other well validated IS regimens as those containing antibody induction followed by the combination of tacrolimus and MMF and with a very favorable safety profile. Obviously, this experience must be validated with double-blind, randomized, prospective and controlled trials, even considering the economic disincentive to conduct a clinical trial that allows saving more than 50% of cytochrome P-450 metabolized agent doses and, in parallel, important quantities of valuable money.

Peer-review

It is an interesting manuscript evaluating the association of cyclosporine and ketoconazole in transplantation.