Safety and Efficacy of Raltegravir in HIV-Infected Transplant Patients Cotreated with Immunosuppressive Drugs

Authors


* Corresponding author: Leïla Tricot, l.tricot@hopital-foch.org

Abstract

Solid organ transplantations (SOT) are performed successfully in selected HIV-infected patients. However, multiple and reciprocal drug–drug interactions are observed between antiretroviral (ARV) drugs and calcineurin inhibitors (CNIs) through CYP450 metabolization. Raltegravir (RAL), a novel HIV-1 integrase inhibitor, is not a substrate of CYP450 enzymes. We retrospectively reviewed the outcomes of 13 HIV-infected transplant patients treated by an RAL + two nucleosidic reverse transcriptase inhibitor (NRTI) regimen, in terms of tolerability, ARV efficacy (plasma viral load, CD4 cell count), drug interactions, RAL pharmacokinetics and transplant outcome. Thirteen patients with liver (n = 8) or kidney (n = 5) transplantation were included. RAL was initiated (400 mg BID) either at time of transplantation (n = 6), or after transplantation (n = 7). Median RAL trough concentration was 507 ng/mL (176–890), which is above the in vitro IC95 for wild type HIV-1 strains (15 ng/mL). Target trough levels of CNIs were promptly obtained with standard dosages of tacrolimus or cyclosporine. RAL tolerability was excellent. There was no episode of acute rejection. HIV infection remained controlled. After a median follow-up of 9 months (range: 6–14), all patients were alive with satisfactory graft function. The use of an RAL + two NRTI-based regimen is a good alternative in HIV-infected patients undergoing SOT.

Introduction

Highly active antiretroviral therapy (HAART) has dramatically improved the survival of HIV-infected patients. HIV-infected patients are now eligible for solid organ transplantation (SOT) and are enrolled on waiting lists. Kidney transplantation (KTx) can give as good results in such patients as in the general population (1,2) and while liver transplantation (LTx) survival rates are not as good as those in the general population, it remains a reasonable option in HIV-infected patients (2). To maintain sustained undetectable plasma HIV-RNA, current guidelines recommend the use of a protease inhibitor (PI) boosted by ritonavir or a nonnucleoside reverse transcriptase inhibitor (NNRTI) backbone treatment combined with two nucleoside reverse transcriptase inhibitors (NRTI) (3). PI and NNRTI require metabolism via the cytochrome P450 family, the same enzyme complex responsible for clearance of the calcineurin inhibitors (CNIs), cyclosporine (CsA) and tacrolimus (FK) (4). Unsurprisingly then, previous publications have reported significant interactions between the agents used in HAART and immunosuppressive drugs (5,6). Multiple and reciprocal drug–drug interactions are difficult to manage and contribute to a high risk of CNI overdosage and toxicity. Ritonavir and the other PI are highly potent CYP3A inhibitors, requiring a drastic decrease in FK dosage (6). Thus, a main challenge to improve transplantation outcome in HIV-infected patients is to manage drug interactions between antiretroviral (ARV) drugs and immunosuppressive treatments.

Raltegravir (RAL) (MK-0518; Isentress, Merck) is a novel HIV-1 integrase strand transfer inhibitor, with potent in vitro and in vivo activity against HIV-1 and HIV-2, with a good safety profile (7). RAL is not a substrate of CYP450 enzymes and is eliminated mainly by metabolism via a UGT1A1-mediated glucuronidation pathway (8).

HAART with RAL + two NRTI could be an attractive alternative in HIV-infected patients eligible for SOT to avoid problems encountered with ARV regimen and CNI interactions.

We report on our first experience of concomitant treatment with CNIs and RAL combined with two NRTIs as ARV therapy in HIV-infected patients who were transplanted with a kidney or a liver.

Patients, Materials and Methods

This observational study was conducted between December 2007 and January 2009 in four French centers (Hôpital Paul Brousse, Hôpital Cochin, Hôpital Foch, Hôpital Pitié-Salpêtrière) and in accordance with French regulatory guidelines. The following criteria were required for inclusion: (i) age of at least 18 years, (ii) controlled HIV infection (plasma viral load [pVL] <40/mL, CD4 cell count >100/mm3) under HAART, (iii) SOT under CNI-based immunosuppression, (iv) switch from PI-based regimen to RAL therapy early after transplantation or at some distance from transplantation, (v) a minimum of 3 months of follow-up after RAL introduction. We retrospectively reviewed the outcomes of these patients in terms of treatment tolerability, ARV efficacy (pVL, CD4 count), drug interactions, RAL pharmacokinetics and transplantation outcome.

Thirteen patients fulfilled the inclusion criteria. RAL was introduced in two settings:

  • (i) In six patients (group A: 3 LTx, 3 KTx), the switch from PI-based HAART to an RAL + two NRTI regimen was decided at the time of transplantation to avoid posttransplant drug interactions. LTx recipients had their HAART stopped on the day of transplantation and an RAL-based regimen (400 mg BID) was introduced 2 to 3 weeks later, when liver graft function had recovered. KTx patients had their PI-based treatment stopped on the day of transplantation and the RAL-based regimen (400 mg BID) was started the next day without ARV treatment interruption.
  • (ii) In seven patients (group B: 5 LTx, 2 KTx), the PI-based regimen was switched to RAL at some distance from transplantation (median: 21 months, range: 3–57) because of PI intolerance (n = 5), HIV resistance (n = 1) or major difficulties in CNI management (n = 1).

Results

A total of 13 consecutive patients (8 LTx and 5 KTx) were included in this study. These 13 patients represent all eligible transplant patients in whom HAART was switched to RAL and who were followed in the four institutions during the study period.

Patients’ characteristics

Patients’ features at the time of switch are summarized in Table 1. Briefly, among the eight patients who received a liver transplant, liver failure was caused by HCV viral cirrhosis in five patients, regenerative nodular hyperplasia without viral coinfection in two and HBV-related cirrhosis in one patient. End-stage renal failure in the five patients who underwent KTx was caused by HIVAN in four and autosomic dominant polycystosis in another HIV-2-infected patient. HIV seropositivity was diagnosed at a median of 16 years (range: 1–22) before transplantation. CDC stage was A in one patient, B in three patients and C in nine. All patients had received several ARV treatments but only two had developed a multiresistant HIV strain. Reasons for multiple switches of ARV treatments were either drug toxicity as previously reported (9), or the availability of new drugs with improved convenience.

Table 1.  Patients characteristics at time of RAL introduction
PatientAge (yrs)GenderCause of end-stage organ failureCDC stageDuration of HIV infectionPrevious ARV drugs (n)HIV resistanceTx date (L or K)IS drugs
  1. HIVAN = HIV-associated nephropathy; K = kidney; L = liver; NNRTI = nonnucleosidic reverse transcriptase inhibitor; PI = protease inhibitor; NRTI = nucleosidic reverse transcriptase inhibitor; P = prednisone; FK = tacrolimus; MMF = mycophenolate mofetil; ATG = antithymoglobulin.

Group A: Early switch to RAL
  144MHCV cirrhosisB16 9016/07/07 (L)P + FK + MMF
  243MNodular regenerative hyperplasiaC1912NNRTI, PI, NRTI07/12/07 (L)P + FK + MMF
  345FHCV cirrhosisC17 6022/02/08 (L)P + FK + MMF
  430FHIVANC 5 2023/02/08 (K)ATG + P + FK + MMF
  537FHIVANC 6 3027/03/08 (K)Basiliximab + P + FK + MMF
  648FRenal polycystosisC20 9NRTI06/05/08 (K)Daclizumab + P + FK + MMF
Group B: Delayed switch to RAL
  738MHCV cirrhosisB22 8027/10/03 (L)P + FK + MMF
  850MHBV cirrhosisB2012018/12/04 (L)P + FK + MMF
  962FHIVANC10>10 NNRTI, PI, NRTI28/09/05 (K)ATG + P + CsA + MMF
 1044MHCV cirrhosisC14 9014/03/06 (L)P + CsA + MMF
 1143MNodular regenerative hyperplasiaC10 7016/02/07 (L)P + FK + MMF
 1243MHCV cirrhosisC19 7002/07/07 (L)P+CsA+ MMF
 1338FHIVANA 1 0013/12/07 (K)P + FK + MMF

Immunosuppressive treatment

All patients received a CNI-based triple therapy including FK (n = 10) or CsA (n = 3), mycophenolate mofetil and steroids.

In LTx, FK target trough levels were 8 to 20 ng/mL from day 0 to week 6 and 5 to 15 ng/mL thereafter. CsA target trough levels (C0) were 300 to 450 ng/mL from day 0 to week 2, and 175 to 350 ng/mL thereafter. Steroids were tapered to 10 mg/day at 3 months followed by gradual withdrawal over the next 3 months. There was no induction.

In KTx, FK target trough levels were 8–12 ng/mL from day 0 to month 3 and 6–8 ng/mL thereafter. CsA target peak levels (C2) were 800–1200 ng/mL during the first 3 months and 600–1000 ng/mL thereafter. Steroids were gradually tapered to 5 mg/day at 3 months. Four patients received an induction therapy with antithymoglobulin serum (n = 2) or anti-Il-2 receptor monoclonal antibodies (n = 2).

Follow-Up

Group A: Early introduction of RAL

Follow-up of patients who had an early switch to RAL are summarized in Table 2. Median follow-up was 11 months (range: 6–14). Postoperative course was uneventful. CNI therapeutic drug monitoring (TDM) was performed at least twice a week during the first 2 weeks, once a week during the first 3 months and then once a month. In all patients, therapeutic trough levels of CNIs were easily maintained within the therapeutic target ranges with usual dosages of FK or CsA. Of note, in the three LTx recipients whose ARV therapy was interrupted in the first 2 to 3 weeks following transplantation, no significant change in CNI dosage was necessary to maintain trough levels within target ranges following the introduction of RAL. The daily dosage of FK was maintained between 6 and 8 mg and the trough level remained between 5 and 10 ng/mL.

Table 2.  Antiretroviral treatment and follow-up after RAL introduction: Group A
PatientF-U after RAL switch (mths)ARV drugsPre-RAL CD4 (cells/mm3 and %)Last F-U CD4 (cells/mm3 and %)Pre-RAL HIV RNA (copies/mL)HIV RNA Last F-U (copies/mL)CNI dose (mg/day)CNI trough levels (ng/mL)S creatinine (μM)ALT/AST (IU/L)Bilirubin (μM) Albumin (g/L) PT (%)
  1. L = liver; K = kidney; F-U = follow-up; ARV = antiretroviral; RAL = raltegravir; 3TC = lamivudine; ABC = abacavir; FTC = entricitabine; T20 = enfuvirtide; CNI = calcineurin inhibitor; FK = tacrolimus; ALT = alanine aminotransferase; AST = aspartate aminotranferase; PT = prothrombin time.

1 (L)14RAL 400 mg BID189 (32%)423 (30%)131 <40FK = 17.4114N 6
3TC 300 mg QD        2.8
ABC 300 mg BID        100%
2 (L)11RAL 400 mg BID190 (32%)164 (21%)22  604<40FK = 89.210822/40N
TDF 300 mg QD         
FTC 200 mg QD         
T20 90 mg BID         
3 (L)6RAL 400 mg BID103 (19%)131 (18%)151<40FK = 83.4 6811/17 8
ABC 300 mg BID        3.7
3TC 300 mg QD        91
T20 90 mg BID         
4 (K)12RAL 400 mg BID       82NN
3TC 300 mg QD427 (19%)1450 (29%) <40<40FK = 10 10.5    
ABC 300 mg QD         
5 (K)11RAL 400 mg BID       90NN
3TC 300 mg QD181 (19%) 90 (20%)<40<40FK = 24 7.1   
ABC 300 mg QD         
6 (K)8RAL 400 mg BID      125NN
3TC 300 mg QD326 (26%)329 (33%)<100 <100 FK = 78.1   
ABC 300 mg QD         

pVL and CD4 count were performed on the day of transplantation, twice a week when ARV withdrawal was transient (LTx patients), at 1 month, and then every 3 months. Viremia remained undetectable following RAL introduction in all patients except for a transient replication (131–22 604 copies/mL) during the short ARV interruption period in the three LTx recipients. The frequency of viral monitoring was based on previous publications showing a risk of reactivation of HIV replication as early as 2 weeks after ARV structured treatment interruptions (10).

CD4 cell count remained above 100/mm3 in all patients except in patient 5 in whom it fell to 90/mm3; in this patient, the ratio of CD4 to total lymphocytes remained unchanged and pVL remained undetectable. Median CD4 count was 190/mm3 (range: 103–427) before RAL introduction and 246/mm3 (90–1450) at last follow-up, after a median time of 11 months (extremes: 6–14).

Graft function was prompt and remained excellent with no episode of acute rejection. No side-effects imputable to RAL, such as hepatitis or rhabdomyolysis, were recorded. At last follow-up, all patients were alive with good graft function.

Group B: Delayed switch to RAL

Follow-up of patients who had a delayed switch to RAL are summarized in Table 3. In these seven patients (5 LTx, 2 KTx), RAL was introduced at a distance from transplantation (median: 21 months, range: 3–57).

Table 3.  Antiretroviral treatment and follow-up after delayed RAL switch: Group B
PatientTime to RAL switch after TxF-U (months) after RAL switchARV drugsPre-RAL CD4 (cells/mm3 and %)Post-RAL CD4 (cells/mm3 and %)Pre-RAL HIV RNA (copies/mL)Last F-U HIV RNA (copies/mL)CNI dose (mg/day)CNI trough levels (ng/mL)S creatinine (μM)ALT/AST (IU/L)Bilirubin (μM) Albumine (g/L) PT (%)
  1. F-U = follow-up; ARV = antiretroviral; ABC = abacavir; 3TC = lamivudine; RAL = raltegravir; TDF = tenofovir; FTC: emtricitabine; T20 = enfuvirtide; CNI = calcineurin inhibitor; FK = tacrolimus; CsA = cyclosporine; T0 = cyclosporine blood trough drug concentration before drug intake; C2 = cyclosporine blood peak concentration 2 h after intake; ALT = alanine aminotransferase; AST = aspartate aminotranferase; PT = prothrombin time.

 7 (L)577RAL 400 mg BID370 (23%)450 (22%)<40<40FK = 0.58.9 6456/5126
3TC 300 mg QD        47
TDF 300 mg QD        100
 8 (L)488RAL 400 mg BID753 (19%)755 (19%)<40<40FK = 29.7 7048/309
3TC 300 mg QD        40
TDF 300 mg QD        100
 9 (K)329RAL 400 mg BID390 (39%)470 (44%)2060<40CsA = 10C2 = 40020021/149
3TC 300 mg QD        37
ABC 300 mg QD        100
10 (L)2114 RAL 400 mg BID170 (17%)100 (21%)<40<40CsA = 175C0 = 218108102/13539
3TC 300 mg QD        28
TDF 300 mg QD        97
11 (L) 39RAL 400 mg BID 70 (18%)140 (17%)<40<40FK = 0.55.210743/3719
3TC 300 mg QD        45
TDF 300 mg QD        100%
12 (L)128RAL 400 mg BID410 (41%)340 (39%)<40<40CsA = 250C0 = 13510651/4267
3TC 300 mg QD        36
ABC 300 mg BID        100%
13 (K) 49RAL 400 mg BID371 (22%)792 (27%)<40<40FK = 107.610823/12N
3TC 300 mg QD         
ABC 300 mg QD         

CNI TDM was performed at least twice a week during the first month following RAL introduction and then once a month. When the ritonavir boosted PI-based regimen was switched to RAL, FK and CsA dosage had to be increased by a factor of 5 to 15 to maintain the CNIs within the therapeutic target concentrations.

HIV infection remained controlled: pVL was undetectable and CD4 remained stable. Median CD4 count was 370/mm3 (range: 70–753) before RAL introduction and 450/mm3 (100–792) at last follow-up, after a median follow-up of 9 months (range 7–14).

Transplant outcome following the introduction of RAL was uneventful. No side-effects imputable to RAL were recorded. At last follow-up, all patients were alive with stable graft function.

The posttransplantation course of patient 9 before RAL introduction was much more complicated. At time of transplantation (2005), RAL was not available. The boosted PI-based regimen for a multiresistant HIV strain was not interrupted in the immediate post-Tx period. Balance between toxicity and efficacy was extremely difficult to obtain. CsA pharmacokinetics was completely modified with no peak concentration and a flat AUC. Trough levels were finally achieved with 20 mg CsA administered every other day. This patient experienced both a severe acute cellular rejection at month 2 and histological signs of CsA toxicity on repeated transplant biopsies. Renal function was suboptimal with a serum creatinine of 200 μmol/L. At month 32, the patient developed HIV resistance and the ARV treatment was switched to RAL, abacavir and lamivudine. After RAL introduction, CsA therapeutic trough levels were easily obtained at a standard dosage of 75 mg BID. HIV pVL became undetectable and CD4 count rose from 390 to 470/mm3. At last follow-up, renal function was stable and HIV infection remained under control.

RAL pharmacokinetics

Results of RAL short pharmacokinetics and trough levels are indicated in Table 4. In group A, four patients had short RAL PK studies for area under curve (AUC). Plasma RAL concentrations were measured with high performance liquid chromatography (HPLC) at a median of 3 months (range: 7 days–7 months) after RAL introduction. Plasma RAL concentrations were measured immediately before morning intake (T0), at 1 h (T60’) and at 3 h (T180’) after RAL intake. Median and mean half-life of RAL was 6 h (range: 2.8–9.9). Median RAL AUC was 20.059 ng.h/mL (range: 9.048–30.108). Mean peak plasma concentration of RAL was 2.332 ng/mL (538–3.097).

Table 4.  Results of RAL short pharmacokinetics (n = 4) and RAL trough levels (n = 5) patients
PatientTime of dosage after RAL introduction (days)T0 (ng/mL)T60 (ng/mL)T180 (ng/mL)Treatment PPI
  1. RAL = raltegravir; T0 = trough level before RAL intake; T60 = dosage 60 min after RAL intake; T180 = dosage 180 min after RAL intake; PPI = proton pump inhibitors.

Group A
  1 (LTx) 33204  Yes
  2 (LTx)21071030063094Yes
  3 (LTx)  6751  Yes
  4 (KTx) 9289040141381Yes
  5 (KTx) 98176 3141682No
  6 (KTx) 62747  Yes
Group B
  9 (KTx)  7229 301 538Yes
 10 (LTx)180298  No
 13 (KTx)117563  Yes

Plasma RAL trough concentration was performed in five other patients. The median trough level was 563 ng/mL (range: 204–751) after the routine daily dose of 400 mg BID whereas the in vitro IC95 for wild type HIV-1 strains is 15 ng/mL (11). Ten of the 13 patients were taking proton pomp inhibitors.

Discussion

RAL is the first compound of a novel class of ARV drugs that inhibits HIV replication through a different mechanism of action (HIV-1 integrase strand transfer inhibition) from that of other ARV drugs. In contrast to PI and NNRTI, RAL is not a substrate of CYP450. This distinct metabolization pathway makes it an attractive ARV drug for use in combination with CNI in SOT HIV-infected patients. The use of RAL in this context has previously been reported in one patient who developed CNI renal toxicity after LTx (11). To our knowledge, our study is the largest to evaluate an RAL-based regimen in HIV-infected transplant patients. The main findings were: (i) the lack of significant interaction between RAL and CNIs, allowing a simplified management of immunosuppressive treatment, (ii) the excellent tolerability of RAL and (iii) the uneventful outcome of transplantation and HIV infection under RAL-based regimen after a median follow-up of 9 months.

CNI dosage was easy to manage after RAL introduction in all patients. In group A, immediate introduction of RAL was associated with prompt and stable target levels of CNIs with standard dosages. Of note, the 2 to 3 weeks delayed introduction of RAL in LTx patients had no effect on FK trough levels and dosage, demonstrating the lack of significant interaction between RAL and CNIs. Thus, therapeutic trough levels of CNIs were promptly achieved without toxicity. This is crucial in the early posttransplant period, when the function of the drug metabolizing organs is recovering and the risk of acute rejection is maximal. Conversely, in group B patients, delayed switch from PI to RAL for drug intolerance resulted in a drastic drop of CNI blood concentrations and CNI doses had to be increased by a factor of 5 to 15. This observation is in accordance with the results of Teicher et al. where FK dose was reduced by 75 to 99% when PI-based regimens were reintroduced several weeks after LTx (6).

To avoid unpredictable CNI concentrations, physicians usually stop HAART during the immediate post-LTx course and reintroduce it several weeks later, once graft function has returned to normal and CNI blood concentrations have stabilized. Although this procedure has been reported to be achievable, it requires careful follow-up of HIV pVL when ARV is stopped and frequent TDM of CNIs when ARV treatments are resumed (6). Using an RAL-based regimen immediately after SOT simplifies treatment management and could be cost-effective by avoiding numerous TDM and virological monitoring in the period when ARV is modified.

As an alternative to conventional HAART, the effective use of three NRTI in HIV-infected patients undergoing LTx has been reported. However, this combination offers low antiviral potency and entails a risk of metabolic acidosis due to mitochondrial toxicity, especially when HCV recurrence requires treatment with pegylated interferon plus ribavirin (4,11).

No adverse events, and in particular no hepatitis or rhabdomyolysis, were observed in these patients. This is in accordance with similar rates of drug-related events reported in the RAL and placebo groups in phase 2 and 3 trials (7).

RAL has previously been shown to be a powerful weapon against HIV infection. However, its genetic barrier to drug resistance is low. In our patients, RAL was associated with two NRTI to avoid incomplete virologic suppression and to decrease the risk of emergence of drug-resistance HIV mutants. Viral load remained undetectable and CD4 cell count remained stable. However, follow-up was less than 9 months in half of the patients and long-term RAL efficacy could be a concern.

It is noteworthy that no episode of acute rejection occurred under the RAL-based regimen. Unexpectedly high rates of acute rejection (22–52%) have been previously reported, especially in HIV-infected KTx recipients under a PI-based regimen (1,2). Such rates could either result from inadequate exposure to immunosuppressive agents related to drug interactions or be due to altered immune response in HIV-infected patients (2). While the small number of patients at risk in our study does not allow us to draw firm conclusions, this observation suggests that the culprits for elevated rates of acute rejection are PI and/or NNRTI–CNI interactions and subsequent difficulties to achieve an adequate CNI exposure.

RAL PK studies in four patients showed a short half-life, allowing a rapid steady state after RAL introduction. RAL AUC wide range reflects the interindividual variability in absorption. Within this range, tolerability was excellent in all 13 patients. Treatment with proton pump inhibitors (PPI) can be used with RAL, which is convenient in the postoperative period. PPIs increase the absorption of RAL (12), which probably contributed to high trough concentrations of RAL in the patients who were taking PPI treatment.

There are several limitations in this study. First, it was retrospective. However, all consecutive eligible HIV-infected patients who underwent an RAL switch in the four participating institutions were included. Outcomes of interest (HIV infection parameters, CNI dosages, RAL side-effects and posttransplant events) were carefully and regularly recorded in every case. Second, the follow-up was below 9 months in half of the patients. However, most of the problems encountered with drug interaction, achieving therapeutic CNI trough levels or acute rejection occur within days to weeks after KTx or LTx. A major concern in the long term could be HIV resistance to RAL because of a low genetic barrier to HIV-resistance. Long-term tolerability will have to be confirmed since a higher incidence of cancer was suggested in the phase 3 study (7).

In conclusion, based on this preliminary study, we would recommend the use of RAL in the context of SOT in HIV-infected patients. In our opinion, the best strategy could be to switch PI and NNRTI for RAL combined with two NRTI at time of transplantation, without any ARV interruption. Long-term evaluation of ARV efficacy and tolerability is necessary.

Acknowledgments

We are particularly grateful to Dr. Valantin, Infectious Diseases Department, Hôpital Pitié-Salpêtrière for referring his patients and to Dr. Hiesse, Kidney Transplantation Department, Hôpital Foch, for his carefully reading of the manuscript.

Ancillary