Portal de Periódicos da CAPES

RAPAMYCIN: CLINICAL RESULTS AND FUTURE OPPORTUNITIES1

2001; Wolters Kluwer; Volume: 72; Issue: 7 Linguagem: Inglês

10.1097/00007890-200110150-00001

1534-6080

Barry D. Kahan, Joseph S. Camardo,

Adenosine and Purinergic Signaling

INTRODUCTION Sirolimus (rapamycin; RAPA) is a macrocyclic lactone with a novel mechanism of immunosuppressive action (1). During the past 7 years, the drug has undergone clinical trials progressing from Phase I safety, tolerability, and pharmacokinetic investigation to Phase II dose-finding studies and limited-sized, multicenter evaluations of drug combination regimens. The completion of Phase III large randomized national and international trials led to approval of the drug to achieve augmented acute rejection prophylaxis in combination with cyclosporine (CsA) and steroids by the Food and Drug Administration of the United States in September 1999. In November 2000, the drug was approved by the European Agency as an alternate to calcineurin antagonists for long-term maintenance therapy. This overview seeks to familiarize the reader with the clinical information that provided the bases for drug approval and with the single-center reports that document alternate approaches to optimize the outcomes of treatment with this immunosuppressive agent. I. Update on Preclinical Findings RAPA, via its c-7 methoxy group (2), cross-links (3) the immunophilin FK binding protein (FKBP) 12, a peptide-prolyl isomerase that acts as a folding catalyst, to the multifunctional serine-threonine kinase, the mammalian target of rapamycin (mTOR) (4). Blockade of mTOR dampens lymphocyte responses to costimulatory signal 2 during the G0 to G1 transition and to cytokine signal 3 during the G1 build-up. By blocking the costimulation signals, RAPA prevents activation of the inhibitory factor kappa kinase necessary for generation of the c-Rel transcription factors of the NF-κB complex (5), and possibly also modulates protein kinase C activity (6). During the later G1 phase, by blocking signal 3, RAPA inhibits four cytokine-driven signaling pathways: a) p27kip1 degradation (7,8) leading to cyclin activation (9,10); b) p70S6 kinase stimulation, a step necessary for the synthesis of endosomal structural proteins (11–13); c) elongation factor 4A release from its association with PHAS-I, thereby facilitating ribosomal protein synthesis (14–16); and d) transcriptional up-regulation of the anti-apoptotic proteins bcl (17,18) and p21Ras (19) (Fig. 1). Both the therapeutic and the toxic effects of RAPA are related to the same cellular actions. The drug's unique effects are complementary to calcineurin antagonists (CNAs) (20) and to interleukin-2 receptor monoclonal antibodies (anti–IL-2R mAbs); a relation that has been called the "cytokine paradigm"(21) (Fig. 2). Figure 1: Sites of enzyme action of mammalian target of rapamycin (mTOR). a) Activation of c-Rel factors downstream from reception of the costimulatory CD28 signal. b) Phosphorylation of p70S6 kinase preceding endosomal structural protein synthesis. c) Release of e-IH-4E from its association with PHAS-I, leading to the 4E activity necessary for elongation of the polypeptide chains on ribosomes. d) Dissociation of p27kip1 from cyclin C kinase, promoting cell division and up-regulated expression of bcl, an anti-apoptotic factor.Figure 2: The cytokine paradigm includes calcineurin antagonists (cyclosporine [CsA] or tacrolimus [TRL]) to block the antigen-driven signal 1 and RAPA to inhibit costimulatory signal 2, thereby mitigating transcriptional activation of cytokine synthesis during progression from G0 to G1. mAbs to IL-2R block the capacity of IL-2 to trigger signal 3 transduction events, which are inhibited by RAPA, preventing G1 progression (modified from reference 21).The preclinical development of RAPA has been extensively reviewed (22). After documentation of its immunosuppressive activity in animal models by the groups of Calne (23) and later by Morris et al. (24), RAPA was demonstrated to display a high degree of synergy with CsA both in vitro (25) and in vivo (26) by the rigorous median effect analysis. These findings provided the foundation for the drug's initial clinical development in renal transplantation, anticipating that synergy with lower doses of CsA would not only more effectively prevent rejection, but also minimize CNA-induced toxicity. Two observations suggest other unique properties of RAPA that may be exploited in future controlled clinical trials. First, high doses of RAPA block the proliferative responses to cytokines by vascular and smooth muscle cells after mechanical injury, such as balloon angioplasty, or allorejection (reviewed in 27). In a nonhuman primate model, supratherapeutic concentrations of RAPA stabilized, and possibly reversed, the intimal vascular lesion caused by the progression of immune injury in aortic allografts (28). Studies are underway to assess the potential contributions of adjunctive agents to potentiate this effect at therapeutic levels of RAPA. Second, by virtue of its inhibition of bcl-2, RAPA may produce a tolerogenic pro-apoptotic effect, in apparent contradistinction to high doses of CNAs. RAPA treatment concomitant with mAb blockade of the costimulatory signal by anti-CD154 in mice induces tolerance (29,30), and the combination of RAPA and anti-B7 in nonhuman primates seems to facilitate graft survival (31). Thus, RAPA may mitigate the vasculopathic response to immune or mechanical injury and facilitate tolerance induction. A compelling aspect of RAPA therapy is the absence of the vasomotor renal side effects exhibited by the CNAs CsA and tacrolimus (TRL). Treatment with RAPA preserves glomerular filtration rates (GFR) and renal blood flow in normal (32), salt-depleted (33), and spontaneously hypertensive (34) rats, as well as in micropuncture preparations (35). Although initial studies in salt-depleted rats suggested that high doses of RAPA potentiate CsA-induced nephrotoxicity (36), recent experiments demonstrate that these adverse effects are caused by pharmacokinetic (PK) interactions that elevate renal tissue CsA concentrations disproportionately to whole blood drug levels (37). Indeed, a median effect analysis based upon renal tissue CsA concentrations suggests that RAPA displays a protective effect, which has been postulated to be related to inhibition of the intrarenal angiotensin II cascade (37). However, RAPA does produce a dose-dependent tubular toxicity in rats, which seems to be caused by delayed recovery of tubular epithelial function after injury (36). II. Clinical Pharmacology An understanding of the PK behavior of immunosuppressants is critical to guide the selection of doses and administration schedules, to predict food and drug interactions, and to assess the impact of ethnicity, age, gender, and organ function on drug exposure. The RAPA data presented in Table 1 were derived from complete concentration-time profiles in 690 subjects, and trough (Cminss) measurements from nearly 1000 patients in 40 clinical studies. The subjects included healthy volunteers, stable and de novo renal transplant recipients, children and adults on dialysis therapy, and patients with hepatic impairment or psoriasis. Table 1: Steady state pharmacokinetic parameters of RAPA in various patient populationsRAPA has been detected in whole blood samples by two high performance liquid chromatography (HPLC) methods specific for parent compound; ultraviolet wavelength (UV) (38,39) and mass spectroscopy (MS) (40). A third automated immunoassay (IMx, Abbott, N. Chicago, IL) is less selective for parent compound because it displays a 42.5% cross-reactivity with metabolites (41). Because the parent compound, not metabolite, concentrations determine biologic activity (42), HPLC/UV and HPLC/MS are the reference measurement methods used at present. RAPA systemic bioavailability (F) is approximately 14%, and the drug shows dose proportionality (43) with a maximal concentration at about 1 hr. RAPA is manufactured as an oral solution and a tablet, which are bioequivalent (44,45). RAPA is widely distributed in tissues (19 L/kg) (46) and more extensively partitions into blood cells (B) compared with plasma (P), with B/P ratios ranging from 36 in renal transplant recipients to 79 in healthy volunteers. The results of in vitro experiments using human liver microsomes suggest that cytochrome P450 3A4 is the major biotransformation system (47), generating inactive hydroxy, di-hydroxy, hydroxy-demethyl, didemethyl, 7-0 demethyl, and 41-0 demethyl metabolites (48). More than 90% of drug-associated radioactivity has been recovered in feces. Urine represents a minor route of elimination (2.2%). The average elimination half-life (t1/2) of 60 hr, albeit dose-independent, shows the greatest interpatient variation, particularly among individuals with hepatic impairment (110 hr) or in the pediatric age group (as low as 10 hr), but not among subjects of African-American versus Caucasian ethnicity. Adult stable and de novo renal transplant recipients display 38% intersubject and 45% intrasubject coefficients of variation in steady-state oral clearance (unpublished data on file, Wyeth-Ayerst Research). Because of this variability, therapeutic drug monitoring is recommended. Cminss determinations provide an adequate index of RAPA exposure for clinical use, displaying a robust correlation with AUC values (r2=0.95) (43,49) (Fig. 3). Figure 3: Correlation between trough level (Cminss) with area under the concentration-time curve (AUC) in de novo renal transplant patients. The solid line shows the equation (Cminss = −0.081 + 0.0294 * AUC), which fits observed values with r2=0.95. Each open circle is a paired observation of AUC and Cminss values. The dotted lines show the 95% prediction interval (reprinted from reference 91 with permission).Although high-fat meals slow the rate of but slightly increase the extent of RAPA absorption (50), administration with either orange juice or water produces equivalent exposures. Prominent interactions occur with other drugs that serve as substrates for CYP450 3A4. RAPA exposure is increased by diltiazem and ketoconazole and decreased by rifamycin and anticonvulsants (Table 1). Adding RAPA to the regimen of patients treated with CsA-Prednisone (Pred) produced a modest increase in steroid concentrations (51), which were not significantly different between patients on RAPA versus CsA base therapy (52). Furthermore, concomitant therapy with RAPA resulted in higher mycophenolate mofetil (MMF) exposure than did CsA and the pharmacokinetic interaction probably explains the exaggerated myelosuppressive side effects (53) that are shared by the two agents. Of greatest interest is the interaction between RAPA and CsA. RAPA concentrations are increased by concomitant versus spaced administration of Neoral, the microemulsion formulation of CsA (54) (Fig. 4) but not by concomitant administration of Sandimmune, a finding of particular benefit in the early posttransplantation period when adequate drug exposure is critical for effective immunosuppression (for review, see 55). Conversely, RAPA increases CsA exposure approximately 2-fold, presumably because of competition for metabolism by CYP 3A4 (56) and possibly drug extrusion by p-glycoprotein. Figure 4: Effect on RAPA AUC of CsA administration Neoral microemulsion either concomitant with (simultaneous, A ▪) or 4 hours after (staggered, B □) CsA using a crossover design in 20 patients. The difference between the two regimens was significant, P <0.001 by two-tailed Wilcoxon signed rank test (reprinted from reference 54 with permission).III. Clinical Development A. Phase I and II studies. The first clinical study employed a blinded randomized design to examine the safety of RAPA (1 to 13 mg/m2) versus placebo added to the CsA-Pred regimen of quiescent renal transplant patients (57) (Table 2). A dose-dependent reduction in mean platelet number and, to a far lesser extent, leukocyte count, was accompanied by increased serum cholesterol and triglyceride values. There were no changes in GFR, blood pressure, or liver function test results. Table 2: Phase I, II, and III clinical trials of RAPA in renal transplantationIn the Phase I/II trial, 40 recipients of living-donor renal transplants were treated de novo with ascending doses of RAPA (0.5, 1.0, 2.0, 3.0, and 5.0 mg/m2 per day; n=4 per group) added to a baseline regimen of full concentration-controlled exposure to CsA and tapering doses of Pred (58). An African-American male recipient of a spousal kidney who was treated in the lowest dose group experienced the only acute rejection episode. Because of the potent immunosuppression displayed by the other 19 patients (5% acute rejection rate), two subsequent cohorts of 10 recipients each were treated with 7 mg/m2 RAPA and full exposure to CsA accompanied by steroid withdrawal at 1 week or 1 month after transplantation. The one acute rejection episode in each group produced an overall acute rejection rate of 7.5% (3/40) among the CsA-RAPA group in comparison to 35% in a historical CsA-Pred cohort, suggesting that early withdrawal of corticosteroids may be feasible in RAPA-treated patients. A multicenter Phase IIB study demonstrated that, despite the protocol-mandated administration of Sandimmune at doses producing reduced drug exposure, the addition of RAPA to the regimen achieved rejection rates among non–African-American recipients of cadaveric renal transplants as low as those displayed by patients who received full CsA exposure with RAPA (59). Two additional Phase II studies explored the use of RAPA as base therapy. Although RAPA-Azathioprine (Aza)-Pred (Phase IIC1) (60) or RAPA-MMF-Pred (Phase IIC2) (53) combinations produced similar acute rejection rates to CsA-Aza-Pred or CsA-MMF-Pred regimens (about 40%), renal transplant function at 12 and 24 months was significantly better among the RAPA groups in both studies. To capture the initial immunosuppressive potency of a CsA-RAPA combination without the risk of long-term nephrotoxic complications, two large open-label trials were conducted in which CsA was withdrawn at 3 months from the regimen of patients who had experienced neither delayed graft function nor an acute rejection episode. As early as 6 months after transplantation, the renal function among patients from whom CsA was withdrawn was significantly better than those remaining on CNA therapy; however, the incidence of acute rejection episodes was numerically but not significantly greater, and all episodes responded to augmented steroid therapy (61,62). In aggregate, these studies suggest a variety of strategies for RAPA use: a RAPA-CsA combination to permit steroid withdrawal despite reduced CsA exposure; CNA avoidance using a RAPA nucleoside-inhibitor-steroid combination; or a 3-month window of RAPA-CsA-steroid therapy followed by CNA withdrawal. Open-label, long-term Phase IV studies are underway to assess the risks and benefits of these alternate approaches. B. Phase III pivotal trials. Large-scale studies using randomized and blinded designs were performed to document the therapeutic efficacy of RAPA. Thirty-eight U.S. transplant centers randomized 719 patients after transplantation once the renal graft displayed initial function. The stratification scheme was based upon ethnicity (63) because of the overwhelming impact of the African-American demographic factor on outcomes (64). Thirty-four centers in Canada, Europe, U.S., and Australia randomized 576 patients before transplantation, stratifying patients based upon living versus cadaveric donor source into a Global trial (65). Both randomization programs enrolled two patients at each dose of RAPA (2 or 5 mg daily) versus one patient for the control treatment, namely Aza (U.S. trial) or placebo (Global trial), in combination with a Cminss-controlled regimen of CsA and a stipulated protocol for tapering of steroids. Routine antimicrobial prophylaxis was mandated for Pneumocystis carinii infection and for cytomegalovirus only in cases of donor-positive to recipient-negative mismatches; otherwise, centers were stipulated to follow their customary policy. Antibody induction therapy was prohibited. Both trials demonstrated that addition of RAPA to CsA-Pred regimens reduced the incidences at 6 and 12 months posttransplantation of the clinical composite endpoint of efficacy failure: a biopsy-proven acute rejection episode, graft failure, loss to follow-up, or death (Fig. 5). The primary component of benefit was the reduced incidence of rejection episodes, which was also significant at 24 months (Table 3). Furthermore, both RAPA groups showed a significant reduction in the occurrence of moderate and severe grades of rejection, as well as in the use of antilymphocyte antibody preparations to treat rejection episodes. The incidences and causes of graft loss and of death were similar among the groups in the U.S. multicenter Phase III trial at 12 (63) and 24 months (66) (Table 4). As has been observed in the general renal transplantation population, the most common causes of patient death in the Phase III trials were infections and cardiovascular or cerebrovascular events. The latter occurred within 12 months at the highest rate among the placebo group in the Global study (65) (not shown), although there was no statistically significant difference among the groups. Because the pivotal trials that led to regulatory approval (67) required a rigorous blinded design, the effects of a protocol-stipulated reduction in CNA exposure could not be tested; CsA had to be administered in sufficient amounts to achieve acceptable rejection prophylaxis in the Aza/placebo groups. However, a synergistic interaction between CsA and RAPA was documented by a post hoc analysis of actual whole blood concentrations in samples drawn at days 1, 2, 3, 4, and 5, and at months 1, 2, 3, 4, 6, 9, and 12 posttransplantation. Logistic regression analyses revealed that RAPA Cminss concentrations of approximately 10 ng/ml (IMx method) in combination with low CsA exposures (Cminss approximately 150 ng/ml) reduced rejection rates to below 10% during the first 75 posttransplantation days, when 86% of all such episodes occurred. In the presence of RAPA, increasing CsA exposure (Cminss approximately 500 ng/ml) produced little increase in immunosuppressive potency (Fig. 6). Moreover, application of the median effect equation (68) revealed that the RAPA-CsA combination rendered 90% of patients free of acute rejection episodes at CsA exposures 2.2-fold lower than those required for the CsA-Aza or CsA-placebo groups, and at RAPA concentrations 5-fold (IMx method) lower than those with RAPA-Aza/MMF combinations. These findings documented therapeutic synergy and confirmed the results of the preclinical studies. Figure 5: RAPA decreases the incidences of composite efficacy failure and of biopsy-confirmed acute rejection episodes at 12 months in both the U.S. (left panel) and the Global (right panel) pivotal trials. The figure shows the incidences of acute rejection (□), lost to follow-up (▨), graft loss (▦), and death (▥) among patients in each treatment group. The number of patients randomized (N) is shown by numbers at the bases of the bars. P values determined by Cochran-Mantel-Haenzel test are shown above the bars for the composite efficacy failure rates and below the bars for the biopsy-confirmed acute rejection rates.Table 3: Graft survival, patient survival, and acute rejection rates using an intent-to-treat analysisTable 4: Multicenter trial intent-to-treat analysis of graft loss and deaths in the U.S. multicenter Phase III trialFigure 6: Three-dimensional representation of the probability of an acute rejection episode during the first 75 days as a function of the concentrations of CsA and RAPA. A logistic regression model was used to estimate the probability of an acute rejection episode (%; ordinate) as a function of RAPA Cminss values measured by the nonselective antibody in the IMx assay (ng/ml, x axis) and the CsA Cminss values measured with a selective mAb in the selective TDx assay (ng/ml, z axis). The dashed line at RAPA=0 ng/ml shows the impact on the occurrence of acute rejection episodes of increases in CsA Cminss exposure among patients in the placebo and Aza groups of the Phase III pivotal trials. The dashed-dotted line shows the impact on the occurrence of acute rejection episodes of increasing RAPA Cminss exposure among patients in the RAPA 2 mg/day and RAPA 5 mg/day groups at CsA Cminss value=150 ng/ml (see text for interpretation).IV. Safety and Toxicity From 1994 to 1999, more than 2500 patients received at least one dose of RAPA. In the Phase III trials alone, 643 patients received RAPA for 6 months, 527 for 12 months, and 396 for 24 months. Discontinuation from therapy within 1 year, an assessment of tolerability, occurred at similar rates: 44% in the 2 mg RAPA group, 44% in the placebo group, 51% in the 5 mg RAPA group, and 53% in the Aza group. Discontinuation rates for efficacy failure were higher in the placebo and Aza groups, lower in the 2 mg RAPA group, and lowest in the 5 mg RAPA group; whereas the discontinuation rate for adverse events was highest among the 5 mg RAPA group, and about equal for all of the other cohorts. Thus, the higher dose was more effective, but somewhat less well tolerated. Infection/malignancy. Except for the increased incidence in the 5 mg RAPA group of apthous mucosal ulcers presumed to represent herpes simplex infections (Table 5), the Phase III data showed no significant difference in infections among all groups at 12 months. In particular, the rates of cytomegalovirus (CMV), herpes zoster, and Epstein-Barr virus (EBV) infections were similar in all dose groups. The three cases of Pneumocystis carinii pneumonia all occurred in patients in whom prophylaxis had been discontinued. Table 5: Transplant-related infections and malignancies among patients in both pivotal trials within 12 monthsIn the Global study (65), the incidence of posttransplantation lymphoproliferative disorders (PTLD) in the 5 mg RAPA group was 1.4%, a value numerically higher than in other groups but neither significantly higher nor outside of the range observed in numerous studies of other immunosuppressive agents (Table 5). At the time of diagnosis of PTLD, most patients had discontinued RAPA and switched to treatment with other immunosuppressants such as antibodies, TRL, or MMF. Interestingly, a recent report suggests that a RAPA analogue has an inhibitory effect on EBV infection (69). In contrast, the incidence of skin cancer was equal among patients in the placebo and the RAPA 5 mg/day groups (n=3 each) in the Global trial and higher in the Aza than the RAPA groups for the U.S. multicenter trial. Because of its antiproliferative effects in vitro on a variety of neoplastic cell lines, RAPA has been reported to be beneficial for patients undergoing transplantation as treatment for liver cancer (70). However, further investigations are necessary before one can conclude that the antiproliferative effects of RAPA can be harnessed to retard the induction or the progression of neoplastic disease. Renal function. As suggested by findings in preclinical models, RAPA does not affect GFR. To avoid the complicating factors associated with transplantation, renal function was assessed in 117 patients diagnosed with recalcitrant psoriasis and treated with 0, 1, 3, or 5 mg/m2 of RAPA for 12 weeks. These groups showed no difference in mean serum creatinine values (Fig. 7a). Furthermore, the Phase IIB (60) and IIC (53) studies (Table 2) revealed significantly better mean creatinine values at 12 and 24 months among RAPA-treated patients compared with CsA-treated patients (Fig. 7b). However, renal tubular abnormalities have been observed among patients treated with a RAPA-based regimen, including hypokalemia and hypophosphatemia (71). Figure 7: Influence of RAPA on renal function in Phase I, II, and III clinical trials. The numbers at the top of the bars are the mean values and those at the bottom are the numbers of patients in the cohort. a) Mean serum creatinine levels after 3 months treatment of psoriasis patients with 0, 1, 3, or 5 mg/m2 RAPA (Unpublished Phase II data on file, Wyeth- Ayerst Research). b) Comparison of mean serum creatinine values among patients treated with RAPA-Aza-Pred or RAPA-MMF-Pred (□) versus CsA-Aza-Pred or CsA-MMF-Pred (▪). P values determined by one way analysis of variance. c) Comparison of mean serum creatinine values of patients treated with CsA-Pred in combination with placebo (empty bars), Aza (lightly shaded bars), RAPA 2 mg (darkly shaded bars) or RAPA 5 mg/day (solid black bars). The difference between the RAPA and the other groups as assessed by one way analysis of variance was significant at 6 and 12 months (+, P <0.001). d) Mean serum creatinine values among patients as a function of CsA discontinuation. Patients treated with RAPA-CsA-Pred either were discontinued at 3 months (▪) or remained on chronic CsA therapy (□). P values determined by analysis of variance included: Δ, P =0.014; +, P <0.001; and •P =0.003.In contrast, the mean serum creatinine levels displayed by RAPA-CsA-Pred patients in both Phase III studies were significantly higher than those of the Aza- or placebo-treated cohorts (Fig. 7c). Because RAPA seems to not have inherent glomerular toxicity by itself, it seems likely that these higher values reflect CsA rather than RAPA toxicity and represent a clinical analog of the PK interaction observed in the rat model (37). Indeed, recent studies suggest that cadaveric kidney function can be modestly, albeit significantly, improved by reduction or elimination of CsA using RAPA as base therapy (61,62,72,73). In a Phase II (II D) study of 246 renal transplant recipients and a subsequent larger Phase III (III C) trial of 525 patients, subjects who were randomized at 3 months to discontinue CsA from a RAPA-Pred regimen displayed significantly better renal function at 6, 9, and 12 months and had significant reductions in systolic and diastolic blood pressures compared with patients who continued on maintenance doses of CsA (62) (Fig. 7d). These results are consistent with the hypothesis that CsA exposure plays an important role in the nephrotoxic effects of the RAPA-CsA combination. Thus, RAPA provides a robust platform for baseline therapy to limit long-term patient exposure to CNAs with their attendant risk of chronic renal dysfunction, the basis for approval by the European Agency. Lipids. RAPA has been observed to elevate blood lipids both in the absence (74) and in the presence (57,58) of CsA. Steroid and/or CsA (75) monotherapy regimens are also known to produce dose-dependent, reversible increases in lipids. Although RAPA dampens lipoprotein lipase activity in cell culture (76), a similar effect was not documented in humans (77). In humans, RAPA significantly raises the content of high-density lipoproteins (HDL) in serum; however, it produces much greater increases in low-density lipoproteins (LDL), cholesterol, and particularly, triglycerides. Metabolic studies suggest that RAPA treatment increases circulating intermediate, low, and very-low-density fractions primarily caused by delayed clearance of lipoprotein remnants (78). The pivotal trials revealed that, during the first 2 months after transplantation, increased mean values of serum cholesterol occurred among both RAPA groups and the Aza- and placebo-treated patients. All groups showed an improvement in lipid values over time, possibly related to decreased CsA and/or steroid exposure, to improvements in dietary lipid restriction, and to an increase in physical exercise. However, members of the RAPA cohorts showed a dose-dependent delay in the return of their values toward baseline. Indeed during the first year after transplantation, RAPA-treated patients more frequently required adjunctive treatment with lipid-lowering drugs: namely, statin therapy in 57.9% (P =0.002) and 51.3% (P =0.03) versus 36.5%; and 55.7% (P =0.001) and 49.7% (P =0.007) versus 29.9% for RAPA 5 mg and RAPA 2 mg versus Aza/placebo groups in the U.S. and Global trials, respectively (63,65). At 24 months the differences were not significant for the U.S. trial, namely, 60.2% (P =0.054) and 55.6% (P =0.20) versus 44.4%, but they continued to be significant for the Global trial, namely 61.4% (P =0.008) and 60.7% (P =0.004) versus 37.3%, respectively (66). Fibrate therapy for hyperlipidemia was less frequent: 13.1% and 11.9% versus 4.9%; and 10.5% and 8.8% versus 3.0%, respectively. Both statins and fibrates were well tolerated, and most important, no increases in serious side effects of these drugs were observed in patients on RAPA versus control therapy. Of interest, box and whisker diagrams show that over the course of 2 years, the most prominent difference between treatment groups was the presence of a cohort of 10–15% of RAPA-treated patients who displayed "outlier" serum cholesterol values ≥240 mg/dL (Fig. 8a) or triglyceride values ≥400 mg/dL (Fig. 8b). With lipid-lowering therapy, the number of affected patients and their outlier lipid values decreased by 12 and particularly by 24 months (66). Figure 8: Mean serum lipid levels over time among patients in the Phase III U.S. multicenter trial. a) Cholesterol values. b) Triglyceride values. The box and whisker diagram shows the mean value as the + sign, the median value as the line, the 25th and 75th percentile as the edges of the box, the 10% and 90% range as the whiskers on the box, and outlier patients as individual asterisks (see text for interpretation).In Phase II trials comparing RAPA without CsA (Phase IIC1 and C2), the mean values of serum cholesterol and triglycerides were similar among RAPA- versus CsA-treated patients at 2 years, although lipid-lowering therapy was required in 53% versus 24% of patients, respectively (79). Although frequently observed, lipid elevations rarely represented the cause for RAPA discontinuation and were not associated with serious ad