Comparison of Neoral and Sandimmun cyclosporin A pharmacokinetic profiles in young renal transplant recipients

A major factor influencing whole blood cyclosporin A levels in young children with renal transplants is the variable absorption of Sandimmun (SIM). Neoral (NEO) is a new microemulsion of cyclosporin A (CYA) that has been reported to have better absorption characteristics. We compared the pharmacokinetics of SIM and NEO in nine renal transplant recipients aged less than 11 years (range 4.8 – 10.9 years) and observed clinical parameters during 6 months of NEO therapy. Median CYA dosage was 149 mg/m² per day (range 98 – 226). We observed an increase in the maximum CYA concentration (C_max) of 114%, an increase in area under the curve (AUC) of 71% and the time to reach C_max was reduced from 1.75 h to 1.25 h with NEO, while 12-h trough levels (C12 h) did not change significantly. AUC correlated with C12 h for SIM (r ² = 0.833) and NEO (r ² = 0.699) and also C1.5 h for NEO (r ² = 0.775). During 24 weeks’ follow-up, the coefficient of variation of CYA levels was lower for NEO (13%) than for SIM (20%). Although CYA dosages at the start and the end of 6 months on NEO were similar, only one patient was maintained on a constant dose. Four patients had acute reversible rises in plasma creatinine which responded to a 11% reduction in NEO dose; their increase in AUC was greater than those patients not showing a rise in plasma creatinine. Overall, median plasma creatinine was unchanged at the end of the study. NEO was well tolerated by the patients; temporary nausea and headache were experienced by three patients and one of them stopped NEO after 20 days. Other biochemical parameters were not significantly different on NEO. Received May 28, 1996; received in revised form and accepted October 24, 1996     

Influence of bile on cyclosporin absorption from microemulsion formulation in primary liver transplant recipients

We analysed the absorption, after oral application, of a new galenic form of cyclosporin A (CyA-NOF) in liver-grafted patients (n=12) during the 1st week (days 2–4) after transplantation. Pharmacokinetic profiling was performed with an open or clamped T tube in situ or with the T tube absent. The pharmacokinetic parameters of CyA-NOF were influenced by T tube clamping and bile diversion. The highest AUC, C_max and earliest T_max values were found in patients without a T tube in situ, indicating that absorption of CyA-NOF in patients during the early course after liver transplantation is not bile-independent. CyA-NOF, at a dose of 7.5 mg/kg, was enterally absorbed with appropriate AUC and C_max levels. Patients receiving a starting dose of 7.5 mg/kg were successfully maintained on CyA-NOF during the subsequent clinical course.     

UK multicentre study to assess the safety and tolerability of Neoral in stable renal transplant patients

Abstract  The safety and tolerability of transferring maintained renal transplant patients from Sandim-mun to Neoral is being assessed in a multicentre, open-label, single-arm study. A total of 250 patients has been enrolled and results are available from 75 patients up to 12 months post-transfer. A slight trend to higher mean cyclosporin trough levels was seen in this cohort, but trough levels were unchanged in the sub-group receiving ± 1 dose changes. The mean dose fell by 13 %. Creatinine levels showed a slight overall upward trend. Blood pressure and uric acid were unchanged and adverse events were typical of those seen with Sandim-mun. Neoral was well-tolerated. Data from the full cohort of 250 patients up to 3 months post-transfer support these findings. These results indicate that transfer from Sandim-mun to Neoral is safe and well-tolerated and provides appropriate im-munosuppression at a lower average dose than Sandimmun. The Neoral dose should be adjusted promptly, as required, to maintain the target trough level.     

Safety and steady-tate pharmacokinetics of a new oral formulation of cyclosporin A in renal transplant patients

Abstract The steady-tate pharmacokinetics of a new oral formulation of cyclosporin A (Sandimmun Neoral, NOF, a microemulsion) was compared with those of the market formulation (Sandimmun, SIM) in stable renal transplant patients. Both formulations were administered as soft gelatin capsules every 12 h with doses adjusted to provide comparable trough concentrations ( C ^ssmin). Whole blood samples were obtained over a steady-tate dosing interval (τ), and the cyclosporin A level was determined by a specific monoclonal RIA. Both formulations were well tolerated. The mean doses were 139 ± 27 mg (SIM) vs. 120 ± 19mg (NOF), indicating a milligram doseconversion factor of approximately 1:1 to yield comparable troughs. NOF exhibited a stronger correlation between AU C ^ss and C ^ssmin ( r ²= 0.821) compared with SIM ( r ²= 0.288), due in part to less variability in the NOF profiles. Average increases of 39% in C ^ssmax and 15% in AU C ^ssT during treatment with NOF were not associated with any safety concerns over the 4-week exposure to Sandimmun Neoral, as evidenced by the absence of changes in blood pressure, hematologic and biochemical parameters (including serum creatinine and blood urea nitrogen, BUN) and ultrasound of the transplanted kidney.     

Neoral induction in pediatric renal transplantation

Neoral was instituted in pediatric renal transplant patients with the hypothesis it would have more predictable kinetics than Sandimmun. However, significant questions have arisen concerning potential toxicity and dosing interval related to its rapid absorption with subsequent high initial peak. This is compounded by the fact that children appear to metabolize cyclosporine at a greater rate than adults. This combination of a rapid peak and rapid absorption may then result in lower trough levels at 12 h. We compared the trough cyclosporine levels of nine children who received Neoral with nine who received Sandimmun at the time of initial transplantation. More frequent dosing (every 8 h) was required in the Neoral population compared with the Sandimmun population for the 1st month in order to obtain comparable trough levels. Beyond the initial 4–6 weeks, trough levels were similar for Neoral and Sandimmun. Whereas 1-month creatinine levels and blood pressures were similar, the number of blood pressure medications was significantly higher in the Neoral group. At 5.5 ± 1.1 months’ followup, a single patient in the current Neoral group and in the retrospective Sandimmun group each experienced a single OKT3 allograft-treated rejection. We suggest that the area under the curve is different in Neoral than Sandimmun, and the initial dosing frequency may need to be adjusted accordingly. Received August 21, 1996; received in revised form June 27, 1997; accepted June 30, 1997     

Pharmacokinetics of a new microemulsion formulation of cyclosporin A (Neoral) in young patients after renal transplantation

Pharmacokinetics of the new galenic formulation of cyclosporin A, Neoral, (Sandoz) was examined in 12 stable young patients after renal transplantation. Six of these patients were tested before and 4 weeks after switching from the standard formulation Sandimmun to Neoral. No significant changes were observed in trough levels, t_max, C_max, and AUC_0–12h, but the absorption rate constant (K_a) increased (P=0.003). Glomerular filtration rate, as assessed by inulin clearance, increased by more than 10% in three patients and decreased in two, and was usually associated with a respective drop and rise in C_max and AUC_0–12h of cyclosporin A. The large interindividual variability in the response to the conversion to the new formulation points to a need for close monitoring of cyclosporin A trough levels and renal function after switching from Sandimmun to Neoral in this age group in order to avoid nephrotoxicity.     

Comparison of Neoral and Sandimmun for induction and maintenance immunosuppression after kidney transplantation

We compared the mean trough level/dose (L/D) ratio, mean coefficient of variation (CV) of individual patients, and graft, patient, and rejection-free survival rates of 40 renal transplant recipients receiving Neoral (CyE) with 103 consecutive renal transplant recipients receiving Sandimmun (CyA). The mean L/D ratio on the 3rd post-transplant day (16.2 vs 11.8, P < 0.04), in the 1st week (24.6 vs 16.1; P < 0.03), and 1st month (39.1 vs 28.7; P < 0.05) were higher in the CyE group. In both groups the L/D ratio improved in proportion to the duration of time post-transplant and reached a maximum in the 3rd post-transplant month. In the early post-transplant period in particular, the number of patients achieving target levels was significantly higher, and the mean dose needed to achieve target levels lower, in the CyE group. The variation in trough levels, demonstrated by the CV, was lower in the CyE group (0.41 ± 0.14) than in the CyA group (0.62 ± 0.21; P < 0.005). Actuarial 1-year patient and graft survival rates in the CyE group were 100 % and 96 %, respectively; these were similar to the 100 % and 95 % in the CyA group. The 1-year rejection-free survival rate in the CyE group was 61 % compared to 43 % in the CyA group (P < 0.02). We conclude that it is possible to obtain higher blood trough levels at lower doses by administering CyE, particularly in the early post-transplant period. The lower variability of trough levels and the higher L/D ratio in the CyE group, which are related to improved bioavailability of CyE, may explain the lower rejection rate among these patients. In this study, the microemulsion formulation of cyclosporin (CyE) was found to be more beneficial and cost-effective as induction and maintenance immunosuppression than the conventional formulation (CyA). Received: 28 January 1997 Received after revision: 13 May 1997 Accepted: 15 May 1997     

Effect of grapefruit juice on Sandimmun Neoral absorption among stable renal allograft recipients.

BACKGROUND: The oral formulation of cyclosporin A (CsA)-Sandimmun-has a highly variable absorption. The development of a CsA microemulsion-Sandimmun Neoral-resulted in increased bioavailability, and decreased variability of absorption. The first oral formulation (Sandimmun) interacted with numerous other drugs and grapefruit juice. Several of these interactions might be explained by decreased pre-systemic metabolism by a cytochrome-enzyme (e.g. CYP3A4) located in the enteral mucosa, and/or via the P-glycoprotein-mediated decreased transport of CsA back from enterocytes into the gut lumen. The purpose of this pharmacokinetic study was to investigate the interaction between Sandimmun Neoral and grapefruit juice. METHOD: Eight stable renal transplant recipients were studied during two 8-h sessions in a randomized cross-over design with 4 weeks interval. Following an overnight fast the patients ingested their habitual morning dose of Neoral either with water or with grapefruit juice. During the 8-h study period 10 blood samples were taken for determination of CsA concentration. These results formed the basis for calculation of area under curve (AUC), and half-life (t(1/2)). Maximum concentration (C(max)) and time until C(max) (t(max)) were obtained from the concentration-time profile. RESULTS: The median AUC increased by 38% (12-194%) (P<0.05) following co-administration of Neoral with grapefruit juice. There were no significant changes in C(max), t(max), and t((1/2)). CONCLUSION: Co-administration of Sandimmun Neoral with grapefruit juice resulted in an increased bioavailability of CsA, indicating unchanged pre-systemic enterocyte first-pass metabolism as compared to Sandimmun. There was no impact of an oral grapefruit juice load on systemic clearance of CsA. It seems prudent to advise renal allograft recipients treated with Sandimmun Neoral not to ingest their medication with grapefruit juice.     

Comparison of Sandimmun with a new cyclosporin derivative (IMM 125) in renal transplant patients with stable renal function

Abstract  A double-blind switchover study was carried out on 70 renal transplant patients to assess the value of a new cyclosporin derivative, IMM 125. Preclinical in vitro and in vivo studies indicated that IMM 125 was as equally immunosuppressive as Sandimmun, but that its therapeutic index should be superior. The duration of the treatment was 24 weeks. The assumption that the dosage of IMM 125 could be 2.5 times lower than Sandimmun proved to be false; three patients suffered acute rejection episodes, probably as a consequence of the low dosage, and dosage adjustments had to be made for all patients receiving IMM 125 after only a few weeks. Although IMM 125 is an effective immunosuppressive agent, it does not appear to offer advantages over Sandimmun with regard to renal function. In addition, IMM 125 causes some disturbances in liver function.     

环孢素新剂型Neoral在肾移植中的应用

对10例肾移植患者作山地明转换Neoral试验进行药效动力学对比研究，30例新移植患者应用Neoral进行临床观察。结果表明，Neoral最高血浓度（Cmax）较山地明高282±84ng／ml，达峰时间（Tmax）短0．79±0．29h，CSA暴露（AUC）大1533±169ng／（h·ml）（P＜0．05）。提示Neoral有较好地抗排斥反应作用，但由于吸收迅速，血药浓度易超过治疗窗水平而引起毒副反应。本组肝中毒达20％。认为Neoral与山地明转换比例应为1：0．8～1：0．9，以避免毒副作用发生；Neoral服药量可较山地明常规剂量减少25％。     

Reduced variability of Neoral pharmacokinetic studies in pediatric renal transplantation

In adult renal transplant recipients the Neoral area under the curve (AUC) displays less inter- and intra- individual variability than Sandimmune, and those renal transplant recipients with reduced intra-individual variability of the AUC have a lower risk for chronic rejection. As variability of Neoral pharmacokinetic (Pk) parameters has not been investigated in pediatric renal transplant recipients, we retrospectively analyzed 453 Pk profiles in 14 pediatric patients who were switched from Sandimmune to Neoral and compared the inter- and intra-individual variability of the Pk profiles on both formulations. After the switch, we observed less inter- and intra-individual variability of AUC, the 2-h concentration, and the oral clearance. As clearance with both formulations is supposedly equal, the significantly lower intra-individual variability of oral clearance is most likely an effect of less variable absorption. While the lower inter-individual variability of the Pk parameters suggests increased success in keeping cyclosporine concentrations on target, the lower intra-individual variability leads to the hypothesis that with Neoral, a lower incidence of chronic rejection might be achieved. Received: 8 February 2000 / Revised: 17 May 2000 / Accepted: 22 May 2000     

A randomized pilot study of cyclosporin G in renal transplantation

Animal studies have suggested that the analogue cyclosporin G (CyG) may be less nephrotoxic than cyclosporin A (CyA). A pilot study was therefore performed in 10 primary cadaveric renal allograft recipients who were randomized to receive posttransplant immuno-suppression with either CyA or CyG. The follow-up time was a minimum of 1 year. One graft was lost in each group. All patients in both groups experienced at least one acute rejection episode. Episodes of acute nephrotoxicity were observed in both groups. Renal function, as assessed by determinations of the serum creatinine level and chromium-ethylene diamine tetra-acetic acid (Cr-EDTA) clearance, did not differ between the two groups. Renal allograft biopsies showed a significantly higher degree of fibrosis in the CyG group than in the CyA group. All CyG-treated patients evidenced laboratory signs of acute liver toxicity, which was dose-dependent and reversible. Today, all CyG-treated patients have been switched to CyA. This study shows that immunosuppression after renal transplantation in man is possible with CyG; however, it does not seem to have any advantages over CyA.     

Reversible renal failure in renal transplant patients receiving oral acyclovir prophylaxis

Two children who developed acute renal failure in the immediate post-renal transplantation period are presented. Each was immunosuppressed with cyclosporin and was also receiving oral acyclovir for prophylaxis against cytomegalovirus infection. Renal biopsy findings suggested drug toxicity. Discontinuation of acyclovir coincided with reversal of renal impairment. As cyclosporin levels were at the lower end of the therapeutic range, we believe acyclovir to be the likely causative agent.     

Acute effect of cyclosporin on renal function following the initial changeover to a microemulsion formulation in stable kidney transplant patients

Potential differences in the acute effect of cyclosporin on renal function when dosed orally as the current market formulation or following a milligram-to-milligram conversion to a new microemulsion formulation were investigated in 14 stable kidney transplant patients. The study consisted of three sequential periods of 2 weeks duration each. Patients entered (period I) and completed (period III) the investigation with the market formulation and received the microemulsion formulation in period II; individualized cyclosporin doses remained unchanged throughout the study. Over one steady-state dosing interval at the end of each study period, whole blood cyclosprin pharmacokinetic profiles were assessed in parallel with endogenous creatinine clearances over sequential 1-to 2-h intervals. The rate and extent of cyclosporin absorption were significantly greater (P<0.01) from the microemulsion formulation with average increases of 73% in peak concentration and 44% in area under the curve compared to the market formulation. Sequential creatinine clearances exhibited a transient decrease with the nadir occurring on average between 4 and 6h post dose followed by a rapid return to baseline. Specifically in period I on the market formulation, clearances decreased from a baseline of 71.7±20.6 to a minimum of 51.1±17.9 ml/min per 1.73 m² (similar values in period III) and from 76.8±24.8 to 53.5±17.5 ml/min per 1.73 m² in period II on the microemulsion. Neither the baseline nor minimum clearances were significantly different among the study periods. Hence, the pharmacokinetic differences between the formulations did not acutely influence the pattern of glomerular filtration rate following the initial milligramto-milligram changeover in stable renal transplant patients.     

Conversion of renal transplant recipients from cyclosporin (Neoral) to tacrolimus (Prograf) for haemolytic uraemic syndrome

Abstract Five patients with cyclosporin-related haemolytic uraemic syndrome (HUS) following cadaveric renal transplantation were converted from cyclosporin- to ta-crolimus-based immunosuppression. All patients had biochemical, haematological and biopsy evidence of HUS at the time of conversion. Four of the patients showed complete resolution of the syndrome within 1 week of conversion with normalisation of haemoglobin, platelets and lactate dehydrogenase levels. In the fifth patient renal function stabilised with slow resolution of the haematological and biochemical parameters. Four of the five patients are still taking tacrolimus, one having converted back to cyclosporin due to marked hair loss. We conclude that conversion to tacrolimus appears to be an effective treatment for cyclosporin-related HUS following renal transplantation.     

Tacrolimus versus cyclosporin A: a comparative study on rat renal allograft survival

Although tacrolimus has been studied in a wide variety of experimental animal models, we are the first group to systematically study the effect of tacrolimus on rat renal allograft survival, as primary therapy and as anti-rejection therapy, in comparison with cyclosporin A (CyA). Renal grafts were transplanted from BN to LEW rats. Tacrolimus and CyA were administrated orally from day 0 for 50 days as primary therapy after grafting. Allografts were rejected after a median survival time (MST) of 8 days. Both tacrolimus und CyA significantly prolonged renal allograft survival, in a dose-dependent manner compared with the allograft controls. The most effective dose was 3.2 mg/kg, per day for tacrolimus, and 10 mg/kg per day for CyA. There was no significant difference in renal function between the group treated with the most effective dose of tacrolimus and the CyA-treated group. The percentage of detectable serum IL-2 level was 45 % in the allograft control group, but was undetectable in groups treated with the most effective dose of tacrolimus or CyA at days 3 and 6 after grafting. On the other hand, no side effects were noted in recipient rats by daily inspection, body weight change, and histological studies, although minimal tubular vacuolation was encountered in the group treated with CyA 32 mg/kg per day. In addition, the most effective doses of tacrolimus and CyA were studied as anti-rejection therapy. All of the 5 recipients treated with tacrolimus from days 2–14, and 3 of the 5 treated from days 4–16 after grafting, survived for more than 50 days. However, the MST was 19 days for recipients treated with CyA from days 2–14, and 13 days for those treated from days 4–16 after grafting. In summary, tacrolimus as primary therapy induced rat renal allograft survival with renal function and side effects comparable with those of CyA. Interestingly, when both agents were used as anti-rejection therapy, tacrolimus, but not CyA, could significantly overcome ongoing renal allograft rejection in the rat. Received: 31 March 1998 Received after revision: 9 September 1998 Accepted: 12 October 1998     

Substitution of conventional cyclosporin with a new microemulsion formulation in renal transplant patients: results after 1 year

BACKGROUND: A new galenic form of cyciosporin A has been developed, basedon microemulsion technology. The bioavailability of the compoundis reiativeiy independent of food intake and bile flow. It wasthe purpose of this prospective clinical trial to study thesafety of the microemulsion form of cyclosporin A. METHODS: Three hundred and two renal transplant patients, stratifiedaccording to transplant age, were switched from the conventionalto the new micro-emulsion formulation of cyclosporin A. A 1:1conversion ratio was used. Measurements included CsA levels,S-creatinine, liver enzymes, uric acid, and blood pressure.Measurements were performed at baseline and on days 4, 8, 15,29 and months 3, 6 and 12 after conversion. Dose adjustmentswere performed to achieve trough levels of 80–120 ng/ml. RESULTS: Within the 12-month observation period the cyclosporin dosewas reduced by 14.7% (from 204±60 mg/day at baselineto 174±51 mg/day after conversion, P<0.001). Acutely,i.e. by day 8, a 1:1 dose conversion resulted in a modest increaseof mean drug trough levels (from 114 ng/ml at baseline to 120ng/ml, P<0.01). This increase was accompanied by an increasein serum creatinine concentration, a decrease in calculatedcreatinine clearance, and an increase in uric acid values (P0.05).Liver enzymes remained unchanged while systolic and mean arterialblood pressure decreased (P<0.05). After 1 month, drug troughlevels had decreased to baseline (112 ng/ml) and remained thereuntil month 6. They were significantly lower after 12 months(102±33 ng/ml, P<0.001). Creatinine clearance valuesincreased to above baseline at 6 and 12 months. Within the 1-yearperiod there occurred 24 (=8%) episodes of biopsy-proven rejectionand seven episodes of cyclosporin-attributed nephrotoxicity. CONCLUSIONS: The 1:1 conversion from conventional cyclosporin A to the microemulsionformulation is efficacious and safe, but an initial dose reductionof 10% is advised in patients with trough levels in the high-normalrange.     

Pharmacokinetics of cyclosporine in pediatric long-term liver transplant recipients converted from Sandimmun to Neoral

Absorption of cyclosporin from the microemulsion formulation Neoral is less variable than from Sandimmun. Because of a lack of data in pediatric liver transplant recipients, the pharmacokinetic profiles with Sandimmun and Neoral were compared in a conversion study. Thirty-eight children with stable graft function were converted 2–12.3 years post-transplant at a 1:1 ratio. The trough-level (C_min) with Neoral was 123 ± 39 ng/ml versus 134 ± 29 ng/ml with Sandimmun (P = NS), the area under the time-concentration curve (AUC) was 3325 ± 1125 ng_*h/ml versus 2423 ± 846 ng_*h/ml (P < 0.001), the peak concentration (C_max) was 650 ± 280 ng/ml versus 337 ± 142 ng/ml (P < 0.001), and the median time to C_max was 2 h (range 0.5–3 h) versus 4 h (range 1–8 h; P < 0.05). The weak correlation between C_min and AUC with Sandimmun (r = 0.5; P = NS) was improved by using Neoral (r = 0.7; P < 0.001). The best predictor of AUC was the 2-h concentration (C_{2 h}) of Neoral (r = 0.9; P < 0.001). Increased absorption and a more predictable pharmacokinetic profile with Neoral permit safer therapeutic monitoring in children. The exclusive measurement of Neoral-C_{2 h} allows one to estimate drug exposure with high precision ( > 90 %). Received: 12 February 1997 Received after revision: 16 May 1997 Accepted: 5 June 1997     

Lymphocyte subsets and assessment of cancer risk in renal transplant recipients

Renal transplant recipients have a well-recognized increased risk of de novo neoplasia. In this study, we investigated whether lymphocyte subset count could predict the risk of developing noncutaneous neoplasia (NCSC) in renal transplant recipients (RTR). Between January 1995 and December 1995, lymphocyte subsets (CD4, CD8, CD19) were measured in 281 RTR. This population was studied until November 1999 for the development of NCSC. The mean follow-up was 42+/-9 months. Neoplasm was diagnosed in 22 patients (7.9%). Patients who developed a cancer were significantly older (53.8+/-6 years vs 38+/-16 years, P<0.0001), had lower CD4 (234+/-126/mm(3) vs 543+/-214/mm(3), P<0.005) and CD19 (19+/-9/mm(3) vs 51+/-22/mm(3), P<0.0001) levels, and more frequently had past histories of skin cancer (24% vs 4%, P<0.01). Cox regression revealed that high CD4 levels (RR 0.73, 95% CI 0.62-0.89 for each 100/mm(3) increase in CD4 cell count) were associated with decreased risk of NCSC, whereas age (RR 2.49, 95% CI 1.12-5.92 for each 10-year increase in age) was predictive of the subsequent development of NCSC. To conclude, CD4 cell depletion is associated with the development of solid cancers and lymphoma in RTR.     

Clinical pharmacokinetics of Neoral in pediatric recipients of primary liver transplants

Pediatric liver transplant recipients constitute a population characterized by a particularly unpredictable and poor bioavailability of cyclosporin (CyA). Even though several adult studies show that the new oral formulation of CyA, Neoral (NEO), produces better bioavailability and blood level predictability, few data describe its pharmacokinetics in children. We performed a complete analysis of the pharmacokinetics of NEO in ten small children after primary liver transplantation. Three pharmacokinetic profiles were set up with data obtained from tests taken during i. v. administration of CyA, after the first oral NEO dose, and after the last NEO dose before discharge from the hospital. The mean half-lives obtained were 8.1, 7.7, and 6.9 h, respectively, and the bioavailabilities were 22 % and 21 % for the first and last NEO doses. A large interpatient variability was observed. This was due, in part, to episodes of diarrhea that interfered with the pharmacokinetic evaluation and, in part, to the variability of post-transplant hepatic function. There was a good correlation between CyA trough levels and their related AUCs for both NEO profiles (r = 0.93 and r = 0.74, respectively). We conclude that, even though the pediatric OLT population remains more unpredictable than that of adults, NEO has a relatively rapid half-life and a remarkably improved bioavailability. Received: 29 November 1996 Received after revision: 10 April 1997 Accepted: 15 May 1997