Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The novel macrolide SDZ-RAD [40-O-(2-hydroxyethyl)-rapamycin] is currently in phase II/III clinical trials as an immunosuppressant coadministered with microemulsion cyclosporine (Neoral; international nonproprietary name: ciclosporin) after organ transplantation. SDZ-RAD is a semisynthetic derivative of rapamycin (international nonproprietary name: sirolimus). Although not yet proven, it is assumed that SDZ-RAD has the same molecular actions as rapamycin. The rapamycin/FK-binding protein (FKBP) complex, and probably also the SDZ-RAD/FKBP complex, binds in T lymphocytes to mTOR, the mammalian target of rapamycin. The results are inhibition of the interleukin-2-stimulated phosphorylation activation of p70-kd S6 protein kinase and blockade of cell cycle progression at the G₁-S-interface (Schuler et al., 1997; Böhler et al., 1998). While the calcineurin inhibitor cyclosporine inhibits interleukin-2 synthesis, SDZ-RAD inhibits interleukin-2-mediated T-lymphocyte proliferation. It is generally thought that inhibition of the subsequent steps of T-lymphocyte proliferation is the main reason for the synergistic immunosuppressive interaction of cyclosporine and SDZ-RAD observed in in vitro and in vivo studies (Schuler et al., 1997; Schuurman et al., 1997, 1998). Major side effects of cyclosporine include nephrotoxicity, hepatotoxicity, neurotoxicity, and hypertension (Kahan, 1989). The side effect pattern of SDZ-RAD in patients can be expected to be similar to that of rapamycin (Murgia et al., 1996): thrombocytopenia, hyperlipidemia, and gastrointestinal disorders.

Access of cyclosporine to the various body compartments is governed by cytochrome P450 (CYP)3A, ATP-binding cassette protein transporters, and binding to immunophilins (Ryffel et al., 1991; Wacher et al., 1998). Like cyclosporine, SDZ-RAD is a substrate of both CYP3A and the ATP-binding cassette protein transporter p170-glycoprotein (Crowe and Lemaire, 1998; Crowe et al., 1999). Because many of the factors that determine cyclosporine pharmacokinetics also are involved in SDZ-RAD pharmacokinetics, we hypothesized that cyclosporine might have a significant effect on SDZ-RAD tissue distribution, thus influencing its immunosuppressive efficacy and its tolerability.

Due to the poor oral bioavailability of microemulsion cyclosporine in cynomolgus monkeys (Schuurman et al., 1996), significantly higher doses than in humans are required to maintain cyclosporine trough blood concentrations in the target range proposed for clinical lung transplantation. The cyclosporine (100 mg/kg) and SDZ-RAD (1.5 mg/kg) doses used in our study were based on those used in the same animal species by Schuurman et al. (1998). These doses resulted in trough blood concentrations in the therapeutic range for patients and were found to be effective in preventing rejection of life-supporting kidney transplants for >50 days. Because coadministration of 100 mg/kg cyclosporine and 1.5 mg/kg SDZ-RAD was tolerated poorly (Hausen et al., 2000), doses of 50 mg/kg cyclosporine and 1.5 mg/kg SDZ-RAD (for comparison with the SDZ-RAD monotherapy group) and of 100 mg/kg cyclosporine and 0.3 mg/kg (for comparison with the cyclosporine monotherapy group) were combined. These doses were based on the results of tolerability studies (Hausen et al., 2000). In addition to these fixed-dose drug regimens, we also included a study group in which, as in the clinical practice, doses were adjusted according to blood trough concentrations (20-40 µg/l for SDZ-RAD and 100-200 µg/l for cyclosporine). The cyclosporine target concentrations were based on those used in patients (Oellerich et al., 1995).

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. SDZ-RAD, cyclosporine, and cyclosporin D were provided by Novartis Pharma AG (Basel, Switzerland). Acetonitrile (HPLC grade), sulfuric acid (American Chemical Society grade), methanol, and methylene chloride were obtained from Fisher Scientific (Fairlawn, NJ). Zinc sulfate, formic acid, and sodium formate (all American Chemical Society grade) were purchased from Sigma Chemical Co. (St. Louis, MO). Extraction columns (bonded phase C18; 1 ml) were from Varian Sample Preparation Products (Harbor City, CA). The internal standard used for quantification of SDZ-RAD, 28-,40-O-diacetyl rapamycin, was synthesized as described by Streit et al. (1996). Analytical columns (250 × 4 mm) filled with Hypersil (Shandon, Chadwick, UK) C8, 3-µm material were from Keystone Scientific (Bellefonte, PA). HPLC micro vials, 100-µl inserts, and Teflon screw caps were purchased from Hewlett Packard (Palo Alto, CA).

Lung Transplantation in Cynomolgus Monkeys. The animals received humane care in compliance with the Principles of Laboratory Animal Care (National Society for Medical Research) and the Guide for Care and Use of Laboratory Animals (National Academy of Sciences, published by the National Institutes of Health). The Institutional Animal Care and Use Committee of Stanford University granted approval.

Extraction of Tissue Samples. Tissue samples were thawed, weighed, and homogenized with 2 ml of KH₂PO₄ buffer [pH = 7.4 (1 M)] with a Teflon-glass manual homogenizer. One milliliter of homogenate was taken for analysis. Cyclosporin D and 28-,40-diacetyl sirolimus (in acetonitrile/sulfuric acid; pH = 3; 90:10 v/v) were added as internal standards, resulting in final concentrations of 100 µg/l of each. After addition of 2 ml of methanol/0.4 M ZnSO₄ (80:20 v/v) for protein precipitation, the samples were vortexed for 30 s and centrifuged at 1500g for 3 min. The organic supernatant was loaded on C18 extraction cartridges by drawing the samples through the columns with a 10 mm Hg vacuum. The extraction columns had previously been primed with 3 ml of water and 3 ml of acetonitrile. Immunosuppressants, metabolites, and internal standards were washed on the columns with 3 ml of water. The extraction columns were dried by drawing air for 5 min. Samples were eluted with 1.5 ml of methylene chloride. The samples were evaporated to dryness under a stream of nitrogen at 50°C. The residues were reconstituted in 120 µl of acetonitrile/0.1% formic acid (75:25 v/v) and were transferred into micro HPLC vials with conical 100-µl inserts and sealed with Teflon screw caps.

HPLC/Electrospray-MS Analysis. Extracts (100 µl) were injected onto a 250- × 4-mm analytical column filled with Hypersil C8-material of 3-µm particle size. SDZ-RAD, cyclosporine, their metabolites, and the internal standards were eluted from the analytical column with methanol and 0.1% formic acid supplemented with 1 µmol/l sodium formate. The following linear gradient was used: 0 min, 70% methanol; and 30 min, 90% methanol. The column was reequilibrated to starting conditions for 5 min before injection of the next sample. The flow rate was 0.4 ml/min and the column temperature was 65°C. The drying gas in the electrospray source was adjusted to a value of 42 (arbitrary units) and heated to 350°C. The pressure of the needle spray gas was 80 psi. The quadrupole was heated to 120°C. The mass spectrometer was run in the positive mode and the multiplier voltage was set to 1750 V and the X-ray voltage to 10 kV. For single ion detection, the mass spectrometer was focused on the [M + Na]⁺ of SDZ-RAD [980 atomic mass units (amu)], hydroxy SDZ-RAD (996 amu), desmethyl SDZ-RAD (966 amu), the internal standard 28-,40-diacetyl sirolimus (1020 amu), cyclosporine (1224 amu), hydroxy cyclosporine (1240 amu), dihydroxy cyclosporine (1256 amu), desmethyl cyclosporine (1210 amu), and the internal standard cyclosporin D (1238 amu). The dwell time per ion was 100 ms.

Assay Validation and Quantification. The assay was validated for SDZ-RAD and cyclosporine in blood with the procedures described in detail by Segarra et al. (1998). The HPLC/MS assay had the following specifications for SDZ-RAD determined in blood: linear range 0.1 to 100 µg/l (y = 0.96x + 0.05; r² = 0.99), mean intra-assay variability 6.9% (n = 10), intraday accuracy +6.8% (n = 10), interassay variability 8.0% (n = 6; 3 days), and mean analytical recovery 83%. In-process stability, freeze-thaw stability, dilution integrity, and partial volume verification were established and have been reported by Segarra et al. (1998). The specifications for cyclosporine were as follows: linear range 1 to 1000 µg/l (y = 0.93x + 15.1; r² = 0.98), intra-assay variability 7.7% (n = 10), intraday precision 3.7% (n = 10), interassay variability 9.8% (n = 5; 3 days), and mean analytical recovery 86%. Abbreviated assay validations were carried out for each of the tissues, including analytical recovery, lower and upper limit of quantitation, linearity, interassay variability, and accuracy. Samples from tissues of untreated donor monkeys were collected. Tissue samples were homogenized and cyclosporine and SDZ-RAD were added and incubated at 37°C for 30 min to allow for distribution and protein binding. The following samples were prepared: blanks (n = 3/tissue), calibration controls (four concentrations; n = 3/concentration), precision controls (three concentrations; n = 3/concentration), samples for determination of the lower (SDZ-RAD, 0.5 µg/l and cyclosporine, 10 µg/l; n = 5) and upper limit of quantitation (SDZ-RAD, 100 µg/l and cyclosporine, 1000 µg/l; n = 5). The samples were extracted and analyzed as described above. Recoveries were calculated from the quality control samples (n = 3 for each concentration). The mass spectrometer responses of the extracted samples were compared with the response after injection of corresponding amounts of internal standard or with standard solutions of the immunosuppressants (in methanol/0.1% formic acid; 9:1 v/v) directly on the analytical column. The lowest concentration that met the following criteria was accepted as the lower limit of quantitation: 80% of the samples analyzed had to be within ±20% of the nominal value, and precision and accuracy variation had to be less than 20%. The upper limit of quantitation was determined similarly.

View this table: [in this window] [in a new window]   TABLE 1 Validation of the HPLC/MS assay for cyclosporine and SDZ-RAD in monkey organ tissues Tissue samples were collected from untreated monkeys. Tissues were homogenized and cyclosporine or SDZ-RAD was added and incubated at 37°C for 30 min to allow for distribution and protein binding. The samples were extracted as described in Experimental Procedures. A concentration was accepted as the lower or upper limit of quantitation when four of five samples (80%) were within ±20% of the nominal concentration and the coefficient of variation of all samples was 20%. The calibration curve consisted of four different concentrations (n = 3 for each concentration). Interassay coefficient of variation (CV) and accuracy (Acc) were measured at three different concentrations (n = 3 per concentration). Means (n = 9) are presented.

Clinical Monitoring of Transplant Animals. Laboratory screening was performed three times a week and included differential blood counts, blood chemistry (serum lipid patterns, serum protein, liver and kidney function parameters), and serum electrolytes. Lung transplant function and rejection was monitored by 1) chest radiographs (two times per week), 2) lung function tests with a Bicore CP-100 pulmonary monitor (Bicore Monitoring Systems, Irvine, CA), 3) bronchoscopy (day 4 and day 14 after surgery), and 4) open lung biopsies 2 and 4 weeks after transplantation. The histological grade of rejection was classified according to the International Society for Heart and Lung Transplantation (A0, no rejection; A1, minimal rejection; A2, mild rejection; A3, moderate rejection; and A4, severe rejection; Yousem et al., 1996).

Data Analysis. Data were processed with ChemStation software revision A04.02 for the HPLC system and C.03.00 for the electrospray interface and MS (all from Hewlett-Packard). Concentrations of the immunosuppressants were calculated from an external standard curve and corrected on the basis of the internal standards. SPSS software, version 9.0, was used for statistical analysis (SPSS Inc., Chicago, IL). Because data were not normally distributed, tissue concentrations in the combination therapy and respective control groups were compared with the nonparametric Mann-Whitney U test and results of distribution statistics are reported as median and range (minimum-maximum). Clinical chemical and biochemical data among groups were compared by multivariate ANOVA. Correlation (Pearson correlation coefficients, two-tailed test of significance) and stepwise regression analysis (probability-of-F-to-enter, .05) were based on the data of all study groups (n = 25).

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Tissue Distribution of SDZ-RAD, Cyclosporine, and Their Metabolites

Tissue Distribution of SDZ-RAD (Group 1.5RAD). After single daily oral SDZ-RAD doses (1.5 mg/kg) for 4 weeks and 24 h after the last dose, the blood concentrations of SDZ-RAD at the time of sacrifice ranged from 3.7 to 44 µg/l (median = 9.2 µg/l) and those of its metabolites from 0 to 7.1 µg/l (median = 2.3 µg/l; n = 4; Fig. 1). SDZ-RAD and its metabolites extensively distributed into tissues (Fig. 1; Table 2). Median concentrations of SDZ-RAD and median total concentrations of its metabolites in most tissues exceeded those in blood. However, concentrations varied widely and, thus, the only organs that reached significantly higher tissue concentrations than in blood were pancreas and gallbladder (P < .05). Although there was a tendency to a higher median SDZ-RAD concentration in the transplant versus the native lung (Fig. 1), this difference was not statistically significant. It must be taken into account that due to the relatively small number of animals in the study groups, the statistical power for this analysis as well as all comparisons of the tissue concentrations reported below was less than 35%.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Distribution of cyclosporine (A), SDZ-RAD (B), and their main metabolites in tissues of lung transplant monkeys. Data are presented as mean ± S.E. (n = 4) and are given in nanograms per gram of wet tissue weight except for blood, which is given in micrograms per liter. The animals in A received 100 mg/kg/day (150 mg/kg the first week) cyclosporine (Neoral) and the animals in B received 1.5 mg/kg/day SDZ-RAD as a single daily oral dose for 4 weeks. TX, transplant.

Tissue Distribution of Cyclosporine (Group 100Cs). Cyclosporine was detected in all organ tissues examined in this study (Fig. 1). Twenty-four hours after the last oral dose, the median blood concentration of cyclosporine was 221 µg/l (range 131-255 µg/l; n = 4). Cyclosporine concentrations in the native and transplant lung were similar (Fig. 1). The cyclosporine metabolites AM1 and AM9 were present in all tissues, except AM9 was not detectable in brain tissues. Median cyclosporine metabolite concentrations, except in gall bladder and cerebrum, were lower than cyclosporine concentrations. The concentrations of cyclosporine and its metabolites exceeded those in blood in most tissues (Table 2). Again, mainly due to the high variability, only cyclosporine concentrations in fat tissue, kidney, spleen, lymph nodes, gall bladder, and colon were significantly higher than in blood (P < .05).

Effect of Cyclosporine on Tissue Distribution of SDZ-RAD (Group 1.5RAD + 50Cs). Compared with the administration of each drug alone, combination of SDZ-RAD and cyclosporine affected the tissue concentrations of both drugs. As shown in Fig. 1B, SDZ-RAD concentrations had the tendency to increase in most tissues when coadministered with 50 mg/kg/day cyclosporine (group 1.5RAD + 50Cs) compared with tissues from monkeys in the corresponding SDZ-RAD monotherapy group (group 1.5RAD). In contrast to most other organs, the average SDZ-RAD tissue concentrations in pancreas and cerebellum were lower in the presence of cyclosporine than in the SDZ-RAD monotherapy group (Fig. 2). This was because, compared with the other animals in this group, one monkey in the SDZ-RAD monotherapy group (1.5RAD) had extremely high concentrations in these organs (Fig. 1). Compared with the monotherapy group, with concomitant cyclosporine the changes of SDZ-RAD concentrations in blood, ileum, duodenum, native lung, kidney, and brainstem (Fig. 1B) reached statistical significance (all P < .04). As indicated by the tissue-to-blood partition coefficients [C_tissue (micrograms per gram)/C_blood (micrograms per liter)], the interaction with cyclosporine led to an increase of SDZ-RAD concentrations in most tissues, the extent of which was not predicted by blood concentrations (Table 2). In most tissues (Table 2; Fig. 2), cyclosporine increased the SDZ-RAD/metabolite ratio. Statistically significant correlations of SDZ-RAD trough blood concentrations with tissue concentrations were only observed in 10 of 20 tissues examined, and significant correlations for blood and tissue concentrations of SDZ-RAD metabolites were only observed in 2 of 20 tissues (Table 3).

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Comparison of the effects of SDZ-RAD on the tissue distribution of cyclosporine and its metabolites and vice versa in lung-transplant monkeys after multiple doses. A, effect of 0.3 mg/kg/day SDZ-RAD on the tissue distribution of cyclosporine and its metabolites. Tissue concentration ratios were calculated by dividing the median cyclosporine tissue concentrations in the SDZ-RAD/cyclosporine combination therapy group 0.3RAD + 100Cs (0.3 mg/kg/day SDZ-RAD + 100 mg/kg/day cyclosporine) by the median concentrations in the cyclosporine monotherapy group (100 mg/kg/day cyclosporine). B, effect of 50 mg/kg/day cyclosporine (Neoral) on the tissue distribution of SDZ-RAD and its metabolites. Tissue concentration ratios were calculated by dividing the median tissue SDZ-RAD tissue concentrations in the SDZ-RAD combination therapy group 1.5RAD + 50Cs (1.5 mg/kg/day SDZ-RAD + 50 mg/kg/day cyclosporine) by the median concentrations in the SDZ-RAD monotherapy group (1.5 mg/kg/day SDZ-RAD). *, missing value (because the metabolite concentrations in the monotherapy group were below the lower limit of quantitation of the liquid chromatography/MS assay, the ratio could not be calculated).

Effect of SDZ-RAD on Tissue Distribution of Cyclosporine (Group 0.3RAD + 100Cs). SDZ-RAD had a smaller impact on cyclosporine tissue distribution than the inverse (Fig. 1, A and B, respectively). The Mann-Whitney U test did not show any statistically significant changes, including concentrations in blood and the transplant lung. Comparing the changes in all tissues, the effect of SDZ-RAD on median cyclosporine tissue/blood ratios lacked a consistent tendency (Table 2). In general, trough blood concentrations of cyclosporine and its metabolites showed a much better correlation with their tissue concentrations than SDZ-RAD and its metabolites (Table 3).

Lung Transplant Rejection, Toxicity, and Clinical Monitoring (Table 4)

Group 1.5RAD (1.5 mg/kg/day SDZ-RAD). At the end of the observation period, three of four animals exhibited severe rejection (A4) of the transplant lung. The median weight loss at the end of the study was 22% (Table 4). There were no signs of anemia or thrombocytopenia. Serum cholesterol concentrations increased during the observation period (Table 4) but were not significantly different from the cyclosporine control group (group 100Cs).

View this table: [in this window] [in a new window]   TABLE 4 Trough blood concentrations of SDZ-RAD, cyclosporine, and their metabolites and clinical parameters The last values before sacrifice are reported (median, minimum-maximum). The histological grade of rejection was classified according to the International Society for Heart and Lung Transplantation (A0, no rejection; A1, minimal rejection; A2, mild rejection; A3, moderate rejection; A4, severe rejection (Yousem et al., 1996). Diarrhea was measured as percentage of days with diarrhea during the observation period. The weight loss was calculated in percentage based on the weight before surgery and at the day of sacrifice. Groups were compared using the Kruskal-Wallis test for independent samples.

Group 100Cs (100 mg/kg/day Cyclosporine). At the end of the study, the transplant lungs in two animals showed moderate (A3) and in the other two animals severe rejection (A4). The median weight loss during the study was 15%. Serum cholesterol concentrations increased during the observation period but the change was not statistically significant. No other changes in clinical chemical and biochemical parameters were detected.

Group 0.3RAD + 100Cs (0.3 mg/kg/day SDZ-RAD + 100 mg/kg/day Cyclosporine). The combination of SDZ-RAD and cyclosporine was more effective in preventing rejection of the lung allografts than monotherapy of either drug. At the end of the study, the biopsies of five animals showed a rejection score of A2 (mild acute rejection) and the biopsy of one animal had a rejection score of A3 (moderate acute rejection). However, in contrast to the monotherapy groups 1.5RAD and 100Cs, the animals developed significant anemia. Three animals required erythropoetin treatment and two animals required blood transfusions. Platelet counts were as low as 37,000/µl. These changes are not reflected in Table 4 because the table only includes the last values before sacrifice. From postoperative day 15 onward, serum cholesterol concentrations were higher in this group than in the control groups. However, the differences did not reach statistical significance. The median weight loss in this group during the observation period was 22% (Table 4). Three of six animals had to be sacrificed early due to anemia and renal failure (n = 2; days 20 and 23) or seizures (n = 1; day 22).

Group 1.5RAD + 50Cs (1.5 mg/kg/day SDZ-RAD + 50 mg/kg/day Cyclosporine). At the end of the study, the lung allografts of three animals had biopsy rejection scores of A0 (no rejection), one animal of A1 (minimal acute rejection), and one animal of A2 (mild acute rejection). The median weight loss was 19% (Table 4). The animals had diarrhea during 59% of the study days (median), which was more than in any other treatment group (P < .05; Table 4). Three animals expired early during the study: one animal died at day 12 due to bilateral pneumonia, and two animals were sacrificed due to intractable diarrhea and renal failure at study days 24 and 25. In the animals with renal failure, including the animal in group 0.3RAD + 100Cs, post-mortem histology showed thrombosis of the kidney blood vessels and without evidence of cyclosporine nephrotoxicity. From postoperative day 21 on, red blood cell count, hemoglobin, hematocrit, and platelets steadily decreased but erythropoetin therapy or blood transfusions were not required.

Group CCRAD + Cs (Concentration-Controlled Dosing). All six animals included in the study survived until the end of the observation period (day 28). For three animals, the observation period was extended to 35 (n = 1) and 49 days (n = 2). At the end of the study, biopsies showed moderate rejection of the transplant lung (A3) in three animals as well as mild (A2), minimal (A1), and no rejection (A0) in one animal each. Among the SDZ-RAD/cyclosporine combination treatment groups, CCRAD + Cs had the lowest incidence of diarrhea (median 26% of the observation period; Table 4). The median weight loss was 21% (Table 4). Red blood cell count, hemoglobin, and hematocrit steadily decreased but did not require therapy. Median serum triglyceride and cholesterol continuously increased during the observation period.

Correlation between Clinical Parameters and Blood and Tissue Concentrations of SDZ-RAD, Cyclosporine, and Their Metabolites. Histological rejection scores of the transplant lung were inversely correlated with SDZ-RAD concentrations in blood (r = 0.68; P < .001; n = 24), lymph nodes (r = 0.58; P < .003; n = 24), thymus (0.63; P < .001; n = 23), and the transplant lung (r = 0.58; P < .005; n = 24). Neither SDZ-RAD metabolite concentrations nor the concentrations of cyclosporine or of its metabolites were significantly correlated with the rejection scores. Stepwise regression analysis, into which the concentrations of SDZ-RAD, cyclosporine, and their metabolites in blood and tissues were entered, revealed a model that included the SDZ-RAD concentrations in lymph nodes and thymus as predictors (r = 0.88; P < .0001). Even the choice of a less stringent F-probability-to-enter of .25 did not result in a change of the proposed model.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Rejection is the most important limitation of long-term survival in clinical lung transplantation. Compared with other solid organ transplants, the incidence of acute rejection is significantly higher in lung transplant recipients and lung allografts are more resistant to immunosuppressive therapy (Trulock, 1993). Because of its sensitivity to rejection, we chose a unilateral lung transplant model to assess the efficacy of combined immunosuppressive therapy with SDZ-RAD and cyclosporine.

Although we used the same cyclosporine and SDZ-RAD doses as Schuurman et al. (1998), their results with kidney transplants were superior to our experience with monotherapy in the more stringent lung transplant model used herein. Cyclosporine trough blood concentrations in our study (Fig. 1) were well within the target concentration range of lung transplant patients.

We demonstrated a significant drug interaction of cyclosporine on SDZ-RAD, but not vice versa. In general, coadministration of cyclosporine resulted in 1) higher SDZ-RAD blood and tissue concentrations, most likely due to an increase of oral bioavailability (Crowe et al., 1999); 2) an increase of the tissue-to-blood distribution coefficients; and 3) an increase of the SDZ-RAD/metabolite ratio. Because both drugs, as mentioned above, are CYP3A as well as p170-glycoprotein substrates (Crowe and Lemaire, 1998; Wacher et al., 1998; Crowe et al., 1999), an interaction at these drug metabolism and transporter sites may explain at least part of the cyclosporine/SDZ-RAD interaction. In spite of the well established lower metabolic capacity of the small intestine compared with the liver, recent studies have demonstrated that the small intestine may play a significant role in first pass metabolism (Benet et al., 1996). The most important unresolved issue is the role of intestinal counter-transport and its functional interaction with intestinal CYP enzymes. It can be expected that counter-transport limits the access of drugs to the CYP enzymes and may prevent CYP enzymes from being overwhelmed by the high drug concentrations present in the small intestine. However, with a drug being repeatedly transported out of the mucosa cells and being reabsorbed, repeated exposure to CYP3A enzymes may lead to more efficient metabolism (Benet et al., 1996; Gan et al., 1996). In addition, the intestinal transporter/CYP barrier may control substrate availability to liver CYP enzymes. Another factor that complicates an estimate of the relative role of intestinal compared with liver first pass metabolism is the high interindividual variability of intestinal CYP3A enzymes and transporters as well as liver CYP3A enzymes (Kolars et al., 1994; Lown et al., 1997). Because in vivo the functional interaction between intestinal transporters and CYP3A enzymes seems important (Benet et al., 1996) and intestinal transporters control substrate availability to the CYP3A enzymes, it is impossible to decide whether the SDZ-RAD/cyclosporine interaction leading to increased SDZ-RAD blood concentrations and, subsequently, tissue concentrations was mainly at the CYP or the P-glycoprotein level.

Because the affinity of both drugs to CYP3A enzymes and p170-glycoprotein transporters is similar (our unpublished data; Crowe and Lemaire, 1998; Crowe et al., 1999), the lack of a significant effect of SDZ-RAD on cyclosporine concentrations can be explained by the >30-fold higher oral doses and in general >10-fold higher blood and tissue concentrations of cyclosporine compared with SDZ-RAD. Because of the wide variability of the data and the number of animals included, a potentially small effect of SDZ-RAD on cyclosporine concentrations could not be detected.

In blood, 58% of cyclosporin is bound to erythrocytes, 10 to 20% to leukocytes, and 33% to plasma proteins and lipoproteins (Lemaire and Tillement, 1982). In the plasma fraction, 46% is bound to high-density lipoproteins, 31% to low-density lipoproteins, 16% to very low-density lipoproteins, and 8% to other proteins, mainly albumin (Mraz et al., 1983; Rödl and Khoshsorur, 1990). Binding to immunophilins plays an important role in the distribution of cyclosporine into erythrocytes and tissues (Ryffel et al., 1991) and also can be expected to affect blood and tissue distribution of SDZ-RAD. However, cyclosporine and SDZ-RAD bind to different immunophilin families, cyclophilins and FKBPs, respectively, and an interaction at this level seems unlikely. Because plasma protein binding and distribution of SDZ-RAD in blood have not yet been reported, it cannot be decided whether protein-binding interactions are involved in the increased tissue-to-blood distribution coefficients for most tissues in the presence of cyclosporine.

Lung allografts in the cyclosporine/SDZ-RAD combination treatment groups showed significantly less histological rejection than in the monotherapy groups. Although the SDZ-RAD trough blood concentrations as well as SDZ-RAD concentrations in lymph tissue, thymus, and the transplant lung were clearly inversely correlated with the histological rejection scores, no statistically significant association between cyclosporine concentrations and transplant rejection could be established. This finding was confirmed by stepwise regression analysis, which did not show cyclosporine concentrations to be predictors for the histological rejection scores in the proposed model.

SDZ-RAD and cyclosporine show synergism in immunosuppression, both in vitro and in vivo (Schuurman et al., 1997). Because cyclosporine increased SDZ-RAD blood and tissue concentrations and SDZ-RAD concentrations were inversely correlated with rejection, our study indicated that the pharmacokinetic cyclosporine/SDZ-RAD interaction significantly contributes to the synergistic immunosuppressive effect in vivo.

Although monotherapy with SDZ-RAD or cyclosporine was well tolerated, several study drug-related adverse events were observed in the combination groups such as drug toxicity or overimmunosuppression. Among the most frequent major adverse events in the SDZ-RAD/cyclosporine combination groups were anemia and thrombocytopenia with hemorrhage, hyperlipidemia, and diarrhea. All of these have been associated with rapamycin-type immunosuppressants (Murgia et al., 1996; Dias et al., 1998). The prevalence of diarrhea was significantly correlated with SDZ-RAD concentrations in blood and several gastrointestinal tissues and SDZ-RAD metabolite concentrations in the liver were significantly correlated with serum cholesterol and triglyceride concentrations. The statistically significant correlation of these adverse events with SDZ-RAD rather than with cyclosporine may be explained by the effect of cyclosporine on SDZ-RAD tissue distribution resulting in increased tissue concentrations and the relatively small effect vice versa.

We also reported the tissue distribution of SDZ-RAD metabolites in the absence or presence of cyclosporine. Because of its structural similarity, SDZ-RAD could be expected to be as unstable as rapamycin (Wang et al., 1994). Therefore, sample handling was critical to obtain reliable results. Tissue samples were snap frozen immediately after collection and stored at 80°C until analysis. None of the recently described (Vidal et al., 1998), ring-open (m/z = 998 [M + Na]⁺) and ring-open, dehydrated degradation products (m/z = 980 [M + Na]⁺) were detected. One demethylated metabolite peak, which according to the fragmentation pattern was 16-O-desmethyl SDZ-RAD (Vidal et al., 1998), was detected in the tissues, and three hydroxylated metabolite peaks (m/z = 996) could be separated by our HPLC method. Because no authentic standard material was available, the exact structures of these metabolites could not be identified and the apparent total concentration of the hydroxylated metabolites was reported. Currently, the immunosuppressive activity of isolated SDZ-RAD metabolites is unknown. In our study, there was no association between SDZ-RAD metabolite concentrations and rejection scores. However, as mentioned above, SDZ-RAD metabolite concentrations in the liver and various gastrointestinal tissues were correlated with serum lipid concentrations and diarrhea, respectively.

From the cyclosporine tissue distribution data in the literature, our study can best be compared with the study by Lensmeyer et al. (1991), which reported the post-mortem tissue distribution of cyclosporine and its metabolites in transplant patients. The distribution pattern was similar to that found in our study; the major metabolites detected in tissues were AM1 and AM9. In earlier studies (Sangalli et al., 1990; Lensmeyer et al., 1991), cyclosporine or metabolites could not be detected in central nervous system fluid and tissues. In our study, cyclosporine and its metabolites were present in all parts of the brain examined and concentrations were approximately 20% of those in blood. In contrast to other investigations, we used an HPLC/MS assay that was 50-fold more sensitive than HPLC/UV assays (Lensmeyer et al., 1991). SDZ-RAD and its metabolites also were present in central nervous system tissue.

Stepkowski et al. (1996) suggested that the pharmacokinetic interaction between rapamycin and cyclosporine contributed to the synergistic effect of both drugs on heart allograft survival in the rat. These data cannot be directly compared with the data from our study because the distribution of sirolimus into rat tissue (Napoli et al., 1997, 1998) was 5- to 10-fold greater than observed herein for SDZ-RAD. Furthermore, potential differences between species and transplant organs must be taken into account. In addition, sirolimus and SDZ-RAD are substrates for different ATP-binding cassette transporters (Crowe and Lemaire, 1998), so that cyclosporine-sirolimus interaction data cannot be extrapolated to the cyclosporine-SDZ-RAD interaction.

Blood concentration-guided dosing regimens are well established clinical practice for cyclosporine (Oellerich et al., 1995). They have been proposed for rapamycin (Yatscoff et al., 1995) and can be expected to be considered for SDZ-RAD. The significant inverse correlation of SDZ-RAD tissue concentrations in lymph tissue, thymus, and transplant lung with the histological rejection scores as well as the correlation of concentrations of SDZ-RAD and its metabolites in several gastrointestinal tissues and the liver with the prevalence of diarrhea and serum cholesterol and triglyceride concentrations, respectively, indicated that a significant correlation between SDZ-RAD blood and tissue concentrations is a requirement for the clinical relevance of blood concentration monitoring. We therefore calculated the tissue-to-blood distribution coefficients (Table 2) and analyzed the correlation of trough blood with tissue concentrations (Table 3). These analyses showed that cyclosporine changed SDZ-RAD tissue/blood distribution by increasing tissue concentrations more than blood concentrations. Most likely as a result, the correlation between SDZ-RAD trough blood and tissue concentration was poorer than that between cyclosporine trough blood and tissue concentrations. Whether this outcome has a negative impact on the clinical relevance of therapeutic drug monitoring requires further assessment.

It can be expected that cyclosporine and SDZ-RAD also will interact in transplant patients. However, due to the lower cyclosporine doses required in patients than in nonhuman primates, the cyclosporine-SDZ-RAD drug interaction may be less significant in the clinical practice than in our study. In addition, the toxicity of rapamycin-type immunosuppressants is known to be species-dependent (Collier et al., 1990, 1991). Therefore, our study cannot be directly extrapolated to predict the potential toxicity of a combined cyclosporine/SDZ-RAD immunosuppressive therapy in transplant patients without restrictions.

Based on our study, we conclude that in addition to the pharmacodynamic rationale for the combination, the pharmacokinetic interaction significantly contributes to the synergistic immunosuppressive effect when SDZ-RAD and cyclosporine are combined, but that the pharmacokinetic interaction also may lead to decreased tolerability.