Objectives: To determine the relationship between clinical outcome, lymphocyte count (LC), and cyclosporine (CsA) lymphocyte maximum level (LTmL) in kidney transplant recipients.
Materials and Methods: CsA LTmL was determined in patients with biopsy-proven graft dysfunction and in patients with normal graft function. Clinical outcome was compared according to CsA LTmL, dosage, blood trough (C0) and maximum (Cmax) levels, hematocrit level, and LC.
Results: Rejecting patients had significantly lower LTmL than did those with normal graft function (27 ± 11 pg/Lc vs 71 ± 79 pg/Lc; P < 0.01) and similar LTmL to those with nephrotoxicity (27 ± 8 pg/Lc). Patients with normal graft function exhibited significantly lower LC (0.001292 ± 696 x 109/L) and serum creatinine levels (88.4 ± 35 µmol/L) when compared with rejecting patients (0.001717 ± 364 x 109/L, 132.6 ± 8.8 µmol/L) and those with nephrotoxicity (0.001884 ± 582 x 109/L, 123.7 ± 8.8 µmol/L) (P < 0.03, P < 0.001). No significant difference was observed among the 3 groups with regard to CsA dosage, C0, Cmax, mycophenolate mofetil (MMF) dosage, and mycophenolic acid (MPA) plasma levels. CsA LTmL closely correlated in an exponential (R2 = 0.98) and linear (R2 = 0.35) fashion with LC and hematocrit level, respectively. Conversely, CsA Cmax failed to correlate with C0 and these 2 latter parameters. Weak correlations were observed between CsA Cmax and its corresponding LTmL.
Conclusions: CsA LTmL appears to correlate better than CsA Cmax with rejection-free outcome and LC. An increase in hematocrit appears to have an adverse effect on CsA lymphocyte binding. CsA LTmL may offer a new alternative for CsA monitoring in kidney transplantation.
Key words : Lymphocyte binding, Immunosuppression, Nephrotoxicity, Bioavailability, Bioactivity
Several cyclosporine (CsA) therapy monitoring strategies such as blood trough (C0), area under the curve, exposure index, and most recently CsA whole blood concentration at 2 hours postdose (C2) [1-3] have been advocated as reliable indicators of CsA efficacy and need for dosage adjustment, despite conflicting results regarding their close correlation with clinical outcome [4,5]. Furthermore, previous studies associate the peak CsA concentration with the greatest intraday decline in renal function  and a higher incidence of acute CsA nephrotoxicity . We have recently reported a close correlation between renal histologic findings, mainly between rejection and CsA lymphocyte trough level (LT0L) . Moreover, we have shown , like others , a poor relationship between these clinical and biological parameters and those routinely used in CsA therapy monitoring such as C0 and CsA dosage.
Interindividual variations in CsA pharmacokinetics and immune responsiveness are well documented [8-12]. This may explain the poor correlation between graft outcome and commonly used CsA pharmacokinetics-based monitoring strategies. These variations probably result from differences among patients in drug binding or incorporation into the lymphocyte, which appear to be partially independent of CsA levels [7,9]. In addition, an inverse correlation between the CsA lymphocyte content and the immune response, defined by the degree of calcineurin inhibition (CNI) , the degree of reduction in lymphocyte proliferation , and the total LC  has been well documented. Therefore, it would be more advantageous to monitor CsA at its site of action, the lymphocyte. This might more accurately reflect the immunologically relevant drug concentration and hence, provide a better correlate with clinical events.
The aims of our study were 1) to determine CsA lymphocyte maximum levels (LTmL) in patients with biopsy-proven graft dysfunction and in matched controls with normal graft function; 2) to compare clinical outcomes defined by clinical diagnosis and serum creatinine (Scr) levels in relation to CsA LTmL, LT0L, C0, Cmax , dosage, MMF dosage, MPA plasma levels, and LC; and 3) to correlate LTmL with all these monitoring parameters.
Materials and Methods
Thirty-five patients were included in the study. Twelve patients with graft dysfunction (Scr, 128.1 ± 8.8 µmol/L) underwent 12 graft biopsies within the first 6 months posttransplantation. A control group of 23 patients with an uneventful course within the same posttransplant period and with normal graft function (Scr, 88.4 ± 35 µmol/L ; P < 0.001) was chosen for comparison. All but 1 was a first-time kidney transplant patient. All patients received similar induction therapy with an anti-thymocyte globulin (ATG-Fresinius) during the first week posttransplantation to maintain a CD4 count of <= 50 for a total of 3-4 mg/kg body weight and were maintained on a CsA-based (initial dosage, 7 mg/kg) triple therapy with MMF and prednisone. MMF dosage was started at 2 g per day (1 g b.i.d.) and adjusted to maintain an MPA plasma level between 1.6 and 2.5 µg/mL. Prednisone dosage was tapered to 0.1-0.2 mg/kg bodyweight at 3 months posttransplantation in a similar fashion for all groups. Both groups were matched for age, sex, weight, donor age and type, number of HLA mismatches, and panel reactive antibody (PRA) negativity. Patients with graft dysfunction were divided into 2 groups: those with biopsy-proven acute rejection (REJ group, n = 7) and those with biopsy-proven CsA nephrotoxicity (TOX group, n = 5). Those with normal graft function were assigned to the control group (NOR group, n = 23).
Patients had their graft biopsies performed within the first 6 months posttransplantation at 1 month (1 REJ, 3 TOX), 3 months (3 REJ), 5 months (2 REJ, 2 TOX), and 6 months (2 REJ). Graft biopsies were performed in cases with no improvement; persistent fluctuation; or continuous or new rise in Scr (>= 30% from baseline) despite CsA tapering or ganciclovir IV therapy in cytomegalovirus-polymerase chain reaction (CMV-PCR) positive patients; and normal graft ultrasound. Rejection and CsA nephrotoxicity were diagnosed according to the Banff criteria .
CsA monitoring parameters in the whole blood (C0, C1 and C2) and in the lymphocyte (LT0L and LTmL) were determined using monoclonal antibodies (Abbott TDX method) for the former and the Masri and colleagues method  for the latter. All parameters in the 2 experimental groups and in the NOR group were obtained at the same time after transplantation (during graft biopsy). Cmax and LTmL, respectively, were considered as representing the highest CsA whole blood concentration and its corresponding maximum lymphocyte content. CsA blood and lymphocyte levels were expressed in ng/mL and pictogram per lymphocyte (pg/Lc) respectively. In addition, CBC, total LC, Scr, and MPA plasma levels (expressed in µg/mL) were determined. CMV-PCR was obtained only in patients with graft dysfunction. Results are expressed as means ± SD and were compared using the Student t and chi-square or Fisher exact test when appropriate. Differences were considered statistically significant at a value of P < 0.05.
Among patients with graft dysfunction, 7 exhibited acute rejection, and the remaining 5 showed CsA nephrotoxicity. Significant interpatient variability was observed in relation to CsA LTmL. Ten (29%), 19 (54%), and 6 (17%) patients had their CsA LTmL >= 40 pg/Lc (level 1), between 20 and 40 pg/Lc (level 2), and <= 20 pg/Lc (level 3) respectively. The CsA monitoring parameters (dosage, C0, Cmax, LT0L, Scr, and LTmL) and LC are summarized in Table 1. While the differences in CsA dosage, C0, and Cmax were not statistically significant among the 3 groups, both CsA LT0L (18 ± 18 pg/Lc) and LTmL (71 ± 11 pg/Lc) were significantly higher in the NOR group (P < 0.05 and P < 0.01 respectively) when compared with both the REJ (10 ± 6 pg/Lc and 27 ± 11 pg/Lc) and the TOX (11 ± 5 pg/Lc and 27 ± 8 pg/Lc) groups, which had comparable values. Interestingly, patients with the highest CsA lymphocyte levels (NOR) had the lowest LC (0.001292 ± 696 x 109/L; P < 0.03) when compared with either group with graft dysfunction (REJ, 0.001717 ± 364 x 109/L; TOX, 0.001884 ± 582 x 109/L). Moreover, differences in MMF dosage (1.9 ± 0.4 g, 1.67 ± 0.5 g, and 1.6 ± 0.7 g) and MPA plasma levels (1.8 ± 0.5 µg/mL, 2.4 ± 0.7 µg/mL, and 2.2 ± 1 µg/mL) in the 3 groups were not statistically significant. Furthermore, only 1 patient (10%) at the LTmL level 1 experienced a mild grade-1 acute rejection versus 21% (4) and 33% (2) at level 2 and level 3 respectively. This suggests an inverse correlation between rejection rate and CsA LTmL levels (Figure 1). In contrast, no relationship was observed regarding the rate of CsA nephrotoxicity. Interestingly, when all patients were analyzed with regard to their renal function, those with a CsA LTmL >= 40 pg/Lc (level 1) were in the majority (90%) of the NOR group and consequently, had the lowest mean Scr levels (91.9 ± 17.6 µmol/L) when compared with their level-3 counterparts with CsA LTmL < 20 pg/Lc (123.7 ± 17.6 µmol/L, P < 0.04).
When the CsA monitoring parameters (dosage, C0, and Cmax ) and the LC were compared according to the CsA LTmL (level 1 >= 40, mean 118 ± 105 pg/Lc; and level 2 < 40, mean 30 ± 5 pg/Lc; P < 0.02), no significant differences were observed except in the LC (0.001027 ± 434 vs 1771 ± 625 x 109/L; P < 0.001), which was lowest in the highest CsA LTmL level (level 1), indicating a strong inverse correlation between the CsA lymphocyte levels and the LC—a rough indicator of immune function. When data from all patients were plotted individually, CsA LTmL closely correlated in a linear fashion with CsA LT0L (R2 = 0.79) and in an exponential manner with the LC (R2 = 0.98) (Figure 2A). In contrast, the analysis of individual values of CsA Cmax failed to show any significant correlation with the LC (Figure 2B) as well as with any of their respective CsA lymphocyte levels (Figure 3) or with their corresponding C0 (Figure 4). As expected, when the data from each patient regarding CsA LTmL were plotted against hematocrit level, it fit best a linear curve, consistent with an inverse correlation (R2 = 0.35), as shown in Figure 5. This relationship was nonexistent when CsA whole blood levels were considered. Surprisingly, while 10 (28%) and 9 patients (26%) exhibited their Cmax at 1 hour after CsA ingestion (T1) and 2 hours after CsA ingestion (T2) respectively, 16 patients (46%) had a Cmax fluctuating between T1 and T2, suggesting a variable temporal relationship between Cmax and C2.
Our results indicate a wide interindividual variation in the CsA LTmL, thus confirming our previously reported findings  on CsA LT0L. These variations may be related to the difference in the plasma level of cyclophilin B (CyPB) or in its binding site’s expression on the membrane of CsA-sensitive T cells, as was recently reported . They also could be the result of a difference in mutation of the MDRD1 gene , which codes for a transmembrane efflux pump (P-glycoprotein) that has, as substrates, numerous important drugs with narrow therapeutic windows such as CsA. These variations may be responsible for the interpatient variability in sensitivity to CsA [9-12]. In fact, it is well established that genetic factors and genetic variations in genes for drug metabolizing enzymes, drug transporters, and drug receptors account for about 40% of the individual variability in the efficacy and toxicity of drugs [17-19].
Denys and coworkers  have demonstrated that an increase in CyPB in vitro, while maintaining the same CsA concentration, potentiates the activity of CsA in restoring a high sensitivity to the immunosuppression by promoting its specific accumulation within the lymphocyte leading to CNI. Another possible mechanism for the observed interpatient variations in CsA lymphocyte levels is the presence of additional extra lymphocytic CsA binding sites such as the erythrocytes [9,20]. While inhibition of calcineurin (CN) activity of transplant patients on CsA therapy is rapidly reversible, the recovery of CN activity in vitro, however, appears to be slow . This is likely due to the lack of extra lymphocytic binding sites such as the red blood cell compartment [9,16,20]. Although the mean hematocrit in our study was higher in the REJ group mainly during the first month posttransplantation (37%), when compared with the NOR group (31%), this difference did not reach statistical significance owing to the small number of patients. The observed inverse linear relationship, however, between CsA LTmL and hematocrit (Figure 5) in all patients combined, is in agreement with these observations and confirms the adverse impact of a high erythrocyte mass on CsA lymphocyte content. This also may explain the difference in sensitivity to CNI between in vivo and in vitro preparations, which could result from the difference between the free fraction of CsA in biological fluids and that in culture medium  for a given dose of the drug.
As expected, a low CsA lymphocyte binding state, as was observed in our LTmL level-2 and level-3 patients (LTmL < 40 pg/Lc), would negate the immunosuppressive effect of CsA through the lack of CNI. This would increase the risk of acute rejection (Figure 1, Table 1). Conversely, an LTmL >= 40 pg/Lc (level 1), an indicator of a high CsA lymphocyte binding state, would ensure adequate immunosuppression and hence minimize the risk for acute rejection. This was clearly demonstrated in 90% of the patients (Figure 1, Table 1) who maintained a CsA LTmL >= 40 pg/Lc during the 6-month posttransplantation period with a CsA dose ranging from 3-7 mg/kg. They had a rejection-free posttransplant course with normal graft function. While in contrast, all but 1 rejecting patient during the same period had a CsA LTmL < 40 pg/Lc with a similar CsA dose range (3-7 mg/kg). Any change in the LC may affect CsA LTmL. This could be due to differences in the dose of immunosuppression during either the induction phase or the maintenance therapy posttransplantation. This is unlikely to be the case in our study, since all patients were subjected to the same induction protocol, same steroid-tapering regimen, and the same MMF dosage with comparable MPA therapeutic levels. This excludes any possible interference of the above-mentioned parameters in the observed results. Furthermore, data from the NOR group were obtained at the same time after transplantation as when the graft biopsies were performed. This was reflected by similar CsA doses, and C0, and Cmax levels in all 3 groups up to the fifth month posttransplantation, indicating that the differences in CsA LTmL observed between the NOR and the REJ groups were mainly related to interpatient variations in CsA binding to the lymphocytes. In fact, patients in the NOR group exhibited a noticeably higher CsA LTmL mainly during the first and third months posttransplantation with either lower or similar total LC respectively, when compared with patients with acute rejection despite similar CsA dosages, C0, and Cmax.
The differences between the groups, although important, did not reach a statistically significant difference when the data were analyzed on a monthly basis. This is due to the small number of patients with graft dysfunction at each posttransplant period. Our current results seem to confirm our previously reported findings  on the relationship between CsA LT0L and the LC. The observed strong exponential relationship between the CsA LTmL and the LC (a rough marker of the immune state, Figure 2A) is suggestive of a close association between the CsA lymphocyte content and the degree of its immunosuppressive activity, and is in agreement with previous reports [10,13]. Interestingly, our results failed to show any association between CsA nephrotoxicity and CsA LTmL (Figure 1), thereby confirming our previous findings  on CsA LT0L. Surprisingly, patients with the highest CsA lymphocyte exposure (level 1, LTmL >= 40 pg/Lc) exhibited the lowest rate of CsA nephrotoxicity.
We recently reported a similar observation with CsA LT0L . Our current results appear to be in contradiction with those reported by other investigators [5,23] on the increased incidence of CsA toxicity in patients with marked whole blood exposure, defined by a high area under the curve and C2. This could be explained by the fact that a high CsA lymphocyte binding state, unrelated to whole blood level (as we have previously and currently show), would lead to a marked CsA incorporation into the lymphocyte. This increase in the intralymphocytic compartment shift of CsA, may reduce renal exposure to the drug and hence, minimize the risk for nephrotoxicity.
Our data clearly demonstrate a poor association between the routinely used CsA monitoring parameters (dosage, C0, and Cmax) and clinical events, mainly acute rejection and CsA nephrotoxicity (Table 1). These results are in agreement with our previously reported work  and those of others [4,5,24]. These findings may be explained by recent observations [10,25] regarding the weak correlation between CsA blood levels and the degree of CNI and by our own previous [7,22] and current observations regarding the lack of association between CsA blood levels (C0, Cmax), poor correlate of the LC (Figure 2B), and the CsA lymphocyte levels (Figure 3)—indicators of CsA lymphocyte binding or content (known to inversely correlate with the CsA immunosuppressive activity) [7,10,13]. In fact, our results confirm the relationship between the CsA LTmL and the LC to fit best an exponential curve as shown in Figure 2A with an R2 = 0.98. The weak association reported in our study between Cmax and C0 (Figure 4), indicators of drug bioavailability, is in accord with recent observations [5,26]. This does contrast, however, with the close linear relationship that we demonstrated between their respective lymphocyte levels, LTmL, and LT0L, suggesting that CsA lymphocyte level measurement may represent a more-consistent and hence more-reliable monitoring method than that of the whole blood. Moreover, the poor correlation between CsA whole blood levels and both CsA lymphocyte levels and LC, as well as clinical outcome, raises the very important question of whether bioavailability is truly representative of bioactivity.
The strong association between CsA lymphocyte levels and both rejection-free outcome and the immune state, indicators of the drug immune responsiveness, clearly confirms the obvious difference between bioavailability and bioactivity and supports the measurement of CsA level at its site of action, the lymphocyte, as being a more-advantageous monitoring technique of graft function. In fact, the ideal noninvasive monitoring method for CsA therapy in organ transplantation short of a graft biopsy, remains controversial. The extensive literature that has been published during nearly 2 decades (more than 1300 articles) represents an unquestionable proof. This also includes the more recent advocation of C2 as a new alternative in CsA monitoring [27-29], which is surprisingly paradoxic to the current tendency toward immunosuppression minimization , which has been prompted by the newly emerging evidence on the major contribution of immunosuppressive drugs’ adverse effects on graft loss [31,32] and on chronic allograft nephropathy .
In the M02ART study , only 60% of the patients achieved the desired C2 target level (C2: 1.6-2.0 µg/mL), with an 11.5% overall incidence of acute rejection. Seventeen percent of the patients either experienced serious CsA-related adverse events or were disqualified from the study. In a Norwegian retrospective study  looking at C2 monitoring in stable transplant recipients 12 months after transplantation, no noticeable difference in Scr levels was observed between patients in the intermediate (450-950 µg/mL) and low (< 450 µg/mL) C2 groups. Interestingly, those with C2 levels greater than 950 µg/mL exhibited significantly higher Scr levels when compared with their intermediate and low C2 counterparts, suggesting an overexposure to CsA. Moreover, given the significant inter- and intraindividual variable temporal relationship between Cmax and C2 that we have shown in the present study, using isolated C2 monitoring only, might predispose many patients to the risk of drug overexposure and hence, to unnecessary and undesirable adverse effects, mainly in those with a CsA peak concentration at T1.
Our data failed to show any correlation between CsA nephrotoxicity and both CsA LTmL and Cmax. These results are in agreement with our previous findings  and seem to suggest that CsA renal susceptibility may be genetically determined  and hence, donor dependant [35,36] with variable degrees of recipient influence related to drug absorption, metabolism, and lymphocyte binding [17,19]. The difference in the genetic makeup between donor and recipient would therefore support this hypothesis and explain the lack of relationship between CsA nephrotoxicity and recipient immune responsiveness to CsA as being predominantly related to the drug concentration within the lymphocyte. This also explains the surprising coexistence of histologic evidence for both acute rejection and CsA nephrotoxicity on many graft biopsies. The transforming growth factor-ß1 (TGF-ß1)-mediated renal toxicity caused by the CsA induction of intrarenal TGF-ß1 gene hyperexpression [37,38] seems to represent an additional argument in favor of CsA nephrotoxicity being donor-related. In fact, preliminary results have revealed high TGF-ß1 gene expression in the donors of kidneys that developed posttransplant biopsy-proven nephrotoxicity in their corresponding CsA-treated recipients who had been low TGF-ß1 gene expressors at baseline. Furthermore, we recently reported  a high prevalence of high TGF-ß1 expression in one of our ethnic communities that is known to be particularly sensitive to the renal side effects of CsA.
In conclusion, the present study clearly demonstrates a close association between CsA lymphocyte binding level (LTmL) and both clinical outcome (reflected by a rejection-free state [LTmL > 40 pg/Lc] within the first 6 months after kidney transplantation) and LC (a rough indicator of immune response). This relationship does not appear to exist with the commonly used CsA therapy monitoring parameters (C0 and Cmax). Our data seem to confirm our previous findings on the poor relationship between all CsA monitoring techniques (CsA lymphocyte and CsA whole blood levels) and nephrotoxic effects, which may be at least partially donor-dependent and probably genetically mediated. The weak correlation between CsA C0 and Cmax, markers of drug pharmacokinetics, and both LC and CsA LT0L and LTmL, markers of pharmacodynamic effect as well as clinical outcome, strongly indicate an unquestionable discord between CsA pharmacokinetic and pharmacodynamic effects and hence, between bioavailability and bioactivity. The observed adverse impact of a higher hematocrit level on CsA lymphocyte binding (which may negate its immunosuppressive effect) should caution physicians against the rapid tapering of CsA, mainly in the early posttransplant period where the fastest rise in hematocrit, and hence of RBC mass, is commonly observed.
We propose CsA LTmL as a new CsA monitoring technique alternative using a receptor-based approach that is rapid, reliable, and cost-effective. It provides links between the drug concentration (bioavailability) and its immunosuppressive response (bioactivity), thereby yielding a new and reliable insight into CsA dosing and providing an important tool for future immunosuppression minimization protocols.
Volume : 3
Issue : 1
Pages : 293 - 300
Address reprint requests to: Antoine G Barbari, Nephrology and Transplantation Unit , Rizk
Hospital, Zahar Str. Ashrafieh, PO Box: 11-3288, Beirut, Lebanon
Phone: 00 96 11 338 931
Fax: 00 96 11 332 044
Table 1. Comparison of the different CsA therapy monitoring parameters and lymphocyte count in relation to the clinical diagnosis at the time of graft biopsy (REJ and TOX). Data in the NOR group were obtained at the same time that graft biopsies were performed.
Figure 1. Correlation between histologic findings and CsA Lc maximum binding (LTmL) levels
Figure 2A- 2B. Relationship between LC and both CsA lymphocyte maximum level (LTmL) (A) and CsA whole blood maximum concentration (Cmax) (B) in all patients. Each point represents 1 patient.
Figure 3. Relationship between CsA whole blood maximum levels (Cmax) and their respective CsA Lc binding levels (LTmL) in all patients.Each point represents 1 patient.
Figure 4. Relationship between CsA whole blood trough (C0) and maximum (Cmax) concentrations in all patients. Each point represents 1 patient.
Figure 5. Relationship between CsA LTmL and hematocrit level in all patients. Each point represents 1 patient.