Lymphocyte-sensitivity to glucocorticoid correlates with the sensitivity to cyclosporin A and tacrolimus in chronic renal failure patients

Abstract

Aims

Association between lymphocyte-sensitivity to immunosuppressants in transplant recipients in vitro and clinical outcomes has been demonstrated. In general, renal transplant recipients are treated with a combination of immunosuppressants such as either glucocorticoid/cyclosporin A (CsA) or glucocorticoid/tacrolimus (FK506) but the pharmacological complementarity of these drugs is still controversial. We examined relationships between the lymphocyte-sensitivities to these immunosuppressants.

Methods

We examined lymphocyte-sensitivities to prednisolone (PSL), CsA, and FK506 in vitro in a total of 190 chronic renal failure (CRF) patients and 140 healthy subjects. The lymphocyte-sensitivity was evaluated from the IC₅₀ value against mitogen-stimulated lymphocyte-blastogenesis in vitro.

Results

Statistically significant correlations of the IC₅₀ values in CRF patients between the following pairs of drugs were observed: PSL and CsA (P<0.0001; n =129, r =0.419), PSL and FK506 (P<0.001; n =54, r =0.441), and CsA and FK506 (P<0.0001; n =45, r =0.608).Similar correlations were also observed in lymphocytes from healthy subjects. The population of CRF patients who exhibited high IC₅₀ values (low sensitivities) to PSL and FK506 was significantly larger than that of healthy subjects (P<0.05).

Conclusions

Patients who showed low lymphocyte-sensitivity to either of the drugs also may exhibit low sensitivity to the others, and thus they may have a high risk of unsatisfactory outcome under combination therapy after renal transplantation. To overcome this risk, the selection of immunosuppressants is recommended to be restricted according to individual lymphocyte-sensitivities to these drugs in vitro, or alternatively, by addition of other drugs with different mechanisms for immunosuppression.

Introduction

The immunosuppressive agents cyclosporin A (CsA) and tacrolimus (FK506) have significantly improved clinical outcome of both solid organ and bone marrow transplantation [1]. So far, a combination of glucocorticoid and CsA (or FK506) has been the most common immunosuppressive therapy following renal transplantation. It is known that sensitivity of peripheral-blood mononuclear cells (PBMCs) to glucocorticoid or CsA in vitro is associated with clinical outcomes of patients after treatment with these agents, i.e. patients who have shown resistance to glucocorticoid or CsA in vitro exhibit higher incidence of acute allograft rejection and decreased graft function [2–7]. Furthermore, there are large individual differences between patients in sensitivities of PBMCs to glucocorticoids and CsA [2–7], and the incidence of patients with chronic renal failure (CRF) who show low sensitivity to glucocorticoid is significantly higher than that of healthy subjects [3]. Thus, variation between individuals of the clinical effects of these drugs cannot simply be explained only by pharmacokinetic parameters [8–10]. The pharmacodynamics of immunosuppressive agents on PBMCs from patients appears to be important in predicting clinical outcomes or selecting an appropriate drug before onset of pharmacotherapy in renal transplantation.

Glucocorticoids, CsA, and FK506 have been reported to inhibit expression and production of certain cytokines which are involved in the regulation of lymphocyte proliferation or differentiation [1, 11], and each of these immunosuppressive agents is able to suppress blastogenesis of PBMCs in vitro. Although there are some differences in immunosuppressive properties between glucocorticoids and CsA or FK506, it seemed that these immunosuppressive agents may share at least a common pathway to suppressing lymphocyte-blastogenesis [1, 11]. However, the relationship between PBMC-sensitivity to glucocorticoid and CsA or FK506 has not been demonstrated. Whether a patient with PBMCs exhibiting resistance to glucocorticoid also exhibits PBMC-resistance to CsA and/or FK506 is unknown. In this study, we examined the PBMC-sensitivities to PSL, CsA, and FK506 in a total of 140 healthy subjects and 190 CRF patients.

Methods

Subjects

After oral informed consent was obtained, heparinized venous blood was taken from healthy and CRF subjects. The collection of blood samples for the study was approved of by the local ethics committee. The study population consisted of 140 healthy adults (39 females and 101 males) 19–74 (31.5±13.5) years of age and 190 CRF patients (74 females and 116 males) 15–87 (43.0±15.1) years of age. None of the patients or healthy subjects had been taking immunosuppressive agents, including glucocorticoids, CsA, and FK506, for at least 6 months before and during the study.

PBMC culture and evaluation of drug effects

PBMCs, including lymphocytes, were separated from venous blood of healthy subjects and CRF patients, and cultured as described previously [3]. In brief, the cells were suspended at 1×10⁶ cells ml⁻¹ in RPMI 1640 culture medium (Gibco BRL, Gaithersburg, USA) containing 10% FBS, 100 IU ml⁻¹ penicillin and 100 μg ml⁻¹ streptomycin. Two hundred microlitres of the cell suspension as prepared above were placed into each of 96 flat-bottom wells of a microtitre plate. Concanavalin A (Seikagakukogyo Co., Tokyo, Japan) was added as a mitogen to each well at 5.0 μg ml⁻¹. This is the optimized lectin concentration for human PBMC-blastogenesis as previously determined in our laboratory [12]. Subsequently, 4 μl of an ethanol solution containing each immunosuppressive agent were added at 10⁻³, 10⁻², 10⁻¹, 10⁰, 10, 10², 10³, and 10⁴ ng ml⁻¹ as final concentrations. Four microlitres of ethanol were added to a control well. The plate was incubated for 80 h in 5% CO₂/air at 37 °C in a humidified chamber. The cells were pulsed with 18.5 kBq well⁻¹ of [³H]-thymidine (New England Nuclear, Wilmington, USA) for the last 16 h of incubation and then collected on a glass fibre filter paper using a multiharvester device and dried. The radioactivity retained on the filter was further processed for liquid scintillation counting. The mean value (d min⁻¹) of the duplicate or triplicate counts for each sample was used for IC₅₀ calculation. The concentration of the immunosuppressive agent which gave 50% suppression of [³H]-thymidine incorporation into PBMCs was considered as the IC₅₀ value. In a previous study, we indicated that there were no important day-to-day or technical variations in the methods of calculating the IC₅₀ values [3]. In addition, we have stated in the previous study that primary disease of CRF and duration of haemodialysis caused no significant effect on the IC₅₀ values in CRF [3].

Statistics

The Mann–Whitney test was carried out to analyse the differences in median IC₅₀ values of the immunosuppressive agents between the two subject groups. Pearson’s correlation coefficient test was used to analyse the association between PBMC-sensitivities to one immunosuppressive agent compared to those of another. Differences in the proportion of drug resistant subjects in the healthy and CRF groups were examined with Fisher’s exact probability test. In each case, P-values less than 0.05 were considered to be significant. Statistical comparisons were supported by 95% confidence intervals for differences.

Results

Correlation between PBMC-sensitivities to PSL, CsA, and FK506

We examined IC₅₀ values for PSL, CsA, or FK506 in a total of 140 healthy subjects and 190 CRF patients. These immunosuppressive agents dose-dependently suppressed blastogenesis of human PBMCs activated by concanavalin A in vitro.

The concentration of each immunosuppressive agent that induced 50% inhibition of PBMC-blastogenesis was determined individually from their dose–response curves. Typical dose–response curves for the immunosuppressive agents in PBMCs from a healthy donor are shown in Figure 1. In this case, the IC₅₀ value for FK506 was estimated to be 0.2 ng ml⁻¹, which was 17.5 and 195.5 times lower than those of CsA (3.5 ng ml⁻¹) and PSL (39.1 ng ml⁻¹), respectively. The IC₅₀ values for these immunosuppressive agents were compared among the immunosuppressive agents and between the two subject groups. As shown in Figure 2 and 3, individual IC₅₀ values for the immunosuppressive agents on PBMCs of the healthy and CRF subjects showed widely varying ranges. However, the IC₅₀ values for these immunosuppressive agents were not significantly associated with age and sex of the subjects (data not shown). Figure 2 a and d show the relationships between PSL and CsA IC₅₀ values in healthy subjects and CRF patients, respectively. There were statistically significant positive correlations in both the healthy and CRF groups (P<0.0001). Moreover, FK506, which has a similar immunosuppressive mechanism to CsA [1], also gave IC₅₀ values significantly correlated with those of PSL in both the healthy (P<0.05) and CRF (P<0.001) groups (Figure 2b,e). As expected, significant correlations were observed in the relationships of IC₅₀ values between CsA and FK506 in both the healthy subjects and CRF patients (P<0.0001) (Figure 2c,f). Thus, the subjects who showed PBMC-resistance to PSL tended to also show PBMC-resistance to CsA and/or FK506 in both the healthy and CRF groups.

Figure 1.

Typical dose–response curves of the immunosuppressive agents including PSL (▴), CsA (•), and FK506 (▪) on concanavalin A-stimulated blastogenesis of PBMCs from a healthy subject.

Figure 2.

Correlation of IC₅₀ values between PSL and CsA (a and d), PSL and FK506 (b and e), and CsA and FK506 (c and f) in healthy subjects (a, b, and c) and CRF patients (d, e, and f).

Figure 3.

Distribution of IC₅₀ value for PSL (a and d), CsA (b and e), and FK506 (c and f) in healthy subjects (a, b, and c) and CRF patients (d, e, and f). Dotted lines indicates mean±2s.d. of log IC₅₀ for each immunosuppressive agent in healthy subjects.

Differences in drug sensitivities between healthy subjects and CRF patients

The medians (ranges) and distribution of individual IC₅₀ values for PSL, CsA, and FK506 in both healthy subjects and CRF patients are presented in Table 1 and Figure 3. Differences of the mean values between the subject groups and 95% Confidence Intervals (CI) were also presented in Table 1. The IC₅₀ values for PSL in CRF patients were significantly higher than those in healthy subjects (P<0.01), while the IC₅₀ values for CsA in the CRF patients were significantly lower than those in the healthy subjects (P<0.05) (Table 1).

Table 1.

IC₅₀ values of PSL, CsA, and FK506 in healthy subjects and CRF patients.

The IC₅₀ values for each immunosuppressive agent in CRF patients had a wide range when compared with the healthy subjects (Table 1 and Figure 3). To define the normal range for IC₅₀ values, mean±2s.d. of the log IC₅₀ values for each drug were calculated from the data for the healthy subjects, and according to these upper and lower thresholds we divided the healthy and CRF subjects into three subgroups, i.e. high-, normal-, and low-sensitivity groups (Table 2). Healthy subjects and CRF patients whose log IC₅₀ values were within a range of mean±2s.d. of log IC₅₀ values calculated from PBMCs of the healthy subjects were defined as the normal sensitivity group. Those giving log IC₅₀ values over or under the normal range were defined as low- or high-sensitivity group, respectively. Consequently, normal ranges of IC₅₀ values for PSL, CsA, and FK506 were defined as 1.5–535.8, 0.6–68.2, and 0.04–0.85 ng ml⁻¹, respectively. The proportion of CRF patients who exhibited PBMC-resistance (low sensitivity) to PSL was larger than those of the healthy subjects (P<0.01) (Table 2). Only 4.1% of the healthy subjects exhibited resistance to PSL, and no healthy subject was detected whose IC₅₀ value for CsA or FK506 exceeded the upper threshold (Figure 3 and Table 2). Despite a median IC₅₀ value for CsA in the CRF patients was significantly lower than that in the healthy subjects (P<0.05), the proportion of CRF patients who exhibited resistance (low sensitivity) to CsA was larger than that of healthy subjects (P =0.0502) (Tables 1 and 2). Moreover, the proportion of FK506-resistant CRF patients was significantly larger than that of the healthy subjects (P<0.05). Although the proportion of CRF patients with PBMCs exhibiting high-sensitivity to CsA or FK506 appeared to be larger than for the healthy subjects, there was no significant difference in the proportion of high-sensitivity subjects between the healthy and CRF groups. However, the population of CRF patients showing normal sensitivity to these immunosuppressive agents was significantly small when compared with that of the healthy subjects (Table 2).

Table 2.

Comparison of proportion in three groups divided according to lymphocyte-sensitivity between healthy subjects and CRF patients.

Discussion

This study has shown significant correlations between the IC₅₀ values for PSL and CsA and FK506 in both healthy and CRF subjects. Although the correlation for CsA and FK506 might have been expected on a mechanistic basis, the correlations with PSL were rather unexpected. Thus, if the PBMCs of CRF patients showed resistance to one immunosuppressive agent they were also likely to exhibit low sensitivity to the other(s). Thus, CRF patients exhibiting PBMC-resistance to glucocorticoid, CsA or FK506 might be at increased risk of allograft rejection if these drugs were used in combination. There is clinical evidence to support this consideration insofar as CRF patients with an impaired preoperative lymphocyte-response to PSL have been shown to have a high incidence of acute allograft rejection under PSL/CsA therapy [3, 5, 6]. A similar correlation between clinical outcome and lymphocyte-response to CsA in vitro has also been reported in patients receiving CsA therapy [2, 4]. Thus, patients whose PBMCs have shown a low sensitivity to either CsA or PSL are more likely to show decreased allograft function and/or allograft rejection.

We have also shown differences between healthy subjects and CRF patients in PBMC-sensitivity to PSL, CsA, and FK506 insofar as a significantly greater proportion of the CRF patients showed resistance to PSL. This is consistant with results of our previous studies [3, 4]. Similarly, a larger population of the CRF patients showed resistance to CsA or FK506 and this is consistant with previous reports that CRF patients have altered immune responses with deficient interleukin (IL)-2 activity, increased numbers of IL-2 receptors, decreased inducibility of INF-gamma, and of IL-6 production by PBMCs [13–15]. Such alterations may be responsible for the wide distribution in CRF patients of individual PBMC-sensitivities to immunosuppressive agents and may, ultimately, also influence the clinical outcome after transplantation [2–6].

Of the 27 patients who were included in the PSL low-sensitivity group, 14 and 6 patients were additionally examined for their IC₅₀ values for, respectively, CsA and FK506. In these patients, 3 of 14 and 3 of 6 also exhibited low sensitivity to CsA and FK506, respectively, whereas others were not cross-resistant to either agent. These results suggest that at least some of the patients who exhibited PSL low-sensitivity do not show low-sensitivity to CsA or FK506. In such cases, using an alternative selection of CsA and FK506 might be possible according to the in vitro PBMC-response data. Thus, CsA might be the drug of choice for a patient classified in the low-sensitivity group for FK506 but not in the low-sensitivity for CsA. Alternatively, on the basis of our preliminary findings that PBMC-sensitivity to PSL did not correlate with PBMC-sensitivity to other types of immunosuppressive drug such as azathioprine or mizoribine, it may be possible to overcome cross-resistance to glucocorticoid and CsA or FK506 by the addition of azathioprine (and/or mizoribine) to create triple or quadruple combination regimens with glucocorticoid/CsA or glucocorticoid/FK506 [16].

The immunosuppressive potency of FK506 has been reported to be approximately 10–100 times superior to CsA [1]. Consistent with this, our results indicate that FK506 is approximately 25–28 times more potent than CsA, as estimated by median IC₅₀ values, although there was wide interindividual variation in the relative potency of the immunosuppressive agents. The highly significant correlation for the IC₅₀ values for CsA and FK506 suggest that the PBMC-resistant mechanism has a common pharmacological pathway(s), e.g. inactivation of Ca²⁺/ calmodulin-dependent serine/threonine phosphatase calcineurin [1]. The significant correlations between PSL, CsA and FK506 also suggest that there may be common pharmacological mechanisms and common resistance mechanisms when glucocorticoid and CsA or glucocorticoid and FK506 are administered.

In conclusion, the suppressive effects of PSL, CsA and FK506 on blastogenesis of PBMCs in vitro are correlated in both healthy subjects and CRF patients. These results suggest that there may be a relative contraindication to combined immunosuppressive therapy based on glucocorticoid/CsA or glucocorticoid/FK506 in patients exhibiting PBMC-resistance to either of these drugs. It must be recognized, however, that lymphocytes from some of the patients did not show cross-resistance and they might be treated with either glucocorticoid/CsA or glucocorticoid/FK506, as appropriate. Alternatively, however, the addition of an antimetabolic agent, such as azathioprine or mizoribine, might be another way to overcome cross-resistance to glucocorticoid and CsA/FK506.