Long-Term Comparison of Tacrolimus- and Cyclosporine-Induced Nephrotoxicity in Pediatric Heart-Transplant Recipients

Authors

  • Robert F. English,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • a Stephen A. Pophal,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • a Silviu-Alin Bacanu,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • d Jay Fricker,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • e Gerard J. Boyle,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • a Demetrius Ellis,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • b Kelly Harker,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • e Robert Sutton,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • a Susan A. Miller,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • a Yuk M. Law,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • a Frank A. Pigula,

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author
  • and c Steven A. Webbera,*

    1. a Division of Pediatric Cardiology, b Division of Pediatric Nephrology and c Division of Pediatric Cardiothoracic Surgery, Children's Hospital of Pittsburgh, USAd Department of Psychiatry, Biostatistics, University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh, PA 15213, USAe Division of Pediatric Cardiology, University of Florida, USA
    Search for more papers by this author

Abstract

Nephrotoxicity is an adverse effect of cyclosporine and tacrolimus. Studies comparing their long-term nephrotoxicities are lacking. This study evaluates the nephrotoxicity of these agents over a 7-year period following heart transplantation. Pediatric heart-transplant recipients receiving cyclosporine or tacrolimus as primary immunosuppression were evaluated at two centers from 1982 to 1998. Data collected included serum creatinine, height and weight prior to transplantation, at 1 and 6 months and 1 years post transplantation, and at yearly intervals thereafter. Creatinine clearance was calculated and compared between the two groups. Glomerular filtration rate was measured using Tc-99 m diethylenetriaminepentacetic acid. In total, 123 patients were evaluated. Demographic data of the two groups were comparable. Creatinine clearance demonstrated a steady decline. This decline did not differ statistically between the two groups: tacrolimus 98.9 and 90.7 mL/min/1.73 m2 at 1 month and 5 years, respectively; cyclosporine 110.7 and 81.7 mL/min/1.73 m2 at 1 month and 5 years, respectively. Four patients developed end-stage renal failure. Calculated creatinine clearance consistently overestimated glomerular filtration rate, the latter being greater than 2 standard deviations below the mean normal in 38% of patients. We conclude that the nephrotoxicities of tacrolimus and cyclosporine are comparable over the medium- to long-term in pediatric heart-transplant recipients.

Introduction

Renal dysfunction is a well-recognized adverse effect of both cyclosporine and tacrolimus (1,2). Few studies have addressed their relative nephrotoxicities (3,4). These have been limited to relatively short-term analyses, usually after kidney and liver transplantation.

To date, there have been no medium- or long-term follow-up studies comparing the nephrotoxic effects of these two agents in pediatric solid-organ recipients. The aims of this study were to assess and compare the medium- and long-term effects of cyclosporine and tacrolimus on renal function in pediatric patients undergoing heart transplantation. We hypothesized that there would be no difference in the nephrotoxicity induced by these two agents.

Patients and Methods

Patient population

We retrospectively studied pediatric patients less than 21 years of age who underwent orthotopic heart transplantation between 1982 and 1998 at two pediatric centers: Children's Hospital of Pittsburgh (cyclosporine immunosuppression for patients transplanted from 1982 to 1989, n = 25; tacrolimus immunosuppression for patients transplanted from 1989 to 1998, n = 70) and Shands Hospital, University of Florida (cyclosporine immunosuppression for patients transplanted from 1986 to 1998, n = 28). No patients were excluded on the basis of pretransplant renal function. A minimum follow-up period of 1 year was required for inclusion in the study. No patient was lost to follow-up. Data collection continued beyond age 21 in the five patients who reached this age. Comparison of the characteristics of the patients receiving cyclosporine immunosuppression from the two centers showed no significant differences. Therefore, these two groups were subsequently combined for further analysis (Table 1). For patients who were switched from cyclosporine to tacrolimus (or vice versa), data collection was terminated at the time of the switch. Target levels in the first 6 months after transplantation were 10–15 ng/mL for tacrolimus (whole-blood microparticle enzyme immunoassay) and 250–500 ng/mL for cyclosporine (whole-blood monoclonal assay). Target drug levels beyond the first 6 months after transplantation depended upon the rejection history of each patient, but had a range, using current assays, of 5–12 ng/mL for tacrolimus and 100–250 ng/mL for cyclosporine at both institutions.

Table 1. : Characteristics of patients receiving cyclosporine
 Cyclosporine (FL)Cyclosporine (Pitt)
Number of patients2825
M:F1 : 11.3 : 1
Mean age at transplant (year ± SD) 6.5 (6.1) 8.7 (5.4)
Mean weight at transplant (kg ± SD)24.0 (21.8)27.8 (20.3)
Mean creatinine at transplant (mg/dL ± SD)0.74 (0.36)0.71 (0.38)
Median follow-up period (year)2.53.0
Indications for transplant:
 Cardiomyopathy1618
 Congenital heart disease12 6
 Other 0 1 (cardiac tumor)

Data collection

Data collection included the following: serum creatinine, height and weight immediately prior to transplantation, 1 and 6 months post transplantation, and at yearly intervals (1–7 years) post transplantation. Creatinine clearance was calculated according to the method of Schwartz (Cr Cl = (ht × k)/Cr) (5). Glomerular filtration rate (GFR) was measured on patients at Children's Hospital of Pittsburgh from 1996 to 1999 using a triple-sample technique with Tc-99 m diethylenetriaminepentacetic acid. Because of the limited serial GFR data, only the GFR at the latest follow-up is included in the present analysis.

Statistical methods

Cyclosporine was administered at both locations, while tacrolimus was only administered at Children's Hospital of Pittsburgh. We therefore first compared creatinine clearances for the cyclosporine groups at the two institutions using a repeated measures analysis. This showed no statistically significant difference between the groups (p = 0.65). We then compared serial creatinine clearance data for the two drugs for the whole study population using repeated measures analysis. Statistical analysis was performed using S-Plus software (Insightful Corporation, Seattle, WA, USA).

Results

In total, 123 patients were evaluated. Characteristics of patients in each treatment group are presented in Table 2. Patient body weight and age were similar between the two groups, but there were more infants in the tacrolimus group. The serum creatinine (mg/dL) and calculated creatinine clearance (ml/min/1.73 m2) over time for each group are shown in Figures 1 and 2. We noted a small but statistically significant difference in renal function between cyclosporine and tacrolimus-treated patients at the time of transplantation, with tacrolimus-treated patients having slightly better renal function (p = 0.049). However, at 1 month post-transplantation, this difference disappeared. For all subsequent statistical inferences, we used only the creatinine clearance measures from1 month onward (when the renal function difference had disappeared), thus eliminating the difference in preoperative renal function as a potential source of variation. The result of the repeated measure analysis then showed that there were no significant differences between treatment groups. Both groups demonstrated a modest decline in calculated creatinine clearance over time.

Table 2. : Characteristics of patients receiving tacrolimus or cyclosporine
 TacrolimusCyclosporine
Number of patients7053
M:F1.7 : 11.1 : 1
Mean age at transplant (year ± SD)6.5 (6.4)7.5 (5.9)
Number of patients < 1 year of age24 (34%)10 (19%)
Mean weight at transplant (kg ± SD)22.4 (20.1)25.7 (21.0)
Mean creatinine at transplant (mg/dL ± SD)0.58 (0.43)0.73 (0.36)
Median follow-up period (year)3.03.0
Indications for transplant:
 Cardiomyopathy36 (51%)34 (64%)
 Congenital heart disease33 (47%)18 (34%)
 Other1 (Kawasaki's)1 (cardiac tumor)
Figure 1.

Mean serum creatinine at each time-point after transplantation.

Figure 2.

Mean calculated creatinine clearance at each time-point after transplantation. (Normal creatinine clearance by the Schwartz method for children aged 2–12 years is 133 ± 27 mL/min/m2) (5).

Four patients developed end-stage renal failure (3%). Two patients transplanted in 1987 and 1989 and who were switched from cyclosporine to tacrolimus immunosuppression at 4 and 1 years post heart transplant underwent renal transplantation at 7.3 and 4.7 years, respectively. A third patient was switched from cyclosporine to tacrolimus 10 years after transplantation, but was switched back to cyclosporine 1 year later and subsequently required renal transplantation. The fourth patient (on cyclosporine immunosuppression) died 12 years after heart transplantation while awaiting renal transplantation. Of the remaining patients, all had a serum creatinine less than or equal to 2 mg/dL at the latest follow-up.

Glomerular filtration rate data at latest follow-up for 69 patients with a serum creatinine less than 2 mg/dL are shown in Figure 3(a). Serum creatinine in this group was less than 1.0 in 49 patients (71%), 1.0–1.5 in 18 patients (26%) and 1.5–2.0 in two patients (3%). Note that 26 of these patients (38%) had GFR values greater than 2 standard deviationsbelow the mean normal, a finding observed at all time-points after transplantation. Figure 3(b) shows the relationship between calculated creatinine clearance and measured GFR. This demonstrates a systematic overestimation of GFR when using calculated creatinine clearance as an estimate of glomerular function. Three patients with serum creatinine levels less than 2 mg/dL had GFR values below 40 mL/min/1.73 m2.

Figure 3.

(a ) Latest glomerular filtration rate for 69 patients with serum creatinine < 2 mg/dL, as measured using Tc-99 m diethylenetriaminepentacetic acid. (b) Comparison of calculated creatinine clearance vs. simultaneously measured glomerular filtration rate. The line of identity (GFR = CrCl) is shown as the solid line.

Discussion

Tacrolimus is a potent calcineurin inhibitor with a mechanism of action similar to that of cyclosporine. It has been shown to have superior efficacy in preventing acute and refractory rejection in pediatric kidney and liver transplant recipients (6–8). It does not cause gingival hyperplasia or hirsuitism, and it permits the use of reduced steroids for prevention of rejection (9). Unfortunately, both drugs have a number of associated morbidities such as nephrotoxicity, neurotoxicity, post-transplant lymphoproliferatve disorder, and diabetes mellitus. Concerns about enhanced nephrotoxicity may be one reason that many centers continue to use cyclosporine as their primary immunosuppressant. It is therefore important that long-term studies of the nephrotoxicity associated with these two agents are performed. Evaluation of extra-renal transplant recipients affords the opportunity to assess renal function without the confounding variable of acute and chronic rejection of the renal allograft.

A number of studies have compared the short-term nephrotoxicities of cyclosporine and tacrolimus in adults, and most of these have shown comparable nephrotoxicity between the two agents. McDiarmid et al. concluded that renal dysfunction, as measured by serial creatinine levels and GFR measurements, was essentially the same for adult liver transplant recipients receiving tacrolimus as for those on cyclosporine (3). Few data exist, however, on the long-term nephrotoxicity of these agents in either pediatric or adult transplant recipients. Ellis demonstrated no significant differences at 1 year between cyclosporine-treated and tacrolimus-treated children undergoing renal transplantation with regard to decline in renal function (4). McDiarmid also showed no significant differences in nephrotoxicity at 1 year in pediatric liver transplant recipients (6). The relatively short-term follow-up in these studies, coupled with the well-known long-term adverse effects that these agents can have on renal function, underscores the need for longer-term comparative data. By extending our data collection period out as far as 7 years, this study addresses the medium- to long-term nephrotoxicity associated with these agents.

Our data demonstrate that the nephrotoxicity induced by cyclosporine and tacrolimus in pediatric heart-transplant recipients are not significantly different over the medium- to long-term. There was a rise in serum creatinine, which occurred slowly and steadily over the first 5 years after transplantation. Of concern, we found that GFR levels greater than 2 standard deviations below the mean normal appear to be common in these patients at all points after transplantation, whether on cyclosporine or tacrolimus. This emphasizes that the use of serum creatinine as the primary marker of renal function may provide an inadequate estimate of the extent of nephrotoxicity in this population.

Importantly, we observed a small group of children (3%) progress to end-stage renal failure in an unpredictable fashion. While the number of patients remains small, it should be noted that most patients in this study were followed for less than 5 years. Because increasing numbers of children are surviving beyond this time, we can anticipate an expanding cohort of patients who will develop end-stage renal failure. The challenge will be to identify who is most at risk and to modify therapy in order to prevent progression of renal disease. The expansion of available immunosuppressive agents in recent years offers an important opportunity to limit nephrotoxicity. The use of rapamycin, for example, may allow for significant reduction in the doses of calcineurin inhibitors (10). The use of multiple immunosuppressive agents with varying toxicities at low doses may also provide effective immunosuppression while limiting end-organ damage. Multi-drug regimens add complexity to the immunosuppressive regimen, however, and may increase costs and significantly erode compliance, especially among adolescents.

The primary limitation of this study is the use of calculated creatinine clearance (based upon serum creatinine and height) as an estimate of renal function. This may provide a poor measure of the degree of tubulo-interstitial fibrosis known to occur with medium- to long-term cyclosporine and tacrolimus therapy (11). Other manifestations of renal toxicity such as abnormal magnesium homeostasis, renal tubular acidosis, impaired secretion of uric acid and phosphate, and hypertension were not evaluated. Also, we did not evaluate and compare the acute nephrotoxic effects that these agents clearly possess (12,13). The retrospective nature of this study precludes uniformity in the use of ACE inhibitors, calcium channel blockers or other agents commonly used post transplantation that may impact upon renal function. Therefore, the confounding effects of these agents on calcineurin inhibitor-induced renal dysfunction could not be evaluated. The nephrotoxicities of certain nonpharmacologic factors such as polyoma virus infection were not evaluated, as these are only now coming to light.

In conclusion, we have shown that chronic nephrotoxicity in pediatric heart-transplant recipients (as assessed by serum creatinine, calculated creatinine clearance, and GFR measurements) is comparable up to 7 years after transplantation in those treated with cyclosporine and tacrolimus.

Ancillary