Differential Effect of Diarrhea on FK506 Versus Cyclosporine A Trough Levels and Resultant Prevention of Allograft Rejection in Renal Transplant Recipients



Diarrhea is the most frequently reported adverse event in patients treated with mycophenolate mofetil. Twenty-six renal transplant patients on a mycophenolate mofetil-based immunosuppressive regime with persistent afebrile diarrhea were examined. Diarrhea caused a significant rise in FK-506 trough levels despite intake of stable doses, necessitating FK-506 dose reductions of 30% to obtain pre-diarrhea trough levels. In contrast, trough levels of cyclosporine A remained stable without dose adjustments. This suggests that absorption and/or metabolism is differentially altered for FK506 compared with cyclosporine A in patients with diarrhea. In nine patients mycophenolate mofetil was reduced or stopped because of persistent diarrhea without identifiable cause. This resulted in end-stage renal disease because of chronic rejection in two patients, and in acute rejection in two patients, all taking FK506 and steroids. Therefore, dose adjustments of FK506 in patients with diarrhea must be carefully monitored, especially when doses of mycophenolate mofetil are also reduced.


Since three pivotal large-scale clinical trials have demonstrated that mycophenolate mofetil (MMF), in combination with cyclosporine A (CsA) and corticosteroids (CS), is a highly effective immunosuppressant with an acceptable safety profile (1–3), MMF has become one of the standard prophylactic immunosuppressive agents in many transplant centers. By analogy, MMF has recently also been used in combination with FK506 and CS to prevent allograft rejection (4). Despite its almost selective inhibition of lymphocytes and its antimicrobial potential, diarrhea is one of the most frequently reported adverse events in patients treated with MMF and the major reason to discontinue MMF in the immunosuppressive regimen.

Materials and Methods

Because the nature of the diarrhea in MMF-treated organ transplant patients is largely unknown, we are prospectively investigating renal transplant recipients treated with an MMF-containing immunosuppressive regimen and persistent (i.e. more than 2 weeks) afebrile diarrhea [i.e. a daily fecal output of more than 200 g, as measured by a 3-day stool collection, with separate collections for each 24 h (5)] for infections, and the morphological and structural integrity of the gastrointestinal tract. Twenty-six patients have been enrolled in the study up to now: three patients were taking MMF + CS, four MMF + CS + CsA, 16 MMF + CS + FK-506 and three MMF + FK-506. All patients enrolled in the study experienced more than 3 loose stools daily, for a period of 5 weeks (median, range 2 weeks−6 months). On inclusion, mean weight loss was 6.2 ± 3.3 kg. The diarrhea started 438 ± 493 days (mean ± SD; range 66–2164) after initiation of MMF; the post-transplant period was 1512 ± 2174 days (mean ± SD; range 66–8219). Before onset of the diarrhea, the dose of MMF was 1.6 ± 0.5 g/day (mean ± SD; range 1–3) [with 12-h trough levels of MPA of 4.0 ± 1.8 mg/L (range 0.3–7.6)] and of CS 5.3 ± 3.8 mg/day (mean ± SD; range 1–12). Six months before hospital admission, 12-h trough levels of FK-506 were 10.7 ± 2.1 μg/L (mean ± SD; range 6–15) and of CsA 104 ± 23 μg/L (mean ± SD; range 75–125).


On hospital admission, 12-h trough levels of FK-506 were significantly increased to 20.0 ± 6.8 μg/L with the intake of stable daily doses of the drug (Figure 1), and without any treatment for the diarrhea. This necessitated a significant reduction of the dose of FK-506 from 8.9 ± 4.1 to 5.6 ± 2.9 mg/day in the following days to obtain pre-diarrhea trough levels. In contrast, 12-h trough levels of CsA remained remarkably stable in patients taking CsA without significant dose adjustments (Table 1). Figure 2 shows the evolution of trough levels and doses of immunosuppressive drugs, expressed as mean monthly values before and after admission to the hospital for persistent afebrile diarrhea. The increase in levels of FK506 was not present during the month prior to hospital admission. Since patients were included only if diarrhea persisted for more than 2 weeks, and the average duration of the diarrhea was 5 weeks, it can be stated that the diarrhea preceded the elevated levels. The pattern of elevation of FK506 was independent of the underlying cause of the diarrhea (infectious – not-infectious). Once the etiology of the diarrhea was substantiated, the patients were treated accordingly: medications affecting the CYP3A4 system and gastrointestinal motility were prohibited.

Figure 1.

Individual FK506 and CsA doses and pre-dose trough levels 6 months before, on the day of hospitalization and 6 months after onset of diarrhea.

Table 1. : FK506 and CsA doses and predose trough levels 6 months before and during in-hospital observation of diarrhea and 6 months after onset of diarrhea (mean ± SD)
 6 Months beforeDuring observation6 Months after
FK506 dose (mg/day) 8.9 ± 4.1 5.6 ± 2.9 5.2 ± 2.9
FK506 pre-dose level (μg/L)10.7 ± 2.120.0 ± 6.810.5 ± 1.9
CsA dose (mg/day) 150 ± 20 175 ± 35 144 ± 13
CsA pre-dose level (μg/L) 104 ± 23  99 ± 44 105 ± 16
Figure 2.

Mean monthly FK506 and CsA doses and pre-dose trough levels and MMF doses of all patients from 6 months before the diarrhea, on hospital admission and until 6 months after onset of diarrhea (mean ± SEM) (* p < 0.0001; ** p < 0.05 paired comparison with hospital admission values).

Because of persistent diarrhea without identifiable cause, the dose of MMF was reduced by 50 ± 25% in nine patients and even stopped in four of these patients; a complete remission of the diarrhea was noted in three out of nine patients within 4 weeks. This led to a significant decrease in MMF dose from month 2 after admission (Figure 2). In two of the patients treated with MMF + FK506 + CS in whom MMF was stopped, however, renal function progressively declined and renal replacement therapy had to be restarted because of chronic allograft nephropathy superimposed by acute rejection (Banff Ib) in one of them. The diagnosis of chronic rejection was made, respectively, 4 and 5 months after onset of diarrhea or 10 and 16 months after transplantation. An acute rejection (Banff IIa) occurred in a third patient treated with MMF + FK506 + CS, necessitating a high dose of CS to restore transplant function (4.5 months after onset of diarrhea or 60 months after transplantation). At the time of renal biopsy confirming the diagnosis of chronic ± acute rejection, the FK506 trough level was 8, 6 and 4 μg/L, respectively.


Although this is an observational study – not able to establish cause and effect definitively – and although high levels of FK506 might cause diarrhea, the temporal relationship (Figure 2) and the lack of reasons for sudden changes in absorption/metabolism of FK506 strongly suggest that mucosal abnormalities during diarrhea are the most plausible explanation for the observed phenomenon. The altered bioavailability of orally administered calcineurin-inhibitors during episodes of diarrhea noted in a few cases, may have different or multiple causes (6–10). It may be due to an altered intestinal absorption because of changes in solubilization of the drugs or intestinal permeability, or because of altered hepatic or mucosal metabolism of the drugs. However, we observed a marked discrepancy between FK506 and CsA that needs further exploration: there was a 100% increase in FK506 trough levels, while CsA trough levels remained stable. In the gastrointestinal mucosa, CYP3A4 is mainly located at the apex of the villus enterocytes of the proximal small intestine, while P-glycoprotein is maximally expressed in the colon. Whether there is a different affinity for P-glycoprotein than for CYP3A4, or whether gut metabolism is more important for FK506 compared with cyclosporine is unknown at present (11). Despite increased FK506 trough levels during diarrhea, lowering the dose of FK506 in patients with diarrhea must be carefully monitored, especially when MMF is also reduced or stopped. After cessation of diarrhea, the dose of FK506 required to obtain similar trough levels increases again in some patients, exposing patients to increased risk of rejection for several weeks.


In conclusion, persistent afebrile diarrhea in MMF-treated renal transplant recipients is associated with increased trough levels of FK506, in contrast to stable levels of CsA. Stopping MMF, resulting in cessation of the diarrhea, may lead to sustained under-immunosuppression in FK506-MMF-treated patients with increased risk for rejection of the renal allograft. Therefore, we feel obliged to inform the transplant community that careful monitoring of FK506 is needed during and after episodes of diarrhea in these patients.


B. Maes holds the Janssens-Cilag Chair for Nephrology at the Catholic University of Leuven. The authors thank the Leuven Collaborative Group for Transplantation for their enthusiastic collaboration.