The relationship between tacrolimus concentration and acute GVHD is not well known, with few published data available for lower target levels. We hypothesized that lower levels of tacrolimus would correlate with higher incidence of acute GVHD and poorer prognosis. Receiver operator characteristic curves (ROC) were used to quantify tacrolimus blood levels as predictors of grade II–IV acute GVHD. A total of 97 pediatric patients with hematological malignancies met the study criteria. On the ROC, a cutoff of 7 ng/ml provided the best balance between sensitivity and specificity (62.8 vs 68.2%, respectively). Cumulative incidence of acute GVHD was 65.9% (range 58.5–73.3%) in patients with mean tacrolimus concentration of ⩽7 ng/ml and 34.8% (range 27.8–41.8%) in patients with mean tacrolimus concentration of >7 ng/ml (P=0.002). Incidence of non-relapse mortality (NRM) was higher in patients with tacrolimus of ⩽7 ng/ml (42.9%; range 35.6–50.2%) than in patients with tacrolimus of >7 ng/ml (28.3%; range 17.4–39.2%; P=0.008). This translated into better EFS in patients with tacrolimus of >7 ng/ml (48.9%; range 39.8–58.0%) than in patients with tacrolimus of ⩽7 ng/ml (31.8%; range 25.0–38.6%; P=0.031). Multivariate analysis showed that tacrolimus concentration was significantly associated with clinical outcomes. Mean whole-blood level of tacrolimus as continuous infusion should be maintained at ⩾7 ng/ml for pediatric patients.
The correlations of tacrolimus blood concentration with clinical efficacy and toxicity have been studied since the introduction of tacrolimus into clinical use. Previous studies have shown a sharp increase in nephrotoxicity for whole-blood levels of >20 ng/ml. In one study, the risk for creatinine >2 mg/dl increased by 84.4% when mean tacrolimus level was >20 ng/ml, compared with a mean level of ⩽20 ng/ml.1 According to those results, tacrolimus concentration should be kept at <20 ng/ml. However, no consensus has been reached regarding the lower limits of effective tacrolimus concentration, as no clear relationship has been identified between blood concentration of tacrolimus and occurrence of acute GVHD. In particular, although Przepiorka et al.2 reported that tacrolimus levels of <5 ng/ml were associated with an increased risk of acute GVHD in adult patients, no reports have addressed this issue in pediatric patients. We therefore analyzed the relationship between tacrolimus blood levels and clinical outcomes after SCT in 97 pediatric patients. Special attention was given to analyzing the relationship between blood concentrations and occurrence of acute GVHD to determine the lower target concentration of tacrolimus.
Patients and methods
This retrospective study included 97 pediatric patients with hematological malignancies who received tacrolimus as part of GVHD prophylaxis from December 1997 to September 2007 at the Japanese Red Cross Nagoya First Hospital. Treatment was approved by the institutional review board, and written informed consent was obtained from the guardians of each patient. Patients were classified with standard- or high-risk disease according to the previously described criteria.3 In brief, 49 patients were categorized as standard risk, based on a diagnosis of ALL, AML or malignant lymphoma in the first or second CR. The remaining 48 patients, including 23 patients who had received a second transplantation, were categorized as high risk. Only eight patients received HLA-matched related donor grafts. All other patients received grafts from alternative donors, including HLA-mismatched related donors, unrelated BM donors or unrelated cord blood donors.
Tacrolimus was administered at 0.02–0.03 mg/kg/day by i.v. continuous infusion from the day before transplantation. When patients were able to tolerate oral intake, tacrolimus was administered orally at three times the i.v. dose in two divided doses. The dose was adjusted to maintain blood levels within 5–15 ng/ml. MTX was administered at 15 mg/m2 on day 1 and 10 mg/m2 on days 3, 6 and 11. Blood samples for tacrolimus assay were taken using phlebotomy and tacrolimus concentrations were measured 2–3 times a week using enzyme immunoassasy (Abbott, Chicago, IL, USA).
The diagnosis of GVHD was graded according to the previously described criteria.4
Mean blood levels of tacrolimus
Mean tacrolimus concentrations from the day before transplantation to 28 days were calculated, while receiving continuous infusion. If patients developed acute GVHD (II–IV) and stopped receiving i.v. tacrolimus before 28 days after SCT due to treatment-related complications, such as transplant-associated microangiopathy (TAM), rejection, veno-occlusive disease, renal failure or early death, we evaluated mean tacrolimus concentrations at a time point before the onset of these events.
All analyses were performed using SPSS software (SPSS, Chicago, IL, USA). Data are presented as mean±s.d. Mann–Whitney tests were applied to examine the association of mean tacrolimus concentrations with different parameters. Receiver operator characteristic curves (ROC) were constructed using mean tacrolimus blood levels as a predictor of the incidence of acute GVHD. Area under the ROC curve was calculated as an overall performance indicator of the blood level. The cutoff values derived from ROC curves were evaluated in terms of sensitivity, specificity and positive and negative predictive values.
Acute GVHD (II–IV), OS, EFS, relapse rate and non-relapse mortality (NRM) were assessed using Kaplan–Meier product limit estimates. Differences between groups were tested using a log-rank test. All variables showing a probable association (P<0.2) with acute GVHD, OS, EFS or TRM were included in the Cox proportional hazards model. Covariates included in the Cox multivariate analysis were: patient age, donor sex, HLA disparity, disease stage, number of transplantations, conditioning regimen and whole-blood levels of tacrolimus. All statistical tests were two sided and differences were considered statistically significant at P<0.05.
Tacrolimus whole-blood levels and patient characteristics
The variables related with the patient, disease and transplant are summarized in Table 1. The subjects comprised 57 males and 40 females with a median age of 7 years (range 0.4–18 years). A total of 80 patients (82.5%) received unrelated donor SCT. In all, 29 patients (29.9%) received a reduced-intensity regimen, whereas the remaining 68 patients (70.1%) received a myeloablative regimen. A total of 93 patients received tacrolimus plus MTX and other patients received methylprednisolone (n=2) or mycophenolate mofetil (n=1) in addition to tacrolimus plus MTX. Only one patient received tacrolimus alone as GVHD prophylaxis. The mean concentration of tacrolimus while receiving continuous infusion during the first 4 weeks was 7.64±3.12 ng/ml. The mean concentrations during days –1 to –7, 7 to 14, 14 to 21 and 21 to 28 after SCT were 7.70±4.21, 6.89±2.53, 7.96±2.82 and 8.93±4.79 ng/ml, respectively. Table 2 shows the associations between mean tacrolimus concentration and various clinical parameters. Except for patient age, no significant differences were shown between variable factors.
A total of 22 patients stopped receiving continuous infusion of tacrolimus at <28 days after transplant, due to either a switch to oral intake (n=1) or transplant-associated complications, such as TAM (n=11), rejection (n=4), veno-occlusive disease (n=2), renal failure (n=1) or early death (n=3). Among 14 patients who stopped receiving tacrolimus at <28 days after transplant because of TAM, veno-occlusive disease or renal failure, 13 patients had earlier developed grade II–IV acute GVHD. For these patients, we evaluated mean tacrolimus concentrations at a time point before the onset of acute GVHD or other complications.
To evaluate whether whole-blood level of tacrolimus was able to predict the incidence of acute GVHD (II–IV), an ROC curve was generated. A cutoff value of 7 ng/ml was chosen as offering the best balance between sensitivity and specificity. Table 3 shows a comparison of sensitivity, specificity, and positive and negative predictive values at different cutoff values of tacrolimus blood levels.
Association between tacrolimus concentration and acute GVHD
Figure 1 shows the cumulative incidence of acute GVHD (II–IV) when patients were grouped using a tacrolimus blood level of 7 ng/ml as a cutoff during the first 2 weeks after SCT. Incidence of grade II–IV acute GVHD was significantly higher in patients with mean blood level of tacrolimus ⩽7 ng/ml (65.9%; range 58.5–73.3%) compared with those with >7 ng/ml (34.8%; range 27.8–41.8%; P=0.002; Figure 1). Incidence of grade III–IV GVHD was also significantly higher in patients with lower blood levels of tacrolimus (48.8%; range 41.0–56.6%) than in patients with tacrolimus >7 ng/ml (17.4; range 11.8–23.0%; P=0.002) (data not shown). During 4 weeks after SCT, the incidence of acute GVHD (II–IV) for patients with tacrolimus ⩽7 ng/ml was also significantly higher (76.7%; range 69.0–84.4%) compared with patients showing >7 ng/ml (35.1%; range 28.8–41.4%; P<0.001) (data not shown).
Using multivariate analysis, the risk for acute GVHD was most significantly increased when mean tacrolimus level was ⩽7 ng/ml compared with a mean level of >7 ng/ml. The relative risks of mean blood tacrolimus levels of ⩽7 ng/ml at 14 days and 28 days after SCT were 3.473 (95% confidence interval (CI) 1.128–4.153; P=0.020) and 3.017 (95% CI 1.619–5.624; P<0.001), respectively.
Associations between blood concentration and other clinical outcomes
For survival analysis, mean tacrolimus blood level was used as a dichotomous variable and patients were grouped after setting the cutoff value for tacrolimus blood levels of 7 ng/ml. Kaplan–Meier estimates of OS were 44.4% (range 38.8–50.0%) for the whole cohort, 49.7% (range 40.3–59.1%) for patients with a mean blood level of >7 ng/ml and 37.6% (range 30.4–44.8%) for ⩽7 ng/ml (P=0.047), using the mean concentration at 14 days after SCT. The probability of EFS was 48.9% (range 39.8–58.0%) and 31.8% (range 25.0–38.6%), respectively (P=0.031; Figure 2). The probability of NRM was 33.1% (range 27.6–38.6%) in the entire study population, 28.3% (range 17.4–39.2%) for patients with a blood level of >7 ng/ml and 42.9% (range 35.6–50.2%) for ⩽7 ng/ml (P=0.008; Figure 3). With regard to possible predictors of clinical outcomes by multivariate analysis using the Cox proportional hazards model, the relative risks for OS, EFS and NRM were 2.240 (95% CI 1.198–4.190; P=0.012), 1.859 (95% CI, 1.044–3.312; P=0.035) and 3.021 (95% CI, 1.339–6.816; P=0.008), respectively (Table 4). In analysis using mean concentration at 28 days after SCT, tacrolimus blood level of ⩽7 ng/ml was also a significant predictor of clinical outcomes such as OS, EFS and NRM (Table 4).
This study analyzed the relationships between mean tacrolimus concentration during 28 days after SCT and clinical outcomes including acute GVHD and TRM. The results suggest that the incidence of acute GVHD after SCT is significantly related to mean tacrolimus concentration.
As a result of interpatient variability and dynamic changes in the metabolism of the drug during the early days after transplantation, a close monitoring of blood concentrations is mandatory in evaluating the efficacy of tacrolimus. According to the literature, the recommended target therapeutic range of tacrolimus is 10–20 ng/ml (Wingard et al.1) or 5–15 ng/ml (Przepiorka et al.2). The upper range of 20 ng/ml was derived from analyses of correlated toxicities.1, 5 In contrast, the relationship between blood concentration of tacrolimus and control of acute GVHD using lower target levels has not been well defined. Przepiorka et al.2 reported that tacrolimus levels of <5 ng/ml were associated with an increased risk of acute GVHD after unrelated donor marrow transplantation. However, the median patient age in that report was 34 years (range 14–53 years), and no studies seem to have analyzed correlations between tacrolimus blood concentration and clinical efficacy in pediatric patients. We analyzed correlations between mean concentration of tacrolimus and clinical outcomes during the first 4 weeks after transplantation in pediatric patients.
Receiver operator characteristic analyses in this study indicated that a cutoff value of 7 ng/ml resulted in the best sensitivity and specificity for predicting the development of acute GVHD. A mean blood level of tacrolimus ⩽7 ng/ml was identified as an independent risk factor for acute GVHD, TRM, EFS and OS in multivariate analysis. Indeed, a 2.164-fold increase in grade II–IV GVHD was observed for tacrolimus levels of ⩽7 ng/ml compared with >7 ng/ml, using mean concentrations at 14 days after SCT and a 3.017-fold increase was observed at 28 days after SCT. Maintaining mean blood levels of tacrolimus >7 ng/ml for 28 days after SCT in continuous infusion may thus be important.
A total of 21 patients stopped receiving tacrolimus at <28 days after SCT because of various complications. The main reason for discontinuation of tacrolimus was TAM, occurring in 11 patients. TAM can be a fatal complication after SCT and immunosuppressants have been indicated as one of the causative agents for microangiopathy.6, 7 To prevent deterioration of TAM, we stopped administration of tacrolimus in patients who were diagnosed with TAM. Grade II–IV acute GVHD had developed at the time of stopping the administration of tacrolimus in all of these patients. In patients who stopped receiving tacrolimus at <28 days after SCT, mean whole-blood level was 5.88±3.07 ng/ml, significantly lower than that of patients without early discontinuation. Whole tacrolimus level of ⩽7 ng/ml at 14 days after SCT represented a significant risk factor for acute GVHD, even though the mean concentration in patients with early cessation of tacrolimus was evaluated at a time point before the onset of acute GVHD or other complications. Acute GVHD caused by lower blood levels of tacrolimus might well contribute to the development of TAM, as some reports have identified grade II–IV acute GVHD as one of the risk factors for TAM.8, 9
Przepiorka et al.2 reported that a tacrolimus whole-blood level of <5 ng/ml in the early post transplantation period is subtherapeutic, and dosages of tacrolimus should be adjusted to maintain whole-blood steady-state or trough levels at 5–15 ng/ml. Since that report, a target range maintaining the level at >5 ng/ml has been recommended.10 We showed that mean whole-blood level of tacrolimus should be maintained at >7 ng/ml in continuous infusion for pediatric patients, as well as in adults. Wong et al.11 also insisted that the optimal range of blood tacrolimus levels during the first 15 days would be 7–9 ng/ml in adult patients with refractory or relapsed myeloid leukemia.
In a recent study using the new pharmacokinetic parameter of area under the trough level, the calculated target blood trough level for orally administered tacrolimus was 0.71 times that for continuous infusion, which was almost equal to the clinical concentration for continuous infusion.12 Blood trough levels of tacrolimus when administered orally should thus be maintained at ⩾5 ng/ml, as our results suggest that mean whole-blood level of tacrolimus should be maintained at ⩾7 ng/ml while receiving continuous infusion.
In summary, although little published data are available regarding relationships between tacrolimus blood levels and risk of GVHD, our study clearly showed that tacrolimus levels at ⩽7 ng/ml with continuous infusion are significantly associated with an increased risk of acute GVHD and lower survival, identifying 7 ng/ml as the lower limit of the therapeutic range.
Wingard JR, Nash RA, Przepiorka D, Klein JL, Weisdorf DJ, Fay JW et al. Relationship of tacrolimus (FK506) whole blood concentrations and efficacy and safety after HLA-identical sibling bone marrow transplantation. Biol Blood Marrow Transplant 1998; 4: 157–163.
Przepiorka D, Saliba R, Cleary K, Fischer H, Tonai R, Fritsche H et al. Tacrolimus does not abrogate the increased risk of acute graft-versus-host disease after unrelated-donor marrow transplantation with allelic mismatching at HLA-DRB1 and HLA-DQB1. Biol Blood Marrow Transplant 2000; 6: 190–197.
Morishima Y, Morishita Y, Tanimoto M, Ohno R, Saito H, Horibe K et al. Low incidence of acute graft-versus-host disease by the administration of methotrexate and cyclosporine in Japanese leukemia patients after bone marrow transplantation from human leukocyte antigen compatible siblings; possible role of genetic homogeneity. The Nagoya Bone Marrow Transplantation Group. Blood 1989; 74: 2252–2256.
Przepiorka D, Weisdorf D, Martin P, Klingemann HG, Beatty P, Hows J et al. 1994 Consensus conference on acute GVHD grading. Bone Marrow Transplant 1995; 15: 825–828.
Przepiorka D, Nash RA, Wingard JR, Zhu J, Maher RM, Fitzsimmons WE et al. Relationship of tacrolimus whole blood levels to efficacy and safety outcomes after unrelated donor marrow transplantation. Biol Blood Marrow Transplant 1999; 5: 94–97.
Nishida T, Hamaguchi M, Hirabayashi N, Haneda M, Terakura S, Atsuta Y et al. Intestinal thrombotic microangiopathy after allogeneic bone marrow transplantation: a clinical imitator of acute enteric graft-versus-host disease. Bone Marrow Transplant 2004; 33: 1143–1150.
Fujino M, Kim Y, Ito M . Intestinal thrombotic microangiopathy induced by FK506 in rats. Bone Marrow Transplant 2007; 39: 367–372.
Nakamae H, Yamane T, Hasegawa T, Nakamae M, Terada Y, Hagihara K et al. Risk factor analysis for thrombotic microangiopathy after reduced-intensity or myeloablative allogeneic hematopoietic stem cell transplantation. Am J Hematol 2006; 81: 525–531.
Oran B, Donato M, Aleman A, Hosing C, Korbling M, Detry MA et al. Transplant-associated microangiopathy in patients receiving tacrolimus following allogeneic stem cell transplantation: risk factors and response to treatment. Biol Blood Marrow Transplant 2007; 13: 469–477.
Przepiorka D, Blamble D, Hilsenbeck S, Danielson M, Krance R, Chan KW . Tacrolimus clearance is age-dependent within the pediatric population. Bone Marrow Transplant 2000; 26: 601–605.
Wong R, Shahjahan M, Wang X, Thall PF, De Lima M, Khouri I et al. Prognostic factors for outcomes of patients with refractory or relapsed acute myelogenous leukemia or myelodysplastic syndromes undergoing allogeneic progenitor cell transplantation. Biol Blood Marrow Transplant 2005; 11: 108–114.
Nakamura Y, Takeuchi H, Okuyama K, Akashi T, Jojima Y, Konno O et al. Evaluation of appropriate blood level in continuous intravenous infusion from trough concentrations after oral administration based on area under trough level in tacrolimus and cyclosporine therapy. Transplant Proc 2005; 37: 1725–1727.
This work was supported by a grant from the Ministry of Health, Labour, and Welfare of Japan, Tokyo.
The authors declare no conflict of interest.
About this article
Therapeutic Drug Monitoring (2019)
Outcomes of matched sibling donor bone marrow transplantation in children using single-agent calcineurin inhibitors as prophylaxis for graft versus host disease
Pediatric Blood & Cancer (2018)
Low incidence of acute graft-versus-host disease with short-term tacrolimus in haploidentical hematopoietic stem cell transplantation
Leukemia Research (2017)
Efficacy of Tacrolimus/Mycophenolate Mofetil as Acute Graft-Versus-Host Disease Prophylaxis and the Impact of Subtherapeutic Tacrolimus Levels in Children after Matched Sibling Donor Allogeneic Hematopoietic Cell Transplantation
Biology of Blood and Marrow Transplantation (2015)
International Journal of Hematology (2015)