|Year : 2021 | Volume
| Issue : 3 | Page : 227-231
Tacrolimus therapeutic drug monitoring and correlation with clinical events – A single-center prospective study
Mayoor V Prabhu1, Akash Nayak Karopadi2, Sreepada V Subhramanyam3, B. H. Santhosh Pai4, KS Nayak5
1 Department of Nephrology, Kasturba Medical College, Mangalore (Manipal Academy of Higher Education, Manipal), India
2 Management Consulting Division, Alira Health GmbH, Munich, Germany
3 Department of Nephrology, Deccan Hospital, Hyderabad, Telangana, India
4 Department of Nephrology, Yenepoya Medical College, (Yenepoya Deemed to be University, Mangalore), India
5 Department of Nephrology, Virinchi Hospital, Hyderabad, Telangana, India
|Date of Submission||12-Aug-2020|
|Date of Decision||27-Nov-2020|
|Date of Acceptance||10-Feb-2021|
|Date of Web Publication||30-Sep-2021|
Dr. B. H. Santhosh Pai
Department of Nephrology, Yenepoya Medical College, Yenepoya Deemed to be University, Mangalore, Karnataka
Source of Support: None, Conflict of Interest: None
Background: Tacrolimus exposure is estimated by trough levels (C0). Recent studies suggest that C0 may not accurately reflect the area under the curve (AUC) and may not correlate with clinical events like acute rejection (AR) and nephrotoxicity. Materials and Methods: In an open, prospective, single-center study, 29 consecutive recipients of renal transplantation underwent C0 along with 2-h (C2), 4-h (C4), and 6-h (C6) estimation of blood tacrolimus levels by enzyme-linked immunosorbent assay, 72 h after initiation with tacrolimus or after a change in its dosage. AUC was estimated by trapezoidal method. C0, C2, C4, and C6 levels were correlated with the AUC. Results: Thirty-six AUC estimations were made over a 2-year period. The best correlate was C6. Correlation coefficients were C0 – 0.868, C2 – 0.788, C4 – 0.839, and C6 – 0.904. C6 values accounted for 79% of the variability of the AUC. Six patients experienced AR, with 5 having C0 within the target range. C6 values correlated best with AUC in these patients (C0 – 0.970, C2 – 0.833, C4 – 0.942, and C6 – 0.970). This was statistically significant. Three patients developed tacrolimus toxicity. In these patients, the correlation coefficients were C0 – 0.551, C2 – 0.556, C4 – 0.77, and C6 – 0.941. By regression analysis, we developed predictive equations. The equation AUC = 12.126 + 2.81 × C0 + 2.92 × C6 best predicted the AUC. Conclusions: Overall C6 levels were more predictive of the AUC, accurately predicting AR and nephrotoxicity. Incorporating C6 may improve tacrolimus therapeutic drug monitoring.
Keywords: Acute rejection, area under the curve, calcineurin inhibitors, nephrotoxicity, tacrolimus, therapeutic drug monitoring, transplantation
|How to cite this article:|
Prabhu MV, Karopadi AN, Subhramanyam SV, Pai BH, Nayak K S. Tacrolimus therapeutic drug monitoring and correlation with clinical events – A single-center prospective study. Indian J Transplant 2021;15:227-31
|How to cite this URL:|
Prabhu MV, Karopadi AN, Subhramanyam SV, Pai BH, Nayak K S. Tacrolimus therapeutic drug monitoring and correlation with clinical events – A single-center prospective study. Indian J Transplant [serial online] 2021 [cited 2021 Nov 29];15:227-31. Available from: https://www.ijtonline.in/text.asp?2021/15/3/227/327381
| Introduction|| |
Calcineurin inhibitors (CNIs) form the backbone of modern solid organ transplantation. Tacrolimus has a narrow therapeutic index with wide intrapatient variability which mandates close attention to therapeutic drug monitoring (TDM). Transplant physicians face a tightrope walk in maintaining serum concentrations of CNIs high enough to prevent rejection of the organ while protecting against systemic overexposure. Since its initial introduction as a therapy for severe refractory liver graft rejection, today, tacrolimus (FK506) is widely used in renal transplantation (RT). With cyclosporine (CsA), prospective studies showed that area under the curve (AUC)-guided dosing achieved better correlation with clinical endpoints like acute rejection (AR) and nephrotoxicity.,, Furthermore, it was found that 2-h postdose (C2) levels were better correlated to the AUC than trough (C0) levels., This led to a paradigm shift in the way CsA levels were monitored in clinical practice. The same issue has not been examined in sufficient detail for tacrolimus. The few studies available have shown conflicting results. While some studies reported a reasonable correlation of trough (C0) levels with the AUC,,, other studies had conflicting results., The relationship between low whole blood trough levels and the incidence of AR has been more clearly demonstrated. It is accepted that the number, timing, and severity of AR episodes are risk factors for earlier graft loss and poor graft and patient survival. Like CsA, the daily dosage of tacrolimus has a more than tenfold difference between extreme cases, which makes accurate and reproducible TDM critical to better patient and graft outcomes., The wide difference in opinion regarding the optimum TDM strategy for tacrolimus and the ambiguity of current literature in this regard prompted us to undertake this study. We aimed to prospectively determine the single-timed tacrolimus blood level that best correlated with the AUC. In addition, we aimed to estimate systemic exposure to tacrolimus and to examine the relation between tacrolimus levels, AUC, and clinical events like AR and nephrotoxicity.
Materials and Methods
This was a prospective, single-center, nonrandomized study performed in a tertiary care center.
All patients admitted for RT (live related/deceased donor) were included in the study, after obtaining appropriate written consent. Exclusion criteria included patients receiving drugs known to alter blood levels of tacrolimus, patients suffering from gastrointestinal or liver disease, which was thought to interfere with tacrolimus absorption, and patients on current therapy with bile acid sequestrants. Protocol Tacrolimus was usually introduced at day 2 post transplant. The initial dose of tacrolimus was 0.1–0.15 mg/kg. All patients received the same generic preparation (Vingraf™, Emcure Pharmaceuticals Pvt Ltd, Pune, India). Tacrolimus doses were adjusted to keep trough levels in the range of 10–12 ng/mL within the 1st month and then 8–10 ng/mL later. After 72 h of the introduction of tacrolimus in the patients' immunosuppressive regimen or 72 h after an adjustment in the tacrolimus dosage, trough levels were estimated prior to the morning dose of tacrolimus (at 8 AM in the morning), after an overnight fast. Patients were allowed to take their other medications including immunosuppressants at the appropriate time, however, they were not allowed to ingest food until at least one hour after the morning tacrolimus dose. For subsequent estimations, the patients were allowed to be in a fed state and fatty meals were not specifically avoided. Blood levels of tacrolimus were measured exactly 2, 4, and 6 h after the morning dose had been taken. The same protocol was followed after every subsequent change in the tacrolimus dosage, after allowing a period of 72 h to allow for a steady state to be achieved. We used the PRO-Trac™ II Tacrolimus ELISA monoclonal antibody kit (Dia Sorin Inc, Stillwater, MN).
Standard statistical software (SPSS Version 17, Chicago, IL, USA) was used for statistical analysis. The AUC was calculated using the parallel trapezoidal rule. Associations between blood concentrations at each sampling time points and the AUC were measured by the Pearson's correlation coefficients. The t-test was used to compare means of continuous variables between patients with or without AR and nephrotoxicity. Pairs of correlation coefficients were compared by an asymptotic Wald-type test. Concomitant immunosuppression used was prednisolone and MMF (500 mg–750 mg twice daily). Nephrotoxicity was defined as a ≥30% increase in the serum creatinine levels not attributable to any other identifiable cause that decreased when the tacrolimus dosage was decreased. Biopsy specimens when evaluated for renal dysfunction were free of AR and exhibited changes attributable to CNIs. Tacrolimus toxicity was diagnosed clinically, with signs and symptoms of any adverse effects of tacrolimus (tremor/ alopecia/diarrhea/saccadic eye movements). Biopsy-proven AR was graded by the Banff criteria.
The patient consent has been taken for participation in the study and for publication of clinical details and images. Patients understand that the names, initials would not be published, and all standard protocols will be followed to conceal their identity.
It was approved by the Institutional Review Board viz GHE 11/2012. Informed written consent was obtained from all participants. The procedure was carried out in accordance with the Declaration of Helsinki and International Council for Harmonization-Good Clinical Practice (ICH-GCP).
| Results|| |
Twenty-nine consecutive RT recipients were included in the study. Demographic details and diagnoses are detailed in [Table 1]. An initial tacrolimus dosage of 0.15 mg/kg was attempted based on the clinical status and biochemistry.
Area under the curve0-12 and correlation with the timed samples
Overall, all the timed samples (C0, C2, C4, and C6) showed a good correlation with the AUCs. The later samples (C4 and C6) showed the best correlation with the AUC, and C2 value showed the least correlation, however, these differences were not statistically significant. The overall best correlation was seen with C6 values [Figure 1]. The details are given in [Table 2].
|Table 2: Correlation of timed samples with the area under the curve of tacrolimus|
Click here to view
Six patients (28%) developed biopsy-proven AR during the course of the study, of which 5 had C0 levels within the desired range (>10 ng/mL, ranging from 7 to 19 ng/mL). The AUCs in these patients showed a wide variance (69.48 ± 19.51).
Three patients developed tacrolimus toxicity during the course of the study. The AUC0-12 in these patients varied from 84 to 95. The correlation coefficients for the timed samples in these patients were C0 – 0.551, C2 – 0.556, C4 – 0.778, and C6 – 0.941 (P < 0.001).
Equation for prediction of area under the curve
We serially used regression to develop the equation with maximum predictive power for the AUC0-12. The best equation included C0 and C6 values and had an r2 value of 0.977.
AUC0-12 = 12.126 + 2.81 × C0 + 2.92 × C6.
| Discussion|| |
We found that all the timed samples we evaluated, namely the trough or C0 levels, C2, C4, and C6 values showed correlation with the AUC0-12. Later samples like the C4 and C6 values showed a better correlation, the best being the C6 values. This also held good for the cases where we saw AR and tacrolimus toxicity, and in all cases, there was a statistical significance. The C6 value was more predictive of the AUC, AR, and tacrolimus toxicity than the trough. In addition, we tested the recent belief that C2 values may be better surrogate markers of the full AUC than the trough levels, as is the case for CsA. We found this was not true, as the trough levels performed better than the C2 levels in predicting the full AUC.
In patients with AR, we found that AR occurred in spite of trough levels being in the target range, in 5 of the 6 patients. In these patients, the C6 values predicted the AUC more efficiently than the trough levels. Thus, there may be a strong case for replacing trough level monitoring with more representative timed samples or with more than one sample. The same held good for patients with nephrotoxicity. The C6 values were the best markers. We also ran a regression analysis to develop equations that accurately predicted the AUC from a limited number of timed samples. In our case, the equation with maximum predictability for the AUC included the trough and C6 values. This, we believe, can help in more accurate and cost-effective monitoring for patients on tacrolimus than a full AUC estimation or trough alone. This may help to make tacrolimus monitoring easier and more cost-effective.
Wong reported that trough levels of tacrolimus were not representative of the full AUC (AUC0-12) in oriental patients. In this study, the correlation between the trough levels and the AUC was not statistically significant. They found the best correlation with the C4 values. The sample size in this study was 18 and all patients were oriental Chinese. The C2 and the C4 values were the best predictors of the AUC and the best equation developed by regression analysis, similar to the one we performed, included the C2 and C4 values. In this study, the authors did not correlate with clinical events like nephrotoxicity and AR. Jorgensen et al. compared C0 and C2 levels as estimates of the AUC in 21 Caucasian RT recipients. They reported that the trough levels performed overall better than C2 levels in predicting the AUC. There was a wide variation in the AUC levels despite near equal dosing, similar to our study and most other studies in this field. Unlike our study, this study was performed at fixed points (day 3 and day 14 post transplantation). These researchers found that C3, C4, and C6 values show the best correlation and the least correlation was with the C1 and C2 values. They also demonstrated that in spite of trough levels being in the target range, some patients received toxic doses of tacrolimus. They concluded that sampling at a point 3–6 h after intake may be at least as good as trough levels. The baseline immunosuppression in these patients was similar to our study (mycophenolate mofetil [MMF] and steroids). In this study, the recipients received organs from deceased donors in an overwhelming majority of the cases. These authors also presented arguments in favor of the trapezoidal method of estimating the AUC over curve fitting to infinity to determine the AUC, assuming first-order kinetics, which we are also in agreement with. If curve fitting is applied in patients taking medications every 12 h, the last estimated value (in this study and in our study, the C-6 value) would have inordinate influence over the determined value. This study did look into rejections, however, they reported that 2 of the 4 rejections occurred with the AUC under 200 on both days 3 and 14, and the other 2 rejections occurred when the dose of tacrolimus had been significantly reduced. Similar to our study, they also reported nephrotoxicity and high AUC occurring in patients where the trough levels were within the target range.
Balbontin used a sparse sampling approach with blood levels at 0, 1, 2, 3, and 4 h post dose and calculation of AUC0-4 in 28 patients. They retrospectively analyzed C0 as a marker of AR and nephrotoxicity. They found, similar to our findings, that C0 did correlate with AUC0-4. However, the best correlation was with the C2 and C3 values and that there was a trend toward lower C3 values in patients with AR.
Chen conducted a trial on 16 oriental RT recipients who received the first dose of tacrolimus on day 3 post RT. Co-prescriptions included MMF and prednisolone and induction agents used were methylprednisolone and antithymocyte globulin. This was almost identical to our study. They used the enzyme-linked immunosorbent assay (ELISA) method to determine blood tacrolimus levels. However, the authors determined a full AUC with eight samples between trough and 12 h. They found, as expected, a wide interindividual variation after the first oral dose. The authors developed regression equations and evaluated the variance in the strength of association between predicted AUC and AUC0-12. Similar to our findings, the authors concluded that samples post 4 h (in this study, C5) were more predictive, and equations incorporating C1.5 and C5 were the best abbreviated AUC for cost-effective tacrolimus monitoring.
Mathew conducted a study on 29 Indian patients, 3–6 months post transplant. They did not include patients earlier in the course of the transplant unlike our study. They calculated a full AUC0-12 with sampling at more frequent intervals than our study. They developed multiple equations by regression analysis and compared them with the calculated AUCs. The authors did not provide detailed information on clinical events like AR and nephrotoxicity. They however did comment that the patients who developed rejection had trough levels within the therapeutic range. The applicability of these regression equations can only be evaluated by prospective studies. They found that a regression equation comprising trough and 1.5 h postdose sample provides a reliable and simple method to estimate exposure by AUC in RT recipients.
Op den Buijsch analyzed 37 Caucasian patients who had full AUCs determined. These patients were stable patients 1 year post RT, unlike our study population. They used the high-pressure liquid chromatography (HPLC) method of estimating tacrolimus blood levels unlike most other studies including ours which used the microparticle immunoassay or ELISA techniques. They developed equations based on multiple values similar to our strategy, and they reported that equations incorporating C3 and C4 values performed best as predictors of the AUC.
The only prospective trial to date that evaluated AUC-based dosing of tacrolimus was performed by Scholten. They developed and validated a two-compartmental population-based pharmacokinetic model with Bayesian estimation of tacrolimus systemic exposure. Later, they applied this model to prospectively dose 15 consecutive RT recipients. By Bayesian forecasting, they developed an equation incorporating the trough level and a second value obtained between 2 and 4 h post dose which significantly improved the squared correlation with the AUC. When compared with trough levels alone, this method improved the correlation significantly and reduced the 95% prediction interval by 50%. This study included patients who were early posttransplant and also a few stable patients longer in the posttransplant course. This study also proved the inadequacy of trough level-based monitoring of tacrolimus, and that the addition of just one sample between 2 and 4 h post dose would significantly improve the correlation with the AUC. This study was a landmark study in which they were the only researchers who applied prospectively AUC-guided dosing of tacrolimus.
Tacrolimus serum levels can be estimated by several methods – clinical laboratory improvement amendments, enzyme multiplied immunoassay technique, affinity column-mediated immunoassay, and HPLC being among the commercially available kits. As generally low trough levels are used clinically, the assay should be able to quantify lower limits, typically 1 ng/ml. We used ELISA-based method due to its lower cost and ease of usage, compared to liquid chromatography–mass spectrometry assays which are more expensive and require more training of laboratory staff.
It was a single centre study.
| Conclusion|| |
To conclude, after a review of available literature, and a comparison with the findings from our study, we believe that the current practice of isolated trough level-based monitoring may not adequately reflect systemic exposure to tacrolimus and may result in either overexposure or underexposure. This assumes more significance with the availability of multiple generic versions of tacrolimus in the market. We found C6 values to be best correlating to the AUC and also to clinical events. Similar findings have also been reported by other researchers. Prospective randomized controlled trials may be a way forward in the quest for an optimum strategy in tacrolimus TDM.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Laskow DA, Vincenti F, Neylan JF, Mendez R, Matas AJ. An open-label, concentration-ranging trial of FK506 in primary kidney transplantation: A report of the United States Multicenter FK506 Kidney Transplant Group. Transplantation 1996;62:900-5.
Demetris AJ, Fung JJ, Todo S, McCauley J, Jain A, Takaya S, et al
. Conversion of liver allograft recipients from cyclosporine to FK506 immunosuppressive therapy – A clinicopathologic study of 96 patients. Transplantation 1992;53:1056-62.
Belitsky P, Mahalati K, West K, Sketris I. Influence of drug formulation on utilization and outcomes: Neoral and monitoring by sparse sample area under the curve. Transplant Proc 1999;31:1667-8.
Mahalati K, Belitsky P, Sketris I, West K, Panek R. Neoral monitoring by simplified sparse sampling area under the concentration-time curve: Its relationship to acute rejection and cyclosporine nephrotoxicity early after kidney transplantation. Transplantation 1999;68:55-62.
David-Neto E, Lemos FB, Furusawa EA, Schwartzman BS, Cavalcante JS, Yagyu EM, et al
. Impact of cyclosporin A pharmacokinetics on the presence of side effects in pediatric renal transplantation. J Am Soc Nephrol 2000;11:343-9.
Mahalati K, Belitsky P, West K, Kiberd B, Fraser A, Sketris I, et al
. Approaching the therapeutic window for cyclosporine in kidney transplantation: A prospective study. J Am Soc Nephrol 2001;12:828-33.
Keown PA. New concepts in cyclosporine monitoring. Curr Opin Nephrol Hypertens 2002;11:619-26.
Jørgensen K, Povlsen J, Madsen S, Madsen M, Hansen H, Pedersen A, et al
. C2 (2-h) levels are not superior to trough levels as estimates of the area under the curve in tacrolimus-treated renal-transplant patients. Nephrol Dial Transplant 2002;17:1487-90.
Kimikawa M, Kamoya K, Toma H, Teraoka S. Effective oral administration of tacrolimus in renal transplant recipients. Clin Transplant 2001;15:324-9.
Braun F, Schütz E, Peters B, Talaulicar R, Grupp C, Undre N, et al
. Pharmacokinetics of tacrolimus primary immunosuppression in kidney transplant recipients. Transplant Proc 2001;33:2127-8.
Wong KM, Shek CC, Chau KF, Li CS. Abbreviated tacrolimus area-under-the-curve monitoring for renal transplant recipients. Am J Kidney Dis 2000;35:660-6.
Tada H, Satoh S, Iinuma M, Shimoda N, Murakami M, Hayase Y, et al
. Chronopharmacokinetics of tacrolimus in kidney transplant recipients: Occurrence of acute rejection. J Clin Pharmacol 2003;43:859-65.
Webster AC, Woodroffe RC, Taylor RS, Chapman JR, Craig JC. Tacrolimus versus ciclosporin as primary immunosuppression for kidney transplant recipients: Meta-analysis and meta-regression of randomised trial data. BMJ 2005;331:810.
Pelletier RP, Cosio F, Henry ML, Bumgardner GL, Davies EA, Elkhammas EA, et al
. Acute rejection following renal transplantation. Evidence that severity is the best predictor of subsequent graft survival time. Clin Transplant 1998;12:543-52.
Balbontin FG, Kiberd B, Squires J, Singh D, Fraser A, Belitsky P, et al
. Tacrolimus monitoring by simplified sparse sampling under the concentration time curve. Transplant Proc 2003;35:2445-8.
Chen YH, Zheng KL, Chen LZ, Dai YP, Fei JG, Qiu J, et al
. Clinical pharmacokinetics of tacrolimus after the first oral administration in combination with mycophenolate mofetil and prednisone in Chinese renal transplant recipients. Transplant Proc 2005;37:4246-50.
Mathew BS, Fleming DH, Jeyaseelan V, Chandy SJ, Annapandian VM, Subbanna PK, et al
. A limited sampling strategy for tacrolimus in renal transplant patients. Br J Clin Pharmacol 2008;66:467-72.
Op den Buijsch RA, van de Plas A, Stolk LM, Christiaans MH, van Hooff JP, Undre NA, et al
. Evaluation of limited sampling strategies for tacrolimus. Eur J Clin Pharmacol 2007;63:1039-44.
Scholten EM, Cremers SC, Schoemaker RC, Rowshani AT, van Kan EJ, den Hartigh J, et al
. AUC-guided dosing of tacrolimus prevents progressive systemic overexposure in renal transplant recipients. Kidney Int 2005;67:2440-7.
[Table 1], [Table 2]