Estimation of Abbreviated Cyclosporine A Area under the Concentration-Time Curve in Allogenic Stem Cell Transplantation after Oral Administration
Measurements of Cyclosporine (CsA) systemic exposure permit its dose adjustment in allogenic stem cell transplantation recipients to prevent graft-versus-host disease. CsA LSSs were developed and validated from 60 ASCT patients via multiple linear regressions. All whole-blood samples were analyzed by fluorescence polarization immunoassay (FPIA-Axym). The 10 models that have used CsA concentrations at a single time point did not have a good fit with AUC0–12 (). and were the time points that correlated best with AUC0–12 h, were respectively 0.848, and 0.897. The LSS equation with the best predictive performance (bias, precision and number of samples) utilized three sampling concentrations was AUC0–12 h. Optimal LSSs equations which limited to those utilizing three timed concentrations taken within 4 hours post-dose developed from ASCT recipient's patients yielded a low bias <5% ranged from 1.27% to 2.68% and good precision <15% ranged from 9.60% and 11.02%. We propose an LSS model with equation AUC0–12 h for a practical reason. Bias and precision for this model are respectively 2.68% and 11.02%.
Cyclosporine is one of the most common immunosuppressants used in allogenic stem cells transplant (ASCT) both in children and adults [1, 2]. This population had more digestive problems compared to solid transplant patient; consisting of mucositis and diffuse inflammation of the intestinal tract related to the preparative regimen. In addition, they frequently develop digestive graft-versus-host disease (GVHD) and intestinal viral disease affecting the absorption of CsA [3–5]. In most case cyclosporine is administered intravenously from day −1 to days 30–50 then orally until day 180 .
Individualization of cyclosporine (CsA) dosage using therapeutic drug monitoring remains an indispensable tool for management of CsA therapy [7, 8]. A number of clinical trials have employed monitoring in solid organ transplantations and demonstrated improvements in rejection rates and acute toxicity compared to trough-level () monitoring [9, 10].
In ASCT Barkholt et al.  have not found a correlation between and either severs GVHD or infection in ASCT patients with guided CsA dosing. Furukawa et al., 2010 , demonstrate that in ASCT patients receiving twice-daily 3 h intravenous administrations, the concentration of CsA at 2 h () and 2 h and 50 min () were best single predictors of exposure () while the trough concentration correlates poorly. Inoue et al.  were observed two close relationships between and the for infusion and between and the for oral administration.
The area under a blood drug concentration-time curve (AUC) is an important index in therapeutic drug monitoring and in pharmacokinetics. Unfortunately, obtaining complete pharmacokinetic profiles from transplant patients is not always an easy task due to ethical, practical, and economical reasons. Therefore, reducing the number of blood samples necessary for determinations would be very practical for patient management and would help in cyclosporine monitoring. Recently, limited sampling strategies (LSSs) have been developed to predict the ; several of them are used to predict cyclosporine AUC in transplanted patients. Up to now, more than 100 equations to estimate AUC using limited sampling strategy have been published, but most of them have not been validated .
ASCT outcomes may be optimized by cyclosporine dose adjustment according to systemic exposure (measured by AUC) rather than trough concentration.
The aim of this study was to develop LSSs for predicting cyclosporine to provide a practical method for a more accurate therapeutic (TDM) after oral dosing in patients undergoing ASCT and validated selected models.
2. Patients and Methods
Following ethics approval by the institutional review board, 60 ASCT recipient’s patients enrolled the study (new graft in 2009). The patient’s characteristics are listed in Table 1. Diseases consisted in acute myeloblastic leukemia ( patients) acquired aplastic anemia ( patients), acute lymphoblastic leukemia (), Fanconi’s anemia ( patients), multiple myeloma ( patient), and Gaucher disease ( patient).
Conditioning regimens consisted on IV Busulfan + Cyclophosphamide (36 patients), Fractionated Total body Irradiation + Etoposide (8 patients), horse antithymocyte globulin + Cyclophosphamide (12 patients), Fludarabine + Cyclophosphamide (4 patients). All patients received oral antimicrobial prophylaxis with amoxicillin or spiramycin, fluconazole, and acyclovir.
CsA (Sandimmun) was started by continuous intravenous (i.v.) infusion from day 1 to 15–21 days after transplantation, and then patients were randomized to oral CsA (equoral or neoral) at twice IV dose. After that all patients received oral CsA (mean of 300 mg/day). The daily dose was taken in two equally divided doses at 12 h intervals.
2.2. Inclusion and Exclusion Criteria
After the conditioning regimes, patients were included in the study, if they were aged between 3 and 50 years, clinically stable after a first ASCT, and receiving oral CsA as graft-versus-host disease prophylaxis. Patients should have stable serum creatinine (<2 baseline value) and no history of hepatic dysfunction (bilirubin and aminotransferases <2 normal value).
Patients excluded if they developed graft-versus-host disease or microangiopathy, received another drug which interferes with CsA pharmacokinetics, and showed evidence of noncontrol digestive problem, renal (2 × creatinine baseline value), or hepatic dysfunction (bilirubin and aminotransferases >2 × normal value).
2.3. Blood Sampling and Measurements of Cyclosporine Concentrations
After written informed consent, blood samples (3 mL) were collected in tubes containing EDTA anticoagulant from the patients at time 0 (; pre dose) and at 0.5 h (), 1 h (), 1.5 h (), 2 h (), 3 h (), 4 h (), 6 h (), 8 h (), and 12 h () after steady-state morning Cs dose. All whole-blood samples were analyzed by fluorescence polarization immunoassay (FPIA-Axym) in Laboratory of Clinical Pharmacology of the National Center of Pharmacovigilance. Precision data of immunoassay was demonstrated previously by comparison of the two methods for CsA therapeutic monitoring HPLC and AxSYM. Result showed that the correlation coefficient is 0.99 .
2.4. Pharmacokinetic and Statistical Analysis
Peak CsA concentration (Cmax) and the time to reach it (Tmax) were recorded for each patient and were reported as the mean ± standard deviation.
Pharmacokinetic parameters were calculated using a noncompartment model. The area under the plasma concentration-time curves was calculated from drug plasma concentration data based on trapezoid rule .
2.4.2. Limited Sampling Strategy (LSS)
CsA LSSs were developed and validated from 60 ASCT patients (30 in the development group and 30 in the validation group) via multiple linear regressions.
Step 1. Multiple linear regressions were performed to estimate abbreviated AUC (dependent variable) and each time point of CsA levels (independent variables) that best fitted the CsA . Selected models are those who have for any sampling time and high correlation coefficient . These analyses produced equations AUC as follows: ; and Cst are coefficients and is the number of samples.
Optimal LSS equations were limited to those utilizing a maximum of four timed concentrations taken within 6 hours after dose, and with a coefficient of determination .
Step 2. To validate the equation, we used a second group of pharmacokinetic profile data obtained from ASCT patients recipients (validation group) with the same characteristics as the initial group. Predictive performances of LSS equations were evaluated by calculating the percentage bias (mean prediction error me %) and percentage precision (root-mean-squared prediction error rmse %) [17, 18]. where is the number of AUC pairs and (% pei) on the AUC, defined as and (% rmse). A positive or negative indicates, respectively, overprediction or underprediction of the concentrations by the model. Precision and mean bias < 15% was acceptable .
2.5. Statistical Analysis
The association between cyclosporine AUC using all available data points (AUC-all) and the AUC values predicted by each LSS was described using Pearson correlation coefficients (); values >0.9 were acceptable. Multiple stepwise regression analysis was used to determine the points of CsA samples that derived model equations that best fitted the CsA .
All statistical analyses were performed using SPSS software for windows (version 11.5.0 Inc., Chicago, Illinois). Data were expressed as mean ± SD.
Thirty full pharmacokinetic profiles of CsA were obtained (Figure 1). The mean CsA dose at the time of the measurement was mg. The peak concentration of CsA generally accrued 2-3 hours after morning dosing. Mean trough concentration was μg·mL−1. The mean was μg·h·mL−1 (Table 2).
The linear regression equations and the concentration at each time point are listed in Table 3. The 10 models that used CsA concentrations at a single time point did not have a good fit (). and were the time points that correlated best with , were, respectively, 0.848 and 0.897. Moreover, the correlation between and the trough concentration ( or ) was poor, for and for (Table 3). The correlations between and CsA concentrations at various points , , and , are demonstrated in Figures 2(a), 2(b), and 2(c). The dose of CsA per kilogram of body weight and total dose of CsA did not correlate with (data not shown).
Pharmacokinetic data from 30 ASCT recipient’s patients were used to test the LSS equations. Multiple linear regression analysis of the correlation between the estimated cyclosporine (CsA) and full hours CsA AUC were listed in Table 4.
From the various equations obtained for the estimation of CsA AUC, the equation delineated from all of the concentration-time data had the highest coefficient of determination (Figure 3(a)).
Among models that used CsA concentrations at 2 time points, best model 1 (), which used and . Among models that used CsA concentrations at 3 time points, best model 12 (), which used , , and . Among models that used CsA concentrations at 4 time points, best model 17 (), which used , , , and (Figure 3(b)). However, the negative bias obtained for this combination (−0.21%) indicates that it tends to underpredict AUC (Figure 2(f)). Optimal LSS equations were limited to those utilizing a maximum of three timed concentrations taken within 4 hours after dose, and with a coefficient of determination ():
Predictive performance of all LSSs models were summarized in Table 5. All three optimal LSSs equations developed from ASCT recipient’s patients in our laboratory yielded a low bias <5% ranged from 1.27% to 2.68% and good precision <15% ranged from 9.60% and 11.02%. The LSS equation with the best predictive performance (bias, precision, and number of samples) was (4), which utilizes , , and (Figure 2(e)). The LSS equation (2) utilizes and demonstrated low bias (2.68%) and good precision (11.02%; Figure 2(d)).
Graft-versus-host disease (GVHD) remains a major limiting factor for a successful result after ASCT; it affects mortality, morbidity, and quality of life. CsA-based immunosuppression has been the most frequently used regimen for prophylaxis and treatment of GVHD in patients undergoing HSCT [19, 20]. Therapeutic monitoring of CsA provides a more accurate measure of posttransplant immunosuppression and this by assessment of pharmacokinetic parameters. The area under concentration-time curve (AUC) is a more important index in therapeutic drug monitoring and in pharmacokinetics. In this investigation patients were received a mean dose of cyclosporine A (neoral or equoral) mg. A recent study demonstrates that two formulations were bioequivalent in ASCT patients .
The time to reach Cmax (Tmax) in this study was ; this is similar to the values reported for other HSCT patients, h  h . As previous studies have reported, Tmax for HSCT recipients has shown that absorption is delayed in comparison to that measured in solid organ allograft recipients . It is thought that the differences are due to the presence of gastrointestinal inflammation caused by mucositis or GVHD .
A number of trials have employed monitoring in solid organ transplantation , monitoring is being increasingly employed in the management of solid organ recipients with a recent survey of renal transplant centres finding that most of the respondents were now measuring concentration . In a stem cells transplant setting, where it is clear that current practice of trough-level monitoring still results in unacceptably high levels of GVHD, therefore is a scope to explore the potential of monitoring the drug (both I.V and oral) via alternative measures of exposure or activity . A recent investigation demonstrate a close relationship between and the after oral administration of CsA in allogenic haematopoietic stem cell transplantation (ASTH) , and between and two concentrations and after twice-daily infusion [12, 13].
Therapeutic monitoring of Tunisian ASTH patients was based on determination; this concentration is weakly correlated with ). That we bring to search what concentrations are well correlated with AUC and how to estimate abbreviated AUC in ASCT.
The major findings of this study are that the concentration measurements at and were the time points that correlated best with after oral administration of CsA in allogenic stem cells transplant patients rather than the trough level. These time points provide a more accurate measure of posttransplant immunosuppression. Therefore, the target concentration of CsA in ASHT patients, after the switch to oral administration using monitoring to adjust the dose, is easily applied the same level of monitoring with oral administration.
Another clinically attractive finding of our study was using two time sampling points and for estimating CsA . The regression equation was , . The bias of this model was 2.68%, and the precision was 11.02%. Additional sampling is valuable to support the individual dosing. To further increase the accuracy and decrease bias of predicted , we added time point sampling in absorption phase (, me% = 1.27, rmse% = 9.60) or (, me% = 1.35, rmse% = 9.76). Indeed, optimal LSSs equations, which are limited to those utilizing three timed concentrations taken within 4 hours after dose, are (4) and (3). Optimal Among models that used CsA concentrations at 4 time points are model 17 (). However, the negative bias obtained for this combination (−0.21%) indicates that it tends to underpredict AUC.
Limited sampling strategy models were selected to predict the full 12-hour CsA AUC in allogenic haematopoietic stem cell transplant patients receiving CsA. We propose an LSS model with equation , which is limited to utilizing two timed concentrations taken within 4 hours after dose and providing a low bias and a good precision, it is for a practical reason.
K. R. Schultz, T. J. Nevill, R. F. Balshaw et al., “Effect of gastrointestinal inflammation and age on the pharmacokinetics of oral microemulsion cyclosporin A in the first month after bone marrow transplantation,” Bone Marrow Transplantation, vol. 26, no. 5, pp. 545–551, 2000.View at: Google Scholar
J. L. M. Ferrara, K. R. Cooke, and T. Teshima, “The pathophysiology of acute graft-versus-host disease,” International Journal of Hematology, vol. 78, no. 3, pp. 181–187, 2003.View at: Google Scholar
A. Devergie, “Graft versus host disease,” in Haemopoietic Stem Cell Transplantation, J. Apperley, E. Carrears, E. Gluckman, A. Gratwohl, and T. Masszi, Eds., pp. 162–177, Forum Service Editore, 2004.View at: Google Scholar
M. Oellerich, V. W. Armstrong, B. Kahan et al., “Lake Louise consensus conference on cyclosporin monitoring in organ transplantation: report of the consensus panel,” Therapeutic Drug Monitoring, vol. 17, no. 6, pp. 642–654, 1995.View at: Google Scholar
B. Nashan, E. Cole, G. Levy, and E. Thervet, “Clinical validation studies of neoral C2 monitoring: a review,” Transplantation, vol. 73, no. 9, pp. S3–S11, 2002.View at: Google Scholar
L. Barkholt, M. Remberger, H. Bodegård, O. Ringdén, and Y. Böttiger, “Cyclosporine A (CsA) 2-h concentrations vary between patients without correlation to graft-versus-host disease after allogeneic haematopoietic stem cell transplantation,” Bone Marrow Transplantation, vol. 40, no. 7, pp. 683–689, 2007.View at: Publisher Site | Google Scholar
T. Furukawa, T. Kurasaki-Ida, M. Masuko et al., “Pharmacokinetic and pharmacodynamic analysis of cyclosporine A (CsA) to find the best single time point for the monitoring and adjusting of CsA dose using twice-daily 3-h intravenous infusions in allogeneic hematopoietic stem cell transplantation,” International Journal of Hematology, vol. 92, no. 1, pp. 144–151, 2010.View at: Publisher Site | Google Scholar
Y. Inoue, T. Saito, K. Ogawa et al., “Pharmacokinetics of cyclosporine A at a high-peak concentration of twice-daily infusion and oral administration in allogeneic haematopoietic stem cell transplantation,” Journal of Clinical Pharmacy and Therapeutics, vol. 36, no. 4, pp. 518–524, 2011.View at: Publisher Site | Google Scholar
B. Koristkova, M. Grundmann, H. Brozmanova, I. Perinova, and K. Safarcik, “Validation of sparse sampling strategies to estimate cyclosporine a area under the concentration-time curve using either a specific radioimmunoassay or high-performance liquid chromography method,” Therapeutic Drug Monitoring, vol. 32, no. 5, pp. 586–593, 2010.View at: Publisher Site | Google Scholar
M. Rowland and T. Tozer, Clinical Pharmacokinetics Concepts and Applications, Lippincott Williams & Wilkins, 3rd edition, 1995.
L. S. L. Ting, N. Partovi, R. D. Levy, K. W. Riggs, and M. H. H. Ensom, “Limited sampling strategy for predicting area under the concentration-time curve of mycophenolic acid in adult lung transplant recipients,” Pharmacotherapy, vol. 26, no. 9, pp. 1232–1240, 2006.View at: Publisher Site | Google Scholar
F. Locatelli, C. Uderzo, G. Dini et al., “Graft-versus-host disease in children: the AIEOP-BMT group experience with cyclosporin A,” Bone Marrow Transplantation, vol. 12, no. 6, pp. 627–633, 1993.View at: Google Scholar
C. Peters, M. Minkov, H. Gadner et al., “Statement of current majority practices in graft-versus-host disease prophylaxis and treatment in children,” Bone Marrow Transplantation, vol. 26, no. 4, pp. 405–411, 2000.View at: Google Scholar
B. O. Tarek, B. F. Nadia, K. Anis et al., “Assessment of bioequivalence of a generic cyclosporine (Equoral) by a prospective randomized controlled trial on allogeneic stem cell transplant recipients,” Transplantation Proceedings, vol. 42, no. 9, pp. 3735–3738, 2010.View at: Publisher Site | Google Scholar
J. S. Choi, S. H. Lee, S. J. Chung, K. H. Yoo, K. W. Sung, and H. H. Koo, “Assessment of converting from intravenous to oral administration of cyclosporin A in pediatric allogeneic hematopoietic stem cell transplant recipients,” Bone Marrow Transplantation, vol. 38, no. 1, pp. 29–35, 2006.View at: Publisher Site | Google Scholar
M. Cantarovich, E. Elstein, B. de Varennes, and J. S. Barkun, “Clinical benefit of neoral dose monitoring with cyclosporine 2-hr post-dose levels compared with trough levels in stable heart transplant patients,” Transplantation, vol. 68, no. 12, pp. 1839–1842, 1999.View at: Google Scholar
D. W. Holt, V. W. Armstrong, A. Griesmacher, R. G. Morris, K. L. Napoli, and L. M. Shaw, “International federation of clinical chemistry/international association of therapeutic drug monitoring and clinical toxicology working group on immunosuppressive drug monitoring,” Therapeutic Drug Monitoring, vol. 24, no. 1, pp. 59–67, 2002.View at: Publisher Site | Google Scholar