Immunosuppressant drugs are indicated to reduce the activity of the immune system to prevent transplant rejection. Circulating concentrations of these compounds should remain within a narrow therapeutic window, as overdosing can cause serious toxicity and long-term morbidity, while underdosing can cause graft rejection . As there is a high pharmacokinetic variability for immunosuppressant drugs between individual patients, TDM is now an established approach to mitigate the risks associated with organ transplantation.
Several commercial immunoassays are available for the TDM of immunosuppressants; however, all immunoassays show a significant positive bias compared to LC-MS/MS methods . Despite the availability of automated immunoassays, each test is restricted to one analyte per test even though multiple immunosuppressants are used with an individual patient in many clinical settings [1,10]. The following paragraphs describe a new streamlined and automated LC-MS/MS method for the routine TDM analysis of immunosuppressants.
Combining individual solutions
To meet the need for high-throughput analysis, a DOSIMMUNE™ kit (Alsachim) was adapted to robotic liquid handling. This kit, validated and CE-IVD marked with manual sample preparation, provides all the necessary reagents and consumables to monitor immuno-suppressants in whole blood by LC-MS/MS. Calibrator and control vials can be directly positioned in robot racks (Fig. 1). However, liquid handlers have a larger dead volume that cannot be sucked up, compared to manual pipetting. So, two vials of each calibrator and control level were pooled prior to installation on the racks. With this adjustment, calibrators and controls can be used for an entire week before being replaced.
In order to reduce the number of consumables and protocol steps, an IS solution was mixed directly with the extraction buffer. The mixture can be then dispensed in a reservoir or through a solvent delivery unit.
Workflow optimization of the liquid handler was conducted to reach the shortest possible preparation time while maintaining the performance obtained with the manual protocol. Most efforts were concentrated on the blood resuspension step. To obtain accurate results, samples should be homogeneous during sampling. Therefore, resuspension of each sample was performed just before sampling. A specific protocol was developed using several sequences of aspiration/delivery using the sample height follow-up function of the robot. With this protocol, samples with volumes ranging from 0.5 to 5 mL and haematocrit values from 30 to 70% were successfully homogenized with a 60% time reduction compared to similar protocols observed in laboratories. Vortex and centrifugation time/speed parameters were also optimized and operations parallelized as much as possible. Finally, a complete 96-well plate (blank plus 6 calibrators, 4 controls and 85 samples) was provided ready-to-be-assayed in 45 minutes. Compared to traditional robotic solutions for immunosuppressant analysis, this corresponds to a 50% reduction in preparation time.
Immunosuppressants are complex hydrophobic compounds that are known to produce carry-over in LC. The new Nexera X3 LC system (Shimadzu) has improved rinsing capabilities that dramatically reduce the time needed to clean inside the needle. Consequently, the LC-MS/MS cycle time was reduced from 2.1 to 1.5 min (injection-to-injection). When considering the number of samples to be assayed, this 30% reduction greatly improves turnaround time.
Evaluation and validation
Precision and stability of the solution were evaluated by sequentially preparing three plates containing only the same control at low–mid levels (tacrolimus, sirolimus and everolimus at approximately 8 µg/L, cyclosporin A at approximately 145 µg/L) as unknown sample. The first plate included calibrators and controls; other plates included controls. Plates were assayed within the same sequence. Finally, 272 data points were obtained across the three plates. Figure 2 shows the measured concentrations for all compounds and the measured IS peak area. The relative standard deviation of concentrations or peak areas were well below the usually accepted criteria of 15%. This demonstrated that the MS calibration was stable, that the sample preparation precision was excellent and that reagents were stable on the robot tray. A typical chromatogram of such low–mid level control samples is shown in Figure 3.
The method including sample preparation was validated according to international guidelines for bioanalytical methods (ICH M10 , CLSI C62-A ). In addition, clinical performance evaluation was evaluated by comparing the results obtained on real patient samples using manual preparation. Statistical analysis demonstrated that the methods were in close agreement and commutable. Figure 4 shows the results of Passing–Bablok regression analysis.
Throughput and compliance
By adjusting the number of liquid handlers and LC-MS/MS systems, it is possible to assay from 300 to almost 900 patient specimens in less than 10 working hours. Figure 5 shows an example of a suggested configuration to reach very high throughput.
In addition to the analytical optimization, informatic connections were studied to be in line with the system speed. The liquid handler automatically generates the working list by scanning the patient tubes and this list is exported directly in the correct format to the Shimadzu LabSolutions software for LC-MS. With a few clicks, the sample assay sequence is created. Using adaptive flagging rules, the software proposes a review-by-exception that highlights the data that need to be looked at by the analyst. Double data validation is possible before sending the results to the laboratory information system through csv files or HL7 bidirectional connection.
All of the instruments and software involved in this time-saving solution are CE-IVD certified. The DOSIMMUNE™ kit has also been certified as CE-IVD with manual or automated sample preparation.
In summary, an efficient and time-saving new method using an optimized and fully compliant combination of instruments, reagents and software now exists for high-throughput analysis