Presentation No. 8 - Increasing the Success Rate of Quantitative LC-MS/MS Bioanalysis in a Discovery Environment
Garnet McRae
Sussex Research Laboratories, 100 Sussex Drive, Ottawa, Ontario, Canada
Novel Aspect: An approach to fast, accurate LC-MS/MS analysis in a discovery environment. The use of generic methods and useful tips for more difficult compounds combined with discovery-stage appropriate Standard and QC preparation and acceptance criteria help to increase the success rate of analysis the first time.
Introduction: In a discovery environment, many New Chemical Entities (NCE’s) are seen in a given day, week or month with many of them seen only once. It is imperative to provide timely, accurate analysis in a very short amount of time to allow go-no go decisions to be made quickly. A generic approach to discovery-grade LC-MS/MS analysis is required to permit fast, accurate analysis of in-vitro and in-vivo samples with little or no prior method development. Standard and QC preparation as well as acceptance criteria are not as stringent as GLP requirements at the discovery stage, however we need to ensure that these criteria are set appropriately for the stage where a particular compound is in the discovery process.
Methods: Compounds were spiked via serial dilution into the appropriate matricies (Caco-2 buffers, microsomal matrix, plasma etc) at the appropriate concentration for the in-vitro or in-vivo test. Extraction was by protein precipitation with MeOH containing internal standard followed by direct injection onto the LC-MS/MS system.
LC-MS/MS Conditions: A Shimadzu Prominence LC System with an Ace 3-C18 30 x 2.1mm, 3µm column and mobile phases consisting of: MP-A: 100:0.1;water:formic acid, MP-B: 100:0.1;ACN:formic acid were used. The gradient was: 2 to 90%B in 1.0 min, 90%B until 1.6 min at a flow of 0.5 mL/min with a total run time of 2.5 min. An MDS/Sciex API-2000 mass spectrometer was used in ESI positive and negative modes.
Results: Typical results using the generic LC-MS/MS method approach to discovery are normally within GLP acceptance criteria for Precision and Accuracy (+/- 15%) as well as specificity and carryover (?20% of LLOQ). Depending on the NCE, application and calibration curve range, carryover may exceed the 20% of LLOQ limit, however this is permissible in a discovery environment where curves are injected in ascending order and samples injected in time-point order. The generic method should be characterized with respect to where known potentially problematic endogenous compounds elute. The Lysophosphatidyl choline and phosphatidyl choline elution times were evaluated relative to NCE and other small molecule retention times using the generic gradient and protein precipitated rat plasma. Retention times of NCE’s can be monitored to determine if they are co-eluting with phospholipids to give a warning of potential method issues. Atenolol, Pindolol, Propranolol, Digoxin and Digitoxin elute prior to the phospholipids, however Domperidone elutes at the start of the phospholipid region indicating possible method issues. MS/MS compound optimization is a necessary step in the analysis of NCE’s. For discovery stage methods, optimize on multiple product ions during infusion to allow a choice of transitions for injection of the run. Inject the LLOQ and Blanks with all transitions and then choose the best one to inject the run. Some compounds are problematic due to low ionization efficiency or poor fragmentation. A few tips can help to increase sensitivity issues in short order. Compounds should be infused into mobile phase at normal method flow rate and mass spec temperatures to get a good idea of sensitivities. Ion Spray voltage should be evaluated as low voltages (mixed mechanism ionization) (1500 V) can provide a signal boost in some cases over higher voltage values (4500 V). When injecting the extracted LLOQ samples, evaluate high and low voltages to ensure that the S/N gain is real ie) that the noise does not increase at the same rate as the compound signal. An ISV of 1500V was chosen in this case to provide a S/N increase of up to 2.5-fold. During infusion, be aware of typical adducts ie) Neg mode: formate or acetate, Pos mode: Na, K, NH4. Some compounds may show low ionization efficiency for the M+H due to high adduct formation and may also show poor fragmentation, resulting in low sensitivity. In these cases, survivor ion scans or adduct to parent transitions may be useful to maintain high sensitivity. Digoxin is an example of a Negative mode formate adduct (m/z 825) fragmented to molecular ion (m/z 779) to overcome poor fragmentation efficiency. The S/N of an LLOQ for Digoxin were 4, 7 and 2 respectively, for formate adduct survivor ion scan (825-825), formate adduct to molecular ion scan (825-779) and molecular ion survivor scan (779-779). The formate adduct to molecular ion scan was chosen in this case as it provided the best S/N. Discovery-grade standard and QC samples are not usually prepared the same as for GLP-grade methods. The number of concentration levels, whether QC’s are included or not and the number of replicates used depends on the stage of the NCE. -New NCE: Curve only (6 concentration levels); -NCE Hit: Curve (5-7 concentration levels) + QC’s (3 concentration levels); -Lead Compound: Curve (7-8 concentration levels) + QC’s (3 concentration levels); -Curve and QC’s (if used) should bracket samples in all cases. Discovery grade acceptance criteria are often more lenient than for GLP methods. The acceptance criteria depend on the stage of the NCE. -New NCE: Precision and Accuracy (P+A) +/- 25%, Specificity: ?50% of LLOQ, carryover: no criteria (use blanks after high standards and inject samples in order); -NCE Hit: P+A, +/- 25%, specificity: ?20% of LLOQ, carryover: ?50% of LLOQ; -Lead Compound: P+A, +/- 15%, specificity: ?20% of LLOQ, carryover: ?20% of LLOQ.
Stabilities and Matrix Effect are normally not evaluated in early discovery, however should be evaluated as a compound progresses: -New NCE: No stability or matrix effect; -NCE Hit: Room Temp (RT) plasma stability, no matrix effect; -Lead Compound: Stock stability, Plasma stability (RT, F/T, L/T at -80 or -20C) and matrix effect.
Summary: To increase the success rate of quantitative LC-MS/MS bioanalysis, the following steps should be performed: -Generic Extraction and LC Parameters;-Use Specificity of MS/MS; -Use tips to increase chances of immediate success; -Prepare curves and QC’s and set acceptance criteria based on stage of NCE; -Evolution of method from Discovery-grade to GLP-grade.
