28 February / March 2016
limit of 10 ng/mL for infliximab (Figure 2). Linearity and accuracy of the standard curves arising from each peptide are summarised in Table 1. The primary, and most sensitive quantitative peptide, SINSATHYAESVK, was linear over 4 orders of magnitude with a mean accuracy (N=3) of >98% for all points on the curve. The additional two peptides were linear over 3.5 orders of magnitude with average accuracies >99% for all curve points.
In addition, the accuracy and precision for the QC samples was excellent with %CVs all <6%. This is summarised in Table 2. In fact, the average %CV for QC samples from the SINSATHYAESVK peptide was ≤3%.
From an assessment of the chromatographic data, it is clear that the quality of the data in terms of peak width and separation from residual endogenous components facilitated both the low level detection and the very high accuracy and precision that were achieved.
Unique challenges in the bioanalysis of antibody- drug conjugates
Antibody-drug conjugates combine the unique targeting capabilities of mAbs with the cancer-killing ability of cytotoxic drug. Due to their complex and heterogeneous nature, ADCs often require multiple bioanalytical assays to determine efficacy, toxicity, and PK/PD response during drug development stages.
The top-down approach, using enzymatic digestion of the ADC/mAb followed by LC-MS/MS analysis, is becoming routine for ADC and mAb quantification. Bioanalytical goals for ADCs include quantification of both the conjugated and unconjugated forms of the ADC, total mAb, cytotoxic payload, and various other catabolites/ metabolites. Of the many experiments required to characterise and quantify ADCs, total antibody measurements are important, and are presented in the work herein.
Trastuzumab is a humanised anti-HER2 monoclonal antibody that was approved by the FDA in 1998. With EU patent expiry in July 2014, and impending U.S. patent expiry in 2019, the focus on this drug as well as next-generation potential as an ADC has steadily increased. Ado-trastuzumab emtansine (T-DM1) is an FDA-approved ADC marketed under the brand name Kadcyla, and is used as treatment for patients with advanced breast cancer [3-5]. ADCs, like T-DM1, are composed of cytotoxic small
Table 3: Linear dynamic range, weighting and average accuracy for standard curves for trastuzumab plasma digested and extracted using the specified kit.
Peptide
IYTNGYTR FTISADTSK
GPSVFPLAPSSK* *Generic IgG signature peptide
molecule drug (payload) covalently bound to an antibody by a linker.
Quantification via direct digestion: T-DM1 and Trastuzumab
To prepare standards and quality control (QC) samples, trastuzumab or T-DM1 was spiked into rat plasma at various concentrations (0.1-500 µg/mL). An intact murine monoclonal antibody standard (Waters Corp, p/n 186006552) was used as a generic internal standard.
Using the commercially available kit from Waters and its optional 5-step digestion protocol (which includes reduction and alkylation), a direct digest of plasma (35 µL) containing either T-DM1 or trastuzumab was performed in preparation for LC-MS/MS analysis. Three signature tryptic peptides were used for quantification: IYPTNGYTR, FTISADTSK, and GPSVFPLAPSSK. (See Appendix for analytical method conditions.) While some commercially available kits eliminate reduction and alkylation (we will call this ‘3-step’), the inclusion of reduction and alkylation steps helps ensure detection of conjugated peptides.
From an analytical perspective, tryptic digestion and the choice of a signature peptide pose a challenge for quantification of T-DM1, since it is a lysine-conjugated ADC. Trypsin cleaves peptides on the C-terminal side of lysine amino acid residues, and if a lysine residue is occupied with the cytotoxic drug, cleavage will not occur (‘miscleavage’). Thus, if one were to choose a lysine-containing peptide to quantify T-DM1, there is potential for miscleavage on the lysine residue when it is conjugated with the small molecule drug. Because the signature peptide IYPTNGYTR lacks a lysine residue, one can confidently and accurately use it to quantify both T-DM1 and trastuzumab. For this same reason one would need to be cautious of using the two
lysine containing peptides, FTISADTSK and GPSVFPLAPSSK for accurate quantification of T-DM1. Both of these peptides have some degree of small molecule drug occupancy and thus, due to potential miscleavage of the lysine residue, may result in lower calculated concentrations than a non-lysine containing peptide.
For this application, sensitivity, linearity, accuracy and precision data met typical method validation requirements [6]. Standard curves were linear over 3.5 orders of magnitude with the average accuracies of 100% for the standard curve points. For the IYPTNGYTR, FTISADTSK, and GPSVFPLAPSSK tryptic peptides, quantification limits between 0.5-1 µg/mL were achieved. Summary statistics from standard curves for trastuzumab are shown in Table 3.
In addition, the accuracy and precision for trastuzumab and T-DM1 QC samples were quantified using the above trastuzumab standard curve specified in Table 3 and were found to be excellent with % CVs <8. The values are summarised in Table 4. Representative QC chromatographic performance and demonstration of sensitive quantification for the IYPTNGYTR peptide is highlighted in Figures 3, Panels A (trastuzumab) and B (T-DM1), respectively.
Due to the hydrophobic nature of the cytotoxic drug molecule attached to the antibody and differences in stereo chemical configurations, conjugated T-DM1 peptides generally will elute later in a reversed-phase chromatographic run as diastereomeric pairs. It is also common for T-DM1 peptides to produce a common fragment (547.2 m/z) by collision induced disassociation (CID). This fragment corresponds to part of the drug molecule broken down by the CID process.
In this application, we were successfully able to detect two conjugated, ‘miscleavage’ peptides of T-DM1 (FTISADTSKNTAYLQMNSLR and GPSVFPLAPSSKSTSGGTAALGCLVK) Figure 4, panels A and B illustrate the presence
Std. curve range (µg/mL)
0.5–500 1.0–500 0.5–500
Weighting
1/X2 1/X
1/X2
Linear fit r2
0.995 0.999 0.990
Mean % accuracy of all points
100.01 100.01 100.00
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60
