Melissa Hoffman (Presenter)
Moffitt Cancer Center/University of South Florida
Bio: Melissa Hoffman is a fourth year Cancer Biology PhD student at Moffitt Cancer Center/University of South Florida. Her primary research interest is the application of mass spectrometry in precision medicine and cancer research. She earned her BS in Cellular/Molecular Biology with a minor in Biochemistry from California State University, Chico in 2013, where she examined the effect of microRNA transcriptional regulation on diabetes progression. After starting the PhD program and completing rotations, she joined Dr. John Koomen’s lab, which focuses on applying quantitative mass spectrometry to translational cancer research. She has developed personalized assays to detect tumor burden in Multiple Myeloma patients and profiled kinase ATP uptake and pY levels to elucidate kinome signaling in lung cancer patient tumors.
Authorship: Melissa Hoffman; Rachid Baz; Kaaron Benson; Sean Yoder; Frederick Locke; Aunshka Collins; Jamie Teer; John Koomen
Moffitt Cancer Center, Tampa, FL
| Short Abstract Multiple myeloma tumor burden is evaluated using the levels of the monoclonal immunoglobulin (M-protein) secreted by the cancer cells; numerous methods have been applied to this clinical biomarker measurement. Treatment strategies have improved both rates and depths of response, requiring improvement in the sensitivity disease detection. Furthermore, treatment with therapeutic antibodies can interfere with clinical assays, resulting in a false positive. Therefore, we use a proteogenomics approach with RNA-sequencing to determine the M-protein variable region sequences from the disease-specific monoclonal Ig followed by quantification using reaction monitoring mass spectrometry. MS quantification demonstrates improved sensitivity and specificity, eliminating biotherapeutic interference. |
Long Abstract
Introduction: Multiple Myeloma (MM), a devastating hematological malignancy formed by clonal expansion of plasma cells in the bone marrow, remains fatal due to persistence of minimal residual disease (MRD) following treatment, which leads to recurrence. Recent improvements in therapeutic strategies have resulted in unprecedented depths of response and correlative improvements in patient outcomes; however, these responses are not quantifiable in serum samples using current clinical methods.[1-2] Because MM tumors are an abnormal clonal outgrowth of one plasma cell, which normally each produce a unique immunoglobulin (Ig), they secrete an excess of one type, termed monoclonal immunoglobulin (M-protein).[3] The amount of secreted M-protein in serum increases in response to tumor growth, resulting in an ideal disease biomarker. Clinical techniques for M-protein detection include serum and urine protein electrophoresis (SPEP/UPEP) and immunofixation electrophoresis (IFE), which have limited sensitivity (~100 mg/dL).[4-5] Other methods including multi-color flow cytometry, RNA-seq, or PCR amplification,[6] require multiple bone marrow samples, placing a heavy burden on the patient. As treatments continue to improve, novel techniques for disease detection are paramount for effective quantification of tumor burden throughout treatment and earlier detection of disease recurrence, supporting prolonged initial treatments or accelerated therapeutic intervention at disease progression.
Therapeutic monoclonal antibodies (mAbs), including FDA-approved Daratumumab[7] (anti-CD38) and Elotuzumab[8] (anti-SLAMF7) have emerged as a new class of treatment, demonstrating enhanced clinical efficacy in relapsed and refractory patients in combination with other MM targeting agents. There are additional therapeutic antibodies yielding promising results in clinical trials, including a new anti-CD38, Isatuximab (SAR650984).[9] Furthermore, mAbs, which are typically IgG-kappa, can interfere with SPEP and IFE assays by co-migrating with the M-protein, resulting in a false-positive M-spike.[10] Detection of the mAb could erroneously be interpreted as a new clonal outgrowth, resulting in unnecessary follow-up testing, or indicative of recurrence or relapse in IgG-kappa disease patients, impacting clinical decision making.[11] Because clinical criteria for complete response (CR) require the patient’s serum is negative by SPEP and IFE, the interference can lead to an underestimation of CRs in clinical trials for biotherapeutics, indicating a need for a different clinical endpoint or a new detection method for determination of efficacy.
In this study, we employ personalized mass spectrometry (MS) quantification to improve sensitivity and specificity of MM disease burden measurements, which mitigates mAB interference in patients treated with a therapeutic antibody. To complement previously developed liquid chromatography – multiple reaction monitoring (LC-MRM) assays quantifying each Ig class,[12] LC-MRM assays specific to unique peptides from each patient’s disease-specific Ig variable region have been developed. In addition, high resolution and accurate mass measurement liquid chromatography-parallel reaction monitoring mass spectrometry (LC-PRM) assays monitoring therapeutic antibodies (e.g. anti-CD38 Daratumumab and Isatuximab), Ig classes, and patient-specific M-protein in a single assay have been characterized.
Methods: A proteogenomics approach is used to design personalized peptide-based mass spectrometry assays for each patient to monitor their disease progression, which have increased sensitivity and specificity compared with current clinical methods. The M-protein light and heavy variable region sequences are determined by RNA-seq (MiSEQ, Illumina) following amplification by 5’-rapid amplification of cDNA ends (RACE). The protein sequence is generated by translation of de novo RNA sequencing to create a full-length Ig variable region sequence, which is then validated using the online ImMunoGeneTics (IMGT) tools.[13] Tryptic VRPs are predicted from the Ig variable region protein sequence using in silico digestion, and the patient serum is screened by UPLC-MRM (RSLC and TSQ Quantum Ultra, Thermo) for the presence of the predicted VRPs. In addition, high resolution and accurate mass UPLC-PRM assays are developed using a targeted approach on a hybrid quadrupole-orbital ion trap instrument (QExactive HF, Thermo) to analyze the longer peptides that are not as amenable to LC-MRM analysis. To ascertain if the peptide is unique to each patient’s M-protein, a cohort of >85 serum samples is screened. The collective Ig repertoire of this group of patients is used as a surrogate for the Ig repertoire of the individual patient to select the most unique variable region sequences. Standard isotope-labeled peptides are synthesized and characterized for absolute quantification. Peptide-based assays are developed in parallel for the following therapeutic mAbs: Daratumumab, Isatuximab, and Elotuzumab (sequences available in chemical databases).[14] Using this approach, we are generating a database of Ig variable region RNA and predicted amino acid sequences to assist with selection of variable region sequences for patient monitoring.
Tumor burden is measured longitudinally in newly diagnosed patients undergoing treatment and patients in remission to evaluate the ability to detect MRD and disease progression using excess sera from serial samples collected as part of the standard of care. Plasma levels of anti-CD38 mAb are quantified in patients undergoing treatment as proof of concept. In parallel, clinical assays including SPEP, IFE and nephelometry (Ig class expression and serum-free light chain) are performed. Results are compared between LC-MRM and LC-PRM along with clinical measurements. Enzyme digestion kinetics is analyzed to optimize digestion efficiency specific to the VRP and to explore the possibility of using the stable isotope-labeled standards as references for relative quantification of the VRPs. Sensitivity, reproducibility, and linearity of the assays are determined.
Results: Longitudinal sample collection is ongoing for a growing cohort of 66 patients that can be divided into two primary groups: newly diagnosed patients beginning treatment and patients in remission that are monitored for disease progression. RNA sequencing provides copy number counts that enable identification of the amplified sequence corresponding to the M-protein. Amplification of the variable region by 5’RACE using class specific primers (from the literature[15] or newly designed) reduces both the cost of sequencing and time for bioinformatics analysis. This workflow increases confidence in Ig sequence assembly, enabling M-protein variable region identification and successful assay development from bulk marrow cells collected at residual disease. LC-MRM assays targeting both common and unique VRPs multiplexed with CRP quantification, increase MM monitoring efficiency by quantifying the M-protein and all Ig isoforms in one analysis. While assays are peptide- and patient-specific, LC-MRM results demonstrate at least a 100-fold increase in sensitivity, with a lower limit of quantification (LLOQ) of ~0.5-1 µg/dL and average CVs of <10%. Assays show a dynamic linear range up to five orders of magnitude. VRP peptides from the M-protein can be quantified when the patient has had a CR according to clinical evaluation, which includes a negative SPEP and IFE, at the time of serum collection. Personalized LC-MRM assays demonstrate an increase in VRPs, before clinical tests identify relapse.
LC-PRM quantification has resulted in improved sensitivity, lowering the lower limit of detection to ~0.1 µg/dl, when compared to our older triple quadrupole instrument. This instrument has also been applied to quantify tumor burden and therapeutic mAb levels in the sample serum sample. As an example, 6 out of 7 VRPs specific to Isatuximab were observed in longitudinal serum samples collected from a patient undergoing treatment, while none were present in control serum. M-protein quantification was performed both with and without VRP specific SIS peptides to assess the use of VRP to CRP ratio with accurate mass measurements in lieu of SIS peptide synthesis, which is both a time and cost bottleneck for assay development. The different analytical strategies were compared and this proteogenomics approach validated using current clinical methods.
Conclusions: Currently, clinical treatment decisions are made based on semi-quantitative results from multiple tests. An increasing number of studies indicate the correlation between depth of response and improved patient outcomes, suggesting that achieving complete response is necessary for long-term survival.[2] VRP assays personalized to each patient have been developed to quantify MM tumor burden using multiple MS methods (LC-MRM and LC-PRM). Results indicate improved sensitivity in detecting MRD, indicating this may be a valuable tool for MRD assessment in the development of new inhibitors and determining responses in clinical trials. If this assay is translated to the clinical lab, it could change the paradigm for patient evaluation and clinical decision-making, increasing the ability of clinicians to continue first line therapy to achieve deeper responses, or to intervene at an earlier time point upon disease recurrence. Here we present a single test resulting in quantitative readouts on all classes of heavy and light chain Igs. The same methodology was used to develop VRP assays specific to therapeutic antibodies, resolving the M-protein and mAb interference. Elimination of mAB interference when quantifying the disease biomarker could improve results from clinical trials by reducing false positives that can mask CRs.
References & Acknowledgements:
References:
1. Lahuerta JJ, et al. J Clin Oncol 2008, 5775.
2. Nishihori T, et al. Curr Hematol Malig Rep 2016, 118.
3. Durie BG, et al. Leukemia 2006, 1467.
4. Katzmann JA, et al. Electrophoresis 1997, 1775.
5. Katzmann JA, et al. Am J Clin Pathol 1998, 503.
6. Billadeau D, et al. Blood 1991, 3021.
7. Nijhof IS, et al. Clin Cancer Res 2015, 2802.
8. Zonder JA, et al. Blood 2012, 552.
9. Martin TG, et al. Blood 2014, 83.
10. Murata K, et al. Clin Biochem 2016.
11. Genzen JR, et al. Br J Haematol 2011, 123.
12. Remily-Wood ER, et al. Proteomics Clin Appl 2014, 783.
13. Brochet X, et al. Nucleic Acids Res 2008, W503.
14. Bento AP, et al. Nucleic Acids Res 2014, D1083.
15. Doenecke A, et al. Leukemia 1997, 1787.
Acknowledgements:
I’d like to thank my PhD mentor, Dr. Koomen, for providing guidance and support for this project. The research for this presentation was funded by the DeBartolo Personalized Medicine Foundation. The Proteomics Core Facility at Moffitt Cancer Center assisted with equipment upkeep in maintenance, supporting mass spectrometry analysis.
| Description | Y/N | Source |
| Grants | no | |
| Salary | no | |
| Board Member | no | |
| Stock | no | |
| Expenses | no |
IP Royalty: no
| Planning to mention or discuss specific products or technology of the company(ies) listed above: | no |