Y. Victoria Zhang, PhD, MBA
February 2019—Mass spectrometry has been proved to be an essential and powerful platform for clinical diagnostics. The value it adds to health care has increased steadily in the past decade, as was showcased at the 8th Annual American Association for Clinical Chemistry Conference on Mass Spectrometry and Separation Sciences for Laboratory Medicine, held last fall in Philadelphia. Stakeholders from academia, hospitals, industry, and regulatory bodies came together to exchange information about developments in the technology and best practices to meet patient care needs.
Pre-conference on-site workshops helped attendees gain firsthand experience with and knowledge of mass spectrometry’s applications in clinical care and omics-based research functions. The University of Pennsylvania, Children’s Hospital of Philadelphia, Temple University, and NMS Labs opened their doors to host the events.
The conference consisted of 18 presentations, among them three by young investigators, and one panel discussion. Here are highlights of six of the 18 presentations.Dennis J. Dietzen, PhD, AACC president and professor of pathology and immunology at Washington University in St. Louis, gave a plenary lecture in which he spoke about the utility of traditional metabolic profiling in pediatric and maternal-fetal pathology. He asked a key question: While these metabolic profiles are used routinely to detect inborn errors of metabolism such as PKU and MCAD deficiency, can they shed light on other disorders with unclear pathogenesis? When considered in the proper clinical context, he said, circulating concentrations of amino acids and carnitine esters provide a wealth of information about nutrition, protein synthesis, metabolic fuel utilization, vascular tone, and nitrogen balance.
Dr. Dietzen talked about animal and human studies to uncover metabolic patterns that might predict liver regeneration, poor pregnancy outcome, and severe neonatal hypoglycemia. Circulating lysine and methionine catabolites reflect liver regeneration and suggest a better prognosis in children with acute liver failure. Such markers of liver regeneration may augment the common biomarkers of hepatocyte dysfunction, such as bilirubin, albumin, and prothrombin time. Elevated maternal concentrations of arginine, alanine, hydroxyproline, and glutamate in the first trimester appear to predict the occurrence of preeclampsia and intrauterine growth retardation. Finally, increased concentrations of acetyl carnitine and branched chain amino acids in neonates characterize a poor ketogenic response to neonatal hypoglycemia and may identify infants who need therapeutic dietary supplementation. The receiver operating characteristics for these profiles yielded areas from 0.75 to 0.85, suggesting that subtle metabolic alterations in these case/control cohorts can provide diagnostic and mechanistic insight to maternal-fetal pathology. Further studies may clarify the metabolic basis of these childhood disorders and lead to more precise prediction, identification, and treatment.
Neil L. Kelleher, PhD, Walter and Mary E. Glass professor of molecular biosciences, professor of chemistry and medicine, and director of the Proteomics Center of Excellence at Northwestern University Feinberg School of Medicine, presented the increasingly important role of proteoforms in human health and disease. The post-translational modification (PTM) of proteins is a critical method for regulating protein activity, he said, but typical proteomic approaches to identifying and localizing PTMs require enzymatic digestion of proteins to generate peptides. These bottom-up proteomic approaches can be useful in high-throughput identification of peptide PTMs, but they suffer from the protein inference problem and are further complicated when amino acid substitutions and multiple possible PTMs can be colocalized onto the same protein molecule.
Top-down proteomics (TDP) circumvents the protein inference problem by identifying the full-length, undigested protein molecule complete with possible amino acid substitutions and PTMs, and these protein forms are called proteoforms. Consequently, top-down proteomics can be used to better understand health and disease by providing a more comprehensive view of the functional proteins encoded by the 20,000 human genes. With top-down proteomics, Dr. Kelleher’s group has mapped 11 proteoforms from the oncogene KRAS in the context of colorectal cancer (Ntai I, et al. Proc Natl Acad Sci U S A. 2018;115:4140–4145). And they have begun to understand how driving mutations in the KRAS gene affect post-translational modifications elsewhere on the expressed protein in a low-bias and isoform-specific manner (i.e. KRAS4b versus KRAS4a).
Furthermore, acute organ rejection after an organ transplant is difficult to predict and even more challenging to detect before acute rejection and organ failure. Dr. Kelleher and colleagues used top-down proteomics to identify proteoforms that could serve as potential biomarkers for acute liver rejection in a pilot study of patients who were transplant excellent or had undergone acute rejection.
Together, these studies have underscored the central role of a proteoform-centric approach when studying human health and disease, and these cases demonstrate the need for the proteomics community to consider making the switch from a peptide-centric approach to protein- and proteoform-based approaches.Henry Rodriguez, PhD, MS, MBA, director of the Office of Cancer Clinical Proteomics Research and deputy director (acting) of the Center for Strategic Scientific Initiatives, National Cancer Institute, discussed the convergence of proteomics and genomics, or proteogenomics, and how this emerging, multidisciplinary field is advancing basic and translational cancer research. He provided a look at the NCI’s Clinical Proteomic Tumor Analysis Consortium (CPTAC) and discussed its roots in addressing analytical rigor and reproducibility in proteomic technologies, which led to the standardization of untargeted (discovery-based) and targeted (confirmatory and clinical-based) protein analyses by mass spectrometry, production of quality reagents and resources (antibodies and assays), and development of community-based proteomic open-data sharing policies.
CPTAC’s initial pilot studies on genomically characterized colorectal (Zhang B, et al. Nature. 2014;513:382–387), breast (Mertins P, et al. Nature. 2016;534:55–62), and ovarian (Zhang H, et al. Cell. 2016;166:755–765) tumors from The Cancer Genome Atlas revealed new insights into the molecular subtypes of cancers that could not be uncovered with next-generation sequencing only. The data, published in three flagship papers, indicate the usefulness of a proteogenomic approach in providing a new paradigm for understanding cancer biology. In its current iteration, CPTAC has two sub-programs. The tumor characterization program uses standardized and harmonized proteomic workflows to further characterize additional cancer types, and the translational research program addresses questions of biology in the context of NCI-sponsored clinical trials.
Dr. Rodriguez highlighted the growing interest in applying proteogenomics to precision oncology. CPTAC has forged partnerships with other federal and international entities to generate multimodal data, accelerate progress in cancer research, ensure greater cooperation and collaboration, share data with the greater research community, and ultimately bridge the gap between cancer genomics and clinical action.
Andrew N. Hoofnagle, MD, PhD, professor and head of clinical chemistry in the Department of Laboratory Medicine, University of Washington, summarized efforts over the past decade to use mass spectrometry to quantify proteins in human samples, which, as he pointed out, overcomes the limitations of many clinical protein immunoassays. Quantitative clinical proteomic methods often use proteolysis to generate peptides that are used as surrogates for the intact proteins, he explained, and he described the projects that his laboratory and others have undertaken to properly calibrate protein mass spectrometry assays to give precise and accurate results.
Early on, his laboratory demonstrated superior results when native human serum was used as the calibration material, rather than calibrators made with spiked proteins, when quantifying apolipoprotein A-I and apolipoprotein B. He described his work with the IGF-1 Working Group, funded by the Partnership for Clean Competition, which built on that study and showed that even when laboratories use different liquid chromatography and mass spectrometry systems, assay results can be precise across laboratories when a single-point calibration approach is used.
Dr. Hoofnagle also presented his work in collaboration with the NCI’s CPTAC, which culminated in a monoclonal antibody that can be used to immunoenrich a specific peptide from thyroglobulin after serum proteolysis and before being quantified by liquid chromatography-tandem mass spectrometry. Many reference laboratories in North America now use the antibody, including the University of Washington, for the diagnosis of recurrent differentiated thyroid carcinoma. Dr. Hoofnagle presented interlaboratory comparison data demonstrating that different laboratories using different liquid chromatography and mass spectrometry systems can achieve better agreement than four immunoassays run in the same laboratory.William Clarke, PhD, professor of pathology at Johns Hopkins University School of Medicine, presented information on the use of clinical mass spectrometry for optimizing cancer therapy. He presented the various ways that ambient ionization techniques coupled with mass spectrometry can be used for tissue imaging and characterization, which is important in driving cancer treatment. This can be done using technologies like MALDI mass spectrometry, which can provide high-resolution images but results in destruction of the original sample. Alternatively, imaging can be performed using DESI and other non-destructive ionization sources, but the images are not of the same quality.
Dr. Clarke also reviewed the utility of LC-MS/MS in more traditional applications such as therapeutic drug monitoring, which can be used to optimize the dose of chemotherapy treatments. TDM in cancer chemotherapy is not routine, but there is much evidence suggesting that it can be a useful tool in treating cancer. Dr. Clarke last presented two promising approaches—iKnife and MasSpec Pen—for using mass spectrometry intraoperatively to assist the surgeon in identifying the margins between benign and malignant tissue.
Jeremy L. Norris, PhD, associate professor of biochemistry at Vanderbilt University School of Medicine and managing director of the National Research Resource for Imaging Mass Spectrometry, presented, on behalf of Richard Caprioli, PhD (director, Mass Spectrometry Research Center), their new developments to advance the field of imaging mass spectrometry. Dr. Norris highlighted three of their many areas of focus: new instrumentation to improve sensitivity and spatial resolution, improving ease of use, and new developments in bioinformatics.
He explained that commercial instruments are capable of imaging at resolutions on the order of 10 microns. Their laboratory had pioneered the development of novel source designs that were capable of imaging mass spectrometry below 1-micron resolution, permitting the imaging of tissue at cellular resolution. To increase the quality of images acquired from the minimal sample amounts at those resolutions, Dr. Norris and his group have developed ion accumulation strategies that increase sensitivity two to three orders of magnitude for selected regions of mass-to-charge.
In addition, new reagents for sample preparation have been developed that significantly improve the stability of the matrix in the high-vacuum conditions of the mass spectrometer. These innovations improve the quality of the result for users with limited training and experience, an important step in translating this technology to a clinical setting.
Dr. Norris explained that using recent advances in computational image fusion, the results from imaging mass spectrometry experiments could be computationally combined with image data emanating from a multitude of imaging modalities. This provides the opportunity to correlate molecular information with microscopy and in vivo imaging modalities to add a molecular dimension to these foundational clinical tools.
Improving the diagnosis of atypical melanocytic lesions and differentiating Crohn’s disease from ulcerative colitis are just two examples among many where pathology alone might not provide the detailed information required for a definitive diagnosis. Early results indicate that the molecular data imaging mass spectrometry provides could improve the diagnostic result significantly compared with the current standard of care.
The CAP is a collaborating partner of the annual AACC Conference on Mass Spectrometry and Separation Sciences for Laboratory Medicine. Frank Schneider, MD, of the Department of Pathology and Laboratory Medicine, Emory University School of Medicine, represented the CAP at the 2018 conference, where he presented the new CAP checklist requirements for imaging mass spectrometry. They are part of the 2018 CAP chemistry and toxicology accreditation program checklist. For details, see “New accreditation program checklist section: imaging mass spec scores its own quality standards” (CAP TODAY, October 2018, captodayonline.com/imaging-mass-spec-scores-quality-standards).
Dr. Zhang is vice chair of clinical enterprise strategy, associate professor, and director of the clinical mass spectrometry and toxicology laboratory, Department of Pathology and Laboratory Medicine, University of Rochester Medical Center, Rochester, NY. She is the chair of the organizing committee of this conference and founding chair of the AACC Mass Spectrometry and Separation Sciences Division. The AACC conference had support for young investigator awards from the AACC Philadelphia local section and from the CAP.