The application of mass spectrometry to leukemia drug discovery

KEYWORDS:

• Acute myeloid leukemia
• mass spectrometry
• drug development
• proteomics

1. Introduction

Acute myeloid leukemia (AML), the most common form of acute leukemia in adults, is a clonal hematopoietic stem cell disorder in which uncontrolled proliferation of leukemic blasts results in life-threatening bone marrow failure. Standard treatment strategies include induction chemotherapy, generally with cytosine arabinoside and an anthracycline, followed by post-remission treatment with chemotherapy or hematopoietic stem cell transplantation. Standard therapy for older AML patients unable to tolerate conventional induction includes low-dose cytarabine and hypomethylating agents (decitabine and azacitidine). While over 70% of younger patients and approximately 50% of selected older AML patients are able to achieve initial complete remission, most eventually relapse and die from their disease. Outcomes are especially dismal for patients unable to achieve remission, those for whom remission is short-lived, and those with relapsed disease.

Significant advances in understanding the biology of AML have resulted in an explosion of different therapeutic agents and strategies, including novel chemotherapeutics, hypomethylating agents, signal transduction inhibitors, molecularly targeted agents, and a variety of cellular and immunotherapies. Mass spectrometry (MS), an analytical chemistry tool that allows qualitative and quantitative assessment of the components of complex mixtures, has been used in all phases of leukemia drug development, including identification of lead compounds and confirmation of their structure and purity, drug manufacturing, bioanalytical studies of pharmacokinetics and pharmacodynamics during early phase clinical trials, and therapeutic drug monitoring [1].

1.1. Basic principles of MS

Mass spectrometers are instruments designed to determine what components, and how much of each component, are present in a sample (analyte). A common feature of all mass spectrometers is that analyte molecules must first be converted into gaseous ions, which are then sorted according to their mass-to-charge (m/z) ratios (not masses). Mass spectrometers generate a beam of gas-phase ions, sort the mixture of ions in electrical and/or magnetic fields or based on the flight times of ions in a field-free region, and provide output signals (peaks) from which the m/z ratio (qualitative analysis, ‘what is present’) and relative abundances (quantitative analysis, ‘how much is present’) of each ionic species may be determined. Detailed structural information about the compound may also be obtained. The five major components of all mass spectrometers are shown in Figure 1 and include the sample introduction system, ionization source, mass analyzer, ion detector, and data system [1]. Complex biological samples generally require separation using high-performance liquid chromatography or capillary electrophoresis for large molecules, gas chromatography–mass spectrometry (GC-MS) for small molecules, or ion mobility spectrometry to separate molecules according to shape. Commonly used methods of ionization in drug development include matrix-assisted laser desorption ionization (MALDI), electrospray ionization (ESI), and chemical ionization (CI). In MALDI, the analyte is first co-crystallized with a matrix. After input of energy from a laser, ions from the matrix ionize the analyte. In ESI, the analyte is dissolved in a conductive solvent, the solution is sprayed through a hollow needle, and evaporation from the charged microdroplets leads to release of gaseous ions. In CI, ionization occurs via ion–molecule reactions. Once ions are generated, they are separated according to their m/z ratios by a mass analyzer. There are several different types of mass analyzers, including time-of-flight (TOF), quadrupole, ion trap, Orbitrap, and others, all of which separate ions either in time or in space. Multianalyzer systems, that is, combinations of two or more similar (tandem) or dissimilar (hybrid) analyzers within a single instrument (known as mass spectrometer/mass spectrometer [MS/MS]), have dramatically extended the capabilities of MS. Depending on the experimental conditions and instrument properties, MS can answer a wide variety of scientific questions related to determination of the molecular mass and structure of compounds, as well as the identification, quantification, and posttranslational modifications of proteins [2].

Figure 1. Components of a mass spectrometer.

1.2. Applications of MS in leukemia therapeutics

Examples of potential applications of MS in drug development are shown in Table 1. MS can be used as a screening tool for novel antileukemic agents. For example, many natural products have been shown to have antileukemic activity and GC-MS has been used to screen for novel agents from the essential oils of algae extracts [3]. MS can also identify new drug targets. For example, identifying and quantitatively profiling phosphoproteins in patient-derived xenografts using MS and stable isotope labeling by amino acids in cell culture identified potential therapeutic targets in pediatric leukemia [4]. MS is well established as a core tool in studies of absorption, distribution, metabolism, and excretion. In leukemia, a particularly useful application of LC-MS/MS is in monitoring compliance and drug–drug interactions with tyrosine kinase inhibitors, for example, imatinib, dasatinib, nilotinib, and sorafenib [5]. MS has also been used to investigate mechanisms of resistance to sorafenib in AML cell lines and patient-derived xenografts [6]. Hydrogen-deuterium exchange MS provides spatial information on protein structure and folding and can be used to identify epitopes, epitope-mapping, and antigen–antibody interactions [7]. Basic LC/MS techniques combined with enhanced analytic strategies (e.g. ion mobility MS) can be used for quantification and characterization of antibody–drug conjugates [8]. A variety of MS techniques are available in proteomics and metabolomics to identify and quantify potential therapeutic targets, diagnostic markers, and markers of therapy resistance [9–11]. The most common general MS techniques are MALDI-MS/MS and LC-ESI-MS; for quantification, the latter is often combined with stable isotope labeling strategies (e.g. isobaric tag for relative and absolute quantification). Quantitative proteomics may provide guidance for the development of novel drugs for AML treatment. For example, this technique was used to reveal the molecular mechanism of action of 2-aza-2ʹ-deoxycytidine, a well-known antileukemic drug [12]. Finally, mass cytometry (CyTOF), a modified flow cytometry technique with readout by TOFMS, has enabled the study of single-cell proteomics. CyTOF has been used to identify functional signaling pathways in antigen-defined subpopulations in primary AML samples [13].

Table 1. Selected applications of mass spectrometry in leukemia drug development.

Application in Drug Development	MS technique	References
Searching for lead compounds
Natural products	GC-MS	[3]
Therapeutic targets	Orbitrap	[4]
ADME
Therapeutic drug monitoring	LC-MS/MS, GC-MS	[5,14]
Drug resistance	LC-MS/MS, Orbitrap	[6]
OMICS
Proteomics	MALDI-TOF/TOF, Orbitrap, ESI-MS/MS	[9,10]
Metabolomics	LC-ESI-MS	[11]
New areas
Antibody–drug conjugates	Ion mobility (IM)-MS, LC-MS	[8]
Protein structural changes	Hydrogen-deuterium exchange MS (HDX-MS)	[7]
Single-cell proteomics/  signaling	Mass cytometry (CyTOF)	[13]

ADME: absorption, distribution, metabolism, and excretion; MS: mass spectrometry; GC: gas chromatography; LC: liquid chromatography; MALDI: matrix-assisted laser desorption ionization; ESI: electrospray ionization.

In AML, the mainstay of treatment is multi-agent chemotherapy, but surprisingly little is known about how the drugs interact with each other or with leukemic cells in vivo. In response to preclinical and clinical data suggesting synergy between the purine analogs cytarabine and clofarabine, investigators developed a new LC-MS/MS method for the simultaneous determination of both agents in human plasma [15]. The technique was then applied to samples collected during the course of an AML clinical trial [14]. The problem is, neither the clinical paper nor the publication describing the MS method attempts to correlate the pharmacokinetic and clinical data. Clinicians could use some guidance on how to optimize combinations of purine analogs. Cytarabine is frequently combined with fludarabine and cladribine, but perhaps we should be mixing and matching the drugs differently. The combination of cytarabine and clofarabine has consistently resulted in improved rates of complete remission, but not in improved overall survival, so interest in the combination has waned. Perhaps further review of the MS data would be helpful.

Unlike with chemotherapy, MS is being actively used in the development of isocitrate dehydrogenase (IDH) 1/2 inhibitors in AML. While research on gain-of-function mutations in IDH1/2 has utilized several sophisticated MS techniques, quantitative monitoring of serum 2-hydroxyglutarate (2HG), an oncometabolite, has been accomplished with simple GC-MS. Ongoing clinical investigations show that IDH1/2 inhibitors can induce complete remission in selected AML patients and that monitoring 2HG may be relevant for prognosis and monitoring therapeutic efficacy [16].

2. Conclusions

Standard treatment algorithms for AML are decades old and, while multiple novel therapeutic approaches are under development, to date, none has emerged as a clear winner. Innovative techniques and applications of MS, a complex analytical tool, may be helpful to optimize and accelerate AML drug development and should be aggressively integrated into real-time evaluation of clinical trial outcomes.

3. Expert opinion

Progress in the treatment of AML has lagged behind that seen in other hematologic malignancies due to molecular and biological heterogeneity, clinical complexity, and the overall aggressiveness of the disease. In our opinion, another major reason for the lack of therapeutic success in AML is the failure to adequately study why patients do and do not respond to a given treatment. Historically, AML drug development has followed a predictable and doomed path: a drug is selected based on preclinical data and/or biological rationale. The drug is then offered to unselected AML patients with relapsed/refractory disease in a phase I trial. Although a phase I trial should only assess safety and drug dose, pharmaceutical companies, investors, clinicians, and patients are always looking for signs of efficacy. This pressure for early success is understandable – there are lives and millions of dollars at stake – but, as innumerable failed trials in AML have shown, unrealistic for this disease. It is extremely difficult to get relapsed/refractory AML patients into remission. Frequently, a clinical trial completes accrual, the results show few or no responses, the investors get nervous, and the drug’s development dies. Even for AML drugs that reach phase II and phase III testing, if they ultimately fail to meet the end points required for drug approval, further clinical or scientific investigation is often not pursued. We feel that several of these ‘failed’ agents could have been rescued if more work had been done to discover why the responders responded and why the nonresponders did not. Sometimes, correlative studies are, in fact, performed (many of which use MS techniques), but the data may never be fully analyzed if the clinical responses are deemed unexciting. Did the drug hit the target? How and to what extent? Did it hit other targets? Were there any significant drug–drug interactions? How did the drug affect the metabolome? The proteome? Are there any relevant biomarkers? Techniques in MS can answer all of these questions, and more, but they are described in review articles much more frequently than in conjunction with primary clinical work. Recently, there have been important flickers of hope that we may be entering a new era of AML therapy. We hope that patients will soon benefit from a ‘perfect storm’ in the clinical and scientific communities of increased understanding of disease pathobiology, technological innovation, and a willingness to revamp clinical trials with more rationally based strategies of patient selection. Novel, real-time applications of MS should be more widely employed during leukemia drug development and clinical trials to investigate why novel therapies are, or are not, helping patients.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This manuscript has not been funded.