National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines for Use of Tumor Markers in Clinical Practice: Quality requirements
Скачать 395.18 Kb.
|
MASS SPECTROMETRY IN CANCER DIAGNOSTICS5 6BackgroundDespite impressive scientific, medical and technological achievements over the past few decades, cancer is still a leading cause of death, largely because most cancer patients are diagnosed when disease is advanced. Accumulating evidence suggests that in the case of many cancers, early detection is associated with improved survival rates (122). Mass spectrometry (MS) has the potential to revolutionize cancer diagnostics by facilitating biomarker discovery, generating proteomic profiles as cancer signatures, enabling tissue imaging and quantifying biomarker levels. The principles of MS as applied to cancer diagnostics are summarized here, together with recommendations for the use of this technique in clinical practice, based on currently published evidence and expert opinions. The main focus will be on matrix-assisted laser desorption/ionization (MALDI) and related MS techniques such as surface-enhanced laser desorption/ionization (SELDI) for proteomic analysis. PRINCIPLES OF DIAGNOSTIC MASS SPECTROMETRY The typical mass spectrometer consists of an ion source, a mass analyzer that measures the mass-to-charge ratio (m/z) of the ionized analytes and a detector that registers the number of ions at each m/z value (123). There are two approaches for biomarker discovery using MALDI/SELDI-TOF MS. One approach uses the differences between MS profiles of the disease and control specimens to generate a diagnostic model. A variation of this approach is to select several discriminate peaks and identify the nature of these protein/peptide peaks. Diagnostics are based on multiplex immuno-MS or ELISA. The other approach is to degrade enzymatically (usually with trypsin) the proteins to peptides, separate the peptides by techniques such as high performance liquid chromatography (HPLC) and direct the eluted fractions into an ion source [electrospray ionization (ESI) or MALDI) where they are converted into ionized species that enter the mass spectrometer followed by identification of the protein fragments and parent proteins comprising the mass spectra by a variety of algorithmic approaches (124). Mass spectrometric measurements are carried out in the gas phase of ionized species. Two commonly used techniques to volatize and ionize the proteins or peptides are electrospray ionization (ESI) and matrix-assisted laser desorption/ionization (MALDI) (125-127). A variant of the latter is SELDI (Ciphergen, ProteinChipTM) (128, 129). The mass analyzer separates ionic species according to their m/z ratios. Four basic types of mass analyzers are commonly used in proteomic research: the ion trap, time-of-flight (TOF), quadrupole and Fourier transform ion cyclotron resonance (FT-ICR), with a potential fifth variant being the new Orbitrap mass spectrometer (Thermo Electron, Inc). These basic types may be variously combined in hybrid instruments. Protein identification is achieved through either peptide mass fingerprinting or peptide sequencing. In the former, peptide masses are compared with mass spectra of proteins listed in databases using appropriate software (130, 131). Peptide sequencing is based on induction of random cleavage of peptide bonds between adjacent amino acid residues, using approaches such as collision-induced dissociation (CID). The resulting ion series is analyzed by software (132-136) to determine the amino acid sequence. APPLICATION OF MASS SPECTROMETRY IN CANCER DIAGNOSTICS Mass spectrometry has been applied to cancer in a variety of contexts, including
Mass spectrometry is considered to be particularly well suited to serve as a diagnostic or biomarker discovery tool in cancer, given emerging evidence that during cancer development, cancer cells and/or the surrounding microenvironment generate proteins and peptides of different type and in different concentrations than normal cells. These abnormal tissue distributions can be analyzed by imaging-based mass spectrometry and the patterns compared with controls to identify cancer-specific changes that may prove to be clinically useful. Should leakage to the circulation occur from the tumor-host microenvironment, then a multiplex of cancer-specific analytes may be detectable in the blood as well, leading to even more widespread clinical utility and convenience of testing (137-140). This concept is graphically illustrated in Figure 1. The identification of cancer-specific protein patterns in blood by mass spectrometry was demonstrated by several investigators including Vlahou et al. (141) for bladder cancer, by Li et al (142) for breast cancer, by Petricoin et al. (139) and Rai et al. (143) for ovarian cancer, and Adam et al. (144) for prostate cancer. Subsequently, many other investigators have used similar approaches to identify multiple markers and informative profiles for many other types of cancer (Table 9). Blood and urine, the most accessible and diagnostically useful body fluids, have been most studied, although other fluids such as nipple aspirate fluid and conditioned media have value as sources for biomarker discovery. In almost every published paper, the profiles generated by MALDI-TOF MS have been shown to yield better diagnostic sensitivities and specificities than the established cancer biomarkers in current use. Because of this, the MALDI-TOF MS approaches have received extensive publicity since they promise to revolutionize early cancer detection, sub-classification, prognosis, prediction of therapeutic response, etc. However, the initial enthusiasm has been tempered somewhat by parallel reports that have identified potential problems with this approach and its clinical reliability (93, 145-162). These issues are not unlike those facing the gene transcript profiling community (163). Future validation studies will determine how ready this technology is for clinical application. CURRENT ADVANCES AND EXISTING LIMITATIONS OF MALDI-TOF MASS SPECTROMETRY-BASED PROFILING FOR CANCER DETECTION If MALDI-TOF profiling is to be successful in the transition from a research technique to a clinical diagnostic tool, then an extensive understanding of pre-analytical, analytical and post-analytical sources of variation must be realized and controlled (93, 145-162). For example, the effect of sample storage and processing, sample type, patient selection and demographic variables (gender, age) on test outcome must be clearly established (164). Analytical performance must improve to the point where sensitivity, specificity and the dynamic range become comparable to those of established techniques such as ELISA. The reproducibility of protein patterns across different batches of chips (when SELDI-TOF is employed), different analysts, different sites and different instrumentation is still under investigation. Robustness of the methodology, in general, is of concern, as are issues related to bioinformatic artifacts, data over-fitting and bias arising from experimental design. However, a large number of these issues relate to inappropriate analysis of publicly available mass spectral data sets that were not meant to be compared. Recently, a large consortium of investigators has shown success at reproducibly obtaining mass spectral signatures, including diagnostically important ones, at multiple sites across time and instruments. This finding establishes a very positive result for those attempting to employ MALDI-TOF type approaches for protein fingerprinting based diagnostics (158). The current limitations and promises of MALDI-TOF, particularly as applied to clinical practice and cancer diagnostics, are addressed more fully in several recently published reviews (93, 138, 140, 145, 147, 159-162). KEY POINTS – MASS SPECTROMETRY PROFILING IN CANCER DIAGNOSTICS Despite numerous publications describing impressive results of MALDI-TOF mass spectrometry as a diagnostic tool (Table 9), the level of published evidence, as described by Hayes et al. (27), is Level IV-V (evidence from either retrospective or small pilot studies that estimate distribution of marker levels in sample population). According to the criteria of Pepe et al. (165), the stage of development of this technology as a biomarker tool is Phase 1 (preclinical exploratory studies). Based on this information, the recommendations shown in Table 10 have been formulated. There is little question that MALDI-TOF MS approaches are promising for biomarker discovery and validation. As for direct profiling of patient specimens for diagnostic use, the issues discussed in this document would need to be resolved. The advantages of proteomic profiling include analysis without the need for a labeling molecule, potentially high specificity, multiparametric analysis, high-throughput, very low sample volume requirements, and direct interface with computer algorithms. The major limitations of the MALDI-TOF technology for MS profiling type work are, at present, a) the cross-platform reliability of the signatures generated; b) dramatic effects on final spectral composition from subtle changes in sample handling and processing; and c) analytical sensitivity, especially when the analyte is present in minute amounts in a highly complex mixture that includes high abundance molecules. However, there are inherent advantages of certain MALDI-TOF approaches. For instance, combining immune isolation prior to MALDI-TOF analysis allows for elimination of secondary antibodies and detection of multiple derivative analytes such as protein isoforms (166, 167). In addition, exciting new research has indicated that many low abundance proteins and low molecular weight analytes exist in a bound state in the serum, and are effectively amplified by carrier protein based sequestration (168-173). These low molecular weight analytes appear to have underpinned many past spectral fingerprints, thus indicating that many of these ions may be generated from low abundant analytes. A list of these low molecular weight carrier protein bound analytes has recently been provided for early stage ovarian cancer patients (171, 174), and the concept verified in an independent study with Alzheimer’s disease detection (172). In that study, high-resolution MALDI-TOF serum proteomic profiling of Alzheimer disease samples reveals disease-specific, carrier-protein-bound mass signatures. These recent findings, together with other recent publications (146, 175-177) describing truncated or fragments of proteins (“fragmentome”) of the circulatory proteome, indicate that MALDI-TOF based approaches may be measuring analytes that are disease specific, of lower abundance than previously thought, and novel. As for all technologies that directly impact patient health, until extensive validation studies are performed, MALDI-TOF MS fingerprinting approaches should not be used as a diagnostic test for cancer in clinical practice. Investigators should perform such validation experiments following CAP/CLIA based codes of good laboratory practice and should provide data in a transparent form for full evaluation by the scientific community. In experiments in which MS fingerprinting is being employed, appropriate independent validation sets should also be employed using inflammatory and benign controls along with high numbers of unaffected controls, since specificity will be an important determining factor of success in the clinic, especially for screening indications. Despite recent difficulties in extending research observations for genomic and proteomic profiling, the field is now evolving with a better understanding of potential sources of bias and instrument variances, as well as the requirements for developing good laboratory practices and standard operating procedures such that clinical adoption and validation could be achieved in the foreseeable future. Table 9. Mass spectrometry for cancer diagnosis and imaging
Table 10. NACB Recommendations for use of MALDI-TOF MS in Cancer Diagnostics
Figure legend Figure 1. Secretion of specific biomarkers into the blood circulation by tumors. Tumor-specific proteins may be actively secreted by tumor cells or released into the circulatory system by necrosis and apoptosis of these cells. Either of these conditions leads to an alteration of the serum protein profile. This may result in detectable differences based either on relative or unique signal intensities when comparing sera from normal and disease samples. [Figure reproduced from Reference (137) with permission from copyright owners] REFERENCES 1 NACB Quality Requirements Sub-Committee Members: Catharine M Sturgeon [Chair], Elizabeth Hammond, Soo-Ling Ch’ng, Györg Sölétormos and Daniel F Hayes. 2 All comments received about the NACB Recommendations for Quality Requirements are included in the on-line supplement. Dr A, Professor B, Professor C and Professor D were invited Expert Reviewers. 3 NACB Microarray Sub-Committee Members: Eleftherios P Diamandis [Chair], Manfred Schmitt and Da-elene van der Merwe. 4 All comments received about the NACB Recommendations for Microarrays in Cancer Diagnostics are included in the on-line supplement. Professor E was an invited Expert Reviewer. 5 NACB Mass Spectrometry Sub-Committee Members: Daniel W Chan [Chair], Oliver J Semmes, Emmanuel F Petricoin, Lance A Liotta, Da-elene van der Merwe and Eleftherios P Diamandis. 6 All comments received about the NACB Recommendations for Mass Spectrometry in Cancer Diagnostics are included in the on-line supplement. Professor F and Professor G were invited Expert Reviewers. |
