KEY POINTS – MICROARRAYS IN CANCER DIAGNOSTICS
There is little doubt that microarrays will eventually become routine diagnostic tools, and the first commercial devices are already on the market (Table 7). However, this is still a relatively new technology and several parameters need to be further optimized and validated prior to their implementation into routine clinical practice, including: selection of optimal capture molecules, standardized hybridization protocols and standardized data collection and interpretation. For DNA and protein microarrays to be reliable tools, they must possess probe sequences that hybridize with high sensitivity and specificity, thereby allowing specific detection of their intended targets. Results must become more reproducible, robust and interchangeable between laboratories, and quality control and quality assurance systems must be established (90). Determining the appropriate level of analytical and clinical validation needed for each application raises new challenges for scientists in industry, academia and regulatory agencies (94).
Two important issues need to be considered when evaluating microarray expression data
Whether the results are valid or accurate for the particular biological system under study, and
Whether the data fundamentally describe the phenomenon being investigated (95).
Introduction of artefacts is possible at any time during an array experiment, therefore each component of the procedure must be carefully considered. The validation process can be divided into three areas: experimental quality control, independent confirmation of data and universality of results (95). Furthermore, before implementation of microarrays into routine practice, it will be preferable to automate the process to minimize variability and increase robustness. Array production, like any other diagnostic device, must meet minimum criteria set by the Food and Drug Administration (FDA) (96). The International Meeting on Microarray Data Standards, Annotations Ontologies and Databases (MGED) focuses on standardization of biochips and proposes appropriate guidelines (97, 98). Despite widespread applications of microarrays in research, the level of evidence of these studies for clinical application, as described by Hayes et al (27), is Level V (evidence from small pilot studies that estimate distribution of marker levels in sample population). Based on the information above, the NACB Panel has formulated the recommendations outlined in Table 8.
Table 6. Microarray applications in cancer diagnostics
Microarray Technology
| Application |
Cancer
|
Reference
|
Comparative genomic hybridization
|
Classification
|
Breast
|
(99, 100)
|
cDNA tissue expression profiling
|
Classification
Therapeutic response
Molecular profiling
|
Breast
Lymphoma
Prostate
|
(101)
(101, 102)
(103)
| Gene expression profiling
|
Prognosis
Classification
Diagnosis
Diagnosis
Diagnosis
Diagnosis
Diagnosis
Diagnosis
Prognosis
Diagnosis
Treatment tailoring
Classification
Classification
Classification
Prognosis
Classification
Classification
Classification
Molecular profiling
Development stages
Mutations
|
Breast
Breast
Ewing sarcoma
Rhabdomyosarcoma
Burkitt lymphoma
Neuroblastoma
GI tumor
Prostate
Prostate
Bladder
Breast
Colorectal
Gastroesophageal
Kidney
Kidney
Ovarian
Pancreas
Lung
Prostate
B-cell lymphomas
BRCA 1 (breast, ovarian)
|
(104, 105)
(104, 106)
(106)
(106)
(106)
(106)
(107)
(108)
(109, 110)
(108)
(108)}
(108)}
(108)
(108)
(111)
(108)
(108)
(108, 112, 113)
(114)
(115, 116)
(105, 117, 118)
| Prognostic signature |
Prognosis
|
Breast
Lung
|
(119)
(120)
| Genome mining |
Biomarker discovery
|
Ovarian
|
(121)
|
Table 7. Some commercially available cancer diagnostic devices based on microarray technology.
Name
|
Intended Use
|
Manufacturer
|
1. Amplichip CYP450
|
Identifies variations in genes CYP2D6 and CYP2C19 for pharmacogenomics
|
Roche
(www.roche.com)
|
2. GeneChip Mapping 100K
|
Whole genome SNP analysis (100,000 SNPs) for establishing disease predisposition
|
Affymetrix
(www.affymetrix.com)
|
3.MammaPrint
CupPrint
|
70-gene signature for breast cancer prognosis
Identifying the primary tumor
|
Agendia
(www.agendia.com)
|
4.p53 GeneChip
|
Sequencing of p53 gene for identifying mutations
|
Affymetrix
|
5. Tumor PSA Array
Tumor Monitoring Array
Colorectal Cancer DNA Array
cDNA Expression Array
|
tPSA, fPSA, CEA
CEA, AFP, hCG, CA19-9, CA125, CA15-3
TP-53, APC, K-ras, BRAF
Ovarian, Breast cancer
|
Randox
(www.randox.com)
|
Table 8. NACB Recommendations for use of microarrays in cancer diagnostics.
Gene expression microarrays are new and promising devices used for cancer diagnosis, prognosis, prediction of therapeutic response, monitoring and selection of therapy. The level of evidence from published studies, according to Hayes et al. (27) is Level V [lowest category]. Consequently, microarrays should continue to be used as research devices, but not as tools for making clinical decisions.
Standardization and clinical validation of expression microarrays is warranted.
Quality control and quality assurance programs for expression microarrays need to be further developed.
Microarray automation is encouraged for improving reproducibility, throughput and robustness.
Tissue microarrays are devices suitable for high-throughput analysis of large numbers of samples and are recommended for use in clinical trials and retrospective studies for evaluating and validating new tumor markers by immunohistochemistry.
Use of microarrays for single nucleotide polymorphism (SNP) analysis is recommended for establishing haplotypes and for correlating these haplotypes to disease predisposition.
Use of microarrays is recommended for high-throughput genotyping and mutation/sequence variation detection for cancer diagnostics and pharmacogenomics. More validation is necessary to ensure equivalent results between standard technologies (such as DNA sequencing) and microarray analysis.
Protein microarrays and other similar technologies are recommended as research tools for multiparametric analysis of large numbers of proteins. The level of evidence is not as yet high enough for clinical applications.
|
