Skip to main content

Interlaboratory comparability study of cancer gene expression analysis using oligonucleotide microarrays.

Dobbin KK, Beer DG, Meyerson M, Yeatman TJ, Gerald WL, Jacobson JW, Conley B, Buetow KH, Heiskanen M, Simon RM, Minna JD, Girard L, Misek DE, Taylor JM, Hanash S, Naoki K, Hayes DN, Ladd-Acosta C, Enkemann SA, Viale A, Giordano TJ

Publication: Clin Cancer Res (2005) • Vol 11 • Page 565-72PMID: 15701842View on PubMed

Purpose

The purpose of this paper was to assess sources of microarray variability among different biological specimens processed and analyzed at four independent laboratories according to standardized protocols.

Conclusion

Based on the level of variation associated with each source (analytical error, interlaboratory, biological) the authors conclude microarray analysis of tumor specimens at multiple laboratories is feasible when using standardized protocols.

Studies (1)

Purpose

The purpose of this study was to assess sources of microarray variability among different biological specimens processed and analyzed at four independent laboratories according to standardized protocols. Surplus tissue from surgical resections were frozen before analysis and included primary lung squamous cell carcinoma (two cases), renal cell carcinoma (two cases), adrenal cortical adenoma, gastric adenocarcinoma, gastrointestincal stromal tumor, uterine leiomyoma, ovarian papillary serous adenocarcinoma, large cell lymphoma, a metastatic melanoma to the lymph node, and a normal liver specimen. DNA and RNA samples derived from these twelve tissues specimens were blinded prior to their distribution and analysis.

Summary of Findings

Within-laboratory gene expression profiles exhibited very strong correlations (0.95-0.97) among tissue, cell pellet, and reference RNA specimens; between -laboratory correlations for specimen replicated were somewhat lower (0.93-0.96). Hierarchical cluster analysis revealed clustering by biospecimen type (frozen tissue, frozen cell pellet, RNA) and by tumor type. For frozen tumor sections, of the three sources of variation among gene expression profiles investigated, biological variation among specimens was the largest contributor followed by analytical error, with inter-laboratory variation having the smallest contribution.

  • Biospecimens

    TissueLung
    TissueAdrenal Gland
    TissueStomach
    TissueKidney
    TissueLiver
    TissueLymph Node
    TissueUterus
    TissueOvary
  • Preservation Types

    Frozen
  • Diagnoses

    NeoplasticCarcinoma
    NeoplasticLymphoma
    NeoplasticMelanoma
    Normal
  • Platforms

    RNADNA microarray
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen AcquisitionBiospecimen location
    Adrenal gland
    Kidney
    Liver
    Lung
    Lymph node
    Ovary
    Stomach
    Uterus