Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses - PubMed (original) (raw)

. 2001 Nov 20;98(24):13790-5.

doi: 10.1073/pnas.191502998. Epub 2001 Nov 13.

W G Richards, J Staunton, C Li, S Monti, P Vasa, C Ladd, J Beheshti, R Bueno, M Gillette, M Loda, G Weber, E J Mark, E S Lander, W Wong, B E Johnson, T R Golub, D J Sugarbaker, M Meyerson

Affiliations

Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses

A Bhattacharjee et al. Proc Natl Acad Sci U S A. 2001.

Abstract

We have generated a molecular taxonomy of lung carcinoma, the leading cause of cancer death in the United States and worldwide. Using oligonucleotide microarrays, we analyzed mRNA expression levels corresponding to 12,600 transcript sequences in 186 lung tumor samples, including 139 adenocarcinomas resected from the lung. Hierarchical and probabilistic clustering of expression data defined distinct subclasses of lung adenocarcinoma. Among these were tumors with high relative expression of neuroendocrine genes and of type II pneumocyte genes, respectively. Retrospective analysis revealed a less favorable outcome for the adenocarcinomas with neuroendocrine gene expression. The diagnostic potential of expression profiling is emphasized by its ability to discriminate primary lung adenocarcinomas from metastases of extra-pulmonary origin. These results suggest that integration of expression profile data with clinical parameters could aid in diagnosis of lung cancer patients.

PubMed Disclaimer

Figures

Figure 1

Figure 1

Hierarchical clustering defines subclasses of lung tumors. Two-dimensional hierarchical clustering of 203 lung tumors and normal lung samples was performed with 3,312 transcript sequences. The normalized expression index for each transcript sequence (rows) in each sample (columns) is indicated by a color code (see EXPRESSION INDEX bar at lower left of figure). (A) Clusters of genes with high relative expression in normal lung (NL, pink branch). (B) Neuroendocrine tumors: small-cell lung cancer (SCLC, gold branch) and pulmonary carcinoids (COID, light blue branch). (C) squamous cell lung carcinomas with keratin markers (SQ, light green branch). (D) Proliferation-related markers. Adenocarcinomas resected from the lung (black branches) and a subset of adenocarcinomas suspected as colon metastases (red branch) are indicated. Color bars on the right correspond to regions displayed in Fig. 10.

Figure 2

Figure 2

Clustering defines adenocarcinoma subclasses. Comparison of classifications derived by hierarchical clustering (dendrogram) and probabilistic clustering (colored matrix) algorithms. The two-dimensional colored matrix is a visual representation of a corresponding numerical matrix whose entries record a normalized measure of association strength between samples. Strong association approaches a value of 1 (red) and poor association is close to 0 (blue). CM, colon metastasis; NL, normal lung; C1 through C4 are adenocarcinoma clusters; I (gold), II, and III are additional groups with weaker association. See Figs. 12–15 for details and sample names.

Figure 3

Figure 3

Gene expression clusters and histologic differentiation within lung adenocarcinoma subclasses. Genes expressed at high levels in specific subsets of adenocarcinomas. The normalized expression index is shown as in Fig. 1. (A) Colon metastases. (B) Proliferation-related gene expression (C1). (C) Neuroendocrine gene expression (C2). (D) Ornithine decarboxylase 1 and surfactant gene expression (C3 and C2). (E) Type II pneumocyte gene expression (C4, C3, and normal lung). (F) Histopathological degree of differentiation (red, poor; yellow, moderate; green, well; white, not available or irrelevant). (G) Estimated nucleated tumor content: white, not determined or irrelevant; gray, 30–40%; blue, 40–70%; black, greater than 70%).

Figure 4

Figure 4

Survival analysis of neuroendocrine C2 adenocarcinomas. Kaplan–Meier curves for C2 versus all other adenocarcinomas. (A) All patients: C2, n = 9; non-C2, n = 117. (B) Patients with stage I tumors only: C2, n = 4; non-C2, n = 72.

Similar articles

Cited by

References

    1. Greenlee R T, Hill-Harmon M B, Murray T, Thun M. CA Cancer J Clin. 2001;51:15–36. - PubMed
    1. Druker B J, Talpaz M, Resta D J, Peng B, Buchdunger E, Ford J M, Lydon N B, Kantarjian H, Capdeville R, Ohno-Jones S, et al. N Engl J Med. 2001;344:1031–1037. - PubMed
    1. Travis W D, Travis L B, Devesa S S. Cancer. 1995;75:191–202. - PubMed
    1. Sorensen J B, Hirsch F R, Gazdar A, Olsen J E. Cancer. 1993;71:2971–2976. - PubMed
    1. Breathnach O S, Kwiatkowski D J, Finkelstein D M, Godleski J, Sugarbaker D J, Johnson B E, Mentzer S. J Thorac Cardiovasc Surg. 2001;121:42–47. - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources