Morphologic assessment of lung tumors is informative but insufficient
to adequately predict patient outcome. We previously identified
transcriptional profiles that predict patient survival, and here
we identify proteins associated with patient survival in lung
adenocarcinoma. A total of 682 individual protein spots were quantified
in 90 lung adenocarcinomas by using quantitative two-dimensional
polyacrylamide gel electrophoresis analysis. A leave-one-out cross-validation
procedure using the top 20 survival-associated proteins identified
by Cox modeling indicated that protein profiles as a whole can
predict survival in stage I tumor patients (P = 0.01). Thirty-three
of 46 survival-associated proteins were identified by using mass
spectrometry. Expression of 12 candidate proteins was confirmed
as tumor-derived with immunohistochemical analysis and tissue
microarrays. Oligonucleotide microarray results from both the
same tumors and from an independent study showed mRNAs associated
with survival for 11 of 27 encoded genes. Combined analysis of
protein and mRNA data revealed 11 components of the glycolysis
pathway as associated with poor survival. Among these candidates,
phosphoglycerate kinase 1 was associated with survival in the
protein study, in both mRNA studies and in an independent validation
set of 117 adenocarcinomas and squamous lung tumors using tissue
microarrays. Elevated levels of phosphoglycerate kinase 1 in the
serum were also significantly correlated with poor outcome in
a validation set of 107 patients with lung adenocarcinomas using
ELISA analysis. These studies identify new prognostic biomarkers
and indicate that protein expression profiles can predict the
outcome of patients with early-stage lung cancer.