An analysis of pathologic markers and gene expression by stage
in colon cancer indicates that between stage II and III disease,
the vast majority of cancer-related genes studied show a
"remarkable similarity," according to authors of the study,
presented at the 2010 ASCO Annual Meeting.
During the Gastrointestinal Cancer Oral Abstract
Session June 6, lead author Michael O'Connell, MD,
Associate Chairman of the National Surgical Adjuvant Breast and
Bowel Project (NSABP), presented the findings of the NSABP
study.1 Dr. O'Connell said the results shed further
light on the "possibility there's a fundamental difference in
biology between stage II and III colon cancer," which was one
explanation for clinical trial results published by the ACCENT
Group in 2009.
Objectives of the new study were to examine stage II and
III colon cancers for pathologic markers and expression of 375
cancer-related genes, including the 12-gene colon cancer recurrence
score, and to determine the distribution of markers and their
relationship to clinical outcome. Previous research of the
recurrence score (Oncotype DX Colon Cancer Assay, Genomic Health,
Redwood City, California) has shown that it predicts recurrence
after surgery in patients with stage II colon cancer.
The investigators examined 644 patients with stage II
cancer and at least 12 lymph nodes assessed as well as 844 patients
with stage III cancer across four independent development
studies (NSABP C-01/C-02, C-04, and C-06, and a study by the
Cleveland Clinic). Of these patients, 1,035 underwent surgery only,
whereas the others also received chemotherapy.
Pathologic Markers
As expected, nodal status strongly predicted disease recurrence,
Dr. O'Connell said. Patients with stage II cancer were more
likely to have mismatch repair (MMR) deficiency than those with
stage III disease (17.2% vs 12%, respectively). They also were
more likely to have mucinous tumor histology (19.5% vs 14.2%).
Analysis of an interaction between stage and grade showed that
high tumor grade was associated with a favorable prognosis in
stage II but an unfavorable prognosis in stage III, Dr.
O'Connell reported.
Gene Expression
The researchers
quantified gene expression using reverse transcription-polymerase
chain reaction (RT-PCR). Nearly all of the 375 genes studied
appeared to be stage-independent (Fig. 1), with only 45
showing some evidence of interaction with tumor stage. However, Dr.
O'Connell said that 37 of these are likely to be false-positive
findings. Only 5 of the 375 genes (1.4%) showed statistically
significant differences in mean expression by stage in all four
studies.
Genes that showed stage-specific differences in mean expression
included FABP4 and SI (related to cancer cell
metabolism), STMY3 (invasiveness), and EFNB2 and
Maspin (tumor-host interactions). Additionally,
MMR-associated genes demonstrated an interaction of gene expression
and stage, which he called "provocative, given the stage-specific
association of MMR deficiency with recurrence."
Dr. O'Connell said further study is needed to determine the
clinical relevance of the markers that differ between
stages II and III. He added, "We cannot rule out
stage-specific differences in other molecular markers that were not
assessed."
The researchers found no evidence of an interaction between the
recurrence score and tumor stage. According to Genomic Health, this
finding suggests that recurrence score may also predict recurrence
risk in stage III colon cancer. Studies are in progress to
evaluate this possibility. ■
Reference
1. O'Connell MJ, Lavery IC, Gray RG, et al: Comparison of
molecular and pathologic features of stage II and stage III colon
cancer in four large studies conducted for development of the
12-gene colon cancer recurrence score. 2010 ASCO Annual Meeting.
Abstract 3503. Presented June 6, 2010.
Dr. O'Connell reported no relevant financial relationships,
but some of his coauthors are employees of Genomic Health.
