A quantitative multigene RT-PCR assay for prediction of recurrence in stage II colon cancer: Selection of the genes in four large studies and results of the independent, prospectively designed QUASAR validation study.
Kerr D., Gray R., Quirke P., Watson D., Yothers G., Lavery IC., Lee M., O'Connell MJ., Shak S., Wolmark N., Genomic Health and QUASAR Colon Teams None.
4000 Background: New clinical tools are needed to improve risk assessment and treatment decisions in stage II colon cancer. Four development studies [Surgery (Sx) alone: NSABP C-01/C-02 (n=270) and CCF study (n=765); Sx+5FU/LV: NSABP C-04 (n=308) and C-06 (n=508)] were performed to select the genes for prediction of recurrence and 5FU/LV benefit. To determine clinical utility of the prespecified assay, we performed a large, independent, prospectively designed, clinical validation study in stage II colon cancer pts from the QUASAR trial. METHODS: Gene expression was quantitated by RT-PCR from 30 μm manually microdissected fixed paraffin-embedded primary colon cancer tissue. Recurrence-free interval (RFI), disease-free survival (DFS), and overall survival (OS) were analyzed using Cox regression. RESULTS: Combined analysis of the four development studies (total n=1,851; 761 candidate genes) identified 48 genes significantly associated with recurrence risk and 66 genes predictive of 5FU/LV benefit. Multivariate analysis, in the context of stage, grade, nodes examined, and MSI status, yielded 18 genes (7 prognostic genes, 6 predictive genes, 5 reference genes) and separate prognostic recurrence score (RS) and predictive treatment score (TS) algorithms. In the QUASAR validation study, tumor blocks were collected for 68% of pts; 1,490 pts with blocks had stage II colon cancer and RT-PCR was successful in 1,436 eligible pts (711 Sx, 725 Sx+5FU/LV). Median FU=6.6 yrs. In the primary analysis of RFI in pts following Sx, the RS predicted recurrence risk (HR/25 units=1.58, 95% CI 1.15-2.15; p=0.004). The RS also predicted DFS (p=0.01) and OS (p=0.04). Recurrence risk increased monotonically with increasing RS. In multivariate analyses, RS retained prognostic significance (p=0.008) independent of mismatch repair (MMR), T stage, nodes examined, grade, and lymphovascular invasion. MMR deficiency (HR=0.31, 95% CI 0.15-0.63; p<0.001) and T4 stage (HR=1.94, 95% CI 1.35-2.79; p=0.005), together ∼25% of pts, also were independently prognostic. 5FU/LV benefit was significant (p<0.001). However, TS was not validated as a predictor of 5FU/LV benefit (interaction p=0.19). CONCLUSIONS: The colon cancer recurrence score is a validated, independent predictor of individualized recurrence risk for stage II colon cancer patients following surgery. [Table: see text].