Prognostic markers for colorectal cancer; estimating ploidy and stroma

Summary

Background

We report here the prognostic value of ploidy and digital tumour-stromal morphometric analyses using material from 2624 patients with early stage CRC.

Patients and Methods

DNA content (ploidy) and stroma-tumour fraction were estimated using automated digital imaging systems and DNA was extracted from sections of formalin-fixed paraffin-embedded (FFPE) tissue for analysis of microsatellite instability (MSI). Samples were available from 1092 patients recruited to the QUASAR 2 trial and two large observational series (Gloucester, n = 954; Oslo University Hospital, n = 578). Resultant biomarkers were analysed for prognostic impact using five-year cancer-specific survival (CSS) as the clinical endpoint.

Results

Ploidy and stroma-tumour fraction were significantly prognostic in a multivariate model adjusted for age, adjuvant treatment, and pathological T-stage in stage II patients, and the combination of ploidy and stroma-tumour fraction was found to stratify these patients into three clinically useful groups; 5-year CSS 90% vs 83% vs 73% (hazard ratio [HR]=1.77 [95% CI 1.13-2.77] and HR = 2.95 [95% CI 1.73-5.03], P < 0.001).

Conclusion

A novel biomarker, combining estimates of ploidy and stroma-tumour fraction, sampled from FFPE tissue, identifies stage II CRC patients with low, intermediate or high risk of CRC disease specific death, and can reliably stratify clinically relevant patient sub-populations with differential risks of tumour recurrence and may support choice of adjuvant therapy for these individuals.