Certain mutations tied to poorer survival in high-risk colorectal cancer
KRAS and BRAF mutations were associated with shorter relapse-free survival in patients with stage II or III colorectal cancer treated with curative intent, particularly in patients who were also microsatellite instability (MSI)-negative, according to a recent study.
“TP53, KRAS, and BRAF mutations were independently associated with poor prognosis, although the association with BRAF and KRAS was restricted to MSI-negative tumours,” said the researchers.
“[W]e have [also] shown that increased mutation burden among non-hypermutated colorectal cancers is associated with favourable outcomes,” they said.
Researchers used targeted next-generation sequencing to identify prognostic biomarkers in patients with stage II or III colorectal cancer treated with curative intent included in the open-label, randomized, phase III QUASAR 2* trial and an independent Australian community-based series (82 and 113 gene panel, respectively). Patients in QUASAR 2 were treated with capecitabine alone or in addition to bevacizumab (no radiotherapy) while those in the community study were treated with adjuvant or neoadjuvant fluorouracil chemotherapy or concurrent chemoradiotherapy.
Based on 511 tumour samples in the QUASAR 2 trial, KRAS [hazard ratio [HR], 1.99, 95 percent confidence interval [CI], 1.37–2.91), BRAF (HR, 2.46, 95 percent CI, 1.51–4.03), TP53 (HR, 1.63, 95 percent CI, 1.12–2.38), and GNAS mutations (HR, 2.76, 95 percent CI, 1.08–7.04) were associated with shorter relapse-free survival (p<0.035 for all). [Lancet Gastroenterol Hepatol 2018;3:635-643]
Conversely, high mutation burden was associated with longer survival (HR, 0.81, 95 percent CI, 0.68–0.96; p=0.014), while there was no association observed between MSI and survival (HR, 1.12, 95 percent CI, 0.57–2.19; p=0.75) in the QUASAR 2 cohort.
Similar findings were observed with the Australian community-based series (253 tumour samples) with KRAS (HR, 1.51; p=0.066), BRAF (HR, 2.18; p=0.029), and TP53 mutations (HR, 1.82; p=0.014) associated with shorter relapse-free survival, higher mutation burden associated with longer survival (HR, 0.78; p=0.014), and MSI having no effect (HR, 0.62; p=0.247).
In an analysis of the combined cohort, a higher mutation burden remained associated with improved survival (n=764; HR, 0.8; p=0.001), including in patients who were MSI-negative with non-pathogenic POLE mutations (n=652; HR, 0.84; p=0.004).
MSI-positivity was associated with good prognosis (HR, 0.45; p=0.00001), while KRAF and BRAF mutations were associated with poor prognosis (HR, 1.22; p=0.035 and HR, 1.53; p=0.004, respectively).
MSI-negative patients with KRAS or BRAF mutations had particularly poor prognosis (HR, 1.35; p=0.003 and HR, 2.02; p=1.20 x 10-5, respectively) compared with MSI-negative patients wild-type for BRAF and KRAS mutations (triple negative), with better prognosis among MSI-positive patients with KRAS or BRAF mutations (HR, 0.28; p=0.028 and HR, 0.55; p=0.017, respectively).
“In view of the modest survival benefits that conventional cytotoxic therapies provide for patients with common solid malignancies, biomarkers of prognosis still have substantial potential clinical importance. Such markers could guide the use of more or less aggressive treatment regimens and enable clinicians to balance expected outcomes against early and late therapeutic toxicities,” said the researchers.
“Our findings show that the use of even a modest-sized gene panel can provide clinically useful information beyond individual driver mutations,” they said, suggesting that larger gene panels could potentially demonstrate stronger associations between mutations and prognosis.