A prospective study of circulating mutant KRAS2 in the serum of patients with colorectal neoplasia: strong prognostic indicator in postoperative follow up

Abstract

Background and aims: Mutant tumour derived DNA has been detected in the sera of colorectal cancer patients. We investigated if mutant serum KRAS2 was detectable preoperatively in a large group of patients with colorectal neoplasia. A prospective study of 94 patients who underwent putative curative resection for colorectal carcinoma (CRC) was performed to ascertain if serum mutant KRAS2 could be used postoperatively as a disease marker.

Methods: Preoperative sera from 78 patients were analysed (group A). Sera from 94 patients were obtained three monthly for up to three years during the postoperative period (group B). Codon 12 and 13 KRAS2 mutations were analysed in matched tumour and serum samples.

Results: In the preoperative group (group A), KRAS2 mutation was found in 41/78 (53%) tumours and in 32/78 (41%) preoperative sera. Of 41 tumour KRAS2 mutation positive cases, 31/41 (76%) had an identical serum mutation detectable. In group B, the postoperative follow up group, 60/94 cases were primary tumour KRAS2 mutation positive. Of these 60, 16/60 (27%) became persistently serum mutant KRAS2 positive postoperatively. Ten of 16 (63%) of these developed a recurrence compared with only 1/44 (2%) patients who remained serum mutant negative (odds ratio 71.7 (95% confidence interval 7.7-663.9; p=0.0000). None of 34 tumour mutation negative cases became serum mutant KRAS2 positive postoperatively, despite recurrence in 9/34 patients. The relative hazard of disease recurrence in postoperative serum mutant KRAS2 positive patients was 6.37 (2.26-18.0; p=0.000).

Conclusions: Serum mutant KRAS2 can be detected preoperatively in all stages of colorectal neoplasia. Postoperatively, serum mutant KRAS2 is a strong predictor of disease recurrence, stronger even than Dukes' stage of disease, and thus shows potential for use in clinical practice as a marker of preclinical disease recurrence.

Figure 1

Representative results of sequencing of the region flanking codon 12 and 13 of the first exon of the k-ras gene. Mutant bands are indicated by arrows. Part (A): (A) Tumour DNA from Dukes' B case P1, with a codon 12 cysteine mutation. (B) Preoperative serum from case P1, showing an identical codon 12 cysteine mutation. (C) Tumour DNA from Dukes' C case P2, showing a codon 13 aspartate mutation. (D) Preoperative serum from case P2, with the mutation matching that of the primary tumour. (E) Tumour DNA from Dukes' A case P3 showing a double aspartate mutation in both codons 12 and 13. (F) Preoperative serum DNA from case P3 showing mutant codons 12 and 13, identical to that found in the tumour. Part (B): (A) Tumour DNA from Dukes' C case P4, with a codon 13 aspartate mutation. (B) Postoperative serum from case P4, with an identical codon 13 aspartate mutation. (C) Tumour DNA from Dukes' C case P5, with a codon 13 aspartate mutation. (D) Postoperative serum DNA from case P5, with a matched aspartate mutation to that in the tumour. (E) Tumour DNA from Dukes' B case P6, with a codon 12 valine mutation. (F) Postoperative serum DNA from case P6, with a codon 12 mutation, identical to that in the tumour.

Figure 2

Graphical representation of the outcome of all patients in group B according to tumour and serum KRAS2 mutation status. These results are presented in tabular form in table 4 ▶.

Figure 3

Kaplan Meier disease free survival curves according to serum mutant KRAS2, tumour mutant KRAS2, and disease stage. Survival analysis included only patients with Dukes' A, B, or C disease, as inclusion of patients with tubulovillous adenoma or Dukes' D may bias the results because of the low and high recurrence rates, respectively. Below, the symbols (+) and (−) refer to KRAS2 mutation status. (A) In n=85 patients, tumour (+) and (−), survival was significantly worse in serum (+) patients (p=0.0000). (B) In tumour (+) patients (n=54), survival was significantly worse in serum (+) patients (p=0.0000). (C) Stratification of n=54 patients who were tumour (+) on the basis of disease stage showed that serum (+) patients had a significantly worse prognosis in each disease stage (p=0.0000 for early disease and p=0.007 for late disease). The disease free survival curves were similar for patients who were serum (+) and who had either early or late disease and for patients who were serum (−) with either early or late disease.