CRC clinical research aims to obtain novel tools to improve prognosis giving light in the decision-making regarding adjuvant chemotherapy. Specifically, identification of reliable factors that improve selection of stage II patients at high-risk of developing recurrence after surgery is of greatest importance
. In the current study, we assessed the relevance of CA 72.4 in prognosis when measured in serum of patients before curative surgery.
DFS rates obtained in this study for TNM stages I-III were highly similar to the ones obtained for Dukes staging, by others and in previous works of us, although rates of death were a bit lower provided this work has not included patients at TNM stage IV and it is a 5-year survival study
Preoperative serum values of CA 72.4 proved its role as prognostic factor, with significantly higher recurrence and death percentages for levels above 7 U/mL. These percentages exceeded the displayed by CEA, the standard for CRC prognosis
[28, 29]. Univariate Cox analysis corroborated the superior value of prognostic information provided by CA 72.4, as recurrence and death are more prone to occur with altered CA 72.4 levels. Nevertheless, CEA complements the prognostic information offered by CA 72.4 since recurrence and death are better predicted when both markers are altered in the patients, being this finding in agreement with the work of Louhimo
Although studies on CA 72.4 in preoperative serum in patients of CRC are very scarce
[24, 30], results agree on the prognostic value of this marker. However, its performance complementing CEA is not well established.
Multivariate Cox analysis was used to generate a model that best explains the recurrence and death relative hazards and test the prognostic independence of the variables that influenced survival in univariate analysis. In addition to TNM, age, tumour location and positivity of tumour markers resulted significant. Positivity of tumour markers is the strongest predictor for recurrence and death.
The current study aimed also to assess the prognostic relevance of CA 72.4 at TNM stage II, to discriminate those patients who are definitely cured after having a curative resection of the tumour, from those that will relapse or even die. Survival analysis confirmed the precise distinction of patients with different DFS and OS by CA 72.4, whilst CEA has a non-significant effect on the recurrence and death hazards, at this early stage. Nevertheless, CEA contribution to prognosis prediction is not negligible, as shows the analysis taking into account the number of positive markers.
For years, efforts have been made on the search for more effective predictors than the traditional staging system not only across all CRC stages, but also regarding early-stage CRC. Genetic and molecular markers that could be useful for detecting who is or not at risk emerged earlier
[3, 4, 31–36]. Nowadays, a different approach based on the identification of clustered genetic alterations and multimarker phenotypes is gaining wider acceptance
Conversely, studies dividing the TNM-stages into subgroups are rare. For instance, in a recent review to study the clinical significance of circulating tumour cells in non-metastatic CRC, Thorsteinsson and Jess
 revised 9 studies and none of them had divided the TNM-stages into subgroups. Recently, a multigene expression assay has been developed to determine the relationship between quantitative tumour gene expression and the risk of cancer recurrence at stage II. This quantitative multigene expression assay is now marketed as Oncotype DX Colon Cancer assay
Unfortunately, to date, all these advances are not yet being translated to the clinical routine, giving arguments for implementation of CA 72.4 into clinical practice. The present research has proved its prognostic relevance that could complement the actual clinical classification of patient prognosis based on the TNM system. This marker has also proved to be valid to determine which patients at stage II are in the high-risk category and therefore should be given chemotherapy, and those who will not have recurrent disease and are therefore in the low-risk category. Moreover, in comparison with genetic markers, determination of CA 72.4 preoperative levels is simple, without necessity of available tissue and cost-effective, all of this corroborated by the fact that is already employed in gastric cancer routine prognosis.