An increasing number of pre-clinical studies have demonstrated a pivotal role of miRA-126 in regulating angiogenesis, a process that has also been related to the delivery and the efficacy of chemotherapy. Little is known, though, about the in vivo localisation of miRNA-126 in human CRC tissue samples, but the present results support the previous studies indicating specificity towards ECs. In the present study we analysed the possible predictive value of miRNA-126 in relation to first line XELOX therapy in patients with mCRC, using a quantitative analysis of the miRNA-126 expression level based on ISH analyses of tumour sections from the primary tumour. This method has recently demonstrated its reliability in predicting short disease-free survival in stage II colon cancer patients [26
A significant relationship between miRNA-126 expression levels and number of metastatic sites was demonstrated. The miRNA-126 expression level was significantly lower in patients with two or more metastatic sites compared to the patients with disease limited to only one metastatic location. This could indicate that the molecular genetic features of the tumour cells originating from the primary tumours with low levels of miRNA-126 are related to the ease by which these circulating tumour cells invade tissues and gives rise to distant metastases. This seems to apply rather well with the conception of miRNA-126 functioning as a tumour suppressor.
In the present study a significant relationship between miRNA-126 expression levels in the primary tumour and response to first line XELOX treatment was demonstrated for patients with mCRC. Previous studies have also reported on predictive value of markers related to angiogenesis in patients with mCRC [28
]. The present marker, miRNA-126, is related to vessel integrity [15
], and correlations between vessel structure and response to chemotherapy have previously been demonstrated [24
] supporting the plausibility of the present results. Furthermore, a study by Zhou et al.
] reported on changes of miRNA expression profiles in colon cancer cell lines following exposure to XELOX. Low expression levels of miRNA-126 may therefore represent tumour vessels with lower integrity and the expected increase in the interstitial tumour pressure that follows may explain the lower response rates seen in these patients.
Dividing the patients by cut-off method, i.e. median miRNA-126 expression level, demonstrated that the observed relationship with response rates was also translated into a corresponding difference in PFS. A difference in OS was also demonstrated, when comparing patients with low and high miRNA-126 expression levels. The relationship with OS may indicate that miRNA-126 has a prognostic importance in addition to the predictive value. We did not find it indicated to perform a multivariate survival analysis adjusting for parameters of prognostic importance in stage I through III disease in this cohort of patients with stage IV disease. In the study by Díaz et al.
no prognostic value of miRNA-126 was demonstrated in patients with stage I to III colon cancer [30
]. The disease stage of the included patients constitutes one of the major differences between their study and ours, which included patients with stage IV disease only. Schepeler et al.
showed that miRNA-126 was up-regulated in patients with subsequent recurrence compared to patients with no recurrence in a cohort of 37 patients operated for stage II colon cancer with microsatellite stable tumours only [31
]. Differences in methods, sample size, and stage of the disease again represent the most obvious explanations of these discrepancies, but the results from the three studies indicate that the possible prognostic value of miRNA-126 remains to be further clarified. If indeed, assessment of miRNA-126 expression in the primary tumour from patients with CRC harbours prognostic value, one would expect a better prognosis for patients with higher expression levels of this miRNA if functioning as a tumour suppressor.
The present study has the standard limitations of a retrospective study; a rather limited sample size and results obtained form a single institution. Having said that it also represents a well argued hypothesis, a reliable method, and results of clinical relevance in line with the preclinical literature.
An alternative technique to quantify angiogenesis is presented based on ISH of miRNA-126 in primary tumour tissue and using a specific LNA based probe and new quantitative systematic image analyses. This approach resulted in an estimate related to disease characteristics, the response to chemotherapy, and the prognosis of patients with mCRC. The new application of the ISH method presented here indicated that miRNA-126 may be an important predictive marker to chemotherapy applied in the clinical setting, but the results call for validation in a prospective trial. Furthermore, it shall be interesting to analyse the possible predictive value of miRNA-126 in patients with mCRC treated with chemotherapy combined with anti-VEGF-A.