Comparison of neuroendocrine differentiation and KRAS/NRAS/BRAF/PIK3CA/TP53 mutation status in primary and metastatic colorectal cancer

Abstract

Neuroendocrine differentiation of tumor tissue has been recognized as an important prerequisite for new targeted therapies. To evaluate the suitability of colorectal cancer (CRC) tissue for these treatment approaches and to find a possible link to pretherapeutic conditions of other targeted strategies, we compared neuroendocrine differentiation and KRAS/NRAS/BRAF/PIK3CA/TP53 mutational status in primary and metastatic CRC. Immunohistochemical expression analysis of neuroendocrine markers chromogranin A and synaptophysin was performed on archival CRC tissue, comprising 116 primary tumors, 258 lymph node metastases and 72 distant metastases from 115 patients. All CRC samples but 30 distant metastases were subjected to mutation analysis of KRAS, NRAS, BRAF, PIK3CA, and TP53. Neuroendocrine marker expression was found significantly less frequently in lymph node metastases compared to primary tumors and distant metastases (20%, 31%, 28%, respectively, P = 0.044). KRAS mutation rates increased significantly from primary tumors to lymph node metastases and distant metastases within the neuroendocrine negative CRC group (44%, 53%, 64%, respectively, P = 0.042). Neuroendocrine differentiation was significantly less concordant than KRAS/NRAS/BRAF/PIK3CA/TP53 mutational status in primary tumor/lymph node metastases pairs (65% versus 88%-99%; P < 0.0001) and primary tumor/distant metastases pairs (64% versus 83%-100%; P = 0.027 and P < 0.0001, respectively). According to these data, therapeutic targeting of neuroendocrine tumor cells can be considered only for a subset of CRC patients and biopsies from the metastatic site should be used to guide therapy. A possible importance of lacking neuroendocrine differentiation for progression of KRAS mutant CRC should be further investigated.

Keywords: KRAS mutations; Neuroendocrine differentiation; colorectal cancer.

Figure 1

Frequency of neuroendocrine differentiation scores 0, 1+ and 2+ in primary and metastatic colorectal cancer. The photo insets display a lymph node metastases without neuroendocrine differentiation (score 0, × 100), a primary tumor with score 1+ synaptophysin expression (× 100) and a distant peritoneal metastasis with score 2+ synaptophysin expression (× 100). Different background color in the table positions indicate primary tumors (white), lymph node metastases (light grey) and distant metastases (dark grey). Frequency of neuroendocrine negativity (score 0) and neuroendocrine positivity (comprising score 1+ and score 2+) are compared separately for mutant (m) cases, considering all five genes analyzed in this study.

Figure 2

Concordance and discordance of neuroendocrine differentiation in matched primary tumor/lymph node metastases pairs (A), primary tumor/distant metastases pairs (B) and distant metastases/lymph node metastases pairs (C), considering all possible combinations of neuroendocrine differentiation scores.

Figure 3

Picture of a KRAS strip assay. Design of the teststrip to the left, real example showing a c.35G < A mutation in codon 12 to the right.