Opioid Receptor Expression in Colorectal Cancer: A Nested Matched Case-Control Study

Abstract

Background: There is growing interest in the possible effect of perioperative anesthetic management on the growth and spread of cancer. The impact of perioperative use of opioids on cancer recurrence remains controversial and an assessment cannot yet be established based on current publications. This study aimed to assess the differential expression of opioid receptors between healthy and tumor tissues in patients with stage II and III colorectal cancer undergoing elective surgery by immunohistochemistry (IHC).

Methods: Propensity-score matched case-control study nested in a retrospective cohort of patients with stage II or III colorectal. The primary endpoint was the difference in µ-opioid receptor (MOR) expression measured by IHC between tumor and healthy tissue in subject with or without recurrence. Secondary endpoints were to evaluate the differences in Opioid Growth Factor Receptor (OGFR), cyclic adenosine monophosphate (cAMP) production and protein kinase A (PKA) in the matched sample and from a from samples of colorectal cancer stored in the Cancer Genome Atlas (TCGA) and Genotype Tissue Expression Project (GTEx).

Results: There was a significant difference in MOR receptor (median 3 [intequartile range IQR: 1-3] and 0 [IQR: 0-2], P<0.001) and OGFR receptor (median 6 [IQR: 5-6] and 2 [IQR: 1-2], P<0.001) in tumor and control tissue respectively. However, there were no significant differences in cAMP nor PKA expression between both types of tissues and in expression in any of the analyzed variables by recurrence status. The MOR and OGFR expression data from TCGA database were similar to our sample size data with lower expression of MOR and higher expression of OGFR in tumoural samples with a skewed distribution for MOR expression in tumor tissue both in patients with and without recurrence.

Conclusion: In patients with stage II and III colorectal cancer, overall expression of MOR and OGFR was significantly increased but was not different between previously matched patients with or without recurrence. No differences were found in the analyzed metabolic pathway of cAMP-PKA: These results were confirmed by an in silico analysis of samples from the TCGA-GTEx database.

Keywords: cancer; immunohistochemistry; neoplasm; opioid receptors; perioperative opioid; surgery; tumor.

Figure 1

Propensity score matching diagnostic plots. Panel (A) jitter plot of propensity scores. The middle lines show the close match between the randomly selected treatment units and the matched control units. The bottom line shows the unmatched control units not included in the analysis. Panel (B) Histogram distribution before and after the matching process.

Figure 2

Immunohistochemical staining examples to describe scoring. All pictures are at 10X magnification. (A) Score 0 control cAMP; (B) score 1 tumor cAMP; (C) score 2 tumor OGFR; (D) score 3 tumor MOR; (E) score 4 tumor OGFR; (F) score 4 tumor MOR; (G) score 5 tumor OGFR; (H) score 5 tumor OGFR; (I) score 6 tumor MOR; (J) score 6 tumor OGFR. MOR, µ opioid receptor; OGFR, opioid growth factor receptor; cAMP, cyclic adenosine monophosphate.

Figure 3

(A–D) Density plots of MOR, OGFR, cAMP and PKA expression determined by IHC by type of tumor. MOR, µ opioid receptor; OGFR, opioid growth factor receptor; cAMP, cyclic adenosine monophosphate; PKA, protein kinase A.

Figure 4

Boxplot of MOR, OGFR, cAMP, PKA expression by recurrence group. Panels (A–D) show results from IHC staining from the nested matched case-control sample. Panels (E, F) show gene expression from the TCGA and GTEx repositories.