Carboxylesterase 2 as a Determinant of Response to Irinotecan and Neoadjuvant FOLFIRINOX Therapy in Pancreatic Ductal Adenocarcinoma

Abstract

Background: Serine hydrolases (SHs) are among the largest classes of enzymes in humans and play crucial role in many pathophysiological processes of cancer. We have undertaken a comprehensive proteomic analysis to assess the differential expression and cellular localization of SHs, which uncovered distinctive expression of Carboxylesterase 2 (CES2), the most efficient carboxyl esterase in activating the prodrug irinotecan into SN-38, in pancreatic ductal adenocarcinoma (PDAC). We therefore assessed the extent of heterogeneity in CES2 expression in PDAC and its potential relevance to irinotecan based therapy.

Methods: CES2 expression in PDAC and paired nontumor tissues was evaluated by immunohistochemistry. CES2 activity was assessed by monitoring the hydrolysis of the substrate p-NPA and correlated with irinotecan IC50 values by means of Pearson's correlation. Kaplan-Meier and Cox regression analyses were applied to assess the association between overall survival and CES2 expression in patients who underwent neoadjuvant FOLFIRINOX treatment. All statistical tests were two-sided.

Results: Statistically significant overexpression of CES2, both at the mRNA and protein levels, was observed in PDAC compared with paired nontumor tissue (P < .001), with 48 of 118 (40.7%) tumors exhibiting high CES2 expression. CES2 activity in 11 PDAC cell lines was inversely correlated with irinotecan IC50 values (R = -0.68, P = .02). High CES2 expression in tumor tissue was associated with longer overall survival in resectable and borderline resectable patients who underwent neoadjuvant FOLFIRINOX treatment (hazard ratio = 0.14, 95% confidence interval = 0.04 to 0.51, P = .02).

Conclusion: Our findings suggest that CES2 expression and activity, by mediating the intratumoral activation of irinotecan, is a contributor to FOLFIRINOX sensitivity in pancreatic cancer and CES2 assessment may define a subset of patients likely to respond to irinotecan based therapy.

© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

Figures

Figure 1.

Analysis of CES2 expression in pancreatic cancer cell lines and tissues. A) Protein expression level of CES2 in pancreatic cancer cell lines based on normalized MS2 spectral counts (MS2#). Pancreatic cancer cell lines in which CES2 was identified with two or more different peptides by MS analysis in total cell lysate (TCL), media, or cell surface (CES2-positive) are listed. All colon, breast, and SCLC cell lines were CES2 negative. Only one out of 45 NSCLC (H2405) and one out of eight ovarian cancer cell lines (Skov3) were positive for CES2 by MS analysis. B-C) Western blot analysis of CES2 expression in pancreatic cancer cell lines (B) and patient-derived xenograft (PDX) tumors (C). β-actin served as a loading control. D) CES2 mRNA expression levels in normal pancreas (n = 4), pancreatic ductal adenocarcinoma (PDAC, n = 112), normal lung (n = 58), lung adenocarcinoma (LuAd, n = 423), normal breast (n = 108), and breast cancer (n = 610) tissues based on The Cancer Genome Atlas (TCGA) RNA-seq data. Boxes indicate 25th and 75th percentiles, and horizontal lines inside the boxes indicate median. Bars indicate 10th and 90th percentiles. Data have normal tissue median value subtracted. P values were calculated by two-sided unpaired t test: normal pancreas vs PDAC P < .001, vs normal lung P = .58, vs normal breast P = .79; normal lung vs LuAd P < .001, vs PDAC P < .001; normal breast vs breast cancer P < .001, vs PDAC P < .001. E) Immunohistochemical analysis, using tissue microarray, of CES2 expression in pancreatic ductal adenocarcinoma (PDAC). Representative micrographs showing strong cytoplasmic staining for CES2 in moderately differentiated PDAC and negative CES2 staining in nontumoral pancreatic tissue (upper panel). Percentages of analyzed PDAC and paired nontumoral pancreatic tissues segregated according to CES2 expression levels. High CES2: strong staining (score 3) in 10% or more cells (bottom panel). Intermediate CES2: moderate staining (score 2) in 10% or more cells. Low CES2: negative or weak staining. Fisher’s exact test was applied to calculate two-tailed P value. LuAd = lung adenocarcinoma; PDAC = pancreatic ductal adenocarcinoma; PDX = patient-derived xenograft.

Figure 2.

Analysis of irinotecan sensitivity in relation to CES2 activity in pancreatic cancer cell lines. A) Half-maximal inhibitory concentration (IC50) values of irinotecan in pancreatic cancer cell lines. B) Percent survival of pancreatic cancer cell lines after treatment with increasing molar concentrations (log scale) of irinotecan. Curves were fitted by nonlinear regression analysis. C) CES2 activity in pancreatic cancer cell lines. Activity was assessed by monitoring transformation of para-nitrophenolic acetate into its hydrolyzed analogue, para-nitrophenol (p-NP), in the presence or absence of CES2-selective inhibitor, loperamide. Activity is expressed in unit of nmoles of p-NP formed per minute per one mg of cell lysate. D) Pearson’s correlation analysis between irinotecan IC50 and CES2 activity values in pancreatic cancer cell lines. Results are the mean ± SD (error bars) of three independent experiments of five replicates. Two-sided P value was calculated by Pearson’s correlation analysis. IC50 = half-maximal inhibitory concentration; p-NP = para-nitrophenol.

Figure 3.

Effect of CES2 knockdown and overexpression on pancreatic cancer cell irinotecan sensitivity. A) Western blot analysis of CES2 expression in: (left panel) Hs 766T, AsPC-1 and CFPAC-1 pancreatic cancer cell lines nontransduced (parental) and stably transduced with a shRNA targeting CES2 (CES2 KD) or a scrambled shRNA as a control (control); and (right panel) SU.86.86 pancreatic cancer cell line nontransduced (parental) and stably transduced with a lentiviral construct overexpressing CES2 (CES2 OE), and the respective empty vector as a control (control). β-actin was used as loading control. B) Quantitative real-time polyermase chain reaction analysis of CES2 transcript levels, C) relative CES2 activity, and D) relative half-maximal inhibitory concentration (IC50) of irinotecan in the above described pancreatic cancer cell lines after CES2 knockdown and overexpression. Results are the mean of three independent experiments ± SD (error bars) of triplicates. P values were calculated by two-sided unpaired t test. *P < .05, ** P < .01, *** P < .001 relative to parental cell lines. CES2 KD = CES2 knockdown; CES2 OE = CES2 overexpression; IC50 = half-maximal inhibitory concentration.

Figure 4.

Kaplan-Meier survival plot of pancreatic cancer patients who underwent first-line neoadjuvant FOLFIRINOX therapy stratified according to CES2 expression. A) All patients (CES2 high, n = 8; CES2 intermediate, n = 8; CES2 low, n = 6) or B) resectable and borderline resectable included in multivariable analysis (CES2 high, n = 7; CES2 intermediate, n = 6; CES2 low, n = 5) were segregated by CES2 expression as assessed trough immunohistochemical analysis. High CES2: strong staining (score 3) in 10% or more cells. Intermediate CES2: moderate staining (score 2) in 10% or more cells. Low CES2: negative or weak staining. Two-sided log-rank test was applied. CI = confidence interval; HR = hazard ratio.