Increased MPO in Colorectal Cancer Is Associated With High Peripheral Neutrophil Counts and a Poor Prognosis: A TCGA With Propensity Score-Matched Analysis

Abstract

Background: Myeloperoxidase (MPO) has been demonstrated to be a local mediator of inflammation in tissue damage in various inflammatory diseases. Given its controversial effect on colorectal cancer (CRC), there has been growing interest in investigating the role of this enzyme in CRC. The mechanism underlying MPO activity and CRC progression requires further clarification.

Methods: The expression and function of MPO in CRC were evaluated using TCGA analysis. TCGA, TIMER, and Human Cell Landscape analyses were used to analyze the correlation between MPO expression and neutrophil infiltration in CRC. Spearman's bivariate correlation analysis was used to verify the correlation between MPO levels in CRC and the peripheral neutrophil count. In the clinical analysis, 8,121 patients who underwent elective surgery for CRC were enrolled in this retrospective cohort study from January 2008 to December 2014. Propensity score matching was used to address the differences in baseline characteristics. The Kaplan-Meier method and Cox regression analysis were used to identify independent prognostic factors in patients with CRC.

Results: MPO was upregulated in CRC tissues, which is related to malignant progression and worse survival in CRC patients from TCGA analysis. MPO was significantly correlated with the infiltration level of neutrophils in CRC in TCGA, TIMER, and Human Cell Landscape analyses. MPO was positively correlated with the peripheral neutrophil count. Data of the 8,121 patients who underwent CRC surgery were available for analysis. After propensity score matching, 3,358 patients were included in each group. Kaplan-Meier survival curves showed that high preoperative neutrophil levels were associated with decreased overall survival (OS; P < 0.001) and disease-free survival (DFS; P = 0.015). The preoperative neutrophil count was an independent risk factor for OS (hazard ratio [HR], 1.157; 95% confidence interval [CI], 1.055-1.268; P = 0.002) and DFS (HR, 1.118; 95% CI, 1.009-1.238; P = 0.033).

Conclusions: Our research indicates that increased MPO levels in CRC are significantly correlated with high preoperative neutrophil counts, and both serve as prognostic indicators for worse survival in CRC patients. Our study suggests that neutrophils may be key players in the mechanism linking MPO levels with poor CRC outcomes.

Keywords: TCGA analysis; colorectal cancer; myeloperoxidase; preoperative neutrophil counts; prognosis; propensity score-matched analysis.

Figure 1

MPO was upregulated in CRC and correlated with the malignant progression of CRC patients. (A) The MPO expression levels in 480 tumor tissues and 41 normal tissues. (B) The expression of MPO in 41 normal and matched tumor tissues. (C) MPO expression in several cancers in the TIMER2 database. (D–F) The association with elevated MPO and clinicopathological characteristics, including M stages, pathological stage, and CEA level. (G–I) The association with elevated MPO and death events, including OS events, DSS events, and PFI events. The difference was considered significant at * P < 0.05, ** P < 0.01, or *** P < 0.001. ns, no significance.

Figure 2

High expression of MPO is related to poor prognosis in CRC patients. (A–C) Kaplan–Meier analyses of OS, PFI, and DSS between the low- and high-MPO groups in TCGA. OS: Logrank=5.72, P=0.018, n(high)=238, n(low)=239, cutoff value=0.26; DSS: Logrank=7.71, P=0.007, n(high)=230, n(low)=231, cutoff value=0.26; PFI: Logrank=8.94, P=0.003, n(high)=238, n(low)=239, cutoff value=0.26. (D–F) Gene Expression Profiling Interactive Analysis2 (GEPIA2), OncoLnc, and TIMER2 databases were used to analyze overall survival between high and low MPO groups, respectively. GEPIA2 OS: P=0.023, n(high)=128, n(low)=133; OncoLnc OS: Logrank=4.947, P=0.0261, n(high)=220, n(low)=220, cutoff value=3.78; TIMER2 OS: P=0.033, n(high)=229, n(low)=229. (G) ROC analysis with respect to the MPO expression.

Figure 3

Univariate analysis of MPO for survival using the Cox regression model and nomogram based on MPO and clinical variables for survival in CRC. (A) Univariate analysis of prognostic factors for OS using the Cox regression model. (B) Univariate analysis of prognostic factors for DSS using the Cox regression model. (C) Univariate analysis of prognostic factors for PFI using the Cox regression model. (D) Nomogram for OS prognosis in CRC. (E) Nomogram for DSS prognosis in CRC. (F) Nomogram for PFI prognosis in CRC.

Figure 4

MPO was significantly correlated with the infiltration level of neutrophils in CRC. (A) MPO expression has a significant correlation with many immune cells infiltration. (B, C) The correlation between MPO and neutrophils by the Wilcoxon signed-rank and Spearman correlation tests. (D) We analyzed the correlation between MPO expression and T cells, B cells, CD8+ T cells, cytotoxic cells, DC, macrophages, mast cells, neutrophils, NK cells, Th1 cells, Th17 cells, Th2 cells, and Tregs using ssGSEA. (E) The relationship between different immune cells and MPO was analyzed in TIMER 2.0, including B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and DCs. (F–H) The relationship between MPO and neutrophils in a single-cell RNA sequencing database. The difference was considered significant at * P < 0.05, ** P < 0.01, or *** P < 0.001. ns, no significance.

Figure 5

MPO was upregulated in CRC and positively correlated with peripheral neutrophil counts. (A) IHC was used to detect MPO protein expression in CRC and non-cancerous tissues (69 pairs); left, scale bar = 250 µm; right, scale bar = 50 µm. (B) Quantitative analysis of MPO IHC scores in CRC and non-cancerous tissues (69 pairs). (C) qPCR was used to evaluate the relative expression of MPO mRNA in CRC and adjacent non-cancerous tissues (69 pairs). (D) The relationship between MPO levels in CRC and peripheral neutrophil counts was determined using Spearman’s correlation test (r = 0.406, P < 0.001, n = 69). Differences were considered significant at * P < 0.05, ** P < 0.01, or *** P < 0.001.

Figure 6

Flow chart of patient selection. Flow chart of patient selection.

Figure 7

Kaplan–Meier analysis of OS and DFS. (A) Kaplan–Meier survival curve for OS according to preoperative neutrophils (pre-neutrophils) in the propensity score-matched cohort (OS, Logrank=13.743, P < 0.001). (B) Kaplan–Meier survival curve for DFS according to preoperative neutrophils (pre-neutrophils) in the propensity score-matched cohort (DFS, Logrank= 5.910, P = 0.015). Statistical significance was set at P < 0.05.