ERCC1 single nucleotide polymorphism C8092A, but not its expression is associated with survival of esophageal squamous cell carcinoma patients from Fujian province, China

Abstract

Esophageal carcinoma is one of the world's deadliest cancers. Esophageal squamous cell carcinoma (ESCC) is more frequent than adenocarcenoma (AC) in China. Platinum-based chemotherapy with surgical resection is a common treatment approach for ESCC; however, the treatment response is uncertain. Evidence suggests polymorphisms in genes encoding excision repair cross-complementing group 1 (ERCC1), a protein involved in nuclear excision repair (NER), may help predict response to cisplatin and other platinum-based chemotherapeutics. Multiple ERCC1 single nucleotide polymorphisms (SNPs) have been associated with platinum chemotherapy response. Two common SNPs occur at the C8092A and C118T loci. Our study aimed to determine if 1) an association exists between ERCC1 tumor expression and patient survival, 2) whether adjuvant therapy influence on survival is related to histological ERCC1 presence in tumor cell nuclei, and 3) whether other clinicopathological characteristics in a cohort of patients following surgery for various stages of ESCC are associated with tumor ERCC1 expression. One hundred eight patients were included in the study, and tumor biopsy was collected for genotyping and immunohistochemical analysis of ERCC1. Sixty-seven patients (62%) received no adjuvant therapy, and the rest had either platinum-based chemotherapy (28.5%), radiotherapy (6.5%) or both treatments (2.8%). Log-rank analysis revealed no significant connection between tumor ERCC1 expression (P = 0.12) or adjuvant therapy (P = 0.56) on patient survival. Also, non-parametric Mann-Whitney analysis showed no significant link between tumor size or nodus tumor formation and ERCC1 presence in patients in the study. Interestingly, C8092A SNP showed significant association with patient survival (P = 0.01), with patients homozygous for the mutant allele showing the most significantly reduced survival (P = 0.04) compared to those homozygous for the dominant allele (CC). Our results provide novel insight into the genotypic variation of patients from Quanzhou, Fujian province China.

Figure 1. Representative immunohistochemical staining for ERCC1 in esophageal squamous cell carcinoma.

Such staining was used for classification of nuclear staining intensity and assignment of a corresponding H-score used in further analyses in the study. A) ERCC1-negative esophageal squamous cell carcinoma tissue (H-score  = 0); B) High nuclear expression of ERCC1 in esophageal tumor cells with labeling intensity scored as 3 (H-score  = 3) (400×).

Figure 2. Genotype analysis for ERCC1 SNP C8092A.

A) Post-digestion PCR of alleles CC, CA, and AA (left to right). B) Sequence analysis of CC genotype. C) Sequence analysis of CA heterozygous genotype. D) Sequence analysis for the AA genotype.

Figure 3. Genotype analysis for ERCC1 SNP C118T.

A) Post-digestion PCR of alleles CT, TT, and CC (left to right). B) Heterozygous sequence analysis of CT genotype. C) Sequence analysis of TT genotype. D) Sequence analysis of CC genotype.

Figure 4. Adjuvant treatment did not significantly influence patient survival probability.

Patient survival probability was not significantly influenced by platinum-based chemotherapy or radiation over no adjuvant therapy. Log-rank (Mantel-Cox) Test; Chi-square  = 2.0; p = 0.56.

Figure 5. Patient survival was not significantly associated with ERCC1 nuclear expression H-score.

There was no significant relationship between high or low intensity ERCC1 nuclear immunohistochemical labeling designated by H-score cutoff of 2.0 and patient survival. Chi- square  = 0.27; p = 0.61. Median patient survival (months: ERCC1+, 23.8; ERCC1-, 20.3, 95% CI of ratio: 0.8341 to 4.227.

Figure 6. Association of ERCC1 polymorphism allelic variation on patient survival.

A) ERCC1 polymorphism C8092A was significantly associated with overall patient survival (p = 0.01). Presence of the mutant allele, especially for the homozygous patients for this allele (AA), was associated with decreased overall and progression-free survival (PFS) compared to homozygous wild-type patients (CC) (p = 0.007 & 0.04, respectively). No significant relationship between polymorphism genotype and C) overall patient survival for C118T (p = 0.78) or D) PFS was observed (P = 0.93).