Overexpression of CDC25B and LAMC2 mRNA and protein in esophageal squamous cell carcinomas and premalignant lesions in subjects from a high-risk population in China

Abstract

Molecular events associated with the initiation and progression of esophageal squamous cell carcinoma (ESCC) remain poorly understood but likely hold the key to effective early detection approaches for this almost invariably fatal cancer. CDC25B and LAMC2 are two promising early detection candidates emerging from new molecular studies of ESCC. To further elucidate the role of these two genes in esophageal carcinogenesis, we did a series of studies to (a) confirm RNA overexpression, (b) establish the prevalence of protein overexpression, (c) relate protein overexpression to survival, and (d) explore their potential as early detection biomarkers. Results of these studies indicated that CDC25B mRNA was overexpressed (>/=2-fold overexpression in tumor compared with normal) in 64% of the 73 ESCC cases evaluated, whereas LAMC2 mRNA was overexpressed in 89% of cases. CDC25B protein expression was categorized as positive in 59% (144 of 243) of ESCC cases on a tumor tissue microarray, and nonnegative LAMC2 patterns of protein expression were observed in 82% (225 of 275) of cases. Multivariate-adjusted proportional hazard regression models showed no association between CDC25B protein expression score and risk of death [hazard ratio (HR) for each unit increase in expression score, 1.00; P = 0.90]; however, several of the LAMC2 protein expression patterns strongly predicted survival. Using the cytoplasmic pattern as the reference (the pattern with the lowest mortality), cases with a diffuse pattern had a 254% increased risk of death (HR, 3.52; P = 0.007), cases with no LAMC2 expression had a 169% increased risk of death (HR, 2.69; P = 0.009), and cases with a peripheral pattern had a 130% greater risk of death (HR, 2.30; P = 0.02). CDC25B protein expression scores in subjects with esophageal biopsies diagnosed as normal (n = 35), dysplastic (n = 23), or ESCC (n = 32) increased significantly with morphologic progression. For LAMC2, all normal and dysplastic patients had a continuous pattern of protein expression, whereas all ESCCs showed alternative, noncontinuous patterns. This series of studies showed that both CDC25B and LAMC2 overexpress RNA and protein in a significant majority of ESCC cases. The strong relation of LAMC2 pattern of protein expression to survival suggests a role in prognosis, whereas the association of CDC25B with morphologic progression indicates a potential role as an early detection marker.

Fig. 1

A low power image of an H&E of the esophageal tumor tissue microarray utilized in this study. The individual tissue cores are 0.6 mm in diameter and grouped in subarrays of 5×5. Not all subarrays were complete. Because of the small size of the tissue specimens, and nature of esophageal tumors, H&E’s of section 1 and 50 of the array were reviewed and the diagnosis of each core was assigned based on this review.

Fig. 2

Representative images of immunohistochemistry for CDC25B. All images are 450X. Representative tumors of: (a) negative, (b) weak, (c) moderate, and (d) strong staining for CDC25B.

Fig. 3

Representative images of immunohistochemistry for LAMC2. All images are 450X. Representative samples of expression for: (a) continuous staining pattern, (b) cytoplasmic staining pattern, (c) diffuse staining pattern, (d) peripheral staining pattern, (e) mixed staining pattern, and (f ) negative staining for LAMC2.

Fig. 4

Kaplan-Meier survival curves (times) for ESCC cases by their LAMC2 protein expression staining patterns: cytoplasmic, diffuse, peripheral, mixed, and negative.

Fig. 5

Design and construction of a biopsy tissue microarray. (a) A schematic representation of the construction of the biopsy tissue microarray. Blocks of previously unsectioned esophageal biopsies were assigned a diagnostic group by initial pathologic diagnosis based on all available biopsies obtained at endoscopy. These donor blocks were then cored with 2.00 mm needles to extact the biopsy in toto for placement in a tissue microarray recipient block. The tissue microarray block was sectioned and H&Es were performed on the 1^st, 25^th and 50^th section. Specific immunohistochemical stains were performed in duplicate on different levels of the array. This approach of review of multiple H&Es and the use of immunohistochemistry on multiple sections mirrors the approach used in clinical practice to ensure the lesion is seen in its entirety and that the immunohistochemistry is performed on the correct histopathology. In calculating the results, only the severest histopathology was considered. (b) A low power image of an H&E of one of the biopsy arrays. The array is constructed with 2.00 mm cores. Because the biopsy specimens are small and embedded at different angles, they do not all appear on a single array at one time, nor do they appear as round cores of tissue as seen in conventional tumor tissue microarrays.