Chromosomal imbalances are uncommon in chagasic megaesophagus

Chagas' disease is a human tropical parasitic disease which occurs in the Americas, particularly in South America. It affects 16 to 18 million people in tropical and subtropical countries of Latin America [1]; in Brazil, the number of cases has reached 6 million [2]. During the chronic phase, 6 to 7% of chagasic patients develop mega syndromes represented by muscular hypertrophy and dilation of the esophagus or colon, in consequence of destruction of the myoenteric and submucous plexus by the protozoan Trypanosoma cruzi[3]. The digestive forms, megaesophagus or megacolon may be observed in advanced stages of the disease [3].

Megaesophagus is consequence of achalasia characterized by the destruction or lack of intramural nerve plexus, which determines the absence of peristalsis and lack of openness of the lower esophageal sphincter in response to swallowing. In consequence, food retention or esophageal stasis occurs, leading to the appearance of chronic esophagitis, acanthosis, paraceratose and leukoplakia, possibly pre-cancerous lesions [3]. One of the severe late consequences of chagasic megaesophagus is the increased risk (3% to 8%) of developing esophageal squamous cell carcinoma (ESCC) compared to when megaesophagus is not present [4–6]. Also, ESCC develops in chagasic megaesophagus patients at a younger age than in those without this disease [5]. The detection of cancer in these patients is difficult because the symptoms are hidden by the severe dysphagia caused by megaesophagus [7]. The diagnosis is frequently late, when the patient is in advanced stage, resulting in a poor prognosis [5].

ESCC has been reported as the ninth most common malignancy and ranks the sixth most frequent cause of death worldwide, but its incidence varies largely among regions [8]. Approximately 16,470 new cases of esophageal carcinoma were expected in the US population in 2009 [9]. The Brazilian National Institute of Research in Cancer (INCA) reported that esophageal cancer was the sixth in the cancer rank mortality in 2000 with 5,307 deaths. The estimated incidence in 2008 was 10550 new cases with an incidence per 100 thousand individuals varying among geographical areas from 1.04 to 19.07 in males and 0.39 to 7.58 in females [8].

In ESCC, molecular cytogenetic techniques have shown common occurrence of unbalanced genomic regions involved in amplification of oncogenes and deletion of tumor suppressor genes [10, 11]. However, studies in benign esophageal lesions with precancerous potential as megaesophagus are scarce. This study used FISH technique to investigate chagasic megaesophagus for genomic status of genes frequently unbalanced in ESCC with the goal of identifying potential markers of risk to cancer. The selected targets included FHIT, TP63, PIK3CA, EGFR, FGFR1, MYC, CDKN2A, YES1 and NCOA3 genes, all of which have been reported as occurring in significantly abnormal numbers in ESCC [12–19].

Data were also subject to an exploratory analysis using three-dimensional plots of the percentages of cells per genomic pattern (loss, balanced and gain) for individual cases as illustrated in Figure 1. Several cases were identified harboring deletions on specifics targets: CM7 (YES1), CM12 (CDKN2A), CM13 (PIK3CA; TP63), CM19 (PIK3CA), CM30 (EGFR), CM36 (FHIT; CEP9), CM16 and CM34 (CEP9). Other cases showed genomic gain for specific targets: CM3 (MYC), CM6 (TP63; MYC), CM7 (TP63), CM11 (PIK3CA; CDKN2A), CM16 (CDKN2A), CM28 (PIK3CA), CM30 (MYC; FGFR1), CM36 (NCOA3), and CM11 and CM21 (CEP9). Figure 2 illustrates unbalanced status of genes and centromeric DNA targets in megaesophagus specimens.

No significant association was detected in CM group between mean copy number per cell of each DNA target tested and parameters such as age, gender and the life style factors tobacco smoking and alcoholism although the small size of each subset impaired robust conclusion. Similarly, no significant differences were observed on mean copy number per cell of each genomic target according the dilation grades of megaesophagus (data not shown). However, the CM cases identified with genomics deletions and gains in the three-dimensional plots were mostly classified as grades III and IV, suggesting that chromosomal imbalances are more likely to occur in more advanced grades.

Esophageal squamous cell carcinoma is one of the most prevalent cancers worldwide and has multifactorial origin, in which environment factors, especially alcohol ingestion and smoking, play significant role [10, 24]. Several genetic alterations are also associated with ESCC, such as chromosomal changes, allelic deletions, activation of oncogenes and inactivation of tumor suppressor genes [10, 11]. In tropical and subtropical countries of Latin America such as Brazil, the risk of ESCC may also be related to megaesophagus due to Chagas' disease. The period between onset of symptoms related to achalasia and detection of cancer ranges from 17 to 28 years [4]. In the current study no patient developed esophageal carcinoma, which can also be due to insufficient follow up.

This study showed that genomic copy number changes were not common events in megaesophagus. Among the tested 12 DNA targets mapped to 6 distinct chromosomes, there was only marginal differences between specimens from chagasic megaesophagus and normal esophageal mucosa. The only statistically significant difference in copy number between groups involved CDKN2A. Loss of function of CDKN2A (also called MTS-1 or P16) prevents the blocking of G1 phase and appears to be a necessary event for the progression of pre-cancerous cells to malignancy [25, 26]. Genetic and epigenetic alterations in CDKN2A have been reported in early stages of esophageal carcinogenesis [27–30]. In primary esophageal carcinomas and cell lines, deletions and point mutations in CDKN2A gene have been reported in 16% to 82% of cases [22, 31], and hypermethylation of promoter region in 20% to 88% of ESCC and precancerous lesions [30, 32]. The current study detected deletion of CDKN2A and CEP 9 sequences in CM compared with NM group. In agreement with these findings, patients with idiopathic achalasia and chagasic megaesophagus, with or without esophageal carcinoma, showed reductions in expression of p16 protein [7] and Bellini et al. [33] observed a marginal decrease in p16 protein expression in chagasic megaesophagus. Additionally, recent subset analyses by our group did not find mutations in CDKN2A (exons 1 and 2) and FHIT (exons 5 and 7) genes, suggesting these events are uncommon in CM [34], and have detected copy number changes of chromosomes 7, 11 and 17 and TP53 deletion by FISH [35].

The tumor suppressor gene FHIT (3p14.2) is deregulated during the development of ESCC. Deletion in both FHIT alleles results in failure in the transcript [36] and consequent absence or reduction of Fhit protein, which act in the cell cycle regulation and apoptosis. Genetic and epigenetic alterations of FHIT are associated with development of several cancer types [37, 38]. Loss of heterozygosity (LOH) in areas of the 3p14-p21 bands has also been reported in low-grade dysplasia and postulated as early events in esophageal carcinogenesis [39]. While hypermethylation of FHIT gene has been described in 33% to 69.4% of ESCC cases, it is also known that deletion and loss of protein expression are frequent in esophageal carcinoma [32, 36]. Deletion of FHIT observed in a single megaesophagus specimen in the present study may associate with early onset of genetic chances in these lesions, but it is worth to note that no significant decrease in the level of Fhit protein expression have been detected in chagasic megaesophagus [33].

All other tested genes are categorized as oncogenes and have been previously shown to associate with ESCC, in which they are commonly amplified [12–19]. However, abnormal copy number patterns were only seen in few individuals in the Chagasic megasophagus cohort and involved both gains and losses. One example is PIK3CA, which encodes the catalytic subunit that uses ATP to phosphorylate phosphatidylinositol, a gene that is frequently amplified in ESCC [40–42]. Yen et al. [40] determined in FISH assays that PIK3CA was amplified in cell lines and primary tumors and detected a positive correlation between amplification and tumor size, lymph node metastasis and clinical stage. Yang et al. [41] showed that PIK3CA gene amplification was highly correlated with protein overexpression, supporting amplification as a major mechanism for overexpression. Nevertheless, in this study PIK3CA was lost rather than amplified in the CM specimens as a group while involved in marginal level of loss two specimens and gained in two other specimens.

Gene amplification was one of the basic mechanisms leading to overexpression of the TP63 in ESCC [18, 43, 44]. The amplification of TP63, whose protein plays in the development and maintenance of stratified epithelial tissues has been described in early stage of ESCC carcinogenesis but down-regulated in advanced [18, 43]. Amplification of the membrane receptor genes EGFR and FGFR1 has been reported in various cancers including ESCC [45, 46], and this phenomenon has been suggested to be a poor prognostic factor in solid tumors [47]. In ESCC, co-expression of both aFGF and FGFR1 was associated with larger tumor area and worse prognosis which suggests that the membrane receptor may promote proliferation of esophageal cancer cells in an angiogenesis-independent and autocrine manner and may contribute to rapid recurrence after esophageal resection [47].

MYC is a transcription factor that binds E-boxes as a heterodimer with Max in about 15% of all genes, and recruits co-activators to regulate gene transcription. MYC is regulated in part through mitogenic stimuli and is activated constitutively in cancer cells through gene amplification, chromosomal translocation, point mutation and mitogenic stimulation [48]. MYC was found to be amplified in ESCC cell lines and in primary tumors [14]. YES1 codes a protein with tyrosine kinase activity that has been previously associated with esophageal carcinogenesis [15, 42]. Similarly, the NCOA3 gene encoding a nuclear receptor coactivator that interacts with nuclear hormone receptors to enhance their transcriptional activator function was found frequently overexpressed and also amplified in 5 to 15% of ESCCs [49–51].

Although is well established that megaesophagus preferentially affects males between the second and fourth decades of life [3], this study has not detected relationship between copy numbers of each tested gene and parameters such as age, gender, life style factors (tobacco smoking and alcoholism) and megaesophagus grade in the CM group. Nevertheless these findings may have been impacted by the limited number of samples in each category.

Interestingly, specimens with genomics deletions and gains exhibited mostly more advanced dilation grades. Megaesophagus patients have high variety of esophageal microbiota, which consists mainly of Gram-positive anaerobic bacteria in concentrations that correlated with the degree of esophageal dilation [6, 52]. Bacteria in fluid stasis can undergo reduction from nitrates to nitrites with the production of N-nitrous compounds, which has potent action in esophageal carcinogenesis [52]. High grade megaesophagus present higher dilation and esophageal stasis associated with bacterial proliferation, thus the data suggest that the level of dilation and inflammation of megaesophagus could promote genetic damages and could be related to increased risk of tumor development [4, 7, 53].

In conclusion, among 40 chagasic megaesophagus comprehensively investigated with a panel of FISH probes, copy number gain was only displayed by three specimens for MYC, by one specimen for FGFR1 and one specimen for NCOA3. Loss of FHIT and YES1 was seen in one specimen and of EGFR in another. For the remainder genes CDKN2A, PIK3CA and TP63, from one to four specimens each showed copy number gain or loss. In every case the difference was only marginal therefore not providing strong support to a biological impact. These findings show that imbalances involving the genomic regions encompassing gene sequences relevant in esophageal carcinogenesis were not found as significant events in chagasic megaesophagus. Consequently, genomic imbalances are not promising markers for assessment of ESCC risk in chagasic megaesophagus.