Lucia Suzuki

The incidences of esophageal adenocarcinoma (EAC) and its precursor lesion, Barrett’s esophagus (BE) are rapidly increasing. At an early stage, endoscopic treatment of EAC is possible with a relatively good chance of cure. In contrast, an advanced stage requires very invasive treatment, often includes neo-adjuvant chemoradiotherapy (nCRT) and holds a poor prognosis. To further improve risk stratification in these patients new and previously described biomarkers were evaluated in this thesis (Figure 1).

MicroRNA (miR) are small non-coding RNAs able to control translation of messenger RNA transcripts of protein-coding genes. This way, they can either function as oncogenes or tumor suppressors. High-throughput miR-profiling can differentiate between tissue types. Distinctive miR signatures are recognized for specific cancer types, both on tissue as well as body fluids. Consequently, miR are increasingly studied as potential cancer biomarkers, both regarding diagnosis as well as prognosis and prediction. Based on our results described in Chapter 3, miR profiles from endoscopic biopsy samples of patients with non-dysplastic BE cannot be used to predict progression to high-grade dysplasia and/or EAC.

In patients with early, submucosal (pT1b) EAC, determining if the risk of metastases outweighs the risk of mortality and morbidity from surgical removal of the entire esophagus and part of the stomach is difficult. Frequently used risk indicators include lymphovascular invasion, tumor differentiation grade and depth of invasion. However, the estimated risk of metastases based on these indicators varies widely in the literature published. Chapters 4 and 5 describe high tumor budding according to the Ohike method and low OLFM4 immunohistochemical expression are independent risk factors for (nodal) metastases. All these factors combined would likely lead to a better risk stratification for this group of patients. However, because of the limited studies related to these topics in EAC so far, additional research is needed to confirm these results.

Advanced EAC carries a poor prognosis, although reported survival rates vary greatly depending on the cohort studied. The TP53 gene and its encoded protein p53, reported to be, at least one of, the guardian(s) of the genome, can be assumed to be the driving force of cancers of various types, including EAC. In spite of this knowledge, still a number of questions need to be answered. Aberrant p53 immunohistochemical staining pattern is an independent risk factor for poor prognosis in a well-defined cohort of EAC patients who were treated with surgery only (Chapter 6). Moreover, it correlates well with TP53 gene mutational status. Yet, neo-adjuvant chemoradiotherapy (nCRT) is currently the gold standard in the treatment of patients with advanced EAC. Moreover, it is currently being addressed if clinically complete responders after nCRT can be kept under close surveillance and only require surgery if evident localized tumor recurs. Therefore, treatment of patients with EAC is increasingly based on small biopsy samples and in a subset of patients, i.e. patients with complete (clinical) response, a surgical tumor sample may even not be available anymore.

To date, the effect of nCRT on biomarker expression such as p53 and SOX2 in EAC and their prognostic value pre- and post-treatment are unknown. This is of relevance in the context of evaluation of these biomarkers in relation with clinical behavior. Chapter 7 describes aberrant p53 and SOX2 IHC are largely concordant before and after nCRT. Nevertheless, results from biomarker studies in nCRT naïve EAC patients cannot simply be extended to patients treated with nCRT. This is most likely due to progression related events during the course of the disease and exposure to nCRT.

In conclusion, optimal treatment strategy in EAC remains a challenge and it is clear EAC pathogenesis is driven by multiple complex, and often interrelated, pathways. Although external, prospective validation is desirable for most results, this thesis further emphasizes that analysis of one single biomarker is unlikely to precisely predict progression, presence of metastases, or recurrence in an individual patient. Hence, it is more likely that a combination of several risk factors, and as such also biomarkers, provides a better estimate. This thesis describes some additional promising markers that could be used for this purpose.

Figure 1. Outline and conclusions of this thesis. GERD, gastro-esophageal reflux disease; NDBE, non-dysplastic Barrett’s esophagus; LGD, low-grade dysplasia; HGD, high-grade dysplasia; miRNA, microRNA; TB, tumor budding; H&E, hematoxylin & eosin; IHC, immunohistochemistry; LVI, lymphovascular invasion; pT1b, limited to the submucosa; nCRT, neo-adjuvant chemoradiotherapy.