Primary Motility  Disorders of the  Esophagus
 The Esophageal
 Esophagogastric  Junction

  Browse by Author
  Browse by Movies
Volume: Barrett's Esophagus
Chapter: Adenocarcinomas

What is the clinical impact of micrometastases in esophageal adenocarcinoma?

G.C. O'Sullivan, D. Maguire, J.K. Collins, F. Shanahan (Cork)

Despite optimal treatment, including multimodal and radical surgical therapy, most patients with clinically manifest esophagogastric cancer die from metastatic disease within three years of presentation [1]. Their outcome is predetermined by the presence of widespread, occult metastatic cancer cells that have disseminated prior to treatment and are in many patients chemoresistant [2]. In 90% of patients who are subjected to curative excisional surgery for esophageal cancer, analysis of haemopoietic bone marrow reveals metastatic cell dissemination. This asymptomatic clinical state, after optimal locoregional therapy, is known as minimal residual disease and is represented by either isolated or microaggregates of malignant tumor cells which have the potential to establish overt metastases. This explains the frequent early tumor recurrences that occur following radical resection.

Bone marrow micrometastases

Cells derived from esophageal adenocarcinomas retain expression of epithelial cell-specific proteins (cytokeratins and carcinoembryonic antigen) that are not constitutively expressed in bone marrow (derived from the embryonic mesoderm). The finding of such markers in bone marrow indicates the presence of metastatic cells [3]. These cells are usually referred to as micrometastases as they are not otherwise identifiable by conventional microscopy and radiology. Strategies to identify these micrometastases within marrow include morphological detection by immunohistochemical staining (Figure 1), antibody-based methods by flow cytometry and gene expression techniques by reverse transcriptasepolymerase-chain reaction. These analytic techniques have been validated with false positive identification in less than 2% of cases [3, 4]. The incidence of false negative analysis is unknown but this may be reduced by concurrent examination for a number of markers and by improved marrow sampling (two or more sites). Currently, analysis of marrow from a resected segment of rib obtained at thoracotomy (prior to manipulation of the tumor) is the "gold standard" for detecting micrometastases in esophageal cancer [2].

Biology of micrometastases

The finding of micrometastases in bone marrow of patients with adenocarcinoma of the esophagus is an independent negative prognostic indicator. The viability and proliferative capacity of micrometastases have been confirmed in vitro using combinations of growth factors in the presence of extracellular matrix proteins [3]. In patients with esophageal adenocarcinoma, tumor cell lines may be generated from rib marrow aspirates. These cell lines establish xenografts when transferred to immunodeficient athymic nude mice, confirming tumorogenicity [2].

The finding of bone marrow micrometastases in the preoperative setting does not universally predict poor outcome. Some micrometastatic cells are possibly transitory and are unable to establish at a secondary site or unable to escape destruction by the immune system, or remain in a dormant state. A study of bone marrow from patients with esophageal adenocarcinoma before and six months after "curative" surgery showed that occasional patients were able to clear their marrow of metastatic cells, but persistence of the cells was noted in most and carried a high risk of developing overt metastatic disease within the subsequent months [5]. The majority of these patients with persistent micrometastases developed overt disease on follow-up but a small number of long-term survivors emerged suggesting dormant disease in some. Often in this context recurrent disease is precipitated by systemic events such as infection, surgery, or immunesuppression suggesting loss of biological control by inhibition of immune containment or by stimulation of pro-tumorigenic mechanisms such as angiogenesis or phenotypic aggressiveness. It is likely that the transition from dormancy to progressive disease involves to some degree cellular genetic changes, angiogenic stimulation and immune evasion. Thus, sequential analysis of bone marrow allows a precise diagnosis of dormancy and study of relevant gene expression and determination of the cell cycle and proliferative status of the metastatic cells.

Figure 1. This is an example of tumor cells in a bone marrow aspirate from a patient with esophageal adenocarcinoma. Alkaline phosphatase anti-alkaline phosphatase staining for cytokeratin 18 demonstrates hyperchromatic tumor cells staining positive (red).

Targets for development of effective systemic therapy

Most cells within the primary tumor are terminally differentiated, and do not possess the intrinsic capabilities necessary for dissemination and propagation. Cells that disseminate have the inherent capability to detach from the primary growth, travel through extracellular matrix and basement membrane into the local microvasculature, survive transit in the circulation, and exit by attaching and extravasating through the endothelium at another site [2, 3]. This complex process involves several interdependent steps regulated by cascades of cytokines, chemokines, growth factors and matrix metalloproteinases [3, 6, 7]. The presence of viable micrometastatic deposits in tissues implies dissemination of cancer cells with the facility for independent survival and growth [2]. These are most likely metastatic stem cells and are the appropriate targets for systemic therapy.

Spread of esophagogastric cancer

Traditionally, esophagogastric cancers have been thought to spread sequentially from mucosa to the esophageal wall, to the regional lymph-nodes and lastly systemically. This model has dominated patient care from preoperative staging investigations, to endeavours at disease control. In the absence of overt metastatic disease, by conventional staging investigations, the lymph node status has been used as a "proxy" indicator of systemic spread, but the majority of patients with node-negative cancer also die of metastases. This suggests that haematogenous dissemination may have occurred independent of lymphogenous spread. This has recently been confirmed by the detection of micrometastases in bone marrow samples from resected rib segments of patients with node negative cancers [2]. Metastatic cells in the marrow of these patients were viable, capable of proliferation, grew independently in tissue culture and formed malignant tumors when injected into immunocompromised mice. Thus, both clinical experience and bone marrow analysis suggest that most patients have residual disease after resection of the primary cancer and the regional lymph nodes and that lymph node status alone is not an accurate marker for systemic spread [1, 2].

Implications for neoadjuvant therapy

To date, improvements in survival from multimodal programs have been modest, with most patients succumbing to metastases [8-10]. There is general acceptance that for chemotherapy or chemoradiotherapy to be effective against esophagogastric cancer it must be given in a neoadjuvant setting, before tumor vascularity and patient fitness are compromised by excisional surgery. The choice of chemotherapeutic agents needs to be determined by the responses of both the primary tumor and the sensitivity/resistance of disseminated metastatic cells. In a recent large scale clinical trial of fluoruracil and platinum the clinical response rate of the primary tumor was not reflected in improved survival [11]. That this was due to resistance of minimal residual disease or micrometastases to the chemotherapeutic agents has been confirmed by the finding of viable tumor cells in bone marrow of patients who received similar treatment [2]. The responsiveness to neoadjuvant therapy should be assessed at the level of micrometastases (accessible in bone marrow) rather than by measurements of shrinkage of primary tumors where the terminally differentiated cells may be induced to die whilst the metastatic stem cells are chemo-resistant and remain viable.

Towards the control of minimal residual disease

The control of minimal residual disease is potentially a realisable goal. Repeated access of bone marrow will help define minimal residual disease status, responsiveness to therapy, and provide a source of micrometastases for genetic modification and manufacture of tumor-specific vaccines. Until more specific antitumor drugs are developed, the residual disease burden in most patients can only be controlled by ablation of the primary tumor, and adjuvant biological approaches at general host and metastatic cell level. This will involve the development of means to control the imunosuppressive and angiogenic responses to surgical trauma and infection.

Recent successes in murine models suggest that effective tumor specific vaccines can be developed for even MHC negative cancers without prior antigen isolation [13]. Data from studies utilising dendritic cells that were primed in vitro by fusion with tumor cells or by pulsing with RNA or membrane components have also been encouraging. Genetic modification of tumor cells to secrete GMCSF and to express the B7 co-stimulatory molecule, permit undefined tumor antigen presentation by dendritic cells, and tumor control by specific natural killer cells. The clinical application of these vaccines is now feasible with the culture of metastatic cell lines from bone marrow aspirates.


Conventional methods for assessment of tumor load understage patients. Bone marrow micrometastases are viable cells with proliferative and tumorigenic potential. Persistence of micrometastases following "curative" surgery indicates minimal residual disease that also carries a poor prognosis, whereas clearance of marrow micrometastases is associated with a favourable outcome.

Since the outcome of patients with esophagogastric and other cancers is determined by the fate of disseminated micrometases, it is time to direct attention to the study of these cells. This may offer the best hope for development of rational treatments which can be based on biological control mechanisms of dormancy or clearance. Whether specific immune, anti-angiogenic or chemo/radiotherapeutic strategies are adopted, the study of systemic micrometastatic disease in esophageal cancer will be critical to predict and assess clinical responsiveness.


1. Bytzer P, Christensen PB, Damkier P, Vinding K, Seersholm N. Adenocarcinoma of esophagus and Barrett's esophagus: a population based study. Am J Gastroenterology 1999;94:86-91.

2. O'Sullivan G, Sheahan D, Clarke A, Stuart R, Walsh T, Kelly J, Kiely MD, Collins JK, Shanahan F. Micrometastases in esophagogastric cancer: high detection rate in resected rib segments. Gastroenterology 1999;116:543-548.

3. Maguire D, O'Sullivan GC, Collins JK, Morgan J, Shanahan F. Bone marrow micrometastases and gastrointestinal cancer detection and significance. Am J Gastroenterol 2000;95:1644-1651.

4. Schlimok G, Funke I, Pantel K, Strobel F, Lindemann F, Witte J, Reithmüller G Micrometastatic tumour cells in bone marrow of patients with gastric cancer: methodological aspects of detection and prognostic significance. Eur J Cancer 1991;27(11):1461-1465.

5. O'Sullivan G, Collins JK, Kelly J, Morgan J, Madden M, Shanahan F Micrometastases: marker of metastatic potential or evidence of residual disease Gut 1997;40:512-515.

6. Banks RE, Gearing AJ, Hemmingway IK, Norfolk DR, Perren TJ, Selby PJ. Circulating intercellular adhesion molecule1 (ICAM-1), E-selectin and vascular cell adhesion molecule-1 (VCAM-1) in human malignancies Br J Cancer 1993;68:122-124.

7. Gearing AJ, Hemingway K, Pigott R, Hughes J, Rees AJ, Cashman SJ. Soluble forms of vascular adhesion molecules E-selectin, ICAM-1, and VCAM-1: pathological significance Ann NY Acad Sci 1992;667:324-331.

8. Altorki NK, Skinner DB. Occult cervical nodal metastases in esophageal cancer: preliminary results of three field lymphadenectomy. J Thorac Cardiovasc Surg 1997;113:540-544.

9. Walsh TN, Noonan N, Hollywood D, Kelly A, Keeling N, Hennessy TPJ. A comparison of multimodal therapy with surgery for esophageal adenocarcinoma. N Engl J Med 1996;335:462-467.

10. Shanahan F, O'Sullivan GC. Progress in treating esophageal adenocarcinoma Gastroenterology 1997;112:1417-1418.

11. Kelsen DP, Ginsberg R, Pajak TF, Sheahan DG, Gunderson L, Mortimer J, Estes N, Haller DG, Ajani J, Kocha W, Minsky BD, Roth JA. Chemotherapy followed by surgery compared with surgery alone for localised esophageal cancer. N Engl J Med 1998;339:1979-1984.

12. Braun, S, Pantel, K, Muller, P. et al. Cytokeratin positive cells in the bone marrow and survival of patients with stage I II or III breast cancer. N Engl J Med 2000;342:525-533.

Publication date: August 2003 OESO©2015