Detection of genomically-tagged cancer cells in different tissues at different stages of tumor development: lack of correlation with the formation of metastasis

Abstract

Genetic detection of tumor cells in blood, lymphatic nodes or bone marrow using reverse transcription and polymerase chain reaction (PCR) is quite attractive because it allows the early diagnosis of cancer dissemination. Unfortunately, this type of detection strategy cannot be applied to solid parenchymas, because they usually share with tumor cells the mRNA markers. To avoid this impediment, we have developed an experimental model of cancer using cells with a genome-associated tag. DHD/K12-PROb cancer cells were stably transfected with pcDNA3.1CAT. Approximately 10⁶ transfected cells (DHD-CAT cells) were injected subcutaneously into the chest of BD-IX rats. Animals were divided into 11 groups according to the time between injection of tumor cells and euthanasia. An additional ‘untagged group’ was injected with untransfected cells (DHD-Wild). Blood and tissues samples were collected after euthanasia. Macroscopic and microscopic analysis was done. To detect circulating tumor cells or their presence in peripheral organs, we performed PCR with nested primers to amplify chloramphenicol acetyl transferase-encoding (CAT-encoding) DNA sequences. The minimum number of cells that yielded detectable cells routinely was 2 in 10⁶. No modification of cancer aggressiveness was observed in DHD-CAT cells. DHD-CAT cells were detected by PCR in lung from the 1st week after inoculation, in liver, spleen and kidney from the 3rd week and in the blood from the 5th week. All animals analyzed 12 weeks after injection showed lung metastases. Metastases in liver, spleen or kidney, either microscopic or macroscopic, were never detected. We have developed an experimental model of cancer based on genomic tagging of tumor cells that allows the detection of small numbers of cells in all organs and the blood. The presence of cancer cells in parenchymas detected with molecular technology does not correlate with the development of clinically relevant metastases.

Introduction

The presence of a very small number of viable tumor cells in an otherwise healthy tissue is, from a clinical point of view, a significant finding with important consequences for the evaluation of therapeutic strategies and prognostic of patients [1], [2], [3], [4], [5], [6]. This is a reflection of the widely accepted assumption that the presence of tumor cells in distant parenchymas correlates with the subsequent development of metastases with clinical relevance. For this reason, many studies have focussed on the development of high sensitivity methods for the detection of tumor cells in distant parenchymas during early stages of tumor progression.

Standard cytology has been the most widely used method for the detection of cell dissemination. However, this method can only detect tumor cells with a limited degree of sensitivity (one tumor cell per 100 normal cells), and is notoriously inaccurate. Immunocytochemistry, although more sensitive (detecting one cell per 10⁵ normal cells), is dependent on the availability of specific antibodies to tumor-associated cell surface antigens, and may be prone to false-positive results due to cross-reactivity with antigens present in normal cells.

In the case of circulating cancer cells, techniques based on molecular technology such as polymerase chain reaction (PCR) provide a more sensitive method for detection. Reverse transcription coupled with amplification by polymerase chain reaction (RT-PCR) of relatively small amounts of mRNAs encoding tumor cell proteins allows the detection of as few as one tumor cell per 10⁷ normal cells [1], [2], [5], [6], [7], [8]. In patients with colorectal cancer, amplification of tumor markers including carcinoembryonic antigen (CEA) [5], [6], [7], CD44 [1] and cytokeratin [8] with this technology has been used to detect the presence of tumor cells in the bloodstream. Unfortunately, these mRNAs encode epithelial markers that are not specifically expressed by tumor cells, and therefore this technology cannot be applied to the detection of tumor cells in parenchymatous tissues.

In the present study we describe the development of an experimental model of cancer using a recombinant DNA-based technology that allows the detection of small numbers of cells in apparently healthy organs and blood, even during very early stages of the disease. Using this technology we have evaluated whether there is a correlation between the presence of cancer cells in any target organ and the subsequent appearance of metastases in that organ.