Pictures of the Future      Fall 2008

How to Fingerprint a Tumor

Researchers are closing in on a diagnostic test that will be able to predict whether a patient with breast cancer can be successfully treated without chemotherapy. Automated analysis of tumor-specific genes is the key to a new world of individually-tailored treatment.

Here’s a distressing figure: Over 150,000 women are treated for breast cancer in Germany each year. However, for those whose disease is detected and treated early, the chances of recovery are good. To ensure that no cancerous cells remain in the body after surgery, the majority of patients undergo chemo­therapy. The severe side effects of this treatment are tolerated better by some women than others. However, medical statistics show that around 70 % do not even need it. "This fact has been confirmed by numerous studies that date mostly from the 1970s and 1980s, in other words, from times when chemo­therapy as we now know it was still in its infancy," says Dr. Christoph Petry, Head of Molecular Research at Siemens Healthcare Diagnostics in Cologne. "In those days, patients were simply sent home after cancer surgery. But over the years, it was shown that around 70 out of 100 patients who had undergone surgery did not develop any recurrence of their cancer. They were permanently cured."

At the time, nobody knew who would belong to the 70 % of those cured and who wouldn’t. It appeared to be a question of fate. Even today, we are still a long way off from correctly identifying all the women in the "70 % low-risk group" who would benefit from  less aggressive, yet equally effective therapy. This is why Petry and 35 of his coworkers from the Cologne research laboratory have been working since the start of 2007 on a new type of breast cancer prognosis test. The test identifies a significant proportion of low-risk patients—and in so doing holds the potential of sparing them from having to undergo chemo­therapy in the future.

Increasing Accuracy. A test of this kind must classify at-risk patients with superlative accuracy. Otherwise, the risk of dying from a tumor would increase dramatically for those patients who had been incorrectly-classified. On the other hand, the test’s ability to correctly assign patients who will later not develop any metastases to the "70 % low-risk group"—a measure known as specificity—need not be as high. This property of the test, however, determines how many patients it can actually be applied to.

In reality, it is already possible to identify patients who are at particularly high or low risk. To do so, a pathologist examines sections of tumor tissue under a microscope. "Based on the cell structure and the properties of the cells, which can be determined by special staining reagents, a pathologist can give the attending physician important information as regards whether the tumor in question is a less aggressive one that will probably not spread into the rest of the body," says Petry.

Current methods, however, experience problems combining sensitivity and specificity at a high level. "But with our breast cancer prognosis test, we want to be able to give pathologists a method that has a previously unattained combination of sensitivity and specificity—in other words, superlative accuracy," says Petry, who holds a doctorate in chemistry. "We’ll probably never achieve 100 % accuracy. However, as appears to be the case, if we are able to say to many of those affected with 90 % certainty ‘You don’t need chemotherapy’, then that’s fantastic. The risk of suffering a recurrence of the disease is then not really any higher for a woman who has received such a diagnosis than that of any healthy woman developing breast cancer at some stage of her life."

The new diagnostic test is expected to be used in concert with current conventional analysis methods. In other words, after breast biopsy or surgery, tumor tissue is sent as usual to the pathologist’s laboratory, where it is placed in formaldehyde for a few hours to preserve it. The pathologist then embeds the pieces of tissue in paraffin and cuts slices from them, each five to 10 µm thick, for investigation under a microscope.

Siemens’ new prognostic test builds on this analytical material. However, rather than analyzing tissue-based cancer markers by means of differing intensities of color reactions in the tissue, it uses—and this is the test’s key feature—an in vitro analysis of the nucleic acids (the carriers of genetic information) found in the tumor tissue.

Magnetic Attraction. This method uses an automatic nucleic acid extraction system developed by Siemens Healthcare and unique worldwide. The system runs on the Versant kPCR analytical system from Siemens. kPCR stands for kinetic PCR, a technique used to replicate and simultaneously quantitively determine nucleic acids using the PCR method (PCR: polymerase chain reaction).

The Versant kPCR, which is fully automated and is expected to enter clinical service at the end of the year, extracts nucleic acids using silicic acid-coated magnetic particles. Nucleic acids bind to the silicic acid and thus to the magnetic particles via the formation of hydrogen bonds.

"Conventional methods result in unevenly-sized particles that contain iron oxide and silicic acid in greater or lesser ratios. However, we have succeeded in optimizing the manufacture of these magnetic particles in such a way that they are evenly-sized and homogeneously-magnetic and are characterized by constant and reproducible chemical and physical properties. The particles are made of an iron oxide core, measuring around 100 nm, and an even, ultra-thin silicic acid shell," explains Petry. "This has the advantage that the particles are easily and highly dispersible, and can also be magnetized very effectively."

In the near future, day-to-day laboratory molecular diagnostics routines may look as follows: A lab technician takes a wafer-thin section of tissue prepared from a tumor sample and inserts it into a small tube. The tube is placed—possibly with a few dozen other tubes containing tumor samples from different patients—into the sample tray of the new Versant kPCR Molecular System. With the click of a mouse, the assistant can select the required test and start the system. From this point onwards, everything runs automatically. "A pipetting robot first adds a solution to the tissue samples. The solution dissolves the samples once they are heated. The nucleic acids in the samples are released in this process and, after magnetic particles have been added on a so-called deep-well plate, are absorbed by the particles," explains Petry.

The robotic system then places the deep-well plate on a magnetic block, which draws the magnetic particles and the nucleic acids downward, whereupon the remaining liquid, which mostly contains protein residues, is aspirated off. Once the magnetic field is switched off, the robotic system washes the particles by briefly shaking the plate—and aspirates off the "dirty" solution again when the magnetic field is reactivated. The cleaned magnetic particle/nucleic acid combination is subsequently separated and the separated nucleic acids are then available for the actual analysis.

Fully-Automated Replication. Now comes the replication of the nucleic acids using PCR. In the first stage, the messenger ribonucleic acid (mRNA) is transcribed into deoxyribonucleic acid (DNA). The DNA is then replicated in 40 PCR cycles and its quantity precisely determined. "With our prognostic test, we look at the volume ratios of 12 tumor nucleic acids, thereby learning which of the 12 tumor-specific genes are expressed more, and which less," says Petry. "These 12 values then provide a type of genetic fingerprint of the tumor, with which we are able to uniquely characterize the type of cancer." If the cancer is a type with a low propensity for recurrence or the formation of metastases, then pathologists and oncologists can recommend that the patient not be required to undergo chemotherapy.

"We are still in the early stages of development for this project," cautions Petry. "It would therefore be premature to predict when we’ll be able to offer this test." He is, however, convinced of one thing: "It’s not science fiction. We’ve com­pleted the research stage. These tests will come, and they will make tumor therapy more effective and less harmful for patients."
                                                                                                          Ulrike Zechbauer