ARIOSTO SILVA / MOFFITT CANCER CENTERTumor environment: healthy cells (in green) protect tumor cells (in red) ARIOSTO SILVA / MOFFITT CANCER CENTER
At one of the major U.S. cancer research centers, the Moffitt Cancer Center in Florida, a Brazilian computer engineer is making progress on a test that uses computer imaging to evaluate the sensitivity of tumor cells to cancer drugs before treatment begins. If the work goes as hoped, this approach could go beyond the tests now available, help select the most effective chemotherapeutic agents, and predict if and when the cancer may return.
“Today, tests similar to ours indicate only whether tumor cells are sensitive or resistant to drugs,” says Ariosto Silva, the man responsible for the design and development of the test, whose preliminary results were presented in an article in the January issue of the journal Cancer Research. “The new test, if it works as expected, will represent an important step towards personalized cancer treatments.”
The test is still undergoing adjustments. On January 27, 2014, Silva’s laboratory received a small glass tube with cells of a type of blood cancer called multiple myeloma, most commonly found in people over age 60. It was a sample of tumor cells extracted from bone marrow, which he intended to use in a test that he has been working on for three years. The next day, the myeloma cells were placed in the compartments of a plastic block mounted on a glass slide. There the diseased cells came into contact with healthy cells from the same patient, immersed in a mixture of nutrients from the blood: this combination tries to reproduce in the laboratory the environment that tumor cells encounter inside bones, where they proliferate.
Then Silva and his team added different dosages of six myeloma drugs to each compartment, as well as combinations of these antitumor compounds. For the next five days, they monitored the rate of cell death through photographs taken every five minutes with the aid of a microscope. Silva and his team have written a computer program that identifies and automatically counts the number of dead and surviving cells and, based on this information, calculates the risk of the cancer recurring after treatment with the most appropriate medication.
Friendly neighborhood
In the initial experiments the researchers used the tumor and healthy cells of seven people with myeloma to evaluate the performance of two compounds (bortezomib and melphalan), which are used in an attempt to reduce the progression of this cancer, for which there is still no cure. They found that when the myeloma cells were in the company of healthy tissue cells (stroma), which provide physical support and nutrients, they were more resistant to the drugs. When in contact with the stroma, more than half of the tumor cells were still alive a day after being treated with melphalan. However when isolated, almost all of them died in just over 12 hours, depending on the concentration of the chemotherapy. The effect of melphalan was restored when the researchers added bortezomib to the medium containing healthy and tumor cells.
Ariosto silva / moffitt cancer center“By reconstructing the tumor environment, we could make a more realistic assessment of the response to treatment,” says Silva. As a matter of fact, plasma cells, the precursors of myeloma cells, are not alone inside the bones. They live next to stromal cells with which they constantly are exchanging chemical information. “The marrow environment protects tumor cells,” he says. “During treatment myeloma cells migrate within the bone marrow to increasingly protected regions.”
With a degree in computer engineering from the Technological Institute of Aeronautics (ITA), Silva specialized in molecular biology for his doctorate at the University of Campinas (Unicamp) and at the Centro Infantil Boldrini, which treats children with cancer. In 2008 he and his family moved to Tampa so he could work at the Moffitt Center with the mathematical oncologist Dr. Robert Gatenby. The author of his own view of the evolution of tumor cells, Dr. Gatenby, like other leaders of the center, sought a comprehensive and multidisciplinary approach, using mathematicians, physicists and computer scientists, working with biologists and oncologists in order to understand cancer as a complex and dynamic system, rather than looking only for diseased cells, which are now seen as only one component of this environment (see Pesquisa FAPESP Issue No. 162, August 2009).
Silva and his team intend to use the new test to study how the tumor environment and the location of the cancer evolve during treatment. Their strategy does not use compounds that damage tumor and healthy cells—a computer program automatically takes pictures and later analyzes them; living cells are recognized, because they pulsate as if they were breathing. Thus, it is easier to identify the effect of chemotherapy and monitor the cells for up to five days, a time period that is considered sufficient. Commercial tests use compounds to mark cells, and these compounds can also damage and kill cells, making it difficult to know whether cellular death came from the treatment or the measurement technique used in the test.
Risk of recurrence
Since the new test allows cell-by-cell observation, Silva and his team were able to determine cell sensitivity to different drug concentrations and measure the size of the resistant population over time. According to Silva, this makes it possible to reconstruct the trajectory of tumor shrinkage and estimate when tumors may recur. “We are trying to customize both the treatment and the monitoring of the tumors as they evolve,” he says.
This strategy has been used experimentally at the Moffitt Cancer Center to determine if what the researchers observed in the laboratory also occurs in clinical practice. The preliminary results look favorable. “We followed some patients for a year and found a correlation between sensitivity and resistance, which we observed in in-vitro tests and in in-vivo measurements; we now need to validate the testing method in 40 or 50 more people,” says Silva, who plans to use the test on other blood cancers, such as lymphoma and leukemia, in collaboration with researchers from Brazil. Silva believes it will not be difficult to validate the method if he is able to work in collaboration with other centers. Furthermore, he adds, the computer programs used in the tests are free to use and, therefore, any molecular biology laboratory could repeat the experiments.
“Due to the nature of the test, many therapeutic regimens could be tested in vitro before exposing patients to treatments that may prove less effective,” says Roger Chammas, a professor at the University of São Paulo School of Medicine and coordinator of the translational research laboratory at the São Paulo State Cancer Institute, which collaborates with Silva and Dr. Gatenby. “The way it’s been set up, the model seems most appropriate for hematological malignancies such as myeloma. However, the model will accommodate future experimental variables that simulate the conditions found in solid tumors. As these details are ironed out, strategies like these should be adopted by clinical practice. In my view, this work represents a very interesting step in that direction,” says Dr. Chammas.
Dr. Gatenby is optimistic. “I think this strategy can quickly go to larger clinical trials because Silva did a really good job working with clinical oncologists, since he always remained focused on the medical applications,” he told Pesquisa FAPESP. “I’m sure this kind of approach is not just necessary, but the only way we will be able to really make progress on cancer control.”
Scientific article
KHIN, Z.P. et al. A preclinical assay for chemosensitivity in multiple myeloma. Cancer Research. V. 74, No. 1, January 1, 2014.
