The paradox was disturbing. Two tumors derived from connective tissues of the brain had diametrically opposed degrees of aggressiveness – one was almost benign, with a slow and restricted growth, and the other was very invasive, expanding quickly and widely. But the pattern of expression (activity) of their genes, instead of being very different, as one would think initially, was extremely alike.
This fact could be an indication that the capacity for spreading through healthy tissues, so marked in the more malignant forms of cancer, was linked to the functioning of a reduced number of genes. To test this hypothesis, researchers from the Clinical Cancer Genome project, funded by FAPESP and the Ludwig Cancer Research Institute, tested roughly 20,000 human genes, and two thirds of the total of the species reacted when coming into contact with the two kinds of tumor.
The behavior of this set of genes, which was found deposited on a special glass slide, technically called a microarray, was monitored in nine different comparisons between the two kinds of tumor.The results of the experiment – carried out at a laboratory of the Albert Einstein Teaching and Research Institute, linked to the Albert Einstein Hospital, in São Paulo, which recently associated itself to the Clinical Genome project, relying on its own resources – reinforced the scientists’ original suspicion.
Only 110 genes worked in a significantly different way in the two tumors: 45 were more expressed in grade 1 astrocytomas, the mildest kind of cancer, and 65 in the glioblastomas (grade 4 astrocytomas), the most devastating form of the tumor. Oddly enough, 27% of the genes most active in the more serious tumors are intimately related to the capacity of the cells for reproducing themselves, an indispensable mechanism for the dissemination of the disease.
“We now need to confirm this data with the use of other methodologies”, says Marco Antonio Zago, from the Ribeirão Preto School of Medicine, of the University of São Paulo, the coordinator of the Clinical Cancer Genome. “And to distinguish in which genes their greater expression is a cause, and not a consequence, of the greater invasiveness of the glioblastomas.”
Nine kinds of tumors
The analyses of gene expression made in brain tumors are part of a pilot study of the Clinical Genome project. Its core objective was to test the methodologies that will be employed in the course of the program. That is why only three of the nine kinds of tumor that will by focused on by the Clinical Genome project were a target for these initial works, carried out by a network of five molecular biology laboratories, with the support of two bioinformatics centers. Besides working with samples of brain tumors, the pilot study measured the workings of groups of genes in cells with cancer (and in healthy ones) taken from the larynx and the colon. “But, for the time being, we only have significant results with brain cells”, Zago says. Started two years ago, the Clinical Genome project works with samples of tumors that are formed in nine different organs or tissues of the human body: the stomach, the esophagus, the bone and bone marrow, besides the brain, head/neck and colon/rectum.
With a budget estimated at US$ 1 million, funded in equal proportions by FAPESP and the Ludwig Institute, the program researches into whether significant differences in the workings of a gene, or of a set of genes, in cells with cancer and their respective healthy tissues can provide useful information for diagnosing or treating the disease. “We want to investigate whether these changes in expression are correlated with the clinical parameters of the people who are ill, such as survival rate, response to treatments, and propensity for metastases (the tumor expanding to other tissues)”, Zago explains.
Not that the target of the program is to develop expensive genetic tests to be applied directly on cancer patients and so to forecast the evolution of the disease or to provide guidance on the best form of treatment. The objective is to relate the behavior of the genes in the tumor tissues with alterations in physiological parameters of the patients, such as the production of proteins and antigens (substances recognized as potentially aggressive by the immune system). The advantage of this approach is that simple and inexpensive tests could be used to detect these physiological alterations. In countries like Brazil, it would be no use to develop very expensive and complex procedures, which would be difficult to introduce into the clinical routine of public hospitals. Furthermore, these proteins and antigens are potential targets for the development of new therapies.
The way how the Clinical Genome project is studying the behavior of the tumor genes is innovative, in that it brings together a varied team of specialists and institutions to pursue its objectives. In spite of its name, which points to the area of genetics, the program is not run by molecular biologists only. The majority of those taking part are clinicians, surgeons, epidemiologists and pathologists, who are engaged in daily contact with cancer patients. These doctors belong to 19 clinical research groups from hospitals and universities in São Paulo.
Their role is of fundamental importance for setting up a structure of primary data that will be used in the course of the whole program in a series of studies. This structure rests on two pillars: good quality samples of nine types of tumors (and their respective healthy tissues) and a detailed epidemiological profile of the patients and healthy people who have supplied these tissues. Up until last December, the Clinical Genome project had clocked up samples from 1,124 patients with tumors and 793 healthy persons, the so-called control cases.
This biological material is the raw material for carrying out studies about gene expression in tumors that are now starting to be done by molecular biologists from the program. Furthermore, the clinicians and surgeons are capable of formulating specific pertinent questions about the behavior of the diseases, which can be answered by molecular analyses. “The Clinical Genome project has built a bridge between laboratory researchers and the clinicians and surgeons who work in hospitals”, says Marcos Brasilino de Carvalho, an oncologist specialized in head and neck surgery from Heliópolis Hospital, in São Paulo’s capital city.
“We present them with the cancer patients, and they show us the genes and the chromosomes.” The coordinator of one of the 19 groups that sent samples of cells with cancer and healthy cells to the program, Carvalho and his team have now provided cell material taken from about 150 patients with head and neck cancer. “The Clinical Genome project was a boost for research in our area”, says neurologist Alberto Alain Gabbai, from the Federal University of São Paulo (Unifesp), whose research group provided samples from 40 brain tumors removed from patients.
To guarantee the quality and the homogeneity of the samples of healthy tissue and tissue with cancer, the program laid emphasis on standardizing the procedures used during the collection of biological material. In surgeries for the removal of tumors, for example, there was always a pathologist in the surgery room. This specialist had responsibility for ensuring the purity of the tumor sample that was to be sent to the Clinical Genome project. Another task for the pathologist is to classify basic characteristics of the tumors (type of cancer, their degree of evolution and invasiveness). “Sometimes, the surgeon removes some 20 centimeters of tissue, but only some 3 centimeters are cancer cells suitable for analysis”, explains Venancio Alves, from USP’s School of Medicine, the coordinator of the pathology part of the program. “So it is up to someone from our area to separate the healthy tissues from those that really show the tumor.”
All this effort for getting quality biological material would be of no use unless there were a computerized database, accessible to all those taking part in the program, guaranteeing updated and reliable information about the patients (and control cases) who supplied almost 2,000 tissue samples to the program. Accordingly, it is possible to know in detail the clinical profile and general information about all these people. “Periodically, we carry out a check on part of our information”, explains Victor Wünsch Filho, from USP’s Faculty of Public Health, who coordinates the epidemiology part of the Clinical Genome project. “This is the only way we can do large epidemiological studies about cancer.”
Computer programs assist the researchers in this task, automatically pointing out any strange data that there may be in the patients’ records. For example, if the on-line form of a cancer patient contains the information that he started to smoke before being 10 years old, the program advises the researchers about the doubtful nature of this item. With the warning, the group that interviewed the patient takes care of checking the veracity of the information. “This on-line system is a great advance”, comments Wilson Silva Jr., one of those responsible for the program’s bioinformatics. “It also gives us a powerful tool for accompanying the evolution of the tumors and establish relationships between the profile of the disease, the genetic load, and the patients’ environmental and behavioral factors.”
Clinical Cancer Genome; Modality Clinical Genome Program; Coordinator Marco Antonio Zago – FMRP; Investment US$ 1 million