Carlos Cordon-Cardo:
A Different View of Cancer

His unique vision may improve the diagnosis and treatment of the disease

By Robin Eisner
Ever since Carlos Cordon-Cardo peered at tumor cells under a microscope as a medical student in Spain, he has wanted to improve cancer diagnosis. His concern then, and now, is that cancer tissue samples from two different patients might appear the same, but one person after treatment will experience a recurrence and may die,

Photo credit: Robert Bean
Carlos Cordon-Cardo, center, with Elizabeth Charytonowicz, right, senior staff associate in Dr. Cordon-Cardo’s lab, and Hanina Hibshoosh, M.D., associate professor of clinical pathology and co-director of the Molecular Pathology Core Shared Resource
Carlos Cordon-Cardo, center, with Elizabeth Charytonowicz, right, senior staff associate in Dr. Cordon-Cardo’s lab, and Hanina Hibshoosh, M.D., associate professor of clinical pathology and co-director of the Molecular Pathology Core Shared Resource
while the other individual may live because the disease does not come back. There is no way a doctor can predict which course the disease will take.
    To understand what causes variation in cancers as a means to provide better diagnoses, prognoses, and treatment for patients has been Dr. Cordon-Cardo’s professional quest for the past 30 years as a research pathologist and cancer biologist. Recently, Dr. Cordon-Cardo, who brought his quest to Columbia University in 2006, has moved a step closer to fulfilling his decades-long aspiration.
    Carlos Cordon-Cardo, M.D., Ph.D., and colleagues have developed a new platform, called systems pathology, which relies on microscopy, optical engineering, clinical parameters, biomarkers and advances from the human genome project, and computer algorithms that analyze the data, to create an objective and quantitative measure of tissue structures and molecular components in tumor cells to help predict prostate cancer behavior.
    In an application of the tool, published in the Journal of Clinical Investigation in 2007, the method could predict recurrence with high accuracy for each of almost 400 men with prostate cancer who had been treated with surgical removal of their prostates.
    “Our approach is a more individualized assessment of what may happen with a tumor,” explains Dr. Cordon-Cardo, vice chairman and professor of pathology; professor of urology; and associate director of the Herbert Irving Comprehensive Cancer Center. Dr. Cordon-Cardo runs his laboratory, and the molecular pathology shared resource, with 27 people located in NewYork-Presbyterian Hospital and the Irving Cancer Research Center, a new Columbia University building that opened in 2005.
    Today, physicians usually are able to give cancer patients only a statistical likelihood their tumor will recur based on criteria developed in populations. The standardized measures include the microscopic appearance of the tumor, biological markers, and the cancer’s spread, either locally or throughout the body. But doctors cannot definitively forecast whether a particular patient’s cancer will recur and spread. Systems pathology integrates clinical and biological knowledge and its aim is to offer personalized information about an individual patient’s tumor. In time, the technology could be enhanced to tailor treatments for specific patients or groups of patients.


The context of history
Dr. Cordon-Cardo puts his new systems approach into the context of the history of pathology: the study of the causes, processes, development, and consequences of human disease. Physicians from the Greeks to about 400 years ago, he explains, first
Two tumors that look alike under the microscope from two patients with similar clinical history and same stage of disease may be molecularly different and have distinct clinical behavior.
believed disease to be due to imbalances in various humors in the body and to supernatural forces. But autopsies through the centuries started to reveal that certain illnesses were associated with different organs in the body, bringing a more rational approach to medicine.
    By the early 19th century, microscopy enabled pathologists to visualize differences between healthy and unhealthy cells to characterize disease closer to the site of the problem. Since the 20th century, advances in biochemistry and genetics have allowed researchers to look deeper into cells and they have discovered biological markers and mutations associated with different diseases.
    With new microarray technologies, using microchips developed in the last decade that allow rapid study of thousands of molecules of DNA, RNA or proteins at the same time, scientists have begun to make associations between common diseases, such as cancer and diabetes, and specific groups of genetic regions and/or proteins found in patients.
    Dr. Cordon-Cardo, however, says he thinks his systems pathology approach may have advantages over microarray approaches for analyzing cancer. “The most wonderful microchip that has been invented is something inside all of us. It is called the tissue. Why take a tissue and destroy it to put it on top of a platform [such as a microarray], when you can put the platform on top of the tissue? Why not use the tissue with the cells, microanatomy, and cartography, which has been well studied, and put the platform on top of it?”
    Microarray analyses require tissue destruction to isolate component biological molecules for analysis. Dr. Cordon-Cardo likens the process of using microarrays to study disease to shredding a priceless oil painting and then trying to figure out what it looks like and its meaning by studying the disordered fragments. That said, these approaches and “other disruptive technologies” are offering great insight into the complexity of cancer and other diseases.


Systems pathology method in prostate cancer
The systems pathology approach maintains tissue architecture but uses technology, such as new microscope lenses, imaging software, and neural network learning algorithms, to quantify and standardize the content within the image. Knowledge from pathologists’ understanding of
Integration of clinical variables, histological features, and molecular profiles achieved by the application of techniques in object-oriented image analysis, pattern recognition, and quantitative biomarker multiplexing. Complex datasets are analyzed by supervised mathematical approaches, including machine learning algorithms and neural networks.
Integration of clinical variables, histological features, and molecular profiles achieved by the application of techniques in object-oriented image analysis, pattern recognition, and quantitative biomarker multiplexing. Complex datasets are analyzed by supervised mathematical approaches, including machine learning algorithms and neural networks.
cancer tissue provides characteristics for the algorithms to assess. Biological and clinical markers, such as prostate specific antigen levels before surgery, also are part of the analysis.
    In the JCI prostate cancer study, Dr. Cordon-Cardo and colleagues first analyzed clinical and tissue data from more than 262 patients with prostate cancer whose glands had been removed surgically. Of those, 37 experienced a recurrence based on rising prostate specific antigen levels in two consecutive tests. The scientists then assessed 93 attributes, including physical characteristics of the tissue, biological markers in the cells, and clinical parameters, in the patients who experienced recurrence. Using statistical methods, the researchers found eight attributes that strongly predicted PSA elevation.
    The extent of lymph node involvement, the surgical margins (a measure of whether all the cancer was removed), the biopsy Gleason score (a pathological assessment of the virulence of the cancer), and invasion in the seminal vesicles outside the prostate were the four clinicopathological characteristics. The three image-based attributes involved irregularity in tissue structures seen in advanced cancer.
    The level of androgen receptor expression in the tumor was the eighth attribute used to predict recurrence. Under normal conditions, the androgen receptor binds male hormones and plays a role in the growth of cells that express it. The androgen receptor is present at low levels in most cells in the prostate but becomes mutated and highly expressed in prostate cancer.
    To see if the eight markers could accurately predict recurrence, they were used to assess tissues and clinical data from another group of 366 patients, whose recurrence status was not made available to the investigators. The method had an 86 percent predictive accuracy.
    The study also showed that higher levels of androgen receptor in the prostate tumor cells increase the likelihood of tumor recurrence. Although prior studies indicated increased androgen receptor levels were associated with prostate cancer severity, the systems pathology method could pinpoint and measure androgen receptor in the tumor itself, removing any confounding amounts from adjacent healthy tissue that also express the receptor.
    For now, the systems pathology method in prostate cancer cannot change the way patients are treated. Clinical trials, in collaboration with Aureon, a Yonkers, N.Y., firm developing and testing the methodology, are under way to further validate the predictive nature of the attributes and biomarkers to determine if certain treatments are more effective in selected patients. Dr. Cordon-Cardo is a scientific founder and member of the board of Aureon Laboratories.
    “Cancer research has shown that the biomarker of today becomes the target of therapy tomorrow,” he says, citing Gleevec and Herceptin. Gleevec is a drug designed to turn off tyrosine kinase enzyme molecules that have become altered in certain cancer cells. Herceptin is a humanized antibody that neutralizes the Her2-neu receptor that is overactive in some breast cancer patients.
    Dr. Cordon-Cardo hopes to apply the systems pathology methodology to other cancers. He also has developed mouse models of cancer. “I want to use the knowledge from humans to produce a disease model in the mouse, but I want to take advantage of the disease in the mouse to better guide the clinical trials of the future and obtain novel biomarkers.”


Trajectory as a physician
Dr. Cordon-Cardo knew he wanted to be a doctor when he was 10 years old. For his birthday, he asked his grandmother for a microscope. She gave him a corner of the family house in Spain that became his first laboratory. By 17, he knew he wanted to study cancer and pathology. “There are people who have a calling to do sports or to do science,” Dr. Cordon-Cardo says. “I had a call to look into the microscope early on. I knew I wanted to study cancer.”
    
Cancer is not just a disease of tumor cells, but also of the host. Integration of profiles and relationships between tumor cells and host blood vessels, as well as inflammatory infiltrates, together with clinical, laboratory, and histopathological parameters, renders a more accurate “individualized” assessment of the disease and gives the patient a better chance of cure by opting for the most appropriate therapeutic intervention.
Cancer is not just a disease of tumor cells, but also of the host. Integration of profiles and relationships between tumor cells and host blood vessels, as well as inflammatory infiltrates, together with clinical, laboratory, and histopathological parameters, renders a more accurate “individualized” assessment of the disease and gives the patient a better chance of cure by opting for the most appropriate therapeutic intervention.
He started medical school in 1975 in Barcelona, Spain, but quickly realized he was more interested in research than clinical care because there “were so many unanswered questions in medicine.” Also, the molecular biology revolution in the 1980s offered enormous promise in helping to answer the questions. He came to America, he said, because of its enormous research resources and to learn more about the basic sciences. He completed his medical degree in 1980 from Autonomous University of Barcelona in a special program for foreign students.
Although he loved his microscope, he was frustrated by its limitations. “You cannot judge a person by what he or she looks like, neither can you judge tissue by its appearance,” Dr. Cordon-Cardo says. “You need to spend time and resources to understand a person and cells.”
    So he decided to pursue graduate studies, receiving a Ph.D. degree in cell biology and genetics from Cornell in 1985. He completed two fellowships at Memorial Sloan-Kettering Cancer Center: in experimental pathology from 1982-1983 and in immunopathology from 1983-1987.
    He became an independent investigator at MSKCC in 1983, staying there for more than 20 years. He created and became the first director of the division of molecular pathology in 1995. While at MSKCC he made many important discoveries about the molecular biology of cancer. (See accompanying article about his research accomplishments.)
    
“There are people who have a calling to do sports or to do science,” Dr. Cordon-Cardo says. “I had a call to look into the microscope early on. I knew I wanted to study cancer."
He came to Columbia University for many reasons, he says. First among them is that Columbia is affiliated with a hospital that specializes in many diseases, including cancer. At a university hospital, such as the Columbia campus of NewYork-Presbyterian Hospital, unlike MSKCC that focuses only on cancer, tissues from many patients are collected with infectious and inflammatory conditions before malignancies arise that might provide important clues to cancer development. MSKCC tissue samples are mostly from cancer biopsies.
    “Columbia University’s mission is not just to help the patient with cancer,” Dr. Cordon-Cardo says. “It is also to understand the chronic processes of many different diseases, including cancer.”
    Also, being in an environment where research is conducted in many disciplines at the highest levels allows Dr. Cordon-Cardo to be among scientists studying related subjects, such as stem cells, but in different applications.
    One of Dr. Cordon-Cardo’s working hypotheses is that adult stem cells, immature cells in the body that have the capacity to develop into a wide range of tissues, play a role in causing cancer. For the past few decades, the prevailing hypothesis about cancer is that it arises due to mutations in adult cells in the body that cause them to revert to a less mature stage and grow uncontrollably. Increasing evidence, Dr. Cordon-Cardo says, suggests that adult stem cells, which are called to help injured organs and tissues in the body to repair them during stress, such as inflammation or infection, become mutated and continue to grow in their immature state and form some cancers.
    Dr. Cordon-Cardo says an additional rationale for joining Columbia was his access to younger students, such as graduate and medical students in his laboratory and during teaching. At MSKCC, he worked mostly with fellows, who “are more advanced in their thinking processes.” One of his goals is to change the future of pathology to become a more objective science, so he wants to influence people earlier in their careers. He hopes systems pathology will help move the discipline from an “eminence-based” system of human expertise subject to variability to an “evidence-based” method.
    Dr. Cordon-Cardo has more than 380 peer-reviewed scientific publications and 85 reviews and book chapters listed on his CV. His combined federal grant support, some of which is shared with other institutions, exceeds $50 million. He is one of the most highly cited authors in biomedical sciences as measured by the Institute of Scientific Information. His many awards include an honorary doctorate from the University of Barcelona in 2006 and “Member of Honor” recognition in 2007 by the Royal Academy of Medicine in Catalonia, Spain.
    In October 2007, however, he won a different type of award — from the Spanish Society of Oncology for an article he co-wrote in a December 2006 Sunday magazine supplement of the Barcelona newspaper La Vanguardia. The topic was cancer, but this time he was the subject, describing in first person his own bout with colon cancer in 2002 and 2003, developed through an inherited genetic condition that predisposes him to cancer.
    He decided to work with a well-known writer, Josep Corbella, to draft the piece about his experiences to help overcome the stigma of cancer in his home country. In the article, he described feelings of fear after the diagnosis, overcoming the dread, and the support from his wife, doctors, nurses, and friends during the ordeal. He also highlighted resources for anyone facing cancer.
    Interestingly, being a cancer survivor has not made Dr. Cordon-Cardo more motivated to do his research. As a scientist, he said, he has been combating cancer for decades. What is different now, with his systems pathology approach, is that he may soon be able to claim major victories against the disease on the professional, as well as personal, fronts.
Selected Research Findings
A theme of Carlos Cordon-Cardo’s studies has been to take findings about cancer from basic science laboratories and reveal their relevance in patients.

Cancer cells exploit existing machinery from cells to survive: Dr. Cordon-Cardo showed that a protein in cancer cells, encoded by the multidrug resistance gene (MDR-1) that enables tumor cells to survive chemotherapy, is actually present in certain healthy endothelial cells, especially cells that line blood vessels, as well as some epithelial cells in different secreting and excreting organs. It had been previously thought that MDR-1 protein was unique to cancer cells. He showed that the MDR-1 gene product, called P-glycoprotein, in normal endothelial cells near the brain pumps out toxins before they can poison the neural tissue. (1989)

Retinoblastoma tumor suppressor plays a role in adult cancers: Retinoblastoma, a rare inherited childhood cancer that affects the eye, is caused by a mutated retinoblastoma gene (RB). The normal gene, which codes for a tumor suppressor, acts to impede cell growth but when altered loses its regulatory power. Children afflicted with the condition are often cured of their eye tumor but then develop other tumors, such as sarcomas that affect connective tissue, nerves, muscles, and blood vessels. Dr. Cordon-Cardo was among the first to show that the RB gene also was altered in adult patients who had sarcomas and no prior history of childhood retinoblastoma. He also found that alterations in the RB gene in adult cancers, such as bladder tumors, were associated with a poor clinical outcome. (1990)

Tumors with mutations in two tumor suppressor genes, p53 and RB, are very aggressive: Dr. Cordon-Cardo was among the first to study both mutations in the same tumor and then understand how they acted together to contribute to a dire prognosis for the patient. Normal RB helps control the growth of a cell. Normal p53 protein acts to survey damage to the DNA in the nucleus, a marker of severe injury in a cell, and targets the cell for self-destruction so the impaired cell doesn’t grow. When both control mechanisms are lost in cancer cells, Dr. Cordon-Cardo found they become very independent of other regulatory mechanisms and resistant to most cancer treatment. (1997)

A mutated regulator of the tumor suppressor p53 becomes an oncogene: It had been known that in normal human cells, levels of the p53 tumor suppressor protein are regulated by the protein product of the hdm2 gene. As p53 can lead to cell suicide, too much of it around would be detrimental to the cell. Under normal conditions, then, the hdm2 protein product reduces the number of p53 molecules. In certain tumors, however, Dr. Cordon-Cardo found that the hdm2 gene becomes amplified and inactivates all the normal p53, allowing the cell to grow uncontrollably. The finding produced, in collaboration with Dr. Arnold Levine, was one of the first examples of how alterations in the regulator of a tumor suppressor can lead to cancer. (1994)

 

| TOP |