Cancerland

by David Scadden

  The year 2007 brought a remarkable scientific achievement and, with it, a freeing of stem cell work from many of the ethical concerns that surrounded it. A team led by a brilliant physician scientist named Shinya Yamanaka reported they had made adult cells into truly pluripotent stem calls without using a donor egg and nuclear transfer. Instead, he used four genes that coded for transcription factors—biochemicals that direct a cell’s differentiation state—to wind back the clock on human cells. What the group produced were truly pluripotent stem cells, which could be used in research without any concern for the controversies around human embryonic tissue or cloning. These would be called induced pluripotent stem cells or iPSC.

  Because he practiced science the way it should be practiced, and therefore shared his previous work in an open way, Yamanaka had provided a roadmap for anyone who wanted to join the pursuit of iPSC. James Thomson at the University of Wisconsin had first generated human ES cells and accomplished the same as Yamanaka at roughly the same time; George Daley followed shortly thereafter. When Yamanaka began publishing his work, our Harvard Stem Cell Institute colleague, Konrad Hochedlinger, discussed it with his former mentor at MIT, Rudolf Jaenisch, who initially expressed doubts about whether the news from Japan would hold up to close scrutiny. Konrad then assigned a doctoral student in his group to replicate Yamanaka’s work. When she succeeded, he decided to move ahead on the idea of creating iPSCs that could be tweaked to become any of the more than two hundred types of cells found in the human body. Jaenisch quietly joined the race and all three— Yamanaka, Hochedlinger, and Jaenisch—published on the same day in June 2007.

  A Hollywood-handsome young man with wavy brown hair, Hochedlinger had been a student at the Research Institute of Molecular Pathology in Vienna when he attended a lecture on nuclear transfer given by Rudy Jaenisch. It was 1999, and the world was still absorbing the news of Dolly the cloned sheep. Hochedlinger then found his way to MIT and threw himself into cloning research. Frustrated with the slow processes that were being used, he helped devise new ones and was soon coauthor on a growing number of papers. A laid-back guy whose demeanor masked intense ambition, he cloned a mouse. Hochedlinger then completed a nifty experiment that showed the therapeutic potential of this science by using stem cells derived from their own skin to cure mice with an immune deficiency. The immune system is implicated in so many diseases, including cancer, that this work would inform a great amount of follow-up experiments.

  Skin cell / iPSC science would eventually lead to the astounding reality of mice created from skin cells in a process called in vitro gametogenesis (IVG). Speculation would rage over the ethical and theological implications of IVG being used to create a baby from any variety of cell donors, or even a single individual. However, in the near term, the development of iPSCs with mature cells foreshadowed the end of the controversy over the use of stem cell lines derived from embryos. It also advanced the cause of cell-based therapies for various cancers and diseases as varied as heart disease, diabetes, and Alzheimer’s. A year after Hochedlinger published on his iPSCs, Amy Wagers reported that she had transplanted muscle stem cells to treat muscular dystrophy in mice. Soon after this discovery, she, Hochedlinger, and Kevin Eggan were among the first recipients of a new “early career” award from the Howard Hughes Medical Institute. The institute created the awards in order to ease the pressure young scientists feel to produce at a breakneck pace in order to keep their positions. The fifty winners were chosen from two thousand applicants, who were judged not on the basis of their proposed project but on their abilities and commitment to science. Each one received six years of salary and benefits plus $1.5 million to defray the cost of his or her research. We had hoped that one of our three applicants would get the prize. When all three were recognized, we were elated.

  Our three young stars were focused mainly on the regenerative power of stem cells. Konrad Hochedlinger even put together a short film on salamanders, which can regenerate limbs lost to amputation. The film explained how stem cells mobilize at the site of an amputation and then differentiate to produce skin, muscle, bone, blood vessels, and more. In the end, he asked, “Why can’t we do that?”

  Although regeneration was an exciting topic, and the manipulation of iPSCs could potentially provide therapies to treat hundreds of illnesses, the work we were doing on stem cells also informed vast amounts of ongoing cancer research. In addition to exciting stem cell science, great strides were being made in the related field of immunology and in all of molecular biology. Intense studies of very specific chemicals were revealing their roles in cancer, and these revelations were being converted into therapies. A good example was work done on epidermal growth factor receptor (EGFR), which is involved in the development of skin (epidermis) and other tissues in the body.

  Like all stories of discovery, the identification of EGFR is a great human tale. The leading investigators, who were on the faculty of Washington University in Saint Louis, were a former dairy bacteriologist named Stanley Cohen and a physician and scientist named Rita Levi-Montalcini. Dr. Levi-Montalcini had lost her job in prewar Italy when the Fascist dictator Mussolini declared that Jews couldn’t work in universities, but she continued her research on nerve fibers in chicken embryos in a lab she set up at home. After the war, she divided her time between research centers in Rome and in Saint Louis, where Cohen also worked. Separately and together, they confirmed that when overexpressed, because of genetic mutation, this EGFR allowed for cancers to develop. (Cancer cells have one hundred times more copies of the receptor than normal cells.) In 1986, Levi-Montalcini and Cohen would be awarded the Nobel Prize for this work.

  Twenty years after the Nobel award, drugs that inhibited the kinds of growth factors identified by Cohen and Levi-Montalcini were moving from trials to clinical use. They first targeted a part of a breast cancer cell called the HER2 receptor, limiting its activity, and also signaled immune cells to come in for the attack. It was given the brand name Herceptin and was found to be effective in some patients who had breast cancer and tested positive for an aberration in one growth factor gene. Women who received it for late-stage cancer got, on average, a 25 percent longer period of survival after diagnosis.

  The first of what would come to be called precision medicines, because it was prescribed for patients with a specific genetic defect, Herceptin is far less toxic than the cancer drugs offered at the time it was introduced. Typical side effects were comparable to the flu, but nothing like the brutal nausea, vomiting, and fatigue associated with old-fashioned chemotherapy. Eventually the drug would be recognized as helpful to patients with early-stage cancer too. Some would live for ten years or more. Since Herceptin targets a particular abnormality, it is suitable for only one in five patients with breast cancer. And as with other drugs, cancers treated with Herceptin can return in forms that are resistant. However, doctors and scientists would continue to look for other types of cancer with similar genetic defects and try Herceptin as a treatment. The drug would work for some people with other types of cancer, most notably cancers of the digestive tract, but it was hard to predict the outcome for an individual patient, and Herceptin remained a drug that was added to other treatments. It did not replace them.

  A second early form of precision medicine, which came into common use under the trade name Avastin, depended in part on Judah Folkman’s science of angiogenesis. As you recall, Folkman had observed that many solid tumors depended on a rich supply of blood, and he theorized that inhibiting the development of blood vessels would slow or even stop many forms of cancers. The substances he developed—endostatin and angiostatin—worked in mice but not humans. Avastin, which depended on a different mechanism, wasn’t a cure, but it added months to people’s lives. Results were less encouraging in trials of breast cancer, but then came good news from doctors using it against colorectal cancer, which is second only to lung/bronchial cancer when it comes to the number of people who die from the disease each year.

  Avastin was such a
significant advance that once again people felt as if the promised land of oncology was in sight. In 2004, Andrew von Eschenbach, director of the National Cancer Institute, talked of cancer as “a chronic disease we will manage much the same way we manage high blood pressure or diabetes.” Eschenbach spoke as oncologists from around the world gathered at a big conference in New York. One of the presentations at the conference was titled “Therapy for Metastatic Colorectal Cancer: What Do We Do with So Many Options?” It was the kind of talk that made laypeople excited but annoyed scientists.

  I happened to serve on an NCI advisory board at the time when Eschenbach delivered his speech, and he had opened each one of our meetings by declaring we would “eliminate [the] suffering and death” caused by cancer before 2015. Whenever he did, the people in the room who understood cancer to be incredibly varied and complex couldn’t help but roll their eyes. This idea was a platitude one might offer to Congress, but saying it to experts resulted in whispers around the table, Perhaps some nuance from the leader of the National Cancer Institute?

  The trouble with public statements that promised so much progress was that they created unrealistic expectations among patients and their families. Physicians were left with the work of explaining to patients that progress was slow and that cancer took so many forms that we were likely to need hundreds of new therapeutic tools to get where the NCI chief said we were heading. Genetic studies of specific cancers could turn up ten thousand or more abnormalities. And sometimes we couldn’t even be sure why a treatment was successful or why it failed. In the case of Avastin, for example, the Food and Drug Administration backed off its early support for the drug’s use against breast cancer, because the results were so mixed. However, the FDA would add approval for Avastin’s use in some lung, kidney, and brain cancers.

  For those of us engaged in the science every day, especially in immunology, hematology, and stem cell research, the tantalizing but limited successes that began the twenty-first century seemed to circle the target but not hit it. The challenge of cancer, as demonstrated by drug resistance, meant that dealing with it would likely require something vastly more complex than interventions that worked against a single genetic problem or acted on one chemical process.

  Cancer arises, generally speaking, when the agreements that govern how cells cooperate in the body break down. In healthy animals, different kinds of differentiated cells regard each other respectfully, working together and behaving themselves. Through chemical agreement, growth is limited, energy is shared, and various duties are parceled out to be accomplished on the organism’s behalf. Malignancies occur when, for lack of a better term, cells “de-differentiate” and revert to a more self-interested state. These cells become cheaters, taking more than their share of available resources, multiplying out of control, and neglecting their responsibilities, including the duty to die and make way for replacements. The one entity that can reliably rein in these rogues, and do it routinely, is the human immune system. It was obvious that we needed to work with the body to provoke the defenses that generally save us from cancer while avoiding unintended consequences.

  SEVEN

  PROVOKING A RESPONSE

  If you stepped back from the bedside, or the lab bench, you could see the progress. In the late 1970s, more than half the people diagnosed with cancer would be dead within five years. The annual rate of new cases was climbing and would continue to rise. In the mid-1990s, it peaked at an annual rate of more than five new diagnoses for every one thousand people. Consult the data and you see that something wonderful began to happen. New diagnoses began to decline, and death rates began dropping by 2 percent per year. Now, two-thirds of all people with cancer will live beyond the five-year mark. The number of people considered cancer “survivors,” which meant their disease was cured or well managed, was approaching fifteen million. This was five times the number who could be counted in 1971.

  The biggest treatment improvements had been made in blood cancers, particularly non-Hodgkin’s and Hodgkin’s lymphoma, chronic myeloid leukemia, and eventually multiple myeloma. In the case of non-Hodgkin’s lymphoma, a drug that is an antibody, called Rituxan, had a very substantial impact and every year since it was approved, increased the number of people cured. It is now estimated that 70 percent of those given the diagnosis will live at least five years, most of them cured. It was one of the first immunotherapy drugs, using a product of the immune system, an antibody, to target the tumor. Gleevec and its relatives have proven to be game changers in chronic myeloid leukemia (CML). Before Gleevec, a CML diagnosis was a death sentence. By 2006, 95 percent of people were living at least five years.

  A host of antibody therapies like Avastin and Herceptin were being introduced and providing at least some benefit for so-called solid (not blood) tumors, though not the curative effects seen with blood cancers. Similarly, drugs targeting specific mutations, based on the model of Gleevec, became available for other solid tumors. The number of therapies increased and options beyond the first treatment choice greatly expanded. Improvement in outcomes could also be credited to technologies that allowed doctors to monitor and screen patients more closely. Colonoscopy, mammography, ultrasound, and other methods allowed for early discovery of new cancers and recurrences, and permitted minimally invasive procedures. In addition to improving survival rates, these technologies reduced complications from treatments. The use of real-time imaging by neurosurgeons, for example, allowed them to operate on brain tumors with more precision. Operations that once carried a significant risk of causing impairments to motor and intellectual functions became safer and more effective.

  The most common form of brain cancer, glioma, originates in the glial cells, which literally glue together the nervous system. High-tech surgery, radiation, and chemotherapy can knock back glioma and even restore brain functions impaired by tumors. With careful monitoring and follow-up therapy, some patients live for five years or more. However, it is a very diffuse, invasive cancer so getting it all with surgery or even radiation is nearly impossible, which means it is likely to recur. And ironically, the return of the disease can involve malignancies with new genetic mutations that were caused by treatment itself. This dynamic was first noticed decades ago by physicians who used radiation against cancers. Later, the same effect was confirmed in people with Hodgkin’s lymphoma who were given chemotherapy, which caused genetic defects that led to leukemia. The problem of new brain cancers related to chemotherapy was reported by physicians and scientists at the University of California–San Francisco’s Brain Tumor Research Center. One of the authors of this study, neurosurgeon Susan Chang, described it as a “natural history” of gliomas that evolve or “over enough time, progress to a higher grade.”

  In literature, works of natural history trace the life span and evolution of living things under observation. Natural histories always consider context, so the study of an animal would include reflections on its environment, predators, prey, and so on. Considered as a natural history, you could say the glioma paper published by the group at UCSF tracked the behavior of a kind of creature—the cancer cell—in the environment of the brain, over the life spans of people who developed this type of cancer. The authors documented the way some malignancies undergo a process of “hypermutation” that makes as many as two thousand genetic changes. These changes can make the cancer resistant or even invisible to both the immune system and medicines.

  Ever since the discovery of the Philadelphia chromosome deformity, which is a defect found in leukemia cells, science had understood that cancer cells have genetic signatures. In some cases, like chronic myeloid leukemia, malignancies are genetically similar from person to person. (This commonality allows for successful treatment of almost all patients with one medicine, Gleevec.) However, some types of cancers are essentially unique in each patient, or they are so rare that little effort is made to find proper treatment.

  The incredible variety of cancers can shock people who get a diagnosis but then discov
er that most of what they think about their illness is wrong. I once treated a middle-aged woman who had a form of leukemia so rare that hers was the only case I ever saw. The word leukemia is so closely associated with successful chemotherapy that many people hear they have it and expect to recover after grueling treatment. This did not happen in this case. Instead, my patient got sicker and sicker. She died, and to this day I’m not certain that she or her family truly understood what we were up against.

  This terrible variety of malignancies explains more than the rare cancers that cannot be treated. It also explains why cancer therapy outcomes vary from person to person. It also suggests a pathway to better care. Knowing the exact genetic profile of a malignancy could permit far more precise treatment. This idea—that we could design therapies to exploit the exact genetic makeup of a malignancy—is what doctors reference when they use terms like precision or personalized cancer care.

  The technology to readily evaluate the genomes of individual cancers didn’t arrive until the start of the twenty-first century and the completion of the Human Genome Project. Then, in the summer of 2006, a group at Johns Hopkins University announced they had sequenced cells from twenty-two breast and colon cancer patients, compared them with normal cells, and identified an average of one hundred mutations in each tumor. Of those, about one-fifth seemed to cause malignancy. These findings, which with poetic justice had been funded in part by proceeds from lawsuits brought against tobacco companies, would contribute to The Cancer Genome Atlas Project.