Cancerland

by David Scadden

The MGH at that time had whole floors of patients in various stages of dying from AIDS. It was horrifying to see such young people in such misery. But MGH was one of the places where AIDS research was a priority. Bruce Walker was asked to head the AIDS Research Center and he invited me to be its codirector. He focused on the immune response to HIV and a vaccine, I focused on the consequences of immune deficiency and how that led to cancers. We shared lab space and constructed a shared office suite. It was a rare moment when professional commitment, friendship, and the opportunity to do something of importance all came together: a true gift.

  The cancers that occurred in AIDS were not those that happened in most other settings. They were due to poor immune control of viruses or the immune system itself. Kaposi’s sarcoma and lymphoma were of extraordinary frequency and we tried to develop new therapies for each.

  Nearly all of our patients were too sick to endure the typical chemotherapy used to treat these cancers. (This despite new drugs that could relieve the nausea, anemia, and low white cells that are among the worst side effects of chemo.) For Kaposi’s sarcoma, my friend at University of Southern California had found that lower doses of a chemical called paclitaxel, which had helped some patients with various malignancies including breast and ovarian cancer. In many patients we saw the cancer retreat and even disappear.

  Paclitaxel, which is known by the brand name Taxol, was derived from the bark of the Pacific yew tree. Long known for producing poisonous leaves, which were sometimes used in suicide, yews have a particular place in history. Yew pollen can cause allergic reactions, and in some cultures, even sleeping beneath a yew is considered to be dangerous. Richard the Lionheart and King William II were supposedly killed by arrows shot from yew bows, and the tree’s boughs were incorporated in funeral rites in ancient Rome.

  The yew bark that produced the first paclitaxel was one of two hundred samples collected in the Pacific Northwest by Arthur Barclay as part of an ongoing effort by the National Cancer Institute to discover potentially useful molecules in plants. Much of what we use as medicine is derived from plants and primitive animals like fungi, which have been producing chemical compounds to defend themselves for roughly four billion years. Aspirin, opium, quinine, and digitalis are among the better-known, plant-derived medicines. Fungi gave us penicillin. Some cardiac medicines are based on substances from lower vertebrates like snakes. The value of natural chemical sources has remained quite constant over the years, with between 12 and 50 percent of new medications credited with origins in the plant or animal kingdoms.

  Years passed between the collection of the yew bark and the discovery that it contained a substance that could kill ovarian and breast cancer cells. This research set off a scramble for the great volumes of yew bark needed to conduct more studies. Loggers in economically depressed communities hoped for a big new business. Environmentalists rallied to protect the forests and the animal species that depended on the yew, including the northern spotted owl. A fierce political debate, which was sometimes framed as a contest between human life and environmental protection, ended when chemists figured out how to synthesize the key molecule.

  The federal government promotes and funds research but isn’t in the business of actually marketing drugs, so when Taxol was eventually approved for use in patients, the National Cancer Institute sought a partner in the industry. The contract was won by Bristol-Myers Squibb (BMS). The drug became a blockbuster, with sales that exceeded $1 billion per year. The company’s profits far exceeded the few millions in royalties paid annually to the federal government, and this sum grew as science showed that additional types of cancer were vulnerable to it.

  When Parkash Gill and I contacted BMS to tell them of early success with low-dose Taxol therapy for Kaposi’s sarcoma, and to ask them for help in conducting a more extensive clinical trial, they scoffed: the dose was almost half that used in other cancers; it couldn’t possibly work. They refused to even allow us discounted access to their extremely expensive drug. We had to find the money ourselves to do the study, and fortunately we did. The results were quite dramatic and even people with severe lung disease recovered promptly. BMS took note. They also realized that their drug was soon to come off patent and our data combined with some from the NCI might get them continued exclusive marketing. Treating an “orphan disease” like Kaposi’s sarcoma could result in seven years of exclusivity from the FDA, keeping any generic from coming out to compete with them. To do so, they needed our data. We were hopelessly naive about the pharmaceutical business. When they asked what we wanted for our data, we thought the right thing to do was to only ask for the reimbursement of our costs. In hindsight, we could have funded our research for seven years. They got their “orphan drug” status for Taxol and made billions.

  As most Americans would eventually learn from political debates over our health care system, and occasionally from personal experience, drug patents produced great flows of revenue. Unless academic centers hold the patents or, like in our case, can pay for and own the information from a clinical study, there is nothing beyond covering the costs. The revenues all go to the drug makers. While they have handsome profit margins, in their defense, the costs associated with successfully moving from research to drug are unfathomably large—estimated to exceed $2.5 billion in a 2016 analysis by the Tufts Center for the Study of Drug Development. The reasons are multiple but include the terrible inaccuracy of research models in predicting what will happen in patients. Much of what is done in research is to reduce the complexity of disease to as simple a set of models as possible. It is the only way to get a handle on the details of what is happening. In drug research, the emphasis is often on molecules: specific enzymes or single receptor molecules that have been implicated in a disease. In cancer, the focus is often on what gene is mutated and whether that mutation causes the molecule the gene makes to be more active or less. Anything that is more active is a more desirable target to go after because it is easier to inhibit something than to make it more active. Finding a good target is often where discovery labs in academia make their contribution. Sometimes they add identification of a chemical compound that might affect the target. But it is the pharmaceutical companies that have the expertise to go from a compound to a drug.

  Getting to a drug involves an enormous team of people who contribute pieces of the process. Chemists modify the original compound to identify variants that might be more specific in what they target (almost all drugs are “dirty” in affecting more than their intended molecular target). Medicinal chemists also define variants that are more easily soluble in water, or more stable in storage, or more able to get inside a cell without being destroyed. Cell biologists assess how the compound alters cancer cells, killing them, just stopping them from growing, or changing their differentiation state. Pharmacologists measure how the compound is handled by the body: whether it is absorbed in oral form, how long it lasts in the bloodstream, and whether it is cleared by the liver or the kidneys. Toxicity is then evaluated in both Petri dish–based assays and multiple different animal types, but none of them completely mimic what will happen in a person. Most drugs end up failing in the late stage of development because of this reason. We just still don’t have decent ways to determine if a drug that passes all the tests will not harm people. Once a drug enters into clinical testing the costs exponentially increase and failure collapses down immense efforts by dozens if not hundreds of people.

  Prices in the U.S., build in virtually all the costs of developing the new drug. They also include the costs of other drug candidates that might have failed. Other countries with national health systems negotiate drug prices. That is not the case in the U.S. where Medicare and Medicaid are forbidden from doing so. That results in pricing strategies by pharmaceutical firms where the U.S. prices are substantially higher than elsewhere (often twice the price of other developed nations) and projected to be the major source of revenue. It is an increasingly contentious debate. Clearly the companies need to cover their rese
arch and development costs and return a profit in order for new drugs to be made. It has to be lucrative enough so that new companies, whose risks are amplified by the limited number of new drugs they can work on, can be attractive to investors. The large pharmaceutical companies are highly averse to risk. It is the small start-ups that are required for innovation. Those start-ups depend on venture capital that has to tolerate the long interval between investing in a company and having a medicine that can be sold—a time estimated to exceed a decade. Investors have lots of options for their money, so to take on the high risk and long duration of new drug development, the possible returns have to be high.

  The true cost of a drug is often obscured from a patient’s view because it is paid by an insurer. This divide means that the insured and their doctors are shielded from the cost-benefit factor and can simply demand treatment, even if it is extremely expensive and of limited benefit. When the insurance system was first devised, few medicines were priced so high that the cost mattered much. If a treatment offered a potential benefit, no matter how limited, it was easy to go ahead with it. But as costs rose, the people who ran the system began casting a wary eye toward medicines that produced only incremental benefits.

  In the case of Taxol, which cost about $35,000 per patient when it was approved, the benefits were impressive in certain advanced breast cancer patients, reducing the risk of progressive tumors. In some cases, it was even prescribed to prevent recurrent cancer in women determined to be largely cured by surgery. The drug works by interfering with the division of cells, including both healthy ones and cancerous ones, and ultimately killing them. Drug resistance was common in patients who received Taxol. Nevertheless, oncologists could prescribe Taxol where other drugs failed and to help some patients live longer. When resistance arose, they could switch agents, though inevitably cells start to become “cross resistant,” resistant to many types of drugs.

  Drug resistance hinges, in part, on a survival-of-the-fittest process that allows a small number of cancer cells that aren’t susceptible to a drug to replicate until the body is repopulated and illness returns. This dynamic mimicked evolution run amok, with drug-resistant cells thriving as susceptible ones became extinct. It is why most drugs are given in combination with others. The idea is that if you first give medicines that work by different mechanisms, a cell may be resistant to one mechanism, but is unlikely to be resistant to the mechanisms of all the drugs. Combinations of drugs with distinct mechanisms of actions are often designed and some have worked spectacularly well. The cure rates for childhood leukemia, testicular cancer, and lymphomas are stunning successes. But most cancers that come from what are called epithelial cells, the cells that make up and are the basis for cancers of the breast, lung, colon, prostate, stomach, pancreas, bladder, and liver, can withstand the multipronged attack. This is also true of cancers from the brain and many from our musculo-skeletal system that are called sarcomas. As this process became better understood, oncologists adjusted the timing, the dosing, and the variety of drugs used to try to find some way around the cancer cells’ defenses. The strategic use of medicines sometimes extended life, but this shuffling of the deck didn’t represent real progress. And in the long run, many cancers came back in more virulent forms, and many patients rightly thought they were playing for time, hoping to live long enough for science to produce a definitive cure It made going to clinical oncology meetings almost unbearable. It seemed that innovation was reduced to just recombining already poor performing drugs and the results were at best tiny increments in patient survival.

  Although hope was ever present, the 1990s brought far too many profound disappointments. One of the most crushing began with a front-page article in the Sunday May 3, 1998 edition of The New York Times that announced that two naturally-occurring proteins—angiostatin and endostatin—seemed to halt the development of blood vessels that tumors required to survive. The scientist leading this research, my Harvard colleague Judah Folkman, had devoted decades to the study of angiogenesis—angio for “blood vessel” and genesis for “creation”—and did impressive work on the way that tumors produce chemicals to stimulate nearby capillaries to develop new blood vessels to nourish the malignancy. Folkman’s team did painstaking work that included processing, literally, gallons of mouse urine to find proteins that could shut off the blood vessel growth. When this success was announced Folkman was cautious, saying, “If you have a cancer and you are a mouse we can take good care of you” but others were not so measured. Nobel laureate James Watson predicted, “Judah is going to cure cancer in two years.”

  In the days after the article appeared a drug company that made the two medicines experienced such a surge in investments that its stock price rose 50 percent. Patients all over the country clamored to be treated with these proteins even though they had never been tried in humans. One oncologist in New York reported that a very wealthy patient called to ask if “a large infusion of cash” could get him treatment based on Folkman’s science. A large contract to produce a book about Folkman’s discoveries was part of the enthusiasm.

  All the attention affirmed what was a lifetime of work for Folkman, who had sometimes been the only prominent scientist devoted to the angiogenesis puzzle. Remarkably empathetic and kindhearted, Folkman was so brilliant and precocious that he was admitted to Harvard Medical School at age nineteen and while there helped develop the first heart pacemaker. By thirty-four he was surgeon-in-chief at Children’s Hospital Boston. His first application for a grant to study angiogenesis was rejected by National Cancer Institute reviewers who thought he had gotten the relationship between tumors and blood vessel backward. The rejection letter offered the blunt assertion that “tumor growth cannot be dependent upon blood vessel growth any more than infection is dependent on pus.”

  For decades Folkman’s research into blood vessel growth as a target for cancer treatment yielded little progress. Folkman was a fabled figure by the time I arrived in Boston, beloved by people who appreciated his enthusiasm, intelligence, and dogged spirit. He was also a living example of the tension between scientific passion and realism. Technical as it may be, science is a human and therefore social endeavor. This fact was illuminated brilliantly by Thomas Kuhn, a physicist, historian, and philosopher who pioneered the study of scientific progress. Kuhn was interested in the ways that communities of scientists respond to new ideas and noted that, at first, concepts that deviate from accepted wisdom are rejected, often at great pain for the men or women who devise them. A classic historical example involved the Hungarian physician Ignaz Semmelwies, who suggested that contamination caused high rates of death among women who gave birth at a hospital where he was chief surgeon. An experiment involving handwashing by doctors cut the fatality rate by 90 percent. However, the concept, which was proposed before germ theory was developed, was roundly rejected and its proponent was scorned by his colleagues. Semmelweis was emotionally devastated by the way he was treated by his peers and by the knowledge women would die because he couldn’t make his case successfully. He died, in 1865, after being forcibly committed to an insane asylum. He was just forty-seven years old.

  Like many rejected scientific revolutionaries Semmelweis would not be affirmed until decades after his death. Today he has been memorialized by statuary and coins, and a university has been given his name. The honors came as his concept attracted support and the accumulated work created a “paradigm shift” (Kuhn coined the term) that changed baseline assumptions about postsurgical disease. The scorned scientist became an acknowledged genius, textbooks were revised, and countless lives were saved.

  Every scientist dreams of making an advance of the sort Semmelweis achieved but dreads the experience he endured. The trick here involves a balance of respect for yourself and respect for others. If you lose confidence and abandon an idea too soon, and it yields a great breakthrough for someone else, you will never forgive yourself. However, persevering in the face of reasonable criticism can also bring you to a place of
regret. In the end, you discover nothing. The tension between confidence and realism is made even more intense by the ever-increasing competition for scarce research funds meted out by government, foundations, and industry. As much as we’d all like to think that data speaks for itself, you must be a good enough salesman—self-confident, energetic, and optimistic—to make funders enthusiastic enough to give you money. If there was a knock on Judah Folkman it was that he sometimes showed an almost messianic zeal for his ideas. However, his zeal had won him the support from funders who helped him create the entire field of angiogenesis research.

  The excitement that came with the big press reports was soon overwhelmed by controversy. Scientists who tried to replicate Folkman’s results, which would be required if his findings were to be accepted, struggled to do it. When this failure was reported, Judah explained that the process his lab used was complicated and, as he explained at the time, the two substances he isolated could fail for some weird reason, but the principles were sound. However, the news about these failed attempts at confirming Folkman’s results was painful for him. At a conference where he was honored for his work he wondered aloud, “How long can one persist without the acceptance of one’s colleagues?” Judah’s experience was not his fault. His had been the first and the loudest voice that insisted it was too soon to talk about cures. Oncology may be the one field where good news travels too fast, and the consequences of this problem can be devastating to scientists. In the case of angiogenesis, the replication problem was solved as other scientists also saw mouse tumors shrink. More ups and downs would follow as hundreds of labs pursued variations on Judah’s original research on the way to chemicals that would ultimately prove useful. However, Folkman’s name would long be associated with a flash of false hope, much of it created by those like James Watson, who seized on a new idea as finally providing some scientific basis for hope in cancer therapy.