In the waning days of World War II, President Franklin D. Roosevelt asked Vannevar Bush, the former dean of the Massachusetts Institute of Technology and his science adviser, to draw up a blueprint for how government should fund science in the postwar world. In Science, the Endless Frontier, which came out after Roosevelt’s death, Bush argued that government should limit its support to basic research, leaving use-oriented applied research to the private sector.1 This would give university-based scientists freedom to pursue their intellectual curiosity. The report assured the president that if scientists came up with something novel and useful, the private sector would be perfectly capable of turning it to commercial use.
There has always been a countervailing view to the Bush paradigm, especially in the health sciences. From the vantage point of elected leaders funneling billions of dollars into medical research and patient populations suffering from incurable ailments, a government that simply funded a pure and detached scientific establishment without taking into account the health needs of the public would be shirking its duty. In their view, the government needed to spend its research dollars on finding cures to specific medical problems. It needed to engage in targeted or directed research.
Over the years, the federal government has done both. If Goldwasser represents the pure science ideal, Roscoe Brady reflects the many targeted research programs inside the National Institutes of Health (NIH). For nearly half a century, Brady has studied rare diseases at NIH’s sprawling campus in Bethesda, outside Washington. An estimated twenty-four million Americans suffer from five to six thousand rare illnesses, many of them like the genetic disorder made famous by the movie Lorenzo’s Oil. Patients with a rare disease—legally defined as affecting fewer than two hundred thousand patients—represent a limited market, and at best can count on spotty attention from the pharmaceutical industry. The search for treatments or cures usually falls to the doctors and researchers who have dedicated their careers to studying that particular disease.
Brady’s work on a subset of rare diseases is highlighted in a museum display on the ground floor of Building 10 of the NIH complex. Building 10 houses NIH’s crown jewel, the renowned Clinical Center, which sits in the middle of a maze of buildings and parking lots on the agency’s 306-acre main campus. NIH is the federal government’s fastest growing nondefense agency. Congress doubled the NIH budget during the 1980s and doubled it again in the 1990s. Misshapen Building 10 stands as an architectural monument to the bipartisan support on Capitol Hill for pouring money into biomedical research. It looks like an amoeba in the center of a petri dish filled with nutrients, sprouting additional hospital wings on every side.
The display in the lobby celebrates Brady’s discovery of the cause and treatment for Gaucher disease, a rare metabolic disorder that affects about ten thousand people around the world. Gaucher sufferers have a defective gene that fails to produce the enzyme needed to break down the fatty remnants of exhausted blood cells. The fats accumulate in the spleen and liver, leaving sufferers in excruciating pain. Most are Ashkenazi Jews. About one-third reside in the United States. The display’s timeline traces Brady’s work at NIH from its fumbling beginnings in 1956 to the FDA approving his enzyme replacement therapy in 1991. The display also mentions how Brady discovered the causes of about a half-dozen similar disorders, including Niemann-Pick, Pompe, Tay-Sachs, and Fabry diseases, which, like Gaucher, are inherited. Though the natural course of each disease differs, they all result in organ destruction, lifelong disabilities, and, in most cases, early and painful deaths.
A handful of drug companies are on the verge of winning FDA approval for treatments for some of these disorders. Notably absent from the display, though, is any mention of those companies. In particular, there’s no mention of Genzyme Corporation, the Cambridge-based biotechnology firm that commercialized Brady’s first breakthrough. The omission is understandable. Brady and his colleagues helped Genzyme—one of the most successful biotechnology firms in the nation with sales well over one billion dollars a year—overcome every obstacle it encountered in the development of enzyme replacement therapy for Gaucher disease.
Yet in the mid-1990s, when confronted with protests over the price on the drug that made it the most expensive in the world, Genzyme took sole credit for its development. Years later, when Genzyme finally got around to developing a very similar drug for Fabry disease, it went around Brady because he didn’t have a key patent. Though in his mid-70s, Brady went around Genzyme and worked with another firm that had a different technology for manufacturing the missing enzyme. The result was that the FDA got to consider the rarest of rarities—two competing therapies for combating a rare disease.
There has been a lot of grousing over the years by Congress, patient groups, and some scientists that NIH spends too much money on basic science and not enough on curing disease. From the polar opposite side of the spectrum, some critics have argued that NIH’s twenty-seven separate institutes and centers should get out of the business of developing treatments altogether.2 The agency’s defenders counter that it must do both. In fact, most of the 80 percent of NIH money spent on outside research at universities and nonprofit institutes—the so-called extramural program—goes toward developing a basic understanding of the biochemistry of the human body. A substantial portion of in-house research—known as intramural research—does as well. But ever mindful of the political pressure emanating from Capitol Hill, NIH officials have always justified their in-house priorities in both basic and applied research—and lobbied for higher budgets—by claiming the government’s medical scientists were working tirelessly to promote the health of the American people. Even science critics have to admit that over the years “fighting disease was clearly a dominant personal motivation in the community of NIH scientists.”3
Brady helped shape that mold. His pioneering work over five decades into the causes of lysosomal storage disorders never lost sight of the goal that his work should one day be used to develop treatments. (The lysosome, sometimes called the police force of the cell, stores powerful enzymes for breaking down old cells, food, and bacteria.) Many of the dozens of researchers who now populate the field passed through his labs, and many of the physicians who treat the patients spent time in the NIH clinics under his tutelage. As he approached his eightieth birthday, he continued to report to his lab nearly every day and consulted with most of the companies pursuing therapies based on his research. “Brady is an amazing character, yet I would never describe him as a driven person,” said John Barranger, a professor of molecular genetics at the University of Pittsburgh and one of Brady’s protégés. “He was always pretty laid back.”4
Brady was born in 1923 to a suburban Philadelphia druggist and spent his after-school hours during the Depression tending the soda fountain in his father’s corner drugstore. Though his father hoped that he would enroll in the local pharmacy college to follow in his footsteps, Brady opted for Pennsylvania State University to study medicine. When war broke out, he rushed through his studies to attend Harvard Medical School on a military scholarship.
Brady encountered the exasperating world of medical research for the first time in Cambridge. His first-year biochemistry professor asked him to replicate a recently published experiment that had claimed that alcoholic mice could be cured with vitamin injections. He thought the experiment would take three months. Four years later, Brady proved that vitamins provided only temporary relief from alcohol dependency. His reward for years of working with tipsy mice was Harvard’s top award for student research. “Ha! I learned something from that experiment,” Brady recalled. A tall, slender man, Brady often precedes his comments with a short, barking laugh. “It taught me how long research takes.”5
After graduating from medical school, he accepted an internship and then a two-year fellowship at the University of Pennsylvania, where he studied under Samuel Gurin, a biochemist researching the body’s lipid system and its relationship to heart disease. Gurin would later consult closely with scientists at Merck and Company as they developed lovastatin, the first cholesterol-lowering drug. After the Korean War started, Brady was called up and given a choice. They wanted the draftee to run a clinic at the Navy Medical Research Institute in Bethesda. But if he wasn’t interested in research, he could report to a base hospital on an island in the South Pacific. He chose Bethesda. NIH, which had just thirteen hundred employees at the end of World War II, was expanding rapidly in the postwar years. Congress created a new institute almost every year. Brady began spending his nights and weekends working on experiments at the National Institute of Neurological Disorders and Blindness. When that war ended and he received his discharge, he joined its staff, where has remained ever since (Blindness having been replaced with Stroke at the institute).
Noting his prior experience at Penn, the director put him to work studying lipids (the insoluble fatty parts of cells) in the brain and nerves. But within a few years Brady gravitated to the disorders caused by excess lipid buildup in the body’s major organs. Many of the lysosomal storage disorders had been discovered in the late nineteenth and early twentieth centuries by scientists who gave them their eponymous names, but little was known about their causes. Did the body create too much of a particular lipid? Was it failing to break it down properly when cells died and were replaced? Was there a genetic defect—these diseases were all known to pass through families—causing them to make the wrong lipid-dissolving enzyme entirely? Whatever the cause, the effects were disheartening to see in the clinic. People with Gaucher disease, for instance, often had distended bellies from their enlarged spleens and livers, where an excess of a lipid called glucocerebroside wound up. It also wormed its way into bone marrow, and symptoms included anemia, bone pain, and a propensity to bleed uncontrollably and suffer bone fractures. Fabry patients built up lipids in the walls of small blood vessels, which led to unbearable pain in the feet and hands, kidney and heart failure, and almost always early death. The average life expectancy for a Fabry disease sufferer was forty-one years.
Brady spent his first two decades at NIH unraveling the mysteries of these diseases. He eventually discovered that each disease was caused by a missing enzyme, which broke down the lipids in normal people. Some people were missing the enzymes due to either genetic inheritance or a mutation at conception. The first breakthrough came in 1964 when Brady and his team of scientists identified the missing enzyme for Gaucher disease. “It was a biochemical Rosetta stone,” he said. “Once we knew this was the basis of Gaucher disease, we had the key to all the single lipid storage disorders.” By the early 1970s, Brady’s team had done similar work on Niemann-Pick, Fabry, and Tay-Sachs diseases and other lysosomal storage disorders. They also worked on mucopolysaccharide diseases, where build-ups of jellylike sugars inhibit normal growth and mental development. “There are about thirteen of them and it was the same principle,” Brady said. “They don’t have enzymes to start the breakdown.”
With much of the basic science under his belt, Brady turned to treatment. “I started with a very simple concept. If this enzyme is not as active as it should be, can you purify it from some source and put it into a patient to make him better? I wanted to try and purify glucocerebrosidase [the missing enzyme in Gaucher disease] from some human source, to reduce the possibility of humans rejecting it when it came in.” One night, while out to dinner with the father of two children with Gaucher disease, it came to him in a flash. Why not human placentas? They were fresh tissues and would likely have higher than normal concentrations of the rare proteins. It took another half-decade to work out the procedure for purifying the enzyme from the placentas. Brady and a team of scientists spent nights and weekends liquefying placentas with a hand-cranked grinder in the cold room next door to their lab. NIH would get a patent for the process in 1975.6
Two years before the patent was granted, Brady and his team had enough enzyme to attempt their first clinical trial. But tests on several patients had spotty results, largely because they didn’t have enough enzyme to maintain large enough doses over a long enough period of time. They also learned that something occurred to the enzyme during purification that made it difficult to absorb. It wasn’t getting into the cells that had built up excess levels of lipids.7
Brady and his colleagues solved the first dilemma by seeking outside help in purifying the enzyme. They contracted with Henry Blair, who had worked at NIH with Brady but left to form the New England Enzyme Center at Tufts University Medical School in Boston. Supported solely by contracts from Brady’s lab, Blair set up a lab for large-scale purification and began collecting fresh placentas. In 1981, with NIH getting ready to move into larger clinical trials and biotechnology fever exploding all around him, Blair privatized his venture. He launched Genzyme, with the NIH contract as its major source of revenue.
But just as the company was getting off the ground, the new venture stumbled. Animal experiments showed the enzyme wasn’t getting to the cells where the excess lipids were stored. Brady assigned several researchers to the problem, including Barranger and Scott Furbish, who later went to work for Genzyme. They discovered that the purification process stripped away the end of the enzyme that stuck to the lipid-storing cells. So they developed a process for restoring the sticky end of the purified enzyme and gave it to Genzyme. The company was back in business. Over the next decade, Genzyme received nearly $10 million in contracts to produce the enzyme for NIH, giving the start-up company a major lift in its formative years. Blair eventually hired a young economist named Henri Termeer to run the firm. A 1992 study by the Office of Technology Assessment (OTA) estimated that at least a fifth of all the direct research costs in developing enzyme replacement therapy for Gaucher disease was represented by that one government contract.8
After solving the purification problem, Brady’s team moved on to the next logical step: discovering the gene. Harvesting placentas would always be a time-consuming and expensive proposition. Manufacturing the enzyme through the brand new recombinant methods of biotechnology would be preferable. “Once we had this pure protein that worked, it was an impetus to have the gene in hand in order to make protein,” Barranger said. “If you had pure protein, and we were sure ours was pretty pure, you could fish out from expression libraries the [gene] sequences that corresponded to the protein.” It didn’t take them very long. Barranger and his colleague Edward Ginns identified the gene that produced glucocerebrosidase in 1984. (Ernest Beutler, now at the Scripps Research Institute in La Jolla, California, deserves credit for similar work, also NIH-funded, which he performed at the City of Hope in the early 1980s.)9 They neglected to patent it. “In those days, you didn’t patent genes,” Barranger said.10
While that work was going on, Brady focused on the clinic. With a properly targeted placenta-derived enzyme in hand, he asked the FDA for permission to use it on a patient. He didn’t have to go far to find one. Robin Ely Berman, an Orthodox Jewish physician from nearby Potomac, Maryland, had quit her practice to volunteer in Brady’s lab a few years earlier. Three of her six children had Gaucher disease, and four-year-old Brian was faring the worst. His overloaded organs had swollen his abdomen to several times its normal size. His desperate parents were about to have his spleen taken out. On a late December morning in 1983, the boy received the first of seven weekly treatments. “It was absolutely amazing. It was movie amazing,” she recalled. “Every week his stomach got smaller and smaller and smaller. He put on some weight. Then we ran out of enzyme. For another seven weeks, he got nothing. He went all the way back down to the bottom, which was absolutely agonizing and wonderful. Agonizing because you watched your kid get ill again. And wonderful because we realized the drug was having some effect.”11
To do a full-blown first-stage clinical trial—which is designed to determine safety and proper dosing—Brady needed a lot more enzyme. NIH began pouring money into Genzyme to produce it. It took until 1986 before they had stockpiled enough to begin enrolling patients. The clinicians in Brady’s lab eventually infused single doses of the drug into nearly two dozen patients at the NIH Clinical Center. Fearful of harming their already ill clientele, the government scientists at first provided extremely low doses of the enzyme. It had no measurable effect. Finally, they infused eight patients—seven adults and one child—with a much higher dose. None of the adults showed any improvement, but the child, like the young Berman, showed a marked reduction in his lipid levels.
Genzyme’s scientific advisory board, focusing on the effects in adults, wanted to cancel the program. But Brady convinced chief executive officer Termeer that the only problem was the dose. Termeer, with Robin Ely Berman in tow, began making the rounds of venture capitalists to raise money for a larger purification facility. She would show slides of her son’s miraculous (albeit short-term) recovery, and Termeer would make the pitch for cash. He raised $10 million to construct a plant that could produce enough placenta-derived enzyme to supply a long-running clinical trial and support eventual commercial production.12
Though Genzyme’s name was now on the paperwork at FDA, the company continued to rely on Brady’s team for the final clinical trial. A dozen patients were given biweekly treatments over a year’s time at NIH. The results at high doses showed adults clearing some of the excess lipids from their livers and spleens, just like the children had done in the earlier tests. The FDA approved Ceredase—Genzyme’s trade name for placenta-derived glucocerebrosidase—in April 1991. The approval gave the company the exclusive rights to market the drug because it had been designated as an orphan drug, under a law signed by President Ronald Reagan in January 1983, which grants seven years of market exclusivity to newly approved drugs for diseases that affect fewer than two hundred thousand people.
Getting drug companies to develop treatments for rare diseases has always bedeviled patient advocacy groups, which are often run by people like Robin Berman. When just several thousand or even a few tens of thousands of people suffer from a disease, most drug companies are unwilling to invest the time and money needed to come up with potential therapies. Even when most of the work on a rare disease—like Brady’s treatment for Gaucher disease—is done in an NIH-funded lab, drug companies often don’t want to get involved. There’s just not that much money in it.
Conditions were worse before passage of the Orphan Drug Act. A survey of NIH in the early 1980s found there were more than one hundred potential therapies for rare diseases languishing in its labs. But the scientists who had come up with them were usually more concerned with publishing their results in medical and academic journals than in rushing off to get a patent. That put the intellectual property in the public domain, available to anyone, so no drug company wanted to get involved. Why commercialize these so-called orphan drugs when another company could make it once it had been shown to work in patients?
Several government panels highlighted the problem in the late 1970s, but legislation to create special incentives for drug companies went nowhere until a 1981 Los Angeles Times article sparked the interest of Jack Klugman, the star of television’s Quincy. He dedicated one of his shows to the woes of a Los Angeles youngster with Tourette’s syndrome. The problem of orphan drugs “was catapulted from an insider’s dialogue conducted mainly in medical journals and federal offices to a nationally recognized tragedy.”13 Rep. Henry Waxman, a liberal California Democrat, introduced legislation in the next session of Congress to give industry special tax breaks for research in rare diseases and a seven-year exclusive market for orphan drugs. Heavy lobbying by the newly created patient advocacy group, the National Organization for Rare Disorders (NORD), and another episode of Quincy showing hundreds of angry patients marching on the Capitol—a case of art substituting for reality—pushed the bill through the House and Senate and onto the president’s desk.
The bill gave the orphan drug field a major boost. The number of companies, especially biotechnology start-ups, willing to get involved in investigating drugs for rare diseases shot up dramatically. The agency made grants to companies to work on orphan drugs and provided technical advice to small firms wending their way through the drug approval maze for the first time. It is tempting to say that the pump priming worked. By the end of 2001 there were 229 FDA-approved drugs and devices aimed at rare diseases, up from just ten when the law was passed. Hundreds more were in the drug development process.14
But those incentives don’t really explain why so many companies became willing to pursue orphan drugs and stick with them through the long development process. The seven years of exclusivity under the new law was far less time than the twenty years of exclusivity granted by a patent, and, as it turned out, most of the new therapies were protected by patents. The extra tax breaks for companies working on orphan drugs didn’t mean much to start-ups that weren’t earning money. And the law hadn’t changed the size of the potential patient population for most of these drugs, which was still quite small.
The real answer was much simpler: price. After Genzyme won FDA approval for Ceredase, it set a price on its new drug that set tongues to wagging across the industry. The initial cost of Ceredase therapy, which didn’t include the office visits for the twice-monthly infusions, was $350,000 a year for an average-sized male (the dose was weight-dependent). After a few years, when the build-up of lipids was under control, the dose could be reduced. But even the maintenance dose cost nearly $200,000 a year.
The lives of Gaucher patients on enzyme replacement therapy—which they would need for the rest of their lives—became a constant scramble for insurance coverage. The company did provide the drug free of charge when people exhausted their benefits, but that left them without medical coverage for any other health problem. “Genzyme’s pricing is similar in its consequences to a policy in which patients are offered a lifetime supply of aglucerase [the generic name for Ceredase] in exchange for the value of their remaining insurance coverage,” the federal OTA study concluded.15 “There was no rationale for Genzyme’s high price. It was beyond belief,” said Abbey Meyers, the mother of three children with Tourette’s syndrome who has run NORD since its inception. “Ceredase was discovered and developed by the NIH.”16 Even Brady was shocked. “I was appalled it was that expensive,” he said.
The outrage over the high price of Ceredase reached Capitol Hill shortly after the drug was approved. Rep. Pete Stark, the California Democrat, introduced legislation in the House to tax windfall profits generated through the Orphan Drug Act. Senators Nancy Kassebaum, a Republican from Kansas, and Howard Metzenbaum, a Democrat from Ohio, introduced similar legislation. Comparing Genzyme’s annual profits, which in 1992 had already soared over $200 million, to the cost of developing the drug, which the federal study had estimated to be less than $30 million, Stark asked, “To stimulate research that we all desire, is it required that we pay any price? Is this sustainable if we are to attack more than one disease afflicting our population? Is this return necessary to stimulate subsequent research?”17 Writing in the Washington Post, the two senators said the Orphan Drug Act was designed “to provide incentives for the development of drugs with small markets, drugs that would otherwise not be produced. Orphan drugs that are, in fact, of tremendous commercial value don’t deserve—and were never intended to receive—seven years’ worth of protection from the price competition that would make them more affordable for victims of rare disease.”18
Genzyme’s Termeer fired back in the pages of the Wall Street Journal. The marketplace will provide an answer, he declared.
Since Genzyme developed Ceredase, other companies have jumped into Gaucher’s disease research. We are now competing with a company working on a variation of our drug, and two others are competing with us to develop gene-therapy approaches. There could be as many as four or five treatments for Gaucher’s disease on the market within the next four years. If we hadn’t taken the first step, there would be no market and no additional research on the disease.19
The published statement was not only disingenuous about who invented the drug, but the company had already taken steps to ensure the promised competition never came to pass—at least, not for a long, long time. Using the unpatented gene uncovered by Barranger and others in Brady’s lab, Genzyme scientists in August 1991 applied for a patent on the recombinant manufacturing of glucocerebrosidase and its use in treating Gaucher disease, and three months later won orphan designation from the FDA for that form of treatment.20 They immediately put this new drug into clinical trials. The FDA approved it a little more than a year after Termeer predicted that intense competition was just around the corner. The new drug effectively precluded other firms from using biotechnology to develop alternatives until well into the twenty-first century. And even though recombinant manufacturing meant the new drug, dubbed Cerezyme, cost far less to manufacture than placenta-derived Ceredase, the price to patients and their insurers didn’t change.
Brady didn’t dwell on the exorbitant price tag placed on his therapy. Once Ceredase was approved, Brady turned his attention to the other lysosomal storage disorders. A quarter-century earlier, his discovery of the Gaucher enzyme defect had been the template for deciphering similar diseases. In 1991, with sales of Ceredase taking off, the senior NIH scientist traveled to Genzyme’s Cambridge headquarters to encourage the firm to use its cash flow from Ceredase to pursue enzyme replacement therapies for the other diseases.
Fabry disease, though it probably struck half as many patients as Gaucher disease, had the harshest impact on patients’ lives. Youngsters often discovered they had it when they experienced shooting pain in their hands and feet while running or jumping. They didn’t sweat and couldn’t stand heat. By the time they were in their twenties, many Fabry sufferers found themselves in dialysis clinics and on kidney-transplant waiting lists. The downhill spiral ended in coronary heart disease, strokes, and heart attacks, usually by the time they were in their early forties. Many victims never knew the cause of their suffering.
Brady’s visit to Genzyme didn’t go well. “I had a list of diseases they should address next, and Fabry was at the top,” Brady said as he reached for one of the file cabinets in his cubbyhole office at NIH. “Ha! I can show you the slide [from his presentation]. If they didn’t know it before, they certainly knew it after I told them. Why they didn’t do something about it is their business.”21 In fact, Genzyme was quite aware of Fabry disease. As early as January 1988, the company had applied to the FDA for orphan drug status for an experimental drug called trihexosidase-alpha, which was its name for the placenta-derived enzyme missing in Fabry patients.22
Brady’s touchy relationship with Genzyme regarding Fabry disease was understandable. He had tried enzyme replacement therapy for a few Fabry patients in the early 1970s, but the Fabry enzyme was much harder to purify than the Gaucher enzyme so he had to put the project on hold. In any case, by the early 1990s the era of enzyme purification from natural sources was nearing its end. Genzyme was already in the process of switching to the recombinant form of the Gaucher enzyme. Any company that wanted to get involved in Fabry disease research was either going to make the missing enzyme through recombinant technology or try gene therapy, which was the hot new item in biotechnology circles. But recombinant manufacturing required a patent on the process of making the enzyme from its gene, and Brady didn’t have it.
The process for manufacturing the missing enzyme in Fabry disease belonged to Robert J. Desnick, the chairman of the human genetics department at Mt. Sinai School of Medicine on Manhattan’s Upper East Side. If Brady was the king of the lysosomal storage disease world, Desnick was its crown prince, towering over the rest of the small field’s roster of academic researchers. A 1971 graduate of the University of Minnesota Medical School, Desnick devoted his career to studying rare genetic disorders. By the early 1980s he had built a large department devoted to the disorders at Mt. Sinai in New York, where he moved because of its proximity to a large number of patients. Desnick, whose broad forehead and dark-knitted eyebrows command immediate respect, depended heavily on the government for support. For more than a decade, he received more than a million dollars a year from NIH to study Fabry disease, grants that lasted well into the 1990s, according to agency records.
Like Brady, he helped train many of the leading clinicians in the field. Unlike Brady, he seems to have made numerous enemies along the way. Desnick turned down many requests for an interview, at first citing the ongoing clinical trials for his patented treatment for Fabry disease, and after they were concluded, his own desire to write a book on the subject. But based on a half-dozen interviews with former colleagues, there can be no doubt that he engenders as much fear as respect from others in the field. “You can’t quote me; I have to make a living,” pleaded one former associate, “but he’s the emperor over there.” Meyers, the blunt-talking head of NORD, was more direct. “He’s not a good-hearted researcher,” she declared. “A lot of people discover genes and don’t patent them. They think scientific discovery belongs to science.”23
The scientific name for the missing enzyme in Fabry disease is alpha-galactosidase. Like Brady, Desnick had purified a small amount of the enzyme in the 1970s. He had even injected it into two patients, although his test trailed Brady’s by a couple of years. But he ran into the same problem as the senior NIH scientist. It was impossible to come up with enough enzyme to pursue effective therapy. By the early 1980s, recombinant technology was making headlines, and Desnick immediately saw the possibilities. In August 1981, David Calhoun, a biochemist at Mt. Sinai, was taking a week off at his home in Leonia, New Jersey, when Desnick called. “Desnick asked me to talk to him about Fabry disease,” Calhoun said. “I knew how to sequence genes. His laboratory had no experience in that area. They were medical geneticists. They worked with lipids, grew cells in culture, and did enzyme assays. Nobody in that department had experience cloning and sequencing genes. It seemed like an interesting project to me. If we cloned, sequenced, and produced it in the laboratory, the gene product could be used therapeutically.”24
Relying on graduate students in his and Desnick’s lab, Calhoun began searching for the gene that produced alpha-galactosidase. They experimented with the probe technology then being used by commercial scientists at Amgen and Genetics Institute to hunt for the erythropoietin gene, but eventually fell back on the traditional chemical means developed in the 1950s by the field’s pioneer, Frederick Sanger. Mary Jean Quinn, a doctoral candidate in Calhoun’s lab, logged long hours testing the results in thousands of test tubes as they mapped the enzyme’s genetic sequence. “I was standing there at the computer as she read me the final sequence,” Calhoun recalled. “We knew we had it.” They uncorked a bottle of champagne kept in the lab’s refrigerator for the occasion to celebrate the culmination of three years’ work.
Unlike Barranger, the NIH scientist who had uncovered the Gaucher gene, Calhoun was very aware of the economic potential of his discovery. The tectonic shifts in the world of biomedical research had reached Mt. Sinai. Calhoun, who’d received his academic training at the University of Alabama, wanted to jump on the plate moving toward commercialization and private gain. But in the mid-1980s, Mt. Sinai as an institution wasn’t ready. The medical school hadn’t yet established an office dedicated to carrying out the 1980 Bayh-Dole Act, which encouraged universities to patent technologies developed in their labs so they could be licensed to commercial partners. “Even though I wanted to do it, there was nobody at Mt. Sinai to work with me,” Calhoun said. Instead, Desnick’s lab, with Calhoun as the lead author, described the gene in a paper published in the proceedings of the National Academy of Science in November 1985.25 Publication, according to U.S. patent law, starts the clock ticking on the year when authors can file for a patent. It passed without event.
NIH continued pouring money into Desnick’s lab for Fabry studies. The senior scientist assigned Calhoun and a slew of other researchers the task of using the gene to develop a recombinant form of the enzyme. Yiannis Ioannou, a native of Cyprus who had come to New York to attend college, began his graduate studies in Desnick’s lab in 1986. He immediately went to work on the project, which would dominate the next eight years of his life. So would Desnick. “He had a very tight leash on everything,” Ioannou said. “He always bothered me. He likes to be on top of everything to the point where he could try and manage every small aspect of day-to-day activity that doesn’t necessarily need supervision.”26 Ioannou, perhaps because he was familiar with the rigid hierarchies of European universities, was able to deal with it. Others couldn’t. Calhoun left in 1987 to join nearby City College, where he began tinkering with producing the Fabry enzyme from insect cell lines.
Desnick’s lab, meanwhile, used the fast-replicating and long-lived cells from Chinese hamster ovaries (CHO cells), the technology developed at nearby Columbia, which by that time had become the standard methodology in the biotechnology field for recombinant manufacturing. Over the next few years, they refined their process, constantly comparing their manufactured enzyme to purified human enzyme and adjusting the process until their hamster-derived molecule came out almost the same. Even the slightest change in the sugars attached to the molecule could render it ineffective. In October 1990—right around the time Brady was concluding his Gaucher clinical trial—Desnick, Ioannou, and David Bishop, who was Desnick’s chief lieutenant in the lab, filed for a patent on a process for producing large quantities of the Fabry enzyme. They also claimed its use in treating the disease.27 Even before they filed for a patent, Desnick contacted the FDA and requested orphan drug status for an investigational new drug he called Fabrase.28 It was granted on July 20, 1990.
Desnick was making all the right moves. He had closely followed Brady’s work with Genzyme and Ceredase, which was nearing FDA approval. Mt. Sinai had one of the largest Gaucher patient populations in the world, and Desnick was their primary care physician (although he usually saw them only once and then turned their care over to clinicians on his staff).29 Desnick’s work on Fabry disease, which Brady had put to the side while pursuing Gaucher therapy, gave his lab a lock on the right to make the drug recombinantly. Indeed, all the pieces were in place for taking Desnick’s Fabrase into clinical trials. All he needed was a company willing to do it.
Clinical trials on CHO cell-derived alpha-galactosidase wouldn’t start for another seven years, even though Mt. Sinai received nearly a half-million dollars from the FDA’s Orphan Drug Development office between 1990 and 1992. Desnick and officials at Genzyme, which eventually licensed his invention, were not willing to talk about the long delay. Frank Landsberger, who launched Mt. Sinai’s technology transfer office in 1992, recalled that Desnick’s demands complicated the negotiations with Genzyme, which dragged on for several years. “There were discussions about the licensee paying to have research done at Mt. Sinai. Desnick wanted to do some of the research at Mt. Sinai, certainly the clinical trials. . . . You have a conflict of interest—this is a generic problem in the industry—because Desnick would benefit financially from the outcome of the clinical trials,” he said.30 With those negotiations moving slowly, Desnick and Landsberger approached other firms. Desnick on his own initiative talked to Genetics Institute, which, despite the Gaucher example, turned him down.31
Landsberger approached Transkaryotic Therapies Inc. (TKT), a small biotechnology firm that was also in Cambridge. Harvard-trained Richard Selden had launched TKT in 1988 and shortly thereafter issued press releases stating it was experimenting with gene therapy cures for Fabry disease. TKT became a logical target for Landsberger’s efforts. But the struggling start-up wasn’t willing to meet Desnick’s demands, and talks broke off after two years of negotiations.32
Desnick and Genzyme finally cut a deal, but Fabry treatment was hardly a high priority at the firm. In 1994, for instance, the company spent $55 million or nearly a quarter of its revenue (which was mostly from Ceredase sales) on research. But in the rare disease arena, its target was cystic fibrosis, a genetic disease of the lungs that had ten times as many potential patients as Fabry disease. It also sought a toehold in the mainstream of medical markets by developing nonscarring products for postsurgical tissue repair. At the time, financial analysts lauded the move as a much-needed effort at diversification. Instead of making Fabry disease a priority, the company gave grants to Desnick and Mt. Sinai, expecting them to carry the ball.
With Genzyme now their main revenue source instead of NIH, Desnick and his team launched a full-scale effort to cure mice. They spent two years developing a strain of mice with Fabry disease, and then treated them with the enzyme. The work was largely done by graduate students. “I had three or four people working full time on the mouse studies, generating data, trying to figure out what would go into the IND [investigational new drug application],” Ioannou said. “There was no question who we were working for. We got to publish our results, and they [Genzyme] got to use them to go to the FDA.”33 The leisurely pace seemed to suit everyone involved, except, of course, Desnick’s Fabry patients, who had no idea what was going on inside their physician’s lab.
Brady, without a patent and cut off by the firm that had brought his Gaucher treatment to market, was hardly ready to throw in the towel. NIH had its own Fabry patient population, and the number of researchers and clinicians abandoning Desnick’s ship was growing rapidly. He first linked up with City College’s Calhoun, who in 1991 earned a patent for producing alpha-galactosidase from insect cells. Calhoun was already working with a small company called Orphan Medical of Minnetonka, Minnesota. They would produce the enzyme, while Brady would conduct the clinical trials. But the start-up company stumbled as it clambered up the learning curve of recombinant manufacturing.
Desperate for a capable partner, Brady called Calhoun and told him he wanted out. He had been approached by Selden of TKT, who wanted to start Fabry disease clinical trials. The offer came totally out of the blue. Brady knew TKT was interested in Fabry disease. But the company had been pursuing treatments that relied on gene therapy, not enzyme replacement. But in the early 1990s, Selden’s firm had stumbled onto an exciting discovery. TKT had learned how to produce proteins by turning on inactive human genes in cell cultures outside the body, a process he called gene activation. If allowed by the patent courts, the humanized process posed a threat to every biotechnology product on the market since the existing products were all made from animal cell lines.
Selden didn’t look the part of a man who wanted to overturn the entire biotechnology field. Sitting at a small table in one of Cambridge’s old brick industrial buildings between Harvard and MIT that have been turned into biotechnology laboratories, he struggled to explain the complicated process of gene activation to a visitor. Though still in his early forties, his curly red hair was already half gray and receding rapidly. He wore topsiders and casual slacks. His knit tie dangled at half-mast.
Since every cell contains every gene, every cell has the potential to make any protein, Selden explained. But in any particular cell, only some genes are turned on. The insulin gene, for instance, is only turned on in the pancreas, not in the brain. Each gene has sequences that keep the gene either turned on or off, depending on where it is in the body. TKT researchers had come up with a way of taking the sequences that turned on a particular gene from active cells and splicing them into other, fast-growing human cells, which could then be cultured en masse outside the body to produce whatever protein the turned-on gene expressed. Biotechnologists would no longer be dependent on hamsters or insects to produce proteins. They could now use human cells.
“There are several advantages,” Selden said. “It’s much more efficient in terms of a production system than working with conventional technology. You’re ending up using the gene in its natural location with all of the synthesis machinery that is supposed to be involved in the first place,” he continued. “The second thing is that many genes are really too big to manipulate in the conventional approach. You can use this method to make enormous proteins that are difficult to manufacture. But probably the most important thing is that the proteins that come out have the potential to be more effective.”34
By 1995, TKT had used gene activation to produce alpha-galactosidase. Selden immediately called Brady to see if he would run the young company’s clinical trials. “I felt and feel to this day that Dr. Brady is the leader in this field,” Selden said. “He has done a phenomenal amount of work since the 1950s to understand Fabry’s disease and other diseases as well. It was an honor when Brady said yes. When I was in college, he was somebody I looked up to.”35 In 1996, TKT and NIH signed a cooperative research and development agreement (CRADA) to develop human-derived alpha-galactosidase, a drug that TKT would soon call Replagal. In late summer, TKT raised $37.5 million in an initial public stock offering. The Securities and Exchange Commission (SEC) document revealed to the world that the upstart company was not only chasing gene therapy cures for Fabry disease—which was well known in Cambridge biotechnology circles—but enzyme replacement therapy as well. It also announced plans to start the first phase of clinical trials before the end of the year.36
The news sent shock waves through Genzyme and Desnick’s Mt. Sinai lab. In the spring of 1997, Genzyme for the first time mentioned in its annual filings with the SEC that it was in “preclinical” research on Fabry disease. “To date, it has successfully produced the recombinant α-Gal enzyme in mammalian cells and has shown it can reduce lipid levels in the plasma and tissues of a Fabry mouse model. The development program is currently focused on producing sufficient quantities of enzyme for pilot clinical studies, which are expected to begin in late 1997 or early 1998,” the company said.37 A year later they revealed they were testing a new drug for Fabry disease dubbed Fabrazyme.
The race was on. Two companies were chasing a therapy for a patient population that was no more than ten thousand persons worldwide. They both quickly leaped over the early safety and dosing hurdles. But then they faced a critical juncture. How should they design the final efficacy trial that would be submitted to the FDA? Measuring improvements from enzyme replacement therapy for Fabry disease was much more difficult than detecting them in patients with Gaucher disease. The lipid buildup in Fabry disease took place in the kidneys, heart, and brain, organs that were difficult to scrutinize without invasive and potentially dangerous biopsies. The reduction in spleen size in Gaucher patients, on the other hand, had been easily measured with magnetic resonance imaging machines. Moreover, reduction in the intermittent pain that afflicted Fabry sufferers, a major potential benefit of treatment, could only be shown through a survey, which the FDA might interpret as a highly subjective measure.
TKT’s clinicians, led by Raphael Schiffmann at NIH, opted to measure pain reduction as the primary test of Replagal after lengthy negotiations with the FDA. In 1999, the government lab enrolled twenty-six patients for the final study. The NIH clinicians also measured lipid levels in the blood as a secondary test of the new drug. Genzyme’s clinicians, meanwhile, led by Christine M. Eng, who worked in Desnick’s division throughout the 1990s before moving to Baylor University Medical School, chose lipid reduction in the kidneys as the primary measure. It was a more limited approach, but one that would more easily pass FDA muster if the agency was willing to accept a surrogate marker. Genzyme tested Fabrazyme on fifty-eight patients recruited from eight clinics in Europe and the United States, including Mt. Sinai. They also surveyed their patients to measure pain reduction. Both studies were double-blind and placebo controlled—neither the patients nor the doctors knew which group was getting the real drug and which the placebo. Both lasted twenty weeks.
The results, according to the academic literature, were extremely positive for both sides. Patients in the TKT test showed reduced pain and lower lipids in the blood compared to those in the placebo group. The Genzyme test cleared lipid deposits from the kidneys, heart, and skin, “the chief clinical manifestations of this disease.” There were major differences between the drugs, however. The TKT/NIH team infused their patients for only forty minutes compared to four to six hours for patients in the Genzyme/Desnick trial. Moreover, TKT’s human-derived protein generated far fewer reactions and didn’t require premedication to curb them.38
Did that make Replagal better? “TKT, by activating a gene in human cells, [has] raised the technology to the next level, the next step,” said Gregory Pastores, a physician who worked with Desnick for a decade before leaving Mt. Sinai in 1997 to join New York University, where he treats hundreds of patients for rare genetic disorders. He participated in neither company’s trial, although he has done work for both companies in the past. “It’s like cow’s milk and mother’s milk. We all feel instinctively that there are advantages to mother’s milk.”39
TKT beat Genzyme to the FDA, filing its license application on June 16, 2000. Genzyme’s application followed a week later.40 The larger company also tried to win the war in patent court. A month after filing with the FDA, Genzyme accused TKT of violating Desnick’s patent. Over their three decades of competition, Desnick had never beaten Brady on an issue of scientific or medical significance except in winning his Fabry patent. Here was his chance to play the trump card. But on December 17, 2001, the U.S. District Court in Delaware dismissed the case.
Less than six months after the two applications arrived at the FDA, Mike Russo discovered he had Fabry disease. A freelance journalist turned manuscript editor, Russo was always aware there was something wrong with him while growing up near Greece, New York. He didn’t sweat as a child. When he joined his high school golf team, he would have to beg off competing on hot days. “What’s the matter now,” his father, the coach, would ask. “It’s too hot,” Russo complained, even though it was barely past eighty degrees. He took up swimming instead. By his early twenties, doctors diagnosed high protein levels in his urine, a sign his kidneys weren’t functioning properly. His doctors took a wait-and-see approach. He suffered periodic bouts of joint swelling and itchy skin. His doctors treated the pain with steroids.
In late 2000, just past his thirty-fifth birthday, Russo received a Friday afternoon call from his nephrologist. The creatinine levels in his urine and blood were soaring, a sign his kidneys were beginning to fail. They had done some extra blood work. Could he come in Monday to discuss the results? “This is my life,” he said, and rushed over with his wife. They told him he had Fabry disease, “which we’re just finding out about.” Russo spent the weekend scouring the Internet to learn more about the rare disorder that was ruining his life. He learned that it was an X-chromosome genetic disorder. His mother had given it to him. His two daughters were carriers. He also learned that researchers had been working on the disease for a long time, and that clinical trials on a possible treatment were underway in Desnick’s lab in New York. He called. He wanted to get involved.41
In 1991, the FDA had approved Ceredase for Gaucher patients just six months after the end of a clinical trial on twelve patients that lasted a year. Nearly a decade later—a decade that has seen several waves of drug-industry-driven liberalization wash over the agency—things had changed, but apparently not in favor of patients when they had two companies competing to treat their rare disease. Six months after TKT and Genzyme submitted applications, the regulators sent letters to both companies indicating their data was inadequate. Desnick assigned Russo to his follow-up trial.
By late 2002, Mike Russo was an angry, frustrated young man. He had been fired from his job for poor performance about a year after he began taking time off every two weeks to go to New York City for his infusions. He was convinced he was on placebo, since none of his symptoms had improved. His kidneys were continuing their downhill slide toward a date with dialysis. While he waited, both drugs had been approved by the European Union (EU) and a half-dozen other countries. With two firms competing to sell in their market, the EU priced enzyme replacement therapy at $160,000 a year to start—substantially below the price Genzyme had placed on its Gaucher treatments. “They can get it in the Czech Republic but not here,” Russo said. “There are people out there who need it more than I do. Some of these people have died since. I don’t understand how other countries can get it approved so quickly, and the United States, where are you would think. . . .” His voice trailed off.42
A few months later, in January 2003, an FDA advisory committee, whose recommendations are invariably followed by the agency, met over two days to consider the two drug applications. A parade of patients testified to the vast improvements in their lives—especially the reduction in pain—since taking the regular infusions of replacement enzyme. But the FDA examiners are rarely swayed by such emotional outpourings. They need numbers. The academics on the advisory committee questioned Genzyme’s assertions that lowering lipid levels would eventually improve patients’ health status. But in the end the committee gave an overwhelming endorsement to the surrogate marker.
The next day the same group came down hard on TKT’s data presentation. The pain studies were not adequately controlled, and the company had not submitted enough data to show that Replagal had the same lipid lowering effect as Fabrazyme. The panel narrowly voted to send TKT back to do more studies, or to retest its slides to show the same thing that Genzyme had showed. “We shot for the moon,” Schiffmann, the NIH scientist in charge of the trial, moaned. “They were just smarter in their approach with the FDA.”43 Desnick had defeated Brady at last.
The competition between Genzyme and TKT presented a startling new paradigm for orphan drug development, indeed, for all drug development in the postgenome era. The U.S. Orphan Drug Law was premised on the idea that no one was interested in developing drugs for tiny markets. Commercial firms, it was believed, wouldn’t touch therapies developed in government labs or in universities that were either unpatented or protected by weak process patents. The law granted seven years exclusivity to overcome that barrier. Of course, it never precluded another company from coming along and getting its drug approved for the same orphan disease if the manufacturer proved its drug was better. In fact, the FDA had twice withdrawn orphan status for interferon treatments for multiple sclerosis when superior versions came along.44
The race to develop a Fabry treatment blew the old paradigm apart. Largely because of the high prices set on orphan drugs, two companies found it worth their while to pursue different versions of the same therapy. In the years ahead, feisty start-up biotechnology firms like TKT will inevitably come up with even better and more innovative ways of building proteins and molecules. Will they be prevented from entering the market? As Senator Dole wrote in the early 1990s, the Orphan Drug Act was never meant to prevent competition.
However, there are forces at work to frustrate the application of new technologies like Selden’s for decades. TKT would have never had a chance to pursue enzyme replacement therapy if Desnick and his team had patented the Fabry gene in the late 1980s. But as the new millennium dawned, the days when scientists thought more about publication than patenting were long over. Across the country, companies and academic researchers ran gene sequencing machines overtime as they sought to stake their claims to the human genome. It was a great gene grab, the fencing of the human commons. And it was all made possible by machines whose invention—the subject of the next chapter—had come at taxpayer expense.