Heiko Jessen was shocked at the headline. THE FIRST AIDS CURE? screamed across the page of B.Z., a tabloid in Berlin. Inside, the pictures were ridiculous. They featured a man pretending to be the Berlin patient, a medical mask covering his face and a surgical cap pulled low. Pictures of Jessen treating a patient, as well as pictures from outside his clinic door, were featured prominently. The article went on to describe the case of Christian, describing him as the Berlin patient and pointing to his remarkable cure. Jessen was not amused. He had carefully avoided using the cure word and felt the tabloid had sensationalized the story. The past year had brought a blizzard of publicity for the young family doctor. He had spoken with countless news outlets, including The New York Times and Newsweek. It was this last interview that caused the most tension among his fellow authors of the New England Journal of Medicine paper.
Originally, the interview with Newsweek was supposed to be a full page, enough space to mention the many collaborators on the project and especially the new institute that Julianna Lisziewicz and Franco Lori, senior author of the paper, had formed a few years earlier. New research institute funding was tight. Most institutes must rely heavily on donations from the private sector, beyond their portion of federal funding. With their Berlin patient, Lisziewicz and Lori had a nice opportunity to raise the profile and capital of their fledgling institute. They pressed Jessen to mention the institute during his interview with Newsweek, and Jessen was happy to oblige, mentioning all his collaborators, especially noting the role of Lori and Lisziewicz’s institute.
Unfortunately, before the article went to press, there was a bigger news story afoot. The Kosovo War had gained momentum, and page space formerly dedicated to Jessen and the Berlin patient now had to be reallocated. What was a full-page article became a single paragraph, a paragraph that did not mention any collaborators. The effect was crushing on what had once been a close relationship. Conversations devolved into angry shouting matches. The friendship that Jessen and Lisziewicz had once shared, a friendship that had resulted in the Berlin patient, was permanently broken.
This fracture would have severe consequences for taking hydroxyurea therapy to the next level. The team was now on shaky ground going to clinical trial. In the wake of the New England Journal of Medicine paper, it seemed that each scientist had a different vision for how to translate the success of the Berlin patient into a viable therapy.
Bruce Walker believed the therapy itself was inconsequential; what mattered was its timing. If they could identify patients soon after becoming infected, before they displayed symptoms, and hit the virus with a heavy dose of antiviral drugs, they could potentially knock it down. Then when they stopped the drugs, the immune system would be at the ready to fight before the virus could get back on its feet. Walker’s thinking on this came from a tiny group of acutely infected patients he and his colleague, Eric Rosenberg, were treating at Massachusetts General Hospital. They had been able to identify three people early in infection, before they displayed symptoms, and treated them aggressively with antiviral drugs. They drew blood from the patients, both before and after treatment, separated out the white blood cells, and stimulated them with purified pieces of HIV. They measured the proliferation of HIV-specific T cell responses, particularly the commanders. When they compared the data from these patients to elite controllers and patients infected with HIV for decades (chronic HIV), they found that the ability of the commander T cells to resist HIV in the acute group was as high as that of the elite controllers, both of which were far higher than chronically infected patients.
When they plotted this data and corresponded it to the amount of virus each person had in his blood, the data formed a perfect curve. The level of response from the commander T cells corresponded perfectly with the amount of virus. The higher the HIV-specific T cell response, the lower the virus. It made intuitive sense that the early therapy had somehow protected these critical cells necessary for the immune system to do its job. The data had some problems, though. Walker lacked a true control group; he hadn’t found any newly diagnosed patient who didn’t want to start therapy. Therefore, he couldn’t directly compare acute patients who took therapy with those who didn’t. Regardless, his data were convincing. He published his findings in Science in 1997. Even then, he didn’t attribute much to the specific drugs given to each patient; they’re not even mentioned in the paper. What was important was the timing.
The next step was obvious. They needed to stop therapy in patients who had been given this early, aggressive treatment. The problem was, there was no way to ethically do so. Walker knew that HIV patients taken off therapy could die. The researchers had no way to know if the strong T cell responses they were measuring were enough to protect the patients. Then the Berlin patient came along. He was the answer to their prayers. No one had documented an HIV patient in remission before. They now had direct proof that a patient identified early and given aggressive treatment, and who had stopped therapy, could control the virus. Even better, the Berlin patient’s T cell responses were incredibly high. His T cells were clearly protecting him from the hidden virus lurking in his body. Walker could take his research to the next level; they would take patients treated in acute infection off therapy. They would, of course, watch them closely, making sure the virus didn’t return. It could be done safely, he argued, as long as they were tested weekly for HIV. The minute the virus came back, they could be restarted on therapy. Walker was not alone. Other HIV researchers were pursuing similar lines of inquiry, all stemming from David Ho’s original campaign for early therapy back in 1995. Where Walker was different was in his unparalleled ability to measure HIV-specific responses of the immune system and compare them to those special patients whose bodies controlled the virus without therapy. Now, with the Berlin patient, the approach made perfect sense; it was surefire.
Lisziewicz, Lori, and Jessen, on the other hand, believed that the hydroxyurea the Berlin patient had taken was the main reason why his body was now able to control the virus. Hydroxyurea was a unique drug. It targeted the ability of the cell to work instead of inhibiting HIV’s enzymes. Hydroxyurea put a wrench in the machinery that created new building blocks of DNA, making a perfect space for fake building blocks, such as the drug didanosine, to infiltrate the viral genetic code. It also froze dividing cells, keeping the virus from establishing a foothold. The only drawback to hydroxyurea was its toxicity. Jessen felt strongly about the safety profile. For the trial in his clinic, he had used half the full dose recommended for cancer patients. Jessen had strong opinions on how a trial of hydroxyurea should be designed in order to minimize the toxicity of the drug.
Unfortunately, as the trial came to fruition, it was clear that Jessen’s concerns would not be taken into account. Several small trials had mimicked Jessen’s safety profile, each finding favorable results, although none included patients who had stopped therapy. It was impossible to know if the results would be similar to the Berlin patient’s.
To know if the Berlin patient’s experience could be replicated for other HIV patients, researchers needed to perform a large-scale trial of the drug, using a similar schedule of early treatment followed by what’s called treatment interruption, a professional-sounding way of saying that patients stop taking therapy, “interrupting” their treatment. These interruptions would become known informally as drug holidays.
Performing such a large-scale trial is expensive. Lisziewicz and Lori partnered with the makers of hydroxyurea, Bristol-Myers Squibb, to get the necessary funding and infrastructure. As you can imagine, Bristol-Myers Squibb was very excited about the way the Berlin patient had catapulted hydroxyurea into the spotlight. They quickly began a trial through the Acute HIV Trials Group (ACTG), called ACTG 5025. The trial tested the same three drugs the Berlin patient received: hydroxyurea, didanosine, and indinavir. The doses, however, while the same, were given on a different schedule. Jessen had given Christian 400 milligrams of hydroxyurea three times daily, carefully balancing the doses with a strict eating schedule designed to promote absorption of the drug and limit toxicity, whereas the Bristol trial ignored these safety concerns and gave a single daily dose of 1200 milligrams. The reason for this was simple: It’s difficult to adhere to a strict schedule as Christian did. Even more challenging is finding a large number of patients capable of following such a schedule. The antiviral drugs present enough of a challenge without adding another scheduled drug. If they enrolled patients incapable of taking the strict regimen, they risked getting no data from the trial. As it was, the trial ended up having worse consequences than a lack of data. It killed two people.
The trial enrolled 202 people, with a goal of 399. Volunteers were not newly infected with HIV. These were deemed too difficult to find and test, requiring a network of clinics like Jessen’s. Instead, ACTG 5025 would recruit chronically infected patients already involved in clinical trials for antiviral drugs. Two patients in the hydroxyurea arm died of pancreatitis. Numerous toxicities were reported, including pancreas, liver, and nervous system damage. The trial was halted. Toxicity was now indelibly associated with hydroxyurea in HIV treatment.
Bristol, however, was not slowed down. At the 39th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) held in San Francisco in September 1999, the corporation used special sessions to promote the use of two of their branded hydroxyurea drugs for treatment of HIV. This flew in the face of FDA regulations that prohibit pharmaceutical companies from promoting off-label uses of their drugs. The labels on both of the company’s hydroxyurea drugs specifically state that it’s useful in treating several cancers, not HIV. At the meeting, they showed data from the Berlin patient, among others, stating that hydroxyurea had been proven as a first-line therapy for HIV. It was a bold, some might say outrageous, move considering that only four days earlier, they had been given notice that the ACTG 5025 trial was terminated. The corporation also listed suggested doses, including a 1200-milligram daily dose. They did not mention the two deaths that had occurred during their trial. In response, the FDA sent a warning letter, asking Bristol to stop using this language in their promotions, to send “Dear healthcare provider” letters to warn physicians of the potential danger of pancreatitis, and to bolster the warnings on the didanosine package insert.
Despite the mounting safety concerns over hydroxyurea, Lisziewicz, Lori, and Jessen were excited to take the lessons they had learned from the Berlin patient and translate them into a new clinical trial. To get enough funding , they turned to Bristol-Myers Squibb again. In particular, they would need the company to donate drugs for their study. As Lori, Lisziewicz, and Jessen were now pioneers in the use of hydroxyurea during HIV infection, receiving this support was not difficult. It did, however, come with a catch. The trial needed two Bristol drugs to replicate Christian’s experience: hydroxyurea and didanosine. Bristol wanted to add a third drug they owned to the proposed clinical trial. This new drug was one that, unlike hydroxyurea, was new on the market. Jessen was convinced that this addition came because hydroxyurea had already been on the market for thirty years and so there were limited profits to be gained from it. A newly patented drug is able to make far more profit than one whose patent has already expired. Indeed, the new drug from Bristol, trademarked as Zerit, brought in sales of $605 million in 1999 alone.
But Zerit, despite its shiny new patent, was not really a new drug at all. In 1966, two years after Jerome Horwitz published data on his failed compound, AZT, he published the methods used to make a similar compound that he called d4T, designed to work the same way AZT worked. It imitates the DNA building block called thymidine. Like a ladder with a faulty rung, the drug sneakily incorporates itself into a growing chain of viral DNA but it’s altered so that the next DNA base can’t attach itself to it. It cuts off the virus from copying itself, thereby protecting more cells from becoming infected. Of course no one knew that d4T would make a potent antiviral drug until two pharmacology professors at Yale University, named William Prusoff and Tai-Shun Lin, dusted off the old compound and tested it against the virus in the early 1990s. The university patented use of d4T for treating HIV, then licensed the drug to Bristol. Bristol conducted several clinical trials of d4T, which was approved by the FDA as a new drug in 1994. The approval raised controversy since it was done by a special procedure for life-threatening diseases that allows a drug to be approved before it has been proven to work. At the time, Deborah Cotton, a Harvard professor who helped approve the drug, said, “I’m not sure how good our advice was today.” She was referring not only to the effectiveness of the drug but also to its safety. Of the 10,000 people given d4T, 21 percent experienced neuropathy, a condition that causes pain and numbness, often in the hands and feet. Like AZT, d4T is toxic; the dose needed to be lowered to make it safer for people with HIV.
This is why, when Jessen learned that d4T, now trademarked as Zerit by Bristol, was being added to the clinical trial, he was not happy. “Here was a recipe for disaster,” he recalls thinking. Hydroxyurea was already a highly toxic drug, but to add d4T was perilous. As Jessen looked around him, he realized there were few clinicians designing this trial. Who was looking after the patients’ interests? He couldn’t condone such an experiment. With a heavy heart, he left the group, disappointed at how things had fallen apart. It seemed to him that the collaborators he once stood in awe of had fallen prey to fiscal motivations.
It’s surprising that following the Berlin patient, the first HIV patient to be in remission, no single investigator was trying to duplicate his unique therapy in a clinical trial. Instead, the treatment Christian had been given was broken up into two different clinical trials. Bruce Walker and his collaborators were testing one unique component of his treatment: giving antiviral drugs aggressively during acute HIV infection. Franco Lori and his collaborators were testing the other unique component: giving hydroxyurea, d4T, and didanosine to chronically infected HIV patients. It was timing versus aggression. In all the world of HIV research, no one was combining these unique components in the hope of replicating the Berlin patient’s cure.
Sadly, neither of the early trials based on the first Berlin patient’s therapy went well. Initially, Walker’s data seemed remarkable. Published in Nature in 2000, Walker and his collaborators identified 16 HIV patients early in infection. All the patients were immediately started on antiviral drugs, most within 72 hours following diagnosis. The antiviral drugs did not include hydroxyurea. Twice because of hospitalizations, Christian had to go off therapy. Although they weren’t planned, these two drug holidays, or treatment interruptions, were intriguing. Walker postulated that the interruptions were training the immune system to recognize the virus. That is, they gave the commander and storm trooper T cells an advance peek at the enemy. With this peek, they could tailor their attack for the virus. Then, once therapy was restarted, the cells were protected, ready for the next fight. If enough HIV-specific T cells were spared from enough rounds of treatment interruptions, they could provide a potent force against the virus. Those receiving treatment interruptions would then become like elite controllers—infected with HIV but able to control it. The strategy was aiming for a functional cure. Although the virus might not be eradicated in these patients, they would be able to live like Christian, free from having to take any antiviral drugs or worry about the virus.
Similar to how the Berlin patient underwent two treatment interruptions, 8 of Walker’s subjects took one to two scheduled breaks. With the breaks scheduled, the trial was almost the opposite of David Ho’s “hit hard, hit early” strategy, fashionable only a few years earlier. Of the 8 patients receiving treatment interruptions, 5 remained off therapy for an average of 2.7 years, with virus undetectable, at less than 500 copies per milliliter of blood. In addition, the HIV-specific T cell responses rose significantly. The results were impressive; the virus was not coming back. In his paper, Walker compared them to an acute HIV cohort not receiving antiviral therapy. In this group, similar to an experimental control, of 109 people, only 4 had less than 500 copies of virus per milliliter of blood at 2.5 years. The results fit perfectly with Walker’s hypothesis—the HIV-specific T cell responses were high, and the virus was low.
The research immediately garnered media attention. Here, it seemed, was the answer to curing HIV. And the answer was so simple: Just stop therapy a few times. It was easy enough that anyone could do it. And people did. The popularity of the drug holiday rose—a well-deserved break from the strict schedules and handfuls of pills. Acute, chronic, young, and old all tried drug holidays, sometimes without telling their doctors. In 2001, Timothy Brown, then a translator in Berlin, took a drug holiday. He couldn’t know that the holiday was inspired by the Berlin patient, a title he would eventually share.
The problem was that the drug holidays didn’t work. In fact, they hurt. Following his drug holiday, the amount of commander T cells in Timothy’s blood dropped to 250 cells per microliter, right on the edge of an AIDS diagnosis. The course of the virus varied from person to person, remaining hidden in some people for days, while in others, it lay quiet for weeks, months, even years. But the virus always returned. It turned out that even when the virus was hidden in the body, it was inflicting damage quietly. Those who took drug holidays experienced high levels of immune activation, a state in which the T cells, and often bystander cells that just happened to be in the wrong place at the wrong time, are overly stimulated, causing them to die. Even worse, in some patients, the HIV strains they now harbored were resistant to antiviral drugs. Similar to how, if you have a bacterial infection and don’t finish your antibiotics, you can develop antibiotic-resistant bacteria. In the HIV patients who took drug holidays, the virus had an advantage when reconfronted with antiviral drugs they had seen once before.
Treatment interruptions were a highly divisive issue in the HIV research community. In an interview in 1999, Anthony Fauci, head of the NIAID, was quick to question the safety of such interruptions, saying, “The strategy needs to be tested. The stop-and-go game can lead to drug resistance even if it looks so far like the wild type strain remains.” Indeed, until the mid-2000s, investigators were still arguing bitterly over the benefits and risks of treatment interruption. Then, one study changed this: The SMART study began in 2002, enrolling patients in 33 countries. It was the largest study of its kind, enrolling 5,472 patients with a goal of 6,000 before it was abruptly stopped in 2006 due to safety concerns. The SMART (Strategic Management of Antiretroviral Therapy) study found that those HIV patients receiving treatment interruptions were twice as likely to progress to AIDS. It was the final nail in the coffin of what had once been heralded as a new course for HIV therapy.
While Walker’s trial of treatment interruptions in newly infected patients soared and then plummeted, Lisziewicz and Lori were conducting a large-scale trial based on the other component of the Berlin patient’s remission: hydroxyurea. Jessen had already left the trial in protest of the inclusion of d4T, a drug he believed was too toxic to be safely used in the study.
Unlike Walker, Lisziewicz and Lori believed that hydroxyurea played a key role in the Berlin patient’s remission. Lisziewicz and Lori were invested in the drug’s success, believing hydroxyurea was able to target the hidden reservoirs of HIV. Therefore, their clinical trial ignored other aspects of Christian’s experience, such as treatment in acute infection and treatment interruptions, and focused on isolating the effect of hydroxyurea. The problem was that hydroxyurea had a tarnished reputation among researchers as well as patients.
The answer to this skepticism didn’t lie in another Bristol-sponsored study, which, surprisingly, included the same drug dosage as the failed ACTG 5025. By the time Lori and Lisziewicz’s study results were released, in 2005, hydroxyurea was a dirty word. This was especially unfortunate because Lori and Lisziewicz’s study made important strides in determining a safe dose for hydroxyurea; they found that cutting the dose in half, from 1200 milligrams daily to 600 milligrams daily, reduced toxicity while maintaining similar levels of virus reduction and increased T cells. Sadly, their study was riddled with problems. Participants taking the higher dose of hydroxyurea, the exact same conditions as the failed ACTG 5025, experienced adverse events, including one death from pancreatitis, the same cause responsible for the two deaths in the earlier hydroxyurea study. For many reading their paper, only one line stood out: “The results of the RIGHT 702 study described here confirm that the use of high-dose hydroxyurea (1200 mg daily) can be associated with fatal pancreatitis.” While other, smaller trials would pursue hydroxyurea as a therapy for HIV, none could surmount the reputation that hydroxyurea wasn’t safe. In a field reluctant to publish negative results, we can never really get at the heart of hydroxyurea.
Lisziewicz believes the problem lies deeper than the safety issues embedded in those early clinical trials. She remains frustrated with the financial aspects of bringing a drug to the market. She believes it’s too difficult to rouse interest in an older drug like hydroxyurea, which represents no major commercial interest and therefore no profit. Lisziewicz sums it up by saying, “If no one can make money, even the best drug in the world will fail.” Even though they are no longer friends, Jessen agrees with her take on the drug that failed before its time, wishing that “it wasn’t all about the money.”