Office of NIH History
In Their Own Words: NIH Researchers Recall the Early Years of AIDS
Previous Page | Next Page (6 of 8) Transcripts  

Harden: One more question along these lines. One of the great criticisms of the NIH is that everybody did not just stop and turn their entire efforts to AIDS, and that seems a very naïve statement to anybody who works here. But, from the point of view of the activist community, it appeared that a lot of people were very callous, they were only career-oriented and they did not care about the public health. But you were one of the people who did work on AIDS. Would you comment in general on whether you think people who did work on AIDS were more altruistic than the others? Or do you have any thoughts about all of this?

Yarchoan: We really reprogrammed our laboratory to work on HIV when the virus was discovered. My impression is that, until the virus was isolated, it was relatively hard for most scientists to get a good handle on how to attack AIDS. You could document how the immune system was going down, and you could show a number of epiphenomena. But people did not have a handle on the pathological root of the disease. If someone was working on cardiac metabolism or lipid metabolism, there would be no reason for them to switch over and do AIDS. They would not necessarily add much to it. And heart disease is also a public health problem.

I think a good number of people who were working in fields where their expertise could be turned to AIDS did it, and often they did it without getting extra money for it or because someone told them to do it. They did it because of the combination of reasons you do science–because you can help people, because opportunities exist to contribute something, and for a number of other reasons.

I think the irony is that I have heard the NIH also being critiqued for too many people working on AIDS in those days. So it seems that whatever we do, we get some criticism. But I do believe that if there are interesting scientific puzzles, a number of people will work on them if they have the tools to work on them; and that discovering the virus made the disease much more amenable to scientific work. In fact, if you look at the papers on HIV, there is first a smattering of papers of all sorts of quality, and as soon as the virus was discovered, you saw an explosion of papers because people could get a handle on how to study it. I am sure that if you graphed the number of scientific papers, there would be a steep break upward in the curve at about that time, or probably a year or so afterwards, as people geared up.

Harden: Now, I have wandered around in the questions. Do you want to go back to the Phase 2 AZT trials?

Hannaway: Yes. How did you get from the Phase 1 to the Phase 2 trials? You have some promising increases in CD4 counts?

Yarchoan: We really did not do the Phase 2 trial. Just to clarify what happened: we did the Phase 1 trial, and we did a number of extensions of it, and small pilot studies. Actually, when the protocol was first written, we had to take people off the drug after a short period of time. Then, when it became evident that this drug was doing something, we badgered the FDA and got permission just to continue people on the drug. We escalated up in doses, and, at a certain point, we felt we had doses that were fairly reliably increasing the CD4 counts of people. We had one patient who really made an impression. A nurse from New York had gotten AIDS through a blood transfusion and had a horrible fungal infection of her fingernail. When we gave her AZT, the infection cleared up, and you could start to see where the normal nail was starting to grow. That was very dramatic for us. She also had severe oral canker sores that cleared up. But to see this normal nail growing out at the start of this drug, that really convinced us that we were doing something.

Hannaway: Was this 1985?

Yarchoan: Yes, November-December of 1985. At that time, both we and the people at Burroughs to whom we were reporting the data were convinced that this drug was doing something. It was worth moving forward as quickly as possible. They felt that unless they showed that they could really improve the clinical lot of patients, that people would be arguing endlessly about whether this drug was worth giving to patients or not. I think that is correct, because even after the results of the Phase 2 trial were released, some scientists challenged us heatedly, saying that the immunologic changes induced by AZT were not enough to make any clinical difference. Some scientists and others still believe that this and related drugs don’t work in HIV infection.

For the Phase 2 trial, the placebo-controlled trial, we had a series of meetings discussing doses and trial design. And the dose that was picked was reasonably high. My understanding was that the people at Burroughs thought that it was better to be a little toxic but to have the drug work than not to have the drug work. They felt they really only had one shot to show that it was working, and so they designed this trial. They had a series of investigators, some of whom they had worked with, who they recruited to do the trial. This trial was then launched in January-February of that year.

Hannaway: This was in 12 centers.

Yarchoan: In 1986, in 12 centers. By September of that year, they had a data and safety monitoring report that looked at the trial data, and at that time they had 19 deaths on the placebo arm and one death on the AZT arm. It was highly statistically significant. It was felt to be unethical to continue the trial, and the trial was stopped. Everyone on the trial was offered AZT, and then, over the next few months, they collected the data and reported it to the FDA. I think it was approved in March of the following year.

Hannaway: Were you surprised by this outcome of the Phase 2 trials?

Yarchoan: No.

Hannaway: Was this a greater success than you had hoped?

Yarchoan: No. I was glad it was as clean as it was, but we were not surprised. The thing that we were struggling with at the time was how long AZT was going to work. We were continuing to follow these patients on AZT. What we were seeing was that the CD4 count was going up and then it was coming down. For me it was something like Flowers for Algernon. My best guess was that we were buying people 20 weeks with AZT.

Hannaway: You did not see this as the solution because you already knew long-term studies showed the CD4 count coming down.

Yarchoan: Right. So it was very distressing for us.

Hannaway: And this was really going on in parallel with this outcome of a Phase 2 trial.

Yarchoan: Right. So we were very pleased that it was working, but we were starting to see the limitations of it.

Harden: If AZT was interrupting the reverse transcriptase phase of viral replication, why did it stop working?

Yarchoan: Part of it is that the drug is toxic to lymphocytes, particularly at higher doses. But most of it is resistance. What happens is that HIV has reverse transcriptase. It is a very sloppy enzyme and it makes a lot of mistakes, and it does not correct these mistakes. By contrast, the human genetic replicative machinery is very accurate, and it has many mechanisms to correct mistakes. If you think of it, we have an enormous genome, and we just cannot tolerate that many mistakes.

The genome of the virus is about 10,000 base pairs long. It basically works on a brute-force mechanism. It produces all sorts of variants, and many of them are not infectious. That is just fine, as long as some of them are infectious. In fact, the ability to mutate turns out to be an advantage for the virus because mutations are selected in response to antibodies that are produced. It is one of its tools. So if you expose it to AZT, there are several mutations that can confer resistance to it. Most of these mutations exist before you give people the drug, just because the virus is so sloppy, Those particular variants now have a selective advantage, and over the course of six months, they can grow out. There are some drugs that can induce highly resistant virus in four weeks in patients.

Harden: So you were aware that the increase in CD4 counts was not going to last, that it was not going to buy more than about 20 weeks.

Yarchoan: But we did not know what the mechanism was. That did not come to light for a few years. But we knew that, for whatever reason, it was not working all that well for a long period of time.

Harden: At that point, were you thinking, then, about alternate drugs of the same type, or of seeing what you could do to combine AZT with other things? What were you thinking about?

Yarchoan: Actually, just taking you back a little–again, I may be off by a month– AZT’s activity was discovered in the laboratory around December, January, February of that year. But we initially did not know its structure, because it was sent under code. Also, during that time, Sam had been going around talking to everyone about what should we try, and he was interested in looking at some other nucleoside analogs. He had come upon a paper by Fermanski in which it had been shown that dideoxythymidine had some activity. He pushed Mitch, saying, “Why don’t you test this?” He would come back the next day and ask, “What are the results with this?” When Mitch looked at related compounds, including dideoxycitidine, dideoxydenicine, and dideoxyinosine, these all worked. Some of these were actually quite highly active.

Harden: What we would call ddI and ddA and ddC.

Yarchoan: And ddG was also one that worked but never got put into patients. These were compounds that, basically, we just ordered from chemical pharmaceutical houses, so they were completely devoid of any drug company support.

So while we were collaborating with Burroughs, Sam presented these results to the NCI and encouraged the parts of the NCI that were involved in drug development to help develop these drugs. Those NCI researchers made a decision to do so, and ddC was the next one chosen, because it seemed to be more straightforward than ddA or ddI. So while this work with AZT was going on, there was animal toxicity work being done within the institute on ddC.

Hannaway: These other drugs just did not have the track record that AZT had with the animal trials and other research already having been conducted.

Yarchoan: Yes. Burroughs had previously been looking at AZT because it had some antibacterial activity, so animal testing had already been done with it. These new ones were really starting from scratch. There was not a drug company involved, so the NCI did it, and the people did a very good job of moving that along quickly and finding out enough about it to get it into clinical trial. And that trial started maybe–the dates are a little bit more...

Hannaway: In 1986?

Yarchoan: Maybe the beginning of 1987, something like that, the end of 1986, the beginning of 1987. So we were moving on to the next drug. There was initially not so much the idea of combining drugs. We just hoped that the next one was going to be better.

There was then some interest in combining AZT with acyclovir, which is an anti-herpes drug and also another Burroughs drug. That drug had some very weak anti-HIV activity, and there was some evidence that combining it with AZT boosted the activity of AZT. We did a small trial of combining AZT and acyclovir. Probably, in retrospect, those effects were relatively minor, and that combination has never really become an established therapy.

But then we tested ddC in the clinic. It was very different from AZT in that, although the drug was eventually licensed to Hoffman-La Roche, it was really an NCI drug.

Hannaway: You found that it had different toxic effects and different positive effects?

Yarchoan: Yes, ddC is still somewhat of a puzzle. It had a fair amount of toxicity but was, in patients at least, extremely active against HIV. In fact, it was active at lower doses than what we would have anticipated. But the toxicity prevented us from going much higher on the doses, and it did not induce as many immunologic benefits as AZT except in doses that over a period of weeks wound up being toxic.

But by that time, Abbott had developed an assay for p24, and we started collaborating with them on looking at our samples with this assay. Dr. Jean-Pierre Allain was the person we collaborated with there. He has since, unfortunately, become embroiled in a controversy in France over the hemophilia issue there.

Hannaway: What did the assay with Abbott show?

Yarchoan: This was an assay looking at one of the viral proteins, the p24 gag protein. It is basically an assay for measuring it in the serum. This was, ironically, a better version of the assay that someone in Bob Gallo’s laboratory had been trying to develop and we had planned to use in the initial suramin trial. It took about two years for the assay to be developed in a way that was usable. The trial showed that ddC could drop the p24 level quite convincingly.

continued on Page 07


Previous Page | Next Page (6 of 8) Office of NIH History | NIH| DHHS