Important note: Information in this article was accurate in April 2000. The state of the art may have changed since the publication date.Adefovir dipivoxil (trade name Preveon) is the first HIV antiviral drug that the Antiviral Drugs Advisory Committee of the Food and Drug Administration (FDA) did not recommend for accelerated approval. (The Committee deadlocked 4-4 on delavirdine, which the FDA subsequently approved.) Perhaps at an earlier time this same set of data would have been sufficient and the Committee's interpretation of its guidelines flexible enough for approval. But the environment is changing. An expanded arsenal of generally effective drugs has shifted the risk-benefit analysis of new drugs from "something is better than nothing" to a more demanding proof of safety and efficacy. In the eyes of the Committee, adefovir fell short on both counts.
It was clear from the discussion that no Committee member saw adefovir as a blockbuster drug for treating HIV. They likewise supported the notion that a less than scintillating therapy might be approved to fill the niche of treating heavily experienced patients with few remaining options. Most Committee members also believed that the side effects are manageable. But, in the end, they felt that the trial designs were so flawed and the proof of efficacy so questionable that the drug did not earn approval on the basis of the data presented.
Product sponsor Gilead Sciences (Foster City, California) listened closely to recommendations for additional data that could lead to approval. It initially vowed to work closely with the FDA to resolve outstanding issues and gain approval for the drug. But, on December 3, Gilead announced that it was stopping development of the drug for HIV infection. Adefovir will continue to be available to patients already in trials or the expanded access program, and development of adefovir for hepatitis B infection will continue.
The Advisory Committee meeting was held November 1, 1999, in Gaithersburg, Maryland. Heidi Jolson, MD, MPH, director of the FDA's Division of Antiviral Drug Products, opened the session by commending Gilead for conducting investigations in treatment-experienced patients, for its ongoing commitment to expanded access, and for developing a sizable database on viral resistance.
"The development of adefovir raised several challenges," Jolson said. "The sponsor elected to pursue marketing a dose that is lower than that studied in most of the phase III development, because of dose-limiting nephrotoxicity."
She reminded the Committee that accelerated approval is based on surrogate or clinical endpoints rather than irreversible morbidity or mortality. Those data must come from two adequate trials at least 24 weeks long, and the sponsor must have a program in place that can lead to full approval based on 48-week data. There should be sufficient study of drug-drug interaction to assure the safety and continued efficacy of all components of a combination therapy. "However," Jolson stressed, "accelerated approval does not change the standards of evidence for efficacy."
Nephrotoxicity that developed in the initial phase III trial of a 120-mg dose led Gilead to consult with the FDA and the Committee. The agency required Gilead to provide 48-week safety data, and when the sponsor decided to reduce the dose from 120 mg to 60 mg, the FDA said Gilead had to "conclusively establish" the "comparability" of the two dosing regimens. That could be demonstrated "by providing either data that the pharmacokinetic difference isn't clinically relevant, or clinical trial data that demonstrate comparability or clinical benefit with the new product," Jolson said.
The FDA is well-versed in issues of comparability, she affirmed. But "we have had very limited opportunity to consider an equivalence trial as the basis of approval of an antiretroviral in the complex setting of combination drug therapy in treatment-experienced individuals."
Jolson posed three key questions for the Committee to consider. Is the trial capable of distinguishing between active and inactive treatment? "There is no agreed upon statistical convention for an acceptable margin of inferiority," Jolson said. "These are matters of clinical judgment considered on a case-by-case basis." She reminded the Committee that the design of equivalency trials "may make treatment arms look more similar." Compliance, dropout, loss to follow-up, overall poor response to the trial drug, and the impact of other drugs in a combination therapy may all play a greater than normal role in comparability trials.
Finally, Gilead requested an adefovir indication "for patients with prior nucleoside analog treatment experience based on safety considerations." Jolson asked the Committee to consider "risk and benefit" in light of both the label indication being sought and for likely off-label use in clinical practice.
"The only anti-HIV therapies were nucleoside RT inhibitors," explained Gilead Sciences presenter Howard Jaffe, MD, in setting the context for design of adefovir's pivotal trial and its initiation in 1994. Now, with several protease inhibitors and nonnucleosides approved, he said that failure of an initial HAART regimen led to progressively quicker failures of succeeding regimens. As a result, he concluded, "there is an urgent need for new treatment options. These are the patients for whom adefovir dipivoxil is intended."
Adefovir dipivoxil is the first drug in the class of nucleotide analog reverse transcriptase (RT) inhibitors. It can be taken once daily without dietary restrictions. Jaffe claimed that adefovir has "a unique resistance profile with activity against HIV resistant to 3TC [lamivudine], as well as virus that combines AZT [zidovudine] and 3TC resistance." He asserted that the initial use of adefovir is unlikely to lead to the development of resistance, a trait that keeps future treatment options open.
"The risk of nephrotoxicity has been well characterized," Jaffe said. "It can be monitored with monthly routine lab tests, and when it does occur, it is largely reversible with drug discontinuation."
Jay Toole, MD, PhD, focused on Gilead's trials, which found little evidence of drug-drug interactions. Phase I trials showed "asymptomatic decreases in serum creatinine," which led Gilead to add a supplement in later studies.
Study 408 was a pivotal phase III trial of the 120-mg dose. It enrolled 442 heavily pretreated patients (a median of 3.5 years) with a mean viral load of 30,000 HIV RNA copies/mL and a mean CD4+ count of 370 cells/mm3. Patients were randomized in a double-blind fashion to add adefovir or placebo to their existing regimen for 24 weeks, followed by an open-label phase. At week 24 the adefovir arm showed a 0.4-log reduction in viral load from baseline. That change was consistently observed across all patient characteristics and combination regimens.
Under Committee questioning, Toole later admitted that Gilead discouraged but allowed new antiretroviral agents to be added during the 24-week trial. About 20 percent of patients did so. "For those that did not add an agent, there was still a decrease of about 0.3 logs after 24 weeks," he said.
Nephrotoxicity, "a Fanconi-like syndrome" of proximal renal tubular dysfunction, emerged in 1 percent of patients after week 28. This grew with time to 50 percent of those who remained on adefovir at week 80. Toole said the problem resolves in more than 95 percent of patients when they stop taking adefovir. The median time to resolution was 17 weeks.
CPCRA 039 was designed to evaluate adefovir or placebo added to background therapy in patients with CD4+ counts below 100 cells/mm3. Only 500 of the planned 2200 patients were enrolled before the trial was discontinued in August 1998, because the introduction of protease inhibitors significantly altered conditions. While this aborted trial did not contribute to an understanding of adefovir's efficacy, it did add to the safety database. Study 420 randomized 47 treatment-naive patients with a viral load above 5000 copies/mL and more than 150 CD4+ cells/mm3. The placebo-controlled study used a 60-mg dose of adefovir as monotherapy. After four week, adefovir induced a 0.3-log decline in viral load.
Study 417 was designed to establish equivalency between the 120-mg and 60-mg doses of adefovir. The 214 patients had to be PI naive and had to have at least four weeks of RT experience, a viral load above 5000 copies/mL, and a CD4+ count above 100 cells/mm3. Patients were further randomized to nelfinavir plus saquinavir, nelfinavir plus one nucleoside RT inhibitor, or saquinavir plus one nucleoside. "Many patients at both doses discontinued after week 24 due to insufficient viral load suppression," Toole said. Yet "equivalence [of 60 mg and 120 mg] was established" by three different methods of analysis.
Norbert Bischofberger, PhD, presented the resistance data. In vitro, adefovir selects for the K65R and K70E RT mutations. The first is associated with resistance to didanosine (ddI), zalcitabine (ddC), and 3TC, while the second is unique to adefovir. Examination of 219 patients who had been on adefovir for five to 12 months revealed no K65R mutations and only two K70E mutations.
"The only viruses that showed reduced susceptibility to adefovir are viruses with high-level resistance to AZT and viruses that contain the multinucleoside insertion," he said. "Viruses that have the K65R mutation in the presence of the 3TC resistance mutation [M184V] all have wild-type susceptibility to adefovir."
FDA review officer Kimberly Struble, PharmD, reminded the Committee that Gilead's dosing trials of 125 mg, 250 mg, and 500 mg "showed similar antiviral activity without evidence of a dose response. However, a dose response was apparent for GI toxicity."
Struble questioned the efficacy of adefovir even at the 120-mg dose. She noted that in the pivotal Study 408, when patients who added a new therapy were excluded from the analysis, there was a 0.28-log viral load reduction in the adefovir arm compared with a 0.06-log reduction in the placebo arm at week 24. There was "considerable overlap" of the viral load plots for adefovir and placebo (Figure 1), with "no statistically significant differences" in the number of patients whose viral load was suppressed below 400 copies/mL. Nor was there a statistically significant difference between the two arms in CD4+ counts (+3 cells/mm3 for adefovir and –5 cells/mm3 for placebo).
 |
|---|
| A comparison of viral load changes in Study 408 participants showed scant difference between the adefovir and placebo arms, according to this daily average (DAVG) analysis by the FDA's Kimberly Struble, PharmD. |
Next the FDA reviewer expressed concern that Gilead submitted only the executive summary of CPCRA 039; thus the agency was not able to review this trial in depth. Struble summarized data from this aborted trial: "Overall there were no differences between adefovir and placebo with respect to HIV RNA or CD4+ cell count."
Similar problems plagued the information submitted for ACTG 359, a double-blind, six-arm salvage study involving different combinations of adefovir, delavirdine, and saquinavir for people whose protease inhibitor regimen had failed. Struble said that the study demonstrated the efficacy of delavirdine but "was unable to demonstrate activity of adefovir in combination with PI-based regimens in treatment-experienced patients." ACTG 359 also raised the specter of a drug interaction between adefovir and saquinavir, a question not raised by other trials. Struble warned that "single-dose studies [as conducted by Gilead] may not be sufficient to assess interactions that may be arising from metabolic induction."
At least one renal-related laboratory abnormality developed in 61 percent of patients taking 120 mg of adefovir. Struble agreed that in most cases these problems were reversible. However, eight patients required dialysis, and continued phosphate wasting may increase the risk of bone fractures. She was also troubled by the small long-term safety database, which included only 73 patients who had received adefovir for 48 weeks or more. "After study day 100," she pointed out, "there is a dramatic increase in the proportion of patients discontinuing from therapy" (Figure 2).
 |
|---|
| The proportion of people who kept taking either 120 mg or 60 mg of adefovir started a steep slide after 100 days, observed the FDA's Kimberly Struble, PharmD. |
Struble said that the lower dose seems to offer a delayed onset of toxicity, but given enough time, the 60-mg dose results in similar levels of renal abnormalities as 120 mg. There were no statistically significant differences between the two doses in recovery from toxicities.
She was skeptical of Gilead's claims of unique action against resistant strains of virus. The presence or absence of the position 184 mutation did not correlate with viral load in either the adefovir or placebo arms. Furthermore, she added, "patients with a high level of AZT resistance alone will demonstrate cross-resistance to adefovir."
Struble concluded that trial data did not support the indication Gilead sought. "Although Gilead has requested approval of 60 mg in treatment of nucleoside-experienced patients," she said, "the more compelling results were found in treatment-naive patients."
During the public comment portion of the meeting, more than a dozen clinicians and patients paraded to the microphone to make brief comments on their experience with adefovir and urge approval. All began by disclosing that Gilead had paid their expenses. While such testimony is common, this display was several times the norm for the Committee.
Committee member Ram Yogev, MD, director of pediatric and maternal HIV infection at Children's Memorial Hospital in Chicago, noted that the personal testimony often involved giving adefovir with far more drugs, so-called mega-HAART, than the number used in Gilead's clinical trials. That made it even harder to tease out the effect of a single drug.
"I am the only person here who has received no financial support" from Gilead, began Michael Marco during the public hearing. Speaking on behalf of the Treatment Action Group, which submitted a tightly reasoned eight-page evaluation of the data, he cited the federally funded CPCRA and ACTG studies, which showed no treatment effect. "I am one who believes that federally funded studies usually yield fairer results than industry-sponsored studies," Marco said.
He criticized the 60-mg and expanded-access data as being "riddled with dropouts." And he noted that of the more than 9000 patients who have taken adefovir, there are 48-week data for only 73 of them. "It is unfortunate," he concluded, "but as current data has shown, we do not believe [adefovir] is effective or safe" for the requested indication.
The FDA posed questions for the Committee to evaluate concerning efficacy of the 120-mg dose, comparability of the 60-mg dose, safety and management of associated nephrotoxicity, and--if adefovir was found lacking--what further effort would be necessary to gain approval. Discussion of these questions wound inexorably toward an ultimate vote of 13-1 against approval.
Opinions were mixed as to the efficacy of the larger dose. The lone vote for approval came from W. Christopher Mathews MD, MSPH, director of the Owen Clinic & AIDS Service and professor of medicine at the University of California at San Diego Medical Center. It was a vote that seemed to be based as much on philosophy as on data.
"Is there a population in need of an agent with this drug's characteristics?" Mathews asked. "I think there very definitely is a need. Physicians are scrambling like wild to find agents with some activity that can be components of salvage regimens." Study 408 left "little doubt in my mind that this drug has modest activity, of the same order of magnitude of agents that were previously approved."
Brian Wong, MD, chief of infectious diseases for the Veteran's Administration Health Care System in Connecticut, was convinced that the 120-mg dose is effective. But he said that comparability of the 60-mg dose had not been demonstrated, and perhaps could not have been because of design problems with the trial.
James J. Lipsky, MD, director of clinical pharmacology at the Mayo Clinic, conceded that, for the 120-mg dose, "there is suggestive evidence that it can produce some degree of viral suppression." But Judith Feinberg, MD, said that Study 408 demonstrated "at best marginally clinically significant" results. She is a professor of medicine at the University of Cincinnati Medical Center.
Wafaa El-Sadr, MD, MPH, expressed concern that even the most favorable trial did not show a statistically significant rise in CD4+ count. Furthermore, she added "we all know that in expanded access [the greatest source of patient data for this application], people who stop taking the drug are lost to follow-up most of the time. There is always a selection bias; you follow the people who have done well." The director of the division of infectious diseases at Harlem Hospital Center was among those pushing for further trials. "There is a dire need for more data," El-Sadr maintained. "I think the way to get it is to compare 60 mg to placebo."
"I'm trying to figure out why 60 mg is the dose of choice," said Feinberg. She was puzzled that adefovir had never shown an antiviral dose response effect at the doses evaluated. Furthermore, Gilead's ongoing hepatitis B trials are using only 30 mg of adefovir. "What is it that this drug does? And why is this the dose?" she asked. Gilead's Jay Toole said that the company was committed to answering those questions in postmarketing studies.
Feinberg urged study of both 60-mg and 30-mg doses "in appropriately controlled, double-blind fashion." That should probably include two to four weeks of study as monotherapy "to know what bang you are getting for this buck, especially since potentially this buck is going to buy you a lot of nephrotoxicity."
"I was taken back by the amount of missing data; 22 to 32 percent in small studies is surprising," said Roger J. Pomerantz, MD, director of the Center for Human Virology at Thomas Jefferson University. "I don't know where those people went, but that is problematic with me." He was concerned that the modest 0.3-log viral load effect could well disappear if the analysis included even a few of the missing patients.
Committee member Jeffery B. Kopp, MD, a senior investigator in kidney disease at NIH, offered a review of creatinine. He explained that the kidneys have a significant reserve or redundancy, a fact that allows for surgical donation of one of those organs. Serum creatinine does not rise until 60 to 70 percent of total kidney function is lost. Thus, "relatively modest changes in creatinine represent significant degradation of filtration." Serum normalization may mask the fact that half of kidney function remains impaired. It narrows the margin for using other drugs that stress the kidneys.
James Lipsky saw that "the inflection points are about the same" on the curves of time to nephrotoxicity for both doses of adefovir. He "wondered if the exposure of some of the patients in the lower-dose group is not comparable to the exposure in some of the patients in the higher-dose group. Perhaps therapeutic monitoring could help avoid toxicity with this agent."
"If you look at the curve [of incidence of nephrotoxicity], it doesn't level off at 48 weeks," said Ram Yogev. "So we need much longer data."
"My experience with adefovir was not quite as cheerful" as that expressed by the public witnesses, said Judith Feinberg. "I have not seen ready reversibility" of nephrotoxicity. "I do not think that the safety profile is adequately characterized. It is frightening to me to see that the number of patients on the 60-mg dose for 48 weeks totals 73 people, 30 in a randomized trial and 43 of the first 1000 in expanded access. I am very anxious about how thin that data set was."
"It is clear from the FDA presentation," Feinberg added, "that there is a subset of patients for whom monthly monitoring of electrolytes is going to be inadequate." Her concern was to understand better which patient populations are less likely to recover and thus will need closer monitoring for signs of nephrotoxicity.
Kopp praised Gilead's management plan for elevated creatinine levels with supplementation as "quite well thought out." But Joseph S. Bertino, Jr, PharmD, from the Clinical Pharmacology Research Center, was less sure. He wanted to see data on how recommended supplementation with carnitine affects drug efficacy. "We don't have any information on that."
Community representative Jeffrey Schouten, MD, was troubled by Gilead's suggestion that elevated creatinine levels might be addressed by dose reduction of adefovir. No data supported that strategy. He would have preferred to see discontinuation rather than lowering the dose of adefovir.
Wafaa El-Sadr was concerned that dose reduction in a clinical setting, absent data from trials, might contribute to the emergence of drug-resistant virus. Kopp joined the chorus of those advising against dose lowering pending validation of that approach through a clinical trial.
Committee Chairman Scott M. Hammer, MD, from Columbia Presbyterian Medical Center in Manhattan, called reversibility of nephrotoxicity and the totality of recovery over time "an open question," but a risk-benefit that can be managed. Schouten pointed to the extensive expanded access program as evidence "that patients and physicians are managing this drug. I would encourage people not to be too paternalistic."
Christopher Mathews was satisfied with the available safety data because he did not expect most of his patients to take adefovir for more than a year. He said he saw adefovir "as a bridge agent for people who need something in the short run, who may not have the option of waiting for two years until the next generation comes along."
"In a situation where there is a therapeutic index question," Lipsky suggested to Gilead, "one would want to know very clearly the dose-response relationship of toxicity to the dose-response relationship of efficacy so as to maximize the therapeutic benefit."
Bertino pushed further. "It is becoming pretty clear that HIV patients are not the same as normal volunteers," he noted. "One speculation is that it is because of cytokine production. So if you are going to do drug interaction studies, there may be a need to stratify by viral load, or actually measure cytokines. Because pharmacokinetics and the potential for drug interactions may change over time as viral load drops." "The company has to decide what they want from this drug," said Roger Pomerantz. "Do they want it to be an up-front drug that is used by many patients who are naive? Do they want it to be for people in their first salvage? Or do they want to try to find a niche where certain patients will get help with a drug when all else has failed? I think they have to decide where their position this."
A month later Roger Pomerantz had his answer. Gilead issued a press release on December 3 announcing termination of US development of adefovir as an anti-HIV agent. An application for approval is still pending in Europe.
US clinical trials of adefovir as an antiretroviral will not enroll new patients, nor will new patients be accepted into the expanded access program. "We will continue to treat patients receiving drug for as long as they and their physicians feel that they are getting benefit from the program," explained Gilead spokesperson Sheryl Meredith.
Why did Gilead retreat? "In subsequent conversations with the FDA," Meredith said, "it was determined that they wanted longer-term placebo-controlled trials of the 60-mg dose. Gilead has decided to focus on other products in our pipeline."
Tenofovir disoproxil fumarate (TDF, formally bis-POC-PMPA) is a second-generation nucleotide analog, a follow-on drug to adefovir, now in phase II/III clinical trials. Preliminary 24-week data presented at ICAAC in September 1999 indicated a 0.75-log decrease in plasma HIV RNA at the highest dose (300 mg once a day) when added to the existing regimen of highly pretreated patients who had detectable levels of viremia. However, as with adefovir, there was no associated rise in CD4+ count in those patients.
Genotypic analysis presented in a poster [740A] at the 7th Conference on Retroviruses and Opportunistic Infections in February 2000 showed that tenofovir is active against virus with the codon 151 multidrug-resistant mutation. A week 24 analysis of 121 patients showed no indication of novel RT resistance mutations to tenofovir. There was no serious increase in serum creatinine levels. Nor was there any sign of nephrotoxicity in the 71 patients who had completed 48 weeks of treatment with the drug.
A phase III trial is currently enrolling 600 patients at as many as 70 sites in the US, Europe, and Australia. Entry requirements for the 48-week study include an HIV RNA between 400 and 10,000 copies/mL and a regimen of no more than three antiretroviral drugs that has been stable for at least eight weeks before entry.
Tenofovir is also available through a compassionate use protocol for up to 300 patients with CD4+ counts below 50 cells/mm3 and viral loads above 10,000 copies/mL. Physicians may contact Gilead Sciences about both the trial and compassionate use at 1-800-GILEAD-5.
Bob Roehr is a medical writer based in Washington, DC (BobRoehr@aol.com).
000410
IA000402
Copyright © 2000 - International Association of Physicians in AIDS Care. All rights reserved. http://www.iapac.org
ÆGiS is made possible through unrestricted grants from Roxane Laboratories, Inc., iMetrikus, Inc., the National Library of Medicine, and donations from users like you. Always watch for outdated information. This article first appeared in 2000. This material is designed to support, not replace, the relationship that exists between you and your doctor.
ÆGiS presents published material, reprinted with permission and neither endorses nor opposes any material. All information contained on this website, including information relating to health conditions, products, and treatments, is for informational purposes only. It is often presented in summary or aggregate form. It is not meant to be a substitute for the advice provided by your own physician or other medical professionals. Always discuss treatment options with a doctor who specializes in treating HIV.
Copyright ©1990, 2000. ÆGiS & the Sisters of Saint Elizabeth of Hungary. All materials appearing on ÆGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of ÆGIS and the Sisters of Saint. Elizabeth of Hungary, or the party credited as the provider of the content.