AIDS TREATMENT NEWS No. 112 - October 5, 1991
John S. James

The idea of approving ddI and ddC now did not begin with activists, however, but among leading physicians and researchers familiar with the drugs. For months, a consensus has been building among the researchers that both drugs would ultimately be approved anyway; why not, then, consider approving them now? What is new is that activists are taking this idea, which might otherwise have remained as conversation at scientific meetings, and placing it on the national agenda.

The issue of ddI and ddC is vitally important because there are tens of thousands of people unable to use AZT, or no longer able to benefit from it. In addition, among gay men, most HIV infections occurred between 1979 and 1985[1], before the cause of AIDS and ways to prevent transmission were widely known; since the median time from infection to AIDS is about 10 years, many people will become ill over the next several years. More treatment options, including early treatment, are essential.

How good are ddI and ddC? The important data generated by the ongoing large-scale controlled trials has not yet been compiled and analyzed, let alone published; therefore, the physicians, scientists, and other professionals involved in the trials are reluctant to make public statements. But informal conversations with those involved suggest a widespread consensus that both drugs will be approved because they do have a role in AIDS/HIV treatment, especially in combination with AZT and/or for patients who cannot successfully use AZT alone. (For results of earlier human studies of ddI, see recent article by Yarchoan and others[2]; also see papers published by Reviews of Infectious Diseases, July-August 1990[3,4,5,6,7]; for background on human trials of ddC, see papers from a February 3, 1990 San Francisco symposium, published in May by The American Journal of Medicine[8,9,10,11,12,13,14,15,16]). The available evidence is not conclusive, but it is consistently positive.

Neither drug is risk free. Both can cause peripheral neuropathy, and must be administered carefully to avoid or control this side effect. In addition, ddI can cause pancreatitis, which in a few cases has been fatal; blood tests are necessary so that if this condition starts to develop, the drug can be stopped or have its dose reduced before it causes illness. ddC does not cause pancreatitis.

The new data, currently being accumulated in formal trials and in the large ddI expanded-access program, seems unlikely to change the current belief that these drugs can be beneficial to many patients, and that side effects are manageable. The new data will be analyzed before drug approval is given. Activists are not seeking instant approval of ddI and ddC, but rather approval in the near future, hopefully before the end of 1990. In the three months until then, the data could be analyzed, if an adequate effort is made.

No one wants unconditional approval regardless of the data -- and presumably no one would deny approval regardless of the data, either. Therefore the real issue is not whether these drugs should be approved, yes or no. The real issue is how to make the decision -- how to evaluate all existing information to best respond to a public-health emergency.

The Traditional Way to Decide

Protocols for the major ddI and ddC studies, like most clinical trials today, have provision for early termination of the trial if early results show definitively that one treatment arm is worse than others. A "data safety monitoring board" (DSMB) meets privately at pre-set intervals and unblinds the study, so that patients will not be continued on a treatment which is clearly worse. Usually, of course, the DSMB decides to continue the study. And in that case it seldom says anything publicly about what results were found, so as not to bias the trial, for example by causing patients on a treatment arm which seemed inferior to drop out. Because of this secrecy, the public cannot monitor what is happening within the DSMB, and must accept its work on faith.

We do not know what criteria the DSMB will use to decide whether or not to terminate the ongoing ddI and ddC studies. But there are several reasons to be seriously concerned that the business-as-usual approach, which is to make this decision according to criteria spelled out long ago, when the protocols were first designed, could lead to the wrong decision. And the decision will be vitally important, because if the trials are not terminated early, they could last for two years or more after enrollment problems are overcome -- meaning that neither drug will be approved for a long time.

The first potential problem is that when these trials were designed, immunological and virological markers (T-helper counts, p24 antigen) were less accepted by researchers than they are today. Therefore the trials were designed statistically to compare two or more arms based on rates of death or major disease progression. That is why these studies require hundreds of volunteers and are likely to last two years. AIDS progresses slowly, so it takes a long time to accumulate enough deaths and major opportunistic infections for statistically definitive comparisons to be made -- no matter how good the drug being tested may be. If the DSMB for each trial evaluates the interim data based on the original protocol, instead of today's knowledge, it is likely that neither study will be stopped, and we will have to wait out the two years for these drugs.

Another concern is that, for statistical reasons, the criteria for early termination for clinical trials are extremely conservative. The statistical methods in common use require that if you take an early look at a study's data, and might have stopped the trial but in fact do nothing, then at the end of the study, your claims for the drug must be weaker than they otherwise could have been. Although this is difficult even for researchers to understand, the unbelievable result is that just looking at the data could cause a drug which otherwise would have been declared effective to be declared unproven instead. (The reason for this is that the overall procedure of the trial must be designed to control the probability of erroneously accepting a worthless drug, and the early look does add to the probability of such an error. Therefore, the final criterion for ruling the drug effective at the end of the study must be tightened, in order to keep the overall probability of error within the limit claimed by the trial's designers.)

The problem for us is that to minimize this effect, to minimize the possibility of having to reject a drug which otherwise would have been accepted, trial designers require extremely severe standards for early termination of a trial, making such termination unlikely. A drug must do much better to cause a trial to end early, than to be considered proven effective if the trial had been scheduled to end at that point.

Another fundamental problem is that the largest trials being conducted now are testing ddI and ddC individually. But today it is becoming increasingly clear that the best way to use these drugs is in combination, such as ddC with AZT. Large studies can take so long to get into operation that they become obsolete even before they begin.

Another Approach: Lasagna Committee Standards

On August 15, the National Committee To Review Current Procedures For Approval Of New Drugs For Cancer And AIDS (commonly known as the Lasagna Committee, after its chairman, Louis Lasagna, M.D.) issued its final report (see "Federal Panel Seeks Drug- Approval Reforms," AIDS TREATMENT NEWS #110, September 7, 1990). The Lasagna Committee had been asked by then Vice- President George Bush to recommend better ways to study and approve drugs for life-threatening conditions, to make them available more rapidly to patients who need them. The Committee's recommendations are based on ten hearings held between January 1989 and April 1990.

Of the Committee's 20 recommendations, the one most relevant to ddI and ddC approval is number four, "The FDA Standard for Effectiveness of New Drugs." Because of its importance, we reproduce the entire recommendation here:

"Because of its special relevance to issues faced today in developing new drugs for cancer and AIDS, the FDA needs to pay particular attention to the congressional intent in requiring substantial evidence of effectiveness prior to approval of a new drug application (NDA), as described in the Senate Report on the Drug Amendments of 1962:

"'The term "substantial evidence" is used to require that therapeutic claims for new drugs be supported by reliable pharmacological and clinical studies. When a drug has been adequately tested by qualified experts and has been found to have the effect claimed for it, this claim should be permitted even though there may be preponderant evidence to the contrary based upon equally reliable studies. There may also be a situation in which a new drug has been studied and its effectiveness established only to the satisfaction of a few investigators qualified to use it. There may be many physicians who would deny the effectiveness simply on the basis of a disbelief growing out of their past experience with other drugs or with the diseases involved. Again, the studies may show that the drug will help a substantial percentage of the patients in a given disease condition but will not be effective in other cases. What the committee intends is to permit the claim for this new drug to be made to the medical profession with a proper explanation of the basis on which it rests. In such a delicate area of medicine, the committee wants to make sure that safe new drugs become available for use by the medical profession so long as they are supported as to effectiveness by a responsible body of opinion and scientific fact.'

"By applying these principles, patients suffering from AIDS and cancer will have available to them new drugs for the treatment of their disease at the earliest stage at which there is responsible scientific evidence to justify marketing.

"The committee recognizes that, by making new drugs available for marketing at this early stage, when there is substantial evidence but not yet definitive evidence of effectiveness, there is an attendant greater risk of serious adverse reactions that have not yet been discovered. Cancer and AIDS patients have made it clear to the committee, however, that in light of the seriousness of the diseases involved, they are willing to accept this greater risk. Earlier approval of new drugs will mean that the patient will bear greater responsibility, along with the physician, for understanding and accepting the risks involved."

It may seem unreasonable to urge that new drugs for life-threatening conditions like cancer and AIDS be approved when "reliable pharmacological and clinical studies" conclude that they work, but "preponderant evidence to the contrary based upon equally reliable studies" concludes that they do not. Why should it be national policy to approve a drug when there is such uncertainty? Other parts of the Lasagna Report provide background for this recommendation. For example, from a section discussing cancer drugs but which also applied to AIDS, "It is only after initial NDA approval of the drug as a single entity that its full potential is realized, because physicians are then free to use it in combination with other drugs in accordance with their best clinical judgment. While still under investigation, such combination uses occur only infrequently and with little opportunity for full clinical exploration."

Before the current law on drug approval was passed in 1962 (the law to which the long quote above applied), drugs only had to be proven safe before they could be sold in the United States; they did not have to be proven effective. Although some analysts have estimated that we would be better off with that system than with the present one (claiming that earlier and less costly access to good drugs would more than balance the harm caused by the useless ones), very few people want to go back to the old system of not requiring efficacy proof, and politically there is no chance of that happening.

But in recent years, proof of efficacy may have been taken too far -- to the extreme of requiring academically satisfying proof, unrelated to real-world concerns such as balancing cost vs. benefit, or the feasibility of actually carrying out some of the trials which are called for. The result is a price tag of over $200,000,000 for each new drug approved -- money the public pays one way or another in drug prices. The more serious price is paid in lives of patients denied new drugs when no satisfactory alternative treatments are available.

The Lasagna Committee has issued an authoritative call for a more balanced and workable approach to developing treatments for life-threatening conditions. "When there is substantial evidence but not yet definitive evidence of effectiveness," critically important drugs should be approved for prescription use, with physicians properly informed of the state of the evidence.

The academic elegance theoretically available from rigidly controlled trials has led to an assumption that all new knowledge about drugs comes from formal trials, and that physicians merely apply that knowledge to patients. In fact, medical progress rests on two legs -- scientific studies and also clinical experience -- and they must work together for best results.

(Note: the Congressional intent and Lasagna Committee recommendation that a drug can be approved even when there is "preponderant evidence" against it does not describe ddI or ddC. The data on both of these drugs, while not conclusive, is overwhelmingly favorable; see references, below.)

Early Approval and Clinical Trials

One argument against early approval of ddI and ddC -- we think it is the strongest argument -- is that early approval would probably result in the ending of the ongoing large-scale controlled trials. As a condition for approval, the manufacturers would of course be required to continue doing studies; the FDA is already moving in the direction recommended by the Lasagna Committee of approving critically important drugs at the earliest possible time, with more of the efficacy research being moved into the period after the drug is available by prescription. But usually such early approval is granted after phase II studies have finished -- either at the end of their scheduled time, often about two years, or in extreme cases by early termination according to the very conservative pre-defined criteria applied by the DSMB, as discussed above. For ddI and ddC, however, activists want the existing data to be evaluated now, without waiting for the ongoing phase II trials to finish; the drugs would be approved if there is substantial evidence of their value.

Theoretically it would be possible to continue the existing large-scale phase II trials of both drugs, even after approval. But in practice, with ddI and ddC approved and with the growing belief that combination therapy with AZT is usually best, many patients, especially those not doing well in the trials, would leave them to choose other treatment options. Therefore, the existing large-scale controlled trials probably could not be completed.

A similar issue arose when AZT was approved; the long-term phase II trial was terminated early by the DSMB. At that time, there were 17 deaths in the placebo group compared to only one in the AZT group, a difference so great that it did meet the very conservative statistical criteria for ending the study early. But today, in retrospect, it seems clear that AZT is not that good; the extreme difference in the death rates must have resulted partly by chance. No one knows why so many patients in the placebo group had died at that time. Some researchers have since argued that they wished the study had been continued longer, so that we could have had more conclusive proof of efficacy, especially regarding long-term use. It has even been suggested that the early end of the AZT trial was a tragedy, since we will never have the data which it could have provided.

But the benefits of early approval of AZT far outweighed the costs. The benefits included not only getting AZT to the study volunteers in the placebo group, but also making the drug available to thousands of others not in the trials. The approval also opened up the field of combination studies, as the FDA almost never approves trials using more than one experimental drug at the same time. And the success of AZT greatly advanced all of AIDS research, by ending the earlier fatalism: the idea that no drug could possibly work to treat AIDS.

And today much more is known about ddI and ddC, especially about dose, side effects, and long-term human use, than was known about AZT when that drug was approved. But perhaps the most significant difference between the current situation and that of AZT is that if the analysis of all existing data (which activists are now calling for) confirms that combination therapies such as AZT and ddC are often better than any of the drugs alone -- and we believe it is almost certain that this result will be confirmed -- then it will mean that the ongoing large-scale controlled trials studying these drugs alone are fundamentally obsolete, since the treatment they are testing would be known to be less than the best. If the benefits of combination treatment are confirmed, then will it be worth sacrificing the interests not only of hundreds of volunteers in the ongoing trials, but of tens of thousands of others for whom existing treatments are unsatisfactory, to get more data on ddI and ddC as single-drug therapies which will never be widely used?

Early Approval and Industry Incentives

Another argument against early approval of ddI and ddC is that approval is not necessary to provide access, which can be done through the proposed "parallel track" system, or a similar expanded access mechanism, to allow drugs to be used while the controlled clinical trials proceed. Parallel track and comparable approaches certainly need to be explored. Some people suspect, however, that early approval of drugs for prescription use, while research continues in post-marketing studies, may work better than expanded access systems such as parallel track, for the following reasons:

* We are in an emergency, and expanded-access systems will take time to work out, including the all-important physician education component. Consider how long it is taking to develop parallel track, since that concept was first proposed. But when drugs are approved for prescription use, the necessary systems are already in place.

* Expanded-access systems may not have the flexibility that physicians and patients want -- for example, to allow use of drugs in combination.

* Paperwork and other difficulties often limit such programs to private physicians, raising equity issues because the treatments are not available to those who receive their medical care in public clinics.

* Expanded-access programs like parallel track will not happen unless the pharmaceutical companies which own the drugs are willing to participate. But the companies may not have enough incentive to put their products on parallel track. They cannot be allowed to earn money on a program to distribute drugs before marketing approval (since if they could, the program would become tantamount to marketing approval, and would have no reason for separate existence). Even recovering their costs from parallel-track distribution is not feasible, as pharmaceutical companies are unwilling to reveal what their costs are. And if companies are pressured to participate, then they have incentives to create restrictions which keep the programs small. Parallel track might turn out to be a good idea, except for one problem -- that it seldom gets used in practice.

What does provide incentive to pharmaceutical companies? One expert on clinical trials described the companies' approach toward the FDA as, "Tell us what we must do, and that if we do it we will get the NDA," (New Drug Application approval, meaning permission to market the drug). The only incentive that counts for pharmaceutical companies is getting their drug approved. (And incidentally, bureaucratic dynamics may favor approval of unimportant drugs which break no new ground over major innovations.)

Therefore, a policy of the earliest possible approval of critically important drugs for life-threatening conditions such as AIDS and cancer -- the policy recommended by the Lasagna Committee -- has a hidden future benefit, in addition to the present benefit of making existing drugs available earlier to patients. It creates incentive which extends throughout the drug development process, all the way back to the earliest theoretical and laboratory studies, for companies to bring their most important drugs forward, to focus their research resources on critical, life-saving drugs, instead of pursuing the me-too products for existing markets, which consume so much effort today.

References

1. Hessol NA, O'Malley P, Lifson A, and others. Incidence and prevalence of HIV infection among homosexual and bisexual men, 1978-1988 [abstract #M. A. O. 27]. Fifth International Conference on AIDS, Montreal, June 4-9, 1989.

2. Yarchoan R, Pluda JM, Thomas RV, and others. Long-term Broadcastoxicity/activity profile of 2',3'-dideoxyinosine in AIDS or AIDS- related complex. The Lancet. September 1, 1990; volume 336, pages 526-529.

3. Rosencweig M, McLaren C, Beltangady M, and others. Overview of phase I trials of 2',3'-dideoxyinosine (ddI) conducted on adult patients. Reviews of Infectious Diseases. July-August 1990; volume 12, supplement 5, pages S570-S575.

4. Cooley TP, Kunches LM, Sanders CA, and others. Treatment of AIDS and AIDS-related complex with 2',3'-dideoxyinosine given once daily. Reviews of Infectious Diseases. July-August 1990; volume 12, supplement 5, pages S552-S560.

5. Dolin R, Lambert JS, Morse GD, and others. 2',3'- Dideoxyinosine in patients with AIDS or AIDS-related complex. Reviews of Infectious Diseases. July-August 1990; volume 12, supplement 5, pages S540-S551.

6. Valentine FT, Seidlin M, Hochster H, and Laverty M. Phase I study of 2',3'-dideoxyinosine: Experience with 19 patients at New York University Medical Center. Reviews of Infectious Diseases. July-August 1990; volume 12, supplement 5, pages S534-S539.

7. Richman DD. Zidovudine resistance of human immunodeficiency virus. Reviews of Infectious Diseases. July- August 1990; volume 12, supplement 5, pages S507-S510.

8. Broder S and Yarchoan R. Dideoxycytidine: current clinical experience and future prospects -- A summary. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B-31S to 5B-33S.

9. Meng TC, Fischl MA, and Richman DD. AIDS Clinical Trials Group: Phase I/II study of combination 2'-3'-dideoxycytidine and zidovudine in patients with acquired immunodeficiency syndrome (AIDS) and advanced AIDS-related complex. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B- 27S to 5B-30S.

10. Bozzette SA and Richman DD. Salvage therapy for zidovudine- intolerant HIV-infected patients with alternating and intermittent regimens of zidovudine and dideoxycytidine. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B-24S to 5B-26S.

11. Skowron G and Merigan TC. Alternating and intermittent regimens of zidovudine (3'-azido-3'-deoxythymidine) and dideoxycytidine (2',3'-dideoxycytidine) in the treatment of patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B-20S to 5B-23S.

12. Pizzo PA. Treatment of human immunodeficiency virus- infected infants and young children with dideoxynucleosides. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B-16S to 5B-19S.

13. Merigan TC and Skowron G. Safety and tolerance of dideoxycytidine as a single agent. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B- 11S to 5B-15S.

14. Richman DD. Susceptibility of nucleoside analogues of zidovudine-resistant isolates of human immunodeficiency virus. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B-8S to 5B-10S.

15. Broder S. Pharmacodynamics of 2',3'-dideoxycytidine: An inhibitor of human immunodeficiency virus. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, pages 5B-2S to 5B-7S.

16. Broder S. Dideoxycytidine (ddC): A Potent antiretroviral agent for human immunodeficiency virus infection -- An introduction. The American Journal of Medicine. May 21, 1990; volume 88, supplement 5B, page 5B-1S.

901005
ATN11201

AEGiS 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 ©1980, 1990. AEGiS. All materials appearing on AEGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of AEGiS, or the party credited as the provider of the content. .