The Body: The Complete HIV/AIDS Resource
Follow Us Follow Us on Facebook Follow Us on Twitter Download Our App 
Professionals >> Visit The Body PROThe Body en Espanol
  • Email Email
  • Printable Single-Page Print-Friendly
  • Glossary Glossary

AIDS Treatment News
November 17, 1995


  1. Protease Inhibitors: Drug Resistance and Cross Resistance Overview
  2. Grapefruit Juice and Saquinavir
  3. FDA Reform: New Senate Bill

Protease Inhibitors: 
Drug Resistance 
and Cross Resistance Overview

by John S. James
Viral resistance to protease inhibitors has become a major concern in the HIV community. The immediate issue is that saquinavir (Invirase(TM)), (an experimental protease inhibitor developed by Hoffmann-La Roche, which was recommended for approval on November 7 and with luck might be approved before the end of this year) will first be approved at a dose of 600 mg three times a day, which is now known to be too low.

Using a drug in a sub-optimal way might make it easier for HIV to develop RESISTANCE to that drug, so that the drug is no longer effective for that particular patient (or for anyone else who acquires a resistant strain). Also, different but similar drugs can sometimes have CROSS RESISTANCE -- meaning that virus which has become resistant to one drug will also be resistant to another drug, even without ever being exposed to it. Because of cross resistance, it is possible that using the wrong drug, or using too low a dose, or not using the best treatment combination, could cause one to permanently lose the ability to benefit from most or all of the new protease-inhibitor drugs.

The issue is complex, because early data suggests that saquinavir even at the low dose may cause less resistance and cross resistance than the other protease inhibitors to which it has been compared. The bottom line is that nobody yet knows whether using the current dose of this drug could cause loss of opportunities to benefit from better treatments later. Sentiment in the community seems to be that persons with more advanced disease may choose to try new treatments as soon as they are available, while others might decide to wait for additional information and options. Fortunately, a small clinical trial could quickly address the immediate questions about saquinavir.


When protease inhibitors were first tried against HIV, many researchers guessed that it would be difficult for the virus to develop resistance against them. The protease enzyme, the target of these new drugs, is fairly small (it consists of only 99 amino acids), and it has a critical function in the production of new copies of HIV.

Researchers had assumed that if the HIV protease changed its structure in response to a drug, the viruses produced would be defective. But in fact, resistance to protease inhibitors has turned out to be a serious problem -- and a complex one. At least 11 of the 99 amino acids in HIV protease can change and still produce a viable virus. And resistance develops very differently to the different protease inhibitors now being tested in people; there seems to be no consistent pattern.

Is cross resistance serious with low-dose saquinavir, the first protease inhibitor likely to be approved? Two of the major companies now developing protease inhibitors - - Hoffmann-La Roche with saquinavir, and Merck & Co. with indinavir (Crixivan(TM)) a rival drug -- publicly disagree about the risk. The bottom line, which came out two days before the FDA Antiviral Advisory Committee meeting (at a drug resistance workshop sponsored by Project Inform and the Gay Men's Health Crisis), is that both sides are guessing; no one knows the answer. But we DO know how to find out; there is widespread consensus that a small clinical trial, combined with the proper laboratory studies, could within a few months answer the practical questions about using saquinavir.

The issue now is whether these companies which are bitter rivals can cooperate enough with each other, or with the National Institutes of Health, to conduct the studies -- when everyone agrees that they need to be done, and everyone agrees that we know exactly how to do them.

Resistance and Cross Resistance:
What Is Known?

While much remains unknown, some facts about resistance are becoming clear. While this information does not answer all questions about practical treatment decisions, it can help to clarify some of the issues.

  • Drug resistance can be more serious with HIV than with most disease-causing organisms, for at least two major reasons. First, unlike bacteria, HIV ordinarily is never completely cleared from the body; it keeps reproducing for years and making new mutants, some of which will be drug resistant in ever-changing ways. Also, HIV reproduces very rapidly and also inaccurately (since retroviruses, unlike human cells, cannot edit the errors which occur when they reproduce), allowing new variants to develop rapidly.

  • A drug can stop working for different reasons -- not only because the virus has become resistant. For example, the body can change its metabolism to resist certain drugs. Unfortunately, this "host resistance" has not been studied enough, and is poorly understood.

  • Conversely, a drug can sometimes still be beneficial even after HIV has become resistant to it. For example, when persons are given too low a dose of the Merck protease inhibitor indinavir the viral load first goes down greatly, and the CD4 count goes up. But then in a few weeks or months the viral load comes back up toward its baseline value, because the virus has become resistant and the drug can no longer stop it from reproducing. However, the CD4 count often stays up, for months or more after the viral load has returned to its starting value. No one knows why this happens; it is possible that the resistant virus is weaker in some way and causes less damage.

  • New resistant variants cannot develop unless the virus is reproducing. Therefore, if it becomes possible to shut off viral reproduction almost completely -- by using better drug combinations -- it might be possible to stop the creation of new resistant variants. But so far, this strategy has been difficult for several reasons. First, the drugs available today are not good enough to shut off viral reproduction sufficiently. Also, there may be reservoirs of virus in the body, such as in the central nervous system, where certain drugs do not penetrate well. In addition, if patients miss doses or take a "drug holiday," the virus can take advantage of this opportunity to start reproducing again. And finally, a person with advanced HIV disease probably has millions of HIV variants already, including variants resistant to the treatments they are using; these then only need to be selected for by the drug, not created by new mutations.

  • Resistant virus is usually (but not always) less "fit" for survival in the absence of the drug. Therefore, if a person uses a drug and develops a resistant virus, and then stops the drug, most of the virus will usually go back to its original form. However, it seems that the resistant virus is not truly gone, but remains at a low level, too low to be directly detected. Therefore, if the drug is started again, the resistant virus will probably come back quickly.

  • As with tuberculosis, it is usually much harder for HIV to develop resistance to a well-chosen combination of drugs which are started together, than to develop resistance to the same drugs one at a time. For this reason, when someone decides to start antiretroviral treatment, it is probably a mistake to start with AZT alone. It seems better to start with a combination, although it is uncertain what combination is best. Persons who have already used AZT extensively may need different combinations than those who are treatment naive (or in some cases they might switch to a single drug such as ddI). Sequences of combinations may be necessary, and some sequences are likely to work better than others. Much more work needs to be done to find out which combinations are best for which patients. (Meanwhile, viral load tests can indicate which combinations are working well for an individual, allowing treatments to be determined by trial and error.) Some treatment combinations are inherently difficult for HIV to develop resistance to. For example, 3TC is especially effective against certain AZT-resistant viruses -- which helps to make this combination a good one. (Today, however, it is known that HIV can become resistant to both AZT and 3TC.)

  • There is probably no cross resistance between nucleoside analogs (AZT, ddI, ddC, d4T, 3TC, etc.) and protease inhibitors (saquinavir, indinavir, ritonavir, etc.) This is because these two drug classes act entirely differently, on different genes in different parts of the virus. When additional classes of drugs such as integrase inhibitors are developed, they also will not be cross resistant to the existing drugs.

  • Because of the very different resistance patters of different protease inhibitors, new protease inhibitors may still be effective even if resistance develops to all the known ones. And sometimes the development of resistance to one protease inhibitor can make HIV more sensitive to another.

  • The three protease inhibitors which are furthest along in human trials have very different patterns of viral resistance. With the Merck drug, the virus needs at least three (and preferably five) mutations to show resistance in laboratory tests. With saquinavir, resistance seems to depend on only one or two mutations, but these tend to develop slowly; after one year, most patients do not have virus with the genes known to cause the resistance. Both a higher dose of saquinavir, and use of that drug in combination with other antiretrovirals, seem to reduce the rate of mutation. The saquinavir mutations are not the ones important for resistance to the Merck drug; whether there is significant cross resistance between these drugs is controversial (Merck says maybe, Roche says no). With the Abbott drug (ritonavir), many different mutations can cause resistance, but one in particular seems likely to develop first. Since that mutation seems to be inconsistent with the major saquinavir mutation, it is possible that combining these drugs could make it hard for HIV to develop resistance. (Caution: people must NOT combine these two drugs until a scientific trial is done, because it is likely that the dose of saquinavir will have to be drastically reduced, probably by many times, because ritonavir greatly raises blood levels of saquinavir by preventing it from being eliminated by the body. Today no one knows how much the dose must be reduced.)

Saquinavir Dose and Formulation

The saquinavir dose now being used in the major Roche clinical trials, and in the Roche lottery (600 mg three times a day, for a total of 1800 mg per day), is the only dose which the FDA could approve at this time. While it is widely agreed that this dose is too low, no one knows what the best dose might be, because the safety and other studies have not been done.

About 20 people have taken twice the current dose (total 3600 mg/day), and 20 others have taken four times the current dose (7200 mg/day), in a small trial in Stanford University. The "maximum tolerated dose" has not yet been reached, meaning that the best dose might be even higher. The low dose was chosen at least two years ago, due to manufacturing supply and cost; a higher dose would have delayed the large phase III clinical trials, because of the time required to manufacture enough saquinavir to supply those trials at the high dose. Also, the high dose would have cost so much to manufacture that people could not have afforded the drug.

More recently, Hoffmann-La Roche has developed a new formulation of saquinavir, which is more efficiently absorbed when given orally, making higher blood levels feasible. This new version of saquinavir is now starting dose-ranging trials, which will determine the optimum saquinavir dose. Unfortunately, current plans call for the large clinical trials to continue with the existing small dose; there is even talk of adding more people to these trials (at the wrong dose), because without more volunteers the trials might not get statistical proof that low-dose saquinavir is beneficial at all. (It has been suggested that everyone in the trials could be switched to a higher dose with the new formulation of saquinavir, once a new dose is determined.)

The immediate practical question facing patients and physicians is whether this low saquinavir dose might cause more harm than good by helping HIV more quickly develop resistance to saquinavir, and possibly cross resistance to other protease inhibitors. With the Merck drug, the answer to this question would clearly be yes; too low a dose causes resistance to develop rapidly, and then the person cannot get the benefit of the higher starting dose even if they shift to that dose, and in addition they have become resistant to most other protease inhibitors. But saquinavir is very different from the Merck drug; resistance develops much more slowly even at the low dose (perhaps because the mutations which cause resistance are harmful to the virus in other ways), and it is unclear if much cross resistance develops. Also, laboratory studies have not shown cross resistance between saquinavir and the Merck and Abbott drugs. (There is cross resistance between saquinavir and AG1343, the experimental protease inhibitor from Agouron, probably because of chemical similarities between the two drugs.)

The FDA Antiviral Advisory Committee, at its meeting on November 7, recommended that saquinavir be approved only for use in combination with one or more nucleoside analog antiretrovirals -- and with the suggestion that these nucleoside drugs should preferably be ones the patient has not used before. Which drugs to use, and when to start treatment, are always individual decisions; some people do not have time to wait but need to start antiretroviral treatment immediately, and may choose to include saquinavir; those who have time may decide to wait until later, when there is more information, and perhaps better protease treatments (either from higher doses of saquinavir, or from other drugs or combinations). The decision whether or not to use saquinavir will be much easier when a study is done to find out whether people who have been treated with low-dose saquinavir are more likely to be resistant to the Merck or other protease inhibitors -- or likely to develop resistance more rapidly -- than those who have never received saquinavir.

As we noted above, there is no mystery about how to do this study. Also,it will not cost much, because it does not need to be a big "clinical endpoint" trial; instead, viral load testing can quickly show in a few volunteers if people who have used low-dose saquinavir are resistant to the Merck drug or not.

At the November 7 meeting, spokespersons for both companies promised FDA Commissioner David Kessler that they would do this research, and would cooperate with each other when necessary. Treatment activists can make an important contribution by continuing to follow this issue, and pushing when necessary to make sure that the study gets organized rapidly.

[Note: For more information about protease inhibitors, and their potential when used properly, see "Protease Inhibitors and Prevention of Cross Resistance," by Jules Levin, AIDS Treatment News #232 (October 6, 1995). [You can contact the Protease Inhibitors Working Group by calling Jules Levin, 718/624-8541.]

Grapefruit Juice and Saquinavir

by John S. James
Saquinavir, the protease inhibitor developed by Hoffmann-La Roche (and recently recommended for approval by the FDA Antiviral Advisory Committee despite general agreement that the dose which will be approved is too low) has very low bioavailability -- only about 4% of the drug taken orally reaches the general circulation. This is partly because the drug is poorly absorbed, and partly because it is rapidly destroyed by an enzyme (cytochrome P450 3A4, also called CYP3A4) which is present in the liver and also in the intestinal wall. Some drugs and even foods inhibit this enzyme, thus increasing the AUC (area under the curve, a measure of total exposure to the drug) of saquinavir.

In studies conducted by Hoffmann-La Roche, the prescription drug ketoconazole increased the AUC -- but ketoconazole would not be used for this purpose because of its toxicity. Ranitidine (Zantac) also increased the AUC of saquinavir, but by an unknown mechanism. [On the other hand, rifampin greatly REDUCED the AUC of saquinavir, and therefore these two drugs should not be combined.]

Chemicals in grapefruit juice also inhibit CYP3A4. Hoffmann-La Roche tested grapefruit juice with saquinavir in 12 HIV- negative volunteers. When they took a 150-ml glass of ordinary reconstituted frozen grapefruit juice with the saquinavir, and then another 150-ml glass an hour later, the AUC of saquinavir was increased to 150% (a 50% increase) [90% confidence interval, 101% to 191%, meaning that if the true value were in fact 150, the variation of the data was such that the observed value would be within that range 90% of the time].

When they used double strength grapefruit juice (reconstituted from the frozen, but with less water) in the same amount and schedule, the AUC of saquinavir was increased to 220% (slightly more than doubled) [90% confidence interval, 161%-300%].

In published studies with other drugs, grapefruit juice usually had comparable effects.(1-7) Some of these studies used 200 ml or 250 ml of fresh-frozen grapefruit juice, taken with the drug, but no dose an hour later. Orange juice had no effect. When the bioflavonoid naringin -- the component of grapefruit juice which was suspected to be the active ingredient in increasing the bioavailability of these drugs -- was tried instead of grapefruit juice, it had much less effect than the juice itself.

Should saquinavir be taken with grapefruit juice, to help make the low dose more effective? Using grapefruit juice might increase the blood level achieved, but also it could have uncertainties and drawbacks. First, the effect of grapefruit juice on drug bioavailability can vary greatly from person to person -- and in addition, each batch of grapefruit juice can be different. Also, there is no way to find out whether or not grapefruit juice is working for oneself, because the test to measure saquinavir in the blood is not commercially available.

Another concern is that some other drugs are also metabolized by the same liver enzyme -- for example, cyclosporine, some calcium channel blockers, theophylline, and others -- and large amounts of grapefruit juice might increase the blood levels of these drugs enough to cause toxicity. Also, the amount of double-strength grapefruit juice required to double the AUC of saquinavir might be irritating to the stomach and hard to tolerate every day. And the "safety database" -- the number of people who already have used saquinavir at higher blood levels -- is small. Due to such concerns, Hoffmann-La Roche decided not to pursue using grapefruit juice, but instead is developing its new formulation of saquinavir, which increases bioavailability about 300%.

Note: Regardless of grapefruit juice, saquinavir must be taken after a meal. The official recommendation, in the Drug Usage Plan of the international compassionate treatment program, is that "Saquinavir should be taken with water within 2 hours after food. Saquinavir should not be taken on an empty stomach." Saquinavir is not affected by stomach acidity; but if taken on an empty stomach, bioavailability is greatly reduced.


1. Hukkinen SK, Varhe A, Olkkola KT, and Neuvonen PJ. Plasma concentrations of triazolam are increased by concomitant ingestion of grapefruit juice. CLINICAL PHARMACOLOGY AND THERAPEUTICS. August 1995; volume 58, number 2, pages 127- 131.

2. Kupferschmidt HH, Ha HR, Ziegler WH, Meier PJ, and Krahenbuhl S. Interaction between grapefruit juice and midazolam in humans. CLINICAL PHARMACOLOGY AND THERAPEUTICS. July 1995; volume 58, number 1, pages 20-28.

3. Ducharme MP, Warbasse LH, and Edwards DJ. Disposition of intravenous and oral cyclosporine after administration with grapefruit juice. CLINICAL PHARMACOLOGY AND THERAPEUTICS. May 1995; volume 57, number 5, pages 485-491.

4. Yee GC, Stanley DL, Pessa LJ, and others. Effect of grapefruit juice on blood cyclosporin concentration. LANCET. April 15, 1995; volume 345, pages 955-956.

5. Merkel U, Sigusch H, and Hoffmann A. Grapefruit juice inhibits 7-hydroxylation of coumarin in healthy volunteers. EUROPEAN JOURNAL OF CLINICAL PHARMACOLOGY. 1994; volume 46, number 2, pages 175-177.

6. Baily DG, Arnold JMO, Strong HA, Munoz C, and Spence JD. Effect of grapefruit juice and naringin on nisoldipine pharmacokinetics. CLINICAL PHARMACOLOGY AND THERAPEUTICS. December 1993; volume 54, number 6, pages 589-593.

7. Baily DG, Arnold JMO, Munoz C, and Spence JD. Grapefruit juice--felodipine interaction: Mechanism, predictability, and effect of naringin. CLINICAL PHARMACOLOGY AND THERAPEUTICS. June 1993; volume 53, number 6, pages 637-642.

FDA Reform: New Senate Bill

by John S. James
A major bill on FDA reform and the entire drug-development process is likely to be introduced by Senator Nancy Kassebaum -- a moderate Republican who has been helpful on important AIDS issues -- probably during the next few weeks. The bill has not yet been made public; we obtained a recent (undated) 76-page discussion draft two days before this issue of AIDS Treatment News went to press, so this article represents our first impressions of the content, before we have had a chance to discuss it widely. The AIDS community should be aware of this bill now because important negotiations may happen before it is publicly introduced. And the position of our community does matter -- especially since the support of Senator Edward Kennedy, the senior Senate Democrat on health policy, will likely be critical for this bill to succeed.

The Kassebaum bill could establish the leadership of a moderate, consensus approach to FDA reform over what might be much more radical proposals from the House. This bill deserves careful evaluation on its merits (keeping in mind that the draft which exists now will change greatly during the legislative process).

Any FDA legislation is likely to be controversial in the AIDS community, because many activists, including this writer, wanted to see FDA and drug-development reform through administrative action, instead of through legislation. Administrative reform can be much faster; also, there is reason to fear what Congress might do once it opens up this issue. But a package of administrative FDA reforms just announced by Vice President Al Gore and FDA Commissioner David Kessler is disappointing.

According to the November 10 and October 30 BIOCENTURY (a weekly faxed newsletter published by BioCentury Publications Inc. in San Carlos, California), "the reforms center on elimination of requirements for establishment licenses and lot release approvals for well characterized biologics." In other words, this highly touted administrative reform concerns technical changes which will save money for business, but not touch the structural issues associated with the astonishing lack of significant improvement in new treatment for patients, despite the great resources devoted to medical research.

There has been no major FDA legislation since 1962. Since that time, the FDA's role has become increasingly important. And the FDA is unusual among government agencies in that there is generally no appeal from even a lower-level individual's decision, and little effective institutional oversight. AIDS does get high-level attention -- and in effect has an informal appeals system which often works well. But away from the limelight of a high-profile issue, when companies feel that they have been treated wrongly, perhaps by an inexperienced employee who did not understand the matter or did not have the resources to address it properly, they have to swallow hard and keep quiet.

What's In the Kassebaum Bill?

The 76-page draft bill is hard to evaluate and summarize, but below are some provisions we noticed. Major bills change greatly during the legislative process; the following should be accurate as of November 14, but later will become obsolete. The points below are in the order in which they appear in the draft of the bill.

  • Mission. The FDA's mission would now explicitly include "facilitating the rapid and efficient development and availability of products subject to its regulations," as well as protecting the public from unsafe or ineffective products.

  • Performance standards and review. The FDA commissioner would be required to publish and use quantifiable performance standards, to measure how well the FDA is doing in processing new drug applications, and in other matters, including responding to citizen petitions.

  • Information. The Secretary (of the Department of Health and Human Services, the FDA's boss in the government) would be required to set up an information system to track all applications, petitions, notifications, and other requests submitted to the FDA -- and permit access by the applicant. In addition, all "policy statements of general applicability" (guidelines, points-to-consider, protocols, recommendations, etc.) must be formally published and compiled. These measures would greatly help the interested public in understanding what actually happens at the FDA.

    [Comment: The three proposals above are important in view of the fact that today nobody really knows why the drug development process works as poorly as it does. In this situation, the first step toward improvement is to clarify and open the process, so that problems and solutions can come into view.]

  • Appeals. The Secretary "shall by regulation establish a system for the appeal," within the FDA, of any decision by an FDA employee (except for administrative or judicial proceedings). Any person could request an evaluation by an appropriate (advisory) scientific review group, and that group would decide which appeals it would hear. Other appeals systems could also be used. Within 60 days of conclusions and recommendations from the appeals process, the appropriate FDA official will make a final determination, and explain in writing the reason for any differences from the appeals recommendation.

  • Expedited access for seriously ill patients. The bill outlines a system which looks to us much like the expanded access systems today. It may help to unify the various different programs for access to experimental treatments, and apply them more equally to all major diseases. It allows companies to charge "cost recovery" for experimental drugs or devices, as they can today under one program, the "treatment IND." [But since the costs to be recovered under the current draft of the bill can include not only manufacturing and handling, but also "research" and "development," it is questionable whether this "cost recovery" language places any real limit on what companies can charge. Clearly this issue needs attention.] At present, large companies do not charge because they do not want to set prices early, or reveal information about their costs. Any reform should be scrutinized to make sure it will not make the existing situation worse.

  • Compromise on new drug research. Apparently Kassebaum's office considered proposals to allow institutional review boards to approve human research in some cases, without going through the FDA -- but then backed off from that approach, and proposed a compromise instead. We are unclear how the compromise differs from current practice.

  • Outside review. The bill would allow (not require) the Secretary to contract for outside expert review on part or all of any application for approval of a drug. The FDA, however, would still make the final decision --within 60 days after the outside review is complete.

  • Information exchange. Under rather strict controls, pharmaceutical companies would be able to tell health professionals about "off label" uses of approved drugs (and treatment uses for experimental drugs, when those uses have been approved by the Secretary). The only information the companies could distribute would be complete peer-reviewed articles or textbook chapters published by an organization independent of the pharmaceutical industry, or text approved by an independent continuing medical education accrediting agency, or information relating "to a use that is recognized under Federal law for purposes of third-party coverage or reimbursement," or information that is "an accurate and truthful summary" of material from the three kinds of sources listed above.

  • Restrictions in evaluating effectiveness. Companies trying to prove the effectiveness of a new drug could not be required to prove relative effectiveness, or cost effectiveness, unless those were claimed. And a "diagnostic device" would not have to prove effectiveness in clinical outcome, unless that were claimed.

    [For example, a test for HIV RNA could be approved based on proof that it accurately measured what it said it did -- without requiring a difficult and time-consuming clinical trial to accumulate statistical proof that patients whose doctors use viral load tests ultimately do better than patients whose doctors do not.]

  • Export reform. This bill takes seriously the responsibility to regulate the export of unapproved drugs to countries which do not have their own regulatory system.

  • New drug approval standard. The bill would add the following language to current law: "Substantial evidence shall consist of at least data from one well-controlled clinical investigation (which may be waived by the Secretary) and confirmatory evidence (obtained either before or after such investigation)." Currently, the FDA almost always requires at least two well-controlled clinical trials to approve a drug; but it does have the power to accept only a single trial in rare circumstances. The new language, which clearly allows but does not require the FDA to accept less than two trials, seems more a clarification rather than a change to current practice.
  • Drugs and biological products. The bill continues the existing movement in the direction of unifying the regulation of "drugs" and "biologics." In the past, these were distinct because biological products often could not be well characterized as to what they contained. As technology advances, the need for this distinction diminishes.

  • Pilot plants. A pharmaceutical company would not necessarily be required to set up a full-scale manufacturing facility to produce the drug supplies needed for clinical trials to prove safety and effectiveness.
  • Other sections include changes in the regulation of biological products, medical devices, and food additives.

  • Email Email
  • Printable Single-Page Print-Friendly
  • Glossary Glossary

This article was provided by AIDS Treatment News. It is a part of the publication AIDS Treatment News.