by Mark Harrington
7 November 1995

In September, 1994, the Treatment Action Group spelled out the following conditions as necessary for our support of an accelerated NDA for an HIV protease inhibitor:

1. Adequate and well-controlled clinical endpoints studies are well underway and likely to show whether the drug can prolong disease-free time or survival;
2. Adequate and well-controlled studies are complete or well underway and demonstrate a favorable combination of changes in CD4 levels and viral load;
3. The safety profile is adequately characterized and acceptable; and
4. The sponsor provided an expanded access program.

One year later, Hoffmann-LaRoche Inc. is the first sponsor to apply for accelerated approval of a protease inhibitor, INVIRASE brand Saquinavir, and is asking for two indications at 600 milligrams (mg) three times daily (t.i.d.):

• Combination therapy for use with approved antiretroviral agents [the nucleoside analogues] in adults with advanced HIV infection, and
• Monotherapy for use in adults with advanced HIV infection who are intolerant to or failing on approved antiretroviral agents.

We believe that Roche has substantially met the conditions we laid out one year ago for accelerated approval for these two indications:

1. Clinical endpoint studies. Two large ongoing studies, NV14256 and SV14604, are examining whether Saquinavir-containing regimens are superior to today’s nucleosides alone. NV14256, with three arms and about 900 participants, will be about as powerful as ACTG 116B/117, which showed ddI superior to AZT as second-line therapy. SV14604, with four arms and about 3,200 participants, will be more powerful than ACTG 175, which showed ddI and two combinations superior to AZT alone as first-line therapy. We believe these studies are likely to demonstrate whether Saquinavir confers clinical benefit.
2. Surrogate marker studies. Data from uncontrolled phase I studies and from the randomized ACTG 229 study demonstrates that Saquinavir has favorable effects on CD4 counts and less dramatic but still favorable effects on viral load. We have not seen today’s interim look at NV14256 which will strengthen or weaken the conclusions from these earlier studies. Assuming that these new data do not contradict the earlier studies, we believe these surrogate marker benefits may confer a clinical benefit.
3. Safety profile. Saquinavir shows little acute toxicity. Long-term data in diverse populations remains inadequate, yet the short-term risk ratio is clearly low.
4. Expanded access program. We are pleased that Roche eventually met the community’s demand for an expanded access program, one which is substantially larger than those of its competitors, but we are also disappointed that the expanded access program did not begin early enough for safety data to be available for review by the FDA and this committee here today. We are also disappointed that Roche did not avail itself of the opportunity to compare outcomes on those receiving Saquinavir under the lottery and those who applied but did not receive the drug. A randomization occurred, but no comparative data will be gathered. This is a regression from the dose-randomizations which took place with ddC and d4T. We hope future expanded access programs will be randomized, and if a lottery is necessary due to supply problems, that comparative survival will be measured.

Proposed indication for Saquinavir

Dose & Formulation. Due to competitive concerns, Roche has submitted what may be a suboptimal dose and what is definitely a suboptimal formulation. This raises troubling issues for the post-marketing period, but does not eliminate the demonstrated surrogate marker benefits.

Population. Roche is, appropriately, asking for approval in the groups for which it has completed or ongoing clinical endpoint studies, those with advanced HIV disease, mostly under 300 CD4 cells.

Regimen. Roche has mainly studied Saquinavir along with AZT, ddC or both. However Roche is asking for approval for use with any approved antiretroviral, presumably including ddI, d4T and (soon) 3TC. These combinations have not been adequately studied, but reflect the real-world conditions under which Saquinavir is likely to be used. FDA should mandate formal studies of these combinations and interactions as part of the post-marketing package.

Resistance and cross-resistance. Because evidence and opinion are contradictory, we believe the initial labeling indication for Saquinavir must include a prominent warning about the possibility of resistance and cross-resistance to other protease inhibitors.

Current issues and future directions

Obviously many outstanding questions remain about how to optimize the use of Saquinavir.

• Pediatric studies. We are disappointed that problems in developing a palatable pediatric formulation have delayed the initiation of safety and activity studies in infants and children. Developing and studying a formulation for children with HIV infection must be among the highest priorities for Roche in the post-marketing period.
• Formulation. Saquinavir is poorly bioavailable (4%). Roche is working on an “enhanced oral formulation” but this is not yet ready. An enhanced formulation must be rapidly developed.
• Dose. The optimal dose is not yet clear. The Stanford study suggested that higher doses may have more dramatic effects on surrogate markers and may delay the emergence of resistance. The optimal dose of Saquinavir needs to be defined.
• Resistance and cross-resistance. There are conflicting reports about whether Saquinavir use induces mutations which may decrease the antiretroviral efficacy of other protease inhibitors such as the Merck and Agouron compounds. Roche must work with academic virologists and with other protease inhibitor sponsors to resolve these issues around the virological and clinical significance of resistance and cross-resistance. We believe the initial labeling for Saquinavir should include a prominent box warning that use of Saquinavir may limit the clinical and virological utility of future protease inhibitors, and that physicians and patients must be educated about this possibility in Roche promotional materials.
• Combination protease therapy. There is also the possibility of antiviral synergy between protease inhibitors, e.g., between Saquinavir and either Indinavir or Ritonavir. The sponsors should work together to define optimal combination or sequential regimens.
• Optimizing clinical utility of protease inhibitors. Currently neither NIH nor industry have developed adequate infrastructure to define the answers to such questions as 1) when is the optimal time to initiate antiretroviral therapy; 2) what are the optimal antiretroviral regimens or treatment strategies; 3) how can we maximize clinical benefit, minimize expense and toxicity and rationally use the plethora of antiretroviral agents which we will soon have at our disposal? NIH, industry and the community should work together to develop structures capable of answering questions about optimizing clinical benefit.
• Price and access to underserved populations. As a chemically new and complex set of compounds, the HIV protease inhibitors are likely to be marketed at prices higher than those now seen with current nucleosides. We are concerned about the impact of an excessive price for Saquinavir and other protease inhibitors on third party reimbursement, particularly in an era when the social safety net (Medicaid, Medicare, Ryan White, and state AIDS drug assistance programs), upon which the great majority of people with HIV depend, is under unprecedented attack. FDA approval will be meaningless if protease inhibitors are priced beyond the ability of programs upon which the vast majority of people with AIDS are dependent are unable to afford them, or are forced to make unacceptable choices between expensive antiviral regimens versus lifeand health-extending opportunistic disease treatment and prophylaxis. We strongly urge Roche to set an example by pricing Saquinavir at an acceptable level and avoid intensifying a reimbursement crisis for people with HIV.
• Continued availability under expanded access in countries lacking accelerated approval. Many countries, e.g., in Europe, will not necessarily approve Saquinavir just because it was approved in the U.S.A. based on surrogate markers. Roche should undertake to assure that an adequate supply of the drug will remain available under expanded access in countries which will await evidence of clinical efficacy before approving Saquinavir.
• Drug interaction studies with commonly-used HIV and AIDS treatments. Surveillance through expanded access will be inadequate for assuring that there are not harmful interactions between Saquinavir, other antiretrovirals and anti-opportunistic disease drugs. We urge Roche to conduct pharmacokinetic interaction and activity studies with Saquinavir and the nucleosides, other protease inhibitors, and other commonly-used AIDS treatments, particularly the rifamycins (Rifampin and Rifabutin), the azoles (Fluconazole, Itraconazole, Ketoconazole, etc.), as well as with other commonly used medications such as birth control pills and methadone.

Potential post-marketing clinical efficacy studies

We believe Roche should commit to post-marketing studies designed to 1) confirm its currently sought indication by demonstrating clinical benefit with other nucleosides, e.g., ddI, d4T and 3TC and 2) extend its application to people with higher CD4 counts. Because it is not randomized the current expanded access program will tell us little about Saquinavir’s efficacy with other nucleosides. Among the most interesting populations to study would be those with medium CD4 counts (between 200-500) using first- or secondline therapy. Possible study designs include:

1. An early versus late trial. People could start, switch or stay on any nucleoside therapy they chose, and be randomized to start Saquinavir immediately of after a CD4 drop, a viral load increase or the development of symptoms;
2. A viral-load switchpoint trial. People could be randomized to start Saquinavir (used with any nucleoside at will) immediately or when their viral load rises past 150,000 copies and by more than three times baseline;
3. Standard-of-care alone versus SOC plus Saquinavir. People concerned about possible cross-resistance could be randomized to start Saquinavir (used with any nucleoside at will) or placebo (used with any nucleoside at will).

These study designs could be adapted for any of the new antiretrovirals currently up for approval, particularly 3TC, and could also be adapted for the protease inhibitors. Insurance could cover the underlying standard-of-care nucleosides, and the sponsors (Roche or Glaxo) could cover the experimental drug, placebo and data gathering costs. Studies could take place in community-based settings, public health clinics, through HMOs or in the approximately 1,200 doctors’ offices used to carrying out expanded access programs. Eventually, protease monotherapy versus combination or sequential therapy approaches should also be studied.