Importantly, enfuvirtide functional monother- apy was largely limited to a small proportion of patients who already had resistance to approved medications and ...
AIDS RESEARCH AND HUMAN RETROVIRUSES Volume 22, Number 11, 2006, pp. 1110–1112 © Mary Ann Liebert, Inc.
Seeking Ethical Designs for HIV Clinical Trials in TreatmentExperienced Patients: An Industry Perspective DOUGLAS L. MAYERS,1 JAIN CHUNG,2 VERONIKA M. KOHLBRENNER,1 DAVID B. HALL,1 RALPH A. DEMASI,3 DIETMAR NEUBACHER,1 NEIL E. BUSS,4 and MIKLOS P. SALGO2
INTRODUCTION
R
OCHE/TRIMERIS AND
BOEHRINGER INGELHEIM WELCOME THIS of the challenges of designing studies for highly treatment-experienced (TE) patients with HIV-1 infection. Studies in advanced patients must be designed in a way that balances the regulatory needs of the investigational compound (proof of safety and efficacy) and the clinical needs of patients coming into the studies. The TORO 1 and 21,2 (Roche/Trimeris) and RESIST 1 and 23,4 (Boehringer Ingelheim) studies were designed with this challenge in mind; the protocols were developed by consensus after extensive discussion with investigators, representatives from patient community groups, and the U.S. and European regulatory agencies. Both studies had strengths and potential limitations, which are discussed below; however, they were considered landmark studies that demonstrated that registrational studies can ensure the very best standard of care for all participating patients including those ultimately randomized to the control arm, meet the highest ethical standards, be scientifically rigorous, and still meet the regulatory requirements for studying the safety and efficacy of new antiretroviral compounds in highly TE patients with HIV. Recognizing the difficulties with designing appropriate clinical trials for TE patients, discussions continue in forums such as the Salvage Therapy II Workshop held on April 16–17, 2004 in Houston, Texas.5 ONGOING DISCUSSION
MATERIALS AND METHODS Minimizing the risk of functional monotherapy Both TORO and RESIST allowed patients to choose an optimized background (OB) regimen based on prior history as well as viral resistance testing made available during screening for the studies. When the TORO studies were designed in 1999–2000 lopinavir/ritonavir and tenofovir were both experimental agents available via expanded access programs and these
were permitted as part of the OB regimen in the TORO studies. The OB arm was therefore designed to ensure access to the best possible standard of care regimen resulting in significant treatment responses in the control arm of the TORO studies. The significant contribution of adding enfuvirtide to this optimized background therefore reinforces the robustness of the study outcome. Importantly, enfuvirtide functional monotherapy was largely limited to a small proportion of patients who already had resistance to approved medications and also had failed lopinavir/ritonavir and tenofovir. These patients would have had to wait for at least 2 years for the tipranavir compassionate use program to open (2003) in order to have the opportunity of adding another potentially active agent to their OB. Similarly, when the RESIST studies were initiated in early 2003 there were no new agents other than enfuvirtide available or likely to be available within a 2-year timeframe. Patients ending up on functional monotherapy did so either because they had extensive multiclass drug cross-resistance or had prior use and/or prior toxicity, which prevented use of some drugs as part of the OB regimen. Would it have been ethical to deny these patients, who were facing possible rapid clinical deterioration, the chance to benefit from a new agent in conjunction with an OB selected by clinical history and drug resistance testing? We agree with the goal of minimizing risks associated with functional monotherapy that result from use of a new antiretroviral in multidrug-experienced patients for whom there are no active drugs with which to pair the new agent. Although intuitively appealing, the major challenge to the modified factorial design proposed by De Gruttola et al. (previous paper) and by others5 is that it requires the simultaneous availability of two or more new active antiretrovirals that can be used together and to which patients’ viruses remain fully susceptible—a situation that rarely occurs in highly TE patients. It is uncommon for several new drugs with activity against drug-resistant viruses to enter Phase III simultaneously. If new drugs were combined in a pivotal trial, the future of each drug becomes dependent on the success of the other drug, and failure of one (as recently experienced with a number of drugs in Phase II/III) could jeopar-
1Boehringer
Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut. Nutley, New Jersey. 3Trimeris Inc., Morrisville, North Carolina. 4Roche, Basel, Switzerland. 2Roche,
1110
ETHICAL DESIGNS FOR HIV CLINICAL TRIALS dize the other. It would also be difficult to ascertain the associated safety profile of each contributing drug, forever linking the two drugs together with a combined safety profile. We accept that there is a need for cooperation across pharmaceutical companies in order to design and implement such studies. There is extensive discussion and cooperation across companies for development of new agents that was originally facilitated by the Intercompany Collaborative Group. In parallel with registrational trials, studies exploring combinations of new drugs using factorial designs may be desirable to further elucidate the role of combining new active agents. Importantly, this would require early collaboration in order to rule out negative drug–drug interactions that would preclude concomitant use. Boehringer Ingelheim has made extensive efforts to obtain pharmacokinetic interaction data for tipranavir combined with investigational drugs being studied in TE patients in order to make tipranavir available for patients enrolling in upcoming Phase II/III trials. The proposed alternative to the RESIST study design, with different randomized treatments based on availability of active drugs in addition to enfuvirtide, requires, for some analyses, that some patients refuse enfuvirtide. This might occur with an injectable drug, but is less likely to occur with a new oral medication. Also, as pointed out by the authors, comparisons of nonrandomized groups (e.g., groups defined by self-selection of use or refusal to use a given agent) can lead to biases as sicker patients may choose to use the drug in question, as clearly happened with enfuvirtide use in the RESIST trials.
RESULTS Early switch options: Impact on efficacy and safety analyses In the TORO and RESIST studies patients randomized to the control arm who met protocol-defined criteria for virologic failure after week 8 were permitted to revise their background regimen and to add enfuvirtide (TORO) or tipranavir (RESIST). The timing of the option to switch and the protocol-defined virologic failure criteria were chosen based on a consensus across the various stakeholders noted above, and reflected the timing and virologic failure criteria that were acceptable in clinical practice for such advanced patients at that time. Furthermore, viral load reductions over time generally demonstrate an initial steep decline followed by a slower rate of reduction from weeks 4 to 8. Therefore the appropriate data-handling rules implemented after week 8 are unlikely to bias the efficacy assessments if the rule is applied to both treatment arms. As indicated by De Gruttola et al. (previous paper), the switch option, in conjunction with the best standard of care for all patients, may have made the studies more acceptable to patients, thus leading to successful accrual while ensuring that all patients would have access to the new investigational agent if needed. This design did indeed result in control arms that dwindled due to patients switching after having achieved virologic failure. However, in both studies, prior to patients switching (week 8) a statistically significant treatment benefit of the investigational agent was already evident.7,8 Several sensitivity analyses based on various data-handling rules and analysis models for week-
1111 24 data also have demonstrated the robustness of the TORO results.9 For safety assessments the number of patients enrolled and the duration of control arm treatment and exposure (patientyears) allowed for meaningful comparisons for safety. In both the TORO and RESIST trials, exposures were diminished but remained generally comparable between the investigational drug and control arms at week 24, but became significantly reduced due to switch and dropouts in the control arms by week 48. Given this changing ratio of exposure due to the switch study design, comparisons of rates of adverse events at week 24 reported percentage of patients with an event, while the week 48 and subsequent analyses additionally presented the exposure-adjusted incidence, expressed as patients with at least one event per 100 patient-years of exposure. De Gruttola et al. (previous paper) suggest that crossover should be delayed for as long as possible and that the decision to switch should be based on CD4 cell counts rather than viral load. It is unlikely that either a delay in switching or the use of immunologic criteria would be acceptable to patients or their providers. Failure based on CD4 cell counts would likely occur many months after virologic failure and keeping patients on a virologically failing regimen is not recommended in current treatment guidelines due to progressive development of drug resistance.10
Controlling potential bias The TORO and RESIST studies were designed to ensure an optimal standard of care for all participating patients thereby minimizing the motivation for patients randomized to the control arm to deliberately fail and switch to enfuvirtide or tipranavir. However, the motivation for patients to deliberately fail and switch to enfuvirtide or tipranavir was discussed with stakeholders during the design phase of the study, and extensive safeguards were incorporated into the studies to help minimize and to document occurrences of potential biases. In addition, criteria for virologic failure and timing for switch that were considered fair and appropriate for the patient population were included. The OB regimen was designed to be the best possible regimen for the patient based on prior history and resistance testing and the choice of OB was committed to and documented prior to randomization. No changes to the OB regimen were permitted except for toxicity management or upon meeting criteria for virologic failure. Further checks of adherence to the OB regimen were carried out based on 4-day recall at every visit (TORO), and also by plasma levels for antiretrovirals (ARVs) drawn prior to virologic failure (TORO and RESIST). In both trial programs, switch to the new drug was allowed only for patients with documented virologic failure, and retrospective analyses (TORO) of the plasma levels demonstrated very low rates of irregularities. Thus, the potential for intentional virologic failure and resulting biases that would have been introduced into the study were recognized and openly discussed during the study design phase. The most important way of preventing intentional virologic failure was to design a study that provided the best possible standard of care in the control arm and was considered fair and patient friendly, both minimizing the motivation to violate protocol criteria, and monitoring and discouraging any protocol violations during the study.
1112
MAYERS ET AL.
DISCUSSION Study endpoints Prior to the TORO studies, treatment guidelines suggested that the goal of therapy for highly TE patients was immunologic stabilization, reflecting that achieving complete virologic suppression was not realistic for the vast majority of such patients. The TORO and RESIST studies themselves helped define a new goal of therapy of viral suppression to 400 or 50 copies/ml as now achievable in a large proportion of highly TE patients. TORO and RESIST also demonstrated the importance of having at least two fully active antiretrovirals—one of which is a boosted PI—in the OB regimen to achieve this suppression. With respect to study endpoints, the TORO studies confirmed that the principles established by Eron et al.11 and Gulick et al.12 apply to HIV-positive individuals throughout the treatment continuum.
Resistance classes As the number of antiretroviral agents grows there is clearly a need to be able to identify those TE patients who will benefit from a particular investigational agent and those who will not. TORO was the first large registration study to include genotypic and phenotypic viral resistance testing for all patients to help investigators choose an effective OB regimen. The RESIST trials offered patients genotypic or virtual phenotypic resistance testing at screening. These studies helped further prove the benefit of such testing in this population.13 In addition, the TORO studies were the first registrational studies to incorporate the concept of the Genotypic or Phenotypic Sensitivity Score (GSS and PSS), first proposed by De Gruttola et al. (previous paper) into the analysis of study results, further establishing the need for two or more active antiretrovirals (GSS or PSS 2) in the OB regimen to achieve the greatest response rates.13 The insights developed in the TORO program were confirmed in the RESIST trials, which again demonstrated the need for additional active OB drugs to be combined with a new agent to obtain durable viral suppression in TE patients. For new agents within existing classes, susceptibility to the new agent must be considered in addition to the availability of active agents for the OB and, as stated by De Gruttola et al.,6 a compromise must be established between excluding patients unlikely to benefit from therapy and including enough patients to define resistance patterns that correlate with treatment failure. The challenge in all of these programs is to find the additional active OB drugs when a new drug becomes available. In summary, the TORO and RESIST studies were designed with input from all the major stakeholders, including investigators (and several authors of the companion editorial), patient advocates, and regulatory agencies. Although some may argue not ideal, they represented the state of the art for studies in TE patients. This design continues to be used for many new compounds in development. Hopefully these companion pieces will result in a renewed dialogue involving industry, regulatory authorities, academia, and the patient community to develop a new generation of clinical trial designs for highly TE patients.
REFERENCES 1. Lalezari JP, Henry K, O’Hearn M, et al.: Enfuvirtide, an HIV-1 fusion inhibitor, for drug-resistant HIV infection in North and South America. N Engl J Med 2003;348:2175–2185. 2. Lazzarin A, Clotet B, Cooper D, et al.: Efficacy of enfuvirtide in patients infected with drug-resistant HIV-1 in Europe and Australia. N Engl J Med 2003;348:2186–2195. 3. Cahn P for the RESIST-2 study team: 24-week data from RESIST2: Phase 3 study of the efficacy and safety of either tipranavir/ritonavir (TPV/r) or an optimized ritonavir (RTV)-boosted standardof-care (SOC) comparator PI (CPI) in a large randomised multicentre trial in treatment-experienced HIV patients. 7th International Congress on Drug Therapy in HIV Infection, Glasgow, UK, November 14–18, 2004. Abstract PL14.3. 4. Hicks C for the RESIST-1 study team: RESIST-1: A phase 3, randomized, controlled, open-label, multicenter trial comparing tipranavir/ritonavir (TPV/r) to an optimized comparator protease inhibitor/r (CPI/r) regimen in antiretroviral (ARV) experienced patients: 24-week data. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, DC, October 30–November 2, 2004. Abstract H-1137a. 5. Struble K, Murray J, Cheng B, Gegeny T, Miller V, and Gulick R: Antiretroviral therapies for treatment-experienced patients: Current status and research challenges. AIDS 2005;19:747–756. 6. De Gruttola V, Dix L, D’Aquila R, et al.: The relation between baseline HIV drug resistance and response to antiretroviral therapy: Re-analysis of retrospective and prospective studies using a standardized data analysis plan. Antiviral Ther 2000;5:41–48. 7. McCallister S, Kohlbrenner V, Villacian J, Mukwaya G, Neubacher D, and Mayers D: 24-week combined analysis of the TPV RESIST studies of 1483 treatment-experienced patients given either tipranavir/ritonavir (TPV/r) or an optimized standard of care regimen using 1 of 4 RTV-boosted comparator PIs (CPI/r). HIV DART 2004, Montego Bay, Jamaica, December 12–16, 2004. Abstract 060. 8. Katlama C, Raffi F, Saag M, et al.: Week 12 response as a predictor of week 24, 48 and 96 outcome in patients receiving the HIV fusion inhibitor enfuvirtide in the TORO studies. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington DC, October 30–November 2, 2004. Poster H-583. 9. Huson L, Chung J, and Salgo M: Comparison of methods of missing-data imputation in analysis of two Phase III trials treating HIV1 infection. 2004 Joint Statistical Meetings, Toronto, Canada, August 8–12, 2004. Abstract 156. 10. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents: Department of Health and Human Services (DHHS). (Accessed October 10, 2005, at http://aidsinfo.nih.gov/ guidelines.) 11. Eron J, Benoit S, Jemsek J, et al.: Treatment with lamivudine, zidovudine, or both in HIV-positive patients with 200 to 500 CD4 cells per cubic millimeter. N Engl J Med 1995;333:1662–1669. 12. Gulick R, Mellors J, Havlir D, et al.: Treatment with indinavir, zidovudine, and lamivudine in adults with human immunodeficiency virus infection and prior antiretroviral therapy. N Engl J Med 1997;337:734–739. 13. Montaner J, Chung J, Guimaraes D, DeMasi R, Gafoor Z, and Salgo M: A treatment staging proposal derived from a post-hoc analysis of the 48 week results from the TORO studies. XV International AIDS Conference, Bangkok, Thailand, July 11–16, 2004. Abstract TuPeB4483.
Address reprint requests to: Douglas L. Mayers Boehringer Ingelheim Pharmaceuticals, Inc. Ridgefield, Connecticut
