Year: 2017
Type: article
Abstract: (See the Major Article by Bonnett et al on pages 46–54.) Tuberculosis is the leading infectious cause of death worldwide [1]. One of the 3 pillars of the World Health Organization End TB Strategy is intensified research and innovation, with a focus on the development of new interventions to end the global epidemic [2]. These interventions involve shorter, safer, more effective combination regimens that result in very high cure rates in program settings, not just in clinical trials. The use of an evidence-based clinical development pathway is essential to ensure that the most effective regimens are quickly identified and advanced into phase 3 definitive evaluation and while dwindling resources for tuberculosis research and development [3] are not wasted on ineffective regimens. What is often described as the traditional clinical development pathway for tuberculosis drug and regimen development—phase 2A to evaluate early bactericidal activity (EBA) over 14 days, phase 2B to evaluate 2-month culture conversion, and phase 3 to evaluate relapse—is relatively new. Indeed only trials for the drugs delamanid and bedaquiline and the pretomanid-based regimens have strictly followed this pathway (with the phase 3 trials for each still ongoing; clinicaltrials.gov identifiers: NCT01424670, NCT02409290, NCT02342886). Historically, the development of regimens for tuberculosis did not follow such a formal pathway. This development was largely led and funded by public agencies such as the British Medical Research Council [4] and the US Public Health Service [5], with a focus on trials that invariably included 2–3 years of follow-up with a primary outcome of relapse. These trials, which today would be described as phase 3, were initiated as soon as there was adequate confidence in the bactericidal and sterilizing activities of the individual drugs. The introduction of rifampicin is a case in point [6]. Nevertheless, the bacteriological endpoints included in the metaanalysis by Bonnet et al in this issue [REF] continue to be the cornerstone of the evaluation of regimens for progression to phase 3, and this excellent and comprehensive review is timely. While the rank order of regimens is as expected—regimens with isoniazid and rifampicin generally perform better than others on all endpoints studied—this is the first study to systematically examine the available data in this fashion and therefore provides invaluable evidence to support regimen choices and inform future regimen development. However, the heterogeneity and variability of the pooled estimates are striking and beg the question: What now is the role of these endpoints in regimen development? No combination regimens were shown to have EBA over 2 or 7 days that was greater than that of isoniazid monotherapy, indicating that these endpoints are not useful for comparing regimens. Fewer regimens evaluated with EBA were over 14 days, but the wide confidence intervals again show that this endpoint is likely only to detect very large differences between regimens in the studies that are typically conducted with 15 patients per arm. Considering the endpoint of percentage culture negative at 8 weeks on solid media, there is very little difference between the results of any of the regimens that contain isoniazid, rifampicin, and at least 1 companion drug. Furthermore, the pooled estimate for the standard regimen of isoniazid, rifampicin, pyrazinamide and ethmabutol (HRZE) is around 90% culture negative at 8 weeks, leaving very little room for improvement. A trial with even 80% power to show a difference from 90% to 95% would require 435 patients per arm. Bonnet et al conclude that current phase 2A and 2B trials are too small to detect modest differences between regimens in these studies. It is an open question whether modest differences in these phase 2 endpoints are likely to reflect clinically relevant differences in clinically meaningful endpoints. In any case, what are the alternatives? The authors rightly call for more efficient adaptive screening trial designs to allow more regimens to be evaluated in phase 2 trials, and sponsors and investigators should also think more broadly about the whole clinical development pathway. This review shows the importance of clinically relevant endpoints in tuberculosis trials—endpoints that measure “directly how a patient feels, functions or survives” [7]. A key objective in clinical development must be to start studies that evaluate relapse as early as is safely possible in the development pathway to select the most promising regimens and reduce the risk of large and expensive phase 3 failures. Approaches to meet this objective include using larger margins of noninferiority to facilitate smaller and more nimble noninferiority phase 3 trials (such as the 12% margin in shortening treatment by advancing novel drugs trial (STAND), NCT02342886), studying a regimen in patients with extensively drug-resistant tuberculosis where there is an even greater need for new treatments, and studying the intended regimen duration prior to phase 3 (the selection trial with extended posttreatment follow-up (STEP) phase 2C design [8]). The authors call for more combinations to be studied across a greater range of phase 2 endpoints in order to provide a more precise evaluation of the role of each endpoint in future metaanalyses. While such activities would allow for a better understanding of the role of phase 2, this metaanalysis has highlighted the limitations of these endpoints, and there will always be a balance between studying more phase 2 endpoints to fill the data gaps and investing more resources to study clinically relevant endpoints that will be more useful for comparing regimens. The authors also highlight the importance of a core outcome set for tuberculosis trials and of sharing individual patient data from tuberculosis trials. These two activities would facilitate further metaanalyses that allow for systematizing knowledge across trials. Only a few groups throughout the world are planning and conducting phase 2 and phase 3 tuberculosis clinical trials, and agreement on core outcomes should not be an impossible task with support from projects such as the core outcome measures in effectiveness trials (COMET) initiative [9]. Similarly, repositories are now available for tuberculosis clinical trial data that investigators and sponsors are encouraged to make use of (eg, the TB-platform for aggregation on clinical TB studies (TB-PACTS); http://c-path.org/programs/tb-pacts/), although the resources and expertise required to ensure proper data aggregation and curation, and appropriate controlled access should not be underestimated [10]. In summary, this well-conducted systematic review provides evidence toward setting tuberculosis regimen development on a firm foundation. Additional work is needed as more data become available. In the interim, clinically relevant endpoints must remain the focus of development of new regimens for tuberculosis. Potential conflicts of interest. The author reports no conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Source: Clinical Infectious Diseases
DOI: 10.1093/cid/cix250
Author: Patrick PJ Phillips
Language: English
Subfield: Infectious Diseases
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2605649728