Efficacy Analysis in Healthy-Volunteer Influenza Challenge Trials: Intention To Treat.
Efficacy Analysis in Healthy-Volunteer Influenza Challenge Trials: Intention To Treat.
复制标题
健康志愿者流感挑战试验的功效分析:治疗意向。
DOI:
10.1128/aac.02018-17
复制
发表时间:
2018
影响因子:
4.9
通讯作者:
Memoli,MatthewJ
中科院分区:
文献类型:
--
作者:
Hunsberger,Sally;Memoli,MatthewJ
Hunsberger and Memoli’s comment (1) challenges the appropriateness of our approach to evaluate only virological endpoints in subjects with evidence of an active influenza A virus (IAV) infection (the intention-to-treat infected [ITTI] population) in the influenza challenge model used in our phase 2 study (2). The authors have cited vaccine trials intended to assess prevention of disease or to diminish the severity of disease of those with virus on board. Had our study been a prophylactic study to prevent influenza virus infection, these concerns regarding the proportion of subjects infected would be appropriate; however, MHAA4549A was not a vaccine but rather a therapeutic intended for treatment of already-active influenza A infections. The treatment model was designed to simulate a clinical setting with natural infection; as such, all subjects were treated approximately 24 h after IAV inoculation. For a prophylactic study, the treatment would have been administered prior to inoculation with virus. It is well appreciated that despite being inoculated, not all healthy subjects are expected to have any detectable infection in these challenge models (3–8). Our intent was not to demonstrate clinical efficacy but rather to understand the pharmacokinetic/pharmacodynamic relationship. These results not only provide proof of activity but also inform MHAA4549A dose selection, which is required for assessing true clinical efficacy in larger, randomized, blind clinical trials in hospitalized-patient and outpatient populations with laboratory-confirmed infection. The ITTI selection was predefined per the study protocol and included subjects with any evidence of viral load or seroconversion on any given day. Rather than the rate of infection, the endpoint proposed by Hunsberger and Memoli (1), the most appropriate outcome measures were area under the viral-load curve (AUC) and the duration of viral shedding to determine if MHAA4549A could reduce the extent and duration of infection within the infected subjects. These measurements included both qPCR results and the 50% tissue culture infective dose (TCID50), two measures of viral burden. If any bias was introduced, ITTI selection may have underestimated the efficacy within our treatment arms by excluding those individuals who, as a result of treatment, did not go on to have detectable virus infection and did not exhibit subsequent seroconversion. ITTI selection in treatment studies is accepted as a standard approach and has been applied when determining proof of activity for novel antivirals or immune measures in IAV or rhinovirus challenge models (3–8). In the Hayden review cited by the Hunsberger and Memoli, both prophylactic and treatment studies are described (9). For the cited prophylactic studies, the proportions of all study subjects that became infected were reported (3, 4). However, in the cited treatment studies, and in additional studies conducted later, the selection of an ITTI population was applied when assessing