Estimation of Stratified Mark-Specific Proportional Hazards Models Under Two-Phase Sampling with Application to HIV Vaccine Efficacy Trials

An objective of preventive HIV vaccine efficacy trials is to understand how vaccine-induced immune responses to specific protein sequences of HIV-1 associate with subsequent infection with specific sequences of HIV, where the immune response biomarkers are measured in vaccine recipients via a two-phase sampling design. Motivated by this objective, we investigate the stratified mark-specific proportional hazards model under two-phase biomarker sampling, where the mark is the genetic distance of an infecting HIV-1 sequence to an HIV-1 sequence represented inside the vaccine. Estimation and inference procedures based on inverse probability weighting of complete-cases and on augmented inverse probability weighting of complete-cases are developed. Asymptotic properties of the estimators are derived and their finite-sample performances are examined in simulation studies. The methods are shown to have satisfactory performance and are applied to the RV144 vaccine trial to assess whether immune response correlates of HIV-1 infection are stronger for HIV-1 infecting sequences similar to the vaccine than for sequences distant from the vaccine.     

Weighted Likelihood Method for Grouped Survival Data in Case–Cohort Studies with Application to HIV Vaccine Trials

Summary Grouped failure time data arise often in HIV studies. In a recent preventive HIV vaccine efficacy trial, immune responses generated by the vaccine were measured from a case–cohort sample of vaccine recipients, who were subsequently evaluated for the study endpoint of HIV infection at prespecified follow‐up visits. Gilbert et al. (2005, Journal of Infectious Diseases 191 , 666–677) and Forthal et al. (2007, Journal of Immunology 178, 6596–6603) analyzed the association between the immune responses and HIV incidence with a Cox proportional hazards model, treating the HIV infection diagnosis time as a right‐censored random variable. The data, however, are of the form of grouped failure time data with case–cohort covariate sampling, and we propose an inverse selection probability‐weighted likelihood method for fitting the Cox model to these data. The method allows covariates to be time dependent, and uses multiple imputation to accommodate covariate data that are missing at random. We establish asymptotic properties of the proposed estimators, and present simulation results showing their good finite sample performance. We apply the method to the HIV vaccine trial data, showing that higher antibody levels are associated with a lower hazard of HIV infection.     

The 2-sample problem for failure rates depending on a continuous mark: an application to vaccine efficacy

The efficacy of an HIV vaccine to prevent infection is likely to depend on the genetic variation of the exposing virus. This paper addresses the problem of using data on the HIV sequences that infect vaccine efficacy trial participants to (1) test for vaccine efficacy more powerfully than procedures that ignore the sequence data and (2) evaluate the dependence of vaccine efficacy on the divergence of infecting HIV strains from the HIV strain that is contained in the vaccine. Because hundreds of amino acid sites in each HIV genome are sequenced, it is natural to treat the genetic divergence as a continuous mark variable that accompanies each failure (infection) time. Problems (1) and (2) can then be approached by testing whether the ratio of the mark-specific hazard functions for the vaccine and placebo groups is unity or independent of the mark. We develop nonparametric and semiparametric tests for these null hypotheses and nonparametric techniques for estimating the mark-specific relative risks. The asymptotic properties of the procedures are established. In addition, the methods are studied in simulations and are applied to HIV genetic sequence data collected in the first HIV vaccine efficacy trial.     

RV144 HIV-1 vaccination impacts post-infection antibody responses

The RV144 vaccine efficacy clinical trial showed a reduction in HIV-1 infections by 31%. Vaccine efficacy was associated with stronger binding antibody responses to the HIV Envelope (Env) V1V2 region, with decreased efficacy as responses wane. High levels of Ab-dependent cellular cytotoxicity (ADCC) together with low plasma levels of Env-specific IgA also correlated with decreased infection risk. We investigated whether B cell priming from RV144 vaccination impacted functional antibody responses to HIV-1 following infection. Antibody responses were assessed in 37 vaccine and 63 placebo recipients at 6, 12, and 36 months following HIV diagnosis. The magnitude, specificity, dynamics, subclass recognition and distribution of the binding antibody response following infection were different in RV144 vaccine recipients compared to placebo recipients. Vaccine recipients demonstrated increased IgG1 binding specifically to V1V2, as well as increased IgG2 and IgG4 but decreased IgG3 to HIV-1 Env. No difference in IgA binding to HIV-1 Env was detected between the vaccine and placebo recipients following infection. RV144 vaccination limited the development of broadly neutralizing antibodies post-infection, but enhanced Fc-mediated effector functions indicating B cell priming by RV144 vaccination impacted downstream antibody function. However, these functional responses were not associated with clinical markers of disease progression. These data reveal that RV144 vaccination primed B cells towards specific binding and functional antibody responses following HIV-1 infection.     

Machine Learning Methods Enable Predictive Modeling of Antibody Feature:Function Relationships in RV144 Vaccinees

The adaptive immune response to vaccination or infection can lead to the production of specific antibodies to neutralize the pathogen or recruit innate immune effector cells for help. The non-neutralizing role of antibodies in stimulating effector cell responses may have been a key mechanism of the protection observed in the RV144 HIV vaccine trial. In an extensive investigation of a rich set of data collected from RV144 vaccine recipients, we here employ machine learning methods to identify and model associations between antibody features (IgG subclass and antigen specificity) and effector function activities (antibody dependent cellular phagocytosis, cellular cytotoxicity, and cytokine release). We demonstrate via cross-validation that classification and regression approaches can effectively use the antibody features to robustly predict qualitative and quantitative functional outcomes. This integration of antibody feature and function data within a machine learning framework provides a new, objective approach to discovering and assessing multivariate immune correlates.     

The Thai phase III trial (RV144) vaccine regimen induces T cell responses that preferentially target epitopes within the V2 region of HIV-1 envelope

The Thai HIV phase III prime/boost vaccine trial (RV144) using ALVAC-HIV (vCP1521) and AIDSVAX B/E was, to our knowledge, the first to demonstrate acquisition efficacy. Vaccine-induced, cell-mediated immune responses were assessed. T cell epitope mapping studies using IFN-γ ELISPOT was performed on PBMCs from HIV-1-uninfected vaccine (n = 61) and placebo (n = 10) recipients using HIV-1 Env peptides. Positive responses were measured in 25 (41%) vaccinees and were predominantly CD4(+) T cell-mediated. Responses were targeted within the HIV Env region, with 15 of 25 (60%) of vaccinees recognizing peptides derived from the V2 region of HIV-1 Env, which includes the α(4)β(7) integrin binding site. Intracellular cytokine staining confirmed that Env responses predominated (19 of 30; 63% of vaccine recipients) and were mediated by polyfunctional effector memory CD4(+) T cells, with the majority of responders producing both IL-2 and IFN-γ (12 of 19; 63%). HIV Env Ab titers were higher in subjects with IL-2 compared with those without IL-2-secreting HIV Env-specific effector memory T cells. Proliferation assays revealed that HIV Ag-specific T cells were CD4(+), with the majority (80%) expressing CD107a. HIV-specific T cell lines obtained from vaccine recipients confirmed V2 specificity, polyfunctionality, and functional cytolytic capacity. Although the RV144 T cell responses were modest in frequency compared with humoral immune responses, the CD4(+) T cell response was directed to HIV-1 Env and more particularly the V2 region.     

CD8 and CD4 Epitope Predictions in RV144: No Strong Evidence of a T-Cell Driven Sieve Effect in HIV-1 Breakthrough Sequences from Trial Participants

The modest protection afforded by the RV144 vaccine offers an opportunity to evaluate its mechanisms of protection. Differences between HIV-1 breakthrough viruses from vaccine and placebo recipients can be attributed to the RV144 vaccine as this was a randomized and double-blinded trial. CD8 and CD4 T cell epitope repertoires were predicted in HIV-1 proteomes from 110 RV144 participants. Predicted Gag epitope repertoires were smaller in vaccine than in placebo recipients (p = 0.019). After comparing participant-derived epitopes to corresponding epitopes in the RV144 vaccine, the proportion of epitopes that could be matched differed depending on the protein conservation (only 36% of epitopes in Env vs 84–91% in Gag/Pol/Nef for CD8 predicted epitopes) or on vaccine insert subtype (55% against CRF01_AE vs 7% against subtype B). To compare predicted epitopes to the vaccine, we analyzed predicted binding affinity and evolutionary distance measurements. Comparisons between the vaccine and placebo arm did not reveal robust evidence for a T cell driven sieve effect, although some differences were noted in Env-V2 (0.022≤p-value≤0.231). The paucity of CD8 T cell responses identified following RV144 vaccination, with no evidence for V2 specificity, considered together both with the association of decreased infection risk in RV 144 participants with V-specific antibody responses and a V2 sieve effect, lead us to hypothesize that this sieve effect was not T cell specific. Overall, our results did not reveal a strong differential impact of vaccine-induced T cell responses among breakthrough infections in RV144 participants.     

Comprehensive Sieve Analysis of Breakthrough HIV-1 Sequences in the RV144 Vaccine Efficacy Trial

The RV144 clinical trial showed the partial efficacy of a vaccine regimen with an estimated vaccine efficacy (VE) of 31% for protecting low-risk Thai volunteers against acquisition of HIV-1. The impact of vaccine-induced immune responses can be investigated through sieve analysis of HIV-1 breakthrough infections (infected vaccine and placebo recipients). A V1/V2-targeted comparison of the genomes of HIV-1 breakthrough viruses identified two V2 amino acid sites that differed between the vaccine and placebo groups. Here we extended the V1/V2 analysis to the entire HIV-1 genome using an array of methods based on individual sites, k-mers and genes/proteins. We identified 56 amino acid sites or “signatures” and 119 k-mers that differed between the vaccine and placebo groups. Of those, 19 sites and 38 k-mers were located in the regions comprising the RV144 vaccine (Env-gp120, Gag, and Pro). The nine signature sites in Env-gp120 were significantly enriched for known antibody-associated sites (p = 0.0021). In particular, site 317 in the third variable loop (V3) overlapped with a hotspot of antibody recognition, and sites 369 and 424 were linked to CD4 binding site neutralization. The identified signature sites significantly covaried with other sites across the genome (mean = 32.1) more than did non-signature sites (mean = 0.9) (p < 0.0001), suggesting functional and/or structural relevance of the signature sites. Since signature sites were not preferentially restricted to the vaccine immunogens and because most of the associations were insignificant following correction for multiple testing, we predict that few of the genetic differences are strongly linked to the RV144 vaccine-induced immune pressure. In addition to presenting results of the first complete-genome analysis of the breakthrough infections in the RV144 trial, this work describes a set of statistical methods and tools applicable to analysis of breakthrough infection genomes in general vaccine efficacy trials for diverse pathogens.     

Nonparametric estimation of the joint distribution of a survival time subject to interval censoring and a continuous mark variable

This article considers three nonparametric estimators of the joint distribution function for a survival time and a continuous mark variable when the survival time is interval censored and the mark variable may be missing for interval-censored observations. Finite and large sample properties are described for the nonparametric maximum likelihood estimator (NPMLE) as well as estimators based on midpoint imputation (MIDMLE) and coarsening the mark variable (CMLE). The estimators are compared using data from a simulation study and a recent phase III HIV vaccine efficacy trial where the survival time is the time from enrollment to infection and the mark variable is the genetic distance from the infecting HIV sequence to the HIV sequence in the vaccine. Theoretical and empirical evidence are presented indicating the NPMLE and MIDMLE are inconsistent. Conversely, the CMLE is shown to be consistent in general and thus is preferred.     

A Novel HIV Vaccine Adjuvanted by IC31 Induces Robust and Persistent Humoral and Cellular Immunity

The HIV vaccine strategy that, to date, generated immune protection consisted of a prime-boost regimen using a canarypox vector and an HIV envelope protein with alum, as shown in the RV144 trial. Since the efficacy was weak, and previous HIV vaccine trials designed to generate antibody responses failed, we hypothesized that generation of T cell responses would result in improved protection. Thus, we tested the immunogenicity of a similar envelope-based vaccine using a mouse model, with two modifications: a clade C CN54gp140 HIV envelope protein was adjuvanted by the TLR9 agonist IC31®, and the viral vector was the vaccinia strain NYVAC-CN54 expressing HIV envelope gp120. The use of IC31® facilitated immunoglobulin isotype switching, leading to the production of Env-specific IgG2a, as compared to protein with alum alone. Boosting with NYVAC-CN54 resulted in the generation of more robust Th1 T cell responses. Moreover, gp140 prime with IC31® and alum followed by NYVAC-CN54 boost resulted in the formation and persistence of central and effector memory populations in the spleen and an effector memory population in the gut. Our data suggest that this regimen is promising and could improve the protection rate by eliciting strong and long-lasting humoral and cellular immune responses.     

Cross-Serotype Immunity Induced by Immunization with a Conserved Rhinovirus Capsid Protein

Human rhinovirus (RV) infections are the principle cause of common colds and precipitate asthma and COPD exacerbations. There is currently no RV vaccine, largely due to the existence of ∼150 strains. We aimed to define highly conserved areas of the RV proteome and test their usefulness as candidate antigens for a broadly cross-reactive vaccine, using a mouse infection model. Regions of the VP0 (VP4+VP2) capsid protein were identified as having high homology across RVs. Immunization with a recombinant VP0 combined with a Th1 promoting adjuvant induced systemic, antigen specific, cross-serotype, cellular and humoral immune responses. Similar cross-reactive responses were observed in the lungs of immunized mice after infection with heterologous RV strains. Immunization enhanced the generation of heterosubtypic neutralizing antibodies and lung memory T cells, and caused more rapid virus clearance. Conserved domains of the RV capsid therefore induce cross-reactive immune responses and represent candidates for a subunit RV vaccine.     

FCGR2C Polymorphisms Associated with HIV-1 Vaccine Protection Are Linked to Altered Gene Expression of Fc-γ Receptors in Human B Cells

The phase III Thai RV144 vaccine trial showed an estimated vaccine efficacy (VE) to prevent HIV-1 infection of 31.2%, which has motivated the search for immune correlates of vaccine protection. In a recent report, several single nucleotide polymorphisms (SNPs) in FCGR2C were identified to associate with the level of VE in the RV144 trial. To investigate the functional significance of these SNPs, we utilized a large scale B cell RNA sequencing database of 462 individuals from the 1000 Genomes Project to examine associations between FCGR2C SNPs and gene expression. We found that the FCGR2C SNPs that associated with vaccine efficacy in RV144 also strongly associated with the expression of FCGR2A/C and one of them also associated with the expression of Fc receptor-like A (FCRLA), another Fc-γ receptor (FcγR) gene that was not examined in the previous report. These results suggest that the expression of FcγR genes is influenced by these SNPs either directly or in linkage with other causal variants. More importantly, these results motivate further investigations into the potential for a causal association of expression and alternative splicing of FCGR2C and other FcγR genes with the HIV-1 vaccine protection in the RV144 trial and other similar studies.     

HIV-1 Envelope Resistance to Proteasomal Cleavage: Implications for Vaccine Induced Immune Responses

Background

Antigen processing involves many proteolytic enzymes such as proteasomes and cathepsins. The processed antigen is then presented on the cell surface bound to either MHC class I or class II molecules and induces/interacts with antigen-specific CD8+ and CD4+ T-cells, respectively. Preliminary immunological data from the RV144 phase III trial indicated that the immune responses were biased towards the Env antigen with a dominant CD4+ T-cell response.

Methods

In this study, we examined the susceptibility of HIV-1 Env-A244 gp120 protein, one of the protein boost subunits of the RV144 Phase III vaccine trial, to proteasomes and cathepsins and identified the generated peptide epitope repertoire by mass spectrometry. The peptide fragments were tested for cytokine production in CD4⁺ T-cell lines derived from RV144 volunteers.

Results

Env-A244 was resistant to proteasomes, thus diminishing the possibility of the generation of class I epitopes by the classical MHC class I pathway. However, Env-A244 was efficiently cleaved by cathepsins generating peptide arrays identified by mass spectrometry that contained both MHC class I and class II epitopes as reported in the Los Alamos database. Each of the cathepsins generated distinct degradation patterns containing regions of light and dense epitope clusters. The sequence DKKQKVHALF that is part of the V2 loop of gp120 produced by cathepsins induced a polyfunctional cytokine response including the generation of IFN-γ from CD4⁺ T-cell lines-derived from RV144 vaccinees. This sequence is significant since antibodies to the V1/V2-loop region correlated inversely with HIV-1 infection in the RV144 trial.

Conclusions

Based on our results, the susceptibility of Env-A244 to cathepsins and not to proteasomes suggests a possible mechanism for the generation of Env-specific CD4⁺T cell and antibody responses in the RV144 vaccinees.     

Gene-based vaccination with a mismatched envelope protects against simian immunodeficiency virus infection in nonhuman primates

The RV144 trial demonstrated that an experimental AIDS vaccine can prevent human immunodeficiency virus type 1 (HIV-1) infection in humans. Because of its limited efficacy, further understanding of the mechanisms of preventive AIDS vaccines remains a priority, and nonhuman primate (NHP) models of lentiviral infection provide an opportunity to define immunogens, vectors, and correlates of immunity. In this study, we show that prime-boost vaccination with a mismatched SIV envelope (Env) gene, derived from simian immunodeficiency virus SIVmac239, prevents infection by SIVsmE660 intrarectally. Analysis of different gene-based prime-boost immunization regimens revealed that recombinant adenovirus type 5 (rAd5) prime followed by replication-defective lymphocytic choriomeningitis virus (rLCMV) boost elicited robust CD4 and CD8 T-cell and humoral immune responses. This vaccine protected against infection after repetitive mucosal challenge with efficacies of 82% per exposure and 62% cumulatively. No effect was seen on viremia in infected vaccinated monkeys compared to controls. Protection correlated with the presence of neutralizing antibodies to the challenge viruses tested in peripheral blood mononuclear cells. These data indicate that a vaccine expressing a mismatched Env gene alone can prevent SIV infection in NHPs and identifies an immune correlate that may guide immunogen selection and immune monitoring for clinical efficacy trials.     

Systems analysis of MVA-C induced immune response reveals its significance as a vaccine candidate against HIV/AIDS of clade C

Based on the partial efficacy of the HIV/AIDS Thai trial (RV144) with a canarypox vector prime and protein boost, attenuated poxvirus recombinants expressing HIV-1 antigens are increasingly sought as vaccine candidates against HIV/AIDS. Here we describe using systems analysis the biological and immunological characteristics of the attenuated vaccinia virus Ankara strain expressing the HIV-1 antigens Env/Gag-Pol-Nef of HIV-1 of clade C (referred as MVA-C). MVA-C infection of human monocyte derived dendritic cells (moDCs) induced the expression of HIV-1 antigens at high levels from 2 to 8 hpi and triggered moDCs maturation as revealed by enhanced expression of HLA-DR, CD86, CD40, HLA-A2, and CD80 molecules. Infection ex vivo of purified mDC and pDC with MVA-C induced the expression of immunoregulatory pathways associated with antiviral responses, antigen presentation, T cell and B cell responses. Similarly, human whole blood or primary macrophages infected with MVA-C express high levels of proinflammatory cytokines and chemokines involved with T cell activation. The vector MVA-C has the ability to cross-present antigens to HIV-specific CD8 T cells in vitro and to increase CD8 T cell proliferation in a dose-dependent manner. The immunogenic profiling in mice after DNA-C prime/MVA-C boost combination revealed activation of HIV-1-specific CD4 and CD8 T cell memory responses that are polyfunctional and with effector memory phenotype. Env-specific IgG binding antibodies were also produced in animals receiving DNA-C prime/MVA-C boost. Our systems analysis of profiling immune response to MVA-C infection highlights the potential benefit of MVA-C as vaccine candidate against HIV/AIDS for clade C, the prevalent subtype virus in the most affected areas of the world.     

Are anti-HIV IgAs good guys or bad guys?

An estimated 90% of all HIV transmissions occur mucosally. Immunoglobulin A (IgA) molecules are important components of mucosal fluids. In a vaccine efficacy study, in which virosomes displaying HIV gp41 antigens protected most rhesus monkeys (RMs) against simian-human immunodeficiency virus (SHIV), protection correlated with vaginal IgA capable of blocking HIV transcytosis in vitro. Furthermore, vaginal IgG exhibiting virus neutralization and/or antibody-dependent cellular cytotoxicity (ADCC) correlated with prevention of systemic infection. In contrast, plasma IgG had neither neutralizing nor ADCC activity. More recently, a passive mucosal immunization study provided the first direct proof that dimeric IgAs (dIgAs) can prevent SHIV acquisition in RMs challenged mucosally. This study compared dimeric IgA1 (dIgA1), dIgA2, or IgG1 versions of a human neutralizing monoclonal antibody (nmAb) targeting a conserved HIV Env epitope. While the nmAb neutralization profiles were identical in vitro, dIgA1 was significantly more protective in vivo than dIgA2. Protection was linked to a new mechanism: virion capture. Protection also correlated with inhibition of transcytosis of cell-free virus in vitro. While both of these primate model studies demonstrated protective effects of mucosal IgAs, the RV144 clinical trial identified plasma IgA responses to HIV Env as risk factors for increased HIV acquisition. In a secondary analysis of RV144, plasma IgA decreased the in vitro ADCC activity of vaccine-induced, Env-specific IgG with the same epitope specificity. Here we review the current literature regarding the potential of IgA – systemic as well as mucosal – in modulating virus acquisition and address the question whether anti-HIV IgA responses could help or harm the host.     

A unified evaluation of differential vaccine efficacy

Many infectious diseases are well prevented by proper vaccination. However, when a vaccine is not completely efficacious, there is great interest in determining how the vaccine effect differs in subgroups conditional on measured immune responses postvaccination and also according to the type of infecting agent (eg, strain of a virus). The former is often called correlate of protection (CoP) analysis, while the latter has been called sieve analysis. We propose a unified framework for simultaneously assessing CoP and sieve effects of a vaccine in a large Phase III randomized trial. We use flexible parametric models treating times to infection from different agents as competing risks and estimated maximum likelihood to fit the models. The parametric models under competing risks allow for estimation of both cumulative incidence-based contrasts and instantaneous rates. We outline the assumptions with which we can link the observable data to the causal contrasts of interest, propose hypothesis testing procedures, and evaluate our proposed methods in an extensive simulation study.     

CD8+ T-cell responses to different HIV proteins have discordant associations with viral load

Selection of T-cell vaccine antigens for chronic persistent viral infections has been largely empirical. To define the relationship, at the population level, between the specificity of the cellular immune response and viral control for a relevant human pathogen, we performed a comprehensive analysis of the 160 dominant CD8(+) T-cell responses in 578 untreated HIV-infected individuals from KwaZulu-Natal, South Africa. Of the HIV proteins targeted, only Gag-specific responses were associated with lowering viremia. Env-specific and Accessory/Regulatory protein-specific responses were associated with higher viremia. Increasing breadth of Gag-specific responses was associated with decreasing viremia and increasing Env breadth with increasing viremia. Association of the specific CD8(+) T-cell response with low viremia was independent of HLA type and unrelated to epitope sequence conservation. These population-based data, suggesting the existence of both effective immune responses and responses lacking demonstrable biological impact in chronic HIV infection, are of relevance to HIV vaccine design and evaluation.     

Molecular Evolution of the HIV-1 Thai Epidemic between the Time of RV144 Immunogen Selection to the Execution of the Vaccine Efficacy Trial

The RV144 HIV-1 vaccine trial (Thailand, 2003 to 2009), using immunogens genetically matched to the regional epidemic, demonstrated the first evidence of efficacy for an HIV-1 vaccine. Here we studied the molecular evolution of the HIV-1 epidemic from the time of immunogen selection to the execution of the efficacy trial. We studied HIV-1 genetic diversity among 390 volunteers who were deferred from enrollment in RV144 due to preexisting HIV-1 infection using a multiregion hybridization assay, full-genome sequencing, and phylogenetic analyses. The subtype distribution was 91.7% CRF01_AE, 3.5% subtype B, 4.3% B/CRF01_AE recombinants, and 0.5% dual infections. CRF01_AE strains were 31% more diverse than the ones from the 1990s Thai epidemic. Sixty-nine percent of subtype B strains clustered with the cosmopolitan Western B strains. Ninety-three percent of B/CRF01_AE recombinants were unique; recombination breakpoint analysis showed that these strains were highly embedded within the larger network that integrates recombinants from East/Southeast Asia. Compared to Thai sequences from the early 1990s, the distance to the RV144 immunogens increased 52% to 68% for CRF01_AE Env immunogens and 12% to 29% for subtype B immunogens. Forty-three percent to 48% of CRF01_AE sequences differed from the sequence of the vaccine insert in Env variable region 2 positions 169 and 181, which were implicated in vaccine sieve effects in RV144. In conclusion, compared to the molecular picture at the early stages of vaccine development, our results show an overall increase in the genetic complexity of viruses in the Thai epidemic and in the distance to vaccine immunogens, which should be considered at the time of the analysis of the trial results.