Prediction of the response to repetitive transcranial magnetic stimulation of the motor cortex in peripheral neuropathic pain and validation of a new algorithm.

CLINICAL TRIAL REGISTRATION: NCT02010281.

Report of similar placebo response in one internet versus onsite randomised controlled trials from the literature.

Objective: The aim of this study was to compare the magnitude and the predictors of the placebo response in an internet versus onsite randomised controlled trials (RCTs) in people with hand osteoarthritis (HOA). Method: This study is a post-hoc analysis based on one internet RCT (RADIANT) and previously published onsite RCTs for HOA identified through a rigorous searching and selection strategy. The magnitude of the placebo response in the two different types of RCTs were compared using heterogeneity statistics and forest plots visualisation. Classic placebo predictors as well as a combined model, defined with data from onsite RCTs, were tested to predict the placebo response. Results: We analysed the dataset from RADIANT and fourteen previously published onsite RCTs. None of the analyses showed a significant difference between the placebo response for the internet versus onsite RCTs. The "classic" placebo predictors combined in a multivariate predictive model correlated significantly with the placebo response measured in RADIANT study. Conclusion: Despite the absence of face-to-face interactions with the study personnel, there is no evidence that either the magnitude or the predictors of the placebo response of this internet RCT differ from those of onsite RCTs. This analysis is considered as a first step towards evaluating the difference between these designs and strengthens the argument that internet RCTs remain an acceptable alternative way to assess the efficacy of an active treatment in comparison to a placebo.

Leveraging historical data to optimize the number of covariates and their explained variance in the analysis of randomized clinical trials.

The amount of data collected from patients involved in clinical trials is continuously growing. All baseline patient characteristics are potential covariates that could be used to improve clinical trial analysis and power. However, the limited number of patients in phases I and II studies restricts the possible number of covariates included in the analyses. In this paper, we investigate the cost/benefit ratio of including covariates in the analysis of clinical trials with a continuous outcome. Within this context, we address the long-running question "What is the optimum number of covariates to include in a clinical trial?" To further improve the benefit/cost ratio of covariates, historical data can be leveraged to pre-specify the covariate weights, which can be viewed as the definition of a new composite covariate. Here we analyze the use of a composite covariate to improve the estimated treatment effect in small clinical trials. A composite covariate limits the loss of degrees of freedom and the risk of overfitting.

Stratification of patients based on the Neuropathic Pain Symptom Inventory: development and validation of a new algorithm.

ABSTRACT: The personalization of neuropathic pain treatment could be improved by identifying specific sensory phenotypes (ie, specific combinations of symptoms and signs) predictive of the response to different classes of drugs. A simple and reliable phenotyping method is required for such a strategy. We investigated the utility of an algorithm for stratifying patients into clusters corresponding to specific combinations of neuropathic symptoms assessed with the Neuropathic Pain Symptom Inventory (NPSI). Consistent with previous results, we first confirmed, in a cohort of 628 patients, the existence of a structure consisting of 3 clusters of patients characterized by higher NPSI scores for: pinpointed pain (cluster 1), evoked pain (cluster 2), or deep pain (cluster 3). From these analyses, we derived a specific algorithm for assigning each patient to one of these 3 clusters. We then assessed the clinical relevance of this algorithm for predicting treatment response, through post hoc analyses of 2 previous controlled trials of the effects of subcutaneous injections of botulinum toxin A. Each of the 97 patients with neuropathic pain included in these studies was individually allocated to one cluster, by applying the algorithm to their baseline NPSI responses. We found significant effects of botulinum toxin A relative to placebo in clusters 2 and 3, but not in cluster 1, suggesting that this approach was, indeed, relevant. Finally, we developed and performed a preliminary validation of a web-based version of the NPSI and algorithm for the stratification of patients in both research and daily practice.

A balanced hazard ratio for risk group evaluation from survival data.

Common clinical studies assess the quality of prognostic factors, such as gene expression signatures, clinical variables or environmental factors, and cluster patients into various risk groups. Typical examples include cancer clinical trials where patients are clustered into high or low risk groups. Whenever applied to survival data analysis, such groups are intended to represent patients with similar survival odds and to select the most appropriate therapy accordingly. The relevance of such risk groups, and of the related prognostic factors, is typically assessed through the computation of a hazard ratio. We first stress three limitations of assessing risk groups through the hazard ratio: (1) it may promote the definition of arbitrarily unbalanced risk groups; (2) an apparently optimal group hazard ratio can be largely inconsistent with the p-value commonly associated to it; and (3) some marginal changes between risk group proportions may lead to highly different hazard ratio values. Those issues could lead to inappropriate comparisons between various prognostic factors. Next, we propose the balanced hazard ratio to solve those issues. This new performance metric keeps an intuitive interpretation and is as simple to compute. We also show how the balanced hazard ratio leads to a natural cut-off choice to define risk groups from continuous risk scores. The proposed methodology is validated through controlled experiments for which a prescribed cut-off value is defined by design. Further results are also reported on several cancer prognosis studies, and the proposed methodology could be applied more generally to assess the quality of any prognostic markers.

A generic cycling hypoxia-derived prognostic gene signature: application to breast cancer profiling.

BACKGROUND: Temporal and local fluctuations in O2 in tumors require adaptive mechanisms to support cancer cell survival and proliferation. The transcriptome associated with cycling hypoxia (CycHyp) could thus represent a prognostic biomarker of cancer progression. METHOD: We exposed 20 tumor cell lines to repeated periods of hypoxia/reoxygenation to determine a transcriptomic CycHyp signature and used clinical data sets from 2,150 breast cancer patients to estimate a prognostic Cox proportional hazard model to assess its prognostic performance. RESULTS: The CycHyp prognostic potential was validated in patients independently of the receptor status of the tumors. The discriminating capacity of the CycHyp signature was further increased in the ER+ HER2- patient populations including those with a node negative status under treatment (HR=3.16) or not (HR=5.54). The CycHyp prognostic signature outperformed a signature derived from continuous hypoxia and major prognostic metagenes (P<0.001). The CycHyp signature could also identify ER+HER2 node-negative breast cancer patients at high risk based on clinicopathologic criteria but who could have been spared from chemotherapy and inversely those patients classified at low risk based but who presented a negative outcome. CONCLUSIONS: The CycHyp signature is prognostic of breast cancer and offers a unique decision making tool to complement anatomopathologic evaluation.

Hypoxia integration in the serological proteome analysis unmasks tumor antigens and fosters the identification of anti-phospho-eEF2 antibodies as potential cancer biomarkers.

The expression by tumor cells of proteins with aberrant structure, expression or distribution accounts for the development of a humoral immune response. Autoantibodies (aAb) directed against tumor-associated antigens (TAA) may thus be particularly relevant for early detection of cancer. Serological proteome analysis (SERPA) aims to identify such circulating aAb through the immunoblotting of 2D-separated tumor cell proteins with cancer patient serum and the consecutive MS identification of proteins in reactive spots. This method has the advantage to use post-translationally modified proteins as a source of potential TAA. Here, we applied this strategy by using colorectal tumor cells pre-exposed to hypoxia in order to promote the expression of a pattern of TAA more likely to represent in vivo conditions. We used two human HCT116 and HT29 colorectal cancer cell lines exposed for 48 hours to 1% O2. Spots positive after immunoblotting of 2D-separated lysates of hypoxic cells with the sera of tumor-bearing mice, were collected and analysed by MS for protein identification. Among the hypoxia-specific immunogenic proteins, we identified a phosphorylated form of eukaryotic translation elongation factor 2 (phospho-Thr56 eEF2). We confirmed the increased phosphorylation of this protein in hypoxic colorectal tumor cells as well as in mouse tumors. Using a specific immunoassay, we could detect the presence of corresponding anti-phospho-Thr56 eEF2 aAb in the serum of tumor-bearing mice (vs healthy mice). We further documented that the detection of these aAb preceded the detection of a palpable tumor mass in mice and validated the presence of anti-phospho-Thr56 eEF2 aAb in the serum of patients with adenomatous polyps and colorectal carcinoma. In conclusion, this study validates a phosphorylated form of eEF2 as a new TAA and more generally, provides evidence that integrating hypoxia upstream of SERPA offers a more relevant repertoire of TAA able to unmask the presence of circulating aAb.