Investigation of age–treatment interaction in the SPACE trial using different statistical approaches

ABSTRACT Selection of treatments to fit the specific needs for a certain patient is one major challenge in modern medicine. Personalized treatments rely on established patient–treatment interactions. In recent years, various statistical methods for the identification and estimation of interactions between relevant covariates and treatment were proposed. In this article, different available methods for detection and estimation of a covariate–treatment interaction for a time-to-event outcome, namely the standard Cox regression model assuming a linear interaction, the fractional polynomials approach for interaction, the modified outcome approach, the local partial-likelihood approach, and STEPP (Subpopulation Treatment Effect Pattern Plots) were applied to data from the SPACE trial, a randomized clinical trial comparing stent-protected angioplasty (CAS) to carotid endarterectomy (CEA) in patients with symptomatic stenosis, with the aim to analyse the interaction between age and treatment. Time from primary intervention to the first relevant event (any stroke or death) was considered as outcome parameter. The analyses suggest a qualitative interaction between patient age and treatment indicating a lower risk after treatment with CAS compared to CEA for younger patients, while for elderly patients a lower risk after CEA was observed. Differences in the statistical methods regarding the observed results, applicability, and interpretation are discussed.

[1]  Lu Tian,et al.  A Simple Method for Detecting Interactions between a Treatment and a Large Number of Covariates , 2012, 1212.2995.

[2]  T. Brott,et al.  Age and Outcomes After Carotid Stenting and Endarterectomy: The Carotid Revascularization Endarterectomy Versus Stenting Trial , 2011, Stroke.

[3]  Lu Tian,et al.  Effectively Selecting a Target Population for a Future Comparative Study , 2013, Journal of the American Statistical Association.

[4]  N. Smith,et al.  With contributions from , 2007 .

[5]  Thomas M. Loughin,et al.  A residual bootstrap for regression parameters in proportional hazards models , 1995 .

[6]  A. Vincent-Salomon,et al.  Designs and challenges for personalized medicine studies in oncology: focus on the SHIVA trial , 2012, Targeted Oncology.

[7]  L. Trinquart,et al.  Endarterectomy Versus Angioplasty in Patients with Symptomatic Severe Carotid Stenosis (EVA-3S) trial: results up to 4 years from a randomised, multicentre trial , 2008, The Lancet Neurology.

[8]  Marc Ratkovic,et al.  Estimating treatment effect heterogeneity in randomized program evaluation , 2013, 1305.5682.

[9]  W. Hacke,et al.  30 day results from the SPACE trial of stent-protected angioplasty versus carotid endarterectomy in symptomatic patients: a randomised non-inferiority trial , 2006, The Lancet.

[10]  David R. Cox,et al.  Regression models and life tables (with discussion , 1972 .

[11]  J. M. Taylor,et al.  Subgroup identification from randomized clinical trial data , 2011, Statistics in medicine.

[12]  Jianqing Fan,et al.  Local partial-likelihood estimation for lifetime data , 2006, math/0605511.

[13]  C. Bulpitt SUBGROUP ANALYSIS , 1988, The Lancet.

[14]  P. Royston,et al.  A new approach to modelling interactions between treatment and continuous covariates in clinical trials by using fractional polynomials , 2004, Statistics in medicine.

[15]  Bingshu E. Chen,et al.  Testing for treatment‐biomarker interaction based on local partial‐likelihood , 2015, Statistics in medicine.

[16]  J. Lewis,et al.  Statistical principles for clinical trials (ICH E9): an introductory note on an international guideline. , 1999, Statistics in medicine.

[17]  B. Freidlin,et al.  Statistical and practical considerations for clinical evaluation of predictive biomarkers. , 2013, Journal of the National Cancer Institute.

[18]  M. Hartmann,et al.  Clinical and angiographic risk factors for stroke and death within 30 days after carotid endarterectomy and stent-protected angioplasty: a subanalysis of the SPACE study , 2008, The Lancet Neurology.

[19]  Marco Bonetti,et al.  A small sample study of the STEPP approach to assessing treatment–covariate interactions in survival data , 2009, Statistics in medicine.

[20]  Patrick Royston MFPI: Stata module for modelling and displaying interactions , 2012 .

[21]  Siwei Zhang An R-package for the Estimation and Testing of Multiple Covariates and Biomarker Interactions for Survival Data Based on Local Partial Likelihood , 2016 .

[22]  M. Sydes,et al.  Practical methods for incorporating summary time-to-event data into meta-analysis , 2007, Trials.

[23]  P. Dorman,et al.  Carotid artery stenting compared with endarterectomy in patients with symptomatic carotid stenosis (International Carotid Stenting Study): an interim analysis of a randomised controlled trial , 2010, The Lancet.

[24]  Paul H. C. Eilers,et al.  Flexible smoothing with B-splines and penalties , 1996 .

[25]  P. Royston,et al.  Regression using fractional polynomials of continuous covariates: parsimonious parametric modelling. , 1994 .

[26]  Robert Gray,et al.  Flexible Methods for Analyzing Survival Data Using Splines, with Applications to Breast Cancer Prognosis , 1992 .

[27]  Ameet Bakhai,et al.  Clinical Trials - A Practical Guide to Design, Analysis, and Reporting , 2006 .

[28]  Brian T. Maurer Regression. , 2020, JAAPA : official journal of the American Academy of Physician Assistants.

[29]  L. Tian,et al.  Analysis of randomized comparative clinical trial data for personalized treatment selections. , 2011, Biostatistics.

[30]  James J. Chen,et al.  Predictive biomarkers for treatment selection: statistical considerations. , 2015, Biomarkers in medicine.

[31]  Xuming He,et al.  Inference for Subgroup Analysis With a Structured Logistic-Normal Mixture Model , 2015 .

[32]  S. Pocock,et al.  Subgroup analysis, covariate adjustment and baseline comparisons in clinical trial reporting: current practiceand problems , 2002, Statistics in medicine.

[33]  Willi Sauerbrei,et al.  Interaction of treatment with a continuous variable: simulation study of power for several methods of analysis , 2014, Statistics in medicine.

[34]  Jian Huang,et al.  A Concave Pairwise Fusion Approach to Subgroup Analysis , 2015, 1508.07045.

[35]  Patrick Royston,et al.  Detecting an interaction between treatment and a continuous covariate: A comparison of two approaches , 2007, Comput. Stat. Data Anal..

[36]  Association between age and risk of stroke or death from carotid endarterectomy and carotid stenting: a meta-analysis of pooled patient data from four randomised trials , 2016, The Lancet.

[37]  Patrick Royston,et al.  Interaction of treatment with a continuous variable: simulation study of significance level for several methods of analysis , 2013, Statistics in medicine.

[38]  T. Friede,et al.  Methods for identification and confirmation of targeted subgroups in clinical trials: A systematic review , 2016, Journal of biopharmaceutical statistics.

[39]  W. Hacke,et al.  Results of the Stent-Protected Angioplasty versus Carotid Endarterectomy (SPACE) study to treat symptomatic stenoses at 2 years: a multinational, prospective, randomised trial , 2008, The Lancet Neurology.

[40]  M. Bonetti,et al.  Patterns of treatment effects in subsets of patients in clinical trials. , 2004, Biostatistics.

[41]  M. Bonetti,et al.  A graphical method to assess treatment-covariate interactions using the Cox model on subsets of the data. , 2000, Statistics in medicine.

[42]  S. Assmann,et al.  Subgroup analysis and other (mis)uses of baseline data in clinical trials , 2000, The Lancet.