Longitudinal assessment of biomarkers for clinical trials of novel therapeutic agents: the Run-In study.

Alton EW, Boyd AC, Cunningham S, Davies JC, Hyde SC, Innes JA, Gill DR, Greening A, Griesenbach U, Higgins T, Porteous DJ

Thorax, Vol 65 S4

British Thoracic Society Winter Meeting, London, 2010

We will be undertaking a phase IIB clinical trial of repeated application of liposome-based gene therapy over a one year period in approximately 100 CF patients (Multidose Trial). In preparation for this, we sought to address two key questions. Firstly, could we define the optimal set of patients in which the therapy could both be delivered (good access to the airways via nebulisation), and in whom any therapeutic effect was measurable (one or more abnormal measures of lung disease). Secondly, in this set of 'can deliverdcan measure' patients, which biomarker(s) could be powered to be the primary outcome measure for the trial. To address both questions, we undertook a study (Run-in), cross-sectionally assessing 'can deliver' and longitudinally assessing a large set of candidate biomarkers for 'can measure'.192 patients from age 10 upwards, with FEV1 >40% were enrolled at two clinical centres; 154 of these remained in the study after four visits spaced at approximately 4e5 month intervals. Biomarkers assessed cross-sectionally included radionucleotide deposition scans, CT and mucocilary clearance. Longitudinal biomarkers included a large series of serum, sputum and exhaled breath inflammatory markers, lung physiology, exercise-related assays and quality of life assessment. 12 patients were judged too severe for adequate delivery and were excluded. A shortlist of 4 biomarkers was generated based on a) showing a CF/non-CF difference, b) response to course of intravenous antibiotics, and c) coefficients of variation. These four were matched against the remaining 142 patients, and a further seven patients excluded in whom none of these short listed biomarkers was abnormal. 89 patients (3 or 4 biomarkers abnormal) have been definitely included to progress into the Multidose Trial, and a further 46 (1 or 2 biomarkers abnormal) are awaiting the final primary outcome selection. The Run-in study has, therefore, been able to a) select a cohort of 'optimal' patients in which to assess gene therapy and b) provide an indication of which may be the more useful biomarkers to use in phase IIB clinical trials of novel therapeutic agents.