Optimising Harvest of Bronchial Brush Biopsy Samples To Maximise Cell and RNA Yield in Gene Therapy Studies.

Vrettou C, Baker E, Tennant P, Baker A, Gordon C, Sumner-Jones SG, Davies JC, Gill DR, McLachlan G, Collie DD

Molecular Therapy, Vol 17, S1, Abstract 322

The American Society of Gene Therapy Annual Conference, San Diego, 2009

Clinically effective gene therapy for cystic fibrosis patients, require safe delivery of the CFTR gene to the airway epithelium and sustained expression of the CFTR protein at the apical surface. From a clinical perspective, one of the challenges is the development and validation of endpoint assays for reliable evaluation of effi cacy. Whilst quantifi cation of vector-derived mRNA and/or protein in patient samples are widely recognised endpoints, effective limits are placed on the size and number of samples that can be harvested from patients using such an invasive procedure. We have noted that the yield of mRNA obtained by airway biopsy sampling in sheep lung is normally insuffi cient to detect endogenous ovCFTR mRNA. As such there is a real clinical need to optimise all aspects of sample collection and processing in order to maximise the likelihood of detection. In this study we investigated whether cell yields differ in brushings derived from sites visualised by bronchoscopy around 5mm in diameter or sites reached by advancing the cytology brush into smaller airways more distal to a wedged bronchoscope. The relationship between the number of brush “passes” and cell yield for these sites was also evaluated. Additionally, protocols for retrieving cells from the cytology brushes to maximise the cell numbers were tested. Component aims included an assessment of the relationship between the gross appearance of brushing samples and their respective cell content Our data indicate that cell yields were increased in the more proximal relative to distal sites although the difference was not statistically signifi cant (2.7+2.38x106 vs 2.1+1.48x106; P=0.743). The number of brush passes showed a strong positive correlation with cell yield (P=0.001). In addition signifi cantly more cells were obtained from proximal sites relative to distal sites after 32 passes (p=0.034). We also noted that a higher cell yield is obtained when the whole bronchoscope is withdrawn and the cytology brush not retracted in its sheath for removal from the lungs [mean cell yield 3.6 x 106 (nonretracted) vs 2.0 x 106 cells (retracted)]. This increased cell yield is accompanied by a corresponding increase in the average mRNA yield in samples collected from the brush when it is not retracted into the sheath 30.3 +/- 5.1ng/μl from a non-retracted brush versus 18.4+/- 3.8ng/μl from a retracted brush. The gross appearance grading scores also strongly correlated with cell yields, number of passes and the extent of gross blood contamination in the cell pellet (spearman’s rho values respectively) but did not differ between sites. These studies have informed procedures for harvesting and processing of human lung samples in our currentclinical trial in CF patients.