Factors associated with success of image-guided tumour biopsies: Results from a prospective molecular triage study (MOSCATO-01)

Introduction MOSCATO-01 is a molecular triage trial based on on-purpose tumour biopsies to perform molecular portraits. We aimed at identifying factors associated with high tumour cellularity. Material and methods Tumour cellularity (percentage of tumour cells in samples defined at pathology) was evaluated according to patient characteristics, target lesion characteristics, operators' experience and biopsy approach. Results Among 460 patients enrolled between November, 2011 and March, 2014, 334 patients (73%) had an image-guided needle biopsy of the primary tumour (N = 38) or a metastatic lesion (N = 296). Biopsies were performed on liver (N = 127), lung (N = 72), lymph nodes (N = 71), bone (N = 11), or another tumour site (N = 53). Eighteen patients (5%) experienced a complication: Pneumothorax in 10 patients treated medically, and haemorrhage in 8, requiring embolisation in 3 cases. Median tumour cellularity was 50% (interquartile range, 30-70%). The molecular analysis was successful in 291/334 cases (87%). On-going chemotherapy, tumour origin (primary versus metastatic), lesion size, tumour growth rate, presence of necrosis on imaging, standardised uptake value, and needle size were not statistically associated with cellularity. Compared to liver or lung biopsies, cellularity was significantly lower in bone and higher in other sites (P < 0.0001). Cellularity significantly increased with the number of collected samples (P < 0.0001) and was higher in contrast-enhanced ultrasound-guided biopsies (P < 0.02). In paired samples, cellularity in central samples was lower than in peripheral samples in 85, equal in 68 and higher in 89 of the cases. Conclusion Image-guided biopsy is feasible and safe in cancer patients for molecular screening. Imaging modality, multiple sampling of the lesion, and the organ chosen for biopsy were associated with higher tumour cellularity.