Safety and feasibility of serial tumour biopsies in phase I clinical trials: the Royal Marsden Hospital Drug Development Unit experience


Session type:

Debashis Sarker, Daniel Tan, Andre Brunetto, Sophia Frentzas, Christina Messiou, Nina Tunariu, Nimrat Mangat, Sarah Stapleton, Nandita Desouza, Jorge Barriuso, Johann de Bono

Royal Marsden Hospital, Sutton, UK



A critical aspect in the successful evaluation of novel molecularly targeted anticancer therapies is the ability to determine modulation of target in tumour tissue. This is usually performed in the context of phase I clinical trials, but there are concerns as to the safety, feasibility and ethics of performing serial tumour biopsies in patients with advanced cancer.


We performed a retrospective review of all biopsies performed at our institution specifically in the context of phase I trials in the Drug Development Unit (DDU) between February 2002 and May 2008. Post biopsy complication rates with regard to pain, bleeding, infection were evaluated and graded according to CTCAEv3.


604 patients were treated on the DDU between Feb 2002 and May 2008.A total of 178 tumour biopsies were performed in 82 patients from 19 different phase I trials. A median of 2 biopsies were performed per patient (range 1-6).A total of 59 patients had successful serial biopsy (72%).

The most common site for tumour biopsy was subcutaneous tissue (28%).Other sites of biopsy were liver (20.7%), lymph node (12.2%), skin (18.3%), abdominal mass (9.8%), oesophago-gastric (6.1%) and prostate (4.9%%). The commonest tumour types were melanoma (25.6%), breast (19.5%), gastrointestinal (19.5%) and sarcoma (13.4%).

The majority of biopsies were obtained under direct radiological guidance (48.7%) or skin punch biopsies (36.6%). Other methods used were surgical excision (4.9%), transrectal biopsy (4.9%) and upper GI endoscopy (4.9%).

Incidence of any grade of pain was 15.9% (13/82; Grade 1 9pts, Grade 2 4 pts).