Comparison of PERCIST to RECIST response assessment for patients recieving chemo-radiation plus nelfinavir for unresectable pancreatic cancer.


Session type:

Cynthia Eccles1, Katherine Vallis1, Thomas Brunner1
1Gray Institute for Radiation Oncology and Biology, Oxford, United Kingdom


Carcinoma of the exocrine pancreas is a serious therapeutic problem worldwide, with 5-year survival rates of 1-4%. A Phase I trial (ARC-I)[i] of the human immonodeficiency virus protease inhibitor, nelfinavir, plus chemoradiotherapy for locally advanced pancreatic cancer showed acceptable toxicity and a promising response rate based on RECIST response criteria.  A recently proposed method for tumour response assessment, PERCIST[ii], uses PET imaging to evaluate statistically significant changes in tumour standard uptake values (SUV).


Ten patients on the ARC-I study underwent CT and FDG-PET imaging at baseline and following the completion of treatment. In a retrospective study, the pre- and post-treatment volumes based on PET imaging, were determined. Additionally, SUV mean and maximum values were used to calculate the changes in SUV up-take over time. The significance of all changes based on PET imaging was compared on a per-patient basis to the changes in anatomic volumes based on CT imaging.


Pre- and post-treatment tumour volumes were determined using anatomic and functional images in seven of ten patients to date. Complete/partial responses were observed in 88% and  57% of patients, based on CT and PET images, respectively. The direction of change (progressive disease vs complete/partial response) was consistent in all patients with both response assessment modalities.  In two patients with discrepant responses, CT demonstrated enhanced response over PET.


PERCIST response assessment was consistent with RECIST in the majority of patients, and shows promise as an emerging method for measuring treatment response.