Comparison of 1-, 2-, and 3-Dimensional Tumor Response Assessment After Neoadjuvant GTX-RT in Borderline-Resectable Pancreatic Cancer.

BACKGROUND Facilitation of margin-negative resection is the goal of neoadjuvant therapy regimens used in the treatment of borderline-resectable pancreatic cancer patients. Multiple treatment approaches have shown efficacy in this setting, including neoadjuvant GTX (gemcitabine [Gemzar], docetaxel [Taxotere], and capecitabine [Xeloda]) and radiotherapy (RT). Three-dimensional tumor response may be a more accurate method of assessment compared to traditional 1- and 2-dimensional techniques. We compared these 3 methods in a series of patients who underwent neoadjuvant GTX-RT and surgical resection. MATERIALS AND METHODS This retrospective review included borderline-resectable pancreatic cancer patients treated with neoadjuvant GTX followed by 5-FU chemoradiotherapy with the intent of downstaging to resectability. Tumor was contoured on computed tomography (CT) scans obtained at the following time points: (A) initial staging, (B) CT simulation, and (C) restaging. These contours were used to determine tumor response according to WHO, RECIST, and volumetric criteria. RESULTS Fourteen patients all experienced a measurable decrease in tumor volume following neoadjuvant therapy and were deemed suitable for at least surgical exploration. Radiotherapy was delivered to a median 50 Gy (range, 45-52 Gy) in 1.8-2.0 Gy fractions via 3-D conformal (21%) or IMRT (79%). The median percent volume changes before and after CT simulation were -3.4% and -52.6%, respectively. The overall median percent change was -54.5%. The corresponding absolute volume changes were -0.42 cm(3) (range, 9.12 to -12.47), -5.31 cm(3) (range, 2.06 to -15.93), and -6.72 cm(3) (range, 0.53 to -15.47), respectively. Response according to WHO, RECIST, and volumetric methods was identical with the exception of 1 patient. CONCLUSION This is the first study to quantify volumetric tumor change objectively as a result of neoadjuvant chemoradiotherapy for the treatment of borderline resectable pancreatic cancer. Our data suggest that tumor response to neoadjuvant therapy is essentially equivalent between 1-, 2-, and 3-dimensional assessment methods.