Morethanadecadeago, investigatorsfromVirginiaMasonMedical Center reported a promising 3-year overall survival rate of 64%(95% CI, 56% to 72%) in a group of 43 patients who received a novel adjuvant regimen comprising interferon alfa, infusional fluorouracil, cisplatin, and external-beamradiationtherapy (45-54Gyover5to6weeks), followedby twomorecyclesoffluorouracil, afterresectionofpancreaticcancer. The GI toxicity in particular was significant, leading to a delay of chemoradiotherapy in 70% of the patients and hospitalization in 42%. 1 Two different institutions and one cooperative group have since reported similar findings with this regimen in single-arm prospective studies: potentially clinically relevant improvement in survival at the cost of significant but reversible acute toxicity. 2-4 The results of the only phase III trial evaluating this regimen, the CapRI (Combined Chemoradioimmunotherapy for Pancreatic Adenocarcinoma) trial, are published in the article that accompanies this editorial. Recognizing the toxicity profile, the study was powered to identifyaneardoublingofthe2-yearoverallsurvivalrate, from35% to 60%, aclinicallymeaningfuldifferencethatmightjustifythehardships of therapy. Despite apparent differences in the mean cancer antigen 19-9 (CA 19-9) level at diagnosis and R1 resections that might have favored the control arm, the patients in the experimental arm had a longer median survival duration (32.1 months v 28.5 months), a difference that did not reach statistical significance. 5 The statistical assumptions at baseline leave open the question of whether a clinically meaningful difference in survival would have been seen in a larger study. Nevertheless, the 3.6-month difference in median survival suggests that some patients probably benefited from the experimental regimen. That makes the search for a biomarker that is predictive of which patients will benefit from this regimen important. Although some investigators have reported that the expression of interferon alfa/beta receptors is prognostic in resected pancreatic cancer, 7 others have shown interferon receptor expression to be of limited predictive value in patients with pancreatic cancer who are treated with chemoradioimmunotherapy. 8 Correlative studies should be planned using the resected specimens from the CapRI trial to look for a predictive marker for responders to radioimmunotherapy. The translational research plan was not detailed in the article, but unless we are able to discover a way to select the likely responders, this once promising regimen should not be studied further. Thus far, surgical pathology specimens from randomized adjuvant trials have proven to be the best source of tissue for correlative studies in pancreatic cancer. For instance, the relevance of human equilibrative nucleoside transporter 1 (hENT1) as a predictor of gemcitabine activity was discovered from translational studies of specimens from the Radiation Therapy Oncology Group (RTOG) 9704 adjuvant trial. 9 Those efforts have led to the Low hENT1 and Adenocarcinoma of the Pancreas (LEAP) trial, the first trial to use an integral biomarker in pancreatic cancer. Patients with low hENT1-expressing metastatic pancreatic cancer, as determined by immunohistochemistry, were randomly assigned to receive gemcitabine versus C0-101, a lipid-conjugated form of gemcitabine that enters the cell independent of hENT1. The LEAP trial has completed accrual and survival results are expected this year.

Personalization of care in other solid tumors has become a reality. Despite what is now a critical mass of motivated and dedicated preclinical and clinical investigators in the field of pancreatic oncology, the …