Background - Older patients with metastatic pancreatic cancer may suffer increased toxicity from intensive chemotherapy. Treatment individualization by geriatric assessment (GA) might improve functional outcome. - Methods - We performed a multicenter, phase IV, open label trial in patients 70 years with metastatic pancreatic adenocarcinoma. Patients underwent GA and were assigned to one of three categories based on their scores: Go-Go, Slow-Go, or Frail. These categories were intended to guide physician's treatment decisions when choosing to treat patients with nab-paclitaxel/gemcitabine (arm A), gemcitabine (arm B), or best supportive care (arm C). Primary objective was a stable (loss of five points or less) Barthel's Activities of Daily Living (ADL) score during chemotherapy; secondary endpoints included GA scores during therapy, safety, quality of life, response and survival rates. - Results - Thirty-two patients were enrolled in the trial in six centers in Germany (out of 135 planned), resulting in termination due to low recruitment. Fifteen patients were allocated to nab-paclitaxel/gemcitabine, fifteen to gemcitabine, and two to best supportive care by their physicians, although according to their GA scores 29 patients (91%) were categorized as Slow-Go and three (9%) as Go-Go. Thus, fifteen of 32 (47%) patients were misclassified and given a course of treatment inconsistent with their GA scores. Median progression-free survival (PFS) were 3.3 months and 9.1 months and median time to quality-of-life deterioration 13 days and 29 days in the nab-paclitaxel/gemcitabine and gemcitabine monotherapy arms, respectively. Serious adverse events were reported in 11 (78.6%) patients in the nab-paclitaxel/gemcitabine and 8 (53.3%) patients in the gemcitabine arm. - Conclusions - Clinical evaluations by investigators differed markedly from geriatric assessments, leading to potential overtreatment. In our modest sample size study, those patients undergoing more intensive therapy had a less favorable course.
