Combination therapy group compared with the monotherapy group objective response rate is higher, the two groups had significant difference. 1-year survival rate increase effective combination therapy group, 95% confidence interval value indicates the highest one-year survival rate of 22%, a minimum of 4%. Another test for heterogeneity ORR values prompt greater heterogeneity, description of each group of clinical trials ORR value in terms of the difference is large, the conclusion is more conservative. This study included trials of 1-year survival analysis of Funnel Plot graphics unpaired symmetrical inverted funnel graph, which shows the existence of bias, may exaggerate the therapeutic effect.

Poor treatment of advanced pancreatic cancer, six months and l, 5-year survival rate is very low. So, find one pair of advanced pancreatic effective drugs and drug combinations are very urgent and important. Gemcitabine is a cell cycle-specific antimetabolite drugs, a major role in S-phase cells, under certain conditions, can prevent G. Progress of the S phase, has been confirmed with Fluorouracil (5 an FU) compared to, can improve survival and clinical benefit response, Gemcitabine monotherapy has been NCCN_2 recommended as one of the main drug for treatment of advanced pancreatic cancer. Many clinical trials are exploring Gemcitabine in combination with other drugs, such as chemotherapy, combined with targeted drug efficacy for advanced pancreatic cancer. Randomized controlled trials compared with other clinical studies have demonstrated a high intensity, can most truly reflect the efficacy of the study drug, and a random sample of large controlled clinical trials, such as Phase Ⅲ clinical trials, its credibility degrees higher. Gemcitabine combined treatment group to improve one-year survival rate has a significant effect, Gemcitabine in combination with other anticancer drugs to improve survival in a certain role. RR values for 1-year survival rate of FunnelPlot funnel plot analysis found significant asymmetry, suggesting the existence of bias in the analysis, which may be exaggerated conclusions.