OBJECTIVE: The role of CA 19-9 in monitoring treatment response in advanced pancreatic cancer remains uncertain. We assessed its value in predicting early failure of first-line chemotherapy. METHODS: Data of 84 patients with advanced pancreatic cancer who had received first-line chemotherapy were analyzed with regard to changes in CA 19-9 during the first 2 months of treatment. RESULTS: Median time to progression and overall survival in patients with a transient increase in CA 19-9 during month 1 (n = 15; 5.5 and 13 months) and in those with no increase during the 2 months (n = 52; 6.5 and 12 months) were comparable and slightly above the median values of the entire study population. The hazard ratios for disease progression for a 20% increase in CA 19-9 during the first and second month of therapy were 1.065 and 1.339 in the univariate- and 1.092 and 1.298 in the multiple Cox regression model, respectively. CA 19-9 did not influence survival. CONCLUSION: Our results suggest that early CA 19-9 measurements are weakly associated with disease progression rather than survival in patients with advanced pancreatic cancer receiving palliative chemotherapy. In view of a possible tumor marker flare, values after the first month of therapy must be interpreted with caution. Copyright 2009 S. Karger AG, Basel.
OBJECTIVE: The role of CA 19-9 in monitoring treatment response in advanced pancreatic cancer remains uncertain. We assessed its value in predicting early failure of first-line chemotherapy. METHODS: Data of 84 patients with advanced pancreatic cancer who had received first-line chemotherapy were analyzed with regard to changes in CA 19-9 during the first 2 months of treatment. RESULTS: Median time to progression and overall survival in patients with a transient increase in CA 19-9 during month 1 (n = 15; 5.5 and 13 months) and in those with no increase during the 2 months (n = 52; 6.5 and 12 months) were comparable and slightly above the median values of the entire study population. The hazard ratios for disease progression for a 20% increase in CA 19-9 during the first and second month of therapy were 1.065 and 1.339 in the univariate- and 1.092 and 1.298 in the multiple Cox regression model, respectively. CA 19-9 did not influence survival. CONCLUSION: Our results suggest that early CA 19-9 measurements are weakly associated with disease progression rather than survival in patients with advanced pancreatic cancer receiving palliative chemotherapy. In view of a possible tumor marker flare, values after the first month of therapy must be interpreted with caution. Copyright 2009 S. Karger AG, Basel.