5% in HCC patients. Rucaparib order Table 2 Sensitivity of tumour markers in patients with primary liver cancer Analysis of ROC curves Receiver operator characteristic curves were constructed to compare the ability of the four markers to differentiate between patients with malignant and benign liver disease. Curves for HCC (Figure 2) and ICC (Figure 3) in distinction to benign liver disease are illustrated. For patients with HCC, the AUC was 0.81��0.03, 0.61��0.04, 0.64��0.03, and 0.58��0.04 for AFP, CEA, CA 19-9, and CYFRA 21-1, respectively. The AUC for AFP differed significantly from that for other markers (P<0.0001). For patients with ICC, the AUC was 0.53��0.08, 0.81��0.05, 0.86��0.06, and 0.95��0.03 for AFP, CEA, CA 19-9, and CYFRA 21-1, respectively. A significant difference was noted between the AUC for CYFRA 21-1 and those for CEA (P=0.

0196) and AFP (P<0.0001); on the other hand, no significant difference was noted between the AUC for CYFRA 21-1 and that for CA 19-9 (P=0.0913). In the ICC group, ROC analysis demonstrated that CYFRA 21-1 was superior to the other markers. Figure 2 Receiver operating characteristic (ROC) curves in patients with hepatocellular carcinoma and benign liver disease were constructed. Figure 3 Receiver operating characteristic (ROC) curves in patients with intrahepatic cholangiocarcinoma and benign liver disease were constructed. CYFRA 21-1 and pathologic variables in patients with ICC For 23 patients with ICC, CYFRA 21-1 concentrations were compared according to disease stage (Table 3, Figure 4). Median and interquartile range of serum CYFRA 21-1 concentrations were 3.

2 (3.1�C3.8), 4.4 (3.1�C7.6), and 11.5 (7.8�C12.2) for stages I/II, III, and IV, respectively. An overall tendency towards an increase in serum concentration was observed from stages I/II to IV, and a significant difference was noted (Kruskal�CWallis test, P=0.0102). No significant difference was evident for serum CEA or CA 19-9 concentration between tumour stages (Kruskal�CWallis test; P=0.0774 and 0.1252, respectively). The serum CYFRA 21-1 concentration also differed significantly according to the T factor by the TNM classification (Table 3, Figure 4). Median and interquartile ranges of serum CYFRA 21-1 concentrations were 3.1 (2.1�C3.7), 4.4 (3.2�C6.4), and 11.5 (7.8�C12.2) for T1/2, T3, and T4, respectively (Kruskal�CWallis test, P=0.0038). Tumour size and vascular invasion were related to serum CYFRA 21-1 concentration, while serum CYFRA 21-1 concentration did not differ according to nodal status. Table 3 Serum CYFRA 21-1 concentrations of 23 patients with intrahepatic cholangiocarcinoma Figure 4 Distribution of individual serum CYFRA 21-1 values Anacetrapib according to stage (upper) and T factor (lower) in patients with intrahepatic cholangiocarcinoma.