Pyogenic liver abscess (PLA) is a severe disease with considerable mortality and is often polymicrobial. Understanding the pathogens that cause PLA is the basis for PLA treatment. Here, we profiled the bacterial composition in PLA fluid by pyrosequencing the 16S ribosomal RNA (rRNA) gene based on next-generation sequencing (NGS) technology to identify etiological agents of PLA and to provide information of their 16S rRNA sequences for application to DNA-based techniques in the hospital. Twenty patients with PLA who underwent percutaneous catheter drainage, abscess culture, and blood culture for isolates were included. Genomic DNAs from abscess fluids were subjected to polymerase chain reaction and pyrosequencing of the 16S rRNA gene with a 454 GS Junior System. The abscess and blood cultures were positive in nine (45%) and four (20%) patients, respectively. Pyrosequencing of 16S rRNA gene showed that 90% of the PLA fluid samples contained single or multiple genera of known bacteria such as Klebsiella, Fusobacterium, Streptococcus, Bacteroides, Prevotella, Peptostreptococcus, unassigned Enterobacteriaceae, and Dialister. Klebsiella was predominantly found in the PLA fluid samples. All samples that carried unassigned bacteria had 26.8% reads on average. We demonstrated that the occurrence of PLA was associated with eight known bacterial genera as well as unassigned bacteria and that 16S rRNA gene sequencing was more useful than conventional culture methods for accurate identification of bacterial pathogens from PLA.

One of the major limitations of curative resection in patients with pancreatic cancer is local tumor extension to the mesenteric vessels. Thus, the purposes of our study were to assess the clinical value of contrast-enhanced spiral computed tomography (CT) in predicting the resectability and survival of patients with pancreatic cancer with suspicious vascular invasion and to assess the influence of curative resection on the survival of these patients. We enrolled 40 patients with pancreatic cancer who were suspected of having an involvement of the adjacent large vessels and who subsequently underwent operation with curative intent in the study. Resectability and survival were correlated with CT findings such as segment length, degree of encasement, and type and number of vessels involved. The survival rate was compared between the curative and palliative resection groups, and survival rate was compared between the resected and unresected groups. Of the 40 patients with adenocarcinoma of the pancreas, 14 had curative resections and 26 had palliative resections. The probability of curative resection was higher in patients with segment lengths less than 2 cm, as compared with segment lengths more than 2 cm. However, there was no difference in survival between the two groups. There were no differences in resectability and survival according to the degree of encasement and type and number of vessels involved. There was no difference in survival between the curative and palliative resection groups. There was no difference in survival between the resected and unresected groups. A survival benefit was not achieved by curative resection in patients with pancreatic cancer with vascular invasion. Therefore, it would be better to avoid aggressive surgery in patients with pancreatic cancer with vascular invasion.