Search

Objective: To assess whether reticular hypointensity on hepatobiliary phase images of gadoxetic acid-enhanced magnetic resonance imaging (EOB-MRI) is a diagnostic finding of sinusoidal obstruction syndrome (SOS) in patients with hepatic metastases who have undergone chemotherapy., Methods: We retrospectively analysed EOB-MRI of 42 patients who had undergone chemotherapy before hepatic resection of colorectal hepatic metastases. Two radiologists, who were unaware of whether or not the patients had SOS, reviewed the hepatobiliary phase images to determine the presence of hypointense reticulation in the liver using a 5-point scale. The sensitivity, specificity and area under the receiver operating characteristics curve (A(z)) were calculated for each reviewer., Results: The sensitivity, specificity and A(z) for the diagnosis of SOS were 75%, 100% and 0.957 for reader 1 and 75%, 96.2% and 0.936 for reader 2, respectively. In one patient who received a false-positive diagnosis by one reader, there was sinusoidal fibrosis on histological examination, but not diagnostic for SOS. False-negative diagnosis occurred in four patients for both readers; histology of these patients showed minimal and localised sinusoidal congestion and fibrosis., Conclusions: Reticular hypointensity on hepatobiliary phase images of EOB-MRI is highly specific for the diagnosis of SOS in patients with treated colorectal hepatic metastases., Key Points: Gadoxetic acid enhanced magnetic resonance imaging (EOB-MRI) can identify the sinusoidal obstruction syndrome (SOS). The diagnosis can be achieved with high specificity and good interobserver agreement. SOS typically demonstrates diffuse hypointensity on hepatobiliary phase images on EOB-MRI. EOB-MRI may be falsely negative in patients with minimal degree of SOS.

Background: Margin-negative pancreatectomy provides only chance to cure pancreatic cancer. However, borderline resectable pancreatic cancer (BRPCa) has the risk of incomplete palliative resection., Materials and Methods: We retrospectively reviewed 32 patients with BRPCa who underwent a pancreatectomy following preoperative chemoradiation therapy (CCRT (+)/Px group) and compared these patients with those with resectable pancreatic cancer (RPCa) who underwent pancreatectomy without preoperative CCRT (CCRT (-)/Px group, n=104)., Results: Eighteen patients (56.2%) showed more than 50% significant pathological response to CCRT. The degree of pathological responses showed a positive relationship between final pT stage (p=0.075). More frequent vascular resection (p<0.001), transfusion (p=0.076), and longer operation time were observed in the CCRT(+)/Px group. However, similar R0 resection rates (p=0.272), lower pT stage (p<0.001), smaller number of metastastic lymph nodes (p=0.002), and lower incidence of lymph node metastasis (p=0.032) were noted in the CCRT(+)/Px group. The overall disease-specific survival were similar (median survival, 30.5 months (95% CI; 23.6-37.4) vs. 26.3 months (95% CI; 15.9-36.7), p=0.709), and no statistical differences in cancer recurrence risks were noted between the two groups (p=0.505)., Conclusion: Pancreatectomy following preoperative neoadjuvant CCRT can be a potential strategy for margin-negative resection in BRPCa patients.