US Findings in Patients at Risk for Pancreas Transplant Failure.

Purpose To determine if ultrasonographic (US) findings, including Doppler US findings, are associated with subsequent pancreas transplant failure. Materials and Methods A cohort of adult patients who underwent pancreas transplantation at a tertiary institution over the course of 10 years (from 2003 to 2012) was retrospectively evaluated for failure, which was defined as return to insulin therapy or surgical graft removal. The institutional review board provided a waiver of informed consent. All US images obtained within the 1st postoperative year were reviewed for three findings: arterial flow (presence or absence of intraparenchymal forward diastole flow), splenic vein thrombus, and edema. These findings were correlated with pancreas graft failure within 1-year after surgery by using Cox proportional hazards models and hazard ratios. Results A total of 228 transplants were included (mean patient age, 41.6 years; range, 19-57 years; 122 men, 106 women). Absent or reversed arterial diastolic flow was identified in nine of 20 failed transplants (sensitivity, 45%; 95% confidence interval [CI]: 23, 68) and in 15 of 208 transplants that survived (specificity, 93% [193 of 208]; 95% CI: 89, 96). The Cox proportional hazard ratio was 6.2 (95% CI: 3.1, 12.4). Splenic vein thrombus was identified in 10 of 20 failed transplants (sensitivity, 50%; 95% CI: 27, 73) and in 25 of 208 transplants that survived (specificity, 88% [183 of 208]; 95% CI: 83, 92). The Cox proportional hazard ratio was 4.2 (95% CI: 2.4, 7.4). Edema had the lowest specificity (Cox proportional hazard ratio, 2.0; 95% CI: 1.3, 2.9). In the multivariate analysis, only absent or reversed arterial diastolic flow remained significantly associated with transplant failure (adjusted hazard ratio, 3.6; 95% CI: 1.0, 12.8; P = .045). Conclusion Absent or reversed diastolic arterial Doppler flow has a stronger association with transplant failure than does splenic vein thrombus or edema. (©) RSNA, 2016.