|
Back to 2015 Annual Meeting Program
Acute Renal Failure Following Liver Resection for Hepatocellular Carcinoma: Prognostic Value of Acute Kidney Injury Network Consensus Criteria.
Alexsander Bressan*, Elijah Dixon, Oliver F. Bathe, Francis R. Sutherland, Chad G. Ball
Department of Surgery, University of Calgary and the Foothills Medical Centre, Calgary, AB, AB, Canada
Background: Postoperative renal failure is reported in approximately 15% of liver resections and is associated with increased morbidity and mortality. Patients with hepatocellular carcinoma (HCC) represent a high-risk population due to the increased prevalence of cirrhosis and potentially poor physiologic reserve. Applicability of Acute Kidney Injury Network (AKIN) criteria (48-hour increase in serum creatinine and urine output) remains poorly studied in this cohort of patients.
Methods: All patients undergoing liver resection for HCC between January 2010 and October 2014 at a high-volume, quaternary care HPB referral center were included. Perioperative care remained constant over the study period, with routine intraoperative fluid restriction to maintain low central venous pressure (CVP) during resection, as well as reestablishment of normovolemia following specimen removal (i.e. continued in the post-anesthesia care unit). Retrospective data collection from electronic medical records utilized 8-hour shift periods to assess urine output from the point of surgery until the morning of postoperative day 2. Institutional laboratory databases were used to extract baseline and postoperative creatinine values. AKIN criteria were employed to diagnose AKI within 48 hours after surgery. Continuous variables were compared using t-test, and statistical analyses were conducted using SPSS, v.19, Chicago, IL.
Results: A total of 71 liver resections was performed (median age: 62 years (IQR: 56-70); male=74.6%; median BMI=25.6 (IQR: 23.1-29.5)). Preoperative patient assessment identified 60.6% were ASA 2 and 28.2% were ASA 3. Underlying liver disease was identified in 58 patients (hepatitis B or C in 48 patients); 57 Child-Pugh A and 1 Child-Pugh B (score 7). Fifty-five surgeries (77.5%) were minor liver resections (2 segments or less), and 12 of these were laparoscopic. Estimated blood loss was 200 ml (IQR: 87 - 400) for minor and 650 ml (IQR: 375 - 800) for major liver resections. Inflow occlusion was used in 14 (19.7%) surgeries, for an average of 18 minutes. Fifty-nine (83.1%) patients had confirmed background cirrhosis in the liver specimen. Median hospital stay was 9 days (IQR: 7-12) and 30-day mortality was 4.2%. The incidence of AKI was 21.1% based on creatinine and 53.5% with urine output criteria. AKI was associated with prolonged hospital stay when defined by serum creatinine criteria (10.4 vs. 19.8 days, p<.01), but not by urine output criteria (10.3 vs. 14.1 days, p=.478).
Conclusion: Urine output criteria results in an overestimation of AKI after liver resection for HCC, and therefore compromises the prognostic value of AKIN criteria in terms of hospital length of stay. Revision of the AKIN criteria to account for the physiologic postoperative reduction in urine output should be considered for patients with HCC undergoing low-CVP liver resection.
Back to 2015 Annual Meeting Program
|
