A propensity score matched analysis of liver transplantation outcomes in the setting of preservation solution shortage

The recent shortage of the University of Wisconsin (UW) solution prompted increased utilization of histidine-tryptophan-ketoglutarate (HTK) solution for liver graft preservation. This contemporary study analyzed deceased donor liver transplant outcomes following preservation with HTK vs UW. Patients receiving deceased donor liver transplantations between January 1, 2019, and June 30, 2022, were retrospectively identified utilizing the Organ Procurement and Transplant Network database, stratified by preservation with HTK vs UW, and a propensity score matching analysis was performed. Outcomes assessed included rates of primary nonfunction, graft survival, and patient survival. There were 4447 patients in each cohort. Primary nonfunction occurred in 60 (1.35%) patients in the HTK group vs 25 (0.54%) in the UW group (P < .001). HTK was associated with lower 90-day graft survival (94.39% vs 96.09%; P < .001) and 90-day patient survival (95.97% vs 97.38%; P = .001). Unmatched donation after cardiac death-specific analysis of HTK vs UW demonstrated respective rates of primary nonfunction of 1.63% vs 0.82% (P = .20), 90-day graft survival of 92.50% vs 95.29% (P = .069), and 90-day patient survival of 93.90% vs 96.35% (P = .077). These results suggest that HTK may not be an equivalent preservation solution for deceased donor liver transplantation.

Concurrent Cholecystectomy Does Not Increase Splenectomy Morbidity in Patients With Hemolytic Anemia: A Pediatric NSQIP Analysis

PURPOSE

Children undergoing splenectomy for hemolytic anemia often have cholelithiasis, which may or may not be symptomatic. It is unclear whether concurrent cholecystectomy increases length of stay or morbidity after splenectomy. The purpose of this study was to compare morbidity among children undergoing laparoscopic splenectomy alone versus splenectomy with concurrent cholecystectomy in patients with hemolytic anemia.

METHODS

We retrospectively evaluated children with hemolytic anemia undergoing non-traumatic laparoscopic splenectomy in the National Surgical Quality Improvement Program-Pediatric database (2012-2020). Outcomes were compared for patients undergoing splenectomy alone (n = 1010) versus splenectomy with cholecystectomy (n = 371). Pearson's Chi-square and Student's t-tests were utilized as appropriate. Propensity score-matching was completed, controlling for eight demographic and clinical variables.

RESULTS

1381 patients were identified, 73.1% undergoing splenectomy alone and 26.9% splenectomy with cholecystectomy. Splenectomy with cholecystectomy patients were older (10.9 years vs. 8.4 years, p < 0.01), more likely to have hereditary spherocytosis (56.1% vs. 40.8%, p < 0.01), less likely to have sickle cell disease (12.1% vs. 33.5%, p < 0.01), more likely ASA class 1 or 2 (49.3% vs. 42.1%, p < 0.01), and had similar preoperative hematocrit levels (29.6 vs. 29.3, p = 0.33). The splenectomy with cholecystectomy group was less likely to receive preoperative blood transfusions (13.5% vs. 25.4%, p < 0.01). There were 360 pairs selected on propensity score-matching, and splenectomy with cholecystectomy was associated with increased operative time (182 min vs. 145 min, p < 0.01) and decreased occurrences of a postoperative transfusion (4.2% vs. 8.9%, p = 0.01). Length of stay after surgery (2.5 days vs. 2.3 days, p = 0.13), composite morbidity (3.9% vs. 3.4%, p = 0.69), and 30-day readmission rates (3.3% vs. 7.4%, p = 0.08) were all similar.

CONCLUSIONS

Splenectomy with cholecystectomy is associated with similar postoperative morbidity, length of stay and readmission rates compared to splenectomy alone. These data support the safety of concurrent cholecystectomy with splenectomy for children with cholelithiasis in the setting of hemolytic anemia.

TYPE OF STUDY

Retrospective Cohort Study.

LEVEL OF EVIDENCE

Level III.

Association of Age at Duodenal Atresia Repair With Outcomes: A Pediatric NSQIP Analysis

PURPOSE

Neonates with duodenal atresia (DA) are often born prematurely and undergo repair soon after birth, while others are delayed to allow for growth until closer to term corrected gestational age (cGA). Premature infants have been demonstrated to experience worse outcomes, but it is unclear whether delaying surgery mitigates the increased morbidity. This study evaluates the association of timing of DA repair with postoperative morbidity.

METHODS

We retrospectively evaluated neonates undergoing DA repair from the National Surgical Quality Improvement Program-Pediatric database (2015-2020). A multivariable regression analyzed factors associated with composite morbidity, including cGA and age in days of life (DOL) at surgery. A propensity score matched analysis was completed in premature neonates born at ≤35 weeks gestation to compare outcomes at similar birth gestational ages (bGA) and birth weight who underwent early (<7 DOL) versus delayed (≥7 DOL) repair.

RESULTS

809 neonates were included with a median bGA of 36 weeks (IQR 34-38), birth weight of 2.46 kg (IQR 1.96-2.95), and DOL at surgery of 2 (IQR 1-5). Infants born ≤35 weeks represented 35.23% of the cohort. On multivariable analysis, increasing cGA at surgery was associated with decreased morbidity (OR: 0.91, CI [0.84, 0.99]), and increasing DOL at surgery was associated with increased morbidity (OR: 1.02, CI [1.00, 1.04]). On propensity score matched analysis, delayed repairs were associated with increased postoperative ventilation (6 days vs. 2 days, p < 0.05); however, there were no differences in composite or surgical morbidity between early and delayed repairs.

CONCLUSIONS

Morbidity after DA repair in neonates ≤35 weeks cGA is primarily driven by non-surgical causes, but delaying surgery does not appear to mitigate the risks associated with prematurity. It seems reasonable to consider repair in neonates around 33-34 weeks gestation without prohibitive risk factors. Optimal timing of DA repair requires a delicate balance between these factors.

LEVEL OF EVIDENCE

Level III.

TYPE OF STUDY

Retrospective Cohort Study.

Assessing Kidney Transplantation Using ECMO-Supported Donors Within a KDPI-Based Allocation System

Background

Organ donors supported by extracorporeal membrane oxygenation (ECMO) have historically been considered high-risk and are judiciously utilized. This study examines transplant outcomes using renal allografts from donors supported on ECMO for nondonation purposes.

Methods

Retrospective review of the Gift of Life (Pennsylvania, New Jersey, Delaware) organ procurement organization database, cross-referenced to the Organ Procurement and Transplantation Network database, assessed kidney transplants using donors supported on venoarterial (VA) and venovenous (VV) ECMO for nondonation purposes. Transplants using VA- and VV-ECMO donors were compared with Kidney Donor Profile Index (KDPI)-stratified non-ECMO donors. Regression modeling of the entire ECMO and non-ECMO populations assessed ECMO as predictive of graft survival. Additional regression of the ECMO population alone assessed for donor features associated with graft survival.

Results

Seventy-eight ECMO donors yielded 128 kidney transplants (VA: 80, VV: 48). Comparing outcomes using these donors to kidney transplants using organs from KDPI-stratified non-ECMO donors, VA- and VV-ECMO donor grafts conferred similar rates of delayed graft function and posttransplant renal function to KDPI-matched non-ECMO counterparts. VA-ECMO kidneys demonstrated superior graft survival compared with the lowest-quality (KDPI 86%-100%) non-ECMO kidneys and similar graft survival to KDPI <85% non-ECMO kidneys. VV-ECMO showed inferior graft survival to all but the lowest-quality (KDPI 86%-100%) non-ECMO kidneys. VV-ECMO, but not VA-ECMO, was associated with increased risk of graft loss on multivariable regression (hazard ratios-VA: 1.02, VV: 2.18). Higher KDPI, advanced age, increased body mass index, hypertension, and diabetes were identified as high-risk features of ECMO donors.

Conclusions

Kidney transplantation using appropriately selected ECMO donors can safely expand the donor pool. Ongoing studies are necessary to determine best practice patterns using kidneys from these donors.

Extracorporeal membrane oxygenation for cardiopulmonary failure in organ donation: Assessing liver transplant outcomes

Liver transplantation continues to face significant organ shortages and efficient utilization of marginal donors is paramount. This study evaluates the practice patterns and outcomes in liver transplantation when utilizing allografts from marginal donors who required extracorporeal membrane oxygenation (ECMO) support. We performed a retrospective review of the Gift of Life (PA, NJ, DE) organ-procuring organization database for transplants performed using donors supported on ECMO for nondonation purposes. These were cross-referenced to the transplant recipients within the Organ Procurement and Transplantation Network database, and the outcomes of liver transplants using donors on ECMO support were compared with those not requiring ECMO. Organ use and nonuse patterns were then evaluated in ECMO-supported donors, identifying the factors associated with nonuse compared with the factors associated with graft failure. Thirty-nine of the 84 ECMO-supported donors contributing at least one intra-abdominal organ for transplant donated a liver. Graft survival and patient survival up to 5 years were comparable between transplants from ECMO and non-ECMO-supported donors, and no cases of primary nonfunction were seen in the ECMO group. ECMO support was not associated with 1-year graft failure on regression modeling. Additional regression analyses within the ECMO donor population identified bacteremia (HR: 19.81) and elevated total bilirubin at donation (HR: 2.44) as predictive of post-transplant graft failure. Livers from donors supported on ECMO before donation appear safe to use in select transplant settings. Better understanding of the impact of predonation ECMO on liver allograft function will help guide the optimal use of these scarcely used donors.

Euglycemic Diabetic Ketoacidosis Due to Sodium-Glucose Cotransporter 2 Inhibitor Use in Two Patients Undergoing Pancreatectomy

Background: Euglycemic diabetic ketoacidosis (euDKA) is a potential side effect associated with inhibitors of the sodium–glucose cotransporter 2 (SGLT-2). This effect is most often recognized during physiologic stress (i.e., sepsis) or in patients who undergo surgery.

Case presentations: Case 1: A 66-year-old woman underwent distal pancreatectomy with en bloc splenectomy after presenting with a biopsy-proven pancreatic adenocarcinoma in the body of the pancreas noted incidentally on a screening magnetic resonance imaging for an ovarian mass. The patient had a history of type 2 diabetes mellitus (T2DM) and used canagliflozin, which she was instructed to hold 24 h before surgery. Case 2: A 75-year-old man underwent a pylorus-preserving pancreaticoduodenectomy after presenting with obstructive jaundice. This patient also had a history of T2DM and was on dapagliflozin, which he was also instructed to hold 24 h before surgery. Postoperatively, both patients were diagnosed with euDKA, which was suspected primarily because of intraoperative and postoperative polyuria.

Conclusions: SGLT-2 inhibitors are associated with euDKA that can be potentiated in patients who undergo surgery. This medication side effect can be easily unrecognized and potentially lead to significant morbidity.