Acceptable short-term outcomes after open heart surgery in kidney transplant recipients: a case control study

Background Cardiovascular disease is common in kidney transplant recipients. They are considered high risk surgery candidates due We assessed short-term results of open-heart surgery in kidney transplant recipients and matched controls between 1989 – 2016 at our center. Methods Ninety-five patients underwent open heart surgery (48 coronary artery bypass grafting, 27 valve replacements or repairs and 29 combined procedures) after kidney transplantation. Controls (n=95) were matched for age, sex, type and year of surgery. Mean follow-up was 5.6 (4.9) years. Independent two-sample t-test and chi-square test were used to compare continuous variables and frequencies, respectively. Logistic regression was used to identify preoperative risk factors for 30-day mortality. Results Included were 76 men and 19 women; mean age 60.3 (11.1) years, 7.1 (5.6) years after transplantation. Kidney transplants had lower renal function, more hypertension, but less pulmonary hypertension than matched controls. Intraoperative data was comparable between kidney transplants and controls. Kidney transplants experienced more frequent acute kidney injury (57% versus 23%, p<0.001), more bleeding (1288 (1081) mL versus 957 (548) mL, p=0.01) and more red cell transfusions (4.9 (5.6) versus 3.2 (5.4) units, p=0.04). Infections were borderline more frequent in kidney transplants (30% versus 20%, p=0.10). Thirty-day mortality was 6.3% in kidney transplants and 2.1% in controls (p=0.14). Independent risk factors for 30-day mortality were acute myocardial infarction last 90 days before surgery (OR 12.5, p=0.02) and current smoking (OR 17.3, p=0.02). Conclusions Kidney transplant recipients undergoing cardiac surgery have acceptable short-term results compared with matched controls; 30-day mortality rates were similar. Careful peri- and postoperative management is, however, warranted as kidney transplant recipients experience more bleeding and higher frequency of AKI.

We present detailed short-term results of heart surgery in KTR and matched controls performed in Norway during a time period of almost three decades. The aim of our study was to assess the shortterm risk of open-heart surgery in KTR and describe the influence of preoperative risk factors.

Methods
Between January 1989 and July 2016 more than 12000 coronary artery bypass graft procedures (CABG) and/or heart valve replacements or repairs were performed at the Department of Thoracic Surgery at Oslo University Hospital, Rikshospitalet. The departmental database, Datacor, was crosschecked with the Norwegian Renal Registry to identify KTR that underwent CABG, valve replacement or a combined procedure after kidney transplantation. The Renal Registry is a nationwide quality registry with 100% coverage of all KTR in Norway. In total, we identified 95 KTR who underwent open heart surgery at Oslo University Hospital, Rikshospitalet. Three of the patients were simultaneous kidney-pancreas recipients. Controls were selected from the thoracic surgery database on a 1:1 basis, matched for age, sex, type of procedure, year of procedure (+/-2 years). All 190 patient records were meticulously manually reviewed to double check the data from the registries, and supplement with relevant additional data such as accurate patient history, total peri-and postoperative blood product transfusions, all postoperative complications up to 30 days after surgery. For procedures performed between 1989-2003, paper records were reviewed, whereas electronic records were available after 2003.
Short-term mortality was defined as mortality within 30 days after heart surgery and is presented as Kaplan-Meier actuarial survival estimates. Log-rank test was used to compare survival rates. Analyses 4 comparing continuous variables were based on independent two-sample t-test (except for age that was matched). Chi-square test was used to compare frequencies between two groups, whereas Fisheŕ s exact test was used if any observed count was less than 5. Binary logistic regression was used to calculate risk factors for 30-day mortality. The multivariable model was based on the backward stepwise (Wald) selection. All statistical analyses were performed using SPSS for Windows, version 23 and 24 (IBM Corp, Armonk, NY) software or R, version 3.3.2 (The R Foundation for Statistical Computing).

Results
Demographic data and baseline characteristics are presented in Table 1. Eighty percent were men and average age for all patients was 60.3 (11.1) years. The procedures were 51% CABG (n=48), 28% (n=27) isolated valve replacements or repairs and 21% (n=20) combined procedures, all performed on average 7.1 (5.6) years after transplantation. The most common valve procedure was aortic valve replacement (AVR). Twenty-four isolated AVRs and 18 combined CABG and AVR procedures were performed. Three KTR and seven controls had active endocarditis at time of surgery (p=0.33).
Thirteen KTR had experienced graft loss and were established on chronic dialysis treatment at time of heart surgery. In those not on dialysis (n=82), the CKD-EPI estimated GFR at time of surgery was lower compared to controls; 45 vs 77 mL/min/1.73m2 (p<0.001). e Heart surgery within 24-hours from diagnosis.
Thirty-day mortality was 6.3% in KTR and 2.1% in controls (p=0.14; Log rank test), see Figure 1. The frequency of acute kidney injury (AKI) was significantly higher in KTR, 57% vs. 23% (p<0.001), driven mainly by a higher incidence of AKI stage 1, i.e. an absolute increase in serum creatinine of ³0.3 mg/dL or 25%. See  Table 2.  There was no statistical difference in intraoperative data and perioperative complications. However, perioperative stroke occurred in 6 KTR (6.3%) and in only 1 control patient (1.1%), p=0.12, see Table   3.   (28). This study is not only interesting due to the significant sample size, but also due to a similar distribution of cardiac procedures performed compared with our study. Others report a varying range of short-term mortality rates (11, 13-20, 22-25, 27, 29, 30), ranging from 1.4% (17) to 14% (30) depending on type of procedures and patient population.
Perioperative stroke was more frequent among KTR than controls, 6.3% vs 1.1% (n=6 vs n=1, As many as 57% of the our KTR experienced AKI, as compared to 23% of the controls. These high numbers were driven by the high incidence of AKI stage 1 in both patient groups, and especially in the former. This can be explained by the readily attainable 0.3 mg/dL increase in creatinine in KTR whose average preoperative creatinine was 1.  (17,28), report much lower frequencies of postoperative infections, approximately 7%, with no difference between groups. Others report rates of 9-21% (14,15,18). A possible explanation for the high frequency of reported infections in the present study was the meticulous manual review of all patient records, thus fewer infections were missed.
The only significant preoperative risk factors for 30-day mortality in our KTR were AMI within 90 days before surgery and active smoking. Although these risk factors are plausible, the interpretation requires caution as the mortality in our cohort was low. Rocha et al (13)  new-onset dialysis and septicemia. We also evaluated potential perioperative risk factors in our KTR cohort. Like Farag et al we confirm that AKI requiring dialysis postoperatively is an independent risk factor for mortality (data not shown).

Conclusions
KTR undergoing cardiac surgery have acceptable short-term results compared with matched controls.
Thirty-day mortality rates were 6.3% and 2.1%, respectively. Careful peri-and postoperative management is warranted as KTR do experience more bleeding and higher frequency of AKI.

List Of Abbreviations
ACEI: Angiotensin II converting enzyme inhibitor AKI: Acute kidney injury