Introduction Patients receiving kidney replacement therapy (KRT) preoperatively who require cardiac surgery with cardiopulmonary bypass support present a particular challenge for the nephrologist. The earliest patient series from 1968 (1) highlights some of the challenges in anuric patients requiring KRT for either kidney failure or AKI who are undergoing cardiac surgery with cardiopulmonary bypass support, and they include the following: (1) Significant volume overload from obligate intravenous fluids and blood products. This can affect gas exchange and oxygenation due to pulmonary edema, delay wound healing, and, in extreme cases, prevent primary closure of the intrathoracic cavity. (2) Increases in both total body potassium and an increase in relative extracellular potassium. Exogenous sources of potassium include the cardioplegia solution ([K] approximately 26 mEq/L) (2) and the supernatant of red blood cell transfusions (approximately 10–60 mEq/L) (3,4). Transcellular shifting also occurs, due to acidemia, during the rewarming phase of iatrogenic hypothermia and from permanent cellular release from hemolysis, cell salvage techniques, and rhabdomyolysis. Inadequate or delayed control of hyperkalemia can lead to persistent cardiac asystole and an inability to come off cardiopulmonary bypass support in a timely manner. (3) Lactic acidosis during anaerobic respiration is common. Taken together, performing cardiac surgery with cardiopulmonary bypass support in patients requiring KRT preoperatively presents a particular challenge for the nephrologist managing these patients. Here, we will share our approach to managing such patients. Patient A 57-year-old woman with type 2 diabetes mellitus, coronary artery disease, and kidney failure on hemodialysis (HD) was diagnosed with mitral valve endocarditis. A combined coronary artery bypass graft and mitral valve repair with cardiopulmonary bypass support was planned. She received HD on the morning of her surgery. Intraoperatively while on cardiopulmonary bypass for 105 minutes, she was managed with convective hemofiltration run in parallel to the cardiopulmonary bypass circuit, which utilizes hydrostatic forces applied across a membrane to create an ultrafiltrate, and an equal weight-based volume of replacement fluid is added to the venous reservoir of the cardiopulmonary bypass circuit (5). At the start of the case, her potassium was 2.9 mEq/L, with a peak concentration of 6.1 mEq/L while on cardiopulmonary bypass, which was successfully lowered to 4.0 mEq/L at the end of cardiopulmonary bypass and remained at 4.0 mEq/L at the end of the 402-minute case. Because of postoperative hemodynamic instability, continuous KRT (CKRT) was initiated before eventually transitioning to HD. Our approach for providing hemofiltration KRT in patients on cardiopulmonary bypass is detailed below. Preoperative Optimization The preoperative management of KRT-dependent patients undergoing cardiac surgery with cardiopulmonary bypass support should focus on optimization of fluid balance and electrolytes and should be tailored to the individual patient. For stable patients undergoing elective procedures, we aim to provide their outpatient modality of KRT within 24 hours of surgery (6). For critically ill patients with hemodynamic instability, with acute coronary syndrome, or who have symptomatic obstructive coronary disease manifesting while on HD, we perform preoperative CKRT. Our main objective in the preoperative period is to provide some form of KRT within 24 hours of surgery in order to be optimized for the operation. Intraoperative Management Although early descriptions of intraoperative management of KRT-dependent patients on cardiopulmonary bypass relied exclusively on medical management, we feel that the risks of the aforementioned potential intraoperative complications necessitate KRT support while on cardiopulmonary bypass. Given the physical space constraints of the operating room for patients with complex cardiac surgical cases, the unpredictable timing of cardiopulmonary bypass support in relation to the operating room case time, and the large intraoperative multidisciplinary teams already caring for these patients, we prefer a strategy relying on hemofiltration on the cardiopulmonary bypass circuit over HD or CKRT (5,7). Utilizing hemofiltration performed by the perfusion team streamlines the coordination of timing between KRT and cardiopulmonary bypass support and minimizes the need for additional personnel and equipment in the operating room. This also allows for improved CKRT and HD resource allocation for other patients in the hospital. Peritoneal dialysis is never utilized intraoperatively at our institution. Perfusionists are already trained to perform hemofiltration and utilize preexisting connections on the cardiopulmonary bypass circuit, as they often use hemoconcentrator filters to produce an ultrafiltrate and manage volume at the end of cases. The cardiopulmonary bypass circuit is depicted in Figure 1A. The hemofiltration procedure utilizes a high-flux polyethersulfone membrane hemoconcentrator (effective surface area of 0.68 m2) (8), which is connected in parallel to the cardiopulmonary bypass circuit (Figure 1A) (5,7). The estimated BP postoxygenator is approximately 200–300 mm Hg, with variable blood flows ranging from 300 to 500 ml/min. Negative pressures ranging from −150 to −200 mm Hg on the effluent side of the hemoconcentrator are achieved via wall suction for a maximum transmembrane pressure (TMP) of 500 mm Hg (TMP = hemoconcentrator BP + negative effluent pressure). At maximum blood flows of 500 ml/min and maximum TMP of 500 mm Hg, ultrafiltration rates >100 ml/min can be achieved if needed (8). A weight-based protocol dictated by the patient’s most recent plasma potassium concentration guides the perfusion team for the appropriate goals for target ultrafiltration volumes every 15 minutes with a commensurate volume of commercially available replacement fluid to be added to the venous reservoir (Figure 1B). This hemofiltration circuit also allows for a negative fluid balance if needed, as is often performed in cardiopulmonary bypass cases. The total clearance of potassium (CK) is, therefore, proportional to the total effluent volume achieved (CK = [K]eff × Qeff, where Qeff = QRF + QUF).Figure 1.: The simplified cardiopulmonary bypass circuit modified for hemofiltration and the cardiopulmonary bypass hemofiltration protocol. (A) Deoxygenated blood (blue) from the patient flows into a venous cardiotomy reservoir before being pumped to an oxygenator membrane, and oxygenated blood is returned to the patient (red). In order to perform hemofiltration, a high-flux polyethersulfone membrane hemoconcentrator is connected in parallel to the cardiopulmonary bypass circuit, with the inlet to the hemoconcentrator coming off of a shunt from the oxygenator and the outlet of the hemoconcentrator returning to the venous line from the patient and ultimately returning to the venous reservoir. Positive hydrostatic pressure is applied to the blood side of the hemoconcentrator, and negative pressure is applied on the effluent side to create a transmembrane pressure gradient, which generates an ultrafiltrate (yellow effluent). An equal volume of replacement fluid (yellow) is added to the venous reservoir. (B) While on cardiopulmonary bypass, plasma potassium concentrations are checked every 15 minutes. The most recent concentration drives a weight-based fluid replacement protocol with lower potassium content and higher volumes of replacement fluid used when hyperkalemia is more severe. ABG, arterial blood gas; iCa, ionized calcium; K, potassium; z-BUF, zero-balance ultrafiltration.Although this therapy allows for streamlining of operational processes and independence of the intraoperative team, it can only be provided while the patient is cannulated for cardiopulmonary bypass, and any emergent indications for KRT after the patient is off cardiopulmonary bypass but still in the operating room would necessitate CKRT or HD. Postoperative Management Similar to the preoperative period, the choice of modality for postoperative KRT and timing will be dictated by each individual patient’s clinical condition. Because of frequent intolerance of a positive fluid balance in this critical postoperative period, we aim to provide KRT by postoperative day 1 for all patients. Patients who are hemodynamically unstable are initiated on CKRT either when indications arise or by 24 hours postoperatively. Patients who are receiving systemic anticoagulation have a lower risk of circuit clotting in this postoperative period; however, when they are not receiving systemic anticoagulation, we use regional citrate or prefilter heparin when possible. Although some groups have reported success in managing patients with peritoneal dialysis postoperatively (9), we do not utilize peritoneal dialysis in the immediate postoperative period given the potential for transdiaphragmatic leaks, increased opposing transdiaphragmatic pressures that may limit ventilation, and the variable perfusion of the peritoneal membrane during severe circulatory shock. Summary Performing cardiac surgery with cardiopulmonary bypass support in KRT-dependent patients presents a particular challenge for nephrologists mainly due to fluid balance, hyperkalemia, and acid-base disturbances. Convective clearance using a weight-based hemofiltration protocol performed in parallel to the cardiopulmonary bypass circuit without the use of a separate KRT machine or personnel allows for streamlining of services and teams already in the operating room. Although many patients can be managed with their outpatient dialytic modality preoperatively, many of these patients require temporary CKRT postoperatively while they remain hemodynamically unstable until they are eventually transitioned to their outpatient modality. Disclosures All authors have nothing to disclose. Funding None.