Victims of Our Own Success…and Failure

Like in many aspects of medicine, including cardiac surgery, general surgery, and trauma, impaired renal function, hyponatremia, and the requirement for renal replacement therapy are well-established risk factors portending morbidity and mortality.^1–3 Cardiorenal interactions in end-stage heart failure are highly complex and have major implications for outcomes of advanced heart failure therapy.^4–6 Indeed, almost all clinical trials in mechanical circulatory support (MCS) for heart failure have listed significant renal impairment as an exclusion criterion for enrollment; therefore, the MCS community has little insight in regard to controlled data.

In retrospective series, MCS has been shown to improve renal function with both pulsatile and continuous-flow left ventricular assist devices (LVADs).^7,8 In this issue of ASAIO Journal, the team from Virginia Commonwealth University describes their single-center experience with SynCardia temporary-Total Artificial Heart (TAH) (SynCardia Systems, Inc., Tucson, AZ) support and associated renal outcomes.⁹ Many end-stage heart failure patients with chronic biventricular failure show evidence of significant end-organ compromise at the time of TAH implantation.¹⁰ In addition to renal dysfunction, comorbidities often include elevated pulmonary vascular resistance, hepatic compromise, malnutrition, skeletal sarcopenia, and overall marked noncardiac frailty. The current study describes renal failure quantitatively in accordance with National Kidney Foundation guidelines, making this report unique and of high quality. The analysis revealed that 75% of patients after successful TAH had improved renal function by these standards, inclusive of patients who required preoperative dialysis. As a result of these data, the authors cautioned that although preoperative renal insufficiency portended a poor prognosis, many patients experienced renal recovery and can be successfully bridged to cardiac transplant alone.

Equally striking, however, is the number of patients with previously normal renal function who progressed to dialysis-dependent renal failure after successful TAH implantation. Even more troublesome is the 50% mortality the authors observed in those patients requiring dialysis after TAH whom had normal preoperative renal function. These findings imply that we have limited understanding of predicting renal function trajectory with TAH support. This is certainly an important topic in LVAD therapy with ongoing dogma, intense investigation, and controversy.¹¹ Indeed, the MCS field seeks the ability to predict renal recuperative potential. Risk models aside, this is most likely accomplished with more clinical experience and better understanding of physiologic principles.

In many heart failure patients, renal insufficiency is often attributed to low cardiac output, hypotension, nephrotoxic drugs, or volume depletion because of overdiuresis. It has also been well-established that venous congestion manifest by elevated right-sided filling pressures has a significant impact (and perhaps the highest impact) on acute and chronic renal function.^12,13 Not only is elevated central venous pressure (CVP) predictive of worsening renal function but higher CVP is also predictive of severity of renal dysfunction. Given the biventricular failure in the majority of patients undergoing TAH implantation, the implications of this observation may be significant.

The unique nature of the cardiectomy required for TAH support completely interrupts inherent physiologic cardiorenal feedback mechanisms. B-type natriuretic peptide (BNP), secreted by ventricular myocytes when stretched, has significant vasculo-epithelial effects, including a decrease in vascular tone, increase in renal blood flow, suppression of the renin–aldosterone axis, and increase in natriuresis. This has led to the observation and common practice of exogenous BNP administration after TAH implantation.¹⁴ Unfortunately, this practice has only demonstrated improved urine output; long-term effects on renal function have not been studied in a systematic fashion. Nonetheless, this is an excellent example of the new physiology imposed on these patients, with effects as yet truly unknown.

The pressure–volume relationships in the TAH lend themselves to excellent study in a simulated environment. The TAH simulator provided by SynCardia Systems, Inc., provides educators and learners a unique insight into basic physiology and device parameters. Specifically lacking, however, is the effect of venous and pulmonary capacitance in the model; rather, it is restricted to changes in resistance only. This, coupled with the lack of compliance in the TAH ventricles, makes for a very narrow euvolemic window—volume enough to fill the device, but not too much volume to result in inadequate emptying. In our experience, this window amounts to approximately 30 ml per cardiac cycle. In real patient management, this window is bracketed by either underfilling (low-flow alarms) or overfilling of the TAH ventricles (lack of ejection flags on the monitor). Our reaction to these alerts are of course patient and scenario specific; in general, low-flow alarms are met with immediate action, whereas poor ejection flags may be treated over the course of hours. Remember, the lack of ejection flags, especially on the right side, may signify central congestion and may contribute to new or ongoing renal dysfunction. These factors are made more salient by the limitations of invasive monitoring in these patients because of concern for catastrophic device failure from wire entanglement in the valvular apparatus.¹⁵ Whether this may be mitigated by transcutaneous monitoring devices remains unknown and worth considering.¹⁶ The limitation of the simulator, and our response to patient conditions, may be an opportunity to advance the appreciation of these parameters.

Survival with the TAH remains well below continuous-flow LVADs. In part, this is readily explained by the inherently more complex nature of patients in need of TAH support. However, as evidenced by these data, many questions regarding predicting postoperative outcomes remain. As technology continues to evolve and we develop novel and improved life-saving devices, we find ourselves facing new challenges that may represent poorly understood physiology. This is a testament to the success achieved in this field—we have the ability to at times recover organ function previously considered contraindications to advanced heart failure therapy. At the same time, along that journey we do not infrequently unmask unintended consequences of our care.