Discussion
CPB is arguably one of the greatest developments to allow for the expansion and advancement of the field of cardiac surgery1,12. Over the decades, preexisting technologies have improved with the goal of reducing CPB-related complications. Some of these technologies include centrifugal pumps, improved oxygenators, and heparin-bound circuit tubing2. These advancements aim to reduce air emboli and reduce thrombus formation within the CPB circuit, and thus eliminate thromboembolic complications. Despite such advancements, MRI-adjudicated studies have revealed shockingly high rates of new, albeit mostly clinically insignificant, brain lesions following on-pump cardiac surgery13,14. Additionally, CPB runs are also linked to elevation of pro-inflammatory cascades, which can result in the development of systemic vasoplegia and end-organ malperfusion15–19. Though necessary for the field of cardiac surgery, CPB technology remains an area for improvement.
Because of the known complications and increased mortality risk with prolonged CPB usage9–11, it is universally accepted that pump runs be kept to a minimum, and that cardiac procedures be planned accordingly to be conducted in such a way to eliminate unnecessary CPB and cross-clamp times. Nissinen and Madhavan have suggested that the optimal CPB time to be under 180-240 minutes in order to minimize risk of severe complications and/or mortality7,11. In the analysis by Nissinen and colleagues, increasing CPB time (per 30 minutes OR 1.47, 95% CI 1.27-1.71 was associated with increased odds of 30-day mortality, independent of case complexity or patient comorbidities. In their series, only 30 patients had a CPB time longer than 300 minutes, and 30-day mortality was 56.7% in this cohort. In our cohort of 293 patients, operative mortality was 22.5%, significantly lower than previously described. In an evolving era of cardiac surgery, increasingly complex procedures are being performed, along with complex redo operations, which not surprisingly require lengthier pump runs. Although prior studies have correlated poorer outcomes with increasing CPB times, the current limitations of our existing technology are not well understood.
As stated, increased CPB times have been linked to worse survival, renal, failure, and increased intensive care unit utilization4,7,9,20, which likely are associated with increased healthcare costs. There are limited studies that have specifically investigated outcomes of patients in which very-long CPB perfusion times were required. One such study evaluated patients with aortic cross-clamp times exceeding 300 minutes with an average CPB time of 420 minutes, and demonstrated a 30-day mortality of 12.4%21. In patients surviving at least 30 days, the 1-year survival was 92% in their series, again supporting the notion that longer term survival is favorable if early mortality is avoided in this patient cohort.
As expected, this population was prone to a high degree of complications in our analysis, namely prolonged mechanical ventilation, renal failure, and need for reoperation and blood transfusion. Furthermore, one fifth of patients died within the perioperative window. However, at one year, more than 60% of patients had survived with roughly one third of patients requiring readmission for any cause. Such findings suggest that although these cases with excessive perfusion times, whether planned or unplanned, are faced with a very high risk of operative mortality and morbidity, mid-term results are acceptable with the majority of patients being alive. Therefore, cardiac operations requiring excessive CPB times should not be considered futile.
The granular causes of very-long CPB times can be multifactorial and were not captured in our data registry. These causes can be categorized as multi-component procedures that require lengthy operative times even if performed efficiently, or cases that should have limited CPB times but where intraoperative complications or issues led to a very prolonged operative course. Nearly 50% of the cases in our series were multi-component cases consisting of coronary revascularization combined with a valve procedure, or a multi-valve operation. The most common procedure was an aortic root replacement. By comparison, the mean CPB times reported for combined CABG with aortic valve replacement or isolated aortic root replacement are 123-203 minutes22,23 and 122-237 minutes24–26, respectively. This suggests that even in these typically lengthier cases, the CPB times typically do not exceed 300 minutes. Factors such as having to revise distal anastomoses in CABG, replacing a valve in which a suboptimal repair was performed, or reapplying a cross-clamp to address a paravalvular leak in valve replacement surgery can add time to the operation. Intraoperative complications, although rare, can add substantial operative time, including such events as atrioventricular groove disruption during mitral valve replacement, root disruption during aortic valve surgery, or poor coronary flow after a root replacement requiring coronary revascularization.
The longer CPB times may also reflect reperfusion time after an unsuccessful attempt at weaning the patient from CPB at the conclusion of the case. Most surgeons, after an initial CPB wean, will empty and reperfuse the heart for a longer period of time if the ventricular function is suboptimal and the patient cannot be weaned successfully and safely from CPB. This can be the result of poor baseline ventricular function, suboptimal myocardial protection, or temporary insults such as air embolism. Protamine reactions can also occur requiring reinstitution of CPB with added time.
Another important factor in achieving reasonable success in these more complex scenarios is early institution of temporary mechanical circulatory support. This is reflected in the 20% of patients in whom we utilized an intraoperative IABP. In cases where the patient cannot be weaned from CPB, we typically will attempt weaning with IABP support, and then escalate to ECMO if the patient is still unable to be weaned. In addition, we prefer to utilize IABP support in marginal cases where the patient can be weaned from CPB but is requiring high levels of inotropic support in the setting of depressed ventricular function. The use of mechanical support depends heavily on the surgeon’s judgement. However, an increasing body of evidence underscores the importance of these devices in reducing native cardiac stress and distension, improving contractility and myocardial recovery, and increasing end-organ perfusion, all of which contribute to the overall mortality and morbidity of the patient27–31.