Minimally invasive cardiopulmonary bypass: does it really change the outcome?

During the last decade, many attempts have been made to reduce the deleterious effects of cardiopulmonary bypass (CPB) in cardiac operations. Basically, these attempts have focused on specific changes to the standard equipment and management, acting on the nature of the materials composing the CPB circuit and oxygenator, on the technical aspects of the pump and the circuit, and on the priming volume and the anticoagulation management.

Biocompatible materials of different types have been applied, but the results of this strategy are still debated. In a large, multicenter, prospective, randomized trial, no impact on postoperative outcome was found in low-risk patients undergoing elective coronary artery bypass graft (CABG) surgery [1]. However, a similar study, which focused on high-risk patients, demonstrated a better outcome in patients treated with heparin-coated materials [2]. Reduction of systemic heparinization coupled with the use of biocompatible materials seems to limit some of the adverse effects of CPB [3‐7]. New developments of coating systems for CPB materials have shown promising results [8‐10]. Many authors [10‐12] have stressed the negative action of open circuits (namely, of the reinfusion of shed blood from the pericardium) and proposed the use of closed circuits with separation of the surgical field suction blood. Finally, the need for limiting the circuit size in order to reduce the hemodilution degree has been thoroughly addressed in recent years, and many authors [13‐17] have demonstrated the deleterious effects of severe hemodilution during CPB in terms of postoperative morbidity and mortality.

In response to the considerable amount of literature stressing the possible impact of CPB technology on postoperative outcome, the industry proposed new 'minimally invasive' CPB (MICPB) systems. Despite differences in biocompatible treatment used and other minor technical issues, these systems all have some points in common: they (a) are closed, (b) are treated with biocompatible coatings, (c) can be primed with a reduced fluid volume, (d) include a centrifugal pump, and (e) are equipped with systems for separation of the shed blood from the circuit.

However, despite enthusiastic reports on the limited number of patients treated with these MICPB systems [18‐21], there is no clear information about the real impact of these technologies on the postoperative outcome of patients undergoing a cardiac operation. The present study aimed to determine the effects of an MICPB strategy on the postoperative outcome of a large population of patients undergoing CABG surgery.

The control group created through the propensity score technique was homogeneous with the MICPB group for all the preoperative variables and comorbidities (Table 1). The logistic Euroscore did not differ significantly between the control (7.2 ± 11.4) and the MICPB (7.5 ± 11.8) groups. Operative variables did not differ significantly between groups. Of course, as a result of the different CPB techniques, patients in the MICPB group had a significantly lower priming volume and consequently a significantly higher nadir HCT value on CPB. As a result of the different anticoagulation protocol, they received significantly less heparin and protamine.

In the univariate analysis, the postoperative outcome (Table 2) was affected by a higher morbidity in the control group. Patients in this group experienced more postoperative bleeding and had higher postoperative peak values of both serum creatinine and bilirubin. MICPB significantly reduced the risk for receiving an IABP and for atrial fibrillation, ventricular arrhythmias, cardiac arrest, and peripheral thromboembolism. Due to the reduced morbidity, patients treated with an MICPB had a significantly shorter ICU stay, and the rates of patients extubated early and discharged early from the ICU or the hospital were significantly higher (1.5 to 2 times) in the MICPB group. No significant difference was observed in terms of hospital mortality.

The main outcome variables significantly associated with the use of an MICPB system were analyzed using a multivariable stepwise forward logistic regression analysis. After correction for the other explanatory variables, MICPB remained independently associated with a better outcome, with the exceptions of early extubation rate and postoperative IABP use (Table 3).

Since in the mid-1980s, many attempts to improve the quality of CPB have been pursued. This trend toward a CPB improvement, which led to the introduction of hollow-fiber oxygenators, centrifugal pumps, and biocompatible treatments of the circuit and oxygenator decreased considerably in the early 1990s due to the emerging off-pump coronary surgery. In recent years, the new challenge has been to improve CPB quality to the point of making this technique competitive even with off-pump coronary surgery. Not by chance, during the last five or six years, the market has been offering an overwhelming number of new products (non-heparin-based biocompatible treatments, new generation centrifugal pumps, and MICPB equipment) aimed at improving the quality of CPB.

At present, at least three major companies are proposing an 'MICPB system.' With the exception of a few differences related mainly to safety devices (aimed at detecting and eliminating air entering the circuit), they share the same philosophy: closed circuit, centrifugal pump, hollow-fiber oxygenator, biocompatible surfaces, separation of the pericardial shed blood suction, and reduced priming volume. Biocompatible treatments of the circuit and oxygenator currently available on the market differ in nature (heparin, phosphorylcholine, sulphate-sulphonate groups, and so on) but exert a well-established action in limiting thrombin formation, platelet-count decrease, and inflammatory reaction [3, 5, 8, 9]. On a practical basis, commercially available integrated MICPB systems require a specific expertise, have a prolonged learning curve, need team work, and are more expensive than conventional circuits. Moreover, some concerns with respect to their safety have been raised [23]. Therefore, their use is justified only if their positive impact on postoperative outcome is well proven.

Unfortunately, this point is far from being clarified. Scientific reports on the use of MICPB systems are limited, represented mainly by case reports [20] or prospective trials enrolling limited numbers of patients [18, 19]. The general feeling is good, and various advantages have been reported: shorter ICU stay, decreased need for inotropes, less myocardial injury, lower rate of atrial fibrillation [18], and blunting of the inflammatory reaction and of activation of the coagulation and fibrinolytic systems [19]. The main risk with these initial trials on new therapeutic or technical systems is that patients in the study group may receive better care and report improved outcomes merely because they are being studied and received a new treatment; this bias is quite common and is known as the Hawthorne effect [24]. In a series of articles, Remadi and colleagues [25‐27] analyzed their experience with one of the commercially available systems. In the first article [25], they basically address the safety issues, practical feasibility, and general outcome of patients treated with an MICPB system. Subsequently, they published a prospective randomized trial on 100 patients undergoing aortic valve surgery [26], where patients treated with the MICPB had less neurologic events and a better preservation of renal function but no differences in mortality. Finally, they recently published a prospective randomized trial on 400 CABG patients [27], where the MICPB group demonstrated a lower rate of low cardiac output syndromes, need for allogeneic blood transfusions, and a better preservation of the renal function (lower peak values of creatinine and urea); mortality was not significantly different between groups.

The present study represents by far the largest experience with MICPB in CABG operations. The main strength of this study is the patient population size: more than 3,200 patients were analyzed. Conversely, the main limitation is that it is not prospective and randomized, the control group having been retrospectively created using a propensity score matching.

The main results of our study are that MICPB does not modify mortality but exerts a considerable beneficial effect on postoperative morbidity. In this respect, our data are in agreement with those of Remadi and colleagues [26, 27], who could not demonstrate different mortality rates in their prospective randomized trials. In our study, myocardial function was not addressed with specific markers; we saw a lower rate of patients receiving IABP support in the MICPB group in the univariate analysis, but this difference lost significance when corrected for confounding factors. Other articles suggest a better myocardial protection exerted by MICPB [18, 27] and a better hemodynamic profile [28], but we cannot confirm this on the basis of our data. However, there are lower rates of atrial fibrillation, ventricular arrhythmias, and cardiac arrest. Atrial fibrillation after cardiac surgery may recognize an inflammatory triggering mechanism [29‐31], and there is evidence of blunting of the inflammatory reaction in patients treated with MICPB [19] or when biocompatible circuits and separation of the shed blood suction are applied [32]. We therefore are inclined to attribute the atrial fibrillation containment to a better control of the inflammatory reaction in MICPB patients, but we recognize that we have no information about the possible role of prophylactic strategies (that is, perioperative beta-blocker use) within our study group.

In agreement with other studies [26, 27], we can confirm that MICPB systems exert a protective effect on perioperative renal function: the postoperative increase of serum creatinine was lower in the MICPB group even if the rate of patients requiring a dialytic treatment was not significantly different. Moreover, the peak postoperative serum bilirubin is lower in MICPB patients. Our MICPB has at least two characteristics that can justify a beneficial effect on the renal side: it guarantees higher HCT values during CPB and is equipped with biocompatible surfaces. Limiting hemodilution during CPB decreases postoperative renal dysfunction rate [13‐17], and biocompatible surfaces have been associated with a reduced rate of renal dysfunction in high-risk patients [2].

Peripheral thromboembolic events are less frequent when the MICPB is used. This can be attributed to the lower coagulation system activation, thrombin generation [32], and the antithrombin-saving effect already demonstrated with this technique [7]. However, we could not demonstrate that the overall rate of thromboembolic events (including myocardial infarction, stroke, pulmonary embolism, and mesenteric infarction) was lower in the MICPB group.

In our series, MICPB is associated with decreased postoperative bleeding but not with a lower transfusion rate. This finding is somewhat surprising, contradicting other studies [27] and our own finding in a published study on a limited number of patients [10]. However, we lack important information about the amount of blood product administered and cannot rule out that patients in the MICPB group may have received less blood product. Moreover, the lower postoperative bleeding detected is significant but clinically trivial (50 ml in 12 hours) and therefore unlikely to determine differences in transfusion rate. Finally, as a result of this better multifactorial outcome, patients in the MICPB group had a shorter ICU stay and higher rates of early ICU and hospital discharge.

Our MICPB system has some differences with respect to other commercially available products. It is not preassembled, and the venous return is not actively drained by a centrifugal pump, but passively gravity-drained into a collapsible reservoir; however, this difference does not result in a higher priming volume. In this respect, it is similar to the one proposed by Aldea and colleagues [5, 6, 32]. In their experience, this group demonstrated that a closed, heparin-bonded circuit with a reduction of systemic heparinization exerted a beneficial effect on postoperative outcome [5], reducing blood loss and transfusions, shortening mechanical ventilation time and ICU and hospital stay, and reducing postoperative complications (namely, thromboembolic events). The main characteristic of our system is that, due to the presence of a collapsible venous reservoir, no specific expertise or prolonged learning curve is required of the perfusionist. Moreover, the risk of air entering the circuit is not different from that of conventional open circuits, and no specific safety devices are required. Conversely, the preassembled circuits available on the market are generally based on an active venous return with the inlet of the centrifugal pump directly connected to the venous line. This almost invariably leads to the risk of air entering the circuit. Therefore, these circuits are equipped with bubble detectors and automated clamps allowing air to be eliminated from the circuit. Of course, this complex extracorporeal circuit needs a specific expertise, and doubts about the safety of the technique have been raised [23].

Heparin dose reduction associated with MICPB is still an open question. In the majority of studies published, the anticoagulation protocol is not changed [18‐21, 25‐27], whereas in our study MICPB was associated with a reduced heparin dose. Aldea and colleagues [5, 6] demonstrated that this approach is associated with a better outcome and does not induce adverse events. Øvrum and colleagues [4], in an impressive series of 5,658 patients, demonstrated that a strategy of reduced heparinization plus heparin-bonded circuits led to a postoperative outcome probably better than what was reported in the literature for off-pump CABG patients.

With respect to previous studies that failed to demonstrate a beneficial effect of heparin-coated, closed circuits, the main difference was the use of a complete MICPB system, including separation of the suction from the surgical field and a reduced priming volume.

The use of MICPB in coronary patients undergoing surgical revascularization is associated with an improvement in postoperative outcome. We did not find any difference in mortality, but given that this figure was approximately 3%, the study was probably underpowered in this respect. Even though our population was large and the selection bias was reduced by the propensity score analysis, our results will need to be confirmed in a large, prospective, randomized, controlled trial.