Low cardiac output syndrome (LCOS) is a potential complication in patients that have undergone cardiac surgery (CS). According to the studies found in the literature, the incidence of the syndrome varies between 3 and 45%, and it is associated to increased morbidity–mortality, a prolongation of stay in the intensive care unit (ICU), and increased consumption of resources.1–4 Low cardiac output syndrome can be regarded as a “peculiar” form of acute heart failure (AHF), characterized by triggering factors, a form of presentation, evolution and treatment that differ from those of AHF of medical origin. The AHF classifications of the European Society of Cardiology and of the American College of Cardiology are not directly applicable to the postoperative period of CS; in fact, they do not contemplate the mentioned syndrome.5 Postoperative LCOS involves the association of cardiogenic, distributive and hypovolemic components that affect patient hemodynamics and imply complex management requiring a dynamic clinical approach. In this context, in the year 2012 a consensus document was published specifically defining this clinical condition and offering a series of global recommendations for its treatment.6

The present study investigates the possible pre- and intraoperative factors associated to LCOS and which can influence its outcome. The complex postoperative course of these patients is examined, and a study is made of the differences among the different clinical diagnostic subgroups (referred to both the association of predisposing factors and evolution and prognosis).

A prospective, multicenter, observational cohort study was made of a group of patients with LCOS in the postoperative period of CS (the ESBAGA Study), with the purpose of evaluating the postoperative course and prognosis (clinical evolution, presence of complications and mortality) of the patients, and the differences according to the three different LCOS diagnostic subgroups defined by the SEMICYUC Consensus 2012. The pre- and intraoperative factors associated to development of the syndrome were also analyzed.

A total of 14 Intensive Care Units belonging to hospitals throughout Spain participated in the ESBAGA study (Table 1). We analyzed a total of 2070 patients subjected to CS, of which 155 presented LCOS in the immediate postoperative period (7.5%). Eighteen of these subjects declined participation in the study; 137 patients were therefore finally evaluated. Fig. 1 shows the flowchart of these patients.

Figure 1.

Patient flowchart. LCOS: low cardiac output syndrome.

(0.11MB).

The general patient characteristics are shown in Table 2. Myocardial revascularization, either isolatedly or accompanied by valve treatment, were the most frequent operations. With regard to the most relevant patient antecedents (Tables 2 and 3), mention must be made of the high percentage of patients presenting serious clinical disease before surgery, with NYHA functional class III and IV, preoperative critical conditions, the need for vasoactive drugs, or ventricular dysfunction with LVEF <35%. Many patients had suffered prior AMI (22 cases in the 90 days preceding surgery), percutaneous coronary intervention (PTCA/stent), previous CS, or moderate to severe pulmonary hypertension.

As intraoperative factors associated to LCOS, we found that 70.1% of the patients required two or more vasoactive drugs on emerging from ECC. Pump reentry to secure definitive suspension of ECC proved necessary in 22 patients (16%). The mean ECC time was 147±60min, and the mean aortic clamping time was 105±43min. Prolonged surgical times, with ECC >120min, were recorded in 62.5% of the patients.

The above factors were compared among the different LCOS clinical diagnostic subgroups. In this regard, 46 patients met hemodynamic criteria of LCOS (group A), 50 complied with the clinical criteria (group B), and the rest (n=41) were diagnosed with cardiogenic shock (group C). No significant differences between groups were observed for most of the factors. However, differences were recorded in those patients with a more serious prior condition, such as the presence of resting angina, presurgical critical condition, NYHA functional class III and IV, the need for emergency surgery, and a high risk EuroSCORE II profile. The percentage of patients with prolonged surgical times (ECC >2h) was greater among those with cardiogenic shock (n=32; p=0.03). Likewise, the need for ECC reentry was more frequent in the cardiogenic shock patients (26.8%; p=0.05). Table 3 shows the distribution of the patients according to diagnostic group and the pre- and intraoperative variables.

Overall, the postoperative evolution of these patients was complicated, with important morbidity–mortality (Table 4).

We examined whether the three clinical scenarios of LCOS proposed in the diagnostic definition of the consensus document6 could imply different clinical courses and prognoses. No statistically significant differences were observed between the subgroups regarding the number of patients requiring repeat surgery (due to bleeding or tamponade) or the incidence of arrhythmias or infectious, neurological, respiratory or gastrointestinal complications. It should be mentioned that although statistical significance was not reached, trends were observed, with inter-group differences (LCOS, LCOS consistent clinical condition and cardiogenic shock) in the rates corresponding to perioperative AMI (p=0.48), acute lung edema (p=0.80) and acute renal impairment (AKI stage 3; p=0.21). Significant differences were recorded in the duration of mechanical ventilation (p<0.001), in the percentage of patients requiring renal replacement therapy (p=0.007), in the proportion of patients who developed MOF over time (p<0.001), and therefore also in percentage mortality (p=0.01). Of note is the fact that the patients who developed MOF (n=32) were more numerous in the cardiogenic shock group (19 versus 7 in the LCOS group and 6 in the clinical diagnostic group; p<0.001). Fig. 2 graphically displays the differences in postoperative complications among the three diagnostic groups.

Figure 2.

Different complications and evolution according to diagnostic subgroups. AKI 3: acute kidney injury 3; MOF: multiorgan failure; Periop. AMI: perioperative acute myocardial infarction; RRT: renal replacement therapy; MV time: duration of mechanical ventilation in minutes; LCOS: low cardiac output syndrome.

(0.21MB).

The analysis of the myocardial damage markers such as ultrasensitive troponin T and I likewise revealed no statistically significant differences. In contrast, the mean maximum lactate concentration (as a perfusion marker) (Fig. 3) and alanine aminotransferase levels were significantly higher in the patients with cardiogenic shock (51.2±49 versus 22.2±21 in group A and 33.7±36 in group B [p=0.002], referred to lactate, and 679.7±1040 versus 215±769 in group A and 167±358 in group B [p=0.008], referred to alanine aminotransferase). Lactate concentration reached statistical significance at all measurement timepoints (admission, p=0.001; 4h, p=0.014; 12h, p=0.012; 24h, p=0.01; 48h, p=0.003; discharge, p=0.01).

Figure 3.

Maximum lactate concentrations (mmol/l) according to diagnostic subgroups. LCOS: low cardiac output syndrome.

(0.05MB).

The analysis of the above data confirms the existence of increased patient severity, a greater percentage of complications, and a poorer prognosis among the patients with cardiogenic shock. It also must be noted that the evolution of the patients in the clinical diagnostic group was similar to that seen in the patients diagnosed through invasive monitoring.

Despite the many patients subjected to CS and the relative incidence of postsurgery LCOS, few studies involving large patient samples have characterized this clinical form of AHF. The published studies analyze concrete aspects, and none have examined the clinical course according to the different clinical diagnostic subgroups. Our prospective multicenter study has analyzed patients with LCOS, differentiating three subgroups with evolutive and prognostic differences.

The literature describes a range of pre- and intraoperative factors related to LCOS. The study of Rao et al.,1 involving patients in the postoperative period of myocardial revascularization surgery, identified LVEF <20%, repeat surgery, the female gender, diabetes, age >70 years, left coronary trunk stenosis, recent AMI and triple vessel disease as independent predictors of the syndrome. Algarni et al.,7 in patients subjected to revascularization, identified repeat revascularization, surgery in the previous year, depressed LVEF (<40%), emergency surgery, cardiogenic shock and the female gender as predictive factors. These authors reported a decrease in the prevalence of LCOS over time. Ding et al.8 analyzed 205 patients with LCOS and found advanced age, depressed LVEF, surgery with ECC, emergency surgery and incomplete revascularization to be the factors associated to LCOS. The authors also described increased morbidity–mortality in these individuals. Very similar predictive factors were documented by Sa et al.9 in 89 patients with LCOS (14.7% coming from revascularization surgery).

Low cardiac output has also been related to the myocardial protection measures adopted during the surgical procedure. In this regard, Algarni et al.10 found the prevalence of LCOS to be significantly lower (p<0.001) in the group of patients revascularized using microplegia (2.7%) than in the group in which standard blood cardioplegia was used (5%). In a study of patients subjected to revascularization in which comparisons were made of the incidence of LCOS using cold (7°C) blood cardioplegia versus warm (29°C) cardioplegy,11 the authors found the incidence of the syndrome to be greater in the latter group. Likosky et al.,12 in a group of patients subjected to elective revascularization with ECC, examined the degree to which the surgeon and surgical practices influence the incidence of LCOS. The authors found the low cardiac output rates to vary among surgeons, and these variations could not be explained only on the basis of the case mix. This suggested that variability in perioperative practices (surgical times, cardioplegia, etc.) influences the risk of LCOS.

In a group of patients subjected to isolated mitral valve surgery, Maganti et al.13 identified the urgency of the operation, surgery in the previous year, LVEF <40%, NYHA functional class IV, ischemic mitral valve disease and clamping time as predictive factors.

In general, the published reviews14 contemplate the abovementioned pre- and intraoperative predictors. In our study, the analysis of the factors associated to the development of LCOS yielded results consistent with those found in the literature. Of note in our series is the observed association to pre- and also intraoperative clinical severity factors.

Knowing the risk factors for LCOS implies the need to identify high risk patients for the adoption of possible preventive measures. In this regard, preconditioning strategies are already being adopted in patients with ventricular dysfunction. The first experiences with the preoperative administration of levosimendan in selected patients have been encouraging.15,16 Trials are underway that will provide definitive results on the use of this drug in patient preconditioning.17 Recently, the LEVO-CTS trial has shown that levosimendan, used on a presurgical prophylactic basis in patients with LVEF ≤35% subjected to CS, does not result in a decrease in the combined target of diminished mortality, need for renal replacement therapy, perioperative AMI or the need for mechanical assist measures.18 On the other hand, cardioplegia and the different myocardial protection measures adopted during CS can modify the possible development of ventricular myocardial involvement due to ischemia/reperfusion or secondary myocardial stunning, and therefore the incidence of LCOS.19–23 In this way, patients at greater risk of suffering LCOS could be amenable to guided studies, with the adoption of specific cardioprotection measures.

On the other hand, the LCOS consensus document of the SEMICYUC 20126 distinguishes three clinical groups of LCOS. We sought to determine whether these three clinical scenarios proposed in the diagnostic definition of the document are associated to differences in clinical outcome. Comparisons were made of the diagnostic groups: (A) LCOS: “classical” definition with direct measurement of the cardiac index; (B) “consistent clinical condition” not based on invasive monitoring data; and (C) cardiogenic shock. In addition to the classical definitions of postoperative LCOS and cardiogenic shock, this classification includes a consistent clinical condition group not contemplated in other definitions. These three subgroups were compared in order to determine possible differences in terms of clinical evolution, complications and postoperative mortality. The results are shown in Figs. 2 and 3. As has been commented under Results, differences between the subgroups were observed in the incidence of perioperative AMI, acute lung edema and acute renal functional impairment on the basis of percentage AKI stage 3 – though statistical significance was not reached, probably because of the limited number of patients involved. In contrast, statistically significant differences were recorded in the duration of mechanical ventilation, the need for renal replacement measures, the development of MOF and patient mortality. In this respect, the prognosis was clearly poorer among the patients with cardiogenic shock.

Two relevant findings were made in our study:

• 1.

  The need to recognize the importance in routine practice of the “consistent clinical condition” diagnostic group. The assessment of clinical data such as oliguria and hypoperfusion, and of laboratory test data available in all centers (even on a point of care basis), such as lactate concentration and central venous oxygen saturation, allows the detection of low cardiac output in these patients. In this respect, advanced invasive monitoring (using a pulmonary artery catheter or transpulmonary thermodilution) might not be strictly necessary in all patients in order to diagnose and treat LCOS.24 In our study this “clinical” group showed a postoperative course characterized by a clear increase in morbidity–mortality very similar to that observed in the group of patients diagnosed on the basis of hemodynamic monitoring. Consequently, in terms of practical clinical management, we always must observe a high level of suspicion in order to start early and adequate treatment in this subgroup of patients (similar to that of the subgroup diagnosed through invasive monitoring).

• 2.

  The clearly poorer clinical course and prognosis in the cardiogenic shock group versus the other two groups. Cardiogenic shock implies increased severity, with a poorer postoperative course, a greater incidence of MOF, and higher mortality. In patients that suffer shock, it must be considered that transesophageal echocardiography reveals right ventricular dysfunction in about 40% of the cases.25 This situation can imply changes in the management approach.26,27

In any case, LCOS is associated to important morbidity–mortality and postoperative complications (as reflected in the corresponding tables and figures). In our series, mortality was similar to that described in the literature, with significant differences depending on the mentioned diagnostic subgroups.

The clinical evolution of patients with LCOS is complex, with important morbidity–mortality.

It is important to know the pre- and intraoperative factors that can lead to LCOS, with the aim of defining possible preventive strategies.

In our series, differences were found in the postsurgery clinical evolution parameters and in mortality among the different LCOS diagnostic subgroups.

Larger patient samples are required in order to confirm the results obtained.

All the signing authors have made substantial contributions to study conception and design, data acquisition, and analysis and interpretation of the results. Likewise, they have participated in drafting of the manuscript or in critical review of its content, and have approved the final version of the article.

As applicable in the case of a working group, the rest of the participating or collaborating authors reflected in the table (ESBAGA study) and listed at the end of the document have participated in data acquisition.

Collaborators such as Dr. Ana Ochagavía-Calvo, Dr. Rafael Hinojosa-Pérez, Dr. Juan José Marín-Salazar, Dr. M. Victoria Boado-Varela and Dr. José Luis Romero-Luján contributed no patients, but participated in the conception and design of the study.

The authors declare that they have no conflicts of interest.