The Impact of Hyponatremia in Aortic Valve Surgery Using Histidine–Tryptophan–Ketoglutarate Cardioplegia

Histidine–tryptophan–ketoglutarate (HTK) or Bretschneider cardioplegia is one of the most widely used methods of cardioplegia owing to its capacity to provide prolonged cardioprotection with a single administration. This solution is hyponatremic compared to blood (15 mmol/l vs. 135–145 mmol/l), and the recommended administration dose is 20 ml/kg within 6–8 min. Consequently, the rapid infusion of a substantial volume of hyponatremic solution will result in a notable reduction in serum natrium (Na) levels. The management of cardiac surgical cases presents a significant challenge in the context of electrolyte imbalances. Serum Na is one of the main electrolytes of concern because of the significant adverse effects associated with its imbalance. It is crucial to consider the timing of hyponatremia and the rapidity of its onset. In severe cases, rapidly induced hyponatremia will precipitate cellular swelling, with significant ramifications, especially for the central nervous system (e.g., cerebral edema, demyelination syndrome) and an increased risk of postoperative mortality.^1,2 Hyponatremia that has developed insidiously is often better tolerated, and it can also be an indicator of heart failure. Hyponatremia after cardiac surgery is typically the result of hemodilution caused by cardioplegia and priming of the cardiopulmonary bypass circuit. Conversely, hypernatremia is less prevalent. In the preoperative period, it is often indicative of renal or hepatic pathology, whereas in the postoperative period, it may result from the administration of hypertonic solutions and renal-free water loss due to diuretic use.

The cardiac biochemical processes involved in cardioplegia have been intensively studied over time, yet the systemic effects of these agents have been largely overlooked.³ There is currently no consensus on the management of hyponatremia secondary to cardiac surgery. In particular, the tolerance limits of serum Na concentration remain unknown, as well as what the potential risk-to-benefit ratio is for correcting hyponatremia in its mild-to-moderate forms.

The aim of this retrospective study was to analyze changes in serum Na concentration following Bretschneider cardioplegia infusion in aortic valve surgery, and to investigate the association between hyponatremia and postoperative outcomes in a group of patients who underwent aortic valve replacement (AVR) between June 2023 and June 2024 at the Emergency Institute for Cardiovascular Diseases and Transplantation of Tîrgu Mureș, Romania. Additionally, we seek to determine whether perioperative hyponatremia necessitates urgent corrective action.

The study included adult patients who had undergone elective surgery for AVR (regardless of prosthesis type), with valvular stenosis or insufficiency, whether congenital or acquired. The exclusion criteria were as follows: concomitant procedures, age under 18 years, retrograde or combined cardioplegia, and emergency surgery. Demographic data included age, sex, and comorbidities. The patients were divided into two groups according to their preoperative serum Na levels: the first group comprised patients with low Na levels (<136 mEq/l), and the second consisted of subjects with normal or elevated Na levels (≥136 mEq/l).

Data collection involved the sampling of blood from all patients at two key time points: before anesthetic induction (T1) and at the time of admission to the intensive care unit (ICU) (T2). Left ventricular ejection fraction (LVEF) and transvalvular gradient were calculated and recorded using transthoracic echocardiography 1 day before and on day 7 after surgery. The echocardiographic data included in the study were documented by a single cardiologist.

Outcomes measured were operative short-term mortality, as well as in-hospital postoperative complications such as neurological impairment, surgical reintervention, new onset atrial fibrillation (AFib), and sustained ventricular fibrillation (VF) or ventricular tachycardia (VT). Postoperative evolution markers, such as duration of mechanical ventilation and need for inotropic support, echocardiographic measurements (LVEF, transvalvular gradient) and blood tests (pH, lactate) were compared between the two groups.

The cardioplegia solution, developed by Bretschneider, was employed as a crystalloid substitute with antegrade administration at a temperature of 4 °C, as previously described by Bretschneider et al., with re-administration at 90 min during aortic cross-clamping. The cardioplegia solution was infused in accordance with the manufacturer’s recommendations.

From our total cohort of 302 patients, 36.3% (110 out of 302) were female, the majority of them being in the age group of 61–70 years (46.0%; 139 out of 302 patients). Regarding comorbidities, 119 patients (39.3%) were diagnosed with dyslipidemia, 60 patients (19.8%) with obesity, 69 (22.8%) with diabetes mellitus, and 218 (71.9%) with arterial hypertension. The mean preoperative LVEF was 52.1 ± 8.0%, the mean diameter of the left ventricle (LV) was 53.8 ± 8.5 mm, the maximum aortic gradient was 82.0 ± 25.9 mmHg, and the medium gradient was 50.1 ± 17.3 mmHg.

The mean Na concentration was 138.0 ± 4.1 mmol/l before the surgery and 135.1 ± 6.6 mmol/l after surgery. Before surgery, the mean Na concentration was 134.0 ± 1.3 mmol/l in the low Na group and 140.5 ± 3.2 mmol/l in the normal/high Na group. After surgery, the mean Na concentration decreased to 134.5 ± 3.4 mmol/l in the low Na group and to 135.5 ± 8.1 mmol/l in the normal/high Na group. More patients were from the normal/high Na group (29.5% vs. 26.1%; p = 0.002) had a lower preoperative pH (p = 0.022) and lactate concentration (p = 0.015) (Table 1).

The mean duration of the AVR procedure was significantly shorter in the low Na group (337.5 min vs. 402.1 min; p < 0.001), with a longer mean duration of bypass (122.5 min vs. 93.8 min; p < 0.001), aortic clamp (81.3 min vs. 66.7 min; p < 0.001), orotracheal intubation (13.7 h vs. 9.9 h; p < 0.001), and inotropic/vasopressor support administration (1.8 days vs. 1.6 days; p = 0.018). Mean ultrafiltrate during cardiopulmonary bypass was 1,862 ml. No significant differences were seen in terms of the other patient characteristics (Table 1).

Logistic regression analysis of the outcomes showed a significant association between low preoperative Na concentrations and surgical reintervention for hemorrhagic cause, being a risk factor (OR = 3.65; 95%CI 1.18–11.34; p = 0.025). No statistically significant association was found between Na levels and new onset of AFib, sustained VT/VF, and postoperative neurological impairment (Table 2).

The 7-day mortality was 7.6% (9 out of 119 patients) in the low Na group compared with 1.1% (2 out of 183 patients) in the normal/high Na group, and 30-day mortality was 10.9% (13 out of 119 patients) in the low Na group vs. 1.6% (3 out of 183 patients) in the normal/high Na group. We found a significant association between low preoperative Na levels and 7-day mortality (OR = 7.40; 95% CI 1.57–34.90; p = 0.011), as well as low preoperative Na levels and 30-day mortality (OR =7.36; 95% CI 2.05–26.42; p = 0.002) (Table 2).

The results of the ROC analysis and maximum Youden’s index suggest that the optimal cut-off value for preoperative Na levels was ≤136 mmol/l for the prediction of 7-day postoperative mortality. This threshold had an area under the curve (AUC) of 0.672 (95% CI 0.615–0.724; p = 0.005), with 81.8% sensitivity and 62.2% specificity (Figure 1).

FIGURE 1.

ROC curve analysis of preoperative Na levels in predicting 7-day postoperative mortality

The same cut-off level (Na ≤ 136 mmol/l) was suggested by the Youden’s index in the prediction of 30-day postoperative mortality, with an AUC of 0.668 (95% CI 0.612–0.721; p = 0.001), with 81.2% sensitivity and 62.9% specificity (Figure 2).

FIGURE 2.

ROC curve analysis of preoperative Na levels in predicting 30-day postoperative mortality

The findings of our study indicate that preoperative hyponatremia has a significant impact on short-term mortality rate. Many patients with aortic valve disease have congestive heart failure with low LVEF, renal failure, and other comorbidities, sometimes requiring complex valve surgery. A common factor in all of these conditions is most often hyponatremia, which is present in 15–30% of patients presenting to hospital, irrespective of the underlying pathology.⁴ The prevalence of hyponatremia in patients with heart disease is estimated to be between 1% and 24%.⁵ Patients with low serum Na levels exhibit increased catecholamine and vasopressin levels, as well as heightened reactivity of the renin–angiotensin–aldosterone system. It is proven that these biochemical alterations result in impaired renal and hepatic microcirculation. Although myocardial injury is a potential consequence, it has not been conclusively demonstrated in human studies. However, an increased susceptibility to the adverse effects of the ischemia–reperfusion phenomenon in conditions of hyponatremia has been demonstrated in isolated rat hearts.^6,7 These laboratory results are consistent with the results of our study. Furthermore, the patients with preoperative hyponatremia in our study experienced prolonged aortic cross-clamp and cardiopulmonary bypass times in comparison to those with normal serum Na levels. This suggests that hyponatremia is predominantly identified in patients with advanced cardiac lesions requiring extensive surgical correction. Despite these observations, it remains unclear whether hyponatremia per se has a negative effect on the myocardium and, implicitly, on postoperative outcome, or it is a marker of organ dysfunction and comorbidities. In clinical practice, hydro-electrolytic and acid–base dysfunctions are corrected preoperatively, if the patient’s condition allows. Furthermore, there are numerous cases in which these corrections cannot be performed without addressing the underlying cardiac lesion.⁸ Accordingly, in the course of our investigation, 39.4% of patients were subjected to surgical intervention while exhibiting a preoperative hyponatremic status. Other studies have found results that are consistent with our findings. It has been stipulated that even a small decrease in pre-operative natrium levels is correlated with an increase in mortality, a correlation that increases proportional to the severity of hyponatremia. Research has demonstrated that even values within the normal range (135–138 mEq/l) are linked to an increased risk of mortality following cardiac surgery.^9,10

Furthermore, we demonstrate that there is no statistical correlation between decreased postoperative Na concentration and either mortality or adverse events. Although there was an average decrease of 2.9 mmol/l compared to preoperative values, this difference had no effect on the postoperative outcome. Other studies that have reported an increased incidence of postoperative adverse events due to hyponatremia have reported much higher mean postoperative Na decreases of 4–15 mmol/l.^11,12 These discrepancies may be attributed to the use of ultrafiltration in the configuration of each cardiopulmonary bypass circuit at our institution, which inhibits dilution and consequently prevents hyponatremia and maintains blood osmolarity.¹³ Other studies with increased postoperative hyponatremia do not specify the use of ultrafiltration, or its use has been confined to a relatively limited number of patients.¹¹ The calculated excess fluid, predominantly derived from cardioplegia (1,600–2,000 ml) and circuit priming (1,100–1,300 ml), is effectively eliminated. The mean volume filtered in our cohort was 1,862 ml, and no further corrective measures were necessary to address hyponatremia, as a slight reduction in Na levels was noted postoperatively. A review of the literature revealed a universal consensus among studies that assessed postoperative osmolarity. Specifically, the findings indicated that osmolarity remains within normal limits when HTK cardioplegia and ultrafiltration are used. As a representative study, in a cohort of 60 patients undergoing cardiac surgery with cardiopulmonary bypass, Yu-Ning Hu et al. showed that there were no postoperative seizures or symptoms of demyelination, as evidenced by quantitative electroencephalography and Mini-Mental State Examination scores.¹⁴

The osmolarity of Bretschneider cardioplegia is slightly higher than that of blood (310 mOsm/kg vs. 275–300 mOsm/kg).^12,15 It should be noted that our study did not record osmolarity values, because the determination of osmolarity is not a standard procedure in our center. It has been demonstrated that the correction of hyponatremia in an isotonic state can have significant adverse effects, as a hyperosmolar state can readily develop with deleterious effects.¹⁶ Furthermore, demyelination syndrome can occur in case of rapid correction of Na.¹⁷

In conclusion, the findings of our study indicate that the occurrence of complications is primarily associated with preoperative rather than postoperative hyponatremia, even when there are minor deviations from the normal range.

It is possible to implement safe and effective methods to prevent intraoperative hemodilution secondary to Bretschneider cardioplegia. These methods are preferred over the correction of iso-osmolar hyponatremia. It is not advised to correct mild forms of hyponatremia, as the treatment itself, rather than the hyponatremia, is associated with an increased mortality rate and a higher incidence of adverse events.

The existing literature contains a paucity of studies that have evaluated the relationship between HTK, hyponatremia, and the incidence of adverse events in adult cardiac surgery. Consequently, the optimal strategy for the management of HTK-related hyponatremia remains a topic of debate. It is imperative that future prospective studies use standardized intraoperative electrolyte monitoring.