The optimal degree of core temperature for hypothermic circulatory arrest in complex aortic arch surgery: results from 1310 patients

Abstract

OBJECTIVES

The optimal core temperature for hypothermic circulatory arrest during aortic arch surgery remains contentious. This study aims to evaluate patient outcomes under various temperatures within a large single-centre cohort.

METHODS

Between 2010 and 2018, patients diagnosed with type A aortic dissection underwent total arch replacement at Fuwai Hospital were enrolled. They were categorized into 4 groups: deep hypothermia group, low-moderate hypothermia group, high-moderate hypothermia group and mild hypothermia group. Clinical data were analysed to ascertain differences between the groups.

RESULTS

A total of 1310 patients were included in this cohort. Operative mortality stood at 6.9% (90/1310), with a higher incidence observed in the deep hypothermia group [29 (12.9%); 35 (6.9%); 21 (4.8%); 5 (3.4%); all adjusted P < 0.05]. Overall 10-year survival was 80.3%. Long-term outcomes did not significantly differ among the groups. Multivariable logistic analysis revealed a protective effect of higher core temperature on operative mortality (odds ratio 0.848, 95% confidence interval 0.766–0.939; P = 0.001). High-moderate hypothermia emerged as an independent protective factor for operative mortality (odds ratio 0.303, 95% confidence interval 0.126–0.727; P = 0.007). Multivariable Cox analysis did not detect an effect of hypothermic circulatory arrest on long-term survival (all P > 0.05).

CONCLUSIONS

High-moderate hypothermia (24.1–28°C) offers the most effective protection against surgical mortality and is therefore recommended. Different hypothermic circulatory arrest temperatures do not influence long-term survival or quality of life.

Open in new tabDownload slide

INTRODUCTION

Aortic arch surgery stands as one of the most intricate and demanding procedures in modern aortic surgery, marked by elevated mortality and complication rates [1, 2]. Utilizing hypothermic circulatory arrest (HCA) enhances the brain’s resilience to ischaemia, while ensuring a bloodless surgical field, thus facilitating distal arch anastomosis. Once deemed indispensable in open arch surgery [3], the integration of antegrade cerebral perfusion (ACP) has significantly diminished operative mortality and brain injury in aortic arch surgery [4–6]. Despite numerous studies affirming the superiority of HCA combined with ACP, the pursuit of lower target temperatures prolongs cooling and rewarming times, consequently escalating cardiopulmonary bypass (CPB) duration. This prolonged CPB time heightens the risk of severe coagulation disorders, inflammation and multiorgan dysfunction [7]. Consensus regarding the optimal core temperature range for arch surgery remains elusive. Moreover, comprehensive data elucidating the long-term outcomes of patients subjected to different temperature regimes are yet to be established. To bridge this gap, we conducted a 13-year follow-up study on patients diagnosed with type A aortic dissection (TAAD) who underwent total arch replacement (TAR) at our institution, aiming to explore patient prognosis under varied HCA strategies.

METHODS

This study received approval from the Ethics Committee of Fuwai Hospital (No. 2022–1745), with written informed consent waived due to its retrospective nature.

Study population and data collection

Between January 2010 and December 2018, TAAD patients who underwent surgical intervention at Fuwai Hospital were included, while those who underwent ascending aorta replacement, hemiarch replacement and hybrid aortic arch repair were excluded from the final cohort. Following the 2023 consensus on hypothermia classification in aortic arch surgery [2], core temperature was categorized into 4 groups: deep hypothermia (DH) (≤20°C), low-moderate hypothermia (LMH) (20.1–24°C), high-moderate hypothermia (HMH) (24.1–28°C) and mild hypothermia (MH) (>28°C). Preoperative baseline data, intraoperative conditions and postoperative outcomes were retrieved from the hospital’s electronic medical records. Long-term results were collected using standardized forms during follow-up clinic visits or via telephone/WeChat follow-ups. The follow-up period concluded in December 2023, with outcomes obtained for all patients surviving the early postoperative phase. Postoperative activities of daily living (ADL) among surviving patients were stratified into 4 levels based on feedback: Level 1 (complete loss of self-care ability), Level 2 (partial self-care requiring assistance), Level 3 (complete self-care but incapacity for general physical work) and Level 4 (engagement in general physical work).

Surgical techniques

Between 2010 and 2018, right axillary artery cannulation and unilateral ACP were routinely combined with HCA as the preferred circulatory management approach for all emergency and elective cases necessitating aortic arch reconstruction at our institution. Median thoracotomy and CPB were universally performed on all patients. Aortic root surgery was conducted during the cooling phase following CPB and tailored according to the extent of injury. Upon reaching the target core temperature, cerebral perfusion commenced, maintaining a flow rate of 8–12 ml/kg/min. Detailed TAR procedures at our centre have been previously described [8]. In brief, upon the completion of distal anastomosis, lower body perfusion was restored through the perfusion limb of the tetrafurcate graft. Sequential reconstruction of the left common carotid artery, left subclavian artery and innominate artery was performed, followed by the initiation of rewarming. Subsequently, the proximal end of the artificial vessel was anastomosed end-to-end with the ascending aorta or repaired aortic root.

Endpoints and definitions

Primary outcomes encompassed operative mortality, long-term mortality, reoperation and the ADL among surviving patients. Operative mortality referred to death within 30-day post-surgery or in-hospital mortality according to the Society of Thoracic Surgeons criteria [9]. Long-term mortality encompassed death from any cause during follow-up. Reoperation encompassed all cardiovascular-related repeat procedures, including but not limited to cardiovascular surgery, pacemaker implantation, radiofrequency ablation and percutaneous coronary intervention. Emergency surgery was defined as surgery performed within 24 hours of admission. Aortic dissection was defined and classified based on the type, entry location, and malperfusion status (TEM) classification [10].

Statistical analysis

Continuous data normality was assessed using the Kolmogorov–Smirnov test, with normally distributed data presented as mean and standard deviation, and non-normal data as median with interquartile range. Analysis of variance and Kruskal–Wallis tests were utilized for normally and non-normally distributed variables, respectively. Categorical variables were reported as counts and percentages compared using chi-squared or Fisher’s exact tests. Multiple comparisons between different groups were further conducted, and P values were adjusted with Bonferroni’s modification. The restricted mean survival time was compared among 4 groups to estimate the long-term survival probability. The cumulative incidence of long-term reoperation and incomplete self-care (ADL ≤ 2) was analysed using the Gray-Fine regression model, with death considered a competing risk.

Restricted cubic spline analysis was conducted to flexibly model and visualize the relationship between core temperature and operative mortality. Collinearity diagnostics were performed for all baseline and operative variables, which showed no significant findings based on a tolerance >0.1 or a variance inflation factor <10. A total of 47 variables for baseline characteristics and intraoperative conditions were then incorporated into univariable logistic and Cox regression analyses, with core temperature and other variables having a P value <0.1 being further included in multivariable logistic and Cox regression analyses through backward stepwise selection methods to identify independent risk factors for operative mortality and long-term mortality, respectively. The proportional hazards assumption was assessed using weighted Schoenfeld residuals, with no violations observed (P > 0.05). Statistical analysis and data visualization were carried out using R 4.2.1 and Microsoft Excel v2402. A two-tailed P value <0.05 indicated statistical significance.

RESULTS

Clinical characteristics

As depicted in Fig. 1, this cohort ultimately comprises 1310 TAAD patients who underwent TAR. Table 1 summarizes the baseline characteristics and intraoperative conditions of the 1310 patients, with stratification as follows: 224 (17.1%) in DH, 505 (38.5%) in LMH, 433 (33.1%) in HMH and 148 (11.3%) in MH. The mean patient age was 46.67 (10.34) years, with no significant difference among groups (P = 0.217). Male patients were most prevalent in the MH group (88.5%; all adjusted P < 0.05). Preoperative imaging data indicated that the percentage of patients with moderate or severe aortic insufficiency was 88 (39.3%), 215 (42.6%), 165 (38.1%) and 44 (29.7%) for each group (P = 0.041). Patients in the DH and LMH groups were more frequently diagnosed with an entry in the ascending aorta (TEM-E1: 76.3% and 70.5%, respectively, all adjusted P < 0.05) and demonstrated a higher incidence of supra-aortic malperfusion (TEM-M2: 76.3% and 68.1%, respectively, all adjusted P < 0.05). Conversely, entry located in the aortic arch (TEM-E2: 42.0% and 45.9%, respectively) and descending aorta (TEM-E3: 6.7% and 12.2%, respectively) was more common in the HMH and MH groups (all adjusted P < 0.05), with malperfusion affecting the spinal, visceral, renal or iliac regions to a greater extent when compared with DH group (TEM-M3: 47.1% and 50.0%, respectively, all adjusted P < 0.05). As depicted in Supplementary Material, Fig. S1, a total of 12 surgeons performed TAR during the period, with our centre conducting more than 100 TAR procedures for TAAD annually since 2012. The trend of core temperature during HCA is illustrated in Supplementary Material, Fig. S2, demonstrating a gradual rise in core temperature over time. Operative time, CPB time and HCA time all significantly decreased from the DH group to the MH group (all adjusted P < 0.05), accompanied by a reduction in intraoperative blood product use (Table 1).

Operative outcomes

Postoperative details are summarized in Table 2. According to core temperature grouping, a higher operative mortality was observed in the DH group [29 (12.9%); 35 (6.9%); 21 (4.8%); 5 (3.4%); all adjusted P < 0.05] (Supplementary Material, Fig. S3). Additionally, the DH group exhibited significantly higher incidences of tracheotomy [15 (6.7%)], acute kidney injury [88 (39.3%)], and hepatic insufficiency [75 (33.5%)], along with a relatively prolonged hospital stay [14.00 (11.00, 19.00)] when compared with HMH or MH groups (all adjusted P < 0.05).

Long-term outcomes

There were no losses during the follow-up period, with a median time of 6.04 [4.22, 8.28] years. The median follow-up time of the DH group was 8.05 [3.94, 10.05] years, the LMH group was 7.46 [4.68, 9.57] years, the HMH group was 5.28 [4.08, 6.60] years, and the MH group was 4.79 [4.33, 5.62] years. The overall 5- and 10-year survival rates were estimated to be 86.6% and 80.3%, respectively. As shown in Fig. 2A, there was a significant difference in long-term survival rates among groups (P < 0.001). The specific distribution of ADL levels at the end of follow-up is depicted in Supplementary Material, Fig. S4.

Effect of core temperature on operative outcomes

A total of 22 variables were included in the final model following univariable logistic analysis, as summarized in Supplementary Material, Table S1. Multivariable logistic analysis unveiled a protective effect of higher core temperature on operative mortality [odds ratio (OR) 0.848, 95% confidence interval (CI) 0.766–0.939; P = 0.001] (Table 3). Other independent risk factors for operative mortality are detailed in Supplementary Material, Table S3. As depicted in Fig. 3, the restricted cubic spline model depicted no nonlinear relationship between core temperature and operative mortality (P = 0.569). However, the overall effect of core temperature was confirmed (P < 0.001). DH appeared to be a risk factor for operative mortality [OR (95% CI) >1, on the left side of the thin dashed line (19.996°C)], corroborated by multivariable logistic analysis when core temperature was dichotomized into below and above 20°C (OR 0.520, 95% CI 0.281–0.977; P = 0.039) (Table 3). Finally, additional logistic regression was conducted to further explore the effect of core temperature, treated as a quadripartite variable, on operative mortality. Results indicated that HMH was an independent protective factor for operative mortality (OR 0.303, 95% CI 0.126–0.727; P = 0.007) compared with DH, while LMH and MH showed no significant effects on operative mortality (P = 0.155 and P = 0.099, respectively), as detailed in Table 3.

Effect of core temperature on long-term outcomes

As detailed in Supplementary Material, Table S2, univariable Cox analysis identified 26 variables showing potential effects (P < 0.1) on long-term survival. As illustrated in Fig. 2B, multivariable Cox analysis did not reveal an effect of LMH, HMH and MH on long-term survival compared with DH (all P > 0.05). In addition, competing risk analyses showed comparable long-term reoperation (P = 0.935) and incomplete self-care (P = 0.055) rates among groups (Fig. 2C and D).

DISCUSSION

In this study, we investigated the prognosis of HCA combined with ACP for TAR treatment of TAAD patients using different core temperature strategies in a large single-centre cohort. We observed that deep HCA increased surgical risk, while higher-moderate HCA provided an optimal core temperature range, with no significant impact on long-term outcomes.

Higher core temperature emerged as an independent protective factor for operative mortality. HCA offers surgeons a safe window and a relatively bloodless surgical field for arch reconstruction procedures. Griepp et al. [11] pioneered HCA technology in 1975 for aortic arch surgery, capitalizing on systemic hypothermia <20°C to significantly reduce metabolism, enabling transient circulatory arrest for surgery. Subsequent clinical and experimental studies, including the incorporation of ACP by Bachet and Kazui [12, 13], have continually improved HCA outcomes, notably reducing neurological complications. However, consensus on the optimal temperature range has remained elusive [4, 14, 15]. While some advocate for deep HCA as the gold standard for end-organ protection during circulatory arrest in aortic arch repair [14], others suggest that moderate HCA combined with ACP is equally effective, rendering deep HCA unnecessary [15]. Lower temperatures theoretically offer better brain protection by reducing the metabolic rate, prolonging the ischaemic tolerance period [16]. Despite the protective effect of ACP (4–6), our study found that for every degree increase in temperature, the risk of death decreased by approximately 18%. By categorizing patients into 2 groups, multivariable logistic regression analysis revealed that moderate-to-mild HCA independently reduced surgical risk, providing clarity amidst previous controversies.

High-moderate HCA emerged as the most effective safeguard against surgical mortality. Prolonged HCA duration is associated with adverse outcomes, with studies indicating increased early mortality, stroke rates, and renal function damage beyond 40 or 50 min [17]. Our median HCA time of 20 min falls within the recommended safe range of 30 min [18], lower than Western studies [2]. Mirroring trends observed elsewhere [19], core HCA temperature at our centre has gradually increased over time (Supplementary Material, Fig. S2). To further assess surgical outcomes across different temperature groups, patients with temperatures above 20° were divided into LMH, HMH and MH groups. Remarkably, the protective effect of higher core temperature, as previously concluded, was most pronounced in the HMH group, with a 3.3-fold reduction in death risk compared with the DH group. These findings bolster the IIa B recommendation in The European Association for Cardio Thoracic Surgery and The Society of Thoracic Surgeons 2023 Guidelines [2], advocating for high-moderate HCA combined with ACP for complex aortic arch procedures. Notably, operation complexity, patient age and medical condition, and surgeon preference influence the ideal HCA level. Hence, we mitigated bias by including patients undergoing the same arch surgery method and disease type in this cohort and conducting various statistical methods on preoperative data. Notably, severe complication indicators such as operative death and continuous renal replacement therapy were significantly higher in the DH group, aligning with results in Table 2. In recent years, several institutions have reported cases of aortic arch replacement under MH conditions and achieved relatively favourable early outcomes [4]. However, the protective effect of MH in our cohort was not evident after conducting multivariable logistic regression analysis. Although the number of cases in the MH group (n = 148) is substantial for TAAD, it remains comparatively smaller than other groups. Given that our cohort exhibited generally satisfactory clinical outcomes, it is plausible that the relatively small sample size may have hindered capturing the statistical significance of the MH group.

Higher temperatures did not compromise intraoperative organ protection. Concerns regarding organ ischaemia during HCA, such as end-organ and spinal cord ischaemia, have been prominent in scholarly discourse [19, 20]. Clinically, the kidney is particularly sensitive to ischaemia, with studies indicating a safe ischaemia duration of up to 60 min even at 30°C [21]. Nevertheless, in our study, despite the overall short HCA time, significant differences persisted in the distribution of continuous renal replacement therapy and acute kidney injury among groups. The heightened incidence of renal injury in the DH group may be attributed to prolonged operation and CPB times, as well as increased blood transfusion volumes. Progressive temperature elevation notably shortens CPB cooling and rewarming periods, potentially mitigating the systemic inflammatory response, all recognized factors contributing to postoperative renal damage. Safe limits for spinal cord ischaemia tolerance during HCA at higher temperatures remain unresolved. Zierer and colleagues [4] reported a paraplegia rate of less than 1% among 1002 patients undergoing arch surgery with ACP and mild HCA. Similarly, although our cohort exclusively comprised TAAD patients, the overall paraplegia rate was only 1%. Notably, there were no significant differences in spinal cord injury rates among the 4 groups (all P > 0.05).

Different HCA temperatures did not influence long-term outcomes. Previously, concerns have been raised that raising HCA temperature may yield unforeseen consequences for long-term prognosis and quality of life [22]. Moreover, TAR patients necessitate vigilant long-term monitoring, particularly in detecting complications and addressing untreated aortic segments, notably in TAAD patients. Yet, large-scale reports on long-term outcomes across various HCA temperatures have been scarce. In our study, comprehensive follow-up data spanning over 13 years were obtained for all patients. Notably, higher temperatures do not compromise the long-term survival of patients. Furthermore, it is noteworthy that in the HMH and MH groups, no patients experienced complete loss of self-care ability. Across all HCA groups, approximately 90% of surviving patients were capable of complete self-care and engaged in general physical work, underscoring the efficacy of TAR in treating TAAD involving the aortic arch.

Limitations

The absence of randomization may have influenced the findings, given the retrospective nature of this study. For instance, within this cohort, our centre’s increasing experience over time led to a reduction in HCA duration, potentially introducing bias into the results. Nonetheless, it stands as one of the largest single-centre studies with the longest follow-up to date. Second, we employed various statistical methods to mitigate bias resulting from baseline data. Furthermore, although a randomized controlled trial may offer superior clinical design, its ethical feasibility warrants careful consideration, especially in light of evidence against deep HCA.

CONCLUSION

In summary, our findings indicate that higher core temperature serves as a protective factor for operative mortality, challenging the notion that deep HCA provides superior surgical protection. HMH emerges as the optimal choice for safeguarding against surgical mortality and should be recommended for complex arch surgery. Importantly, different HCA temperatures do not impact long-term survival or quality of life.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

This study was supported by the National High Level Hospital Clinical Research Funding (2022-GSP-GG-24; 2023-GSP-GG-17) and the Beijing Municipal Natural Science Foundation (No. 7224341).

Conflict of interest: none declared.

DATA AVAILABILITY

The data underlying this paper cannot be shared publicly because of relevant data protection regulations. However, the data will be shared on reasonable request to the corresponding author with permission from the ethics committee.

Author contributions

Kai Zhang: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Resources; Visualization; Writing – original draft; Writing—review & editing. Chenyu Zhou: Conceptualization; Data curation; Formal analysis; Investigation; Resources; Visualization; Writing—original draft; Writing—review & editing. Shiqi Gao: Software; Visualization; Writing—original draft; Writing—review & editing. Yumeng Ji: Data curation; Investigation. Jiawei Qiu: Data curation; Resources. Fangfang Cao: Data curation; Resources. Juntao Qiu: Conceptualization; Funding acquisition; Project administration; Resources; Supervision; Validation; Writing—review & editing. Cuntao Yu: Conceptualization; Funding acquisition; Methodology; Project administration; Resources; Supervision; Validation; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Thomas Schachner, Shekar L.C. Reddy, David Spielvogel and the other anonymous reviewers for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• ACP

  Antegrade cerebral perfusion

 
• ADL

  Activities of daily living

 
• DH

  Deep hypothermia

 
• HCA

  Hypothermic circulatory arrest

 
• HMH

  High-moderate hypothermia

 
• LMH

  low-moderate hypothermia

 
• MH

  Mild hypothermia

 
• TAAD

  Type A aortic dissection

 
• TAR

  Total arch replacement

 
• TEM

  The Type, Entry location, and Malperfusion

 
• OR

  Odds ratio

Author notes