DISCUSSION
The major finding of this study is that DHCA+DR has the same prevalence of operative mortality, NEs (any or permanent) and association of death+PND than MHCA+ACP. On the other side, lower body protection, evaluated by need of RRT, is significantly better with DHCA+DR. Furthermore, the composite endpoint (death, PND and RRT need) is significantly lower in patients where arch surgery has been performed with DHCA+DR. CPB time and CA time were similar in both groups.
NEs prevalence during aortic arch surgery remains not negligeable and is highly variable in different studies. A report from STS5 analyzed the cerebral outcome in chronic aortic arch surgery (CA time 25.8 min, T 21.1°C) according to any CP or no CP. PND rate was 6.3%, 7.6% without CP, 6.2% with ACP and 4.6% with RCP (p<0.001), favoring any CP. The relationship between temperature and stroke rate has also been widely studied. Urbanski et al6 in 1000 patients who underwent chronic aortic arch surgery (T 31.1°C, rectal, CA time 18.4 min, all UCP) reported a PND rate of 1% and TND rate of 4.9%. A similar experience was reported by Jabagi et al7: in 66 patients with hemiarch replacement (86% elective, T 32°C, CA time 17 min) PND rate was 2% and TND rate 3%. Damberg et al8 in 613 patients (86.3% elective, CA time 29.7 min, T 18-20 °C) reported a stroke rate of 2% without any CP.
NEs are generally perceived to be linked to the strategy of cerebral protection applied and ACP is considered the safest technique of cerebral protection, but both concepts are not completely true. The great majority of stroke is due to embolism9-11, which arises from atherosclerosis or thrombi at sites of ascending aorta, aortic arch, aortic arch branches, aortic clamping, vascular anastomosis, and aortic or CP cannulation10. ACP allows continuous CP throughout MHCA; however, the manipulation of arch branch vessels, such as dissecting and clamping the arch branch vessels or directly cannulating the ostia of the arch branch vessels, could dislodge debris from the vessels or introduce air into the cerebral circulation, causing cerebral embolism.
DHCA+DR combines different benefits. The epiaortic vessels are not dissected and/or clamped, the aortic arch is not manipulated and no cannula is inserted blindly inside the cerebral vessels. The brain protection is enhanced by DR, that can be defined a cerebral protection after CA. The rationale can be summarized as follows. Glucose is the fuel of the brain and comes from the blood by crossing the blood–brain barrier (BBB). Even if neurons have the capacity to take up glucose directly, due to the specific anatomy of the BBB (astrocytes are in direct contact with BBB, while neurons are not so close) a good part of the glucose that enters the brain does so through astrocytes12,13 (Fig. 1), from where it can be shuttled to neurons in the form of lactate that, after being converted into pyruvate, enters the Krebs cycle. Oligodendrocytes can transport lactates14, produced by themselves or by astrocytes, or glucose, if necessary, to neurons through the myelin (Fig. 1). Astrocytes have the possibility as well to store glucose as glycogen, a fuel reserve for the brain metabolism when glucose uptake is reduced or in stress condition.
The metabolic interdependence among neurons, astrocytes and oligodendrocytes allows the neurons to be supported by different energetic sources. This interdependence becomes more important with respect to metabolism of glutamate, the main neurotransmitter in the adult central nervous system, released into the synaptic cleft in a process called exocytosis. Glutamate, after being released, is taken up by surrounding astrocytes, and, after being converted to glutamine, is recycled to neuronal terminals, where it is converted again into glutamate, to replenish the glutamate pool15-17.
When the glutamate homeostatic balance is disrupted and levels become elevated in the extracellular fluid18, the excess of glutamate leads to influx of Na+ and Cl– into the postsynaptic cell, causing intracellular hyperosmolarity, and influx of water into the cell, contributing to intracellular oedema and neuronal death (exitotoxicity). Furthermore, glutamate stimulates glutaminergic receptors, but their excessive activation leads to the opening of Ca++ channels. Efflux of Ca++ into neurons, which activates plasmatic proteolytic enzymes, results in neuronal death via apoptosis or necrosis19-21.
After circulatory arrest, astrocytes being their glycogen reserve exhausted, cannot send lactate to neurons and cannot maintain low the glutamate level in the extracellular fluid. When the circulation starts again, but the temperature is still low (delayed rewarming), astrocytes use the glucose to rebuild their glycogen stores during a period of time when the metabolism of the neurons is not stimulated by a higher temperature. They start immediately to remove glutamate from the extracellular fluid, eliminating the danger of neuronal death, and to send lactate to neurons to face their metabolic needs. When rewarming starts, metabolic reserves have been restored and the glutamate in the extracellular fluid removed. In general, all organs can benefit from a period of cold perfusion before rewarming, to re-establish their nutrients store.
The progressive increase in the temperature of lower body circulatory arrest reduced the possibility of organs protection, that, in our study, had as target the kidney. Acute kidney injury (AKI) by the KDIGO criteria3, is common after cardiac surgery (81.2% of the patients: score 1 20.5%, score 2 48.7%, score 3 12%)22. In the literature RRT prevalence after CA is variable. Vekstein et al23 did not find any difference in RRT between DHCA and MHCA. It is likely that RRT need could be related more to CA time rather than to temperature, as it ranges from 7.3%24 with a CA time of about 23 min to 18%25 when the mean CA time was 50±27 min. In our experience any AKI occurred in 38.6% of the patients, 17% in DHCA+DR group and 47% in MHCA+ACP group (p<0.001). Deep hypothermia seems to be protective for renal function, especially for RRT need, 0 in DHCA+DR group versus 15.2% in MHCA+ACP, p=0.001. In this latter group there was a correlation between CA time and CPB time and RRT need (cut points 29 min and 127 min, respectively), suggesting that, when temperature is higher, longer CA and CPB can affect lower body protection.
Primary endpoints, after weighted logistic regression, failed to show any significant difference between groups, with the exception of RRT need and composite endpoint, both lower in DHCA+DR group. Secondary endpoints were all in favor of DHCA+DR group, in particular time to extubation and ICU time. However, it is possible that these were related to different anesthetic and surgical protocols.
The main limitation of this study is that reflects the experience of several Centers, that used different surgical protocols. Anesthetic techniques were not uniform and outcomes in the postoperative periods can reflect different strategies.
In conclusion, the results of our study demonstrate that DHCA+DR has the same prevalence of operative mortality, NE (any or permanent) and association of death+PND than MHCA+ACP. However, the data seems to suggest that DHCA+DR when compared with MHCA+ACP provides better renal protection and reduced prevalence of composite endpoint. These advantages have to be taken into account when the best surgical strategy has to be chosen for aortic arch surgery