The effect of whole-body cooling on renal function in post-cardiac arrest patients

BackgroundTo evaluate the incidence of Acute Kidney Injury (AKI) during therapeutic hypothermia (TH) and rewarming in comatose patients resuscitated from Cardiac Arrest (CA).MethodsWe have performed a pilot study of consecutive comatose patients resuscitated from CA and admitted to our Intensive Care Unit (ICU) from January 2013 to March 2015. The surface cooling devices used were: 1) Arctic Sun® 5000; 2) Blanketrol® III. Data obtained at baseline and during TH included: temperature trend and rate, serum creatinine, interleukin 1-beta, interleukin 6 (IL-6), urinary Interleukin-18 (uIL-18), diuretic use, urine output, fluid balance (FB). AKI was defined according to Kidney Diseases Improving Global Outcomes (KDIGO) criteria.ResultsThirty-six patients were treated with TH out of 46 ICU admissions (78%). According to KDIGO classification, 21 (58%) had no evidence of AKI while 15 (41.7%) presented AKI during TH. In particular, the incidence of AKI was 2.8% at 24 h, 33.33% at 48 h and 30.6% at 72 h from the onset of cooling. Slower rewarming (above 600 min) was associated with with a non-significant lower incidence of AKI and with a non-significant lower levels of IL-6 and IL-18u. Only two patients required renal replacement therapy during TH (7.6%). Median cumulative FB was 2441 [437–4043] ml for all patients; 3140 [1421–4347] and 1332 [−131–3772] specifically for AKI and not-AKI patients.ConclusionsThe hypothermia treatment, if not well performed, could be a double-edged sword for kidneys: whereas hypothermia may confer protection by reducing metabolism and oxygen consumption, rapid rewarming could nullify benefits leading to a worsening of kidney function and AKI. Additional clinical studies are needed to determine the optimal rewarming rate and strategy.


Background
Cardiac Arrest (CA) and resuscitation are an example of a whole-body Ischemia/Reperfusion (I/R) injury characterized by a multi-organ dysfunction and systemic activation of the innate immune system causing a 'sepsis-like syndrome' [1,2]. The restoration of spontaneous circulation (ROSC) after prolonged, complete, whole-body ischemia is an unnatural pathophysiological state created by successful cardiopulmonary resuscitation (CPR) [1]. The complex phase of resuscitation which begins when patients regain spontaneous circulation after CA has been defined by the International Liaison Committee on Resuscitation (ILCOR) as "post-cardiac arrest syndrome" [1]. The ischemia and reperfusion during CA, resuscitation and postresuscitation phases can particularly affect the kidney: I/R injury leads to damage of epithelial cells by increased oxygen radical production [2] and peroxidation of lipids [3,4] but also to interstitial inflammation and interstitial "micro-vasculopathy" that are involved in abnormal repair processes including incomplete repair of tubular cell and fibrosis development [4]. The cross-talk between the kidney and peripheral immune system in humans after whole-body I/R injury and the changes over time in the inflammatory response are speculative [5] and aside from cardiogenic shock limited information are available regarding predisposing factors and outcome of acute kidney injury (AKI) after CA [6,7]. Although therapeutic hypothermia (TH), also termed targeted temperature management, is an established intervention (ILCOR/American Heart Associatio) used for neuroprotection [3,4] though recent data challenge the benefit of TH using target temperature lower than 36°C [8,9]. The effect of TH on renal protection is uncertain and poorly investigated.
The purpose of this study was to investigate the effect of TH on the development of AKI and on renal outcomes in comatose patients resuscitated from CA and treated with surface cooling techniques. We hypothesized that TH could be a protective strategy against renal I/R injury while the shift from hypo-to normothermia may cause rewarming injury.

Study design
This was a single center observational study. The Institutional Review Board approved the protocol and written informed consent was obtained from the nearest relatives. Data were collected prospectively on 36 patients between January 2013 and March 2015.

Patients selection criteria
All patients admitted to the emergency department for out-of-hospital CA and treated with TH in the intensive care unit (ICU) with a standardized protocol [10] were considered eligible for the study. Inclusion criteria were: 18 years or older post-cardiac arrest patients (asystole or pulseless electrical activity or ventricular fibrillation or non-perfusing ventricular tachycardia as initial rhythm) and ST-segment elevation myocardial infarction. Exclusion criteria were: time from ROSC to initiation of cooling >6 h; body temperature below than 30°C on admission; major trauma; severe burns; major surgery less than 72 h; sepsis and septic shock; active bleeding or known pre-existing coagulopathy; intracranial bleeding; pregnancy; aggressive care. Patients were treated either by the Arctic Sun® or standard cooling blankets Blanketrol® IIIas decided by the treating physician. The time between collapse and start of cooling was calculated based on data provided by the paramedic team.

Patient management
In all patients, the resuscitation attempt followed the European Resuscitation Council 2010 guidelines for basic and advanced cardiac life support and postresuscitation care [11]. In conjunction with resuscitation procedures, the patient was assessed for study inclusion. ROSC was defined as an organized rhythm and palpable pulse sustained for at least 20 min. In prehospital setting, infusion of ice-cold normal saline, the use of cold packs or the trans-nasal cooling was permitted and it was according to choose of the physician involved in the first aid. If necessary, a coronary angiogram and a percutaneous coronary intervention were performed before the admission to ICU. Once the patient fulfilled the inclusion criteria and in agreement with our protocol, sedation, analgesia, and paralysis were started with: a Propofol bolus of 2 mg/kg, followed by continuous infusion (0.5-3 mg/Kg/h) or Midazolam bolus of 0.03-0.04 mg/kg, followed by continuous infusion (0.2 mg/kg/hr); continuous infusion of Remifentanil (0.02-0.2 mcg/kg/min) and of Cisatracurium continuous infusion of 1 mcg/kg/min. Other treatments such as magnesium, fluids, and vasopressors were left to the discretion of treating physicians. All patients had a bladder temperature probe inserted for a continuous measurement of core temperature, which provides feedback to the cooling devices, freely chosen by the physician involved in the treatment. The cooling devices used were: 1) Blanketrol® III, a cold blanket wrapped around the upper torso. Automatic control mode, designed to limit the magnitude of the difference between the bladder and circulating water temperature (gradient modes), was set with a bladder temperature of 33°C; If the patient's temperature was lower than the target temperature, the device heated the circulating water to the highest allowable water temperature (42°C); on the contrary, if the temperature was higher than the target temperature, the device cooled the circulating water to the lowest allowable water temperature (4°C) until the achieving the body target temperature. The circulation of the water without heating or cooling characterizes the reach of the target temperature. 2) Arctic Sun® 5000, the cooling pads applied on the back, chest and thighs. Automatic mode was set to a target temperature of 33°C, and the maximum cooling rate was used. The decision of the cooling device was performed by treating physician based on his preference. Target temperature was achieved and maintained for 24 h. At the end of the maintenance phase, rewarming phase started with a target temperature of 36°C and with a rewarming rate between 0.25-0.5°C per hour. Once the patient's temperature reached 36°C, the device was turned off.

Data collection and blood sampling
Demographic variables and pre-hospital data were collected prospectively for all patients at the admission utilizing the Utstein criteria. Hemodynamic parameters and temperature were measured continuously. Data obtained on admission and 6, 24, 48 and 72 h included: temperature trend and rate, serum creatinine (sCr), interleukin 6 (IL-6), interleukin 1-beta (IL-1beta), urinary interleukin −18 (uIL-18), urine output, fluid balance. AKI was defined according to Kidney Diseases Improving Global Outcomes (KDIGO) criteria [12,13]. Estimated Glomerular Filtration Rate and urinary output criteria were not used for AKI diagnosis and staging in this study. Baseline sCr was calculated using the Modification of Diet in Renal Disease (MDRD) equation (back-estimation). In particular, in absence of previous values, the baseline creatinine has been calculated as follow: GFR(Glomerular Filtration Rate) = (75/[186 × (age -0.203 ) × (0.742 if female) × (1.21 if black)]) -0.887 ). For each time point, sCr considered for AKI diagnosis was corrected for fluid balance according to recent evidence [14]. The Sequential Organ Failure Assessment (SOFA) score was calculated using standard methods. Quantitative determination of uIL-18 was performed by Human Instant ELISA kit (eBioscience, San Diego, CA, USA). Cytokines determinations were performed according to manufacturer's protocol and instructions. Optical density was read by using a VICTORX4 Multilabel Plate Reader (PerkinElmer Life Sciences, Waltham, MA, USA) at 450 nm. The concentration values for these molecules were calculated from standard curves, according to the manufacturer's protocols. All tests were performed in triplicate.

Outcomes
The primary outcome was the development of AKI during all treatment period in patients treated with either one of the two cooling devices. The secondary outcomes were: the evaluation of association between duration of each phase of TH (i.e. cooling, hypothermia, rewarming) and AKI; the variation of sCr, urine output, fluid balance and serum cytokine during TH; the effects of rewarming phase on development of AKI. The outcomes were assessed by study personnel who were aware of the type of external cooling assigned for the subject.

Statistical analysis
The Shapiro-Wilk test was used to test the data for normality. The differences between AKI and not-AKI patients were tested using Mann-Whitney-Wilcoxon for continuous variables, according to the not-normal distribution of data, and presented as median [I-III interquartile]. Categorical variables were analyzed using the chi-square test and given as a percentage. Differences between AKI and not-AKI have also been evaluated through univariate and multivariate analysis; results are presented in tables as p values, Odds ratio (OR) and 95% confidence interval (95% CI). A p value less than 0.05 was considered for statistical

Baseline demographics
From January 2013 till March 2015, 36 patients with successful CPR were treated by TH out of 46 ICU admissions for CA (78%) to the Department of Anesthesia and Intensive Care Medicine, San Bortolo Hospital (Fig. 1) and followed for the development of AKI during ICU stay (Table 1). Ten (28%) patients were treated with Blanketroll III (Blanket Group), and twenty-six (72%) patients were treated with the Arctic Sun Cooling Machine (Artic Sun Group).

Temperature trend and duration of each treatment phase
At the initiation of cooling, the temperature dropped from a median pre-treatment temperature of 35. 6 (Fig. 2). The duration of each phase and the time In overall population, the median fluid balance from 24 h to 72 h was +2441 [432-4044] ml. Only two patients required RRT during the treatment (7.6%) ( Table 2).

Cytokines concentration during the treatment period
For each time point, the plasmatic concentration of IL-6 and IL-1beta and the urinary concentration of uIL-18 have been described for overall population and for AKI and no AKI patients in Table 4. Figure 5 shows cytokines levels at different timepoint in AKI and not-AKI patients with slow rewarming (Panel A) and with rapid rewarming (Panel B).

Discussion
To the best of our knowledge, this is the first study suggesting that the rewarming phase after TH applied to post-CA patients may have an impact on the development of AKI. During the TH period, among all 36 patients, 21 (58%) had no evidence of AKI according to KDIGO classification. AKI is a common occurrence, in particular from  out-of-hospital-CA patients. Indeed, post-cardiac arrest is characterized by a whole-body ischemia followed by reperfusion with activation of a systemic inflammatory response and release of injury products into the circulation [5,6] responsible of additive stimuli for ischemic injury [15]. Following resuscitation from spontaneous CA, AKI is a common complication with increased mortality risk, dialysis requirement and prolonged hospital stay [7,16,17]. An intriguing field of research shows the capability of a decrease in whole-body temperature to limit the ischemic injuries. Although TH has been used from 1950s to protect the brain against a global ischemia after CA, there is scant data on the potential benefit of this strategy on kidney endpoints [18,19]. Experimental studies   [20]. These findings in literature showed that in animals hypothermia is able to reduce the risk of renal failure after renal I/R injury [21,22]. Renal function is eventually depressed during hypothermia owing to a fall in systemic blood pressure and the effect of hypothermia on organ metabolism itself. As renal blood flow progressively decreases, renal vascular resistance rises, promoting a further decrease in renal flow and a subsequent decrease in glomerular filtration. Increased blood viscosity and vasoconstriction were both responsible for this reduction of flow. In the setting of TH, this effect is maintained for 24 h. In this hypothermic phase, from admission to 24 h, there is a decrease of sCr probably due to decrease in muscle metabolism. Clinical practice is based on sCr measurements to monitor renal function and classify AKI. For this reason, we used a validated criteria to define and classified AKI: KDIGO. It is presumed that TH increases urine output particularly in the induction phase, this phenomenon is called cold-induced diuresis [23]. For this reason, we decide to not consider urinary output criteria for diagnosis of AKI. Our results suggested that in the rewarming phase, which started after 24 h, there is a slight increase of sCr associated with an increase in urine output that did not correlate with diuretic dose. Rewarming is a delicate phase of therapeutic hypothermia. Similar to the brain [24], the rewarming phase seems to be very important for the kidney and we strongly believe that could be dangerous if it is not performed slowly and with adequate time duration. Our results showed that with a rewarming time more than 600 min the risk of AKI is decreased. However, it does not mean that longer rewarming times exclude any risk of AKI. Rewarming of the patient is begun 24 h after the initiation of cooling. The literature recommends rewarming slowly at a temperature of 0.3-0.5°C every hour. It means that rewarming will take approximately 8 h [25]. Differences in the time of rewarming could impact microinflammation and AKI risk. In our study, patients who did not develop AKI had median rewarming time of 630 min (10.5 h) whereas those who did develop AKI had a median time of 540 (9 h). Rewarming is a delicate phase of TH. Adverse consequences of rewarming on the whole body may seriously limit the protective effects of hypothermia, leading to secondary injury. In literature, previous studies showed that the rise in body temperature accompanies a vasodilation causing a rewarming shock and then careful fluid monitoring during rewarming is crucial [26]. Probably, controlled rewarming of 0.15°C per hour should be recommended [27] Our data support this recommendation showing less AKI with rewarming time > 600 min which would translate to a rewarming rate of less than 0,4°C/h. It should be considered that in patients with severe oliguria or anuria despite volume resuscitation, RRT should be started before rewarming to avoid hyperkalemia or metabolic acidosis due to transcellular electrolyte shifts associated with this phase [28]. Hypothermia suppresses ischemia-induced inflammatory reactions and release of pro-inflammatory cytokines [29]. In our study, we assessed inflammatory response dosing IL-6, IL-1beta and IL-18 and results supported the evidence. In addition, uIL-18, a 22 KDa pro-inflammatory cytokine, produced by immune cells, macrophages and proximal tubular cells, has a role in inflammatory processes that exacerbate renal injury during the extension phase of AKI [30,31]. This cytokine is an early biomarker for AKI in a variety of clinical scenarios indicating the activation of inflammation in kidney [32][33][34]. Our results showed that with a rewarming time less than 600 min, there is a more increase in concentration respect to those with a slower rewarming (above 600 min) supporting the hypothesis that rapid rewarming is associated with increase risk of AKI (Table 4). The study has some limitations. First, this is a single center trial with only a relative small number of patients included. Secondly, the patients were assigned to two different cooling methods in a nonrandomized fashion. Thus, a bias introduced by the cooling method cannot be excluded. Our results, suggest possible harm by an insufficiently controlled and or too short rewarming period. As feasibility study, there was not a comparison with a normothermic control group. The study is too small to make conclusions about outcomes, although the results are hypothesis generating.

Conclusions
The hypothermia treatment, if not well performed, could be a double-edged sword for kidneys: whereas hypothermia may confer protection by reducing metabolism and oxygen consumption, rapid rewarming could nullify benefits leading to a worsening of kidney function and AKI. Additional clinical studies are needed to determine the optimal rewarming rate and strategy.

Key messages
In our prospective study, a rewarming time of more than 600 min is associated to a decrease of the risk of AKI; The uIL-18 levels has a role in inflammatory processes that exacerbate renal injury during the extension phase of AKI. During rewarming time less than 600 min, there is a more increase in concentration respect to those with a slower rewarming (above 600 min) supporting the hypothesis that rapid rewarming is associated with increase risk of AKI. Whereas hypothermia may confer protection by reducing metabolism and oxygen consumption, rapid rewarming could nullify benefits leading to a worsening of kidney function and AKI.