Advice

Evidence review

Evidence review

Clinical and technical evidence

Three clinical studies in which the RhinoChill system was used as an intervention for cardiac arrest were identified from the literature search. These were a randomised controlled trial, the PRINCE trial (Castrén et al. 2010); a single-arm observational study (case series; Busch et al. 2010); and a case study (Gordic et al. 2013).

The PRINCE trial was conducted in a pre-hospital emergency setting (before resuscitation) in 200 patients who had a cardiac arrest (Castrén et al. 2010). The primary outcomes were return of spontaneous circulation, survival until discharge (of those admitted to hospital alive), and survival neurologically intact (of those admitted to hospital alive). None of these outcomes were found to be significantly improved by the intervention. However, in patients whose spontaneous circulation returned within 10 minutes, and who were admitted to hospital alive, post hoc subgroup analysis showed that the intervention was associated with a statistically significant increase in survival compared with control (56.5% compared with 29.4%, p=0.04) and neurologically intact survival (43.5% compared with 17.6% p=0.03). For secondary intermediate outcomes, the intervention was found to have a statistically significant cooling effect on arrival at hospital as measured by tympanic membrane temperature (34.2°C compared with 35.5°C, p=0.001) and core temperature (35.1°C compared with 35.8°C, p=0.01). Adverse events are reported in table 1.

The case series recruited 84 patients whose spontaneous circulation had returned (Busch et al. 2010). After using the intervention for 1 hour, the median temperature reduction was 2.3°C (tympanic membrane temperature) and 1.1°C (core temperature). Thirty-four of 84 patients (40%) survived until discharge, with 76% of these (26/34) having a favourable neurological outcome. However, without a comparator group it is not possible to fully attribute these effects to the intervention. Adverse events are listed in table 2.

A single case report (Gordic et al. 2013) did not provide usable efficacy or safety data, but included the observation that after migration into the pharynx and paranasal sinuses the RhinoChill coolant appears as a hyperdense liquid (resembling contrast media) in whole-body CT. In the case reported, this confused emergency physicians, anaesthetists and radiologists and led to a delay in patient treatment.

In addition to the published studies, a number of clinical trials of the RhinoChill system for cardiac arrest, stroke, traumatic brain injury and other clinical applications were highlighted by the manufacturer. At the time of publication of this briefing, 2 studies are complete and pending results, 5 studies are currently recruiting and a further 6 studies are in development. The ongoing Prehospital Resuscitation Intra Nasal Cooling Effectiveness Survival Study (PRINCESS) will report on neurologically intact survival as the primary outcome measure.

The manufacturer has also provided information on 2 in-service evaluations of the RhinoChill system:

  • London's Air Ambulance Service (a charity that provides advanced trauma services to critically injured people in London) studied the RhinoChill system for cardiac arrest. Starting in May 2012, the device was carried in the Physician Response Unit, to start therapeutic hypothermia during active resuscitation and before return of a pulse. The aim of the pilot study, on 20 patients, was to assess the feasibility of using the RhinoChill system before arrival in hospital in the UK (Lyon et al. in press).

  • The South East Coast Ambulance Service NHS Foundation Trust, with the Accident and Emergency Department of the Royal Sussex County Hospital, have begun an in-service evaluation of the RhinoChill system, and have treated 4 patients up to June 2013. A planned total of 25 patients with cardiac arrest will start on the RhinoChill system as soon as they have been resuscitated and when treatment has been started by a critical care paramedic. Brain cooling with the system will be maintained from when patients arrive at hospital until they are transferred to the intensive care unit. The evaluation will assess the ease of use of the RhinoChill system in an ambulance, the time it takes to reduce the patient's brain temperature to the optimal range of 32–34°C, the number of days the patient spends in intensive care, the percentage of patients surviving to discharge from hospital and the neurological status of surviving patients at their discharge.

The manufacturer has made available a statement of reported adverse events related to the RhinoChill system, on which the following summary is based.

Risks associated with the use of the RhinoChill intranasal cooling system include those related to the use of the device as well as those related to systemic hypothermia. The following device- or procedure-related adverse events were reported in the 213 patients in whom the RhinoChill has been used:

Serious adverse events

  • Cold-related tissue damage: n=1 (0.47%). The patient was in irreversible cardiogenic shock when enrolled and cooled with the RhinoChill system. Tissue discoloration appeared approximately 2 hours after RhinoChill cooling was halted following an 80‑minute cooling period. The patient died from persistent cardiogenic shock approximately 36 hours after RhinoChill use without resolution of tissue discoloration.

  • Epistaxis: n=1 (0.47%). The patient was enrolled under emergency conditions during resuscitation from cardiac arrest with an undiagnosed coagulopathy. Bleeding began 16 minutes after RhinoChill cooling was started. Cooling was halted, and resuscitation efforts persisted for 8 more minutes. The patient did not regain spontaneous circulation despite 40 minutes of basic and advanced cardiac life support efforts. It was learned that the patient was in late stage hepatic failure at the time of cardiac arrest.

  • Hypertension: n=1 (0.47%). The patient presented unconscious with no apparent need for additional sedation. Mean arterial pressure rose from 75 to 94 [mmHg] within the first 15 minutes of RhinoChill cooling. An oral anti-hypertensive was administered. No reduction in pressure was observed over the subsequent 15‑minute period of RhinoChill cooling. Cooling was then halted, an intravenous line was placed, and the patient was sedated; mean arterial pressure normalised within 30 minutes.

  • Hypoxia: n=1 (0.47%). A loss of airway protection was noted during RhinoChill cooling with a ventilation setting of 50% inspired oxygen when the pulse oximetry reading fell to 94%. Airway protection was re-established and cooling was halted per protocol 15 minutes later. Arterial oxygen saturation fell to 85% over the ensuing 2‑hour period. Inspired oxygen was then increased to 100% and positive end-expiratory pressure was increased. Arterial oxygen saturation rose from 85 to 97% over the subsequent 30 minutes. Arterial saturation was maintained at 98% on 50% oxygen within 2 hours.

Non-serious adverse events

  • Discoloration of the nasal tissue – approximately 10%

  • Epistaxis – approximately 5%

  • Peri-orbital emphysema – approximately 1%

Additional potential risks exist that have not been attributed to the use of the RhinoChill in past patient series. These include:

  • Acute myocardial infarction

  • Barotrauma to the nasal cavity/nasopharynx

  • Cardiac rhythm disturbances including ventricular fibrillation

  • Death

  • Emphysema, pulmonary

  • Haemorrhage (not epistaxis)

  • Infection

  • Pneumonia

  • Pulmonary embolism

  • Sepsis

Post-marketing experience

The manufacturer states that the following non-serious adverse events have been reported since marketing the RhinoChill system:

  • Possible device-related mechanical trauma to intranasal surface; this could have been a result of inserting the RhinoChill catheter or other devices into the nasal cavity (n=1).

  • Device-related pneumocephalus, which resolved without medical intervention and without sequelae to the patient (n=1).

Table 1 Summary of the PRINCE trial: Castrén et al. (2010)

Study component

Description

Objectives/
hypotheses

To determine the safety and feasibility of transnasal evaporative cooling by pre-hospital EMS personnel during ongoing resuscitation before achieving ROSC

Study design

Randomised controlled trial

Setting

Pre-hospital setting

Inclusion/exclusion criteria

Inclusion criteria: adults aged ≥18 years, with witnessed cardiac arrest, eligible for advanced cardiac life support irrespective of cardiac rhythm, when cardiopulmonary resuscitation was initiated by EMS within 20 minutes of collapse

Exclusion criteria: patients with trauma, drug overdose, cerebrovascular accident, known coagulopathy, asphyxia or known requirement for supplemental oxygen, electrocution, very cold temperature on arrival of EMS personnel, those in whom successful ROSC was achieved before randomisation and those with a do-not-attempt resuscitation order or intranasal obstruction

Primary outcomes

All adverse events within 24 hours, serious adverse events within 7 days, neurological outcome via the CPC score

Statistical methods

Continuous variables that were not normally distributed reported as median and IQR. Categorical variables reported as counts and percentages. Primary analyses for the efficacy end points were conducted with Pearson chi-squared tests and comparison of binomial proportions. Relative risks (expressed as treatment divided by control) were computed to further characterise the effect sizes. Other analyses were performed with 2-group t tests or Wilcoxon rank sum tests for continuous variables and Pearson chi-squared tests for categorical variables. All probability values were 2-sided, with values less than 0.05 regarded as statistically significant. No statistical adjustments were made to account for multiple comparisons

Participants

n=200 (6 lost during follow-up, 3 from each group)

Results

There was no increase in serious adverse events within 7 days in the treatment group. Among device-related adverse events, nasal whitening was the most common event, occurring in 14% of patients. It resolved spontaneously in all resuscitated patients. Epistaxis occurred in 3 treated patients and was serious in 1 patient with an underlying coagulopathy secondary to hepatic failure. This was the only device-related serious adverse event

Conclusions

Pre-hospital intra-arrest transnasal evaporative cooling is feasible and safe. Early use of cooling is associated with a significant improvement in the time interval required to cool patients. No improvement in the rate of ROSC was observed for the intra-arrest cooled group

RhinoChill + advanced cardiac life support

Advanced cardiac life support only

Analysis

Randomised

n=93

n=101

Not applicable

Efficacy

Primary outcome:

ROSC achieved

35 (37.6%) (n=93)

43 (42.6%) (n=101)

p=0.48

Primary outcome:

Survival until discharge (of those admitted to hospital alive)

14 (43.8%) (n=32)

13 (31.0%) (n=42)

p=0.26

Primary outcome:

Favourable neurological outcome (of those admitted to hospital alive) (CPC=1, CPC=2)

11 (34.4%) (n=32)

9 (21.4%) (n=42)

p=0.21

Selected secondary outcomes:

Mean (SD) tympanic temperature at ROSC, °C

35.5 (0.9) (n=93)

35.8 (1.5) (n=101)

p=0.40

Mean (SD) tympanic temperature at arrival at hospital, °C

34.2 (1.5) (n=93)

35.5 (0.9) (n=101)

p<0.001

Time to tympanic temperature of 34°C, median [IQR], minutes

102 [81:155] (n=93)

291 [183:416] (n=101)

p=0.03

Mean (SD) core temperature, °C

35.1 (1.3) (n=93)

35.8 (0.9) (n=101)

p=0.01

Time to core temperature of 34°C, median [IQR], minutes

155 [124:315] (n=93)

284 [174:471] (n=101)

p=0.13

Safety:

Adverse events within 24 hours (all patients)

n=93

n=101

Nasal whitening

14.0% (13/93)

0% (0/101)

Not reported

Epistaxis

3.2% (3/93)

0% (0/101)

Not reported

Peri-orbital emphysema

1.1% (1/93)

0% (1/101)

Not reported

Safety:

In-hospital data (those admitted to hospital alive)

n=32

n=42

Cardiogenic shock cause of death

9.4% (3/32)

26.2% (11/42)

p=NS

Length of hospitalisation, days

24.1

26

p=NS

Length of ICU stay, days

8

11

p=NS

Time on ventilator, days

4.2 

8.8

p=NS

Patients reporting serious adverse events (within 7 days)

22% (7/32)

33.3% (14/42)

p=0.23

Acidosis

0% (0/32)

4.8% (2/42)

Not reported

Acute myocardial infarction (non-fatal)

0% (0/32)

2.4% (1/42)

Not reported

Bleed

(device-related epistaxis in patient with an underlying coagulopathy)

3.1% (1/32)

2.4% (1/42)

Not reported

Cardiac arrest (new)

9.4% (3/32)

4.8% (2/42)

Not reported

Convulsions

3.1% (1/32)

2.4% (1/42)

Not reported

Lethal/long-lasting arrhythmia

3.1% (1/32)

4.8% (2/42)

Not reported

Renal failure

3.1% (1/32)

4.8% (2/42)

Not reported

Sepsis/multi-organ failure

0% (0/32)

7.1% (3/42)

Not reported

Abbreviations: CPC, cerebral performance category; EMS, emergency medical services; IQR, interquartile range; n, number of patients; NS, not significant; ROSC, return of spontaneous circulation; SD, standard deviation.
Table 2 Summary of the case series: Busch et al. (2010)

Study component

Description

Objectives/
hypotheses

Primary aim to demonstrate safety, feasibility and cooling effectiveness of nasopharyngeal evaporative cooling in comatose patients after successful resuscitation from cardiac arrest

Study design

Case series

Setting

11 European intensive care units and emergency departments

Inclusion/exclusion criteria

Inclusion criteria: patients aged over 18 years, who did not obey any verbal command at any time after ROSC or if received chest compressions for any duration, tympanic temperature >34°C, oxygen saturation >95% on 50% oxygen

Exclusion criteria: if trauma or severe bleeding occurred subsequent to cardiac arrest, existence of terminal disease, pregnancy, known coagulopathy, or a barrier inhibiting placement of the intranasal catheter (for example septum deviation, skull base fracture)

Primary outcomes

Cooling rate, time needed to achieve mild hypothermia (34°C), time needed to achieve target temperature (33°C), side effects of evaporative cooling in the nasopharynx and elsewhere occurring between enrolment and discharge, neurological outcome via the CPC score

Statistical methods

Descriptive statistics were performed. Continuous variables presented as median and IQR. Binary variables presented as number and percentage

Participants

n=84

Results

The RhinoChill device effectively lowered tympanic and core temperatures in patients after cardiac arrest. The device proved feasible in an emergency department setting, and safe during 1 hour's use, with the exception of persistent tissue damage in 1 patient at the high oxygen flow rate of 60–80 litres/min. At the lower oxygen flow rate of 40–50 litres/min, no persistent cold-related tissue damage was observed. The reduction in flow rate from 60–80 litres/min to 40–50 litres/min did not affect the cooling rate. Essential safety measures that prevent tissue damage include uncovering the face and keeping the mouth open during cooling, so that coolant vapour can escape from mouth and nostrils. No evidence was obtained that the coolant might have caused lung damage following aspiration. Smell tests demonstrated that cooling via the nasal cavity did not affect the olfactory epithelium. The mortality rate of 60% was not unexpected for this unselected patient population, and good neurologic recovery was observed in a comparably high percentage

Conclusions

Nasopharyngeal cooling for 1 hour using the RhinoChill device is effective in reducing core temperature in cardiac arrest survivors. The device is safe at oxygen flow rates of 40–50 litres/min

RhinoChill

Randomised

Not applicable

Efficacy

Duration of cooling (median [IQR], [min–max] minutes)

60 [50:90], (25–195) (n=84)

Tympanic cooling rate in 1 hour of cooling (median [IQR] °C)

2.3 [1.6:3.0] (n=82)

Core cooling rate during cooling (median [IQR] °C/h)

1.1 [0.7:1.5] (n=84)

Cooling rate for central core temperature measurements from oesophageal and arterial sites (median [IQR] °C/h)

1.4 [0.9:2.0] (n=36)

Cooling rate for peripheral core temperature measurements from bladder and rectal sites (median [IQR] °C/h)

0.9 [0.5:1.2] (n=48)

Time to tympanic temperature 34°C (median [IQR] minutes)

27 [14:58] (n=82)

Time to tympanic temperature 33°C (median [IQR] minutes)

60 [36.5:117.5] (n=82)

Time to core temperature 34°C (median [IQR] minutes)

52 [26:86] (n=36)

Time to core temperature 33°C (median [IQR] minutes)

180 [120:285] (n=36)

Number reaching target tympanic temperature 33°C

55 (66%) (n=76)

Number reaching target core temperature 33°C

16 (19%) (n=80)

Number with favourable neurological outcome (CPC=1, CPC=2)

26 (76%) (n=34)

Safety

Mortality

50 (59.5%) (n=84)

Total number of adverse events related to device

15 (17.9%) (n=84)

Nasal discoloration

10 (11.9%) (n=84)

Cold-induced tissue damage

1 (1.2%) (n=84)

Epistaxis

2 (2.4%) (n=84)

Coolant in sinus

1 (1.2%) (n=84)

Peri-orbital gas emphysema

1 (1.2%) (n=84)

Patients reporting serious adverse events

n=1 (1.2%)

Unspecified event

After the observation of a severe device-related adverse even at the oxygen flow rate of 60–80 litres/min, the flow rate during cooling was lowered to 40–50 litres/min. After reduction of the flow rate, no more serious device-related adverse events were observed

Abbreviations: CPC, cerebral performance category; IQR, interquartile range; n, number of patients; ROSC, return of spontaneous circulation.

Costs and resource consequences

The RhinoChill intranasal cooling system is intended for use as an adjunct to systemic temperature control systems and is an additional procedure, and cost, in the emergency care pathway.

The RhinoChill intranasal cooling system has only been used in the NHS to date in clinical trials or in-service evaluations. No published information on the NHS costs of adopting the technology was found.

Other than medical-grade oxygen or breathing air and a power source to charge the control unit, the RhinoChill is a standalone system that does not need any supporting technology to operate.

The only training resources needed for the RhinoChill system are the instructions for use.

The manufacturer states that patient temperature monitoring is needed whenever the device is used. Tympanic or oesophageal, but not nasopharyngeal, thermometers are appropriate for use.

No published evidence on resource consequences was identified.

Strengths and limitations of the evidence

One randomised controlled trial that described the Rhinochill system as the intervention was identified by the literature search. This was the Pre-ROSC IntraNasal Cooling Effectiveness trial (PRINCE trial; Castrén et al. 2010) which recruited 200 patients. Overall, this was a well-conducted trial that set out to investigate the feasibility, safety and efficacy of the intervention in a pre-hospital emergency setting. This was a challenging trial to perform, and there were inevitable practical difficulties that could lead to some forms of selection, performance and detection bias, but which were reasonably well controlled for. Additionally, this was a relatively small trial that was not powered to detect the primary clinical outcomes of achieving successful return of spontaneous circulation, overall survival and discharge without neurological deficit. The clinical outcomes reported in the trial are therefore subject to significant uncertainty and should be viewed with caution. Although the trial failed to conclusively demonstrate any clinical benefit from the device, the intervention significantly reduced cerebral and core temperatures relative to control (standard care).

One case series that described RhinoChill as the intervention was identified by the literature search (Busch et al. 2010). This was a prospective, multicentre, single-arm observational study. A total of 84 patients were recruited into this study, all of whom had suffered a cardiac arrest but had since achieved return of spontaneous circulation (in contrast to the PRINCE trial). Although this study was well conducted it lacked a control group, and consequently was unable to demonstrate the safety or efficacy of the intervention over that of standard care.

Two service evaluations are currently being performed in UK NHS settings. As these are uncontrolled observational studies they are likely to provide low-level evidence for the RhinoChill system.