Intraosseous administration of hydroxocobalamin after enclosed structure fire cardiac arrest

Abstract

Smoke inhalation is the most common cause of acute cyanide poisoning in the developed world. Hydroxocobalamin is an antidote for cyanide poisoning. There is little published about human intraosseous antidote administration. We present a case of intraosseous hydroxocobalamin administration in an adult smoke inhalation victim, found in cardiac arrest inside her burning manufactured home. Return of spontaneous circulation was achieved after 20 min of cardiopulmonary resuscitation. Five grams of hydroxocobalamin were subsequently given intraosseously. On hospital arrival, patient was found to have a respiratory-metabolic acidosis. She had red-coloured urine without haematuria, a known sequela of hydroxocobalamin administration. Patient’s neurological status deteriorated, and she died 4 days after admission. This case highlights that intraosseously administered hydroxocobalamin seems to adequately flow into the marrow cavity and enter the circulatory system despite the non-compressible glass antidote vial. This appears to be only the second reported human case of intraosseous hydroxocobalamin administration.

BACKGROUND

Hydrogen cyanide inhalation is a well-known concern associated with smoke inhalation in structure fires.¹ Smoke inhalation is the most common cause of acute cyanide poisoning in the developed world.² Hydrogen cyanide is one of many toxic byproducts formed from incomplete combustion of household synthetics such as polyurethane insulation and foams, furniture and carpeting.¹ Hydroxocobalamin is a precursor to vitamin B₁₂, cyanocobalamin. Hydroxocobalamin binds cyanide, forming cyanocobalamin, which is then excreted in the urine.³ Intravenous (IV) hydroxocobalamin has been used in France since the 1980s for out-of-hospital cyanide poisoned patients.² Hydroxocobalamin was approved by the Food and Drug Administration for use in the USA in 2006. A prospective study in 2007 concluded that hydroxocobalamin is safe to administer to critically ill smoke inhalation victims.² A 2016 study by Nguyen et al found that while hydroxocobalamin did not significantly affect mortality rates when compared with a control group, hydroxocobalamin use led to lower pneumonia rates, fewer days on the ventilator and shorter duration spent in the intensive care unit.⁴ A three-component cyanide antidote kit containing sodium thiosulphate, sodium nitrite and amyl nitrite is available but is not recommended for use in smoke inhalation victims. This is because the nitrites create methaemoglobin, decreasing the blood’s oxygen carrying capacity in a victim that is already relatively hypoxic due to inhalation of carbon monoxide and cyanide from the fire smoke. There is a relative paucity of human research evaluating intraosseous (IO) administered antidotes.^{5 6} We report a case of IO administered hydroxocobalamin in a prehospital adult cardiac arrest smoke inhalation victim.

Case presentation

Patient is a 53-year-old woman with a history of asthma, bipolar disorder, chronic obstructive pulmonary disease, rheumatoid arthritis and a smoking history of 8 pack-years who was found in cardiac arrest inside her burning manufactured home by a Basic Life Support fire department.

At 14:54, the local fire department was dispatched for a reported structure fire and arrived on-scene 4 min later to find a manufactured home with light smoke from the side door and windows (table 1). The initial entry search and rescue crew found heavy smoke conditions and fire in the living room. As the crew continued a search with a thermal imaging camera, they found an unresponsive woman sitting in a hallway chair. She was extricated from the home and was found to be without pulses or spontaneous respirations. High-performance cardiopulmonary resuscitation (CPR) was initiated in a 30:2 ratio, with ventilations provided by a bag valve mask (BVM) with an oropharyngeal airway adjunct.

Table 1.

Prehospital timeline*

Time	Event/Intervention
14:54	Fire department dispatched
14:58	Fire department on-scene
15:03	Entry into structure
15:06	Victim located
15:07	Victim extricated and CPR initiated
15:10	Advanced Life Support ambulance on-scene
15:14	Supraglottic airway placed
15:20	Tibia IO line placed
15:21	EMS physician response unit on-scene
15:26	Return of spontaneous circulation
15:27	Intubation; hydroxocobalamin prepped and administered
NR	Humerus IO line placed
15:50	Advanced Life Support ambulance en route to hospital
16:07	Advanced Life Support ambulance at hospital

*Times obtained from fire department and advanced life support ambulance reports

CPR, cardiopulmonary resuscitation; EMS, Emergency Medical Services; IO, intraosseous; NR, Exact Time Not Recorded.

Subsequently, an Advanced Life Support ambulance arrived and a supraglottic airway device was placed. End-tidal carbon dioxide (EtCO₂) was monitored and initially found to be 99 mm Hg. Patient was noted to have superficial burns to the mouth and face along with soot around her nasal and oral airways. Paramedics attempted to gain peripheral IV access without success. IO access was then obtained in the left tibia. The patient was noted to be in a tachycardic sinus rhythm with pulseless electrical activity during rhythm checks. After approximately 20 min of resuscitation, including two rounds of epinephrine, return of spontaneous circulation (ROSC) was achieved. Vitals signs were obtained and included a pulse of 126 beats/min (bpm), blood pressure of 206/115 mm Hg, respiratory rate of 9 assisted breaths/min, oxygen saturation of 99% and an EtCO₂ of 97 mm Hg.

Following ROSC, an on-scene Emergency Medical Services (EMS) physician successfully placed an endotracheal tube and a left humeral IO line. Out of concern for possible cyanide poisoning, 5 g of hydroxocobalamin were prepared using a commercially available kit (CyanoKit) that includes vented IV tubing and a glass antidote vial. The hydroxocobalamin freely infused from the glass vial through the left humeral IO site and completed its infusion during the 17-minute transit time to a Level-I Trauma Centre. During transport, patient began to take spontaneous breaths, bite down on the endotracheal tube and open her eyes, although was not following commands. Vital signs following the administration of hydroxocobalamin as reported by the paramedic at 16:05 were, a pulse of 107 bpm, blood pressure of 181/118 mm Hg, respiratory rate of 15 breaths/min, oxygen saturation 94% and an EtCO₂ of 62 mm Hg.

Investigations

On arrival to the emergency department resuscitation bay (table 2), patient’s vital signs included a temperature of 97.5° Fahrenheit, pulse of 104 bpm, blood pressure of 151/123 mm Hg, respiratory rate of 23 breaths/min being assisted via an endotracheal tube, oxygen saturation of 95% and EtCO₂ of 66 mm Hg. She had bilateral diminished breath sounds. Her Glasgow Coma Scale score was 6. She occasionally opened her eyes but would not follow commands. The patient seemed to withdraw to pain; however, this was difficult to assess due to spontaneous tremulous movements. She was given 5 mg midazolam with resolution of this seizure-like activity. A primary and secondary trauma assessment were completed. Patient was noted to have soot around the nostrils and mouth along with 4.5% body surface area partial thickness burns isolated to the face around the nose and mouth. An initial arterial blood gas demonstrated a pH of 7.05, pCO₂ of 89 mm Hg, pO₂ of 218 mm Hg and HCO₃ of 24 mEq/L. Her lactic acid was 8.5 mmol/L. She had a carboxyhaemoglobin of 4.9%, and a methaemoglobin of 0.0. Her troponin was negative. A foley catheter was placed and the urine was sent to the laboratory for urinalysis. The EMS physician specifically noted a distinct purple–red hue to the urine. The urine colour was officially reported by the laboratory as red. There were no red blood cells present and no bacteria.

Table 2.

Hospital timeline*

Time	Event/intervention
Hospital day 1
16:09	Arrival in emergency department resuscitation bay
16:15	Peripheral IV inserted, left hand
16:18	Transitioned from manual ventilation to mechanical ventilation: mode: continuous mandatory ventilation, rate: 20 breaths/min, tidal volume: 500 mL, positive end expiratory pressure: 8 cm H2O, inspiratory time: 0.9 s, FiO2: 100%
16:19	Nasogastric tube placed, left naris; medical critical care consult
16:22	Propofol infusion started
16:29	Urethral Foley catheter placed with return of purple–red urine, as witnessed by EMS physician, emergency department attending physician and trauma surgery resident
16:30	Trauma surgery service consult
16:40	Quadruple lumen (right subclavian vein) central line placed
16:57	Patient transferred to trauma burn unit
18:41	Arterial line (right femoral artery) placed
19:46	Second dose of hydroxocobalamin administered: 5 g IV piggyback over 60 min
NR	Bronchoscopy revealed carbonaceous sputum
Hospital day 2
08:38	All sedation discontinued, non-reactive pupils, cough reflex present, CT of the head demonstrated loss of grey–white differentiation and the presence of diffuse cerebral oedema
16:29	Gift of life contacted
Hospital day 3
NR	Next of kin arrived; patient transitioned to comfort care in anticipation of organ donation after cardiac death
Hospital day 4
15:31	Patient extubated
15:46	Patient pronounced dead and transferred to operating room for organ donation (liver and kidneys)

*Times obtained from patient’s electronic medical record

EMS, Emergency Medical Services; IV, intravenous; NR, Exact Time Not Recorded.

Treatment

The patient was transferred to the trauma intensive care unit for further management of her post-cardiac arrest pathophysiology, acute respiratory failure, inhalation injury and burns. She was transitioned to a ventilator from the bag value mask in the Emergency Department with initial settings of continuous mandatory ventilation mode, tidal volume 500 mL, rate of 20 breaths/min, positive end expiratory pressure of 8 cm H₂O and an FiO₂ of 100%. An arterial line and a central venous catheter were placed. She was treated for inhalational injuries with inhaled heparin, albuterol, ipratropium bromide and N-acetylcysteine. A second dose of 5 g of hydroxocobalamin was given as an IV piggyback. Vital signs obtained before this second administration of hydroxocobalamin were: pulse of 66 bpm, blood pressure of 123/90 mm Hg, respiratory rate of 26 breaths/min and oxygen saturation of 100%. Following the completion of the hydroxocobalamin infusion, repeat vital signs were: pulse of 69 bpm, blood pressure of 127/90 mm Hg, respiratory rate of 26 breaths/min and oxygen saturation of 98%. Patient also underwent bronchoscopy on the day of admission, which demonstrated carbonaceous sputum.

Outcome and follow-up

With aggressive medical management and ventilator support, patient’s lactic acidosis and hypercarbia (pCO₂) improved to 1.8 mmol/L and 35 mm Hg, respectively, by the first night of hospitalisation. However, on day 2 of admission, patient was no longer reacting to painful stimuli and was found to have anisocoria. A CT scan of the head was obtained and demonstrated diffuse loss of the grey–white differentiation and diffuse cerebral oedema concerning for anoxic brain injury. Over the next 2 days, patient’s neurological status continued to decline. Her pupils became fixed and dilated. The family elected to pursue organ donation given the patient’s poor prognosis for recovery. The patient was extubated and expired on hospital day 4.

Discussion

The mechanism of cyanide toxicity is well established, whereby the cyanide molecule binds to the mitochondrial cytochrome C oxidase, arresting the electron transport chain. This inhibits cellular respiration and the production of adenosine triphosphate by oxidative phosphorylation.^{3 7} Tissues, which consume the greatest amount of oxygen, most notably the brain and heart, are subsequently most affected by this interruption in the electron transport chain.

With the inhibition of oxidative phosphorylation, energy production at the cellular level is significantly reduced. The shift from aerobic metabolism to anaerobic metabolism leads to increased glycolysis, ending with conversion of pyruvate to lactate, causing the associated lactic acidosis often seen in cyanide poisoning patients. A plasma lactate level greater than or equal to 10 mmol/L, in the context of smoke inhalation, has been shown to be both a sensitive and specific marker of cyanide poisoning.² Cyanide toxicity may manifest as altered mental status, dizziness, dyspnoea, cardiac dysrhythmias, seizures and/or cardiovascular collapse.¹ Although the patient in this case report, with smoke exposure from an enclosed structure fire, had a lactate level less than 10 mmol/L, this level was not drawn at the scene, and is potentially confounded by interventions prior to the hospital venipuncture, including administration of IO fluid and hydroxocobalamin.

Carbon monoxide poisoning is also of a major concern in enclosed structure fires. Our patient was a known tobacco smoker and was found to have a carboxyhaemoglobin level of 4.9% while in the emergency department. A carboxyhaemoglobin of 3%–5% is considered to be in the expected physiological range for smokers, with a level greater than 10% (>3% in non-smokers) considered to provide evidence for carbon monoxide poisoning.⁸ Our patient received 100% normobaric oxygen delivered by a BVM through a supraglottic airway/endotracheal tube in the field and by a mechanical ventilator in the hospital. The elimination half-life of carboxyhemoglobin while breathing 100% normobaric oxygen is approximately 74 min.⁸ Given that our patient had been on oxygen for greater than 1 hour prior to her venipuncture in the emergency department, this theoretically should have decreased her carboxyhaemoglobin level, but since a blood sample was not collected in the field at the time of patient contact, the initial level is not known.

Another important variable to consider when making treatment decisions and interpreting carboxyhaemoglobin results is to take into account the interference from the hydroxocobalamin within the serum. Several studies have demonstrated that due to the red chromogenic feature of hydroxocobalamin and its light absorption at the wavelengths of 274 nm, 351 nm, 500 nm and 526 nm, certain spectroscopy laboratory results might not be accurate.^9–12 With respect to the carboxyhaemoglobin percentage reported for a given blood sample, the magnitude of hydroxocobalamin interference appears to vary among published studies and across diagnostic analytical instruments. Carlsson et al reported up to a 70% underestimation of the true carboxyhaemoglobin percentage, while Lee et al reported up to a 15% overestimation.^{10 11} This overestimation is statistically significant, but not clinically important due to the minimal difference between the real and reported values, as discussed by Baud, and therefore, should not change the clinician’s management.¹³ In a symptomatic patient with a known exposure to carbon monoxide, one should err on the side of treatment. The standard treatment for carbon monoxide toxicity is normobaric 100% oxygen therapy until the presenting symptoms resolve and the carboxyhaemoglobin level falls to 3% or less, with hyperbaric oxygen therapy recommended for carboxyhaemoglobin levels >25% or >15% during pregnancy.^{8 13} A reported carboxyhaemoglobin result that is an underestimation is a more significant issue as it could give the treating physician a false impression that the patient’s condition is better than it truly is, leading to early inappropriate discontinuation of normobaric or hyperbaric oxygen therapy.^{11 13} In the case of our patient, if the reported carboxyhemoglobin was indeed a 70% underestimate, similar to the finding by Carlsson et al, then the true carboxyhaemoglobin level may have been as high as 7%. This would certainly indicate the need for further 100% normobaric oxygen therapy.

Given the variable effect of serum hydroxocobalamin on the reliability of laboratory reported results and on the interference with measurements across different brands of medical laboratory analytical instruments, it is important for the clinician to understand that this chromogenic interference may occur and produce inaccurate results. Additionally, it is important to communicate to the laboratory technicians if a patient has received hydroxocobalamin in the past 2–3 days, in order to potentially minimise chromogenic interference bias.^{9 11} Because many conventional laboratories cannot directly measure hydroxocobalamin concentrations, Fueyo et al, proposed the use of the haemolysis index as a method to detect and evaluate the magnitude of hydroxocobalamin interference.¹² They found that the haemolysis index was positively correlated with the hydroxocobalamin concentration, thereby allowing the laboratory to alert clinicians of the possibility of chromogenic interference with certain test results, the magnitude of the suspected interference and if test results should be rejected.¹² While it is recommended that a blood sample be collected prior to administering hydroxocobalamin (for smoke inhalation victims), the administration of hydroxocobalamin should not be delayed in cases of suspected cyanide poisoning, though providers do need to be aware of the possible test result alterations, secondary to hydroxocobalamin chromogenic interference, when making subsequent treatment plans.^{4 11 13}

The IO route has become a common method for administration of medications due to the ability to obtain rapid reliable access in multiple sites independent of a patient’s haemodynamic status.¹⁴ However, there remains a paucity of published data regarding efficacy of IO administered antidotes in humans in situations of chemical poisoning. A systematic review published in 2017 demonstrated several human case reports and one case series of antidote administration for chemical poisoning via the IO route.⁵ With regard to hydroxocobalamin specifically, there has been only one human case abstract published in which a 9-month-old female received hydroxocobalamin intraosseously for suspected cyanide poisoning secondary to smoke inhalation.⁶ The authors of this case abstract note that the patient was discharged with no neurological or other sequelae. We are aware of no published details surrounding the efficacy or systemic absorption of IO administered hydroxocobalamin in humans. However, several animal models have demonstrated similar bioavailability of IO and IV administered hydroxocobalamin.^14–16 One may be able to, therefore, extrapolate that IO administered hydroxocobalamin would be bioavailable as an antidote in humans. As Petitpas et al suggest in their IO access systematic review publication, theoretically any medication that can be given intravenously can also be given intraosseously.¹⁷

In our patient, testing on hospital admission suggested evidence of systemic absorption of hydroxocobalamin after IO administration. The EMS physician, trauma surgery resident and emergency department attending physician explicitly noticed an odd purple–red colour to the first urine taken from the patient in the emergency department. The official laboratory report indicated that the urine colour was red. A red hue to the urine is an expected and reported side effect of hydroxocobalamin according to the package insert from the manufacturer.² Blood and urine hydroxocobalamin and cyanocobalamin levels were not obtained by the treating physicians at the hospital as it would not have altered management. Although we do not have those levels to report and cannot say with 100% confidence that the IO administered hydroxocobalamin was bioavailable to peripheral tissues and within the central circulation after infusion, our patient had no other potential aetiologies for the red urine, including haematuria, haemoglobinuria, myoglobinuria or other medications.

There are several cyanide antidote kits available, including a three-component kit containing sodium thiosulphate, sodium nitrite and amyl nitrite and a kit containing hydroxocobalamin. The latter is recommended for smoke inhalation patients over the former due to the creation of a more hypoxic state when haemoglobin is converted to methaemoglobin by the nitrites in the three-component kit. The cyanide antidote kit used for our patient contained a glass vial from which the hydroxocobalamin was infused into the patient using supplied vented IV tubing. Typically, when infusing a medication through an IO line, pressure must be maintained on the medication bag to keep the medication moving into the marrow space. One cannot place a pressure bag on or squeeze a glass vial to assist with increasing the pressure within the vial. Nonetheless, we witnessed that the antidote easily flowed from the glass vial through the supplied vented IV tubing into the proximal humerus by way of an IO line.

It is important to note, however, that the duration of the IO infusion exceeded the manufacturer’s recommendation that the first dose of hydroxocobalamin be administered over 15 min. According to patient’s EMS chart, the first dose was given IO at 15:27 and finished infusing during transport to the emergency department. Unfortunately, the completion time for the first infusion was not documented. Based on the reported scene departure time and hospital arrival time, the total infusion time for the first dose of hydroxocobalamin would have been between 23 min and 40 min. It is unclear if the extra 8–25 min resulted in a less fruitful treatment effect than would have been seen with a 15-minute infusion time. Of note, patient did begin to show some signs of neurologic improvement en route to the hospital in that she began to breath spontaneously, bite down onto the endotracheal tube and occasionally open her eyes. Although the IO infusion route used in our patient resulted in a slower infusion rate than recommended by the manufacturer, we feel that delaying treatment for suspected cyanide poisoning until IV access could be obtained at the hospital would be a disservice to the patient. One may question, why not transfer the diluted hydroxocobalamin to a normal saline bag, which would then allow for the application of a pressure bag for quicker infusion through an IO line? We do not recommend this approach. After querying English language publications within PubMed and reviewing the manufacturer’s prescribing information package insert, we could find no data to support this practice. Additionally, due to the chelating nature of hydroxocobalamin, it is possible that a percentage of the antidote would adhere to the normal saline bag and, therefore, result in some degree of under-dosing. Furthermore, there is the risk that aggressively transferring the diluted solution to a normal saline-bag could transform a portion of the solution into foam, rendering it then unable to be infused into a patient. Transferring the diluted hydroxocobalamin from the glass vial into a normal saline bag also wastes time, when one could be starting the infusion into the patient. Lastly, the hydroxocobalamin is designed to flow into the patient over 15 min from the glass vial via vented IV tubing. Using a normal saline bag with a pressure bag may result in too fast of an infusion, placing the patient at a theoretically increased risk for adverse effects such as significant hypertension (systolic blood pressure: >180 mm Hg or diastolic blood pressure: >110 mm Hg), headache, nausea or vomiting.¹²

The second dose of hydroxocobalamin was given as an IV piggyback at the hospital and infused over the course of 60 min. There is no documentation in this patient’s record that specifically explains the decision to infuse this dose over the course of 60 min versus the usual 15-minute infusion time. However, according to the manufacturer, the acceptable rate of infusion for a second dose may range from 15 min to 2 hours.¹⁸

In closing, based on the case abstract from Fortin et al, and our experience, it would seem reasonable that in the case of suspected cyanide toxicity and without quick IV access, that the antidote, hydroxocobalamin, could be initiated through an IO line.

Learning points.

• Acute cyanide and carbon monoxide poisoning should be considered when treating structure fire smoke inhalation patients.

• Hydroxocobalamin administration should be considered in fire smoke inhalation victims with altered mental status, dizziness, dyspnoea, seizure, shock or cardiac arrest.

• In an emergent situation when intravenous access is unattainable, intraosseous (IO) administration may be a viable option for the administration of hydroxocobalamin.

• Hydroxocobalamin can be readily infused from the CyanoKit glass vial via an IO route when using the supplied vented tubing.

• Providers should consider the possibility that hydroxocobalamin may alter laboratory results, such as carboxyhaemoglobin, due to the deep red hue chromogenic interference with spectroscopy analysis.