DOWNLOAD PDF

Abstract

Introduction

Methods

Discussion

References

SHARE

Cardiac Involvement during the Hospital Course of Carbon Monoxide Poisoning and Follow-up

 

Sébastien Champion, MD; Jérôme Aboab, MD; Raphaël Landau, MD; Nathanaël Lapidus, MD; Olivier Vignaux, MD, PhD; Mathilde Champion4, MD; Denis Duboc, MD, PhD; Djillali Annane1, MD, PhD

 

Service de Réanimation, Hôpital Raymond Poincaré, AP-HP, Université de Versailles Saint Quentin, Garches, France. Service d’anesthésie et réanimation, Hôpital de Bourg-en-Bresse, 01000 Bourg-en-Bresse, France. Université Pierre et Marie Curie-Paris 6, Paris, France. Service d’imagerie médicale et cardio-vasculaire, hôpital Cochin, Paris, France. Service de cardiologie, hôpital Cochin, Paris, France.

 

Abstract

Aims: description of cardiac involvement during carbon monoxide (CO) poisoning and its prognostic role in short and long term. Methods: this is a single center retrospective prognosis study, made from 2005 to 2009. We made a systematic definition of cardiac involvement for every adult patient addressed to our center in order to perform hyperbaric oxygen therapy. Patients with CO-related severe cardiac impairment were prospectively followed-up. Results: during 4 years, 393 patients were included. Thirteen percent were intubated, 4% received catecholamines, Simplified Acute Physiology Score (SAPS II) was 17 (+/- 16) and the mortality was 5.6%. Arrhythmias were seen in 16% and Brugada pattern in 2% (8 cases). Cardiac involvement was absent, moderate and severe in 76%, 10%, and 14% respectively. Severe cardiac involvement was associated with a 23% mortality rate (OR 11; p < 0.0001). In multivariate analysis, these associations were no more significant but SAPS II, Sequential Organ Failure Assessment, shock and neurological impairment remained independently associated with a worse prognosis. Ten patients had an initial cardiac arrest and 2 survived without neurological impairment. Late cardiac death accounted for 5/23 patients with severe cardiac involvement but cardiac magnetic resonance did not show any anomaly among survivors. Conclusions: Severe cardiac involvement is a frequent complication of CO poisoning leading to an 11 fold risk of death because of more severe poisoning and/or neurological impairment (J clin invasive cardiol 2015;2:15-22 ).

Key Words: poisoning, carbon monoxide, cardiotoxicity, Brugada, hyperbaric oxygen therapy, cardiac magnetic resonance.

 

Introduction

Carbon monoxide (CO) poisoning is the principal cause of toxic death in France. A recent study reports 4200 intoxications in 2007 (without smoke inhalation from fire exposures) with 300 yearly deaths.1 Hyperbaric oxygen therapy (HBOT) is indicated in case of neurological involvement in order to prevent neurological sequelae and should be performed in 2500 cases each year in France.2,3 Since it has been described for at least a century, cardiac involvement is incompletely understood.4 Cardiac manifestations occur in about 35% of moderate to severe poisoning,5 and appear as a powerful prognosis factor for short,6 and long term outcome.7 We describe herein the cardiac involvement after carbon poisoning, during the hospital course of and the follow-up.

 

Methods

We conducted a cohort study of consecutive adult patients admitted to the intensive care unit of Hospital Raymond Poincaré, between junuary 2006 and december 2009 for CO poisoning. The study protocol is in accordance with the Declaration of Helsinki and was approved by our Ethics Committee. Data analysis was performed retrospectively for hospital stay and prospectively for follow-up. All patients of both sexes, victim of CO poisoning, and admitted in the hyperbaric center in the 24 hours following the exposure were included. All admitted patients received normobaric oxygen therapy, started since the extraction from the area of CO exposure. The indications for HBOT were defined according to the 7th conference of European consensus recommendations validated by the French High Authority of Health in 2007.3,8 These authorities recommend the realization of a hyperbaric oxygen therapy in CO poisoned patients who had a loss of consciousness or an organ failure (persistent coma, cardiogenic shock). It is important to note that the CO score does not include cardiovascular symptoms or findings. Pregnant women were systematically treated, taking into account the risk for the fetus. The HBOT session lasted 2 hours (including 1 hour at 2 ATA) in a multiplace chamber (Comex Pro®) with a fraction of inspired oxygen of 100%. Before HBOT, 10 mg intramuscular diazepam was injected in order to prevent hyperoxic seizure.9 From each patient were collected: age, sex, and simplified acute physiology score II (SAPS II) at the time of admission to the intensive care unit (ICU).10 The cardiovascular history was collected. The source of the intoxication and the severity CO score (0: no symptoms, 1: discomfort, tiredness, headache, 2: nausea, vomiting, asthenia, giddiness, 3: transitory loss of conscience quickly reversible, 4: Organ failure (coma, shock, respiratory distress), 5: death) were also collected.11 The following events were assessed daily during the course of the hospitalization: need for tracheal intubation, the use of vasoactive drugs, acute respiratory distress syndrome (ARDS), renal failure, hepatic failure, and coagulopathy. ARDS was defined according to the criteria of the American-European consensus conference.12 Shock was defined according to the criteria of the American College of Chest Physicians-Society of Critical Care Medicine consensus conference.13 Renal, hepatic, coagulopathy, and respiratory failure were defined to allow calculation of the Single Organ Failure Assessment (SOFA) score.14 After HBOT, neurological impairment was reported following the Glasgow Outcome Scale (GOS). It was graded as absent, minimal (GOS 4 or 5) or severe (GOS 3 or less).15 The first arterial blood gas measurement, the arterial lactate rate, the rate of HbCO, the liver, kidney and hemostasis assessment, were recorded. For all patients chest radiography was performed and a radiological score was calculated.16 An electrocardiogram (ECG) was also performed for all patients upon admission and a troponin or creatine kinase (CK) assay if needed, based on clinical judgment. Each ECG was reviewed by a cardiologist (SC) and interpreted by a standard way.17 A score of the cardiac involvement was defined leading to three groups: severe, moderate or no cardiac involvement. Severe myocardial involvement was arbitrary defined by a troponin Ic >1 ng/ml, or CK > 800 UI/L without muscular sign, or extensive ECG ischemic pattern. Extensive electrical ischemia was defined by two or more ischemic territories (anterior, lateral or inferior). Moderate cardiac involvement was defined by troponin above threshold (99th percentile or troponin Ic >0.1 ng/ml in our laboratory as the definition of myocardial infarction),18 or CK >270 UI/L without muscular sign, or any ECG ischemic pattern. The others patients were considered without cardiac involvement. Though, moderate and severe patients were considered with cardiac involvement. Every cardiac clinical manifestation (chest pain and pulmonary edema) or explorations were recorded (echocardiography, coronary artery angiography). Prognosis parameters (mortality in the ICU, length of stay in intensive care and duration of mechanical ventilation) were also collected. All patients were followed up until ICU discharge and patients with severe cardiac involvement were followed up after discharge by a phone call. Functional status was recorded and a cardiologic consultation with cardiac magnetic resonance (CMR) was proposed, then patients gave informed consent. Statistical analysis Data were expressed as mean and standard deviation, median and interquartile range for continuous variables as appropriate, and as number and percentage for binary variables. Comparisons were made using Student’s t or Mann–Whitney tests for continuous variables (as appropriate), and chi-square or Fischer’s exact test (as appropriate for binary variables). Variables selected by the univaried analyses were evaluated by using logistic regression models to estimate the odds ratio of dying along with the 95% confidence interval (CI). Variables with significant association with mortality were included in the multivariate analysis. Statistical analysis was performed with R version 2.9.2 (R Development Core Team, Vienna, Austria, 2010). Values of p < 0.05 were considered to be statistically significant.

Results

During 4 years, 393 adult patients were admitted in our institution to perform a HBOT for CO poisoning. Initial description of the population is summarised in table 1 and follow-up in figure 2.

Table 1

Figure 1

The indication of HBOT was pregnancy in 22%. These women had mild intoxication without any complication or fatality. None of them had abnormal cardiac findings (ECG, biomarkers or echocardiograms). Sixty percent had domestic (pure) intoxication (236 patients), 29% inhaled smoke from fire exhaust (114 patients) and the last 11% had intentional poisoning by car exhaust inhalation (43 patients). We collected 29 co-intoxications in the intentional CO poisoning. The other toxics were almost always alcohol or benzodiazepines. Only 7 patients had taken drugs that could interfere with cardiac involvement (various cardiotoxicants). The medical support and evolution are summarised in table 1. The hospital course was quite short because patients were transferred to the previous hospital. HBOT could not be performed in 16 cases: 12 patients had ear-pain with only 1 proven pressure-related otitis, 1 claustrophobia, 2 hemodynamic instability leading to a quick death and 1 unavailability of the hyperbaric box. No seizure or adverse event was noticed. Hydroxocobalamine (Cyanokit®) is the French antidote of cyanide poisoning and was used in 56 patients: 93 % of the smoke inhalation with hemodynamic compromise or lactic acidosis (maximum 18.7 mmol/L). The neurological impairment was considered minimal (nystagmus, depressive trend, deep tendon reflex anomaly leading to a GOS at 4 or 5) in the survivors but severe (GOS 2 or 3 before death) in every fatal case. The patients could not be followed by a neurologist to look for the delayed neurological sequelae. ECG findings The ECGs were interpreted in all medical form and were reviewed in 94%. Slight abnormalities were noticed in more than 35% of the ECG (incomplete right bundle-branch block, early repolarization or straight ST-segment). Brugada pattern was noticed in 8 patients in whom 3 were type 1 Brugada. Only 3 patients had to be treated immediately for rhythm disturbance. A patient with rapid ventricular tachycardia received emergent direct current cardioversion and 2 patients with bradycardia (sinus bradycardia and complete heart block) received immediate care. Heart rate, QRS and QT corrected durations were associated with arrhythmia and death. Cardiac involvement Cardiac involvement was moderate in 39 patients (10%), severe in 56 patients (14.2%), and absent in 298 patients (75.8%). Cardiac involvement elements (without ECG) are summarised in table 2.

Figure 1

Figure 1

Six patients had atheromatous known past medical history and 8 complained from cardiac symptoms (7 chest pain and 1 pulmonary edema). Troponin Ic was measured in 36% of the patients and reached a maximal of 100.8 ng/ml. It was >0.1 ng/ml in 34 % and >1 ng/ml in 17% of the measures. Past medical atheromatous history was slightly associated (p = 0.048) with cardiac involvement, ECG ischemia, arrhythmia or death. Left ventricle ejection fraction value was associated with initial lactates, troponin and cardiac arrest (p < 0.01). The same associations were found in the sub-groups of smoke inhalation and domestic CO patients. Prognosis analysis The univariate prognosis analysis is summarised in table 3. HbCO and expired CO measures, HBOT, co-intoxication, hospital stay and mechanical ventilation lengths were not associated with death. Initial cardiac arrest, shock, moderate or severe neurological impairment, intubation, ARDS, smoke inhalation, SOFA and Murray scores, blood creatinine, bicarbonates and hepatic enzymes were associated with increased mortality (p < 0.01). Pregnancy was associated with a better outcome. The cause of death divided as follows: cardiogenic shock in 6 (27%), secondary cardiac arrest in 3 (14%), ARDS in 2 (9%), unknown in 5 (23%) and withdrawal of care (mainly due to severe neurological impairment) in 6 (27%) of the 22 deaths. In the smoke inhalation sub-group, mortality was 17.5% (20/114). The same data were associated with death in the univariate or multivariate analysis. In the domestic CO sub-group, only corrected QT interval prolongation was associated with death (survivors: QT = 375 ms; non- survivors QT = 440 ms; p = 0.038). The mortality was 1% (2/236) in this population with pure, less severe intoxications. Moderate cardiac involvement was not associated with a worse outcome. Any cardiac involvement (moderate and severe) and severe cardiac involvement were associated with a mortality of 15.8 % and 23.2 % respectively. The Odd's Ratio of death were 7.8 (CI 3-20) and 11 (CI 4-27) respectively. Cardiac involvement was no more associated with death in the sub-groups of mechanically ventilated patients or not ventilated. Cardiac involvement was strongly associated with minimal or severe neurological impairment (p < 0.01). Within the smoke inhalation sub-group, troponin Ic had a strong prognosis value but was measured in only 44% (50/114) of the patients. A threshold of 3 ng/ml could indicate 100% mortality. The area under the ROC curve was 0.851 (CI 95%: 0.724-0.979) (figure 1 and 1bis). Follow-up Three years follow-up (figure 2) found 8 deceased patients (5 heart diseases after 1, 7, 22, 24 and 50 months ) and 15 alive patients, from 42 patients with initial severe CO-related cardiac impairment. Twelve had normal life and 3 patients complained from cardiac symptoms (1 myocardial infarction 28 months after poisoning, 2 heart failure class II and III of NYHA), 36 (+/-18) months after poisoning. Unfortunately, 19 were lost of follow-up and 11 refused the cardiologic examination with CMR. CMR was performed in four asymptomatic patients with severe ischemic ECG pattern and enzymatic elevation (troponin Ic = 6 and 9 ng/ml for two patients and CK 30000 and 44000 for the two others), 3 years (31, 31, 42 and 50 months) after poisoning. CMR did not show any anomaly. In different multivariate models (see Appendix, table 4), cardiac involvement (any or severe) was not associated with death. A worse outcome was found with higher age, higher SAPS2 or SOFA score, lower paO2/FiO2 ratio or Coma Glasgow Score (CGS), shock, cardiac arrest, neurological impairment. Smoke inhalation was not associated with highest death rate but a trend (p = 0.07). There was no severe burn patient and hydroxocobalamine (Cyanokit®) administration was not associated with better outcome. Severe neurological impairment and total SOFA scores were the most strongly associated with death in different multivariate models.

Figure 1

Discussion

Cardiac involvement during CO poisoning is known for at least a century but has been thoroughly explored only recently.4 Table 5 (see Appendix) summarises the literature data about cardiac involvement during CO poisoning. The thresholds for biologic measurements are different and ECG findings are not standardised. Although arbitrary, our definition of moderate and severe cardiac involvement obeys to international guidelines and is reproducible.18 Unfortunately our analysis is retrospective and data are lacking: only 36% of the patients were screened for troponin Ic and very few had an echocardiography which shows abnormalities in up to 35% of severely poisoned patients in large series.19 In the present study, univaried analysis showed an association between cardiac involvement and death, but it was no more significant in the multivariate analysis. One may hypothesize that cardiac involvement is associated with more severe CO poisoning and leads to death by neurological impairment. The latter is strongly independently associated with death. This is questionable because the cause of death was cardiogenic shock, secondary cardiac arrest, or pulmonary edema in most of the cases and withdrawal of care because of severe neurological impairment in 27% of the deaths. The prolongation of the QT interval has previously been described during CO poisoning.20 Our study shows for the first time that QTc prolongation is a valuable prognostic indicator during domestic (pure) CO poisoning. Careful examination of the ECG is mandatory during CO poisoning. In order to limit cardiac death, attention must be paid with catecholamines and anti-arrhythmic drugs. The mechanism of CO cardiotoxicity is partly due to ischemia (which could be reversed by HBOT) and partly due to catecholamine cardiotoxicity.21 A description of Takotsubo cardiomyopathy secondary to CO poisoning illustrate this well-known experimental toxicity of the catecholamines.22 In case of cardiogenic shock, non bêta-1 stimulating inotropes such levosimendan should be discussed. The death usually follows severe bradycardia so anti-arrhythmic drugs should be used with caution. The presence of 8 cases (2%) and 3 cases of type 1 (0.8%) Brugada pattern is much higher than expected in the population (less than 1/1000) and could not be attributable to other medications. A sodium channel slur has never been noticed even in experiments in cardiac cell. A membrane-stabilizing effect can even be hypothesised to explain some of our described data. HBOT seems to be efficient in decreasing neurologic sequelae even if this effect is debated.23 The efficiency of HBOT to decrease cardiac involvement during CO poisoning is far from proven. Hyperoxic state is associated with worse outcome in the ischemic-reperfusion model of cardiac arrest and in the intensive care unit.24,25 But the association of hyperbaric and hyperoxic state for the infarcted myocardium do not seem to lead to worse (or better) outcome.26 To prove the evidence of the efficiency of HBOT in isolated cardiac involvement, a randomised prospective study with more than 10000 patients should be performed in order to decrease significantly the relative mortality rate about 5%. This study is far from being done and a particular attention should be paid on association between cardiac and neurological impairment in our study. Even if the patients were not screened for neurological sequelae, the cardiac involvement seen in more severe CO poisoning can indicate unseen neurological involvement. On that basis, treating with HBOT isolated cardiac involvement is still debated.9 Follow-up A major recent discovery is the possible presence of a cardiac “scar” of CO poisoning with cardiac involvement. The “scar” could be a fibrotic lesion better seen on CMR by a late Gadolinium enhancement (LGE).27,28 By that way, cardiac involvement could lead to a 3-fold late death risk.7 That kind of cardiac lesion is best treated with beta-blockers (BB) and angiotensin converting enzyme (ACE) inhibitors during ischemic or dilated cardiomyopathy. The screening of that kind of lesion can be a reasonable surrogate marker to prove the efficiency of HBOT during CO poisoning and indicate eventually BB or ACE inhibitors. We did not observe any image evocative of myocardial fibrosis in patients with severe initial cardiac involvement however with many loss of follow-up. No CMR was performed before 31 months and earlier CMR could have screened such abnormality. As late cardiac mortality was a concern in our and other studies,7 follow-up is mandatory for these patients with particular attention to cardiac symptoms. Prognosis In our study, smoke inhalation was associated with death only in univaried analysis. This could be attributed to more severe poisonings, different toxics in the smoke or skin or lung burns. Cyanokit® was largely prescribed and had no effect on mortality. The specific cardiac or neurologic manifestations of CO, cyanide or other gas could not be differentiated in our study.29 In our study, in case of initial cardiac arrest, the mortality was 80% with 20% of recovery without sequelae. Those results are encouraging and make the resuscitation of that kind of cardiac arrest a care far from futile. The 100% mortality reported by Hampson is challenged.30 In case of smoke inhalation, 100% mortality was seen if the troponin was over 3 ng/ml but in a burden of 50 patients. Troponin may be a useful prognosis tool during smoke inhalation. Limitations This retrospective study has many limitations. Only a part of the patients had cardiac investigations and the impact of the treatments could not be investigated. Then follow-up is only parcel and late after poisoning. Finally, our data are from single centre and only hypothesis generating.

 

Conclusion

Cardiac involvement is a frequent and serious complication of CO poisoning accounting for one quarter of patients addressed to a hyperbaric centre for neurological reasons. It is associated with an 11 fold risk of death (mainly cardiac deaths) in case of severe cardiac involvement but this association is no more significant in multivariate. The association between more severe CO poisoning and neurological impairment, a well-known prognosis factor, can explain the lack of significance in different multivariate models. Serial ECG and a troponin dosage should nevertheless be performed to define prognosis for short and long term after moderate to severe CO poisoning. Late cardiac death might occur after such poisonings but we did not find any cardiac anomaly on CMR during follow-up.

 

Conflicts of interest

The author has no financial disclosures to declare and no conflicts of interest to report.

 

References

1. Verrier A, Delaunay C, Coquet S, et al. Les intoxications au monoxyde de carbone en France métropolitaine en 2007. InVS. BEH n°1. 12 janvier 2010. p 1-5. www.invs.sante.fr/beh/2010/01/beh_01_2010.pdf last accessed 30/07/14 2. Weaver LK. Carbon Monoxide Poisoning. N Engl J Med 2009;360:1217-1225. 3. Biga J. Rapport de la Haute Autorité de Santé concernant l’oxygénothérapie hyperbare. 2007. 4. Gandini C, Castoldi AF, Candura SM, et al. Carbon monoxide cardiotoxicity. J Toxicol Clin Toxicol 2001;39:35–44. 5. Satran D, Henry CR, Adkinson C, et al. Cardiovascular manifestations of moderate to severe carbon monoxide poisoning. J Am Coll Cardiol 2005;45:1513–1516. 6. Kao HK, Lien TC, Kou YR, et al. Assessment of myocardial injury in the emergency department independently predicts the short-term poor outcome in patients with severe carbon monoxide poisoning receiving mechanical ventilation and hyperbaric oxygen therapy. Pulm Pharmacol Ther 2009;22(6):473-7. 7. Henry CR, Satran D, Lindgren B, et al. Myocardial injury and long-term mortality following moderate to severe carbon monoxide poisoning. JAMA 2006;295:398–402. 8. Mathieu D (2004). 7th conference of European consensus recommendations. http://www.echm.org/documents/ECHM%207th%20Consensus%20Conference%20Lille%202004.pdf last accessed 30/07/14 9. Annane D, Chadda K, Gajdos P, et al. Hyperbaric oxygen therapy for acute domestic carbon monoxide poisoning: two randomized controlled trials. Intensive Care Med 2011;37(3):486-92. 10. Le Gall JR, Neumann A, Hemery F, et al. Mortality prediction using SAPS II: an update for french intensive care units. Crit Care 2005;9(6):R645-52. 11. Cevik AA, Unluoglu I, Yanturali S, et al. Interrelation between the Poisoning Severity Score, carboxyhaemoglobin levels and in-hospital clinical course of carbon monoxide poisoning. Int J Clin Pract 2006;60(12):1558-64. 12. Bernard GR, Artigas A, Brigham KL, et al. The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 1994;149(3 Pt 1):818-24. 13. Muckart DJ, Bhagwanjee S. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference definitions of the systemic inflammatory response syndrome and allied disorders in relation to critically injured patients. Crit Care Med 1997;25(11):1789-95. 14. Vincent JL, de Mendonça A, Cantraine F, et al. Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on "sepsis-related problems" of the European Society of Intensive Care Medicine. Crit Care Med 1998;26(11):1793-800. 15. Kaye AH, Andrewes D. Glasgow Outcome Scale: research scale or blunt instrument? Lancet 2000;356(9241):1540-1541. 16. Offner PJ, Moore EE. Lung injury severity scoring in the era of lung protective mechanical ventilation: the PaO2/FIO2 ratio. J Trauma 2003;55(2):285-9. 17. Wagner GS, Macfarlane P, Wellens H, et al.; American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; American College of Cardiology Foundation; Heart Rhythm Society. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part VI: acute ischemia/infarction: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol 2009;53(11):1003-11. 18. Thygesen K, Alpert JS, White HD, et al. Universal definition of myocardial infarction. Circulation 2007;116(22):2634-53. 19. Kalay N, Ozdogru I, Cetinkaya Y, et al. Cardiovascular Effects of Carbon Monoxide Poisoning. Am J Cardiol 2007;99:322–324. 20. Yelken B, Tanrıverdi B, Cetinbaş F, et al. Carbon monoxide and QT intervals. Ana do lu Kar di yol Derg 2009;9:397-400. 21. Fineschi V, Agricola E, Baroldi G, et al. Myocardial findings in fatal carbon monoxide poisoning: a human and experimental morphometric study. Int J Legal Med 2000;113:276 –282. 22. Kawashima T, Okashiro M, Ishii N, et al. Can magnetic resonance imaging of the brain in the acute phase of carbon monoxide intoxication forecast delayed encephalopathy? Chudoku Kenkyu 2007;20(2):117-24. 23. Buckley NA, Juurlink DN, Isbister G, et al. Hyperbaric oxygen for carbon monoxide poisoning. Cochrane Database Syst Rev 2011;4:CD002041. 24. Kilgannon Jh, Jones AE, Shapiro NI, et al., for the Emergency Medicine Shock Research Network (EMShockNet) Investigators. Association Between Arterial Hyperoxia Following Resuscitation From Cardiac Arrest and In-Hospital Mortality. JAMA 2010;303:2165-2171. 25. De Jonge E, Peelen L, Keijzers PJ, et al. Association between administered oxygen, arterial partial oxygen pressure and mortality in mechanically ventilated intensive care unit patients. Crit Care 2008;12(6):R156. 26. Stavitsky Y, Shandling AH, Ellestad MH, et al. Hyperbaric oxygen and thrombolysis in myocardial infarction: the 'HOT MI' randomized multicenter study. Cardiology 1998;90(2):131-6. 27. Leroy A, Cuisset T, Quilici J, et al. An unusual “winter case” of myocarditis. Int J Cardiol 2011;148(3):e68-9. 28 - Henry TD, Lesser JR, Satran D. Myocardial Fibrosis From Severe Carbon Monoxide Poisoning Detected by Cardiac Magnetic Resonance Imaging. Circulation 2008;118;792. 29. Baud FJ, Barriot P, Toffis V, et al. Elevated blood cyanide concentrations in victims of smoke inhalation. N Engl J Med 1991;325(25):1761-6. 30. Hampson NB, Zmaeff JL. Outcome of patients experiencing cardiac arrest with carbon monoxide poisoning treated with hyperbaric oxygen. Ann Emerg Med 2001;38(1):36-41. 31. Dueñas-Laita A, Burillo Puzte G, Nogué Xarau S, et al. Cardiovascular manifestations of carbon monoxide poisoning. J Am Coll Cardiol 2006;47(3):690-1.