Author + information
- Received January 22, 2018
- Revision received March 8, 2018
- Accepted April 10, 2018
- Published online August 20, 2018.
- David Sparv, PhDa,∗ (, )
- Robin Hofmann, MD, PhDb,
- Annika Gunnarsson, BScc,
- Stefan James, MD, PhDd,
- Camilla Hedberg, BSce,
- Jörg Lauermann, MDe,
- Petronella Torild, BScf,
- Elmir Omerovic, MD, PhDf,
- Kristina Bergström, BScg,
- Espen Haugen, MD, PhDh,
- Camilla Bergström, BSci,
- Rikard Linder, MD, PhDi,
- Pia Borg, BScj,
- Urban Haaga, BSck,
- Anneli Olsson, BScl,
- Elin Böving, BScm,
- Ollie Östlund, PhDn,
- Rebecca Rylance, MSca,
- Nils Witt, MD, PhDb,
- David Erlinge, MD, PhDa,
- for the DETO2X-SWEDEHEART Investigators
- aDepartment of Cardiology, Clinical Sciences, Lund University, Lund, Sweden
- bDepartment of Clinical Science and Education, Division of Cardiology, Karolinska Institutet, Södersjukhuset, Stockholm, Sweden
- cDepartment of Cardiology, Uppsala University Hospital, Uppsala, Sweden
- dDepartment of Medical Sciences and Uppsala Clinical Research Center, Uppsala University, Uppsala, Sweden
- eDepartment of Internal Medicine, Division of Cardiology, Ryhov Hospital, Jönköping, Sweden
- fDepartment of Cardiology, Sahlgrenska University Hospital, Gothenburg, Sweden
- gDepartment of Cardiology, Gävle Sjukhus, Gävle, Sweden
- hDepartment of Cardiology, Sundsvall Regional Hospital, Sundsvall, Sweden
- iDepartment of Cardiology, Danderyd University Hospital, Stockholm, Sweden
- jSvensk PCI Värmland AB, Karlstad, Sweden
- kDepartment of Cardiology, Central Hospital Karlstad, Karlstad, Sweden
- lDepartment of Cardiology, Skane University Hospital, Lund, Sweden
- mDepartment of Cardiology, Södersjukhuset, Stockholm, Sweden
- nUppsala Clinical Research Center, Uppsala University, Uppsala, Sweden
- ↵∗Address for correspondence:
Dr. David Sparv, Department of Cardiology, Lund University, Skane University Hospital, SE-221 85 Lund, Sweden.
Objectives In this substudy of the DETO2X-AMI (An Efficacy and Outcome Study of Supplemental Oxygen Treatment in Patients With Suspected Myocardial Infarction) trial, the authors aimed to assess the analgesic effect of moderate-flow oxygen supplementation in patients with suspected acute myocardial infarction (AMI) treated with percutaneous coronary intervention (PCI) and to study the effect of oxygen supplementation on the use of opiates and sedatives during PCI.
Background Routine oxygen in normoxemic patients with AMI does not provide clinical benefit. However, oxygen may relieve ischemic pain.
Methods Patients were randomly allocated to oxygen or ambient air according to the main study protocol. After PCI, peak level of pain during PCI was measured by the Visual Analogue Scale. The total amount of opiates and sedatives was reported.
Results A total of 622 patients were enrolled: 330 in the oxygen group and 292 in the ambient air group. There was no significant difference in peak level of pain (oxygen 4.0 [1.0 to 6.0] vs. air 3.0 [0.6 to 6.0]; p = 0.37), use of opiates (mg) (oxygen 0.0 [0.0 to 3.0] vs. air 0.0 [0.0 to 3.0]; p = 0.31), or use of sedatives between the groups (median [interquartile range]) (oxygen 2.5 [0.0 to 2.5] vs. air 2.5 [0.0 to 2.5]; p = 0.74).
Conclusions In the present study, the authors did not find any analgesic effect of routine oxygen as compared with ambient air, and no differences in the use of sedatives and opiates during PCI. Our results indicate that moderate-flow oxygen supplementation does not relieve pain in normoxemic patients with suspected AMI undergoing treatment with PCI and should thus not be used for this purpose.
The use of routine oxygen therapy in acute myocardial infarction (AMI) has recently been evaluated in several studies (1–3). Even though the use of oxygen in the treatment of hypoxemia is well established (4), the evidence in other indications remain sparse (5). The potential hazard of hyperoxemia has been investigated, and the risk-benefit ratio of routine oxygen remains unclear (6–8). A common use for oxygen in AMI is to relieve pain (5,9). Ischemic pain represents a visceral pain sensation primarily mediated by chemosensitive nociceptors in the heart muscle (10). In the presence of myocardial ischemia, a number of chemical mediators are released, which triggers the production of prostaglandins through a cyclooxygenase pathway (11). The chemical response stimulates nociceptors, mainly located in the epicardium, which results in a depolarization of the cardiac visceral spinal afferent fibers (12,13).
Oxygen has been used to relieve ischemic symptoms for more than 100 years, which was the very first indication of oxygen therapy in suspected AMI (14). The rationale behind the use of oxygen as an analgesic agent is that oxygen is considered to improve oxygenation of the ischemic heart muscle and by this, diminish the effects of the chemical and mechanical response to ischemia (6). A small number of studies used the amount of analgesic pharmaceuticals as a proxy variable for pain estimation (15,16), but to our knowledge, only 2 randomized clinical trials have conducted a direct measurement of oxygen therapy in pain management, where the most recent study only measured pain as a screening tool for pre-hospital inclusion before intervention (1). The other trial was our previous study, the randomized, double-blind, placebo-controlled OXYPAIN (Analgesic effect of OXYgen during PercutAneous coronary INtervention) trial (17), in which 300 patients with coronary artery disease were treated by percutaneous coronary intervention (PCI). In this trial, oxygen was not superior to air in pain reduction. However, this was a single-center trial in which patients with ST-segment elevation myocardial infarction (STEMI) were excluded. In the present substudy, we aimed to assess the analgesic effect of moderate-flow oxygen supplementation in a cohort of patients undergoing treatment with PCI and to study the effect of oxygen supplementation on the use of opiates and sedatives during PCI.
The present study was a pre-specified substudy of the DETO2X-AMI (An Efficacy and Outcome Study of Supplemental Oxygen Treatment in Patients With Suspected Myocardial Infarction) trial (18). Patients with suspected AMI included in the main trial were considered eligible for inclusion if PCI was carried out within the time frame of the main study intervention. All Swedish cardiology departments with capacity for primary PCI and part of the main study (n =14) were asked for participation. Eight high-volume centers accepted.
Ethics approval and consent to participate
The main study was approved by the ethical review board in Gothenburg (DNR 287-12) as well as by the Swedish Medical Products Agency (MPA) (EudraCT 2013-002882-20) and was registered at clinicaltrials.gov (NCT01787110). This substudy was approved by the ethical review board in Gothenburg as an amendment to the main study (DNR T510-14). All included patients gave oral consent followed by written confirmation within 24 h as described in the main publication (18).
Procedure and data collection
All patients followed the protocol of the main study (19). In the present cohort, 70 to 100 U/kg heparin was administered by an intra-arterial injection directly after the insertion of the arterial sheath. Oxygen saturation was measured continuously during the intervention by pulse oximetry. Five minutes after removal of the guiding catheter from the arterial sheath, the patients were asked to estimate the peak level of chest pain during the time span of the intervention, defined as the period from insertion to removal of the arterial sheath. The Visual Analogue Scale (VAS), a well-established and validated tool for determination of pain/discomfort (20–22), was used to estimate the level of pain. Furthermore, the total amount of opiates and sedative agents administered during the study period was reported.
Sample size and statistical analysis
Based on previous VAS data from the OXYPAIN study (17), an estimated sample size of 150 in each group would result in 80% power with a significance level of 0.05 (2-tailed) to detect a relative difference of 15% between the study groups. Because we also included STEMI patients with possibly more severe pain and thus more variation in VAS score (23), we decided to increase the sample size to 250 in each arm. VAS was considered a continuous variable. VAS and the amount of analgesic agents were calculated and reported as median with interquartile range. The comparison between the supplemental oxygen and ambient air groups was computed by the Mann-Whitney U test (Wilcoxon rank-sum test). Categorical variables were analyzed using the chi-square test or Fisher exact test as appropriate. A p value of <0.05 was considered statistically significant. Explorative subgroup analyses were performed in the STEMI, non–ST-segment elevation myocardial infarction (NSTEMI), and sex groups, using the same statistical calculations/methods.
Patient characteristics and procedural data
Between December 1, 2014, and December 29, 2015, a total of 624 patients were enrolled (9.4% of the main study population and 14.7% of the main study population treated by PCI) (18). Of the 622 patients, 330 were randomized to receive supplemental oxygen at a moderate flow rate and 294 to ambient air. In the ambient air group, 2 patients were excluded from the analysis due to missing data (1 patient with a gastrointestinal bleeding without evidence of ischemia and 1 with inaccurate personal identification number). Five patients in the oxygen group and 21 in the ambient air group received additional oxygen due to development of hypoxemia. Thirteen additional patients did not complete the study intervention for miscellaneous reasons (Figure 1). A total of 594 patients (95%) were enrolled because of chest pain, and of those, 582 (93.6%) were diagnosed with myocardial infarction (Table 1). The most common finding on coronary angiography was 1-vessel disease (34.4%) involving the left anterior descending coronary artery as the culprit lesion. Ninety-four percent of the patients were classified as Killip class 1 (24), and interventional success rates were similar between the groups (Table 2).
There was no significant difference in the peak level of pain measured by VAS between the groups (median [interquartile range]: oxygen 4.0 [1.0 to 6.0] vs. air 3.0 [0.6 to 6.0]; p = 0.37) (Figure 2). Subgroup analyses of peak level of pain were performed in the groups of STEMI, NSTEMI, men, and women. There were no significant differences in any of the subgroup analyses (Table 3). Among all patients, 225 (36.1%) received opiates. The median dosage (mg) did not differ between the groups (oxygen 0.0 [0.0 to 3.0] vs. air 0.0 [0.0 to 3.0]; p = 0.31). Six hundred and twenty-one patients (99.8%) received sedatives (diazepam). There was no difference in dosage (mg) between the groups (oxygen 2.5 [0.0 to 2.5] vs. air 2.5 [0.0 to 2.5]; p = 0.74 (Table 3).
In this clinical trial, we assessed the analgesic effect of moderate-flow oxygen supplementation in patients with suspected AMI and normal oxygen saturation. Our results demonstrate no significant differences in the peak level of pain between the groups randomized to receive oxygen or ambient air, a finding consistent with previous studies (1,5,15–17). Furthermore, there were no differences in the use of opiates or sedatives between the groups.
The role of oxygen in the treatment of ischemic pain has been studied since the early 20th century (14,25,26). Existing evidence has been mainly empirical from small case series, based on the assumption that in the case of impaired oxygen supply in the myocardium, supplemental oxygen would increase oxygen delivery to the ischemic tissue by collateral circulation and thus, decrease ischemic pain (27,28). Similarly, oxygen therapy was believed to decrease myocardial injury and thereby reduce the risk of arrhythmias and, ultimately, morbidity and mortality (29). This assumption was further supported by animal models (30,31), studies of hyperbaric oxygen (32,33), and nonrandomized trials conducted before the era of myocardial revascularization (34,35). Russek et al. (36) showed in 1950 that hyperoxemia causes a prolonging of the ischemic characteristics of the electrocardiogram, and does not relieve ischemic pain in AMI. Moreover, routine oxygen in normoxemic patients might be associated with potentially harmful effects that can worsen myocardial ischemia (1,37). The effects are considered to derive from the vasoactive properties of oxygen and caused by inadvertent hyperoxemia (8). Different underlying mechanisms have been proposed: increased production of reactive oxygen species decreasing the bioavailability of nitric oxide leading to subsequent vasoconstriction (38,39), direct vasoconstriction caused by closure of ATP-sensitive K+-channels (40), direct effects on L-type Ca2+-channels (41), increased production of the vasoconstrictor metabolite 20-HETE (42), and an increased engagement of the angiotensin I/II system (43). Furthermore, vasoconstriction and the subsequent hemodynamic responses may be associated with aggravated ischemia, as demonstrated by increased lactate levels during hyperoxemia (28,44). Even though the exact mechanisms are yet to be determined, the vasoactive capacity of oxygen is beyond doubt. On the basis of theoretical data, hyperoxemia could actually result in a more severe pain experience (7). However, our study demonstrates that supplemental oxygen at a moderate flowrate does not increase or decrease ischemic pain during AMI.
The estimation of pain remains a difficult task in the diagnostics and treatment of coronary artery disease (45). Nevertheless, in the recently updated European Society of Cardiology guidelines for AMI presenting with ST-segment elevation (46), the authors emphasize the importance of pain relief. Pain is associated with an increased sympathetic activity with a subsequent vasoconstriction and increased workload for the heart, and thus is important to manage. In the guidelines, titrated intravenous opiates and sedatives in anxious patients are recommended for pain relief (Class IIa, Level of Evidence: C). Routine oxygen is not recommended for patients with an arterial oxygen saturation (SaO2) ≥90% (Class III, Level of Evidence: B) due to the possible harmful effects of hyperoxemia. In hypoxemia, defined as a partial arterial oxygen pressure of <60 mm Hg or an arterial SaO2 of <90%, supplemental oxygen is recommended (Class 1, Level of Evidence: C). In addition, the American College of Cardiology Foundation/American Heart Association guidelines (47) state that oxygen is appropriate in SaO2 <90%, and recommend that oxygen be used with caution due to the vasoconstrictive properties. Hence, our present results are in agreement with current clinical guidelines (46,47).
Our main study, the DETO2X-AMI trial (18), did not demonstrate any significant beneficial effect of oxygen as compared with ambient air on all-cause mortality at 1 year, infarct size measured by troponin T release, or on the incidence of rehospitalization with myocardial infarction. The results from this substudy further strengthen the recommendation that it is safe and reasonable to withhold oxygen from patients with suspected AMI and normal oxygen saturation, and that supplemental oxygen at a moderate flowrate is not associated with an analgesic effect. Also, combined with the results from our previous publication (17), the present findings are consistent in stable angina, unstable angina, NSTEMI, and STEMI. Even though we did not perform a pooled analysis of the results, we consider it reasonable and in accordance with guidelines to limit the use of oxygen in all patients during PCI to hypoxemic patients with a SaO2 ≤90% only, irrespective of the level of pain.
First, pain is a complex outcome to measure, mainly due to the high degree of subjectivity, but also because of a lack of objective methods to quantify individual pain experience. There are a considerable number of tools available for self-reported estimation of pain, where the VAS is associated with high reliability and validity for chronic pain conditions as well as in acute pain (20,21,48). However, pain estimation by any tool might be affected by interviewer bias where the research staff influences the response of the subject. Therefore, we used a standardized procedure that in some extent should adjust for this. Another limitation is that some patients in our study were already treated with opiates in the pre-hospital setting, which could influence pain experience during PCI. Furthermore, as the study was not double blinded or placebo controlled, we cannot rule out a potential placebo effect of oxygen. However, in our previous study (17) as well as in the SOCCER (Supplemental Oxygen in Catheterized Coronary Emergency Reperfusion) trial (2), we used a placebo-controlled design with similar results. Also, in this study, we used oxygen with a moderate flowrate of 6 l/min. In several publications, higher doses of oxygen have been associated with an increase of vascular resistance, decrease in coronary blood flow, aggravated cardiac injury, and worsen long-term prognosis (1,8,49,50). The results of this study do not refute the risk of a dose-dependency, and it remains unclear whether the results would have been similar in a study of higher oxygen concentration. However, we designed this study to reflect current clinical practice in patients with suspected AMI without hypoxemia at baseline. By using a moderate dose of oxygen combined with a lower limit of SaO2 of 90%, we increased oxygen saturation enough to diminish the risk of hypoxemia while avoiding excessive hyperoxemia and potential harm. Finally, in clinical practice, oxygen is commonly used to alleviate nausea. We did not specifically investigate this variable, which poses a limitation.
In this pragmatic, registry-based randomized clinical trial, we did not find any analgesic effect of oxygen as compared with ambient air, and no differences in the use of sedatives and opiates during PCI. Our results indicate that moderate-flow oxygen supplementation does not relieve pain in patients with suspected AMI and normal oxygen saturation undergoing treatment with PCI and thus should not be used for this purpose.
WHAT IS KNOWN? Oxygen in normoxemic patients with AMI does not provide clinical benefit but may relieve ischemic pain.
WHAT IS NEW? Our findings do not demonstrate a significant analgesic effect of moderate oxygen flowrate supplementation. Oxygen should therefore not be used in normoxemic patients.
WHAT IS NEXT? The generalizability of our findings should be studied in high-concentration oxygen flowrate.
The authors would like to acknowledge all staff members at the participating hospitals, and the DETO2X-SWEDEHEART investigators for making it possible to perform this important substudy.
Dr. Östlund has received grants from AstraZeneca and The Medicines Company. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- acute myocardial infarction
- non–ST-segment elevation myocardial infarction
- percutaneous coronary intervention
- arterial oxygen saturation
- ST-segment elevation myocardial infarction
- Visual Analogue Scale
- Received January 22, 2018.
- Revision received March 8, 2018.
- Accepted April 10, 2018.
- 2018 American College of Cardiology Foundation
- Stub D.,
- Smith K.,
- Bernard S.,
- et al.
- Khoshnood A.,
- Carlsson M.,
- Akbarzadeh M.,
- et al.
- Cabello J.B.,
- Burls A.,
- Emparanza J.I.,
- Bayliss S.E.,
- Quinn T.
- Sepehrvand N.,
- Ezekowitz J.A.
- Cabello J.B.,
- Burls A.,
- Emparanza J.I.,
- Bayliss S.,
- Quinn T.
- Fu L.W.,
- Longhurst J.C.
- Steele C.
- Rawles J.M.,
- Kenmure A.C.
- Zughaft D.,
- Bhiladvala P.,
- Van Dijkman A.,
- et al.
- Hofmann R.,
- James S.K.,
- Jernberg T.,
- et al.
- Abdikarim A.B.D.I.,
- Basgut B.
- Rizer R.
- Lovelace W.R.
- Maroko P.R.,
- Radvany P.,
- Braunwald E.,
- Hale S.L.
- Sterling D.L.,
- Thornton J.D.,
- Swafford A.,
- et al.
- Horvat M.,
- Yoshida S.,
- Prakash R.,
- Marcus H.S.,
- Swan H.J.,
- Ganz W.
- Mouren S.,
- Souktani R.,
- Beaussier M.,
- et al.
- Welsh D.G.,
- Jackson W.F.,
- Segal S.S.
- Winegrad S.,
- Henrion D.,
- Rappaport L.,
- Samuel J.L.
- Ibanez B.,
- James S.,
- Agewall S.,
- et al.
- American College of Emergency Physicians,
- Society for Cardiovascular Angiography and Interventions,
- O'Gara P.T.,
- Kushner F.G.,
- Ascheim D.D.,
- et al.
- Fonnes S.,
- Gogenur I.,
- Sondergaard E.S.,
- et al.