Clinical characteristics in STEMI-like aortic dissection versus STEMI-like pulmonary embolism

Methods: Reports were culled from a literature search covering the period January 2000 to March 2020 using Googlescholar, Pubmed, EMBASE and MEDLINE. Reports were included only if there was a specifi cation of the location of ST segment elevation and an account of the clinical signs and symptoms. Search terms were “ST segment elevation”,”aortic dissection”, “pulmonary embolism”, “myocardial infarction”, and “paradoxical embolism”. Fisher’s exact test was utilised for two-sided comparison of proportions. Proportion was calculated for each group as the number of patients with that parameter relative to the total number of patients assessed for that parameter.


Introduction
The syndrome of ST segment elevation encompasses acute myocardial infarction (AMI) [1], dissecting aortic aneurysm (DAA) [2], and pulmonary embolism(PE) [3], respectively. When faced with a patient with an electrocardiogram (ECG) characterised by ST segment elevation, the challenge which confronts front line doctors is whether or not to fast-track the patient to the catheter lab for percutaneous coronary intervention (PCI) [4] or to evaluate the patient for alternative diagnoses, which include DAA [2] and PE [3], respectively. Although clinical decision rules (CDRs) explicitly spell out algorithms for the differential diagnosis of ST segment elevation myocardial infarction (STEMI)[1], DAA [5], and PE[6], those CDRs do not take into account those cases of DAA and PE, respectively, who have a STEMI-like presentation. In the absence of guidelines for management of STEMI-like DAA and STEMI-like PE most clinicians rely on gestalt (ie unstructured clinical decision-making) for the work-up of STEMI-like clinical presentation. This review of the clinical characteristics of STEMI-like patients with DAA and STEMI-like patients with PE is, in essence, a re lection of the degree to which the index of suspicion for either DAA or PE operates in real world clinical medicine, where clinicians rely mainly on gestalt for clinical decision making.

Clinical features in 138 DAA patients with ST segment elevation
These 138 patients [2,7-124] were subdivided into 91 who had initially been managed by percutaneous coronary intervention(PCI) or by thrombolysis, arguably on the premise that the underlying cause of their clinical presentation was acute myocardial infarction (AMI), and 47 who had experienced alternative strategies for diagnostic workup.

Clinical features in 91 patients managed according to the AMI protocol
There were 91 patients [2,7-85,105,124] managed according to an AMI-oriented protocol encompassing either PCI or thrombolysis, the latter administered in 10 of the subjects [18,29,31,32,34,39,41,42,75,82]. Mean age of the 91 patients was 58. Sixty-eight were males, 23 females. Chest pain was reported in 85(98%), and back pain in 17(20%) of the 87 patients in whom those parameters could be ascertained. In 4 instances it could not be ascertained whether or not there was back pain nor chest pain. Breathlessness occurred in 11(12%) subjects [20,27,32,34,35,37,42,58,62,80,84], including one patient with a clinically audible murmur of aortic regurgitation. Focal neurological symptoms were present in six subjects, comprising unilateral upper limb pain, lower limb sensory symptoms, hemiparesis, unilateral headache, unilateral lower limb sensory symptoms, and paraparesis, respectively. The murmur of aortic regurgitation was detected in six, and mediastinal enlargement in two. The caveat is that there was no routine evaluation for the murmur of aortic regurgitation, focal neurological symptoms, or mediastinal widening in "all comers" with ST segment elevation. In the period leading up to operative intervention, cardiac arrest occurred in 16 subjects [2,31,[40][41][42][43][44][46][47][48][49][50][51]66]. Nine survived the episode of cardiac arrest. Electrocardiographic manifestations in these 91 patients were the following:-ST segment elevation in the inferior leads II,III,AVF( singly or in various combinations, with or without concurrent ST elevation in AVR or precordial leads) was documented in 45 patients. ST elevation in precordial leads was documented in 34 patients (without concurrent ST elevation in II,III,AVF, and also without concurrent ST elevation in AVR). ST elevation in AVR was documented in 12 patients(with or without concurrent elevation in precordial leads, and exclusive of concurrent ST elevation in II, III, or AVF).
In the entire group of 91 subjects managed according to the AMI-oriented algorithm, 46(51%) were also evaluated by TTE. Three of these 46 patients were evaluated by TTE as well as by transoesophageal echocardiography (TOE). Thirty-two (70%) of the 46 TTE evaluations showed stigmata of DAA. Ten subjects were evaluated by transoesophageal echocardiography(including 3 evaluated by TTE as well as TOE), which showed stigmata of DAA in each instance.

Clinical features in 47 patients who were not managed according to the ST segment elevation myocardial infarction algorithm
These 47 patients  had a mean age 58. Thirty-two were male, 15 were female. Chest pain occurred in 34(72%), and back pain 13(27.6%) of the 44 patients in whom those parameters could be ascertained. However, in 3 instances it could not be ascertained whether or not there was presence or absence of either back pain or chest pain. In 8 other instances [107][108][109]111,113,114,122], the patients speci ically denied presence of both chest pain and back pain. Breathlessness occurred in 14(30%) [22,45,63,88,91,92,99,100,104,111,1 14,118,122] subjects, including 4 with the clinically audible murmur of aortic regurgitation. Focal neurological symptoms occurred in 14, comprising paraparesis, unilateral lower limb weakness, unilateral upper limb pain, unilateral lower limb pain, and unilateral lower limb pain, respectively. The murmur of aortic regurgitation was elicited in 12, and mediastinal enlargement was documented in 10. However, none of the latter three parameters were routinely evaluated in every patient not fast tracked. In the period leading up to surgical intervention cardiac arrest occurred in 9 patients [ Electrocardiographic manifestations in these 47 patients were the following:-In 22 patients ST segment elevation was documented in leads II,III,AVF (with or without ST concurrent segment elevation in AVR or precordial leads). In 19 patients, ST segment elevation was documented in precordial leads, without concurrent ST segment elevation in II,III,AVF, or AVR. In 6 patients ST segment elevation was documented in lead AVR, with or without ST segment elevation in precordial leads, but exclusive of concurrent ST elevation in II,III,AVF.
In the entire subgroup of 47 subjects not managed according to an AMI-related algorithm 31(66%) were evaluated by TTE. One of these 31 patients was evaluated by TTE as well as by TOE. Twenty six (84%) of the 31 TTE evaluations showed stigmata of DAA. Three others were evaluated only by TOE, which was diagnostic of DAA in all three. (Table 1)

Clinical features in 102 PE subjects with STEMI-like presentation
These 102 subjects [3,  included 71 in whom the initial management consisted of fast-tracking to PCI. Among these were 32 patients with paradoxical coronary embolism [134,136,211,212,214,215,217,[219][220][221]. The other subgroup consisted of 31 subjects who were not fast-tracked to PCI Among these were 4 subjects with paradoxical coronary embolism [210,213,216,218).

PE patients in the PCI subgroup (71 patients; 33 males, 38 females)
In the PCI subgroup mean age was 60. Thirty-three were males, thirty-eight were females. Among those 71 patients chest pain was documented in 56, and breathlessness in 47. In 4 instances the presence or absence of chest pain and/or breathlessness could not be ascertained. Thirty two patients had paradoxical coronary artery embolism, associated, in 9 instances, with stroke and, in 2 instances, with transient ischaemic attack. Other manifestations of paradoxical embolism(PDE) were emboli to arteries in the limbs [195], and emboli to the renal artery (2 cases) [195,206], and to the splenic artery (1 case) [206]. All 32 PDE subjects had coronary angiography but thrombotic occlusion was absent in 2 cases [189,195]. In both of the latter 2 cases there were concurrent extracardiac embolic manifestations. Thirty-one of the 71 subjects with STEMI-like PE had documentation of deep vein thrombosis (DVT). Twenty of the 31 cases of DVT were in the PDE subgroup. Cardiac arrest was a feature of the clinical presentation in 15 patients (21%) [151][152][153][154][155][156][157][158][159][160]173,174,180,203], including one with PDE [203]. Six patients survived the cardiac arrest episode. Three of the survivors had received thrombolysis, one had embolectomy, and one had been managed solely by anticoagulants.
Electrocardiographic manifestations in these 71 patients were as follows:-In 30 patients, the ECG showed ST segment elevation in the precordial leads without concurrent ST elevation in AVR or in inferior leads.
Lead AVR(with or without involvement of precordial leads was the location of ST segment elevation in 6 cases. In 35 patients ST segment elevation occurred in leads II,III,AVF (singly or in various combinations) with or without concurrent ST segment elevation in AVR or precordial leads. Twenty-two of the 35 cases of ST segment elevation with inferior location were attributable to PDE. Among those with cardiac arrest, that complication was associated with pulseless electrical activity in 9, ventricular ibrillation in 2, ventricular tachycardia in one, and shock in 2.
Fifty-four (76%) of the 71 patients were evaluated by TTE. Right ventricular dilatation was documented in 43 (79.6%) of Chest pain was signi icantly (p = 0.0003) more prevalent in patients fast-tracked to the myocardial infarction pathway than in counterparts managed differently. Furthermore, all 8 of the DAA patients with pain-free presentation belonged to the subgroup of patients not fast-tracked to PCI [107-109, 111,113,114,122]. Breathlessness was signi icantly (p = 0.018) more prevalent in patients not fast tracked to the myocardial infarction pathway. Back pain, cardiac arrest were each just as prevalent in subjects fast-tracked to the AMIrelated algorithm as in counterparts not fast-tracked to that algorithm. TTE use was comparable in the two subgroups.
the 54 TTE evaluations. Five patients were evaluated by TOE [189,200,206,212,214], which showed embolus-in-transit across the interatrial septum in 4 cases.

PE subgroup not managed with PCI (31 patients)
In the subgroup managed without PCI, there were 31 patients, including 4 with PDE. Their mean age was 54. Twenty were males, eleven females. Chest pain and breathlessness occurred in 13 and in 23 patients, respectively. In 4 instances the presence or absence of chest pain and/or breathlessness could not be ascertained . Twenty-one had TTE, which showed right ventricular dilatation in twenty. Two others, both with PDE, had transoesophageal echocardiography (TOE), which showed embolus-in-transit in one instance [213]. Coexistence of DVT was documented in ten (32%) of the 31 subjects. Six of the ten [161, 162,167,175,179,185] had TTE, and in each case TTE showed right ventricular dilatation. Two patients, each with right ventricular dilatation on TTE [175,185] had cardiac arrest, and were successfully resuscitated. Altogether 14 had cardiac arrest [175,[177][178][179][180][181][182][183][184][185][186][187][188]. Seven patients survived the cardiac arrest episode. Four of the survivors had received thrombolysis, one had been managed by embolectomy, and two wad been managed solely by anticoagulants. In these 31 patients, electrocardiographic manifestations were as follows:-In 16 patients the electrocardiogram showed ST elevation in the precordial leads. In 10 patients the ECG showed ST segment elevation in inferior leads. In 2 instances inferior ST segment elevation was associated with paradoxical embolism. In 6 patients the ECG showed ST segment elevation in AVR. Among those with cardiac arrest, that complication was associated with pulseless electrical activity in 6, ventricular ibrillation in 2, shock in 2, unknown mechanism in 3. (Table 2) Overall Comparison between STEMI-like DAA and STEMI-like PE (Table 3)

Summary of STEMI-like DAA vs STEMI-like PE
The proportion of men in the STEMI-like DAA subgroup was greater than the proportion of men in STEMI-like PE.
Chest pain was signi icantly more prevalent in STEMI-like DAA than in STEMI-like PE.
Breathlessness was more prevalent in STEMI-like PE than in STEMI-like DAA..
Prevalence of cardiac arrest was lower in PE patients fasttracked to PCI than in PE patients not fast tracked to PCI.
Prevalence of use of TTE was signi icantly greater in STEMI-like PE than in STEMI-like DAA.
Irrespective of the underlying cause of ST segment elevation, and also irrespective of presence of risk factors, either for DAA (such as back pain) or for PE (such as DVT), patients with STEMI-like presentation who had chest pain were more likely to be fast-tracked to coronary angiography than to be initially evaluated with alternative strategies.

Discussion
The pattern which emerges from this review is that chest pain is the principal clinical feature which raises the index of suspicion for AMI in DAA patients who present with ST segment elevation. Conversely, when a patient with STEMIlike DAA presents with dyspnoea that symptom appears to raise the index of suspicion for a diagnosis (such as DAA) other than AMI. Although not all patients with STEMI-like clinical presentation were evaluated for the murmur of aortic regurgitation, focal neurological symptoms, and mediastinal widening, it may well be that, in some patients, breathlessness was a symptom of aortic regurgitation, thereby indirectly raising the index of suspicion for DAA.
Back pain had comparable prevalence in patients fasttracked to the AMI-related pathway vs patients managed by alternative strategies, arguably because none of the patients described their back pain as having the "tearing" quality attributed to DAA in the clinical decision rules [5]. When back pain does not have that distinctive quality it can be indistinguishable from the back pain which sometimes occurs in AMI [224].
In the context of STEMI-like PE, presenting features included chest pain, breathlessness, and cardiac arrest. The higher prevalence of chest pain in the patients fast-tracked to PCI was, arguably, indicative of a heightened index of suspicion for AMI in that subgroup. Cardiac arrest, on the other hand had signi icantly lower prevalence in fast-tracked patients then in their non fast-tracked counterparts. This phenomenon might have been attributable to the fact that 32 of the fast tracked patients had paradoxical coronary embolism, the latter a risk factor for cardiac arrest which can be mitigated by PCI. Indeed, 20 of the 32 PDE subjects in the fast-track subgroup were managed by thrombectomy, thereby mitigating the mortality risk associated with coronary occlusion. The opportunity to mitigate the risk of sudden death in this manner was missed in a 75 year old patient with PDE and angiographically documented right coronary artery embolism, and in whom coronary thrombectomy was not undertaken. Instead, after coronary angiography, she was urgently evaluated by computed tomography for suspected DAA. Soon after that procedure(which showed massive PE but no DAA), he suffered a fatal cardiac arrest [203]. Arguably, in her case, the risk of cardiac arrest could have been mitigated by the hybrid approach of combined coronary thrombectomy and pulmonary artery thrombectomy [192,202,211]. Breathlessness, on the other hand, had comparable prevalence in patients fast tracked to PCI vs patients managed according to a PE-related algorithm. The same was also true of the comparable prevalence of DVT in the two subgroups, the caveat being that not all subjects were evaluated for presence or absence of DVT. The suboptimal inclusion of PE in the differential diagnosis of patients in whom ST segment elevation coexisted either with breathlessness or DVT, and the suboptimal use of point-of care-Doppler ultrasonography of the lower limbs (to identify DVT) might have been attributable to underrecognition of the entity of PE related ST segment elevation [3].
The management of STEMI-like DAA would, arguably, have been vastly improved by point-of care ultrasonography, especially in the subgroup with breathlessness. In that context the use of point-of-care ultrasonography is exempli ied by the Vivid-7 system(GE Medical, Milwauwkee, Wisconsin, USA) which can also be taught to paramedical personnel [225]. Use of this system would have facilitated detection of heart murmurs including the wide variety of heart murmurs associated with aortic dissection [226]. Mere detection of a heart murmur would then have raised the index of suspicion for DAA, prompting more sophisticated evaluation by means of point of care TTE. The corollary is that routine lower limb ultrasonography would have raised the index of suspicion for PE, thereby prompting more sophisticated evaluation of suspected PE by point-of-care TTE.
This review was not only a re lection of "real world" clinical practice but also a re lection of the demographics of DAA and PE, respectively, exempli ied by the observation that, in this review, STEMI-like DAA had a male preponderance which mirrored the male preponderance of DAA documented in epidemiological studies [227]. For STEMI-like PE this review did not show a distinctive male preponderance, arguably re lecting the observation that epidemiological studies document only a 1.2:1 male: female ratio in the incidence of venous thromboembolism [228]. This review also created an opportunity to compare the rate at which STEMI-like DAA and STEMI-like PE have been reported in the medical literature. Remarkably, notwithstanding the fact that DAA(with age and sex adjusted incidence of 4.4 per 100,000 person years [227] has a much lower incidence than PE(with age and sex adjusted incidence 100-130 per 100,000 (persons/years) [228], the number of DAA-related reports of ST segment elevation was much higher than the number of PE-related reports of ST segment elevation. This anomaly might be a re lection of a greater propensity for Type II myocardial infarction to occur in DAA than in PE. That propensity might, in turn, be attributable to differences in the mechanism for generating the equivalent of Type II myocardial infarction in the two disorders. In DAA, myocardial ischaemia is attributable to the mechanical disruption of myocardial blood supply as a result of involvement of the coronary ostium by the tear in the ascending aorta [2,229]. That aetiopathogenic mechanism is the underlying cause for the high prevalence of myocardial malperfusion in DAA. In PE the operative factor might be the mismatch between oxygen demand and oxygen supply to the myocardium as a result of PE-related haemodynamic changes. The latter aetiopathogenic mechanism appears to have only a modest likelihood of generating the occurrence of myocardial infarction with nonobstructive coronary arteries.