Spontaneous Coronary Artery Dissection

Spontaneous coronary artery dissection (SCAD) is a pathophysiologically distinct cause of acute coronary syndromes (ACS). It is increasingly recognized that optimal management is different from that for atherosclerotic ACS and that a SCAD diagnosis has specific long-term prognostic and therapeutic implications. Accurate diagnosis is therefore essential to ensure the best treatment of patients. At present this relies on the recognition of typical features of SCAD identified on invasive coronary angiography. Although most SCAD can be readily distinguished angiographically from alternative causes of ACS, false positive and false negative diagnoses remain common. In particular, sometimes non-SCAD presentations, including atherothrombosis, takotsubo cardiomyopathy, coronary embolism, coronary vasospasm, contrast streaming, and myocardial infarction with nonobstructive coronary arteries, can mimic angiographic features usually associated with SCAD. The authors present the combined experience from European and US SCAD referral centers reviewing the classical angiographic appearances of SCAD, presenting potential diagnostic pitfalls and exemplars of SCAD mimickers. The authors further review the benefits and limitations of intracoronary imaging in the context of SCAD. Finally, the authors discuss the investigation of ambiguous cases and an approach to minimize misdiagnosis in difficult cases.

S pontaneous coronary artery dissection (SCAD) is recognized as an important cause of acute coronary syndrome (ACS) leading to myocardial infarction (1)(2)(3)(4). It is caused by hematoma formation within the tunica media of the coronary vessel wall, leading to the development of a false lumen that tracks both longitudinally and circumferentially.
There is increasing evidence that in most cases, this hematoma arises de novo within the vessel wall, rather than as a consequence of a primary endothelial-intimal tear or flap (3,5,6). Compression of the true lumen leads to coronary insufficiency, myocardial infarction, and in some cases ventricular arrhythmia (7). Accurate diagnosis is critical, as management of SCAD differs compared with that for ACS of atherosclerotic etiology both in the cardiac catheterization laboratory and afterward. For example, percutaneous coronary intervention (PCI) in SCAD is associated with high rates of complications and lower rates of angiographic success, whereas conservative management is associated with complete coronary healing in most cases (7)(8)(9). For this reason, both European and American consensus documents recommend a conservative strategy when practicable (1,2). The diagnosis of SCAD can be challenging. At present, there is no biomarker that reliably differentiates SCAD from atherosclerosis. Noninvasive diagnosis by computed tomographic coronary angiography (CTCA) is not routinely recommended, because of its lower spatial resolution, which limits assessment of the more distal coronary territories that are frequently affected by SCAD (10). As such, the diagnosis continues to rely upon recognition of characteristic features on invasive angiography.
The aim of this review is to revisit the diagnostic features of SCAD, present potential pitfalls including SCAD mimickers, and suggest an approach to optimize accurate diagnosis.

PRETEST PROBABILITY: BEFORE ANGIOGRAPHY
Before the patient arrives in the cardiac catheterization laboratory, there are a number of characteristics that influence the pretest probability of SCAD. For example, almost all patients with SCAD present with ACS. Biomarkers of myocardial injury (especially serial monitoring of high-sensitivity troponin [11]) are almost invariably elevated, except perhaps in cases in which presentation is very early or has been delayed.
A nonacute presentation should therefore raise the level of diagnostic doubt.
Patients with SCAD are overwhelmingly female, with male SCAD occurring in about 10% of cases in most series (9,12,13). SCAD has been reported as the cause of up to 35% of ACS events in women younger than 50 years (14)(15)(16) and 23% to 68% of pregnancyassociated ACS (17,18). The relative rarity of men diagnosed with SCAD compared with those with atherosclerosis warrants a higher index of suspicion of potential SCAD in men. Cases of SCAD in older patients are increasingly recognized (3), with a mean age in a recent prospective series of 52 years (12).
However, SCAD is uncommon in very young adults (<25 years of age), especially outside the context of pregnancy or hereditary connective tissue disorders, and is also uncommon in very old patients (>80 years of age). Presentations falling outside this age range should therefore be more carefully scrutinized before a diagnosis of SCAD is confirmed. It is worth noting that up to 90% of SCAD cases reportedly occur in women between 47 and 53 years of age (4). The presence or absence of atherosclerotic risk factors is not very useful as a guide to the likelihood of SCAD. It is important to appreciate that although risks are lower than in atherosclerotic patients, patients with SCAD are not "free" of risk factors, as is sometimes reported. Hypertension occurs in about 30% of patients with SCAD, although established diabetes is rare (12). Conversely, because atherosclerosis is at least an order of magnitude more common than SCAD, it remains the most probable cause of ACS, even in patients with few risk factors (3).
SCAD is associated with a small number of known genetic disorders (19,20). A recent gene sequencing study showed that 3.5% of patients with SCAD had causal or likely pathogenic rare genetic variants, mostly in genes associated with other known disorders (eg, vascular Ehlers-Danlos, Loeys-Dietz, or adult polycystic kidney disease) (21). Patients presenting with ACS who are known to have these disorders or with family histories or suggestive clinical features should trigger consideration of a potential SCAD diagnosis.
Symptoms at the time of SCAD presentation are similar to those occurring with other causes of ACS and therefore are not a useful diagnostic discriminator (22). In some patients, potential trigger exposures have been identified, such as emotional or physical stressors (23). Where clear-cut, such as the HIGHLIGHTS Accurate diagnosis of SCAD is critical as recommended interventional management differs from atherosclerotic ACS. For SCAD, a conservative approach is favored over percutaneous coronary intervention where possible.
SCAD diagnosis is based on invasive angiography. While the existing angiographic classification helps with recognition of common presentations, some SCAD appearances are not easily classified and sometimes other non-SCAD diagnoses can generate angiographic SCAD mimickers.
The authors present the potential pitfalls in the angiographic diagnosis of SCAD to help clinicians recognize areas of diagnostic uncertainty. They propose an approach centered around the careful assessment of pre-and postangiographic probability coupled with the judicious use of intracoronary imaging and other follow-up investigations to maximize diagnostic accuracy.  Recent research has demonstrated that the common genetic risk variants associated with SCAD provide some protection from atherosclerotic disease (33)(34)(35). Extensive atheroma on angiography is correspondingly rare, although nonobstructive plaque may occur (36). Other coronary abnormalities, including increased tortuosity (29), have been described in patients with SCAD and may represent associated coronary arteriopathies (37), although an angiographic string of beads akin to the appearance of extracoronary fibromuscular dysplasia (FMD) appears rare (36).
Intramural hematoma in patients with SCAD is frequently bounded at its proximal and distal extent by branch points, which seem to provide some resistance to further axial extension (26,30). This is distinct from atheroma, which has a predilection for bifurcation points. One additional key feature that can help in the differential diagnosis of SCAD is the degree of luminal thrombus. SCAD arises from external compression of the true lumen, and luminal thrombus is a less common feature when this has been assessed angiographically (9) Figures 3C1 to 3C3). Administration of intracoronary nitrate, when blood pressure permits, will usually relieve spasm ( Figure 3C2), but caution is required, as there is often an element of vasospasm associated with SCAD.
ATHEROSCLEROSIS. Because atherosclerotic disease is the most common cause of ACS, it is therefore also the most common differential diagnosis for SCAD.
Rupture with fissuring can lead to contrast penetration of the atherosclerotic plaque core, sometimes giving an appearance akin to contrast penetration of a type 1 SCAD false lumen and even evolving into a localized plaque-associated dissection ( Figure 4B).
When present, these features are usually confined to the plaque location. Recanalized coronary thrombus from atherosclerotic plaque rupture can also sometimes generate multiple channels, giving an (A1) Contrast streaming in the mid left anterior descending coronary artery (LAD) after a short stenosis gives the appearance of type 1 spontaneous coronary artery dissection (SCAD) (circled and arrows). (A2) A few frames later, as more contrast is injected, this appearance is lost upstream, but an apparent dual-lumen appearance now appears more distally (arrow). (B) A mid right coronary artery (RCA) probable type 1 SCAD is shown for comparison; note that the lack of heterogeneity in the dark contrast-opacified vessel makes streaming unlikely. This patient also had confirmed LAD SCAD. (C1) Catheter-induced coronary spasm of the RCA is relieved by intracoronary nitrate (C2). Subsequent assessment of the distal vessel confirms a type 2b dissection of the posterior descending branch (arrows, C3). Particularly diagnostically challenging is the differential diagnosis between SCAD and highly localized rupture or erosion of noncalcified, lipid-rich atherosclerotic plaque leading to coronary thrombus formation. This is a cause of ACS in young patients and women (38,39) and can, like SCAD, be provoked by rigorous exercise (40). Furthermore, these lesions are often nonobstructive before the acute event (38,39,41) and, when managed without revascularization, may heal, leaving only minor residual stenosis (again like SCAD) (Figures 4B to 4D). can also give the appearance of multiple channels, akin to type 1 SCAD (Figures 5A2, 5B2, and 6).
Furthermore, clot resorption means that coronary embolus, like SCAD, will often resolve over time with  Figure 5B), or proximal atherosclerotic plaque (as a source of plaque-associated thrombus; Figure 6).
There may also be truncation of multiple coronary branches, which is highly suggestive of an embolic source (although this must be carefully distinguished from multivessel SCAD, which occurs in about 15% of cases [9,42]). Sometimes additional investigations may be required to look for evidence of paradoxical embolism, cardiac source of embolus, or a predisposing hypercoagulable state (Figures 5C1 and 5C2).
TAKOTSUBO CARDIOMYOPATHY. Both SCAD and takotsubo cardiomyopathy affect predominantly women, with takotsubo affecting an older but overlapping population (43). Because SCAD has a predilection for more distal coronary territories and for the LAD, apical regional wall motion abnormalities akin to those seen in takotsubo cardiomyopathy are common ( Figures 7A1 to 7A3, Supplemental Figure 2).
Cardiac magnetic resonance imaging in convalescence may demonstrate a pattern of late gadolinium enhancement suggestive of infarction in a coronary territory, which might suggest SCAD. However, 40% of SCAD events resolve without evidence of lasting late gadolinium enhancement (44). Therefore, an apical regional wall motion abnormality that resolves However, it is well established that SCAD is associated with an increased risk for iatrogenic dissection (46) ( Figures 8A1 to 8A3, Supplemental Figures 4A1 to 4A3).
What is less clear is to what extent iatrogenic dissection in patients with ACS with otherwise normal coronary arteries occurs either because of preexisting proximal coronary SCAD (with the catheter penetrating the thin intimal-medial membrane and  Figures 8B and 8C). In some cases, contrast injection before coronary intubation can demonstrate a preexisting proximal hematoma, confirming SCAD prior to iatrogenic dissection ( Figure 8D). In other cases, SCAD may be demonstrated elsewhere in the coronary tree ( Figures 8A1 to 8A3). Rarely, dissection may arise from trauma or even from an active fixation pacing wire (Supplemental Figures 4B1 to 4B3). Dissections occurring in this context would not be considered as SCAD.
ODDITIES: IS THIS EVEN SCAD?. There are some angiographic appearances that appear different from classical SCAD and may not be part of the same clinical syndrome. These include those with a chronic dual-lumen or multichannel appearance in which the differential seems to lie between recanalization of an occlusion and some form of chronic dissection (Supplemental Figure 5). Likewise, it is unclear if ectasia-associated dissection is part of the SCAD spectrum of coronary arteriopathies or is pathophysiologically distinct (Supplemental Figure 6).

INTRACORONARY IMAGING: USE AND LIMITATIONS
Although invasive coronary angiography provides diagnostic images for most patients with SCAD, there remain cases in which angiography alone leaves uncertainty. In these cases, intracoronary imaging is frequently helpful but carries a small added procedural risk in the already fragile arteries in SCAD. The largest published coronary imaging series confirm a small number of complications (5 of 63 cases) directly attributable to imaging, all of which were managed For this reason, when intracoronary imaging is required for diagnosis, OCT with its much higher spatial resolution is generally better (Figures 9D to 9F).
Theoretical risks for dissection extension with contrast injection have not been borne out in described series to date, although caution in very proximal type 1 dissections may be sensible, and it is not usually necessary to image the entire SCAD length.
It is worth remembering that light penetration of the false lumen is variable, particularly in type 2 cases in Spontaneous coronary artery dissection (SCAD) is associated with an increased risk for iatrogenic dissection (A1 to A3). Initial angiography demonstrating type 4 SCAD in the mid left anterior descending coronary artery (A1). During subsequent angiography, a linear filling defect is seen in the left mainstem (A2), followed on the subsequent injection by complete occlusion (A3). (B) In this case there is just a hint of a nonobstructive hematoma causing a minor stenosis (arrow) to suggest preexisting SCAD prior to extensive iatrogenic dissection (not shown) occurring on the next coronary injection. It can be difficult to be certain when dissection from the ostium is evident on the first contrast injection such as this right coronary artery injection (C), unless there are preexisting inferior electrocardiographic changes or if a nonselective injection (D, arrow) confirms dissection before catheter engagement.
Adlam et al. which contrast has not penetrated into the false lumen ( Figure 9E) or luminal blood clearance with contrast injection is difficult. Careful image review for the typical features of SCAD is therefore needed, but it is usually possible to distinguish SCAD from lipid-rich atheroma with OCT ( Figure 9F).
In some situations, however, intracoronary imaging in patients with SCAD may not be feasible, particularly in cases with severe tortuosity or when SCAD occurs in distal, small-caliber arteries.

AFTER ANGIOGRAPHY
When diagnostic uncertainty remains after angiog- A demonstration of complete healing is consistent with SCAD and can be helpful for some cases (although coronary thrombus including emboli may also similarly heal, albeit on a shorter time scale) ( Figure 6). Persisting stenosis, even can refute a diagnosis of SCAD in favor of an atherosclerotic event or prompt percutaneous intervention when deemed appropriate. When repeat coronary assessment is contemplated, it is important to allow sufficient time for healing. SCAD studies with angiographic follow-up suggest that most will have healed by 1 month, but when imaging is driven by diagnostic considerations rather than symptoms, it may be sensible to wait up to 6 months (9,47,48).
SCAD has a strong association with extracoronary arteriopathies, particularly FMD, which occurs in at least one third of patients (8,49). Brainto-pelvis cross-sectional imaging to screen for coexistent aneurysm, extracoronary dissection, or FMD is currently recommended (1,2). Some centers also advocate fluoroscopic angiography of the renal arteries at the time of coronary angiography.
The presence of FMD in a patient whose angiogram is nondiagnostic is supportive of a SCAD diagnosis, as FMD appears to be uncommon in the general population (50). However, atherosclerotic ACS has also been described in the context of FMD (51).

CONCLUSION AND SUGGESTED APPROACH
Most SCAD can be confidently diagnosed angiographically, and for these patients further investigations beyond extracoronary arteriopathy screening and an assessment of left ventricular function are unnecessary. However, some cases are challenging, and in these patients further investigations may help establish or refute a firm diagnosis of SCAD. This can be critical to determine the best long-term management. A suggested approach to investigations for different angiographic scenarios is shown in the Central Illustration.