Percutaneous interventional cardiovascular medicine: The PCR-EAPCI Textbook

Summary

Spontaneous coronary artery dissection (SCAD) is an important cause of acute coronary syndromes (ACS). It predominantly affects young or middle-aged women including a small proportion who are pregnant or post-partum. Accurate diagnosis is critical as optimal management differs from atherosclerotic ACS. Angiographic appearances of SCAD are usually distinct but ambiguous cases may require other investigations, particularly intracoronary imaging. Percutaneous coronary intervention (PCI) is associated with an increased risk of complications in SCAD and should be reserved for cases where a conservative revascularisation strategy is not possible. Mortality after SCAD is low and myocardial injuries are mostly small. However, recurrent SCAD is common. SCAD is also associated with extra-coronary arteriopathies such as fibromuscular dysplasia (FMD) for which brain-pelvis screening is recommended.

Management after SCAD requires consideration of key issues including menorrhagia, contraception and future pregnancy risk alongside questions around the risk/benefit of exercise and the psychological consequences of SCAD. Clinical trials are needed to provide greater certainty on optimal management.

What is SCAD?

SCAD is a cause of acute coronary syndromes (ACS). It is caused by the generation of a haematoma within the tunica media of the coronary arterial wall ( Figure 1) [1, 2, 3]. As pressure rises within this false lumen and exceeds diastolic blood pressure, the true lumen becomes progressively compressed leading to coronary insufficiency and myocardial infarction. Blood in the false lumen may also track both longitudinally and circumferentially ( Video 1) [4]. Coronary branch points are more often found at the proximal and distal limits of the false lumen suggesting these may provide a partial barrier to further haematoma tracking [5, 6].

The inside-out versus outside-in hypotheses

The source of the false lumen haematoma in SCAD has been the subject of two hypotheses ( Figure 2) [7]. The “inside-out” hypothesis proposed the development of an endothelia-intimal disruption or “tear”, allowing blood from the true lumen to enter the space between the internal and external elastic laminae. The “outside-in” hypothesis suggested the primary pathophysiological event is a de novo bleed within the vessel wall itself, perhaps from disruption of traversing microvessels. Increasing evidence has suggested the outside-in mechanism predominates in SCAD. Firstly, optical coherence tomography (OCT) studies have confirmed many SCAD cases at presentation have no initial connection between the true and false lumens (sometimes described as an intramural haematoma) [5, 8, 9]. Further image analysis has shown that at the site of the dissection, these cases have an expanded external elastic lamina, larger false lumen and thinner intimal-medial membrane, suggesting pressurisation of the false lumen. These findings suggest connections between true and false lumen, where they exist, arise from rupture of the pressurised false lumen into the true lumen rather than vice versa [5]. This is supported by serial angiographic studies which show that contrast penetration of the false lumen occurs later in the disease course and that angiographic lesions corresponding to intramural hematoma are more likely to present with early progression or re-infarction than those depicting a double-lumen [10, 11].

Pathophysiology of SCAD - genetics

Considerable progress has been made in our understanding of the pathophysiology of this condition. Recent studies have investigated both common and rare genetic variants associated with SCAD.

Rare genetic variants in SCAD

Rare genetic variants reportedly associated with SCAD are detailed in Table 1 [12, 13, 14, 15, 16, 17, 18, 19]. SCAD occurs rarely in patients with other known hereditary conditions, particularly adult polycystic kidney disease, Loeys Dietz syndrome and Vascular Ehlers Danlos syndrome [12, 13, 15, 17, 18]. Familial SCAD is rare [20] and in in a genome sequencing study likely pathogenic variants were identified in only 3.5% of unselected patients suggesting most cases are sporadic [12]. Decision-making on genetic testing with either a panel of arteriopathy genes or by sequencing is a balance between the potential benefits arising from identifying patients and families with a rare causal variant and the psychological morbidity and cost arising from testing a population where such variants are rare. At present genetic testing should be considered for all SCAD patients with a clinical phenotype or family history suggestive of a hereditary arteriopathy but is not recommended in all cases of SCAD.

Common genetic variants in SCAD

Variations in common genes associated with SCAD are now increasingly understood. The first described risk locus was rs9349379 on chromosome 6q24 close to the PHACTR1 and EDN1 genes [22]. The rs9349379-A allele is not only associated with SCAD but interestingly also with fibromuscular dysplasia (FMD) [23], cervicocerebral arterial dissection (CeAD) [24] and migraine [25], all conditions which occur at increased frequency in SCAD patients. Furthermore, the risk locus is also recognised from ischaemic heart disease genome-wide association studies (GWAS) but in this case the opposite allele rs9349379-G confers increased risk [26]. Two small single centre GWAS have now reported further common variants associated with SCAD ( Table 2) [27, 28]. These risk loci identified for SCAD are associated with a lower the risk of atherosclerotic disease but an increased risk of migraine. A larger multi-centre study will report findings soon.

Epidemiology – who gets SCAD?

Incidence

The precise incidence of SCAD is difficult to estimate due to issues of under-diagnosis and mis-diagnosis (with some series including atherosclerotic-associated dissections [30]). Estimates range from 0.8-4% [30, 31, 32]. SCAD fatalities are uncommon but again the incidence is unknown, due to challenges with accurate post-mortem diagnosis [33, 34].

Age, sex and ethnicity

SCAD shows a strong female predominance with around 90% of cases occurring in women[1, 2, 3, 35]. It accounts for up to 23-36% of ACS events in younger women (variously defined as under 50-60 years) [6, 36, 37, 38]. White Caucasian ethnicity predominates in most observational series but cases have been described in many racial groups[1, 2, 3, 35]. Most cases occur outside the context of pregnancy with the median age at presentation around 50 years[39]. One review article suggested up to 90% of SCAD cases occur in women between the ages of 47 and 53 years [35]. This narrow age-range is likely over-stated but, whilst SCAD is reported in early adulthood and in the elderly (patients <30 or >80), events in this context are rare.

Pregnancy-associated SCAD

Pregnancy-associated SCAD (P-SCAD; usually defined as SCAD occurring during gestation or within 12-months of delivery) accounts for around 5-10% of SCAD cases[1, 2, 3, 35, 39, 40]. SCAD reportedly occurs in 1.81 per 100,000 pregnancies, accounting for 10% - 22% of ACS events in pregnancy and 23-67% of post-partum ACS[41, 42, 43] with most post-partum events occurring within the first few weeks of delivery[40]. There is growing evidence that P-SCAD is associated with a more severe phenotype with more proximal and extensive dissections and larger infarcts[40, 44, 45]. Fatal P-SCAD is an uncommon but recognised cause of maternal death[46]. SCAD has also been observed in association with multi-parity and pre-eclampsia in some series[40].

Risk factors for ischaemic heart disease

It is sometimes stated that SCAD occurs in patients without conventional risk factors of ischaemic heart disease (IHD). Whilst this can occur, observational series show most risk factors occur at equivalent prevalence to the general population with some suggestion that hypertension may be more prevalent in SCAD and advanced diabetes less common[39, 47, 48]. For example a recent prospective series reported hypertension in 32%, dyslipidaemia in 20%, smoking in 12% and diabetes in 4.5% [39]. So whilst risk factors are less common in SCAD patients than atherosclerotic patients, SCAD should not be excluded from diagnostic consideration when risk factors are present.

Conditions associated with SCAD – causation versus co-incidence?

Many other conditions have been described in association with SCAD (summarised in Adlam et al [1] Supplementary table2). However, few associations extend beyond case descriptions or anecdote. All such reported associations should be approached with some caution.

Inflammation

Illustrating this point, a recent publication comparing rates of inflammatory disease in SCAD with those in an appropriately matched control population showed no clear link, despite many previous observational reports suggesting an association[49, 50].

Female sex hormones

The strong female sex predominance of SCAD and the association with pregnancy and perhaps multiparity strongly suggest some association between female sex hormones and the pathophysiology of SCAD. Furthermore, there is evidence that sex hormones may modulate the severity of SCAD with P-SCAD a more severe phenotype than non-P-SCAD and pre-menopausal SCAD more severe than post-menopausal SCAD[40, 44, 45, 51]. However the precise relationship remains unknown and SCAD has been described in men, post-menopausal and nulliparous women. SCAD has been described in patients taking hormonal contraception and hormone replacement therapy. However, it has never been demonstrated that exogenous hormone use is more prevalent in SCAD or that continuing to take hormones after SCAD increases the risk of recurrence [3].

Emotional stress

SCAD is frequently described in association with increased emotional stress, particularly in women [52]. This is clearly difficult to control for as emotional stress is highly prevalent in the population. However, a recent Dutch study restricted to female patients confirmed significantly higher odds of an emotional precipitant in the 24-hours prior to SCAD (56%) than prior to atherosclerotic ACS (39%)[53].

Exercise

SCAD is also frequently associated with exercise although again quantification of any effect size is challenged by the absence of an appropriate control population [1, 2, 3]. An association with exercise reportedly occurs more commonly in men [52]. In some cases symptom onset coincides with isometric or unusually intense exercise, reported in one series as occurring in as many as 9.8% and 28.9% respectively [39]. Much more frequently SCAD may occur in people who have a higher performance status but without a clear temporal link between an exercise episode and symptom onset. In these cases, it is unclear if exercise is a precipitant with delayed onset, a risk factor or if there is a non-causal link between SCAD risk and exercise performance status. Exercise after SCAD is further discussed below.

Presentation

Almost all SCAD presents acutely with an ACS event [1, 2, 3]. Serially measured high sensitivity biomarkers (Troponin) are therefore normally elevated, except in very rare cases with a staggered onset presentation where timing of first symptoms and Troponin testing are separated in time. Symptoms at presentation are in keeping with the typical spectrum for other ACS events but may be less intense for non-occlusive dissections [54]. Diagnosis of SCAD is frequently missed or delayed predominantly because a diagnosis of ACS is often not considered in the younger, predominantly female population affected by SCAD because of their low risk of atherosclerotic cardiovascular events.

At presentation around a quarter of patients with SCAD present with STEMI, three quarters with NSTEMI and 5% with cardiac arrest [35, 39, 55]. SCAD is a cause of sudden unexplained cardiac death and should be considered in the differential for all such cases, particularly as the diagnosis is easily missed at post mortem [33].

Diagnosis

Accurate diagnosis of SCAD requires invasive angiography, supplemented by other imaging in selected cases [1, 3, 35]. SCAD is associated with an increased risk of catheter-induced dissection (occurring in 3.4% of cases compared to ~0.2% in the general angiographic population in one series) [56]. When SCAD is clinically suspected, angiography should be approached accordingly with careful co-axial technique and avoidance of aggressive catheter shapes. Assessment of the coronary ostia through non-selective injection prior to catheter engagement allows assessment for proximal dissections prior to intubation.

Angiographic classification

The angiographic findings in SCAD are generally characteristic and in the context of a patient of the typical demographic profile, usually sufficient for diagnosis [57]. SCAD has a predilection for the mid-distal coronary arteries and the left anterior descending coronary artery [39, 58, 59]. It is associated with increased coronary tortuosity[60].

The Yip-Saw angiographic classification greatly aided recognition of the angiographic features of SCAD cases, particularly those without overt contrast penetration of the false lumen ( Figure 3) [57]. Type 1 SCAD has a dual lumen appearance due to contrast penetration of the false lumen ( Figure 3 A). There may be associated dye hang-up after cessation of contrast injection. Type 2 SCAD manifests as a long smooth stenosis. In Type 2a SCAD this may taper at the distal extent of the haematoma with recovery of a normal vessel diameter distal to this ( Figure 3 B). In Type 2b SCAD the stenosis extends into the distal branches without recovery of a normal vessel architecture ( Figure 3 C). Type 3 SCAD describes shorter dissections where it is not possible to definitely distinguish SCAD from focal atherosclerotic disease without recourse to intracoronary imaging ( Figure 3 D). Type 4 SCAD has been used to describe coronary occlusions [4, 35]. These are frequently distal but may show some upstream vessel tapering allowing distinction from coronary embolism ( Figure 3 E). In some such cases the diagnosis only becomes clear after restoration of flow during PCI ( Figure 4). Multivessel SCAD (defined as more than one discontiguous SCAD affected American Heart Association segment [61]) occurs in around 10% of cases at presentation ( Figure 3 F) [1, 2].

Like any angiographic classification, although a useful aide-memoire, there are SCAD cases which do not fit well into these groupings and it is important to also recognise these anatomies ( Figure 5). Proximal, extensive dissections ( Figure 5 A) and extensive minimally-obstructive dissections defy easy classification. There are also some hybrid cases ( Figure 5 B) where long dissections with a Type 2 appearance have short segments of contrast penetration (a Type 1 feature) and others where a long narrowed segment of intramural haematoma culminates in an occlusion.

An alternative aid-memoire of the angiographic appearances of SCAD has been proposed by Motreff et al who describe an angiographic “stick insect” and “radish” appearance as indicative of the characteristic features of SCAD [6].

In reality, the main pathophysiological distinction lies between dissections with fenestrations between true and false lumens and those without, as Type 1 appearances have been associated with a lower risk of complications with PCI and a lower risk of early and in-hospital events [10, 11].

Limitations of angiography and grey cases

Whilst the angiographic appearances of SCAD are generally characteristic, there are some SCAD mimickers which should be considered [3]. These include:

Contrast streaming ( Figure 6). This can give a linear appearance similar to Type 1 SCAD but can usually be distinguished by increasing the contrast injection pressure and further imaging is not usually required.
Coronary vasospasm. This can usually be relieved by administration of intracoronary nitrates where haemodynamics allow. However superadded spasm on segments of non-obstructive SCAD is not uncommon so care should be taken to ensure complete resolution to a normal coronary dimension before SCAD is excluded.
Iatrogenic dissection ( Figure 7). It can be challenging to distinguish catheter-induced iatrogenic dissection from proximal SCAD where the catheter penetrates the intimal-medial membrane of a pre-existent haematoma ( Figure 7 A). Previous symptoms and ECG changes and non-selective injection ( Figure 7 B) would support SCAD, whereas sudden symptoms with corresponding ECG changes should point to iatrogenic dissection. SCAD is known to be associated with an increased risk of additional iatrogenic dissection[56, 62].
Coronary embolism. Type 4 SCAD and coronary embolic occlusion can give very similar appearances. Truncation of multiple branches due to a shower of embolic material can be a diagnostic feature but needs to be carefully distinguished from multivessel SCAD. Non-occlusive embolic clot can sometimes cause a filling defect giving an apparent dual lumen appearance mimicking Type 1 SCAD.
Atherosclerotic plaque rupture ( Figure 8) and athero-embolism ( Figure 9). It is important to remember that atherosclerosis is orders of magnitude more common than SCAD. The commonest SCAD mimickers are atherosclerotic. These may include non-obstructive sites of plaque erosion and rupture where contrast penetration of ruptured plaque or a tongue of thrombus can give a linear appearance or sometimes embolization of thrombus into the distal vessel can give a dual lumen appearance distally.
Takotsubo syndrome ( Figure 10). The apical left anterior descending coronary artery is a common site for SCAD. This can lead to an apical regional wall motion abnormality similar in appearance to Takotsubo syndrome [63]. As 40% of SCAD does not cause demonstrable lasting myocardial injury [44], both conditions can also lead to complete recovery of heart function.

FOCUS BOX 1Approach for diagnostically ambiguous cases

In cases of diagnostic uncertainty, the following approach can be useful:

Reconsider the pre-test probability. For example, although SCAD can affect any age or sex, a higher index of suspicion is merited in male cases and patients at the extreme ends of the age spectrum (<30 or >80).
Check for true lumen thrombus. SCAD does not appear to be highly thrombogenic and the presence of substantial true lumen thrombus should raise suspicion of an alternative diagnosis.
Consider intracoronary (OCT) imaging.
Assess for healing or evidence of atherosclerosis with follow-up CT Coronary Angiography (where spatial resolution allows).
Consider other investigations (e.g. echo-bubble, prolonged ECG monitoring and thrombophilia screen – to assess probability of paradoxical embolism or intracardiac thrombus).

Intracoronary imaging – use, risks and limitations

Diagnostic accuracy is critical for both the acute and long-term SCAD management as treatment strategies are increasingly divergent from those for atherosclerotic ACS. Instrumentation of the coronaries should be avoided where a diagnosis of SCAD is angiographically certain and conservative management is practical. However, where there is diagnostic doubt, intracoronary imaging is recommended where practicable [1, 3]. It may also be useful to guide PCI where this is essential. The risk of intracoronary imaging in SCAD is low. In a retrospective study of 63 patients undergoing OCT imaging for SCAD, imaging-related complications occurred in 5 cases, none of which led to adverse clinical outcomes although 2 required stenting [5].

SCAD can be diagnosed by either intravascular ultrasound (IVUS) or OCT ( Figure 12.png" data-toggle="modal" data-target="#popup-media" class="media-link" data-media_id="4676" data-folder="pcr-textbook" data-chapterid="362"> Figure 11) ( Figure 12) [8, 64]. IVUS has the theoretical advantage of a greater depth of imaging and not requiring contrast injection for image acquisition. These putative advantages are more than offset by the lower spatial resolution of imaging with IVUS compared to OCT [3, 4]. The higher spatial resolution of OCT allows better delineation of the intimal-medial membrane and other key features such as fenestrations connecting the true and false lumen [5, 8]. Therefore, although light depth penetration does not always allow complete delineation of the entire path of the external elastic lamina around the false lumen ( Figure 12), images are usually diagnostic. With IVUS, the intimal-medial membrane, when seen, is pathognomonic ( Figure 12.png" data-toggle="modal" data-target="#popup-media" class="media-link" data-media_id="4676" data-folder="pcr-textbook" data-chapterid="362"> Figure 11) but the limitations of spatial resolution can make demonstration difficult and scrutiny of multiple frames is frequently required. The ultrasound properties of the false lumen are variable and often heterogeneous within a case. In general, the false lumen's grayscale appearances with IVUS are not sufficient alone to distinguish SCAD from lipid-rich atherosclerotic plaque, frequently a critical differential diagnosis. For these reasons, currently, OCT is the preferred imaging modality for SCAD where available, except perhaps in cases with a proximal type 1 dissection where the potential risk of haematoma extension during contrast injection should be weighed-up on an individual case basis [3, 4]. However, no case has yet been reported where contrast injection during OCT imaging of SCAD led to extension of dissection [5].

Role of CTCA – use and limitations

Given the increased risk of iatrogenic catheter-induced dissection in SCAD [56, 62], it can be tempting to consider the alternative of CTCA. However, although CT appearances for acute SCAD have been described and can be characteristic, the lower spatial resolution of CTCA particularly impacts assessment of the smaller calibre mid-distal coronary territories for which SCAD has a predilection [65, 66, 67]. A negative CTCA can therefore not rule-out SCAD. It is also currently challenging to distinguish non-contrast penetrant intramural haematoma from lipid-rich atherosclerotic plaque and CTCA does not allow further interrogation of ambiguous lesions with intracoronary imaging. For these reasons CTCA is not recommended as the primary diagnostic stratagem for SCAD [1, 2, 3, 35].

CTCA can, however, be a useful adjunct where the diagnosis remains uncertain following initial angiographic assessment (for example in cases where intracoronary imaging was not practicable), particularly when the disease is affecting larger more proximal coronary segments [3]. SCAD typically heals completely within 3-6 months of follow-up although apparent healing can also occur with resorption of thrombus whether associated with minimally stenotic plaque/erosion or embolic in origin. Confirmation of complete healing of the affected (usually proximal-mid) coronary segment by follow-up CTCA can provide diagnostic reassurance whereas persistent stenosis, evidence of positively remodelled plaque or coronary calcium raises suspicion of an alternative (atherosclerotic) diagnosis.

Special considerations in pregnancy

P-SCAD has been demonstrated to be a more malignant phenotype with more proximal and more extensive dissection and worse outcomes [68, 69, 70]. Accurate diagnosis remains critical for the management of both patient and the pregnancy and P-SCAD does not alter the rationale for invasive angiography as the diagnostic test of choice. X-ray doses associated with angiography (or PCI if required) are acceptable in pregnancy and adjustment of fluoroscopy angles and shielding, where practical, can further reduce foetal exposure [71]. Cognisance of the risk of proximal dissections and iatrogenic dissection is critical and a careful technique with non-selective assessment of the coronary ostia before intubation is particularly important to minimise procedural risk. The pregnancy heart team should be integral to the management of all cases of SCAD in pregnancy [72]. Questions around management of delivery should be individualised but operative delivery after prior SCAD in pregnancy is not necessarily mandatory and supported vaginal delivery is often possible and may be preferable [71].

Cath lab management

Conservative management and healing

Angiographic follow-up series show most SCAD managed conservatively will heal with restoration of a normal coronary architecture ( Figure 13) [73, 74, 75, 76]. For example, a series by Saw et al of 74 cases having follow-up angiography on clinical grounds ≥26 days after the index event showed all had healed completely. Likewise, a series by Rogowski et al showed healing in all but one of 30 conservatively managed cases undergoing routine angiographic follow-up after 6-months. These studies suggest the overwhelming majority of SCAD cases managed conservatively heal completely within 3-6 months of the index event. Because of the potential risk of iatrogenic dissection [56, 62], follow-up angiography simply to confirm vessel healing is not recommended [1, 2]. Use of CTCA to confirm healing of proximal and mid vessel dissections (where the spatial resolution of CT is likely to be adequate) has been proposed ( Figure 13) but it remains unclear if the additional X-ray exposure is merited for follow-up screening in asymptomatic patients.

Percutaneous coronary intervention – risks and benefits

PCI in SCAD patients is associated with an increased risk of complications ( Figure 14) and lower reported rates of “procedural success” [58, 76, 77]. Complications include iatrogenic dissection, haematoma extension ( Figure 14) and vessel or side-branch occlusion. Furthermore infarct sizes after SCAD are generally small [44]. For this reason, a conservative approach to revascularisation is advocated in SCAD where possible [1, 2, 3]. This is distinct from guideline-based management of atherosclerotic ACS and serves to emphasise the critical importance of accurate diagnosis outlined above.

Whilst most SCAD can be managed conservatively, infarct sizes are increased in patients with proximal disease, occluded vessels and STEMI presentations [44]. Not all such patients will require PCI but in some cases where there is a significant amount of myocardium in jeopardy (e.g. proximal and mid vessel occlusions), revascularisation will be necessary.

Where PCI is undertaken, the aims of revascularisation may be different from PCI in the context of atherosclerosis, with the focus on restoring flow rather than restoring a normal coronary architecture. A study in 215 SCAD-PCI cases showed 64% had some residual un-covered dissection at the end of the procedure [78]. However, as this will heal over time, this should settle in most cases. As iatrogenic dissection is a significant risk (affecting 8.4% of SCAD-PCI cases in the same study [78]), selection of non-aggressive guiding catheter shapes and cautious intubation of the coronary ostia is important. It is critical to ensure the coronary guidewire is sited within the true lumen as stenting into the false lumen can be catastrophic [79]. One study showed 5 of 63 SCAD-OCT cases showed at least one pullback in which the guidewire was in the false lumen indicating the potential utility of intracoronary imaging where there is uncertainty [5].

There are currently no clinical trials to investigate the best PCI strategy in SCAD and little data beyond anecdotal case presentations. Approaches suggested include:

Cutting balloon angioplasty (see illustrative case): this approach has some logic as observational studies suggest Type 1 (contrast penetration of the false lumen) SCAD appearances have lower PCI complications and a lower risk of early re-infarction. Fenestrating into the false lumen serves to decompress the haematoma and if stenting is required, could reduce the likelihood for haematoma extension. However, clearly a cutting balloon is not a subtle tool to achieve this end and there may be a theoretical risk of the vessel injury being more extensive than intended.
Limited wiring or plain old balloon angioplasty (POBA): given SCAD will heal over time, there is a rationale to limiting PCI to what is necessary to achieve restoration of coronary blood flow. If this can be achieved by wiring alone or gentle POBA with an undersized balloon, this may be adequate in some cases to restore flow without necessitating stenting.
Oversized stenting: if stenting is needed, using a longer stent length may distribute stent expansion forces over a wider area and mitigate the risk of haematoma migration.
Upstream and/or downstream stenting: the main risk of stent expansion is haematoma extension. Placing a stent just upstream and/or downstream of the haematoma has been advocated before stenting the main area of stenosis. An alternative strategy of limited localised stenting sufficient to restore flow, allowing distal non-occlusive residual dissection, has been proposed in selected patients.

Outcomes of stenting

When stenting is necessary, it has been shown that longer lengths of smaller calibre stents will be required when compared to atherosclerotic practice. Stent strut mal-apposition has been reported following haematoma resorption [80] but this does not seem to be associated in a higher signal for stent thrombosis in reported series. Despite the risk of complications, in SCAD-PCI cases with reduced TIMI flow at presentation, improvements in coronary flow are reportedly achieved in 84% (with worsening in only 7%) and medium-term outcomes in terms of MACE and left ventricular function are good [78]. Drug-eluting stents should currently be used [81] although bioresorbable scaffolds have also been studied in SCAD and appear to provide favourable short term and medium term outcomes [82].

Coronary artery bypass grafting (CABG)

CABG for SCAD is usually an option of last resort in cases where conservative management is not an option and the risk of PCI is felt to be exceptionally high or a serious complication has arisen during PCI which requires surgical bail-out [1, 2, 3, 35]. There are significant challenges to CABG in this context. Ideally, the anastomosis should be conducted distal to the dissected segment but this may not always be possible for extensive dissections. If anastomosis onto a dissected coronary is required, it can be challenging to distinguish the true from false lumen. Follow-up of grafted SCAD cases has shown a high medium-term graft failure rate, likely due to healing of the native coronaries leading to competitive flow with the graft conduit, a well-recognised risk for graft occlusion [76]. CABG should therefore be primarily regarded as a short-term emergency revascularisation solution where no other option is viable. Although arterial grafting and off-pump CABG have been described [83], it may be best to keep operative techniques simple and aim for rapid revascularisation with high flow conduits (saphenous vein grafts) unless there is a particular reason to do otherwise.

Cardiac assistance and critical care

Cardiogenic shock in association with SCAD is uncommon but well recognised and described. All forms of bridging assistance have been described including standard critical care, intra-aortic balloon counter-pulsation, extra-corporeal membrane oxygenation (ECMO) and left ventricular assist devices[84, 85]. If cardiogenic shock is intractable, cardiac transplantation can be successfully performed [86]. There are no reports of SCAD subsequently occurring in the arteries of the transplanted heart.

FOCUS BOX 2Revascularisation in SCAD

PCI in SCAD is associated with an increased risk of complications, particularly iatrogenic dissection and haematoma propagation.
Conservatively managed SCAD will usually heal with complete restoration of luminal architecture.
A conservative approach is favoured where this is possible but PCI should be considered where there is significant myocardial jeopardy.
Where PCI is required, care should be taken to minimise the risk of guiding catheter induced iatrogenic dissection and to ensure a correct luminal guidewire position.
Despite the need for longer stent lengths and a higher risk of complications, improvements in coronary flow and good medium term outcomes can be achieved where PCI is required.
CABG should be reserved for an emergency bail-out for extremely high risk scenarios.

Length of stay

Further ischaemia can occur after SCAD or as a complication of revascularisation. This needs to be distinguished from (non-ischaemic) post-SCAD chest pain using careful assessment of serial electrocardiographs and biomarkers of myocardial injury. In general such episodes settle with a medical/conservative approach but significant re-infarction can occasionally occur requiring repeat angiography in some cases[87]. 30-day outcomes from a prospective observational study showed of 6.1% of patients with recurrent ACS, two-thirds occurred within a median 4-day in-patient stay [39]. For this reason, a longer in-patient stay than is now usual practice for atherosclerotic ACS is recommended for SCAD, particularly in conservatively managed and more extensive dissections [1, 2, 3].

Medical management

There have been no randomised clinical trials in SCAD to date. Recommendations are therefore based on limited observational data and expert opinion only.

Acute therapies

Thombolytics are not recommended for SCAD patients or where SCAD is suspected. Whilst there is little evidence for potential harm in this population, substantial true luminal thrombus is an uncommon feature of SCAD and these drugs are unlikely to be very effective [88, 89, 90].

Anticoagulation (heparin or bivalirudin) should be used as per standard practice for coronary procedures [91]. There is no current evidence of harm specific to SCAD and the risk of procedure-related coronary thrombosis without anticoagulation is well established.

GPIIb/IIIa inhibitors and Cangrelor are not recommended for SCAD patients although there are only anecdotal reports of either benefit or harm [1, 2, 3]. Again the relatively low levels of true luminal thrombosis in SCAD suggests these medications are unlikely to be helpful.

Long term therapies

Side-effects related to medications initiated after SCAD such as menorrhagia, low blood pressure and excessive fatigue are common in this younger population. Rationalisation and minimisation of medications for which there is no clear benefit is frequently helpful.

Antiplatelet therapy: in patients who have had PCI with stenting, management should remain in accordance with guideline-based recommendations[92, 93]. For conservatively managed patients the role of antiplatelet therapy and the duration of treatment remain controversial. The rationale behind use of antiplatelet treatment is largely based on longstanding evidence of benefit in the context of athero-thrombotic ACS. For SCAD where the underlying pathophysiology is thought to be a spontaneous intramural bleed, rather than a thrombotic event, the rationale for giving medications which prolong bleeding time is uncertain. Increased menstrual loss may also be an issue in women of childbearing age. Practically, many authors advocate an initial period of dual antiplatelet therapy with Aspirin and Clopidogrel (the latter is suggested in preference to more potent P2Y12 inhibitors) [3] to mitigate against the theoretical risk of true lumen thrombus relating to endothelial-intimal fenestrations acting as a nidus for clot formation. Clopidogrel can then be discontinued after 2-4 weeks[3].
The duration of antiplatelet monotherapy (usually aspirin) after SCAD is likewise controversial. The only available observational data is from a single underpowered Canadian observational study and showed no evidence of benefit of harm[94]. One approach is to complete post-SCAD imaging to ensure there is no other relative or absolute indication for aspirin (e.g. an extra-coronary chronic dissection or extensive cervical FMD[95]) before making a final decision to discontinue treatment.
Anticoagulation: there is no established increased risk of anticoagulants in SCAD beyond the risk of menorrhagia in women of menstrual age and a theoretical and unconfirmed risk of increasing the probability of recurrence (as for antiplatelet therapies). If anticoagulation is indicated (e.g. for post infarct left ventricular thrombus), it should be used according to current clinical practice[92, 96].
Betablockade: patients with significant left ventricular systolic dysfunction (LVSD) after SCAD should be treated with beta blockers in accordance with guidelines for post myocardial infarction LVSD[92, 96]. For patients without LVSD after SCAD, it has been argued that beta blockers may favourably modify coronary shear stress. There is also limited data from the same Canadian observational study referenced above that beta-blockers (and prevention of uncontrolled hypertension) were associated with lower rates of recurrent SCAD[94]. These data have not been validated in other series and are really hypothesis generating rather than a guide to treatment[97]. At present, where tolerated, it is reasonable to continue maintenance beta-blockers after SCAD. However, many patients in this younger population have lower baseline haemodynamic indices and are unable to tolerate even small doses of beta-blockers.
ACE-inhibitors and angiotensin receptor antagonists: these medications should be part of guideline based treatment in patients with LVSD after SCAD[92, 96] and considered for the treatment of hypertension. There is no clear role in normotensive patients without LVSD.
Statins: there is no current evidence of a pathophysiological link between SCAD and dyslipidaemia (in contrast to atherosclerosis). For this reason statins are not routinely recommended after SCAD but should be prescribed in patients with abnormal lipid levels according to current clinical practice guidelines [98]. Any role in patients with extensive stenting or saphenous vein grafts to prevent neo- or accelerated-atherosclerosis is unclear.

Contraception and hormone replacement therapy

Hormonal contraceptives and hormone replacement therapy (HRT) have been considered as predisposing to SCAD by some authors[99], largely on the basis of the female sex predominance of SCAD and the known association with pregnancy. However, to date it has not been demonstrated that hormonal contraceptive and HRT use is more prevalent in SCAD patients than appropriately matched controls, nor that continued use of hormonal therapies after SCAD increases the risk of recurrence[3]. Appropriately controlled observational analyses are awaited[50]. In the interim, the most important message is to avoid unplanned pregnancy in women of child-bearing age. If this can be reasonably and securely achieved without hormonal contraception (e.g. through partner vasectomy), this is the simplest approach. However if hormonal contraception is required, progesterone-based approaches are preferred where possible over oestrogen containing preparations[3]. Likewise, if the menopause can be negotiated without intrusive symptoms, HRT is not indicated. However HRT can reasonably be considered for the management of intrusive menopausal symptoms at the lowest dose and for the shortest duration required for symptom relief. In premenopausal women the use of the levonorgestrel-releasing intra-uterine system can be an effective and safe option for both contraception and mitigation of menstrual bleeding [100].

Migraine

Migraine occurs at increased frequency in patients with SCAD[101] and these conditions have shared genetic elements[28, 102]. For acute management, triptans are usually avoided[101] although reports of SCAD occurring immediately after triptan use are anecdotal and very rare[103]. Any associated risk of recurrent SCAD with triptan use therefore remains unclear and unquantified. However, alternative approaches such as high dose aspirin and anti-emetics may be helpful. Beta blockers are the migraine prophylaxis of choice where tolerated. Where control of migraine frequency and severity is difficult, there are a range of alternative approaches (including topiramate, riboflavin, magnesium, flunarizine, lamotrigine, amitriptyline and valproate). The choice should be individualised and discussed with a migraine specialist with careful consideration given to contraceptive issues where there are potential issues of teratogenicity.

Outcomes

Most observational studies focus on survivors of SCAD so data on fatal SCAD is limited. Although SCAD may be under-represented in autopsy series[33], fatal acute SCAD is likely very rare. Both echocardiographic[104] and CMR[44] studies show myocardial injuries after SCAD are generally small. 40% of SCAD patients undergoing follow-up research CMR had no measurable late-gadolinium enhancement and significant impairment of LV function was rare[44]. Those with STEMI presentation, reduced TIMI flow on angiography, more proximal and more extensive dissections had larger injuries as did those with P-SCAD and a high Beighton score (a measure of hypermobility) [44]. Mortality after SCAD is very low although MACE rates are high, (1.2% and 19.9% respectively in a prospective series followed up for a median 3.1 years), driven primarily by the risk of recurrent SCAD[94].

Device therapy

About 4-5% of SCAD presents with cardiac arrest[39, 55]. There is no evidence to suggest that such patients are at significantly higher risk of subsequent life threatening arrhythmia outside the context of recurrent AMI and/or LV systolic dysfunction. One study showed an incidence of cardiac arrest of 3.7% of 1056 patients presenting with a first SCAD event. During follow-up 6/1056 patients (0.6%) had cardiac arrest, 8/1056 died (0.8% - cause of death not given) and 2/1056 suffered ventricular arrhythmias not leading to cardiac arrest. There was a higher risk of cardiac arrest or ventricular arrhythmia in patients with recurrent AMI, impaired LV systolic function and ventricular arrhythmia during the first SCAD event[105]. These findings require validation in other series. Despite the recognised risk of recurrence, all observational registries report prevalent patients with previous SCAD have very low mortality rates[1, 2, 3, 35]. In this context, current data do not support extending the defibrillator use beyond the indications laid out in current guidelines for protection in the context of post-infarct LVSD[106] but more data is needed. In selected cases where there is early concern about arrhythmia in the context of persisting dissection, a temporary life vest may be considered. However this is not necessary for most patients.

Remote arteriopathies – role of imaging

SCAD is associated with a range of other arterial abnormalities ( Figure 15) and shares the common PHACTR1 genetic risk variant with both FMD and CeAD [22]. For this reason a single cross sectional imaging study from brain to pelvis by computed tomography angiography (CTA) or magnetic resonance angiography (MRA) is recommended in all SCAD patients [1, 2]. The most frequent finding is the radiographic “string-of- beads” characteristic of FMD. This occurs in 11-86% of cases with predominant involvement of the cervical, renal and ileofemoral arteries [1, 2]. Multiple vascular beds are involved in 29-49% [77, 107, 108, 109]. The reported prevalence is highly variable due to differences in the imaging modalities used, completeness of screening and the criterion used to define FMD. One recent study even reported extracoronary vascular abnormalities in 100% of screened patients [48]. Other abnormalities reported include mild aortic root dilation, arterial dissections (both asymptomatic and symptomatic), focal stenoses and aneurysms[77, 107, 108]. Intracerebral aneurysms have been reported in 11-23%[77, 101, 108]. Importantly although remote arteriopathies have a relatively high reported prevalence in SCAD, related vascular events and/or associated hypertension seem very uncommon[94]. From what is currently known, these lesions mostly seem stable and serial imaging is only required for follow-up of aneurysms or dissections in accordance with current guidelines. Lifelong aspirin may be considered in patients with remote dissections or extensive fibromuscular dysplasia (because of the perceived increased risk of associated dissection), although firm evidence of benefit in this context is lacking[95].

Post-SCAD chest pain syndrome

Non-ischaemic chest pain after SCAD is very common[1, 2, 3]. It occurs reportedly more frequently in females, patients with migraine and those with prior psychological or chronic pain issues. Symptoms are episodic and may be clustered pre-menstrually or at times of emotional stress[4, 110]. Re-admissions to hospital are common and it is important, when this occurs, to assess for objective evidence of ischaemia or infarction before contemplating invasive angiography (given the increased risk of iatrogenic dissection[56, 62]). Evidence for vasospasm or microvascular dysfunction as an underlying explanation for these symptoms is mixed and limited[111, 112, 113]. Where uncertainty exists, non-invasive coronary imaging with CTCA or functional imaging with stress CMR may be helpful to exclude a structural coronary issue. Overtime patients can often learn to distinguish post SCAD chest pain and manage episodes pragmatically without the need for admission. In most patients things improve over time but it can take 18-months to 2-years after the index event for symptoms to have faded into the background in most patients and even longer in a minority of cases.

Treatment of post-SCAD chest pain can be challenging and at present there is no clearly favoured treatment option. Pre-menstrual chest pain may respond to progesterone-based contraceptives (including the progesterone-releasing intrauterine device). Vasodilators (such as calcium channel blockers, particularly Diltiazem, or long acting nitrates) can be trialled but response to treatment is variable and often incomplete. Anecdotally ranolazine may be helpful in some patients. Where possible polypharmacy should be avoided and the aim should be for temporary rather than lifelong treatment (perhaps trailing treatment withdrawal after 12-18 months). Non-medical strategies including cardiac rehabilitation [114, 115] can also be a helpful element in some patients.

Recurrent SCAD

Recurrence is well recognised after SCAD. This should be distinguished from early relapses due to extension in the previously dissected segment. Around 10% of patients will experience recurrent SCAD within 5 years of the first event[1, 2, 3, 35]. Longer follow-up data is limited and whilst the rate of events may fall, very late recurrences ~10 years after the first event are recognised. When they do occur recurrences usually affect a different coronary location suggesting there is some aspect of the healing process which protects against future dissection[59]. Increased coronary tortuosity, FMD, migraine, uncontrolled hypertension and non-use of beta-blockers have been associated with increased odds of recurrence in observational studies [60, 94, 97]. Occasionally recurrent AMI may occur without a clear site of dissection. Recurrence SCAD needs to be carefully distinguished from post SCAD chest pain syndrome. Despite the recurrence risk and the associated additional myocardial injury, prognosis, even in those with recurrent SCAD, remains good.

Pregnancy after SCAD

Unplanned pregnancy should be avoided after SCAD and appropriate contraceptive measures (as above) taken to ensure this[1, 2, 3]. The risk of pregnancy after SCAD should be individualised in pre-conception counselling. This will take into consideration: medications (including potentially teratogenic drugs such as ACE inhibitors), left ventricular function, extra-coronary remote arteriopathies (such as aneurysms and dissections), potential for undiagnosed hereditary connective tissue disorders as well as the risk of recurrent SCAD[71]. Current data suggest there is a risk of recurrent SCAD associated with a subsequent pregnancy (including pregnancy and the post-partum period), occurring in perhaps 1:10 cases[69]. This cannot be predicted or prevented. As with all SCAD, the chances of a fatal or highly damaging event are low, although as outlined above, P-SCAD is recognised as being associated with a more severe SCAD phenotype and larger SCAD-associated infarcts. Pregnancy after SCAD is therefore neither absolutely contraindicated nor risk free. Careful pre-conception counselling and early involvement of the pregnancy heart team are therefore recommended to ensure fully informed decision-making and optimal care[72].

FOCUS BOX 3Pregnancy and contraception in SCAD

P-SCAD is associated with a more severe phenotype but the approach to diagnosis (including careful invasive angiography) and management is the same as non-P-SCAD.
Early involvement of the pregnancy heart team is recommended and delivery decisions and management individualised (vaginal delivery is not precluded in many cases).
Recurrent SCAD can occur during a subsequent pregnancy with around a 1:10 risk. This risk cannot be mitigated but many patients will still wish to consider pregnancy. Preconception counselling by the pregnancy heart team is recommended to ensure individualised assessment of risk and adequate patient information prior to pregnancy.
Unplanned pregnancy should be avoided after SCAD. Secure contraception is recommended. When required, hormonal contraception, favouring progestogenic preparations is acceptable.

Exercise and cardiac rehabilitation after SCAD

The reported association of SCAD with exercise (see above) has led to some concerns about the safety of exercise after SCAD. There is currently no evidence that exercising after SCAD increases the risk of recurrence or that restricting exercise prevents recurrence. This stands against the extensive evidence-base that exercise is beneficial for general cardiovascular as well as psychological health. Several studies have addressed the role of cardiac rehabilitation after SCAD, confirming both safety and efficacy in terms of clear evidence of psychological benefit[114, 115]. A pragmatic approach is to recommend against isometric exercise or lifting weights which requires an alteration to a regular breathing pattern (the Valsalva manoeuvre) but to permit toning weights (without specific limits), emphasising repetitions over maximal exertions. For other exercising, the aim should be to avoid exercise to exhaustion, again emphasising regular repeated exertion over occasional or extreme activities[3].

Mental health and post traumatic health disorder

In view of the younger, female and sometimes post-partum population affected, SCAD-survivors are at high risk of adverse psychological impacts following myocardial infarction[53, 116, 117]. Clinicians and cardiac rehabilitation practitioners should remain aware of this risk and consider early referral for counselling or other appropriate intervention if this is required.

Personal Perspective - David Adlam

Understanding of SCAD is advancing rapidly. Once considered “rare” and largely a disease of pregnancy, the advent of high sensitivity cardiac biomarkers, early angiography and intracoronary imaging has demonstrated SCAD is a very important cause of ACS in women and although most patients are not pregnant or post-partum, SCAD remains a key cause of ACS in this context. Large national observational registries have greatly enhanced our understanding of the presentation, diagnosis and management of SCAD and more recently international collaborations have started to unravel the fascinating genetic and molecular basis of this condition. Much more work is needed, in particular to build the collaborative partnerships required to deliver clinical trials in SCAD focused, in particular, on reducing the high recurrence rates particular to this condition. In the meantime, a key focus remains on educating the multidisciplinary health care team to improve the timely and accurate diagnosis of SCAD and therefore to ensure appropriate and optimal care for our patients with this condition.

Illustrative case: Cutting balloon angioplasty in SCAD

A 46-year-old woman was admitted for a non-ST segment elevation acute myocardial infarction. A type 1 spontaneous coronary artery dissection (SCAD) was detected involving the mid segment of the left anterior descending coronary artery (LAD). As the anterograde coronary flow was normal (TIMI 3) and she was completely asymptomatic during the angiographic study, a conservative medical strategy was selected with initial success. However, 5-days later she was admitted again, this time with an anterior ST-segment elevation myocardial infarction.

Figure 16: Angiography (A) revealed progression of the angiographic findings (white arrows) with a 95% diameter stenosis and a TIMI 1 coronary flow that did not improve despite repeated intracoronary nitroglycerin administration. OCT imaging (B-C) of the most proximal aspect of the lesion confirmed the presence of an intramural hematoma (+ signs) with a reduced coronary lumen and also a proximal double lumen with a well-defined intimal-medial membrane.

Figure 17: Angiographically, distal coronary flow drastically improved after dilation at multiple levels with a 2.5 mm diameter scoring balloon (A) with complete resolution of the ST segment changes. A large lumen with multiple sites of “type 1” angiographic dissections (sites of contrast penetration of the false lumen) were visualized along the entire LAD with TIMI 3 coronary flow throughout (white arrows). OCT (B-D) demonstrated multiple ruptures of the intimal-medial membrane in a spiral configuration and a large true lumen. Asterisk = wire artefact.

Spontaneous coronary artery dissections