Spontaneous coronary artery dissection: the emerging role of coronary computed tomography

Abstract

Spontaneous coronary artery dissection (SCAD) is a cause of acute coronary syndrome and myocardial infarction, more frequent among young women. Invasive coronary angiography (ICA) is the gold standard for the diagnosis of SCAD, although the risk of propagating dissection flap is considerable. Therefore, coronary computed tomography angiography (CCTA) is an emerging alternative modality to diagnose SCAD with the advantage of being a non-invasive technique.

Clinicians should be aware of the predisposing conditions and pathophysiology to raise the pre-test probability of SCAD and select the most appropriate diagnostic tools.

In recent times, improvements in spatial and temporal resolution and the use of semi-automated software providing quantitative assessment make CCTA a valid alternative to ICA also for the follow-up. Moreover, CCTA may be helpful to screen and evaluate extra-coronary arteriopathies closely related to SCAD.

In this review, we illustrate the current and the potential role of CCTA in the diagnosis of SCAD, highlighting advantages and disadvantages of this imaging modality compared to ICA.

Graphical Abstract

Open in new tabDownload slide

Introduction

Spontaneous coronary artery dissection (SCAD) is a well-recognized cause of myocardial infarction, and it accounts for at least 4% of all acute coronary syndromes (ACSs), although the real incidence seems to be underestimated.¹ Indeed, the incidence of SCAD raises up to 35% of ACSs in women <60 years old, in pregnant female, and in patients with history of fibromuscular dysplasia or neuropsychiatric disorders.^2,3 In 1931, Pretty⁴ firstly described this condition in an autopsy performed in a 42-year-old woman who suddenly died. Although no randomized controlled trials are available, nowadays, there are large multicentre studies investigating various aspects of this disease, especially regarding diagnosis and therapeutic management. Some observational studies highlighted that invasive treatment with percutaneous coronary intervention (PCI) is associated with higher rate of complications, even in cases presenting with preserved vessel flow. Moreover, myocardial revascularization strategy does not reduce risk of long-term target vessel revascularization or recurrent SCAD, underscoring the need for standardized angiographic and clinical follow-up.^5,6

The diagnosis of SCAD is routinely made by invasive coronary angiography (ICA) in the contest of patients presenting with ACS, even if the diagnostic suspicion of SCAD is high. Recently, some authors highlighted the role of coronary computed tomography angiography (CCTA) in this setting, emerging as an alternative imaging modality for stable patients with SCAD in routine practice and second line diagnostic tool in ambiguous cases.⁷ The purpose of our review is to describe the potential role of CCTA in this condition and to evaluate key aspects and advantages regarding diagnosis and follow-up of SCAD.

Pathophysiology

SCAD is defined as a spontaneous (i.e. not-iatrogenic and not-traumatic) coronary wall injury due to acute detachment of one of the components of the arterial wall (intima), which leads to the development of a false lumen within the coronary artery. Thus, in dissected vessels, two lumens are present: (i) a true lumen, inside the tunica intima and (ii) a false lumen, outside the intima or intima-media complex. The presence of the false lumen may propagate longitudinally, and/or axially, leading to true lumen compression, which can result in blood flow obstruction and myocardial ischaemia.

The primary cause of SCAD is unclear, but two potential mechanisms have been proposed. The first one, called ‘inside-out or intima tear hypothesis’, is based on the primary creation of a dissection flap in the intima surface; this condition constitutes an entry point for coronary blood which propagates into the false lumen. The second one is called ‘outside-in or media haemorrhage hypothesis’ in which the intimal wall is healthy, but intramural microbleed occurs and leads to the development of haemorrhage into the tunica media.^8–10 Regardless, clear pathophysiologic mechanism leading to SCAD remains unclear, and reports regarding histopathological aspects are limited to case reports and small case series. A central role of inflammation and connective tissue abnormalities has been proposed.^11,12 In support of this hypothesis, peri-coronary inflammation and eosinophilic infiltrates have been described in some autopsies, even though the pathophysiological significance of these findings is controversial.^11,12 Moreover, the well-known association between SCAD and fibromuscular dysplasia (FMD) highlights a vascular fragility as primary condition predisposing to SCAD. Similarly to SCAD, FMD is a non-atherosclerotic and non-inflammatory vascular disease that manifests itself as arterial stenosis, aneurysm, tortuosity, and dissection.¹³ The histopathology of FMD is characterized by extensive disorganization of smooth muscle fibres and alterations of myofibroblasts in the intima and media.

Epidemiology and predisposing factors

The true prevalence of SCAD remains uncertain, and near 90% of SCAD were described in female patients.^1,14 As known, middle-aged women, with no or few traditional cardiovascular risk factors, represent the ‘typical’ SCAD patients, and the average age of presentation ranges from 45 to 53 years,¹⁴ On the contrary, SCAD is very uncommon among very young patients (<15 years of age) and in very old people (>85 years of age), especially outside the context of pregnancy and hereditary connective tissue disorders.¹⁵ SCAD seems to be the most common cause of pregnancy and peripartum-related myocardial infarction.¹⁶ The majority of pregnancy-associated SCAD occurs in the third trimester or in the early post-partum period, therefore a potential role of hormonal fluctuations must be considered regarding the pathogenesis.¹⁷ Additionally, use of contraceptive and post-menopausal hormones is a controversial predisposing factor.^18,19

Fibromuscular dysplasia is the most common co-existing condition, and this association has been described up to 86% of SCAD patients in at least one non-coronary territories. Extra-coronary vascular territories hit by FMD are renal, iliac, and cerebrovascular arteries, in order of prevalence.¹³ The real prevalence of FMD among SCAD patients could be also underdiagnosed because patients are often clinically asymptomatic, and routine screening in extra-coronary territory is underperformed in clinical practice.

Concerning genetic factors, SCAD has been frequently reported in association with inherited connective and vascular diseases, such as Marfan syndrome, Ehlers–Danlos syndrome, and pseudoxanthoma elasticum, but, only few studies described the association between familiar history of SCAD and the occurrence of this condition.²⁰

Several physical and emotional triggers are linked to the development of SCAD.²¹ Anxiety, depression, and neuropsychiatric disease may also predispose to SCAD. The frequency of external stressors is relatively high compared with atherosclerotic ACS. Moreover, intense exercise could be a mechanical trigger predisposing to SCAD in vulnerable subjects,^5,21 although the mechanism of this association needs to be clarified.

Diagnosis of SCAD

The clinical presentation is widely variable, ranging from simple chest pain to sudden cardiac death. Indeed, asymptomatic spontaneous coronary dissection is extremely rare. Chest pain is the most common presenting symptom; sustained or non-sustained ventricular arrhythmias were also described.²²

Almost all SCAD patients present with increase of myocardial injury biomarkers, with the exception of patients in which clinical presentation is very early or delayed.²³ The proportion of patients presenting with ST-segment elevation is considerable, and it depends on the amount of false lumen compression and the occurrence of total vessel occlusion. Therefore, the clinical presentation is similar to atherosclerotic ACS. In this setting, recognizing the typical patient phenotype suggestive for SCAD is of paramount importance to avoid misdiagnosis and delays.

Clinicians should identify the above-mentioned risk factors to raise the pre-test probability of SCAD.²⁴ It is true that in the emergency setting, it is difficult to distinguish SCAD from typical atherosclerotic myocardial infarction, therefore, ICA is up to now the predominant first-line imaging modality. Specific intracoronary imaging tools (i.e. optical coherence tomography [OCT] and intravascular ultrasound [IVUS]) could be used to improve and confirm diagnosis, although some operators prefer to avoid these modality due to the risk of propagating dissection flap.²⁵

A SCAD angiographic classification has been proposed,²⁶ as showed in Figure 1:

• Type 1: classic angiographic appearance with contrast dye staining of the coronary wall with multiple radiolucent lumens (Figure 1A);

• Type 2: diffuse stenosis of varying severity, in which the angiographic appearance consists of diffuse and smooth narrowing with an abrupt reduction in arterial caliper. SCAD commonly involves the mid to distal segments of coronary arteries. According to SCAD extension, type 2 could be divided into 2A if culprit vessel has normal caliber proximally and distally to dissection (Figure 1B), and 2B if the lesion extents up to the tip of the vessel (Figure 1C); and

• Type 3: focal or tubular stenosis; this pattern is the most challenging because of its likeness with atherosclerotic stenosis (Figure 1D).

Anyhow, this proposed SCAD classification has some limitations. One of the most important is that the SCAD-related total vessel occlusion does not fit into the above classification. For this reason, some authors proposed to add another SCAD type (i.e. type 4, Figure1E and F) defined as vessel occlusion that not met criteria for type 1, 2, or 3.

The prognostic implication of this proposed classification remains under-investigated, although data derived from a multicentre European registry showed a higher rate of adverse cardiovascular events among types 2A and 3 SCAD.²⁷

Role of computed tomography in acute setting

The use of CCTA in patients with low or intermediate risk for coronary artery disease is considered safe and appropriate.²⁸ In comparison to ICA, CCTA has the advantage of being a non-invasive imaging test with relatively acceptable spatial and temporal resolution.²⁹ In the last years, the role of CCTA in acute chest pain assessment has increased, and prominent data regarding the use of this imaging modality have been published (Table 1).

The diagnosis of SCAD may be challenging, and the ICA shows some limitation especially regarding focal lesions and vessel occlusion SCAD (i.e. type 4, Saw classification). Therefore, in recent years, improvements in CCTA techniques and protocols have expanded its indication. Retrospective protocol is preferable to the prospective in the SCAD setting, as it allows interpretation from different phases of cardiac cycle and cine-view, with the expense of higher radiation dose.

Nevertheless, the main limitation of CCTA is its low spatial resolution in evaluating small coronary arteries compared to ICA,³⁰ so, normal findings at CCTA cannot exclude diagnosis of SCAD.^14,31,32

Using CCTA, Tweet et al.,³³ in a small retrospective study, found four primary SCAD-related features, in order of prevalence: (i) abrupt luminal stenosis >50% over a length of 0.5 mm; (ii) intramural haematoma, defined as thickening within wall vessel and outside true lumen, corresponding to type 2 or 3; (iii) tapered luminal stenosis >50%; and (iv) dissection flap visualized as linear hypodensity, consistent with type 1 SCAD.

Although dissection flap occurs in minority of cases, this is the easiest angiographic pattern to recognize on CCTA, since culprit vessel appears with typical ‘double lumen’ pattern. Contrastingly, in the context of abrupt luminal stenosis pattern, it is challenging to distinguish between SCAD and other causes of vessel occlusion. In addition, it is also important to observe other findings. While coronary calcification and soft plaques are almost always absent in a context of SCAD,³⁴ epicardial fat stranding, beading (defined as >50% luminal change which increase and decrease in sequence), coronary ectasia, and myocardial bridge are additional high-risk features reported in SCAD cases.^33,35 Most importantly, as demonstrated by a large retrospective study, a severe tortuosity in coronary anatomy is associated with higher risk of SCAD³⁶ (Figure 2).

Over direct evaluation of coronary vessels, myocardial perfusion defects related to vessel injury may be present, although up to half of patients may not report this finding and it is not specific of SCAD.³³

A recent study retrospectively analysed consecutive in-hospital CCTA performed in a cohort of SCAD patients.³⁷ The study blindly compared CCTA findings with ICA and OCT data. CCTA was able to recognize SCAD in 78% of the patients, with four different morphological patterns. The most common finding was diffuse luminal narrowing described as a sleeve-like wall thickening; this pattern corresponded to intramural haematoma at OCT evaluation. All lesions that were not detected by CCTA scan were located on distal segment, highlighting the intrinsic limit of this methodic on evaluating small coronary arteries. Focal plaque-like stenosis (i.e. SCAD type 3 on ICA) mimics atherosclerotic pattern, so distinction from soft plaque is the most challenging aspect. However, the authors showed that plaque-like pattern was characterized by mid attenuation (68 Hounsfield Unit (HU) [56–95]), negative vessel remodelling, and absence of calcification spot; conversely, high-risk atherosclerotic plaques are frequently characterized by low attenuation (<30 HU), positive remodelling, spotty calcification, and napkin-ring sign³⁸ (Figures 3 and 4). Therefore, quantitative analysis is a fundamental aspect that may improve diagnostic accuracy of CCTA on recognizing SCAD lesions.³⁹

A summary of SCAD and atherosclerotic disease features is described in Table 2.

A novel emerging field of search concerns the role of peri-coronary adipose tissue inflammation in the context of SCAD disease. The amount of these adipose depots can be easily investigated by CCTA scan, and some studies demonstrated a positive association with mortality and coronary artery disease.^29,40 Regarding SCAD, small cohorts of patients underwent CCTA were analysed with controversial results,^41,42 so further larger studies are warranted to better clarify a possible role of peri-coronary fat tissue attenuation in this subset.

In conclusion, CCTA may be helpful in the acute setting of SCAD, especially when dissection flap and ‘double-lumen’ pattern is detected. Operator experience and quantitative analysis are needed in challenging case. Nevertheless, at the current state of art, the role of CCTA remains unclear and not recommended as first-line investigation because no dedicated large studies are available in this setting. We illustrated advantages and disadvantages of ICA and CCTA in Table 3.

Role of computed tomography during follow-up

Recurrent de novo SCAD has been frequently reported following the index event, so patients suffering from SCAD have a higher risk of recurrence compared to general population. Some studies showed that up to 30% of patients suffering from de novo SCAD reported at least one previous event.^6,43 Saw et al.⁴⁴ described that at median follow-up of 3.1 years, recurrent myocardial infarction occurred in 16.8% of patients, of which recurrent SCAD in 10.4%. Hypertension is associated with increases risk of recurrent SCAD, conversely beta-blocker therapy demonstrated a protective role in this setting.⁴⁴ Moreover, patients suffering from SCAD present a higher rate of hospital readmission compared with non-SCAD myocardial infarction. Interestingly, conservative management for SCAD is associated with lower rate of hospital readmission among SCAD patients.⁴⁵

In most cases, the treatment of SCAD is often conservative for stable patients, whereas percutaneous coronary intervention or coronary artery bypass grafting is reserved for patients with ongoing ischaemia or haemodynamic instability. Among patients enrolled in DIssezioni Spontanee COronariche (DISCO) registry, nearly two third of patients were treated conservatively,⁴⁶ but angiographic follow-up was routinely made in a minority of patients. In fact, in current clinical practice, there is no clear recommendation about angiographic and clinical follow-up in patients suffering from SCAD. Observational data demonstrated that SCAD lesions treated conservatively achieved angiographic healing in most cases, but nowadays, studies aimed to evaluate angiographic follow-up remain scarce.^47,48 Hence, currently, there are two main issues regarding the diagnostic work-up after the index event: how, and when to perform angiographic follow-up in SCAD patients. As we have already pointed out, ICA with intracoronary imaging represents the gold standard for the diagnosis of acute SCAD, but the risk of propagating dissection flap due to intrinsic vessel fragility limits its feasibility for angiographic follow-up.⁴⁹ On the other hand, safety of CCTA makes this technique the preferred modality for follow-up, both in asymptomatic and in symptomatic patients. The sensitivity and specificity of CCTA for assessment of dissection healing were 72% and 53.8%, respectively.⁵⁰ Roura et al. reported data on feasibility of non-invasive follow-up among a relatively small cohort of SCAD patients. CCTA was performed at a median follow-up of 121 days, showed complete healing of the culprit vessel in 83% of patients. Moreover, patient without signs of dissection had excellent prognosis at long-term follow-up.⁵¹ As regards when to perform angiographic follow-up, available data are even more limited. A single study investigated the correct timing of CCTA in SCAD patients during follow-up. It demonstrated that the optimal timing to assess SCAD healing was 80 days, with sensitivity of 76.9% and specificity of 84%. Moreover, when CCTA was performed within 80 days from the index event, the occurrence of healing was only 12.5%, conversely, when CCTA was performed after 80 days from SCAD event, the rate of healing raises to 71.4%.⁵⁰

Given the highlighted intrinsic limitations of the CCTA in detecting SCAD, especially in the context of intramural haematoma and plaque-like morphologies, a comparison with the CCTA performed at index event should be performed (Figure 5). Therefore, it would be useful to perform CCTA even in the acute phase, regardless of ICA, in order to increase report accuracy and reproducibility during follow-up period. Although these data are probably too limited to make specific recommendations, we propose a simple algorithm for the diagnosis and follow-up of SCAD (Figure 6).

Extra-coronary imaging

The association between SCAD and other arteriopathies was described in many studies.^52–54 As already described, FMD is the common vascular coexisting condition with SCAD. During index diagnostic procedure, as ICA as CCTA, routine evaluation of abdominal aorta and collateral major vessels should be made. The typical FMD angiographical appearance is described as multifocal vascular disease with areas of alternating stenosis and dilatation, the so-called string-of-bed pattern.

Moreover, cerebrovascular aneurism and carotid disease have also been described in association with SCAD, but the real prevalence is unclear. Of course, the prevalence of extra-coronary disease involvement is most common among patients with genetic connective diseases, thus in this subset is useful to screen extra-coronary district.

Conclusions

SCAD is a not-so-rare cause of myocardial infarction and sudden cardiac death, especially in middle-age female patients and pregnant women. Although current recommendations indicate ICA as gold standard modality for diagnosis, CCTA has emerged as potential alternative in haemodynamically stable patients. Quantitative vessel analysis can help to raise the accuracy of CCTA in recognizing SCAD lesions, also in plaque-like morphologies. Moreover, its role during follow-up period is of utmost importance considering the high rate of recurrence in SCAD patients.

Acknowledgements

None.

Funding

This study was endorsed by the Italian Association of Hospital Cardiologists (ANMCO) and by the Italian Society of Echocardiography and Cardiovascular Imaging (SIECVI).

Data availability

No new data were generated or analysed in support of this research.

References

Author notes