Coronary artery disease Original research Risks and benefits of percutaneous coronary intervention in spontaneous coronary artery dissection Deevia Kotecha,1 Marcos Garcia-­Guimaraes ‍ ‍,2,3 Diluka Premawardhana,1 Dario Pellegrini ‍ ‍,4,5 Clare Oliver-­Williams,6 Vasiliki Bountziouka,1 Alice Wood,1 Nalin Natarajan,1 Robert Jackson,1 Nathan Chan,1 Jan Ziaullah,1 Roby D Rakhit,7 Stephen P Hoole,8 Tom W Johnson,9 Jacek Kadziela,10 Peter Ludman ‍ ‍,11 Nilesh J Samani,1 Angela H E M Maas,4 Robert-­Jan van Geuns ‍ ‍,4 Fernando Alfonso,2 David Adlam ‍ ‍1 ►► Additional supplemental ABSTRACT women.1–3 Percutaneous coronary intervention (PCI) material is published online Objective To investigate percutaneous coronary in acute SCAD has been associated with an increased only. To view, please visit the journal online (http://​dx.​doi.​ intervention (PCI) practice in an international cohort of risk of procedural complications compared with org/​10.​1136/​heartjnl-​2020-​ patients with spontaneous coronary artery dissection atherosclerotic ACS, with ‘PCI technical failure’ 318914). (SCAD). To explore factors associated with complications reported in 36%–53% of cases.4 5 Complications and study angiographic and longer term outcomes. include iatrogenic dissection, haematoma exten- For numbered affiliations see sion and emergency coronary artery bypass grafting Methods SCAD patients (n=215, 94% female) who end of article. underwent PCI from three national cohort studies were (CABG).4–6 When managed conservatively, most Correspondence to investigated and compared with a matched cohort of SCAD heals completely with restoration of apparently Dr David Adlam, Department conservatively managed SCAD patients (n=221). normal coronary architecture at follow-­up.7 8 For this of Cardiovascular Sciences and Results SCAD-­PCI patients were high risk at reason, the American and European Consensus State- NIHR Biomedical Research presentation with only 8.8% undergoing PCI outside ments on SCAD both advocate conservative manage- Centre, University of Leicester, ment over PCI, where this is possible.1 2 the context of ST-­elevation myocardial infarction/cardiac Leicester LE3 9QP, Leicestershire, UK; ​da134@​leicester.​ac.​uk arrest, thrombolysis in myocardial infarction (TIMI) Accepting that conservative management is 0/1 flow or proximal dissections. PCI complications favoured ‘where possible’, there remains a signif- DK and MG-­G contributed occurred in 38.6% (83/215), with 13.0% (28/215) icant population of SCAD patients where conser- equally. serious complications. PCI-­related complications were vative management risks extensive infarction.9 10 RG, FA and DA are joint senior associated with more extensive dissections (multiple vs For these higher risk cases, it remains unclear to authors. single American Heart Association coronary segments, what extent complications are a necessary cost of OR 1.9 (95% CI: 1.06–3.39),p=0.030), more proximal improving coronary perfusion and if a more proac- Received 30 December 2020 dissections (proximal diameter per mm, OR 2.25 tive approach to revascularisation may be important Revised 6 April 2021 for selected patient groups. (1.38–3.67), p=0.001) and dissections with no contrast Accepted 12 April 2021 Published Online First penetration of the false lumen (Yip-­Saw 2 versus 1, In this study, our objective was to investigate 18 May 2021 OR 2.89 (1.12–7.43), p=0.028). SCAD-­PCI involved outcomes in a large observational SCAD-­PCI cohort long lengths of stent (median 46mm, IQR: 29–61mm). derived from three national registries. In particular, Despite these risks, SCAD-­PCI led to angiographic to investigate the balance of risks and benefits of improvements in those with reduced TIMI flow in 84.3% PCI in a SCAD population. (118/140). Worsening TIMI flow was only seen in 7.0% (15/215) of SCAD-­PCI patients. Post-­PCI major adverse METHODS cardiovascular and cerebrovascular events (MACCE) and Study population left ventricular function outcomes were favourable. The study was conducted in accordance with the Conclusion While a conservative approach to Declaration of Helsinki. Patients and the public revascularisation is favoured, SCAD cases with higher (http://​ beatscad.​ org.​ uk/) were involved in the ►► http://​dx.​doi.​org/​10.​1136/​ risk presentations may require PCI. SCAD-­PCI is funding, concept and dissemination of the study heartjnl-​2021-​319524 associated with longer stent lengths and a higher risk findings. Patients were recruited from the UK, of complications but leads to overall improvements Dutch and Spanish SCAD registries with SCAD in coronary flow and good medium-­term outcomes in events from 2003 to 2019. More details of these © Author(s) (or their patients. registries are provided in the online supplemental employer(s)) 2021. Re-­use data. Patients in all studies gave written informed permitted under CC BY. consent to data collection and contact for assess- Published by BMJ. ment of outcomes. To cite: Kotecha D, INTRODUCTION Garcia-­Guimaraes M, Spontaneous coronary artery dissection (SCAD) is an Definition of PCI Premawardhana D, et al. important cause of acute coronary syndromes (ACS), PCI was defined by an apparent angiographic intent Heart 2021;107:1398–1406. predominantly affecting young to middle-­ aged to intervene through, at minimum, the passage of a 1398   Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 Coronary artery disease coronary guidewire except where guidewire passage was solely (LMS) stenting or CABG. There is no established definition of for the purpose of diagnostic intracoronary imaging. SCAD-­PCI complications, although these are acknowledged to All SCAD patients with PCI were included. To fully charac- be somewhat different from those in atherosclerotic ACS. The terise SCAD in the PCI cohort, each national registry selected interventional representatives of the three national registries a matching number of consecutively recruited conservatively therefore adopted these consensus definitions, largely based on managed controls for comparison. As the aim was to investi- those reported in related studies.1 2 gate the clinical and angiographic differences between PCI Postoperative outcomes were major adverse cardiovascular and conservative cohorts (rather than to investigate outcomes and cerebrovascular events (MACCEs; death, stroke, myocar- in patients with equivalent presentations), cases were selected dial infarction or revascularisation occurring at any point after blinded to clinical and angiographic findings and without prior discharge from the index SCAD event) and recurrence (a new matching for demographic or angiographic features. Any cases angiographically confirmed SCAD occurring after discharge or controls not felt to be definite SCAD by any observer during from the index episode and either anatomically or temporally the subsequent angiographic analysis were excluded. This small separated from the first event). Time to recurrence and time to number of exclusions means the final analysed groups are of MACCE were also recorded. slightly different sizes. Iatrogenic dissection cases were only included if there was clear evidence of SCAD preceding and/ or anatomically remote from the iatrogenic injury. Iatrogenic Statistical analysis dissections were further classified according to whether they Patient, clinical and intervention characteristics were summarised occurred during the diagnostic or interventional phase of the by median and IQR for continuous measures or number and procedure. percentage for categorical variables. Comparisons of number and total length of stents between SCAD-­ PCI patients who Confirmation of SCAD diagnosis developed complications and uncomplicated SCAD-­PCI patients All patients had an angiographically confirmed diagnosis of were made using Kruskal-­Wallis test as the variables were not SCAD on image review. All angiograms, with intracoronary normally distributed. imaging where available, were assessed by experienced inter- Logistic regression was used to calculate the OR and 95% ventional cardiologists representing at least two of the national CIs of the risk of any complication and serious complications registries with any initial disagreements resolved by consensus. associated with patient, clinical and intervention characteris- Patients with atherosclerotic, traumatic or iatrogenic dissec- tics. Unadjusted and two adjusted models were created. The tion (except where the latter complicated definite SCAD) were first adjusted model included age, sex and ethnicity. The second excluded. adjusted model included additionally all patient, clinical and intervention characteristics significantly associated with the outcome of interest. The latter was performed to assess which of Patient, SCAD and intervention characteristics the characteristics remained significant when all variables were Demographic information, medical history and a detailed adjusted for (eg, total stent length and number of stents). Due history of the SCAD event were obtained from the patient’s to the inclusion of variables pertaining to stents, only stented medical record. PCI procedure details were collected from individuals were included in the analyses. An a priori decision procedure reports. Clinical characteristics included: myocar- was made that interaction terms would not be included, as we dial infarction type (cardiac arrest, non-­ST-­elevation myocar- did not believe that the joint effect of any two of these variables dial infarction (NSTEMI) or ST-­elevation myocardial infarction would be higher than expected from the sum of their individual (STEMI)) - STEMI was defined as typical ECG changes occur- effects. ring at any stage prior to angiography; cardiac arrest as any Two Kaplan-­Meier plots were constructed to compare: (1) cessation of circulation requiring resuscitation occurring before time to MACCE and (2) time to recurrence for SCAD-­PCI and the diagnostic angiographic procedure), pregnancy-­ associated SCAD-­non-­PCI patients. As the proportional hazards assump- SCAD (P-­SCAD) as SCAD occurring during gestation or within tion was met, as assessed by Kaplan-­Meier plot and the test of 12 months of delivery. Not all registries held the full contem- Grambsch-­Therneau,11 Cox proportional hazards regression poraneous medical records to allow an accurate determination models were used to calculate the HR and 95% CI for the risk of admission blood pressure to reliably determine cardiogenic of MACCE and recurrence associated with patient, clinical shock according to accepted standard definitions. These data and intervention characteristics, with adjustment for age, sex were therefore not included. and ethnicity. To assess differences in risk between SCAD-­PCI and SCAD-­non-­PCI patients, an interaction term between clin- Angiographic analysis ical management (PCI vs conservative) and the predictor was Details of SCAD findings were recorded from angiographic included. images. Additionally, see online supplemental data. All statistical analyses were performed using STATA release V.16 (Stata Corp, College Station, Texas, USA) and R for statis- Outcomes tical computing.12 PCI complications were recorded from angiographic images and defined as: aorto-­ostial iatrogenic dissection; haematoma extension; loss of flow in the stented vessel or a significant side RESULTS branch (>2 mm diameter); and vessel perforation as shown Cohort demographics by extravasation of contrast (contrast penetration of the false Two hundred and fifteen SCAD-­ PCI patients were included. lumen during PCI was not considered a complication). Serious Patient, clinical and intervention characteristics are summarised PCI complications were defined as aorto-­ostial iatrogenic dissec- in table 1. Patients were predominantly female and white Euro- tion, complications resulting in reduced flow in a proximal pean with a median age at the time of SCAD of 48 years. In the epicardial coronary vessel or leading to unplanned left mainstem SCAD PCI cohort there were 12 (5.6%) P-­SCAD cases. Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 1399 Coronary artery disease Table 1 Descriptive characteristics of the SCAD cohort, by clinical Table 1 Continued management Conservative PCI Conservative PCI (n=221) (n (%)) (n=215) (n (%)) (n=221) (n (%)) (n=215) (n (%))  Wiring 0 0.0% 14 6.5% Patient characteristics Maximum stent diameter, mm (median, – – 3.0 (2.5–3.5) Total 221 50.7 215 49.3 IQR) Age at first SCAD event, years (median, 49 (43–55) 48 (42–54) Total number of stents (median, IQR) – – 2 (1-3) IQR) Total length of stents, cm (median, IQR) – – 46 (28-61) Ethnicity* Proximal diameter, mm (median, IQR) – – 2.53 (2.15–3.1)  White European 212 95.9 197 92.5 Length of lesion, mm (median, IQR) – – 38.6 (26.6–56.2)  Not white European 9 4.1 16 7.5 Total volume of haematoma, mm3 – – 57.0 (30.0– Sex (median, IQR) 102.7)  Female 203 91.9 203 94.4 Final TIMI grade flow  Male 18 8.1 12 5.6  0 (no flow) – – 16 7.4% Pregnancy status  1 – – 11 5.1%  Not pregnant (female) 192 86.9 191 88.8  2 – – 23 10.7%  Pregnant (female) 11 5.0 12 5.6  3 (good flow) – – 165 76.7% Smoking Outcomes  Never smokers 140 (63.3) 146 (67.9) Any complication* – – 83 38.6%  Ex-­smokers 55 (24.9) 44 (20.5) Serious complication* – – 28 13.0%  Current smokers 26 (11.8) 25 (11.6) Time to MACCE (median, IQR) 2.10 (1.06–3.61) 2.32 (1.04–4.28) Diabetes mellitus MACCE 22 9.5% 31 14.4%  No >216 (98.2) >210 (99.1) Time to recurrence* (median, IQR) 2.5 (1.2–4.7) 4.4 (2.7–6.3)  Yes <5 (1.8) <5 (0.9) Recurrence* 15 6.8% 13 6.1% Hypertension *Missing values: ethnicity/race: 2 (0.5%), maximum stent diameter: 74 (34.4%), total number of stents: 1 (0.6%), total stent length: 14 (9.0%), proximal diameter: 47 (21.9%),  No 162 (73.3) 166 (77.2) lesion length: 74 (34.4%), volume of haematoma: 79 (36.7%), tortuosity index: 1 (0.2%),  Yes 59 (26.7) 49 (22.8) time to recurrence: 2 (0.5%), aspirin: 5 (1.2%), DAPT: 8 (1.8%), beta-­blocker: 5 (1.2%), ACE: Dyslipidaemia 4 (0.9%), statin: 5 (1.2%), recurrence: 2 (0.5%). †Of those not taking DAPT at discharge, three in the PCI group and two in the non-­PCI  No 185 (83.7) 187 (87.0) group were anticoagulated. Of those taking DAPT, 65%, 30% and 4% in the non-­PCI group  Yes 36 (16.3) 28 (13.0) and 37%, 49% and 14% in the PCI group were taking clopdiogrel, ticagrelor or prasugrel, respectively. Clinical characteristics DAPT, dual antiplatelet therapy; NSTEMI, non-­ST-­elevation myocardial infarction; PCI, NSTEMI 131 59.3 74 34.4 percutaneous coronary intervention; SCAD, spontaneous coronary artery dissection; STEMI, STEMI 77 34.8 119 55.3 ST-­elevation myocardial infarction; TIMI, thrombolysis in myocardial infarction. Cardiac arrest 13 5.9 22 10.2 Left main stem vessel affected <5 2.3 14 6.5 Left anterior descending artery affected 112 50.7 155 72.1 Clinical and angiographic presentation Left circumflex artery affected 83 37.6 48 22.3 Two hundred and twenty-­one blindly selected SCAD-­non-­PCI Right coronary artery affected 58 26.2 29 13.5 controls were included for comparison of clinical presentation AHA coronary segment involved as shown in figure 1 and table 1. SCAD-­PCI patients were more  Proximal 17 7.7% 64 29.8% likely to present with STEMI and with TIMI 0/1 than SCAD-­  Mid 52 23.5% 78 36.3% non-­PCI (figure 1A, table 1). In keeping with this, while Yip-­Saw  Distal 84 38.0% 45 20.9% classification type 2 SCAD was the most common angiographic  Branch 68 30.8% 28 13.0% class at presentation, SCAD-­PCI patients were more likely to More than one vessel involved* 32 14.5% 19 8.8% present with Yip-­Saw classification type 4 (occlusions) compared More than one segment in the vessel 60 27.1% 83 38.6% with SCAD non-­ PCI patients (table 1, online supplemental involved efigure 1A). Tortuosity index* (median, IQR) 4 (2-6) 3 (2-5) A radial initial approach was undertaken in 74.3% of Type 1 28 12.7% 26 12.1% SCAD-­PCI patients and in 84.2% of SCAD non-­PCI patients. Type 2 153 69.3% 118 54.9% The left anterior descending coronary artery was the most Type 3 26 11.8% 14 6.5% affected vessel. A coronary heatmap of affected coronary Type 4 14 6.3% 57 26.5% segments is shown in figure 1B (a graph of all affected Amer- Taking aspirin* 205 94.5% 210 98.1% ican Heart Association (AHA) coronary segments is shown in Taking DAPT* 161† 74.5% 193† 91.0% online supplemental efigure 1B). There was greater involve- Taking beta-­blocker* 173 79.7% 189 88.3% ment of proximal and midvessel lesions in the SCAD-­ PCI Taking ACE inhibitors* 146 67.0% 158 73.8% compared with the SCAD-­non-­PCI group (table 1, figure 1C). Taking statins* 155 71.1% 171 80.3% There was more multisegment disease but not multivessel Intervention quantitative coronary analysis disease in SCAD-­PCI patients than SCAD-­non-­PCI patients Type of intervention (table 1, figure 1D).  Conservative 221 100.0% – – SCAD-­PCI lesion characteristics, as assessed by quantitative coronary angiography (3D-­QCA) are shown in online supple-  Stent 0 0.0% 156 72.6% mental efigure 2. The median upstream vessel diameter was  Balloon 0 0.0% 45 20.9% 2.7 mm (IQR: 2.3–3.2, n=168) with a downstream diameter Continued of 1.8 mm (IQR: 1.5–2.3, n=126) (online supplemental figure 1400 Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 Coronary artery disease Figure 1 Clinical presentation of SCAD-­PCI (n=215) and SCAD-­non-­PCI patients (n=221), including (A) initial presentation, (B) coronary heatmap of affected segments (see also online supplemental efigure 1 for full segment-­by-­segment analysis), (C) TIMI flow and (D) multisegment and multivessel disease. PCI, percutaneous coronary intervention; SCAD, spontaneous coronary artery dissection; TIMI, thrombolysis in myocardial infarction. 2A). The median minimal luminal diameter was 0.6 mm (IQR: the maximum bending angle for affected vessels (median 55°; 0.4–0.9, n=124), and the median minimal luminal area was IQR: 43–70°, n=168) (online supplemental efigure 2C). The 0.5 mm2 (IQR: 0.3–1.1 mm2, n=121; online supplemental median volume of haematoma displaced by stenting was 54.0 efigure 2B). The median of the measured lesion bending mm3 (IQR: 27.3–113.9 mm3, n=86) (online supplemental angles was 47° (IQR: 34–61°, n=148), which was close to efigure 2D). Figure 2 Details of the PCI procedure in SCAD intervention patients (n=215): (A) interventional strategy, (B) number of stents deployed, (C) stent length compared with lesion length and (D) coronary heat map of stented AHA coronary segments. AHA, American Heart Association; PCI, percutaneous coronary intervention; SCAD, spontaneous coronary artery dissection. Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 1401 Coronary artery disease Nine patients underwent intracoronary imaging only and of the smallest deployed stent in each case was 2.5 mm (IQR: were included in the SCAD-­non-­PCI group, 62 patients under- 2.5–3.0 mm). Figure 2D shows a coronary heatmap of sites of went intracoronary imaging as part of SCAD-­PCI. stenting also shown by AHA coronary segment. Areas of residual dissection mostly occurred in distal coronary locations such as Characteristics of SCAD revascularisation AHA coronary segments, 8, 4, 14 and 16, while initially unaf- Of patients undergoing PCI, 7.0% (15/215) had failed an initial fected segments that were more frequently stented were mainly trial of conservative management. Four patients underwent in proximal segments (1, 5 and 11). 10.3% (16/156) of stented emergency CABG, two as a planned revascularisation strategy cases required stenting into the LMS. after angiography and two after failed or complicated PCI. The Of all SCAD-­PCI cases, 33.0% (71/215) required one or more nature of the PCI procedure undertaken is shown in figure 2A, of either unplanned LMS stenting, excessive stent number (≥4) 72.6% (156/215) of patients underwent stenting, mostly with or stent length (≥50 mm). drug eluting stents, 20.9% (45/215) had balloon angioplasty, 5 with cutting balloons and 6.5% (14/215) underwent wiring Outcomes of SCAD-PCI only. The number of stents deployed is shown in figure 2B Angiographic exemplars of typical complications occurring (mean 2.3 (range 1–8) per stented case). Twenty-­ three cases during PCI are shown in figure 3. (10.6% of all SCAD-­PCI cases) required four or more stents. The impact of PCI on TIMI flow is shown in figure 4A (with In the 142 patients with both total number of stents and total findings for the SCAD-­STEMI subpopulation shown in online stent length recorded, the median total length of deployed stents supplemental efigure 4A). In those with reduced TIMI flow, was 46 mm (IQR: 28.5–60.8 mm; mean 51.5 mm) with 64 improvements were seen in 84.3% (118/140) of patients. In all cases (29.8% of all SCAD-­PCI cases) requiring ≥50 mm stents patients undergoing PCI, worsening of TIMI flow was seen in (figure 2C). 64.1% (100/156) of stented cases were left with 7.0% (15/215) (figure 4A). Changes in TIMI flow according to residual unstented areas of dissection. The median diameter initial TIMI are shown in online supplemental efigure 3. Figure 3 Angiographic exemplars of complications (both serious and not serious) occurring during percutaneous coronary intervention. Iatrogenic dissection: A1: extensive dissection of mid-­right coronary artery (solid arrows) with further haematoma distally (dotted arrow), A2: iatrogenic dissection from ostium to mid vessel and A3: extensive stenting from proximal to distal vessel; Haematoma extension: B1: proximal-­mid left anterior descending haematoma (solid arrows), B2: treated with stenting to the ostium, initially with only small prestent residual haematoma (dotted arrow), B3: but later leading to propagation of haematoma into circumflex (solid arrows); Side-­branch occlusion: C1: mid-­left anterior descending stenosis (solid arrows), C2: confirmed on OCT to be due to SCAD and C3: stenting leads to occlusion of diagonal branch (dotted arrow). OCT, optical coherence tomography; SCAD, spontaneous coronary artery dissection. 1402 Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 Coronary artery disease Figure 4 (A) Changes in thrombolysis in myocardial infarction (TIMI) flow for SCAD-­PCI patients and (B) PCI complications in the SCAD-­PCI cohort (n=215). CABG, coronary artery bypass grafting; PCI, percutaneous coronary intervention; SCAD, spontaneous coronary artery dissection. PCI complications occurred in 38.6% (83/215) of SCAD-­PCI Long-term outcomes patients with the the most common being haematoma extension Follow-­u p information was available on 436 patients (58/215, 27.0%) and iatrogenic dissection (18/215, 8.4%; nine (median follow-­ u p 900 days, IQR: 440–1590). Most occurring during the diagnostic phase and nine after initiation patients retained good cardiac ejection fraction at follow-­up of PCI) (figure 4B). Two complications occurred during intra- regardless of revascularisation strategy (online supplemental coronary imaging, the remainder during SCAD-­PCI. A similar efigure 5). There was no difference in MACCE events frequency and distribution of complications was noted in the between SCAD-­ PCI and SCAD-­ non-­ PCI patients (online SCAD-­STEMI subpopulation (online supplemental efigure 4B). supplemental efigure 6A, PCI: 14.4%, 31/215, non-­ PCI: Stenting into the false lumen was noted in one case. 9.5%, 21/221; MACCE components - recurrent AMI (PCI: Compared with patients without complications, patients who 9.3%, 20/215, non-­PCI: 7.7%, 17/221); revascularisation developed complications had a longer stent length (median 52 (PCI: 4.7%, non-­PCI: 1.4%); stroke (PCI: 1.5%, non-­PCI: (IQR: 38–64) PCI complications; median 35 (IQR: 26.8–56) 0.7%); death (PCI: 1.4%, non-­PCI: 0.5%)). There was also PCI no complications; p=0.004). Despite this, 47.0% (39/83) no difference in the incidence of recurrent SCAD between showed improvements in TIMI flow grade at procedure end, the PCI and non-­PCI groups (online supplemental efigure 12.0% (10/83) showed a deterioration and 41.0% (34/83) no 6B, PCI: 6.1%, 13/214, non-­PCI: 6.8%, 15/220). change in TIMI flow grade. Of those with impaired TIMI flow grade at the start of the procedure, 33/45 (73.3%) showed an improvement in flow in at least one AHA coronary segment DISCUSSION of a major epicardial coronary territory, while only 5 (11.1%) We present the largest international observational study of PCI showed a deterioration. Subjectively in terms of overall angio- in SCAD. We report first that PCI in current SCAD practice graphic coronary perfusion, 44 of 83 (53.0%) cases with is appropriately reserved for more serious clinical and angio- complications showed an improved situation at procedure end, graphic presentations including STEMI, reduced TIMI flow and 17 (20.5%) showed a deterioration and 22 (26.5%) showed no proximal disease. Second, that complications in SCAD-­PCI are overall change. high, occurring in 38.6% of cases, one-­third of which are serious 13.0 (28/215) of SCAD patients undergoing PCI suffered (defined as iatrogenic dissection, unplanned LMS stenting, loss subjectively serious complications (16 cases of iatrogenic dissec- of flow in proximal AHA coronary segment and emergency tion, 16 cases requiring unplanned LMS stenting, 3 cases demon- CABG). Third, that despite the higher rates of complications strated loss of flow in a proximal AHA coronary segment and 2 and longer stent lengths required for SCAD-­PCI, intervention patients required CABG because of PCI failure). in this context generally resulted in improvements in measures Predictors of any PCI complications and serious compli- of coronary flow. Finally, that measures of more extensive and cations are shown in table 2 and online supplemental etable more proximal dissections as well as those without contrast 1, respectively. Predictors of complications grouped into penetration into the false lumen are associated with a higher risk markers of more extensive dissections (lesion length, volume of complications. of haematoma and multisegment dissections), more prox- The first important risk of PCI in SCAD is the requirement imal dissections (proximal vs distal, proximal diameter and for multiple, often small calibre stents and longer stent lengths. maximum stent diameter) and a lack of angiographic contrast SCAD-­PCI on average required an extra 0.7 stents and an extra penetration of the false lumen (Yip-­Saw classification type 2). 21.4 mm of stent compared with UK national audit data (mean Of these multisegment dissections, LMS dissections and the 1.64 stents; 30.1 mm stented length per stented case), where use of larger diameter stents remained predictors of serious atherosclerotic disease will predominate.13 Overall, one-­third of complications. SCAD-­PCI patients required at least one of unplanned LMS PCI, When all variables associated with the outcome were included ≥4 stents or ≥50 mm stent length. Furthermore, despite exten- in the model, along with age, sex and ethnicity, the total number sive stenting, there remained unstented dissection in 64.1% of of stents remained associated with the risk of any complication SCAD-­PCI cases. Despite this, over the duration of available in SCAD-­PCI patients (OR=1.90, 95% CI 1.26 to 2.85), and MACCE follow-­up (median 30 months), there was no overall the maximum stent diameter remained associated with risk of increase in MACCE in the SCAD-­PCI group (despite the more serious complications in SCAD-­PCI patients (OR=2.62, 95% CI severe characteristics at presentation). Importantly, longer term 1.28 to 5.39) (online supplemental etable 2 and 3). adverse sequelae of the more extensive stenting in this relatively Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 1403 Coronary artery disease Table 2 Association between patient, clinical and intervention characteristics and odds of any complication in a SCAD cohort Unadjusted Age, sex and ethnicity adjusted Cases/n OR (95% CI) P value Cases/n OR (95% CI) P value Patient characteristics Age at first SCAD event, per year 83/215 1.01 (0.98 to 1.04) 0.631 n/a n/a n/a Ethnicity (white European vs non-­white European)* 83/213 0.81 (0.29 to 2.26) 0.684 n/a n/a n/a Male versus female 83/215 0.78 (0.23 to 2.69) 0.658 n/a n/a n/a Pregnant female versus non-­pregnant female 79/203 0.50 (0.13 to 1.92) 0.316 79/201 0.51 (0.12 to 2.15) 0.360 Grading of tortuosity for all vessels imaged, per unit* 83/214 0.95 (0.85 to 1.07) 0.437 83/213 0.94 (0.83 to 1.06) 0.349 Clinical characteristics Type of myocardial infarction   STEMI versus NSTEMI 83/215 0.92 (0.51 to 1.67) 0.794 83/213 0.91 (0.50 to 1.65) 0.753   Cardiac arrest versus NSTEMI 0.68 (0.25 to 1.88) 0.462 0.74 (0.27 to 2.08) 0.573 Left main stem vessel affected 83/215 1.64 (0.56 to 4.87) 0.369 83/217 1.67 (0.55 to 5.12) 0.366 Left anterior descending artery affected 83/215 0.92 (0.50 to 1.70) 0.794 83/217 0.90 (0.49 to 1.66) 0.736 Left circumflex artery affected 83/215 1.05 (0.55 to 2.03) 0.874 83/217 1.05 (0.55 to 2.02) 0.879 Right coronary artery affected 83/215 1.14 (0.52 to 2.53) 0.742 83/217 1.18 (0.54 to 2.59) 0.679 AHA coronary segment involved   Mid versus proximal 83/215 0.79 (0.40 to 1.54) 0.485 83/213 0.74 (0.37 to 1.46) 0.383   Distal versus proximal 0.32 (0.14 to 0.76) 0.010 0.29 (0.12 to 0.71) 0.007   Branch versus proximal 0.73 (0.30 to 1.81) 0.501 0.65 (0.26 to 1.66) 0.367 More than one vessel involved 83/215 0.92 (0.35 to 2.44) 0.869 83/213 1.03 (0.38 to 2.78) 0.960 More than one segment within the vessel involved 83/215 1.77 (1.01 to 3.11) 0.046 83/213 1.90 (1.06 to 3.39) 0.030 Yip-­Saw classification based on appearance when first imaged  Type 2 versus type 1 83/215 2.90 (1.13 to 7.42) 0.026 83/213 2.89 (1.12 to 7.43) 0.028  Type 3 versus type 1 2.71 (0.70 to 10.57) 0.150 2.70 (0.69 to 10.64) 0.155  Type 4 versus type 1 0.44 (0.14 to 1.39) 0.163 0.42 (0.13 to 1.33) 0.141 Intervention details Type of intervention  Balloon versus stent 83/215 0.08 (0.02 to 0.26) <0.001 83/213 0.07 (0.02 to 0.24) <0.001  Wiring versus stent 0.60 (0.19 to 1.87) 0.379 0.53 (0.17 to 1.68) 0.278 Maximum stent diameter, per mm* 65/141 1.83 (1.04 to 3.25) 0.037 65/140 1.93 (1.06 to 3.49) 0.030 Total number of stents, per additional stent* 74/155 1.39 (1.06 to 1.80) 0.015 74/154 1.48 (1.12 to 1.95) 0.006 Total length of stents, per mm* 66/142 1.01 (1.00 to 1.02) 0.137 66/141 1.01 (1.00 to 1.02) 0.049 Proximal diameter, per mm* 65/168 2.11 (1.32 to 3.37) 0.002 65/166 2.25 (1.38 to 3.67) 0.001 Length of lesion, per mm* 59/141 1.02 (1.00 to 1.03) 0.068 59/139 1.02 (1.00 to 1.03) 0.060 Volume of haematoma, per mm3* 59/136 1.01 (1.00 to 1.01) 0.016 59/135 1.01 (1.00 to 1.01) 0.014 TIMI flow  1 versus 0 (no flow) 83/215 0.65 (0.14 to 3.04) 0.582 83/213 0.66 (0.14 to 3.10) 0.596  2 versus 0 (no flow) 0.85 (0.24 to 3.06) 0.802 0.86 (0.24 to 3.12) 0.817  3 (good flow) versus 0 (no flow) 0.41 (0.15 to 1.16) 0.093 0.42 (0.15 to 1.21) 0.107 Only individuals who underwent stenting were included in analyses of the maximum of stent diameter, the total number of stents and the total stent length. ARB, angiotensin II receptor blockers; DAPT, dual antiplatelet therapy; NSTEMI, non-­ST-­elevation myocardial infarction; SCAD, spontaneous coronary artery dissection; STEMI, ST-­elevation myocardial infarction; TIMI, thrombolysis in myocardial infarction. young SCAD population are currently unclear, and further data with complications, suggesting perhaps that connections between are needed to assess this. true and false lumen allow decompression of the false lumen The high rate of complications is another risk of SCAD-­PCI, with stent inflation, making haematoma propagation and there- which has been widely reported4 5 and is confirmed in this study. fore PCI complications less likely. Stent number and stent length However, in keeping with a previous study in SCAD-­STEMI,14 also correlated with PCI complications, but for these factors, it most patients undergoing PCI in this study were high risk at is difficult to determine cause versus effect. Coronary tortuosity, presentation with only 8.8% undergoing PCI outside the context known to be increased in SCAD,15 did not increase the risk of PCI of STEMI/cardiac arrest, TIMI 0/1 flow or proximal dissec- complications. tions. Conservative management in these patients would likely There were also important benefits of SCAD-­ PCI. Improve- incur higher jeopardy. Furthermore, this analysis confirms that, ments in TIMI flow grade were seen after PCI in 84.3% of cases even within this group, complications are more likely to occur with 38.6% improving three grades, 23.6% improving two grades in patients with more extensive and more proximal dissections, and 22.1% improving one grade. Only 7.0% of cases saw a dete- where larger volumes of false lumen haematoma pose a greater rioration in TIMI flow grade with PCI. Interestingly, these gains risk from displacement or propagation during stenting. Also of were mostly seen in cases where stenting rather than more limited potential clinical utility is the finding that angiographic features wiring or plain old balloon angioplasty strategies was adopted. of intramural haematoma without contrast penetration of the Taken together these data suggest that PCI should remain false lumen (Yip-­Saw classification type 2) were more associated the reserve of high-­ r isk SCAD presentations. Operators 1404 Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 Coronary artery disease should be aware that the risk of complications is higher analysis phase. Criteria for both complications and serious compli- than for atherosclerotic cases and that long lengths of stent cations were selected by experienced SCAD interventional cardiol- may be required. Cases with extensive dissections, prox- ogists but are not established definitions. Benchmarking data from imal dissections and those with no contrast penetration into the BCIS audit are discussed to allow a meaningful comparison of the false lumen may be at particularly high risk. However, stenting data. These data are from a single national jurisdiction, despite this, it is possible in most of these high-­risk cases although no significant transnational differences were observed to achieve subjective improvements in coronary perfusion. online supplemental etable 4. Data on cardiogenic shock were not It should be noted that there is also potential to further available, although the reported incidence in larger SCAD series is improve the risk–benefit balance in SCAD-­P CI by consid- low (2%17). ering measures to reduce potentially avoidable complica- tions such as the use of less aggressive more coaxial catheter shapes that might reduce the risk of iatrogenic dissection in CONCLUSIONS patients with clinically possible SCAD. The point of equi- While a conservative approach to revascularisation in SCAD poise between PCI and a conservative approach to revascu- is optimal where possible, sometimes the clinical presentation larisation remains unclear. A previous propensity matched (STEMI, cardiac arrest, poor TIMI flow and proximal occlusive analysis of angiographically and clinically matched cases dissection) mandates intervention to improve coronary perfu- showed no difference in convalescent infarct size between sion and reduce myocardial injury. This study demonstrates PCI and conservatively managed cases.16 This analysis was that although more extensive stenting may be required, with an inevitably largely limited to intermediate risk anatomies. A elevated risk of procedural complications, improved coronary flow clinical trial in predefined higher risk anatomies is required and good medium-­term outcomes can be achieved with PCI. More to address this question definitively. It was also not possible extensive dissection, proximal segment location and an absence of from this analysis to assess the efficacy of different PCI tech- contrast penetration of the false lumen associate with the highest niques. Future studies should be designed to investigate this. risk of PCI complications. Author affiliations Limitations 1 Department of Cardiovascular Sciences and NIHR Biomedical Research Centre, This is an observational study, and therefore, we cannot conclude University of Leicester, Leicester, UK 2 that the associations demonstrated are causative. Our data are Department of Cardiology, Hospital Universitario de La Princesa, Instituto de derived from a SCAD survivor cohort and do not include infor- Investigación Sanitaria Princesa (IIS-­IP), Universidad Autónoma de Madrid, CIBER-­CV, Madrid, Spain mation on early SCAD-­ PCI non-­ survivors. Although the three 3 Department of Cardiology, Hospital del Mar, Barcelona, Spain national SCAD registries recruit all comers and include patients 4 Department of Cardiology, Radboud University Medical Center, Nijmegen, The referred from hospital clinicians, primary care physicians and Netherlands 5 self-­referrals, there remains the potential for selection bias. The Department of Cardiology, ASST Papa Giovanni XXIII, Bergamo, Italy 6 non-­PCI cohort were selected blinded to clinical and angiographic Department of Biostatistics, University of Leicester, Leicester, UK 7 Department of Cardiology, Royal Free Hampstead, London, UK findings. The difference in cohort sizes is due to a small number 8 Royal Papworth Hospital and NIHR Cambridge Biomedical Research Centre, of exclusions arising after case selection. Exclusions arose when a Cambridge, UK 9 case was felt by any observer not to be definite SCAD on review of University Hospitals Bristol and Weston NHS Foundation Trust, Bristol, UK 10 angiography (plus intracoronary imaging when present) during the Department of Interventional Cardiology and Angiology, National Institute of Cardiology, Warsaw, Poland 11 Institute of Cardiovascular Sciences, University of Birmingham, Birmingham, UK Key messages Correction notice Since Online First publication, author name ’Vasiliki Boutziouka’ has been corrected to ’Vasiliki Bountziouka’. What is already known on this subject? Acknowledgements We are grateful for the support of spontaneous coronary ►► Percutaneous coronary intervention (PCI) in spontaneous artery dissection (SCAD) survivors and our clinical colleagues throughout the UK, coronary artery dissection (SCAD) is associated with high Spain and the Netherlands and the leadership of the ESC-­ACCA SCAD Study Group. rates of procedural complications. We specifically acknowledge the support of Jenny Middleton, Jane Plume, Donna ►► A conservative approach to revascularisation is recommended Alexander, Sue Sterland, Daniel Lawday, Tara Maitland and Andrea Marshall for all their support for SCAD research. We are grateful to the British Cardiovascular where possible. Intervention Society (BCIS) National Audit of Adult Interventional Procedures for PCI outcome data presented in the discussion to help readers benchmark the stenting What might this study add? data presented. ►► The ‘cost’ of PCI in SCAD is longer stent lengths and more Contributors DA conceived the study and is responsible for the overall content as procedural complications. guarantor. DK, MG-­G, DiP, DaP, AW, NN, RJ, NC and JZ collected and analysed data, ►► The ‘benefit’ of PCI in SCAD is improved coronary flow. RJvG and FA provided oversight of data collection and undertook image analysis, PL ►► Outcomes from SCAD-­PCI in terms of left ventricular function supported provision of comparator BCIS audit data, CO-­W and VB led the statistical and major adverse cardiovascular and cerebrovascular events analysis. RR, SPH, TJ, JK NJS and AM provided senior overview of SCAD registries. All authors reviewed and modified the manuscript. are generally good. Funding This study was supported by the BeatSCAD, the British Heart Foundation PG/13/96/30608, the National Institute for Health Research (NIHR) rare disease How might this impact on clinical practice? translational collaboration and the Leicester NIHR Biomedical Research Centre. ►► Case selection for PCI in SCAD is critical. Most cases can be Competing interests DA has received research funding to support a clinical managed conservatively but for high-­risk cases (eg, proximal research fellow from Abbott Vascular Inc. He has also received funding from to mid vessel occlusions) intervention is required. AstraZeneca Inc. for unrelated research and has undertaken consultancy to support ►► While PCI is associated with a significant risk of procedural research funds for General Electric Inc. TJ has received consultancy/speaker fees from complications requiring long stent lengths, improved flow Abbott Vascular, Boston Scientific, Medtronic and Terumo. All other authors have and good long-­term outcomes can be achieved. reported that they have no relationships relevant to the contents of this paper to disclose. Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914 1405 Coronary artery disease Patient consent for publication Not required. 2 Hayes SN, Kim ESH, Saw J, et al. Spontaneous coronary artery dissection: current state of the science: a scientific statement from the American heart association. Circulation Ethics approval The UK SCAD study (ISRCTN42661582) was approved by the 2018;137:e523–57. UK National Research Ethics Service (14/EM/0056) and the UK Health Research 3 Hayes SN, Tweet MS, Adlam D, et al. Spontaneous coronary artery dissection: JACC Authority. The Spanish SCAD registry (​ClinicalTrials.​gov NCT03607981) was state-­of-­the-­art review. J Am Coll Cardiol 2020;76:961–84. approved by national and institutional ethical review boards (Hospital Universitario 4 Tweet MS, Eleid MF, Best PJM, et al. Spontaneous coronary artery dissection: La Princesa) and is endorsed by the Spanish Society of Cardiology Working Group on revascularization versus conservative therapy. Circ Cardiovasc Interv 2014;7:777–86. Cardiac Catheterization and Interventional Cardiology. The Dutch SCAD registry was approved by the institutional ethical review board of Radboud University Medical 5 Saw J, Aymong E, Sedlak T, et al. Spontaneous coronary artery dissection: association Centre (Radboudumc, Nijmegen, the Netherlands). with predisposing arteriopathies and precipitating stressors and cardiovascular outcomes. Circ Cardiovasc Interv 2014;7:645–55. Provenance and peer review Not commissioned; externally peer reviewed. 6 Prakash R, Starovoytov A, Heydari M, et al. Catheter-­Induced Iatrogenic Coronary Data availability statement Data from this study are available on reasonable Artery Dissection in Patients With Spontaneous Coronary Artery Dissection. JACC request. Cardiovasc Interv 2016;9:1851–3. 7 Alfonso F, Paulo M, Lennie V, et al. Spontaneous coronary artery dissection: long-­term Supplemental material This content has been supplied by the author(s). It follow-­up of a large series of patients prospectively managed with a "conservative" has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have therapeutic strategy. JACC Cardiovasc Interv 2012;5:1062–70. been peer-­reviewed. Any opinions or recommendations discussed are solely those 8 Rogowski S, Maeder MT, Weilenmann D, et al. Spontaneous coronary artery of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and dissection: angiographic follow-­up and long-­term clinical outcome in a predominantly responsibility arising from any reliance placed on the content. Where the content medically treated population. Catheter Cardiovasc Interv 2017;89:59–68. includes any translated material, BMJ does not warrant the accuracy and reliability 9 Jackson R, Al-­Hussaini A, Joseph S. Spontaneous coronary artery dissection: of the translations (including but not limited to local regulations, clinical guidelines, pathophysiological insights from optical coherence tomography. JACC Cardiovasc terminology, drug names and drug dosages), and is not responsible for any error Imaging 2019;12:2475–88. and/or omissions arising from translation and adaptation or otherwise. 10 Franco C, Starovoytov A, Heydari M, et al. Changes in left ventricular function after Open access This is an open access article distributed in accordance with the spontaneous coronary artery dissection. Clin Cardiol 2017;40:149–54. Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits 11 Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on others to copy, redistribute, remix, transform and build upon this work for any weighted residuals. Biometrika 1994;81:515–26. purpose, provided the original work is properly cited, a link to the licence is given, 12 Team RC. R: a language and environment for statistical computing. Vienna, Austria: R and indication of whether changes were made. See: https://​creativecommons.​org/​ Foundation for Statistical Computing, 2020. https://wwwR-projectorg/ licenses/​by/​4.​0/. 13 BCIS. BCIS national audit of adult interventional procedures. Available: https:// wwwbcisorguk/tag/bcis-audit/ [Accessed 5 Oct 2020]. ORCID iDs 14 Lobo AS, Cantu SM, Sharkey SW, et al. Revascularization in Patients With Spontaneous Marcos Garcia-­Guimaraes http://​orcid.​org/​0000-​0001-​8509-​6184 Coronary Artery Dissection and ST-­Segment Elevation Myocardial Infarction. J Am Coll Dario Pellegrini http://​orcid.​org/​0000-​0002-​5545-​585X Cardiol 2019;74:1290–300. Peter Ludman http://​orcid.​org/​0000-​0002-​7237-​0946 15 Eleid MF, Guddeti RR, Tweet MS, et al. Coronary artery tortuosity in spontaneous Robert-­Jan van Geuns http://​orcid.​org/​orcid.​org/​0000-​0003-​2943-​5932 coronary artery dissection: angiographic characteristics and clinical implications. Circ David Adlam http://​orcid.​org/​0000-​0002-​0080-​9884 Cardiovasc Interv 2014;7:656–62. 16 Al-­Hussaini A, Abdelaty AMSEK, Gulsin GS, et al. Chronic infarct size after spontaneous coronary artery dissection: implications for pathophysiology and clinical REFERENCES management. Eur Heart J 2020;41:2197–205. 1 Adlam D, Alfonso F, Maas A, et al. European Society of cardiology, acute 17 Saw J, Starovoytov A, Humphries K, et al. Canadian spontaneous coronary cardiovascular care association, SCAD Study group: a position paper on spontaneous artery dissection cohort study: in-­hospital and 30-­day outcomes. Eur Heart J coronary artery dissection. Eur Heart J 2018;39:3353–68. 2019;40:1188–97. 1406 Kotecha D, et al. Heart 2021;107:1398–1406. doi:10.1136/heartjnl-2020-318914