Introduction

Cardiac arrest in the cardiac catheterization laboratory (CCL), or in-lab cardiac arrest (ILCA), while uncommon, represents the most severe complication of invasive cardiovascular procedures. ILCA occurring during coronary angiography, due to ventricular fibrillation (VF), is managed with immediate defibrillation, usually without the need for cardiopulmonary resuscitation (CPR). However, ILCA due to pulseless electrical activity (PEA) or asystole in the setting of severe cardiogenic shock (CGS), or due to recurrent pulseless ventricular tachycardia (VT) or VF, can become refractory to standard therapy and is associated with poor outcomes. Regardless of the etiology, ILCA requires thoughtful and systematic management to prevent mortality. This article provides a concise summary of the contemporary management of cardiac arrest occurring in the CCL (as opposed to out-of-hospital cardiac arrest managed in the CCL).

Incidence

The incidence of ILCA in patients undergoing percutaneous coronary intervention (PCI) is quite low. Studies from the 1990s report an incidence of 0.2% (among 51,985 patients treated at the Mayo Clinic between 1990 and 2000) and 1.3% (among 4363 patients treated at St. Paul’s and Vancouver General Hospitals between 1996 and 1999) [1, 2]. The Mayo Clinic investigators noted a decrease in the incidence of ILCA from 0.35 to 0.13% from 1990–1994 to 1995–2000 [1]. More recent studies report an incidence of 0.36% in 8738 patients who underwent coronary angiography and PCI between 2009 and 2013, and 0.5% in 13,122 patients who underwent coronary angiography between 2012 and 2016 [3, 4]. The incidence is expected to vary based on patient characteristics and the indications for the procedure. A significant proportion of patients in the above referenced studies were undergoing emergency coronary angiography and PCI for acute coronary syndromes (ACS), which are more likely to be associated with ILCA. ILCA may occur intra-procedurally or following PCI. In one study, 54% of reported cases of ILCA occurred after the completion of PCI. Repeat coronary angiography demonstrated significant findings, such as occlusion of a major side-branch, no-reflow, or stent thrombosis in more than 90% of cases [2]. In studies with a high proportion of patients undergoing PCI for acute myocardial infarction (AMI), VF and pulseless VT were the cause of ILCA in 2/3rds of the cases, and the majority of cases occurred after initial balloon inflation [1, 2]. In other studies which included a high percentage of patients with CGS, PEA arrest was reported more commonly [3, 4].

Risk factors

ILCA occurs in the setting of sustained and recurrent ventricular arrhythmias, severe bradycardia, asystole, and PEA due to severe/global myocardial ischemia and severely reduced myocardial contractility. These pathophysiological mechanisms usually occur in the setting of ST elevation MI (STEMI) and CGS, which account for the majority of cases of ILCA. In the study by Webb et al., the majority of patients who suffered ILCA were undergoing emergency ad hoc PCI for STEMI and/or CGS [1]. The incidence of ILCA in patients with STEMI and CGS in this study was 15% and 10% respectively. Conversely, ILCA was uncommon in patients undergoing PCI for unstable angina (0.5%) and rare in those undergoing elective PCI for stable angina (0.02%). In the study by Wagner et al., the incidence of ILCA in patients with CGS was 17%, and CGS and STEMI were present in 62% and 75% of patients experiencing ILCA respectively [3]. Cardiac arrest and CGS frequently occur together and the prognosis for a patient with both is worse than the presence of either cardiac arrest or CGS alone. Patients undergoing high-risk PCI on the last remaining vessel, or for complex multi-vessel CAD with poor cardiac reserve, can develop ILCA due to ventricular arrhythmias and/or pump failure from severe global ischemia. In the PROTECT-2 trial, 2.2–3.2% of patients undergoing high-risk PCI required CPR or experienced ventricular arrhythmia requiring treatment [5].

Other conditions associated with or resulting in ILCA include prior cardiac arrest, severe left ventricular systolic heart failure (especially with associated severe right ventricular dysfunction), severe pulmonary hypertension, massive pulmonary embolism, cardiac tamponade, PCI complications (type 3 perforation, massive air embolism, left main dissection, no-reflow, abrupt vessel closure), and exsanguination from left ventricular free wall or pulmonary artery rupture.

ILCA can also occur during structural interventions such as transcatheter aortic valve replacement (TAVR), due to acute severe aortic insufficiency, occlusion of left and/or right coronary ostia from bulky valve tissue or calcium, ventricular arrhythmias, valve embolization, refractory cardiogenic shock post-rapid cardiac pacing, annular rupture, and perforation of the ventricle from a pacemaker or stiff wire with resultant cardiac tamponade. The incidence of emergent cardiopulmonary bypass (CPB) for hemodynamic support during TAVR was 5.2%, based on data from the PARTNER A and B trials and non-randomized registries between 2011 and 2012. The incidence of emergent CPB was much higher for valve-in-valve procedures (25.4%) compared with non-valve-in-valve procedures (4.7%); p < 0.0001 [6]. In a single-center study of patients undergoing TAVR between 2012 and 2014, 4.3% (10/230) of patients required extracorporeal membrane oxygenation (ECMO) for ILCA [7].

Predictors of mortality and outcomes following in-lab cardiac arrest

In an observational study of 114 patients who developed ILCA during coronary angiography or PCI, a prolonged procedure, emergency catheterization, and prior coronary artery bypass graft (CABG) surgery predicted lower procedural and hospital survival on multivariate analysis [1]. Advanced age, intra-procedural cardiac arrest (compared with cardiac arrest in the recovery room) and hypotension prior to cardiac arrest predicted lower hospital survival, whereas an initial rhythm of PEA as opposed to VF was associated with lower procedural survival. Advanced age, cardiogenic shock, no-reflow, intra-procedural cardiac arrest, and side-branch occlusion were predictors of poor survival in unadjusted analysis in another observational study [2].

A systematic management strategy can result in higher survival of patients with ILCA compared with in-hospital cardiac arrest occurring outside the CCL [8]. Recent studies have reported 50–60% survival following treatment of ILCA with a programmatic approach, including ECMO [9•, 10•]. Importantly, it has been noted that patients who survive to hospital discharge following ILCA typically do well with 80–90% 1-year survival, with full neurological recovery in all survivors noted in one study [1, 9•]. Earlier institution of ECMO after onset of refractory cardiac arrest has been associated with improved outcomes. In an observational study of 35 patients, the median duration (25th–75th quartiles) of CPR in the CCL was 18 (10–41) min in patients who achieved return of spontaneous circulation (ROSC), compared with 50 (33–60) min in those who did not achieve ROSC (p = 0.007) [11]. CPR duration was also significantly lower in survivors (10 (8–25) min) compared with non-survivors (45 (30–60) min) (p = 0.001). In this study, survival to discharge was 0% among patients who arrived to the CCL already in cardiac arrest survived compared with 33% survival in those that developed cardiac arrest in the CCL. Median aortic diastolic blood pressure was significantly higher among those that obtained ROSC compared to non-ROSC patients (30 (22–40) mmHg vs 19 (14–28) mmHg; p = 0.012). While some studies have shown an improved rate of survival in patients who have an initial shockable rhythm such as VT/VF compared with PEA, most studies have shown no impact of the presenting rhythm on survival [1, 9•, 10•, 11]. In the study by Parr et al., higher baseline serum creatinine and 24-h lactate were independently associated with increased 30-day mortality, and patients with an initial serum creatinine >150 μmol/L (> 1.7 mg/dL) had 100% mortality [9•]. Observational studies have shown that survival to hospital discharge following institution of ECMO in the CCL depends on the ability to reverse the cardiovascular pathology that resulted in cardiac arrest. Patients who are placed on ECMO without long delay regain a stable cardiac rhythm and undergo correction of reversible causes of cardiac arrest, such as with percutaneous or surgical revascularization, and have better outcomes than patients without recovery of a perfusing rhythm and uncorrected/uncorrectable cardiac pathology [10•, 12, 13].

Management of in-lab cardiopulmonary arrest

Prevention and planning

The possibility of ILCA must be considered in any critically ill patient referred to the CCL (Fig. 1). In such patients, who are usually in severe CGS (SCAI shock stage C or D) [14] or undergoing high-risk PCI, preventative and planning steps should be undertaken to prevent the development of ILCA. These include the following:

  1. 1.

    A mechanical circulatory support (MCS) device should be placed pre-PCI in patients undergoing PCI on the last remaining vessel supplying viable myocardium such that occlusion of the vessel would result in catastrophic consequences. In addition, MCS should be strongly considered in patients prior to complex PCI of coronary arteries supplying large myocardial territory in the presence of moderate to severe left ventricular systolic dysfunction, especially in the presence of poor hemodynamics (hypotension, low cardiac output, severely elevated left ventricular end diastolic pressure) or severe CGS (SCAI shock stages C, D). By maintaining sufficient cardiac output, mean arterial and coronary perfusion pressure during PCI, MCS devices can prevent the downward spiral of progressive myocardial dysfunction which results in cardiac arrest.

  2. 2.

    Endotracheal intubation may need to be delayed until establishment of vascular access and placement of an MCS device. Patients in CGS with severe tachycardia and a narrow pulse pressure may develop ILCA due to loss of adrenergic drive after being administered deep sedation for intubation. Rather, non-invasive ventilation, such as bi-level positive airway pressure (BiPAP), can be used to support oxygenation while awaiting establishment of vascular access.

  3. 3.

    Early notification of anesthesia and perfusion services about the possible need for endotracheal intubation and ECMO when critically ill patients are being transported to the CCL. The presence of these services in the CCL before the case begins minimizes critical delays in the event the patient sustains ILCA.

Fig. 1
figure 1

Algorithm for management of in-lab cardiopulmonary arrest. AHA, American Heart Association; ACLS, Advanced Cardiovascular Life Support; MCS, mechanical circulatory support device; V-A ECMO, veno-arterial extracorporeal membrane oxygenation; ROSC, return of spontaneous circulation. *Telephone consultation with ECMO team should be considered to determine candidacy for ECMO and advanced heart failure therapy.

Full size image

Resuscitation considerations

The primary objective during cardiopulmonary arrest is to maintain vital organ perfusion, and therefore, the CCL team should focus on the delivery of high-quality CPR, immediate defibrillation for shockable rhythms, airway management, medication administration, and adjunctive therapies based on AHA guidelines (Table 1). The Interventional Cardiologist should delegate defined roles to CCL staff to initiate CPR, perform defibrillation, and administer medications, so that he/she can focus on tasks to reverse the cause of cardiac arrest. If a vascular sheath is in place, it should be attached to a pressure transducer, and the goal should be to achieve a systolic arterial pressure of 90–100 mmHg with each chest compression. Additional help should be called for which may be limited if the cardiac arrest occurs after hours. If cardiac arrest is refractory to initial therapies, ancillary support from cardiac anesthesia, perfusion, and cardiac surgery should be requested as appropriate. The major challenge of ILCA is to ensure the delivery of uninterrupted high-quality CPR while attempting to perform tasks intended to reverse the immediate cause of cardiac arrest. In this context, while mechanical chest compression devices have not been shown to improve survival compared with manual chest compressions, they may be better suited to allow simultaneous CPR and procedures which require fluoroscopy, such as PCI and placement of MCS. If a mechanical chest compression device is unavailable, interruptions in manual chest compressions should be minimized as much as possible.

Table 1 Practical considerations for management of refractory in-lab cardiac arrest. CPR cardiopulmonary resuscitation, AHA American Heart Association, ACLS Advanced Cardiovascular Life Support, MCD mechanical chest compression devices, ECMO extracorporeal membrane oxygenation, CCL cardiac catheterization laboratory
Full size table

Mechanical chest compression devices

Mechanical chest compression devices (MCDs), including the mechanical piston driven and the load-distribution band types, have been extensively studied in out-of-hospital cardiac arrest (OHCA). An example of the mechanical piston device is the Lund University Cardiac Arrest System or LUCAS™ (Physio-Control Inc./Jolife AB, Lund, SWE)—a gas- (oxygen or air) or electric-powered device that produces a consistent chest compression rate and depth. It incorporates a suction cup at the end of a piston that attaches to the sternum and returns the sternum to the starting position when it retracts. The load-distributing band device, such as the Autopulse™ (Zoll Corporation), is a circumferential chest compression device composed of a pneumatically or electrically actuated constricting band and backboard [15]. Randomized controlled trials (RCT) have shown no benefit of either form of MCD compared with manual chest compressions for CPR for OHCA [16, 17].

There have been no RCT comparing MCD with manual chest compressions for ILCA. In a case series of 32 patients with ILCA who underwent CPR with the LUCAS™ MCD, PCI was attempted in 87% of cases and was successful in 81%. CGS was present in 62% of patients and 31% had a culprit lesion in the left main coronary artery. 15 out of 32 patients (47%) achieved return of spontaneous circulation (ROSC) and were alive on leaving the CCL. 8 out of 32 (25%) survived to hospital discharge with good neurological outcome, and all but one patient were alive at 1 year [3]. In a retrospective study, 31 patients, of which 19 suffered refractory in-hospital cardiac arrest (15 in the CCL and 4 in the Emergency Room), underwent CPR with the LUCAS™ MCD between 2014 and 2016 [18•]. PCI was attempted during CPR with a MCD in 16/31 (52%) patients and was successful in all attempted cases. 23/31 (74%) patients achieved return of spontaneous circulation (ROSC) compared with 5/12 (42%) historical control patients who received manual chest compressions. There was no difference in survival to hospital discharge between the two groups. Since the subjects were not randomized to either treatment, the difference in ROSC rates was likely due to significant differences in patient characteristics. The highest rate of ROSC (95%) was noted in patients who underwent CPR with a MCD and also received a MCS (predominantly ECMO). Patients who underwent CPR with a MCD but did not receive a MCS achieved ROSC in only 11% of cases. Survival to hospital discharge was <15% and did not differ among patients receiving manual chest compressions or CPR with a MCD (with or without MCS). These studies demonstrate the feasibility of performing successful PCI during CPR with a MCD. Conversely, performing successful PCI during manual chest compressions is extremely challenging and increases radiation exposure to the person performing compressions. At the same time, there are concerns for a higher incidence of serious injuries with MCDs compared with manual chest compressions, and the time delay in correct application of the MCD can significantly prolong the period of absent CPR.

Because of lack of data from randomized controlled trials, the 2015 American Heart Association Guidelines Update for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care state that evidence does not demonstrate a benefit with the use of either piston driven or load-distribution band devices versus manual chest compressions in patients with cardiac arrest [15]. The guidelines state that manual chest compressions remain the standard of care for the treatment of cardiac arrest, but piston driven and load-distribution band devices may be a reasonable alternative for use by properly trained personnel (class IIb, LOE B-R). The use of piston driven or load-distribution band assisted CPR may be considered in specific settings where the delivery of high-quality manual chest compressions may be challenging or dangerous for the provider (e.g., limited rescuers available, prolonged CPR, during hypothermic cardiac arrest, in a moving ambulance, in the angiography suite, during preparation for ECPR), provided that rescuers strictly limit interruptions in CPR during deployment and removal of the devices (class IIb, LOE C-EO). The use of a LUCAS device in a patient who sustained ILCA pre-TAVR and survived following emergent TAVR during ongoing mechanical chest compressions has also been described [19].

Impella

There is limited clinical data regarding use of the Impella CP (Abiomed Inc. Danvers, MA) MCS for management of ILCA. An elegant animal study was performed by Lotun et al to assess the effect of an Impella 2.5 device, either without chest compressions or combined with a MCD (LUCAS) for CPR, on survival with favorable neurological outcome when placed prior to the onset of VF arrest in swine [20•]. The control group consisted of animals that received conventional manual chest compression CPR following onset of arrest. The combination of Impella and MCD was associated with the highest neurologically intact survival (56%) compared with 0% survival in animals receiving only manual chest compressions. The integrated area of coronary perfusion pressure (CPP), calculated as the difference between aortic and right atrial diastolic pressure over time, was highest in the animals receiving uninterrupted chest compressions with a MCD and Impella 2.5 in place. CPR provides higher aortic systolic and lower right atrial diastolic pressures than Impella alone, and moves blood across the lungs to fill the left heart, which is then unloaded by the Impella. The authors hypothesized that the combination of Impella and a MCD therefore produces a higher integrated CPP, which in turn results in a higher neurologically favorable survival following cardiac arrest. Frequent interruptions of manual compression to simulate real-world CPR resulted in a very poor outcome in this animal model.

In a small case series of 8 patients with refractory PEA arrest, due to acute MI or arrhythmia in the setting of decompensated heart failure, persisting after 10 min of mechanical chest compressions and ACLS, an Impella CP device was placed in the CCL during ongoing CPR [21]. 5 out of 8 patients experienced intermittent ROSC before Impella placement and 6 out of 8 underwent PCI; 4 patients underwent PCI following and 2 patients underwent PCI prior to Impella placement. In 6 out of 8 patients mechanical chest compressions were stopped as sufficient Impella flow was achieved, but 2 patients required extended manual chest compressions before sufficient flow was achieved. Duration of support was 63 + 25 h, and 50% survived to hospital discharge. There was a high rate of vascular complications (50%, of which 38% required surgery) and renal failure requiring dialysis (63%). Impella can result in ROSC in the setting of refractory ILCA which persists despite ECMO or CPR and PCI. This may be due to acute unloading effect of a distended severely hypocontractile left ventricle [22].

Extracorporeal membrane oxygenation

The use of veno-arterial (V-A) extracorporeal membrane oxygenation (ECMO) to support CPR and ACLS during cardiac arrest has been variously labelled as ECLS or ECPR. While some observational studies have reported improved rate of survival to discharge and 1-year survival following ECPR compared with conventional CPR, others have shown no difference in outcome, and all are affected by significant bias due to confounding variables [23, 24]. There are no randomized controlled studies of the use of ECPR for in-hospital cardiac arrest. Several case series reporting outcomes of in-hospital and out-of-hospital cardiac arrest patients, treated with ECMO placed in the CCL, have been published. The CCL represents a uniquely suited environment for rapid institution of ECMO due to the availability of fluoroscopy, supplies, and support personnel.

In an observational study of 127 patients with cardiac arrest, long-term survival was significantly higher when ECMO was initiated in the CCL compared with initiation elsewhere in the hospital (50% versus 14.7%, p < 0.001) [25] In a recent study, ROSC was achieved in 100% of 14 patients who were treated with a MCD and ECMO, compared with 53% in 17 patients treated with a MCD without ECMO [18•]. However, of the 14 patients, only 5 patients survived to ECMO decannulation, and only 1 patient (7%) survived to hospital discharge. In this study, the incidence of STEMI in the total cohort of 31 patients was 32% and PCI was performed in only 52%, which may explain the poor overall survival. In another recently reported series, 62 patients with witnessed in-ambulance or in-hospital cardiac arrest or CGS were managed with ECMO in the CCL of a large tertiary care hospital between 2010 and 2018 [9•]. 39 out of the 62 patients had developed ILCA. ACS was present in the majority of the patients (STEMI 51%, NST-ACS 18%). Among patients with ILCA, PEA was noted in 51% and VT/VF in 49%, median duration of CPR before initiation of ECMO was 38 (interquartile range or IQR 20–48) min, and duration of ECMO support was 48 (IQR 17–97) h. Approximately 10% received a ventricular assist device or Impella, 20% underwent CABG and 5% received cardiac transplantation. 29 out of 62 (46.8%) patients survived to hospital discharge and the majority of survivors were alive at 1 year. There was no difference in survival between patients who had ECMO placed during CPR versus those who had ECMO placed for CGS after ROSC. The most common causes of in-hospital death were multi-system organ failure and anoxic brain injury. ECMO was associated with a high rate of complications. More than one in three required renal replacement therapy, one in three patients suffered a stroke, one in five developed impaired lower extremity perfusion requiring revision, and eight out of ten patients required blood transfusion. A study from the Mayo Clinic evaluated results in 25 consecutive patients receiving ECMO in the CCL between 2010 and 2017 [10•]. The majority (17/25) had undergone attempted PCI for ACS, and the remainder were treated for other indications such as massive pulmonary embolism, VT, and mitral regurgitation. A third of the patients had an IABP or Impella in place. 84% of patients developed cardiac arrest in the CCL and the median duration of CPR prior to ECMO initiation was 36 min. The initial rhythm was PEA in 67% and VT/VF in 33%. 16 out of 25 (64%) patients were undergoing active CPR during ECMO insertion. The 30-day survival was 40%, with 80% showing cardiac recovery after prolonged ECMO support (mean duration 116 h), and 20% requiring a durable ventricular assist device or cardiac transplantation. Severe bleeding was common with an average of 21 PRBC units transfused per patient and more than half of the patients requiring vascular surgical intervention. Institution of ECMO in 14 patients with in-hospital cardiac arrest, of whom a high percentage underwent PCI, was associated with a 57% survival rate, and time from arrest to ECMO initiation was predictive of survival [26].

These studies demonstrate that ECMO is very effective in achieving ROSC, and can result in 40–50% survival rates when used relatively early in patients with a potentially reversible cause of cardiac arrest, such as severe myocardial ischemia correctable by PCI. However, due to absence of randomized controlled data, the 2019 American Heart Association Focused Update on Advanced Cardiovascular Life Support concluded that there is insufficient evidence to recommend the routine use of ECPR for patients with cardiac arrest. The guidelines state that ECPR may be considered for selected patients as rescue therapy when conventional CPR efforts are failing in settings in which it can be expeditiously implemented and supported by skilled providers (class 2b; level of evidence C-LD) [27]. In several situations, even after initiation of ECMO, a perfusing rhythm cannot be obtained, unless the underlying cause of cardiac arrest is addressed by other procedures. Examples include PCI of the last remaining vessel/conduit, restoration of flow into right ventricular branches in a patient who develops refractory VF following primary PCI of a thrombotic right coronary artery occlusion, placement of an Impella device to decompress the left ventricle, pericardiocentesis to treat cardiac tamponade, or correction of severe acid-base and electrolyte disturbances. While ECMO maintains perfusion of vital organs like the brain, it increases LV afterload and wall stress which may adversely affect recovery of LV function. Absolute contraindications for ECMO include an unrecoverable heart and a patient who is not a candidate for heart transplant or ventricular assist device (VAD), advanced age (typically older than 75 years), chronic organ dysfunction (emphysema, cirrhosis, renal failure), compliance limitations, and prolonged CPR without adequate tissue perfusion.

Challenges, unanswered question, and future directions

The biggest challenge in the management of refractory ILCA is the lack of high-quality data such as RCT to support the routine use of devices and strategies to improve survival. The challenge is to identify refractory ILCA patients who are good candidates for ECMO early in the resuscitation process and not place patients who are marginal or poor candidates for advanced heart failure therapies with prolonged CPR time on ECMO. This is a difficult decision and urgent consultation from an advanced heart failure specialist and/or transplant/VAD surgeon is recommended to avoid the situation of a patient who is kept alive on ECMO but has an unrecoverable brain/heart and no other options but withdrawal of care. The routine use of MCD for CPR, Impella, and post-arrest targeted temperature management needs further investigation [28]. The impact of a formal institutional protocol for management of in-lab cardiac arrest, incorporating multidisciplinary care and systematic use of ECMO in selected patients, on outcomes needs to be prospectively studied on a larger scale.

Conclusion

While ILCA remains a relatively uncommon event, it is challenging to predict and manage. The key objectives are to support the patient till ROSC and prevent irreversible neurological damage. As the patient population presenting to the CCL becomes older and sicker, and with a greater complexity of coronary and structural interventions, we can expect the incidence of ILCA to increase. There have been enormous advances in our understanding of the pathophysiology of cardiac arrest, cardiogenic shock and mechanical circulatory support especially ECMO with a large volume of reported data. The application of this data with a systematic multidisciplinary team-based approach is essential to achieve the best outcomes for patients with cardiopulmonary arrest in the cardiac catheterization laboratory.