Downloaded from http://heart.bmj.com/ on January 16, 2016 - Published by group.bmj.com Education in Heart

INVASIVE IMAGING: AND Right heart catheterisation: indications and interpretation Paul Callan, Andrew L Clark

Department of Cardiology, INTRODUCTION Hull and East Yorkshire Significant improvements in the diagnostic power Learning objectives Hospitals NHS Trust, Hull, UK and availability of non-invasive cardiac imaging techniques, in addition to evidence of potential Correspondence to ▸ To gain an overview of the history and Dr Paul Callan, Department of harm associated with pulmonary artery (PA) cath- development of right heart catheterisation. Cardiology, Hull and East eterisation in patients in critical care, have led to a ▸ To learn how to perform a right heart study Yorkshire Hospitals NHS Trust, decline in right heart catheterisation (RHC) over Hull, HU16 5JQ UK; tailored to answer a specific clinical question. recent years.1 RHC, however, remains an import- [email protected] ▸ To gain a better understanding of the role of ant tool in a cardiologist’s diagnostic armoury, pro- right heart catheterisation as a diagnostic tool viding direct haemodynamic data that can be used in specific conditions, including pulmonary to determine cardiac output (CO), evaluate intra- hypertension, valvular heart disease and cardiac shunts and valve dysfunction. It is the gold differentiating between constrictive pericarditis standard method for diagnosing pulmonary hyper- and restrictive cardiomyopathy. tension (PH) and an essential component in the evaluation of patients prior to heart and/or lung transplantation.23RHC can be also used to assess pulmonary capillary wedge pressure (PCWP) under the haemodynamic effects of treatments directly fluoroscopic guidance, which added to the value of and provides an entry route for intracardiac biopsy. the test.8 The European Society of Cardiology (ESC) core Use of a self-guiding was first described curriculum 2013 states that trainees should possess in 1953 by Lategola and Rahn.9 Further modifica- the skills to ‘carry out right heart catheterization in tions ultimately led to the development of a the catheterization laboratory and at the bedside, balloon flotation catheter developed by Jeremy and measure cardiac output, intravascular pressure, Swan and William Ganz.9 The ‘Swan–Ganz’ cath- and oxygen saturation’.4 eter had an average time to catheterisation of 35 s, This article covers the history of RHC, how to and could be left in situ for prolonged periods, perform a complete right heart study and a review which led to a surge in its use in critical care. At of its current place as a diagnostic tool in a range the peak of their popularity in the mid-1980s, of cardiovascular disorders. Swan–Ganz were used in 45% of American patients admitted to the critical care BACKGROUND unit.10 Although balloon flotation catheters are The first reported RHC was performed on a horse more reliable than end-hole catheters in making it by the physiologist Claude Bernard in 1844.5 Glass possible to measure a wedge pressure, the balloon tubes were inserted via the jugular and carotid poses some risk of damage to the pulmonary vascu- artery in order to measure the temperature in both lature, particularly if it is left inflated inadvertently. ventricles. Bernard et al subsequently used the tech- Concerns over the safety of seemingly indiscrimin- nique to measure intracardiac pressures. The first ate use of invasive monitoring led to a number of RHC in a human was performed by Werner large studies that showed PA catheterisation-guided Forssmann in 1929 in Eberswalde, Germany. Using management was associated with increased mortal- an approach that is typically frowned on by ity and length of intensive care unit stay and has research ethics committees, he performed self- appropriately led to a substantial decline in the use catheterisation using a urethral catheter through his of PA catheters.1 left antecubital vein into his right ventricle (RV), and confirmed placement with X-ray.6 The tech- TECHNIQUE nique was subsequently refined by Andre Cournand Patient preparation and Dickinson Richards, who were able to demon- Patients must be fully informed of the indication strate its safety, even when the catheter was left in for, and risks of, the procedure. A large review of situ for over 24 h.7 They developed the direct Fick complications during RHC, incorporating both method for measuring CO and gained a greater retrospective and prospective data, found a serious understanding of right heart haemodynamics in adverse event rate of 1.1%, and mortality rate of both healthy subjects and those with pulmonary 0.05%.11 The most common complications are disease. Drs Forssman, Cournand and Dickinson access site haematoma, vagal reaction, pneumo- received the Nobel Prize in medicine in recognition thorax and arrhythmias. Where available, individ- To cite: Callan P, Clark AL. of their contributions to medical science. Hellems ual or departmental complication rates should be – Heart 2016;102:147 157. et al subsequently developed a way of measuring quoted.

Callan P, Clark AL. Heart 2016;102:147–157. doi:10.1136/heartjnl-2015-307786 147 Downloaded from http://heart.bmj.com/ on January 16, 2016 - Published by group.bmj.com Education in Heart

Catheterisation is frequently performed without the patient’s mid-chest. Lines and manifolds should interruption of anticoagulation. This is safe in be flushed to remove bubbles that would otherwise patients with an international normalised ratio of result in pressure damping. Each transducer should <3.5 undergoing RHC via either the internal be zeroed during set up, with zeroing repeated jugular vein (IJV) or antecubital .12 prior to recording pressures. The patient should Fasting prior to a procedure will depend on local have ECG monitoring throughout the procedure, policy, although it should be borne in mind that ensuring the pressure and ECG traces are timed overzealous fasting protocols may lead to volume correctly. The scale and sweep speed should be depletion, thus making more adjusted for each waveform measurement to facili- challenging. tate subsequent analysis. Recordings are best taken at end-expiration for consistent and stable traces. Venous access Box 1 gives a structured protocol for obtaining The route of access depends on a number of RHC measurements that is used at our institution. factors, including operator experience, the presence Clearly, the choice of measurements taken will of cardiac devices and in-dwelling catheters, and depend on the clinical question that one wishes to prior history of venous cannulation and associated answer, but a plan should be made in advance in complications. Femoral vein (FV) access is com- order to reduce the risk of a missed trace or satur- monly used if left heart catheterisation is per- ation measurement. formed concurrently, although a number of small studies have demonstrated the feasibility and safety Oxygen saturation measurements of performing RHC and left heart catheterisation Direct sampling of blood from intracardiac cham- via an antecubital fossa vein and radial artery, bers and great vessels enables the detection and – respectively.13 16 quantification of shunts between the systemic and A meta-analysis of ultrasound-guided versus pulmonary circulations. In RHC via the superior landmark-based venous access demonstrates a clear vena cava (SVC), samples are taken from the SVC benefit of ultrasound for IJV cannulation, with a and main PA (MPA). A difference of >8% may higher success rate, fewer complications and faster indicate the presence of a left to right shunt and access.17 There are very limited data on the use of should prompt a full saturation run for more ultrasound for FV and subclavian vein (SCV) precise localisation. Samples are then taken from cannulation. high and low inferior vena cava (IVC), high and Strict aseptic measures should be applied. The low SVC, high, mid-RA and low RA, mid-RV, RV procedure is performed using local anaesthetic; outflow, MPA, left and right PAs, left ventricle (LV) sedation is rarely required. A Seldinger technique and distal aorta. Additional samples are taken from should be used to gain venous access. the pulmonary vein and LA when the interatrial Balloon flotation catheters, such as the Swan– septum is crossed. A disadvantage of this method is Ganz catheter, have a balloon at the distal end to that it lacks sensitivity to detect small shunts, but facilitate passage through the right heart. They are the majority of haemodynamically significant designed to be placed without the need for fluoros- shunts will be identified.18 It also assumes steady copy, although screening is often helpful if the state blood flow throughout the entire sampling patient has marked right heart dilation or severe run, which may not always be the case, particularly tricuspid regurgitation. The catheter is inserted into if arrhythmias occur. Table 1 provides calculations the right atrium (RA) (15–20 cm via IJV, 10–15 cm for the determination of shunt size and figure 1 via SCV, 25–30 cm via FV) and the balloon is provides an example saturation run from a patient inflated. The catheter then follows the direction of with a large secundum atrial septal defect. blood flow towards the PAs. Advancing further should allow the performer to obtain the PCWP.It CO measurement is important to avoid leaving the balloon inflated in The two most commonly used techniques for the wedge position for longer than necessary, as determining CO are the thermodilution and Fick there is a risk of pulmonary infarction or rupture. methods. Thermodilution measurements involve Catheterisation from the FV is commonly per- injecting a fluid bolus at a known temperature into formed using a multipurpose end-hole catheter the proximal of a PA catheter, and recording under direct fluoroscopy. It requires greater the change in temperature at the distal end of the manipulation than the balloon flotation catheters to catheter with a thermistor. CO is calculated based navigate through the right heart, and a guidewire on the temperature and specific gravity of blood, may be required to improve steerability. and the temperature, specific gravity and volume of Multipurpose catheters can be used to cross directly injected fluid. into the left atrium (LA) in patients with a patent The Fick method for determining CO requires the foramen ovale for direct pressure recordings. following measurements: ▸ Oxygen consumption, or VO2 Pressure tracings ▸ The oxygen content of arterial blood It is vital to make sure that pressure lines and trans- ▸ The oxygen content of mixed venous blood ducers are correctly set up, as any inaccuracies in The Fick method requires measurement of measurement will be magnified in subsequent cal- oxygen consumption during cardiac catheterisation culations. Transducers are positioned at the level of using a spirometer and rebreathing bag, which is

148 Callan P, Clark AL. Heart 2016;102:147–157. doi:10.1136/heartjnl-2015-307786 Downloaded from http://heart.bmj.com/ on January 16, 2016 - Published by group.bmj.com Education in Heart

CO monitoring using Swan–Ganz catheters was Box 1 Right and left heart catheterisation pressure recording protocol once commonplace in critical care patients, but as described earlier, randomised controlled trials sub- sequently demonstrated that this was associated ▸ PA pressure with harm. This appears to be due to a combin- ▸ PCWP pressure ation of inaccuracies in measurement, incorrect ▸ Simultaneous PCWP and LV pressures interpretation of the data and responses to mea- ▸ Withdrawal from PCWP to PA surements that may have been harmful. PA catheter- ▸ Simultaneous PA and LV pressures isation remains useful in specific situations such as ▸ Withdrawal from PA to RV acute RV infarction and in patients with PH under- ▸ Simultaneous RV and LV pressures going surgical intervention.20 ▸ Withdrawal from RV to RA ▸ Simultaneous RA and LV pressures Pulmonary angiography LV, left ventricle; PA, pulmonary artery; PCWP, pulmonary capillary wedge Pulmonary angiography can be used to diagnose pressure; RA, right atrium; RV, right ventricle. pulmonary embolism (PE), PA stenosis and arterio- venous malformations. An angled pigtail catheter time-consuming and often impractical. Oxygen with side holes is commonly used to deliver con- consumption is thus often assumed based on age, trast, in order to reduce the risk of damage to the sex and body surface area. Mixed venous saturation pulmonary vasculature. The tip of the catheter is is measured from the MPA, where there is com- positioned in the proximal left and right PAs, and plete mixing of blood returning from the SVC, images should be acquired during a breath-hold at – IVC and coronary sinus. Arterial oxygen saturation end inspiration. Contrast is injected at a rate of 15 can be taken during simultaneous left heart cath- 20 mL/s for 2 s. Images should be obtained in at eterisation, or via pulse oximetry, or peripheral least two projections, typically anteroposterior, and fi arterial sampling. A recent haemoglobin level is posterior-oblique views ( gure 2). Diagnosis of PE requires the demonstration of also required for determination of blood oxygen 21 content. The formula for calculating CO using the thrombus in two radiographic projections. CT Fick method is shown in table 1. pulmonary angiography is now the preferred imaging modality for diagnosing PE, and may be Studies have demonstrated a reasonable correl- 22 ation between thermodilution and Fick methods better at identifying sub-segmental emboli. for the estimation of CO, although there may be Invasive pulmonary angiography enables the simul- significant variation in individual patients.19 taneous collection of haemodynamic data which Thermodilution tends to overestimate in low CO can help in determining the severity of PE. RHC states, and it is also inaccurate in patients with sig- can also be used for the treatment of life- nificant tricuspid regurgitation. threatening emboli, particularly when there is an absolute contraindication to thrombolysis.23 A number of techniques can be used, including mechanical clot disruption and aspiration, which Table 1 Commonly used haemodynamic calculations during right heart frequently result in acute haemodynamic improve- catheterisation ment and early recovery of RV function.24 25 Cardiac output (Fick) (L/min) O2 consumption (mL=min) Pulmonary angiography is considered the gold : (SaO2 MVO2saturation) 1 36 Hb 10 standard for diagnosing chronic thromboembolic Cardiac index (L/min/m2) Cardiac Output pulmonary hypertension (CTEPH), and assessing BSA the feasibility of pulmonary .26 This Cardiac Output Stroke volume (mL/beat) 1000 should be carried out in specialist centres by opera- HR Cardiac Index tors experienced in the management of CTEPH. Stroke volume index (mL/beat/m2) 1000 HR Delayed image acquisition once contrast has Mean PA pressure Mean PCWP passed into the LA (termed the levo-phase) can be Pulmonary vascular resistance (Wood Units) Cardiac output used to assess pulmonary venous drainage and may Mean arterial pressure Mean RA pressure Systemic vascular resistance (Wood Units) be helpful in detecting LA thrombus and Cardiac output myxoma.27 28 (To convert from Wood Units to dynes/s/cm multiply by 80) RV stroke work index (g/m2/beat) SVI×(Mean PA pressure−Mean RA pressure)×0.0136 PULMONARY ARTERIAL HYPERTENSION Intracardiac shunt calculations RHC is the gold standard for measuring PA pres- = O2 consumption (mL min) sure. The 2009 ESC guidelines on diagnosis and Pulmonary blood flow PVO2 PAO2 treatment of PH state that ‘RHC is required for the O consumption(mL=min) Systemic blood flow 2 diagnosis of pulmonary arterial hypertension SaO2 MVO2 Pulmonary blood flow (PAH), to assess the severity of haemodynamic Qp/Qs Systemic blood flow impairment, and to test the vasoreactivity of the 2 3 SVC O2 þ IVC O2 pulmonary circulation’. PAH is defined as a mean In the presence of a left to right shunt, MVO should be calculated as 2 4 PA pressure of ≥25 mm Hg at rest or ≥30 mm Hg

BSA, body surface area; HR, heart rate; IVC, inferior vena cava; MVO2, mixed venous oxygen saturation; during exercise. Measurement of PCWP allows dif- PA, pulmonary artery; PCWP, pulmonary capillary wedge pressure; RA, right atrium; SaO2, systemic arterial ferentiation between PH due to left-sided heart oxygen saturation; SVC, superior vena cava; SVI, stroke volume index. disease and pre-capillary PAH—conditions that

Callan P, Clark AL. Heart 2016;102:147–157. doi:10.1136/heartjnl-2015-307786 149 Downloaded from http://heart.bmj.com/ on January 16, 2016 - Published by group.bmj.com Education in Heart

A 400 mL fluid challenge will help discriminate between the two conditions.31 Acute vasodilator testing should be performed to identify a subset of patients with type 1 PAH that are likely to respond to treatment with long-term calcium channel blockers.32 Short-acting therapies such as inhaled nitric oxide and intravenous adeno- sine should be used for testing. A positive response is defined as reduction in mean PA pressure of >10 mm Hg to reach a mean of <40 mm Hg with an increased or unchanged CO. This is seen in approximately 10% of patients with PAH, who have a more favourable prognosis. Vasodilator testing should only be performed in centres experi- enced in the investigation and management of PH. Figure 1 Oxygen saturation recordings from a patient with a large secundum atrial The haemodynamic data also provide important fi ‘ ’ septal defect. This demonstrates a signi cant step up in saturations at the mid-atrial prognostic information. Data from international level, indicating the presence of a left to right shunt. registries consistently show that higher mean RA pressure, low cardiac index and higher mean PA require markedly different therapeutic approaches pressure are independently associated with worse – (figure 3). A PCWP of >15 mm Hg excludes the outcomes.33 35 The REVEAL registry of 2716 diagnosis of pre-capillary PAH. patients from the USA found that a pulmonary vas- The transpulmonary pressure gradient (TPG) is cular resistance (PVR) >32 Wood units and mean defined as the difference between the mean PA RA pressure >20 mm Hg were associated with pressure and PCWP. In a patient with PH and reduced survival.36 These are incorporated into a PCWP >15 mm Hg, a TPG >12 mm Hg suggests prospectively validated risk scoring system for ‘out of proportion’ PH; in other words, the raised newly diagnosed PH.37 Such elevated PVR mea- pulmonary arterial pressure cannot be fully surements were seen in a small number of patients, explained by an increase in LA pressure and is in predominantly with PH secondary to congenital part due to adaptive changes within the pulmonary heart disease (R Benza. personal communication). vasculature.29 The diastolic pulmonary gradient (ie, the gradient between pulmonary arterial diastolic VALVULAR HEART DISEASE pressure and mean PCWP, DPG) is less dependent Advances in echocardiography have resulted in a on changes in CO and PA vessel distensibility than marked reduction in the use of RHC for the assess- TPG, and might thus be a better measure in diag- ment of valvular heart disease. RHC was previously nosing ‘out of proportion’ PH. Both a TPG a ubiquitous part of the investigative work-up prior >12 mm Hg and DPG >7 mm Hg are associated to surgery, but current ESC guidelines recommend with lower median survival in patients with PH due that RHC should be reserved for cases in which to left-sided heart disease.30 ‘non-invasive evaluation is inconclusive, or discord- Wedge pressure is dependent on volume loading; ant with clinical findings’.38 hypovolaemia may cause an inappropriately low PCWP, leading to a false diagnosis of PAH. Mitral valve disease The severity of mitral stenosis can be assessed by RHC through measurement of the diastolic pressure gradient across the mitral valve or calculation of valve area. Measurement of mitral valve area by planimetry on echocardiogram is often limited by image quality, and valve area estimation by pressure half time may not be valid in calcific mitral stenosis or in the presence of abnormal LV relaxation.39 RHC and left heart catheterisations allow simultan- eous recording of PCWP (an indirect measure of LA pressure) and LV pressure. The Gorlin formula can then be used to calculate valve area, as shown below.

Mitral valve area ðcm2Þ = ¼ Cardiac output (L min) 1000 : 44 3 heart ratepffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffidiastolic filling period mean gradient:

Figure 2 Pulmonary angiography demonstrating multiple filling defects within the The diastolic filling period is measured from the right pulmonary artery and lobar arteries (starred), due to acute pulmonary embolism. onset of diastole, at the point when the PCWP

150 Callan P, Clark AL. Heart 2016;102:147–157. doi:10.1136/heartjnl-2015-307786 Downloaded from http://heart.bmj.com/ on January 16, 2016 - Published by group.bmj.com Education in Heart

Figure 3 Pressure traces taken from a patient with pulmonary arterial hypertension. Trace A shows very high pulmonary artery pressure (mean 75 mm Hg). Trace B shows the normal pulmonary capillary wedge pressure, confirming that the elevated pulmonary pressures are due to precapillary pulmonary vascular disease. Note that the wedge pressure is under-damped with marked artefact.

exceeds LV pressure, to the end of diastole, when Mitral regurgitation (MR) can be assessed during LV pressure exceeds PCWP, measured in seconds/ a left ventriculogram, with semiquantitative scoring beat. tools used to determine severity. Care should be The formula incorporates a constant to account taken to avoid getting the catheter caught in the for the loss of velocity due to friction, which was mitral valve apparatus, which may accentuate the calculated based on the difference between the esti- degree of MR. Large v-waves on the PCWP trace mated and observed valve areas in the original are most commonly due to MR (figure 4), although study by Gorlin and Gorlin.40 Catheter-based mea- the differential diagnoses includes atrial septal surements can lead to an overestimation of valve defects and LV volume overload. A v-wave greater area by up to 50%, due to a delay in the transmis- than twice the mean PCWP suggests severe MR, sion of LA pressure through the pulmonary vascu- although this is neither a particularly sensitive or lature.41 Furthermore, accurate wedge tracings may specific finding.43 be difficult to obtain in patients with significant PH or right heart dilation. Direct measurement of LA pressure via trans-septal puncture obviates these Pulmonary and tricuspid stenosis problems, but greatly adds to the procedural risk. Valve area calculation using the Gorlin formula has PH significantly increases perioperative morbid- not been validated for the assessment of tricuspid ity and mortality in patients undergoing mitral or pulmonary stenosis. Tricuspid stenosis can be valve surgery.42 Right heart catheter measurements assessed using simultaneous RA and RV pressure may therefore enable a more individualised assess- recordings (figure 5). Pulmonary stenosis is typic- ment of operative risk, thus helping informed dis- ally measured by catheter pull-back from the MPA cussion between patient and clinician. into the RV, although simultaneous recordings can

Figure 4 Simultaneous left ventricular (LV) and pulmonary capillary wedge (PCW) traces in a patient with atrial fibrillation and significant mitral regurgitation are shown. Large v-waves are