Patients and methods

Patient population

This study was approved by the Mayo Institutional Review Board. All patients provided informed consent. The study population consisted of patients who were referred to the Mayo Cardiac Catheterisation Laboratory for comprehensive haemodynamic evaluation between September 2004 and May 2007 for the investigation of heart failure, and subsequently found to have severe tricuspid regurgitation (group I). Among these patients, 14 fulfilled criteria for inclusion in the study: (a) severe tricuspid regurgitation during right ventriculography or echocardiography; (b) surgical confirmation of the absence of constrictive pericarditis during subsequent surgery for the repair or replacement of the tricuspid valve; (c) absence of severe pulmonary hypertension as an explanation for tricuspid regurgitation (mean pulmonary arterial pressure <40 mm Hg). The aetiology of severe tricuspid regurgitation was tethering of the valve leaflets by a pacemaker or defibrillator lead in eight patients, damage to the valve from a prior ventricular septal defect correction surgery in one patient, a flail leaflet in one patient and idiopathic tricuspid annular dilatation in the remainder.4,5

A cohort of 14 patients with constrictive pericarditis was used for comparison (group II). In each case of constrictive pericarditis, surgical confirmation of constrictive pericarditis was obtained by the presence of an obliterated pericardial space, an adhesive pericarditis with bulging of the heart out of the pericardial incision at pericardiectomy and pathological confirmation.2 Each of the patients with constrictive pericarditis had trivial or only mild tricuspid regurgitation by two-dimensional colour flow Doppler echocardiography and intraoperative transoesophageal echocardiography.6

Cardiac catheterisation

All patients underwent a comprehensive haemodynamic study using standardised methodology.2,3 Simultaneous right and left heart catheterisation using high-fidelity micromanometer catheters (Millar Instruments, Austin, Texas, USA) was performed during shallow breathing and deep inspiration. Digital pressure records at 4 ms intervals from each cardiac chamber were taken with shallow breathing, then with deep respiration. Three to five cardiac cycles were analysed and averaged during offline analysis. For patients with relatively low right atrial pressure (<15 mm Hg), 1 litre of intravenous normal saline bolus was administered before haemodynamic assessment.

Pressure analysis

Analysis of pressure waveforms was performed by investigators blinded to the clinical diagnosis. The following diastolic pressure variables were analysed at end inspiration and end expiration: end-diastolic pressure (EDPs), height of the early rapid filling wave and diastasis pressure (fig 1). End-diastole was defined at the peak of the first QRS deflection on the electrocardiogram. The interventricular gradient at end diastole (LVEDP−RVEDP) was termed ΔP. End equalisation was defined as ΔP <5 mm Hg. The respiratory variation in ΔP was calculated as the absolute value of inspiratory ΔP minus the absolute value of expiratory ΔP. Early rapid filling wave (RFW) was defined as the first positive deflection after the minimal ventricular diastolic pressure point (fig 1). The height of the RFW in the left and right ventricles was defined as the pressure change from minimal pressure to the plateau phase in diastasis (y in fig 1). The slope of this RFW was calculated (fig 1). The peak inspiratory beat was defined as that which was preceded by the lowest early diastolic nadir of the LV pressure.3 Selection of the peak inspiratory beat required that the early diastolic nadir was at a minimum for the diastolic filling period before and after the systolic pressure contours. The peak expiratory beat was selected as the systolic impulse that was preceded by the highest early diastolic nadir of the LV pressure. The RV index and systolic area index, two indices previously described to assess ventricular interaction from the RV and LV systolic pressures during respiration were calculated.2,3 The RV index was the ratio of the inspiratory to expiratory RV maximal systolic pressure, divided by the ratio of the inspiratory to expiratory LV maximal systolic pressure.2 The systolic area index was the ratio of the RV to LV systolic pressure–time area during inspiration divided by that ratio during expiration.3 To assess the dissociation of intrathoracic and intracardiac pressures, the difference between the pulmonary capillary wedge pressure (PCWP) and the minimum early LV diastolic pressure during expiration was subtracted from that same difference in inspiration. Continuous variables were expressed as mean (standard deviation). Comparisons were made using the Student t test.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Typical ventricular pressure tracing showing a rapid early diastolic filling wave (RFW, block arrow) after the minimal pressure point. The height of the RFW (y) referred to in the text is the pressure difference between the plateau after the rapid early diastolic filling and the minimal ventricular pressure. EDP, end diastolic pressure.

Results

Conventional haemodynamic criteria

At the time of the study, groups I and II had similar heart rates, systolic blood pressures, mean right atrial pressures and pulmonary artery systolic, diastolic and capillary wedge pressures (table 1). During expiration and shallow breathing, both groups I and II had elevation and equalisation of LVEDP and RVEDP (table 1, fig 2), with a prominent RFW on both RV and LV pressure tracings. Eleven (79%) group I and 13 (93%) group II patients had an RVEDP to RV systolic pressure ratio of more than 1:3. All but one of the group I patients had equalisation of LV and RV end-diastolic pressures on expiration. The right atrial pressure waveform differed between the two groups. Overall, the height of the v wave in group I was greater than that seen in group II (table 1). Both groups had a deep, prominent y descent, but the x descent was blunted in group I in comparison with group II.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Examples of simultaneous left ventricular (LV) and right ventricular (RV) pressure tracings in a patient with tricuspid regurgitation (left) and a patient with constriction (right) showing prominent early rapid filling waves and elevation and equalisation of LV and RV end-diastolic pressures in both cases.

Dynamic respiratory changes

Subjects in group I showed many of the characteristic dynamic respiratory changes seen in constrictive pericarditis, although the magnitude of theses changes tended to be less marked than in group II. Measurements of both respiratory enhancement of interventricular dependence and intracardiac–intrathoracic dissociation of pressures for group I and group II are shown in fig 3 in group I albeit to a lesser extent than that seen in group II. The group means for both systolic area index and RV index were greater in group II patients than in group I patients, but there was considerable overlap of these two indices between the two groups. The systolic area index was >1.0 in all group II patients, but also was >1.0 in five of 14 group I patients. The absence of ventricular interdependence (area index <1.0) was thus 100% specific for diagnosing tricuspid regurgitation versus constriction, but only 64% sensitive. The PCWP-LV diastolic pressure gradient significantly decreased during inspiration in all group II patients, but also decreased >5 mm Hg in 4/14 group I patients (fig 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Scatter plots of individual indices in patients with severe tricuspid regurgitation compared with patients with constriction. (A) The systolic area index (the ratio of the right ventricular (RV) to left ventricular (LV) systolic pressure–time area during inspiration divided by that ratio during expiration was >1 (meaning an enhancement of ventricular interdependence) in all patients with constrictive pericarditis and in five patients with tricuspid regurgitation.3 (B) RV index (the ratio of the inspiratory to expiratory RV maximal systolic pressure, divided by the ratio of the inspiratory to expiratory LV maximal systolic pressure showing considerable overlap between the two groups.2 (C) Dissociation of intrathoracic–intracardiac pressures (expiratory minus inspiratory gradient between pulmonary capillary wedge pressure and minimal LV diastolic pressure) was prominent in constrictive pericarditis but also present in many patients with tricuspid regurgitation.

Left ventricular and right ventricular diastolic pressures

During deep breathing, the changes in diastolic LV and RV pressures behaved very differently in the two groups (table 1, fig 4). In group I, ΔP became wider and more negative during inspiration, with RVEDP exceeding LVEDP in the majority of patients (fig 5A). In group II, ΔP increased more with expiration, and throughout the respiratory cycle tended to remain positive (fig 5B). The respiratory variation in ΔP (fig 5C) was positive (+2.6 (3.7)) in group I and negative (−4.8 (3.1)) in group II. A respiratory variation in ΔP >0 was 93% specific and 78% sensitive for diagnosing tricuspid regurgitation versus constrictive pericarditis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

(Lower panel) Simultaneous left ventricular (LV) and right ventricular (RV) tracings in a patient with constriction. Note that during both inspiration and expiration, there is elevation and equalisation of diastolic pressures (arrowhead). The LV diastolic pressure does not drop significantly with inspiration, and RV rapid filling is not accentuated on inspiration. This is in contrast to a patient with severe tricuspid regurgitation (upper panel), where on deep inspiration, the diastolic pressures separate with a higher RV diastolic pressure (arrow), and the rapid RV filling wave (arrowhead) becomes deeper and steeper.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

(A) Scatter plot of the difference between left ventricular end-diastolic pressure (LVEDP) and right ventricular (RV) EDP (ΔP) during deep inspiration, being on average lower in patients with tricuspid regurgitation (LVEDP lower than RVEDP). (B) ΔP at end expiration, showing an almost equalisation of LVEDP and RVEDP in most patients with tricuspid regurgitation, with overlap with the constrictive pericarditis group. (C) The change in the absolute difference between LVEDP and RVEDP from expiration to inspiration (respiratory variation in ΔP). Positive numbers mean that the absolute value of ΔP widens on deep inspiration (seen mostly in group I), and negative numbers means that ΔP shrinks on inspiration (seen mostly in group II).

Early rapid right ventricular filling

The height of the RFW in RV diastolic pressure was accentuated with inspiration in group I but did not change for group II (fig 6A). The slope of the RV RFW increased significantly on inspiration in group I, but decreased or showed little change in all but one of the group II patients (fig 6B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

(Upper panel) Change from expiration to inspiration in the right ventricular height of the diastolic rapid filling wave (RFW). As opposed to the constrictive pericarditis group, the RFW is significantly accentuated on inspiration in the tricuspid regurgitation group. (Lower panel) Inspiratory-to-expiratory ratio in the slope of the RFW. The slope is significantly accentuated on inspiration in the tricuspid regurgitation group, but not in most of the constrictive pericarditis group.

Discussion

The principal findings of this study are (a) many traditional haemodynamic findings in patients with severe tricuspid regurgitation can mimic those seen in constrictive pericarditis; (b) respiratory enhancement of interventricular dependence is observed in severe tricuspid regurgitation, similar to that seen with constrictive physiology; (c) tricuspid regurgitation can be differentiated from constrictive pericarditis by examining the relationship between LV and RV diastolic pressures during deep respiration. On inspiration, the difference between the RV and LV end-diastolic pressures becomes more prominent in tricuspid regurgitation, and less prominent in constrictive pericarditis. RV diastolic pressure exceeds LV diastolic pressure during inspiration, and the height and slope of the RV RFW are also more accentuated in tricuspid regurgitation, but not constrictive pericarditis. The similarities and novel differences reported here are critically important when considering the differential diagnosis and evaluation of patients presenting with chronic right-sided heart failure with indeterminate aetiology.

Similarities between tricuspid regurgitation and constrictive pericarditis

Traditional diagnostic criteria for constrictive pericarditis such as increase and equalisation of end-diastolic pressures in all four cardiac chambers, a dip and plateau pattern in the ventricular pressure curves, rapid x and y descents in the atrial pressure curves, pulmonary artery systolic pressure <55 mm Hg and RV end-diastolic pressure >1/3 RV systolic pressure are not specific to constrictive pericarditis.2,7,8 These findings are also present in patients with restrictive cardiomyopathy and as shown herein, they can also be seen in severe tricuspid regurgitation. Signs of enhancement of systolic ventricular interdependence are more specifically seen in constrictive pericarditis and help to differentiate it from restrictive cardiomyopathy. However, this finding of enhanced ventricular interaction as assessed by the relationship of LV and RV pressures during respiration is also seen in patients with severe tricuspid regurgitation.2 In constrictive pericarditis, the fact that all cardiac chambers are confined within a constricting pericardium creates a haemodynamic interdependence between the two ventricles. As inspiration increases venous return to the right ventricle, the left ventricle is “compressed” with attendant reduction in diastolic filling, reduced venous return and lower inspiratory stroke volume, while the RV diastolic filling and stroke volume increase. The reverse occurs during expiration. This enhanced ventricular interdependence creates a LV pressure–time tracing with a lower area under the curve on inspiration versus expiration, with the opposite seen on the RV pressure tracing.2,3 The constrictive pericardium also “shields” transmission of the negative intrathoracic pressure seen on the PCWP tracing to the ventricles, resulting in respiratory fluctuations in PCWP tracing that are much more prominent than those of the LV diastolic pressures. In severe tricuspid regurgitation, a dilated and failing right ventricle can constrict the left ventricle and create a haemodynamic picture similar to that in constrictive pericarditis, with elevation and equalisation of diastolic pressures, systolic enhancement of ventricular interdependence and possibly, some intrathoracic–intracardiac dissociation of pressures, though these findings are less prominent than in constrictive pericarditis, and out of proportion to the degree of right heart failure with which patients with severe tricuspid regurgitation present.

Differentiating features of tricuspid regurgitation and constrictive pericarditis

The critical difference between constrictive pericarditis and tricuspid regurgitation centres upon changes in the diastolic flow into the right ventricle and left ventricle during inspiration. In patients with constrictive pericarditis, the flow is significantly limited by the rigid pericardium, resulting in very little increase in flow through the tricuspid valve during inspiration.9 Conversely, in severe tricuspid regurgitation, there is no limitation of blood flow into the right ventricle, resulting in significant enhancement of RV diastolic filling during inspiration with a marked increase in RV diastolic pressures and rapid flow waves.

The indices shown in fig 6(upper panel) and 6(lower panel) (change from expiration to inspiration in the height of the RV RFW, and the inspiratory-to-expiratory ratio of the slope of the RFW) simply mean that the accentuation in RV RFW on inspiration versus expiration is more likely to be seen in tricuspid regurgitation than in constriction. Figure 5C demonstrates that RV and LV diastolic pressure tracings that diverge on deep inspiration, with RV pressure > LV pressure are most consistent with severe tricuspid regurgitation. The RV and LV diastolic pressure tracings that come together during deep inspiration are more consistent with constriction. Systolic signs on the pressure tracings should also be used for a comprehensive assessment: a very prominent enhancement of systolic ventricular interdependence is more consistent with constrictive pericarditis as the primary pathology than with tricuspid regurgitation, since the magnitude of dissociation was much greater in constrictive pericarditis. The right atrial pressure waveform can also be helpful: the v wave is higher and the x descent is more blunted in patients with tricuspid regurgitation. Conversely, signs such as elevated and equal diastolic pressures, prominent early rapid filling waves and low pulmonary arterial pressures have little discriminatory value and should not be used to differentiate tricuspid regurgitation from constrictive pericarditis.

Clinical implications

There has been an increasing recognition of severe tricuspid regurgitation due to a variety of disorders.4,10,11,12 Such patients may present with symptoms and signs of right heart failure. The differentiation between severe tricuspid regurgitation and constrictive pericarditis as an aetiology for right heart failure in indeterminate cases is of critical importance, because misdiagnosis could lead to referral for an inappropriate corrective operation.13 Tricuspid regurgitation related to a pacemaker or defibrillator lead impinging on the tricuspid valve has been increasingly reported as the number of device implantations continues to grow in the heart failure population.4 These patients have frequently had prior cardiac instrumentation, constrictive pericarditis is often suspected when they present with symptoms related to increased right-sided pressures. Echocardiography, while helpful in most cases, can be limited, especially in definitively ruling out concomitant constrictive pericarditis.

Non-invasive imaging can identify tricuspid regurgitation in the majority of patients, particularly in those with good echocardiographic windows. However, recent data4 have demonstrated limited sensitivity of transthoracic echocardiography in detecting severe tricuspid regurgitation due to pacemaker or defibrillator leads. Given the current limitations of non-invasive techniques, this investigation emphasises the need for absolute pressure measurement with high-fidelity catheters as a means of differentiating these two disorders in cases where the cause of right heart failure is indeterminate.

Limitations

This study is limited by its retrospective nature, relatively small sample size and the potential bias in case selection. This study used high-fidelity micromanometer catheters, which are not widely available in all catheterisation laboratories, but qualitative differences in RV and LV diastolic pressure tracings can still be interpreted with fluid-filled systems, particularly during diastole when ringing artefacts are less problematic.

Conclusion

Invasive haemodynamic findings in patients with severe, symptomatic tricuspid regurgitation can mimic those of constrictive pericarditis. Opposing changes in right and left ventricular diastolic pressure tracings during respiration allow for reliable differentiation of the two clinical entities and provide additional diagnostic utility.