What should be called a brisk 'Y' descent?B Dalvi, P Kerkar, A Vora, J Singh
Dept of Cardiology, KEM Hospital, Parel, Bombay, Maharashtra.,
Patients with pericardial constriction show a prominent 'y' descent in right atrial and vena caval pressure traces. In all earlier hemodynamic descriptions of constrictive pericarditis, the 'y' descent has been described as 'brisk', 'sharp' or 'rapid' but no effort has been made to quantify the same. In this study, we have tried to objectively evaluate and describe this 'y' descent by measuring its negative slope (-dy/dt) at its steepest portion. Forty one patients were studied hemodynamically, 9 with constrictive pericarditis (Group I) and 32 normals (Group II). The negative slope of the 'y' descent in patients with constrictive pericarditis (69.95 +/- 23.04 mm Hg) was found to be significantly greater than normals (35.13 +/- 7.84 mm Hg, p < 10(-6). Discriminant analysis was used to determine its sensitivity, specificity, predictive value and overall accuracy, in the diagnosis of pericardial constriction. Value of > or = 45 mm Hg/sec was found to have the highest overall accuracy (0.88). The correlation between the right ventricular end diastolic pressure and the slope of 'y' descent in patients with pericardial constriction (r = 0.66) and in normals (r = 0.60) was fair. It is concluded that -dy/dt is significantly different in patients with constrictive pericarditis as compared to normals. The diagnostic utility of this parameter needs to be evaluated in patients with equivocal clinical and hemodynamic data, in those with occult pericardial constriction and in post-pericardiectomy cases where the pressures do not normalise immediately after adequate pericardial resection.
Keywords: Adult, Bias (Epidemiology), Blood Pressure, Diagnosis, Differential, Diastole, Discriminant Analysis, Heart Catheterization, Hemodynamics, Human, Pericarditis, Constrictive, diagnosis,epidemiology,physiopathology,Retrospective Studies, Sensitivity and Specificity,
Pericardial constriction is characterized by impairment of diastolic filling of all chambers of the heart secondary to the symmetrical constricting effects of the pericardium. Hemodynamics of constrictive pericarditis have been described since the early 1950s. Elevation and equalization of the diastolic pressures in all four cardiac chambers with a 'dip and plateau' in the right ventricular pressure curve are the main characteristics of pericardial constriction. However, early diastolic filling of the ventricles characterized by "brisk", "deep", "sharp", or "rapid" 'y' descent has been the clinical and hemodynamic hallmark of constrictive pericarditis,,,,,.
The aim of this study, was to quantity this subjective "brisk" or "rapid" 'y' descent and assess its utility in the diagnosis of constrictive pericarditis.
Patient Population: In a retrospective study of 41 patients, nine had clinical and hemodynamic evidence of pericardial constriction and 32 were normal. These 32 patients were referred for coronary angiography during the course of their investigations for chest pain and were found to have normal coronary arteries. All the patients had intact interatrial septum and normal tricuspid valve on clinical and non-invasive assessment.
Catheterization protocol: The patients were subjected to right heart catheterization after obtaining informed consent. An 8F endhole Cournard catheter was used for recording the pressures from the right atrium, right ventride, pulmonary artery and pulmonary artery wedge. The pressures were recorded on Rectigraph - 8K (SAWEI, Tokyo, Japan) using a fluid filled manometric system. Care was taken to adjust the zero point 5 cm posterior to the sternum and the instrument was calibrated at the start of every case.
Diagnostic criteria and measurements: The diagnosis of constrictive pericarditis was confirmed on the basis of right heart pressure recordings fulfilling the following criteria: evidence of elevation and virtual identity (within 5 mm Hg) of right atrial, right ventricular diastolic and pulmonary capillary wedge pressures before 'a' wave; characteristic wave form of pressure pulse in the right ventricle, Le early diastolic dip and plateau; and right atrial pressure waves characterised by preserved systolic Y descent, a prominent early diastolic 'y' descent and 'a' and 'v' waves which are small and equal resulting in typical "M" or "W'' pattern.
After obtaining an optimum pressure tracing from the right atrium, the negative slope of the 'y' descent was measured at its steepest portion as shown in [Figure:1]). Five beats were analysed and the value of -dy/dt was computed by averaging them to minimize the error.
Statistical analysis: All data is expressed as mean + standard deviation. The negative slope of 'y' descent in patients of constrictive pericarditis was compared with normals using Student's 't' test and -dy/dt was compared with the right ventricular pressure in normals and in those with pericardial constriction by linear regression.
Hemodynamic data in nine patients with surgically proven constrictive pericarditis was compared with 32 normals. The age, heart rate and body surface area in Group I and Group II patients were comparable (p - ins) as shown in [Table - 1].
Right ventricular end diastolic pressures and -dy/dt in patients with constrictive pericarditis were significantly more than normals (p< 10,-6) [Table - 2].
Although -dyldt of ? 35 mm Hg/sec was found to be most sensitive for differentiating the normals from patients of constrictive pericarditis, it had a low specificity (62.5) with an overall accuracy of 0.7. Similarly negative slope of 'y' of k 50 mm Hg/sec was found to be more specific but at the cost of sensitivity. Negative slope of the 'y' descent of greater than or equal to 45mm Hg/sec has been found to have the best overall accuracy of 0.88 ([Table - 3] on pg. 173).
Negative slope of the 'y' descent was found to correlate only fairly with the right ventricular end diastokc pressure both in patients with pericardial constriction (r = .66) and in normals (r = .60) (See [Figure:2] and [Figure:3]).
Clinically and hemodynamically, constrictive pericarditis is characterized by abnormally prominent 'y' descent in both the venae cavae and the right atrium. This sharp 'y' descent corresponds to the early diastolic dip on the ventricular pressure curve and pericardial.
Previous hemodynamic studies have referred to the y' descent in right atrial trace in patients of constrictive pericarditis as being 'brisk', 'rapid' and 'sharp'. This appears to be a purely subjective description and thus assigning a label 'rapid' or 'brisk,' becomes difficult especially when the character of the 'y' descent is equivocal.
Determinants of -dyldt: Factor, which governs the slope of the 'y' descent in the right atrium, is the instantaneous pressure gradient between right atrium and right ventricle during the early diastole. This in turn will be determined by the state of tricuspid valve, intactness of the interatrial septum, volume of diastolic blood flow across the tricuspid valve, duration of the diastole, rate of right ventricular relaxation and the state of pericardium. In the absence of any abnormality of the tricuspid valve and the interatrial septum, as in our patients, the rate of diastolic flow and right ventricular diastolic properties are the predictors of -dy/dt. Although the effects of heart rate on the slope of 'y' descent were not analysed in the present study, Brecheril has convincingly shown that the rate of diastolic flow in the venae cavae is not significantly affected by heart rate. Thus in the presence of normal ventricular myocardium, state of pericardium becomes the sole determinant of -dy/dt. The present study, however, showed only a fair correlation between the right ventricular end diastolic pressure and -dy/dt (r = 0.6). This may be due to lack of correction of -dy/dt for heart rate and diastolic blood flow across the tricuspid valve. In our patients, the slope of 1 y' descent of greater than or equal to 45mm Hg/sec has best overall accuracy in the diagnosis of constrictive, pericarditis [Table - 3], It is probable that the sensitivity, specificity, predictive value and overall accuracy could be improved by defining and applying the correction factors for heart rate and diastolic flow volumes.
Potential implications of -dyldt: Although it may not have a significant role to play in cases of frank pericardial constriction, -dy/dt should find a place in establishing the diagnosis of occult or subclinical pericardial constriction. Similarly, in patients of constrictive pericarditis in whom pressures in the cardiac chambers and the venae cavae do not normalise immediately after pericardiectomy, dy/dt may be helpful in differentiating residual pericardial constriction or epicarditis from a subset with delayed or incomplete recovery without surgically correctible abnormality. Prospective studies will be necessary to define the utility of -dy/dt in prediction of future constriction in those with pericardial effusion. If the slope of 'y' descent can be correlated to the rate of early diastolic flow using non-invasive modalities like doppler, it may help in the follow up and early detection of pericardial constriction.
Potential Errors and Limitations of the Study: Although Millard catheters, would give more accurate readings, fluid manometry is most commonly employed for pressure recordings. These measurements are subject to distortions and inaccuracies inspite of every effort being made to design an ideal system with high sensitivity uniform frequency response and optimum damping. Since most of the cardiac catheterization laboratories all over the world make use of fluid filled systems for pressure recording, these results will be of practical value inspite of the problems inherent to such recording.
It is likely that heart rate and diastolic flow across the tricuspid valve may change the slope of 'y' descent. However to calculate the correction factors for these variables, may not amount to merely computing a regression of -dy/dt against heart rate or diastolic flow because rate of change of -dy/dt will also depend on the basal -dy/dt. This has probably resulted in only a fair correlation between the right ventricular end diastolic pressure and the slope of 'y' descent in the present study.
Authors wish to thank Dr. PM Pai and Dr PA Kale for their help and guidance, Mr Karan Chavan for technical assistance and Mrs. Supriya Deshpande for preparing the manuscript.
[Table - 1], [Table - 2], [Table - 3]