Objective: Doppler myocardial imaging is increasingly being used to evaluate regional and global cardiac function. Quantitative measurements of tissue deformation obtained during ejection as well as isovolumic contraction have been proposed as new indices of contractility; however, their load-sensitivity remains a matter of controversy. Maximum strain rate (SRmax) and isovolumic strain acceleration (ISAmax) were compared with regard to sensitivity for inotropic state, heart rate and loading conditions in the right ventricle (RV), using pressure-volume analysis as the reference method.
Design: Prospective animal study.
Setting: University hospital laboratory.
Interventions: RV contractility was measured at baseline, after inotropic modulation with esmolol and dobutamine, at different atrial pacing rates and during controlled alterations of RV preload and afterload.
Main outcome measures: RV contractility was assessed with the slope (Mw) of preload recruitable stroke work and longitudinal SRmax and ISAmax.
Results: SRmax and ISAmax reflected the drug-induced changes in contractility, while only ISAmax increased with higher pacing rates. Acute lowering of RV preload did not affect either of the indices studied. In contrast, an increase in RV afterload consistently decreased SRmax (from 1.05 (SD 0.41) to 0.73 (SD 0.26) s-1,p = 0.03) but had variable effects on ISAmax and Mw. However, a significant correlation was found between proportional changes in ISAmax and Mw during high-afterload conditions (r2 = 0.89, p = 0.005).
Conclusions: Both SRmax and ISAmax reflected changes in RV contractility. ISAmax was less sensitive to changes in RV afterload than SRmax and may therefore be a more robust index of global RV contractility.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Because of its complex geometry and load-sensitive performance, the clinical assessment of right ventricular (RV) function remains difficult.1 There is increasing interest in the quantification of regional tissue velocities and tissue deformation to assess RV function since this can be performed with non-invasive imaging techniques and obviates the need for volume determinations. Experimental studies show that maximum strain rate (SRmax), an ejection-phase index of regional myocardial tissue deformation, corresponds with global contractility.2 3 The value of SRmax to assess RV myocardial function will ultimately depend on its ability to distinguish loading-induced RV pump failure from RV contractile dysfunction; however, studies addressing its load-sensitivity are scarce and have produced controversial findings.4 5
Vogel et al reported that isovolumic indices of tissue velocity were less load-sensitive than ejection phase indices and showed that myocardial acceleration during isovolumic contraction (IVA) was a useful index of RV contractility.6 These findings were based on Doppler-derived measurements of tissue velocities, which not only reflect local muscular events but are also influenced by cardiac rotation and translation and, more importantly, by tethering from adjacent structures such as the left ventricle (LV). Tissue deformation analysis overcomes this limitation by calculating spatial velocity gradients within a small cardiac region. An experimental study showed that maximum isovolumic strain rate acceleration (ISAmax) was indeed more sensitive to inotropic changes than IVA, but the data also suggested a higher sensitivity to load alterations.7 Clearly, tissue deformation analysis shows promise for the study of RV function, but important information with regard to the effects of alterations in preload and afterload is still lacking.
The present study was therefore designed to evaluate the sensitivity of myocardial tissue deformation to changes in inotropic state, heart rate and loading conditions in the RV, using pressure-volume analysis as the reference method. We specifically tested the hypothesis that isovolumic deformation, quantified with ISAmax, is less sensitive to external cardiac loading conditions than the ejection phase index SRmax.
This investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996) and was approved by the ethics committee of the Katholieke Universiteit Leuven.
Six adult Lovenaar ewes (weight 53 (SD 5) kg) were included in this study. The animals were premedicated with ketamine hydrochloride 10 mg/kg. Anaesthesia was induced with intravenous sodium pentobarbital 8 mg/kg and piritramide 1 mg/kg and maintained with sodium pentobarbital 3 mg/kg/h and piritramide 1 mg/kg/h. Endotracheal intubation was performed and the lungs were mechanically ventilated with a mixture of oxygen and room air. Arterial blood gases were measured at regular intervals and the ventilation was adjusted to maintain normocapnia and normoxia. Lactated Ringer’s solution was administered at a rate of 5 ml/kg/h.
A lateral cutdown was performed in the cervical region. A triple-lumen catheter was inserted into the right jugular vein. A fluid-filled catheter was advanced into the proximal aorta via the right carotid artery for monitoring of systemic arterial pressure. Via a midline sternotomy, a tourniquet was placed around the inferior vena cava (IVC) for controlled alterations of preload. The pericardium was opened, and a 20 mm perivascular flow probe (Transonic Systems Inc., Ithaca, NY, USA) and a tourniquet were placed around the main pulmonary artery (PA). Pacemaker wires were sutured to the right atrium. A combined pressure-conductance catheter (Millar Instruments, Houston, TX, USA) was inserted into the RV through a small stab wound in the pulmonary outflow tract. One pair of sonomicrometry crystals (Sonometrics Corporation, London, Ontario, Canada) was sutured along the long axis of the RV free wall with the basal crystal placed at 10 mm distance from the atrioventricular groove with an approximate inter-crystal distance of 3 cm.
After achievement of haemodynamic steady state, baseline measurements were performed with the ventilation suspended at end-expiration. Data were acquired during steady state conditions (for general haemodynamics and sonomicrometry), and during a brief period of IVC occlusion (for the calculation of the slope (Mw) of the global preload recruitable stroke work (PRSW) relationship). Alterations of inotropic state, loading conditions and heart rate were performed in random order.
Esmolol was administered intravenously (loading dose of 1000 μg/kg followed by a continuous infusion titrated to achieve a decrease in heart rate of approximately 25%; mean dose 66 μg/kg/min). Measurements were performed at least 15 min after the start of the infusion, when the heart rate had stabilised. The esmolol infusion was then stopped. After the heart rate had returned to baseline values, dobutamine was started at a dose of 3 μg/kg/min. Measurements were performed at least 15 min after the start of the infusion, again after heart rate had stabilised.
Sensitivity to alterations in preload and afterload
For preload reduction, the IVC was partially occluded with a tourniquet until end-diastolic RV volumes as indicated on the oscilloscope decreased to about 50% of baseline values. This position was maintained while haemodynamics were allowed to stabilise for at least 5 minutes. After this period, new measurements were obtained and a brief additional IVC occlusion was performed to calculate global Mw. For afterload increase, the main PA was partially occluded with a tourniquet to obtain an increase of RV pressures by about 50% of baseline values. The development of acute tricuspid valve insufficiency, characterised by the disappearance of an isovolumic contraction phase in the instantaneous pressure-volume (PV) loops, was considered a cut-off sign: the banding was then slightly untightened until the PV loops resumed their expected shape on the oscillograph. The tourniquet was fixed in this position and haemodynamics were allowed to stabilise for at least 5 minutes. After this period, measurements were repeated and an IVC occlusion was again performed in order to calculate Mw. In three animals, the measurements in baseline and during afterload increase were repeated after pharmacological blockade of the autonomic nervous system (ANS) with atropine methyl nitrate (3 mg/kg), propranolol hydrochloride (2 mg/kg) and hexamethonium bromide (20 mg/kg) (Sigma-Aldrich NV/SA, Bornem, Belgium).
Sensitivity to alterations in heart rate
After stabilisation of all haemodynamic parameters, measurements were performed as described above during atrial pacing at 100, 120 and 140 beats per minute.
The conductance catheter was connected to a signal-processing unit (Sigma 5 DF, CDLeycom, Zoetermeer, The Netherlands).8 Parallel conductance (Gp) and blood resistivity (σb) were measured at regular intervals using the hypertonic saline method (injection of 5 ml NaCl 10% into the right atrium) and the CDLeycom resistivity meter respectively.9 The correction factor α was recalculated for each measurement using corresponding data from the PA flowprobe.10 Instantaneous segmental ventricular volumes (V(t)) were thus derived from conductance measurements G(t) using the formula:
V(t) = (1/α) (L2/σb) [G(t)−Gp]
where L represents the interelectrode spacing of the dual-field conductance catheter. All traces obtained from ECG, pressure, flow and conductance monitors were digitised at 300 Hz and stored for off-line analysis with algorithms written in Matlab® (The Mathworks Inc., Natick, MA, United States).
Leads from the 10 MHz ultrasound crystals were connected to a sonomicrometry system (Sonometrics Corporation, London, Ontario, Canada). These parameters were digitised at 960 Hz and stored for off-line analysis (Cardiosoft®, Sonometrics Corporation, London, Ontario, Canada).
Global RV contractility was quantified using the slope of the PRSW relationship as described previously.11 RV afterload was determined as effective pulmonary arterial elastance (Ea), and calculated as the ratio of end-systolic pressure to stroke volume.
Langrangian Strain (∊L) was computed from longitudinal segment shortening as ∊L = (SL-SL1)/SL1
where SL represents the actual segment length and SL1 the end-diastolic segment length in the first analysed cardiac cycle.4 End-diastole was defined as the peak R wave of the ECG. The onset of ejection was derived from the pulmonary flow signal.
SR was calculated by inverting and differentiating ∊L over 20 msec and SR-acceleration was obtained by differentiation of SR over 10 msec. SRmax represents the maximal SR during the ejection phase. SRmax was averaged from three consecutive cardiac cycles during steady state conditions in the different protocols (fig 1).
ISAmax was calculated as previously described using the acceleration of SR during isovolumic contraction.7 Calculations of myocardial acceleration were averaged from three consecutive cardiac cycles during steady state conditions in the different protocols (fig 1).
Results were statistically analysed using commercially available software (Statistica® for Windows, Statsoft, Tulsa, OK, USA). Data are expressed as mean (SD). The effects of alterations in inotropic state and heart rate were statistically analysed using repeated measurements analysis of variance (RMANOVA). The Tukey post-hoc test was performed to correct for multiple comparisons. Comparison of the baseline and the altered loading conditions was performed using the paired Student t test. Linear regression analysis was used to compare the changes in Mw and the changes in peak acceleration during the high-afterload protocol. Finally, a forward stepwise (F to enter = 5) multiple regression analysis with Mw, RV end-diastolic volume (EDV) and Ea as independent variables and SRmax and ISAmax as dependent variables was performed. A p value <0.05 was considered to be statistically significant.
The experimental protocol was completed in all six animals. General haemodynamic parameters measured during the inotropic modulation protocol, during acute changes in preload and afterload, and during changes in heart rate are shown in table 1.
Inotropic sensitivity (table 2)
Esmolol significantly decreased Mw, SRmax and ISAmax, while dobutamine produced significant increases in Mw, SRmax and ISAmax.
Effect of increased heart rate (table 2)
ISAmax increased significantly when atrial pacing rate was increased from 100 to 140 beats per minute. Global Mw and SRmax did not change during modulation of heart rate.
Effect of decreased preload (table 2)
Partial occlusion of the IVC resulted in an average reduction of RV end-diastolic volume of 40%, while global Mw, SRmax and ISAmax remained unchanged.
Effect of increased afterload (table 2)
Partial constriction of the PA resulted in an average RV pressure increase of 50%. This was associated with a decrease in SRmax, while ISAmax and Mw remained unchanged. Although individual responses for ISAmax and Mw to high afterload were variable, there was a significant correlation between the changes in ISAmax and in Mw (r2 = 0.89; p = 0.005), but not between changes in ISAmax and the magnitude of RV pressure rise (r2 = 0.18; p = 0.405) (fig 2). Pharmacological blockade of the autonomic nervous system in three animals produced a 20% decrease in heart rate, mean aortic pressure, cardiac output and peak RV pressure and a 30% decrease in contractility.
The forward stepwise multiple regression analysis with SRmax and ISAmax as the dependent variables and Mw, EDV and Ea as the independent variables showed that ISAmax is determined by Mw and EDV, while SRmax is determined by Mw, EDV and Ea (table 3).
The results of this study confirm that both SRmax and ISAmax correlate directly with the global contractility of the RV, as quantified independently by the slope of the RV preload recruitable stroke work relationship. More importantly, our data also show that SRmax consistently decreases in response to pulmonary artery banding, whereas ISAmax appears to be less sensitive to such acutely imposed changes in afterload. The latter observation supports previous data based on tissue velocity measurements and suggests that an isovolumic index of tissue deformation may be a more robust index of global RV contractility than indices obtained during the ejection phase.
Although SRmax is widely used in experimental and clinical studies to describe global LV and RV function, surprisingly little information is available on how this rate of cardiac tissue deformation during ejection is affected by altered loading conditions. Given the well-established inverse relationship between afterload and shortening velocity of isolated muscle preparations,12 the decrease of SRmax in response to PA banding as observed in our intact animal model is not surprising. Until now, little attention has been dedicated to the load-dependency of SRmax. Several groups have claimed SRmax to be a strong index of LV global contractility.3 This conclusion was based on experimental data on inotropic and chronotropic sensitivity only, ignoring the effects of loading conditions. Also for the RV, Jamal et al reported that longitudinal SR quantitates regional RV contractile function, although the sensitivity of SRmax to inotropic drugs was not studied and no reference index of contractility was used.4 Interestingly, their data showed that longitudinal SRmax increased after PA banding. We have no explanation for the apparent discrepancy with our data but speculate that a more pronounced PA banding may have caused tricuspid insufficiency or an exaggerated positive inotropic response to high afterload. We excluded the presence of significant tricuspid insufficiency by inspecting the pressure-volume loops and in addition used the slope of PRSW as an independent reference index of contractility. Our findings appear more consistent with a clinical study by Kjaergaard et al, who also observed a decrease in RV-SRmax in the presence of hypoxia-induced pulmonary vasoconstriction in volunteers.5
Vogel et al6 were the first to hypothesise that isovolumic indices should be less sensitive to altered cardiac loading conditions than ejection phase indices. Using Doppler myocardial imaging, they demonstrated that isovolumic acceleration corresponds well with contractile state of both the LV and RV, with minimal sensitivity to changes in afterload and preload. Load-sensitivity was tested by assessing beat-to-beat changes in IVA during brief occlusions of the PA and IVC, respectively. This technique does not allow a direct comparison of IVA with an independent reference method for assessment of contractile state during altered loading. In addition, IVA reflects the acceleration of tissue motion rather than tissue deformation, and is affected by tethering of adjacent structures and by cardiac translation and rotation. Although we evaluated load sensitivity during steady state conditions and focused on tissue deformation characteristics, our primary conclusions are consistent with the data of Vogel et al. Also, with regard to the effects of heart rate, our data confirm their findings that ISAmax directly correlates with chronotropic state. In the traditional concept, an ideal index of contractility allows quantification of inotropic state, independently of heart rate and loading state. Indices based on pressure-volume loop analysis were shown to meet this criterion over a wide range of experimental conditions and for that reason are still considered the gold standard approach to derive information on muscular cell properties from the physical behaviour of a beating heart.11 On the other hand, heart rate is a leading determinant of cytosolic calcium concentration, and is strictly linked to contractility in adult mammalian myocytes.13 In this respect, Vogel et al stated that the sensitivity to inotropic changes is higher for IVA than for pressure-volume-based indices, and claimed that only the former is able to detect the presence of a force–frequency relationship in intact hearts.6 14 Our data may provide indirect support for this hypothesis; the changes in response to positive and negative inotropic drugs were indeed much more pronounced for ISAmax than for the slope of PRSW and SRmax. Nevertheless, we cannot exclude the possibility that heart rate had an independent effect on the time-derived variable ISAmax. The solution to this important dilemma, however, requires a more complex study design with specific focus on the interaction between inotropic and chronotropic effects on contractility indices.
Clinical studies using tissue Doppler imaging have produced conflicting results with regard to the load-sensitivity of IVA, and clinical data on ISAmax are not, to our knowledge, available. In human volunteers, Kjaergaard et al found that RV-IVA was not affected by afterload changes caused by hypoxic pulmonary vasoconstriction.5 Likewise, rapid infusion of saline to increase preload did not alter IVA in this study. In contrast, Andersen et al reported significant changes in LV-IVA to occur even with discrete (10%) changes of end-diastolic volume in volunteers subjected to Trendelenburg position and nitroglycerin infusion respectively.15 None of these studies employed an independent method to assess contractility, and the reason for the reported discrepancies therefore remains elusive. Technical issues such as the temporal resolution used, the location of the sampling area, the timing with regard to the cardiac cycle and the alignment of the Doppler beam are all potential sources of observer variability and error. We used transit-time sonomicrometry from epicardially attached crystals to circumvent such limitations. Data were obtained at high frequency and related to ECG and pulmonary flow signals for an exact assessment of isovolumic and ejection phases of contraction. A similar invasive approach was used by Lyseggen et al,16 who observed a complex, biphasic relationship between ISA and preload and suggested that this variable may reflect diastolic properties rather than an active isovolumic systolic event. Unfortunately, no attempts were made to study the interaction with afterload. Our study focused specifically on the RV, where pathological increases in afterload are a common clinical problem. In these conditions, the differentiation between loading-induced RV dysfunction and reduced RV contractility has important therapeutic and prognostic implications. A non-invasive technique that allows this differential diagnosis would be an important clinical tool. From our data it appears that ISAmax but not SRmax may have this potential, although further studies with regard to the precise physiological correlate of ISA and further validations are clearly needed.
We acknowledge that our study has several limitations. We did not subject our model to bidirectional changes in loading conditions but only assessed the response to elevated afterload and decreased preload respectively. We selected these interventions because they are easily reversible and not associated with confounding effects. Augmenting preload with rapid infusions of colloid or crystalloid solutions often results in haemodilution, while afterload reduction is difficult to achieve in the RV, which is maximally unloaded in normal conditions. All measurements were obtained in open-chest experimental conditions. Opening of the thorax and pericardium undoubtedly affects overall cardiac motion and may therefore significantly change tissue velocities; however, its influence on cardiac tissue deformation is less pronounced.17
In conclusion, our experimental data show that regionally derived ISAmax and SRmax correspond to the global contractile state of the RV. ISAmax appears to be less sensitive to increased afterload conditions. Although further studies with regard to the precise physiological correlate of IVA are required, this study suggests that, for the study of RV function, isovolumic indices of myocardial tissue deformation are superior to ejection-phase indices.
Funding: Supported by research grants from the University Hospitals KULeuven (CM) and the Medizinische Fakultät, RWTH Aachen (SR). This funding source was not involved in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
Competing interests: None.