OUP user menu

End-systolic pressure/volume relationship during dobutamine stress echo: a prognostically useful non-invasive index of left ventricular contractility

Aurelia Grosu, Tonino Bombardini, Michele Senni, Vincenzo Duino, Mauro Gori, Eugenio Picano
DOI: http://dx.doi.org/10.1093/eurheartj/ehi444 2404-2412 First published online: 16 August 2005


Aims Left ventricular end-systolic pressure–volume relationship (PVR) provides a robust, relatively load-insensitive evaluation of contractility and can be assessed non-invasively during exercise echo. Dobutamine might provide an exercise-independent alternative approach to assess inotropic reserve. The feasibility of a non-invasive estimation of PVR during dobutamine stress in the echo lab and its relationship with subsequent clinical events was assessed.

Methods and results We enrolled 137 consecutive patients referred for dobutamine stress echo. To build the PVR, the force was determined at different heart rate increments during stepwise dobutamine infusion as the ratio of the systolic pressure/end-systolic volume index. The PVR at increasing heart rate was flat-biphasic in 65 and up-sloping in 72 patients: 42 patients underwent surgery and 95 patients were treated medically (median follow-up, 18 months; interquartile range, 12–24). Events occurred in 18 patients (death in eight, acute heart failure in 10); a flat-biphasic PVR was independent predictor of events (RR=10.16, P<0.01).

Conclusion PVR is feasible during dobutamine stress. This index of global contractility is reasonably simple, does not affect the imaging time, and only minimally prolongs the off-line analysis time. It allows unmasking quite different, and heterogeneous, contractility reserve patterns underlying a given ejection fraction at rest. The best survival is observed in patients with up-sloping PVR, whereas flat-biphasic pattern is a strong predictor of cardiac events.

  • Bowditch treppe
  • Force–frequency relationship
  • Dobutamine stress echocardiography
  • Heart failure


Contractility is the inherent capacity of the myocardium to contract independently of changes in the pre-load or afterload. Whatever the problems of measuring it, contractility remains an essential corner concept to separate the effects of a primary change in loading conditions from an intrinsic change in the force of contraction.1,2 Recently, a non-invasive echocardiographic method has been proposed to assess the changes in contractility, as mirrored by end-systolic pressure–volume relationship (PVR), during physical or electrical stress, such as semi-supine exercise echo3 or external pacing echo in patients with permanent pacemakers.4 This novel method is based on the proven assumption that positive inotropic interventions are mirrored by smaller end-systolic volumes and higher end-systolic pressures. Dobutamine might provide an exercise-independent alternative approach to assess inotropic reserve in patients unable to exercise or without a permanent pacemaker. During dobutamine, the assessment of PVR involves both the force–frequency relationship and the effect of inotropic stimulation. Experimental studies showed that beta-adrenergic receptor stimulation produces an important enhancement of myocardial contractility, which is impaired in heart failure.5,6 In failing hearts in which the basal cAMP content of cardiomyocytes is decreased because beta-adrenergic receptors are down-regulated, the force–frequency relationship is altered.7,8

The aim of the study was to assess the feasibility of a non-invasive estimation of PVR at increasing heart rates during dobutamine stress in the echo lab and to relate PVR to clinical events in the subset of medically treated patients.


Study patients

In the Bergamo Cardiology Department, we enrolled 137 consecutive patients (102 males, age 66±11 years) referred for dobutamine stress echo (up to 40 µg/kg per minute) (Table 1). All patients gave written consent before performing the test.

View this table:
Table 1

Characteristics of the study patients

Age (years)66±11
Gender (M/F)102/35(75/25%)
LVEF (%)40±14
Previous MI58(42%)
PTCA/bypass (previous)11(8%)
Valvular disease12(9%)
NYHA class(I–IV)

LVEF, left ventricular ejection fraction; CAD, coronary artery disease; DCM, dilated cardiomyopathy; MI, myocardial infarction.

The underlying diagnosis was coronary artery disease in 100 patients, idiopathic dilated cardiomyopathy (DCM) (with angiographically normal coronary arteries) in 13, valvular heart disease in 12, and hypertension in 12. Coronary artery disease was defined by the presence of angiographically assessed coronary stenosis (≥50% quantitatively assessed diameter reduction in at least one major coronary vessel) or previous myocardial infarction (MI). Patient's functional classification was based on the NYHA class.

Data acquisition

Baseline and dobutamine stress echocardiography

All patients underwent transthoracic echocardiography at baseline. Intravenous access was secured, and dobutamine was infused with 3 min dose increments starting from 5 µg/kg per minute and increasing to 10, 20, 30, and 40 µg/kg, with atropine co-administration up to 1 mg.

Left ventricular end-diastolic (LVEDV) and end-systolic volumes (LVESV) were measured from apical four- and two-chamber view by an experienced observer using the biplane Simpson method.9 Only representative cycles with optimal endocardial visualization were measured and the average of three measurements was taken. The endocardial border was traced, excluding the papillary muscles. The frame captured at the R-wave of the ECG was considered to be the end-diastolic frame, and the frame with the smallest LV cavity the end-systolic frame. Images were acquired at baseline and at each 10-beat frequency increase during stress.

Data analysis

Regional wall motion analysis

Regional wall motion was assessed according to the recommendations of the American Society of Echocardiography from 1 (normal) to 4 (dyskinetic) in a 16-segment model of the left ventricle.10

Blood pressure analysis

One investigator recorded all blood pressures at rest and during each individual study step. The blood pressure recording was made using a sphygmomanometer and the diaphragm of a standard stethoscope.11

Volume analysis

After completion of the study, echocardiographic images were read by one experienced cardiologist unaware of the identity of the patient. The EDV and the ESV were measured at rest and at each step. All volume measures were normalized by dividing by body surface area.

Ejection fraction analysis

Left ventricular ejection fraction (LVEF) was measured at rest and at each stress step. LVEF contractile reserve (≥5% vs. baseline) was calculated as the difference (in absolute value) from baseline to peak stress.

End-systolic pressure–volume determination

To build the force–frequency relationship, the force was determined at each step as the ratio of the systolic pressure (cuff sphygmomanometer)/end-systolic volume index (biplane Simpson rule/body surface area). At each heart rate step, three cardiac cycles were memorized and the end-systolic volume was calculated as the mean value. The force–frequency relationship was built off-line. The slope of the relationship was calculated as the ratio between SP/ESV (systolic pressure/end-systolic volume) index increase (from baseline to peak stress) and heart rate increase (from baseline to peak stress). The PVR was defined up-sloping when dobutamine-induced SP/ESV index increase was higher than 25th percentile values of the entire study group (Figure 1A), flat when within the 25th percentile values (Figure 1C); biphasic, with an initial up-sloping followed by a later down-sloping trend, when peak dobutamine SP/ESV index was lower than intermediate stress values12,13 (Figure 1B). The critical heart rate was defined as the heart rate at which SP/ESV index reached the maximum value during progressive increase in heart rate.12 In biphasic pattern, the critical heart rate was the heart rate beyond which SP/ESV index declined by ≥5%.

Figure 1 Methodology of the force–frequency curve with dobutamine stress echo. Left, from upper to lower rows: systolic blood pressure by cuff sphygmomanometer (first row); LVESVs calculated with the biplane Simpson method (second row); heart rate increase during dobutamine infusion (third row); in the lowest row, the force–frequency curve built off-line with the values recorded at baseline (second column), and at different steps (third, fourth, fifth columns) up to peak exercise (seventh column). (A) Normal subject. An increased heart rate is accompanied by an increased systolic pressure with smaller end-systolic volumes (normal up-sloping PVR). (B) A subject with LV dysfunction (EF%=32%) without dilation, no stress-induced ischaemia. The PVR was biphasic, with an initial up-sloping followed by a later down-sloping trend, the critical heart rate (90 b.p.m.) was the heart rate beyond which SP/ESV index declined by ≥5%. The test was stopped at 20 gammas for limiting symptoms (dyspnoea). (C) A subject with post-MI depressed baseline LV function (EF%=30%). An increased heart rate at peak exercise is accompanied by no changes of end-systolic volumes (abnormal flat PVR).

Quality control of stress echocardiographic readings

All studies were performed by an experienced cardiologist with documented experience in stress echocardiography and who passed the quality-control procedures of stress echocardiography reading.14 Thirty randomly selected patients were analysed by a second less experienced observer who was blind to all other data. The first observer reanalysed 30 echocardiograms in a new random order after a minimum interval of 8 weeks.

Limits of agreement between readings were estimated as mean difference (bias) ±2 SD of the differences, as described by Bland and Altman.

Comparing the readings of two observers, the limits of agreement for baseline echo were −18.6–17.5 mL (ESV) and −11.6–12.7% (EF%). At intraobserver analysis, the limits of agreement were −7.1–7.5 mL (ESV) and −6.8–8.1% (EF%). Mean inter- and intraobserver differences for ESV were −0.6 and −0.3%, respectively. Mean inter- and intraobserver differences for EF% were 0.7 and 0.6%, respectively. Both for intraobserver and interobserver analysis, >95% of the differences were between d−2 SD and d+2 SD.

Follow-up data

Follow-up visits occurred at 6-month steps in medically treated patients. Cardiac events (total mortality, heart failure-related hospitalization) were the primary endpoints. Hospital and physician records and death certificates were used to ascertain the cause of death, which was attributed to a cardiac aetiology if a cardiac illness provoked the final presentation or if death was sudden and unexpected. Patients with NYHA class improvement were also considered.

Statistical analysis

SPSS 11 for Windows was utilized for statistical analysis. The statistical analyses included descriptive statistics (frequency and percentage of categorical variables and mean and standard deviation of continuous variables). Pearson's χ2 with Fisher's exact test for categorical variables and the Mann–Whitney test for continuous variables for intergroup comparisons were performed to confirm significance (using Monte Carlo method for small sample comparisons). Our significance tests were two-sided, in the sense that sufficiently large departures from the null hypothesis, in either direction, were judged significant.

Spearman rank correlation and partial Spearman rank correlation tests were used for correlation analyses.

The follow-up analyses included Kaplan–Meier survival curves and Cox proportional hazards models.

The following covariates were analysed: demographics (age, sex, BSA), incoming disease, medical therapy (beta-blockers, Ca-blockers, digoxin, diuretics, nitrates, ACE-inhibitors), echocardiographic data [resting and peak stress LVEF, LVEF contractile reserve (≥5% vs. baseline), wall motion score index (WMSI) at rest and peak stress, change in wall motion score index (the difference in wall motion score index from rest to peak stress), resting and peak stress SP/ESV index, flat-biphasic or up-sloping PVR, critical heart rate]. To check the proportional hazard assumption, i.e. that the hazard ratio for two subjects with fixed predictors is constant over time, log(−log[survival probability]) for different categories was plotted against time to ensure that the curves were reasonably parallel. In general, all proportionality assumptions were appropriate.

Assumption of linearity was assessed by the ANOVA test of linearity and by visual inspection of events' incidence rates in deciles of the continuous variables. Obvious deviations from linearity were seen in resting and peak stress LVEF, peak WMSI, resting and peak stress SP/ESV index, CHR. On the basis of this, we used these variables as dichotomous variables in all analyses: the deciles 1–5 vs. deciles 6–10 for rest LVEF, cut-off value 0.36; the deciles 1–4 vs. deciles 5–10 for peak LVEF, cut-off value 0.41; the deciles 1–4 vs. deciles 5–10 for peak WMSI, cut-off value 1.64; the deciles 1–6 vs. deciles 7–10 for rest SP/ESV index, cut-off value 3.17; the deciles 1–2 vs. deciles 3–10 for peak SP/ESV index, cut-off value 1.74; the deciles 1–2 vs. deciles 3–10 for rest CHR, cut-off value 100 b.p.m.

To assess independent correlates of variations, groups of variables (demographics, incoming disease, medical therapy, resting and stress echocardiographic data) were first examined in separate regression models with simultaneous introduction of covariates, as depicted in Table 3. Those variables, significant at the P<0.1 level in these initial models, were simultaneously entered in a summary regression model.

For the events during the follow-up, Kaplan–Meier survival curves were constructed using complete follow-up both for standard LVEF contractile reserve and for up-sloping vs. flat-biphasic PVR.

A subgroup of 62 patients with symptoms at the time of testing (NYHA class=II or III) was analysed for symptom improvement in the follow-up using the same covariates. In this subgroup, obvious deviations from linearity were seen only in peak SP/ESV index (cut-off value 2.98, deciles 1–5 vs. deciles 6–10) and in critical heart rate (cut-off value 100 b.p.m., deciles 1–3 vs. deciles 4–10). Because the assessment was only being done every 6 months and these are not just right-censored data, but also left-censored data, independent predictors were further analysed by means of follow-up life tables in which the period of observation was divided into smaller time intervals. For each interval, all people who have been observed at least that long are used to calculate the probability of a terminal event occurring in that interval. The probabilities estimated from each of the intervals are then used to estimate the overall probability of the event occurring at different time points. Life tables and Gehan's generalized Wilcoxon test were used to compare patients with and without symptom improvement. A P-value less than 0.05 was considered to be statistically significant.


Resting and dobutamine echocardiographic findings

Technically adequate images were obtained in all patients at baseline (by selection) and during stress. Twenty-one patients interrupted the test for occurrence of limiting symptoms (dyspnoea in four, angina in 12, non-sustained ventricular tachycardia in five). The conventional echocardiographic data for the two groups are described in Table 2.

View this table:
Table 2

Conventional clinical, echocardiographic data, and force–frequency relationship during dobutamine echo

Group 1: flat-biphasic PVR, 65 patientsGroup 2: up-sloping PVR, 72 patients
LVEDV index (mL/m2)103±36*83±34
LVMI (g/m2)144±43131±44
DOB (µg)33±933±10
Atropine (patients/yes)22(34%)18(25%)
WMSI rest1.95±0.61*1.68±0.66
WMSI peak2.04±0.61*1.69±0.60
Ischaemia during DOB, n(%)27(42%)23(32%)
Negative DOB (patients)23(35%)33(46%)
Viability during dob (patients)15(23%)16(22%)
HR rest (b.p.m.)74±1672±13
HR peak (b.p.m.)122±17123±19
Systolic pressure (rest)122±20127±20
Systolic pressure (peak)129±26*151±30
LVEF (rest %)36±13*43±14
LVEF (peak %)41±14*52±15
ΔLVEF (peak rest)5.0±7.1*9.1±6.5
SP/ESV index (rest)2.53±2.17*3.85±2.77
SP/ESV index (peak)3.24±2.92*6.43±5.03
PVR slope (×10−2)1.40±2.35*5.23±5.36
Critical heart rate (b.p.m.)100±19*125±18
On therapy at time of stress
 Beta-adrenergic blockers27(42%)25(36%)
 Ca-channel blockers3(5%)*12(17%)

LVMI, left ventricular mass index; DOB, dobutamine; LVEDV, left ventricular end-diastolic volume.

*P<0.05 between groups.

Pressure–volume relationship

The PVR (or modified force–frequency relation) response was abnormal (i.e. flat or biphasic) in 65 patients (Group 1) and normal (i.e. steep up-sloping) in 72 patients (Group 2). In Group 1, 34 patients had a flat pattern and 31 had a biphasic pattern. Examples of a normal steep up-sloping, an abnormal biphasic, and an abnormal flat response are shown in Figure 1A–C, respectively. Critical heart rate was lowest in Group 1 and highest in Group 2 (Table 2). The biphasic response was present in 27 out of 50 (54%) patients with positive and in 31 out of 87 patients (36%) with negative stress echo (χ2 P=0.05). Prevalence of biphasic PVR was higher in patients with previous MI (P=0.013). Peak WMSI was higher (2±0.6 vs. 1.75±0.63, P=0.015) and peak stress LVEF was lower (42±15 vs. 50 ±16%, P=0.02) in patients with biphasic PVR.

Relation between PVR and EF

Resting EF was normal in 27 patients and abnormal in 110 patients. Patients with normal resting EF had frequently abnormal (flat-biphasic) PVR, and a good proportion of patients with abnormal resting EF had normal PVR (Figure 2). Peak EF was also unable to separate patients with normal and abnormal PVR, although in this case there was a clear trend to more abnormal PVR in patients with lowest values of peak EF.

Figure 2 Upper rows, from left to right: pie charts of flat-biphasic (dark coded) vs. up-sloping (white coded) PVR in patients with rest normal (>55%, first row), slightly abnormal (EF%=54–45%, second row), abnormal (EF%=44–30%, third row), and severely abnormal (EF%=29–15%, fourth row) baseline EF%. Lower row: the same pie charts for peak stress EF%. Resting EF was normal in 27 patients and abnormal in 110 patients. Patients with normal resting EF had frequently abnormal (flat-biphasic) PVR, and a good proportion of patients with abnormal resting EF had normal PVR. Peak EF was also unable to separate patients with normal and abnormal PVR, although in this case there was a clear trend to more abnormal PVR in patients with lowest values of peak EF.

The prognostic value of PVR

Four patients were lost to follow-up (3%), therefore the final population consisted of 133 patients: 42 underwent clinically driven early intervention and were excluded from subsequent analysis.

The remaining 91 patients were treated medically. Patients were followed for a median of 18 months (range 1–24, interquartile range 6–24). Events occurred in 18 (cardiac death in eight, hospital readmission for heart failure symptoms in 10) patients (Figure 3).

Figure 3 Plot of the force–frequency relationship in Group 1 (flat-biphasic PVR, left panel) and Group II (up-sloping-PVR, right panel) medically treated patients. Black circles: cardiac death at follow-up. Black triangles: follow-up heart failure related hospitalization. Events prevalence is higher in flat-biphasic PVR.

Predictors of total events in the initial logistic regression models and in the ‘summary’ model are reported in Table 3. Independent predictors of total mortality were flat-biphasic PVR (RR=10.16, 95% CI=2.06–50.12, P=0.004) and ACE-inhibitors therapy (RR=0.26, 95% CI=0.08–0.82, P=0.022).

View this table:
Table 3

Predictors of total events in medically treated patients

Total eventsInitial regression modelsSummary model
RR (95% CI)P-valueRR (95% CI)P-value
 Age (years)1.04 (0.98–1.10)0.14
 Gender (M/F)0.53 (0.16–1.72)0.29
 BSA0.65 (0.16–9.78)0.76
Incoming disease
 Hypertension1.19 (0.45–3.08)0.72
 CAD3.26 (0.59–17.86)0.17
 Valvular disease1.19 (0.12–11.41)0.87
 DCM6.79 (1.07–42.82)0.041.09 (0.33–3.59)0.887
Medical therapy
 Beta-blockers1.22 (0.39–3.78)0.72
 Ca-blockers2.82 (0.57–14.01)0.20
 ACE-inhibitors0.36 (0.11–1.13)0.080.26 (0.08–0.82)0.022
 Nitrates1.92 (0.59–6.19)0.27
 Diuretics4.14 (1.01–16.97)0.051.31 (0.35–4.81)0.683
 Digoxin3.34 (1.07–10.42)0.042.69 (0.81–8.87)0.104
Echocardiographic variables
 LVEF (rest)1.05 (0.16–6.53)0.95
 LVEF (peak)0.74 (0.07–7.86)0.80
 ΔLVEF≥5%0.50 (0.15–1.70)0.27
 WMSI (rest)0.89 (0.88–3.81)0.88
 WMSI (peak)1,14 (0.13–9.68)0.90
 ΔWMSI2.77 (0.47–16.43)0.26
 SP/ESV index (rest)0.47 (0.05–4.09)0.50
 SP/ESV index (peak)0.21 (0.05–0.93)0.040.40 (0.11–1.40)0.155
 Flat-biphasic PVR4.05 (0.80–20.55)0.0910.16 (2.06–50.12)0.004
 Critical heart rate1.35 (0.47–3.84)0.57

Association (relative risk, RR and 95% confidence intervals, CI) of groups of variables with total events in medically treated patients, as calculated by the initial logistic regression models and by the ‘summary’ model. All the covariates were entered simultaneously into the regression models.

The overall event-free survival for the 37 patients with ΔLVEF <5% was 65% compared with 91% for the 54 patients with ΔLVEF >5% (Log rank=8.4, P=0.004). The overall event-free survival for the 43 patients with flat-biphasic PVR was 63% compared with 96% of the 48 patients with up-sloping PVR (Log rank=13, P=0.0003) (Figure 4).

Figure 4 Kaplan–Meier survival curves (considering total events as an endpoint) in medically treated patients stratified according to the presence of a flat-biphasic vs. up-sloping PVR or stress LVEF-based contractile reserve (ΔLVEF ≥5% vs. rest as cut-off value). The flat-biphasic pattern is a worse prognosis predictor. Beta-blocker treatment at the time of testing was similar (36 vs. 42%) in the 48 patients with up-sloping vs. the 43 patients with flat-biphasic PVR.

Clinical improvement in symptomatic patients at the time of testing

At follow-up, 35 of the 62 patients with NYHA class II or III at the time of testing improved symptoms (16 from NYHA class III to II, four from III to I, 15 from II to I).

At the initial regression models, predictors of improved symptoms were age (RR=0.96, 95% CI=0.93–1.004, P=0.074) and critical heart rate (RR=3.14, 95% CI=1.07–9.17, P=0.036); at the ‘summary model,’ the only independent predictor was critical heart rate (RR=2.90, 95% CI=1.12–7.49, P=0.028).

The best cut-off value for improved symptoms was ≥100 b.p.m.; 30 (67%) of the 45 patients with critical heart rate ≥100 b.p.m. vs. five (29%) of the 17 patients with critical heart rate <100 b.p.m. improved symptoms (Gehan's generalized Wilcoxon test=11.3, P=0008) (Figure 5).

Figure 5 Survival curves (considering improved NYHA class as an endpoint) in the 62 patients with NYHA class II or III at the time of testing. Thirty (67%) of the 45 patients with critical heart rate ≥100 b.p.m. vs. five (29%) of the 17 patients with critical heart rate <100 b.p.m. improved symptoms. The contraction frequency at which the PVR begins its descending limb (‘critical heart rate’) declines progressively with worsening outcome. In patients with events, the critical heart rate (95±20 b.p.m.) was significantly lower than in patients with improving symptoms (115±18 b.p.m.) or NYHA I class (122±23 b.p.m.).

Beta-blockers treatment at the time of testing was similar (18 patients, 40%) in the 45 patients with critical heart rate ≥100 b.p.m. vs. eight (47%) of the 17 patients with critical heart rate <100 b.p.m.


We found that LV end-systolic PVR can be obtained at different frequency steps during dobutamine stress echo, providing a relatively load-insensitive index of contractility. The three major merits of the present approach are the non-invasive, exercise-independent, and dynamic assessment, which allows to monitor PVR at increasing heart rates. The dynamic approach allows to identify in each patient the critical heart rate, a variable of established pathophysiological importance in the failing heart.2,12,13

Previous studies have used the PVR to assess LV contractility with invasive7,12,1518 or non-invasive3,4,1921 measurements. The approach used in the present study extends to dobutamine stress, the methodology previously proposed with exercise3 and pacing stress.4 All the studies are based on non-invasive volume and pressure measurements. LVEDV and LVESV were measured from apical four- and two-chamber view by an experienced observer using the modified Simpson's rule. This method of LV volume calculation during echo is widely used and accepted: the intraobserver and interobserver reproducibility of the method is high9,22 with improved results with the use of harmonic imaging.

Calculation of the systolic PVR requires measurement of the LV pressure in systole. Because only non-invasive measurements were available, systolic cuff pressure was used as a surrogate for end-systolic pressure. This certainly introduces an approximation; nevertheless, there is a tight relationship between peak and end-systolic pressure, and furthermore, any error is systematically distributed along the whole PVR, probably not affecting the slope values.20,21,23

Comparison with previous studies

Several experimental and clinical studies suggest that latent LV dysfunction may be manifested as limited contractile reserve in response to adrenergic stimuli5,6 represented by exercise12 or direct catecholamine infusion.7,12

Barghava et al.7 studied the response to progressive increases in heart rate by atrial pacing before and during dobutamine infusion in three normal subjects and in five patients with severe DCM. The slopes of the relations between frequency and force in control subjects were positive at baseline, and the mean slope increased substantially during dobutamine infusion. In patients with heart failure, the PVR was depressed and flattened. Dobutamine infusion shifted this relation upward slightly, without increase in mean slope, indicating lack of amplification.

Inagaki et al.12 demonstrated that during atrial pacing, the force–frequency relationship was biphasic in seven patients with severe LVH irrespective of LV function, but was preserved in 10 patients with less severe LVH and in 10 control subjects. The biphasic PVR was reversed to a normal ascending slope in the seven patients with severe LVH during isoproterenol-induced tachycardia. The authors concluded that a biphasic PVR at physiological pacing rates is one of the earliest markers of the transition from physiological to pathological LVH in patients with hypertension.

The results of the present study are broadly consistent with the previous evidences suggesting that the challenge of inotropic reserve with catecholamine infusion is useful to unmask depressed contractile reserve in a left ventricle with latent dysfunction.

PVR and prognostic implications

Several attempts have been done in the past to transfer the force–frequency relationship from the experimental lab to clinical applications.3,4,7,12,1519 Extensive data exist demonstrating that the force–frequency relation is profoundly modified in heart failure,8 ischaemic,7 dilated,7,16,18 or hypertrophic17 cardiomyopathy, and hypertensive12,15 and valvular13,19 diseases. However, prognostic data are conspicuously lacking to date. Our data demonstrate that patients with lack of normalization by dobutamine of an abnormal flat-biphasic PVR had worse prognosis. Patients able to express continuously increasing contractility reserve at heart rates >100 b.p.m. (despite symptoms on physiological effort) are more likely to improve clinically under optimal medical treatment.

Limitations of the study

The assessment of PVR during dobutamine infusion involves both the force–frequency relationship and the effect of inotropic stimulation. Previous reports demonstrated that beta-adrenergic stimulation-induced enhancement of the PVR is impaired in heart failure,68 whereas in initial, subtle, myocardial dysfunction, a biphasic PVR during pacing can be reversed to up-sloping PVR by dobutamine.12 To calculate the actual force–frequency relationship (i.e. a ‘pure’ index of contractility) one should perform this measurement with pacing to increase the heart rate, in the absence of inotropic stimulation. This is certainly feasible, even in a totally non-invasive way, in patients with permanent pacemakers.4 The changes in the adrenergic level during dobutamine infusion can be the base to get a clinical comparison in failing hearts of the PVR in the basal state (i.e. with pacing) and during adrenergic stimulation, as occurs during beta-agonists infusion.12

Near future technical improvement such as use of tonometry and transfer function for pressure measurement and 3D echo for volume measurement could improve assessment in non-geometric ventricles.


Assessment of the PVR during dobutamine stress testing is feasible. This index of global contractility is reasonably simple, does not affect the imaging time, and only minimally prolongs the off-line analysis time. It allows unmasking quite different, and heterogeneous, contractility reserve patterns underlying a given EF at rest. The pathophysiological advantage of using a relatively load-independent index translates into a more powerful prognostic stratification of PVR when compared with simpler, but more load-dependent, EF. This extra-information can be clinically useful in some patients' subsets.


T.B. is funded by a PhD programme on Cardiovascular Pathophysiology of the Scuola Superiore S. Anna, Pisa, and conceived this study, performed the data analysis, and drafted the manuscript. A.G., M.S., V.D., and M.G. were responsible for data collection and revised the manuscript, and E.P. gave a contribution to preparation of study design, data discussion, and critical revision of the manuscript.

Conflict of interest: none declared.


View Abstract