Blunted increase in plasma adenosine levels following dipyridamole stress in dilated cardiomyopathy patients


Eugenio Picano, MD, PhD, CNR, Institute of Clinical Physiology, via Moruzzi, 1, 56123 Pisa, Italy (fax: 0039-050-3152374; e-mail:


Background.  Heart failure is characterized by chronically increased adenosine levels, which are thought to express a protective anti-heart failure activation of the adenosinergic system. The aim of the study was to assess whether the activation of adenosinergic system in idiopathic dilated cardiomyopathy (IDC) can be mirrored by a blunted increase in plasma adenosine concentration following dipyridamole stress, which accumulates endogenous adenosine.

Methods.  Two groups were studied: IDC patients (n = 19, seven women, mean age 60 ± 12 years) with angiographically confirmed normal coronary arteries and left ventricular ejection fraction <35%; and normal controls (n = 15, six women, mean age 68 ± 5 years). Plasma adenosine was assessed by high-performance liquid chromatography methods in blood samples from peripheral vein at baseline and 12 min after dipyridamole infusion (0.84 mg kg−1 in 10 min).

Results.  At baseline, IDC patients showed higher plasma adenosine levels than controls (276 ± 27 nm L−1 vs. 208 ± 48 nm L−1, P < 0.001). Following dipyridamole, IDC patients showed lower plasma adenosine levels than controls (322 ± 56 nm L−1 vs. 732 ± 250 nm L−1, P < 0.001). The dipyridamole-induced percentage increase in plasma adenosine over baseline was 17% in IDC and 251% in controls (P < 0.001). By individual patient analysis, 18 IDC patients exceeded (over the upper limit) the 95% confidence limits for normal plasma adenosine levels at baseline, and all 19 exceeded (below the lower limit) the 95% confidence limits for postdipyridamole plasma adenosine levels found in normal subjects.

Conclusion.  Patients with IDC have abnormally high baseline adenosine levels and – even more strikingly – blunted plasma adenosine increase following dipyridamole infusion. This is consistent with a chronic activation of the adenosinergic system present in IDC, which can be measured noninvasively in the clinical theatre.


Dipyridamole is a pharmacological stress widely used in cardiac stress imaging [1–4]. According to a haemodynamic classification, it is a coronary vasodilator stress. It determines arteriolar dilation and perfusion heterogeneity [1] and/or absolute subendocardial underperfusion with ischaemic wall motion abnormalities [2]. If a biochemical classification is used, dipyridamole can be considered an ‘adenosinergic’ stress [5, 6], as it is an adenosine cellular reuptake inhibitor that determines endogenous adenosine accumulation in the interstitial space [7, 8]. In patients with coronary artery disease, dipyridamole infusion determines a threefold increase in plasma adenosine levels [9]. IDC patients are characterized by an activation of the adenosinergic system [10], and adenosine is thought to exert a protective effect on failing hearts [11] for its anti-inflammatory [12], antiapoptotic [13], antiadrenergic [14] and antioxidant effect [15, 16]. We hypothesized that the activation of adenosinergic system in IDC can be mirrored by a blunted increase in plasma adenosine concentration following dipyridamole stress. We assessed plasma adenosine concentrations at baseline and following dipyridamole stress in IDC patients and normal controls.


Study patients

Nineteen patients with IDC were recruited from the Cardiology Clinic of the Institute of Clinical Physiology of the National Research Council of Pisa. By selection, all patients had left ventricular ejection fraction <35% and angiographically normal coronary arteries. A control group of 15 normal subjects was also evaluated: they were either asymptomatic healthy controls with maximal negative exercise stress test (n = 11) or patients with atypical chest pain, negative maximal stress test and angiographically normal coronary arteries (n = 4). The main clinical characteristics of the two groups are reported in Table 1. All patients gave their written informed consent before testing.

Table 1.  Demographic, clinical and echocardiographic features of the study patients
 ICDControlP value
  1. IDC, idiopathic dilated cardiomyopathy; NYHA, New York Heart Association.

Age (years)60 ± 1268 ± 50.03
Gender (M/F)12/79/6NS
NYHA class2.421.270.000
Ejection fraction (%)30 ± 562 ± 40.000

All the IDC patients were on medical therapy at time of testing, most of them on a combined therapy of ACE-inhibitors and diuretics (15 of 19, 79%); of these 10 were on digoxin and six on beta-blockers. Only three patients were not receiving ACE-inhibitors and only two patients were not receiving diuretics. Only two patients received beta-blocking agents and diuretics. No patient was on inotropic agents.

Dipyridamole stress echocardiography

The standard protocol for dipyridamole stress echocardiography (cumulative dose 0.84 mg kg−1 over 10 min) was used [17]: 0.56 mg kg−1 over 4 min, followed by 4 min of no dose, and – if negative – additional 0.28 mg kg−1 over 2 min. During the procedure, two-dimensional echocardiographic, 12-lead electrocardiographic and blood pressure monitoring were continuously performed, according to the recommendations of the American Society of Echocardiography [4]. In all studies, segmental wall motion was semiquantitatively graded as follows: normal – normal wall motion at rest, with normal/increased wall motion after dipyridamole (score 2); akinetic – virtual absence of inward motion (score 3); dyskinetic – paradoxical wall motion away from the left ventricular centre in systole (score 4). The wall motion score index was derived by dividing the sum of individual segment by the number of interpretable segments [4]. Aminophylline (up to 240 mg over 3 min) was given at 5 min after the end of the test. Echocardiographic monitoring was performed throughout dipyridamole infusion and up to at least 5 min after the end of the infusion. Two-dimensional echocardiographic images were digitized at baseline, and 12 min after starting of dipyridamole infusion.

Adenosine assay

Adenosine plasma levels were analysed by an high-performance liquid chromatography (HPLC) method with Bio-Rad Isocratic UV Analyzer (Cambridge Scientific Products, Cambridge, MA, USA) and C-R Shimadzu recorder (Shimadzu Biotech, Kyoto, Japan), as previously described in detail [6, 9, 18]. Blood samples anticoagulated with heparin (100 UI mL−1 of blood) were collected in ice-cooled syringes, containing a stopping solution consisting of: eritro-9-(2-hydrossy-3nonil)adenine hydrocloride (EHNA 10 μmol L−1) + dipyridamole 50 μmol L−1 in order to block the adenosine uptake and production from other cellular sources.

The intra- and inter-assay coefficients of variation were 3.8 and 2%, respectively, as previously described [6, 9]. In previous studies, we assessed plasma adenosine values every 2 min during dipyridamole infusion, and consistently found the peak value at 12th minute – in keeping with the known temporal variation of the coronary vasodilatory effect [6, 9]. Therefore, in the present study, we decided to assess peak plasma adenosine at the 12th minute.

Statistical analysis

Continuous variables were expressed as mean value ± 1 standard deviation (SD). Differences were tested for significance by anova for repeated measures. Upper and lower 95% confidence limits for each variable were calculated from the two tails of the Student t-test distribution using the following formulas: mean value ± (2.042 × SD) and mean − (2.042 × SD), respectively. Relations between rest-stress variations in plasma adenosine and systemic haemodynamic variables are expressed in terms of linear regression analysis. A P value of <0.05 was considered significant.


The main baseline clinical and echocardiographic characteristics of the study patients are reported in Table 1.

Dipyridamole test: clinical and haemodynamic response

No complications or limiting side-effects occurred during dipyridamole infusion. Blood pressure and heart rate changes are reported in Table 2. No significant differences in systemic haemodynamics were observed between the two groups both at baseline and during stress, except for a higher resting heart rate (P < 0.03) and lower systolic blood pressure (P < 0.03) in IDC patients.

Table 2.  Haemodynamic and echocardiographic parameters at rest and after dipyridamole infusion
 ICDControlP value
  1. NS, not significant.

Heart rate
 Rest (b.p.m.)78 ± 2060 ± 40.03
 Dipyridamole (b.p.m.)91 ± 2185 ± 14NS
Systolic blood pressure
 Rest (mmHg)124 ± 21145 ± 100.032
 Dipyridamole (mmHg)117 ± 23140 ± 22NS
Diastolic blood pressure
 Rest (b.p.m.)68 ± 1788 ± 20.000
 Dipyridamole (b.p.m.)61 ± 1282 ± 100.01
Wall motion score index
 Rest2.1 ± 0.31.03 ± 0.060.000
 Dipyridamole1.8 ± 0.31.03 ± 0.060.000
Plasma adenosine
 Rest (nm L−1)276 ± 27208 ± 480.000
 Dipyridamole (nm L−1)322 ± 56732 ± 2500.000

Dipyridamole test: echocardiographic findings

Wall motion score index values at rest and following dipyridamole are reported in Table 2. No patient developed regional or global left ventricular dysfunction during dipyridamole stress.

Dipyridamole test: plasma adenosine findings

Values of plasma adenosine at baseline and during dipyridamole are reported in Table 2. When compared with normal controls, IDC patients had higher baseline adenosine values and lower values after dipyridamole administration (Fig. 1). At individual patient analysis, 18 IDC patients had abnormal values (outside 95% confidence intervals of the distribution in normal patients) at baseline; 19 had abnormal values following dipyridamole stress. There was no correlation between extent of blunting of adenosine increase and ejection fraction (r = 0.036; P = NS) in IDC patients.

Figure 1.

Individual patients values in group 1 [idiopathic dilated cardiomyopathy (IDC), left panel] and in group 2 (controls, right panel) at baseline (rest) and following dipyridamole (dipy). As a group, IDC patients show higher baseline values and lower postdipyridamole values when compared with control subjects.


Patients with IDC show higher resting plasma adenosine levels than normal controls. Following dipyridamole infusion, the increase in plasma adenosine is markedly lower in patients with IDC. These data are consistent with an activation of adenosinergic system in heart failure, with blunted adenosinergic reserve, which can be unmasked by dipyridamole stress used as a biochemical probe.

Dipyridamole as a biochemical stress

The principle of stress is widely used in medicine in general, and in cardiac imaging in particular. In this study, dipyridamole was used as a biochemical stress probing the adenosinergic system. Conceptually, it is similar to the methionine-loading test, which identifies patients with normal fasting levels of homocysteine and a propensity for the achievement of higher concentrations under certain circumstances [19]. Similarly, dipyridamole stress may identify patients with exhausted adenosinergic reserve and a clearly reduced capability to increase plasma adenosine concentration following dipyridamole, which inhibits cellular reuptake of endogenous adenosine mostly produced by the myocardium [7]. In our patient population only one of 19 IDC patients had a normal plasma adenosine value at baseline but all a clearly subnormal increase following dipyridamole stress. The derangement of the adenosinergic system can be more profound and frequent than can be assessed by a simple evaluation in resting conditions, exactly as it happens when a patient with coronary artery disease is tested with functional imaging. Abnormalities in myocardial perfusion and/or function can be absent at rest, and unmasked by application of the stress. It is tempting to use the same drug to explore both sides of myocardial activity: the functional side at the organ level evaluating myocardial contraction and/or perfusion and the metabolic side at the cellular level evaluating the plasma adenosine concentration at baseline and following dipyridamole infusion.

It is possible to speculate that the blunted increase in plasma adenosine levels following dipyridamole administration to dilated cardiomyopathy patients could be the expression of a reduced nucleoside transporters activity. This may explain either the reduced response to dipyridamole infusion and the high basal level of adenosine, the latter being the consequence, not only of a high production rate but also of a low cellular reuptake.

Our findings do not allow us to clarify how the reduced dipyridamole sensitivity occurs, nor the specific cellular target(s). As far as we know, the inhibitory effect of dipyridamole on nucleoside cell uptake is ubiquitous: we can only speculate that the blunted increase in plasma adenosine levels following dipyridamole administration to dilated cardiomyopathy patients could be the expression of a reduced nucleoside transporters activity. This may explain either the reduced response to dipyridamole infusion and the high basal level of adenosine, the latter being the consequence, not only of a high production rate but also of a low cellular reuptake. Cardiomyocytes and endothelial cells are putatively involved in such blunted dipyridamole response, as indirect evidence suggests that changes occurring in these cells in CHF could negatively affect nucleoside transporters activity.

Comparison with previous studies

Our data are consistent with the results of Funaya et al., who found increased adenosine levels in the systemic venous blood of patients with nonischaemic and ischaemic chronic heart failure [10]. They also hypothesized that because adenosine counteracts catecholamine-, renin-, angiotensin- and cytokine-induced cellular injury, increased adenosine levels may be endogenous compensatory mechanisms for heart failure [10]. These same mechanisms – largely unrelated to the classic antiplatelet effect – are thought to be crucial in the anti-ischaemic neuroprotective [20, 21] and cardioprotective [22, 23] effect of dipyridamole. The blunted increase in plasma adenosine that we observed after dipyridamole infusion may be consistent with this hypothesis, witnessing an exhaustion of the ‘adenosinergic reserve’ (i.e. the organism capability to increase the baseline value of plasma adenosine), which is already exploited at rest.

In the control subjects the resting adenosine levels was threefold higher than the value of 62 nmol L−1 reported by Funaya et al. in their own control group [10]. Two reasons may account for this difference. First, we used an HPLC-based method, which may have larger inter-individual variability than radio immuno assay (RIA) method. Secondly, our ‘controls’ had a clinical indication to stress testing and were substantially older than the patients of Funaya et al.

Study limitations

Adenosine is generally believed to be the biochemical mediator linking dipyridamole infusion to physiological effects on perfusion and function. However, this is a clear pathophysiological oversimplification, as dipyridamole has complex ‘ancillary’ effects, including increase in prostacyclin, potentiation of nitric oxide, and direct antioxidant effect [14, 16]. All these actions may co-mediate the effects of dipyridamole infusion, but we restricted our evaluation to circulating plasma adenosine. The adenosine independent effects of dipyridamole may also explain why the adenosine increase is blunted in cardiomyopathy patients, whereas the haemodynamic effects on heart rate and blood pressure were not.

Another study limitation is that plasma adenosine does not necessarily mirror the myocardial and, even less, the regional interstitial adenosine concentration – which is several-fold higher than plasma concentrations during ischaemia [17] and is really pivotal for the regulation of local coronary haemodynamics [18]. Nevertheless, obvious practical reasons favour the plasma adenosine assay. With all these limitations, plasma adenosine showed a peculiar pattern following dipyridamole administration in the different patient groups – in a study design in which each patient acted as his/her own control.

We did not evaluate, in the same patient, changes in adenosine levels during dipyridamole stress and following appropriate long-term treatment. In fact, theoretically, an improvement of adenosine metabolism (with the restoration of the blunting adenosine increase) might mirror the functional improvement. In patients with heart failure followed longitudinally after heart transplantation, alterations of peripheral A2A adenosine receptors progressively normalized to control values within 6 months, suggesting that improvement of cardiac performance is indeed accompanied by progressive restoration of a normal adenosynergic system [24].

Conflict of interest statement

No conflict of interest was declared.