Author + information
- Received January 10, 2001
- Revision received April 23, 2001
- Accepted June 26, 2001
- Published online October 1, 2001.
- ↵*Reprint requests and correspondence:
Dr. John D. Parker, Division of Cardiology, Department of Medicine, Mount Sinai Hospital, Suite 1609, 600 University Avenue, Toronto, Ontario, Canada M5G 1X5
We studied the effects of nitroglycerin (GTN) therapy on the response to endothelium-dependent and independent vasoactive agents in the forearm circulation of healthy subjects.
Recent evidence suggests that therapy with GTN may induce specific changes in endothelial cell function, including increased superoxide anion production and sensitivity to vasoconstrictors. Additionally, continuous GTN therapy worsens endothelial function in the coronary circulation of patients with ischemic heart disease.
Forearm blood flow was measured with venous occlusion, mercury-in-silastic strain gauge plethysmography.
Sixteen male volunteers (26 ± 6 years) were randomized to no therapy (control) or GTN, 0.6 mg/h/24 h, for six days in an investigator-blind, parallel-design study. The flow responses to brachial artery infusions of acetylcholine ([Ach] 7.5, 15.0, 30.0 μg/min), N-monomethyl-L-arginine (L-NMMA) (1, 2, 4 μmol/min) and sodium nitroprusside (SNP) (0.8, 1.6, 3.2 μg/min) were recorded. The vasodilator responses to Ach were blunted in the GTN group as compared with the control group (p < 0.05). The vasoconstrictor responses to L-NMMA were also blunted in the GTN group (p < 0.001). In the GTN group, paradoxical vasodilation was observed in response to the lowest infused concentration of L-NMMA. The vasodilator responses to SNP did not differ between groups.
The response to Ach confirms the hypothesis that continuous GTN causes endothelial dysfunction. The responses to L-NMMA suggest that GTN therapy causes abnormalities in nitric oxide synthase (NOS) function; the vasodilation observed at the lowest infused concentration of L-NMMA in the GTN group also suggests that continuous GTN therapy is associated with a NOS-mediated production of a vasoconstrictor.
Organic nitrates remain important compounds in the therapy of coronary artery disease and congestive heart failure. Recent data suggest that therapy with nitroglycerin (GTN) causes specific biochemical changes at the level of the endothelium, including increased endothelin and superoxide anion production (1,2). These changes are associated with increased sensitivity to vasoconstrictors such as angiotensin II and phenylephrine and appear to play an important role in the development of tolerance (1,3). Furthermore, it is recognized that nitrate therapy induces abnormalities in
endothelium-dependent vasodilation (4–6). This was confirmed by a report from our laboratory demonstrating that continuous GTN therapy causes increased evidence of endothelial dysfunction in the coronary circulation of subjects with ischemic heart disease (7). The mechanism(s) of these vascular responses to nitrate therapy remain uncertain. However, there is now evidence from in vitro and animal experiments that the induction of nitrate tolerance is associated with abnormalities in endothelial nitric oxide (NO) synthase (NOS) function causing increased superoxide anion production (8). Similar changes in NOS function were demonstrated in an animal model of endothelial dysfunction induced by hypertension (9). The purpose of this investigation was to explore the presence and the mechanism of GTN-induced endothelial dysfunction further using a human, in vivo model. We set out to investigate whether continuous exposure to GTN was associated with abnormalities in endothelium-dependent vasodilator and vasoconstrictor responses and to investigate the impact of nitrate treatment on NOS function.
Sixteen healthy male subjects (age 20 to 41 years) were enrolled. Subjects were nonsmokers and abstained from medications, supplemental vitamins and alcohol or caffeine-containing beverages during the study period. The Human Subjects Review Committee of the University of Toronto approved the protocol, and written informed consent was obtained in all cases before the start of the study.
Measurement of forearm blood flow
Forearm blood flow (FBF) was determined simultaneously in both arms by venous occlusion strain-gauge plethysmography (D. E. Hokanson Inc., Bellevue, Washington) with calibrated mercury-in-silastic strain gauges using techniques previously reported by our laboratory (10). Briefly, circulation of the hand was excluded by inflating wrist cuffs to 200 mm Hg during measurement periods. The upper arm cuffs were inflated to 40 mm Hg and deflated at 10 s intervals (Hokanson rapid cuff inflator, D. E. Hokanson Inc.), and FBF was recorded as the average of five consecutive measurements. The same investigator who was blinded to the randomization of the subjects performed all FBF measurements.
On visit 1, only basal FBF was measured. On visit 2, FBF was measured at baseline and in response to normal saline and vasoactive agents as described in the following text.
Intra-arterial drug infusions
Under local anesthesia (lidocaine 1%), a 21-gauge catheter (Cook Inc., Bloomington, Indiana) was inserted into the brachial artery of the nondominant arm; at least 15 min were allowed to pass before FBF responses to the endothelium-dependent vasodilator acetylcholine (Ach) (7.5, 15, 30 μg/min CIBA, Mississauga, Ontario, Canada) and to the NOS-inhibitor N-monomethyl-L-arginine (L-NMMA) (1, 2 and 4 μmol/min, Clinalfa AG, Läufelfingen, Switzerland) were evaluated. To test endothelium-independent vasorelaxation, sodium nitroprusside (SNP) (Hoffman-La Roche, Mississauga, Ontario, Canada) was infused at 0.8, 1.6 and 3.2 μg/min. Infusion rate was kept constant at 0.4 ml/min with a precision pump (Harvard apparatus, South Natick, Massachusetts). Each concentration was infused for 5 min, and FBF measurements were performed during the last 2 min. Intra-arterial blood pressure was recorded after each infusion (Horizon 2000, Mennen Medical Inc., Clarence, New York) using the average of at least 15 cardiac cycles. The electrocardiogram was monitored continuously. The re-control interval between different drugs was at least 20 min.
On visit 1, after screening for admission into the study, baseline heart rate and blood pressure were measured after the patient had been sitting for 10 min and then standing for 5 min. Measurements were performed by an automatic sphygmomanometer (Critikon Company LLC, Tampa, Florida), and results represented the average of three measurements. Basal FBF was measured as described (10). Subsequently, subjects were randomized in an investigator-blind, parallel design to receive either transdermal GTN 0.6 mg/h (Novartis Pharmaceuticals, Dorval, Quebec, Canada; GTN group) or no treatment (control group). A nurse not involved in other study procedures randomized all volunteers and instructed them to wear the patch continuously for 24 h a day for the following six days. Three hours later, blood pressure measurements were repeated. During all the procedures of the study, investigators were blinded to the allocation group of the subjects.
Six days later, subjects returned to our laboratory. Blood pressure and heart rate measurements were repeated as above, and FBF was measured at baseline and in response to serial infusions of Ach, L-NMMA and SNP. At the end of the visit, the GTN patch was removed, and subjects were discharged from the laboratory.
Data and statistical analysis
Results are expressed as mean ± SE. Comparisons between groups for baseline values were performed with unpaired ttest. Differences in heart rate and blood pressure between visits were analyzed with a one-way analysis of variance (ANOVA) for repeated measures.
Forearm blood flow is expressed in units of ml/100 ml forearm tissue/min. The change in FBF from baseline (saline infusion) in the infused arm was calculated. Changes in FBF in the infused arm were also expressed as the ratio of FBF measured in the infused and the noninfused arm (I/N). Differences within and between groups were analyzed with a two-way ANOVA with repeated measures for the increasing concentrations of each infused drug. All post-hoc comparisons were made with the Student-Newman-Keul’s test. Differences were considered to be significant when p <0.05. Statistical analysis was performed using Sigmastat (Jandel Scientific, San Rafael, California).
Blood pressure and heart rate responses
On visit 1, standing heart rate and systolic blood pressure did not differ between groups. A significant decrease in standing systolic blood pressure was observed in the GTN group 3 h after the administration of the first patch (p < 0.02, Table 1). On visit 2, after six days of transdermal GTN treatment, systolic blood pressure returned to baseline values. Blood pressure did not change significantly in the control group (Table 1).
Heart rate increased significantly (p < 0.005, Table 1) 3 h after the first dose of transdermal GTN in the treatment arm and remained significantly higher after six days in the GTN group. No significant change in heart rate was observed in the control group. Blood pressure and heart rate did not change significantly in response to any of the intra-arterial drug infusions.
Effect of GTN treatment on FBF
On visit 1, FBF was similar between the two groups. On visit 2, baseline FBF remained similar between groups (Table 2), and within each group it did not significantly differ from the values obtained on visit 1.
Effect of GTN treatment on endothelium-dependent vasodilation
In the GTN group, the increase in FBF caused by the infusion of Ach was significantly blunted compared with the control group (Table 2). The ANOVA demonstrated a significant difference in the responses to Ach between the two groups (effect of group: p < 0.05). There was also a significant dose response relationship (effect of infused concentration: p < 0.001). Finally, the analysis revealed that there was no interaction between the effect of group and infusion rate. This demonstrates that the dose response relationship does not vary as a function of the group. During the infusion of the highest concentration of Ach, the increase in FBF observed in the GTN group was 4.9 ± 2.0 ml/min/100 ml, while in the control group it was 14.5 ± 3.7 ml/min/100 ml (Table 2). Thus, Ach-induced vasodilation was significantly impaired in the GTN group. When data were analyzed using the I/N ratio, similar results were obtained (effect of group: p < 0.02; effect of infused concentration: p < 0.001; interaction: p = NS; Fig. 1, Table 2).
Effect of GTN on the inhibition of NOS
In the GTN group, a paradoxical vasodilation in response to the lowest concentration of L-NMMA was manifested by an increase in FBF of 0.4 ± 0.1 ml/min/100 ml (p < 0.05, one-way ANOVA). By contrast, in the control group, the same infusion of L-NMMA caused the expected vasoconstriction with a reduction in FBF of 0.5 ± 0.2 ml/min/100 ml. At the two higher concentrations, L-NMMA caused vasoconstriction in both groups. However, this response was blunted in the GTN group as compared with the control group. At the peak infused concentration of L-NMMA, the observed decrease in FBF in the GTN group was 0.5 ± 0.1 ml/min/100 ml and 1.2 ± 0.2 ml/min/100 ml in the control group (Table 2). The differences between groups were highly significant (effect of group: p < 0.001; effect of infused concentration: p < 0.001; group by infusion interaction: p = NS). When data were analyzed as the I/N ratio, the differences remained highly significant (effect of group: p < 0.001; effect of infused concentration: p < 0.001; interaction: p = NS; Fig. 2, Table 2).
Effect of GTN on the endothelium-independent vasodilation
Sodium nitroprusside caused large increases in FBF in both groups. In the GTN group, the increase in FBF at the maximum concentration was 11.9 ± 2.6 ml/min/100 ml, while in the control group it was 16.5 ± 2.3 ml/min/100 ml (Table 2). Despite the presence of a trend toward a greater response in the control group, there was no significant overall difference in the FBF responses to SNP (ANOVA: effect of group: p = NS; effect of infused concentration: p < 0.001; group by infusion interaction: p = NS). When data are expressed as the I/N ratio (Fig. 3), the responses of the two groups are essentially identical (ANOVA: effect of group: p = NS; effect of infused concentration: p < 0.001; group by infusion interaction: p = NS).
During the past decade, there has been increasing recognition of the importance of the vascular endothelium as a mediator of responses to the organic nitrates. Ex vivo experiments from animal models have documented that tolerance can be reversed by the removal of the endothelium (2). Different laboratories have reported that tolerance to organic nitrates, particularly GTN, is associated with increased endothelial production of reactive oxygen species such as superoxide anion and peroxynitrite (1,11,12). The mechanism of this effect remains uncertain, although some studies suggest that membrane-bound oxidases present in vascular smooth muscle cells and regulated by angiotensin II are involved (13,14).
Summary of the findings
This study confirms that continuous GTN treatment induces significant changes in the function of the endothelium and NOS. The blunted vasodilator response to Ach is consistent with the existence of cross-tolerance to a NO-mediated vasodilator but is also an accepted manifestation of endothelial dysfunction. The observed abnormalities in the response to Ach might be due to decreased endothelial synthesis or to decreased bioavailability of endogenous NO, as suggested by previous reports in animal models and in humans with atherosclerosis (5,7). A decreased vasodilation to Ach might also be explained with an increased sensitivity of smooth muscle cells to vasoconstrictors (3)or to an increase in sympathetic activity. However, these hypotheses do not seem compatible with the responses to L-NMMA and SNP observed in this study.
The GTN group displayed very abnormal responses to L-NMMA with significantly blunted vasoconstriction in response to the two higher doses. These findings confirm the observation suggested by the response to Ach. They demonstrate that GTN therapy causes endothelial dysfunction mediated, at least in part, by abnormalities in NOS function. This, once again, is consistent with the major finding of this study, that GTN causes either reduced production or net bioavailability of NO. The significant vasodilation observed in response to the lowest concentration of L-NMMA was not expected and cannot be explained by a simple decrease in NOS activity. Although our experimental approach does not allow us to discern mechanism(s), this observation is consistent with the view that GTN therapy is associated with important abnormalities in NOS activity. We cannot establish the nature of such abnormalities; however, given the demonstration that after continuous GTN treatment, NOS has been shown to produce increased amounts of superoxide anion (8), one may hypothesize that this is the mechanism. Whether the vasodilation mediated by the low concentration of L-NMMA was caused by a greater decrease in superoxide anion generation, as compared with any change in NO generation, can only be a matter of speculation. Whatever the mechanism, it appears that at this concentration L-NMMA might paradoxically lead to greater bioavailability of NO.
The responses to SNP infusion confirm that the results obtained with Ach and L-NMMA are not attributable to a decreased vasodilatory effect of NO, despite the presence of a trend toward a reduced effect of SNP in the GTN group. This reduced response might be due to the quenching of SNP-derived NO by superoxide anion or to a decreased activity of the soluble guanylate cyclase in smooth muscle cells. However, the differences in flow responses to SNP were not significant. When flows in the infused arm were normalized to physiologic systemic changes in blood flow, which is considered a more reliable way to analyze FBF data, absolutely no difference was observed. Thus, whatever the mechanism of partially reduced sensitivity to SNP, it does not explain the abnormal responses to Ach and L-NMMA. The absence of a significant impairment in the responses to exogenous NO is not inconsistent with the hypothesis of GTN-induced increased oxidative stress. Indeed, it has been repeatedly demonstrated that SNP responses are not blunted in the setting of conditions such as long-term smoking and hypercholesterolemia, felt to be associated with increased production of free oxygen radicals. In those conditions, a difference was demonstrated between the responses to a direct, versus a NOS-mediated, NO donor (15,16).
Evidences for a NOS-dependent superoxide production in nitrate tolerance
Münzel et al. (8)have recently reported that the induction of GTN tolerance in rats is associated with increased expression of NOS, decreased NO and increased superoxide anion production. The observed increase in superoxide anion production was mediated, at least in part, by NOS, as inhibition of the enzyme reduced it. The findings of this study are consistent with an L-NMMA-mediated inhibition of superoxide synthesis. This might be considered unlikely since, in the presence of normal NOS function, L-NMMA has been found to induce an uncoupling of the enzyme, resulting in higher superoxide anion production (17). However, there is now clear evidence, both in the setting of GTN tolerance and in a model of angiotensin II-induced NOS dysfunction (18,19), that L-NMMA can inhibit the NOS-mediated production of superoxide anion.
The mechanism(s) of the abnormal NOS expression and function in the setting of nitrate tolerance remain uncertain. In his ex vivo study, Münzel et al. (8)demonstrate that it is a protein kinase C-dependent process, as tolerance could be reversed in vitro using specific protein kinase C inhibitors. They hypothesize that protein kinase C-mediated phosphorylation of NOS caused the observed abnormalities in its function (8). More recently, Abou-Mohamed et al. (20)demonstrated in cultured bovine aortic endothelial cells that GTN induces abnormalities in cellular L-arginine transport mechanisms leading to intracellular arginine depletion. They hypothesize that such L-arginine depletion results in an uncoupled NOS function leading to an increased superoxide anion production from molecular oxygen.
Importantly, Gruhn et al. (21)recently confirmed the existence of GTN-induced abnormal responses to L-NMMA. In their animal model of nitrate tolerance, a bolus infusion of L-NMMA caused a blunted increase in blood pressure compared with nontolerant rats. This effect was reversed by tetrahydrobiopterin, a cofactor for NOS, suggesting that the oxidation of this cofactor might be the cause of impaired NOS function in nitrate tolerance. Our findings are compatible with the same mechanism in humans.
Previous findings in animal models (8,20)suggest that abnormal NOS function is responsible, at least in part, for the increase in endothelial superoxide anion production associated with GTN treatment, and our results seem to be compatible with this hypothesis. A number of reports documented that antioxidant therapy can modify the development of tolerance (22,23), but, in our experience, the acute administration of vitamin C did not reverse tolerance to GTN (24). We do not believe that this latter observation is inconsistent with the present findings. If the trigger for NOS dysfunction or nitrate tolerance is a free radical-mediated process, it is possible that antioxidant supplementation can prevent this effect but be unable to reverse it.
Continuous versus intermittent GTN therapy
In this investigation, subjects underwent continuous transdermal GTN, and it is possible that the observed responses to intra-arterial infusions would have been different if an intermittent regimen had been employed. Nevertheless, continuous GTN therapy is used in a significant number of patients, particularly in the setting of unstable angina. Furthermore, since there is some documentation that intermittent regimens have been associated with changes in exercise tolerance and rebound ischemic events (25,26), it is possible that intermittent therapy might lead to NOS dysfunction and superoxide production, as recently confirmed by an ex vivo animal study (16). Whether similar responses would be observed during therapy with other long-acting nitrates is not known.
Clinical relevance of our findings
We believe these observations have direct clinical relevance. The fact that GTN therapy can cause profound endothelial dysfunction in otherwise normal individuals is surprising, given the traditional concept that NO donors might have beneficial effects in the setting of decreased NO bioavailability (27). This finding confirms our previous observation that therapy with continuous GTN worsens endothelial dysfunction in the coronary arteries of patients with coronary artery disease (7). An impaired endothelium-dependent production of NO is now considered per se a risk factor for cardiovascular disease.
The demonstration that therapy with GTN causes abnormalities in NOS function in humans in vivo provides support for the hypothesis generated by animal experiments that nitrate-induced abnormalities in NOS function might play an important role in the development of nitrate tolerance. As such, it is an important observation since mechanisms to prevent this effect and modify the development of tolerance can now be explored.
☆ Supported by a grant-in-aid from the Heart and Stroke Foundation of Ontario (B3160). Dr. Gori is supported by a granting arrangement between the University of Siena, Italy, and the University of Toronto. Dr. Mak is a Research Fellow of the Canadian Institute for Health Research.
- analysis of variance
- forearm blood flow
- ratio of forearm blood flow, infused to noninfused arm
- nitric oxide
- nitric oxide synthase
- sodium nitroprusside
- Received January 10, 2001.
- Revision received April 23, 2001.
- Accepted June 26, 2001.
- American College of Cardiology
- Münzel T,
- Giaid A,
- Kurz S,
- Stewart D.J,
- Harrison D.G
- Heitzer T,
- Just H,
- Brockhoff C,
- et al.
- Buga G.M,
- Griscavage J.M,
- Rogers N.E,
- Ignarro L.J
- Laursen J.B,
- Boesgaard S,
- Poulsen H.E,
- Aldershvile J
- Caramori P.R,
- Adelman A.G,
- Azevedo E.R,
- Newton G.E,
- Parker A.B,
- Parker J.D
- ↵Münzel T, Li H, Mollnau H, et al. Effects of long-term nitroglycerin treatment on endothelial nitric oxide synthase (NOS) gene expression, NOS-mediated superoxide production, and vascular NO bioavailability. Circ Res 2000;:E7–12.
- Dikalov S,
- Fink B,
- Skatchkov M,
- Stalleicken D,
- Bassenge E
- Griendling K.K,
- Minieri C.A,
- Ollerenshaw J.D,
- Alexander R.W
- Ueda S,
- Matsuoka H,
- Miyazaki H,
- Usui M,
- Okuda S,
- Imaizumi T
- Münzel T,
- Mollnau H,
- Hartmann M,
- et al.
- Pueyo ME, Arnal JF, Rami J, Michel JB. Angiotensin II stimulates the production of NO and peroxynitrite in endothelial cells. Am J Physiol 1998;274:C214–20.
- Watanabe H,
- Kakihana M,
- Ohtsuka S,
- Sugishita Y
- Parker J.D,
- Parker A.B,
- Farrell B,
- Parker J.O
- Harrison D.G,
- Bates J.N