The Phosphodiesterase 5 Inhibitor Sildenafil Has No Effect on Cerebral Blood Flow or Blood Velocity,but Nevertheless Induces Headache in Healthy Subjects

Abstract

Cyclic nucleotides are important hemodynamic regulators in many tissues. Glyceryl trinitrate markedly dilates large cerebral arteries and increases cGMP. Here, the authors study the effect of sildenafil, a selective inhibitor of cGMP-hydrolyzing phosphodiesterase 5 on cerebral hemodynamics and headache induction. Ten healthy subjects were included in a double-blind, placebo-controlled crossover study where placebo or sildenafil 100 mg (highest therapeutic dose) were administered on two separate days. Blood velocity in the middle cerebral artery (V_mca) was recorded by transcranial Doppler, and regional cerebral blood flow in the perfusion area of the middle cerebral artery (rCBF_mca) was measured using single photon emission computed tomography and ¹³³xenon inhalation. Radial and temporal artery diameters were studied using high-frequency ultrasound. Blood pressure and heart rate were recorded repeatedly. Headache responses and tenderness of pericranial muscles were scored verbally. Sildenafil caused no significant changes in rCBF_mca, V_mca, or in temporal or radial artery diameter, but heart rate increased and diastolic blood pressure decreased significantly compared to placebo. Despite the lack of cerebral arterial dilatation, sildenafil caused significantly more headache than placebo. The present results show that sildenafil 100 mg does not dilate cerebral or extracerebral arteries but nevertheless causes headache, which may be attributed to nonvascular mechanisms.

Keywords

Cerebral artery dilatation Cerebral blood flow Headache Phosphodiesterase inhibitors Sildenafil Nitric oxide

Sildenafil is a new highly selective inhibitor of the cGMP-degrading intracellular enzyme phosphodiesterase 5 (PDE5) widely used as treatment for erectile dysfunction. Sildenafil potentiates the actions of endogenous nitric oxide released during sexual stimulation from the endothelial cells or parasympathetic nerves (Ballard et al., 1998). Through inhibition of PDE5, the major enzyme responsible for cGMP degradation in that tissue, the accumulation of cGMP leads to relaxation of the smooth muscle cells, which in the penile tissue causes erection.

In cerebral arteries from different animals, the presence and activity of the Ca²⁺/calmodulin-stimulated PDE1, the cGMP-inhibited PDE3, the AMP-degrading PDE4, and the cGMP-degrading PDE5 have been described (Kruuse et al., 2001; Willette et al., 1997). Several studies have investigated the role of PDEs in the regulation of cerebral blood flow and cerebral artery diameter in animals using selective inhibitors of PDE (Kawamura et al., 1985; Parfenova et al., 1993). Of the selective PDE inhibitors, only PDE3 inhibitors have been studied using cerebral blood flow measurements, and no in vivo studies have investigated the ability of selective PDE inhibitors to dilate cerebral arteries.

Since species variability has been shown for the distribution and function of PDEs and since PDEs and their regulatory effects on cyclic nucleotide levels may be potentially important in cerebral hemodynamics, we studied the effects of sildenafil on cerebral hemodynamics. Furthermore, the headache-generating potential of sildenafil was evaluated since headache is one of the major side effects of this compound (Morales et al., 1998). Cerebral vasodilatation due to the accumulation of cGMP in the smooth muscle cells of the arteries has previously been proposed as the mechanism of headache induction by sildenafil (Morales et al., 1998).

MATERIALS AND METHODS

The study design was a double-blind, placebo-controlled crossover study with balanced, random oral administration of sildenafil 100 mg or placebo on 2 separate days, at least 1 week apart. Ten healthy subjects were included, 4 women and 6 men, mean age 24 years (range 20 to 31 years), with mean body weight of 77.3 kg (range 73.5 to 87 kg), and with no medical history of cerebrovascular or cardiovascular disorders, migraine, hypotension, or hypertension. When asked, no subjects reported having first-degree relatives (parents, siblings, children) suffering from migraine. All subjects gave informed consent before inclusion, and a medical examination, including an electrocardiogram, was performed before the study. The study was approved by the ethical committee of Copenhagen County and the Danish health authorities and was conducted according to the Helsinki II declaration.

Procedure on the day of the study

No food, medication, coffee, tea, alcohol, chocolate, or tobacco was allowed for the last 8 hours before the examination that began at 8:30 a.m. A brief medical examination was performed. After 30 minutes of rest, baseline values of blood velocities, blood pressure, muscle tenderness, heart rate, and regional cerebral blood flow in the perfusion area of the middle cerebral artery (rCBF_mca) were measured. A nontransparent gelatin capsule containing either sildenafil or placebo was ingested in a double-blinded fashion. The subject was kept for observation and measurements for a period of 120 minutes. The subjects remained in the supine position during the entire examination in quiet surroundings.

Single photon emission computed tomography (SPECT) acquisitions were performed at baseline, at 60 minutes, and at 120 minutes. Simultaneous transcranial Doppler (TCD) and radial artery diameter measurements were performed at baseline and every 15 minutes during the observation period. Each TCD measurement was followed by measurements of temporal artery diameter. TCD measurements were carried out immediately after SPECT acquisitions. Blood pressure and heart rate were measured repeatedly. Headache including side effects were recorded every 15 minutes, and the tenderness of the pericranial muscles was scored every 60 minutes during the observation period (120 minutes).

Transcranial Doppler

A time-averaged mean of the maximal blood velocities in the middle cerebral artery (V_mca) was recorded bilaterally by transcranial Doppler (TCD) (2 MHz, Multidop X Doppler, DWL; Lübech & Sipplingen, Germany; Transscan, EME, Uberlingen, Germany) using handheld probes (Thomsen and Iversen, 1993). The average of 4 measurements comprising approximately 4 cardiac cycles each over a time interval of 30 seconds was used. A fixed point for measurements of V_mca was chosen along the middle cerebral artery as the point that was as close as possible to the bifurcation between the middle cerebral and the anterior cerebral artery. This fixed point was then used throughout the study in each individual, and every measurement was done after carefully optimizing the signal from this point.

Measurements of end-tidal Pco₂ were performed simultaneously with the TCD measurements (POET, Criticare Systems Inc., Waukesha, WI, U.S.A.) using a mask placed lightly over the nose and mouth.

Single photon emission computed tomography

Measurements of CBF were performed with ¹³³Xe inhalation and SPECT using a brain-dedicated camera—Ceraspect (DSI, Inc., Waltham, MA, U.S.A.) with a stationary annular NaI crystal and a fast rotating collimator system. Each rotation took 10 seconds, acquiring 1 frame in a 30-frame dynamic protocol of ¹³³Xe inhalation (3 background, 9 wash-in, 18 washout) using the Kanno-Lassen algorithm (Kanno and Lassen, 1979). Thirty-two slices were reconstructed in a 64 × 64 matrix with each pixel measuring 0.33 × 0.33 cm. A set of eight transaxial slices was generated by adding four slices together to a total slice thickness of 1.33 cm. Attenuation correction using the Chang algorithm (μ = 0.05 cm) and nose-artifact correction were performed. The output in each pixel was the k1 value. The k1 was corrected with a partition coefficient (Λ) = 0.85 mL/g to get CBF values (mL·100 g⁻¹·min⁻¹).

A Datex Normocap 200 (Dameca, Rødovre, Denmark) was used for end-tidal Pco₂ measurements during CBF acquisition. A Ceretronic (Xenon Administration System, XAS SM 32C, Randers, Denmark) was used for Xe administration. A standard vascular region of interest (ROI) set of five slices positioned 3.6, 5.0, 6.3, 7.6, and 9.0 cm above the canthomeatal plane was used.

Since transcranial Doppler was performed on the middle cerebral artery, the distribution area for the middle cerebral artery was chosen as ROI for further analysis. Mean blood flow in the rCBF_mca for each side was calculated, and a mean of left and right rCBF_mca values was analyzed. Global CBF (gCBF) was calculated as an average of all regions of interests.

Diameter of temporal and radial artery

The temporal and radial artery diameters were measured by a high-resolution ultrasound unit (20 MHz, bandwidth 15 MHz; Dermascan C, Cortex Technology, Hadsund, Denmark) (Nielsen et al., 1993). Marks were drawn on the skin to ensure that the repeated measurements were performed in the same place and to ensure reproducibility in measurements from day to day. The diameter of the temporal artery was measured at the frontal branch of the superficial temporal artery on the side of the dominant brain hemisphere assessed by hand dominance (on left side when right-handed).

Measurements of temporal artery diameter were not performed in 3 of the 10 subjects (hence, n = 7 in the results) because of technical difficulties.

Measurements of the radial artery were made on the left arm proximal to the distal volar wrist crease where maximal pulsation of the artery was felt by palpation.

Headache recordings

Headache intensity was scored every 15 minutes on a verbal scale from 0 to 10, where “1” represents a very mild headache (including a feeling of pressing or throbbing), “5” a headache of moderate intensity, and “10” the worst headache imaginable (Iversen et al., 1989). Headache characteristics, location, and other side effects were recorded. All subjects were asked to record their headache, headache characteristics, and any intake of medication as well as other side effects every hour at home on a headache diary card from 3 to 12 hours after drug administration.

Tenderness score

Bilateral palpation of 8 pairs of muscle and tendon insertion (masseter, temporal, frontal, sternocleidomastoid and trapezius muscles, coronoid and mastoid processes, and neck muscle insertions) was performed at baseline, 60 minutes, and 120 minutes. Tenderness of pericranial muscles was recorded according to the Total Tenderness Scoring system on a 4-point scale (0–3) (Langemark and Olesen, 1987). Summed values of left and right side were summed to a total tenderness score for analysis.

Statistics

All values are presented as means ± SDs. P < 0.05 was considered significant.

The area under the curve for each variable, V_mca, CBF, rCBF_mca, blood pressure, heart rate, and Pco₂ on the separate days was chosen as summary measure when analyzing difference in response between treatments. The calculated values were compared using a paired t-test. Changes over time for the variables were analyzed with two-way analyses of variance with factors time and subjects. The multivariate analysis of variance was repeated for each single variable. If an overall effect of time was found, further analysis was performed using a multiple range test for location of significant changes (Stat-graphics 3.3). Difference in headache response and total tenderness score after placebo and sildenafil (area under the curve and median peak response) were analyzed using Wilcoxon rank sum test.

RESULTS

Hemodynamics

There was no significant difference between the area under the curve for rCBF_mca (P = 0.43) or gCBF (P = 0.58) between the two treatments. Baseline values for gCBF were 46.5 ± 8.7 mL·100 g brain tissue⁻¹ · min⁻¹ on placebo and 47.3 ± 9.5 mL·100 g brain tissue⁻¹ ·min⁻¹ on sildenafil and showed no change over time compared to baseline on either day (P = 0.67 and P = 0.94, sildenafil and placebo, respectively). Baseline value for rCBF_mca was 45.8 ± 10.1 mL·100 g brain tissue⁻¹·min⁻¹ on placebo and 45.7 ± 7.5 mL·100 g brain tissue⁻¹·min⁻¹ on sildenafil, also without changes from baseline after the two treatments (P = 0.70 after sildenafil and P = 0.66 after placebo).

End-expiratory Pco₂ was not affected by either placebo or sildenafil, and there was no significant difference between treatments (P = 0.73).

Blood velocity in the middle cerebral artery (V_mca) did not change significantly over time on the placebo day (P = 0.33) or the day of sildenafil (P = 0.30), and the area under the curves did not differ between treatments (P = 0.91) (Fig. 1).

Fig. 1.

Mean (n = 10) blood velocity in the middle cerebral artery (V_mca) over time. Absolute values (± SD) of the mean blood velocity are presented after sildenafil (represented by filled squares, ■) and after placebo (represented by open squares, □). No significant difference in responses between the two treatments was seen. Since regional cerebral blood flow remained unchanged, sildenafil did not dilate the middle cerebral artery.

According to the following equation: rCBF = velocity x lumen area, V_mca may be taken as an indirect measure of change in diameter of the middle cerebral artery when rCBF remains unchanged (Sorteberg, 1992). Sildenafil 100 mg induced a mean 0.8 ± 3.3% increase in diameter after 1 hour and mean 1.9 ± 3.6% increase after 2 hours. The mean dilatory response after placebo was 1.9 ± 4.5% after 1 hour and 2.3 ± 4.2% after 2 hours (P = 0.58 and P = 0.82, 1 and 2 hours, respectively, in comparisons between treatments).

Radial artery diameter was unchanged by sildenafil compared to placebo (P = 0.42). The temporal artery diameter increased compared to baseline at 60 minutes (P = 0.03) after sildenafil, but not at other time points, and the difference between the area under the temporal diameter curve for sildenafil and placebo was not significant (P = 0.39)(Fig. 2).

Fig. 2.

Temporal (n = 7) and radial (n = 10) arterial diameter over time. Mean absolute values (± SD) of temporal (represented by triangles, ▲) and radial artery diameter (represented by squares, ■) are shown after intake of placebo (by open symbols, Δ and □) and after sildenafil (by filled symbols, ▲ and ■). The temporal artery increases significantly (P = 0.03) at time 60 minutes after sildenafil; however, the change was very minor and was not significantly different from placebo.

Diastolic blood pressure decreased from mean values of 60 ± 5 mm Hg at baseline to 56 ± 6 at 90 minutes (P = 0.01) and heart rate increased from a mean value of 55 ± 9 beats per minute at baseline to a maximum of 64 ± 10 (P = 0.02) beats per minute 60 minutes after sildenafil. Although the change was small, it was significantly different from that seen on the day of placebo (P = 0.01 for diastolic blood pressure and P = 0.02 for heart rate). Systolic blood pressure was not significantly affected compared to placebo (P = 0.16) (Fig. 3).

Fig. 3.

Mean values (n = 10) of heart rate (●), systolic (■), and diastolic (▲) blood pressure are shown in percent of baseline over time after placebo (open symbols) and sildenafil (filled symbols). Systolic blood pressure remained unchanged (P = 0.16), diastolic blood pressure decreased slightly (P = 0.01), and heart rate increased (P = 0.02) after ingestion of sildenafil compared to placebo.

Headache and other side effects

Sildenafil induced headache in nine subjects compared to three after placebo (P = 0.031). Median peak headache score after placebo was 0 (range 0 to 3) and after sildenafil 2.5 (range 0 to 6) (P = 0.007)(Figs. 4A and B). Area under the headache curve was significantly different between sildenafil and placebo (P = 0.008). Median time-to-peak headache score was 2 hours from administration of sildenafil, but a second peak was seen also at 6 hours (Fig. 4B).

Fig. 4.

Individual (thin lines) and median headache scores (thick line with filled circle, ●) after placebo (A) and after sildenafil 100 mg (B). Nine of 10 subjects reported headache after sildenafil and 3 of 10 after placebo (P = 0.031). The headache was mild and transient, although after sildenafil in three subjects the symptoms fulfilled the criteria for migraine without aura after 5 to 7 hours with moderate to severe, mostly throbbing aggravated headache, accompanied by nausea. None of the subjects had first-degree relatives having migraine.

Three subjects fulfilled the IHS criteria for a single attack of migraine without aura including accompanying symptoms such as nausea 5 to 7 hours after sildenafil administration but none after placebo. Headache characteristics and accompanying symptoms for each subject on the day of sildenafil are shown in Table 1.

Table 1.

Headache characteristics in healthy subjects after the intake of sildenafil

Subject no.	Age/sex (yrs)	Headache characteristics (localization/peak intensity/quality)	Time to peak (hours)	Aggravated	Accompanying symptoms(nausea/photophobia/phonophobia)	Fulfillment of migraine criteria
1	31/M	Bilat./2/pressing	0.75	No	No/no/no	No
2	23/M	None	0	No	No/no/no	No
3	23/F	Bilat./4/throbbing	5	Yes	Minor/minor/no	Yes
4	20/F	Bilat./5/throbbing and pressing	6	Yes	Minor/no/no	Yes
5	26/M	Right/3/throbbing and pressing	2	Yes	Minor/no/no	No*
6	25/M	Left/6/throbbing	7	Yes	Minor/no/no	Yes
7	23/F	Left/2/throbbing and pressing	0.5	No	No/no/no	No
8	22/M	Bilat./2/pressing	2	No	No/no/no	No
9	25/M	Right/3/stabbing	2	No	No/no/no	No
10	22/F	Bilat./2/pressing	2	No	No/no/no	No

The headache was scored on a verbal scale from 0 to 10; accompanying symptoms were graded none, minor, moderate, or severe. Three subjects fulfilled the criteria for migraine without aura during the 12-hour period after ingestion of sildenafil 100 mg. The most commonly reported headache quality was “pressing”; a few subjects, however, reported short-lasting throbbing headache alternating with pressing headache (herein listed as “throbbing and pressing”). M, male; F, female; bilat., bilateral.

The nausea was reported at a time without concomitant headache.

Flushing was present in all subjects after sildenafil and in three after placebo. Nine reported a feeling of warmth after sildenafil and three after placebo, while two had palpitations with sildenafil and one had palpitations with placebo.

The subjects were asked to report any effects on sexual functions. Only two of six male subjects reported increased frequency of erection on the day of sildenafil, and none reported any effects on the placebo day. The female subjects reported no sexually related effects of sildenafil.

Tenderness of pericranial muscles

No difference in total tenderness score was seen between placebo and sildenafil (P = 0.14). The median baseline values were 4.5 (range 2 to 29) on the day of placebo and 4 (range 1 to 31) on the day of sildenafil.

DISCUSSION

The effects of inhibitors of phosphodiesterases on cerebral blood flow have been studied during the last decades in relation to treatment of stroke, dementia, and cerebral vasospasm, using nonselective and nonspecific inhibitors such as pentoxifylline and papaverine (Kruuse et al., 2000a). More recently, selective PDE3 inhibitors such as cilostazol, olprinone, and milrinone (Arakawa et al., 2001; Kobayashi et al., 1985; Yu et al., 2000) and to smaller degree PDE5 inhibitors, such as dipyridamole (Kruuse et al., 2000b) have been investigated in cerebral vascular disease.

In the present study, we demonstrate that inhibition of cGMP degradation by the selective PDE5 inhibitor sildenafil had no significant effects on cerebral blood flow and diameters of cerebral and extracerebral arteries. However, sildenafil was able to induce headache in healthy subjects without concomitant cerebral large artery dilatation.

Phosphodiesterase 5 distribution and sites of action of sildenafil

Phosphodiesterase 5 is present in penile vascular smooth muscle cells and has recently been reported to be present in brain tissue, mostly cerebellum and hippocampus and in the superior cervical ganglion (Giordano et al., 2001; Giorgi et al., 1994; Loughney et al., 1998). PDE5 has also been shown to be present and active in the guinea pig basilar artery (Kruuse et al., 2001) and in human cerebral arteries (unpublished data).

Oral administration of sildenafil 100 mg results in a peak plasma concentration of 1 μg/mL after approximately 1 hour (Jackson et al., 1999). The previously described intracellular increase of cGMP in smooth muscle cells and enhancement of relaxation by nitric oxide (NO) occurs well within this plasma concentration of sildenafil (Ballard et al., 1998; Jeremy et al., 1997). Sildenafil is the most selective PDE5 inhibitor available for human use, the mean IC₅₀ being 3.5 nmol/L for the isolated PDE5 enzyme. For the retinal PDE6, the mean IC₅₀ was 33 nmol/L and 281 nmol/L for PDE1, but the potency against PDE2, PDE3, and PDE4 is very low (Ballard et al., 1998).

Apart from the effect on smooth muscle cells, sildenafil may also have a central effect in the brain. Inhibitors of PDE5 increase cGMP concentrations in certain brain areas (Van Staveren et al., 2001) and it was found that sildenafil enhances auditory attention in healthy subjects suggesting a central effect of this lipophilic drug (Schultheiss et al., 2001).

This is consistent with previous studies, showing that sildenafil also acts on sensory nerve fibers in stimulating penile erection (Ballard et al., 1998; Medina et al., 2000) and modulates central NO-cGMP pathways in the rat brain involved in erection (Sato et al., 2001). It remains to be investigated whether activation of sensory nerve fibers is also taking place in humans. However, in light of the aforementioned data this seems very likely.

Cyclic nucleotides, phosphodiesterases, and cerebral vasodilatation

The endothelium of cerebral arteries has a basic production of NO, which through activation of soluble guanylate cyclase produces cGMP. Modulation of both cGMP and cAMP elicits essential regulation of cell signaling and probably cerebral vasodilatation through protein kinases, ion channels, and intracellular Ca²⁺ (Faraci and Sobey, 1999; Parfenova et al., 1993). The intracellular concentration of the cyclic nucleotides is regulated through their rate of production by adenylate cyclases or guanylate cyclases, through degradation by phosphodiesterases, or through efflux from the cell (Beavo, 1995).

Nitric oxide may cause an increase not only in cGMP but also in cAMP in some tissue, because the increase in cGMP inhibits PDE3 and thus increases intracellular cAMP in vitro (Sandner et al., 1999). This effect of NO has not been verified in cerebral arteries so far, and, at least in guinea pig cerebral arteries, the NO donor, sodium nitroprusside, did not increase cAMP levels (Kruuse et al., 2001). Upregulation of phosphodiesterases may also modulate the effects of NO. Thus, continuous infusion of NO donors in rats induced tolerance to the vasodilating properties of NO on isolated rat aortas by upregulation of PDE1A1, thus increasing the Ca²⁺/calmodulin-stimulated cGMP degradation (Kim et al., 2001).

As mentioned previously, the presence of several PDE families, PDE1, PDE3, PDE4, and PDE5 have been found to be active in cerebral arteries of animals (Kruuse et al., 2001; Willette et al., 1997). The presence and importance of the different PDE families has, however, not yet been established in human cerebral arteries. In studies on animals, most nonselective and selective PDE inhibitors induce cerebral artery dilatation, although with different potency depending on species and vessels tested. Selective PDE3 inhibitors show effective endothelial independent relaxation, whereas PDE1, PDE4, and PDE5 inhibitors elicit endothelial-dependent relaxation (Kruuse et al., 2001; Parfenova et al., 1993; Rosenblum et al., 1993).

Table 2 (Kruuse et al., 2000a, b ; Bowton et al., 1988; Nishimura et al., 1993; McHenry et al., 1983; Wang and Obrist, 1976; Schmidt and Waldemar, 1990; Yu et al., 2000; Ito et al., 1999; Beavo, 1995; Stoclet et al., 1995; Fujishige et al., 1999; Fawcett et al., 2000) gives an overview of the previously reported effects of PDE inhibitors on CBF, blood velocity, and general hemodynamics as well as the occurrence of headache in healthy subjects. Recently, milrinone, which is mainly a PDE3 inhibitor, was shown to reverse cerebral vasospasms and possibly improved clinical outcome in patients with sub-arachnoidal hemorrhage (Arakawa et al., 2001).

Table 2.

Effects of phosphodiesterase inhibitors on cerebral hemodynamics in healthy subjects

	Pentoxifylline	Theofylline	Papaverine	Nimodipine	Olprinone	Dipyridamole	Sildenafil
Main PDE inhibited	Nonselective	Nonselective	Nonselective	PDE1	PDE3	PDE5, 10, 11	PDE5
cAMP/cGMP change	Mainly cAMP ↑	cAMP↑	Mainly cAMP ↑	cAMP↑	cAMP↑	Mainly cGMP↑	cGMP↑
Other effects	Adenosine	Adenosine	Adenosine	Ca²⁺-channel		Adenosine
	receptor	receptor	receptor	blocker		reuptake
	antagonist	antagonist	inhibitor			inhibitor
Effects on:
CBF	No change	↓	↑	↑	No change	No change
V_mca	No change	ND	ND	ND	ND	↓	No change
Temporal artery	No change	ND	ND	ND	ND	ND	No change
Radial artery	No change	ND	ND	ND	ND	ND	No change
Blood pressure	No change	No change	Not reported	↓	Systolic ↓, no change in diastolic	No change	Diastolic ↓, no change in systolic
Heart rate	↑	No change	Not reported	Note reported	No change	↑	↑
Pco₂	↓	↓	No change	Pco₂ reactivity↓	ND	↓	No change
Headache	No	Not reported	Yes	Not reported	Not reported	Yes	Yes
Reference	Kruuse et al. 2000a	Bowton et al.1988; Nishimura et al. 1993	McHenry et al. 1983, Wang and Obrist, 1976	Schmidt and Waldemar, 1990	Yu et al., 2000	Ito et al., 1999; Kruuse et al., 2000b	Present article

The above compounds are considered inhibitors of phosphodiesterases (PDE) (Beavo, 1995; Stoclet et al., 1995), although some are not only nonselective and inhibit more than one PDE family but also nonspecific and thus have other effects than PDE inhibition. Table lists reported significant changes in different hemodynamic parameters and headache after administration of the PDE inhibitors compared with placebo in healthy subjects. Some studies have not investigated (ND, not done), or do not report, all parameters. PDE1 is the Ca²⁺/calmodulin-stimulated PDE; PDE3 is inhibited by cGMP and hydrolyzes cAMP at a higher rate than cGMP. PDE5 is specific for cGMP hydrolysis. Dipyridamole is considered an inhibitor of PDE5, but it has recently also been described as inhibiting the isolated enzymes of PDE10 (present in brain; Fujishige et al., 1999) and PDE11 (present in skeletal muscle; Fawcett et al., 2000), which hydrolyze both cAMP and cGMP.

cAMP, cyclic adenosine monophosphate; cGMP, cyclic guanine monophosphate; CBF, cerebral blood flow; V_mca, middle cerebral artery blood velocity; Pco₂, partial pressure of carbon dioxide; ↑, increase; ↓, decrease.

Theoretically, sildenafil should be an equally effective vasodilator but, unexpectedly, it had no effect in the present study. The lack of a dilatory response to sildenafil in cerebral arteries and arterioles in healthy subjects may indicate that cGMP is not accumulated in cerebral smooth muscle or that a compensatory regulatory mechanism very effectively decreases cGMP. This could be through a negative feedback mechanism, where an increased activity of PDE5 overruled the effect of PDE5 inhibition or where an upregulation of PDE1 is initiated. Another PDE may also be responsible for lack of dilatation, most likely PDE10A that degrades both cAMP and cGMP, with cGMP degradation being inhibited by cAMP. It is also inhibited by dipyridamole and is present in brain tissue (Fujishige et al., 1999). Another explanation may be that the PDE5 in smooth muscle cells is a splice variant, which only to a minor degree is inhibited by sildenafil, but not enough to elicit vasodilatation. The expression of different splice variants of PDE5 is known in the human penile cavernosum (Lin et al., 2000), but the functional importance of this remains to be investigated.

Mechanisms of nitric oxide- and sildenafil-induced headache

Nitric oxide is proposed to induce headache either by stimulating pain-sensitive nerve fibers around cerebral arteries or by arterial dilatation (Olesen et al., 1993). Similar to NO donors, dipyridamole, an adenosine reuptake inhibitor and a much less selective inhibitor of PDE5 than sildenafil, induces headache concomitant to a dilatation of large cerebral arteries in healthy subjects, but there was a temporal dissociation between headache and arterial dilatation. It was not possible to conclude which effect of dipyridamole was responsible for the headache, except that it was due to mechanisms related to cyclic nucleotides, either directly by increased cGMP or indirectly by adenosine-stimulated increase in cAMP (Kruuse et al., 2000b).

Sildenafil induced headache in 9 of 10 healthy subjects compared with 3 of 10 after placebo. The major difference in the sildenafil-induced headache compared with headache induced by NO donors was the apparent lack of effects on cerebral hemodynamics in the number of subjects studied. Thus, no clear dilatation of the cerebral arteries was seen after sildenafil on the same number of subjects used previously to show dilation by NO donors and histamine.

Sildenafil may theoretically have an effect on veins; however, in previous studies it seems to be equally modest to the effect on arteries (Jackson et al., 1999; Wallis et al., 1999). Also, in a previous study on migraine patients during attack of migraine, Queckenstedt's maneuver was performed to increase the pressure on the venous sinuses, without causing an increase in the headache pain. This speaks against a role of dilating venous sinuses in the migraine pain (Daugaard et al., 1998).

Three subjects fulfilled the IHS criteria for one attack of migraine without aura, although they had never experienced migraine before, and reported having no first-degree relatives suffering from migraine. The only major difference compared to migraine was in the quality of headache. In migraine a throbbing headache is dominant, while sildenafil predominantly induced headache of pressing quality.

The frequency of induced headache by sildenafil was considerably higher in the present study than reported in previous safety data (Morales et al., 1998). This is most likely due to a lack of focus on headache in the previous studies.

In parallel with the present study, we have studied the effects of sildenafil in migraine sufferers. It caused no change in cerebral blood flow or cerebral artery diameter, but nevertheless induced migraine attacks.

The present results, where headache without significant dilatation of cerebral arteries was induced by an inhibitor of a cGMP degrading enzyme, make it likely that NO induces headache through cGMP-related mechanisms and not by a direct action of NO on the pain-sensitive perivascular nerves. The marked NO-induced dilatation of large cerebral arteries may be an epiphenomenon rather than a pain-activating mechanism, because the headache-generating effects of sildenafil do not seem to be via intracerebral or extracranial arterial dilatation.

Footnotes

Acknowledgements

We thank the subjects for participating and enduring the provoked pain. We also thank L. Elkaer, K. Enghave, and O. Jonassen for technical assistance, and H. Dige-Petersen for generous use of equipment. Pfizer A/S, Denmark, generously supplied the sildenafil tablets.

References

Arakawa

Kikuta

Hojo

Goto

Ishii

Yamagata

(2001) Milrinone for the treatment of cerebral vasospasm after subarachnoid hemorrhage: report of seven cases. Neurosurgery 48:723–728; discussion 728–730.

Ballard

S.A.

Gingell

C.J.

Tang

Turner

L.A.

Price

M.E.

Naylor

A.M.

(1998) Effects of sildenafil on the relaxation of human corpus cavernosum tissue in vitro and on the activities of cyclic nucleotide phosphodiesterase isozymes. J Urol 159:2164–2171.

Beavo

J.A.

(1995) Cyclic nucleotide phosphodiesterases: functional implications of multiple isoforms. Physiol Rev 75:725–748.

Bowton

D.L.

Haddon

W.S.

Prough

D.S.

Adair

Alford

P.T.

Stump

D.A.

(1988) Theophylline effect on the cerebral blood flow response to hypoxemia. Chest 94:371–375.

Daugaard

Thomsen

Olesen

(1998) No relation between cephalic venous dilatation and pain in migraine. J Neurol Neurosurg Psychiatry 65:260–262.

Faraci

F.M.

Sobey

C.G.

(1999) Role of soluble guanylate cyclase in dilator responses of the cerebral microcirculation. Brain Res 821:368–373.

Fawcett

Baxendale

Stacey

McGrouther

Harrow

Soderling

Hetman

Beavo

J.A.

Phillips

S.C.

(2000) Molecular cloning and characterization of a distinct human phosphodiesterase gene family: PDE11A. Proc Natl Acad Sci USA 97:3702–3707.

Fujishige

Kotera

Michibata

Yuasa

Takebayashi

Okumura

Omori

(1999) Cloning and characterization of a novel human phosphodiesterase that hydrolyzes both cAMP and cGMP (PDE10A). J Biol Chem 274:18438–18445.

Giordano

De Stefano

M.E.

Citro

Modica

Giorgi

(2001) Expression of cGMP-binding cGMP-specific phosphodiesterase (PDE5) in mouse tissues and cell lines using an antibody against the enzyme amino-terminal domain. Biochim Biophys Acta 1539:16–27.

10.

Giorgi

Squitti

Bonsi

Paggi

Toschi

(1994) Activities of 3′:5′ cyclic nucleotide phosphodiesterases in the superior cervical ganglion of rat: characterization, compartmentalization and observations in young and old animals. Neurochem Int 25:493–500.

11.

Ito

Kinoshita

Tamura

Yokoyama

Iida

(1999) Effect of intravenous dipyridamole on cerebral blood flow in humans: a PET study. Stroke 30:1616–1620.

12.

Iversen

H.K.

Olesen

Tfelt Hansen

(1989) Intravenous nitroglycerin as an experimental model of vascular headache: basic characteristics. Pain 38:17–24.

13.

Jackson

Benjamin

Jackson

Allen

M.J.

(1999) Effects of sildenafil citrate on human hemodynamics. Am J Cardiol 83:13C–20C.

14.

Jeremy

J.Y.

Ballard

S.A.

Naylor

A.M.

Miller

M.A.

Angelini

G.D.

(1997) Effects of sildenafil, a type-5 cGMP phosphodiesterase inhibitor, and papaverine on cyclic GMP and cyclic AMP levels in the rabbit corpus cavernosum in vitro. Br J Urol 79:958–963.

15.

Kanno

Lassen

N.A.

(1979) Two methods for calculating regional cerebral blood flow from emission computed tomography of inert gas concentrations. J Comput Assist Tomogr 3:71–76.

16.

Kawamura

Watanabe

Kimura

(1985) Effect of cilostazol, a new antithrombotic drug, on cerebral circulation. Arzneimittelforschung 35:1149–1154.

17.

Kim

Rybalkin

S.D.

Wang

Zhang

Munzel

Beavo

J.A.

Berk

B.C.

Yan

(2001) Upregulation of phosphodiesterase 1A1 expression is associated with the development of nitrate tolerance. Circulation 104:2338–2343.

18.

Kobayashi

Yamaguchi

Katsube

Kitani

Okada

Tsunematsu

(1985) Long-term effect of cilostazol on cerebral blood flow in chronic cerebral infarction. Arzneimittelforschung 35: 1193–1197.

19.

Kruuse

Jacobsen

Thomsen

Hasselbalch

Frandsen

Dige-Petersen

Olesen

(2000a) Effects of the non-selective phosphodiesterase inhibitor pentoxifylline on regional cerebral blood flow and large arteries in healthy subjects. Eur J Neurol 7:629–638.

20.

Kruuse

Jacobsen

T.B.

Lassen

L.H.

Thomsen

L.L.

Hasselbalch

S.G.

Dige-Petersen

Olesen

(2000b) Dipyridamole dilates large cerebral arteries concomitant to headache induction in healthy subjects. J Cereb Blood Flow Metab 20:1372–1379.

21.

Kruuse

Rybalkin

S.D.

Khurana

T.S.

Jansen-Olesen

Olesen

Edvinsson

(2001) The role of cGMP hydrolyzing phosphodiesterases 1 and 5 in cerebral artery dilatation. Eur J Pharmacol 420: 55–65.

22.

Langemark

Olesen

(1987) Pericranial tenderness in tension headache: a blind, controlled study. Cephalalgia 7:249–255.

23.

Lin

C.S.

Lau

Lue

T.F.

(2000) Expression of three isoforms of cGMP-binding cGMP-specific phosphodiesterase (PDE5) in human penile cavernosum. Biochem Biophys Res Commun 268:628–635.

24.

Loughney

Hill

T.R.

Florio

V.A.

Uher

Rosman

G.J.

Wolda

S.L.

Jones

B.A.

Howard

M.L.

McAllister-Lucas

L.M.

Sonnenburg

W.K.

Francis

S.H.

Corbin

J.D.

Beavo

J.A.

Ferguson

(1998) Isolation and characterization of cDNAs encoding PDE5A, a human cGMP-binding, cGMP-specific 3′,5′-cyclic nucleotide phosphodiesterase. Gene 216:139–147.

25.

McHenry

L.C.

Jr Stump

D.A.

Howard

Novack

T.T.

Bivins

D.H.

Nelson

A.O.

(1983) Comparison of the effects of intravenous papaverine hydrochloride and oral pavabid HP capsulets on regional cerebral blood flow in normal individuals. J Cereb Blood Flow Metab 3:442–447.

26.

Medina

Segarra

Torondel

Chuan

Domenech

Vila

J.M.

Lluch

(2000) Inhibition of neuroeffector transmission in human vas deferens by sildenafil. Br J Pharmacol 131:871–874.

27.

Morales

Gingell

Collins

Wicker

P.A.

Osterloh

I.H.

(1998) Clinical safety of oral sildenafil citrate (Viagra) in the treatment of erectile dysfunction. Int J Impot Res 10:69–74.

28.

Nielsen

T.H.

Iversen

H.K.

Tfelt Hansen

Olesen

(1993) Small arteries can be accurately studied in vivo, using high frequency ultrasound. Ultrasound Med Biol 19:717–725.

29.

Nishimura

Yoshioka

Yamamoto

Akiyama

Miyamoto

Kawakami

(1993) Effect of theophylline on brain tissue oxygenation during normoxia and hypoxia in humans. J Appl Physiol 74:2724–2728.

30.

Olesen

Iversen

H.K.

Thomsen

L.L.

(1993) Nitric oxide supersensitivity: a possible molecular mechanism of migraine pain. Neuroreport 4:1027–1030.

31.

Parfenova

Shibata

Zuckerman

Mirro

Leffler

C.W.

(1993) Cyclic nucleotides and cerebrovascular tone in newborn pigs. Am J Physiol 265:H1972–H1982.

32.

Rosenblum

W.I.

Shimizu

Nelson

G.H.

(1993) Interaction of endothelium with dilation produced by inhibitors of cyclic nucleotide diesterases in mouse brain arterioles in vivo. Stroke 24:266–270.

33.

Sandner

Kornfeld

Ruan

Arendshorst

W.J.

Kurtz

(1999) Nitric oxide/cAMP interactions in the control of rat renal vascular resistance. Circ Res 84:186–192.

34.

Sato

Zhao

Christ

G.J.

(2001) Central modulation of the NO/cGMP pathway affects the MPOA-induced intracavernous pressure response. Am J Physiol Regul Integr Comp Physiol 281:R269–R278.

35.

Schmidt

J.F.

Waldemar

(1990) Effect of nimodipine on cerebral blood flow in human volunteers. J Cardiovasc Pharmacol 16:568–571.

36.

Schultheiss

Muller

S.V.

Nager

Stief

C.G.

Schlote

Jonas

Asvestis

Johannes

Munte

T.F.

(2001) Central effects of sildenafil (Viagra) on auditory selective attention and verbal recognition memory in humans: a study with event-related brain potentials. World J Urol 19:46–50.

37.

Sorteberg

(1992) Cerebral artery blood velocity and cerebral blood flow. In: Transcranial Doppler ( Newell

D.W.

Aaslid

, eds), New York: Raven Press, pp 57–66.

38.

Stoclet

Karavis

Komas

Lugnier

(1995) Cyclic nucleotide phosphodiesterases as therapeutic targets in cardiovascular diseases. Exp Opin Invest Drugs 4:1081–1100.

39.

Thomsen

L.L.

Iversen

H.K.

(1993) Experimental and biological variation of three-dimensional transcranial Doppler measurements. J Appl Physiol 75:2805–2810.

40.

Van Staveren

W.C.

Markerink-Van Ittersum

Steinbusch

H.W.

De Vente

(2001) The effects of phosphodiesterase inhibition on cyclic GMP and cyclic AMP accumulation in the hippocampus of the rat. Brain Res 888:275–286.

41.

Wallis

R.M.

Corbin

J.D.

Francis

S.H.

Ellis

(1999) Tissue distribution of phosphodiesterase families and the effects of sildenafil on tissue cyclic nucleotides, platelet function, and the contractile responses of trabeculae carneae and aortic rings in vitro. Am J Cardiol 83:3C–12C.

42.

Wang

H.S.

Obrist

W.D.

(1976) Effect of oral papaverine on cerebral blood flow in normals: evaluation by the xenon-133 inhalation method. Biol Psychiatry 11:217–225.

43.

Willette

R.N.

Shiloh

A.O.

Sauermelch

C.F.

Sulpizio

Michell

M.P.

Cieslinski

L.B.

Torphy

T.J.

Ohlstein

E.H.

(1997) Identification, characterization, and functional role of phosphodiesterase type IV in cerebral vessels: effects of selective phosphodiesterase inhibitors. J Cereb Blood Flow Metab 17:210–219.

44.

Mizushige

Ueda

Nishiyama

Seki

Aoyama

Ohkawa

Matsuo

(2000) Effect of olprinone, phosphodiesterase III inhibitor, on cerebral blood flow assessed with technetium-99m-ECD SPECT. J Cardiovasc Pharmacol 35:422–426