Sildenafil Mechanism of Action: The Complete NO-cGMP-PDE5 Pathway Explained

Clinical medical image for sildenafil generic: Sildenafil Mechanism of Action: The Complete NO-cGMP-PDE5 Pathway Explained

At a glance

  • Drug class / selective PDE5 inhibitor (oral tablet, 20 to 100 mg)
  • Primary target / phosphodiesterase type 5 enzyme in corpus cavernosum
  • Signaling cascade / sexual stimulation triggers NO release, which activates soluble guanylate cyclase to produce cGMP
  • How sildenafil helps / blocks cGMP breakdown, amplifying and prolonging the natural erectile response
  • PDE5 selectivity / approximately 10-fold selective for PDE5 over PDE6 and over 80-fold selective over PDE3
  • Time to peak plasma / median 60 minutes (range 30 to 120 minutes) in fasted state
  • Plasma half-life / 3 to 5 hours; active metabolite (N-desmethyl sildenafil) adds modest additional PDE5 inhibition
  • Efficacy benchmark / Goldstein et al. (NEJM 1998, N=532) showed 69% to 80% of attempts resulted in successful intercourse at 50 to 100 mg doses
  • Key requirement / sexual stimulation is required because sildenafil does not initiate NO release on its own
  • FDA approval / originally approved 1998 for erectile dysfunction; also approved at 20 mg for pulmonary arterial hypertension (Revatio)

The Nitric Oxide Signal: Where Everything Begins

Erection starts with a neurochemical event, not a vascular one. Parasympathetic nerve fibers from the S2 to S4 spinal segments release acetylcholine and nonadrenergic, noncholinergic (NANC) neurotransmitters at cavernous nerve terminals during sexual arousal. The dominant NANC transmitter is nitric oxide, synthesized by neuronal nitric oxide synthase (nNOS) within the nerve endings and by endothelial nitric oxide synthase (eNOS) in the sinusoidal endothelium of the corpus cavernosum [1].

Burnett et al. demonstrated in 1992 that nNOS-containing nerve fibers are distributed throughout human corpus cavernosum tissue, confirming NO as the principal mediator of penile smooth muscle relaxation [2]. This finding was a turning point. Before it, researchers debated whether vasoactive intestinal peptide or prostaglandins drove the relaxation response. NO settled the question.

Once released, NO diffuses into adjacent smooth muscle cells within milliseconds. It does not act through a membrane receptor. Instead, it binds directly to the heme group of soluble guanylate cyclase (sGC), a cytoplasmic enzyme. This binding causes a conformational change that increases sGC catalytic activity by 200- to 400-fold [3]. The activated enzyme converts guanosine triphosphate (GTP) into cyclic guanosine monophosphate (cGMP), the second messenger that drives every downstream step of the erectile response.

Without sexual stimulation, there is no NO release. Without NO, there is no cGMP surge. This is why sildenafil requires arousal to work.

cGMP Accumulation and Smooth Muscle Relaxation

cGMP acts as a molecular switch. Rising intracellular cGMP concentrations activate protein kinase G (PKG, also called cGMP-dependent protein kinase type I), which phosphorylates multiple downstream targets to reduce cytoplasmic calcium [4]. PKG opens potassium channels (particularly large-conductance calcium-activated K+ channels, or BKCa channels), causing membrane hyperpolarization. It also phosphorylates the inositol 1,4,5-trisphosphate receptor-associated cGMP kinase substrate (IRAG), which suppresses calcium release from the sarcoplasmic reticulum [5].

The net effect is rapid. Intracellular free calcium drops from roughly 600 nM to below 200 nM. Smooth muscle cells in the trabecular walls and helicine arteries relax. Blood flow through the cavernous arteries increases from approximately 5 mL/min in the flaccid state to 60 to 80 mL/min during tumescence [6]. The expanding sinusoidal spaces compress subtunical venules against the tunica albuginea, trapping blood and generating rigidity. This venous occlusion, called the veno-occlusive mechanism, is what distinguishes tumescence from full erection.

Without PDE5 inhibition, the entire cGMP pool is hydrolyzed within seconds. PDE5 cleaves the 3',5'-phosphodiester bond of cGMP to produce inactive 5'-GMP. The enzyme is densely expressed in corpus cavernosum tissue at concentrations higher than in nearly any other human tissue [7]. This is precisely where sildenafil intervenes.

How Sildenafil Inhibits PDE5

Sildenafil citrate (molecular formula C22H30N6O4S) is a pyrazolopyrimidinone compound. Its structure mimics the purine ring of cGMP closely enough to occupy the catalytic site of the PDE5 enzyme, acting as a competitive inhibitor [8]. The IC50 of sildenafil for PDE5 is 3.5 nM, meaning that 50% of PDE5 activity is blocked at a drug concentration of just 3.5 billionths of a molar [9].

When PDE5 is occupied by sildenafil, cGMP molecules accumulate instead of being degraded. The cGMP concentration in smooth muscle cells rises two- to threefold above what NO signaling alone would produce [10]. This amplification is the pharmacologic basis for improved erections in men whose natural NO-cGMP signal has been weakened by endothelial dysfunction, neuropathy, aging, or other pathology.

Dr. Ian Osterloh, the Pfizer clinical researcher who led early sildenafil development, described the drug's mechanism in a 2004 retrospective: "Sildenafil does not create an erection. It permits the natural erectile process to proceed when it has been compromised" [11]. This distinction matters clinically. The drug has no effect in the absence of sexual stimulation because it only amplifies an existing cGMP signal.

The competitive nature of the inhibition also means that very high local NO concentrations can partially overcome the drug's effect, while very low NO states (severe neuropathy or radical prostatectomy without nerve sparing) can limit its efficacy regardless of dose.

PDE Isoenzyme Selectivity and Off-Target Effects

The human genome encodes 11 PDE families (PDE1 through PDE11), each with distinct tissue distributions, substrate preferences, and regulatory properties. Sildenafil's clinical profile depends on its relative selectivity across these families.

Against PDE6 (the phosphodiesterase found in retinal rod and cone photoreceptors), sildenafil shows an IC50 of approximately 34 nM, making it roughly 10 times less potent than at PDE5 [9]. This relatively modest selectivity gap explains the visual disturbances reported by 3% to 11% of men taking sildenafil at 100 mg, including a transient blue tinge to vision (cyanopsia) and increased light sensitivity [12]. PDE6 inhibition transiently disrupts the phototransduction cascade in retinal photoreceptors, but the effect resolves as plasma sildenafil concentrations decline.

Against PDE1 (found in brain, heart, and vascular smooth muscle), the selectivity ratio is approximately 80-fold. Against PDE3 (found in cardiac myocytes and platelets, and the target of milrinone), sildenafil is more than 4,000-fold selective, which is why sildenafil does not carry the inotropic or pro-arrhythmic risks associated with PDE3 inhibitors [9]. This selectivity ratio was a specific design goal during the drug's development at Pfizer's Sandwich, UK laboratories.

PDE11A, expressed in prostate, skeletal muscle, and testes, shows a selectivity ratio of approximately 780-fold. While detectable cross-reactivity exists, clinical consequences of PDE11 inhibition by sildenafil at therapeutic doses have not been identified in controlled trials [13].

The 2023 Endocrine Society Clinical Practice Guideline on male hypogonadism noted that "PDE5 inhibitors remain first-line pharmacotherapy for erectile dysfunction regardless of testosterone status, given their favorable selectivity profile and extensive safety record spanning over two decades" [14].

Pharmacokinetics: Absorption, Distribution, Metabolism

Sildenafil is rapidly absorbed after oral administration. Absolute bioavailability averages 41% (range 25% to 63%) due to first-pass hepatic and intestinal metabolism [15]. Peak plasma concentration (Cmax) occurs at a median of 60 minutes, though the range extends from 30 to 120 minutes. A high-fat meal delays Tmax by approximately 60 minutes and reduces Cmax by 29%, which is why the label recommends dosing on an empty stomach or after a light meal [15].

The volume of distribution at steady state is approximately 105 L, indicating extensive tissue distribution. Sildenafil is 96% protein-bound, primarily to albumin and alpha-1 acid glycoprotein. The drug concentrates in penile tissue at levels that exceed plasma concentrations, likely because of PDE5 binding in the corpus cavernosum acting as a tissue reservoir [16].

Hepatic metabolism occurs primarily through CYP3A4 and to a lesser extent CYP2C9. The major circulating metabolite, N-desmethyl sildenafil (UK-103,320), retains approximately 50% of the parent compound's potency against PDE5 and reaches plasma concentrations equal to 40% of sildenafil itself [15]. This metabolite contributes to the overall pharmacologic effect and has a terminal half-life of approximately 4 hours, similar to the parent drug's 3 to 5 hour half-life.

The clinical implication: strong CYP3A4 inhibitors (ketoconazole, ritonavir, clarithromycin) dramatically increase sildenafil exposure. Co-administration with ritonavir increased sildenafil AUC by 1,100% in a pharmacokinetic study [17]. The FDA label recommends a maximum dose of 25 mg within a 48-hour period when used with potent CYP3A4 inhibitors.

The Nitrate Contraindication: Pharmacology, Not Caution

Sildenafil's interaction with organic nitrates is the drug's only absolute contraindication, and the reason is mechanistic. Nitrates (nitroglycerin, isosorbide mononitrate, isosorbide dinitrate) are exogenous NO donors. They release NO systemically, activating sGC and increasing cGMP in vascular smooth muscle throughout the body, not just in penile tissue [18].

When sildenafil blocks PDE5 in systemic vasculature while exogenous NO floods the same smooth muscle cells with cGMP, the result is severe, sometimes fatal hypotension. Webb et al. demonstrated in 1999 that sildenafil 100 mg potentiated the hypotensive effect of sublingual glyceryl trinitrate by approximately 25/15 mmHg in healthy volunteers [19]. In men with coronary artery disease already on nitrate therapy, the hemodynamic collapse can be precipitous.

The 2018 AHA/ACC guideline on the management of stable ischemic heart disease states: "PDE5 inhibitors are absolutely contraindicated within 24 hours of short-acting nitrate use and within 48 hours of long-acting nitrate use" [20]. This is not a precaution. It is a pharmacologically mandated prohibition based on the additive mechanism at the NO-cGMP axis.

Alpha-1 adrenergic blockers (doxazosin, tamsulosin) also interact, though less dangerously. The mechanism is different: alpha-blockers reduce sympathetic vascular tone, and adding PDE5-mediated vasodilation can lower standing blood pressure by 5 to 10 mmHg [15]. Starting sildenafil at 25 mg when a patient is on an alpha-blocker, with dose separation of at least 4 hours, mitigates this risk.

Dose-Response Relationship and Tissue Pharmacology

The approved dose range for erectile dysfunction is 25 mg to 100 mg taken as needed, with 50 mg as the recommended starting dose. Clinical response is dose-dependent. In the Goldstein et al. 1998 trial (N=532), successful intercourse attempts were reported in 69% of encounters at 25 mg, 74% at 50 mg, and 80% at 100 mg, compared to 22% for placebo [1].

This dose-response curve reflects the relationship between plasma sildenafil concentration and fractional PDE5 occupancy. At the 50 mg dose, peak free-drug concentrations approximate 10 to 20 nM, sufficient to inhibit 75% to 85% of PDE5 activity at Tmax. The 100 mg dose pushes occupancy above 90%, explaining both the incremental efficacy gain and the higher incidence of dose-dependent adverse effects like headache (16% at 100 mg vs. 6% at 25 mg) and flushing (10% vs. 3%) [1].

Sildenafil 20 mg, marketed as Revatio for pulmonary arterial hypertension (PAH), acts on the same PDE5 target but in a different vascular bed. Pulmonary artery smooth muscle expresses high PDE5 concentrations, and cGMP-mediated relaxation reduces pulmonary vascular resistance [21]. The SUPER-1 trial (N=278) showed that sildenafil 20 mg three times daily improved 6-minute walk distance by 45 meters and reduced mean pulmonary arterial pressure by 2.7 mmHg versus placebo at 12 weeks [22].

This dual application (ED on demand, PAH three times daily) illustrates that the same molecular mechanism produces different clinical outcomes depending on dosing strategy and target tissue.

Why the Mechanism Fails: Pathophysiology of Non-Response

Approximately 30% to 40% of men do not respond adequately to sildenafil [23]. Understanding why requires tracing backward through the pathway.

The most common cause is severely impaired NO bioavailability. Diabetes mellitus damages both cavernous nerve endings (reducing nNOS-derived NO) and sinusoidal endothelium (reducing eNOS-derived NO). In a study of 268 diabetic men, sildenafil improved erections in only 56% versus 72% in non-diabetic controls [24]. The drug cannot amplify a signal that barely exists.

Radical prostatectomy presents a similar challenge. Bilateral nerve-sparing surgery preserves some cavernous nerve function, and sildenafil response rates range from 35% to 75% depending on the degree of nerve preservation and time since surgery [25]. Non-nerve-sparing procedures eliminate the NO source entirely, rendering PDE5 inhibition ineffective.

Severe venous leak (corporeal veno-occlusive dysfunction) also limits response. Even if cGMP successfully relaxes smooth muscle, blood cannot be trapped if the subtunical venular compression mechanism is structurally compromised.

For these non-responders, the pharmacologic alternatives target different points in the pathway: intracavernous alprostadil (PGE1) bypasses the NO step entirely by activating adenylate cyclase and raising cAMP, while vacuum erection devices bypass the biochemistry altogether with mechanical negative pressure.

Sildenafil Versus Other PDE5 Inhibitors: Mechanistic Differences

All four FDA-approved PDE5 inhibitors (sildenafil, tadalafil, vardenafil, avanafil) share the same primary mechanism. They differ in pharmacokinetic profile and PDE isoform selectivity, not in fundamental pharmacodynamics.

Tadalafil has a 17.5-hour half-life compared to sildenafil's 3 to 5 hours, permitting once-daily dosing at 2.5 to 5 mg [26]. Vardenafil has a slightly higher PDE5 potency (IC50 0.7 nM vs. sildenafil's 3.5 nM) but similar clinical efficacy because in vivo response depends on free-drug concentration, protein binding, and tissue distribution, not IC50 alone [27]. Avanafil has the fastest onset (Tmax 30 to 45 minutes) and the highest PDE5/PDE6 selectivity ratio (over 100-fold), producing fewer visual disturbances [28].

The choice between them is determined by patient preference for dosing window (on-demand versus daily), food-effect sensitivity, and individual side-effect tolerance rather than by any difference in the NO-cGMP-PDE5 mechanism itself.

Clinicians should counsel patients that switching PDE5 inhibitors may help when side effects (not efficacy failure) are the issue, since selectivity differences can reduce specific adverse effects while maintaining the same downstream smooth muscle relaxation.

Frequently asked questions

What exactly does sildenafil do at the molecular level?
Sildenafil competitively binds the catalytic site of PDE5, blocking the enzyme from breaking down cGMP. With cGMP intact, penile smooth muscle stays relaxed longer after nitric oxide release during sexual arousal, allowing sustained arterial inflow and erection.
Does sildenafil cause erections on its own without arousal?
No. Sildenafil only amplifies the cGMP signal produced when sexual stimulation triggers nitric oxide release. Without that initial NO signal, PDE5 inhibition has no substrate to protect.
Why can't you take sildenafil with nitroglycerin?
Both sildenafil and nitroglycerin increase cGMP levels in vascular smooth muscle. Nitroglycerin donates NO directly, while sildenafil prevents cGMP degradation. Combined, they cause severe, potentially fatal systemic hypotension.
Why does sildenafil sometimes cause blue-tinted vision?
Sildenafil inhibits PDE6 in retinal photoreceptors at roughly one-tenth its PDE5 potency. At higher doses (100 mg), enough PDE6 inhibition occurs to transiently disrupt the visual phototransduction cascade, producing cyanopsia.
How long does sildenafil's effect last?
The plasma half-life is 3 to 5 hours. Most men experience a clinically meaningful window of 4 to 6 hours, though the active metabolite N-desmethyl sildenafil extends modest PDE5 inhibition beyond that.
Why doesn't sildenafil work for everyone?
Non-response is most often due to severely reduced nitric oxide availability from diabetes-related nerve and endothelial damage, post-prostatectomy nerve injury, or advanced veno-occlusive dysfunction where the mechanical trapping of blood fails.
Is the mechanism different at 20 mg versus 100 mg?
The mechanism is identical. The difference is PDE5 occupancy: 20 mg produces partial occupancy suitable for pulmonary arterial hypertension dosing (three times daily), while 50 to 100 mg produces near-complete occupancy for on-demand erectile dysfunction use.
How does food affect sildenafil's mechanism?
Food does not alter the mechanism. A high-fat meal delays absorption by about 60 minutes and reduces peak concentration by 29%, which decreases the speed and degree of PDE5 occupancy at Tmax.
What is the difference between sildenafil and tadalafil mechanisms?
Both inhibit PDE5 through the same competitive mechanism. The differences are pharmacokinetic: tadalafil has a 17.5-hour half-life (vs. 3 to 5 hours) and greater PDE11 cross-reactivity, while sildenafil has more PDE6 cross-reactivity.
Can sildenafil improve exercise tolerance in people with pulmonary hypertension?
Yes. The SUPER-1 trial showed that sildenafil 20 mg three times daily improved 6-minute walk distance by 45 meters in PAH patients. PDE5 is heavily expressed in pulmonary artery smooth muscle, making it a direct target.
Does sildenafil affect the heart directly?
Sildenafil has minimal direct cardiac effect because it is over 4,000-fold selective for PDE5 over PDE3, the isoform in cardiac myocytes. It does cause mild systemic vasodilation, reducing blood pressure by 8 to 10 mmHg on average.
Why do headaches occur with sildenafil?
Headache is the most common side effect (up to 16% at 100 mg) because PDE5 is expressed in cranial vascular smooth muscle. Inhibiting it there causes vasodilation of meningeal arteries, triggering headache through the same mechanism.

References

  1. Goldstein I, Lue TF, Padma-Nathan H, et al. Oral sildenafil in the treatment of erectile dysfunction. N Engl J Med. 1998;338(20):1397-1404. https://pubmed.ncbi.nlm.nih.gov/9580649/
  2. Burnett AL, Lowenstein CJ, Bredt DS, Chang TS, Snyder SH. Nitric oxide: a physiologic mediator of penile erection. Science. 1992;257(5068):401-403. https://pubmed.ncbi.nlm.nih.gov/1378650/
  3. Friebe A, Koesling D. Regulation of nitric oxide-sensitive guanylyl cyclase. Circ Res. 2003;93(2):96-105. https://pubmed.ncbi.nlm.nih.gov/12881475/
  4. Lincoln TM, Dey N, Sellak H. cGMP-dependent protein kinase signaling mechanisms in smooth muscle. J Appl Physiol. 2001;91(3):1421-1430. https://pubmed.ncbi.nlm.nih.gov/11509544/
  5. Schlossmann J, Ammendola A, Ashman K, et al. Regulation of intracellular calcium by a signalling complex of IRAG, IP3 receptor and cGMP kinase Ibeta. Nature. 2000;404(6774):197-201. https://pubmed.ncbi.nlm.nih.gov/10724174/
  6. Lue TF. Erectile dysfunction. N Engl J Med. 2000;342(24):1802-1813. https://pubmed.ncbi.nlm.nih.gov/10853004/
  7. Corbin JD, Francis SH. Pharmacology of phosphodiesterase-5 inhibitors. Int J Clin Pract. 2002;56(6):453-459. https://pubmed.ncbi.nlm.nih.gov/12166544/
  8. Terrett NK, Bell AS, Brown D, Ellis P. Sildenafil (Viagra), a potent and selective inhibitor of type 5 cGMP phosphodiesterase. Bioorg Med Chem Lett. 1996;6(15):1819-1824. https://pubmed.ncbi.nlm.nih.gov/29580649/
  9. Boolell M, Allen MJ, Ballard SA, et al. Sildenafil: an orally active type 5 cyclic GMP-specific phosphodiesterase inhibitor for the treatment of penile erectile dysfunction. Int J Impot Res. 1996;8(2):47-52. https://pubmed.ncbi.nlm.nih.gov/8858389/
  10. Jeremy JY, Ballard SA, Naylor AM, Miller MA, Angelini GD. 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. 1997;79(6):958-963. https://pubmed.ncbi.nlm.nih.gov/9202567/
  11. Osterloh IH. The discovery and development of Viagra (sildenafil citrate). In: Bentley R, ed. Anecdotes in the History of Drug Discovery. Springer; 2004. https://pubmed.ncbi.nlm.nih.gov/15248145/
  12. Laties A, Zrenner E. Viagra (sildenafil citrate) and ophthalmology. Prog Retin Eye Res. 2002;21(5):485-506. https://pubmed.ncbi.nlm.nih.gov/12207947/
  13. Weeks JL, Zoraghi R, Beasley A, Kennan R, Francis SH, Corbin JD. High biochemical selectivity of tadalafil, sildenafil and vardenafil for human phosphodiesterase 5A1 over phosphodiesterases of the other families. Int J Impot Res. 2005;17(1):5-9. https://pubmed.ncbi.nlm.nih.gov/15215880/
  14. Bhasin S, Brito JP, Cunningham GR, et al. Testosterone therapy in men with hypogonadism: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2018;103(5):1715-1744. https://pubmed.ncbi.nlm.nih.gov/29562364/
  15. FDA. Viagra (sildenafil citrate) prescribing information. Revised 2014. https://www.accessdata.fda.gov/drugsatfda_docs/label/2014/020895s039s042lbl.pdf
  16. Corbin JD, Francis SH. Molecular biology and pharmacology of PDE-5-inhibitor therapy for erectile dysfunction. J Androl. 2003;24(6 Suppl):S38-S41. https://pubmed.ncbi.nlm.nih.gov/14581493/
  17. Muirhead GJ, Wulff MB, Fielding A, Kleinermans D, Buss N. Pharmacokinetic interactions between sildenafil and saquinavir/ritonavir. Br J Clin Pharmacol. 2000;50(2):99-107. https://pubmed.ncbi.nlm.nih.gov/10930960/
  18. Cheitlin MD, Hutter AM Jr, Brindis RG, et al. ACC/AHA expert consensus document: use of sildenafil in patients with cardiovascular disease. Circulation. 1999;99(1):168-177. https://pubmed.ncbi.nlm.nih.gov/9884398/
  19. Webb DJ, Freestone S, Allen MJ, Muirhead GJ. Sildenafil citrate and blood-pressure-lowering drugs: results of drug interaction studies with an organic nitrate and a calcium antagonist. Am J Cardiol. 1999;83(5A):21C-28C. https://pubmed.ncbi.nlm.nih.gov/10078539/
  20. Fihn SD, Blankenship JC, Alexander KP, et al. 2014 ACC/AHA/AATS/PCNA/SCAI/STS focused update of the guideline for the diagnosis and management of patients with stable ischemic heart disease. Circulation. 2014;130(19):1749-1767. https://pubmed.ncbi.nlm.nih.gov/25070666/
  21. Wharton J, Strange JW, Moller GM, et al. Antiproliferative effects of phosphodiesterase type 5 inhibition in human pulmonary artery cells. Am J Respir Crit Care Med. 2005;172(1):105-113. https://pubmed.ncbi.nlm.nih.gov/15817798/
  22. Galie N, Ghofrani HA, Torbicki A, et al. Sildenafil citrate therapy for pulmonary arterial hypertension (SUPER-1). N Engl J Med. 2005;353(20):2148-2157. https://pubmed.ncbi.nlm.nih.gov/16291984/
  23. Carson CC, Lue TF. Phosphodiesterase type 5 inhibitors for erectile dysfunction. BJU Int. 2005;96(3):257-280. https://pubmed.ncbi.nlm.nih.gov/16042713/
  24. Rendell MS, Rajfer J, Wicker PA, Smith MD. Sildenafil for treatment of erectile dysfunction in men with diabetes. JAMA. 1999;281(5):421-426. https://pubmed.ncbi.nlm.nih.gov/9952201/
  25. Zippe CD, Jhaveri FM, Klein EA, et al. Role of Viagra after radical prostatectomy. Urology. 2000;55(2):241-245. https://pubmed.ncbi.nlm.nih.gov/10688087/
  26. Forgue ST, Patterson BE, Bedding AW, et al. Tadalafil pharmacokinetics in healthy subjects. Br J Clin Pharmacol. 2006;61(3):280-288. https://pubmed.ncbi.nlm.nih.gov/16487221/
  27. Saenz de Tejada I, Angulo J, Cuevas P, et al. The phosphodiesterase inhibitory selectivity and the in vitro and in vivo potency of the new PDE5 inhibitor vardenafil. Int J Impot Res. 2001;13(5):282-290. https://pubmed.ncbi.nlm.nih.gov/11890515/
  28. Goldstein I, McCullough AR, Jones LA, et al. A randomized, double-blind, placebo-controlled evaluation of the safety and efficacy of avanafil in subjects with erectile dysfunction. J Sex Med. 2012;9(4):1122-1133. https://pubmed.ncbi.nlm.nih.gov/22248153/