Long‐acting glucagon‐like peptide 1 receptor agonists boost erectile function in men with type 2 diabetes mellitus complaining of erectile dysfunction: A retrospective cohort study

The pharmacological management of erectile dysfunction (ED) in type 2 diabetes (T2D) is challenging as ED has a multifactorial etiology. The therapeutic potential of certain antihyperglycemic medications, such as glucagon‐like peptide 1 receptor agonists (GLP‐1RAs), has yet to be entirely studied in this setting.

Discussion: GLP-1RAs plus metformin over metformin alone improved ED regardless of different background characteristics of patients and partially irrespective of therapeutic targets achieved after 12 months of treatment.GLP-1RAs may have induced positive vasculature effects, resulting in improved erectile function in T2D.

Conclusion:
Due to the retrospective nature of the study, a potential cause-effect relationship between the use of GLP-1RAs plus metformin over metformin alone in improving ED cannot be verified and confirmed.Randomized clinical trials are needed to provide evidence supporting the use of GLP-1RAs for treating ED in T2D.

K E Y W O R D S
erectile dysfunction, glucagon-like peptide 1 receptor agonists, international index of erectile function, retrospective cohort study, type 2 diabetes

INTRODUCTION
Men complaining of erectile dysfunction (ED) usually present with hypogonadism and concomitant chronic cardiometabolic diseases. 1,2 the other hand, ED is generally considered strong predictor of poor health, worse quality of life, and increased mortality. 3pe 2 diabetes mellitus (T2D) is one of the major public health diseases in the Western world and is frequently associated with ED and male hypogonadism.Indeed, more than 50% of men with T2D complain of ED, 4 and more than 30% of T2D men are affected by secondary hypogonadism, with around 20% of them having overt hypogonadism (serum testosterone levels < 8 nmol/L). 5Hyperglycemia per se may facilitate the onset of ED, given that it is associated with oxidative stress and the overproduction of reactive oxygen species leading to a cascade of events such as blunted synthesis of nitric oxide and increasing prothrombotic factors.All these microvascular alterations play an essential role in endothelium dysfunction that, in turn, may precipitate ED. 6 Symptomatic treatment of ED with phosphodiesterase 5 inhibitors (PDE5i) often fails in T2D due to concomitant male hypogonadism. 7wever, testosterone supplementation in T2D with low testosterone levels is ineffective, provided its levels are in the typical range of overt hypogonadism. 8centered-patient approach including a comprehensive management of chronic diseases can improve ED in those patients. 9,10raglutide, a once-daily long-acting glucagon-like peptide 1receptor agonist (GLP-1RA), significantly improves metabolic control and erectile function in men with T2D and overt hypogonadism who underwent testosterone replacement and reached eugonadal status. 11laglutide, a once-weekly long-acting GLP-1RA, has recently been shown to improve the erectile function in T2D men and high background cardiovascular risk, as suggested by a subgroup analysis. 12sides this evidence, to our knowledge, no specific clinical trials have been conducted to assess the efficacy or effectiveness of GLP-1RAs, as an add-on to metformin, in improving the erectile function in men with T2D irrespective of the functional testicular status.

Study design, approval, and institution
This is a 1-year retrospective cohort study evaluating the effect of GLP-1RAs (Liraglutide and Dulaglutide) in add-on to metformin versus metformin alone on metabolic parameters and erectile function in men with T2D complaining of ED.All the parameters were collected at baseline (T0) and after 12 months of consecutive and unmodified treatment (T12).
All the procedures were completed at the Outpatients Clinic of Endocrinology and Metabolic Disease of the Conversano Hospital.
Metformin was started at 500 mg/day and progressively titrated up to the maximum tolerated daily dose according to baseline patients' characteristics and concomitant chronic diseases or confirmed if ongoing in case of baseline (T0) HbA1c ≤ 7%.A GLP-1RA (Liraglutide or Dulaglutide according to patient's preference) in add-on to metformin at the maximum tolerated dose was prescribed if baseline (T0) HbA1c levels were greater than 7% in patients with established cardiovascular disease (CVD) or having two or more cardiovascular risk factors, including overweight/obesity syndrome, dyslipidemia, and arterial hypertension.
Liraglutide was started at 0.6 mg daily and titrated up to 1.We classified ED severity according to the IIEF5 score.Signs and symptoms of urogenital autonomic dysfunction and peripheral neuropathy were also checked to rule out purely neuropathic ED.None complained of or had suspicious signs and symptoms or neuropathy.

Study participants
One hundred fifty-four patients with T2D complaining of ED were admitted to ambulatory care in that period.Among them, 20 patients were excluded as they denied informed consent and 26 due to the lack of interest data at T0, T12, or both.

Study outcomes
The primary study outcome was to assess changes in glucose control, body weight, serum testosterone concentration (total and free), and IIEF5 score by comparing T0 to T12.
Moreover, we examined predicting factors, including antihyperglycemic treatments, explaining gain in the penile function after 1 year of treatment.To rule out possible confounding factors, such as the levels of HbA1c, the evolution of diabetes, and the presence of diabetes-related comorbidities and complications, we carried out first a univariate analysis and then a multivariate analysis of the covariance model for testing the independent effect of parameters on the IIEF5 score at T12.

RESULTS
One hundred eight consecutive outpatients with T2D and ED were retrospectively analyzed.Among them, 53 (49%) were eugonadal, and 55 (51%) had hypogonadism.Forty-five patients were treated with metformin at a mean daily dose of 2,000 mg/day, 25 (56%) were eugonadal, and 20 (44%) had hypogonadism.Sixty-three men were on GLP-1RAs in add-on to metformin 2,000 mg: 34 (54%) on Liraglutide The background gonadal status did not affect the distribution of ED severity among the study population.The exhaustive baseline (T0) characteristics of the examined population split by treatments are shown in Tables 1 and 2.
The mean age was around 60 years, and patients received an established diagnosis of T2D around 6 years before.Concomitant comorbidities were arterial hypertension, which was diagnosed in 72 (67%) patients, carotid stenosis in 57 (52.7%), and CVD in 37 (34.2%).
None had signs or symptoms of peripheral neuropathy.There were no differences between the two groups.
Weight loss was clinically irrelevant, and, most importantly, a considerable decline in serum total but not FT was also observed (Table 2).
Focusing on ED, the frequency of men who gained points in the IIEF5 score (T12 IIEF5 − T0 IIEF5 > 0) was also calculated.Sixty out of 63 patients (95.2%) on GLP-1RAs plus metformin had an improvement in the IIEF5 score that was significantly greater than that observed in the other group (23/45, 51.1%), and the difference was statistically significant (p < 0.0001).
Background variables of study participants, such as the BMI, WC, FPG, gonadotropins, and TT and FT, were found to influence the clinical outcome (Table 3).However, the independent effect of the combination of GLP-1RAs plus metformin over metformin alone resulted in a significant weight loss (−5.82 ± 0.69 kg, p < 0.0001), reduction in WC (−4.99 ± 0.6 cm, p < 0.0001), HbA1c (−0.56% ± 0.13%, p < 0.0001), and FPG (−25.54 ± 3.09 mg/dL, p < 0.0001).Moreover, the former treatment was associated with a significant increase in total (+41.41± 6.11 ng/dL, p < 0.0001) and free (+0.44 ± 0.09 The baseline characteristics of the study population were subdivided into two categories according to antihyperglycemic medications.In a secondary analysis, the weight of single variables in predicting the gain of the erectile function was calculated (Table 4).Age, diabetes evolution, arterial hypertension, chronic renal disease and established CVD, background glucose control, body weight, WC, and gonadal status did not have a role in predicting gains in IIEF5 scores.On the contrary, a background history of carotid stenosis was positively associated with a gain of IIEF5 score, while baseline BMI was negatively associated with it.

Parameters
After excluding from the model age, change in BMI, baseline TT, and FT, which did not significantly affect the IIEF5 score after 12 months of treatment, the parameters associated with IIEF5 gain regardless of baseline treatment were a higher baseline IIEF5 score (estimated coefficient: 0.16 ± 0.08, p = 0.045), evidence of carotid stenosis (0.50 ± 0.24, p = 0.045), weight loss from baseline (−0.08 ± 0.03, p = .013),and baseline FSH levels (Table 5).The leading determinant of the final (T12) IIEF5 score was a 1-year treatment with GLP-1RAs plus metformin over metformin alone (2.74 ± 0.53, p < 0.0001).

DISCUSSION
ED is a multifactorial complication of T2D with an estimated prevalence of more than 60% of T2D men, which is more than triplicated TA B L E 3 Independent effect of antihyperglycemic treatment and background variables (T0) on anthropometric, glycemic, and hormonal changes from T0 to T12. than in healthy controls. 4The frequency of ED increases along with aging and burdens of comorbidities and related chronic complications. 2 The actual prevalence of ED is probably underestimated in everyday practice due to, at least in part, patient-related complaints in raising concerns or underrating the clinical relevance of ED during ambulatory visits.The results of an Italian study suggested that 1 in 2 men discuss ED with their physician, most delaying the first consultation after months of symptom persistence.wall are the main contributors to atherosclerosis, including ED, which could be a warning sign of hidden CVD, especially in young patients. 19crovascular complications, including neuropathy and microangiopathy, are other common contributors to ED in T2D men who express low intra-cavernous nitroxide and vascular endothelial growth factor levels. 20[22] Besides promptly recognizing ED, the adequate treatment is necessary to minimize cardiovascular and microvascular risks and possibly restore or improve penile function.The GLP-1RAs have been demonstrated to improve glucose control, promote weight loss, reduce background systemic inflammation and oxidative stress, improve insulin sensibility, optimize lipid control, and minimize the risk of diabetesrelated complications. 23Although GLP-1RAs could play an important role in the management of ED, only a few focused studies have been published so far to our knowledge. 11,12is retrospective study explored the role of GLP-1RAs (Dulaglutide and Liraglutide) in add-on to metformin, compared to metformin alone in T2D men with ED irrespective of hypogonadism in a realworld scenario.As expected, GLP-1RAs plus metformin compared to metformin alone significantly improved glucose control and provided more evident weight loss despite relevant differences in baseline char-acteristics of patients prescribed with the two different therapeutic strategies.The serum TT levels were improved significantly by around 11% after 12 months of treatment with GLP-1RAs plus metformin compared to baseline, while patients treated with metformin alone had normal levels of FT between the two points of the examination (T0 and T12), as indicated in Table 2.The IIEF5 score was higher at T12 than T0

Effect of 1-unit increase in baseline parameters
in almost all patients (95.5%) on GLP-1RAs plus metformin, with only 3 of 63 men who had exhibited stable or worsening ED.In the GLP-1RAs group, the frequency of men meeting the criteria for mild-to-moderate ED (IIEF5 score < 17) at T0 (82.5%) dropped to 2% (1 in 63) at T12.In the metformin group, only 23 of 45 (51.1%) men at T12 scored greater than T0.The frequency of men with IIEF5 score < 17 at T0 (66.7%) dropped to 48.9% (22 in 45) at T12.
Testosterone plays a crucial role in regulating the erectile function by acting in several districts with direct and indirect effects.
Testosterone may cross the blood-brain barrier and enter the central nervous system, but it may also be synthesized directly in the brain and undergoes a further transformation into dihydrotestosterone and estradiol. 24A strong relation between testosterone and sex drive is well known, as male hypogonadism results in a low or absent sex drive, while testosterone replacement restores it.Normal testosterone levels are essential to improve the penile function in animals and men.
Testosterone exerts genomic-and non-genomic effects involved in the relaxation of the corpus cavernosum, such as the activation of inward rectifying potassium channel and blockage of L-type calcium channel, both involved in the relaxation of smooth muscle cells. 24In addition, testosterone increases the gene expression regulating the vascular wall in the circulation of the corpus cavernosum, such as nitroxide synthase and cyclic guanidyl monophosphate, and reduces the expression of type 5 phosphodiesterase. 25Penile response to type 5 phosphodiesterase inhibitors is blunted in hypogonadal men, but it improves with testosterone replacement soon after achieving serum TT values within the normal ranges. 26 but not in FT at T12 (Table 2).In addition, the IIEF5 score was also higher in the group of GLP-1RAs at T12 compared to T0.This possible improvement of the self-reported erectile function was not observed in those who had assumed metformin alone, and it was the same regardless of the gonadal status at T0 and T12.In fact, the IIEF5 score increased with the same magnitude in both hypogonadal and eugonadal patients.To confirm this observation, neither background TT and FT nor improvement in TT from baseline to T12 have been found to be related to IIEF5 gain in patients who had assumed GLP1-RAs.Overall, GPL-1RAs added to metformin over metformin alone contributed most to the self-reported improvement of the erectile function.
The effect of metformin on ED is controversial.On one hand, insulin resistance negatively affects the erectile function in diabetic and nondiabetic obese patients.Nitroxide production is impaired in insulin resistance, and treatment with metformin restores endothelial nitroxide synthesis by improving insulin signaling.In one pre-clinical study carried out in male Sprague-Dawley rats, metformin improved the angiotensin II-induced ED by reestablishing normal cavernosal smooth muscle tone. 27Similar results were found in another study in a murine model. 28In one prospective, randomized, double-blind pilot study, patients were randomized to receive metformin (n = 17) or placebo (n = 13).After treatment with metformin, patients with ED showed a significant increase in the IIEF5 score and a substantial decrease in the HOMA index, both occurring at month 2 (IIEF5 score: 17.0 ± 6.0 vs.
14.3 ± 3.9, p = 0.01; HOMA-IR: 3.9 ± 1.6 vs. 5.5 ± 2.4, p = 0.01) to 4 of treatment (IIEF5 score: 19.8 ± 3.8 vs. 14.3 ± 3.9, p = 0.005; HOMA-IR: 4.5 ± 1.9 vs. 5.5 ± 2.4, p = 0.04), with no changes in these parameters in patients receiving placebo. 29 the other hand, there is evidence that T2D men treated with metformin display erectile complaints and a significant decline in serum TT and FT concentrations over time.In one Iraqi case-control study, metformin, compared to glycyclamide induced a significant decrease in serum total and FT concentration, sex drive, and ED. 30 In a Taiwanese retrospective study on a cohort of around 260,000 men with a new-onset T2D followed for more than 5 years, patients treated with metformin compared to those who had not displayed a higher risk of incident ED (hazard ratio: 1.35; CI 95% 1.26-1.44)and infertility (hazard ratio: 1.39; CI 95% 1.08-1.81). 31In insulin-resistant nondiabetic obese men with hypogonadism, the combination of metformin plus testosterone replacement treatment compared to testosterone and metformin administered separately provided a more significant increase in serum TT and FT.However, it did not further ameliorate the erectile function. 32In a sub-study of a randomized, double-blind, crossover, placebo-controlled trial, 24 men with impaired glucose tolerance or T2D diagnosed with hypogonadism (serum TT ≤10.4 nmol/L) were randomized to receive metformin 2 g/day plus clomiphene or placebo (12 weeks apart by 6 weeks of pharmacological wash out).The overall sexual function assessed by the IIEF15 and qADAM scores was not modified significantly. 33anks to the glycemic and pleiotropic effects of GLP-1RAs, this class of antihyperglycemic medication has gained success, as they positively affect the pathophysiology of CVDs from which ED is strictly correlated. 30,34In line with the evidence, carotid stenosis, a direct

LIMITATION AND STRENGTH
The retrospective nature of the study does not allow us to conclude that results explain a potential cause-effect relationship between the use of GLP-1RAs in add-on to metformin and amelioration in ED over metformin alone.Another limitation of this study is the lack of a third group of patients on GLP-1RAs only.The latter point was affected by the real-life setting of the study, as guidelines and local recommendations did not allow the prescription of a GLP-1RA alone (as an alternative to metformin).Therefore, in most cases, a GLP-1RA was added-on to metformin, and only a few patients with documented intolerance or contraindication to receiving the biguanide could have been considered ideal candidates for a GLP-1RA only. 36e IIEF5 score is a validated method to screen, diagnose, and assess the severity of ED; however, a direct instrumental assessment of the erectile function to confirm the results in terms of ED and potential improvements is lacking.Globally, the results of our study should be supposed from a reliable but self-reported method of the erectile function assessment.Lastly, all men had a mild or mild-to-moderate ED, and no cases of more severe ED were therefore diagnosed and consequently examined.Thus, the results of the present study should be addressed with caution to patients with moderate or severe ED.
The strengths are in the study design.Patients were managed using a standardized approach based on background glucose control and diabetes-related comorbidities.Moreover, as per habitual practice, the IIEF5 score, a validated tool for assessing ED, was systematically administered to men with T2D to screen ED and, in those diagnosed with ED, to estimate the severity and follow it up.Hence, the estimation of ED frequency among T2D outpatients is accurate.In addition, serum gonadotropins and Testosterone were measured according to standardized procedures, including validated assay methods.

For a
Type III F-test of the treatment predictor adjusting for the other seven predictors [baseline IIEF5 score, carotid stenosis (yes vs. no), change in body weight (T12-T0), baseline WC, change in FPG, baseline LH, and FSH] in a multivariate analysis of covariance model with a significance level of 0.05, assuming a conditional model with fixed predictors and an R-square of 0.55 in the full model, a sample size of 108 has powered higher 0.95 to detect an R-square difference of 0.12.Continuous variables were expressed as mean and standard deviation (SD) for normally distributed parameters.Median and interquartile range (IQR) were considered in the case of data with the skewed distribution.The Shapiro-Wilk statistic was used to test normality.Comparison between T0 and T12 in both groups was analyzed by a Student paired t-test, or nonparametric Wilcoxon signed-rank test.The comparison between treatments of fixed parameters (age and diabetes evolution) was carried out with the nonparametric Mann-Whitney test.The Chisquare test or Fisher's exact test was used to compare comorbidities between the two groups of treatment.The independent effect of antihyperglycemic treatment on parameters measured at T12 was tested by pre-post analysis of covariance after adjusting by the measurement of the same parameter at time T0.Univariate analyses of covariance were also used to test the independent effect of fixed and repeated parameters on the IIEF5 score at T12 after adjusting it for the IEFF-5 score at T0. Parameters with p-values < 0.25 at the univariate analysis were included in the multivariate analysis of the covariance model, and the stepwise selection was used to estimate the final model by the Akaike information criterion.All tests of statistical significance were two-tailed, and p-values ≤ 0.05 were considered statistically significant.Statistical analyses were performed using the SAS/STAT Statistics, Version 9.4 (2013), SAS Institute Inc., Cary, NC, USA.

The
IIEF5 score is a validated and simple-to-use tool to collect information about erectile performance over the last 6 months.Patients are briefly asked to respond to five simple questions (items) with five possible answers exploring the penile function throughout a sexual activity and sexual intercourse.It is widely used as a screening tool to diagnose ED and stratify its severity in men complaining of or at risk of ED.Endothelial dysfunction, oxidative stress, inflammation, immune response, and ultimately atherosclerotic degeneration of the vascular Moreover, testosterone enhances insulin sensitivity, promotes weight loss, and reduces visceral adipose tissue and systemic inflammation.Thanks to these mechanisms, testosterone replacement improves ED, especially in hypogonadal men.In our study, patients with T2D and ED who had assumed GLP-1RAs plus metformin for 12 months exhibited better the penile function than patients treated with metformin alone.Patients had different background characteristics regarding glucose control, body weight, and WC that were clinically consistent with the two therapeutic choices.As expected, GLP-1RAs provided more significant amelioration of glucometabolic control and weight loss, which are well-recognized factors associated with improvements in the erectile function.Our study found an inverse relationship between change in body weight and FPG from T0 to T12 and a boost in the final IIEF5 score, suggesting that better glucose control and weight loss contributed, at least in part, to the enhancement of erectile function.Serum TT and FT levels were higher at T12 than baseline (T0) in patients on GLP-1RAs plus metformin, whereas those on metformin alone exhibited a slight reduction in TT sign of atherosclerosis, predicted a positive response regarding erectile function (gain in IIEF5 score at T12 compared to T0).In a previously published retrospective study, adding Liraglutide to metformin and testosterone replacement in T2D men with inadequate glucose control, hypogonadism, and ED induced a relevant increase in serum testosterone concentrations and an overall improvement of the erectile function.11Interestingly, in an animal model of diabetes-induced ED (streptozotocin-induced diabetes), Liraglutide improved smooth muscle dysfunction in the corpus cavernosum, reduced oxidative stress and autophagy, and lastly, provided improvement in ED, which was independent of glycemic control.35In our study, most results are linked to the therapeutic association between GLP-1RAs and metformin over metformin alone, regardless of different background characteristics and partially irrespective of therapeutic targets achieved after 12 months of treatment.GLP-1RAs per se may induce additive macro-and microvascular effects, resulting in relavant gain in erectile function.
Liraglutide and Dulaglutide added to metformin in overweight/obese and T2D patients with inadequate glucose control and ED, regardless of their gonadal status, reduce symptoms and related complaints of ED after 12 months of treatment.The mechanisms coupled with this improvement could be partly attributable to glycemic and pleiotropic effects of GLP1-RAs, such as improvement in glucose control and weight loss.As mechanistic studies suggested a potential therapeutic effect of both metformin and GLP-1RAs, with the latter keeping a direct and positive impact on cavernosal microvasculature and smooth musculature dysfunction, specific prospective and randomized trials are needed to provide evidence supporting an ampler and safer use of GLP-1RAs in patients with ED.
*Thirty-four patients were on Liraglutide, and 29 were on Dulaglutide.ˆChi-square test.$ Fisher exact test.Changes in anthropometric, glycemic, and hormonal parameters between T0 and T12 by treatment.