The observations of the essentially neutral effects of tadalafil (vs placebo) on objective measures of endothelial function in the current study are somewhat at odds with data from previous clinical studies involving treatment with tadalafil. For example, Rosano et al (2005) studied alternate-day tadalafil (vs placebo) in men with increased cardiovascular risk and found improved endothelial function, as measured by flow-mediated dilatation (FMD) and endothelin-1 levels, in the tadalafil treatment group.
In another study, treatment with tadalafil 20 mg 3 times per week was associated with improved brachial artery FMD and increased progenitor cells in men with ED (Foresta et al, 2006). In general, PDE5 inhibitors are not effective in improving endothelial function in healthy individuals (Dishy et al, 2001). This observation is supported by a study conducted by Bocchio and colleagues (2008), in which changes in circulating angiogenic cells and improved brachial artery FMD in men receiving tadalafil 20 mg every other day (vs placebo) were associated with the presence of baseline endothelial dysfunction.
In the current study, we chose a range of putative cellular markers as exploratory end points in an effort to better understand the potential impact of tadalafil once daily on endothelial function. However, there is no consensus concerning which of these markers should be evaluated in such an analysis. In addition, we do not yet have reliable population data on normative values for each of these markers, the sensitivities and specificities of abnormal levels in detecting endothelial dysfunction, and the levels of changes in these parameters that might be clinically meaningful. In a recent article, for example, Costa and Virag (2009) observed that, “To date, many circulating biomarkers have been proposed for the evaluation of endothelial function; however, none have been considered the ideal or more specific, and most are unavailable for current practice.” Despite these limitations, we felt that it was important to contribute our results to this growing area of scientific investigation.
Given that eligible participants could not have had previous exposure to PDE5 inhibitors, a “ceiling” effect may have limited the degree to which once-daily tadalafil for 12 weeks could further improve objective measures of endothelial function compared with placebo. The fact that our study population involved men with ED naive to PDE5 inhibitors (and hence with mainly mild ED) helps to explain why tadalafil did not result in higher proportions of men with normal IIEF-EF domain scores at end point: 43.6% compared with 30.3% with placebo. In this context, Rosen and colleagues (2011) demonstrated that minimum clinically important differences in the IIEF-EF domain are directly related to baseline ED severity, with patients having mild ED exhibiting a minimum clinically important difference of 2; moderate ED, 5; and severe ED, 7. Hence, given the large proportion of our patient population with mild ED, it is not entirely surprising that tadalafil-associated improvements in normalization of erectile function according to the IIEF-EF domain would be modest.
Evidence that patients analyzed in the present study had mild overall vascular disease and experienced a ceiling effect with tadalafil included the fact that 122 participants (57.5%) in the ITT population had RHI data that were considered a priori to be either inconclusive for endothelial dysfunction (RHI-PAT index ≥1.67 and ≤2.07) or normal (>2.07), where a value of ≤1.67 has been associated with endothelial dysfunction in a population at risk for ischemic heart disease (Yinon et al, 2006). Although the proportion of participants with hypertension (30.7%) was not dissimilar in our study and previous trials, a smaller proportion of our patient sample had diabetes: approximately 12% in the total population compared with 14% to 21% in previous studies of tadalafil once daily and as needed (Carson et al, 2004; Porst et al, 2006; Rajfer et al, 2007; Donatucci et al, 2008). Corresponding data for CAD were 0.9% of participants in the present study compared with 5% to 14% of those in previous tadalafil trials.
Other baseline data in the current study support the explanation of a ceiling effect as having limited potential improvements in endothelial dysfunction with tadalafil (vs placebo). The mean participant age was 52 years, approximately 75% of study participants had mild (44.2% of the total ITT population) or moderate (32.1%) ED, and 29.3% of the entire study population had organic ED.
Most other parameters of endothelial dysfunction were also not profoundly elevated at baseline. The baseline RHI-PAT index value in the total ITT population of the present study was higher than the previously reported cutoff of ≤1.67 for endothelial dysfunction (mean ± SD, 1.89 ± 0.57 in all participants) and consistent with an inconclusive finding about endothelial dysfunction.
On the other hand, it is also possible that the daily dose of tadalafil evaluated in the present study (5 mg) was too low to expect consistent beneficial effects on biomarkers of endothelial dysfunction in a low-risk population when compared with previous studies (Foresta et al, 2006, 2009; Aversa et al 2007). In the future, it might be of interest to compare PAT findings with results from other forms of testing, including FMD and other, penile hemodynamic and anatomic measures, such as cavernous artery intima-media thickness (Caretta et al, 2009). FMD remains the first-choice noninvasive method to measure endothelial function and has the most data relating to CAD risk (Tamler and Bar-Chama, 2008). Considering that endothelial dysfunction of penile arteries is a different entity from systemic endothelial dysfunction as detected by other modalities (ie, veno-occlusive plethysmography and FMD), the utility of evaluating penile arterial dysfunction as an early marker of atherosclerotic disease seems to be more promising than fingertip PAT, especially in men with ED (Aversa, 2012).
Similarly, the baseline mean ± SD hsCRP level of 2.12 ± 3.44 mg/L (and median of 1.19 mg/L) in the present study is not consistent with elevated cardiovascular risk (Ridker et al, 2008; Genest et al, 2009). In a randomized, open-label crossover study of 20 individuals with ED in whom alternate-day tadalafil 20 mg treatment significantly reduced hsCRP, the baseline value was considerably higher (x̄ ± SD, 3.1 ± 0.4 mg/L; Aversa et al, 2007).
A 12-week randomized, double-blind, placebo-controlled study of 298 men with comorbid ED and diabetes showed that once-daily tadalafil 2.5- to 5-mg treatment did not significantly reduce hsCRP, intercellular adhesion molecule (ICAM-1), or vascular cell adhesion molecule (VCAM-1), but the baseline values were within normal limits (Hatzichristou et al, 2008). Two studies involving alternate-day treatment with tadalafil 20 mg for 4 weeks showed significant reductions in endothelin-1 on active therapy in study participants with baseline mean values of 3.3 ng/L in both studies (Rosano et al, 2005; Aversa et al, 2007).
Our post hoc analyses demonstrated that changes in levels of the cellular adhesion molecule soluble E-selectin correlated inversely with ED severity on the IIEF. To our knowledge, this is the first reported observation of such a relationship. The origins of circulating adhesion molecules are not entirely clear, but these cells may derive from shedding or proteolytic cleavage from endothelial cells, and thus reflect increased expression of membrane-bound adhesion molecules (an initiating event in atherogenesis). Men with ED have increased levels of adhesion molecules, including soluble P-selectin (Bocchio et al, 2004), which recruits leukocytes to the site of arterial injury, and hence is intimately involved in the inflammatory pathway. In the present study, tadalafil treatment significantly reduced soluble E-selectin levels only in participants with baseline soluble E-selectin values in the upper tertile (or above the median; data not shown). This finding suggests that in such study participants, soluble E-selectin may represent an early marker of endothelial repair during progression of the atherogenic process.
The number of EPCs is considered to be an indicator of the presence of vascular injury as well as an expression of reparative mechanisms and vascular protection, including angiogenesis and vasculogenesis (Rauscher et al, 2003; Foresta et al, 2010a). In previous studies by Foresta et al (2005a,b, 2006, 2007, 2009a,b, 2010a,b), patients with increasing degrees of endothelial dysfunction exhibited progressive decreases in numbers of cEPCs. However, age-specific normative values for cEPCs have yet to be established, and larger studies are needed to determine the clinical significance of cEPC numbers in the general population. In a previous study involving an Asian population, most of whom did not have cardiovascular risk factors, there was no significant correlation between numbers of cEPCs (and surface marker expression) and blood pressure, cholesterol or glucose levels, or a history of smoking (Chen et al, 2010). It is also important to note that not all studies of EPCs have focused on the same cell types, including progenitors expressing CD34, CD133, KDR, and/or VEGFR2. Taken together, these factors may have contributed to uncertainty in the evaluation of our results. In this context, Esposito et al (2009) stated that “…there is no clear consensus on which antigenic profile best identifies progenitor cells with the potential to repair the endothelium.” They concluded that “literature data evaluating the putative causative role of EPCs in ED are scanty and discordant.”
In the current study, there were no statistically significant differences between tadalafil and placebo in terms of mean changes in levels of cEPCs. However, a post hoc analysis showed a reduction in cEPCs with tadalafil compared with placebo in men with 2 or more cardiac risk factors. The results of cEPC analyses in the current study must be interpreted with great caution given the mentioned discussion on the limitations introduced by the relatively mild ED population in this study, the current understanding of how cEPCs should best be studied in endothelial function, and the relatively small number of participants in the subgroup analysis involving those with 2 or more risk factors. It is possible, for example, that decreases in cEPCs observed in our study may represent reductions in ligand-mediated activation of cEPCs, with stabilization of the cEPC pool and potential long-term beneficial outcomes, especially in study participants at high cardiovascular risk. Further, it is plausible that PDE5 inhibitors do not increase cEPCs and other salutary biomarkers to the same extent as other medications for vascular diseases; in one small study of patients with angiographically documented stable CAD, daily treatment with an HMG-CoA reductase inhibitor (statin) for 4 weeks increased numbers of cEPCs by a factor of 1.5 to 3 (Vasa et al, 2001). Work in one of our institutions (Aversa, 2012), which included both PAT and cEPC measures, demonstrated that low-dose once-daily tadalafil did not affect endothelial function in men with mild ED.