Study Type – Aetiology (case control) Level of Evidence 3b
What’s known on the subject? and What does the study add?
Very little is known about the aetiology of premature ejaculation. This analysis shows that many PE patients have a heightened penile sensitivity. This information could result in the design and development of new drugs.
To investigate the correlation between penile hypersensitivity and premature ejaculation (PE), as defined by the criteria identified by the International Society of Sexual Medicine (ISSM). Penile hypersensitivity as a cause of PE is based on historical clinical neurophysiological data and clinical efficacy of the topical desensitizing agent PSD502 in the treatment of PE. PSD502 is a eutectic-like mixture of two local anaesthetics, lidocaine and prilocaine, whose primary action is to reduce neuronal conduction in sensory afferents.
Historical neurophysiological data was reviewed, together with data from the recent PSD502 clinical trials, including the first published double-blind clinical trial data evaluating a topical desensitizing agent in a population of men with PE, as per the new ISSM definition. The clinical profile of PSD502, based on its local anaesthetic properties, is used as a surrogate index of the role of sensory afferents in the ejaculatory reflex.
The published data does not support unequivocally penile hypersensitivity as the cause of PE. Interpretation of the data is hampered by the variability of the populations described as having PE across studies. Data from the PSD502 clinical trials clearly shows that PSD502 increases ejaculatory latency, and improves control and sexual satisfaction when applied topically to men with PE 5 min before intercourse, enabling subjects to delay ejaculation up to six times longer than those who used a placebo.
The clinical profile of PSD502 lends credibility to the penile hypersensitivity hypothesis for PE. The predominant action of local anaesthetics is to reduce neuronal firing in sensory afferents; the clinical profile of PSD502, which shows improvement of ejaculatory function in the absence of a generalized reduction in penile sensitivity, can most readily be explained based on an underlying hypersensitivity in patients with PE.
eutectic mixture of lidocaine and prilocaine (cream).
Premature ejaculation (PE) is the most prevalent male sexual problem, generally considered to affect up to 30% of the world’s male population, although some studies claim it to be as high as 75%.
The determination of the prevalence of PE was hindered until the recent introduction of the International Society of Sexual Medicine’s (ISSM) evidence-based definition: ‘Premature ejaculation is a male dysfunction characterized by ejaculation which always or nearly always occurs prior to or within about one minute of vaginal penetration; an inability to delay ejaculation on all or nearly all vaginal penetrations; and, negative personal consequences, such as distress, bother, frustration and/or the avoidance of sexual intimacy’.
While there has been an increase in men diagnosed and self-diagnosing  with PE over the last decade, there remains a dearth of clear, evidence-based research as to its aetiology. Penile hypersensitivity has been implicated as one of various causative factors, generating its own targeted research. However, the link between penile hypersensitivity with other potential causes has been conflicting and lacking in definitive conclusions. One potential reason for the uncertainty is that most of the studies predated the new ISSM definition of PE and, as such, the definitions of ‘normal’ and PE are somewhat arbitrary and variable across study centres. This review focuses on the correlation between penile hypersensitivity and PE, as defined by the criteria identified by the ISSM definition of PE.
PHYSIOLOGY OF EJACULATION
Ejaculation consists of emission and expulsion phases; two distinct but coordinated sequential neurological reflexes stimulated by sensory input to the penis (Fig. 1). Pudendal sensory nerve fibres provide information from nerve endings in the glans penis to the sacral cord and sensory cerebral cortex. The ejaculatory reflex is modulated by the brain and also at the lumbosacral spinal cord level, and is not dependent upon central neural control (depending on the location of the lesion) patients can ejaculate with vibratory stimulation of the penis (reviewed in ).
AETIOLOGY OF PE
No organic disease causes PE, although there is evidence that prostatitis and hyperthyroidism play a role in the aetiology of PE in some men [5,6].
There are many overlapping theories on the aetiology of PE. One contemporary view is that primary PE (being present since the onset of sexual activity) is part of the normal biological variability of ejaculatory latency time (ELT) in men, with a possible familial genetic vulnerability [7,8].
The large natural variation in ELT in men might be due to differences in penile sensitivity, with hypersensitivity causing, or contributing to, the very short latency of men with PE. If this is the case, drugs that selectively produce some degree of penile desensitization, or which act within the afferent-efferent reflex arc, should benefit men with PE.
The measurement of intravaginal ELT (IELT) in the male population shows a skewed distribution, and the question of where to apply a particular threshold that defines a particular IELT as being premature or not has been largely academic, as previous definitions of PE did not include such a threshold point. As a result, the population of men with PE in many clinical trials and neurobiological evaluations [9–11] have been largely self-diagnosed, are somewhat heterogeneous and, in many cases, poorly defined, making comparative evaluations difficult.
This paper re-evaluates the theory of penile hypersensitivity as a possible aetiology of PE in the light of its contemporary definition, which includes an IELT of ‘≈1 min’ as one of the features defining PE.
MOTOR PATHWAY IMPAIRMENT
The bulbocavernosus reflex is a contractile response of the bulbocavernosus and ischiocavernosus muscles (internal and external anal sphincter) when the glans penis is compressed. This reflex can also be tested electrophysiologically by stimulating the penis and recording evoked potentials (the electrical potential produced as a result of the external stimulus) from the anal sphincter . In healthy males, the evoked potential appears simultaneously in the bulbocavernosus muscle and in the anal sphincter, with an average latency of 45 ms. In men with PE (undefined), the average delay was only 15 ms between the response in the bulbocavernosus muscle and the anal sphincter .
Based on this evidence, Vignoli  suggested that in men with PE the shortening of the latency in the bulbocavernosus muscle could be due to a shunting in the reflex arc, possibly mediated by facilitatory impulses from supraspinal centres. However, it is not clear whether this physiological parameter represents a consequence or cause of PE, and the value of the study, in the absence of information defining the patient population, is limited.
A similar, larger study (85 men with undefined PE) reported no differences in the latencies of the evoked responses, although they did show there was significantly higher amplitudes of evoked responses in men with PE than in controls . The authors suggest that men with PE have a reflex hyperexcitability, or an impaired ‘modulation’ of the motor neurones of the pudendal nucleus by the regulating upper centres. PE was not clearly defined in that study and there was no mention of IELT having been measured.
Genital somatosensory evoked potentials (SEPs) are cerebral cortex recordings that reflect cortical representation from genital sensory stimuli.
(a) Amplitude of SEPs
The amplitudes of the SEPs in the cortical area were significantly higher in men with (undefined) PE than in ‘normal’ controls [14,15], suggesting a greater cortical representation of the sensory stimuli from the genital area (penile shaft and glans).
The authors postulate that men with PE have a greater cortical representation of the sensory stimuli from the glans penis, indicating that excessive excitation from the glans penis to the ejaculation centres has a role in rapid, uncontrolled ejaculation [14,15]. Perretti et al.  reported the opposite when using SEPs of the dorsal nerve, which were significantly less in men with PE (n= 14, mean IELT <3 min, range 0–2.7) than in controls .
(b) Latency of SEP
The mean latency of both the dorsal nerve and the glans penis SEP was shorter in the PE group (mean IELT 1.2 min, range 0.5–3.0) than in the control group, although both were within normal limits . In addition, in normal subjects the mean latency of the SEP from the glans was prolonged by 0.99 ms, compared with the dorsal nerve; postulated by the authors to be related to the longer distance from the glans penis compared with the penile shaft, as well as the different distribution of nerve fibres. In men with PE this was reversed, with the mean latency from the glans penis being shorter by 4.30 ms than that in the dorsal nerve .
The authors suggested a peripheral cause for primary PE; their hypothesis included a thickening of the glans penis nerves, increased myelinization, and an increased ratio of the thicker and faster A nerve fibres compared with the C nerve fibres. Alternatively, sensory stimuli in men with primary PE were conducted by a different pathway from the dorsal nerve. Neither hypothesis has been proven to date.
In another study, there was no significant difference in the mean latency values of the dorsal nerve SEPs between men with PE (n= 14, mean IELT of <3 min, range 0–2.7) and controls. Only three patients underwent glans penis stimulation but, in contrast with the earlier study , the glans penis SEP latency in all cases was longer than that from the penile shaft .
Taken together, individual differences in the neurophysiological properties of penile receptors and effectors may contribute to the speed of the ejaculatory reflex.
(c) Vibratory thresholds
The sensory perception threshold of the penis can be evaluated and measured using penile biothesiometry, in which a fixed frequency variable amplitude vibration is used. This test evaluates the afferent somatic dorsal pathway and is used to quantify penile sensory levels in men with PE.
Rowland et al.  tested empirically whether penile sensitivity to physical stimulation could distinguish premature ejaculators from men with normal sexual function, by measuring penile vibrotactile stimulation in four small groups of men: primary PE (IELT 0–2 min); secondary PE and erectile dysfunction (ED; IELT 0.5–2 min); ED alone (IELT 5–7 min); and controls with normal sexual function (IELT 7–10 min). A biothesiometer, modified with variable amplitude, was used to determine the sensory thresholds at various locations on the flaccid penis (without retraction of the foreskin).
Penile sensory thresholds varied in the four groups, with the highest threshold in the control group followed by the PE, PE + ED, and ED groups, in descending order. However, the difference between the control and PE group (as defined in this small study) was not significant and the authors concluded that the subjective threshold to vibrotactile stimulation of the penis in men with PE did not differ from that of sexually functional men . However, interpretation of these results in light of the contemporary ISSM definition of PE is somewhat different, as the PE group was small and included men with an IELT of >1 min and was estimated rather than measured with a stopwatch.
This study formed a strong correlation between self-reported IELT and penile sensory thresholds, with low thresholds related to short latencies; this is independent of the definition of PE used. This study does provide evidence that ELT is related to penile sensitivity. Ejaculation occurs during penile erection and in one study  was used as a surrogate of penile sensitivity in men with PE in the erect state . A precision digital vibrometer evaluated the penile sensitivity thresholds of 18 men with primary PE (defined as an IELT of <1 min) and in 15 controls with normal sexual function, at a range of locations on the penis when flaccid, and also with an ‘artificial’ erection after an injection of prostaglandin E1 (10 µg).
There was no significant difference between the two groups in vibratory threshold at the glans penis, dorsum of the penile shaft or frenulum of the penis, in either the flaccid or erect state. Also, in contrast to other studies, there was no difference in penile vibratory thresholds between the flaccid and erect states between the PE and control groups. The authors acknowledge that these findings can be attributed to the few subjects and variations caused by age differences . In addition, the authors indicate that penile sensitivity does not necessarily represent the sensory threshold that ultimately induces ejaculation . Little is known about the intensity of stimulus required to induce ejaculation, or the modulating influence from the cerebral cortex on the ejaculatory reflex.
A different approach was used in one of many studies by Xin et al. : a biothesiometer was placed on the index finger, right and left lateral aspects of the penile shaft, glans penis and mid scrotum of 120 men with primary PE (with an men IELT of 1.1 min, range 0–3) and 66 controls with normal sexual function . The patients informed the examiner of the first sensation of vibration as the amplitude of vibration was slowly increased, or the disappearance of the sensation as the amplitude was decreased.
There was a statistically significant decrease in vibratory threshold on the glans penis and penile shaft in men with PE (P < 0.001), but not at the index finger or scrotum . In the normal control group, but not the PE group, the vibratory threshold at the glans penis and penile shaft showed a significant age-dependent increase . This study may have some inherent operational bias as the vibrator was hand-held (allowing possible variation in the pressure applied) and the stimuli were presented in a non-random order .
(d) Correlation between penile sensitivity and ELT in normal men
In men without PE, i.e. presumably ‘normal’ men, no correlation has been found between penile sensitivity at any of the penile surface areas and IELTs, masturbatory or laboratory ELTs , indicating that ELT variability in normal men cannot be explained by differences in penile sensitivity. However, this does not exclude the possibility that penile sensitivity may have a role in a lowered ejaculatory threshold in men with PE.
Various electrophysiology studies have shown that men with PE have a heightened sensory response to penile stimulation, with a vibratory threshold significantly lower than that of normal men and abnormal ejaculatory reflex pathways, including shorter bulbocavernosal latency time and higher bulbocavernosal evoked potentials. Some researchers have shown no significant difference between these variables in men with PE and normal controls. In many studies, variables such as age and circumcision complicate the interpretation. Correlation between IELT and vibratory thresholds are more useful as they are independent of self-diagnosis or the application of arbitrary IELT thresholds. Only one study to date has shown a significant correlation between IELT (albeit estimated) and penile sensory threshold .
The major issue of these studies is that most were performed before a uniform and contemporary (ISSM) PE definition was established.
DATA FROM CLINICAL TRIALS EXPLORING DRUG TREATMENTS
Treatment with daily fluoxetine or clomipramine in two independent studies showed significantly increased penile sensory thresholds in men with PE [9,10]. In both cases, there was no change in either the amplitude or latencies of sacral evoked response and cortical SEPs. This is in contrast to the decrease in amplitude of SEPs after the use of secret severance (SS) cream . Increased penile sensory thresholds may be the mechanism of action for selective serotonin reuptake inhibitors in increasing IELT in men with PE.
SS cream from Korea consists of extracts of nine natural products of herbal and animal origin, several of which are reported to have a local desensitizing effect . Evaluation of penile biothesiometry and SEPs in men with PE showed that SS cream increased the penile sensory perception threshold, prolonged the latencies of SEPs and decreased amplitudes of SEPs [22,24], suggesting that SS cream has a local desensitizing effect on penile hypersensitivity and/or hyperexcitability in men with PE.
The same group investigated the effect of various doses of SS cream and placebo (double-blind) on the vibratory threshold of the glans penis in 53 men with primary PE (estimated IELT of <3 min, mean IELT 1.45 ± 0.57 min) 1 h after application of the cream . After 153 tests with various doses of SS cream, there was a significant dose-dependent increase in the mean vibratory threshold of the glans penis from baseline. The authors concluded that SS cream increases the penile sensory threshold .
Several clinical studies on the eutectic mixture of lidocaine and prilocaine (EMLA) cream have shown that the local anaesthetic-containing medication can prolong ELT . The main concern from these relatively small population studies is whether this can be achieved in the absence of generalized penile ‘numbing’ or hypoesthesia.
Late in 2008, phase III data emerged on the use of a new aerosolized topical form of lidocaine-prilocaine, PSD502 . The study was conducted on 300 men with a mean age of 35 years who were clinically diagnosed with PE across 32 investigational centres in the UK, Czech Republic, Hungary and Poland.
Data from the study showed that PSD502 produced a highly clinically and statistically significant increase from baseline in all three co-primary study endpoints, and also in all secondary endpoints. The IELT for PSD502 treatment was 4 min, compared with 1 min for placebo (P < 0.001). There was a 7-point difference between PSD502 and placebo in Ejaculatory Control (P < 0.001), and a 6-point difference between PSD502 and placebo in ‘sexual satisfaction’ (P < 0.001), where a 2-point difference in a 16-point range is considered clinically significant. There was a 3-point difference between PSD502 and placebo in the Index of Premature Ejaculation domain for distress (P < 0.001).
The number of patients in the PSD502 group who rated the quality of their orgasm as ‘good’ or ‘very good’ increased from 20% at baseline to 62% after treatment. In comparison, the number of placebo-treated patients with this rating decreased from 21% to 19%.
PSD502 was well tolerated and there were no serious adverse effects, with only 2.6% of patients reporting treatment-related adverse effects in the PSD502 group, compared with 1% in the placebo group. Of these localized adverse effects, one patient who received PSD502 (0.5%) reported temporary numbness of the penis, which was described as mild. The incidence of systemic side-effects such as headache was <1%.
An additional phase Ibis/III clinical trial has recently been completed and preliminary data is available (Carson/Wyllie Sexual Medicine Society of North America abstract, 2009). In general, a qualitatively and quantitatively similar picture emerged; there were clinically and statistically significant changes in the IELT and all domains. Adverse side-effects during this study indicated that only four patients (2.4%) reported penile hypoesthesia.
The evidence that penile hypersensitivity is the underlying cause for PE has been hindered in part by the methodology and definition of PE used in the analysis of data. All studies examining the link were completed before the new ISSM PE definition. The recent clinical profile of PSD502 does provide considerable additional evidence to support the hypersensitivity concept.
The clinical profile of PSD502, i.e. prolongation of ELT without penile desensitization (in all but one patient), arises from the local anaesthetic actions of (lidocaine and prilocaine). Local anaesthetics act on sodium channels within sensory nerve endings  to alter axonal conduction to spinal and supra-spinal centres (Fig. 1).
The nerve receptor density of the glans penis is higher than in any other region of the body and has a lower tactile threshold , therefore accounting for its higher sensitivity compared with other parts of the body . Topical treatments providing local anaesthesia reduce the sensitivity of the glans penis and delay ejaculation. In other dysfunctions, there are clinical parallels as to why this can be achieved without a generalized ‘numbing’ or hypoaesthesia. In hypertension, most effective drugs only restore the status quo, i.e. the patient becomes normotensive without becoming hypotensive . It is assumed therefore that PSD502 acts to address the imbalance in PE (i.e. hypersensitivity without producing the clinically effective dose range of hypoaesthesia). However, it is possible that the selectivity of the effect is a result of the localized administration of the spray.
Although animal research may give some clues, the precise mechanism of the ejaculatory reflex in man is unclear and not fully elucidated. Undoubtedly, various peripheral afferent-efferent reflexes and spinal and supra-spinal pathways (Fig. 1) are involved. The underlying pathophysiology of PE is likewise unclear. The clinical benefit of ‘off-label’ selective serotonin reuptake inhibitors and dapoxetine is likely to arise from the ability of these drugs to rectify an imbalance in serotonin within the brain. The action of desensitizing creams, e.g. EMLA, on peripheral sensory afferent pathways, alters signal transmission to the brain; a selective sensory ‘dampening’ action at this level would also be effective independent of the aetiology of PE .
Overall, although circumstantial, the clinical profile of PSD502 supports penile hypersensitivity as being a major contributor in the manifestation of PE. PSD502 acts by restoring penile sensitivity to a normal level. If there were no underlying hypersensitivity, a generalized hypoaesthesia or ‘glandular numbing’ would occur.
Published data provides some degree of circumstantial evidence that penile hypersensitivity is an underlying cause of PE. Additional support for the peripheral hypersensitivity ‘concept’ comes from the clinical profile of PSD502, where increased ELT is achieved in the absence of generalized penile desensitization. A direct association remains to be shown. A major problem with previous investigations is that no uniformly accepted definition of PE was used. Prospective studies using the ISSM definition to stratify patients and ‘normals’ will be required to resolve completely the extent to which penile hypersensitivity contributes to PE.
CONFLICT OF INTEREST
Michael G. Wyllie is a Director and Shareholder in Plethora Solutions Ltd. and a Consultant to Shionogi Pharma Inc. Wayne J.G. Hellstrom is a Consultant and Investigator for Plethora Solutions Ltd. and a Consultant to Shionogi Pharma Inc. Source of funding: Plethora Solutions Ltd. and Shionogi Pharma Inc.