Topical eutectic mixture for premature ejaculation (TEMPE): a novel aerosol-delivery form of lidocaine-prilocaine for treating premature ejaculation


Michael Wyllie, Urodoc Ltd, Maryland, Ridgeway Road, Herne, Kent, CT6 7LN, UK. e-mail:



To evaluate, in a phase II study, the efficacy and safety of a topical eutectic mixture for premature ejaculation (TEMPE), a metered-dose aerosol spray containing a eutectic mixture of lidocaine and prilocaine, as a treatment for PE.


Men with PE (Diagnostic and Statistical Manual-IV definition) aged 18–75 years were randomized into a double-blind, placebo-controlled study in the UK and the Netherlands. Efficacy variables included the mean change in intravaginal ejaculatory latency time (IELT) from baseline and the proportion of patients who achieved an IELT of ≥ 4, ≥ 3 or ≥ 2 min on two occasions, and the effect of TEMPE on the index of ejaculatory control (IEC) and sexual quality-of-life (SQoL) scores of patients and their partners. Safety and adverse event data were also collected.

Fifty-four patients were randomized and received study treatment.


The observed mean change in IELT from baseline to the end of the treatment period was 3.8 min in the TEMPE group and 0.7 min in the placebo group, and when adjusted for baseline and centre was 2.4 times higher in the TEMPE than the placebo group (P < 0.01). The efficacy of TEMPE in increasing IELT was further supported by positive trends in the other efficacy endpoints. The proportion of men who had an IELT time ≥ 2, ≥ 3 or ≥ 4 min on two occasions after treatment was 11/20 (55%), 8/20 (40%) and 5/25 (20%) in the TEMPE group, and 8/23 (35%), 3/23 (13%) and 3/23 (13%) in the placebo group, respectively, although these differences were not statistically significant. Improvements in IEC and SQoL (male and female) scores also showed trends towards greater efficacy for TEMPE than placebo. In all, 35 of 42 (83%) patients considered the spray easy to use. Mild to moderate local numbness occurred in three (12%) of the TEMPE-treated patients but did not lead to discontinuation.


Topical treatment with TEMPE produced a statistically and clinically significant increase in IELT compared with placebo, and resulted in positive trends in ejaculatory control and SQoL. TEMPE was considered easy to use and was well tolerated. The data support the conduct of further large-scale studies to establish the utility of TEMPE as a first-line treatment for PE.


topical eutectic mixture for (premature ejaculation)


Diagnostic and Statistical Manual IV


intravaginal ejaculatory latency time


sexual quality of life (male) (female)


index of ejaculatory control


adverse event




analysis of covariance


Premature ejaculation (PE) is the most common sexual problem experienced by men [1]. Whilst reports of the incidence of PE vary, it is generally estimated that PE symptoms are experienced by 25–30% of the male population [1,2], but is paradoxically a problem for which they are unlikely to seek help.

There are no biomarkers for PE and the diagnosis of the condition according to the Diagnostic and Statistical Manual-IV (DSM-IV) definition is a subjective evaluation made by the patient. However, in clinical trials PE is evaluated in a time-dependent manner to characterize the study population and quantify the benefit of intervention, with the most robust results obtained in patients with a baseline stopwatch-measured intravaginal ejaculatory latency time (IELT) of <1 min [3].

The subjective diagnosis and wide range of underlying causes make PE a difficult condition to treat. Behavioural therapy is commonly used, although with limited success, as most benefits are lost within 3 years of treatment [4]. Systemic treatments have included adrenergic antagonists [5], γ-amino butyric acid and selective serotonin reuptake inhibitors [6–10]. Success with these agents has been variable and is associated with side-effects.

Men with PE generally have abnormal autonomic reflex pathways for the ejaculatory process, including a lower vibratory threshold to ejaculation [11]. Reducing this heightened sensitivity of the glans penis with topical desensitizing agents might therefore be a way of improving IELT, without adversely affecting the sensation of ejaculation. Topical local anaesthetics have been used in the treatment of PE, including lidocaine-prilocaine cream (EMLA®, AstraZenenca, Luton, UK) which has been claimed to be effective in improving IELT [12–14]. However, this technique has not become widely accepted, possibly due to the long time to onset of effect (20–30 min) [13], the rather messy use of cream in a condom, and that the whole penis is anaesthetized [12–14].

The topical eutectic mixture for premature ejaculation (TEMPE, Plethora Solutions Ltd, London, UK) is a metered-dose spray of lidocaine and prilocaine under development as a first-line therapy for PE [15]. The physiochemical characteristics of the eutectic mixture and the spray delivery have been designed to optimize tissue penetration so that the onset of effect should be more rapid than that obtained with a cream formulation, and a condom is not required. Also the spray only penetrates, and hence anaesthetises, the mucosa of the glans penis (and not the keratinized skin of the shaft). The concept of using TEMPE to prolong IELT for patients was assessed previously in a preliminary, open-label study in 11 men with PE, and resulted in significant increases in IELT [15].

The primary objective of the present study was to evaluate the efficacy of TEMPE compared to placebo in treating men with PE. Secondary objectives included evaluating the effect of TEMPE on ejaculatory control and sexual quality of life (SQoL), and to assess the safety and tolerability of treatment in men with PE and their sexual partners.


The study included men aged 18–75 years with a history of primary PE according to DSM-IV criteria [16] of ≥ 6 months duration; all had to be in a stable heterosexual and monogamous relationship of ≥ 3 months, and the couples had to be willing to attempt sexual intercourse on at least seven occasions during the 10-week study. The exclusion criteria were as usual for studies in men with PE [17] but in addition, men with known drug sensitivity to amide-type local anaesthetics were also excluded. All patients and their partners provided written, informed consent to participate. The study recruited men from seven centres across the UK and the Netherlands.

Patients were expected to attend three visits (screening, randomization and follow-up) over a maximum of ≈ 10 weeks; the study design is shown in Fig. 1. Patients and their partners attended a screening visit at which they both provided consent. Eligible couples were provided with diary cards, including an index of ejaculatory control (IEC, a study-specific questionnaire containing six questions relating to a subject’s control over his ejaculation) and study-specific male and female SQoL questionnaires (SQoL-m and SQoL-f; each containing 10 statements relating to how the individual felt about the quality of his/her sex life and how it affected his/her sense of well-being), and a stopwatch. After screening, the patient or his partner measured his IELT with the stopwatch on three consecutive occasions during the first study month (baseline). Patients and their partners completed their respective questionnaires in their diary cards. At the second study visit, continuing eligibility for the study was evaluated. Patients who continued to fulfil the eligibility criteria were randomized to TEMPE or placebo spray to use before sexual intercourse on four consecutive occasions during the second month (double-blind treatment phase).

Figure 1.

The study design.

A treatment regimen of three self-administered actuations of TEMPE (delivering a total of 22.5 mg lidocaine and 7.5 mg prilocaine) onto the glans penis 15 min before sexual intercourse was chosen for this study. This dosage regimen was previously used in the preliminary, open-label study using a prototype of the TEMPE spray, and was found to be both effective and well tolerated [15]. Both active drug and placebo were presented in a metered-dose aerosol spray device and contained a hydrofluorocarbon gas as a propellant. The placebo spray was identical in appearance to the TEMPE spray, but did not contain active drugs. Patients were instructed not to use the spray more than once in any 24-h period, to avoid possible bias in IELTs resulting from too frequent ejaculation.

The patient or his partner measured his IELT with a stopwatch whenever the study spray was used. At the end of the second month, patients and their partners completed their respective questionnaires and recorded their overall impression of study treatment in their diary cards. The patients were re-assessed at a third visit ≈ 2 weeks after the end of the second month or after the patient’s last IELT recording during the double-blind treatment phase, whichever was earlier.

Efficacy was evaluated by the stopwatch-recorded IELT and the questionnaires included in the diary cards, the IEC and SQoL. There were two primary efficacy variables: (i) the change from baseline in mean IELT during the double-blind treatment phase (second month) for TEMPE-treated vs placebo-treated patients; and (ii) the proportion of patients who responded to the study treatment. A responder was defined as a patient who had at least two sexual encounters where the IELT was ≥ 4 min during the second month.

Secondary efficacy variables included the proportion of patients who had at least two IELT readings of ≥ 2 and ≥ 3 min during the treatment period, the change from baseline in the total IEC score, the effect of TEMPE on the SQoL of patients and their partners as assessed by the change from baseline in the SQoL scores, overall impression of effectiveness of study treatment, and ease of use of the study spray.

Safety was assessed by the reporting of the nature and incidence of adverse events (AEs) by patients and their partners during the study. Standard laboratory evaluations, vital signs (heart rate and blood pressure), 12-lead electrocardiogram and physical examination of the patient were recorded.

All patients who used study medication for at least one sexual encounter and who had at least one efficacy measurement after randomization, irrespective of any major protocol deviations, were included in the intent-to-treat (ITT) population; this was the primary patient group used to evaluate the efficacy of TEMPE. All patients with documented use of at least one dose of study medication were included in the safety population, this being used for all safety and tolerability analyses.

The baseline IELT for each subject was calculated as the mean IELT of the three consecutive attempts during the first month. The IELT during the double-blind treatment phase was calculated as the mean IELT of all attempts during the second month. The change from baseline was analysed using analysis of covariance (ancova), with factors of treatment and centre included in the model and baseline IELT as covariate. The interaction between centre and treatment was investigated in the model and small centres were pooled for the purposes of analysis. Adjusted least-squares means were presented for each treatment group, with the 95% CI for the estimated difference between them. Where the data did not satisfy the usual assumptions of normality and constant variance, the data were log-transformed and analysed using a similar ancova model to that described above, with the second month observation used as the dependent variable, rather than the change.

The proportions of responders in each treatment group were calculated and compared using a chi-square test (with 95% CIs for the differences). An exploratory binary logistic regression analysis was also used to examine potential prognostic factors such as subject age, mean baseline IELT and centres pooled. Odds ratios and 95% CIs were presented as appropriate.

The numbers of patients who had at least two IELT readings of ≥ 2 min and ≥ 3 min in the second month were each calculated. Data are presented as percentages, with no formal statistical analysis. A single index was calculated from each of the IEC and SQoL-m and SQoL-f questionnaires, such that low values indicated a poor outcome and high values a good outcome. The change from baseline in the IEC and each of the two SQoL questionnaires was analysed using ancova, with factors of treatment and centre included in the model and baseline scores as covariates. The results from small centres were pooled as described for the analysis of IELT above.

The sample size was based on the proportion of patients who responded to the study treatment (as defined above), using data from an early study [15]. The study was designed to have ≥ 70% power to show a difference in response of ≈ 40% between TEMPE and placebo.


Sixty-two men gave informed consent and were eligible to enter the study at seven study sites. Patient accountability, including reasons for withdrawal, is summarized in Fig. 2. Patients and their partners were reasonably comparable across the two treatment groups for all demographic and baseline characteristics (Table 1). The IELTs at baseline were similar for the two groups, with a mean IELT of ≈ 1 min. The IEC and SQoL-m and SQoL-f scores were also similar at baseline for the two groups.

Figure 2.

The flow of patients through the study.

Table 1.  The characteristics of the patients and partners at baseline, the efficacy variables and the treatment-emergent AEs
  • *

    One study site did not report ethnic origin; this is tabulated as missing;

  • adjusted for baseline + centre;

  • ‡higher score reflects a better outcome.

Number of patients 26 28
Mean (sd):
 Age, years 38.5 (9.8) 39.3 (10.7)
 Height, m  1.76 (0.07)  1.78 (0.08)
 Weight, kg 80.5 (9.05) 84.19 (13.01)
 Months since diagnosis130.6 (116.8)102.4 (100.8)
Median (range) months since diagnosis 91.0 (8–415) 59.0 (7–324)
Ethnic origin, n (%)
 White 19 (73) 24 (86)
 Afro-Caribbean  2 (8)  1 (4)
 Asian (Indian subcontinent)  1 (4)  1 (4)
 Missing*  4 (15)  2 (7)
Mean (sd) age, years 37.2 (9.3) 36.9 (9.6)
Ethnic origin, n (%)
 White 19 (73) 23 (82)
 Afro-Caribbean  2 (8)  1 (4)
 Asian (Indian subcontinent)  1 (4)  2 (7)
 Missing*  4 (15)  2 (7)
Number of patients 20 23
Mean (sd) IELT, min
 Baseline  1.0 (1.2)  0.9 (0.7)
 Follow-up  4.9 (4.9)  1.6 (1.6)
 Change  3.8 (4.5)  0.7 (1.4)
Geometric mean (95% CI) change  2.50 (1.6–3.9)  1.04 (0.7–1.6)
 difference (95% CI), P ×2.4 (1.3–4.4), <0.01 
Mean (sd) IEC
 Baseline  2.6 (3.0)  2.7 (2.7)
 Follow-up  9.6 (6.3)  5.6 (4.4)
 Change  6.7 (7.4)  3.0 (4.0)
Geometric mean (95% CI) change  8.1 (6.4–10.2)  6.3 (5.0–7.9)
 difference (95% CI), P ×1.3 (0.9–1.8), 0.12 
SQoL-m, mean (sd)
 Baseline 18.1 (13.6) 13.8 (8.6)
 Follow-up 25.0 (14.9) 18.8 (10.7)
 Change  7.0 (14.1)  5.5 (9.4)
Mean (95% CI) change  7.9 (2.7–13.1)  5.4 (0.5–10.3)
 difference, P ×2.5 (−4.7, 9.7), 0.48 
SQoL-f, mean (sd)
 Baseline 27.6 (14.8) 25.4 (12.3)
 Follow-up 29.8 (13.5) 27.2 (12.6)
 Change  3.3 (10.1)  1.8 (10.6)
Mean (95% CI) change  3.5 (−1.0, 8.0]  1.7 (−2.4, 5.8)
 difference (95% CI), P ×1.8 (−4.3, 7.9), 0.56 
Treatment-emergent AEs, n (%), as AE preferred term
Number of patents 26 28
Hypoaesthesia  3 (12)  0
Influenza  1 (4)  2 (7)
Erectile dysfunction  1 (4)  0
Respiratory tract infection  1 (4)  0
Skin burning  0  1 (4)
Headache  0  1 (4)

The observed mean change in IELT from baseline to the end of the treatment period was higher in the TEMPE than in the placebo group, at 3.8 min vs 0.7 min, respectively (Table 1). Overall, the mean change in IELT from baseline (adjusted for baseline and centre) was 2.4 times higher in the TEMPE than the placebo group (P < 0.01; Table 1).

The other primary efficacy variable, the proportion of patients with an IELT ≥ 4 min on at least two occasions, and the secondary variables, the number of patients who had IELTs of ≥ 3 or ≥2 min during at least two sexual encounters also showed a trend towards greater efficacy with TEMPE than placebo (Fig. 3).

Figure 3.

The percentage of patients who achieved IELTs of ≥ 2, ≥ 3 or ≥ 4 min on at least two occasions during the treatment period.

There was a clinically meaningful improvement in the IEC score in the TEMPE group over baseline. The mean change from baseline in the total IEC score was 6.7 and 3.0 points in the TEMPE and placebo groups, respectively, although adjusted for baseline and centre this difference was not statistically significant (Table 1).

The SQoL scores for patients and their partners at the end of the second month showed a trend to a greater improvement over baseline in the TEMPE than in the placebo group. The mean change from baseline for the SQoL-m total score was 7.0 points, vs 5.5 points in the placebo group, and for female partners the mean change from baseline was 3.3 points in the TEMPE vs 1.8 points in the placebo group. The estimated differences between treatments in the change in score, when adjusted for centre and baseline, were 2.5 points and 1.8 points in favour of TEMPE for the patients and partners, respectively, although these differences were not statistically significant (Table 1).

Of the 42 patients who used the spray (either active or placebo) and answered the question, 35 (83%) found the spray easy to use. Most patients in the TEMPE group believed that the spray prolonged the time to ejaculation and most partners believed that the TEMPE spray improved their partner’s ability to control ejaculation (Fig. 4).

Figure 4.

The overall impression of effectiveness; the percentage of patients who believed that the treatment prolonged the time to ejaculation, and the percentage of partners who believed that the treatment improved their partner’s ability to control his time to ejaculation.

The topical administration of three actuations of TEMPE spray 15 min before sexual intercourse was well tolerated. In all, only four (15%) TEMPE-treated men had a treatment-related AE, all of which were either mild or moderate in severity; three of these patients (12%) had hypoaesthesia (numbness) of the penis, whereas the fourth reported erectile dysfunction. All treatment-emergent AEs are also summarized in Table 1. TEMPE was also well tolerated by the female partners; four AEs were considered to be related to study medication, all of which occurred in the partner of one TEMPE-treated patient. She recorded a mild burning sensation during intercourse each time the spray was used, although this did not result in treatment discontinuation. There were no adverse effects of TEMPE on variables monitored for patient safety, i.e. vital signs, physical findings, 12-lead electrocardiogram, haematology, biochemistry or urine analysis during the study.


The objectives of this phase II study were to evaluate the efficacy and tolerability of TEMPE compared to placebo as a topical treatment for men with PE. The study showed, with clinical and statistical significance, that TEMPE was 2.4 times more effective at prolonging IELT than placebo. There were positive trends in all secondary endpoints, although there were too few patients to give statistical significance.

There appears to be little consensus as to how to define PE (or indeed normality) in terms of time to ejaculation, and PE has been variously defined as IELTs of <1 min [3] to <4 min [18], which overlaps considerably with a ‘normal’ duration of sexual intercourse of 2–7 min [3,15]. The present study recruited men with PE based on the DSM-IV criteria and did not require specific IELT thresholds for inclusion. As such, it is likely that the population studied here is representative of the general population with PE [17].

In the present study, treatment with TEMPE resulted in a mean IELT of 4.9 min, i.e. within the ‘normal’ range (vs 1.6 min in the placebo group). The study was not powerful enough to detect a statistically significant difference in the response rate defined in this study (at least two sexual encounters where the IELT was ≥ 4 min during treatment), largely due to the lower than anticipated response to TEMPE, which arose because there were four patients for whom the response could not be determined (because they had insufficient IELT recordings). However, although analysed as non-responders, where IELT was recorded for these four patients, the times were all >4 min, apart from one reading of 3.8 min.

It is now generally accepted that arbitrary fixed times for a ‘normal’ IELT are not necessarily the most relevant aspect of patient satisfaction. Prolongation of previously unsatisfactory IELT, however short or long that might be, is considered more relevant to patient satisfaction than achieving predefined IELT goals or indeed aiming to reach a ‘normal IELT’[3,15,19]. Overall, this phase II study showed that TEMPE produces a clinically and statistically significant improvement in patients’ IELT when compared with placebo, and the results are comparable to those reported in recent reports for effective PE products [10,20–22], including a topical anaesthetic cream [14] and an oral treatment for PE [21].

Whilst prolonging IELT is an accepted goal of PE therapy, more subjective measures, such as improvements in SQoL and ejaculatory control, are also relevant and might be as important to men with this condition [23,24]. The effect of PE on the patient’s partner is also relevant [23]. In the present study, patients used TEMPE on four occasions, which is probably insufficient usage to expect a large change in SQoL and IEC scores. However, even after such limited use there was an encouraging trend to improvements in ejaculatory control and SQoL in men with PE and in their partners. Longer-term studies are required to draw more meaningful conclusions about changes in SQoL and ejaculatory control. It might be prudent to consider using a variable not focused on time, e.g. the IEC score (or other validated measure of ejaculatory control) as a primary efficacy variable in future studies, to evaluate the treatment in a way that might be more meaningful to patients.

The components of TEMPE have an established long-term safety profile and, as would be expected, TEMPE was well tolerated in the study. Although TEMPE contains a local anaesthetic, numbness was infrequent, mild and did not lead to discontinuation of use. The product has the potential advantage over oral drugs taken every day, as is it applied locally when needed (i.e. as an ‘on-demand’ therapy) and is effective within minutes. The topical formulation minimizes the risk associated with systemic exposure, so there is less likelihood of side-effects compared to oral, long-acting treatment [6–9]. The refusal of the USA Food and Drug Administration to approve dapoxetine (a novel selective serotonin reuptake inhibitor, effective as an on-demand therapy [10]) for treating PE, is possibly due to concern over the possible (albeit low) incidence of syncope and psychotropic effects, and has highlighted the importance of the risk : benefit ratio in the approval of drugs for treating this condition.

Currently, only a small percentage of men with PE seek or receive treatment from a healthcare professional [25] and the lack of effective pharmacological treatment is a contributing factor in this. The encouraging data from the present phase II study suggest that TEMPE ‘as required’ has the potential to offer a convenient, novel treatment option for men with PE. The study results suggest that TEMPE might be useful as a first-line treatment for men with PE.


The authors acknowledge members of the Plethora Clinical Team for their contributions to this study.


W. Dinsmore, G. Hackett, D. Goldmeier, M. Waldinger, J. Dean, P. Wright, M. Callander and K. Wylie are study investigators funded by sponsor; W. Dinsmore and C. Keywood are paid consultants to sponsor; C. Novak and P. Heath are employess of sponsor; P. Heath is also a stock holder for the mentioned product; M. Wyllie is a member of the board of sponsor. Source of funding: Plethora Solutions PLC.