SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

BACKGROUND Botulinum toxin type A (BoNT-A) is frequently used to treat crow's feet; the optimal dose for each toxin preparation should be established.

OBJECTIVE To compare the efficacy and safety of three doses of BoNT-A with placebo to treat crow's feet.

METHODS Subjects with moderate to severe crow's feet at maximum smile (mild to severe at rest) were randomized to a single bilateral BoNT-A treatment (15, 30, or 45 U) or placebo. Outcome measures included evaluation of crow's feet by an independent panel from blinded photographs at maximum smile, investigator assessment, and patient satisfaction.

RESULTS Independent panel assessments (Week 4) showed that all BoNT-A doses resulted in significant improvements in crow's feet severity at maximum smile (p<.001); a clear dose–response effect was seen. Improvement over placebo was seen in the 30-U and 45-U groups to Week 12. Investigator assessment showed significant improvement for all doses for 12 weeks at maximum smile and rest (p≤.01). Patient satisfaction was significantly greater for all doses than for placebo for 16 weeks (all p<.05). All doses were well tolerated.

CONCLUSION BoNT-A (15, 30, or 45 U) is an effective and safe treatment for mild to severe crow's feet.

Over time, the actions of the facial musculature, coupled with loss of skin elasticity and a reduction in subcutaneous tissue,1 result in the development of permanent facial lines. Contraction of the orbicularis oculi muscle forms crow's feet, which are usually the result of aging. Over the past 2 decades, there has been increasing demand for aesthetic procedures to reverse the effects of aging, particularly in the facial area, and treatment of crow's feet is among the most frequent of corrective treatment requests.2,3

A number of techniques have been developed to meet the increasing demand for aesthetic procedures to reverse the signs of aging, including surgery, various resurfacing procedures, and the use of filling agents.4 Injection of botulinum toxin type A (BoNT-A) is now one of the most common aesthetic procedures performed,2 and there is a considerable body of medical literature demonstrating its tolerability and efficacy in aesthetic indications.2,3,5–10 When injected, BoNT-A causes dose-dependent weakness or paralysis in skeletal muscles by blocking the calcium-mediated release of acetylcholine from motor nerve endings.

BoNT-A is an effective treatment for crow's feet and is associated with high subject satisfaction and only occasional mild side effects,11–15 but the published literature shows that a wide range of doses (20–75 U) of BoNT-A has been used to treat crow's feet, with injections given at a varying number of sites.16 No placebo-controlled trials have established the optimal dose of this preparation of BoNT-A (Dysport, Ipsen Biopharm Ltd., Wrentham, UK) for this indication. The aim of this multicenter, randomized, double-blind, placebo-controlled study was to compare the efficacy and safety of three doses of BoNT-A with placebo for the treatment of crow's feet.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

Study Design

This was a multicenter, randomized, double-blind, parallel-group study comparing three doses of BoNT-A (Dysport, Ipsen Ltd., Slough, UK) with placebo. The study was conducted at nine sites in Europe; generally, there was only one physician injector at each site. Subjects received a single treatment with study medication at the baseline visit. Efficacy and safety evaluations were conducted at Week 2 and then every 4 weeks for 6 months.

Subject Population

Male and female subjects aged 18 to 65 with moderate to severe (Grades 2 or 3) crow's feet during maximum smile and mild to severe (Grades 1, 2, or 3) crow's feet at rest (on both sides of the face), as determined by the investigator, were included in this study. All subjects gave written, informed consent before the study, which national and local ethics committees approved and was conducted according to the principles of Good Clinical Practice and the Declaration of Helsinki.

Subjects were excluded if they had any prior surgery affecting the area around the eye or any prior botulinum toxin treatment. Subjects taking medication, with facial skin conditions affecting neuromuscular function, or with psychiatric illnesses that in the investigator's opinion could affect the safety, conduct, or outcome of the study were also excluded. Women of child-bearing potential were required to have a negative pregnancy test before study entry.

Study Medication and Dosage

Subjects were randomly assigned to one of four treatment arms (BoNT-A 15, 30, or 45 U per eye or placebo) according to a computer-generated randomization list. Injecting physicians were blinded to treatment (medication vs placebo). An independent reconstitutor reconstituted the study medication (active treatment or placebo) to maintain the study blind. Only the independent reconstitutor was aware of the treatment, and all study treatments were provided to the injecting physicians in a final volume of 0.3 mL, regardless of dose, to further preserve the blind. The injection volume of study medication was identical (0.05 mL/injection point) regardless of the dose administered. Study medication was reconstituted with 0.9% preservative-free sterile saline to give 5, 10, or 15 U of BoNT-A per 0.05-mL injection aliquot (or equivalent excipients in the case of placebo). Subjects received three injections per eye into the lateral canthal area, as shown in Figure 1. Care was taken to avoid injecting close to the inferior margin of the zygomaticus major muscle and to avoid any intraorbital diffusion.

image

Figure 1.  Precision injection points (courtesy of Benjamin Ascher, MD).

Download figure to PowerPoint

Outcome Measures and Procedures

Scoring System and Standardized Digital Photographs: All subjects were assessed twice at each study time point—one “live” assessment by the investigating clinician and one standardized photograph assessment by an independent panel of four expert clinicians blinded to treatment group and study time point. All photographs of the subjects were taken under standard conditions (identical camera equipment, positioning, lighting, and settings).

Clinicians and the independent panel assessors assessed the severity of the subject's crow's feet at rest and at maximum smile using the following 4-point rating scale: 0=none, 1=mild, 2=moderate, 3=severe. This scale had been validated before this study and has shown good inter- and intra-observer reproducibility.17 To standardize interpretation of the scale, the independent review panel members and all clinicians responsible for grading the severity of crow's feet were given the same training on the use of the scale and provided with a photographic atlas containing photographs to represent each grade of severity.

Subjects' own assessment of satisfaction with the overall appearance of their crow's feet was made using a 4-point rating scale: 0=very satisfied, 1=satisfied, 2=dissatisfied, and 3=very dissatisfied.

Study End Points

For assessments at maximum smile, response was defined as an improvement in severity of crow's feet from moderate or severe (Grade 2 or 3) at baseline to none or mild (Grade 0 or 1) on both sides. For assessments at rest, response was defined as an improvement in crow's feet severity of at least one grade from baseline on both sides. The primary efficacy end point was the percentage of responders at maximum smile at Week 4, as assessed by blinded independent panel review of the standardized photographs. Additional outcome measures were independent panel assessment of percentage responders (maximum smile and at rest) at all other time points, “live” assessment of percentage responders (maximum smile and at rest) at all time points, and subject satisfaction with appearance of their crow's feet at all time points. Adverse events (AEs) were recorded throughout the study using the Medical Dictionary for Regulatory Activities preferred terms.

Statistical Methods

The primary efficacy analysis, and all analyses of data from photographic assessments, were performed on the modified intention-to-treat (mITT) population. This data set excluded subjects who did not meet the moderate or severe (Grades 2 or 3) criteria for crow's feet during maximum smile and mild to severe (Grades 1, 2 or 3) during rest after review of baseline photographs by the independent panel. End points relating to live assessments provided by the investigator or the subjects were analyzed using the intention-to-treat (ITT) population. The safety population included all subjects who received at least one dose of treatment and who provided any follow-up data.

The number of responders in each treatment group were compared with placebo for each variable at each visit using the chi-squared test. If the criteria of validity for a chi-squared test were not fulfilled, then the Fisher exact test was used. Hochberg's procedure was used to correct for multiple testing.18 For each photograph, separate ratings were obtained from each of the four independent assessors and the majority opinion (mode value) of the ratings taken to give one score per subject. In the event that there was an even split between two or more grades, the most severe grade was taken.

Determination of Sample Size

Sample size calculation was based on the primary criterion for which an estimate of the placebo responders rate was 25% and estimates of the BoNT-A responders rates were 60%, 65%, and 70% for 15, 30, and 45 U, respectively. Under these conditions, it can be calculated that 42 subjects per treatment group would allow 85% power to detect a 35% difference in responder rates using a two-sided chi-squared test at the required adjusted significance level. To allow for subjects dropping out of the study or not being assessable for the primary criterion, it was planned to recruit 52 subjects per treatment arm into the study, making a total of 208 subjects.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

Subjects

A total of 220 subjects were enrolled into the study (12 more than planned because of technical problems with photographic equipment during the early stages of the trial); 200 subjects completed the study. Demographic data for the safety population (n=218) are shown in Table 1. The majority of subjects were female, Caucasian, and in their mid-40s. Because of a high incidence of protocol violations at one center, efficacy data for all subjects from that site were excluded, resulting in an ITT population of 193 subjects. The mITT (maximum smile) population consisted of 162 subjects; 31 were excluded because severity ratings were missing or were rated none or mild at baseline by the independent panel. The mITT (rest) population consisted of 188 subjects; five subjects were excluded because severity ratings were missing or were rated none at baseline by the independent panel.

Table 1. Demographics—Safety Population
CharacteristicBotulinum Toxin Type A Dose
15 U30 U45 UPlacebo
(n=55)(n=54)(n=55)(n=54)
  • Two subjects were excluded from the safety population because they did not provide any follow-up safety or efficacy data.

  • *

    One subject had no date of birth recorded, so age could not be calculated.

Age, mean ± standard deviation (range)47.6 ± 7.6 (30–64)47.8 ± 9.5* (30–65)48.3 ± 10.0 (28–69)47.6 ± 8.2 (28–62)
Sex, n (%)
 Male7 (13)6 (11)9 (16)4 (7)
 Female48 (87)48 (89)46 (84)50 (93)
Race, n (%)
 Caucasian54 (98)54 (100)55 (100)54 (100)
 Non-Caucasian1 (2)0 (0)0 (0)0 (0)

Efficacy

Independent Panel Assessments: The severity of crow's feet at maximum smile, assessed by the independent panel at Week 4 (primary efficacy criteria) is summarized in Table 2. The treatment difference between BoNT-A and placebo at Week 4 was statistically significant for all doses (p<.001). At Week 4, response rates were 57%, BoNT-A 15 U group; 68%, BoNT-A 30 U group; 78%, BoNT-A 45 U group; and 16%, placebo group, showing a clear dose-related treatment response.

Table 2. Severity of Crow's feet at Maximum Smile Assessed by Independent Panel at Week 4—Modified Intention-to-Treat Population
Severity and Response RateBotulinum Toxin Type A Dose
15 U30 U45 UPlacebo
(n=42)(n=37)(n=40)(n=43)
  • *

    p<.001 between botulinum toxin type A and placebo using Hochberg's correction for multiple testing.

Severity at baseline, n (%)
 Moderate23 (55)19 (51)15 (38)24 (56)
 Severe19 (45)18 (49)25 (63)19 (44)
Severity at Week 4, n (%)
 None2 (5)5 (14)8 (20)0 (0)
 Mild22 (52)20 (54)23 (58)7 (16)
 Moderate14 (33)9 (24)6 (15)14 (33)
 Severe4 (10)3 (8)3 (8)22 (51)
Response rate at Week 4; n (%; 95% confidence interval)24 (57; 41–72)*25 (68; 50–82)*31 (78; 62–89)*7 (16; 7–31)

The independent panel assessments of severity of crow's feet at maximum smile at all time points are shown in Figure 2A. There was a statistically significantly greater proportion of responders in all BoNT-A groups than with placebo at Weeks 2, 4, and 8 (all p≤.001). A statistically significantly greater proportion of responders in the BoNT-A 30-U (p≤.05) and BoNT-A 45-U (p≤.05) groups than with placebo was also seen at Week 12. Maximum response was seen 2 weeks after treatment for BoNT-A 15 U (69%) and BoNT-A 30 U (78%) and 4 weeks after treatment with BoNT-A 45 U (78%). When age of subjects was considered, response rate at Week 4 was higher in subjects aged 50 or younger than those aged over 50 in all treatment groups (Figure 3).

image

Figure 2.  Percentage responders (independent panel assessment at maximum smile (A), and at rest (B))—modified intention-to-treat population.

Download figure to PowerPoint

image

Figure 3.  Percentage responders (independent panel assessment at maximum smile) at Week 4 according to age—modified intention-to-treat population.

Download figure to PowerPoint

Independent panel assessments of severity of crow's feet at rest (mITT population, Figure 2B) showed a statistically significant difference between BoNT-A 15 U and placebo at Week 2 and Week 12 (p<.05), between BoNT-A 30 U and placebo at Weeks 2, 8, and 12 (p<.05), and between BoNT-A 45 U and placebo from Week 2 to Week 16 (p<.05). The number of responders at Week 4 (at rest) was dose related: 28%, 15-U group; 43%, 30-U group; 53%, 45-U group; 23%, placebo group.

Investigator Assessments: Investigator-assessed severity of crow's feet at maximum smile (ITT population) showed a statistically significant difference between all BoNT-A doses and placebo for up to 12 weeks follow-up (p≤.01) (Figure 4A). Maximum response was seen 2 weeks after treatment for all BoNT-A doses (46%, 67%, and 65%, respectively). At Week 4, response rates at maximum smile were: 42%, 15-U group; 60%, 30-U group; 57%, 45-U group; 10%, placebo group. Investigators' assessments of severity of crow's feet at rest (ITT population) are presented in Figure 4B. These results showed a statistically significant difference between all BoNT-A doses and placebo at Weeks 2, 4, 8, 12 (p<.001), and 16 (p≤.05).

image

Figure 4.  Percentage responders (investigators'; assessment at maximum smile (A), and at rest (B))—intention-to-treat population

Download figure to PowerPoint

Subject Assessments: Subjects' assessment of satisfaction with improvement in the appearance of their crow's feet is shown in Figure 5. All BoNT-A doses resulted in statistically significantly higher satisfaction ratings than placebo from Week 2 to Week 16 (all p<.05). The highest rates of satisfaction (very satisfied or satisfied) occurred at 4 weeks of follow-up for 15 U and 45 U and at 8 weeks of follow-up for 30 U.

image

Figure 5.  Subject satisfaction with treatment—intention-to-treat population.

Download figure to PowerPoint

Safety

AEs were reported for 57 of 218 subjects in the safety population. The most common AEs reported by two or more subjects in either group were nasopharyngitis, periorbital hematoma, depression, and bronchitis. The incidence of treatment-emergent AEs, considered to be possibly or probably associated with the study treatments, was similar for all doses of BoNT-A (13%, 11%, and 13% for the 15-U, 30-U, and 45-U groups, respectively). The most frequently reported treatment-related AE was mild periorbital hematoma (10% of all subjects). All treatment-related events appear to be local to the injection site or related to the treatment procedure. Only one patient in the 45-U group reported eyelid ptosis. The patient first reported this 1 week after treatment, and the investigator assessed it as mild in intensity at Week 2. The patient reported that this event had resolved on Day 31 after treatment, and at Week 8 the investigator recorded the event as having resolved without sequelae. No ptosis was reported in any other treatment groups. Only one treatment-related AE was considered to be severe in intensity, and none led to study withdrawal. Six subjects reported eight serious AEs, all of which were assessed to be unrelated to treatment.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

The results of this study demonstrate that BoNT-A is an effective treatment for crow's feet, as has been previously demonstrated with this product in the glabellar area.3,9,19,20 All doses of BoNT-A resulted in a statistically significantly better response in the appearance of crow's feet than placebo on the primary end point—assessment of crow's feet at maximum smile by an independent panel at 4 weeks. A dose-related increase in treatment response was observed at Week 4 according to independent panel assessment of crow's feet at maximum smile and at rest. Significant improvement over placebo was maintained for up to 8 weeks with 15 U and for up to 12 weeks with 30 U and 45 U, rated by independent panel assessment. A clear differentiation in dose response between the 30-U and 45-U doses was not observed here, and is similar to previous findings using a different preparation of BoNT-A.14 Investigator assessments of severity of crow's feet at maximum smile produced similar results, with statistically significant improvement over placebo lasting up to 12 weeks. Improvement with all BoNT-A doses was rapid and occurred within 2 weeks.

Investigator assessments of severity of crow's feet at rest showed significantly greater improvement than with placebo for up to 16 weeks for all BoNT-A doses, although results from independent panel assessment were less consistent. Although all doses produced statistically significant improvement at Week 12, the lower doses (15 and 30 U) were not significantly different at all previous time-points, and only BoNT-A 45 U remained significant at Week 16. Furthermore, investigator assessments of severity of crow's feet at Week 4 did not show a clear dose–response relationship. The small difference in findings between the different assessors probably reflects methodological differences between “live” and photographic assessment. In this study, independent panel assessment was conducted using double-blind retrospective evaluation of standardized digital photographs of each subject, whereas investigator assessment was made using “live” evaluation. Although both methods used the same validated rating scale and a photographic scoring guide to aid wrinkle assessment, live assessment allows for assessment of additional factors, such as the effects of muscular movement, not detectable using still photography. Furthermore, flash photography can flatten details of the face and reduce the apparent severity of fine lines and wrinkles, resulting in poorer sensitivity than with live assessment.

Published data regarding crow's feet treatment with botulinum toxin show that a wide variety of assessments methods have been adopted.11–15 These vary from assessment by subjects, clinicians, or trained observers to more objective methods such as the use of photography and measurement of wrinkle length. The methods used in this study were designed to be as rigorous as possible, and the observed differences between independent panel assessment and investigator assessment highlight the importance of more objective primary evaluation criteria for assessment of severity of crow's feet. In this study, the objective, blinded evaluation of crow's feet severity by an independent panel probably provided better consistency and standardization than observations made by the individual clinicians at each study center.

Independent panel assessment of the beneficial effects of treatment at rest may also have been less pronounced than at maximum smile because the genesis of crow's feet is a muscular process, and these lines are therefore most evident during muscular activity (when smiling). Because BoNT-A exerts its aesthetic effect in the treatment of dynamic wrinkles by reducing underlying muscular contraction, it could be predicted that the greatest improvements in severity of crow's feet would be seen at maximum smile rather than at rest. Given these observations, the relevance of assessment of severity of crow's feet at rest should be questioned. Indeed, previous studies in crow's feet using a different botulinum toxin preparation were restricted to assessment of severity of crow's feet at maximum smile.14,15 Furthermore, subjects with severe crow's feet at rest could be included in this study; given that botulinum toxin reduces the appearance of dynamic wrinkles, severe crow's feet at rest would be less likely to improve after treatment.

More subjects in the 45-U BoNT-A group than other treatment groups had crow's feet ratings of severe at baseline (investigator assessment at maximum smile). Although this has the potential to bias the findings of this study in favor of lower doses, this seems unlikely, because the study observed a dose-related increase in the proportion of responders at Week 4 (placebo, 16%; 15 U, 57%; 30 U, 68%; 45 U, 78%) by independent panel assessment at maximum smile.

A further observation in this study was a trend toward greater treatment response rates for younger subjects (≤50 years) than older subjects (>50 years) at Week 4. Changes in elasticity of the skin with aging may have contributed to this observation, because the benefits of treatment with BoNT-A may be less apparent if the skin is lax. Given these observations, preventive treatment of crow's feet in patients younger than 50 may be appropriate, with the aim of decreasing the incidence of periocular wrinkles in the early stages followed by maintenance of effect.

As with any cosmetic procedure, patient satisfaction with outcome is an important consideration. Despite being a subjective assessment, subject ratings of satisfaction with treatment mirrored more objective assessments by the independent panel and investigator and showed high levels of patient satisfaction with all BoNT-A doses. Subjects perceived that the beneficial effects last for at least 16 weeks post-treatment.

All doses of BoNT-A were well tolerated in this study, with a safety profile similar to that of placebo. The reported AEs were similar across the range of doses and were largely attributable to the injection per se rather than BoNT-A treatment. The most commonly reported treatment-related AE was periorbital hematoma. Eyelid ptosis is a known side effect associated with botulinum toxin treatment in the periorbital region3 and is caused by diffusion of the injected toxin through the orbital septum, whereby it paralyzes the levator palpebrae superioris.10 Ptosis was rare in this study, reported in only one patient, and mild in intensity.

Conclusions

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

This study has established that BoNT-A 15, 30, and 45 U are all highly effective in the treatment of subjects with mild to severe crow's feet, and most patients will benefit for a period of 12 to 16 weeks. Although BoNT-A 15 U was effective, a dose-response relationship was observed for independent assessment of severity scores at maximum smile, showing that BoNT-A 30 and 45 U were optimal doses. Younger patients benefited more from BoNT-A in this region than older patients. This study has also shown that live assessment of facial rhytides may result in different observed benefits from photographic assessments, especially when the subject is at rest. Despite differences between independent panel and investigator assessment, there was broad agreement in terms of efficacy and duration between the results obtained at maximum smile. Treatment with BoNT-A also provided high patient satisfaction across the range of doses. All doses of BoNT-A were well tolerated, with few associated side effects.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  • 1
    Ascher B, Coleman S, Alster T, et al. Full scope of effect of facial lipoatrophy: a framework of disease understanding. Dermatol Surg 2006;32:105869.
  • 2
    Carruthers A, Carruthers J. Botulinum toxin type A: history and current cosmetic use in the upper face. Semin Cutan Med Surg 2001;20:7184.
  • 3
    Rzany B, Dill-Müller D, Grablowitz D, et al. Repeated botulinum toxin A injections for the treatment of lines in the upper face: a retrospective study of 4,103 treatments in 945 patients. Dermatol Surg 2007;33 (1 Spec No.):S1825.
  • 4
    Ascher B, Bui P, Halabi A. Fillers in Europe. In: GoldbergD, editor. Fillers in Cosmetic Dermatology. Abingdon, UK: Informa Publications; 2006. p. 12750.
  • 5
    Ascher B, Klap P, Marion MH, Chanteloub F. Botulinum toxin in the treatment of frontoglabellar and periorbital wrinkles. An initial study. Ann Chir Plast Esthet 1995;40:6776.
  • 6
    Feller G, Bayerl C, Jung EG, Rzany B. Treatment of dynamic facial wrinkles with Botulinum Toxin Type A (Dysport®), a pilot study. Akt Dermatol 2000;26:659.
  • 7
    Becker-Wegerich P, Rauch L, Ruzicka T. Botulinum toxin A in the therapy of mimic facial lines. Clin Exp Dermatol 2001;26:61930.
  • 8
    Ascher B, Rossi B. Botulinum toxin and wrinkles: few side effects and effective combining procedures with other treatments. Ann Chir Plast Esthet 2004;49:53752.
  • 9
    Ascher B, Zakine B, Kestemont P, et al. A multicenter, randomized, double-blind, placebo-controlled study of efficacy and safety of 3 doses of botulinum toxin A in the treatment of glabellar lines. J Am Acad Dermatol 2004;51:22333.
  • 10
    Ascher B, Zakine B, Kestemont P, et al. Botulinum toxin A in the treatment of glabellar lines: scheduling the next injection. Aesthetic Surg J 2005;25:36575.
  • 11
    Keen M, Kopelman JE, Aviv JE, et al. Botulinum toxin A: a novel method to remove periorbital wrinkles. Facial Plast Surg 1994;10:1416.
  • 12
    Ellis DA, Tan AK. Cosmetic upper-facial rejuvenation with botulinum. J Otolaryngol 1997;26:926.
  • 13
    Matarasso SL. Comparison of botulinum toxin types A and B: a bilateral and double-blind randomized evaluation in the treatment of canthal rhytides. Dermatol Surg 2003;29:713; discussion.
  • 14
    Lowe NJ, Ascher B, Heckmann M, et al. Double-blind, randomized, placebo-controlled, dose-response study of the safety and efficacy of botulinum toxin type A in subjects with crow's feet. Dermatol Surg 2005;31:25762.
  • 15
    Lowe NJ, Lask G, Yamauchi P, Moore D. Bilateral, double-blind, randomized comparison of 3 doses of botulinum toxin type A and placebo in patients with crow's feet. J Am Acad Dermatol 2002;47:83440.
  • 16
    Moore PM, Naumann M. Cosmetic Indications. Handbook of Botulinum Toxin Treatment. 2nd ed. Malden, MA: Blackwell Science; 2002.
  • 17
    Hund T, Ascher B, Rzany B. Reproducibility of two four-point clinical severity scores for lateral canthal lines (crow's feet). Dermatol Surg 2006;32:125660.
  • 18
    Chi George YH. Multiple testings: multiple comparisons and multiple endpoints. Drug Inform J 1998;32:1347S62S.
  • 19
    Rzany B, Ascher B, Fratila A, et al. Efficacy and safety of 3- and 5-injection patterns (30 and 50 U) of botulinum toxin A (Dysport) for the treatment of wrinkles in the glabella and the central forehead region. Arch Dermatol 2006;142:3206.
  • 20
    Monheit G, Carruthers A, Brandt F, Rand R. A randomized, double-blind, placebo-controlled study of botulinum toxin type A for the treatment of glabellar lines: determination of optimal dose. Dermatol Surg 2007;33 (1 Spec No.):S519.