ARTICLE
Auteur(s) : Keith D Kaufman1,
Cynthia J Girman2, Elizabeth M Round1, Amy O
Johnson-Levonas1, Arvind K Shah1, Jennifer
Rotonda1
1Merck Research Laboratories, 126 East Lincoln
Avenue, RY34-A248, Rahway, NJ 07065, USA
2Merck Research Laboratories, 351 North Sumneytown Pike,
UG1D-60, North Wales, PA 19454, USA
accepté le 13 Février 2008
Androgenetic alopecia (AGA), or male pattern hair loss, occurs
in men with an inherited sensitivity to the deleterious effects of
androgens on scalp hair growth [1, 2]. This condition is
characterized by visible loss of scalp hair due to progressive
miniaturization of hair follicles [3-5]. Dihydrotestosterone (DHT),
a metabolite formed via the reduction of testosterone by the enzyme
5α-reductase, plays a key role in the pathogenesis of AGA.
Few studies have examined the long-term progression of scalp
hair loss in untreated men with AGA [6-9]. We previously reported
the effects of treatment with finasteride, a type 2 5α-reductase
inhibitor, over 5 years in two replicate, double-blind,
placebo-controlled Phase III studies in men with AGA [10]. In this
previous report, pooled placebo group data from these two studies
served as a control relative to the data from groups receiving
active treatment (finasteride 1 mg). In the present report, we
provide a detailed assessment of the long-term changes in scalp
hair growth occurring in men with untreated AGA, based on analyses
of the pooled placebo group data from these two studies over 5
years.
Methods
Study protocols
Full methodological details of the two Phase III base studies and
their extensions have been reported previously [9, 10]. Two
replicate, 1-year, double-blind, placebo-controlled, randomized,
multicenter studies were initiated, and upon completion continued
as four consecutive, 1-year, double-blind, placebo-controlled,
randomized extension studies. Men aged 18 to 41 years at
randomization, with mild to moderately severe vertex male pattern
hair loss according to a modified Norwood/Hamilton classification
scale (II vertex, III vertex, IV or V) [11, 12], were eligible to
enroll in these studies.
After screening, patients (N = 1553) were randomly assigned
(1:1) to one of two treatment groups, either finasteride
1 mg/day (N = 779) or placebo (N = 774) for 1 year. Patients
completing the 1-year base studies were eligible to enroll in four
consecutive, 1-year, double-blind, placebo-controlled extension
studies. As defined at initial randomization, patients entering the
extension studies were randomized (9:1 during the first extension,
9.5:1 thereafter) to treatment with finasteride 1 mg/day or
placebo to form four unique treatment groups: continuous
finasteride; finasteride in Year 1 → placebo in Year 2 →
finasteride in Years 3-5; placebo in Year 1 → finasteride in Years
2-5; and continuous placebo. Previous reports have documented the
results from the initial and first extension studies [9] and the
entire 5-year study period [10]. The present report summarizes the
results derived from the treatment group randomized to receive
placebo over the entire 5-year period of observation. The
intention-to-treat (ITT) population for this cohort (N = 68)
consisted of those patients who were randomized to receive
continuous placebo treatment over 5 years, who received at least
one day of randomized placebo therapy, and who had both baseline
and at least one post-baseline hair growth assessment, and
represents slightly less than the calculated 10% of 774 patients
randomized to placebo at baseline who were assigned at initial
randomization to continue on placebo treatment over 5 years.
Evaluation procedures
Four predefined efficacy endpoints provided a comprehensive
assessment of changes in scalp hair from baseline:
- 1) hair counts, obtained from color macrophotographs of
a 1-inch diameter circular (5.1 cm2) target area of
clipped hair (length 1 mm), centered at the anterior leading
edge of the vertex thinning scalp;
- 2) independent assessment of standardized clinical
global photographs of the vertex scalp by a panel of 3
dermatologists who were blinded to treatment and experienced in
photographic assessments of hair growth, using a standardized
7-point rating scale;
- 3) investigator assessment of scalp hair growth, using
the same standardized 7-point rating scale as used for the global
photographic assessment;
- 4) patient self-assessment of scalp hair growth or loss,
using a validated, self-administered hair growth questionnaire
[13].
Statistical analysis
The primary efficacy population assessed in this post-hoc analysis
was the cohort of patients who completed 5 years of placebo
treatment (completers cohort; N = 16). This cohort is a subset of,
and was compared with, the ITT placebo cohort (N = 68), which
consisted of patients randomized to receive continuous placebo
treatment over 5 years who received at least one day of randomized
placebo therapy and who had both baseline and at least one
post-baseline hair growth assessment.
Of the 16 patients in this completers cohort, 15 had hair count
and patient self-assessment data, 13 had investigator assessment
ratings, and 16 had global photographic assessment ratings at the
end of the fifth year.
Hair counts were assessed as the difference between the count at
each time point and the accompanying baseline count. Mean hair
count values were determined using the least squares means method.
Analysis of variance (ANOVA) was used for within-group
comparisons.
The expert panel assessment of global photographs and the
investigator assessment of scalp hair growth were analyzed by
comparison of mean rating scores at each time point, based on the
7-point rating scale of change in hair growth from baseline
(minimum value = –3.0 [greatly decreased]; maximum value = 3.0
[greatly increased]) [6]. Patient self-assessment was based on a
self-administered hair growth questionnaire that was designed and
validated for use in clinical trials of men receiving treatment for
hair loss [6, 7, 13]. Four questions in this questionnaire provided
a measure of treatment-related changes in hair growth (Questions 1,
2, 3, and 4) and three questions were designed to evaluate patient
satisfaction with appearance of hair since study start (Questions
5a, 5b, and 5c). Patient responses to each question at each time
point were taken as assessments of changes from baseline. Positive
changes indicated improvement in hair growth, negative changes
indicated deterioration, and a score of 0 (neutral response)
indicated no change from baseline; one question in the
questionnaire (Question 4) did not include 0 (neutral response) as
an option in the rating scale.
Results
Baseline characteristics
The baseline characteristics for all patients in the ITT cohort (N
= 68, or slightly less than 10% of 774 patients randomized at
baseline to placebo, who were assigned to receive placebo
continuously for 5 years), and those in the completers cohort (N =
16), are summarized in table 1. A
greater proportion of patients in the completers cohort had more
severe baldness (as measured by the Norwood/Hamilton
classification) at baseline compared to the ITT placebo cohort (81%
[13/16] versus 63% [43/68], respectively; table
1). Consistent with having more advanced hair loss at
baseline, men in the completers cohort had approximately 90 fewer
hairs (~10% lesser hair density) in the target area of the vertex
scalp at baseline compared to the ITT cohort (844 ± 225 vs. 931 ±
250 hairs, respectively; table 1).
Table 1 Baseline characteristics
|
Baseline characteristic
|
- Original
- intention-to-treat (ITT) placebo cohort*
- (N = 68)
|
- 5-year completers placebo cohort
- (N = 16)
|
|
Age (mean ± SD)
|
32 ± 5
|
33 ± 6
|
|
Age at which hair loss began (mean ± SD)
|
24 ± 4
|
23 ± 5
|
|
Baseline hair count (mean ± SD)
|
931 ± 250
|
844 ± 225
|
|
Number (%) of patients with a particular hair loss
pattern1:
|
|
II vertex
|
11 (16)
|
0
|
|
III vertex
|
14 (21)
|
3 (19)
|
|
IV
|
21 (31)
|
8 (50)
|
|
V
|
22 (32)
|
5 (31)
|
*Consists of patients randomized to receive placebo
treatment over 5 years who received at least one day of randomized
placebo therapy and who had both baseline and at least one
post-baseline hair growth assessment.
1According to a modified Norwood-Hamilton
scale.
Hair counts
Treatment with placebo led to progressive loss of scalp hair over 5
years (p ≤ 0.001 vs. baseline at Months 24, 36, 48, and 60),
culminating in a mean decrease from baseline of 239 hairs [95% CI =
304, 173] at 5 years (figure 1A). The change
from baseline in hair counts over time was generally similar
between the ITT and completers cohorts, with a slightly greater
hair loss rate observed in the completers cohort (figure 1A). There was no
significant change in hair count at Month 6, but thereafter,
patients displayed progressive hair loss over 2 years (the mean ±
SE change from baseline for the ITT (N = 68) and completers (N =
16) cohort was –15 ± 10 (N = 64) and –25 ± 22 (N = 15) hairs at
Month 12; and –51 ± 15 (N = 47) and –80 ± 27 (N = 14) hairs at
Month 24, respectively). Thus, during the first 2 years of the
study, treatment with placebo resulted in a loss of hair in the
target area of 5.8% for the ITT placebo cohort and 9.5% for the
completers cohort. This initial loss of scalp hair continued to
progress during the following 3 years (Months 25-60), resulting in
a 26.3% loss of hair in the target area by 5 years. Approximately
73% (11/15) of patients in the completers cohort had fewer hairs
after 1 year of treatment than at baseline, compared to 93% (13/14)
at Month 24. By 5 years, hair loss by hair count was demonstrated
in all men with hair count data who received placebo over the
entire 5-year observation period.
Global photographic assessment
Global photographic assessment demonstrated progressive loss of
visible hair over 5 years in placebo-treated men (figures 1B and 2). The
change from baseline in global photographic assessment score over
time was generally similar between the ITT and completers cohorts,
with a slightly greater rate of hair growth deterioration being
evident in the completers cohort. A significant decline in mean
score was observed in the completers cohort at Month 24 (p <
0.050 vs. baseline), which continued to worsen through 5 years (p
< 0.001 vs. baseline). The greatest deterioration in scalp hair
coverage occurred between Months 24 and 36, after which time the
mean change from baseline in global photographic assessment score
stabilized (figures 1B and 2).
After 1 and 2 years of placebo treatment, 13% (2/16) and 38% (6/16)
of men, respectively, in the completers cohort were rated as having
worsened (slightly, moderately, or greatly) compared to baseline.
By 5 years, 75% (12/16) of placebo patients demonstrated further
visible hair loss from baseline. Hair growth in the majority of
these men was rated as having slightly (25%; 4/16) or moderately
(31%; 5/16) decreased at 5 years. Only 1 patient (6%; 1/16) had
hair growth that was rated as slightly improved by global
photographic assessment after 5 years.
Investigator assessment
Based on the investigator assessment, treatment with placebo
resulted in progressive deterioration in hair growth over 5 years.
In contrast to scalp hair counts and global photographic
assessment, worsening of hair growth from baseline in placebo
patients was not demonstrated by investigator assessment until
after Month 24, with further deterioration occurring through 5
years (figure
3). At 5 years, 39% (5/13) of patients in the completers
cohort were assessed as worsened based on the investigator rating;
46% (6/13) of patients in the completers cohort were rated as
unchanged relative to baseline and 15% (2/13) were rated as
slightly improved. The change from baseline in the investigator
assessment score over time was generally similar between the ITT
and completers cohorts, with a slightly greater rate of
deterioration evident in the completers cohort (figure 3).
Patient self-assessment
In general, treatment with placebo over 5 years led to decreases in
mean self-assessment scores for questions relating to patient
satisfaction with appearance of hair (Questions 5a-c), and size of
the bald spot (Question 1) for both ITT and completers cohort.
Overall, a greater proportion of placebo-treated patients reported
either dissatisfaction or no change in the appearance of their hair
at 5 years (table 2). A small percentage
of patients reported satisfaction with the appearance of their hair
(Questions 5a-c) or improvements in the size of their bald spot
(Question 1) at 5 years. In contrast, a greater percentage of
patients had a more positive assessment of their overall appearance
of hair (40%), hair growth (40%), and rate of hair loss (67%) at 5
years (Questions 2, 3, and 4, respectively).
Table 2 Percentage of placebo-treated patients with
negative, neutral, or positive self-assessment ratings at Year
5†
|
Patient self-assessment questionnaire
|
Negative
|
Neutral
|
Positive
|
|
Q1: Size of bald spot
|
8/15 (53%)
|
4/15 (27%)
|
3/15 (20%)
|
|
Q2: Appearance of hair
|
4/15 (27%)
|
5/15 (33%)
|
6/15 (40%)
|
|
Q3: Growth of hair
|
4/15 (27%)
|
5/15 (33%)
|
6/15 (40%)
|
|
Q4: Slowing of hair loss
|
5/15 (33%)
|
--
|
10/15 (67%)
|
|
Q5a: Satisfaction with frontal hairline
|
7/15 (47%)
|
7/15 (47%)
|
1/15 (7%)
|
|
Q5b: Satisfaction with hair on top
|
7/15 (47%)
|
6/15 (40%)
|
2/15 (13%)
|
|
Q5c: Satisfaction with hair overall
|
6/15 (40%)
|
6/15 (40%)
|
3/15 (20%)
|
†Fifteen of the 16 patients who completed 5 years had
self-assessment ratings at the final timepoint.
--Not applicable; Question 4 did not include 0
(neutral score) as an option in the rating scale.
Discussion
The pooled placebo data from the two Phase III studies included
herein represent the longest reported observations in
placebo-treated men with AGA and provide physicians with valuable
information regarding the progression of hair loss in these
patients. Although the number of patients randomized to placebo who
remained in these studies through 5 years was limited, our results
are in agreement with those of previous placebo-controlled clinical
trials of shorter duration, which demonstrated that lack of
treatment led to progressive loss of scalp hair in men with AGA [7]
Analysis of data from the completers cohort revealed the continued
loss of scalp hair by all predefined hair growth endpoints over 5
years. The observed deterioration in hair density based on hair
counts in a representative target area was associated with visibly
significant deteriorations in scalp coverage as assessed by
comparison of pre- and post-treatment global photographs by an
expert panel. Although the completers cohort had a more severe hair
loss profile at baseline compared to that of the ITT cohort, the
deterioration from baseline over time for all hair growth endpoints
was generally similar between these two cohorts. Hair count
methods, such as the one used in our studies, generally detect
alterations in hair density more rapidly than other methods, due to
a high sensitivity for detecting change. Nonetheless, the
relationship between assessment of hair loss as measured by hair
count and by global photographic assessment observed in this
analysis is consistent with a recent report demonstrating that
quantification of hair density using macrophotographic methods
similar to our own may offer advantages over other hair count
methods, such as digital image analysis [14].
Assessment of hair growth by investigators also demonstrated a
progressive decline in scalp hair growth over 5 years. The
investigators’ assessments were based on observations of patients
seen in the clinic and provided a clinical assessment of the
patient’s scalp hair since the study start. In general, the
investigators reported continued hair loss between Years 2 and 5 in
placebo-treated patients. However, no deterioration in hair growth
was observed with this instrument at earlier time points (up to 2
years) in the study. This finding most likely represents a placebo
effect, which is not unusual in double-blind, placebo-controlled
clinical trials. This effect can be explained in part by
investigator recall and in part by expectation biases involving
physician-patient interaction where both parties anticipate
improvement in hair growth, obscuring the detection of ongoing hair
loss as documented by more objective, and potentially more
sensitive endpoints, such as hair count and global photographic
assessment. The use of different endpoints should be considered
when comparing results across studies, given the differences
observed in hair growth assessments using the predefined endpoints
in our study.
Patient self-assessment of hair growth in these studies provides
a mechanism by which each patient evaluated changes in hair growth
under controlled and blinded conditions. Placebo patients generally
perceived a lack of improvement or a worsening of their scalp hair
growth with time, with the majority of patients reporting either no
change or dissatisfaction with the appearance of scalp hair after 5
years (Questions 1-3 and 5a-c). The larger than expected proportion
of patients self-reporting neutral scores was probably due to
expectation bias obscuring their own observation of ongoing hair
loss, as patients did not know whether they were receiving active
therapy or placebo throughout the 5-year studies.
As is often seen in long-term clinical trials, a large
proportion (44/68 or 76.5%) of the randomized patients in the ITT
cohort withdrew from the studies before completing five years of
placebo treatment. The 16 patients who completed five years of
placebo treatment were characterized by lower baseline hair counts
compared to the cohort of patients in the ITT cohort who did not
complete 5 years of treatment. The selective and potentially
non-representative nature of the completers cohort is a limitation
of the current analysis and should be taken into consideration when
relating the findings of the present analysis to the general
population of men with AGA.
It should also be noted that patients who completed the study
were recruited from a clinical population of men seeking treatment
for hair loss who were required to maintain compliance with a study
protocol (e.g., willingness to participate in routine examinations,
agreement not to alter hair style during the study, etc.). Thus,
this report documents the long-term progression of hair loss in a
relatively small number of untreated men with AGA who were seeking
treatment for hair loss, and may not reflect fully the natural
history of hair loss in a general population of men with AGA.
Nonetheless, our findings are in agreement with those from
population-based epidemiologic studies demonstrating that hair loss
is a progressive condition that worsens over time. A 2-year,
longitudinal analysis of hair loss in a community-based population
of untreated men [15] examined the potential likelihood of further
hair loss over time from both the perspective of untreated men with
AGA and that of a trained, independent observer. Both the
self-reported and trained observer ratings in this epidemiologic
study used the same 7-point rating scale of change in scalp hair
growth as that used in the analyses presented in this report. The
epidemiological study reported a gradual, measurable change in the
magnitude of hair loss for all men, regardless of age, based on
observer rating data. Similarly, a trend of increasing hair loss
over time was reported for all age groups based on patient
self-report data. In contrast to our 5-year studies [16], no
placebo effect was reported in the epidemiological study. This was
presumably due to the fact that, unlike in our clinical trials,
patients and investigators in the epidemiological study were not
receiving active treatment or placebo under blinded conditions and
thus had no reason to anticipate treatment-related improvements in
hair growth.
In summary, treatment of men with AGA with placebo over 5 years
led to progressively decreased scalp hair. These findings are in
agreement with results from population-based epidemiology studies,
which are more representative of the general population and less
prone to placebo effects, and lend support to the conclusion that
progressive hair loss is an inevitable consequence of AGA and that,
if left untreated, will lead to cosmetic worsening over time.
Acknowledgements
We are grateful to Dr. Alan G. Meehan for his assistance with this
manuscript. This study was supported by a grant from Merck &
Co., Inc. Keith D. Kaufman, Cynthia J. Girman, Elizabeth M. Round,
Amy O. Johnson-Levonas, Arvind K. Shah, and Jennifer Rotonda are
employees of and hold stock in Merck & Co., Inc.
References
1 Hamilton JB. Male hormone stimulation is prerequisite and an
incitant in common baldness. Am J Anat 1942; 71: 451-80.
2 Hamilton JB. Patterned loss of hair in man: types and
incidence. Ann N Y Acad Sci 1951; 53: 708-28.
3 Price VH. Testosterone metabolism in the skin. Arch
Dermatol 1975; 111: 1496-502.
4 Whiting DA. Diagnostic and predictive value of horizontal
sections of scalp biopsy specimens in male pattern androgenetic
alopecia. J Am Acad Dermatol 1993; 28: 755-63.
5 Olsen EA. Androgenetic alopecia. In: Olsen EA, ed.
Disorders of hair growth: diagnosis and treatment. New York:
McGraw-Hill, Inc., 1994: 257-83.
6 Lopez-Bran E, Robledo A, Aspiolea F,
Moragas JM, Yu A, Quintanilla E, Sambrico F,
del Pino J. Multicenter comparative study of the efficacy of
topical 2% minoxidil (Rogaine) versus placebo in the treatment of
male baldness. Adv Ther 1990; 7: 159-68.
7 Olsen EA, Buller TA, Weiner S, Delong ER.
Natural history of androgenetic alopecia. Clin Exp Derm 1990; 15:
34-6.
8 Rushton DH, Ramsay ID, Norris MJ,
Gilkes JJH. Natural progression of male pattern baldness in
young men. Clin Exp Derm 1991; 16: 188-92.
9 Kaufman KD, Olsen EA, Whiting D, Savin R,
DeVillez R, Bergfeld W, Price VH, Van-Neste D,
Roberts JL, Hordinsky M, Shapiro J,
Binkowitz B, Gormley GJ. Finasteride in the treatment of
men with androgenetic alopecia. Finasteride Male Pattern Hair Loss
Study Group. J Am Acad Dermatol 1998; 39: 578-88.
10 Finasteride Male Pattern Hair Loss Study Group. Long term (5
year) multinational experience with finasteride 1 mg in the
treatment of androgenetic alopecia. Eur J Dermatol 2002; 12:
38-49.
11 Norwood O. Male pattern baldness: classification and
incidence. South Med J 1975; 68(11): 1359-65.
12 Takashima I, Iju M, Sudo M. Alopecia
androgenetica - its incidence in Japanese and associated
conditions. In: Orfanos CE, Montagna W, Stüttgen G,
eds. Hair research. Status and future aspects. New York:
Springer-Verlag, 1981: 287-93.
13 Barber BL, Kaufman KD, Kozloff RC,
Girman CJ, Guess HA. Hair growth questionnaire for use in
the evaluation of therapeutic effects in men. J Dermatol Treat
1998; 9: 181-6.
14 Aktan S, Akarsu S, Ilknur T,
Demirtasoglu M, Ozkan S. Quantification of female pattern
hair loss: a study in a Turkish population. Eur J Dermatol 2007;
17: 321-4.
15 Rhodes T, Girman CJ, Savin RC,
Kaufman KD, Guo S, Lilly FR, Siervogel RM,
Chumlea WC. Prevalence of male pattern hair loss in 18-49 year
old men. Dermatol Surg 1998; 24: 1330-2.
16 Kaufman KD, Rotonda J, Shah AK,
Meehan AG. Long-term treatment with finasteride 1 mg
decreases the likelihood of developing further visible hair loss in
men with androgenetic alopecia (male pattern hair loss). Eur J
Dermatol 2008; 18(4): 400-6.
|
Printable version
Version PDF
