Fluridil,
a Rationally Designed Topical
Agent for Androgenetic Alopecia:
First Clinical Experience
MILOS SOVAK, MD, ALLEN L. SELIGSON, PHD, RENATA
KUCEROVA, MD, MARIE BIENOVA, MD, MARIAN HAJDUCH, MD, AND MILAN BUCEK, MD
*Radiology Research, University of California, San Diego, and Biophysica
Foundation, La Jolla, California, Department of Dermatology and Laboratory for
Experimental Medicine, Pediatrics Department, Palacky University of Olomouc,
Czech Republic
BACKGROUND. Fluridil, a novel topical
antiandrogen,
suppresses ( Comment: not blocks ) the
human androgen receptor. While highly hydrophobic and hydrolytically
degradable, it is systemically nonresorbable. In animals, fluridil
demonstrated high local and general tolerance.
OBJECTIVE. To evaluate
the safety and efficacy of a topical anti androgen, fluridil, in male
androgenetic alopecia.
METHODS. In 20 men, for
21 days, occlusive forearm patches with 2, 4, and 6% fluridil, isopropanol,
and/or vaseline were applied. In 43 men with androgenetic alopecia (AGA),
Norwood grade II.-Va, 2% fluridil was evaluated in a double-blind,
placebo-controlled study after 3 months clinically by phototrichograms,
hematology, and blood chemistry including analysis for fluridil, and at 9
months by phototrichograms.
RESULTS.Neither
fluridil nor isopropanol showed sensitization /
irritation potential, unlike vaseline. In all AGA subjects, base line anagen /
telogen counts were equal. After 3 months, the average anagen percentage did
not change in placebo subjects, but increased in fluridil subjects from 76% to
85%, and at 9 months to 87%. In former placebo subjects, fluridil increased
the anagen percentage after 6 months from 76% to 85%. Sexual functions,
libido, hematology, and blood chemistry values were normal throughout, except
that at 3 months, in the spring, serum testosterone increased within the
normal range equally in placebo and fluridil groups. No fluridil or its
decomposition product, BP-34, was detectable in the serum at 0, 3, or 90 days.
CONCLUSION.Topical
fluridil is nonirritating, nonsensitizing, nonresorbable, devoid of systemic
activity, and anagen promoting after daily use in most AGA males.
M. SOVAK, MD, A. L. SELIGSON, PHD, R. KUCEROVA, MD, M. BIENOVA, MD, M. HAJDUCH,
MD, AND M. BUCEK, MD
HAVE INDICATED NO SIGNIFICANT
INTEREST WITH COMMERCIAL SUPPORTERS.( Comment: Dr. Sowak has definetely a
commercial interest because he is a leading head of Biophysica, CA )
The Pathophysiology of androgenetic alopecia (AGA) is not yet fully understood
and there is no optimal therapy.
1 AGA is androgenetic because it
can be either an autosomal or polygenic trait,
2 and because
whatever the underlying mechanism, androgens play a pivotal role in first
promoting and later suppressing hair growth. Androgens act in numerous ways,
but the common denominator is binding to a protein transcription factor, the
androgen receptor (AR). The resulting complex, after entering the cell
nucleus, binds to DNA to initialize the pathway via RNA polymerase II.
3
A specific and limited interference with androgens thus should be useful in
AGA treatment.
To this end, inhibitors of type II 5a -reductase, an enzyme that converts
testosterone into much more potent dihydrotestosterone (DHT), were explored.
Finasteride at 1 mg/day in men was shown to increase the growth rate and
thickness of hair.
4 Side effects such as decreased libido and
erectile function are reportedly minimal and reversible.
1 Limited
to blocking DHT formation, finasteride does not affect other androgens and has
to be taken daily. Although subjects with a congenital 5a-reductase deficiency
do not manifest any side effects,
5 there is no proof yet that
systemic manipulation of hormone balance in normal subjects would be
innocuous.
Developed originally for the treatment of prostate cancer, the nonsteroidal
oral antiandrogens were considered for topical use in AGA. Antiandrogens
contain a phenyl moiety substituted with a nitro or cyano adjacent to a
trifluoromethyl group, which closely mimics the steroidal androgen’s
binding region to ARs (Figure 1).
6 Systemic antiandrogens block all AR
indiscriminately and they adversely affect libido and male sexual functions.
Flutamide applied topically to bald human scalp grafted into nude mice was
shown to induce hair growth and longer hair shafts.
7 It is,
nevertheless, probable that in humans, topical flutamide induces systemic
antiandrogenic effects since at least 16% is transcutaneously resorbed.8
Preclinical data of another oral antiandrogen, an N-substituted aryl hydantoin,
RU-58841 (Figure 1), suggested topical safety since the
cutaneous resorption was low, 9,10 however, one of the metabolites
proved stable and strongly anti androgenic, 11 (
comment:
this is contradicted by the cited study ) and further development was
abandoned. Another agent used in AGA is minoxidil, an antihypertensive drug
which, by a mechanism yet to be clarified, increases hair thickness and anagen
count. Some minoxidil is resorbed systemically, but the effect on hemodynamics,
if any, is very low.
12
|
|
| Figure 1. Testosterone,
dihydrotestosterone, flutamide, nilutamide, and RU-58841. |
Design
and Primary
Evaluation
of Fluridil
Androgens are continuously synthesized
from the circulating pool of dehydroepiandrosterone in a balanced way.
5a-reductase converts testosterone into DHT, while aromatases convert
testosterone into estradiol. An approach to AGA treatment should neutralize
the effects of all androgens, not only DHT, on the hair and its appendages,
while maintaining an unperturbed androgen physiologic balance. As an
alternative, selectively inhibiting AR binding in the scalp thus appears
preferable. In reviewing the subject, Sawaya and Hordinsky
13 noted
that "the antiandrogen will also have to be safe, have minimal
systemic absorption, be locally metabolized, and have minimal or no effects
on other target tissues, sex organs, and gonadotropin levels. Such an
antiandrogen or compound may still be thought of as a ‘fantasy’ drug."
It occurred to us that a compound would meet these criteria if it, upon
encountering the aqueous milieu of the microcirculation, would degrade
into nontoxic, excretable fragments devoid of antiandrogenic activity.
Incorporation of a hydrolyzable bond seemed a plausible solution, and we
noted that the electron-withdrawing nature of perfluoroalkyl moieties
destabilize the otherwise stable carboxamide bond. We also surmised that
a highly hydrophobic moiety at the terminal region of the molecule would
provide for strong binding. Applying this design, we synthesized a series
of
analogues of which 2-hydroxy-2-methyl-N-[4-nitro-3 -(trifluoromethyl)phenyl]
-3 -(2,2,2-trifluoroacetylamino) propanamide—fluridil (Figure 2)—was
selected for further development.
14,15
Synthesis, analysis, and preclinical evaluation of fluridil is described in
detail elsewhere
14 (Seligson AL and Sovak M, manuscript
submitted). Briefly, to simulate the shelf life of 2% fluridil in anhydrous
isopropanol, an accelerated stability study was conducted at 50°C for 8
weeks, equivalent to 5 years at 20°C.
16 Only traces of
decomposition were observed. After 8 months at ambient temperature, no
decomposition could be identified by high-performance liquid chromatography
(HPLC).
Fluridil’s in vitro biodegradability in human serum (0.5 mg/mL) was
assessed after incubation at 38°C; its half life was found to be about
6 hours and only traces were seen after 48 hours. In vivo, the decomposition
must be much faster, as no fluridil or its decomposition products could
be
found in the serum. The oral toxicity of fluridil was determined by median
lethal dose (LD50 using doses of 1500, 2000, and 2500 mg/kg. In NMRI mice,
LD50 was 2871.7 mg/kg (males) and 2232.0 mg/kg (females). In Wistar rats,
the LD50 could not be calculated since only one male rat died (at the 1500
mg/kg dose) and none of the female rats died, suggesting an LD50 greater
than 2500 mg/kg.
The systemic toxicity of fluridil was orientationally evaluated in mice by
seven daily intraperitoneal injections of doses increased from 300 to 500
mg/kg. The LD50 was estimated as 450 mg/kg and the maximum tolerated dose (MTD)
as 300 mg/kg. Fluridil therefore does not fall into the category of a
harmful or toxic substance as defined by the National Institute of
Environmental Research.
17
 |
| Figure 2. Fluridil and its hydrolytic
decomposition. |
The degradation products of fluridil are BP-34 and
trifluoroacetic acid (Figure 2). In mice, intraperitoneal 100—300
mg/kg BP-34 produced no mortality. Morbidity was observed at 300 mg/kg,
suggesting
a murine MTD of about 250 mg/kg, corresponding to about 25 mg/kg or 1730
mg in an average human. Tolerance of trifluoroacetic acid is reportedly
high.
18
The interaction of fluridil and/or BP-34 with human AR was studied using
immortalized human LNCaP cells and a standard Western blot. At 10 mm, AR
suppression by the antiandrogens bicalutamide and hydroxyflutamide was
only 2-3% compared to 97% for fluridil. The degradation product BP-34 was
inactive, and the results were validated in a transfected cell line which
expresses stable levels of human ARs and MMTV-CAT sequences, as described
by
Fuhrmann et al.
19 Since BP-34 from 0.3 nM/L to 1 mM/L in the
assay, using 2 nm of tritium-labeled R-1881, demonstrated no measurable
affinity (Thierauch KH, Schering AG, Berlin, personal communication; we
thank Dr. Thierauch for kindly providing the data to us), it is not expected
to be an antiandrogen.
A standard contact sensitization test was conducted
on 20 guinea pigs exposed once a week to the test material soaked into
filter paper and applied for 6 hours. After 2 weeks, fluridil or its
vehicle (anhydrous isopropanol) applied with an occlusive patch elicited
no allergic skin or adverse systemic responses. A patch test on covered
and noncovered sites in 10 rabbits found fluridil to be equal to its
vehicle according to the irritation index (with nonirritability defined as
less than 0.4), which was 0.14 for the noncovered and 0.11 for the covered
group, indicating a negligible potential to irritate rabbit skin.
To simulate the intended human cosmetic use (0.6 mg/kg/day), fluridil was
applied topically twice a day for 10 days on two separate 10 cm2 areas of
closely shaved skin in four rabbits. Serum was obtained at 2, 5, and 21
hours after the first application, and then once every other day. Using
spiked and blank serum to ascertain linearity, the limits of reliable
quantification by HPLC of fluridil and BP-34 were approximately 20 ng/ml,
about three times the detectability level. Peaks representing either
compound could not be discerned in the chromatograms, although the
cutaneous absorption in rabbits is known to be about five to six times
greater than in humans.
20 The quantity of fluridil, if
resorbed systemically and remaining intact, thus would have to be less
than its detectability limit of 5 ng fluridil/ml serum, and therefore, at
most, its total amount in the circulation could not exceed 0.25 mg/kg,
that is, 17.5 mg in a 70 kg human. Ames reversion assay determined that
fluridil and its metabolites have no mutagenic potential.
21
Based on experimental efficacy proof, acceptable topical and systemic
tolerance, and extremely low-if any-systemic resorption from the skin, we
concluded that fluridil as a cosmetic topical hair growth agent qualified
for clinical safety and efficacy evaluation.
Materials and Methods
Cumulative Irritancy Assay in Human Subjects
(This
study was conducted by the laboratories of Professor H. Maibach, MD, San
Francisco, CA.) A group of 20 men, 18-60 years of age, were recruited,
examined, and found free of active skin pathology. Medical histories and
informed consents were obtained from all subjects. The study protocol of
Phillips et al.
22 was used. Briefly, 2, 4, and 6% fluridil in
anhydrous isopropanol was applied 5 days a week (excluding weekends) for
21 days to the same site using occlusive patches. The patch was an
occlusive plastic chamber held in place by paper tape. As controls,
isopropanol and Vaseline Intensive Care Lotion (Chesebrough-Ponds) were
used. There were 15 days of readings that were made at each removal of the
patch using an erythema score scale of 0-4.
Efficacy and Safety of Fluridil in Male AGA
Two percent fluridil was formulated in
anhydrous isopropanol under Good Manufacturing Practice (GMP) and filled
into 5O ml class I glass penicillin vials capped with a polypropylene/
Teflon-coated, inert stopper. identical vials contained neat anhydrous
isopropanol as placebo. A 2 ml syringe was provided with each coded vial.
Forty-three male subjects of phototype II-IV and AGA of degree II-Va (II
and IIa [ 2 patients], III [ 10 patients], IIIa [1 patient], IV [9
patients], IVa [ 6 patients], V [ 5 patients], and Va [ 8 patients])
according to the Hamilton-Norwood classification were recruited from the
Department of Dermatology, Ambulatory Services (20-56 years of age,
average 33.6 years) for this double-blind study. The institutional review
board approved the study and all subjects signed an informed consent. None
of the subjects received systemic or local alopecia therapy for at least 1
month prior to the recruitment.
| Table 1. Hematologic and
Blood Chemistry Parameters |
|
| Hematology |
Blood Chemistry |
|
| Leukocytes |
Urea |
| Erythrocytes |
Creatinine |
| Hemoglobin |
ALT |
| Thrombocytes |
AST |
| Lymphocytes |
ALP |
| Monocytes |
Ca2+ |
| Granulocytes |
Mg2+ |
| Eosinophils |
Na+ |
| Basophils |
K+ |
| Total Protein |
CI- |
|
The subjects were randomized: 23 received fluridil
and 20 placebo. On day 0, a baseline phototrichogram was obtained by
closely shaving a permanently tattooed round area 1.2 cm in diameter,
corresponding to about 1.1 cm2.23 The area was photographed with a
close-up lens using a digital camera, a global view of the scalp was also
recorded, and the data stored. On day 3, the area was again imaged. The
follicles/hairs were counted manually using a televised display; anagens
and/or telogens were calculated as a percentage of the total. On day 3, 10
ml of venous blood was taken for hematology and blood chemistry (Table 1).
Subjects were instructed to keep the fluridil vials
in a cool, dry place and open them only to draw 2 ml to be applied to the
skin by circular movement of a blunt syringe over the scalp before bed.
The subjects were also instructed to complete a clinical questionnaire, to
avoid direct sun exposure, and to reduce the use of liquid shampoos. They
were advised to limit hair washes to twice a week using only warm water
and were provided with dry shampoo to use ad libitum. ( comment: this is a
completely unrealistic scenario but foundation for the study success - see
later comment )
On day 90, the tattooed scalp area was again closely shaved and imaged,
and hematology and blood chemistry reevaluated. The subjects were queried
specifically about any changes in sexual function and libido. The study
protocol envisioned follow-up at 6, 9, and 12 months for clinical
assessment, at 9 and 12 months for phototrichograms, and at 12 months for
final blood chemistry and hematology. Further, a provision was made for
breaking the code at 90 days should AGA manifestly improve in at least 40%
of all subjects, to offer the fluridil treatment to all placebo subjects,
and to continue the study as an open label.
Blood collected from subjects on day 0, 3, and 90 was analyzed for
fluridil and its hydrolysis product, BP-34, in the following fashion: For
fluridil, 1.0 ml samples were treated with 1.0 ml 0.05 M NaH2PO4 (pH 3.5),
1.0 mg/ml flutamide (internal standard) and loaded on a 200 mg phenyl
extract clean column pretreated with acetonitrile (2 X 1 ml), methanol (2
X 1 ml), and 1% methanol in water (2 X 1 ml). After elution of the serum
sample, the cartridge was washed with 1% methanol in water (2 X 1 ml) and
the analyte and internal standard eluted with methanol (2 x 1 ml). The
eluent was filtered through a 0.45 mm polytetrafluoroethylene (PTFE)
filter. The samples were evaporated under N2 flow and reconstituted in
acetonitrile for HPLC analysis on a Waters Symmetry C-18 column at a flow
rate of 1.0 mI/min using a 0.005 M KH2PO4 buffer (pH 3.5)/acetonitrile
gradient at 290 nm. Using spiked and blank serum samples, a linear
concentration relationship was observed for fluridil with a quantification
limit of 20 ng/ml and a detection limit of about 5-7 ng/ml.
For BP-34, 1.0 ml samples were treated with 600 ng/ml flutamide (internal
standard) and loaded on a 200 mg extraction column pretreated with 1 ml
acetonitrile, 1 ml methanol, water (2 X 1 ml), 1 ml 1.0 N NaOH, and 0.05 m
NaH2PO4 (pH 3.0) buffer (4 x 1.0 ml or until eluent from the cartridge is
acidic). After elution of the serum sample, the cartridge was washed with
1 ml 95% 1.0 mm HCl methanol, 1 ml 95% 0.1 m HCl/5% methanol and the
analyte and internal standard eluted with 1 ml 90% methanol/ 10% 1.0 m HCl.
The samples were evaporated under N2 flow and reconstituted in mobile
phase for HPLC analysis on a Waters Symmetry C-18 column at a flow rate of
1.0 ml/min using a 0.005 m KH2PO4buffer (pH 4.0)/acetonitrile gradient at
290 nm. Using spiked and blank serum samples, a linear concentration
relationship was observed for fluridil with a quantification limit of 20
ng/ml and a detection limit of about 5 ng/ml.
Results
Cumulative Irritancy Assay in Human Subjects
Fluridil, in all three concentrations in
its vehicle, isopropanol, as well as neat isopropanol, proved
nonirritating to human forearm skin. Response to the induction was
minimal, with a score of 0.5 for isopropanol and the three fluridil
concentrations, and up to 4 for the vaseline control. On the delayed
challenge, only 6% fluridil induced macular erythema, and then only to a
slight degree (1.4 of the attainable score). Two percent and 4% fluridil
and/or isopropanol proved nonirritating and devoid of sensitization
potential.
Efficacy and Safety of Fluridil in Male AGA
Blood Chemistry, Hematology, and General
Observation. At 90 days, substantial individual variation and an increase
in average serum testosterone, statistically equal in both groups, and
within normal range, was seen (3.8 nM/L in the fluridil group and 3.2 nM/L
in the placebo group) (Table 2). Upon direct questioning and using a
questionnaire, neither placebo nor fluridil subjects indicated any change
in their health, libido, or sexual performance, nor were there any
statistically significant differences in their hematology and blood
chemistry parameters.
| Table 2. Serum Testosterone
Levels After Topical Application of Fluridil and/or Placeboa |
|
|
Testosterone
levels (nM/L)
|
| |
|
|
Fluridil
(day 0) |
Fluridil
(day 90) |
Placebo
(day 0) |
Placebo
(day 90) |
Fluridil
(day 0) |
Placebo
(day 0) |
Fluridil
(day 90) |
Placebo
(day 90) |
|
| n |
23 |
23 |
20 |
20 |
23 |
20 |
23 |
20 |
| Mean |
15.53 |
19.34 |
13.75 |
16.91 |
15.53 |
13.75 |
19.34 |
16.91 |
| SD |
9.5 |
8.29 |
7.24 |
7.27 |
9.5 |
7.24 |
8.29 |
7.27 |
| t
test paired |
|
0.0046 |
|
0.0150 |
|
0.2463 |
|
0.1555 |
|
n
= number of subjets.
p < 0.05 signifies statistically significant difference between
two groups.
aNote: 12-month testosterone data recently obtained were
15.15 +/- 5.64 nM/L for the placebo and 17.88 +/- 7.85 nM/L for the
fluridil group. |
Hair Growth. The
phototrichogram data show no statistical difference in
the average percentage of anagens/telogens in the fluridil
and
placebo groups
at the study’s outset, indicating the clinical material’s
homogeneity (Table 3).
| Table 3. Initial
Trichologic Assessment of AGA in Male Subjects Randomized into
Fluridil and Placebo Groups. |
|
|
Anagens
|
Telogens
|
| . |
|
|
|
Placebo
(0 months) |
Fluridil
(0 months) |
Placebo
(0 months) |
FLuridil
(0 months) |
|
| n |
20 |
23 |
20 |
23 |
| Mean |
70.13 |
75.68 |
29.51 |
25.41 |
| Median |
71.45 |
78.1 |
30.6 |
19.8 |
| SD |
13.26 |
14.27 |
11.42 |
12.95 |
| t
test paired |
0.2189 |
|
0.3301 |
|
|
n
= number of subjets.
p < 0.05 signifies statistically significant difference between
two groups.
Anagens as a percentage of total hairs/follicles counted. |
Since practically all participants in the study
followed the fashion of shorn hair, it was not possible to evaluate
effluvium clinically, but in the fluridil group three subjects observed a
substantial decrease. The effects of fluridil and/or placebo at 3 months,
derived from phototrichograms, and comparison of the two groups is shown
in Table 4. It can be seen that, after 3 months, there was no significant
difference in the anagen count/telogen count in the placebo group, but an
increase in anagen count and a decrease in telogen count in the fluridil
group was significant and so was the differnect between fluridil and
placebo.
| Table 4. Effect of
Fluridil and Placebo on the Average Anagen/Telogen Counts as a
Percentage of the Total Hair Count |
|
|
Anagens
|
Telogens
|
| . |
|
|
| Panel
A |
Placebo
(0 months) |
Placebo
(0 months) |
Fluridil
(0 months) |
Fluridil
(3 months) |
Placebo
(0 months) |
Placebo
(0 months) |
Fluridil
(0 months) |
Fluridil
(3 months) |
|
| n |
20 |
20 |
23 |
23 |
20 |
20 |
23 |
23 |
| Mean |
70.13 |
77.25 |
75.68 |
85.09 |
29.51 |
21.87 |
25.41 |
14.61 |
| Median |
71.45 |
79.8 |
78.1 |
87.5 |
30.6 |
21.65 |
19.8 |
12.3 |
| SD |
13.26 |
11.17 |
14.27 |
8.73 |
11.42 |
11.46 |
12.95 |
9.189 |
| t
test paired |
0.0738
|
0.0099
|
0.0413
|
0.0021
|
|
|
Anagens
|
Telogens
|
|
|
|
|
| . |
|
|
. |
. |
. |
. |
| Panel
B |
Placebo
(3 months) |
Fluridil
(3 months) |
Placebo
(3 months) |
Fluridil
(3 months) |
|
|
|
|
| . |
|
|
. |
. |
. |
. |
| n |
20 |
23 |
20 |
23 |
|
|
|
|
| Mean |
77.25 |
85.09 |
21.87 |
14.61 |
|
|
|
|
| Median |
79.8 |
87.5 |
21.65 |
12.3 |
|
|
|
|
| SD |
11.17 |
8.73 |
11.46 |
9.189 |
|
|
|
|
| t
test unpaired |
0.0136
|
0.0264
|
|
|
|
|
|
A
double-blind study of male subjects with AGA, randomized into two
groups, evaluated by phototrichograms before and after 3 months
(panel A) and compared to each other at 3 months (panel B).
n = number of subjets.
p < 0.05 signifies statistically significant difference between
two groups.
Anagen/telogen averages as a precentage of ttoal hairs/follicles
counted. |
In accordance with the study protocol, the obvious
improvement seen clinically and gleaned from the phototrichograms in
the majority of the participants justified discontinuation of placebo and
administration of fluridil to all participants for another 6 months. 
