ARTICLE
Tetracyclines are the most common oral antibiotic treatments prescribed
for acne vulgaris and the agents of choice for moderate to moderately
severe inflammatory acne that fails to respond to topical therapy [1].
Lymecycline (Tetralysal®) and minocycline (Minocin®)
are semi-synthetic tetracycline antibiotics and as a class, display similar
efficacy, but differ in their safety profile [2]. Lymecycline is distributed
to the skin, penetrates in the lipid-rich layers of the epidermis, where
it binds to epithelial cells before being excreted in the sebum where
it acts to reduce the inflammatory effect of "oleosa-produced acid"
[3]. Previous research work on minocycline has demonstrated its efficacy
and safety in acne vulgaris [4-6]. In addition to the typical adverse
effects of tetracyclines, minocycline may be at the origin of vestibular
disturbances and blue-black hyperpigmentation affecting the skin, mucous
membranes and permanent teeth. It has also been implicated in immunologically-mediated
reactions, with reports of systemic lupus erythematosus and auto-immune
hepatitis, hypersensitivity syndrome, Sweet's syndrome and eosinophilic
pneumonia [7-10].
The Cochrane Database System Review [11] evaluated the efficacy and safety
of minocycline in acne vulgaris throughout 27 randomised controlled
trials and concluded that even if minocycline is widely perceived by clinicians
to have a faster onset of action than oxytetracycline, none of the studies
has shown any important clinical difference between minocycline and other
tetracyclines, but pointed out the risk of minocycline related adverse
events such as lupus erythematosus syndrome.
A multicentre, randomized, double-blind and double-dummy study conducted
in 144 patients confirmed that lymecycline and minocycline were equally
effective in reducing acne lesion counts, with no significant safety difference
between the two drugs [12]. This study, however, compared the two drugs
at the treatment previously recommended in France: lymecycline 300 mg/day
for 2 weeks then 150 mg/day for 10 weeks; and minocycline
100 mg/day for 2 weeks then 100 mg every other day for
10 weeks. Even though the efficacy and safety objectives have been
met by both drugs the trial did not take into consideration the costs
for such a treatment.
To investigate the drug regimen of 300 mg day of lymecycline, currently
administered in the United Kingdom and Belgium along with a cost analysis,
we conducted a randomized, investigator masked study comparing lymecycline
and minocycline in two parallel groups of 134 patients with acne
vulgaris.
Materials and methods
To be considered for inclusion in this multicenter, randomized, investigator
masked, parallel group trial, patients had to meet the following criteria:
male or female between the ages of 12 and 30 years, with at
least 15 and at most 120 inflammatory facial lesions (papules,
pustules, nodules) including at most 2 facial nodules (diameter > 1 cm),
a maximum of 60 non-inflammatory facial lesions (open and closed
comedones) and an acne severity grade between 1 and 5 (Leeds grading
scale [13]). Women of childbearing age were to use contraception throughout
the study and for one month after completing the trial. Women on oral
contraceptives were to have been using the same method for 3 months
before enrolment, or for at least 12 months for contraceptive pills
containing cyproterone acetate. Use of cosmetics was allowed during the
course of the study. Contraceptives and cosmetics had to be listed as
concomitant medication.
Patients could not be included if they were pregnant or lactating women,
suffering from acne conglobata, acne fulminans or secondary acne. Moreover,
patients using topical anti-acne or anti-inflammatory medications on the
face or systemic anti-inflammatory drugs or antibiotics, with the exception
of short-courses of penicillin during the previous 4 weeks, or using
systemic retinoids during the previous 6 months, were to be excluded
from the trial. Patients with known renal or hepatic disease, known or
suspected allergy to tetracyclines, known or suspected systemic lupus
erythematosus were also not to be considered for this study.
Qualified patients were randomized to receive either oral lymecycline
300 mg (Tetralysal® 300) or oral minocycline modified
release 100 mg (Minocin® MR). Both treatments were administered
for 12 consecutive weeks as one capsule once a day and patients were
asked to ingest the treatment before, during or after meals but always
about the same time of the day. Randomization was balanced by center and
by block.
Patients were not to take any other anti-acne treatment during the study,
systemic antibiotics and corticosteroids, as well as any treatment likely
to interfere with tetracyclines.
Patients were seen at baseline for inclusion, then at week 4, 8 and
12. At inclusion informed consent was obtained from all patients and demographic
data, medical history, previous treatment data, and baseline disease data
were collected. Lesion counts, broken down into non-inflammatory lesions
(open and closed comedones) and inflammatory lesions (papules, pustules,
nodules) were taken from the face, including forehead, cheeks and chin.
Global disease severity based on the Leeds scale (from 0, clear to 10,
nodulo-cystic acne or acne conglobata with severe, painful cystic lesions)
was to be graded.
Global assessment of improvement was rated by patients and investigators
on the following scale: - 1, worsened; 0, unchanged; 1, improved;
2, much improved, 3, cleared. Global tolerance was assessed by both investigators
and patients, on the three-point scale - 1, poor tolerance;
0, fair tolerance; 1, good tolerance. Adverse events were collected at
each post-baseline study visit.
A pharmacoeconomical analysis was performed on the Intent to treat (ITT)
population at week 12.
Statistical methods
The study was designed to demonstrate the non-inferiority of lymecycline
treatment compared to minocycline in patients with acne vulgaris. We estimated
that with a power of 80 %, based on an one-tailed alpha risk of 0.025 the
difference of at most 15 % in the reduction in inflammatory lesion
counts between the minocycline and the lymecycline groups would demonstrate
comparability between the lymecycline treatment and the minocycline treatment.
To achieve this the required number of patients was 64 per group.
A total recruitment of 150 patients was therefore planned to allow
patient attrition and non-evaluability.
The primary efficacy criterion was the mean percent reduction in number
of inflammatory lesions at week 12 among those patients who completed
the study as planned (per protocol population [PP]). Mean percent decrease
in the inflammatory lesion counts and in the total lesion counts were
submitted to a covariance analysis, using the baseline counts as covariates,
and country and treatment as factors.
A non-parametric approach was used for non-inflammatory lesions; medians
of the difference between treatments were calculated, as well as the two
tailed 95 % confidence interval.
To take into account a potential bias in the PP analysis, an ITT analysis
was conducted on all randomized patients using the last observation carried
forward (LOCF) method for missing data (end-point analysis). For patients
with no post-baseline data, the baseline values were carried forward to
endpoint. All other data were analyzed descriptively.
Safety data were summarized for all randomized patients who took the treatment
at least once.
Pharmacoeconomic evaluation
We performed a pharmacoeconomical analysis on the ITT population at week
12. The perspective of this analysis was that of the UK National Health
Service as the payer. Due to the short time horizon of this study, no
cost discounting was performed. The global improvement assessment at week
12 as scored by the investigator was the outcome measure to allow
a treatment effectiveness comparison. Depending on the efficacy results
obtained, a cost-effectiveness (efficacy difference) or a cost-minimisation
(no efficacy difference) analysis was performed. The denominator used
was the "investigator's global assessment of improvement", i.e.
the number of patients who scored at least 1. The numerator was the full
treatment cost. Direct costs of treatment were included in this study.
Drug unit costs were based on the NHS treatment costs based on the Monthly
Index of Medical Specialities (April 2002 issue). Consultation costs
and dispensing fees were not included as those costs are common to both
treatments and will, therefore, be incurred irrespective of the treatment
selected. No direct, indirect or intangible costs that accrue to patients
or to any other entity other than the NHS were considered.
Results
The study was conducted from March 10, 1999 to February 8, 2001 in
11 centers (4 in Belgium and 7 in the UK). A total of 136 patients
were enrolled, of whom 2 were screened but not treated. Sixty-six
(66) patients received lymecycline and 68 received minocycline. The
two treatment groups were comparable for demographic data and baseline
characteristics at entry (Table
I). More than 70 % of the patients in both groups reported previous
anti-acne therapy, mostly antibiotics (macrolides and tetracyclines) and
benzoyl peroxide; 14.2 % reported previous tretinoin or isotretinoin
use. From the 134 patients receiving medication 30 withdrew
from the study. Reasons for withdrawal are listed in Table
II.
Lesion counts
The mean number of inflammatory lesions decreased progressively in the
two groups over the duration of treatment (Fig.
1). The adjusted mean (LS-mean) percent reduction at week 12 versus
baseline in the inflammatory lesion count was 63 % for the lymecycline
group and 65 % for the minocycline group (Table
III) with a difference of 1.74 and a confidence interval of 95 %
[ - 9.07; 12.55], showing that lymecycline is comparable in
efficacy to minocycline. The number of non-inflammatory lesions decreased
in both groups at all time points (Fig.
2). Non-inflammatory lesion counts decreased after 12 weeks of
treatment by 53 % for lymecycline and by 49 % for minocycline.
It should be noted that due to a few outliners in the minocycline group,
a non-parametric approach had to be applied for this criterion and medians
were used instead of LS-means. Reductions of inflammatory and non-inflammatory
lesions counts were confirmed by the results found for the ITT population.
The total lesion count followed the same pattern: both groups showed decreased
total lesion counts with no significant difference after 4, 8 and
12 weeks of treatment (Table
III).
Global severity grade
At baseline, mean global severity grade (± SE) was rated 1.9 (± 0.10),
and 2.0 (± 0.13) in the lymecycline and minocycline group respectively
and decreased then in the two treatment groups down to 0.9 (± 0.11)
and 1.0 (± 0.11) respectively after 12 weeks of treatment.
Global assessment of improvement
The endpoint analysis of the ITT population showed that investigators
rated the condition "much improved" or "cleared" in
53 % of the lymecycline treated patients compared to 37 % of
the minocycline treated patients (Table
IV). Patients' assessments go along with those of the investigators:
42 % of the patients under the lymecycline treatment and 38 %
of the patients under the minocycline treatment considered their condition
having "much improved" or "cleared". Results for the
PP population were found similar to those of the ITT population.
Safety
A total of 18 (27.3 %) patients in the lymecycline group and 20 (29.4 %)
patients in the minocycline group experienced at least one adverse event.
The most frequently reported adverse events were headache (lymecycline
4; minocycline 6), pharyngitis (lymecycline, 4; minocycline, 2) and nausea
(lymecycline, 3; minocycline 2). All were mild or moderate in intensity.
In the lymecycline group, 8 adverse events (headache, nausea, convulsion,
dermatitis, diarrhea, dyspepsia, myalgia and pharyngitis) were considered
to be treatment related. All other adverse events were reported as unrelated
to the trial medication.
In the minocycline group, 5 adverse events (oral moniliasis, dermatitis,
nausea, urticaria and vaginitis) were reported as related to the trial
medication. All other adverse events were reported as unrelated to the
trial medication.
One serious adverse event of severe kidney failure was reported after
two days of minocycline treatment in a 23-year-old male with a history
of untreated hypertension and who had previously received minocycline
without any untoward effect. The event was therefore considered to have
an unlikely relationship with the trial medication.
In the minocycline group, vaginitis (definitely related to treatment),
dermatitis (relationship to treatment unlikely) and urticaria (probably
related to medication) each led one patient to discontinue treatment.
Both lymecycline and minocycline were well-tolerated by the majority of
patients (87 % in both groups). Assessment of global tolerance yielded
similar results whether performed by the investigator or patients (Table
V).
Pharmacoeconomic analysis
Improvement (score at least 1 in the investigators' global improvement
assessment scale) was demonstrated in 98.4 % of the patients treated
with lymecycline compared to 91.5 % of patients treated with minocycline.
Since this difference was statistically not significant a cost-minimisation
analysis was performed. The total cost per treatment was £17.88 for
lymecycline and £70.46 for minocycline. To achieve this clinical
success over a treatment period of 12 weeks, two boxes of minocycline
containing 56 capsules or three boxes of lymecycline containing 28 capsules
were needed. The total cost per treatment for a patient cured was calculated
as £18.17 for lymecycline and £77.00 for minocycline (Table
VI).
Discussion
Being the most common oral antibiotic treatment prescribed for acne vulgaris,
tetracyclines are the agents of choice for moderate to severe inflammatory
acne that fails to respond to topical therapy [1].
The Cochrane Database System Review [11] has evaluated the efficacy and
safety of minocycline in acne vulgaris, and has shown that although there
are a lot of published studies with minocycline, widely perceived by clinicians
to have a faster onset of action than oxytetracycline, none of them has
conclusively shown any important clinical difference between the various
tetracyclines and that a significant number of the trials analysed were
generally small and of poor quality.
We confirmed with this trial that lymecycline 300 mg day is comparable
to minocycline in terms of percent decrease in lesion counts and slightly
superior in terms of efficacy compared to lymecycline 150 mg day
as shown in a previous clinical study [12].
As a result of our trial we could prove that the percent decrease for
inflammatory lesions after 12 weeks was similar in both groups. Decrease
for the inflammatory lesion count was also demonstrated all along the
treatment period while showing global improvement as assessed by the investigators
and patients. Global severity decreased in both groups, from 1.9 to
0.9 for lymecycline and from 2.0 to 1.0 for minocycline.
Final evaluation of tolerance performed by both the investigator and patient
showed that lymecycline and minocycline were well tolerated by the majority
of patients.
No treatment related serious adverse event was reported either for lymecycline
or for minocycline.
The most common drug related adverse events reported, with no significant
difference between lymecycline and minocycline, concerned the body as
a whole.
Given the rising pressure on health care budgets, it is becoming increasingly
important to provide evidence of value for money in the selection and
utilisation of treatments. Regarding the results of this pharmacoeconomic
analysis, both treatments provided the same outcome, with lymecycline
being 4 times less costly than minocycline and it is surprising that,
with regard to the above, no results on pharmacoeconomical data for oral
antibiotics in the treatment of acne vulgaris have been previously published.
The present study results confirm the observations made by Cunliffe et
al. [12] and by the Cochrane Database System Review [11]: lymecycline
as part of the tetracycline group has an equivalent efficacy and safety
profile to minocycline while being 2.9 to 4.8 times more cost-effective
than minocycline (variances depending on countries).
CONCLUSION
Lymecycline is an undeniable alternative to minocycline, offering a comparable
efficacy and safety profile. Furthermore, from the perspective of the
UK NHS, the results of this study have shown that lymecycline is more
cost-effective than minocycline, providing therefore good value for money.
We acknowledge the participation of Dr Giles Dunnill, Bristol Royal Infirmary,
United Kingdom and Galderma Belgilux N.V./S.A. and Galderma UK Limited
for organising and funding the study.
Article accepted on 23/12/2002REFERENCES
1 - Sykes NL, Webster GF. Acne. A review of optimum treatment. Drugs
1994; 48: 59-70.
2 - PiErard-Franchimont C, Goffin V, Arrese JE, Martalo O, Braham C, Slachmuylders
P, PiErard GE. Lymecycline and minocycline in inflammatory acne. Skin
Pharmacol Appl Skin Physiol 2002; 15: 112-9.
3 - Cunha BA. Clinical Uses of Tetracyclines. Handbook of Experimental
Pharmacology, The Tetracyclines, (J.J. Hlavka, J.H. Boothe) 1985;
78: 393-404.
4 - Katsambas, A Towarky AA and Stratigos J. Topical Clindamycine phosphate
compared with oral tetracycline in the treatment of acne vulgaris. Br
J Dermatol 1987; 116: 387-91.
5 - Olafsson J, Gudgeirson J. Doxycycline versus Minocycline in the treatment
of acne vulgaris: a double blind study. J Dermatol Treat 1989;
1: 15-7.
6 - Dare S, Papworth-Smith J, Cunliffe WJ. A double blind comparison of
topical clindamycin and oral minocycline in the treatment of acne vulgaris.
Acta Dermatol Venerol 1990; 70: 534-7.
7 - Allen JC. Minocycline. Ann Intern Med 1976; 85: 482-7.
8 - Eisen D. Minocycline-induced oral hyperpigmentation. Lancet
1997; 349: 400.
9 - Okada N, Sato S, Sasou T, et al. Characterisation of pigmented
granules in Minocycline-induced cutaneous pigmentation: observations using
fluorescence microscopy and high performance liquid chromatography.
Br J Dermatol 1993; 129: 403-7.
10 - Siller GM, Tod MA, Savage NW. Minocycline-induced oral pigmentation.
J Am Acad Dermatol 1994; 30: 350-4.
11 - Garner SE, Eady EA, Popescu C, Newton J, Li Wan Po A. Minocycline
for acne vulgaris: efficacy and safety. Cochrane Database System Review
(2) 2000: CD002086 (2000).
12 - Cunliffe WJ, Grosshans E, Belaich S, Meynadier J, Alirezai M, Thomas
LA. comparison of the efficacy and safety of lymecycline and minocycline
in patients with moderately severe acne vulgaris. European Journal
of Dermatology 1998; 8: 161-6 (1998).
13 - Burke BM, Cunliffe WJ. The assessment of acne vulgaris - the
Leeds technique. British Journal of Dermatology 1984; 3: 83-92.
|
Summary
Pictures