doi: 10.1111/1346-8138.

12687 Journal of Dermatology 2015; 42: 64–69

ORIGINAL ARTICLE
Propranolol, doxycycline and combination therapy
for the treatment of rosacea
Jung-Min PARK,1 Je-Ho MUN,1 Margaret SONG,1 Hoon-Soo KIM,1 Byung-Soo KIM,1,2
Moon-Bum KIM,1,2 Hyun-Chang KO1,3
1
Department of Dermatology, School of Medicine, Pusan National University, 2Biomedical Research Institute, Pusan National University
Hospital, and 3Research Institute for Convergence of Biomedical Science and Technology, Pusan National University Yangsan Hospital,
Busan, Korea

ABSTRACT
Doxycycline is the standard systemic treatment for rosacea. Recently, there have been a few reports on b-adren-
ergic blockers such as nadolol, carvedilol and propranolol for suppressing flushing reactions in rosacea. To our
knowledge, there are no comparative studies of propranolol and doxycycline, and combination therapy using
both. The aim of this study was to investigate and compare the efficacy and safety of monotherapy of proprano-
lol, doxycycline and combination therapy. A total of 78 patients who visited Pusan National University Hospital
and were diagnosed with rosacea were included in this study. Among them, 28 patients were in the propranolol
group, 22 the doxycycline group and 28 the combination group. We investigated the patient global assessment
(PGA), investigator global assessment (IGA), assessment of rosacea clinical score (ARCS) and adverse effects.
Improvement in PGA and IGA scores from baseline was noted in all groups, and the combination therapy was
found to be the most effective during the entire period, but this was statistically insignificant. The reduction rate
of ARCS during the treatment period was also highest in the combination group (57.4%), followed by the doxycy-
cline group (52.2%) and the propranolol group (51.0%). Three patients in the combination group had mild and
transient gastrointestinal disturbances but there was no significant difference from the other groups. We con-
clude that the combination therapy of doxycycline and propranolol is effective and safe treatment for rosacea and
successful for reducing both flushing and papulation in particular.

Key words: combination, doxycycline, propranolol, rosacea, treatment.

INTRODUCTION Beta-adrenergic blockers nadolol, carvedilol and propranolol

Rosacea is a common chronic dermatological condition char-
acterized by recurrent episodes of exacerbation and remission.
It usually affects individuals between the ages of 30 and
50 years and women are more affected than men.1,2 Classifi-
cation of rosacea includes erythematotelangiectatic (ETR), pap-
ulopustular (PPR), phymatous and ocular subtypes.3–5 ETR is
characterized by flushing and persistent central facial erythema
and PPR presents with persistent facial erythema and transient
papules or pustules, or both, in a central facial distribution.3
The etiology of rosacea is still unknown and this condition
has led to a therapeutic challenge. Although there is no cura-
tive treatment for rosacea, tetracycline compounds have been
the mainstay therapy and among them, tetracycline and doxy-
cycline are the standard systemic therapy of rosacea. Doxycy-
cline shows anti-inflammatory effects and antioxidant
properties. It exhibits superior pharmacokinetic advantages
and lesser toxicity than tetracycline, so it is widely used for
rosacea.6
have been reported to suppress flushing reactions, particularly
when associated with anxiety.7,8 Its therapeutic mechanism is
to block the b2-adrenergic receptor on the smooth muscle of
cutaneous arterial blood vessels resulting in vasoconstriction.
To our knowledge, there is no comparative study between
propranolol and doxycycline for rosacea and also no previous
report on the efficacy of combination therapy of propranolol
and doxycycline. Therefore, we investigated the efficacy and
safety of the monotherapies of propranolol and doxycycline,
and the combination therapy of propranolol and doxycycline.
METHODS
Patients
Rosacea patients aged above 18 years who visited the outpa-
tient clinic of the Department Of Dermatology at Pusan
National University Hospital from August 2008 to August 2012
were enrolled. The exclusion criteria were patients who had
been treated with topical and systemic medications which can

Correspondence: Hyun-Chang Ko, M.D., Department of Dermatology, School of Medicine, Pusan National University, Geumoh-ro 20,
Mulgeum-eup, Yangsan-si, Busan, Gyeongsangnam-do 626-770, Korea. Email: hcko@pusan.ac.kr
Received 23 January 2014; accepted 29 September 2014.

64 © 2014 Japanese Dermatological Association

 Propranolol and doxycycline for rosacea

affect the symptoms of rosacea (e.g. other antibiotics, isotretin- Methods

oin, corticosteroid, cyclosporin) or with laser that targeted the
vasculature, such as flash pumped pulsed dye laser and
intense pulsed light, for the previous year. For the doxycycline
group, pregnant or lactating women and patients with accom-
panying chronic renal failure, hepatic failure and myasthenia
gravis were excluded. For the propranolol group, patients with
bronchial asthma, hypotension, bradycardia, atrioventricular
block, sinoatrial block and congestive heart failure were
excluded.
The study protocol was approved by the Pusan National Uni-
versity Hospital institutional review board.
At their first visit, the patients’ age, sex and disease duration
were recorded and the severity of the rosacea was assessed.
Subtypes of rosacea (ETR and PPR), distribution, aggravating
factors and symptomatology were also checked.
The global change assessment in the rosacea condition,
as assessed by the patient global assessment (PGA) and
investigator global assessment (IGA), was compared with

Table 1. Demographics and clinical manifestations of the patients

Propranolol Doxycycline Combination Total

group (n = 22) group (n = 15) group (n = 26) (n = 63)
Age, years (mean ! SD) 55.7 ! 12.7 47.4 ! 11.8 48.4 ! 12.6 50.6 ! 12.8
Sex (M : F) 2:9 4:11 4:9 16:47
Subtype (ETR : PPR) 19:3 4:11 9:17 32:31
Duration (months, mean ! SD) 33.3 ! 46.5 25.3 ! 30.1 29.2 ! 32.6 29.7 ! 37.0
Frequency in a day 3.0 ! 2.4 2.0 ! 1.7 2.0 ! 1.0 2.3 ! 1.8
Distribution, n (%)
Whole face 2 (9.1) 1 (6.7) 16 (61.5) 19 (30.2)
Cheek 20 (90.9) 14 (93.3) 21 (80.8) 55 (87.3)
Nose 2 (9.1) 7 (46.7) 15 (57.7) 24 (38.1)
Chin 3 (13.6) 5 (33.3) 16 (61.5) 24 (38.1)
Forehead 4 (18.2) 2 (13.3) 14 (53.8) 20 (31.7)
Periocular 3 (13.6) 2 (13.3) 3 (11.5) 8 (12.7)
Aggravation factor (n, compound factor) (%)
Heat 11 (50.0) 8 (53.3) 13 (50.0) 43 (68.3)
Emotional change 8 (36.4) 9 (60.0) 13 (50.0) 30 (47.6)
Exercise or bathing 6 (27.3) 6 (40.0) 6 (23.1) 18 (28.6)
Alcohol 5 (22.7) 3 (20.0) 6 (23.1) 14 (22.2)
Cold 4 (18.2) 3 (20.0) 3 (11.5) 10 (15.9)
Sun exposure 3 (13.6) 0 4 (15.4) 7 (11.1)
Other 1 (4.5) 0 0 1 (1.6)
Symptom (n, compound factor) (%)
Flushing 18 (81.8) 12 (80.0) 21 (80.8) 51 (81.0)
Itching 2 (9.1) 3 (20.0) 3 (11.5) 8 (12.7)
Tingling 5 (22.7) 1 (6.7) 4 (15.4) 10 (15.9)
Burning 1 (4.5) 0 0 1 (1.6)

ETR, erythematotelangiectatic; PPR, papulopustular; SD, standard deviation.

(a) (b)

Figure 1. (a) Mean physician global assessment and (b) inves-

tigator global assessment scores of rosacea patients through Figure 2. Mean assessment of rosacea clinical score (ARCS)
12 weeks. through 12 weeks.

© 2014 Japanese Dermatological Association 65

J.-M. Park et al.

Table 2. Mean scores of primary features in assessment of rosacea clinical score

Baseline 4 weeks 8 weeks 12 weeks P*

Propranolol group (mean ! SD)
Flushing 2.4 ! 0.5 1.7 ! 0.5 1.3 ! 0.6 0.7 ! 0.6 <0.05
Non-transient erythema 2.1 ! 0.4 1.8 ! 0.6 1.4 ! 0.6 1.1 ! 0.6 <0.05
Papules and pustules 1.7 ! 0.5 1.6 ! 0.5 1.6 ! 0.5 1.5 ! 0.5 0.23
Telangiectasia 1.9 ! 0.4 1.7 ! 0.5 1.4 ! 0.7 1.2 ! 0.7 <0.05
Doxycycline group
Flushing 2.1 ! 0.4 2.0 ! 0.4 1.7 ! 0.5 1.5 ! 0.5 <0.05
Non-transient erythema 2.4 ! 0.5 1.9 ! 0.5 1.7 ! 0.5 1.2 ! 0.4 <0.05
Papules and pustules 2.5 ! 0.5 1.6 ! 0.6 1.3 ! 0.8 0.8 ! 0.7 <0.05
Telangiectasia 2.4 ! 0.6 1.9 ! 0.5 1.8 ! 0.4 1.6 ! 0.5 <0.05
Combination group
Flushing 2.2 ! 0.5 1.5 ! 0.6 1.3 ! 0.7 0.7 ! 0.5 <0.05
Non-transient erythema 2.1 ! 0.3 1.7 ! 0.5 1.5 ! 0.5 1.0 ! 0.6 <0.05
Papules and pustules 2.3 ! 0.5 1.5 ! 0.6 1.1 ! 0.6 0.5 ! 0.5 <0.05
Telangiectasia 1.9 ! 0.4 1.7 ! 0.5 1.5 ! 0.6 1.2 ! 0.6 <0.05

*Baseline vs 12 weeks. SD, standard deviation.

baseline and scored on a 7-point scale, with +3 being mark-
edly improved, +2 moderately improved, +1 mildly improved,
0 unchanged, "1 mildly worse, "2 moderately worse and
"3 markedly worse.9 The assessment of rosacea clinical
score (ARCS) was also checked.4 PGA and IGA were
checked at weeks 2, 4, 8 and 12 and ARCS at baseline and
at weeks 4, 8 and 12. Laboratory tests were done for com-
plete blood cell count, liver and renal functions, and urinaly-
sis before and during the treatment.
The patients with rosacea were divided into three groups:
28 patients treated with propranolol 10 mg three times a day
(propranolol group); 22 patients treated with doxycycline
100 mg two times a day (doxycycline group); and 28 patients
treated with propranolol 10 mg three times a day and doxycy-
cline 100 mg two times a day (combination group).
Statistical analysis
The Kruskal–Wallis test was performed to evaluate differences
between the three groups using the PASW for Windows (IBM,
Armonk, NY, USA). Student’s two-sample t-test was performed
to estimate the differences of the score for the primary features
of ARCS between baseline and after 12 weeks of treatment.
Statistical significance was defined as P < 0.05.
RESULTS
Of the 78 subjects enrolled, 63 completed the study. In the
propranolol group, 78.6% (22/28) of patients completed the
study. Among the six patients who dropped out, other sys-
temic agents were added in the treatment of five patients
(three with doxycycline, one with minocycline, one with isotre-
tinoin) and one patient decided to change to doxycycline
because of the unsatisfactory effects of the propranolol on the
erythema and papules. In the doxycycline group, 68.2% (15/
22) of the patients completed the study. Three patients added
propranolol, one patient added laser therapy, and three
patients changed to propranolol because of the unsatisfactory
effect on flushing during the study period. In the combination
group, 92.9% (26/28) of the patients completed the study. One
patient changed doxycycline to roxithromycin and the other
added laser therapy because of unsatisfactory effects.
Mean age was 50.6 years (range, 16–76), 47 patients were
female and 16 were male. According to the subtypes, 32
patients were ETR patients and 31 PPR. Mean duration of dis-
ease was 29.7 months (range, 1–200) (Table 1).
Mean PGA scores in propranolol, doxycycline and the com-
bination group were 1.7, 1.9 and 2.0 after 12 weeks of treat-
ment, respectively. Mean IGA scores were the same as the
mean PGA scores at the end of the study. Propranolol and the
combination group tended to show rapid improvement within
4 weeks but the doxycycline group caught up with the
improvement of these groups between 4 and 8 weeks (Fig. 1).
However, there were no statistically significant differences
among the three groups.
Mean ARCS were 10.2, 11.3 and 11.6 in the propranolol,
doxycycline and combination groups, respectively, at base-
line and decreased to 5.0, 5.4 and 5.5 after the 12-week
treatment. Reduction ratio of ARCS between baseline and
the end of the study was 51.0%, 52.2% and 57.3% in the
propranolol, doxycycline, and combination groups, respec-
tively (Fig. 2). There also were no statistically significant dif-
ferences among the three groups at baseline and during the
treatment period.
Table 3. Change of percentage in total assessment of rosacea
clinical score from baseline
Propranolol Doxycycline Combination
group (%) group (%) group (%) P
4 weeks 14.6 20.7 24.5 0.037
8 weeks 29.2 29.6 36.4 0.436
12 weeks 42.4 43.1 60.0 0.003

66 © 2014 Japanese Dermatological Association

 Propranolol and doxycycline for rosacea

(a)

(b)

(c)

Figure 3. Serial changes of rosacea patients after 12 weeks of treatment. (a) A 72-year-old woman (erythematotelangiectatic) in the
propranolol group. (b) A 41-year-old man (papulopustular) in the doxycycline group. (c) A 55-year-old woman (papulopustular) in the
combination group.

© 2014 Japanese Dermatological Association 67

J.-M. Park et al.
Table 4. Adverse effects during the study
Propranolol group (%) Doxycycline group (%)
Adverse effect 2 (9.0): dyspepsia (n = 1) 3 (20.0): gastrointestinal
headache (n = 1) disturbance
The primary features in ARCS were analyzed separately to
assess the specific effects of each treatment regimen. In the
propranolol group, the papules and pustules score showed a
partial reduction while the flushing score showed the biggest
decrease after the 12-week treatment, with statistical signifi-
cance. In the doxycycline and combination groups, all primary
feature scores showed a significant decline after the 12-week
treatment, and the papules and pustules scores revealed the
biggest drop at the end of the period (Table 2). In Table 3,
changes of percentage in total ARCS from baseline were ana-
lyzed. At weeks 4 and 12, the combination group showed a
significantly higher percentage change than other groups. The
propranolol group demonstrated a significantly lower percent-
age change than the doxycycline group at week 4, but they
showed similar percentage change after week 8 (Fig. 3).
Adverse events were reported in 12.7% (8/63) of the
patients. Dyspepsia and headache were experienced in 4.5%
(1/22) of the patients in the propranolol group. Gastrointestinal
disturbances were found in 20.0% (3/15) and 11.5% (3/26) of
the patients in the doxycycline and the combination group,
respectively. These side-effects were transient and self-limited
and there were no serious events or discomfort that made the
patients stop the treatment (Table 4).
DISCUSSION
Rosacea is a chronic inflammatory skin disease and yet the
underlying pathophysiology is not entirely known.10 It is char-
acterized by persistent erythema, telangiectasia, papules and
even pustules on the face.11 Various causes have been found
to act as aggravation factors of rosacea, such as emotional
change, heat, exercise, bathing, alcohol, cold and sun expo-
sure, and it is difficult to avoid all provocative stimuli.12,13 Thus,
the therapeutic approach of rosacea depends more on the
clinical subtype than a known etiology.14
Treatment for rosacea includes topical anti-inflammatory
agents, topical or systemic antibacterials, retinoids and laser
therapy.14 Oral tetracycline, particularly tetracycline and doxy-
cycline, have been the mainstay of treatment of rosacea for a
long time. They are especially effective in treating PPR but also
ETR through inhibition of leukocyte-derived matrix-degrading
metalloproteinases.15,16 Flushing usually does not respond to
conventional rosacea treatment. Therefore, treatment of ETR
with severe flushing is challenging although some successes
with beta-blockers such as nadolol, carvedilol and propranolol
have been reported.7,17 Propranolol has not demonstrated
objective evidence for direct effects on cutaneous blood ves-
sels in flushing, but a previous study reported that 88.9% (8/9)
of patients showed improvement of their symptoms and had
fewer flushing episodes while taking propranolol.7 The mechanism
68
Combination group (%) Total (%)
3 (11.5): gastrointestinal 8 (12.7)
disturbance
of propranolol in treating ETR is supposed to be by blocking
the b2-adrenergic receptor on the smooth muscle of cutaneous
arterial blood vessels, resulting in vasoconstriction.5 Moreover,
reactive oxygen species released by local inflammatory cells
which contribute to the inflammation in rosacea can be
controlled by the antioxidant properties of propranolol.18,19
Although ETR is characterized by flushing and PPR is char-
acterized by papules and pustules, symptoms of both sub-
types can be shared in mild form. As previously described, it is
known that doxycycline is more effective in PPR and proprano-
lol seems to be more effective in ETR. Hence, we conducted
this study to compare the efficacy between the monotherapies
of doxycycline and propranolol and the combination of both.
In the present study, the propranolol group showed a faster
response than the doxycycline group regarding PGA and IGA
within the first 4 weeks. Both groups demonstrated the same
effectiveness at week 8 and, finally, the doxycycline group had
higher PGA and IGA scores than the propranolol group by the
end of the study. The doxycycline group showed drastic
improvement between weeks 4 and 8. The combination group
showed the best effect among the three groups during the
entire period and also rapid improvement within the first
4 weeks.
Regarding ARCS, the combination group had the highest
reduction ratio among the three groups. In the analysis of pri-
mary features, after 12 weeks of therapy, with the exception of
papules and pustules in the propranolol group, all of the
parameters showed statistically significant improvement com-
pared with baseline. The propranolol group showed an espe-
cially rapid response in flushing, and the doxycycline group
showed a notably faster effect in papules and pustules. The
combination treatment was effective in both flushing and pap-
ules and pustules. Regarding percentage change of total ARCS
from baseline, the combination group demonstrated signifi-
cantly higher change at first visit and after 12 weeks. This
result indicates that combination therapy of doxycycline and
propranolol can show fast and constant improvement of ARCS
in rosacea patients during 12-week treatment compared with
monotherapy.
Our results show that the combination of doxycycline
200 mg/day and propranolol 30 mg/day significantly reduced
both flushing (70%) and papulation (82%) in rosacea over a
period of 12 weeks. Wise20 reported that doxycycline 40 mg/
day improved 80–100% of inflammatory lesions and reduced
erythema by 50%. Ertl et al.21 demonstrated that isotretinoin
10 mg/day for 16 weeks showed 75% improvement of papular
lesions and 38% reduction in erythema. Compared with previ-
ous studies, combination therapy of doxycycline and propran-
olol is appropriate for patients with both flushing and
papulation.
© 2014 Japanese Dermatological Association

The side-effects were minimal and well-tolerated in all
groups. There were no side-effects of hypotension and brady-
cardia although all of the patients who were treated with pro-
pranolol were normotensive. No cases of photosensitivity were
reported in both the doxycycline and combination group.
This study is subject to a number of limitations: there was a
disproportionate number of patients in each subtype; the
cohort size of the patients was small; and the study was non-
randomized, non-blinded and non-placebo controlled. Particu-
larly, patient groups in this study were divided according to
treatment modality and rosacea subtypes are not equally dis-
tributed among the three treatment groups. It will be valuable
to attempt to analyze treatment groups classified by subtype
because the rosacea treatments are chosen by subtypes.
However, the number of patients per subtype group in this
study was small and it was difficult to perform a statistical
analysis. However, we believe our results add further data to
demonstrate the high effectiveness of the combination therapy
of doxycycline and propranolol in rosacea patients. Random-
ized controlled studies with a large series will be helpful to
more accurately investigate the efficacy and safety of the com-
bination therapy of doxycycline and propranolol. This study is
the first to present data about the combination therapy of
doxycycline and propranolol, showing that it is more effective
than monotherapy, especially in 4 weeks and after 12 weeks
of treatment. The treatment regimen was also well tolerated.
CONFLICT OF INTEREST: The authors have no conflict of
interest to declare.
REFERENCES
1 Powell FC. Clinical practice. Rosacea. N Engl J Med 2005; 352:
793–803.
2 Berg M, Liden S. An epidemiological study of rosacea. Acta Derm
Venereol 1989; 69: 419–423.
3 Wilkin J, Dahl M, Detmar M et al. Standard classification of rosa-
cea: report of the National Rosacea Society Expert Committee on
the classification and staging of rosacea. J Am Acad Dermatol
2002; 46: 584–587.
© 2014 Japanese Dermatological Association
Propranolol and doxycycline for rosacea
4 Wilkin J, Dahl M, Detmar M et al. Standard grading system for rosa-
cea: report of the National Rosacea Society Expert Committee on
the classification and staging of rosacea. J Am Acad Dermatol
2004; 50: 907–912.
5 Lee JY. Rosacea: clinical aspects, pathogenesis and treatment.
Dermatol Sinica 2005; 23: 121–130.
6 Valent!ın S, Morales A, Sa !nchez JL, Rivera A. Safety and efficacy of
doxycycline in the treatment of rosacea. Clin Cosmet Investig
Dermatol 2009; 2: 129–140.
7 Craige H, Cohen JB. Symptomatic treatment of idiopathic and rosa-
cea associated cutaneous flushing with propranolol. J Am Acad
Dermatol 2005; 53: 881–884.
8 Hsu CC, Lee JY. Carvedilol for the treatment of refractory facial
flushing and persistent erythema of rosacea. Arch Dermatol 2011;
147: 1258–1260.
9 Draelos ZD, Ertel K, Berge C. Niacinamide-containing facial mois-
turizer improves skin barrier and benefits subjects with rosacea.
Cutis 2005; 76: 135–141.
10 Conde JF, Yelverton CB, Balkrishnan R et al. Managing rosacea: a
review of the use of metronidazole alone and in combination with
oral antibiotics. J Drugs Dermatol 2007; 6: 495–498.
11 Marks R. The enigma of rosacea. J Dermatolog Treat 2007; 18:
326–328.
12 Lee SY, Choi JH, Sung KJ, Moon KC, Koh JK. A clinical study of
112 patients with rosacea. Korean J Dermatol 2001; 39: 636–642.
€s E et al. The role of free oxygen radicals in
13 Oztas MO, Balk M, Ogu
the aetiopathogenesis of rosacea. Clin Exp Dermatol 2003; 28: 188–
192.
14 Alikhan A, Kurek L, Feldman SR. The role of tetracyclines in rosa-
cea. Am J Clin Dermatol 2010; 11: 79–87.
15 Sapadin AN, Fleischmajer R. Tetracyclines: nonantibiotic properties
and their clinical implications. J Am Acad Dermatol 2006; 54: 258–
265.
16 Webster G, Del Rosso JQ. Anti-inflammatory activity of tetracy-
clines. Dermatol Clin 2007; 25: 133–135.
17 Hsu CC, Lee JY. Pronounced facial flushing and persistent
erythema of rosacea effectively treated by carvedilol, a nonselec-
tive b-adrenergic blocker. J Am Acad Dermatol 2012; 67: 491–
493.
18 Mak IT, Weglicki WB. Potent antioxidant properties of 4-hydroxyl-
propranolol. J Pharmacol Exp Ther 2004; 308: 85–90.
19 Jones D. Reactive oxygen species and rosacea. Cutis 2004; 74
(Suppl): 17–20, 32–4.
20 Wise RD. Submicrobial doxycycline and rosacea. Comp Ther 2007;
33: 78–81.
21 Ertl GA, Levine N, Kligman AM. A comparison of the efficacy of top-
ical tretinoin and low-dose oral isotretinoin in rosacea. Arch Derma-
tol 1994; 130: 319–324.
69