JtothaK
10th October 2007, 10:40 PM
An oldie but a goodie!
Systematic review of rosacea treatments
sther J. van Zuuren MDa
Aditya K. Gupta MD, PhD, FRCP(C)b, c, Melissa D. Gover BScc, Mark Graber MDd and Sally Hollis MSce
aFrom the Department of Dermatology B1-Q, Leiden University Medical Center
bDivision of Dermatology, Department of Medicine, Sunnybrook and Women's College Health Sciences Center and the University of Toronto
cMediprobe Research Inc, London
dEmergency Medicine and Family Medicine, University of Iowa College of Medicine
eUniversity of Lancaster
Leiden, The Netherlands; Toronto and London, Ontario, Canada; Iowa City, Iowa; and Lancaster, United Kingdom. Available online 7 November 2006.
Background
Rosacea is a common chronic skin and ocular condition. It is unclear which treatments are most effective. We have conducted a Cochrane review of rosacea therapies.1 This article is a distillation of that work.
Objective
We sought to assess the evidence for the efficacy and safety of previous rosacea therapies.
Methods
Multiple databases were systematically searched. Randomized controlled trials in people with moderate to severe previous rosacea term were included. Study selection, assessment of methodologic quality, data extraction, and analysis were carried out by two independent researchers.
Results
In all, 29 studies met inclusion criteria. Topical metronidazole is more effective than placebo (odds ratio 5.96, 95% confidence interval 2.95-12.06). Azelaic acid is more effective than placebo (odds ratio 2.45, 95% confidence interval 1.82-3.28). Firm conclusions could not be drawn about other therapies.
Limitations
The quality of the studies was generally poor.
Conclusions
There is evidence that topical metronidazole and azelaic acid are effective. There is some evidence that oral metronidazole and tetracycline are effective. More well-designed, randomized controlled trials are required to provide better evidence of the efficacy and safety of other rosacea therapies.
Abbreviations: CI, confidence interval; OR, odds ratio; RCT, randomized controlled trial
Article Outline
Methods
Search strategies
Selection criteria
Study design quality assessment and data extraction
Outcome measures
Analysis
Results
Description of studies and methodologic quality of included studies
Analysis
Metronidazole
Topical metronidazole versus placebo
Topical azelaic acid versus topical metronidazole
Topical metronidazole versus oral tetracycline
Metronidazole plus sunscreen (sun protection factor 15) versus placebo
Topical metronidazole versus topical permethrin versus placebo
Benzoyl peroxide 5%/erythromycin 3% gel versus metronidazole gel
Oral antibiotics
Oral metronidazole versus oral oxytetracycline
Tetracycline versus placebo
Clarithromycin and omeprazole versus placebo
Azelaic acid
Azelaic acid versus placebo
Benzoyl peroxide with antibiotics
Benzoyl peroxide 5%/clindamycin 1% gel versus placebo
Other
Benzoyl peroxide acetone versus placebo
Oral metronidazole and topical hydrocortisone 1% cream versus oral placebo and topical hydrocortisone 1% cream
Rilmenidine versus placebo
Sodium sulfacetamide/sulfur versus placebo
Discussion
References
Rosacea is a chronic condition characterized by recurrent episodes of facial flushing, erythema, papules, pustules, and telangiectasia in a symmetrical, facial distribution.1, 2, 3 and 4 Several well-defined types of rosacea are described including erythematotelangiectatic rosacea, papulopustular rosacea, phymatous rosacea, ocular rosacea, and the variant granulomatous rosacea.3 and 4 Ocular rosacea can develop without involvement of other areas of the skin and may wax and wane.2 and 5 Rosacea usually presents in the second or third decade of life and has a prevalence of up to 10%.6 It is especially common in fair-skinned people of Celtic and northern European heritage, with women more often affected than men.7, 8 and 9 However, men will more often progress to the later stages.9
Traditionally, rosacea has been managed with a treatment tailored to the specific symptoms presented.7 A brief overview of these therapies is presented in Table I.7, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23 and 24 Other treatments tried include facial massage (for edema), spironolactone, beta-blockers, dapsone, oral contraceptives, benzoyl peroxide, bifonazole cream, and treatment of Helicobacter pylori.16 and 24 Unfortunately, many of these remain poorly studied. This review was performed to systematically evaluate rosacea treatments including the potential impact of nonpharmacologic agents such as foods (eg, spicy food), certain cosmetics, and sunscreens.10
Table I.
Rosacea therapies8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23 and 24
Signs/symptoms Treatments
Limited no. of papules/pustules Topical therapies Metronidazole (0.75%, 1%)
Clindamycin lotion
Permethrin 5% cream
Tretinoin cream
Sulfacetamide 10%/sulfur 5%
Azelaic acid (15% gel, 20% cream)
Proposed therapies Tacrolimus
Topical NADH
More extensive skin lesions Oral antibiotics Tetracycline Possible side effects including gastrointestinal symptoms, photosensitivity, candidal vaginitis, reduction in oral contraceptive efficacy Metronidazole side effects may include neuropathy and mutagenicity
Ampicillin
Metronidazole
Erythromycin
Oral/topical therapy combination Discontinue oral treatment once sufficient efficacy noted Maintenance therapy with topical medications
Vascular symptoms Pulse dye laser, intense pulsed light
Severe or persistent rosacea Oral isotretinoin 13-cis-retinoic acid Possible side effects include: dry sensitive skin, dry mucosae, dry eyes, pruritis, dermatitis, myalgia, elevated liver enzymes, cholesterol and triglyceride elevation Routine monitoring of liver functions, cholesterol, triglycerides required
Possible fetal abnormalities for women who become pregnant
Control of flushing Oral hypotensives Clonidine
Rilmenidine
Rhinophyma Oral Low-dose isotretinoin
Laser therapy
Surgical intervention
Ocular rosacea Oral antibiotics Tetracycline
Topicals Metronidazole
Fusidic acid gel
NADH, reduced form of ß-nicotinamide adenine dinucleotide.
Unfortunately, there is no universally accepted clinical definition of rosacea, and there are no standard validated tools for assessing the severity of rosacea. As rosacea can cause shame, embarrassment, low self-esteem, anxiety, lack of confidence, and depression, our primary outcome was the patients' self-assessment of rosacea, and their perception of their quality of life.8, 9 and 10
Methods
A systematic review of randomized controlled trials (RCTs) was performed according to a prespecified protocol.1
Search strategies
Two reviewers performed independent searches of the following 6 electronic databases: The Cochrane Skin Group Specialized Trials Register (February 2005), The Cochrane Central Register of Controlled Trials (February 2005), MEDLINE (1966-February 2005), EMBASE (1980-February 2005), Biosis (1970-March 2002), and Science Citation Index (1988-February 2005). In addition, the reference lists of all identified RCTs and key review articles were searched. Attempts were made to obtain details of unpublished and ongoing RCTs and grey literature through correspondence with authors and pharmaceutical companies.
Selection criteria
We considered all RCTs evaluating any type of intervention used to treat rosacea. Study participants had to be older than 19 years with moderate to severe rosacea as assessed by a physician. Two reviewers independently assessed these articles for eligibility. Any disagreement was resolved by discussion.
Study design quality assessment and data extraction
Study design was assessed by two reviewers as per the criteria in Table II. Studies meeting all the criteria were considered high quality, whereas studies meeting some, but not all, were generally classified as intermediate. Studies classified as low quality were excluded from analysis. Supporting methodology descriptions for each criterion had to be present in the published text to merit the grading. Details of eligible trials were extracted and summarized using structured data collection forms.
Table II.
Criteria used to assess the methodologic quality of randomized controlled trials for rosacea therapies
Quality assessment criteria:
* Was the randomization procedure used appropriate?
* Was the allocation concealment adequate?
* Was an intention-to-treat analysis used?
* Were health workers and study personnel blind to treatment?
* Were participants blind to treatment?
Aside from the intervention, were groups treated equally?
* Was the study duration fixed/adequate (at least 4 weeks)?
Were number and timing of assessment points fixed?
* Was there an acceptable description or definition of rosacea?
* Was the site of evaluation recorded?
*,† Were concomitant medications permitted and recorded?
Was previous oral and topical rosacea therapy stopped a minimum of 4 weeks before the initial assessment?
Were the therapeutic interventions adequately described?
Were adequate details about how to use/take the medication given to all participants?
* Was the dropout rate less than 5%?
Modified1 and used with permission.
* All these criteria must be “yes” to be high quality.
† Study must not allow concomitant medications that might change outcome.
Outcome measures
The primary outcome measures included impact on quality of life and participant-assessed changes in rosacea severity. Secondary outcome measures were physician-assessed changes in rosacea severity, physician's global evaluation (improvement defined as ≥50% change), lesion counts (treatment success defined as >50% reduction), time needed for improvement, and duration of remission. Other outcomes included dropout rates and incidence of adverse events.
Analysis
Quantitative pooling was performed using odds ratio (OR) for categorical measures or weighted mean differences for continuous measures. Where study results were heterogeneous, the reasons for this were explored (eg, treatment or participant factors) and a random effects model was used to reflect the increased uncertainty. Investigation of the robustness of the conclusions according to the methodologic quality of the contributing studies was not practical because there were only a few studies contributing to each comparison; study quality was considered qualitatively when drawing conclusions.
Some studies used a split-face, within-patient design, where two interventions were allocated randomly to the left and right side of the face. Where possible, a conditional OR (based on the discordant cases only) was calculated; this can be interpreted in the same way as the ORs from parallel group studies.25 However, the paired data necessary for this were sometimes unavailable, in which case marginal ORs (based on the overall rates for each treatment) were calculated and reported. These marginal ORs should be interpreted cautiously, because they differ from conditional ORs when there is correlation between the outcomes of the two treatments.
Results
Description of studies and methodologic quality of included studies
Searches identified 71 possible RCTs. A total of 29 RCTs were included.21, 22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50 and 51 Breneman et al34 and Leyden et al51 described different outcome measures of the same study and Thiboutot et al49 reported two RCTs in one publication. Most of the participants in the included studies had papulopustular rosacea and were between 40 and 50 years old; only two studies27 and 28 addressed ocular rosacea. Of the 71 studies, 41 were excluded because allocation concealment was inadequate, the study was not blinded, the dropout rate was more than 10%, or other major methodologic flaws,11, 12, 18, 19, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84 and 85 or because they were awaiting assessment.13, 86 and 87 Of the 29 included studies, 8 were classified as high quality.23, 26, 28, 29, 30, 32, 34 and 36 The remaining 21 trials were of intermediate quality.21, 22, 27, 31, 33, 35, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49 and 50 In only 14 of the 29 trials23, 26, 28, 29, 30, 32, 34, 35, 36, 40, 42, 47 and 49 was there adequate blinding of treatment allocation. Blinding of outcome assessment was demonstrated in all except two studies.27 and 43 Intention-to-treat analysis was used in 17 of the 29 trials.21, 22, 23, 26, 28, 29, 30, 31, 32, 34, 35, 36, 37, 38, 39 and 49 For 14 studies the variability (SD or SE) of continuous measurements were completely or partially lacking, making these data unusable in a meta-analysis.21, 30, 32, 33, 34, 36, 37, 38, 41, 42, 43, 46, 48 and 50
Analysis
The treatments could be categorized into 5 groups: topical metronidazole (15 trials),21, 22, 23, 27, 29, 30, 32, 33, 36, 38, 39, 42, 43, 48 and 50 oral antibiotics (8 trials),23, 26, 28, 40, 44, 45, 47 and 50 topical azelaic acid (6 trials),22, 31, 35, 36 and 49 topical benzoyl peroxide combined with topical antibiotics (2 trials),34 and 43 and other therapies (4 trials).37, 41, 44 and 46 Five trials included comparisons in more than one category.22, 23, 36, 43 and 44 Even within these therapeutic categories, making comparisons and pooling of data was problematic because of heterogeneous study designs, skewed data, missing variability, and differences in comparators or dosing regimens. Only data on outcome measures from trials on topical metronidazole, topical azelaic acid, and oral tetracycline could be pooled. Most studies used numbers of papules or pustules as an outcome measure rather than a more clinically relevant measure, such as participant assessment of appearance. Below is a summary of the most important conclusions. For details and full reporting of the data, please refer to the complete Cochrane review as published in the Cochrane Library.1
Metronidazole
Topical metronidazole versus placebo
Nine trials assessed the efficacy of topical metronidazole versus placebo.21, 27, 29, 30, 32, 33, 38, 39 and 42 The treatment period ranged from 8 to 9 weeks in each trial, except for that of Dahl et al,21 which was 6 months. Three studies addressed self-assessed improvement of rosacea severity.30, 32 and 42 Only data from two studies30 and 42 could be pooled (Fig 1, A) and there was clear evidence that metronidazole was more effective than placebo. Bleicher et al32 confirmed these data (OR 7.0; 95% confidence interval [CI] 2.5-20.0). Data on physician's global evaluation concerning improvement were similar to the patient-assessed measures in favor of metronidazole (OR 7.01; 95% CI 3.56-13.81).30, 33 and 42 The other studies showed comparable data.21, 27, 29, 32 and 38
Fig 1. Meta-analytic comparisons of participant-assessed improvement between topical metronidazole and placebo (A), oral tetracycline and placebo (B), and topical azelaic acid and placebo (C). Modified1 and used with permission. CI, Confidence interval; OR, odds ratio.
Most of the adverse events mentioned were mild, including pruritus, skin irritation, and dry skin. There were no significant differences in the number of adverse events between groups.
Topical azelaic acid versus topical metronidazole
There was no statistically significant difference in the patient self-assessment between topical azelaic acid and topical metronidazole.22 and 36 However, physicians rated the azelaic acid group more improved (OR 1.84; 95% CI 1.10-3.09).36 The number of adverse events was lower in the metronidazole group (OR 4.56; 95% CI 2.07-10.03).36 However, the severity of adverse events in both groups was reported as mild to moderate and mostly transient.
Topical metronidazole versus oral tetracycline
In two 8-week studies23 and 50 no statistically significant treatment difference was seen between metronidazole cream and (oxy)-tetracycline.
Metronidazole plus sunscreen (sun protection factor 15) versus placebo
A poorly designed study favored metronidazole plus sunscreen over placebo.48
Topical metronidazole versus topical permethrin versus placebo
Koçak et al39 investigated the efficacy and safety of permethrin for the treatment of rosacea. Permethrin was inferior to topical metronidazole because it showed no effect on pustules.
Benzoyl peroxide 5%/erythromycin 3% gel versus metronidazole gel
No significant difference was shown between the two therapies in 4 weeks (OR 0.92; 95% CI 0.21-4.11).43
Oral antibiotics
Oral metronidazole versus oral oxytetracycline
In one study, oral metronidazole and oral oxytetracycline were not statistically different at 12 weeks by both physician and patient assessment.45 No adverse events were reported in either group.
Tetracycline versus placebo
One trial28 compared oral oxytetracycline with placebo, and in two trials40 and 47 oral tetracycline was compared with placebo. These are both (older) tetracyclines with a similar molecular structure and the same pharmacokinetic and pharmacodynamic profile and so the results were pooled. Study duration ranged from 4 to 6 weeks. Bartholomew et al28 addressed treatment efficacy in ocular rosacea.
There was insufficient evidence of any advantage of tetracycline over placebo according to patients' assessment (Fig 1, B).40 However, the dropout rate was unclear and the data were skewed with large variability. By physician assessment, tetracyclines are far more effective than placebo in the treatment of rosacea (OR 6.06; 95% CI 2.96-12.42). Repeated courses of treatment with the same dose achieved lasting remission 3 to 6 months after stopping treatment.28
Clarithromycin and omeprazole versus placebo
These data were skewed with large variability and, thus, it is impossible to draw conclusions about this therapy.26
Azelaic acid
Azelaic acid versus placebo
Four trials compared azelaic acid with placebo.31, 35 and 49 The treatment period ranged from 9 to 12 weeks. Three studies31 and 49 showed a clear improvement in the azelaic acid group as rated by both physicians and patients (Fig 1, C). A split-face, within-patient study35 confirmed these results (marginal OR 30.1; P < .0003).
The data on lesion counts did not include variability and the data in the study by Carmichael et al35 were skewed.
More side effects were reported in the azelaic group (11.5%) versus the placebo group (5.7%) (OR 1.61; 95% CI 0.89-2.92).31 and 49 The same holds true for the study of Carmichael et al.35 Side effects were considered mild and transient with burning, stinging, and irritation being reported most frequently.
Benzoyl peroxide with antibiotics
Benzoyl peroxide 5%/clindamycin 1% gel versus placebo
The mean scores at 12 weeks for patient's global assessment in the study of Breneman et al34 were 1.54 (much to slightly better) in the benzoyl peroxide and clindamycin group versus 2.50 (slightly better to same) in the placebo group (authors state P = .0002). The mean scores at 12 weeks for physician's global assessment were 1.85 (marked to definite improvement) versus 2.96 for placebo (minimal improvement) (authors state P = .0026).
The data showed large variability and some data were missing. Most data were skewed. Treatment-related adverse events included site burning and itching, both well-known side effects of benzoyl peroxide.34 The same study using photographic assessments as outcomes came to similar same conclusion.51
Other
Benzoyl peroxide acetone versus placebo
At 4 weeks, benzoyl peroxide showed an improvement on the physician's global evaluation score compared with placebo (OR 3.17; 95% CI 1.08-9.31).41 The other measurements were also in favor of benzoyl peroxide (P < .05). Irritation and burning were frequently reported in both groups.
Oral metronidazole and topical hydrocortisone 1% cream versus oral placebo and topical hydrocortisone 1% cream
The physicians considered 10 of 14 participants treated with oral metronidazole improved versus only 2 of 13 participants on placebo (OR 13.75; 95% CI 2.05-92.04).44 Only limited data were given in this study.44
Rilmenidine versus placebo
Both the patients and the physicians believed that there was no significant difference between rilmenidine and placebo; neither treatment was effective.37
Sodium sulfacetamide/sulfur versus placebo
The percentage of participants who considered themselves improved was 90% in the sodium sulfacetamide 10%/sulfur 5% group versus 58% in the placebo group (authors state P < .001).46 The physicians shared this opinion. Adverse events were reported in 38% versus 29%, respectively. Application site reactions such as dryness, erythema, and pruritus were the most commonly reported adverse events. It was unclear how many participants started in each group or how improvement was defined, and for continuous measurements the variability was large and the data skewed.46
Discussion
There were significant limitations in the quality of evidence available for most treatments. Although the clinical design of the included studies was in theory adequate, closer examination revealed that the quality of reported data was often low. Table III and Table IV summarize recommendations for future rosacea studies.88 It should be noted that although split-face studies can be efficient, they are subject to potential biases. Contamination may occur if active cream is accidentally transferred onto the placebo side. Furthermore, a treatment may have systemic effects, beneficial or harmful, which will affect both sides.
Table III.
Data to be included in future rosacea studies
Well-designed RCT with reporting the following88
Proper description of randomization procedure and allocation concealment
Data presented with appropriate summaries and analysis (including variability)
The number of participants who started in and dropped out from each group
Outcomes primarily based on: patient's opinion of treatment efficacy, quality of life, and patient-assessment
Physician's opinion, reflected by global evaluation, lesion counts, and assessment of telangiectasia
Use of intention-to-treat analysis
RCT, Randomized controlled trial.
Modified1 and used with permission.
Table IV.
Questions for which evidence is lacking in the literature
1. What is the efficacy and safety of commonly used treatments for rosacea (eg, tetracycline, minocycline, doxycycline, isotretinoin, and laser therapy)?
2. What is the efficacy and safety of treatments for ocular rosacea?
3. Is there any efficacy of dietary measures and/or sun-protective measures in the treatment of rosacea?
4. What is the efficacy and safety of benzoyl peroxide alone or in combination with topical antibiotics for rosacea?
5. Is permethrin effective and safe for rosacea treatment?
Studies to answer these questions should meet the criteria mentioned in Table III.
Our principal outcome measure, quality of life, was not assessed in any of the studies and only a few studies assessed the participant's own opinion. It is interesting to note that the investigators were more satisfied at the end of the study than the participants.35, 36 and 40 For other diseases it is often the reverse. This may have implications for clinicians, as a patient's perception of a lack of sufficient efficacy can impact compliance and may lead to the use of alternative therapies. Topical metronidazole and azelaic acid appear to be effective and safe for short-term use, with the rate of adverse events in the placebo groups being similar to the active treatment groups. With regard to tetracycline, only 3 studies28, 40 and 47 could be included in this review, only one of which assessed the opinion of the participants; however, this study failed to detect any difference from placebo. It is possible that in this case the study duration of 6 weeks was too short.
There were no studies evaluating other treatment options, such as erythromycin, dapsone, and topical tretinoin,7, 16, 89, 90, 91 and 92 that met the inclusion criteria. Three studies were included using benzoyl peroxide alone or in combination with topical antibiotics.34, 41 and 43 Unfortunately, the quality of these studies was suboptimal. The same holds true for the study with permethrin.39 Both benzoyl peroxide and permethrin are well-known drugs and further investigation in the treatment of rosacea may be beneficial.
No studies could be included addressing dietary or sun-protective measures; however, two studies did combine treatment with a sun protection factor.39 and 48 Although not really substantiated, most people with rosacea are given the advice to avoid trigger factors, (eg, spicy foods, alcohol, and sunlight).
Only two trials could be included on treatment of ocular rosacea,27 and 28 even though almost 60% of people with rosacea have ocular involvement.27, 66, 79 and 91 Although often mild, the ocular presentation can be both severe and debilitating. There was insufficient evidence for the efficacy of topical metronidazole.27 Oral oxytetracycline seems to be effective for ocular rosacea,28 although only the opinion of the physician was reported.
A very interesting treatment seems to be low-dose doxycycline (20 mg twice a day),77 and 93 which is a subantimicrobial dose that reduces inflammation. Other potential advantages of this treatment include lessening the risks of Propionibacterium acne's resistance to tetracyclines and lowering the incidence of tetracycline-induced adverse events. Unfortunately, even though they are commonly used to treat rosacea, no RCTs evaluating doxycycline, minocycline, isotretinoin, or laser therapy could be included in this review (most often because of inadequate study design). There is an urgent need for better quality, adequately designed RCTs on the commonly used treatments for rosacea.
Systematic review of rosacea treatments
sther J. van Zuuren MDa
Aditya K. Gupta MD, PhD, FRCP(C)b, c, Melissa D. Gover BScc, Mark Graber MDd and Sally Hollis MSce
aFrom the Department of Dermatology B1-Q, Leiden University Medical Center
bDivision of Dermatology, Department of Medicine, Sunnybrook and Women's College Health Sciences Center and the University of Toronto
cMediprobe Research Inc, London
dEmergency Medicine and Family Medicine, University of Iowa College of Medicine
eUniversity of Lancaster
Leiden, The Netherlands; Toronto and London, Ontario, Canada; Iowa City, Iowa; and Lancaster, United Kingdom. Available online 7 November 2006.
Background
Rosacea is a common chronic skin and ocular condition. It is unclear which treatments are most effective. We have conducted a Cochrane review of rosacea therapies.1 This article is a distillation of that work.
Objective
We sought to assess the evidence for the efficacy and safety of previous rosacea therapies.
Methods
Multiple databases were systematically searched. Randomized controlled trials in people with moderate to severe previous rosacea term were included. Study selection, assessment of methodologic quality, data extraction, and analysis were carried out by two independent researchers.
Results
In all, 29 studies met inclusion criteria. Topical metronidazole is more effective than placebo (odds ratio 5.96, 95% confidence interval 2.95-12.06). Azelaic acid is more effective than placebo (odds ratio 2.45, 95% confidence interval 1.82-3.28). Firm conclusions could not be drawn about other therapies.
Limitations
The quality of the studies was generally poor.
Conclusions
There is evidence that topical metronidazole and azelaic acid are effective. There is some evidence that oral metronidazole and tetracycline are effective. More well-designed, randomized controlled trials are required to provide better evidence of the efficacy and safety of other rosacea therapies.
Abbreviations: CI, confidence interval; OR, odds ratio; RCT, randomized controlled trial
Article Outline
Methods
Search strategies
Selection criteria
Study design quality assessment and data extraction
Outcome measures
Analysis
Results
Description of studies and methodologic quality of included studies
Analysis
Metronidazole
Topical metronidazole versus placebo
Topical azelaic acid versus topical metronidazole
Topical metronidazole versus oral tetracycline
Metronidazole plus sunscreen (sun protection factor 15) versus placebo
Topical metronidazole versus topical permethrin versus placebo
Benzoyl peroxide 5%/erythromycin 3% gel versus metronidazole gel
Oral antibiotics
Oral metronidazole versus oral oxytetracycline
Tetracycline versus placebo
Clarithromycin and omeprazole versus placebo
Azelaic acid
Azelaic acid versus placebo
Benzoyl peroxide with antibiotics
Benzoyl peroxide 5%/clindamycin 1% gel versus placebo
Other
Benzoyl peroxide acetone versus placebo
Oral metronidazole and topical hydrocortisone 1% cream versus oral placebo and topical hydrocortisone 1% cream
Rilmenidine versus placebo
Sodium sulfacetamide/sulfur versus placebo
Discussion
References
Rosacea is a chronic condition characterized by recurrent episodes of facial flushing, erythema, papules, pustules, and telangiectasia in a symmetrical, facial distribution.1, 2, 3 and 4 Several well-defined types of rosacea are described including erythematotelangiectatic rosacea, papulopustular rosacea, phymatous rosacea, ocular rosacea, and the variant granulomatous rosacea.3 and 4 Ocular rosacea can develop without involvement of other areas of the skin and may wax and wane.2 and 5 Rosacea usually presents in the second or third decade of life and has a prevalence of up to 10%.6 It is especially common in fair-skinned people of Celtic and northern European heritage, with women more often affected than men.7, 8 and 9 However, men will more often progress to the later stages.9
Traditionally, rosacea has been managed with a treatment tailored to the specific symptoms presented.7 A brief overview of these therapies is presented in Table I.7, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23 and 24 Other treatments tried include facial massage (for edema), spironolactone, beta-blockers, dapsone, oral contraceptives, benzoyl peroxide, bifonazole cream, and treatment of Helicobacter pylori.16 and 24 Unfortunately, many of these remain poorly studied. This review was performed to systematically evaluate rosacea treatments including the potential impact of nonpharmacologic agents such as foods (eg, spicy food), certain cosmetics, and sunscreens.10
Table I.
Rosacea therapies8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23 and 24
Signs/symptoms Treatments
Limited no. of papules/pustules Topical therapies Metronidazole (0.75%, 1%)
Clindamycin lotion
Permethrin 5% cream
Tretinoin cream
Sulfacetamide 10%/sulfur 5%
Azelaic acid (15% gel, 20% cream)
Proposed therapies Tacrolimus
Topical NADH
More extensive skin lesions Oral antibiotics Tetracycline Possible side effects including gastrointestinal symptoms, photosensitivity, candidal vaginitis, reduction in oral contraceptive efficacy Metronidazole side effects may include neuropathy and mutagenicity
Ampicillin
Metronidazole
Erythromycin
Oral/topical therapy combination Discontinue oral treatment once sufficient efficacy noted Maintenance therapy with topical medications
Vascular symptoms Pulse dye laser, intense pulsed light
Severe or persistent rosacea Oral isotretinoin 13-cis-retinoic acid Possible side effects include: dry sensitive skin, dry mucosae, dry eyes, pruritis, dermatitis, myalgia, elevated liver enzymes, cholesterol and triglyceride elevation Routine monitoring of liver functions, cholesterol, triglycerides required
Possible fetal abnormalities for women who become pregnant
Control of flushing Oral hypotensives Clonidine
Rilmenidine
Rhinophyma Oral Low-dose isotretinoin
Laser therapy
Surgical intervention
Ocular rosacea Oral antibiotics Tetracycline
Topicals Metronidazole
Fusidic acid gel
NADH, reduced form of ß-nicotinamide adenine dinucleotide.
Unfortunately, there is no universally accepted clinical definition of rosacea, and there are no standard validated tools for assessing the severity of rosacea. As rosacea can cause shame, embarrassment, low self-esteem, anxiety, lack of confidence, and depression, our primary outcome was the patients' self-assessment of rosacea, and their perception of their quality of life.8, 9 and 10
Methods
A systematic review of randomized controlled trials (RCTs) was performed according to a prespecified protocol.1
Search strategies
Two reviewers performed independent searches of the following 6 electronic databases: The Cochrane Skin Group Specialized Trials Register (February 2005), The Cochrane Central Register of Controlled Trials (February 2005), MEDLINE (1966-February 2005), EMBASE (1980-February 2005), Biosis (1970-March 2002), and Science Citation Index (1988-February 2005). In addition, the reference lists of all identified RCTs and key review articles were searched. Attempts were made to obtain details of unpublished and ongoing RCTs and grey literature through correspondence with authors and pharmaceutical companies.
Selection criteria
We considered all RCTs evaluating any type of intervention used to treat rosacea. Study participants had to be older than 19 years with moderate to severe rosacea as assessed by a physician. Two reviewers independently assessed these articles for eligibility. Any disagreement was resolved by discussion.
Study design quality assessment and data extraction
Study design was assessed by two reviewers as per the criteria in Table II. Studies meeting all the criteria were considered high quality, whereas studies meeting some, but not all, were generally classified as intermediate. Studies classified as low quality were excluded from analysis. Supporting methodology descriptions for each criterion had to be present in the published text to merit the grading. Details of eligible trials were extracted and summarized using structured data collection forms.
Table II.
Criteria used to assess the methodologic quality of randomized controlled trials for rosacea therapies
Quality assessment criteria:
* Was the randomization procedure used appropriate?
* Was the allocation concealment adequate?
* Was an intention-to-treat analysis used?
* Were health workers and study personnel blind to treatment?
* Were participants blind to treatment?
Aside from the intervention, were groups treated equally?
* Was the study duration fixed/adequate (at least 4 weeks)?
Were number and timing of assessment points fixed?
* Was there an acceptable description or definition of rosacea?
* Was the site of evaluation recorded?
*,† Were concomitant medications permitted and recorded?
Was previous oral and topical rosacea therapy stopped a minimum of 4 weeks before the initial assessment?
Were the therapeutic interventions adequately described?
Were adequate details about how to use/take the medication given to all participants?
* Was the dropout rate less than 5%?
Modified1 and used with permission.
* All these criteria must be “yes” to be high quality.
† Study must not allow concomitant medications that might change outcome.
Outcome measures
The primary outcome measures included impact on quality of life and participant-assessed changes in rosacea severity. Secondary outcome measures were physician-assessed changes in rosacea severity, physician's global evaluation (improvement defined as ≥50% change), lesion counts (treatment success defined as >50% reduction), time needed for improvement, and duration of remission. Other outcomes included dropout rates and incidence of adverse events.
Analysis
Quantitative pooling was performed using odds ratio (OR) for categorical measures or weighted mean differences for continuous measures. Where study results were heterogeneous, the reasons for this were explored (eg, treatment or participant factors) and a random effects model was used to reflect the increased uncertainty. Investigation of the robustness of the conclusions according to the methodologic quality of the contributing studies was not practical because there were only a few studies contributing to each comparison; study quality was considered qualitatively when drawing conclusions.
Some studies used a split-face, within-patient design, where two interventions were allocated randomly to the left and right side of the face. Where possible, a conditional OR (based on the discordant cases only) was calculated; this can be interpreted in the same way as the ORs from parallel group studies.25 However, the paired data necessary for this were sometimes unavailable, in which case marginal ORs (based on the overall rates for each treatment) were calculated and reported. These marginal ORs should be interpreted cautiously, because they differ from conditional ORs when there is correlation between the outcomes of the two treatments.
Results
Description of studies and methodologic quality of included studies
Searches identified 71 possible RCTs. A total of 29 RCTs were included.21, 22, 23, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50 and 51 Breneman et al34 and Leyden et al51 described different outcome measures of the same study and Thiboutot et al49 reported two RCTs in one publication. Most of the participants in the included studies had papulopustular rosacea and were between 40 and 50 years old; only two studies27 and 28 addressed ocular rosacea. Of the 71 studies, 41 were excluded because allocation concealment was inadequate, the study was not blinded, the dropout rate was more than 10%, or other major methodologic flaws,11, 12, 18, 19, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84 and 85 or because they were awaiting assessment.13, 86 and 87 Of the 29 included studies, 8 were classified as high quality.23, 26, 28, 29, 30, 32, 34 and 36 The remaining 21 trials were of intermediate quality.21, 22, 27, 31, 33, 35, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49 and 50 In only 14 of the 29 trials23, 26, 28, 29, 30, 32, 34, 35, 36, 40, 42, 47 and 49 was there adequate blinding of treatment allocation. Blinding of outcome assessment was demonstrated in all except two studies.27 and 43 Intention-to-treat analysis was used in 17 of the 29 trials.21, 22, 23, 26, 28, 29, 30, 31, 32, 34, 35, 36, 37, 38, 39 and 49 For 14 studies the variability (SD or SE) of continuous measurements were completely or partially lacking, making these data unusable in a meta-analysis.21, 30, 32, 33, 34, 36, 37, 38, 41, 42, 43, 46, 48 and 50
Analysis
The treatments could be categorized into 5 groups: topical metronidazole (15 trials),21, 22, 23, 27, 29, 30, 32, 33, 36, 38, 39, 42, 43, 48 and 50 oral antibiotics (8 trials),23, 26, 28, 40, 44, 45, 47 and 50 topical azelaic acid (6 trials),22, 31, 35, 36 and 49 topical benzoyl peroxide combined with topical antibiotics (2 trials),34 and 43 and other therapies (4 trials).37, 41, 44 and 46 Five trials included comparisons in more than one category.22, 23, 36, 43 and 44 Even within these therapeutic categories, making comparisons and pooling of data was problematic because of heterogeneous study designs, skewed data, missing variability, and differences in comparators or dosing regimens. Only data on outcome measures from trials on topical metronidazole, topical azelaic acid, and oral tetracycline could be pooled. Most studies used numbers of papules or pustules as an outcome measure rather than a more clinically relevant measure, such as participant assessment of appearance. Below is a summary of the most important conclusions. For details and full reporting of the data, please refer to the complete Cochrane review as published in the Cochrane Library.1
Metronidazole
Topical metronidazole versus placebo
Nine trials assessed the efficacy of topical metronidazole versus placebo.21, 27, 29, 30, 32, 33, 38, 39 and 42 The treatment period ranged from 8 to 9 weeks in each trial, except for that of Dahl et al,21 which was 6 months. Three studies addressed self-assessed improvement of rosacea severity.30, 32 and 42 Only data from two studies30 and 42 could be pooled (Fig 1, A) and there was clear evidence that metronidazole was more effective than placebo. Bleicher et al32 confirmed these data (OR 7.0; 95% confidence interval [CI] 2.5-20.0). Data on physician's global evaluation concerning improvement were similar to the patient-assessed measures in favor of metronidazole (OR 7.01; 95% CI 3.56-13.81).30, 33 and 42 The other studies showed comparable data.21, 27, 29, 32 and 38
Fig 1. Meta-analytic comparisons of participant-assessed improvement between topical metronidazole and placebo (A), oral tetracycline and placebo (B), and topical azelaic acid and placebo (C). Modified1 and used with permission. CI, Confidence interval; OR, odds ratio.
Most of the adverse events mentioned were mild, including pruritus, skin irritation, and dry skin. There were no significant differences in the number of adverse events between groups.
Topical azelaic acid versus topical metronidazole
There was no statistically significant difference in the patient self-assessment between topical azelaic acid and topical metronidazole.22 and 36 However, physicians rated the azelaic acid group more improved (OR 1.84; 95% CI 1.10-3.09).36 The number of adverse events was lower in the metronidazole group (OR 4.56; 95% CI 2.07-10.03).36 However, the severity of adverse events in both groups was reported as mild to moderate and mostly transient.
Topical metronidazole versus oral tetracycline
In two 8-week studies23 and 50 no statistically significant treatment difference was seen between metronidazole cream and (oxy)-tetracycline.
Metronidazole plus sunscreen (sun protection factor 15) versus placebo
A poorly designed study favored metronidazole plus sunscreen over placebo.48
Topical metronidazole versus topical permethrin versus placebo
Koçak et al39 investigated the efficacy and safety of permethrin for the treatment of rosacea. Permethrin was inferior to topical metronidazole because it showed no effect on pustules.
Benzoyl peroxide 5%/erythromycin 3% gel versus metronidazole gel
No significant difference was shown between the two therapies in 4 weeks (OR 0.92; 95% CI 0.21-4.11).43
Oral antibiotics
Oral metronidazole versus oral oxytetracycline
In one study, oral metronidazole and oral oxytetracycline were not statistically different at 12 weeks by both physician and patient assessment.45 No adverse events were reported in either group.
Tetracycline versus placebo
One trial28 compared oral oxytetracycline with placebo, and in two trials40 and 47 oral tetracycline was compared with placebo. These are both (older) tetracyclines with a similar molecular structure and the same pharmacokinetic and pharmacodynamic profile and so the results were pooled. Study duration ranged from 4 to 6 weeks. Bartholomew et al28 addressed treatment efficacy in ocular rosacea.
There was insufficient evidence of any advantage of tetracycline over placebo according to patients' assessment (Fig 1, B).40 However, the dropout rate was unclear and the data were skewed with large variability. By physician assessment, tetracyclines are far more effective than placebo in the treatment of rosacea (OR 6.06; 95% CI 2.96-12.42). Repeated courses of treatment with the same dose achieved lasting remission 3 to 6 months after stopping treatment.28
Clarithromycin and omeprazole versus placebo
These data were skewed with large variability and, thus, it is impossible to draw conclusions about this therapy.26
Azelaic acid
Azelaic acid versus placebo
Four trials compared azelaic acid with placebo.31, 35 and 49 The treatment period ranged from 9 to 12 weeks. Three studies31 and 49 showed a clear improvement in the azelaic acid group as rated by both physicians and patients (Fig 1, C). A split-face, within-patient study35 confirmed these results (marginal OR 30.1; P < .0003).
The data on lesion counts did not include variability and the data in the study by Carmichael et al35 were skewed.
More side effects were reported in the azelaic group (11.5%) versus the placebo group (5.7%) (OR 1.61; 95% CI 0.89-2.92).31 and 49 The same holds true for the study of Carmichael et al.35 Side effects were considered mild and transient with burning, stinging, and irritation being reported most frequently.
Benzoyl peroxide with antibiotics
Benzoyl peroxide 5%/clindamycin 1% gel versus placebo
The mean scores at 12 weeks for patient's global assessment in the study of Breneman et al34 were 1.54 (much to slightly better) in the benzoyl peroxide and clindamycin group versus 2.50 (slightly better to same) in the placebo group (authors state P = .0002). The mean scores at 12 weeks for physician's global assessment were 1.85 (marked to definite improvement) versus 2.96 for placebo (minimal improvement) (authors state P = .0026).
The data showed large variability and some data were missing. Most data were skewed. Treatment-related adverse events included site burning and itching, both well-known side effects of benzoyl peroxide.34 The same study using photographic assessments as outcomes came to similar same conclusion.51
Other
Benzoyl peroxide acetone versus placebo
At 4 weeks, benzoyl peroxide showed an improvement on the physician's global evaluation score compared with placebo (OR 3.17; 95% CI 1.08-9.31).41 The other measurements were also in favor of benzoyl peroxide (P < .05). Irritation and burning were frequently reported in both groups.
Oral metronidazole and topical hydrocortisone 1% cream versus oral placebo and topical hydrocortisone 1% cream
The physicians considered 10 of 14 participants treated with oral metronidazole improved versus only 2 of 13 participants on placebo (OR 13.75; 95% CI 2.05-92.04).44 Only limited data were given in this study.44
Rilmenidine versus placebo
Both the patients and the physicians believed that there was no significant difference between rilmenidine and placebo; neither treatment was effective.37
Sodium sulfacetamide/sulfur versus placebo
The percentage of participants who considered themselves improved was 90% in the sodium sulfacetamide 10%/sulfur 5% group versus 58% in the placebo group (authors state P < .001).46 The physicians shared this opinion. Adverse events were reported in 38% versus 29%, respectively. Application site reactions such as dryness, erythema, and pruritus were the most commonly reported adverse events. It was unclear how many participants started in each group or how improvement was defined, and for continuous measurements the variability was large and the data skewed.46
Discussion
There were significant limitations in the quality of evidence available for most treatments. Although the clinical design of the included studies was in theory adequate, closer examination revealed that the quality of reported data was often low. Table III and Table IV summarize recommendations for future rosacea studies.88 It should be noted that although split-face studies can be efficient, they are subject to potential biases. Contamination may occur if active cream is accidentally transferred onto the placebo side. Furthermore, a treatment may have systemic effects, beneficial or harmful, which will affect both sides.
Table III.
Data to be included in future rosacea studies
Well-designed RCT with reporting the following88
Proper description of randomization procedure and allocation concealment
Data presented with appropriate summaries and analysis (including variability)
The number of participants who started in and dropped out from each group
Outcomes primarily based on: patient's opinion of treatment efficacy, quality of life, and patient-assessment
Physician's opinion, reflected by global evaluation, lesion counts, and assessment of telangiectasia
Use of intention-to-treat analysis
RCT, Randomized controlled trial.
Modified1 and used with permission.
Table IV.
Questions for which evidence is lacking in the literature
1. What is the efficacy and safety of commonly used treatments for rosacea (eg, tetracycline, minocycline, doxycycline, isotretinoin, and laser therapy)?
2. What is the efficacy and safety of treatments for ocular rosacea?
3. Is there any efficacy of dietary measures and/or sun-protective measures in the treatment of rosacea?
4. What is the efficacy and safety of benzoyl peroxide alone or in combination with topical antibiotics for rosacea?
5. Is permethrin effective and safe for rosacea treatment?
Studies to answer these questions should meet the criteria mentioned in Table III.
Our principal outcome measure, quality of life, was not assessed in any of the studies and only a few studies assessed the participant's own opinion. It is interesting to note that the investigators were more satisfied at the end of the study than the participants.35, 36 and 40 For other diseases it is often the reverse. This may have implications for clinicians, as a patient's perception of a lack of sufficient efficacy can impact compliance and may lead to the use of alternative therapies. Topical metronidazole and azelaic acid appear to be effective and safe for short-term use, with the rate of adverse events in the placebo groups being similar to the active treatment groups. With regard to tetracycline, only 3 studies28, 40 and 47 could be included in this review, only one of which assessed the opinion of the participants; however, this study failed to detect any difference from placebo. It is possible that in this case the study duration of 6 weeks was too short.
There were no studies evaluating other treatment options, such as erythromycin, dapsone, and topical tretinoin,7, 16, 89, 90, 91 and 92 that met the inclusion criteria. Three studies were included using benzoyl peroxide alone or in combination with topical antibiotics.34, 41 and 43 Unfortunately, the quality of these studies was suboptimal. The same holds true for the study with permethrin.39 Both benzoyl peroxide and permethrin are well-known drugs and further investigation in the treatment of rosacea may be beneficial.
No studies could be included addressing dietary or sun-protective measures; however, two studies did combine treatment with a sun protection factor.39 and 48 Although not really substantiated, most people with rosacea are given the advice to avoid trigger factors, (eg, spicy foods, alcohol, and sunlight).
Only two trials could be included on treatment of ocular rosacea,27 and 28 even though almost 60% of people with rosacea have ocular involvement.27, 66, 79 and 91 Although often mild, the ocular presentation can be both severe and debilitating. There was insufficient evidence for the efficacy of topical metronidazole.27 Oral oxytetracycline seems to be effective for ocular rosacea,28 although only the opinion of the physician was reported.
A very interesting treatment seems to be low-dose doxycycline (20 mg twice a day),77 and 93 which is a subantimicrobial dose that reduces inflammation. Other potential advantages of this treatment include lessening the risks of Propionibacterium acne's resistance to tetracyclines and lowering the incidence of tetracycline-induced adverse events. Unfortunately, even though they are commonly used to treat rosacea, no RCTs evaluating doxycycline, minocycline, isotretinoin, or laser therapy could be included in this review (most often because of inadequate study design). There is an urgent need for better quality, adequately designed RCTs on the commonly used treatments for rosacea.