Medicine.org.hk

Evidence Based Medicine in Dermatology:
Minocycline and Acne: from Clinical to Literature Review
Social Hygiene Service (Dermatology), Department of Health, Hong Kong ABSTRACT
Minocycline is one of the tetracyclines frequently used in the treatment of acne. There is a lack of consensus
over the relative risk and benefits of minocycline in the treatment of acne. Starting from a clinical scenario, a
literature search was initiated for the best available evidence on this subject. A Cochrane systemic review
located 27 randomized controlled trials on the efficacy and risk of minocycline in treating mild to moderate
acne. The trials were generally of insufficient size and quality to meet requirements of the systemic review.
Although minocycline is an effective treatment for mild to moderate acne vulgaris, there was no reliable
randomized controlled trial evidence to justify its continued first line use, given the price and concerns about
safety that still remains.
Keywords: Acne, minocycline, randomized controlled trials
INTRODUCTION
Steps in practicing Evidence Based Medicine
Evidence based medical practice can be divided Minocycline is one of the tetracycline antibiotics frequently used in the treatment of acne vulgaris. It canbe taken in a more convenient once or twice daily dose 1. To ask the clinical question in a format that can be compared with the generally more frequent dosing of answered. This identifies gaps or area of deficiencies other tetracyclines, but it is more expensive. There have in clinical knowledge that further search of evidence also been concern about the safety of minocycline following case reports of death after taking the drug.1 2. Search for the best external evidence.
As there is a lack of consensus among dermatologists 3. Critically appraise evidence for the validity and on the relative risk and benefits of minocycline in treating acne, a review on this subject should be based 4. Apply the evidence into clinical practice.
on the best available evidence. The material presented 5. Evaluate self-performance in the practice of evidence in this article was based on journal presentation of articles by author in April 2001, including CochraneSystemic Review on this subject.2 In this review article, we shall concentrate on the first three steps in evidence based medical practice.
CLINICAL SCENARIO
Dr. C. W. SuTang Shiu Kin Social Hygiene Clinic1/F Tang Shiu Kin Hospital A 20-year-old university student had mild acne for several years, requiring no therapy or intermittent 2.5% benzoyl peroxide aqueous obtained over the counter.
Vol.9 No.4, December 2001 159
Two months ago he experienced a flare up of acne while SEARCH FOR THE BEST EXTERNAL
preparing for his final examination. He visited a general EVIDENCE
practitioner who prescribed minocycline 100 mg dailyand referred him to the dermatology clinic. Now his The types of evidence available can be broadly acne has improved, but he wishes to know whether to 1. Systemic review of well designed studies, including ASKING AN ANSWERABLE
2. Well designed studies – randomized controlled trials, CLINICAL QUESTION
The questions that may arise from daily clinical practice can be divided broadly into the following areas:4 Evidence resource
External evidence from the literature may come Hand searched review articles from recent or current issues of major dermatology journals, such as the Archives of Dermatology, Journal of American Academy of Dermatology, British Journal ofDermatology, and International Journal of Dermatology.
From the clinical scenario above, the clinical Although these may provide useful information of problem and question is obviously about therapy, harm topical interest, they do not necessarily cover the and cost. The patient had received two months of intended search topic unless by chance.
minocycline already. Now he wishes to know whetherto continue with oral minocycline for full six months S t a n d a r d t e x t b o o k s g e n e r a l l y p r o v i d e or to substitute with an alternative, such as another oral comprehensive coverage on a wide range of topics, tetracycline or topical antibiotic therapy, after taking unfortunately the information they contained become into consideration the relative efficacy, side effects, and Electronic databases such as the Medline allow rapid search over many journals indexed by the US Elements of the clinical question
National Library of Medicine in the Index Medicus.
The components of a clinical question on therapy Other electronic databases available on CD-ROM or on can be divided into the following areas:4 (1) The patient the internet world wide web include Cochrane Systemic is a male student with mild acne who had recent flare Review of Randomized Controlled Trials, Journal of up. (2) The intervention being investigated is acne Evidence Based Medicine, and American College of therapy with oral minocycline. (3) Comparative intervention will, for example, be other oral tetracyclines,or topical antibiotics such as clindamycin. (4) The Even with librarian or experienced searchers of outcome measures will be improvement in acne, with electronic databases, a Medline search does not always objective and subjective scores as assessed by the locate all the relevant articles indexed. The sensitivity physician and the patient himself, as well as quality of of search unfortunately decreases further with lack in life scores. In addition, the safety and tolerability of 160 Hong Kong Dermatology & Venereology Bulletin
A search of the Medline using the keywords acne The studies included in the systemic review were and minocycline therapy yielded over 200 published randomized controlled trials, assessing the efficacy of articles, mostly uncontrolled trials or case reports. To minocycline at any dose in comparison with control, narrow down the search to randomized controlled trials which may be a placebo or another active acne only, Medline search yielded two articles. One article treatment. The participants should have inflammatory was a systemic review of randomized controlled trials acne vulgaris on face and/or upper trunk, which may on minocycline for acne vulgaris: efficacy and safety be papulopustular, polymorphic or nodular acne. The (Cochrane Review) published by Cochrane Skin Group outcome measures evaluate clinical efficacy and patient in Cochrane library.2 The abstract was available on the acceptability in a defined way (for example, lesion following web address: http://www.update-software.
counts, acne severity scores, physician's global com/default.htm. The second article was a randomized evaluation and patients' self-assessment).
contr olled trial compar ing doxycy cline withminocycline in the treatment of acne vulgaris.5 It was Trials were not excluded on the basis of language.
one of the studies included by Cochrane Systemic Non-English studies were translated to English if it was Review mentioned above for analysis.
not apparent from their original language whether theywere randomized control trials or not. The resources Cochrane Collaborative Review Groups consist of and databases on which the studies were located include a panel of international experts initiated and coordinated MEDLINE, EMBASE, Biosis, Biological Abstracts, by health care epidemiologists at Oxford University UK, International pharmaceutical abstracts, Cochrane Skin to perform systemic reviews of randomized controlled Groups Trial Register, Thesis Online, BIDS ISI Science trials in a specific field. There were 50 different groups Citation Index, BIDS Index to Scientific and Technical starting alphabetically from acute respiratory infections proceedings. Other methods to ensure that important to wound care. Cochrane Skin Group (Group 46) was studies were not missed included scanning references responsible for systemic reviews in dermatology.
of articles already retrieved, hand searching of majordermatology journals; last but not least personalcommunication with trialist and drug companies on CRITICAL APPRAISAL OF EVIDENCE
unpublished data to reduce publication bias.
FOR VALIDITY AND IMPORTANCE
The criteria used to assess the validity of a systemic Description of studies
There were a total of 72 studies of minocycline in acne, 32 studies were randomized control trials, two The objectives of the Cochrane systemic review studies were duplicate, two were interim report, and on minocycline for acne vulgaris: efficacy and safety, one study compared minocycline with streptokinase were to collate and evaluate evidence on the clinical versus placebo. Therefore only 27 studies met the efficacy of minocycline in the treatment of inflammatory primary inclusion criteria, with 3031 subjects in total, acne vulgaris. It also compared the efficacy of and sample sizes varying from 18 to 325 subjects minocycline with other drug treatments for acne and estimated the incidence of adverse drug reactions.
Methodological quality of studies
Table 1. Assessing the validity of systemic overview
Two reviewers independently assessed each study 1. Did the overview address a focused question? to see if it met the inclusion criteria for review. The methodology and validity of included studies were 2. Were the criteria used to select articles for inclusion appraised according to assessment criteria listed in 3. Is it unlikely that important relevant studies were missed? Tables 2 and 3.7,8 The methodological components 4. Were the methodology and validity of included studies concern overall trial design and execution.7 The substantive components are specific to the topic under 5. Were the results similar from study to study? Vol.9 No.4, December 2001 161
Table 2. Appraisal of included studies-methodological
Five studies failed to mention stopping previous components
medications prior to entry into trial. Nine trials specifically disallowed concomitant therapy that might 2. Correct randomization protocol, allocation concealed 3. Baseline comparability of groups4. Withdrawals (number and reason) clearly stated; all Most trials tried to show that different treatment patients enrolled in the trial accounted for groups were comparable at baseline. For example, age 5. Appropriate method of analysis (for example, Intention (16/26 trials), sex (15/23 trials), weight (6/26 trials), lesion count or scores (15/26 trials), duration of acne(5/26 trials), and acne grade or severity (11/26 trials).
Table 3. Appraisal of included studies-substantive
In the 27 studies a total of 50 different outcome components
measures were used. Most trials used more than one outcome measure. Ten trials used some acne grade or 2. Explicit and appropriate inclusion/exclusion criteria overall severity score, 20 trials used some form of lesion 3. Concomitant medication prohibited, monitor patient count, 15 trials included separate counts for inflamed and non-inflamed lesions. Categorical outcome 4. Standardize skin hygiene routine, control for ultraviolet measures such as physicians global evaluation and patients global assessments were reported respectively 5. Uniform site of evaluation6. Number and timing of assessments standardized in 10 trials. Three trials used visual analogue scale to 7. Evaluation of inter-assessor variability obtain patients assessment, two trials used quality oflife questionnaire, and one trial evaluated patientsatisfaction.
The main theme was heterogeneity among the 27 It was unclear how withdrawals or patients who trials, with variety and differences rather than consensus failed to attend one or more visits were dealt with. Few and standardization. Fourteen studies were conducted studies specified how many of the patients enrolled had in more than one center (Dermatology clinics, Air Force, been included in the final analysis. Most studies were college and university volunteers, and general practice).
analyzed on a per protocol basis, only 7/26 trials used Only two studies performed power calculation to intention to treat analysis, and three used both methods.
estimate sample size of trial. Twenty of 27 trials wereof insufficient size to detect any real difference between Twenty-six trials reported data on adverse events, treatments if one existed. Only five trials mentioned side effects or tolerance. How unwanted effects were how the randomization procedure was carried out.
identified were often not given or rarely adequate.
Eleven studies were not blinded, these were not excluded Sometimes they were obtained by asking the patient but analyzed in consideration of bias associated with directly, by patients' spontaneous reporting of subjective open trials. The duration of trials varied from five to 24 symptoms such as dizziness, or physicians observation weeks, the majority of trials (14/27) were for 12 weeks of objective signs such as urticaria. There was confusion about definition, with arbitrary decisions about whichadverse reactions were possibly drug related. In six Entry (inclusion) criteria were reported in all trials, studies, side effects were only reported if it led to but were not standardized across trials. Eight trials specified whether mild, moderate, severe and nodularacne were included, six trials had no statement of diseasestatus as an entry criterion, and one trial simply stated Minocycline comparators
acne that merited antibiotic therapy. Exclusion criteria were mentioned in all but five trials. These included minocycline versus placebo, another trial was a hypersensitivity, pregnancy, and lactation.
minocycline dose response study. Seven trials comparedminocycline with other tetracyclines or oxytetracycline.
The washout period of previous acne treatments Five trials compared minocycline with doxycycline, and on entry to trials varied from 48 hours to four months.
three trials compared minocycline with topical 162 Hong Kong Dermatology & Venereology Bulletin
clindamycin. Other comparators include isotretinion, Minocycline versus topical clindamycin
Diane, topical fusidic acid, topical erythromycin and Two trials had similar results for both minocycline zinc. The methodological deficiencies of individual and clindamycin, but it was uncertain whether the trials were discussed in the Cochrane Review.4 product was applied to all of the affected areas of facewhere spots were.18,19 One trial showed superiority oftopical clindamycin applied to entire face. But this did Minocycline versus placebo
not reach statistical significance because large range of This study compares minocycline 200 mg daily lesion counts and small number of patients were for one week followed by 100 mg daily for four weeks with placebo treatment of identical appearance. It wasa randomized double-blind cross-over study of fiveweeks for each arm, with no washout period in between.9 Adverse reactions
During the first phase, minocycline demonstrated There were 1230 patients from 22 studies who significant reduction in summed weighted acne lesion received minocycline. The total adverse reactions were 137 (11.1%), with 36 (2.9%) led to withdrawal oftherapy. The most common were gastrointestinaldisturbance, followed by vertigo or dizziness, vaginal Dose response of minocycline
candidiasis, and abnormal pigmentation. However the This was a randomized double-blind control trial reported incidence of common side effects may not be comparing minocycline 100 mg daily for eight weeks reliable due to inadequacies in collection and reporting with minocycline 100 mg daily for two weeks followed methods. The lack of a denominator in nearly all studies by 50 mg daily for six weeks.10 The study found no means that risk for minocycline compared to other significant difference between dosage regimens in tetracyclines cannot be reliably compared.
outcome measures using either per protocol orintentional to treat analysis. However, due to the short Rare but serious side effects such as autoimmune duration of study (eight weeks only), inference could disorders may not be detected. The study might not be not be made concerning the relative efficacy in long large enough to detect rare adverse reactions (with term treatment. There were no adequate dose response incidence <1 in 1000) and was not controlled. Although studies to confirm that 200 mg and 100 mg per day case reports suggested that minocycline had greater risk were equivalent in terms of clinical efficacy.
of severe side effects, this might reflect current interestand selective reporting, for example, minocyclineinduced auto-immune hepatitis and LE-like syndromes.1 Minocycline versus other tetracyclines
One trial found that significantly more patients A case control study involving 27,688 acne patients show improvement in their acne after receiving four from a primary care research database found that 29 weeks of minocycline instead of oxytetracycline.11 Two (0.1%) developed LE-like syndromes, 27 of whom were trials showed statistically significant difference in favour females.21 Comparing with age and sex matched of minocycline over tetracycline in acne after six controls, minocycline was associated with 8.5 fold risk weeks.12,13 In all cases where initial response to (95% confidence interval 2.1-35) of developing lupus minocycline was faster, the magnitude of reduction in erythematosus, and other tetracyclines with 1.7 fold risk acne severity at the end of treatment (12-24 weeks) was only (95% confidence interval 0.4-8.1). The absolute risk was 52.8 cases per 100,000 prescriptions.
All five trials that compared patients receiving minocycline and doxycycline showed no overall Limitations of overview
difference in acne improvement between the drugs.5,14-17 Systemic overview attempts to review individual There was no evidence of earlier onset of acne studies objectively, minimizing subjective bias in the improvement with minocycline compared with selection of studies, analysis of data, and in drawing doxycycline. However, pooling of data was impossible conclusions. It has clearly focused objectives, due to variability of dosage and methodological design.
predetermined selection criteria to retrieve studies, Vol.9 No.4, December 2001 163
exhaustive search of literature to avoid publication bias, the combination of individual's clinical experience with and translation of foreign languages to avoid language best available external evidence is important. This bias. At least two independent reviewers critically personal experience is required to make judgement as appraised the validity of studies, and where possible to to whether the external evidence found may be pool study results in a systemic fashion (for example, appropriately applied to the clinical situation in hand, mathematically in meta-analysis before drawing taking into account individual patient characteristics and This Cochrane systemic overview was limited by the quality of the individual studies it could find; Conclusion
including heterogeneity of primary studies due to Evidence based medical practice begins by asking methodological insufficiencies, inadequacies of reported an answerable question arising from daily clinical data, insufficient numbers of patients. The studies were practice, continues with searching for the best available generally of inadequate duration, majority lasted only external evidence and critically appraising this evidence 12 weeks, so that assumptions could not be reliably for its validity and importance, and eventually its made on long-term therapy. The poor characterization of patients made subgroup analysis impossible. Therewas a lack of adequate outcome data for analysis.
Starting from a clinical scenario, a literature search Standardized outcome measures were not available, and was initiated for the best available external evidence on pooling of results was impossible. It was also not the relative risk and benefits of minocycline in the possible to examine the impact of study design on results treatment of acne vulgaris. A Cochrane systemic review (especially the degree of blinding), as many of the located 27 randomized controlled trials. They were generally of inadequate size and quality to meetrequirements of the systemic review. Althoughminocycline is an effective treatment for mild to Conclusions of overview
moderate acne vulgaris, there was no reliable The systemic review concluded that there was no randomized controlled trial evidence to justify its clear cut and unbiased evidence to support the routine continued first line use, given the price and concerns first-line use of minocycline in the treatment of acne.
Minocycline 100 mg daily is an effective treatment ofmoderate acne, but no study has shown conclusivelyany important clinical difference in efficacy between Table 4. Applying evidence based medicine in clinical
the various tetracyclines in acne therapies. There was practice
insufficient information to make any recommendations Therapy
concerning the appropriate dose of minocycline that 1. Can the results be applied to my patient care? should be used. The relative safety of tetracyclines could Is my patient so different from those in the trial that its not be adequately determined, and there was an inherent inability of the studies to detect rare events. However, How great would be the potential benefit of therapy one case control study suggested minocycline in acne therapy was associated with higher risk of lupus 2. Were all clinically important outcomes considered? erythematosus syndromes than other tetracyclines.
Are the benefit worth the harms and the costs?What alternative treatments are available? 3. Is my patients' values and preferences satisfied by the APPLYING EVIDENCE IN
Do my patient and myself have a clear assessment of their CLINICAL PRACTICE
values and preferences?Are they met by this regimen and its consequences? When applying the findings and conclusions of external evidence to patient care in the clinical setting, 164 Hong Kong Dermatology & Venereology Bulletin
References
vulgaris. Arch Dermatol 1982;118:989-92.
14. Olafsson JH, Gudgeirsson J, Eggertsdottir GE, Kristjansson 1. Gough A, Chapman S, Wagstaff K, Emery P, Elias E.
G. Doxycycline versus minocycline in treatment of acne Minocycline induced auto-immune hepatitis and systemic lupus vulgaris: a double blind study. J of Dermatologic Treatment erythematosus-like syndrome. BMJ 1996;312:72-3.
2. Gardner SE, Eady EA, Popescu C, Newton J, Li WP.
15. Lorette G, Belaich S, Beylot MC, Ortonne JP. Doxycycline Minocycline for acne vulgaris: efficacy and safety. Cochrane 50mg/day versus minocycline 100mg/day in the treatment of acne vulgaris [French]. Nouvelles Dermatologiques 1994;13: 3. Sackett DL, Haynes RB. On the need for evidence-based medicine. Evidence- Based Medicine 1995;1:4-5.
16. Waskiewicz W, Grosshans E. Treatment of acne vulgaris with 4. Sackett DL, Richardson WS, Rosenberg W, Haynes RB. How cyclines of second generation: a comparison of doxycycline to answer clinical questions you can answer? In: Evidence Based 50mg daily versus minocycline 100mg daily [French]. Nouvelles Medicine. How to practice and teach EBM. New York: Churchill 17. Schollhammer M, Alirezai M. Compar ative study of 5. Laux B. Treatment of acne vulgaris. A comparison of doxycycline lymecycline, minocycline and doxycycline in the treatment of versus minocycline [German]. Hautarzt 1989;40: 577-81.
acne vulgaris. Realites Therapeutiques en Dermato-Venerologie 6. Oxman AD, Cook DJ, Guyatt GH. Users' Guides to the Medical Literature. VI. How to use an overview. JAMA 1994;272:1367- 18. Drake L. Comparative efficacy and tolerance of Cleocin T topical gel (clindamycin phosphate topical gel) versus oral minocycline 7. The Standards of Reporting Trials Group. A proposal for the in the treatment of acne vulgaris. Data on file (Pharmacia and structured reporting of randomised controlled trials. JAMA 1994; 19. Peacock GE, Price C, Ryan BE, Mitchell AD. Topical 8. Eady EA, Cove JH, Joanes DN, Cunliffe WJ. Topical antibiotics clindamycin compared to oral minocycline in treatment of acne for the treatment of acne vulgaris: a critical evaluation of the vulgaris. A randomized observer-blind controlled trial in three literature on their clinical benefit and comparative efficacy. J university health centres. Clin Trials J 1990;27:219-28.
Dermatological Treatment 1990;1:215-26.
20. Sheehan-Dare RA. Papworth-Smith JW, Cunliffe WJ. A 9. Hersle K, Gisslen H. Minocycline in acne vulgaris: a double- comparative study between topical clindamycin and oral blind study. Curr Ther Res Clin Exp 1976;19:339-42.
minocycline in the treatment of acne vulgaris. Round Table Series 10. Dreno B. Personal Communication. 1998.
11. Blechschmidt J, Engst R, Hoting E, et al. Treatment of papulo- 21. Sturkenboom MC, Meier CR, Jick H, Stricker HC. Minocycline pustular acne- comparison of the efficacy and tolerance of and lupuslike syndrome in acne patients. Arch Intern Med 1999; minocycline and oxytetracycline. [German]. Munch Med 22. Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Can 12. Khanna N. Treatment of acne vulgaris with oral tetracyclines.
you apply this valid, important evidence about a treatment in Indian J of Dermatol Venereol and Leprol 1993;59:74-6.
caring for your patient? In: Evidence Based Medicine. How to 13. Hubbell CG, Hobbs ER, Rist T, White JW. Efficacy of practice and teach EBM. New York: Churchill Livingstone, 1997: minocycline compared with tetracycline in treatment of acne Vol.9 No.4, December 2001 165

Source: http://www.medicine.org.hk/hksdv/journal/200112-04.pdf