38 Practical Dermatology June 2005

W ith the continuing rise in clinical reportsof cutaneous MRSA (Methicillin-resist-ant Staphylococcus aureus) infections inthe community and public health offi-cials issuing consistent warnings about

the risks of antibiotic overuse and misuse, dermatologists wellknow the importance of selective and appropriate use of oralantibiotics. Yet dermatologists continue to depend on systemicantibiotics to manage acne and rosacea. Despite compellingevidence of the danger of resistance, antibiotics remain a safe,effective, and very efficient treatment for certain patients withacne and rosacea. In ongoing efforts to capitalize on the thera-peutic benefits of oral antibiotics in acne and rosacea whileminimizing the risks associated with long-term use, interest hasfocused on the potential benefits associated with the use of sub-

antimicrobial-dose antibiotics, specifically doxycycline.Though not approved for the treatment of acne or rosacea, sub-antimicrobial-dose doxycycline (SDD) demonstrates effective-ness in both conditions.1,2,3,4 I’ll review what we know aboutSDD and how that knowledge may influence the provision ofdermatologic care right now.

An Historical PerspectiveSystemic antibiotics, particularly those of the tetracycline class,are well known for their efficacy in treating acne and rosacea.In the case of acne, early interest in the use of systemic antibi-otics aimed at diminishing P acnes colonization. While the roleof P acnes in the pathogenesis of acne vulgaris is well docu-mented and universally accepted, research and clinical experi-ence have established the importance of other pathogenic

June 2005 Practical Dermatology 39

By Joseph B. Bikowski, MDSewickley, PA

I nflammatory response has beenimplicated in the pathogenesisof both acne and rosacea.

Though our understanding of theinflammatory process at the molec-ular level is not complete, researchhas elucidated many of the keysteps of the inflammatory cascade.A general explanation of chronicinflammation as it relates to acne,rosacea, and SDD follows.

In response to inflammatorystimuli, the body recruits activatedneutrophils and macrophages to thearea. Stimuli may include bacteria(proposed in both acne and rosacea)or infestation by other micro-organ-isms (such as Demodex in rosacea).Additionally, elastic and collagentissue damage may initiate inflam-matory processes in rosacea. Otherunproven or as yet unidentifiedinflammatory instigators may alsoplay a role in both diseases.

Activated neutrophils mayexpress matrix metalloproteinases(MMPs). Cytokines or T-helper cellsare also produced in the early stagesof inflammatory response. Theactivity of hydrolytic MMPs pro-duces tissue destruction and is sus-

pected to lead to free-radical gener-ation that can in turn produce fur-ther inflammation.

A number of MMPs have beenidentified, and they are necessaryfor normal development and woundhealing. But MMPs are also implicat-ed in a number of pathologicalprocesses, including the spread ofmetastatic cancer cells, arthriticconditions, atherosclerosis, andneuroinflammation. Two types ofMMPs—collagenases and gelatinas-es—have been implicated in theinflammatory processes of acne androsacea.

Tetracyclines and tetracyclineanalogs demonstrate an ability tointerfere with MMPs and down-reg-ulate cytokines. In the case ofrosacea, inhibition of gelatinasesmay protect capillary wall leakage,reduce swelling, and improve con-nective tissue integrity.

Perhaps the greatest body of evi-dence for this anti-inflammatoryeffect of tetracyclines and tetracy-cline analogs comes from the studyof periodontal disease. Collagenaseand gelatinase are implicated in theinflammation and bone loss associ-

ated with periodontitis. Studiesdemonstrate that various non-antimicrobial tetracycline analogsinhibit collagenase and gelatinaseand lead to decreased expression ofcytokines, including TNF, IL-1, andIL-6. This activity accounts for theclinical effect of Periostat (doxycy-cline 20mg BID, CollaGenex) indecreasing the inflammation asso-ciated with periodontitis and delay-ing progression of the disease.

Inflammatory Response: A Thumbnail Sketch

1. Mandal M, Mandal A, Das S, et al. Clinical impli-cations of matrix metalloproteinases. Molecular andCellular Biochemistry 2003.252:305–329.

2. Rifkin BR, Vernillo AT, Golub LM. Blocking periodon-tal disease progression by inhibiting tissue-destructiveenzymes: a potential therapeutic role for tetracyclinesand their chemically-modified analogs. J Periodontol.1993 Aug;64(8 Suppl):819-27.

3. Milliken L. The Proposed InflammatoryPathophysiology of Rosacea: Implications for Treatment.SKINmed 2(1):43-47, 2003.

4. Bikowski JB. Subantimicrobial dose doxycycline foracne and rosacea. Skinmed. 2003 2(4):234-45.

5. Golub LM, Ramamurthy NS, Llavaneras A, et al. Achemically modified nonantimicrobial tetracycline (CMT-8) inhibits gingival matrix metalloproteinases, periodon-tal breakdown, and extra-oral bone loss in ovariec-tomized rats. Ann N Y Acad Sci. 1999 30;878:290-310.

6. Ramamurthy NS, Rifkin BR, Greenwald RA, et al.Inhibition of matrix metalloproteinase-mediated periodon-tal bone loss in rats: a comparison of 6 chemicallymodified tetracyclines. J Periodontol. 2002Jul;73(7):726-34.

factors, including inflammation, sebum production, and faultykeratinization. The fact that antibiotics confer anti-inflamma-tory benefits in addition to their antimicrobial action against P acnes further supported their use in acne management. Thefirst reports of P acnes resistance to antibiotics emerged in the1980s with subsequent reports drawing worldwide attention tothe issue.5,6

Working on the assumption that rosacea—once termedacne rosacea—was related to acne vulgaris, clinicians began touse antibiotics with good results in managing this condition.

We now believe that the clinical benefit that results from theuse of antibiotics in rosacea is due at least in part to the anti-inflammatory action of these agents. Some have proposed abacterial etiology in rosacea, speculating that rosacea may berelated to helicobacter pylori infection, P acnes, or bacteria fromDemodex mites.7,8,31,32,33 Several studies fail to support the pro-posed link between H. pylori and rosacea.9,10 In fact, one studyfound the prevalence of H. pylori was similar between rosaceapatients and controls and showed that treatment with amoxi-cillin, clarithromycin, or omeprazole failed to significantly

Subantimicrobial Dose Doxycycline

40 Practical Dermatology June 2005

decrease erythema or provide more than temporary improve-ment in papulopustules in rosacea patients.9 There is indirectevidence for a link between P acnes and rosacea. Studies showthat once-daily benzoyl peroxide/clindamycin topical gelreduces the papulopustular component of rosacea.32,33

The Trouble with AntibioticsDespite their utility, oral antibiotics at standard doses (at orabove minimal inhibitory concentration) pose challenges in themanagement of acne and rosacea. Most notable is the risk ofantibiotic resistance. Studies document the phenomenon ofantibiotic resistant acne and, increasingly, resultant treatmentfailure.11 Concerns about resistance have prompted new guide-lines in acne management thatemphasize topical antimicrobials andretinoids as well as shortened coursesof systemic antibiotics in efforts todiminish dependence on oral antibi-otics and thus limit risk of developingresistance.12

Importantly, resistance is not lim-ited to P acnes. Researchers have iden-tified resistant strains ofStaphylococcus epidermidis,13,14,15

among acne patients treated with oralerythromycin. Another study of acnepatients showed that systemic antibi-otic therapy was associated withStreptococcus pyogenes colonization andresistance in the oropharynx. Whileonly 20 percent of S pyogenes culturesfrom individuals not treated withantibiotics were resistant to at leastone tetracycline, 85 percent of cul-tures from antibiotic-treated patientsdemonstrated resistance.16

Concern about long-term antibi-otic use and subsequent resistance riskis especially high in light of the increase in reports of commu-nity-acquired MRSA (Methicillin-resistant Staphylococcusaureus) skin and soft-tissue infections.17,18 Price et al report thatMRSA accounted for just 1.5 percent of all strains of S. aureusat select dermatology outpatient clinics in 1988. In 1998 atthese same clinics, 11.9 percent of all S. aureus strains wereMethicillin-resistant.19

A published report linking long-term antibiotic use bywomen with increasing risk of breast cancer20 has engenderedadditional concern among patients regarding systemic antibiot-ic therapy. Of interest, the authors write that, “Among womenwith the highest levels of tetracycline or macrolide use, risk of

breast cancer was not elevated in those using these antibioticsexclusively for acne and rosacea.” Due to flaws in this epidemi-ological study, any conclusions—whether favorable or unfavor-able—are unreliable. Nonetheless, there was wide reporting ofthe study, and dermatologists may field questions from anxiouspatients.

Finally, systemic antibiotics (erythromycin, tetracyclineclass) may be associated with changes in normal flora, GI upset,and Candida vaginitis. In addition to these general concernsrelated to antibiotic use, there are specific drawbacks associatedwith each of the agents administered systemically for acne androsacea.

P acnes resistance, as noted above, has led to general aban-donment of erythromycin in favorof the tetracycline class for themanagement of acne.21 However,tetracycline has also been associat-ed with high rates of resistant acne,increasing dependence on minocy-cline and doxycycline.22 At stan-dard doses, doxycycline is associat-ed with more GI upset than isminocycline. But minocycline isassociated with CNS side-effects,i.e. vertigo, while doxycycline isnot. Minocycline poses the addi-tional risk of bluish hyperpigmen-tation of the sclera, nails, teeth,mucous membranes, and skin.There is a dose-dependent risk ofphototoxicity related to doxycy-cline.

Dermatologists may minimizebut not obviate the risks associatedwith standard antibiotic therapy byvarying: the dose, administrationschedule, and formulation ofantibiotic used. For example, slow-

release minocycline (Dynacin, Medicis) may produce fewerCNS side effects than standard formulations, and administer-ing doxycycline in the morning with food or administeringdoxycycline pellets (Doryx, Warner-Chilcott) can minimize GIside effects without hindering absorption.

Seeking AlternativesIn light of the concerns associated with standard-dose systemicantibiotics and the specific side effects of each of the common-ly-used agents, dermatologists are naturally interested in effec-tive alternatives. Topical antimicrobials—particularly benzoylperoxide and/or benzoyl peroxide combination products for

June 2005 Practical Dermatology 41

Dermatologists mayminimize but not obviate the risks associated with

standard antibiotictherapy by varying:

the dose,administration schedule, and formulation of antibiotic used.

acne—are obvious alternatives to systemic antibiotics in bothacne and rosacea. Furthermore, use of topical retinoids to reg-ulate keratinization and inflammation in acne is standard, andthere is growing use of topical retinoids in rosacea manage-ment. Nonetheless, there remains a subset of patients withmoderate to severe inflammatory acne or rosacea for whomtopical therapy alone is insufficient and who are not candidatesfor or will not pursue oral isotretinoin therapy. For these indi-viduals, subantimicrobial dose doxycycline may be an effectiveand convenient option with fewer side effects than standarddose antibiotics.

The Theoretical Bases. “Sub-antimicrobial” does not mean“sub-therapeutic.”3 In fact, though the tetracyclines are gener-ally classified as antibiotics, they demonstrate a range of thera-peutic effects independent ofantimicrobial mechanisms. As such,these agents have current or theoret-ical applications in periodontitis,arthritis, osteoporosis, and can-cer.25,26,27 Chemically-modified tetra-cycline (CMT) analogs, of whichthere are at least 10, provide noantimicrobial effect. These analogs,modified so that the dimethylaminogroup from carbon-4 position (theside-chain required for antimicrobialactivity) is removed, inhibit synthe-sis of collagenase and matrix metal-loproteinases and down-regulatecytokines in animal models.25,26,27

Based on this evidence thatCMTs confer anti-inflammatorybenefits even when no antibioticeffect is possible, researchers investi-gated subantimicrobial dose doxycy-cline for the management of peri-odontal disease, which was shown toinvolve an increase in matrix metal-loproteinases. A 20mg dose of doxycycline hyclate is below theminimum inhibitory concentration; there is no antibioticaction.28 Yet, twice-daily dosing of doxycycline hyclate(Periostat, CollaGenex) is widely and effectively used for themanagement of periodontitis. Its action is exclusively anti-inflammatory and long-term SDD therapy (up to 18 months)produced no changes in antimicrobial susceptibility in patientsduring the treatment period or up to six months post-treat-ment.28,29 Periostat received FDA approval in 1998 for treat-ment of adult periodontitis and has been widely used for thatindication.

Applications in Dermatology. The efficacy of subantimicro-

bial dose doxycycline in reducing the inflammation of peri-odontitis without simultaneously providing an antibacterialeffect naturally prompted interest in the potential for SDD totreat inflammatory acne and rosacea. The ability to reduceinflammatory lesions and target the inflammatory componentof either disease without the side effects of standard doseantibiotics or risk of resistance would be clearly beneficial.

Data from 40 patients in a six-month double-blind, ran-domized, placebo-controlled study assessing the efficacy oftwice daily dosing of 20mg doxycycline in moderate facial acneshow that SDD (20mg per dose) significantly reduced thenumber of inflammatory and non-inflammatory lesions oversix months.4 Importantly, SDD was well-tolerated, and therewas no detectable antimicrobial effect on skin flora or increase

in the number or severity of resist-ant organisms. There was nodecrease in P acnes counts from thestart of the study through its conclu-sion. Inflammatory lesion countsdecreased by 50.1 percent in thetreatment group, comedonesdecreased by 53.6 percent, and over-all lesion counts were reduced by52.3 percent. Clearance rates in theplacebo group were 10.3 percent,10.6 percent, and 17.5 percent,respectively.

Results of an open-label studyinvolving 50 patients with all stagesof rosacea receiving 20mg doxycy-cline BID show a 80 to 100 percentreduction in inflammatory lesionsand a 50 percent reduction in ery-thema after an average duration offour weeks of therapy.3

One hundred thirty-fourpatients participated in a double-blind, randomized, placebo-con-

trolled study that confirmed the safety and efficacy of SDD inthe treatment of rosacea.2 Treatment consisted of 20mg doxy-cycline twice daily for 16 weeks. At the close of the study, 13.1percent of treated patients demonstrated complete clearing ver-sus 1.5 percent of those in the placebo group.30

There is also evidence to support the combination of SDDwith topical therapy, with the oral agent enhancing the effica-cy of the topical treatment. A double-blind, randomized, place-bo-controlled study involving 40 patients with rosacea witheight to 30 total inflammatory lesions compared SDD plustopical metronidazole lotion 0.75% to placebo plus topicalmetronidazole lotion 0.75%. The results document clinically

Subantimicrobial Dose Doxycycline

42 Practical Dermatology June 2005

A 20mg dose ofdoxycycline hyclate

is below the minimum inhibitory

concentration; there is

no antibioticaction...Its action is

exclusively anti-inflammatory.

significant differences in Clinician’s Global Severity Scoresbetween individuals in the SDD/topical metronidazole groupand those in the placebo/topical metronidazole group by weekfour of treatment.1 These differences became statistically signif-icant by week 12, at which time the topical agent was with-drawn and patients continued either oral therapy or placebotreatment. SDD-treated patients maintained changes throughmonotherapy.

Current ImplicationsWhile the surge in reports of antibiotic resistance and mount-ing concerns about community-acquired MRSA infectionsforce all dermatologists to judiciously consider the use of sys-temic antibiotics, no other FDA approved agent has yet provenable to fully replace these agents in the treatment of certaincases of acne and rosacea. The cumulative evidence fromresearch studies and clinical practice continues to suggest thatthe primary benefit of oral antibiotics in the management ofacne and rosacea comes via their anti-inflammatory rather thantheir antimicrobial actions.3,30,34

Evidence thus far indicates that subantimicrobial dose doxy-cycline safely and effectively treats acne and rosacea but with-out the risks of developing resistance or the side effects associ-ated with standard doses. Currently Periostat is approved onlyfor the treatment of adult periodontitis, but CollaGenex hascompleted enrollment for phase three trials of once-dailyOracea (40mg doxycycline hyclate) and could file a new drugapplication before year’s end.

Whenever possible, treatment for mild to moderate diseaseshould begin with topical agents alone or in combination,based on the particular characteristics of the patient and thepresentation. If topical therapy proves insufficient, considersystemic therapy with SDD. SDD is preferable to standard-dose antibiotics based on the numerous benefits describedabove—documented efficacy, few if any side effects, and obvi-ated risk of resistance. If necessary, SDD may be effectivelycombined with topical therapy. If after three months SDDtherapy does not yield improvement, consider switchingpatients to minocycline or higher doses of doxycycline.

1. Sanchez J, Somolinoas AL, Almodovar PA, et al. Combined effect of doxycyclinehyclate 20mg tablets and metronidazole 0.75 % topical lotion in the treatment ofrosacea. 2004 62nd Annual Meeting of the AAD, Washington, DC.2. Thiboutot D, Skidmore R, Fowler J, et al. Efficacy and Safety of subantimicrobial-dosedoxycycline for the treatment of rosacea. 2005 63rd Annual Meeting of the AAD, NewOrleans, LA.3. Bikowski JB. Subantimicrobial dose doxycycline for acne and rosacea. Skinmed.2003 2(4):234-45.4. Skidmore R, Kovach R, Walker C, et al. Effects of subantimicrobial-dose doxycyclinein the treatment of moderate acne. Arch Dermatol. 2003 139(4):459-64.5. Leyden JJ, McGinley KJ, Cavalieri S, et al. Propionibacterium acnes resistance to antibi-otics in acne patients. J Am Acad Dermatol. 1983 8(1):41-5.6. Eady EA, Gloor M, Leyden JJ. Propionibacterium acnes resistance: a worldwide prob-lem. Dermatology. 2003;206(1):54-6.

7. Diaz C, O’Callaghan CJ, Khan A, et al. Rosacea: a cutaneous marker of Helicobacterpylori infection? Results of a pilot study. Acta Derm Venereol. 2003;83(4):282-6. 8. Utas S, Ozbakir O, Turasan, et al. Helicobacter pylori eradication treatment reducesthe severity of rosacea. J Am Acad Dermatol. 1999 Mar;40(3):433-5. 9. Herr H, You CH. Relationship between Helicobacter pylori and rosacea: it may be amyth.J Korean Med Sci. 2000 Oct;15(5):551-4.10. Bamford JT, Tilden RL, Blankush JL, et al. Effect of treatment of Helicobacter pylori infec-tion on rosacea. Arch Dermatol. 1999 Jun;135(6):659-63.11. Eady AE, Cove JH, Layton AM. Is antibiotic resistance in cutaneous propionibacte-ria clinically relevant? Implications of resistance for acne patients and prescribers. Am JClin Dermatol. 2003;4(12):813-31.12. Gollnick H, Cunliffe W, Berson D, et al. Global Alliance to Improve Outcomes inAcne. Management of acne: a report from a Global Alliance to Improve Outcomes inAcne. J Am Acad Dermatol. 2003 Jul;49(1 Suppl):S1-37.13. Nishijima S, Akamatsu H, Akamatsu M, et al. The antibiotic susceptibility ofPropionibacterium acnes and Staphylococcus epidermidis isolated from acne. J Dermatol.1994 Mar;21(3):166-71.14. Nishijima S, Kurokawa I, Katoh N, Watanabe K. The bacteriology of acne vulgarisand antimicrobial susceptibility of Propionibacterium acnes and Staphylococcus epider-midis isolated from acne lesions. J Dermatol. 2000 May;27(5):318-23.15. Dreno B, Reynaud A, Moyse D, et al. Erythromycin-resistance of cutaneous bacterialflora in acne. Eur J Dermatol. 2001 Nov-Dec;11(6):549-53. 16. Levy RM, Huang EY, Roling D, et al. Effect of antibiotics on the oropharyngeal florain patients with acne. Arch Dermatol. 2003 Apr;139(4):467-71. 17. Frazee BW, Lynn J, Charlebois ED, et al. High prevalence of methicillin-resistantStaphylococcus aureus in emergency department skin and soft tissue infections. AnnEmerg Med. 2005 Mar;45(3):311-20.18. Fridkin SK, Hageman JC, Morrison M, et al. Methicillin-resistant Staphylococcusaureus disease in three communities. N Engl J Med. 2005 Apr 7;352(14):1436-44.19. Price MF, McBride ME, Wolf JE Jr. Prevalence of methicillin-resistant Staphylococcusaureus in a dermatology outpatient population. South Med J. 1998 Apr;91(4):369-71.20. Velicer CM, Heckbert SR, Lampe JW et al. Antibiotic use in relation to the risk ofbreast cancer. JAMA. 2004 18;291(7):827-35.21. Eady EA, et al. Erythromycin resistant in antibiotic treated acne patients: Associationwith therapeutic failure. Br J Dermatol 1989;121:51-7.22. Eady EA, Jones C, Gardner K, et al. Tetracycline-resistant propionibacteria from acnepatients are cross-resistant to doxycycline but sensitivie to minocycline. Br J Dermatol1993;128:556-560.23. Golub LM, Lee HM, Ryan ME, et al. Tetracyclines inhibit connective tissue breakdownby multiple non-antimicrobial mechanisms. Adv Dent Res. 1998 Nov;12(2):12-26.24. Lokeshwar BL. MMP inhibition in prostate cancer. ANN NY Sci 1999;878:271-89.25. Golub LM, Ramamurthy NS, Llavaneras A, et al. A chemically modified nonantimi-crobial tetracycline (CMT-8) inhibits gingival matrix metalloproteinases, periodontal break-down, and extra-oral bone loss in ovariectomized rats. Ann N Y Acad Sci. 199930;878:290-310.26. Ramamurthy NS, Rifkin BR, Greenwald RA, et al. Inhibition of matrix metallopro-teinase-mediated periodontal bone loss in rats: a comparison of 6 chemically modifiedtetracyclines. J Periodontol. 2002 Jul;73(7):726-34. 27. Lokeshwar BL, Selzer MG, Zhu BQ, et al. Inhibition of cell proliferation, invasion,tumor growth and metastasis by an oral non-antimicrobial tetracycline analog (COL-3) ina metastatic prostate cancer model. Int J Cancer. 2002 10;98(2):297-309.28. Walker C, Thomas J, Nango S, Lennon J, Wetzel J, Powala C. Long-term treatmentwith subantimicrobial dose doxycycline exerts no antibacterial effect on the subgingivalmicroflora associated with adult periodontitis. J Periodontol. 2000;71(9):1465-71. 29. Thomas J, Walker C, Bradshaw M. Long-term use of subantimicrobial dose doxycy-cline does not lead to changes in antimicrobial susceptibility. J Periodontol. 2000Sep;71(9):1472-83.30. Del Rosso JQ. A status report on the use of subantimicrobial dose doxycycline: Areview of the biologic and antimicrobial effects of the tetracyclines. Cutis2004;74(8):118-122.31. Powell FC. Rosacea and the pilosebaceous follicle. Cutis. 2004 Sep;74(3 Suppl):9-12, 32-4.32. Breneman D, Savin R, VandePol C, et al. Double-blind, randomized, vehicle-con-trolled clinical trial of once-daily benzoyl peroxide/clindamycin topical gel in the treat-ment of patients with moderate to severe rosacea. Int J Dermatol. 2004 May;43(5):381-7. 33. Leyden JJ, Thiboutot D, Shalita A. Photographic review of results from a clinical studycomparing benzoyl peroxide 5%/clindamycin 1% topical gel with vehicle in the treatmentof rosacea. Cutis. 2004 Jun;73(6 Suppl):11-7.34. Milliken L. The Proposed Inflammatory Pathophysiology of Rosacea: Implications forTreatment. SKINmed 2(1):43-47, 2003.

June 2005 Practical Dermatology 43