Isotretinoin treatment for acne and risk of depression: A systematic review and meta-analysis Yu-Chen Huang, MD,a,b and Ying-Chih Cheng, MDc,d Taipei and New Taipei City, Taiwan Background: The relationship between isotretinoin treatment for acne and depression is controversial. Quantitative analysis has not yet been conducted. Objective: To conduct a meta-analysis, evidence-based examination of the relationship between isotretinoin and depression. Method: Asystematicreviewandmeta-analysisoftheliteraturepublishedfrominceptiontoSeptember30, 2016,wasconducted.Controlledorprospectivenon-controlledtrialson$15acnepatientsreceivingisotretinoin treatmentwereincluded.Theprevalenceofdepressionandchangeindepressionscoreswerecalculated. Result: Thirty-one studies met the inclusion criteria. In the controlled studies, the change in depression scores from baseline was not significantly different between patients receiving isotretinoin treatment and those receiving an alternative treatment (standardized mean difference [SMD] (cid:1)0.334, 95% confidence interval [CI] (cid:1)0.680 to 0.011). The prevalence of depression after isotretinoin treatment significantly declined(relativerisk[RR]0.588,95%CI0.382e0.904).Themeandepressionscoressignificantlydecreased from baseline (SMD (cid:1)0.335, 95% CI (cid:1)0.498 to (cid:1)0.172). Limitations: No randomized controlled trials were reviewed; a large inter-study variation was observed. Conclusions: Isotretinoin treatment for acne does not appear to be associated with an increased risk for depression. Moreover, the treatment of acne appears to ameliorate depressive symptoms. (J Am Acad Dermatol 2017;76:1068-76.) Keywords:acne;depression;isotretinoin;meta-analysis;psychologicalimpact;systemicreview. A cneisacommon,chronicskinconditionthat Abbreviationsused: affects nearly all adolescents. Isotretinoin is the most effective treatment available for CI: confidenceinterval recalcitrant nodulocystic acne.1 The possible RCT: randomizedcontrolledtrial RR: riskratio induction of depressive symptoms by isotretinoin SD: standarddeviation treatment for acne was first reported in 1983.2 In SMD: standardizedmeandifference 1998,theUSFoodandDrugAdministrationissueda warning regarding the possible associations of isotretinoin with depression, psychosis, suicidal based studies in 20003 and 2003,4 as well as several ideation, and suicide. However, 2 largepopulation- controlled5-8andnoncontrolledstudies,9-14failedto FromtheDepartmentofDermatology,WanFangHospital,Taipei Reprintsnotavailablefromtheauthors. Medical Universitya; Department of Dermatology, School of Correspondence to: Yu-Chen Huang, MD, Department of MedicineandCollegeofMedicine,TaipeiMedicalUniversityb; Dermatology, Wan Fang Hospital, Taipei Medical University, Department of Psychiatry, Cardinal Tien Hospital, New Taipei 111,Hsing-LongRoadSec3,WenshanDistrict,TaipeiCity116, Cityc; and Department of Public Health and Institute of Taiwan.E-mail:[email protected]. Epidemiology and Preventive Medicine, College of Public PublishedonlineMarch10,2017. Health,NationalTaiwanUniversity,Taipei.d 0190-9622/$36.00 Fundingsources:None. (cid:1)2016bytheAmericanAcademyofDermatology,Inc. Conflictsofinterest:Nonedeclared. http://dx.doi.org/10.1016/j.jaad.2016.12.028 AcceptedforpublicationDecember20,2016. 1068 JAMACADDERMATOL Huang and Cheng 1069 VOLUME76,NUMBER6 demonstrate an increased risk for depression or English. The search parameters included the suicide associated with isotretinoin. In 2008, a case terms ‘‘depression’’ combined with ‘‘isotretinoin,’’ cross-overstudybyAzoulayetal15revealedastatis- ‘‘accutane,’’or‘‘13-cis-retinoicacid.’’ tically significant association between isotretinoin anddepression.However,alargepopulation-based Study selection study16 found acne alone to be significantly associ- We primarily focused our literature search on ated with depression and suicidal ideation. Despite randomized controlled trials (RCTs). However, the controversy surrounding in the absence of an RCT, isotretinoin, the potential in- we included large-scale CAPSULE SUMMARY creaseintheriskforpsycho- population-based studies, logical problems associated non-RCT, and prospective withsevereacneshouldalso d The relationship between isotretinoin open-label studies with $15 treatment for acne and depression is be considered. Multiple patients with acne who had controversial. studies on the relationship received isotretinoin therapy. between isotretinoin and d Meta-analysis did not show a positive Onlystudiesthatprovidedthe depression have been con- associationbetweenisotretinoinuseand prevalence of depression or ducted, some of which depression. In fact, incidence of depression scores were demonstrated that treating depression declined after isotretinoin included.Articlesfromadverse acne with isotretinoin use. event reporting systems, re- improved depressive symp- Although individual susceptibility to viewarticles,casereports,cor- d toms.17-23 Considering depression during isotretinoin use respondence, and conference whether sex, isotretinoin cannot be ruled out, the available reportwereexcluded.Quality dose, treatment time, and evidence suggests that patients with assessmentwasperformedus- the patient baseline psycho- nodulocystic acne can safely be treated ing methodological index for logic condition affected the withisotretinoinwithoutincreasingtheir nonrandomizedstudies.27The results of these studies is risk for depression. studieswithqualityscores\12 crucial. However, with only werealsoexcluded. limited data and small sam- ple sizes, assessing these confounding factors is Outcomes difficult. Further, 3 systematic reviews by Strahan Theprimaryoutcomesofthepresentstudywerethe et al in 2006,24 Marqueling et al in 2007,25 and prevalenceofdepressionandchangeinthedepression Bremner et al in 201226 did not achieve consistent scorefollowingisotretinointherapy. results. Strahan et al24 and Marqueling et al25 concluded that the current literature does not sup- Data extraction port a causative association between isotretinoin Data were independently extracted by 2 authors and depression. However, Bremner et al26 (Dr Huang and Dr Cheng). Any disagreement was concluded that isotretinoin has a causal link with resolved by consensus. Data on the following depression. Here we sought to determine the measures were extracted: study design, inclusion relationship between isotretinoin and depression criteria, sample size, treatment regimen, study by providing a thorough, evidence-based meta- results, and quality scores (Supplemental Table I; analysis of this controversy. available at http://www.jaad.org). For population- based studies, we extracted the outcome (relative METHODS risk[RR])(SupplementalTableII;availableathttp:// Thismeta-analysiswascarriedoutinaccordance www.jaad.org). Age, the proportion of male with the Preferred Reporting Items for Systematic patients, follow-up time, cumulative isotretinoin Reviews and Meta-Analyses (http://www.prisma- dose, depression scale, and depression score with statement.org/). standard deviation (SD) before, during, and after treatmentwerealsoextracted(SupplementalTables Data source and search strategy III and IV; available at http://www.jaad.org). We We identified studies indexed in PubMed, extracted the number ofdepression cases (Table I), MEDLINE, EmBase, and the Cochrane Library data- ifprovidedbythestudy.Thenon-RCTsconsistedof bases from inception of isotretinoin treatment to 2 groups of patients with acne who received September 30, 2016, (last literature search day isotretinoin or alternative therapy within the same October4,2016).Allarticlesincludedinthepresent study.Depressionwasdefinedbytheoriginalstudy. study involved human clinical studies written in The cumulative isotretinoin dose was calculated 1070 Huang and Cheng JAMACADDERMATOL JUNE2017 Table I. The prevalenceofdepression before, during,and aftertreatment Casesofdepression Casesofdepressionduring Casesofdepressionatfinal Definitionof Study atbaseline,n(%) treatment,n(%) follow-up,n(%) depression Brunoet al,198428 1(1.1)major depression 10 (11)minordepression N/A Nodefinition Kellettet al,199910 6(18) N/A 0(0) HADS-D[10 Hullet al,200029 1(0.5)had history of 4(4.4) 2(4) Self-reported depression Strauss et al,200114 N/A, 15(5)had history of 10(3.3) N/A BDI[13 psychiatric medication Kellettet al,200511 8(25.81) 4(19.05) 4(18.19) BDI[13 Chiaet al,20055 9(15.3) N/A 5(8.2) CES-D[16 Kaymak etal, 20069 0(0) 1(1) 0(0) HSRD[13 Cohenetal, 20076 1(1) N/A 2(2) CES-D[15 Kaymark etal, 200930 12(33.3) 7(19.4) 4(11.1) BDI[13 Rehnetal, 200922 9(7.1) 7(5.6) 4(3.2) BDI$ 10 Ergun etal, 201117 5(7.9)had depression 5(7.9) 2(40)nofollow-up, Self-reported history, 4 ofthem 3(60)improved taking antidepressants Marron etal, 201320 12(3.5) N/A 6(1.7) HADS-D[10 Webster et al,201431 N/A 4(0.9) 0(0) Self-reported Suarez etal, 201632 1(2.8) 3(8.3) 3(8.3) Zungscale[50 BDI, Beck Depression Inventory; CES-D, Center for Epidemiological Studies Depression; HADS-D, Hospital Anxiety Depression Scale- Depression;HRSD,HamiltonRatingScaleforDepression;N/A,notavailable. accordingtothedoseprovidedintheoriginalstudy. groups.Wecomputedthemeanchangeinthescores If the follow-up time was provided in weeks, we from baseline and SD for both the isotretinoin and calculatedaweektobe7days;ifgiveninmonths,we controlled groups. The data were combined in the calculatedamonthtobe30days.Thedetailsofthe same analysis as the studies that reported only the depression scales are shown in Supplemental final scores of the 2 groups. This computation TableV(availableathttp://www.jaad.org). requires the preepost correlation value to be input. However, no studies reported the preepost Data analysis correlation values necessary to conduct an For the controlled trials, we produced a pooled appropriately paired analysis. We used a preepost estimate of the mean difference in the depression correlation of 0.5.33 The same computations were score change for patients treated with isotretinoin conducted for all meta-analyses in prospective compared with patients who received alternative open-label studies with before and after data. therapy. For all studies that provided depression Sensitivity analyses were conducted for each case scores,weperformedapooledestimateforthechange by using preepost correlations of 0.2 and 0.8. If indepressionscore.Iftherewere$3studiesusingthe studieslackednecessarydata,theywerenotpooled same depression scales, a pooled estimate was also in the meta-analysis. Similarly, none of the performed. Data were then stratified by follow-up uncontrolled studies that measured dichotomous timeinto3groups:1)1-2months;2)3-4months;and outcomes before and after treatment reported 3) past 4 months. Separate meta-analyses were then data in a matched format; therefore, these conducted for each subgroup, as well as a pooled outcomeswereanalyzedasunmatcheddata,shown analysis of the prevalence of depression before and to be similar and easier to interpret than after treatment for studies with a clear definition of matched analyses.34 Homogeneity testing was depression and a depression case number. performed using the I2 test. A fixed-effects Continuous data were analyzed using the weighted modelwasusediftherewasalackinheterogeneity. mean difference when comparing grouped studies WhenI2was[60%,arandom-effectmodelwasused. usingthesamescaleorstandardizedmeandifference A meta-regression was performed to determine (SMD) when comparing variable and non- the effect of the cumulative isotretinoin dose, pro- standardizedoutcomescalesreportedacrossstudies. portion of male patients, high baseline depression DichotomousanalyseswereconductedusingRR. scores (studies with baseline scores exceeding For meta-analyses of controlled trials, some the definition of depression, categorical variable), studies reported the baseline and final scores for 2 and quality scores on the change of depression JAMACADDERMATOL Huang and Cheng 1071 VOLUME76,NUMBER6 Fig1. Selectionofstudiesincludedinthepresentsystematicreviewandmeta-analysis. scores for patients who received isotretinoin. after depression scores. A summary of the trial Publication bias was tested by the Egger test. All of characteristicsispresentedinTableI. the above analyses were performed using the Ofthe31studies,1population-basedstudyfound Comprehensive Meta-Analysis software version 3 that isotretinoin significantly increased the risk of (Biostat,Inc,Englewood,NJ). depression, while 2 open-label studies found increased depression scores following isotretinoin RESULTS therapy (1 significantly and the other nonsignifi- Search results and trial characteristics cantly increased). The other 10 controlled studies Ofthe172studiesscreened,theinclusioncriteria and 15 open-label studies found no association were met in 31 studies (Fig 1): 3 population-based betweenisotretinoinuseandtheriskfordepression. studies, 8 controlled studies, and 20 prospective Eleven out of 25 studies (2 controlled5,30 and 9 open-label studies, of which 4 provided only the open-label studies10,11,17,19-23,35) found a significant prevalence of depression without before and improvement in the depression scores or a 1072 Huang and Cheng JAMACADDERMATOL JUNE2017 4 151111 reduced frequency of depression following 6 030080 P .0 \.0.0.0.0.0.0 isotretinointherapy.Theimproveddepressionscore wassignificantlygreaterthanthecontrolgroupin1 6 971946 2I 83. 93.93.92.93.95.93. ce. controlled study and nonsignificant in the other n controlled study. The remaining 3 studies only pro- 5%CI) 029) 0.170)0.074)1.028)0.167)0.021)0.152) ndiffere vdiedperdestshioenpraefvtearlen1cemoofndtheproesfsiothne:ra1p)y1;1%2)m4il%d Correlation0.8 estimatemean(9 (cid:1)0(0.103to0. (cid:1)(cid:1)6(0.461to(cid:1)(cid:1)8(2.041to(cid:1)(cid:1)6(4.005to(cid:1)(cid:1)5(0.643to(cid:1)(cid:1)4(0.369to(cid:1)(cid:1)0(0.589to D,weightedmea TpcreoarsembovllpeavlleIee;dntiacovwena;iiotlhafabndlcedeopnar3ttei)hnstsut0ipeo.9d:n/%/twfrreowdamuwtmri.tnjehagneatdst.t(hoaSerrugrtap)t.ppoylethmweerhnaiptcahyl Effect (cid:1)0.50 (cid:1)0.31(cid:1)1.05(cid:1)2.51(cid:1)0.40(cid:1)0.17(cid:1)0.37 e;WM Cishoatrreatcitneoriiznatuisoen of patients with acne and c n 5 192315 e Clinical and outcome data for patients with acne 1 040080 er es P .0 \.0.0.0.0.0.0 diff aresummarized(TableI,SupplementalTablesII-IV; r n available at http://www.jaad.org). Excluding the 3 sco 2I 0.9 1.37.58.86.96.47.8 mea population-based studies, 2932 acne patients on-controlledstudiesaboutdepression Correlation0.2 2Effectestimatemean(95%CI)IP (cid:1)(cid:1)(cid:1)62.4.0580.257(0.465to0.049)4olledandopen-labelstudies)\(cid:1)(cid:1)(cid:1)87.2.0010.342(0.512to0.169)8(cid:1)(cid:1)(cid:1)85.6.0461.087(2.169to0.006)7(cid:1)(cid:1)(cid:1)80.4.0012.337(3.809to0.864)6(cid:1)(cid:1)(cid:1)85.4.0020.386(0.639to0.134)7(cid:1)(cid:1)(cid:1)90.6.0780.210(0.446to0.026)8(cid:1)(cid:1)(cid:1)84.2.0020.328(0.555to0.101)7 DepressionScale-Depression;SMD,standardizedw-upaftercompletionofisotretinointreatment. wcwdckeodSTscouxtgoweaeaoammcb.csTsrntreerleIiehpuee0esnsseall.dtcIcae5ositI4eahtrcrt-e.fiee1edlaavdFes0dsaneltmo..utbut1geahirTledgs-tetdneh0thsol/ishoe.eekat2oerdwsflgwe2rf,aye6/sotfitstftidmtttmhhnhiuciehnaseoeoegdryint/ibirdsmniemkiiotacesnesrgdeosnoseo.f/eiatourtalnadrelllTosnlieel-ianettahfdtissyiinesnortendcuascaanowotenrdlnuieyirppgisneddeeasoorer,isisseest2deiernsebsedoasfmnt,xretirfoiooitceifgnnhnoetnm/oewpprtks1hier1uitndghgee5ssf/eitsnotcir-d3peunr1eihraefnrdg5aoiye2gctt0yte.smiah2twmsmdhTna6eieoneihet1rgtnilsinndhnye1/rt. ultsofmeta-analysisofcontrolledstudiesandn Correlation0.5Noofstudy/EffectpatientmeasureEffectestimatemean(95%CI) (cid:1)(cid:1)(cid:1)6/210vs158SMD0.334(0.6800.011)beforeandafterisotretinointreatment(includecontr(cid:1)(cid:1)(cid:1)19/1411SMD0.335(0.498to0.172)(cid:1)(cid:1)(cid:1)8/812WMD1.068(2.120to0.017)(cid:1)(cid:1)(cid:1)4/132WMD2.422(3.897to0.948)(cid:1)(cid:1)(cid:1)8/544SMD0.399(0.647to0.152)(cid:1)(cid:1)15/1002SMD0.199(0.420to0.022)(cid:1)(cid:1)(cid:1)5/645SMD0.347(0.571to0.123) Inventory;CI,confidenceinterval;HADS-D,HospitalAnxiety3618d2studies(SimicetalandGnanaraietal)withoutfollo 0i0pdl9sBDuTodw(cid:1)sic..5nepfioheoe050ifo%tpefcsemTtp18.hrer3tkare18hr-remrea3teteCs),eteeisl4igsesenIansw1s,soDat9ittl-oiaopsr(cid:1)tho25(es9r-doipnabenn%recp05letrmonostes.i%reig4eictesrngSvothir9ensosCmrocnean8ssetoricaIislsimietpanefiooflfshtesot(unicntpnoms0ocehrsatco,fr.oiniei3eenoniDrg(cid:1)desa8ntvrgcnIoeefln0e2natyoefiintp.dav-finvrdr1co0dcreerecsi)70ee.feeetaifan9a2.iagsolcnl5nn0atls)menrltoit.oti4tleendcoiliodro)fiy0ioraotTynnw.etnen.csnrh,dh0preaarg-t1aeveeFvrru(tdnti0eetcaHlaoHvphtarw7saalmllreAteyeusot,ihhmiapeo[eDspo9Citsteapanghd5ehan(SeInnsiSh%nt]eee-dt(tiMDaacFetrdp,f(cid:1)al4slniC)Dripegfrpst,noto0hIoeprsymem2(cid:1)A.leeps0fr6ltd)os-oeron.o08.(me0eblswnx.Sr0ld0as3aotaiMe(tiiit1e3hsRnoownebat5et5DsRddnonngyy---., TheresTableII. ControlledstudyDepressionscoresTotalBDIHADS-D1-2months3-4months[4months* BDI,BeckDepression*Thisgroupcontaine 00ssdpic..eug02opbn14rrlie90iefcs93isc.a;,satPiinoT9o=tn5hnly%e.0br2sieafcC2lusoa)Intr(aeenEn0dseg.d0lgto6aeip3nrhl8letio-ges0ts1sh.t94d,(q1FePu0cisag8=rtle;uit.md35yP)1iese5ncs=8foto)ri.rens.0h(d0coch7eow6pae)enrfefdgiswcesiieeonrninnoet JAMACADDERMATOL Huang and Cheng 1073 VOLUME76,NUMBER6 Study name Statistics for each study Std diff in means and 95% CI Std diff Standard Lower Upper in means error Variance limit limit Z-Value p-Value Chia et al, 2005 -0.080 0.199 0.040 -0.470 0.311 -0.400 0.689 Kaymak et al, 2009 -0.836 0.260 0.068 -1.346 -0.326 -3.215 0.001 Simie et al, 2009 -0.195 0.218 0.047 -0.622 0.231 -0.898 0.369 McGrath et al, 2010 0.110 0.271 0.074 -0.421 0.642 0.407 0.684 Suarez et al, 2016 -0.744 0.272 0.074 -1.277 -0.210 -2.733 0.006 -0.334 0.176 0.031 -0.680 0.011 -1.899 0.058 -1.00 -0.50 0.00 0.50 1.00 Isotretinoin Control Fig 2. Forest plot. Standardized mean difference in depression score change from baseline (post-score minus pre-score) between the isotretinoin group and the control group (isotretinoingroupminuscontrolgroup).Themorenegativeastandardizedmeandifference is the greaterthe improvement in depression scores in theisotretinoin group in comparison withthecontrolgroup. Fig 3. Funnel plot analysis for potential publication bias. The funnel plot for change in depressionscoresin19studiesshowednoasymmetry,whichmeansnopublicationbias.Each dot represents a study; the y-axis represents the standard error of the study and the x-axis shows the study result (standardized mean difference of pre- and post-treatment depression scoresineachstudy). DISCUSSION increase depression-related behavior in mice and The relationship between isotretinoin treatment ratsinsomestudies39,40butnotinothers.41,42 for acne and depression remains controversial both This meta-analysis found no association between clinically and scientifically. Because isotretinoin is isotretinoin and depression. Moreover, symptoms of a fat-soluble compound, it can easily cross the depression improved following isotretinoin treat- bloodebrain barrier and interact with brain tissue ment, but this effect was not significantly different whereverintracellularretinoidreceptorsarepresent. from alternative therapy. Among studies that could It can affect the dopaminergic36 and serotonergic notbeusedinthismeta-analysis,onlythepopulation- systems,37 hippocampal neurogenesis,38 and basedstudybyAzoulayetal15concludedthatisotret- frontal orbital activity.8 The chronic administration inoin significantly increased the risk for depression of 1-mg/kg dose of isotretinoin has been shown to in acne patients. All 3 population-based studies 1074 Huang and Cheng JAMACADDERMATOL JUNE2017 underestimatedtheincidenceofdepressionbecause In our study, similar to that in the study by diagnostic codes and antidepressant prescriptions Halvorsen et al,16 the patients with acne had high were used as inclusion criteria and patients with baseline depression scores, which exceeded the inadequate data were excluded. Moreover, this un- depression cut-off points of 4 studies. However, derestimation might be particularly problematic for meta-regression results revealed that high baseline thestudybyAzoulayetal15becausetheRRestimate scores did not affect depression score changes and wasbasedonasmallsubsetofpatientswhometstrict male patients were as capable of improving depressionanddataavailabilitycriteria;therefore,the depressionscoresasfemalepatients. results might not be applicable to a general popula- From the view of prevalence of depression, the tionofpatientstreatedwithisotretinoin.43 risk was significantly decreased after treatment. The systematic review by Bremner et al26 However, some studies described newly developed concluded that isotretinoin had a causal link with depression during treatment. In the controlled trial depression. Despite their finding that multiple conductedbySuarezetal,32newonsetofdepression studies demonstrated no association between was noted in both the isotretinoin and antibiotic isotretinoin and depression or an improvement in groups, implying that depression is associated with depressive symptoms following isotretinoin acne,independentlyofisotretinoin.Thus,physicians treatment, they concluded that the sample sizes shouldconsiderthepossibilityofdepressionamong were too small to confirm that most patients allacnepatientsregardlessofthetreatmentmethod. receiving isotretinoin did not develop depression. WealsoagreewithBremneretal26thatsomepatients The systematic reviews by Strahan et al,24 are more susceptible to depression. However, if Marqueling et al,25 and Bremner et al26 did not there is a link between isotretinoin and depression, includeapoolinganalysistoderivetheirconclusion. no predictive tests exist for quantifying the level of In our meta-analysis, we pooled the results of 1411 risk to patients; perhaps it is an idiosyncratic patients who received depression evaluations at reaction.45 baseline and after treatment, which revealed a Ourstudyhassomelimitations.Thefirstlimitation significant improvement in the depression scores. was that no RCTs were included; however, we Bremner et al26 used the study by Meysken et al44 included recent studies with the highest level of to support a dose-response relationship of up to evidence, all of which demonstrated a weak 3 mg/kg/day isotretinoin for advanced cancer relationship between isotretinoin and depression. patients, 25% of whom exhibited depression. Thesecondlimitationwashighinter-studyvariability However,highdosesthatmightcausesymptomsof (eg, different evaluation tools). Therefore, we used depression are not usually prescribed for acne. In thestandardizedmeandifferenceoutcomemeasure our study, the typical dose prescribed for acne for the meta-analyses of continuous data and (0.5-1mg/kg)wasnotfoundtohaveadose-related performed meta-regression for sensitivity analysis. riskfordepression,consistentwiththeresultsofthe Thequalityscoreaffectedthechangeindepression studybyAzoulayetal.15 scores: when omitting the studies with the lowest Bremner et al26 stated that depression usually scores, the improvement in depression score develops 1-2 months or sometimes around remainedsignificant. 2-4 months after treatment and further emphasized In conclusion, this meta-analysis demonstrated that challengeerechallenge cases support a causal that isotretinoin treatment for acne at the typical link.Inourstudy,thedepressionscoressignificantly therapeutic dose is not associated with an decreased within the first 1-2 months and after increased risk for depression. Moreover, the 4 months and tended to decrease (but not treatmentofacneimprovedsymptomsofdepression significantly) within 3-4 months. Although 4 studies formostpatients.Somepatientsmightbemoreprone found increased scores at 3-4 months, 2 of them todepressionregardlessofacneorotherconditions. foundthatscoresat6monthshaddecreasedbelow Thus,closelymonitoringacnepatientsfordepression baseline despite continued treatment. The other 2 isessentialtoidentifypatientsatahighrisk. studies involved follow-up for only 3-4 months. Webster et al31 found that in all patients who REFERENCES reporteddepression,theconditionresolvedwithout 1. WebsterGF.Acnevulgaris.BMJ.2002;325:465-469. discontinuing isotretinoin. Increased depression 2. Hazen PG, Carney JF, Walker AE, et al. Depressionea side scores or newly reported depression cases might effect of 13-cis-retinoic acid therapy. J Am Acad Dermatol. 1983;9:278-279. beexplainedbythepersistenceofacneorotherside 3. JickSS,KremersHM,Vasilakis-ScaramozzaC.Isotretinoinuse effectsof isotretinoin, which mightresolveoverthe and risk of depression, psychotic symptoms, suicide, and courseoftreatment. attemptedsuicide.ArchDermatol.2000;136:1231-1236. JAMACADDERMATOL Huang and Cheng 1075 VOLUME76,NUMBER6 4. Hersom K, Neary MP, Levaux HP, et al. Isotretinoin and treatment: a 12-week follow-up study of male Finnish antidepressant pharmacotherapy: a prescription sequence military conscripts. J Eur Acad Dermatol Venereol. 2009;23: symmetryanalysis.JAmAcadDermatol.2003;49:424-432. 1294-1297. 5. Chia CY, Lane W, Chibnall J, et al. Isotretinoin therapy and 23. Yesilova Y, Bez Y, Ari M, et al. Effects of isotretinoin on moodchangesinadolescentswithmoderate-to-severeacne: obsessivecompulsivesymptoms,depression,andanxietyin acohortstudy.ArchDermatol.2005;141:557-560. patients with acne vulgaris. J Dermatolog Treat. 2012;23: 6. CohenJ,AdamsS,PattenS.Noassociationfoundbetween 268-271. patientsreceivingisotretinoinforacneandthedevelopment 24. Strahan JE, Raimer S. Isotretinoin and the controversy of of depression in a Canadian prospective cohort. Can J Clin psychiatricadverseeffects.IntJDermatol.2006;45:789-799. Pharmacol.2007;14:e227-e233. 25. MarquelingAL,ZaneLT.Depressionandsuicidalbehaviorin 7. NgCH,TamMM,CeliE,etal.Prospectivestudyofdepressive acnepatientstreatedwithisotretinoin:asystematicreview. symptomsandqualityoflifeinacnevulgarispatientstreated SeminCutanMedSurg.2007;26:210-220. withisotretinoincomparedtoantibioticandtopicaltherapy. 26. Bremner JD, Shearer KD, McCaffery PJ. Retinoic acid and AustralasJDermatol.2002;43:262-268. affective disorders: the evidence for an association. J Clin 8. BremnerJD,FaniN,AshrafA,etal.Functionalbrainimaging Psychiatry.2012;73:37-50. alterations in acne patients treated with isotretinoin. Am J 27. Slim K, Nini E, Forestier D, et al. Methodological index for Psychiatry.2005;162:983-991. non-randomized studies (minors): development and valida- 9. Kaymak Y, Kalay M, Ilter N, et al. Incidence of depression tionofanewinstrument.ANZJSurg.2003;73:712-716. related to isotretinoin treatment in 100 acne vulgaris 28. Bruno NP, Beacham BE, Burnett JW. Adverse effects of patients.PsycholRep.2006;99:897-906. isotretinointherapy.Cutis.1984;33:484-489. 10. KellettSC,GawkrodgerDJ.Thepsychologicalandemotional 29. HullPR,Demkiw-BartelC.Isotretinoinuseinacne:prospec- impactofacneandtheeffectoftreatmentwithisotretinoin. tiveevaluationofadverseevents.JCutanMedSurg.2000;4: BrJDermatol.1999;140:273-282. 66-70. 11. Kellett SC, Gawkrodger DJ. A prospective study of the 30. Kaymak Y, Taner E, Taner Y. Comparison of depression, responsiveness of depression and suicidal ideation in acne anxiety and life quality in acne vulgaris patients who were patients to different phases of isotretinoin therapy. Eur J treated with either isotretinoin or topical agents. Int J Dermatol.2005;15:484-488. Dermatol.2009;48:41-46. 12. Rubinow DR, Peck GL, SquillaceKM, et al. Reducedanxiety 31. Webster GF, Leyden JJ, Gross JA. Results of a phase III, and depression in cystic acne patients after successful double-blind, randomized, parallel-group, non-inferiority treatmentwithoralisotretinoin.JAmAcadDermatol.1987; studyevaluatingthesafetyandefficacyofisotretinoin-Lidose 17:25-32. in patients with severe recalcitrant nodular acne. J Drugs 13. SchulpisK,GeorgalaS,PapakonstantinouED,etal.Psycho- Dermatol.2014;13:665-670. logical and sympatho-adrenal status in patients with cystic 32. Suarez B, Serrano A, Cova Y, et al. Isotretinoin was not acne.JEurAcadDermatolVenereol.1999;13:24-27. associatedwithdepressionoranxiety:atwelve-weekstudy. 14. Strauss JS, Leyden JJ, Lucky AW, et al. Safety of a new WorldJPsychiatry.2016;6:136-142. micronizedformulationofisotretinoininpatientswithsevere 33. AbramsKR,GilliesCL,LambertPC.Meta-analysisofhetero- recalcitrant nodular acne: a randomized trial comparing geneously reported trials assessing change from baseline. micronized isotretinoin with standard isotretinoin. J Am StatMed.2005;24:3823-3844. AcadDermatol.2001;45:196-207. 34. ZouGY.Onerelativeriskversustwooddsratios:implications 15. AzoulayL,BlaisL,KorenG,etal.Isotretinoinandtheriskof formeta-analysesinvolvingpairedandunpairedbinarydata. depression in patients with acne vulgaris: a case-crossover ClinTrials.2007;4:25-31. study.JClinPsychiatry.2008;69:526-532. 35. Bozdag KE, Gulseren S, Guven F, et al. Evaluation of 16. Halvorsen JA, Stern RS, Dalgard F, et al. Suicidal ideation, depressive symptoms in acne patients treated with isotret- mentalhealthproblems,andsocialimpairmentareincreased inoin.JDermatologTreat.2009;20:293-296. inadolescentswithacne:apopulation-basedstudy.JInvest 36. DiehlDJ,GershonS.Theroleofdopamineinmooddisorders. Dermatol.2011;131:363-370. ComplPsychiatry.1992;33:115-120. 17. Ergun T, Seckin D, Ozaydin N, et al. Isotretinoin has no 37. O’ReillyKC,TrentS,BaileySJ,etal.13-cis-retinoicacidalters negativeeffectonattention,executivefunctionandmood.J intracellularserotonin,increases5-HT1Areceptor,andsero- EurAcadDermatolVenereol.2012;26:431-439. tonin reuptake transporter levels in vitro. Exp Biol Med 18. GnanarajP,KarthikeyanS,NarasimhanM,etal.Decreasein (Maywood).2007;232:1195-1203. ‘‘Hamiltonratingscalefordepression’’followingisotretinoin 38. BremnerJD,McCafferyP.Theneurobiologyofretinoicacidin therapy in acne: an open-label prospective study. Indian J affectivedisorders.ProgNeuropsychopharmacolBiolPsychia- Dermatol.2015;60:461-464. try.2008;32:315-331. 19. Hahm BJ, Min SU, Yoon MY, et al. Changes of psychiatric 39. O’Reilly KC, Shumake J, Gonzalez-Lima F, et al. Chronic parameters and their relationships by oral isotretinoin in administrationof13-cisretinoicacidincreasesdepression-re- acnepatients.JDermatol.2009;36:255-261. lated behavior in mice. Neuropsychopharmacology. 2006;31: 20. MarronSE,Tomas-AragonesL,BoiraS.Anxiety,depression, 1919-1927. quality of life and patient satisfaction in acne patients 40. Trent S, DrewCJ, MitchellPJ, etal. Chronictreatmentwith treatedwithoralisotretinoin.ActaDermVenereol.2013;93: 13-cis-retinoic acid changes aggressive behaviours in the 701-706. resident-intruderparadigminrats.EurNeuropsychopharma- 21. NevoralovaZ,DvorakovaD.Moodchanges,depressionand col.2009;19:876-886. suicide risk during isotretinoin treatment: a prospective 41. Ferguson S, Cisneros FJ, Gough B, et al. Chronic oral study.IntJDermatol.2013;52:163-168. treatmentwith13-cis-retinoicacid(isotretinoin)orall-trans- 22. Rehn LM, Meririnne E, Hook-Nikanne J, et al. Depressive retinoicaciddoesnotalterdepression-likebehaviorsinrats. symptoms and suicidal ideation during isotretinoin ToxicolSci.2005;87:451-459. 1076 Huang and Cheng JAMACADDERMATOL JUNE2017 42. FergusonS,CisnerosFJ,HanigJP,etal.Oraltreatmentwith 47. McGrathEJ,LovellCR,GillisonF,etal.Aprospectivetrialof Accutane does not increase measures of anhedonia or the effects of isotretinoin on quality of life and depressive depressioninrats.NeurotoxicolTeratol.2007;29:642-651. symptoms.BrJDermatol.2010;163:1323-1329. 43. Bigby M. Does isotretinoin increase the risk of depression? 48. FerahbasA,TuranMT,EselE,etal.Apilotstudyevaluating ArchDermatol.2008;144:1197-1199. anxietyanddepressivescoresinacnepatientstreatedwith 44. Meyskens FL. Short clinical reports. J Am Acad Dermatol. isotretinoin.JDermatologTreat.2004;15:153-157. 1982;6:732. 49. OrmerodAD,ThindCK,RiceSA,etal.Influenceofisotretinoin 45. Goodfield MJ, Cox NH, Bowser A, et al. Advice on the safe onhippocampal-basedlearninginhumansubjects.Psycho- introductionandcontinueduseofisotretinoininacneinthe pharmacology(Berl).2012;221:667-674. UK.BrJDermatol.2010;162:1172-1179. 50. FakourY,NoormohammadpourP,AmeriH,etal.Theeffect 46. Simic D, Situm M, Letica E, et al. Psychological impact of of isotretinoin (roaccutane) therapy on depression and isotretinointreatmentinpatientswithmoderateandsevere qualityoflifeofpatientswithsevereacne.IranJPsychiatry. acne.CollAntropol.2009;33(Suppl2):15-19. 2014;9:237-240. SupplementalTable I. Summary ofincluded studies VJ O A LU M Study Studydesign Patientinclusioncriteria Samplesize Isotretinoindose Studyresult Score M A E C Population-based studies 7A 6D Jicketal, 20003 Case controlstudy Between6 monthsand 7195isotretinoin vs N/A Noassociation ,ND 5yearsbefore, and at 13,700antibiotic UM ERM least12months after, 340isotretinoin vs676 BE A R T theirfirstisotretinoin antibiotic 6OL orantibiotic prescription. Hersomet al,20034 Sequence symmetry analysis 12-49years 2821isotretinoinvs7360 N/A Noassociation Isotretinoinprescriptions antibiotic betweenJune 1,1999, andMarch 31,2000 Azoulayetal, 200815 Case crossover study Received$1isotretinoin 30,496(126depression N/A Statistically significant prescription during patients) associationbetween 1984-2003 isotretinoinand depression Nonrandomized controlledtrials Nget al,20027 Prospective controlled Severecystic acne 174(171) vs41 Startedat40mg/day, Noassociation 1417 study Nocurrent diagnosis of antibiotic/topical increasedto a dosage depression, of1.0mg/kg/day over concomitant 1monthand antidepressants, continuedto total corticosteroids, cumulativedose of anabolicsteroids, or 120 mg/kg(over otherdepression- 5-6 months) inducing medications Chiaet al,20055 Prospective controlled 12-19years 59isotretinoinvs 73 1mg/kg/day Bothtreatments 1417 study,intent-to-treat Moderate-to-severe conservative therapy associatedwith a analysis inflammatory and (49vs52) decreased depression cystic acne,no history symptoms butno H oforcurrent DSM-IV significant difference u a AxisI diagnosis n g Continued a n d C h e n g 1 0 7 6 .e 1
Description: