Side Effects of Anabolic Androgenic Steroids: Pathological Findings and Structure–Activity Relationships AndreasBu¨ttnerandDetlefThieme Contents 1 MethodologicalLimitations............................................................... 460 2 MechanismsofActionandSubstanceSpecificity ........................................ 460 3 AnabolicEffects .......................................................................... 463 3.1 CorrelationofSteroidStructureandAnabolicEffects.............................. 463 3.2 PathologicalFindings ............................................................... 464 4 BioconversionofSteroidstoCharacteristicAndrogensandEstrogens ................... 468 4.1 ConversiontoAndrogens ........................................................... 468 4.2 ConversiontoEstrogens ............................................................ 469 4.3 SuppressionofEndogenousSteroidBiosynthesisandPathologicalEffectsonthe EndocrineandReproductiveSystem ............................................... 470 4.4 MetabolicSuppressionandHepatotoxicEffects .................................... 471 5 NeurosteroidsandPsychiatricEffects .................................................... 474 5.1 BiosynthesisofNeurosteroids ...................................................... 474 5.2 Behavioral–PsychiatricSideEffects ................................................ 476 Reference ..................................................................................... 479 Abstract Sideeffectsofanabolicsteroidswithrelevanceinforensicmedicineare mainlyduetolife-threateninghealthriskswithpotentialfataloutcomeandcasesof uncertain limitations of criminal liability after steroid administration. Both pro- blems are typically associated with long-term abuse and excessive overdose of anabolic steroids. Side effects may be due to direct genomic or nongenomic activities (myotrophic, hepatotoxic), can result from down-regulation of endoge- nousbiosynthesis(antiandrogenic)orbeindirectconsequence ofsteroidbiotrans- formation(estrogenic). Logically, there are no systematic clinical studies available and the number of causally determined fatalities is fairly limited. The following compilation reviews typicalabundantobservationsincaseswherenonnaturaldeaths(mostlyliverfailure A.Bu¨ttner(*) InstituteofLegalMedicine,St.-Georg-Str.108,18055,Rostock,Germany D.ThiemeandP.Hemmersbach(eds.),DopinginSports, 459 HandbookofExperimentalPharmacology195, DOI10.1007/978-3-540-79088-4_19,#Springer‐VerlagBerlinHeidelberg2010 460 A.Bu¨ttnerandD.Thieme and sudden cardiac death) were concurrent with steroid abuse. Moreover, frequent associationsbetweenstructuralcharacteristicsandtypicalsideeffectsaresummarized. Keywords Sideeffects • Heart • Liver • Arteriosclerosis • Psychiatry 1 Methodological Limitations Theevaluationofsideeffectsofanabolicandrogenicsteroid(AAS)abusecontains several methodological problems. Firstly, the exorbitant dosages, which are up to 40 times higher than typical medical applications, prohibit any ethically justified clinical studies. Observations of side effects of therapeutic applications are only partially valid due to their comparatively low dosages. Moreover, the therapeutic relevanceofsteroidsisreducedtoafewtestosteroneanalogsfortreatmentof,e.g., malehypogonadism,renalfailureassociatedwithanemia,cancerassociatedprotein wastingdiseases,burns,AIDSandhereditaryangioedema. Reliableempiricaldatafrombodybuildingarevirtuallyunavailable,becauseself- reportsofadministereddosagesandcorrespondingsideeffectsarecertainlybiased. Finally,observationsofpathologicalfindingsinautopsycasesarelessfrequentand thereforenotstatisticallysignificant.Moreover,thecausalitybetweensteroidabuses, the cluster of pathological findings and a presumptive cause of death needs to be challenged,becausesteroideffects(includingsideeffects,healthrisksandhazards) do not coincide with their analytical identification at the time of death. Therefore, these observations do reflect a general steroid-associated elevation of health risks inthebodybuildercohortbuttypicallycannotclarifytheindividualsituation. 2 Mechanisms of Action and Substance Specificity Genomic steroid action consists of direct regulation of gene transcription by ligand–receptor interaction or indirect modulation of coactivators. Biochemical effects (including side effects) of steroids may be categorized according to the various receptortypes, i.e., androgen(hAR), estrogen, glucocorticoid(hGR), pro- gesterone(hPR)andmineralcorticoidreceptors.Inaddition,nongenomiceffectsof steroidsare well described, e.g.the direct activation ofionotropicGABA recep- A torsbyneurosteroids. Both mechanisms are characterized by a high specificity of receptor–ligand binding,requiringstructuralcongruenceofsteroidsandcorrespondingreceptors. Differentsteroidreceptortypesarestructurallyrelated;thesequencehomologyof theligandbindingdomainsofhARisrelativelyhigh(e.g.55%withhPR,51%with hGR; Gao et al. 2005) which suggests the potential of cross-reaction in particular afteradministrationofhighdosesofsteroids.Extratothishighhomologyofrelevant 19 PathologicalFindingsandStructure–ActivityRelationships 461 receptors,thesteroidalligandsarecharacterizedbyaconsiderablestructuralsimilar- ityandmayoftenbebiotransformedintoeachother.Therefore,astrictdifferentiation ofsteroidsintofunctionalsubgroupsisneitherexpectednorobserved. All relevant endogenous steroid hormones are synthesized and biotransformed within the biochemical pathway shown in Fig. 1 (compare Kicman 2009). Con- sequently,biologicalphenomenacanresultfromdirectsubstancesupplementation (e.g. abuse of anabolic steroids) or indirect effects associated with inhibition (suppression of gonadotropic hormones) or induction (formation of estrogens from testosterone) of endogenous biosynthesis. The availability of steroidogenic enzymesintargettissues(e.g.5a-steroid-reductaseinprostateorskin)ishenceof primaryimportancefortheregio-specificityofsteroidaleffects. Regulationofsteroidbiochemistryisacomplexprocessandsubjecttopositive enzymaticamplificationandnegativefeedbackmechanisms.Anyadministrationof asteroid–especiallyinamountsfarbeyondendogenouslevels–willcompromise thebalanceofitsendogenoussynthesis.Theadministrationoftestosteroneleadsto a rapid down-regulation of gonadotropic hormones (e.g. luteinising hormone), followed by termination of endogenous production of steroids and atrophy of steroidogenicorgans(testis).Thismayresultinasuppressionofotherendogenous steroids(e.g.glucocorticoidsorneurosteroids)producedwithinthesamebiochem- icalpathway. Specific correlations between steroid structure and corresponding side effects are based on exemplary animal or clinical studies in therapeutic dose ranges and remainvague. With respect to endogenous steroids, androgenic side effects may clearly beattributedtoreductionofa4-doublebondto5a-dihydrotestosterone (DHT)by 5a-steroid-reductase and estrogenic effects are due to aromatisation of the steroid A-ring by steroid-aromatase (CYP19). However, a simplified extension of these principlestosyntheticsteroidsisnotgenerallyvalid. Steroid side effects may either result fromdirect receptoragonism (whether or not in combination with metabolic activation) or may be due to a suppression of steroidbiosynthesis.Factorsdeterminingtheefficiencyofsteroidactionarethere- forethestructuralspecificityaswellastheavailabilityofactivatingenzymesinthe respectivetargetcells(e.g.skin,brainregion). As a simplified review, main side effects may be associated with compounds frequentlyabusedinsportsandbodybuildingandtypicalstructuralcharacteristics, i.e.: l Anabolicsideeffectsoftestosteroneandanalogs l Enhancedandrogeniceffects,oftenassociatedwith5a-dihydrogenationofster- oids l Estrogenicsideeffects,dependingonsusceptibilitytoA-ringaromatisation l Antiandrogenic effects based on a suppression of the hypothalamus–pituitary– adrenal/gonadal(HPA/HPG)axes,withoutclearstructuraldetermination l Hepatotoxicity,clearlycorrelatedwith17a-alkylationandtherelatedinhibition ofmetabolicdeactivationofsteroidsbyoxidationofthe17b-hydroxygroup. 462 A.Bu¨ttnerandD.Thieme HO Cholesterol Hypothalamus GnRH Pituitary LH P450scc O O O HO Progesterone Pregnenolone Aldosterone, O O Cortisol 3β-HSD HO O Androstene- DHEA dione 17β-HSD 17β-HSD OH OH 3β-HSD HO O Androstenediol Testosterone Estradiol 5α-DHT Fig.1 Biochemicalsynthesisandmetabolismofsteroids l Psychiatriceffectsofneurosteroids(characterizedby3a-hydroxylationand5a- conformation of the A-ring) synthesized locally in the brain. The availability, inductionorblockingofsteroidogenicenzymesinrelevantbrainregionsseems tobemorerelevantthanperipheralsteroidlevels. 19 PathologicalFindingsandStructure–ActivityRelationships 463 3 Anabolic Effects 3.1 Correlation of Steroid Structure and Anabolic Effects Theinactiveintracellularreceptorproteinsarelocatedinthecytosol.Afterforma- tion of a receptor–ligand complex with suitable steroids and dimerisation, the receptor–ligand complex migrates into the cell nucleus and binds to a specific sequence of the DNA, leading to transcriptional activation of protein synthesis. Moreover, anticatabolic effects based on glucocorticoid receptor inhibition may contributetoanaboliceffects. TheprocessiscontrolledbythespecificityofligandbindingandDNAbinding domainofthesteroidreceptors.Owingtothehighsimilarityoftheligandbinding domainsofallreceptors,steroidligandscross-reactwithdifferentreceptors. 17b-Hydroxylation and an A-ring carrying a 3-keto or equivalent substitutions are thought to be the essential structural characteristics of anabolic effects and therefore affected by its biotransformation (Fig. 2). The time-limiting metabolic reactionoftestosteronedegradationisanA-ringreductionby5a/b-hydrogenation. Furthermetabolicdegradationleadingto(partial)deactivationofanabolicsteroids areoxidationof3-and17a/b-hydroxysteroidsbyvarioushydroxysteroiddehydro- genase (HSD) isoforms. Moreover, anabolic steroids are partially deactivated by aromatisationoftheA-ringtoyieldestrogens. Therefore, a protection of the respective groups contributes to the efficacy of anaboliceffectsandmodifiesthespectrumofpotentialsideeffects.Themostpotent measuretoprolongitsactivityconsistsofalkylationinposition17aprotectingthe 17b-hydroxygroupandenhancesbioavailabilityandanaboliceffectsinparticular afteroraladministration. Typical modifications to prevent aromatisation and/or saturation of the A-ring arealkylationinpositions1or2,chlorinationinposition4,insertionofthe2-oxa group (oxandrolone), introduction of a 1–2 double bond, reduction of the 4–5 doublebondtoyield5a-dihydrosteroidsandremovaloftheC19-methylgroup. OH OH 17 CH 19 3 1 2 O 4 Fig.2 Modificationsofbiologicaleffectsofsteroidsandtheirbioavailabilitymaybeachievedby structuralmodifications,particularlyatpositions1,2,4and17orremovaloftheangularC19.The b-conformationofthe17-hydroxy-groupisessentialforanaboliceffectsofsteroidswhile3-keto groupsareconsideredtobesupportivebutnotessentialandmaybereplacedbya2–3doublebond (e.g.,desoxymethyltestosterone) 464 A.Bu¨ttnerandD.Thieme Generally,androgenicanabolicefficacyofsteroidsisgovernedbytheirreceptor binding and susceptibility to biotransformation. DHT, metandienone and oxy- metholone are supposed to be strong anabolic agents. Remarkably, the anabolic effectsofdesoxymethyltestosterone(1.6timessuperiortotestosterone),stanozolol orfurazabolaresignificantinspiteoflacking3-ketogroups. 3.2 Pathological Findings TheuseofAAS inathletesiswidespread,bothforperformanceimprovementand for augmenting muscular development and strength. However, AAS use is not limited to elite athletes and seems to be more extensive among recreational and amateur athletes (Catlin and Hatton 1991; Hartgens and Kuipers 2004). Strength athletes mainly use AAS to increase muscle mass and strength. Weightlifters and power-lifters strive primarily for strength, whereas bodybuilders train for muscle massandbodydimensions(Fig.3). Thelong-termsideeffects arerelatedtostructure ofsteroids, dosage, frequency of use, age at initiation and concurrent illicit drug use and have not been fully elucidated (Yesalis and Bahrke 1995). Nevertheless, there is a broad spectrum of possibleadverseeffectscausedbysupraphysiologicalAASlevelswhicharesummar- izedinTable1(CatlinandHatton1991;Eklo¨fetal.2003;Evans2004;Grahamand Kennedy1990;HartgensandKuipers2004;HauptandRovere1984;Hicksonetal. 1989;Karila2003; LaBree1991; Maraveliasetal.2005; Melniketal.2007; Mod- linskietal.2006;Narduccietal.1990;YesalisandBahrke1995).Ofthese,someare reversibleandoftenconstitutecosmeticproblemsonly.However,othersareirrevers- ibleandmayleadtoseriousharm.Ofthese,thebestdocumentedeffectsarethoseon thecardiovascularsystem,serumlipids,liverandthereproductivesystem. Although life-threatening side effects seem to be rare, nearly 100% of persons usingAASexperiencesubjectivesideeffectsasaresultofAASuse(Evans2004; ParkinsonandEvans2006).However,therearenumerouscasestudiesandreports which reported serious adverse effects and deaths. Obviously, even many well designed studies do not reflect the AAS problem in real life. Many studies lack a control group and do not account for individual, exercise and environmental Fig.3 Externalaspectofaprofessionalbodybuilder(left)andofarecreationalAASuser 19 PathologicalFindingsandStructure–ActivityRelationships 465 Table1 PossibleadverseeffectsofAASabuse Cardiovascularsystem Psychic/behavioraldisturbances Disturbedlipidmetabolism Moodswings Elevatedbloodpressure,hypertension Aggressiveness Cardiacarrhythmias Depressivesymptoms Myocardialhypertrophy Manicsymptoms Cardiomyopathy Sleepdisturbances Thrombosis Withdrawal,dependence Earlymyocardialinfarction Psychosis Suddencardiacdeath Endocrine/reproductivesystem Kidney Decreaseoflibido Pollakisuria Decreaseoffertility Increaseofserumcreatinine DecreasedLHandFSH Kidneystones Alteredglucosemetabolism Male-specificsideeffects Skin Testicularatrophy Acne Erectiledysfunction,impotence Urticaria Subfertility Striae Prostatichypertrophy Alopecia Impairedspermatogenesis Gynecomastia Female-specificsideeffects Gastrointestinaltract Hirsutism,virilization Queasiness Voicedeepening Emesis Menstrualirregularities Diarrhoe Clitoralenlargement Hematemesis Reducedbreastsize Liver Musculoskeletal Cholestasis,jaundice Tendondamage Peliosishepatis Bonepain Neoplasia Prematureepiphysealclosure(adolescents) Gallbladderstones Injection-related Various Hematoma Edema Infection Fever,shivers Fibrosis Anaphylacticshock Neuro-vascularinjury HepatitisBorCandHIVinfection variables. Furthermore, the covert nature of their misuse may explain the limited andconflictingdataconcerningtheirsideeffects.Therefore,adverseeffectsmight bemuchmoreseverethanreportedintheliterature. In a prospective study, a Finnish cohort of former elite power-lifters with presumed,butnotverified,earlieruseofAASexhibiteda4.6-foldhighermortality compared to a control population, with myocardial infarction (Fig. 4) asthe main causeofdeath(Pa¨rssinenetal.2000).ASwedishstudysuggestedthatAASabuseis associated with an increased risk of premature death, especially among persons withadditionalsubstanceabuseand/orpsychiatricdisease(Peterssonetal.2006a). 466 A.Bu¨ttnerandD.Thieme Fig.4 Oldmyocardialinfarctionina30-yearoldAASabuser After long-term AAS abuse (sudden) cardiac death is the most common mani- festation of AAS toxicity (Di Paolo et al. 2007; Dickerman et al. 1995; Fineschi etal.2007;Hausmannetal.1998;Lukeetal.1990;ownobservation). Upon autopsy as well as in living persons, a dose-dependent left-ventricular hypertrophy has been observed (De Piccoli et al. 1991; Dickerman et al. 1997; Karila2003;Karilaetal.2003;Kriegetal.2007;McKillopetal.1986;Nottinetal. 2006;Sachtlebenetal.1993;Saderetal.2001;Thiblinetal.2000;Urhausenetal. 2004; Yeater et al. 1996; own observation). Similar findings have been described forthecardiacventricularseptum(Saderetal.2001).Accordingtoanechocardio- graphic study this concentric left-ventricular myocardial hypertrophy can still be demonstratedalongtimeaftercessationofAASabuse(Urhausenetal.2004). Other cardiac complications of AAS abuse include myocardial fibroses and myocardial necroses (Halvorsen et al. 2004; Luke et al. 1990; Sullivan et al. 1998; own observation), as well as early coronary sclerosis (Santora et al. 2006; ownobservation,Fig.5). As the cause of these alterations, coronary spasm, increased atherogenesis, impairmentofcoagulationandfibrinolyisaswellasdirectcardiotoxic(arrhythmo- genic)effectsofAAShavebeenproposed. The effects of AAS on blood pressure are still discussed controversially (Hart- gens and Kuipers2004;Sullivan etal. 1998).Insomestudiesthe development of hypertensioncouldnotbedemonstratedafterthechronicintakeofAAS(Hartgens et al. 2003; Karila et al. 2003; Kuipers et al. 1991). In contrast, other studies 19 PathologicalFindingsandStructure–ActivityRelationships 467 Fig.5 MicrophotographofmyocardialfibrosisinanAASabuser Fig.6 Markedcoronaryarterysclerosis observed the occurrence of hypertension after long-term AAS abuse (Grace et al. 2003;Rockhold1993). Besides morphological alterations, functional changes of the vascular system have been described after AAS abuse. These consisted of an increased aortic stiffness(Kasikciogluetal.2007)andanimpairedvascularreactivity(D’Ascenzo etal.2007;Ebenbichleretal.2001;Laneetal.2006).Althoughthealterationsof 468 A.Bu¨ttnerandD.Thieme impaired vascular reactivity seem to be partially reversible after abstinence, con- sumersareatriskforvasospasmwithsubsequentmyocardialinfarction(Fig.4).In anownautopsycase(Fig.6),a41-yearoldmandiedduetosuddencardiacdeath. The heart weighted 482 g and showed a severe coronary artery sclerosis. Toxico- logicalanalysesrevealedhighlevelsofmetandienone. 4 Bioconversion of Steroids to Characteristic Androgens and Estrogens 4.1 Conversion to Androgens Althoughandrogenicandanaboliceffectsaremediatedthroughthesamereceptor (hAR),thebalancebetweenbotheffectsofsteroidsvariessignificantlyandseems tobeassociatedwithcertainstructuralcharacteristics.DHTinparticularexhibitsa hARbindingwhichis2–10timessuperiortotestosterone.Therefore,conversionof TtoDHTinandrogen-responsivetissues(prostate)amplifiestheandrogeniceffects considerably. Since5a-DHTisapotentandrogenicsubstance,5a-reductionofsteroidA-rings is supposed to be a characteristic structural feature of androgens. Similarly, syn- thetic steroids with original 5a structure, e.g. mesterolone or methenolone, are potent androgenic compounds. Steroids with A-ring modifications stabilizing the A-ring conformation, e.g. stanozolol, oxandrolone or oxabolone, appear to be comparatively less androgenic. Another promising approach to partial separation ofandrogenicandanaboliceffectsistheremovalof19-methyltoyield19-norster- oids. Nandrolone (19-nortestosterone) exhibits reduced androgenic effects but retainsanabolicactivitycomparabletotestosterone(Gaoetal.2005). The ratio of anabolic (myotrophic) to androgenic effects of nandrolone is significantly higher than the corresponding index of testosterone, although both steroids are biotransformed by analog enzymes and pathways (Fig. 7). This is assumed to be due to the fact that 5a-dehydro-19-nortestosterone (relative to nortestosterone) binds with weaker affinity to the hAR, opposite to DHT which exhibitsahigheraffinitythantestosterone. Consequently,adrenergiceffectsofsteroidsareinconsistentanddependonthe particular chemical structure. The administration of metandienone, nandrolone decanoate or testosterone cypionate stimulated male sexual behavior after gonad- ectomy, while stanozolol, oxymetholone or methyltestosterone showed no effect (ClarkandHarrold1997;Clarketal.1997). InfemalestheadministrationofAASwillinducemasculinization.Acne,reduc- tion of libido and voice deepening was reported in the first weeks of abuse. After long-term AAS abuse menstrual irregularities, enlargement of the clitoris and reduction of breast size usually develops (Evans 2004; Hartgens and Kuipers 2004; Wu 1997). In contrast to males, some changes, e.g. voice deepening, are notfullyreversible(Maraveliasetal.2005).