Growth hormone deficiency in treated acromegaly

Growth

hormone

deficiency

in

treated

acromegaly

Gherardo

Mazziotti

,

Paolo

Marzullo

,

Mauro

Doga

,

Gianluca

Aimaretti

,

and

Andrea

Giustina

1

Endocrinology,UniversityofBrescia,Brescia,Italy

2

Endocrinology,DepartmentofTranslationalMedicine,Universita`delPiemonteOrientale‘A.Avogadro’,Novara,Italy

3_{DivisionofGeneralMedicine,OspedaleS.Giuseppe,IstitutoAuxologicoItaliano,Verbania,Italy}

Growthhormonedeficiency(GHD)oftheadultis char-acterized byreduced quality of life(QoL) and physical fitness,skeletalfragility,andincreasedweightand car-diovascular risk. Hypopituitarism may develop in patients after definitive treatment of acromegaly, but an exact prevalence of GHD in this population is still uncertainowingtolimitedawarenessandthescarceand conflicting data available on this topic. Because acro-megaly and GHD may yield adverseconsequences on similartargetsystems,thefinaloutcomesofsome com-plicationsofacromegalymaybefurtheraffectedbythe occurrenceofGHD.However,itisstilllargelyunknown whetherpatientswithpost-acromegalyGHDmay bene-fit from GH replacement. We review the diagnostic, clinical,andtherapeuticaspectsofGHDinadultpatients treatedforacromegaly.

Introduction

Acromegalyis a chronic disease characterized by excess secretionofgrowthhormone(GH),generallycausedby a pituitarymacroadenoma(70%ofcases),whichresultsin theelevationof circulatinglevels ofGHandinsulin-like growth factor (IGF)-I [1]. The estimated prevalence of acromegaly isapproximately 40–70 patients per million, withanincidenceof3–4newcasespermillioneveryyear

[2,3]. Although a relatively rare disease, acromegaly is associatedwithreducedlifeexpectancyinstrict relation-ship with GH hypersecretion and comorbidities such as cardiovascular,respiratory,metabolic,andneoplastic com-plications[4,5].

Therapies and treatment modalities for acromegaly aimtoreduceorcontroltumorgrowth,inhibitGH hyper-secretion,andnormalizeIGF-Ivaluestoimprovequality oflife(QoL)andreducemorbidityandmortality associat-edwithGHandIGF-Iexcess[6].Asamatteroffact,the therapyshould ideallybe directed to the restoration of physiologicalGH secretion,which isachieved whenthe tumorisremoved,suchthattheresponseofGHtodynamic stimulianditsintegrateddailysecretionarenormalized.

However, several acromegaly patients receiving treat-mentdonotachievecompletenormalizationofGH secre-tion[7].SomepatientsmaintainhighserumGHandIGF-I values,whereas othersmaydevelop GHDas aresultof overtreatmentofacromegaly.

GHDof theadult is nowrecognized asa well-defined clinicalcondition,characterizedbyreducedQoLand phys-ical fitness, skeletal fragility, adiposity, and increased cardiovascular risk [8]. All these complications of GHD may be clinically relevant in patients with a history of acromegaly who have already developed cardiovascular, metabolic,andskeletalcomplications[9].AlthoughGHD has beenextensively characterized inrelation toits pri-marycauses,theexactprevalenceofthisclinicalcondition in patients treated for acromegaly is still uncertain be-causeoflowawarenessandscarceandconflictingdatain theliteratureonthistopic[10–20].Moreover,itislargely unknown whether patients with post-acromegaly GHD maybenefitfromreplacementwithGH[21–24].

Thisreviewdealswiththeemergingclinicalchallenge ofGHDinadultpatients undergoingtreatmentfor acro-megaly, focusing on diagnostic, clinical, and therapeutic aspectsofthiscondition.

Currentmanagementoptionsinacromegaly

Multimodal treatment isoften necessaryto control acro-megalybysuppressingGHhypersecretion,reducingIGF-I levels,andcontrollingtumorgrowth,leadingtosymptom control and minimizingthe associated clinical signs and comorbidities [6]. The three approaches to therapy are surgery, medical management, and radiotherapy. Each treatmentmodalityhasspecificadvantagesand disadvan-tages,buttheoptimaluseofthesetreatmentsshould theo-retically result in reducing mortality in the acromegaly patientpopulationtothatofthegeneralpopulation[4,5].

Transsphenoidalsurgeryisthetreatmentofchoicefor intrasellar microadenomas, noninvasive macroadenomas (i.e.,thosewithoutcavernoussinusorboneinvasion),and when the tumor is causing compression symptoms

[6,25,26].Inpatientswithintrasellarmicroadenomas, sur-gical removalprovidesbiochemicalcontrol with normali-zation of IGF-I in 75–95% of patients, whereas control rates are much lower in patients with macroadenomas

[6].Optionsforsuchlattertumorsincludeprimarymedical therapyorprimarysurgicaldebulkingfollowedbymedical

1043-2760/

ß2014ElsevierLtd.Allrightsreserved.http://dx.doi.org/10.1016/j.tem.2014.10.005

Correspondingauthor:Giustina,A. (a.giustina@libero.it).

Keywords:growthhormonedeficiency;IGF-1;acromegaly;neurosurgery; radiotherapy;mortality.

therapyforhormonalcontroland/orradiationtherapyfor treatmentofresidualtumor[6].

Three formsofmedical therapyhavebeen usedinthe treatmentofacromegaly[27].Long-actingformulationsof somatostatinanalogscurrentlyapprovedforclinicaluse,in other wordsoctreotidelong-acting releaseand lanreotide autogel,aretheprimarymedicaltreatmentoptionifsurgery is notappropriate and arethe primary first-line therapy aftersurgery[27,28].Thesedrugssignalviasomatostatin receptor subtype 2, and to a lesser extent by targeting receptorsubtype5,leadingtoadecreaseinGHsecretion and to tumor shrinkage [29,30]. About 50% of patients treated with thesedrugsachieve fullbiochemicalcontrol ofacromegaly,althoughthispercentagewasshownto de-clinewhendatafromregistriesofunselectedpatientswere considered[7,31].Thedopamineagonistcabergoline,which targetstype2dopaminereceptorsonpituitaryadenomas, mayretainsomeadvantageintreatingacromegalicpatients withbiochemicallymilddisease[32,33].Patients unrespon-sivetosomatostatinanalogtherapyareswitchedto pegvi-somant, a drug capable of blocking GH receptor and reducingliverIGF-Iproduction[34,35].Ininitial multicen-tertrials,serumIGF-1levelswerenormalizedinmorethan 90%ofpatientstreatedwithpegvisomant,whiledrug effec-tivenesswassomewhatlowerinopen-labelor postmarket-ing studies performed in clinical settings or based on retrospective analysis of disease-specific databases

[34]. In some patients, who do not respond to medical monotherapyand/orrequiretumormasscontrol, combina-tiontherapieswithsomatostatinanalogspluscabergoline, somatostatin analogs plus pegvisomant, or pegvisomant pluscabergolinehavebeenproposed[27,34,35].

Radiation therapy should generally be scheduled as third-line treatment, occasionally as second-line treat-ment,butrarelyasfirst-linetreatment[6].Patientswho do not have tumor growth control or normalization of hormone levels with surgeryand/ormedicaltherapy are possible candidates for radiation therapy. Conventional radiotherapy can lower GH levels and normalize IGF-I inover60%ofpatients,butmaximalresponseisachieved 10–15yearsafterradiotherapyisadministered[6].Before thegenerationofmodernmedicaltherapies,conventional radiotherapywasusedas asecond-lineoptionwhen sur-gery failed to control GH/IGF-I hypersecretion, but this approachwas burdenedbyvariable efficacy,longtimeto reachcompleteeffectiveness,highprevalenceof hypopitu-itarism, increased cerebrovascular mortality, and in-creased risk of secondary brain tumors in recipient patients[6].Morerecently,stereotacticradiosurgery tech-niques, suchas gKnife, have beenused in patients with acromegalywiththeaimofavoidingirradiationofnormal brain andminimizing thelong-term consequences of ra-diotherapy while improvingits effectiveness[6,13,16,19], but very long-term data on safety and efficacy of these newerapproachesarestilllacking.Moreover,stereotactic neurosurgerymaycauseopticneuropathymoreoftenthan conventionalradiotherapyinpatientswithtumorremnant too close tothe optic pathways[6]. In fact,choice ofthe technique is dependent upon the tumor characteristics: conventional radiotherapy is preferred for large tumor remnantsortumors thataretooclose toopticpathways,

whereasgKnifeispreferredwhenthereisasmallertumor sizeorwhenimprovedpatientconvenienceisdesired[6]. Criteriaforcureofacromegaly

Before 2000 there was a widevariability in the criteria arbitrarily used fordefining biochemical control of acro-megaly in different settings. The first historical step to definethebiochemicalcontrolofacromegalywasthe Cor-tinaConsensusConferencewhich,forthefirsttime, estab-lished general criteria for universal use based on the concept that both GH and IGF-I should be ‘normalized’ foracompletecontrolofdisease[36].Thereafter, several subsequent studies and reappraisals did challenge the validityoftheconsensuscriteriaandcalledfortheir revi-sion [37]. Optimal disease control (i.e., post-treatment remission of acromegaly) is now defined as IGF-I level (determined by a reliable standardized assay) in the age-adjusted normal range and a GH level less than 1.0mg/LfromarandomGHmeasurement(usingan ultra-sensitiveassay)[37].Inpatients withacromegaly under-goingsurgicalmanagementofGH-secreting tumors,oral glucose-tolerancetestcanbe usedtoassess theoutcome andanadirGHlevelslessthan0.4mg/L(with ultrasensi-tive assays) may define control in these circumstances

[37,38].NormalizationofIGF-Iistheonlyreliablemarker ofdiseasecontrolunderpegvisomant[34,37].

RiskofGHDinacromegalywithdifferenttreatments In aneffort toachievebiochemicalremission inpatients with acromegaly, it is predictable that a proportion of patientsmayberenderedGHDwhen subjectedto proer-adicativetreatments.Infact,asthecurecriteriafor acro-megalyhavebecomestricter,thespacebetweendefinitive cureononesideandsubnormalGHsecretionontheother has become narrower. GHD is not expected to occur in acromegaly patients undergoing medical therapies be-causedosingcanbefinelyadjustedonthebasisofserum GHand/orIGF-Ivalues[27,34,39].However,itis concep-tuallypossiblethatastateoffunctionalGHDmayoccurin some medically treated patients, mainly when pegviso-mantisused.Nevertheless,theriskofdrug-induced func-tionalGHDinacromegalyisstillunknown,whereasthere hasbeenconvincingevidencethat radiotherapyandtoa lesserextentneurosurgerymaycauseGHDinthisclinical context, as has been demonstrated for other pituitary diseases [8]. Overtreated acromegaly may be considered a distinct category of disease outcome, and attention shouldbepaidtopreventitwhenpossibleandparticularly duringpharmacologicaltreatment.

GHDpost-neurosurgery

Hypopituitarismmaybepresentatdiagnosisof acromeg-alyowingtotheeffectsofcompressionofamacroadenoma on the portal vessels in the pituitary stalk and/or the surrounding pituitary gland [1]. Successful surgery is expected to resolve GH hypersecretion and at the same time to immediately restore normal pituitary function

[40].However,experiencefrom patientswithother histo-typesofpituitaryadenomassuggeststhatGHislesslikely torecover compared togonadotropins, corticotropin, and thyrotropinoncethetumorbulkisremoved[41].Insome

patients, hypopituitarism may persist in the long-term with potential consequences on the QoL and survival of the patient [8,42]. In published surgical series,the inci-dence of hypopituitarism after removal of GH-secreting tumorwasvariable,butonlyfewstudiesinvestigatedthe GHreserveinacromegalypatientsundergoing neurosur-gery (Table 1). In some studies the prevalence of post-surgicalGHDwaslessthan 10%[17,20],whereasothers reported greater frequencies of GHD exceeding 50% of acromegalypatients whoweredefinedascured fromthe diseaseafterneurosurgery[15].Differencesinthereported ratesofsevereGHDmaybeattributabletodifferencesin selectioncriteriafordynamictesting,tumorsize,methods used to evaluate GH secretory function, interval after surgicalprocedure,metabolicbackground,and manipula-tionofthepituitaryglandduringsurgery.It wasargued that the incidence of severe GHD in surgically cured acromegalics was lower than that observed in patients withnonfunctioningpituitaryadenomas,likelyin relation-ship to the growth direction of somatotropinomas that oftengrowinferiorly, withoutdislocation or compression ofthepituitarygland.Infact,whenthetumorwasshown toinvolvebilaterallythepituitarygland,theriskofGHD tendedtobesignificantlyhigher[20].TheriskofGHDwas alsoshowntocorrelatewiththedimensionofthe somato-tropinoma and serum GH levels before and soon after neurosurgery[15,20].

The introduction of the endoscope to transsphenoidal pituitarysurgerycouldofferthesurgeonbetter visualiza-tionaswellasanimprovedrangeofmotioncomparedto the operating microscope, although it is still unclear whether the incidence of post-surgical hypopituitarism maybereducedbyusingthisnewprocedure[25,43]. GHDpost-irradiation

Hypopituitarismisawell-recognizedsequelaof radiother-apy for pituitary tumors, and GH is usually the first hormoneaffected[44].Itwasshownthat30–69%of acro-megaly patients developedGHDafter conventional frac-tionedradiotherapy[10,15].GHstatusevolves withtime aftercranialradiotherapyanddependsondose,andthe

likelihood of GHD is greater than 50% if the biological effectivedoseisgreaterthan40Gy[45].Theyoungerthe patient,thelongertheintervalafterradiotherapy,andthe higher the dose,the greater is the chance of developing GHD after irradiation. The risk of hypopituitarism was shown to be lower in patients treated by stereotactic radiosurgery compared to conventional fractioned radio-therapy [13,46], althoughin some cases the incidenceof hypopituitarisminducedbystereotacticradiosurgerywas notlow,andappearedtobeincloserelationshipwiththe radiationdosetothetumormarginandthewayhormonal deficiencieswereascertained [16,19].

Diagnosisof GHDin(over)treatedacromegaly (acroGHD)

Assuming that no clinical feature can be relied onas a guide,asforauxologicalcriteriainchildren,theapproach todiagnosingGHDinadultsmustincludeahighindexof suspicion. The diagnosis is often not easy, especially in clinicalconditions,suchasacromegaly,inwhichthereare persistentqualitativeabnormalitiesinGHsecretion espe-ciallyfollowingirradiation[47,48](Table1).

As in the general hypopituitary population, having normal age-adjusted levels of IGF-I does not exclude a diagnosisofGHDalsoinpatientswithtreatedacromegaly

[39,49]. However, a low IGF-I level, in the absence of catabolic conditions such as poorly controlled diabetes, liverdisease,andoralestrogentherapy,isgoodevidence for significant GHD, and may be useful in identifying patients who may benefit from treatment andtherefore require GHstimulation testing [39]. Whenexpressed as standarddeviationscore,serumIGF-Ilevelslessthantwo standard deviations below the age-matched mean are highly suggestive of GHD [50]. Analysis of the Pfizer International Metabolic Database (KIMS) suggests that, intheclinicalpractice,serumIGF-Ivalueshavebeenoften used by clinicians to guide the selection of acromegaly patients to be investigated for GHD [24]. In fact, low IGF-I valueswere foundtobemorefrequent inpatients with post-acromegaly GHD compared to those with a history of non-functioning adenoma [24]. Mean IGF-I

Table1.Resultsofstudiesinvestigatingtheriskofgrowthhormonedeficiency(GHD)inpatientswithacromegalyundergoing

surgeryand/orradiotherapy(RT)and/orradiosurgery(RS)a

Ref. Median/mean follow-up (months) Patients (n) Sex (M/F) Median/mean age(years) Treatments Stimulating testsfordiagnosis ofGHD Frequency ofsevere GHD Median/mean serumIGF-Iin GHD

[11] 131 90 44/46 61 Surgery GHdayprofile 22%

[15] 60 56 23/33 54 Surgery/RT GHRH+arginine 61% 1.1SDS [17] 186 81/105 NS Surgery GHRP-2/Arginine 9.1% 1.180SDS [18] NS 72 25/47 51 Surgery ITT 12.5% 0.38SDS [20] 49 123 60/63 44 Surgery ITT 9.8% 176ng/ml [10] 180 36 23/13 44 Surgery+RT ITT 36% NS [12] 108 33 11/22 49 Surgery+RT/RS Ghrelin 27.3% 146.2ng/ml [13] 66 61 32/29 47 RS ITT 5.7% NS [14] 167 57 38/19 38 Sugery+RT NS 5.8% NS [16] 114 35 17/18 45 Surgery+RS GHRH+arginine 13% NS [19] 61 136 62/74 44 Surgery+RS ITT NS NS

a_{Abbreviations:F,female;GHRH,growthhormonereleasinghormone;GHRP-2,growthhormonereleasingpeptide2;IGF-I,insulinlikegrowth-factor-1;ITT,}

insulin-tolerancetest;M,male;n:number;NS,notspecified;SDS,standarddeviationscore.

levelsandz-scoresincuredacromegalicpatientswithGHD are also significantly decreased compared to those mea-suredincuredpatients withGHsufficiency[17,42].Also relevantisthefindingofalongerdurationofGHDinsome patientswithprevioushistoryofacromegaly,whichlikely reflects low clinician awareness of the risk of GHD in patients inremission [21,24].By contrast, some authors based theirdiagnosticassumptionofGHDexclusivelyon IGF-I levels when threeor morepituitary hormone defi-ciencieswerepresent[50–53].

In other cases, stimulation testing is mandatory for diagnosis of GHD in acromegaly patients, as well as in those with other pituitary diseases. GH production is largelyunderhypothalamiccontrolincludingGHreleasing hormone(GHRH)andsomatostatin,whichstimulateand inhibit secretion, respectively [47]. In the early 1980s a classofsyntheticpeptidylcompounds,collectivelytermed GHsecretagogs,wereshowntostimulateGHreleaseboth in vivo and in vitro [54]. Ghrelin, which is the natural ligandofGHsecretagogreceptor,mediatesthereleaseof GH by increasing GHRH and decreasing somatostatin neuronal activity within the hypothalamus [54]. It also directly stimulates GH release from pituitary somato-tropes.Inadditiontoclassicand‘non-classic’hypothalamic peptides,manyotherneuropeptides(galanin), neurotrans-mitters (acetylcholine), metabolic signals (fasting, hypo-glycemia, aminoacids) andperipheral hormones(thyroid andsexhormones)areinvolvedinthemodulationofGH secretion[47,55–60].

Theinsulin-tolerancetest(ITT)isthegoldstandardin establishingthediagnosisofGHDinadultsandtheGHRH in conjunction with arginine (GHRH+arginine) test was showntogivecomparableresults[52,53,61,62]. Notwith-standingpossible variabilityrelatedtoindividualfactors (age,bodycomposition,etc.)anddifferentGHassays,3mg/ L is traditionally considered the cut-off for diagnosis of severeGHDinadultsasassessedbyITT[52,53],whereas differentcut-offshavebeenproposedforthe GHRH+argi-ninetestinrelationshiptobodymassindex(BMI).Corneli etal.[63]showedthattheappropriateGHcut-offpointsfor diagnosingGHDwere 11.5mg/Lforpatients with aBMI lessthan25kg/m2_{,8.0mg/LforaBMIwithin25–30kg/m}2_,

and 4.2mg/L for those with a BMI greater than 30kg/ m2.Both ITTandGHRH+argininetestshave beenused toinvestigateGHsecretioninpatientswithtreated acro-megaly, with positive correlation between the peak GH responsestothetwotests[64].AsanalternativetoITTand GHRH+arginine, some authors proposed to use GHRH alone or ghrelin-mimetic GH secretagogs to stimulate GHsecretion inpatients with acromegaly [12,65].These tests were shown to be reliable in patients previously treatedbysurgery,whereastheresultscouldbe mislead-inginpatientstreatedwithconventionalradiotherapywho couldshowabnormalresponsetoITTandarginine,butcan retain a normal or low-normal response to GHRH and ghrelin-mimeticGHsecretagogs[65].Infact,theresponse toITTandargininerequiresanintactsomatostatinaxis, which indeed is more vulnerable to radiation-induced damage than the GHRH axis and pituitaryparenchyma

[66].Asamatteroffact,GHRH+argininetestwasshownto beunreliablefordiagnosisofGHDwhenperformedinthe

first5yearsafterconventionalradiotherapy[67],whereas data in patients treated with gKnife are still lacking. EvolutionofspontaneousGHsecretionmaybeofhelpin patientswithequivocalGHresponsetostimulatingtests

[68,69].Infact,patientsdevelopingGHDafteracromegaly show abnormalities of spontaneous GH secretion, with specificattenuationofthesizeofGHburstsandahighly irregularpattern,but withretained circadianproperties

[70],whicharecomparabletothoseoccurringinpatients developing GHD after radiotherapy for other pituitary adenomas [64]. By contrast, some cranially irradiated adultpatients withisolatedabnormalresponseto stimu-latingtestsmaytransientlyshownear-normal spontane-ousGHsecretionlikelyreflectinganinitialcompensatory increaseinhypothalamicstimulatoryinputwithina par-tiallydamagedhypothalamic–pituitaryaxis[71].

Clinicalfeaturesof GHDin(over)treatedacromegaly (acroGHD)

TheclinicalsignificanceofGHDinovertreated acromegal-icpatientsisexplainedbythepleiotropiceffectsofGHand IGF-Ionlipolysis, gluconeogenesis,boneformation, mus-cleanabolism,andpsychologicalwell-being.Patientswith acromegaly have a considerable burden of complications andcoexistingillnesses, such ascardiomyopathy, hyper-tension,diabetesmellitus,sleepapnea,osteoarthropathy, andfragility vertebral fractures, which lead todifferent extentstoanimpairmentofQoLandshortenedlife expec-tancy[9,72–74].Ontheotherside,adultGHDisassociated with high cardiometabolic risk due alterations in body composition, lipoprotein profile, fibrinolytic activity, and endothelialfunction[75–77].Moreover,GHDisassociated with feelings ofsocial isolation,fatigue, andpoorer QoL

[78],andhighriskoffragilityfractures[79].IfGHDisleft untreated, mortalityrates doublecompared tothe refer-encepopulation[80].Nonetheless,controversialopinions exist on beneficial effects of long-term GH therapy on mortalityratesinGHD[81].

GHDinpatientstreatedforacromegaly(acroGHD): pathophysiologicalaspects

GH excess and defects share some clinical features (Figure1).Thus,itisconceivablethatthetransitionfrom active acromegaly to GHD may have severe impact on several targets, such as the cardiovascular system and theskeleton,withconsequentprogressionofsomeclinical complicationsalreadycausedby GHexcess.

Hypertensionandinsulinresistancemaybepersistent inseveral patients withacromegaly even after adequate biochemicalcontrolofGHhypersecretion [82,83].Inthis specificclinicalcontext,itisexpectedthattheoccurrenceof GHD, withconsequentnegativeeffectsonbody composi-tionandendothelialfunction,mayfavorhypertensionand insulinresistance,withfurtherincreaseofcardiovascular riskandmortality.ItisworthmentioningthatGHDalso favors accelerated atherosclerosis and coronary artery disease[75].

GHexertsdirecteffectsonmyocardialgrowth[84],and theoccurrenceofGHDisexpectedtodecreaseventricular mass in patients with persistent acromegalic cardiomio-pathy. Although GH may also exert positive effects on

myocardialfunction[84,85],thedevelopmentofGHDmay theoreticallyleadtoheartfailureinpredisposedpatients with pre-existing cardiac abnormalities [69,86], such as thosewithacromegalic cardiomyopathy[87].

GHexcessanddeficiencyhaveoppositeeffectsonbone turnover,butbothleadtoskeletalfragilityandhighriskof vertebralfractures[88].Interestingly,althoughappropriate and effective treatment of acromegaly improves skeletal health[89–91],theriskofvertebralfracturesmaypersist highin somepatients withwell-controlled orcured acro-megaly, in close relationship with pre-existing vertebral fracturesanduntreatedhypogonadism[91,92].Inthis spe-cific clinical context,the development ofhypopituitarism andGHDmayfurtherincreasethefracturerisk[93–95]. GHDinpatientstreatedforacromegaly(acroGHD): clinical/epidemiologicalstudies

Notwithstanding these pathophysiological premises, the clinicalconsequencesofnaı¨veGHDinpreviously acrome-galic subjects have been analyzed only interms of body compositionandQoLoutcomesinasofarlimitednumber ofstudieswith smallpopulationsamples.Race,personal andfamilyhistory,age,durationofacromegaly,degreeof GH/IGF-Iexcessexposure,treatmentstrategies,and du-rationofGHDmay be specificrelevant confounding fac-tors. Current studies suggest that BMI may be slightly increased[12,96,97]orsimilar[17,23,98,99],inacroGHD patients ascompared toreferenceGHDandacromegalic populations.Bodycompositionanalysisbydual-energy X-rayabsorptiometry (DXA)producedconflictingresults.A studyon10acroGHDpatients(nineofwhomwerewomen) found no difference in body fat and lean mass, while quadricepsmuscleendurancewasreducedincomparison to 10 reference GHD patients [22]. It is worth noting, however, that acroGHD patients included in the study weremore hypertensive,dyslipidemic,anddiabetic than their counterparts [22]. In a larger study, analysis by

abdominalcomputedtomographyandDXAof31acroGHD patients found an increase in percent body fat, visceral adiposetissue,andhigh-sensitivityC-reactiveprotein con-centrationscomparedbothtocuredandactiveacromegalic patients [98]. This may explain the increase in leptin levels, a fat-derived hormone, seen in males with GHD compared with non-GHD cured acromegalic subjects

[12].Conversely,the prevalenceofglucose abnormalities and dyslipidemia was found to be comparable between acroGHD and reference GHD groups [12,15], a circum-stance which suggests that these populations may be metabolicallysimilar.

There is, at present, very limited data on cardiac function, withone studydocumentingslightalterations indiastolicfunctioninvestigatedbytissueDoppler imag-inginacroGHDpatientsascomparedtohealthysubjects

[96].

DespitetheadditionalburdenofGHDinthissubsetof patients, it remains to be convincingly demonstrated whether GHD can further aggravate the mortality risk beyondthatoriginatingfrompreviousacromegaly[100].

Results on QoL are more uniform, and support the conceptthatlowGHlevelsmaycontributetoanimpaired QoLinpreviouslyacromegalicpatients.AnalysisofKIMS datadocumentedthat QoLassessedontheQoL-AGHDA (QoLAssessmentofGrowthHormoneDeficiencyinAdults) scalewasdecreasedinwomenwithacroGHDcomparedto thosewithGHDfrom otherpituitarycauses[21].Wexler etal.[42]foundastrongcorrelationbetweenQoLandGH peakafterGHRH+argininetestusingQoL-AGHDA(36 in-depth interviews), the symptom questionnaire (92 item questionnaire), andtheshort-termhealthsurvey (SF-36, 36itemsonphysicalandpsychologicalwell-being).Inthis cohort of 45 acromegaly subjects, 26 GHD and 19 GH sufficient, QoL was significantly poorer in subjects with GHDbyallscalesandsubscales(exceptthe anxietyand anger/hostility subscales of the symptom questionnaire) also after controlling for potentially confounding factors suchasgender,BMI,radiationtherapy,orotherhormone deficiencies[42].

ReplacementtherapyofGHDin(over)treated acromegaly(acroGHD):prosandcons

Innon-acromegalicadultpopulationswithhypopituitarism, replacementtherapywithrecombinanthumanGH(rhGH) has been shown to reverse many systemic abnormalities associated with GHD. Body composition and QoL can improvewithinafewmonthsafterstartingrhGHtherapy

[101],while increasesinbone mineraldensity (BMD)are usually achievedin longertime-periods[102,103]. Reduc-tionsintotalbodyfatandvisceraladiposetissuehavebeen observedinhypopituitarypatientstreatedwithrhGH[104], andareusuallyparalleledbyimprovementsinserum lipo-proteins[105],alsowhenrhGHtherapyisaddedontopof maintenancestatintherapy[106].Conversely,ifGHDisleft untreated,orifrhGHtreatmentissuspended,deterioration ofQoL,cardiovascularriskfactors,andcardiacperformance isobserved[105,107].Importantly,insulin-resistanceand raised bloodglucose levelsmaydevelop atthe beginning of rhGH therapy, particularly in patients exhibiting an adverseriskprofileatbaselinesuchashighBMI[108]. rhGH treatment Untreated GHD Acromegaly Quality of life Fracture risk Cardiovascular risk Mortality Diabetes Hypertension Neoplasc risk Cerebrovascular risk

TRENDS in Endocrinology & Metabolism

Figure1.Clinicaldomainssharedbyacromegaly,untreated,andreplacedgrowth hormonedeficiency(GHD).UntreatedGHDandacromegalyshareimpairedquality oflifeandhighfracture,cardiovascular,andmortalityrisk.Thereplacementof GHDbyhumanrecombinantgrowthhormone(rhGH)maycausecomplications (i.e.,hypertension,diabetes,highneoplastic,andcardiovascularrisk)whichare similartothoseoccurringinacromegaly.

WhenGHDsubjectswithpreviousacromegalyare tak-en into consideration, studies onthe effectsof short- or long-termrhGHtherapyhavegeneratedconflictingresults (Table2).Inasub-analysisoftheKIMSdatabase,6month rhGH therapydeterminednostatisticallysignificant im-provement of QoL, BMI, waist circumference, or blood pressure inacroGHDpatients, althougha positivetrend was registered for these clinical end-points [21]. In a subsequent6monthprospectivestudyon15patients ran-domizedtorhGHcomparedto15patientsrandomizedto placebo,theauthorsreportedabout10%reduction (treat-mentvsplacebo)inbodyfatmass,14%and7.8%decrease in visceral and subcutaneous abdominal adipose tissue, respectively, and an increase in fat-free mass of 1.4kg

[51]. In the same study, QoL measured by the AGHDA improvedwithrhGH,asdidfiveofeightSF-36subscales (physical health,vitality, mental health, social function-ing, andgeneral health)andone offour symptom ques-tionnaire subscales (somatic symptoms). In the rhGH-treated arm, improvements in bodycomposition and ab-dominaladiposetissueweregreaterinpatientswithlower

baseline GH peak after the GHRH+arginine test and higher final rhGH dosing [51]. The same researchers obtained similar favorable effects of rhGH therapy on QoL in a group of GHD women with prior acromegaly

[107]. Conversely, a 2 year rhGH replacement therapy studyfailedtodocumentsignificantchangesinbody com-position,whileisometrickneeflexorstrengthsignificantly improvedandatrendtowardsreductionwasseenfortotal andLDLcholesterollevels[22].In linewith theselatter data, a 1 year open-label prospective study found that replacement with rhGH therapy was able to increase IGF-Iconcentrationstotheage-andgender-adjusted nor-mal range, while neither echocardiographic parameters noranyofthecardiovascularriskfactorsor QoL param-eterschangedsignificantlyduringrhGHtreatmentinthis clinical setting [109]. As expected, longer-term studies found moreremarkable effectsof rhGH therapy [97]. In a 3 year prospective controlled study, Giavoli et al. ob-servedaprogressiveimprovementofQoL,totalandLDL cholesterol levels, andpercent bodyfat in rhGH-treated subjects,while afurtherdeteriorationofthelipidprofile

Table2.Resultsofstudiesinvestigatingtheeffectsofrecombinanthumangrowthhormone(rhGH)treatmentinpatientswith

growthhormonedeficiency(GHD)causedbyovertreatmentofacromegalya,b

Refs n Sex (M/F) Median/mean age(years) Startingdaily doseofrhGH Duration (months)

Results Undesiredeffects

[21] 40 14/26 49.3 0.006mg/kg 6 "(QoL)

#(Waist,diastolicBP)

NS

[22] 10 1/9 57.6 0.36mg 24 ""(BCM,kneeflexion608) #(BF,waist,BMI,HbA1c,LDL-C) "(HDL-C,TBN) $(LBM,ECW,BP,TG) MI(10%)** CI(20%)** [23] 10 4/6 46 0.2-0.3mg 36 ##(BF,LDL-C,Total-C) ""(QoL) #(HDL-C) "(TG,BP,2hglucose) $(BMI,waist,HbA1c)

None [24] 115 44/71 49.8 0.2mg 60 ##(Total-C,LDL-C) ""(QoL,HDL-C,HbA1c) "(Waist,BF) $(BMI,BP,TG) Neoplasms(8.7%)* Cardiovasculardisease(8.7%)* Cerebrovasculardisease(6%)* Diabetes(16.5%) Cardiovascularmortality(6%)** [51] 15 6/9 0.003–0.006mg/kg 6 ##(BF,VAT) ""(FFM,REE,QoL) #(hsCRP) "(TBW)

$(BMI,CIMT,total-C,BP, HDL-C,LDL-C)

Arthralgia(33%)* Edema(33%)*

Carpaltunnelsymptoms(6.7%)*

[97] 65 27/38 53 0.28mg 37 $(BMI,waist,HDL-C,TG) ##(BP,total-C,LDL-C) ""(HbA1c) Cardiovasculardisease(3.8%)* Stroke(1.7%)* [107] 9 0/9 44 0.003–0.006mg/kg 6 ""(QoL) NS [109] 16 8/8 56 0.2mg 12 ""(PINP,CTX) ##(FN-BMD) #(Total-C,LDL-C,BF,LVM) "(TG,glucose,LBM,LVEF) $(HDL-C,BP,Waist,LS-BMD,QoL) NS a_Symbols:

"",statisticallysignificantincrease;##,statisticallysignificantdecrease;",trendtowardsincrease;#,trendtowardsdecrease;$,unchanged;*,non-significant differencewithrespecttothecontrolgroup;**,significantlydifferencecomparedtothecontrolgroup.

b_{Abbreviations:BCM,bodycellmass;BF,bodyfat;BMI,bodymassindex;BP,bloodpressure;CI,cerebralinfarction;CIMT,carotidintimal-mediathickness;CTX,}

b-crosslaps;PECW,extracellularwater;F,female;FFM,fat-freemass;FN-BMD,femoralneck-bonemineraldensity;HbA1c,glycatedhemoglobin;HDL-C,high-density lipoprotein;hsCRP,high-sensitivityC-reactiveprotein;PINP,N-terminalpropeptidesoftypeIcollagen;LBM,leanbodymass;LDL-C,low-densitylipoprotein;LVEF,left ventricularejectionfraction;LVM,leftventricularmass;LS-BMD,lumbarspine-bonemineraldensity;M,male;n,numberofpatientstreatedwithrhGH;NS,notspecified; MI,myocardialinfarction;QoL,qualityoflife;REE,restingenergyexpenditure;TBN,totalbodynitrogen;TBW,totalbodywater;TG,triglyceride;total-C,totalcholesterol; VAT,visceraladiposetissue.

wasnotedintherhGH-untreatedGHDgroup[23].Similar effects of rhGH treatment were observed after 3 years of follow-up in acroGHD compared to patients with GHDpost-(over)-treatment of non-functioning adenomas

[97].rhGHwas alsoshowntostimulatecirculating mar-kers of bone turnover without changes in BMD at the lumbarspineandhipafter1[109]and2yearsoftherapy

[22].InarecentanalysisofdataextractedfromtheKIMS databaseon115acroGHDpatientstreatedforupto5years withrhGH,significantreductionoftotalandLDL choles-terollevels,aswellasasignificantincreaseofHDL cho-lesterollevels,werereportedduringrhGHtherapy[24].In thisanalysisthechangeinlipidswasnotassociatedwith changes in BMI or body adiposity, a circumstance that supportsthebeneficialroleofrhGHtherapyonlipoprotein kinetics. In a subanalysis on fasting blood glucose and glycated hemoglobin levels, rhGH therapy was found to impairglucoselevels inrelationtoattainedBMI values, thus substantiating the BMI-dependent risk of incident diabetes in GHD treated with rhGH [24]. Collectively, current data seem to suggest that long-term (3 years) rhGH treatment might be needed to achieve significant improvements in body composition and lipid profile in acroGHD subjects. QoL status may motivate patients andclinicianstostartrhGH therapyandcanbeauseful indexoftherapyeffectivenessevenintheshortterm.

SideeffectsinacroGHDpatients areinfrequent,mild, and comparable between the rhGH-treated and placebo groups or other reference groups [21,23,51,64,107]

(Table2).Thesemayincludemusculoskeletalpain,carpal tunnel symptoms, or aggravation of hypertension

[110].Nonetheless,safetyissueshavebeenrecentlyraised in prospective and long-term retrospective intervention studiesonrhGHtherapy.In10acroGHDandtheir refer-ence controls, Norrman et al. [22] witnessed a case of myocardial infarction in oneacroGHD woman 5 months afterinitiationofrhGH,andcerebrovascularaccidentsin onewomanandoneman(bothpreviouslyirradiated)after 2and6weeksofstartingrhGHtherapy,respectively.Inan analysisoftheKIMSdatabase,Feldt-Rasmussenetal.[21]

recognized a higher risk of stroke in acroGHD patients before starting rhGH therapy, suggesting the potential impact ofprevious GH excess and/or cranial irradiation onthisevent.Inarecentscrutinyoflong-termKIMSdata, Tritosetal.[24]identifiedhigherstandardmortalityrates (SMR)for all-cause mortality in patients with acroGHD treatedwithrhGHthaninthereferenceGHDpopulation (SMR = 1.88, P <0.05), with women showing a slightly increasedriskovermen(SMR=1.51,P=0.04).Compared tothegeneralpopulation,rhGHtreatmentdidnotincrease theriskofall-causemortality,northeprevalenceof cere-brovascular accidents, malignant and benign brain tumors,ortheincidenceofdiabetesmellitus[24]. Cardio-vascularmortalityappearedtobeincreasedinacroGHDon rhGHcomparedtoboththegeneralpopulation(observed/ expected=2.89) and the reference GHD population (SMR=4.23, P=0.004), with female gender acting as the second-best predictor of cardiovascular mortality. Based on these data, it seems reasonable to suggest a prudent approach in considering rhGH therapy in acroGHDpatientswitholderage,elevatedcardiovascular

risk, resistanthypertension,uncontrolled diabetes melli-tus,andalong-termhistoryofmetabolicdiseasesor pre-vious cranial irradiation. Because the neoplastic risk is intrinsicallyincreasedinacromegaly[110],whilefataland nonfatal malignancies are as prevalent in GH-treated adults as in the generalpopulation [111], aprudent ap-proachwouldalsosuggestthattheriskofneoplasiashould beassessed inacroGHDpatients whoarecandidates for replacementtherapy.Infact,someclinicianshavestarted rhGHtherapyinpatientswithnegativecolonoscopyandin women aged40 years andolder with negative mammo-gramswithin 1yearbefore baselinevisit [51].However, extensiveinterventionalstudiesarerequiredto definitive-lyclarifythebenefit/riskratioofrhGHtreatmentinthis setting.

Recommendationsformanagementof GHDin (over)treatedacromegaly(acroGHD)

There are no specific guidelines for the diagnosis and treatmentofGHDinpatientswithhistoryofacromegaly. However, severalrecommendations providedforthe gen-eral population of adults with pituitary diseases

[52,53,112,113] may be translated also to patients with acroGHD, taking into account some peculiar aspects in-herenttothislatterclinicalcondition.

Basedontheavailabledata,weshouldtestallpatients who meet the biochemicalcriteria ofcure of acromegaly after surgeryand/orradiotherapyand radiosurgery,and whohaveclinicalfeaturessuggestiveofGHD,suchaspoor exercise capacity, impaired QoL, heart failure, adverse lipid profiles, and low bone mass and/or fractures. An IGF-I value below the reference range for age is highly suggestiveforsevereGHD,particularlyifotherpituitary deficiencies are alreadyknown [39,50].However, the ab-sence of other pituitary deficiencies and normal IGF-I valuesdonotexcludethediagnosisofGHD.

ITT is the gold standard for diagnosis especially in irradiatedpatients(Box1).Inpatientstreatedbysurgery, GHRH+arginine may be used as valuablealternative to ITT (Box 1), especially in patients at risk for undesired effects of hypoglycemia (i.e., high cardiovascular risk). Patients should be tested 6 months after neurosurgery. In patients treated with conventional radiotherapy, the

Box1.Protocolsofclinicalinterest

Protocolofinsulin-tolerancetest

Regularinsulinisintravenouslyinjectedinadoseof0.10IU/kg. Thereafter blood samples for GH and glucose are drawn every 15minutesfor1handthenevery30minforthesecondhour.The pharmacological stimulus is considered appropriate if plasma glucosedeclinetovaluesbelow2.2mmol/Land/orpatients show symptomsofhypoglycemia.SevereGHDisdefinedwhenGHpeak isbelow3mg/L.

ProtocolofGHRH+argininetest

Argininehydrochlorideisintravenouslyadministeredatthedose of0.5g/kg(i.v.,uptoamaximumof30g)over30minfrom0to 30min.Thereafter,GHRHisintravenouslyadministeredatdosage of1mg/kg.BloodsamplesforGHevaluationaretakenat0,30,45, 60,90,and120min.SevereGHDisdefinedbyaGHpeakbelow 11.5ng/mlforpatientswithaBMIlessthan25kg/m2_,below8.0ng/

mlforaBMIof25–30kg/m2_{,andbelow4.2ng/mlforthosewitha}

BMIgreaterthan30kg/m2_.

GHRH+argininetestbecomesreliableaftera5year inter-val since treatment [67]. Before testing GH reserve, all other possible pituitary deficiencies should be corrected becausefunctionalGHDmaybecausedbyhypoadrenalism (Giustina effect), hypogonadism, and hypothyroidism

[47,57,58].Irradiatedpatientsshouldbeevaluatedat reg-ular intervalsgiven thathypopituitarism andGHDmay developaslateas10–15yearsaftertreatment.

rhGH hasbeenlicensed foruseinadultpatients with GHD for over15 years. Early weight-and surface area-based dosing regimens, derived from pediatric practice, resulted in supraphysiological levels of IGF-I and in-creased incidenceof side effects, such as arthralgia and peripheral edema [112,114]. These symptoms are also common in activeacromegaly, and subside after cure of thedisease.A‘low-dose’approach(0.7–3.5mg/week)isin linewithcurrentguidelines[52,112,113]andisalso advis-ableinthecontextofacroGHDpatients.Thisapproachis basedonatailoreddose-titrationstrategy,whichaccounts forinter-individualdifferencesinGHsensitivityrelatedto age,gender,andvariousbaselinecharacteristics.Because doserequirementsaregreaterinyoungerthanolder indi-viduals,inwomenthaninmen,andinobesecomparedto lean individuals [52,112], the starting dose is higher in young(0.2and0.3mg/dayinmenandwomen, respective-ly)than oldersubjects (0.1mg/day)[115],andistitrated according toindividual IGF-I levels andclinical respon-siveness.Atpresent,thereisnoconsensusontheoptimal dosetitration,finalreplacementdose,oridealIGF-Ivalues to achieve onrhGH therapy in GHD patientswith prior acromegaly.Interventionalstudieshaveusedweight-based

[21,109] or fixed-dose regimens of rhGH therapy [22– 24,51].TheachievedIGF-I levelswere variableand com-prised between –1.1 and +1.7 IGF-I standard deviation scores.ThelikelihoodofreachingtargetIGF-Ilevels recom-mendedbycurrentguidelineswasgreaterinmid-and long-terminterventionalstudies[24,64]thaninshort-termones

[51,107].AsintreatingGHDfromothercauses,dose titra-tionofrhGHinpatientswithacroGHDshouldbebasedon age,gender,anduseoforalestrogens.AttheMassachusetts GeneralHospital[51,107]thestartingdoseswerereported tobeasfollows:3mg/kgdailyinmen50yearsandwomen 50yearsnotonoralestrogens;5mg/kgdailyand6mg/kg daily,respectively,inwomen<50yearsnotonoral estro-gensandinwomen<50yearsonoralestrogens;and4mg/kg dailyinmen<50years.

TreatmentgoalsofrhGHtherapyinacroGHDare(not differentfromthoseofGHDfromothercauses):(i)correcting theclinicalalterationsassociatedwithGHD,(ii)achieving IGF-Iinthenormalrange,and(iii)avoidingovertreatment, asreflectedinsideeffectsandhighIGF-Ilevels. Replace-menttherapywithrhGHisexpectedtoyieldclinicalbenefits in acroGHD patients, and positive outcomes such as de-creased body fat, increased lean mass, improvement in C-reactive proteinlevelsand QoLhavebeen documented tooccurassoonasafter6monthsoftherapy[51,107].After atleast1yearoftherapy,significantimprovementsintotal as wellas LDLand HDLcholesterol become measurable

[23,24], while after thesame interval the effectof rhGH onBMD,echocardiographicparameters,and carotidIMT is modest [96]. Objective parameters used to monitor

responsivenesstorhGHtherapy,oncemaintenancerhGH doseisachieved,shouldincludebodycompositionand fast-inglipidlevelsafter6monthsandthenyearly,in conjunc-tion with evaluation of arterial blood pressure and electrocardiogram. Particular caution is warranted in patientswithseverecardiometabolicimpairment.Fasting glucose levels should be carefully monitored during the initialtitrationphaseofGHreplacementtherapyandthen yearlyduringthelong-termtoidentifythedevelopmentof hyperglycemia and/or diabetes mellitus. Indeed, GH re-placementwasshowntoaffectinsulinsensitivityinGHD patients,especiallythosewithhighstartingBMIand met-abolicsyndrome[112,116].AspecificquestionnaireforQoL should beused before treatment, at 6 months, and then annuallyafterstartingtreatment.QoL mayimprovetoa lesser extent in patients suffering from musculoskeletal disabilitiescausedbylong-termexposuretoGHexcess.If theinitialboneDXAscanisabnormal,repeatedscansare recommendedevery18–24 monthstoassesstheneedfor additionalbone-treatmentmodalities.Anindependent as-sessmentofvertebralfracturesbyaradiologicaland mor-phometricapproach[117]alsocouldbeusefulbecauseBMD isnotagoodpredictoroffracturesinGHD[91,118].Infact, earlydiagnosisofvertebralfracturesisimportanttoassess thefutureriskofclinicalosteoporoticfractures[92,119]. Be-cause acromegaly is associated with high risk of colon tumors [9] and rhGH could favor theirdevelopment and growth,colonoscopyshouldberegularlyperformedduring long-termtreatmentofacroGHD.

Concludingremarks

CurrentguidelinesrecommendGHtreatmentinGHDadult patientsinanappropriateclinicalcontext,inotherwords, documentedpituitarydisease.Thefirstquestionthatnow needstobeaddressedis:aretheacromegalypatientsmade GHDby (over)treatmentan appropriateclinicalcontext? Obviously,thestoryisslightlydifferentcomparedtoother somewhatsimilarconditions,suchasthyroidhormone sup-plementationinhypothyroidthyroidectomizedpatientsor eveninhypoadrenalex-Cushingpatients,becausenodoubt existsinthescientificcommunityconcerningthenecessity of thyroxine/cortisol replacement, whereas no definitive consensus exists on the risk–cost/benefit ratio of rhGH treatmentinthegeneralGHDpopulation[120].However, thepatientwithacromegalycouldbethoughttobeevena moreappropriatecandidateforrhGHtreatmentthan the other GHD patients because, paradoxically, excess and deficiencyofGHsharemanyclinicalcharacteristicswhich maymaketheacromegalypatientmoreclinically vulnera-blethantheotherpatientstotheonsetofGHD.

Havingsaidthis,twootherquestionsmayarise:isthere enough clinical evidence on the positive effects of GH replacementtorecommenduniversal screening forGHD in‘cured’acromegalypatients?Hastheissueofthedoseof GH,alreadylargelydebatedinthegeneralGHD popula-tion,incombinationwithwhatIGF-1leveltotarget,been adequately addressed in acroGHD? These questions are legitimate considering the risk of double overtreatment bringingthepatientbacktoastatusof‘subtleacromegaly’. OutstandingquestionsaresummarizedinBox2.Weknow fromepidemiologicaldatathatacroGHDisnotmarginal,

if we proactively look for it. Acromegaly patients with features likely to be complicated by GHD (osteoporosis, heartfailure,impairedQoL,andadverselipidprofile)are probablythebestcandidatestobebiochemicallytestedfor GHD. However, literaturedata onthe positive effectsof GHsubstitutioninthisclinicalsettingarepreliminaryat bestandcannotsustaintherecommendationfortestingall theseacroGHDpatients.Decisionsshouldbemadeonan individualbasis,takingintoaccounttheriskprofileofGH substitutioninthesepatients,becauseitisalsotruethat some acromegaly features could be worsened by rhGH (oncologic and diabetogenic risk). In this context, a non evidence-baseddecision–butasuggestionbasedongood clinicalsense–istousestartingdoseslowerthaninother GHDpatients,targetinginitiallythelowIGF-Irangefor age. Finally, biochemical and clinical monitoring of the rhGH-treated acroGHDpatient should bestrict to allow carefulassessmentofshortandlong-termpositiveresults versuspossiblesideeffects.

References

1 Melmed,S.(2006)Medicalprogress:acromegaly.N.Engl.J.Med.355,

2558–2573

2 Holdaway, I.M. and Rajasoorya, C. (1999) Epidemiology of

acromegaly.Pituitary2,29–41

3 Daly,A.F.etal.(2006)Highprevalenceofpituitaryadenomas:a

cross-sectionalstudyintheprovinceofLiege,Belgium.J.Clin.Endocrinol.

Metab.91,4769–4775

4 Rajasoorya,C. etal.(1994) Determinantsof clinicaloutcomeand

survivalinacromegaly.Clin.Endocrinol.41,95–102

5 Holdaway, I.M. et al. (2004) Factors influencing mortality in

acromegaly.J.Clin.Endocrinol.Metab.89,667–674

6 Melmed,S.etal.(2009)AcromegalyConsensusGroup.Guidelinesfor

acromegalymanagement:anupdate.J.Clin.Endocrinol.Metab.94,

1509–1517

7 Giustina,A.etal.(2014)Investigationaldrugstargetingsomatostatin

receptorsfortreatmentofacromegalyandneuroendocrinetumors.

Exp.Opin.Invest.Drugs23,1619–1635

8 Schneider,H.J.etal.(2007)Hypopituitarism.Lancet369,1461–1470

9 Melmed,S.etal.(2013)Aconsensusonthediagnosisandtreatmentof

acromegalycomplications.Pituitary16,294–302

10 Biermasz, N.R. et al. (2000) Long-term follow-up results of

postoperativeradiotherapyin36patientswithacromegaly.J.Clin.

Endocrinol.Metab.85,2476–2482

11 De,P.etal.(2003)TranssphenoidalsurgeryforacromegalyinWales:

resultsbasedonstringentcriteriaofremission.J.Clin.Endocrinol.

Metab.88,3567–3572

12 Pekic,S.etal.(2006)GhrelintestfortheassessmentofGHstatusin

successfullytreated patientswithacromegaly.Eur.J.Endocrinol.

154,659–666

13 Vik-Mo,E.O.etal.(2006)Gammaknifestereotacticradiosurgeryfor

acromegaly.Eur.J.Endocrinol.157,255–263

14 Mullan, K. et al. (2009) Long term effect of external pituitary

irradiation on IGF1 levels in patients with acromegaly free of

adjunctivetreatment.Eur.J.Endocrinol.161,547–551

15 Ronchi, C.L. et al. (2009) Prevalence of GH deficiency in cured

acromegalicpatients:impactofdifferentprevioustreatments.Eur.

J.Endocrinol.161,37–42

16 Ronchi,C.L.etal.(2009)Efficacyand tolerabilityofgammaknife

radiosurgery in acromegaly: a 10-year follow-up study. Clin.

Endocrinol.71,846–852

17 Yamada, S. et al. (2011) GH deficiencyin patients aftercure of

acromegalybysurgeryalone.Eur.J.Endocrinol.165,873–879

18 Fujio,S.etal.(2013)Severegrowthhormonedeficiencyisrarein

surgically-curedacromegalics.Pituitary16,326–332

19 Lee,C.C.etal.(2014)Stereotacticradiosurgeryforacromegaly.J.

Clin.Endocrinol.Metab.99,1273–1281

20 Ku,C.R.etal.(2014)Clinicalpredictorsofgrowthhormonedeficiencyin

surgicallycuredacromegalicpatients.Eur.J.Endocrinol.171,379–387

21 Feldt-Rasmussen,U.etal.(2002)KIMSInternationalStudyBoardof

KIMSStudyGroup.Growthhormonedeficiencyandreplacementin

hypopituitarypatientspreviouslytreatedforacromegalyorCushing’s

disease.Eur.J.Endocrinol.146,67–74

22 Norrman,L.L.etal.(2008)Baselinecharacteristicsandtheeffectsof

twoyearsofgrowthhormone(GH)replacementtherapyinadultswith

GHdeficiencypreviouslytreatedforacromegaly.J.Clin.Endocrinol.

Metab.93,2531–2538

23 Giavoli,C.etal.(2012)GHreplacementimprovesqualityoflifeand

metabolic parameters in cured acromegalic patients with growth

hormonedeficiency.J.Clin.Endocrinol.Metab.97,3983–3988

24 Tritos, N.A. et al. (2014) Effects of long-term growth hormone

replacement in adults with growth hormone deficiency following

cureof acromegaly:a KIMSanalysis.J.Clin. Endocrinol.Metab.

99,2018–2029

25 Mamelak,A.N. (2014) Pro: endoscopic endonasal transsphenoidal

pituitarysurgery issuperior tomicroscope-based transsphenoidal

surgery.Endocrine47,409–414

26 Mortini, P. (2014) Cons: endoscopic endonasal transsphenoidal

pituitary surgery is not superior to microscopic transsphenoidal

surgeryforpituitaryadenomas.Endocrine47,415–420

27 Giustina,A.etal.(2014)Expertconsensusdocument:aconsensuson

themedicaltreatmentofacromegaly.Nat.Rev.Endocrinol.10,243–248

28 Giustina, A. et al. (2011) Current management practices for

acromegaly:aninternationalsurvey.Pituitary14,125–133

29 Giustina,A.etal.(2012)Meta-analysisontheeffectsofoctreotideon

tumormassinacromegaly.PLoSONE7,e36411

30 Giustina, A.et al. (2013) Octreotidefor acromegaly treatment: a

reappraisal.ExpertOpin.Pharmacother.14,2433–2447

31 Carmichael,J.D.etal.(2014)Acromegalyclinicaltrialmethodology

impactonreportedbiochemicalefficacyratesofsomatostatinreceptor

ligandtreatments–ameta-analysis.J.Clin.Endocrinol.Metab.99,

1825–1833

32 Marazuela,M.etal.(2014)Cabergolinetreatment inacromegaly:

pros.Endocrine46,215–219

33 Kasuki,L.etal.(2014)Cabergolinetreatmentinacromegaly:cons.

Endocrine46,220–225

34 Giustina, A.et al. (2014) Useof Pegvisomant in acromegaly. An

Italian Society of Endocrinology guideline. J. Endocrinol. Invest.

37,1017–1030

35 Giustina,A.(2014)Optimaluseofpegvisomantinacromegaly:arewe gettingthere?Endocrine PublishedonlineOctober28,2014, http:// dx.doi.org/10.1007/s12020-014-0462-0

36 Giustina,A.etal.(2000)Criteriaforcureofacromegaly:aconsensus

statement.J.Clin.Endocrinol.Metab.85,526–529

37 Giustina,A.etal.(2010)AcromegalyConsensusGroup.Aconsensus

oncriteriaforcure ofacromegaly.J.Clin.Endocrinol.Metab.95,

3141–3148

38 Giustina,A.etal.(2013)Commentary:post-surgicalmonitoringof

acromegaly.Neurosurgery73,E746–E748

39 Aimaretti, G.etal.(2003) Diagnostic reliabilityofa single IGF-I

measurementin237adultswithtotalanteriorhypopituitarismand

severeGHdeficiency.Clin.Endocrinol.(Oxf.)59,56–61

40 Arafah,B.M.etal.(1994)Immediaterecoveryofpituitaryfunction

aftertranssphenoidalresectionofpituitarymacroadenomas.J.Clin.

Endocrinol.Metab.79,348–354

41 Arafah,B.M.(1986)Reversiblehypopituitarisminpatientswithlarge

nonfunctioningpituitaryadenomas.J.Clin.Endocrinol.Metab.62,

1173–1179

42 Wexler,T.etal.(2009)Growthhormonedeficiencyisassociatedwith

decreasedqualityoflifeinpatientswithprioracromegaly.J.Clin.

Endocrinol.Metab.94,2471–2477

Box2.Outstandingquestions

 WhoaretheacromegalypatientstobetestedforGHD?

 Whoarethepatientswithpost-acromegalyGHDwhoshouldbe treatedwithrecombinantGH?

 WhichisthedoseofrecombinantGHtobeusedinpatientswith post-acromegalyGHD?

 Whichisthebiochemicaltarget ofGHreplacementinpatients withpost-acromegalyGHD?

43 Goudakos,J.K. etal.(2011) Endoscopicversusmicroscopic

trans-sphenoidalpituitarysurgery:asystematicreviewandmeta-analysis.

Clin.Otolaryngol.36,212–220

44 Littley, M.D. et al. (1989) Hypopituitarism following external

radiotherapyforpituitarytumoursinadults.Q.J.Med.70,145–160

45 Gleeson,H.K.etal.(2004)Reassessmentofgrowthhormonestatusis

requiredatfinal heightinchildren treated withgrowth hormone

replacementafterradiationtherapy.J.Clin.Endocrinol.Metab.89,

662–666

46 Losa,M.etal.(2008)Theroleofstereotacticradiotherapyinpatients

with growth hormone-secreting pituitary adenoma. J. Clin.

Endocrinol.Metab.93,2546–2552

47 Giustina, A. and Veldhuis, J.D. (1998) Pathophysiology of the

neuroregulation of growth hormone secretion in experimental

animalsandthehuman.Endocr.Rev.19,717–797

48 Peacey, S.R. et al. (1998) Hypothalamic dysfunction in ‘cured’

acromegalyistreatmentmodality dependent.J.Clin.Endocrinol.

Metab.83,1682–1686

49 Marzullo,P.etal.(2001)Usefulnessofdifferentbiochemicalmarkers

oftheinsulin-likegrowthfactor(IGF)familyindiagnosinggrowth

hormoneexcessanddeficiencyinadults.J.Clin.Endocrinol.Metab.

86,3001–3008

50 Hartman,M.L.etal. (2002)Which patientsdo notrequireaGH

stimulationtestforthediagnosisof adultGHdeficiency?J.Clin.

Endocrinol.Metab.87,477–485

51 Miller,K.K.etal.(2010)Growthhormonedeficiencyaftertreatmentof

acromegaly: a randomized, placebo-controlled study of growth

hormonereplacement.J.Clin.Endocrinol.Metab.95,567–577

52 Molitch,M.E.etal.(2011)EndocrineSociety.Evaluationandtreatment

of adult growthhormone deficiency:an Endocrine Society clinical

practiceguideline.J.Clin.Endocrinol.Metab.96,1587–1609

53 Giustina, A.et al.(2008) Guidelines for thetreatment of growth

hormone excess and growth hormone deficiency in adults. J.

Endocrinol.Invest.31,820–838

54 Ghigo,E.(2001)Biologicactivitiesofgrowthhormonesecretagogues

inhumans.Endocrine14,87–93

55 Giustina, A. and Wehrenberg, W.B. (1995) Influence of thyroid

hormones ontheregulation ofgrowth hormonesecretion. Eur.J.

Endocrinol.133,646–653

56 Giustina,A.etal.(1993)Effectsofsexandageonthegrowthhormone

responsetogalanininhealthyhumansubjects.J.Clin.Endocrinol.

Metab.76,1369–1372

57 Mazziotti, G. and Giustina, A. (2013) Glucocorticoids and the

regulation of growth hormone secretion. Nat. Rev. Endocrinol. 9,

265–276

58 Giustina, A. and Mazziotti, G. (2014) Impaired growth hormone

secretionassociatedwithlowglucocorticoidlevels:anexperimental

modelfortheGiustinaeffect.Endocrine47,354–356

59 Brogan,R.S.etal.(1997)EffectsoffooddeprivationontheGHaxis:

immunocytochemicalandmolecularanalysis.Neuroendocrinology65,

129–135

60 Giustina,A.etal.(1992)Argininenormalizesthegrowthhormone

(GH)responsetoGH-releasinghormoneinadultpatientsreceiving

chronic daily immunosuppressive glucocorticoid therapy. J. Clin.

Endocrinol.Metab.74,1301–1305

61 Aimaretti, G. et al. (1998) Comparison between insulin-induced

hypoglycemia and growth hormone (GH)-releasing hormone +

arginineasprovocativetestsforthediagnosisofGHdeficiencyin

adults.J.Clin.Endocrinol.Metab.83,1615–1618

62 Ghigo,E.etal.(2001)Growthhormone-releasinghormonecombined

witharginineorgrowthhormonesecretagoguesforthediagnosisof

growthhormonedeficiencyinadults.Endocrine15,29–38

63 Corneli,G.etal.(2005)Thecut-offlimitsoftheGHresponseto

GH-releasinghormone-argininetestrelatedtobodymassindex.Eur.J.

Endocrinol.153,257–264

64 vanderKlaauw,A.A.etal.(2006)GHdeficiencyinpatientsirradiated

foracromegaly:significanceofGHstimulatorytestsinrelationtothe

24hGHsecretion.Eur.J.Endocrinol.154,851–858

65 Murray,R.D.etal.(2001)Thediagnosisofgrowthhormonedeficiency

(GHD)insuccessfullytreatedacromegalicpatients.Clin.Endocrinol.

(Oxf.)54,37–44

66 Toogood,A.A. (2014)Endocrineconsequencesofbrainirradiation.

GrowthHorm.IGFRes.14(Suppl.),118–124

67Darzy, K.H. etal. (2003)The usefulness ofthe combinedgrowth

hormone (GH)-releasinghormoneandargininestimulationtestin

thediagnosisofradiation-inducedGHdeficiencyisdependentonthe

post-irradiationtimeinterval.J.Clin.Endocrinol.Metab.88,95–102

68Giustina,A.etal.(1991)Effectsofpyridostigmineonspontaneousand

growth hormone-releasing hormone stimulated growth hormone

secretion in children on daily glucocorticoid therapy after liver

transplantation.Clin.Endocrinol.(Oxf.)35,491–498

69Giustina, A. et al. (1996) Impaired spontaneous growth hormone

secretioninseveredilatedcardiomyopathy.Am.HeartJ.131,620–622

70vanderKlaauw,A.A.etal.(2007)Attenuatedpulsesize,disorderly

growthhormoneandprolactinsecretionwithpreservednyctohemeral

rhythm distinguishirradiated from surgically treated acromegaly

patients.Clin.Endocrinol.(Oxf.)66,489–498

71Darzy,K.H.etal.(2009)Craniallyirradiatedadultcancersurvivors

may have normal spontaneous GH secretion in the presence of

discordant peakGH responses to stimulationtests (compensated

GHdeficiency).Clin.Endocrinol.(Oxf.)70,287–293

72Yedinak, C.G.and Fleseriu, M.(2014)Self-perception ofcognitive

function among patients with active acromegaly, controlled

acromegaly, and non-functional pituitaryadenoma: a pilot study.

Endocrine46,585–593

73Anagnostis,P.etal.(2014)Psychologicalprofileandqualityoflifein

patientswithacromegalyinGreece.Isthereanydifferencewithother

chronicdiseases?Endocrine47,564–571

74Arosio,M.etal.(2012)ItalianStudyG-roupofAcromegaly.Predictors

ofmorbidityandmortalityinacromegaly:anItaliansurvey.Eur.J.

Endocrinol.167,189–198

75Doga, M. et al. (2006) Growth hormone deficiency in the adult.

Pituitary9,305–311

76Gola,M.etal.(2005)Growthhormoneandcardiovascularriskfactors.

J.Clin.Endocrinol.Metab.90,1864–1870

77Gazzaruso,C.etal.(2014)Cardiovascularriskinadultpatientswith

growthhormone(GH)deficiencyandfollowingsubstitutionwithGH–

anupdate.J.Clin.Endocrinol.Metab.99,18–29

78Rose´n, T.etal.(1994)Decreasedpsychologicalwell-beinginadult

patientswithgrowthhormonedeficiency.Clin.Endocrinol.(Oxf.)40,

111–116

79Mazziotti,G.etal.(2006)Increasedprevalenceofradiologicalspinal

deformities inadult patientswith GHdeficiency:influenceof GH

replacementtherapy.J.BoneMiner.Res.21,520–528

80Sherlock,M.etal.(2010)Mortalityinpatientswithpituitarydisease.

Endocr.Rev.31,301–342

81vanBunderen,C.C.etal.(2011)Doesgrowthhormonereplacement

therapyreducemortalityinadultswithgrowthhormonedeficiency?

Data from the Dutch National Registry of Growth Hormone

Treatmentinadults.J.Clin.Endocrinol.Metab.96,3151–3159

82Maison,P.etal.(2007)Impactofsomatostatinanalogsontheheartin

acromegaly:ametaanalysis.J.Clin.Endocrinol.Metab.92,1743–1747

83Mazziotti,G.etal.(2009)Effectsofsomatostatinanalogsonglucose

homeostasis: a metaanalysis of acromegaly studies. J. Clin.

Endocrinol.Metab.94,1500–1508

84Sacca`,L.etal.(1994)Growthhormoneandtheheart.Endocr.Rev.15,

555–573

85Volterrani, M.etal. (1997) Haemodynamic effectsof intravenous

growthhormoneincongestiveheartfailure.Lancet349,1067–1068

86Giustina,A.etal.(1999)Endocrinepredictorsofacutehemodynamic

effectsofgrowthhormoneincongestiveheartfailure.Am.HeartJ.

137,1035–1043

87Bihan, H. et al. (2004) Long-term outcome of patients with

acromegaly and congestive heart failure. J. Clin. Endocrinol.

Metab.89,5308–5313

88Giustina,A.etal.(2008)Growthhormone,insulin-likegrowthfactors,

andtheskeleton.Endocr.Rev.29,535–559

89Bonadonna,S.etal.(2005)Increasedprevalenceofradiologicalspinal

deformities in active acromegaly: a cross-sectional study in

postmenopausalwomen.J.BoneMiner.Res.20,1837–1844

90Mazziotti,G.etal.(2008)Prevalenceofvertebralfracturesinmen

withacromegaly.J.Clin.Endocrinol.Metab.93,4649–4655

91Mazziotti,G.etal.(2014)Boneturnover,bonemineraldensityand fractureriskinacromegaly:ameta-analysis.J.Clin.Endocrinol.Metab. PublishedonlineNovember3,2014, http://dx.doi.org/10.1210/jc.2014-2359

92 Mazziotti, G. et al. (2013) Vertebral fractures in patients with

acromegaly:a3-yearprospectivestudy.J.Clin.Endocrinol.Metab.

98,3402–3410

93 Mazziotti,G.etal.(2008)Effectofgonadalstatusonbonemineral

density andradiological spinaldeformities in adultpatients with

growthhormonedeficiency.Pituitary11,55–61

94 Mazziotti,G.etal.(2010)Glucocorticoidreplacementtherapyand

vertebralfracturesinhypopituitaryadultmaleswithGHdeficiency.

Eur.J.Endocrinol.163,15–20

95 Mazziotti,G.etal.(2014)Associationbetweenl-thyroxinetreatment,

GHdeficiency,andradiologicalvertebralfracturesinpatientswith

adult-onsethypopituitarism.Eur.J.Endocrinol.170,893–899

96 van derKlaauw,A.A. etal. (2008)Cardiac manifestationsofGH

deficiencyaftertreatmentforacromegaly:acomparisontopatients

with biochemicalremissionand controls.Eur.J.Endocrinol.159,

705–712

97 van Bunderen, C.C. (2014) Effect of long-term growth hormone

replacement therapy on cardiovascular outcomes in growth

hormone deficient patients previously treated for acromegaly -a

sub-analysis fromtheDutch nationalregistry of growthhormone

treatmentinadults.Eur.J.Endocrinol.PublishedonlineSeptember

16,2014,http://dx.doi.org/10.1530/EJE-14-0515

98 Lin,E.etal.(2012)Effectsofgrowthhormonedeficiency onbody

composition andbiomarkersofcardiovascular riskafterdefinitive

therapyforacromegaly.Clin.Endocrinol.(Oxf.)77,430–438

99 Wexler, T.L. et al. (2010) Growth hormone status predicts left

ventricular mass in patients after cure of acromegaly. Growth

Horm.IGFRes.20,333–337

100Bex,M.etal.(2007)AcroBel–theBelgianregistryonacromegaly:a

surveyofthe‘real-life’outcomein418acromegalicsubjects.Eur.J.

Endocrinol.157,399–409

101Woodhouse,L.J.etal.(2006)Theinfluenceofgrowthhormonestatus

onphysicalimpairments,functionallimitations,andhealth-related

qualityoflifeinadults.Endocr.Rev.27,287–317

102Go¨therstro¨m,G.etal.(2007) Ten-yearGHreplacement increases

bonemineraldensityinhypopituitarypatientswithadultonsetGH

deficiency.Eur.J.Endocrinol.156,55–64

103Barake, M. et al. (2014) Effects of recombinant human growth

hormone therapy onbone mineral density inadults with growth

hormone deficiency: a meta-analysis. J. Clin. Endocrinol.Metab.

99,852–860

104Beauregard,C.etal.(2008)Growthhormonedecreasesvisceralfat

and improves cardiovascular risk markers in women with

hypopituitarism: a randomized,placebo-controlled study.J. Clin.

Endocrinol.Metab.93,2063–2071

105Colao,A.etal.(2002)ThecardiovascularriskofadultGHdeficiency

(GHD)improvedafterGHreplacementandworsenedinuntreated

GHD:a12-monthprospectivestudy.J.Clin.Endocrinol.Metab.87,

1088–1093

106Monson, J.P. et al. (2007) Growth hormone (GH) replacement

decreases serum total and LDL-cholesterol in hypopituitary

patients on maintenance HMG CoA reductase inhibitor (statin)

therapy.Clin.Endocrinol.(Oxf.)67,623–628

107Valassi,E.etal.(2012)Effectofgrowthhormonereplacementtherapy

onthequalityoflifeinwomenwithgrowthhormonedeficiencywho

haveahistoryofacromegalyversusotherdisorders.Endocr.Pract.

18,209–218

108Luger,A.etal.(2012)Incidenceofdiabetesmellitusandevolutionof

glucose parameters in growth hormone-deficient subjects during

growth hormone replacement therapy: a long-term observational

study.DiabetesCare35,57–62

109vanderKlaauw,A.A.etal.(2009)Limitedeffectsofgrowthhormone

replacementinpatientswithGHdeficiencyduringlong-termcureof

acromegaly.Pituitary12,339–346

110Giustina, A. etal.(2003) Diagnosis and treatmentof acromegaly

complications.J.Endocrinol.Invest.26,1242–1247

111van Bunderen, C.C. et al. (2014) Efficacy and safety of growth

hormone treatment in adults with growth hormone deficiency: a

systematicreviewofstudiesonmorbidity.Clin.Endocrinol.(Oxf.)

81,1–14

112Ho,K.K. etal.(2007)Consensusguidelinesforthediagnosisand

treatmentofadultswith GHdeficiencyII:astatementoftheGH

Research Society in association with the European Society for

Pediatric Endocrinology, Lawson Wilkins Society, European

SocietyofEndocrinology,JapanEndocrineSociety,andEndocrine

SocietyofAustralia.Eur.J.Endocrinol.157,695–700

113Doga,M.etal.(2005)CurrentguidelinesforadultGHreplacement.

Rev.Endocr.Metab.Disord.6,63–70

114Cuneo,R.C.etal.(1998)TheAustralianmulticentertrialofgrowth

hormone(GH)treatmentinGH-deficientadults.J.Clin.Endocrinol.

Metab.83,107–116

115Cook,D.M.etal.(2009)Medicalguidelinesforclinicalpracticefor

growth hormone use in growth hormone-deficient adults and

transitionpatients–2009update.Endocr.Pract.15(Suppl.2),1–29

116Attanasio, A.F.et al. (2010) International Hypopituitary Control

Complications Study Advisory Board. Prevalence of metabolic

syndrome in adult hypopituitary growth hormone (GH)-deficient

patients before and after GH replacement. J. Clin. Endocrinol.

Metab.95,74–81

117Griffith, J.F.and Genant, H.K.(2012) New advances in imaging

osteoporosisanditscomplications.Endocrine42,39–51

118Mazziotti,G.etal.(2012)Newunderstanding andtreatments for

osteoporosis.Endocrine41,58–69

119Cauley, J.A. et al. (2007) Long-term risk of incident vertebral

fractures.JAMA19,2761–2767

120Giustina, A. (2007) Does growth hormone replacement therapy

improve cardiacfunction? Nat. Clin. Pract.Endocrinol.Metab. 3,

614–615