SEOM clinical guidelines in advanced and recurrent breast cancer (2018)

Vol.:(0123456789) 1 3 Clinical and T ansla ional Oncology (2019) 21:31–45 h ps://doi.o g/10.1007/s12094-018-02010-w CLINICAL GUIDES INONCOLOGY SEOM clinical guidelines inad anced and ecu en b eas cance (2018) J.I.ChacónLópez‑Muñiz1,2 · L.delaC uzMe ino3· J.Ga iláG ego i4· E.Ma ínezDueñas5· M.Oli ei a6· M.A.SeguíPalme 7· I.Ál a ezLópez8· S.An olinNo oa9· M.Belle Ezque a10· S.López‑Ta uellaCobo11 Recei ed: 3 Decembe 2018 / Accep ed: 5 Decembe 2018 / Published online: 8 Janua y 2019 © The Au ho (s) 2019 Abs ac Al hough he me as asic b eas cance is s ill an incu able disease, ecen ad ances ha e inc eased signi ican ly he ime o p og ession and he o e all su i al. Howe e , oo much in o ma ion has been p oduced in he las 2 yea s, so a well-based guideline is a aluable documen in ea men decision making. The SEOM guidelines a e in ended o make e idence-based ecommenda ions on how o manage pa ien s wi h ad anced and ecu en b eas cance o achie e he bes pa ien ou comes based on a a ional use o he cu en ly a ailable he apies. To assign a le el o ce ain y and a g ade o ecommenda ion he Uni ed S a es P e en i e Se ices Task Fo ce guidelines me hodology was selec ed as e e ence. Keywo ds Ad anced· Loco- egional ecu ence· B eas cance · SEOM· Guidelines * J. I. Chacón López-Muñiz [email p o ec ed] L. dela C uz Me ino [email p o ec ed] J. Ga ilá G ego i jga[email p o ec ed]g E. Ma ínez Dueñas edua do.ma inez@hospi alp o incial.es M. Oli ei a moli [email p o ec ed] M. A. Seguí Palme [email p o ec ed] I. Ál a ez López ISABELMANUELA.ALV[email p o ec ed] S. An olin No oa sil ia.an olin.no [email p o ec ed] M. Belle Ezque a [email p o ec ed] S. López-Ta uella Cobo slopez a [email p o ec ed] 1 Se icio de Oncología Médica, Hospi al Vi gen de la Salud, A da. de Ba be , 30, 45004Toledo, Spain 2 GEICAM, Mad id, Spain 3 Se icio de Oncología Médica, Medicine Depa men , Hospi al Vi gen Maca ena, Uni e si y o Se illa, Se ille, Spain 4 Se icio de Oncología Médica, Fundación Ins i u o Valenciano de Oncología, Valencia, Spain 5 Se icio de Oncología Médica, Hospi al P o incial de Cas ellón, Cas ellón, Spain 6 Vall d’Heb on Uni e si y Hospi al andVall d’Heb on Ins i u e o Oncology, Ba celona, Spain 7 Se icio de Oncología Médica, Co po ació Sani a ia Pa c Tauli, Sabadell, Ba celona, Spain 8 Se icio de Oncología Médica, Hospi al Uni e si a io Donos ia/Donos iako Unibe si a e Ospi alea, Donos ia, Spain 9 Se icio de Oncología Médica, Hospi al Uni e si a io ACo uña, ACo uña, Spain 10 Vall d’Heb on Uni e si y Hospi al andVall d’Heb on Ins i u e o Oncology (VHIO), Ba celona, Spain 11 Se icio de Oncología Médica, Hospi al Gene al Uni e si a io G ego io Ma añón,Ins i u o de In es igación Sani a ia G ego io Ma añón (IiSGM), Uni e sidad Complu ense,Cibe Onc, GEICAM, Mad id, Spain 32 Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 In oduc ion B eas cance (BC) ep esen s he i s cause o in asi e can- ce in he Spanish women, accoun ing o 29% o all emale cance s. Acco ding o ecen da a, 26,730 new cases and 6477 dea hs a e es ima ed annually in Spain [1]. Me as a ic BC (MBC) emains i ually an incu able dis- ease, wi h epo ed median o e all su i al (OS) o app oxi- ma ely 2yea s. Howe e , imp o ed OS (up o 5yea s) has been obse ed ecen ly o ce ain sub ypes, pa icula ly in HER2-posi i e disease. De no o me as a ic disease has be - e 5-yea OS ha ecu en MBC and p ognosis appea o imp o e o e ime [2]. Pa ien s wi h loco- egional ecu ence (LRBC) may o may no be amenable o adical local ea men . O e all, acco ding o a Spanish s udy, an inc ease in OS has been obse ed in he ecen yea s in LRBC pa ien s [3]. Me hodology The SEOM guidelines ha e been de eloped wi h he con- sensus o en b eas cance oncologis s om he coope a i e g oups GEICAM (Spanish B eas Cance Resea ch G oup) and SOLTI (Spanish Collabo a i e G oup o he S udy, ea men and o he expe imen al s a egies in solid umo s). To assign a le el o ce ain y (LC) and a g ade o ecom- menda ion (GR) o he di e en s a emen s desc ibed in he clinical guidelines, he Uni ed S a es P e en i e Se ices Task Fo ce (USPSTF) guidelines me hodology was selec ed as e e ence, as he p e iously adop ed o he o me e sion o SEOM ecommenda ions [4] (Table1). Gene al o e iew o ad anced b eas cance Goals o  ea men The wo main goals o he ea men o MBC pa ien s a e o imp o e su i al and o op imize he quali y o li e [5]. Fo LRBC pa ien s o whom a adical app oach is no ea- sible, he aims o he ea men a e simila o ha in MBC pa ien s. Con e sely, o hose LRBC amenable o a local ea men he objec i es will be o e adica e all mac oscopic diseases and o imp o e bo h disease- ee su i al and o e - all su i al. • Since he diagnosis o ad anced o ecu en BC is made, pa ien s should also be o e ed app op ia e mul idisci- Table 1 S eng h o ecommenda ion and le el o ce ain y acco ding o he Uni ed S a es P e en i e Se ices Task Fo ce (USPSTF) a e July 2012 [4] Ca ego y De ini ion S eng h o ecommenda ions (g ade) A The USPSTF ecommends he se ice. The e is high ce ain y ha he ne bene i is subs an ial B The USPSTF ecommends he se ice. The e is high ce ain y ha he ne bene i is mode a e o he e is mode a e ce ain y ha he ne bene i is mode a e o subs an ial C The USPSTF ecommends selec i ely o e ing o p o iding his se ice o indi idual pa ien s based on p o essional judgmen and pa ien p e e ences. The e is a leas mode a e ce ain y ha he ne bene i is small D The USPSTF ecommends agains he se ice. The e is mode a e o high ce ain y ha he se ice has no ne bene i o ha he ha ms ou weigh he bene i s I The USPSTF concludes ha he cu en e idence is insu icien o assess he balance o bene i s and ha ms o he se ice. E i- dence is lacking, o poo quali y, o con lic ing, and he balance o bene i s and ha ms canno be de e mined Le els o ce ain y ega ding ne bene i High The a ailable e idence usually includes consis en esul s om well-designed, well-conduc ed s udies in ep esen a i e p ima y ca e popula ions. These s udies assess he e ec s o he p e en i e se ice on heal h ou comes. This conclusion is he e o e unlikely o be s ongly a ec ed by he esul s o u u e s udies Mode a e The a ailable e idence is su icien o de e mine he e ec s o he p e en i e se ice on heal h ou comes, bu con idence in he es ima e is cons ained by such ac o s as he numbe , size, o quali y o indi idual s udies; inconsis ency o indings ac oss indi idual s udies; limi ed gene alizabili y o indings o ou ine p ima y ca e p ac ice; lack o cohe ence in he chain o e i- dence. As mo e in o ma ion becomes a ailable, he magni ude o di ec ion o he obse ed e ec could change, and his change may be la ge enough o al e he conclusion Low The a ailable e idence is insu icien o assess e ec s on heal h ou comes. E idence is insu icien because o : he limi ed numbe o size o s udies; impo an laws in s udy design o me hods.; inconsis ency o indings ac oss indi idual s udies; gaps in he chain o e idence; indings no gene alizable o ou ine p ima y ca e p ac ice; lack o in o ma ion on impo an heal h ou - comes. Mo e in o ma ion may allow es ima ion o e ec s on heal h ou comes 33Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 plina y ca e, as i may ha e an impac in OS, including symp om- ela ed in e en ion (LC high; GR A). • Resea ch is a p io i y in his se ing. Pa icipa ion in well-designed, independen , p ospec i e ials should be o e ed o all eligible pa ien s, whene e possible (LC high; GR A). Fo all indica ions and b eas cance ypes, pallia i e ea men is s ongly ecommended when indica ed. Diagnosis o  ecu ence andme as a ic disease • Clinical loco- egional ecu ence should be con i med by biopsy in all LRBC be o e planning any he apeu ic s a egy (LC high; GR A). • Bo h in LRBC and in MBC (p ima y o elapsed) he his ologic analyses should be pe o med i possible. In ecu ences, i will se e o con i m neoplas ic ecu ence and o echeck his ologic sub ype, as changes in ho mone ecep o s (HR) and HER2 be ween p ima y umo and ecu ences ha e been epo ed [6] (LC high; GR A). S aging • Fo bo h MBC and LRBC clinical e alua ion should include medical his o y and physical examina ion. Mini- mal s aging wo kup should include imaging echniques o ches , abdomen and bone, hema ology and biochemis- y. The ecommended imaging echniques a e body TC and bone scin ig aphy (LC mode a e; GR B). • The ole o umo ma ke s (TM) in he ollow-up o ea ly BC pa ien is con o e sial [7] (LC low; GR C). Once ecu ence is diagnosed, basal TM (CEA, CA15.3 and/ o Ca 27.4) may be pe o med. Because, i ele a ed, may help o moni o disease esponse o he apy, especially in he p esence o non-measu able disease (LC low; GR C). Howe e , ea men decisions should no be based only on he a ia ion o TM le els (LC mode a e; GR A). • O e all, he use o sys ema ic b ain image in all pa ien s in he absence o suspicious symp oms is no ecom- mended, e en in HER2-posi i e disease (LC mode a e; GR B). • The use o Posi on emission omog aphy (PET) is con- o e sial in MBC, bu can be used ins ead o TC and bone scan, i a ailable (le el o e idence low, GR B) [8]. In he pos ope a i e su eillance PET/TC is ecom- mended only in cases o equi ocal and con lic ing ind- ings [9] (LC low; GR C). Loco‑ egional ecu ence managemen Local he apy Local only ecu ence Pa ien s wi h local and egional disease a e di ided in o h ee g oups. • Ini ial ea men wi h lumpec omy + adia ion he apy (RT): ea ecu ence wi h o al mas ec omy + axilla y lymph node s aging i le el II/III axilla y dissec ion no p e iously done [10] (LC mode a e; GR B). Limi ed da a sugges ha a epea ed sen inel node biopsy may be success ully pe o med in pa ien s who ha e p e i- ously unde gone b eas -conse ing he apy and sen inel node biopsy [11] (LC low; GR C). Fo isola ed ipsila e al b eas cance ecu ences, b eas conse a i e su ge y plus pa ial b eas i adia ion is an al e na i e op ion [12] (LC mode a e; GR B). • Ini ial ea men wi h mas ec omy and no p io RT: ea ecu ence wi h su gical esec ion i possible + RT [13, 14] (LC low; GR B). • Ini ial ea men wi h mas ec omy + le el I/II axilla y dissec ion and p io RT: su gical esec ion i possible [13] (LC low; GR B). Limi ed da a ega ding addi ional i adia ion [15] (LC low; GR C). Regional onlyo local and egional ecu ence • Axilla y ecu ence: su gical esec ion i possible + RT i possible [16] (LC mode a e; GR B). • Sup acla icula and in e nal mamma y node ecu ence: RT i possible [17] (LC mode a e; GR B). Pa ien s wi h disease no amenable o adical local ea - men should be ea ed wi h induc ion chemo he apy (CT), endoc ine o an i-HER2 he apy when indica ed, and hen pallia i e adia ion, which is manda o y i he pa ien is adia ion naï e [14, 18] (LC low; GR B). Sys emic he apy CT a e i s local o egional ecu ence imp o es long- e m ou comes p ima y in ER nega i e disease. Endoc ine he apy in his se ing imp o es long- e m ou comes o ER posi i e disease [19] (LC mode a e; GR B). In case o HER2-posi i e disease and he absence o p e ious an i- HER2 adju an ea men , as uzumab is indica ed [20] (LC mode a e; GR B). Howe e , he op imal s a egy in case 34 Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 o p e ious adju an an i-HER2 ea men is unknown. CT is indica ed o endoc ine- esis an disease, as i s line in iple-nega i e disease and combined wi h an i-HER2 d ugs in HER2-posi i e disease. In pa ien s wi h disease no amenable o adical local ea men , he choice o pallia i e sys emic he apy should be made acco ding o he p inciples de ined o me as a ic BC (LC high; GR A). Endoc ine he apy inad anced HR‑posi i e/ HER2‑nega i e b eas cance Since he las SEOM guideline o MBC on 2015 se e al ad ances ha e occu ed in he ea men o endoc ine he - apy (ET) o luminal MBC [21]. Endoc ine he apy includes ER- a ge ing d ugs as single agen s o in combina ion wi h d ugs a ge ing pa hways in ol ed in ho mone esis ance such as he mTOR inhibi o , e e olimus (PIK3/AKT/mTOR pa hway) and CDK4/6 inhibi o s (cell cycle pa hway). ER- a ge ing d ugs can ac by lowe ing he le els o ci cula ing es ogens o hose ac ing di ec ly in he ER (Table2). Gene al conside a ions The p e e ed ea men o luminal ABC is ET in he majo - i y o cases. Sequen ial ET should be used as long as he pa ien seems o be bene i ing om ET and does no ha e e idence o immedia ely li e- h ea ening disease o apid p og ession o isce al disease wi h o gan dys unc ion ( isce al c isis), o when he e is an e idence o endoc ine esis ance. The p esence o isce al in ol emen alone is no a con aindica ion o he use o ET (LC high; GR A). The choice o which ET o use in each si ua ion should ake in o conside a ion: (1) p io (neo) adju an ET s “de no o” ABC; (2) Disease- ee in e al; (3) esponse o p io ET; (4) bu den o disease and symp oms; (5) menopausal s a us; (6) como bidi ies; (7) pa ien p e e ences and (8) cos s and a ailabili y. Since e en ually all pa ien s de elop esis ance o p ima - ily esis an o ET, his mus be conside ed be o e s a ing any he apy ( i s o nex lines). Al hough a clea and con- sis en de ini ion o esis ance o ET is lacking, endoc ine esis ance has been de ined by consensus (Table3), as well as he ype o ET ha should be used in i s and second lines (Table4) [5, 22]. The op imal sequence o endoc ine-based he apy is unce ain. A ailable op ions o p e- and pe imenopausal women wi h o a ian unc ion supp ession (OFS)/o a ian Table 2 Common classes o endoc ine he apy Mechanism o ac ion SERM selec i e es ogen ecep o modula o , SERD selec i e es ogen ecep o down egula o (Ful es an 500mg/mon h wi h loading dose is he ecommended dosage), GnRH gonado opin-ho mone eleasing- ho mone, NSAI non-s e oideal a oma ase inhibi o s (3 d gene a ion), SAI s e oidal a oma ase inhibi o s (3 d gene a ion) Mechanism o ac ion Class Agen Es ogen ecep o blockage SERM Tamoxi en, o emi en SERD Ful es an Es ogen dep i a ion O a ian abla ion Su ge y, adia ion O a ian supp ession wi h GnRH analogs Gose elin T ip o elin Leup olide NSAI Anas ozole Le ozole SAI Exemes ane Unknown P oges ins Meges ol ace a e Med oxyp oges e one ace a e High-dose es ogens Die hyls ilbes ol (DES) Table 3 De ini ion o endoc ine esis ance le els [5, 22] P ima y endoc ine esis ance Relapse while on he i s 2yea s o adju an endoc ine he apy (ET), o p og essi e disease (PD) wi hin i s 6mon hs o i s -line ET o ABC, while on ET Seconda y endoc ine esis ance Relapse while on adju an ET bu a e he i s 2yea s, o elapse wi hin 12mon hs o comple ing adju an ET, o PD ≥ 6mon hs a e ini ia ing ET o ABC, while on ET 35Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 unc ion abla ion (OFA), and pos -menopausal women include a oma ase inhibi o (AI), amoxi en, ul es an , AI/ ul es an plus CDK 4/6 inhibi o , and ET plus e e olimus. In la e lines, also meges ol ace a e and oes adiol, as well as epe i ion o p e iously used agen s, may be used. Besides he ER/PR posi i i y, as de ined by in e na ional guidelines [23], we do no ha e ano he use ul bioma ke o selec ET. As he blockage o he es ogen signal is he mains ay o he ea men and he es ogen le el a ies depending on he menopausal s a us, i is impo an o de ine he si ua ion o each pa ien (Table5) [24]. Pos menopausal women Fi s ‑line se ing • Thi d gene a ion AIs Anas ozole, le ozole (non- s e oidal a oma ase inhibi o s, NSAI) and exemes- ane (s e oidal a oma ase inhibi o , SAI) a e supe- io o amoxi en. The e a e no di e ences in e icacy be ween he h ee AIs [25] (LC high; GR A). • Ful es an 500mg has be e p og ession ee su i al (PFS) han anas ozol (18 s 13mon hs), wi hou di - e ences in OS in pa ien s wi hou p io exposu e o ET, pa icula ly in hose wi h non- isce al disease [26]. • The combina ion o an NSAI o ul es an plus CDK4/6 inhibi o has consis en ly shown o inc ease PFS in abou 10mon hs compa ed o ET alone, al hough wi h mo e oxici y, and he e ec is consis en in all ials and in all clinical subg oups. The h ee op ions (AI, ul es an 500mg and AI/ ul- es an + CDK4/6 inhibi o ) a e alid o he i s line. Tamoxi en is also an op ion, i o he s canno be used. The combina ion o ET plus CDK4/6 inhibi o is he p e e ed one i no o he con aindica ions a e p esen . [5, 22, 25, 26] (LC high; GR A). • Fo pa ien s ha ecei ed chemo he apy as i s line o ea men , main enance ET is a easonable op ion, al hough no assessed in clinical ials [5, 22]. The e is somewha less e idence o combina ion o ET wi h CDK4/6 inhibi o s in his si ua ion; so, ou ecommen- da ion is o conside he use o ET alone, aking in o accoun he oxici y and QoL a iables (LC low; GR B). Second‑line se ing The second line will depend on whichd ughas been used as i s line. • AI is be e han p oges in, and simila o ul es an 250mg. In second line ul es an 500mg is be e han 250mg [21]. • Fo hose pa ien s ea ed wi h p io AIs, ul es an 500 would be he op imal op ion. [27]. Also, he al e na e Table 4 Consensus de ini ion o 1s and 2nd lines o endoc ine he apy (ET) [5, 22] 1s line ET: (endoc ine sensi i e pa ien s) Newly diagnosed (de no o) ABC Relapse > 12mon hs om comple ion o (neo) adju an endoc ine he apy wi h no ea - men o ad anced o me as a ic disease ( ea men naï e in he ad anced se ing) 2nd line ET Relapse on o wi hin 12mon hs om comple ion o (neo) adju an endoc ine he apy wi h no ea men o ad anced o me as a ic disease (ea ly elapse) P og ession a e 1s line o endoc ine he apy o ad anced disease (as desc ibed be o e) Table 5 De ini ions and menopausal s a us [24] Menopause Is he pe manen cessa ion o menses Reasonable c i e ia o de e mining menopause include any o he ol- lowing P io bila e al oopho ec omy Age ≥ 60yea s Age < 60yea s and ameno heic o 12 o mo e mon hs in he absence o chemo he apy, amoxi en, o emi ene, o o a ian supp ession and ollicle-s imula ing ho mone (FSH) and es adiol in he pos meno- pausal ange I aking amoxi en o o emi ene, and age < 60yea s, hen FSH and plasma es adiol le el mus be in pos menopausal anges I is no possible o assign menopausal s a us o women who a e ecei ing an LHRH agonis o an agonis In he apy-induced ameno hea, oopho ec omy o se ial measu emen o FSH and/o es adiol a e needed o ensu e pos menopausal s a us i he use o a oma ase inhibi o s is conside ed as a componen o endoc ine he apy 36 Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 class o AI can be conside ed. Swi ching be ween s e- oidal and non-s e oidal AIs p oduces modes addi ional clinical bene i s, sugges ing pa ial non-c oss- esis ance be ween he classes o inhibi o . Howe e in hese ci - cums ances, he esponse a es o he second AI ha e gene ally been low [28]. • The combina ion o CDK4/6 inhibi o s + ul es an has p o en o be be e han ul es an alone in pa ien s wi h- ou p io exposu e o CDK4/6 inhibi o s a e p og es- sion o AIs, wi h inc ease in PFS [29–31] (LC high; GR A). • Also, he combina ion o he mTOR (mammalian a ge o apamycin) inhibi o e e olimus wi h ET (exemes- ane in a phase III ial, and also wi h amoxi en and ul es an in phase II ials) a e p og ession o an AI leads o an inc eased PFS compa ed wi h ET alone, bu wi h wo se ole ance (see e iew in nex sec ion). The e a e no da a o de ine he bes choice o ea men a e p og ession o a CDK4/6 inhibi o in i s line. Subsequen lines o  he apy The e is e y limi ed in o ma ion om p ospec i e ials in pa ien s wi h p io exposi ion o mo e han wo lines o ET. In cases whe e a posi i e e ec has been achie ed wi h p io ET, hose ET no p e iously used and p oges ins and o he ET (Table2) can be es ed [5, 22] (LC mode - a e; GR B). P emenopausal women • The e a e less da a o p emenopausal women ea ed wi h endoc ine he apy alone, as many ials wi h ET ha e included mainly pos menopausal women. Howe e , he old da a wi h OFS in combina ion wi h amoxi en, small ials wi h OFS and AI o ul es an and da a o p emenopausal pa ien s included in new ials o ET wi h CDK4/6 inhibi o s ha e p o en ha he bene i in p e- menopausal women is compa able o ha ob ained in pos menopausal [32, 33]. Fo all hese easons he in e - na ional consensus is ha he op imal managemen o p e/pe imenopausal pa ien s wi h luminal ABC consis s o he induc ion o OFS o OFA in combina ion wi h ano he endoc ine agen . Once he pa ien has been en- de ed pos menopausal, ecommenda ions o pos meno- pausal apply (LC high; GR A). • Adequa e OFS o ABC p emenopausal pa ien s can be ob ained h ough bila e al o a iec omy, con inuous use o LHRH agonis s o OFA h ough pel ic RT ( his la e is no always e ec i e, and he e o e is he leas p e e ed op ion) [5, 22] (LC high; GR A). • Fo hose pa ien s ha do no desi e o a ian supp ession, amoxi en is a easonable op ion (LC high; GR B). Endoc ine he apy inmen The e a e ew da a in hese popula ions. In e na ional guide- lines ecommenda ions a e ha ea men should be chemi- cal cas a ion wi h GnRH analogs and hen combina ion wi h ET as in pos menopausal women. Also, o hose ha do no wan cas a ion amoxi en is a easonable op ion [5] (LC mode a e; GR A). Chemo‑endoc ine he apy The e is no clea e idence ha concomi an use o ET plus chemo he apy esul s in imp o emen in OS. The e o e, his combina ion should be discou aged ou side a clinical ial [34] (LC low; GR D). Du a ion o ET ET should be con inued un il p og essi e disease o ox- ici y. Fo hose pa ien s ea ed wi h combined he apy (CDK4/6 inhibi o s o e e olimus) who ha e se e e ox- ici y o he non-ho mone componen o he combina ion ET alone can be used un il p og ession (LC high; GR A). Ta ge ed he apy inad anced b eas cance CDK4/6 inhibi o s: palbociclib, ibociclib andabemaciclib The combina ion o a CDK 4/6 inhibi o wi h an AI is he p e e ed i s -line op ion o mos pa ien s wi h endo- c ine-sensi i e HR-posi i e/HER2-nega i e me as a ic b eas cance (Table2) [35–37] (LC high; GR A). Recen e idence sugges s he e icacy o he combina- ion o CDK4/6 inhibi o wi h ul es an in i s -line se - ing and endoc ine-sensi i e disease [30] (LC high; GR A). The op ion o CDK4/6 inhibi o wi h an AI would also be app op ia e o hose pa ien s who ha e no ecei ed p io ea men wi h a CDK4/6 inhibi o (LC mode a e; GR B). Fo pa ien s wi h endoc ine- esis an HR-posi i e/ HER2-nega i e disease (Table2) ABC, he combina ion o a CDK4/6 inhibi o wi h ul es an is he p e e ed op ion. [29–31, 38] (LC high; GR A). 37Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 Bo h combina ions (CDK/AI and CDK/ ul es an ) a e applicable ega dless o he pa ien ’s menopausal s a us, al hough p e/pe imenopausal pa ien s addi ionally will equi e o a ian unc ion abla ion o supp ession. Conside ing all hese da a, a CDK4/6 inhibi o should be added o endoc ine he apy, a he la es when s a ing sec- ond-line endoc ine he apy. Because he side e ec p o ile o he CDK4/6 inhibi o s appea s subs an ially mo e ole able han ha seen wi h e e olimus, he panel o expe s ecom- mends using CDK4/6 inhibi o s a he han e e olimus as he ini ial- a ge ed he apy o pa ne wi h ET. Wi h he excep ion o HR s a us, a p esen he e a e no alida ed p edic i e bioma ke s o iden i y hose women who could bene i om endoc ine-based he apy wi h a CDK4/6 inhibi o (Table6). mTOR inhibi o s: e e olimus The combina ion o an AI wi h e e olimus, an mTOR inhibi- o , can be a alid ea men op ion o pa ien s wi h endo- c ine- esis an HR-posi i e/HER2-nega i e me as a ic b eas cance , since hey a e likely o ob ain a signi ican ly longe median PFS compa ed o a oma ase inhibi o mono he apy (7.8 s 3.2mon hs) [39, 40]. Howe e , as an OS bene i could no be p o ed [41], when conside ing his a ge ed he apy, special a en ion mus be paid o he inc eased oxici y including po en ially se e e side e ec s (e.g., non-in ec ious pneumoni is) (LC high; GR B). P ima y p ophylac ic measu es (e.g. mou hwashes wi h dexame hasone and me iculous o al hygiene) a e ecom- mended o p e en oublesome side e ec s [42]. Elde ly pa ien s should be closely moni o ed du ing ea men wi h p oac i e managemen o side e ec s. Wi h he excep ion o HR s a us, a p esen he e a e no ali- da ed p edic i e ma ke s o iden i y hose women who could bene i om endoc ine-based he apy wi h a mTOR inhibi o . PARP inhibi o s: olapa ib and alazopa ib The poly-(ADP- ibose)-polyme ase (PARP) inhibi o s olapa ib and alazopa ib a e bo h use ul ea men op ions o pa ien s wi h ad anced ge mline BRCA1/2-mu a ed iple-nega i e b eas cance (TNBC) o HER2-nega i e luminal-like b eas cance [43, 44]. Speci ically, olapa ib has been ecen ly app o ed by EMA o ea me as a ic b eas cance a e p og ession on chemo he apy. Pa ien s should ha e ecei ed p e ious ea men in he o m o (neo) adju an he apy o up o wo lines o chemo he apy wi h an h acyclines and axanes o me as a ic disease, and mus no ha e pla inum- esis an disease. This ecommenda ion is based on longe PFS (app oxi- ma ely 3mon hs), highe o e all esponse a e, good side e ec p o ile and an imp o ed quali y o li e wi h bo h PARP inhibi o s compa ed o physician’s choice o s anda d chem- o he apy (capeci abine, e ibulin, o ino elbine) in wo an- domized phase III ials (OlympiAD and EMBRACA ials) [43, 44] (LC high; GR A). T ea men o HER2‑posi i e ad anced b eas cance Fi s ‑line he apy The ini ial ea men app oach o HER2-posi i e MBC mus include a combina ion o chemo he apy and an i-HER2 he - apy [45]. The same HER2 a ge ing agen should con inue beyond p og ession h ough subsequen lines o ea men [46] (LC high; GR A). Dual-blockade wi h as uzumab, pe uzumab and axanes is he ea men o choice in he i s -line se ing ollowing he esul s o phase III CLEOPATRA ial ha demons a es a s a- is ical signi ican imp o emen in PFS and OS om adding pe uzumab o as uzumab and doce axel [47] (LC high; GR A). Replacing axane wi h ino elbine may be conside ed in ce ain ci cums ances [48] (LC low; GR C). O no e, only 10% o CLEOPATRA pa ien s we e p e i- ously exposed o as uzumab. In he PHEREXA ial es ing as uzumab plus capeci abine wi h o wi hou pe uzumab, a smalle bene i om addi ion o pe uzumab o as uzumab- exposed MBC pa ien s in second-line se ing was epo ed [49]. T as uzumab em ansine (T-DM1) was non-in e io o as uzumab- axane in he phase III i s -line MARIANNE s udy. Howe e , he e a e no da a ega ding a head- o-head compa ison be ween T-DM1 and dual HER2-blockade wi h Table 6 Compa ison o he s a us o au ho iza ion o CDK4/6 inhibi- o s HR +/HER2 − ABC ho mone ecep o -posi i e and HER2-nega i e Ad anced B eas Cance , AI a oma ase inhibi o , ET endoc ine he apy a Endoc ine he apy mus be combined wi h a lu einizing ho mone– eleasing ho mone (LH–RH) agonis in p e o pe imenopausal women EMA Indica ion PALBOCICLIB HR +/HER2 − ABC in combina ion wi h: An AIa Ful es an , in women p e iously ea ed wi h ETa RIBOCICLIB Women wi h HR +/HER2 − ABC, in combina- ion wi h an AI o Ful es an as ini ial ET o in women who ha e ecei ed p io ETa ABEMACICLIB Women wi h HR +/HER2 − ABC in combina ion wi h an AI o Ful es an , as ini ial ET o in women p e iously ea ed wi h ETa 38 Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 as uzumab, pe uzumab and a axane. Consis en ly, T-DM1 is gene ally ese ed o second-line se ing [50]. One excep- ion could be in cases o as p og ession on/a e adju an as uzumab (6–12mon hs) o i he pa ien is no sui able o axanes and dual blockade [51, 52] (LC mode a e; GR B). Second‑line he apy T-DM1 is he ecommended egimen o second-line he apy, as in he phase III EMILIA ial. I demons a ed supe io i y o e lapa inib and capeci abine in e ms o PSF and OS [53] (LC high; GR A). Howe e , i should be no ed ha he e a e limi ed da a abou T-DM1 e icacy in pe uzumab exposed pa ien s. Pe uzumab, as uzumab and chemo he apy may be con- side ed as second line in pa ien s p e iously no exposed o pe uzumab, and lapa inib and capeci abine a e sui able op ions i T-DM1 was used as i s line o i is con aindica ed [51, 52] (LC mode a e; GR B). Thi d‑line he apy andbeyond Regimens cu en ly ecommended o i s o second line should be conside ed o he la e lines, i no used p e iously [5] (LC low; GR C). T-DM1 demons a ed supe io i y o e ea men o physi- cian’s choice in hi d and la e lines in phase III TH3RESA ial [54] (LC high; GR A). Despi e he p o en ac i i y o lapa inib and capeci abine in he second line [55], his combina ion was in e io o T-DM1 in EMILIA ial, so i p e e ably should be used a e wa ds (LC mode a e; GR B). T as uzumab plus di e en chemo he apies ( ino elbine, capeci abine, gemci abine) may be an op ion i no used p e i- ously (LC low; GR C). T as uzumab and lapa inib a e also an op ion [56] (LC mod- e a e; GR B). The numbe and du a ion o each ea men line wi h HER2- a ge ed he apy and chemo he apy combina ions canno be es ablished. The chemo he apy should con inue o app oxima ely 6mon hs (o longe ) and/o o he ime o maximal esponse, depending on oxici y and in he absence o p og ession. When chemo he apy is s opped, clinicians should con inue he HER2- a ge ed he apy. A ailable da a sugges ha he bene i is main ained in hi d line and u he he apy [51, 52] (LC mode a e; GR B). HR‑posi i e HER2‑posi i e MBC The i s ea men app oach in his popula ion is HER2- a - ge ed he apy plus chemo he apy (LC high; GR A). In selec ed cases (including hose wi h con aindica ions o chemo he apy, pa ien ’s wi h a s ong p e e ence agains chemo he apy o hose wi h a long disease- ee in e al, mini- mal disease bu den, in pa icula in e ms o isce al in ol e- men , and/o s ong ER/PR exp ession) ET plus as uzumab o lapa inib can be an op ion [57, 58] (LC mode a e; GR B). I a egimen o HER2- a ge ing he apy and chemo he apy is s a ed, ET may be added o he HER2- a ge ed he apy when chemo he apy ends [51, 52] (LC low; GR C). T ea men o  iple‑nega i e ad anced b eas cance TNBC is cha ac e ized by he absence o exp ession o ER, PR, and HER2. TNBC is a he e ogeneous en i y. The e is a signi ican o e lap o TNBC wi h basal-like sub ype by PAM50, al hough hey a e no synonyms [59]. Chemo he apy (CT) is he s anda d ea men o pa ien s wi h TNBC [60]. The choice o he s a egy and cy o oxic agen s is condi ioned by a la ge numbe o ac o s and mus be conside ed indi idually. In gene al, sequencing single agen chemo he apy is p e e ed [61], limi ing combina ion he apies o pa ien s wi h agg es- si e, symp oma ic o li e- h ea ening disease [51] (LC high; GR A). The op imal du a ion o CT is no well es ablished, bu gene ally a gi en egimen should be used un il p og es- sion o disease o unaccep able oxici y [62] (as de ined oge he wi h he pa ien ) (LC mode a e; GR B). Gi en he agg essi eness o he disease and he limi ed e ec i e ea men op ions, pa ien s wi h me as a ic TNBC should always be o e ed pa icipa ion in well designed, p ospec i e, independen ials whene e such ials a e a ailable, and he pa ien is willing o pa icipa e (LC high; GR A). Fi s ‑line ea men • In pa ien s ha a e CT-naï e, an h acyclines o axanes, ei he alone o in combina ions a e conside ed as i s - line ea men [63] (LC high; GR A). This ecommen- da ion is also alid o pa ien s wi h la e ecu ences (> 1yea ) a e comple ing (neo) adju an an h acy- clines and/o axanes. • In pa ien s wi h axane-naï e and an h acycline- esis - an MBC, o wi h an h acycline cumula i e dose o ox- ici y who a e being conside ed o u he CT, axane- based he apy, p e e ably as single agen would usually be conside ed as he he apy o choice (LC high; GR A). In pa ien s p e ea ed wi h adju an axanes and an h acyclines, o he op ions such as ino elbine [64, 65] and capeci abine [66] a e also a ailable (LC mode - a e; GR B). 39Clinical and T ansla ional Oncology (2019) 21:31–45 1 3 • Be acizumab, a humanized an i-VEGF monoclo- nal an ibody, imp o es PFS and o e all esponse a e (ORR), bu no OS, when combined wi h axa- nes o capeci abine in HER2-nega i e MBC pa ien s [67–70] and may be conside ed o selec ed pa ien s wi h agg essi e o symp oma ic disease (LC mode - a e; GR C). In he absence o p edic i e bioma ke s, his bene i mus be weighed agains i s oxici y p o ile (hype ension, p o einu ia, and hemo hagic e en s) [71]. Main enance he apy wi h be acizumab plus capeci abine (compa ed o be acizumab alone) a e induc ion i s -line ea men wi h be acizumab plus doce axel imp o es PFS and OS o HER2-nega i e MBC pa ien s (PFS 11.9 s 4.3mon hs) [72]. Gi en hese esul s, capeci abine plus be acizumab may be conside ed o selec ed cases a e ini ial CT wi h doc- e axel plus be acizumab (LC low; GR C). • The combina ion o ca bopla in and gemci abine has been accep ed as con ol a m by EMA and FDA in an- domized ials, and ac ually showed a signi ican ac i - i y (PFS o a ound 5mon hs and median OS o a ound 1yea as i s -line he apy) [73]. The combina ion is ac i e in pa ien s esis an o an h acyclines and axa- nes, and i is an accep able op ion in young pa ien s wi h agg essi e symp oma ic disease (LC mode a e; GR B). • Ca bopla in as i s -line ea men o pa ien s wi h TNBC and/o ge mline BRCA1- o BRCA2-associa ed MBC (gBRCA) was as e ec i e as doce axel in he and- omized phase III TNT ial [74]. Al hough he e we e no di e ences in ORR o PFS in he unselec ed popula ion (ORR, ca bopla in 31.4% s doce axel 34.0%), gBRCA mu a ion ca ie s had an absolu e di e ence o 34.7% in ORR (68% s 33%, p = 0.03; bioma ke , ea men in e - ac ion p = 0.01), which also ansla ed in signi ican di - e ences in PFS (6.8 s. 3.1mon hs, p = 0.04). Based on hese esul s, ca bopla in can be conside ed as an op ion bo h o unselec ed TNBC pa ien s (LC mode a e; GR B) and o gBRCA MBC (LC high; GR A). • A ezolizumab in combina ion wi h nab-pacli axel has shown o imp o e PFS in pa ien s wi h me as a ic TNBC when compa ed o nab-pacli axel alone [75]. This was achie ed bo h in he in en - o- ea popula ion (7.2 s 5.5mon hs; HR 0.80) and among pa ien s wi h PD-L1-exp essing umo s in he in il a ing immune cells. In he in en ion- o- ea analysis, he median OS was 21.3mon hs wi h a ezolizumab plus nab-pacli axel and 17.6mon hs wi h placebo plus nab-pacli axel (HR 0.84; p = 0.08), and among pa ien s wi h PD-L1-pos- i i e umo s, he median OS was 25.0mon hs and 15.5mon hs, espec i ely. Al hough encou aging, addi- ional esul s o ials es ing immuno he apy agen s plus chemo he apy mus be awai ed be o e a ecommenda ion can be made o inco po a e immuno he apy in i s -line ea men o TNBC. Second and u he lines o  ea men The e is no limi o he numbe o he apy lines o be p o- posed o me as a ic TNBC pa ien s, as long as a good quali y o li e is main ained [51]. In pa ien s p e ea ed wi h an an h acycline and a ax- ane, and who do no need combina ion CT, single-agen capeci abine, ino elbine (o al o IV), and e ibulin a e he p e e ed choices [64, 65, 76, 77] (LC mode a e; GR B). Addi ional choices include an al e na i e axane (s anda d o nab-pacli axel), echallenge wi h an h acyclines (liposo- mal o mula ion), gemci abine, o pla inum agen s [78] (LC mode a e; GR B). Chemo he apy inluminal‑ad anced b eas cance Chemo he apy should be he s anda d ea men o HR-pos- i i e MBC pa ien s ha a e e ac o y o endoc ine he apy, and o hose women wi h iple-nega i e MBC [5] (see ec- ommenda ions in he TNBC sec ion). An h acyclines, axanes, ino elbine, capeci abine, gem- ci abine and e ibulin a e easonable and a ailable op ions. The choice o he s a egy and cy o oxic agen s mus be con- side ed indi idually, aking in o accoun p e ious he apies and hei oxici ies, umo bu den, ime un il ecu ence a e p io he apy, biological age, pe o mance s a us, como - bidi ies, es ima e li e expec ancy, need o a apid disease/ symp om con ol and pa ien ’s p e e ences [5, 79]. No pa - icula chemo he apeu ic agen o egimen has been able o p o ide consis en gains in su i al in hese pa ien s [79]. In gene al, sequencing single-agen chemo he apy is p e e ed [61, 80], limi ing combina ion he apies o pa ien s wi h agg essi e, symp oma ic o li e- h ea ening disease (a leas same e icacy, less oxici y) [5] (LC high; GR A). E alua ion o esponse o chemo he apy should gene ally occu a e wo o ou cycles depending on he dynamics o he disease, he loca ion and ex en o me as a ic in ol e- men and ype o ea men . Imaging o a ge lesions may be su icien in many pa ien s (LC mode a e; GR B) [5]. Subjec- i e and objec i e oxici ies mus be e alua ed epea edly, as well as he symp oms and pe o mance s a us. Du a ion o each egimen and he numbe o egimens should be ailo ed o each indi idual pa ien .