Infiltration of M2 Macrophages and Regulatory T Cells Plays a Role in Recurrence of Renal Cell Carcinoma

KidneyCancer

Infiltration of M2 Macrophages and Regulatory T Cells Plays a Role in Recurrence of Renal Cell Carcinoma

Sabina Davidsson^a,*,Michelangelo Fiorentino^b,Francesca Giunchi^b,Margareta Eriksson^c, AnnErlandsson^d, PernillaSundqvist^a,y,Jessica Carlsson^a,y

aDepartmentofUrology,FacultyofMedicineandHealth,ÖrebroUniversity,Örebro,Sweden;^bDepartmentofPathology,MaggioreHospitalandS.Orsola- MalpighiHospital,UniversityofBologna,Bologna,Italy;^cDepartmentofClinical PathologyandCytology,CentralHospitalKarlstad,Karlstad,Sweden;

dDepartmentofEnvironmentalandLifeSciences/biology,KarlstadUniversity,Karlstad,Sweden a v a i l ab l e a t w w w . s c i e n c e d i r e c t . c o m

j o u r n a l h o m e p a g e : w w w . e u -o p e n s c i en c e . eu r o p e an u r o l o g y . c o m

Articleinfo

Articlehistory:

AcceptedJune11,2020 AssociateEditor:AxelBex Keywords:

Renalcellcarcinoma Immunohistochemistry CD163⁺M2macrophages FOXP3⁺regulatoryTcells

Abstract

Background: Ithasbeenhypothesizedthat M2macrophages andregulatoryTcells (Tregs)maycontributetotumorprogressionbysuppressionofantitumorimmunity.

Objective: To investigate theassociationbetweeninfiltrationofCD163⁺M2macro- phagesandCD4⁺FOXP3⁺Tregswithclinicaloutcomesinrenalcellcarcinomapatients.

Design,setting,andparticipants: Acohortof346patientsdiagnosedwithrenalcell carcinomaat ÖrebroUniversityHospitalbetween1986 and 2011was evaluatedfor CD163⁺M2macrophageandCD4⁺FOXP3⁺Treginfiltrationbyimmunohistochemistry.

Outcomemeasurementsandstatisticalanalysis: Associationsbetweenclinicopatho- logicalfeaturesandinfiltrationofCD163⁺M2macrophagesand/orCD4⁺FOXP3⁺Tregs wereestimatedwithchi-squareorFisher’sexacttests.Forsurvivalanalyses,Kaplan- Meiercurveswithlog-ranktestsandmultivariateCoxproportionalhazardsregression modelswereused.

Resultsandlimitations: WefoundthatinfiltrationofCD163⁺M2macrophagesand CD4⁺FOXP3⁺Tregswereassociated withadverse clinicaloutcomes.Our datafurther demonstratethatCD163⁺M2macrophagesandCD4⁺FOXP3⁺Tregscolocalizeintumor andnormaltissue,andthatthiscolocalizationmayhavesynergisticeffectsontumor aggressiveness.Theuseoftissuemicroarraysratherthanwholesectionsmaybeviewed asalimitation.

Conclusions: InfiltrationofCD163⁺M2macrophagesandCD4⁺FOXP3⁺Tregsisassoci- atedwithrecurrenceofrenalcellcarcinoma,andcolocalizationofthesecelltypesmay haveanassociationwithclinicaloutcome.

Patientsummary: Theaimofthisstudywastoinvestigatetheassociationbetween infiltrationofM2macrophagesandregulatoryTcellswithclinicaloutcomesinrenalcell carcinoma.Wedemonstratedthatrenalcellcarcinomapatientswithhighinfiltrationof boththesecelltypesareatanincreasedriskofpoorclinicaloutcomes.

©2020TheAuthor(s).PublishedbyElsevierB.V.onbehalfofEuropeanAssociationof Urology.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creati- vecommons.org/licenses/by-nc-nd/4.0/).

y Theseauthorscontributedequallytothiswork.

*Correspondingauthor.Department ofUrology,O¨ rebroUniversityHospital, SE-70185O¨ rebro, Sweden.Tel.:+46196026604;Fax:+46196026650.

E-mailaddress:sabina.davidsson@regionorebrolan.se(S.Davidsson).

http://dx.doi.org/10.1016/j.euros.2020.06.003

2666-1683/©2020TheAuthor(s).PublishedbyElsevierB.V.onbehalfofEuropeanAssociationofUrology.Thisisanopenaccessarticle undertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Oneofthehallmarks ofcanceristhecapabilityoftumor cells to evade immune destruction, and one mechanism utilizedbytumorsisto recruitimmunesuppressivecells, suchasM2macrophagesandregulatoryTcells(Tregs),into themicroenvironment[1].

Macrophages are the most abundant innate immune cells within the microenvironment of most tumors.

Dependingonthemicro milieu,macrophagescanacquire distinctsubtypes,eitherpro-oranti-inflammatory[2].M1 macrophages are immunostimulatory and antitumoral, while M2 macrophages are immunosuppressive and contributestoaprotumoralenvironment[3].Highinfiltra- tionofCD163⁺M2macrophageshasbeenassociatedwith worse clinical outcome in a number of malignancies includingrenalcellcarcinoma(RCC)[4–8].

It has been suggested that Tregs are a key player in antitumor immune suppression and thereby promote tumor progression [9,10]. The most specific marker to identify Tregs is the transcriptionfactor FOXP3 [11]. The prognostic valueofintratumoralFOXP3⁺ Tregshasinten- sivelybeenevaluatedpreviously,andpositiveassociations werefoundforanumberofdifferentmalignancies[12–14]. An association between infiltration of FOXP3⁺ Tregs and worse clinical outcome has also been reported in themajorityofstudiesinvestigatingRCCpatients[15–17]. However,conflictingdataexist[18].

Inthemajorityofstudiesevaluatingtheclinicalimpact ofM2 macrophagesand Tregsincancer progression, the prognostic value has been assessed based on separate infiltrationof either CD163⁺ M2 macrophages orFOXP3⁺ Tregs.Thecorrelationbetweenthesynergisticeffectandthe clinical outcome has been reported only for a limited numberofmalignancies.Recently,Sunetal[19]foundthat laryngeal squamous cell carcinoma patients with a high numberofCD163⁺M2macrophagesandFOXP3⁺Tregshave shorteroverallsurvivalthanpatientswithlowinfiltration.

To our knowledge, no previous study has evaluated the prognosticvalueandsynergisticeffectonclinicaloutcome ofsimultaneousinfiltrationofCD163⁺M2macrophagesand FOXP3⁺Tregsintotumorandtumor-adjacentnormaltissue inRCC.

The aim of thisstudy was to evaluate the prognostic valueofseparateinfiltrationandcolocalizedinfiltrationof CD163⁺M2macrophagesandCD4⁺FOXP3⁺Tregsfordisease recurrenceandcancer-specificdeath.

2. Patientsandmethods 2.1. Caseandtissuecollection

Thepresentstudyincludes 346patients diagnosedwith RCCand treated with radical nephrectomy (n=303) or nephron-sparing surgery(n=43)attheÖrebroUniversityHospital,Örebro,Sweden, between January 1986 and December 2011. The study cohort was followed for cancer-specific and all-cause mortality until December 2015. The cohort has previously been described by Grabowskaetal[20].

Formalin-fixed,paraffin-embedded(FFPE)blocksandcorresponding hematoxylinandeosin(H&E)slidesfromallcaseswereretrievedand re-reviewedbytwoexperiencedgenitourinarypathologists(M.F.andF.

G.).Thepathologistsre-evaluatedtumorspecimensbasedonthetumor, node, metastasis (TNM) classification according to American Joint Committee on Cancer (AJCC) 2010 eighth edition. The histological subtype of RCCwas classified accordingto the 2008 World Health Organization(WHO) tumorclassification.Moreover, eachtumorwas gradedaccordingtotheWHOnucleolargradingsystem.Inaddition,the pathologists circledtumorandtumor-adjacentnormalareas onH&E slidescorrespondingtotheFFPEblocks.

Thisstudywasapprovedbytheregionalethicsreviewboard(ethical approvalnumber2010/135and2015/353).

2.2. Immunohistochemistry

Tissuemicroarrays(TMAs)wereconstructedpreviouslyforthiscohort, includingthreetissuecoreswithadiameterof0.6mmfromeachtumor andtumor-adjacentnormalarea.Four-micrometersectionswereusedfor immunohistochemistry.Theanti-CD163antibody(LeikaBiosystems,USA) wereusedtoidentifyM2macrophagesandanti-CD4andanti-FOXP3 antibodies(Agilent Dako,USA,andThermoFisherScientific,USA) to identifyTregs.ImmunohistochemicalstainingofCD163andCD4/FOXP3 was performed onAutostainer Link 48 (AgilentDako). The optimal conditionsfortheprimarymousemonoclonalantibodyanti-CD163(clone 10D6)weretoperformantigenretrievalbysteamheatat97^CwithHIER FLEXTRSlowbuffer,EnvisionFlexTargetRetrievalSolutionlowpH,for 20min(K805;AgilentDako).Theslideswerethenincubatedwiththe antibodyat1:200dilutionfor30minfollowedbyvisualizationwithan amplification system including EnVision FLEX/rabbit linker, EnVision FLEX/HRP,andFLEXDABSub-Chromogen(AgilentDako).

Theoptimalconditionsfortheprimaryantibodiesusedfordouble staining,mousemonoclonalantibodyanti-CD4(clone4B12,RTU)and mouse monoclonal antibody anti-FOXP3 (clone 236A/E7), were to perform antigenretrieval bysteamheatat97^Cwith EnvisionFlex Targetretrievalsolutionfor20minatpH9(DakoK8004).Theslides werethenincubatedwiththeanti-CD4antibodyfollowedbytheanti- FOXP3 antibody diluted 1:25. Revelation was performed with an amplificationsystem includingEnVision FLEX/rabbitlinker, EnVision FLEX/HRP,andFLEXDABSub-Chromogen(AgilentDako)forCD4and Vina Green Chromogen (Biocare, USA) for FOXP3. All slides were counterstainedwithMayer’shematoxylinandmountedusingTissue-Tek coverslippingfilm(SakuraFinetek,USA).Tissuefromtonsilswasusedas a positive control. The Panoramic250 Flash IIsystem (3DHISTECH, Hungary)wasusedtoconvertallCD163-,CD4-,andFOXP3-stainedslides intohigh-resolutiondigitalslidesusingthesoftwareCaseviewerversion 2.1(3DHISTEC).

2.3. EvaluationofCD163andCD4⁺Foxp3⁺expression

We quantified M2 macrophagesby CD163expression andTregs by simultaneousCD4andFOXP3expression.AllTregscellswerecounted.

ForM2macrophages,upto200cellswerecounted;agreaternumberof positivecellswererecordedinasinglecategoryas>200.Theobservers (S.D.andA.E.)wereblindedtotheclinicaldata.Themediannumberof CD163⁺M2macrophagesforthethreecoreswascalculated.Patients withCD⁺FOXP3⁺TregsweredefinedasTreg-positivecasesandpatients withoutCD4⁺FOXP3⁺TregsasTreg-negativecases.

2.4. Statistics

TheassociationsofCD163⁺M2macrophagesandCD4⁺FOXP3⁺Tregswith clinicopathologicalcharacteristics wereevaluatedby chi-squaretest,

Fisher’s exact test, or Mann-Whitney U test as appropriate. The correlationbetweenCD163⁺M2macrophagesandCD4⁺FOXP3⁺Tregs was evaluated by Spearman’s rank coefficient test. Recurrence and cancer-specificdeathwereusedasendpoints,andaunivariateanalysis oftheassociationofoutcomewasperformedusingtheKaplan-Meier methods.Thesignificancebetweenthecurveswasassessedbythelog- ranktest. For the multivariate analysis, aCox proportional hazards regressionmodelwasused,adjustedfornucleolargrade,AJCCstage,and primary tumor size (40mm). Two-side p<0.05 was considered statistically significant.The statistical analysis wasperformed using SPSSversion22(IBM,USA).

3. Results

In this study, we evaluated tissue obtained from 346 patients diagnosed with RCC (195 males and 151females). Theclinical pathological characteristics of the patients are summarized in Table 1. The median follow-uptimewas81mo(range,0–325)andthemedian time to recurrence was 104 mo (range, 2–346). At the last follow-up, 174 of the 346 study participants were deceased,ofwhom79diedfromRCCwithinthefollow-up time.

3.1. InfiltrationofCD163⁺M2macrophagesandCD4⁺FOXP3⁺ TregsispositivelycorrelatedwitheachotherinRCCtissueand associatedwithadverseclinicaloutcomes

Here, we examined the infiltration of CD163⁺ M2 macrophages and CD4⁺FOXP3⁺ Tregs into tumor and tumor-adjacent normal tissue obtained from 346 RCC patients(Fig.1).

ThemediannumbersofCD163⁺M2macrophageswere 65 (interquartilerange [IQR]58.5)and141 (IQR116.0)in tumor-adjacentnormalareaandtumorarea,respectively.

ThemeannumbersofCD4⁺FOXP3⁺Tregsincorresponding areaswere1.2(standarddeviation[SD]2.8)and2.3(SD4.9), respectively(Figs.2and3).Whenassessingtherelationship betweenCD163⁺M2macrophagesandCD4⁺FOXP3⁺Tregs, we observed that patients with CD163⁺ M2 macrophage infiltration had a significantly higher number of CD4⁺FOXP3⁺ Tregs. The density between the two cell populations was positively correlated in both tumor- adjacent normal tissue (R=0.31, p<0.001) and tumor tissue(R=0.31,p<0.001).

To determine the clinical importance of CD163⁺ M2 macrophages and CD4⁺FOXP3⁺ Tregs in tumor-adjacent normal tissue and tumor tissue, we analyzed their associations with clinicopathological characteristics. In tumor-adjacenttissue,thelevelsofinfiltrationofCD163⁺ M2 macrophages were significantly higher in patients presenting with higher T (p=0.03) and AJCC (p=0.02) stages, andin patients with cancer recurrence (p=0.02).

Moreover,thelevelsofCD4⁺FOXP3⁺Treginfiltrationwere significantlyhigherinpatientswithnuclear grades3and 4 than in those with nuclear grades 1 and 2 (p=0.03).

Table2summarizestheassociationsbetweenclinicopatho- logical characteristics and the results of the immuno- stainingintumor-adjacentnormaltissue.No associations betweenCD4⁺FOXP3positivityorCD163positivityintumor tissue andclinicopathological characteristicswere found, even though a trend was observed for CD4⁺FOXP3⁺ Treg infiltration and cancer recurrence (p=0.06). Table 3 summarizes the associations betweenclinicopathological characteristics and the results of the immunostaining in tumortissue.

3.2. CD163⁺M2macrophageandCD4⁺FOXP3⁺Treginfiltration andclinicaloutcome

The medianfollow-up of the cohortused inthe present studywas81mo,and79patientsdiedfromRCC.According totheKaplan-Meieranalysis,patientswithahighnumber of intratumoral CD163⁺ M2 macrophages had a shorter mediantimetorecurrencethanthosewithalownumberof CD163⁺M2macrophages(82vs127mo,p=0.038;Fig.4).A shorter median time to recurrence was also found for patientswithahighnumberofCD163⁺M2macrophagesin tumor-adjacent normal tissue than in those with a low number of CD163⁺ M2 macrophages (93 vs 127 mo, p<0.01;Fig.5).Further,therewasasignificantdifference inmediantimetorecurrencebetweenpatientswithhigh intratumoral infiltration ofCD163⁺ M2 macrophagesand Table1–Characteristicsofrenalcellcarcinomapatients

Totalcohort(N=346),n(%)

Smoking

Yes 96(36.1)

No 170(63.9)

Missing 80

Primarydiameter(mm)

Median(min–max) 60(10–180)

WHOnucleargrade

1 22(6.9)

2 151(47.3)

3 108(33.9)

4 38(11.9)

Missing 27

AJCCstage

1 183(52.9)

2 62(17.9)

3 62(17.9)

4 39(11.3)

pTstage

T1 193(55.8)

T2 78(22.5)

T3 64(18.5)

T4 11(3.2)

Nstage

N0 333(96.2)

N1 12(3.5)

Nx 1(0.3)

Mstage

M0 316(91.3)

M1 30(8.7)

Recurrence

Yes 92(27.1)

No 248(72.9)

Missing 6

AJCC=American Joint Committee on Cancer; WHO=World Health Organization.

CD4⁺FOXP3⁺Tregspresent,comparedwiththosewithlow infiltrationofCD163⁺M2macrophagesandnoCD4⁺FOXP3⁺ Tregspresent(79versus197mo,p<0.01).Toevaluatethe impactofbothCD163⁺M2macrophagesandCD4⁺FOXP3⁺ Tregs in progression of RCC, we compared patients with concurrent infiltration of CD163⁺ M2 macrophages and CD4⁺FOXP3⁺Tregswiththosewithonlyhighinfiltrationof CD163⁺ M2 macrophages. A shorter median time to

recurrencewasseeninpatientswithconcurrentinfiltration (79 vs 91 mo), although it did not reach statistical significance(p=0.08;Fig.6).

3.3. PrognosticvalueofCD163⁺M2macrophagesand CD4⁺FOXP3⁺TregsinRCC

WethenevaluatedthesignificanceofCD163⁺M2macro- phage andCD4⁺FOXP3⁺ Treginfiltration, inaddition to a number of clinicopathological variables, as predictors of

Fig.1–RepresentativeimmunohistochemicalimagesofRCCcoresstainedtovisualizetheTregmarkersCD4/FOXP3(brownandgreen)andtheM2 macrophagesmarkerCD163(brown):(A)ahighnumberofTregs,(B)alownumberofTregs,(C)ahighnumberofM2macrophages,and(D)alow numberofM2macrophages.Diameterofcoresis0.6mm.RCC=renalcellcarcinoma;Treg=regulatoryTcell.

Fig.3–ComparisonofCD163⁺M2macrophageinfiltrationintotumor tissueinRCCpatientswithandwithoutCD4⁺FOXP3⁺Treginfiltration.

RCC=renalcellcarcinoma;Treg=regulatoryTcell.

Fig.2–ComparisonofCD163⁺M2macrophageinfiltrationintumor- adjacentnormaltissueinRCCpatientswithandwithoutCD4⁺FOXP3⁺ Treginfiltration.RCC=renalcellcarcinoma;Treg=regulatoryTcell.

recurrence and cancer-specific death (Tables 4 and5). A univariate Cox regression analysis revealed a significant positiveassociationoftimetorecurrencewithhighCD163⁺ M2 macrophageinfiltration,into both tumorand tumor- adjacent normal tissue (hazard ratio [HR] 1.62; 95%

confidence interval [CI] 1.02–2.56 and HR 1.86; 95% CI 1.16–2.98,respectively).Asignificantpositiveassociationof time to recurrence was also found with simultaneous infiltration of CD163⁺ M2 macrophages and CD4⁺FOXP3⁺ Tregs into tumor tissue(HR 3.12; 95%CI 1.43–6.81).The associationforinfiltrationofCD163⁺M2macrophagesinto tumortissue(HR1.77;95%CI1.02–3.07)andforsimulta- neous infiltration of CD163⁺ M2 macrophages and CD4⁺FOXP3⁺ Tregs into tumor tissue (HR 3.32; 95% CI 1.45–7.63) remained significant afteradjusting for WHO nucleolargrade,AJCCstage,andprimarytumorsize.

WhenevaluatingthesignificanceofCD163⁺M2macro- phagesandCD4⁺FOXP3⁺Tregsaspredictorsofmediantime tocancer-specificdeath,CD163⁺M2macrophagesintumor- adjacentnormaltissuewere foundtobeanindependent prognosticmarkerformediantimetocancer-specificdeath inthemultivariateanalysis(p=0.02).

4. Discussion

Inthepresentstudy,wedemonstrate,forthefirsttimein kidney cancer, that CD163⁺ M2 macrophages and CD4⁺FOXP3⁺Tregscolocalizeintumorandtumor-adjacent normal tissue, and that this colocalization may have an associationwithtumorrecurrence.

M1macrophagesaresuggestedtobeimmunostimula- tory andantitumoral,while M2 macrophagesare immu- nosuppressive and contribute to a protumoral environment [2]. We found that high infiltration of CD163⁺M2macrophages,intobothtumorandthenormal adjacent tissue, was associated with worse clinical prognosis.Thepresentstudyalsoconfirmedtheprognostic roleofCD163⁺ M2macrophagessince apositiveassocia- tion of time to recurrence was found. In addition, our results showed that high infiltration of CD163⁺ M2 macrophages into tumor-adjacent normal tissue was associated with shorter time to RCC-specific lethality.

TheprognosticroleofCD163⁺M2macrophageinfiltratein theadjacentnormaltissue,ratherthaninthetumortissue, issomewhatunexpectedbutnotsurprising.Inaprevious Table2–AssociationsbetweenCD4/FOXP3andCD163immunoreactivityandclinicalcharacteristicsintumor-adjacentnormaltissuein patientswithrenalcellcarcinoma

CD4⁺FOXP3⁺Tregs CD163⁺M2macrophages

Negative Positive pvalue Lowinfiltration Highinfiltration pvalue

Gender 0.05^a 0.03^a

Male 94(61) 61(39) 95(56) 75(44)

Female 56(48) 60(52) 54(43) 71(57)

WHOnucleargrade 0.03^b 0.27^a

1 10(53) 9(47) 12(63) 7(37)

2 53(47) 61(53) 65(50) 66(50)

3 56(68) 30(32) 47(51) 45(49)

4 20(69) 9(31) 11(36) 20(64)

AJCCstage 0.34^a 0.02^a

1 73(52) 67(48) 89(57) 66(43)

2 30(64) 17(36) 29(54) 25(46)

3 27(51) 26(49) 19(35) 35(65)

4 20(65) 11(35) 12(38) 20(62)

pTstage 0.95^b 0.03^b

pT1 80(55) 65(45) 92(56) 72(44)

pT2 35(57) 26(43) 35(52) 33(48)

pT3 30(56) 24(44) 20(36) 35(64)

pT4 5(63) 3(37) 2(25) 6(75)

Nstage 0.76^b 0.99^a

N0 141(55) 116(45) 141(50) 140(50)

N1 7(64) 4(36) 6(55) 5(45)

Nx

Mstage 0.29^a 0.54^a

M0 134(54) 113(46) 138(51) 132(49)

M1 16(67) 8(33) 11(44) 14(56)

Recurrence 0.89^a 0.02^a

Yes 41(56) 32(44) 29(38) 47(62)

No 105(54) 88(46) 116(55) 97(45)

Causeofdeath 0.48^a 0.13^a

Renalcellcancer 41(63) 24(37) 21(33) 42(67)

Otherreason 39(56) 31(44) 38(47) 43(53)

AJCC=AmericanJointCommitteeonCancer;Tregs=regulatoryTcells;WHO=WorldHealthOrganization.

a Fisher’sexacttestwasused.

b Chi-squaretestwasused.

Table3–AssociationsbetweenCD4⁺FOXP3andCD163immunoreactivityandclinicalcharacteristicsintumortissueinpatientswithrenal cellcarcinoma

CD4⁺FOXP3⁺Tregs CD163⁺M2macrophages

Negative Positive pvalue Lowinfiltration Highinfiltration pvalue

Gender 0.40^a 0.91^a

Male 66(41) 97(59) 77(50) 77(50)

Female 57(46) 67(54) 63(51) 61(49)

WHOnucleargrade 0.11^a 0.75^b

1 11(65) 6(35) 10(56) 8(44)

2 54(42) 76(58) 60(47) 68(53)

3 36(42) 50(58) 39(46) 46(54)

4 9(28) 23(72) 16(55) 13(45)

AJCCstage 0.82^a 0.91^a

1 61(41) 88(59) 75(51) 71(49)

2 22(42) 31(58) 23(47) 26(53)

3 26(48) 28(52) 28(53) 25(47)

4 14(45) 17(55) 14(47) 16(53)

pTstage 0.39^b 0.82^b

pT1 65(42) 91(58) 80(52) 74(48)

pT2 28(42) 39(58) 28(45) 34(55)

pT3 28(51) 27(49) 28(52) 26(48)

pT4 2(22) 7(78) 4(50) 4(50)

Nstage 0.99^b 0.99^b

N0 118(43) 155(57) 134(51) 131(49)

N1 5(46) 6(54) 5(50) 5(50)

Nx

Mstage 0.28^a 0.99^a

M0 110(42) 153(58) 128(50) 126(50)

M1 13(54) 11(46) 12(50) 12(50)

Recurrence 0.06^a 0.08^a

Yes 26(33) 52(67) 32(43) 45(57)

No 94(46) 110(54) 105(54) 91(46)

Causeofdeath 0.99^a 0.99^a

Renalcellcancer 27(43) 36(57) 33(52) 30(48)

Otherreason 33(42) 45(58) 44(51) 42(49)

AJCC=AmericanJointCommitteeonCancer;Tregs=regulatoryTcells;WHO=WorldHealthOrganization.

a Fisher’sexacttestwasused.

b Chi-squaretestwasused.

Fig.4–Kaplan-MeieranalysisofmediantimetorecurrenceintermsofintratumoralCD163⁺M2macrophagesshowingshortermediantimeto recurrenceinpatientswithahighnumberofCD163⁺M2macrophages.Lowinfiltrationspecifiedasunder141CD163+M2macrophages.

Cum=cumulative.