CD4:CD8 T-Cell Ratio Differs Significantly in Diffuse Large B-Cell ...

INTERNATIONAL TRENDS IN IMMUNITY VOL.1 NO.2 APRIL 2013ISSN 2326-3121 (Print) ISSN 2326-313X (Online) http://www.researchpub.org/journal/iti/iti.htmlCD4:CD8 T-Cell Ratio Differs Significantly inDiffuse Large B-Cell Lymphomas from OtherLymphoma Subtypes Independently fromLymph Node LocalizationCécile Daussy 1,2,3 , Diane Damotte 1,3,4,5 , Thierry J. Molina 3,4 , Mikael Roussel 6 , Thierry Fest 6,7 , AudreyVarin 1,2,3 , Jean-Yves Perrot 8 , Lamia Ouafi 3,4 , Hélène Merle-Béral 9 , Pierre Julia 3,10 , Wolf HermanFridman 1,2,3 , Catherine Sautès-Fridman 1,2,3 , and Sylvain Fisson 1,2,3,11,12,13, *Abstract— Although the identification of newly described T-cellsubpopulations in human B-cell lymphoma is under investigation,contradictory data still persist regarding the relative proportionsof CD4 + and CD8 + T-cells in tumoral lymph nodes (LN).CD4:CD8 T-cell ratio in LN from Hodgkin (HL) and non HodgkinB-cell lymphoma subtypes at diagnosis was analyzed by flowcytometry based on three independent cohorts (n=112). Asignificantly lower median CD4:CD8 ratio (0.82±1.06) wasobserved only in diffuse large B-cell lymphomas (DLBCL). Ratioswere unchanged by the localizations suggesting that thelymphoma subtype influences the relative proportions of CD4 +and CD8 + T-cells in the tumoral LN.Index Terms— CD4:CD8 T-cell ratio, B-cell lymphoma,Hodgkin lymphoma, lymph node, flow cytometryAbbreviations— DLBCL, diffuse large B-cell lymphoma; FL,follicular lymphoma; HL, Hodgkin lymphoma; LL, lymphocyticlymphoma; LN, lymph node; NHL, non Hodgkin B-celllymphoma; Tc, cytotoxic T-cells; Th, helper T-cells; TIL, tumorinfiltrating lymphocytes; Treg, regulatory T-cells.Paper submitted for review on Feb 24 2103 and accepted March 10, 2013.This work was supported by the Institut National du Cancer (GrantRC013-C06N631-2005), the Association pour la Recherche contre le Cancer,the Institut National de la Santéet de la Recherche Médicale, the UniversityPierre and Marie Curie, the University Paris-Descartes, the pole decompétitivitéIle de France (ImmuCan project). C.D. was recipient of a grantfrom the Institut National du Cancer.1 Institut National de la Santé et de la Recherche Médicale (INSERM),UMRS872, Centre de Recherche des Cordeliers, Paris, F-75006, France;2 UniversitéPierre et Marie Curie-Paris 6, UMRS 872, Paris F-75006, France;3 Université Paris Descartes, UMRS 872, Paris F-75006, France; 4 Serviced’Anatomie Pathologique, Hôtel Dieu, AP-HP, Paris F-75005, France; 5 Serviced’Anatomie Pathologique, Hôpital Européen Georges-Pompidou (HEGP),AP-HP, Paris F-75015, France; 6 Laboratoire d’Hématologie, UniversityHospital Pontchaillou, Rennes, France; 7 INSERM U917, Rennes, France;8 Service d’Hématologie Biologique, Hôtel Dieu, AP-HP, Paris F-75005,France; 9 Service d’Hématologie Biologique, Hôpital de la Pitié-Salpêtrière,AP-HP, Paris F-75013, France; 10 Service de Chirurgie Cardiaque et Vasculaire,Hôpital Européen Georges-Pompidou, AP-HP, Paris F-75015, France.11 Généthon, Evry, France; 12 INSERM, U951, Evry, France; 13 Université d’EvryVal d'Essonne, UMRS 951, Evry, France.* Correspondance: Sylvain Fisson. Généthon, INSERM UMRS 951, 1 bisrue de l’Internationale, Evry F-91002, France. Phone: +33 169 473 440 ; Fax :+33 160 778 698 ; e-mail: sylvain.fisson@univ-evry.fr45II. INTRODUCTIONNTEREST regarding the importance of the local adaptiveimmune microenvironment in cancer is growing [1-3].Although CD4 and CD8 respective proportions and functionwere reported as predictors of patient outcome in cancer,studies were often unclear or contradictory in lymphomas [4-7].Some data suggest that increased numbers of cytotoxic CD8 +T-cells could be an unfavourable marker [8]. Some of thesecontroversial results could be due to the heterogeneity of thelymphoma subtypes and lymph node (LN) localization,methods of study, reagents used and time of analysis (ie. beforeor after treatment) [6-7]. Our aim was to analyze extensively byflow cytometry the CD4:CD8 T-cell ratio in LN from Hodgkinlymphomas and different subtypes of non Hodgkin B-celllymphomas at diagnosis and to compare them to normal andreactive LN. 113 LN from three independent cohorts of 112untreated patients were analyzed both by flow cytometry andhistopathology, and were classified in 6 entities: normal (N);reactive (R); Hodgkin Lymphoma (HL); lymphocyticlymphoma (LL); follicular lymphoma (FL); diffuse large B-celllymphoma (DLBCL). The cohorts included 24 primarymediastinal HL and DLBCL. Our results clearly showed alower CD4:CD8 ratio in DLBCL, including the mediastinallocalization, compared to other lymphomas subtypes.II. MATERIALS AND METHODSA. Patients and controlsControl and tumoral LN were selected at diagnosis fromthree independent cohorts (Table 1). This study was performedin accordance with the Declaration of Helsinki. All data weretreated confidentially. The study was performed with theapproval of hospital review boards. All specimens wereanalyzed in the pathology department from Paris-HEGP,Paris-Hôtel Dieu, and Rennes-Pontchaillou hospitals and werereviewed blindly by expert hematopathologists (DD, TJM).Normal LN corresponded to intra-thoracic non inflammatoryand non tumoral LN harvested during cardiac surgery (n=11,

INTERNATIONAL TRENDS IN IMMUNITY VOL.1 NO.2 APRIL 2013ISSN 2326-3121 (Print) ISSN 2326-313X (Online) http://www.researchpub.org/journal/iti/iti.htmlcohort #1). Reactive LN were non tumoral but inflammatory oradjacent to a lymphomatous LN for one DLBCL patient (n=7,cohort#1). Tumoral LN were HL, LL, FL or DLBCL (n=27,cohort#1; n=22, cohort#2; n=46, cohort#3).TABLE ISUMMARY OF PATIENT CHARACTERISTICS AT DIAGNOSIS FOR COHORTS #1(PARIS-HEGP HOSPITAL), #2 (PARIS-HÔTEL DIEU HOSPITAL), AND #3(RENNES-PONTCHAILLOU HOSPITAL).Cohort#1#2#3Lymph nodeB. Cell isolationFresh unfixed LN were minced with a sterile blade in RPMImedium and were either digested with 2mg/mL Collagenase(Worthington, Roche, Meylan, France) and 0.1mg/mL Dnase I(Roche, Meylan, France) at 37°C for 50min, filtered and rinsedin PBS with 2mM EDTA and 2% FCS (cohort#1), or weresubjected to a mechanical dissociation in petri dishes using25-gauge needles (cohort#2), or with the Medimachine (BectonDickinson, San Jose, CA) (cohort#3).C. Flow cytometryCells were preincubated with PBS + 2% human AB serum toblock non-specific binding to Fc receptors, then 5.10 5 cells perwell were stained with ECD anti-CD3 (UCHT1), PE-cyanin5anti-CD4 (13B8.2), PE-cyanin7 anti-CD8 (SFCI21Thy2D3)from Beckman Coulter, or with PE-cyanin5 conjugatedanti-CD3 (UCHT1, Dako, Trappes, France) and/or with thefollowing BD-Biosciences (Le Pont-de-Claix, France) mAb:Alexa fluor 700 anti-CD3 (UCHT1), PE anti-CD4 (SK3),pacific blue anti-CD4 (SK3), FITC anti-CD8 (SK1),APC-cyanin7 anti-CD8 (SK1) or the corresponding isotypicmAb controls. Data were acquired using a nine-colours LSRII(BD-Biosciences) or a four-colors FACS calibur(BD-Biosciences) or a five-colors FC500 (Beckman Coulter),and analyzed with the Diva (BD-Biosciences), Cellquest(BD-Biosciences), or CXP softwares (Beckman Coulter) forcohorts #1, #2 and #3, respectively.D. ImmunohistochemistryNumber(% of total)AgeMin-max (median)SexM/FNormal 11 (9.7%) 55-81 (72) 9/2Reactive 7 (6.2%) 25-70 (32) 3/4HL :non mediastinal localizations7 (6.2%) 21-80 (35) 3/4LL 4 (3.5%) 64-86 (75) 4/0FL 9 (8%) 27-83 (56) 5/4DLBCL : mediastinal 2 (1.8%) 25-55 (40) 2/0Other localizations 5 (4.4%) 53-86 (73) 3/2DLBCL : mediastinal 10 (9.9%) 18-68 (33) 5/5Other localizations 4 (3.5%) 40-79 (66.5) 4/0HL: mediastinal 7 (6.2%) 25-50 (32) 4/3Other localizations 1 (0.9%) 42 (42) 1/0DLBCL : mediastinal 4 (3.5%) 20-25 (23.5) 1/3Other localizations 26 (23%) 37-96 (77.5) 15/11HL: mediastinal 1 (0.9%) 36 (36) 1/0Other localizations 15 (13.3%) 15-80 (32) 6/9HL, hodgkin lymphoma ; LL, lymphocytic lymphoma ; FL, follicularlymphoma ; DLBCL, diffuse large B-cell lymphoma ; M, male ; F, female.Immunohistochemistry staining was performed on paraffinembedded sections with the following mAb specific to CD3,(clone SP7, Microm Neomarker, Francheville, France), CD8(114B, Dako), TiA1 (TiA1, Beckman Coulter, Villepinte,France), Granzyme B (GranzB, Tebu, Le Perray-en-Yvelines,France) for the T-cell subtypes. Briefly, immunostaining onparaffin section was performed after antigen retrieval,avidin/biotin and Fc receptors blocking treatment. Staining wasrevealed with streptavidin-peroxidase and DAB.E. Statistical analysisMann-Whitney tests were performed using StatView (SASInstitute, Cary, NC). P values less than 0.05 were consideredstatistically significant.III. RESULTS AND DISCUSSIONLN from 7 HL and 21 B-NHL subtypes were analyzed byflow cytometry for the presence and the relative distribution ofCD4 + and CD8 + T-cell subsets. These lymphomas werecompared to 11 normal and 7 reactive LN (Table 1, cohort #1).Non tumoral LN as well as HL, LL and FL samples showed amedian CD4:CD8 ratio > 2 (Fig. 1). Surprisingly, an invertedmedian CD4:CD8 T-cell ratio was specifically observed inDLBCL (0.76±0.33; DLBCL vs normal LN: P

INTERNATIONAL TRENDS IN IMMUNITY VOL.1 NO.2 APRIL 2013ISSN 2326-3121 (Print) ISSN 2326-313X (Online) http://www.researchpub.org/journal/iti/iti.htmlshown in Fig. 1 (open and light-blue diamonds, cohorts #2 and#3 respectively) confirmed the low median CD4:CD8 ratio(0.79±0.57 and 1.22±1.27, respectively) without statisticaldifference between the three cohorts (P=0.37). Moreover, intwo DLBCL cases, tumoral LN from cohorts #1 and #2 weretreated simultaneously but independently with differenttechniques and the respective CD4:CD8 T-cell ratios werequite similar (data not shown). Interestingly, a tumoral LN andan adjacent non tumoral LN (confirmed by pathologicalexamination) from one patient were analyzed simultaneouslyand different CD4:CD8 ratios were found (Fig. 1, arrows). Inthis case, the CD4:CD8 ratio from the LN without presence oftumor cells was in the range of the set of reactive LN whereasthe tumoral LN displayed the lowest CD4:CD8 ratio (0.11).Therefore, the low CD4:CD8 T-cell ratio for DLBCL samplesseems to be restricted to the tumoral LN and may be induced bythe tumour cells. Moreover, the CD4:CD8 ratio is not linked tothe T-cell number variation (data not shown).In an extensive study on T cell lymphoma, Gorczyca et coll.noticed a low CD4:CD8 ratio in 10 DLBCL cases [5]. Anotherstudy showed that DLBCL patients displaying the lowestCD4:CD8 ratios showed a worse overall survival,independently from the clinical stage at diagnosis [9]. However,no extensive comparison between normal and tumorallymphoid tissues, nor between LN from HL and non-Hodgkinlymphoma subtypes was performed in these studies. In thepresent work we analyzed 67 LN from three independentcohorts classified in 6 entities and we demonstrate that DLBCLLN exhibit an inverted CD4:CD8 ratio before any treatment.To investigate the influence of the mediastinal (ie. thymic)localisation on the CD4:CD8 ratio, we compared DLBCL andHL which are the most common lymphomas of themediastinum. Identification of these tumors (DLBCL; HL,respectively) was assessed by immunohistochemistry for theexpression of the following markers: CD20 + (12/12; 0/7),CD79a + (12/12; 0/7), CD23 + (10/12; 1/7), CD30 + (10/12; 7/7),CD15 + (0/12; 5/7), and LMP-1 + (0/12; 1/7). The medianCD4:CD8 ratios are below 1 in primary mediastinal (0.69±1.15)and non mediastinal (0.91±1.03) tumoral LN from DLBCLpatients, and are not significantly different (Fig. 2; P=0.10).Contrasting with DLBCL, the median CD4:CD8 ratio for HLwas over 3, independently from the mediastinal and nonmediastinal localisations (3.93±4.29 and 5.61±10.98,respectively; P=0.73). Altogether these results suggest that theCD4:CD8 ratio seems to depend more on the lymphomasubtype than on tumoral LN localization. The CD4:CD8 ratioevaluation in LN could be an interesting additional tool for thedifferential diagnosis between some primary mediastinalDLBCL and mediastinal HL.The inverted median ratio in DLBCL could be explained byan increased number of CD8 + T-cells, a decreased number ofCD4 + T-cells, or both. Using immunohistochemistry, we founda high proportion of CD8 + cells compared to the CD3 stainingof T-cells in DLBCL (Fig. 3). In contiguous slides, these cellsexpressed cytotoxic proteins like TiA1 and Granzyme B.Moreover, the DLBCL T-cell proportion did not differ from HLand FL, nor reactive and normal LN (data not shown).DLBCLHLCD3CD8100nonmediastinalmediastinalnonmediastinalmediastinaliiitioral10e-cT8D:C4 1DCiiiTiA1ivGz-B0.1***********Fig. 2. Analysis of the CD4:CD8 T-cell ratio for different subgroups ofDLBCL patients. CD4:CD8 T-cell ratios are represented on a logarithmicordinate in order to respect the proportionality of distribution between valuesover or below 1. The horizontal red bar corresponds to the median value.Statistical analysis was performed using the Mann-Whitney test.Significativity values are indicated as follow: *: P < 0.05 ; **: P < 0.01 ; ***:P < 0.001.47Fig. 3. Immunostaining of (i) CD3, (ii) CD8, (iii) TiA-1 and (iv) GranzymeB on a lymph node from a representative mediastinal DLBCL patient.Magnification: 200. Bars: 20µm.

INTERNATIONAL TRENDS IN IMMUNITY VOL.1 NO.2 APRIL 2013ISSN 2326-3121 (Print) ISSN 2326-313X (Online) http://www.researchpub.org/journal/iti/iti.htmlIV. CONCLUSIONAltogether, these data suggest an increased number ofcytotoxic T-lymphocytes, associated to a decreased number ofCD4 + T-cells. Since this inverted median ratio is characteristicof DLBCL, and differs significantly from low grade lymphomasubtypes, the T-cell subset modulations may be involved in thelymphomagenesis processes of DLBCL.ACKNOWLEDGMENTThe authors thank the facility core of cellular imaging (CICC,Centre de Recherche des Cordeliers, Paris F-75006, France),the cytometry facility from Rennes Hospital and Sylvie Baudetfor technical assistance. We are grateful to Jo Ann Cahn for hercareful reading of the manuscript.REFERENCES[1] S. S. Dave, G. Wright, B. Tan, A. Rosenwald, R. D. Gascoyne, W. C.Chan, R. I. Fisher, R. M. Braziel, L. M. Rimsza, T. M. Grogan, T. P.Miller, M. LeBlanc, T. C. Greiner, D. D. Weisenburger, J. C. Lynch, J.Vose, J. O. Armitage, E. B. Smeland, S. Kvaloy, H. Holte, J. Delabie, J. M.Connors, P. M. Lansdorp, Q. Ouyang, T. A. Lister, A. J. Davies, A. J.Norton, H. K. Muller-Hermelink, G. Ott, E. Campo, E. Montserrat, W. H.Wilson, E. S. Jaffe, R. Simon, L. Yang, J. Powell, H. Zhao, N.Goldschmidt, M. Chiorazzi, L. M. Staudt, “Prediction of survival infollicular lymphoma based on molecular features of tumor-infiltratingimmune cells,” N. Engl. J. Med., vol. 351, pp. 2159-2169, Nov. 2004.[2] F. Pagès, A. Berger, M. Camus, F. Sanchez-Cabo, A. Costes, R. Molidor,B. Mlecnik, A. Kirilovsky, M. Nilsson, D. Damotte, T. Meatchi, P.Bruneval, P. H. Cugnenc, Z. Trajanoski, W. H. Fridman, J. Galon,“Effector memory T cells, early metastasis, and survival in colorectalcancer.” N. Engl. J. Med., vol. 353, pp. 2654-2666, Dec. 2005.[3] T. Alvaro, M. Lejeune, M. T. Salvadó, C. Lopez, J. Jaén, R. Bosch, L. E.Pons, “Immunohistochemical patterns of reactive microenvironment areassociated with clinicobiologic behavior in follicular lymphomapatients.” J. Clin. Oncol., vol. 24, pp. 5350-5357, Dec. 2006.[4] D. M. Knowles, J. P. Halper, F. A. Jakobiec, “T-lymphocytesubpopulations in B-cell-derived non-Hodgkin's lymphomas andHodgkin's disease.” Cancer, vol. 54, pp. 644-651, Aug. 1984.[5] W. Gorczyca, J. Weisberger, Z. Liu, P. Tsang, M. Hossein, C. D. Wu, H.Dong, J. Y. Wong, S. Tugulea, S. Dee, M. R. Melamed, Z. Darzynkiewicz,“An approach to diagnosis of T-cell lymphoproliferative disorders byflow cytometry.” Cytometry, vol. 50, pp. 177-190, Jun. 2002.[6] O. Hernandez, T. Oweity, S. Ibrahim, “Is an increase in CD4/CD8 T-cellratio in lymph node fine needle aspiration helpful for diagnosing Hodgkinlymphoma? A study of 85 lymph node FNAs with increased CD4/CD8ratio.” Cytojournal, vol. 2, p. 14, Sep. 2005.[7] S. D. Hudnall, E. Betancourt, E. Barnhart, J. Patel, “Comparative flowimmunophenotypic features of the inflammatory infiltrates of Hodgkinlymphoma and lymphoid hyperplasia.” Cytometry B Clin. Cytom., vol. 74,pp. 1-8, Jan. 2008.[8] J. Muris, C. Meijer, S. Cillessen, W. Vos, J. Kummer, B. Bladergroen, M.Bogman, M. Mackenzie, N. Jiwa, L. Siegenbeek Van Heukelom,G.Ossenkoppele, J. Oudejans, “Prognostic significance of activatedcytotoxic T-lymphocytes in primary nodal diffuse large B-celllymphomas.” Leukemia, vol. 18, pp. 589-596, Mar. 2004.[9] Y. Xu, S. H. Kroft, R. W. McKenna, D. B. Aquino, “Prognosticsignificance of tumour-infiltrating T lymphocytes and T-cell subsets in denovo diffuse large B-cell lymphoma: a multiparameter flow cytometrystudy.” Br. J. Haematol., vol. 112, pp. 945-949, Mar. 2001.48