terminal deoxynucleotidyl transferase staining of malignant lymphomas in paraffin sections: a useful...

, . 182: 86–91 (1997)

TERMINAL DEOXYNUCLEOTIDYL TRANSFERASESTAINING OF MALIGNANT LYMPHOMAS IN

PARAFFIN SECTIONS: A USEFUL METHOD FOR THEDIAGNOSIS OF LYMPHOBLASTIC LYMPHOMA

1,2*, 1, 1, 1, 1 1

1First Department of Pathology, Fukuoka University, School of Medicine, 7-45-1 Nanakuma, Jonan-ku, Fukuoka 814-01, Japan2First Department of Internal Medicine, Fukuoka University, School of Medicine, 7-45-1 Nanakuma, Jonan-ku, Fukuoka 814-01,

Japan

SUMMARY

Terminal deoxynucleotidyl transferase (TdT) is a DNA polymerase located in the cell nucleus which catalyses the polymerization ofdeoxynucleotides at the 3* hydroxyl ends of oligo- or polydeoxynucleotide initiators without a template. TdT is known as a useful markerfor the diagnosis of acute lymphoblastic leukaemia/lymphoma, but its detection usually requires fresh tissue specimens or cellsuspensions, using either an enzyme analysis or immuno-fluorescence or -peroxidase staining. Until the recent development of the use ofmicrowave-treated paraffin sections for immunoperoxidase staining, detection of TdT in paraffin sections required rather complicatedprocesses. This new simple technique was applied to paraffin sections from the tumour tissue specimens of 16 patients with lymphoblasticlymphoma and of seven patients with non-endemic Burkitt’s lymphoma, which is sometimes difficult to differentiate from lymphoblasticlymphoma because of their similar clinicopathological characteristics. In addition, as a control, ten cases each were examined of adultT-cell leukaemia/lymphoma (ATLL) and angioimmunoblastic lymphoma (AILD), which are both peripheral T-cell lymphomas. Thetumour cells from 15 of the 16 (94 per cent) patients with lymphoblastic lymphoma were found to be TdT-positive. The specificity of theanti-TdT antibody used was confirmed by immunoblot and the specific 60 kD band was detected only in a specimen of lymphoblasticlymphoma. These results show that the immunostaining of TdT on paraffin-embedded sections is a useful method for differentiatinglymphoblastic lymphoma from other lymphomas. This method is applicable to a routine diagnostic service. ? 1997 by John Wiley &Sons, Ltd.

J. Pathol. 182: 86–91, 1997.No. of Figures 2. No of Tables 2. No. of References 29.

KEY WORDS—malignant lymphoma; TdT; lymphoblastic lymphoma; immunostaining; paraffin sections

INTRODUCTION

Terminal deoxynucleotidyl transferase (TdT) is usu-ally expressed in primitive T- and B-lymphocytes ofthe normal thymus and bone marrow.1–5 TdT hasbeen demonstrated in most cases of acute lympho-blastic leukaemia, regardless of their immunologicalphenotype.6–12 TdT positivity has also been seen in alymphoblastic crisis of chronic myelogenous leu-kaemia,3,5,9,13 though it has rarely been detected ineither acute myelogenous or acute myelomonocyticleukaemia.2,5,8–14 Mature T- and B-cell leukaemias arealmost invariably TdT-negative. In cases of malignantlymphoma, the expression of TdT was consistentlyobserved in lymphoblastic lymphoma3,13,15–17 but rarelyin other subtypes of malignant lymphoma.15,16 Thedetection of TdT is therefore considered to be usefulin the diagnosis and classification of malignantlymphomas.9,13,15,16Unfortunately, the methods for detection of TdT

usually require fresh or frozen specimens, because of

antigenic deterioration of TdT protein in conventionallyprocessed tissues. Although immunoperoxidase stainingof TdT in paraffin sections has been repeatedlyreported,4,5,17 the method is rather complicated, includ-ing pretreatment with DNase I with/without trypsiniz-ation, repeated immunostaining cycles, and post-fixationwith sodium cacodylate and osmium tetroxide.Recently, however, Orazi et al.18 reported excellentresults with a simpler method for detecting TdT usingroutine paraffin-embedded sections. The purpose of thepresent study was to confirm the usefulness of thisTdT-detection method in the diagnosis of lymphoblasticlymphoma.

MATERIALS AND METHODS

Cells and tissuesA diagnosis of lymphoblastic lymphoma was made

on the basis of the following histological characteristics:(1) monotonic growth pattern of lymphoma cells (2)frequent occurrence of a ‘starry sky’ pattern; (3) lym-phoma cells having small- to medium-sized nuclei withfinely dispersed nuclear chromatin; (4) small, pinpoint or

*Correspondence to: Dr J. Suzumiya, MD, PhD.

Contract grant sponsors: Fukuoka Cancer Society; Ministry ofEducation, Science and Culture, Japan.

CCC 0022–3417/97/050086–06 $17.50 Received 31 January 1996? 1997 by John Wiley & Sons, Ltd. Revised 28 September 1996

Accepted 10 October 1996

invisible nucleoli; (5) predominantly convoluted, butoccasionally non-convoluted nuclei; (6) many mitoticfigures; and (7) the finding of blasts in touch smearspecimens; using these criteria, we selected 16 cases oflymphoblastic lymphoma. As a control group for B-cellswe selected seven cases of non-endemic Burkitt’s lym-phoma and for peripheral T-cells we selected ten cases ofadult T-cell leukaemia/lymphoma (ATLL), ten cases ofangioimmunoblastic lymphoma (AILD), and ten casesof non-specific lymphadenitis. All cases were diagnosedin our department, according to the updated Kiel classi-fication.19 All specimens obtained either from biopsiesor at surgery were fixed with 10 per cent formalin andembedded in paraffin. Histological sections were stainedwith haematoxylin and eosin (H & E) and Giemsa.Imprint smears were prepared and stained with Giemsain all cases of lymphoblastic lymphoma.

Immunohistochemical staining for TdT

TdT was detected using an affino-purified polyclonalrabbit anti-TdT antibody (Supertec, Bethesda, MD,U.S.A.) by a streptavidin–biotin complex alkaline phos-phatase immunohistochemical method (Vector Lab.,Burlingame, CA, U.S.A.). The staining methods werethose described by Orazi et al.,18 with slight modifi-cations. In brief, the slides were deparaffinized, rinsedtwice with phosphate-buffered saline (PBS), transferredto jars filled with citrate buffer (pH 6·0), and heated in amicrowave oven (H2500 Microwave Processor: EnergyBeam Science, Inc., Agawam, MA, U.S.A.) for 10 minat a power of 700 W. After cooling at room temperaturefor 15 min, the slides were covered with normal goatserum for 20 min, incubated with the anti-TdT antibodyat a dilution of 1:10 at room temperature for 1 hr, andthen stained with a biotin-conjugated goat anti-rabbitantibody for 30 min followed by alkaline phosphatase-conjugated streptavidin for 30 min. New fuchsin andnaphthol AS-B1 phosphate were used as substratechromogen reagents. Cases were recorded as positivefor TdT when the antigen was expressed by at leat10 per cent of the malignant cells and the positivereaction was found in the nuclei. Labelling indices weredetermined by counting 500 tumour cells in each case.

Immunohistochemical staining for the determination ofcell lineage

A portion of the specimen obtained from the lymphnode or tumour tissue was snap-frozen in Tissue TekOCT compound (Miles, Elkhardt, IN, U.S.A.). Fromthe frozen tissue, 5 ìm sections were cut, acetone-fixed,and stained with a panel of monoclonal antibodies. Theantibodies used were specific for CD2 [T11, CoulterClone (CC), Hialeah, FL, U.S.A.], CD3 [Leu 4; Becton-Dickinson (BD), San Jose, CA, U.S.A.], CD4 (OKT4;Ortho, Raritan, NJ, U.S.A.), CD8 (OKT8; Ortho),CD10 (CALLA; BD), CD19 (B4; CC), CD20 (B1; CC),CD56 (Leu19; BD), TCRâ (âF1: T Cell Diagnostics,Cambridge, MA, U.S.A.), and TCRä (ä1; T Cell Diag-nostics). A streptavidin–biotin complex alkaline phos-phatase immunohistochemical method was performed,

according to the manufacturer’s instructions (VectorLab.).

Immunoblot analysis

Pieces of tumour tissue and cell pellets were hom-ogenized in the extraction medium composed of50 m Tris–HCl/150 m NaCl/2 m EDTA, pH 7·5,containing 2 m PMSF, 1 m leupeptin, 2 mN-ethylmaleimide, 2 per cent NP-40, and 2 per centTriton X-100; centrifuged at 10 000 g at 4)C; and thenthe supernatants were collected. Samples of 150 ìg pro-teins were run on SDS-PAGE under reducing conditionsusing a 10 per cent separating gel. After electrophoresis,the proteins were transferred electrophoretically ontoan Immobilon membrane (Millipore, Bedford, MA,U.S.A.). After transfer, the membrane was blocked with5 per cent non-fat milk protein in 50 m Tris–HCl/200 m NaCl/0·05 per cent Tween-20, pH 7·6 (TBS-T),for 3 h at 37)C and then incubated with 1:200 dilutedrabbit polyclonal TdT antibody in TBS-T supplementedwith 1 per cent BSA overnight at 4)C. After three washesin TBS-T, the membrane was incubated in 1:5000peroxidase-conjugated goat anti-rabbit IgG in TBS-Tcontaining 1 per cent BSA for 1 h at room temperature.A chemiluminescence reagent (DuPont NEN, Boston,MA, U.S.A.) was used to visualize the labelled proteinbands according to the manufacturer’s instructions.

Southern blotting

Southern blotting was performed to determine the celllineage and to examine the HTLV-1 proviral inte-gration. Part of the frozen material was used for DNAisolation and gene analysis. Details of the examinationmethod and gene analysis have been reported pre-viously.23 The T-cell receptor gene Câ, Jã and theimmunoglobulin heavy chain (JH) gene, and proviralDNA of human T-cell lymphotropic virus type I, fulllength (HTLV-I) were examined by Southern blotanalysis.23–26 In addition, the DNA was digested withrestriction enzymes EcoRI, HindIII or BamHI.

RESULTS

Expression of TdT in various lymphomas was exam-ined by an immunoblot analysis. The polyclonal rabbitanti-TdT antibody used in this study gave a major bandwith a molecular weight of 60 kD, compatible with thatof TdT, in a specimen of lymphoblastic lymphoma. Incontrast, no such bands were recognized in the samplesof non-endemic Burkitt’s lymphoma, AILD or ATLL(Fig. 1).The results of TdT staining in paraffin sections of

various lymphomas are summarized in Table I. TdT waspositive in 15 of 16 (94 per cent) cases with lympho-blastic lymphoma. Although one negative case wasexamined further using a higher concentration of theanti-TdT antibody (1:5 dilution), it remained negativefor TdT. In contrast to lymphoblastic lymphoma, thespecimens from non-endemic Burkitt’s lymphoma,

87TDT IN MALIGNANT LYMPHOMAS

? 1997 by John Wiley & Sons, Ltd. J. Pathol. 182: 86–91 (1997)

ATL, AILD, and non-specific lymphadenitis wereentirely negative for TdT staining. Positive TdT stainingin lymphoblastic lymphoma and negative results in otherlymphomas in paraffin sections are shown in Fig. 2,together with their H & E appearances. The localizationof TdT was essentially restricted to the nuclei of thelymphoblastic lymphoma cells. Although the positivestaining for TdT was sometimes encountered in thecytoplasm of non-endemic Burkitt’s lymphoma cells,this seemed to be a non-specific binding because thespecific band for TdT was not detected in immunoblotanalysis. Cytoplasmic staining for TdT was also occa-sionally observed in mitotic lymphoblastic lymphomacells, as described in a previous report.18Table II summarizes the results of immunostaining

and Southern blot analyses of the immunoassociatedgenes in lymphoblastic lymphomas. Of 16 cases, 14 had

a T-cell phenotype and only one was of B-cell pheno-type. One case reported elsewhere27 was considered to beof natural killer (NK) cell type, because the tumour cellswere CD56-positive, cytoplasmic CD3-positive, havingazurophilic granules and no rearrangement of TCRs.Together with the results shown in Table I, TdT waspositive in 13/14 (93 per cent) cases of T-lymphoblasticlymphoma and in 1/1 case of B-lymphoblastic lym-phoma. One lymphoblastic lymphoma expressing NKcell marker was positive for TdT. The tumour cells ofpatient 8, who had a TdT-negative T-lymphoblasticlymphoma, showed nuclear convolution and expressedCD2, 3, 7 and a rearrangement of TCR Câ2. Thus,no difference was observed between the phenotypesof TdT-positive and TdT-negative T-lymphoblasticlymphomas.Monoclonal integration of HTLV-1 was present in all

cases of ATLL, but none of the other groups showedmonoclonal integration of HTLV-1.

DISCUSSION

Lymphoblastic lymphoma is a high-grade malignantlymphoma20–22,28 and a proper intense and continuousregimen of chemotherapy is necessary. Patients withlymphoblastic lymphoma are, at present, treated withthe same protocols of either chemotherapy or bonemarrow transplantation as patients with acute lympho-blastic leukaemia:29 the early and precise diagnosisof lymphoblastic lymphoma is therefore crucial.T-lymphoblastic lymphoma accounts for about 3–4 percent of all cases of non-Hodgkin’s lymphoma and occursmainly in the young.20,21 Although lymphoblastic lym-phoma is histologically distinguishable from Burkitt’slymphoma, they share some histological characteristics,such as monotonic proliferation of lymphoma cellswith medium-sized and hyperchromatic nuclei and astarry-sky appearance.Some cases of lymphoblastic lymphoma are indeed

difficult to differentiate from Burkitt’s lymphoma.13Furthermore, variant lymphoblastic lymphoma (L2variant) cells also have relatively large nuclei and promi-nent nucleoli and the diagnosis of this type is consideredto be difficult. In this respect, staining for TdT hasproved useful in the diagnosis and classification ofmalignant lymphoma. Unfortunately, however, themethods previously used to detect TdT have beenlimited to fresh tissues and require complicated pre-treatment.9,13,15,16 In contrast, the simpler method estab-lished by Orazi et al.18 and used in this study enabled usto detect TdT successfully in paraffin sections. In thepresent study, the specificity of anti-TdT antibody wasconfirmed by immunoblotting. Furthermore, controlcases of non-specific lymphadenitis, ATLL, AILD ornon-endemic Burkitt’s lymphoma were entirely negativefor TdT. In addition, except for mitotic cells, positivestaining was restricted to the nuclei of lymphoblasticlymphoma cells. Thus, even if fresh tissue or smearsamples were not available, TdT staining on paraffinsections may be included in the diagnostic methods formalignant lymphomas.

Table I—TdT in lymphoblastic lymphoma and other lym-phomas

Cases Positivity (%)

Lymphoblastic 15/16 94T-cell type 13/14 93NK-cell type 1/1 100B-cell type 1/1 100

Burkitt’s 0/7 0AILD 0/10 0ATLL 0/10 0Lymphadenitis 0/10 0

Fig. 1—Western blot analysis for TdT. LBL=lymphoblastic lym-phoma; Burkitt’s=Burkitt’s lymphoma; AILD=angioimmunoblasticlymphoma; ATLL=adult T-cell leukaemia/lymphoma. Major band of60 kD for TdT is observed only in LBL (arrow-head)

88 J. SUZUMIYA ET AL.


Table II—Immunophenotype and immunogenotype of lymphoblastic lymphoma

Patients 1 2 3 4 5 6 7 8

Age, Sex 41, M 9, M 22, M 26, M 3, F 15, M 41, M 10, FSite LN LN LN LN Mediast. T LN LN LNNuclear convolution " " + + + + + +

TdT + + + + + + + "Positive cells >80% >80% 20–50% 50–80% >80% 20–50% 20–50% 0%

T-cell markersCD2 (T11) " " " " ++ & & +CD3 (Leu4) + + + ++ ++ ++ ++ +CD4 (OKT4) " " " " " " " &CD8 (OKT8) & " " " + " " "TCR-â " " " " ++ ND " +TCR-ä " " & " " ND " "

B-cell markerCD19 (B4) " " " " " " " "CD20 (B1) " " ND " " " " "

Other markersCD10 (CALLA) " ++ " " ++ " " "CD56 (Leu19) " " " " " ND " &

Southern blotCâ1/Câ2/Jã/JH ND G/G/G/G G/G/R/G G/R/G/G R/R/R/G G/G/d/G G/G/R/G G/r/G/G

Phenotype T T T T T T T T

Patients 9 10 11 12 13 14 15 16

Age, sex 7, M 9, M 9, M 30, M 46, M 32, M 3, M 19, MSite LN Mediast. T LN LN Nose LN LN LNNuclear convolution + + + " " + + +

TdT + + + + + + + +Positive cells 20–50% >80% >80% >80% >80% >80% >80% >80%

T-cell markersCD2 (T11) + ++ " + " " + +CD3 (Leu4) + ++ " + + ++ + +CD4 (OKT4) + " " " " " " "CD8 (OKT8) + " " " " " + "TCR-â + " " ND " " + "TCR-ä " " " ND " " " "

B-cell markersCD19 (B4) " " + " " " " "CD20 (B1) " " " " " " " "

Other markersCD10 (CALLA) ++ " + " " & + "CD56 (Leu19) " " " " + " " "

Southern blotCâ1/Câ2/Jã/JH ND ND G/G/G/R ND G/G/G/G G/G/G/G D/R/r/G ND

Phenotype T T B T NK T T T

++=Strongly positive; +=positive; &=weakly positive; "=negative; ND=not done. Mediast. T=mediastinal tumour.Câ1=TCRCâ1; Câ2=TCR; Jã=TCRJã; JH=immunoglobulin JH; G=germ line; R, r= rearranged; D, d=deleted.



In the present study, among 16 lymphoblastic lym-phomas selected by histological characteristics, one caseof T-cell phenotype was negative for TdT. Similarresults have been reported by others.18 Whether suchcases truly belong to lymphoblastic lymphoma should

now be examined in a larger group of TdT-negativecases, with reference to possible clinical, biological, andgenetic differences.The present results confirm previous obser-

vations4,5,17,18 that a majority of lymphoblastic

Fig. 2—Histology with H & E stain (left column) and immunohistochemical stain with TdT antibody (right column).(a, b) Lymphoblastic lymphoma; (c, d) Burkitt’s lymphoma; (e, f) AILD; (g, h) ATLL. Note: positive reactions onlyin the nuclei of lymphoblastic lymphoma cells

90 J. SUZUMIYA ET AL.


lymphomas exhibit a T-cell phenotype and are TdT-positive. In addition, we found that TdT was notexpressed in peripheral T-cell lymphomas such as ATLLor AILD. In non-specific lymphadenitis, TdT waspresent in only a few lymphocytes. To our knowledge,expression of TdT in peripheral T-cell lymphomas hasnever been explored in detail and there has been onlyone case of ATLL17 positive for TdT. Taken together,expression of TdT is restricted to cases with lympho-blastic morphology. Thus, staining for TdT in paraffinsections is considered to be a useful method for thedelineation of T-cell lymphomas.Since TdT expression is also present in some cases of

acute myelogenous leukaemia,8,11,12,14 some caution isnecessary when we encounter TdT-positive tumour cells.In practice, there are some non-lymphoid tumourswhich may mimic the histological patterns of lympho-blastic lymphoma. One such tumour is a neuroblastomain which the cells possess blastic features. When weexamined five patients with this tumour by TdT stainingin paraffin sections, they were entirely negative for TdT(data not shown), again indicating the usefulness of TdTstaining in the differential diagnosis of lymphoblasticlymphomas.In conclusion, using paraffin sections, we could suc-

cessfully demonstrate that TdT reactivity was restrictedto lymphomas of lymphoblastic type, while other dif-fuse, aggressive non-Hodgkin’s lymphomas were TdT-negative. Staining for TdT is thus considered useful forthe diagnosis and classification of human lymphoidneoplasms.

ACKNOWLEDGEMENTS

We would like to thank Professor Y. Nawa, MiyazakiMedical College, for useful discussions; Mr H.Nishimura for photographic assistance; Misses Y.Hirose, M. Yuuki, K. Saga, and K. Takasu for immuno-histochemical staining; and Ms K. Miyazaki for assist-ance in the Southern blot analysis. This work wassupported in part by a Grant-in-Aid from the FukuokaCancer Society and a Grant-in-Aid for ScientificResearch (C) from the Ministry of Education, Scienceand Culture, Japan.

REFERENCES1. Chang LMS. Development of terminal deoxynucleotidyl transferase activity

in embryonic calf thymus gland. Biochem Biophys Res Commun 1971; 44:124–131.

2. Coleman MS, Hutton JJ, de Simone P, Bollum FJ. Terminal deoxyribo-nucleotidyl transferase in human leukemia. Proc Natl Acad Sci USA 1974;71: 4404–4408.

3. Bollum FJ. Terminal deoxynucleotidyl transferase as a hematopoietic cellmarker. Blood 1979; 54: 1203–1215.

4. Janossy G, Thomas JA, Pizzolo G, et al. Immunohistological diagnosis oflymphoproliferative diseases by selected combinations of antisera andmonoclonal antibodies. Br J Cancer 1980; 42: 224–242.

5. Halverson CA, Falini B, Taylor CR, Parker JW. Detection of terminaltransferase in paraffin sections with the immunoperoxidase technique. Am JPathol 1981; 105: 241–254.

6. Penit C, Brouet J-C, Rouget P. Terminal deoxynucleotidyl transferase inacute lymphoblastic leukemias and chronic T cell proliferations. LeukemiaRes 1977; 1: 345–350.

7. Shaw MT, Dwyer JM, Allaudeen HS, Weitzman HA. Terminal deoxyribo-nucleotidyl transferase activity in B-cell acute lymphocytic leukemia. Blood1978; 51: 181–187.

8. Janossy G, Hoffbrand AV, Greaves MF, et al. Terminal transferase enzymeassay and immunological membrane markers in the diagnosis of leukemia:a multiparameter analysis of 300 cases. Br J Haematol 1980; 44: 221–234.

9. Bearman RM, Winberg CD, Maslow WC, et al. Terminal deoxynucleotidyltransferase activity in neoplastic and nonneoplastic hematopoietic cells. AmJ Clin Pathol 1981; 75: 794–802.

10. Peiper SC, Stass S. Markers of cellular differentiation in acute lympho-blastic leukemia. Arch Pathol Lab Med 1982; 106: 3–8.

11. Lanham GR, Melvin SL, Stass SA. Immunoperoxidase determination ofterminal deoxynucleotidyl transferase in acute leukemia using PAP andABC methods: experience in 102 cases. Am J Clin Pathol 1985; 83: 366–370.

12. Erber WN, Mason DY. Immunoalkaline phosphatase labeling of terminaltransferase in hematologic samples. Am J Clin Pathol 1987; 88: 43–50.

13. Kung PC, Long JC, McCaffrey RP, Ratliff RL, Harrison TA, Baltimore D.Terminal deoxynucleotidyl transferase in the diagnosis of leukemia andmalignant lymphoma. Am J Med 1978; 64: 788–794.

14. McCurley TL, Greer JP, Glick AD. Terminal deoxynucleotidyl transferase(TdT) in acute nonlymphocytic leukemia: a clinical morphological, cyto-chemical, immunologic, and ultrastructural study. Am J Clin Pathol 1988;90: 421–430.

15. Donlon JA, Jaffe ES, Braylan RC. Terminal deoxynucleotidyl transferaseactivity in malignant lymphomas. N Engl J Med 1977; 297: 461–464.

16. Braziel RM, Keneklis T, Donlon JA, et al. Terminal deoxynucleotidyltransferase in non-Hodgkin’s lymphoma. Am J Clin Pathol 1983; 80:655–659.

17. Said JW, Shintaku IP, Pinkus GS. Immunohistochemical staining forterminal deoxynucleotidyl transferase (TdT). Am J Clin Pathol 1988; 89:649–652.

18. Orazi A, Cattoretti G, John K, Neiman RS. Terminal deoxynucleotidyltransferase staining of malignant lymphomas in paraffin sections. ModPathol 1994; 7: 582–586.

19. Stansfeldt AG, Diebold J, Noel H, et al. Updated Kiel classification forlymphomas. Lancet 1988; 8580: 292–293.

20. The non-Hodgkin’s lymphoma pathologic classification project. NationalCancer Institute sponsored study of classifications of non-Hodgkin’s lym-phomas: summary and description of a working formulation for clinicalusage. Cancer 1982; 49: 2112–2135.

21. Lennert K, Feller AC. Histopathology of Non-Hodgkin’s Lymphomas:based on the Updated Kiel Classification. Berlin: Springer-Verlag, 1992;244–250.

22. Medeiros LJ. In: Jaffe ES, ed. Surgical Pathology of the Lymph Nodes andRelated Organs. Philadelphia: W. B. Saunders, 1994; 306–312.

23. Yoshida T, Kikuchi M, Ooshima K, et al. Anti-human T-cell lympho-trophic virus type I antibody-positive adult T-cell leukemia/lymphoma withno monoclonal proviral DNA. Cancer 1989; 64: 2515–2524.

24. Kimura N, Takihara Y, Akiyoshi T, et al. Rearrangement of T-cell receptorä chain gene as a marker of lineage and clonality in T-cell lymphoprolifer-ative disorders. Cancer Res 1989; 49: 4488–4492.

25. Ohshima K, Kikuchi M, Eguchi F, et al. Analysis of Epstein–Barr viralgenomes in lymphoid malignancy using Southern blotting, polymerasechain reaction and in situ hybridization. Virchows Arch B [Cell Pathol]1990; 59: 383–390.

26. Ohshima K, Kikuchi M, Masuda Y, et al. Defective provirus form ofhuman T-cell leukemia virus type I in adult T-cell leukemia/lymphoma:clinicopathological features. Cancer Res 1991; 51: 4639–4642.

27. Koita H, Suzumiya J, Takeshita M, et al. Lymphoblastic lymphomaexpressing natural killer cell phenotype with involvement of the mediasti-num and nasal cavity; a case report. Am J Surg Pathol 1997; 21: 242–248.

28. Nathwani BN, Diamond LW, Winberg CD, et al. Lymphoblasticlymphoma: a clinicopathologic study of 95 patients. Cancer 1981; 48:2347–2357.

29. Morel P, Lepage E, Brice P, et al. Prognosis and treatment of lymphoblasticlymphoma in adults: a report on 80 patients. J Clin Oncol 1992; 10:1078–1085.



terminal deoxynucleotidyl transferase staining of malignant lymphomas in paraffin sections: a useful...

Documents