Terminal Deoxynucleotidyl Transferase Expression in Acute Non-lymphoid Leukaemia: an Analysis by Immunofluorescence

BritishJoiirtial qfHarmatology, 1981, 47, 133-143.

Terminal Deoxynucleotidyl Transferase Expression in Acute Non-lymphoid Leukaemia: an Analysis by Immunofluorescence K. F. BRADSTOCK, A. V. HOFFBRAND, K. GANESHAGURU, P. LLEWELLIN, K. PAITERSON,* B. WONKE,A. G. PRENTICE,t M. BENNEXT,$G. PIZZOLO, F. J. B O L L U M AND ~ G. JANOSSY

Royal Free Hospital, London; *University College Hospital, London; tWestern Inzrmary, Glasgoui; $North Middlesex Hospital, London; §Department ofBiochemistry, Uniformed Services University of the Health Sciences, Bethesda, U.S.A. (Received 2 4 March 1980; acceptedfor publication 21/uly 1980)

SUMMARY. Indirect immunofluorescence for terminal transferase enzyme (TdT) was used to study the blasts of 64 patients with acute non-lymphoid leukaemia (ANLL). In 32 patients no TdT positive cells were seen. In 19 cases a small subpopulation of cells expressing TdT was detected; these constituted up to 5% oftotal nucleated cells, and it was not clear whether these TdT positive cells were part of the leukaemic process or represented residual normal bone marrow lymphoid cells. The remaining 13 patients had TdT positive cells accounting for 7-90% of the total. In two of these cases TdT was expressed on blasts with myeloid features, representing an aberrant expression of TdT by myeloid cells; in contrast, in three cases mixed populations of TdT positive lymphoid blasts and TdT negative myeloid blasts were observed. I n the remaining cases it was not possible to determine whether the TdT positive cells had definite lymphoid or myeloid features. Cytogenetic analysis showed no evidence of the Philadelphia chromosome. Response to treatment was assessed in 11 of the 13 patients. Only one patient remitted with the initial choice of therapy (DAT); four failed to respond to initial regimes of vincristine and prednisone (V &- P) while the other five patients did not respond to myelotoxic combinations (DAT). Only one patient subsequently entered complete remission on second line therapy (V & P). This group of patients with TdT+ ANLL had a particularly bad prognosis, and appeared to differ from cases of TdT positive acute undifferentiated leukaemia, which often respond to V & P therapy. Terminal deoxynucleotidyl transferase (TdT) is an unusual D N A polymerase found in normal thymus and a small subpopulation of normal bone marrow lymphoid cells (Bollum, 1979). Clinical interest in TdTwas first aroused by the discovery of high levels of the enzyme in leukaemic blasts from a case of childhood acute lymphoblastic leukaemia (ALL) (McCaffrey et Correspondence: Dr K. F. Bradstock, Department of Immunology, Royal Free Hospital, Hampstead, London

N W 3 2QG. 0007-1048/81/0100-0133$02.00

01981 Blackwell Scientific Publications 133

134

K . F. Brndstock et al

nl, 1973) This finding has been confirmed by numerous investigators, who have demonstrated

high levels of T d T enzyme activity in the majority of cases of non-T, non-B ALL (common ALL; c-ALL) and thymic or T-cell ALL (Thy-ALL), as well as some cases of acute undifferentiated leukaemia (AUL) (Hutton et nl, 1979; McCaffrey et nl, 1975; H o a r a n d et al, 1977). In contrast, other haematological malignancies are usually T d T negative, including most cases of acute non-lymphoid leukaemia (ANLL; consisting of acute myeloid leukaemia and its variants), Philadelphia chromosome (Ph’) positive chronic myeloid leukaemia (CML) in chronic phase and myeloid blast transformation, as Lvell as most chronic lymphoproliferative disorders such as chronic lymphocytic leukaemia, hairy cell leukaemia, and non-Hodgkin’s lymphomas (Hofirand et n l , 1977; Kung er a!, 1978). High level< of T d T are, however, occasionally found in myeloid leukaemias. Approximately 20-309” of cases of Ph’ positive CML in blast crisis have high T d T levels (Kung ef al, 1978); these cases, frequently of lymphoblastic morphology, express the surface membrane antigenic phenotype of Ph’ negative common-ALL (Janossy et al, 197%). Furthermore, isolated cases ofbona fide Ph’ negative ANLL have been reported to express high levels o f T d T (Srivastava er n l , 1976, 1978; Hoffbrand et al, 1977; Gordon et 01, 1978; Kung et nl, 1978; Hutton er a / , 1979); in a recent review of the large series of ANLL cases analysed for T d T activity, 6% of cases wcre T d T positive (Hoffbrand er al, 1979). Since these T d T determinations were done by bulk biochemical assay, it was not possible to conclusively establish whether T d T was expressed ‘aberrantly’ by myeloid cells, or Xvas present on a subpopulation of lymphoid cells. Although heterogeneous mixtures of lymphoid and myeloid blasts, as determined by morphology and immunological analysis, have been clearly demonstrated in CML blast crisis (Janossy ct a / , 1978), only one case report has suggested that similar mixtures may occur in Ph’ negative ANLL (Mertelsman er al, 1978). However, Lvith the availability of a highly specific antiserum to TdT, we have analysed a large number of cases of ANLL for T d T expression. From 64 cases studied, 13 with heterogeneous populations o f T d T positive and T d T negative blasts have been identified, and the clinical and laboratory features of these cases are presented here. METHODS Specimens were obtained from 64 patients Lvith untreated ANLL attending the Royal Free Hospital and University College Hospital, London, and a number of collaborating hospitals in the U.K. The diagnosis of ANLL w a s madc on the basis of morphological appearances on Romanowsky-stained bone marro\v and peripheral blood smears, and the cytochemical reactivity of the blast cells with Sudan black (and in some cases myeloperoxidase), periodic acid SchiR, chloroacetate esterase and alpha-naphthyl acetate esterase stains. For immunological and biochemical analysis, heparinized specimens were collectcd in Eaglc’s medium; bone marrow aspirate was analysed in 51 cases, while in the remaining 13 cases with high circulating blast counts peripheral blood was used. Mononuclear cells were separated on Ficoll-Isopaque, washed and assessed for viability. Cells were then analysed in suspension for expression of surface membrane leukaemia-associated antigens by indirect immunofluorescence, using previously described methods (Janossy et al, 1980);antisera used included anti-acute lymphoblastic leukaemia serum (Greaves et 01, 1975), anti-human p38,33 Ia-like serum (Schlossman et 01, 1976), and anti-human T lymphoid antiserum (Janossy et al,

TdT in Acute Non-lymphoid Leukaemia

335

1980). Cells were also examined by direct immunofluorescence for surface membrane immunoglobulin, using goat anti-human-Ig (FITC) (obtained from Behringwerke). For T d T staining, cells previously labelled in suspension with anti-la serum (labelled with tetraethyl-rhodamine isothyocyanate; TRITC), were pelleted onto glass slides in a cytocentrifuge. Slides were air dried, fixed in cold methanol (4"C, 15 min) and transferred to phosphate buffered saline (PBS). Rabbit anti-calf T d T antibody (Bollum, 1975) was added over the cell pellet a t 0.1 pg/p1, and incubated for 15 min a t 20°C in a humidified chamber. Slides were then washed in PBS for 30 min and F(ab')? fragment of goat-anti-rabbit IgG conjugated with fluorescein isothyocyanate (FITC) added for 15 min. After further washing in PBS, cells were sealed in formol-glycerol solution under a glass coverslip. The preparations were then examined using a standard Zeiss 18 microscope equipped with IV/F epifluorescence, selective excitation filters for FITC and TRITC, and a 63 Ph oil objective. For biochemical T d T analysis, 5 x lohcells were analysed per test, using previously described methods ( H o a r a n d et al, 1977), and activity expressed as units/lOx cells. Cytogenetic analysis was performed on unstimulated cells using standard methods. RESULTS Sixty-four patients with ANLL were studied; 27 were female and 37 male, with age range 5-89 years (mean 45 and 51 years respectively). In all cases the blasts were found to be ncgative for common-ALL antigen and T cell differentiation antigens, and failed to express monoclonal surface membrane immunoglobulins. Ia-like antigens were detected on a variable proportion of blasts in 43 of the 55 cases studied; the remaining 12 cases failed to express Ia. Analysis of thc leukaemic cells with anti-TdT antibody revealed considerable heterogeneity and the cases will be considered in three separate groups. 1. No T d T p o s i t i v e cells detected. In 32 cases no T d T positive cells were seen by immunofluorescence; more than 1000 cells were scanned in each case. Twenty-two cases were specimens of bone marrow, and 10 were from peripheral blood. Biochemical T d T levels were low in all cases (range 0-0.7 U/lOxcells; mean 0.15 U/lOx). 2. T d T p o s i t i v e cells detected in corzceritratiorir lip to 5% oftotal. Nineteen cases were included in this group; 17 were bone marrow specimens, and two were peripheral blood. In these cases a minor subpopulation of T d T positive cells was present; concentrations of positive cells ranged from 1 in 10'nucleated cells to 5%. In 17 cases, the proportion ofpositive cells was less than 1 %, and biochemical T d T values were low (range 0-0.8 U/lOxcells; mean 0.2 U/lOH).The other two cases had 2% and 5% positive cells, with biochemical T d T values of0.6 and 1.6 U/lOxcells respectively. In all cases phase contact appearance of the positive cells was extremely variable; some cells had the appearance of small lymphoid cells less than 10 p m diameter, while largcr cells up to 15 p m diameter with one or two nucleoli and more abundant cytoplasm lacking granules were also seen. 3. TdT positive cells detected in coriceritratioris greater thari 5% of total. Thirteen patients were included in this category; in three, peripheral blood cells were analysed (cases 1, 7 and 13), while in the other 10 cases bone marrow cells were used. Details of clinical and laboratory features, and response to treatment protocols, are given in Tables I and 11. A number ofpoints of special interest concerning these cases are, however, outlined below.

K.F. Bradstock et a1

-rc 2 +l

3 L

c

.rL

Y

r. c

si

Ki

57

18

25

45.6 (87)

2.2 (50)

5.6 (69)

34(91)

6.5

6.0

9.9

8.6

10

11

12

13

NAD

NAD

80

>90

NAD

NAD

NAD

>90

91

>95

Chroniosomes

>90% SB+ >90% MPX+ PAS45% MPX+ 22% NSE+7 PAS90% SB+ NSEPASMPX 12%+ PASNSE lo%+ 7%' SE 15%'

SBPASNSE n.t.

Cytochemical reartiorrs

S. lysozyme 15 &/ml (normal 2-5 pg/ml)

Aucr rods++

EM monoblasts' S. lysozyme 100 pg/ml (normal 5-15) urine lysozyme 184 pg/ml (normal A 1- x 1 N o rc\poiitc

V&P (2 weeks)

5

T R A M P C O x 2-treduced numbers o f circulating blasts; no remission D A T x 3-+partial remission

N o response No responge

3

V &. P (2 wccks)+no recponsc

Reported in detail elsewhere (Prentice ef a / , 1980).

t M T X : mcthotrcxate; L-asp: L-asperaginase; ara-C: cytosine arabinoside; 6TG: Gthioguanine.

*

3+ 2

ATxI 75 Hi\tory o f paro.;y\ni.il dtrial td~hyl.irdia,trcdtcd \kith procainc-.~midc v ti I’ 80 ( 2 week\) 75 v 8i I’ (3 wcckc) 77 None DAT x 2 57 D A Tx 2 55 Trcatcd for ostcoarthritis for 2 years with Indonicthacin I 3 ’ I for thyrotoxicosis X years prcviously DAT x 2 64

18

F

1

62

3.5

M

F

I

-

TAHLI. 11. Clinical fcaurcs o f patient\ with > 5% TdT.+cell\

EL

2

L

TdT iii Aciite Noti-lytizplioid Lerikactiiia

FIG 1. Bone marrow from patient C.R. (case 2). Left: May-Griinwald-Giernsa stain. A mixture of largc myeloblasts and smaller lymphoblasts is seen ( x 630). Right: Sudan black stain. A large Sudan black positive blast is seen (arrow) ( x 630).

FIG 2. Bone marrow from patient C.R. (case 2), stained with rabbit-anti-TdT-FITC. The same field has been photographed with phase contrast (left) and using selective filters for FITC (right). Four TdT positive blasts (arrowed) are seen, which are generally smaller than the TdT-negative blasts ( X W ) .

( F a c i q p . 138)

Fic. 4. Bone niarro\t from patient I.M. ( c a x 12). 5tainc.d n i t h anti-TdT-FITC and anti-la-TRITC. Ttic. \.imc ticld h d \ hccn photographed by phaw contrdst (left) and using sclcctivc filtcrc for FITC (Lcxtrc) 2nd TRIT'C (right), Four TdT p o w i v c blast\ (arm\\ cd) cxpres\ surface Ia antigcns ni