Expression of the Tn antigen in myelodysplasia, lymphoma, and leukemia D.J. ROXBY,M.B. PFEIFFER,A.A. MORLEY,AND M.A. KIRKLAND Expression of the normally cryptic blood group antigen Tn has occasionall been reported in hematologic disease, but the true frequency of this change is not nown. A mouse monoclonal antibody (FBT3) and immunohistochemistry were used to examine expression of the Tn antigen. Expression was not detected in 35 normal bone marrow aspirates examined, but it was detected in 5 of 725 abnormal bone marrow aspirates, including 2 (3.6%) of 55 cases of de novo acute nonlymphocytic leukemia and 2 cases that terminated in acute nonlymphocytic leukemia. In two patients, one with acute myeloblastic leukemia and the other in blast transformation of chronic myeloid leukemia, the Tn antigen was expressed on 2 percent of blast cells. In one case of non-Hodgkin's lymphoma, 4 percent of normal myeloid cells expressed the antigen. In the other two cases, one of acute myelomonocytic leukemia and the other of myelod splasia, only 2 to 8 percent of myeloid and erythroid cells initially were Tn positive. {ubsequent serial immunohistochemical studies of bone marrow aspirates and peripheral blood in these two cases showed increasing numbers of Tn-positive erythroid and myeloid cells 8 to 12 months before polyagglutination was detected serologically.Tn-positive cells increased to >90 percent in the terminal phase in both cases of both diseases.The results suggest that Tn expression in these two patients may have conferred a growth advantage to the cells and could be related to disease progression. TRANSFUSION 1992;32:834-838.
l
Abbrevlatlons: MoAb = monoclonal antlbody; RBC(s) = red cell(s); TBS = Trls-buffered sallne.
surface of platelets, granulocytes, and lymphocytes of affected individuals.'O Recently, it has been recognized by ourselves and others that the Tn antigen is expressed in a wide range of solid tumors, including breast, colon, lung, stomach, pancreas, and ovary,'l-lZ and that the degree of expression may be related to the degree of tumor differentiation and aggressiveness and, hence, prognosis. l 3 The true incidence of the Tn phenomenon and the true frequency of Tn-positive cells in individual patients are uncertain because Tn-positive cells have previously been detected only by polyagglutination, which is insensitive to low levels of Tn expression and requires antigen expression on a large proportion of the RBCs. We therefore used a monoclonal antibody (MoAb) against Tn, developed in our laboratory, for systematic examination of normal and abnormal bone marrow aspirates so as to detect Tn expression on hemopoietic cells.
Tn POLYAGGLUTINATIONor persistent mixed-field agglutination is an acquired condition that was first described by Moreau et al.' in 1957 and that has been reported in association with a number of hematologic abnormalities, including acute leukemia, myelofibrosis, non-Hodgkin's lymphoma, and hemolytic anemia.2-6 Common characteristics of Tn red cells (RBCs) include agglutination by most human adult sera irrespective of blood group, reduced membrane sialic acid and galactose levels, and agglutination by the Tn-specific lectin, Salvia sclarea ,' The precise nature of the Tn antigen on RBCs has been identified as terminal N-acetylgalactosamine linked a-glycosidically to a serine or threonine residue of glycophorin A.8 The defect that leads to the in vivo production of Tn RBCs is due to a somatic mutation that occurs at the level of the pluripotent hemopoietic stem cell and results in a selective deficiency of the enzyme 3-P-~-galactosyltransferase.~ This gives rise to the heterogeneous cell population characterized by both Tnpositive and Tn-negative RBCs in the blood of the same individual. The Tn antigen can also be exposed on the
Case Reports Case 1
F.C., a 63-year-old woman (group B, Rh-positive), was admitted to the hospital in July 1988 with a diagnosis of acute
From the Hematology Department, Flinders Medical Centre, Bedford Park, South Australia, Australia. Rcccivcd for publication Dcccmbcr 16, 1991; rcvision rcceivcd May 10, 1992, and acceptcd May 20, 1992.
myeloblastic leukemia. Following induction and consolidation chemotherapy, the patient remained in remission. In October 1989, a follow-up bone marrow aspirate showed a sideroblastic
834
TRANSFUSION
835
TN ANTIGEN IN HEMATOLOGIC DISEASE
1992-Val. 32. No. 9
/
e f
80
-
40
8 F o
s
,
c
5
I
i
20 0 n
4Nov I S
Jan I 4
Junm (I6
Jan (IB
Mar I 6
Nov I S Jan I 4 Nzv
~~
Jun
Jan I 6
Mar Ma). II
Peripheral Blood (AT)
Bone Marrow Aapirationa (AT)
r--f ~
0.0
(I0
~
~~
NW (I7
0.0
(I9 Juna 90
Bone Marrow Aspirations (FMc)
Peripheral Blood (FMc)
FIG. 1. Percentage of Tn-positive erythroid (0)and myeloid (+) cells in bone marrow and peripheral blood from patients A.T. and F.Mc.
picture with 25-percent ringed sideroblasts present. Two percent of the blasts in this bone marrow aspirate were Tn positive. On the most recent clinical review (August 1991), she remained in remission and was receiving no chemotherapy. Case 2 E.M., a 58-year-old woman (group 0, Rh-positive), presented in December 1984 with diffuse, large-cell non-Hodgkin’s lymphoma with bone marrow involvement. This bone marrow was Tn negative. Remission was induced with combined chemotherapy and radiotherapy. In October 1989, she presented with a small bowel obstruction that was surgically corrected. This bowel obstruction was due to relapse of the lymphoma, which was Tn negative. A bone aspirate was performed in November 1989 and showed no evidence of involvement by lymphoma, but 4 percent of myeloid cells (including blasts and immature forms) were Tn positive. It is interesting that, on retesting of the marrow after chemotherapy, the Tnpositive cells had disappeared.When last seen in February 1991, the patient was in remission. Case 3 D.D., a 33-year-old man (group A, Rh-positive), was diagnosed as having chronic myeloid leukemia in March 1987. He was treated with hydroxyurea and interferon. In March 1988, the leukemia underwent acute myeloid transformation, and the patient died the following March of septicemia. Tn expression was detected in 2 percent of blast cells at the time of transformation.
Case 4 A.T., a 68-year-old man (group 0,Rh-positive), has been described previously.6 Briefly, he presented with acute myelomonocytic leukemia in November 1983 and was treated with standard chemotherapy. His disease relapsed in February 1985 and he died of septicemia in March 1985. Tn polyagglutination of his RBCs was noted 2 days prior to his death. Retrospective examination of archival bone marrow smears by immunohistochemistry showed Tn-positive cells in small numbers in April 1984 (Fig. lA), 8 months before relapse and 11 months before Tn polyagglutination was detected as a blood group anomaly. There was a progressive increase in the number of erythroid cells and, to a lesser extent, of myeloid cells expressing the Tn antigen. Similar immunohistochemical results were seen in his peripheral blood (Fig. 1B). Case 5 F.Mc., a 69-year-old man (group A, Rh-positive), presented in April 1986 with pancytopenia, splenomegaly, and some clinical features suggestive of an autoimmune disorder. Bone marrow examination was inconclusive. A second bone marrow examination was carried out in June 1986, and a diagnosis of myelodysplasia (subtype: refractory anemia with excess blasts) was made. Cytogenetic studies at that time showed a deletion occurring on a long arm of chromosome 11 in the banding region 13 to 23. In August 1986, he underwent splenectomy for persistent refractory thrornbocytopenia. He developed overt acute myeloid leukemia in January 1990 and died of septicemia in June 1990. Treatment involved administration of cyclo-
836
TRANSFUSION
ROXBY ET AL.
phosphamide, vincristine, danazol, and prednisolone and regular transfusions of RBCs and platelets. The original bone marrow sample, screened as part of this study, showed a small population of Tn-positive RBCs and white cells. With time, the percentage of Tn-positive cells within the marrow gradually increased (Fig. 1C). As in the case of A.T., peripheral blood smears from F.Mc. were also screened. The first available peripheral blood smear (December 1986) showed a small population of Tn-positive cells (Fig. 1D). Repeated peripheral blood examinations showed an increase in circulating Tn-positive RBCs over the period from December 1986 to December 1987 (Fig. lD), which was followed by a decrease over the next 12 months. From August 1989, there was a dramatic increase in the number of circulating Tn-positive RBCs; this preceded the transformation to acute leukemia in January 1990. The level of circulating Tnpositive RBCs had reached 95 percent at the time of the patient's death. Immunohistochemical examination of paraffinembedded spleen sections from F.Mc. showed that RBCs in the splenic sinusoids were Tn positive, but that splenic lymphatic tissue was negative. Using peripheral blood samples from F.Mc., we compared the sensitivity of S. sclureu and FBT3 in hemagglutination tests with the immunohistochemical technique for detection of Tn RBCs. As can be seen from Table 1, immunohistochemical staining using FBT3 was far more sensitive than hemagglutination.
Materials and Methods MoAb production The production of the MoAb, FBT3, was as described previously." In brief, FBT3 was cloned by using spleen cells from a male Balb/c mouse immunized with intact Tn-positive RBCs. We confirmed antibody specificity by demonstrating reactivity against seven examples of Tn-positive RBCs and lack of reactivity against Tn-negative RBCs which expressed between them various blood group specificities (ABO, Rh, Ii, Fy, K, MNSs, PI, Lu, Jk, Le, Xg, Wr, En(a-), Ma+, SD+, Sd(a+ +), St(a + ), T). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of Tn-positive and -negative RBC membranes and immunoblotting further confirmed Tn specificity.
Immunohistochemistly After obtaining informed consent, we made smears from bone marrow aspirates collected intraoperatively from volunTable
1.
Date (monthlyear) 12/86 2/87 3/87 4/87 1 1 I87 1 I88 1 1 189 1 190 2/90 5/90
Sensitivity of tests to detect Tn red cells in peripheral blood from F.Mc. Hemagglutination reaction score' S. sclarea FBT3 0 0 0 0 0 0 2 4 7 7
* Reaction score scale from 0 to
lmmunohistochemical stainingt loercentaoe of TN-positive"ce1ls)
0 0 0 1 6
5 12 12 12 12 12.
t This test used monoclonal anti-Tn (FBT3)
.r-
~~
3 4 5 7 23 14 52 47 66 95
Vol. 32, No. 9-1992
teers who were undergoing elective orthopedic or gynecologic surgery. They were presumed to have normal hemopoiesis and not to have cancer. Smears were also made from patients with documented hematologic abnormalities who were undergoing a diagnostic marrow aspirate. We used both fresh and previously stained (Jenner-Giemsa) bone marrow aspirate or peripheral blood smears for immunohistochemical staining. Previously stained smears were briefly destained in acid alcohol without artifactual problems being encountered. Fresh smears were fixed in methanol. We treated all smears with H,O,/methanol for 30 minutes at room temperature to block endogenous peroxidase, incubated them with normal horse serum for 20 minutes to reduce nonspecific binding, and then incubated them with either FBT3 or HO-2-2 for 60 minutes at room temperature. HO-2-2 is an IgM MoAb that is directed against the Lyt-2-2 murine T-lymphocyte differentiation antigen and that acted as a negative control. We then washed the smears three times each for 5 minutes in 0.02M (0.02 mol/L) Trisbuffered saline (TBS) (pH 7.6) and incubated then with biotinylated horse anti-mouse IgM for 60 minutes at room temperature, after which we again washed them in TBS. Avidinbiotin-peroxidase complex (Vector Laboratories, Burlingame. VT) was then applied to the smears for 45 minutes at room temperature. Following further washing in TBS, the smears were incubated in 1 mM (1 mmol/L) diaminobenzidine containing 0.015-percent (vol/vol) H202in TBS (pH 7.6) for 10 minutes. After rinsing them with TBS, we counterstained the smears in chloroform-extracted methyl green and examined them by light microscopy.
Results FBT3 was isotyped as IgM and was shown to react specifically with Tn-positive RBCs by hemagglutination and indirect immunofluorescence. We further confirmed Tn specificity by immunoblotting in which binding to a,6, and dimers of a and S sialoglycoproteins from Tn-positive RBC membranes was observed with FBT3, but none was evident with normal RBC membranes. We examined aspirates from 35 normal patients and 725 patients with hematologic disorders (Table 2). No hemopoietic cells from the 35 normal bone marrow aspirates examined expressed the Tn antigen, whereas membrane and cytoplasmic staining of mature and immature erythroid cells and some myeloid cells was seen in five pathologic bone marrow aspirates (Fig. 2) and in one case each of myelodysplasia, acute myelomonocytic leukemia, acute myeloblastic leukemia, chronic myeloid leukemia in blast phase, and non-Hodgkin's lymphoma. Table 2. Expression of Tn antigen in normal and hematologic disease Diagnosis Positive Total Normal 0 35 Acute lymphoblastic leukemia 0 27 Acute nonlymphoblastic leukemia 2 55 Myeloproliferative disorders 1 84 Myelodysplasia 1 60 Chronic lymphocytic leukemidymphoma 1 206 Others' 0 293 ' Includes autoimmune hemolytic anemia, iron deficiency and megaloblastic anemia, erythroid hypoplasia and hyperplasia, neutropenia, idiopathic thrombocytopenia purpura, and pancytopenia.
TRANSFUSION
837
TN ANTIGEN IN HEMATOLOGIC DISEASE
1992-Vol. 32, No. 9
FIG. 2. Bone marrow aspirate showing Tn-positive (1))and Tn negative (4)red cells.
Discussion
There is now much evidence that blood group-related antigens may be inappropriately expressed or deleted in patients with cancer, although the mechanisms responsible are poorly understood. However, the documentation of altered expression and the evaluation of distribution of these antigens in neoplastic tissues may provide further insights into mechanisms of tumor growth, metastasis, and the immune recognition process of the host that enhance or impair tumor proliferation. The Tn syndrome is a clonal disorder that affects granulocytic, lymphoid, erythroid, and megakaryocytic lineages. The biochemical defect in RBCs is a gross deficiency or absence of 3-P-~-galactosyltransferase,~ and its absence results in the exposure of terminal N-acetylgalactosamine serine or threonine residues that make up the Tn antigen on these cells. Bigbee et al.14 showed that normal individuals appear to have very low levels (c 1 x of circulating Tn-positive RBCs, and they gave several possible explanations for the presence of low numbers of variants in normal individuals, including the loss of both 3-P-~-galactosyltransferase alleles, the elimination of Tn-positive RBCs in vivo by serum antibodies against Tn that are present in most normal individua1s,l2 and a mutation that requires two genetic events to produce the Tn phenotype. However, the true prevalence of Tn expression in hematologic disorders is unclear, as Tn expression has previously been detected only by polyagglutination. Overall, our results suggest that the Tn antigen is not commonly expressed in hematologic malignancies, as compared to other tumors.”J* In a previous study,ll we showed that Tn was expressed in 97 percent of colon tumors, 92 percent of breast tumors, 88 percent of ovarian tumors, 86 percent of gastric tumors, and 56 percent of bladder tumors. Nevertheless, the finding of five cases of Tn expression in 725 bone marrow aspirates and, in
particular, in two patients with new acute leukemia and two patients whose disease terminated in acute leukemia suggests that this phenomenon is more common than has previously been recognized. In Cases 1, 2, and 3, only a small percentage of cells in the marrow showed Tn expression. In these cases, that expression may have represented clonal evolution of the underlying hemopoietic malignancy. At no time were Tn-positive cells detectable in the peripheral blood of Patients 1, 2, or 3. Cases 4 and 5 were the most interesting, in that Tnpositive cells became detectable in peripheral blood in low numbers and slowly increased in both the bone marrow and peripheral blood, until, at the time of the patients’ deaths, nearly the entire population of RBCs was Tn positive, as was a proportion of myeloid cells (Fig. 2). In Cases 4 and 5, there appeared to be a relationship between Tn expression and disease progression, which suggests that this is an epi-phenomenon and that the unmasking of this determinant could be of biologic importance in disease progression and may have provided a proliferative advantage to the leukemic clone. Our data indicate that Tn is expressed in a small proportion of patients with acute leukemia, and it seems possible that, in such patients, monitoring with FBT3 may be of value in the detection of small numbers of leukemic cells and in predicting relapse. Although this study used immunohistochemistry to detect Tn expression, in other studies (Roxby DJ, Burpee My Morley AA. Unpublished observations, 1985) we have found that RBCs expressing Tn can be detected with equally with flow cytometry. high sensitivity, down to References 1. Moreau R, Dausset J, Bernard J. Antmie htmolytique acquist avec polyagglutinabilitt des htmaties par un nouveau facteur prisent dans le serum humain normal (anti-Tn). Bull Soc Mtd Hop Paris 1957;73:569-87. 2. Bird GW, Wingham J, Pippard MJ, Hoult JG, Melikian V. Erythrocyte membrane modification in malignant diseases of myeloid and lymphoreticulartissues. I. Tn-polyagglutinationin acute myelomonocytic leukaemia. Br J Haematol 1976;33:289-94. 3. Bird GW, Wingham J, Chester GH, Kidd P, Payne RW. Erythrocytc membrane modification in malignant diseasc of myeloid and lymphoreticular tissues. 11. Erythrocyte “mosaicism” in acute erythroleukaemia. Br J Haematol 1976;33:295-9. 4. Bird GW, Wingham J, Richardson SG. Myelofibrosis, autoimmune haemolytic anaemia and Tn-polyagglutinability. Haematologia (Budap) 1985;18:99-103. 5. Wallner M, Waldner R. Tn polyagglutinability occurring in a patient with B cell lymphoma. Blut 1985;51:355-60. 6. Roxby DJ, Morley AA, Burpee M. Detection of thc Tn antigen in leukacmia using monoclonal anti-Tn antibody and immunohistochcmistry. Br J Haematol 1987;67:153-6. 7. Bird GW, Wingham J. Haemagglutinins from Sufviu. Vox Sang 1974;26:163-6. 8. Dahr W, Uhlenbruck G, Gunson HH, Van Dcr Hart M. Molecular basis of Tn-polyagglutinability. Vox Sang 1975;29:36-50. 9. Cartron JP, Cartron J, Andreu G, Salmon C, Bird GW. Selective deficicncy of 3-P-D-galactosyltransferase(T-transfcrase) in Tnpolyagglutinable erythrocytes. Lancet 1978;1:856-7.
838
ROXBY ET AL.
10. Ness PM. Garratty G, Morel PA, Perkins HA. Tn polyagglutination preceding acute leukemia. Blood 1979;54:30-4. 11. Roxby DJ. Skinner JM, Morley AA, Weeks S. Burpee M. Expression of a Tn-like epitope by carcinoma cells. Br J Cancer 1987;56:734-7. 12. Springer GF. T and Tn, general carcinoma autoantigens. Science 1984;224: 1198-206. 13. Nishiyama T, Matsumoto Y, Watanabe H, Fujiwara M, Sato S. Detection of Tn antigen with Kciu villosu agglutinin in urinary bladder cancer: its relevance to the patient’s clinical course. J Natl Cancer lnst 1987;78:1113-8. 14. Bigbee WL. Langlois RG, Stanker LH. Vanderlaan M, lensen RH. Flow cytometric analysis of erythrocyte populations in Tn
TRANSFUSION
Vol. 32. No. 9-1941
syndrome blood using monoclonal antibodies to glycophorin A and the Tn antigen. Qtometry 1990;11:261-71. David J. Roxby, MSc, Principal Hospital Scientist, Hematology Department, Flinders Medical Centre, Flinders Drive, Bcdford Park, South Australia 5042. Australia. [Reprint requests] Marie B. Pfeiffer, BSc, Research Assistant, Hematology Department, Flinders Medical Centre. Alec A. Morley. MD, Professor and Head, Hematology Department, Flinders Medical Centre. Mark A. Kirkland, BMedSc, MBBS, PhD, Registrar. Hematology Department, Flinders Medical Centre.
Corrections Goodnough LT, Verbrugge D, Vizmeg K, Riddell J IV. Identifying elective orthopedic surgical patients transfused with amounts of blood in excess of need: the transfusion trigger revisited. TRANSFUSION 1992;32:648-53.An error occurred in the last sentence of the abstract. The last sentence of the abstract should have read: With this method, it was found that the transfusion trigger is different for women and for men as well as for autologous blood donors and those who dld not donate autologous blood undergoing elective orthopedic surgery.
Ramos RR, Curtis BR, Chaplin H. A latex particle assay for platelet-associated IgG. TRANSFUSION 1992;32:235-8.An error occurred on page 236, column 1, lines 12-15.The sentence should have appeared as: Alternatively, the normal platelets were frozen at -80°C in dimethyl sulfoxide ( 5 % ) and PGE, (prostaglandin El) 1 Fg per mL (Sigma).
l
Abbrevlatlons: MoAb = monoclonal antlbody; RBC(s) = red cell(s); TBS = Trls-buffered sallne.
surface of platelets, granulocytes, and lymphocytes of affected individuals.'O Recently, it has been recognized by ourselves and others that the Tn antigen is expressed in a wide range of solid tumors, including breast, colon, lung, stomach, pancreas, and ovary,'l-lZ and that the degree of expression may be related to the degree of tumor differentiation and aggressiveness and, hence, prognosis. l 3 The true incidence of the Tn phenomenon and the true frequency of Tn-positive cells in individual patients are uncertain because Tn-positive cells have previously been detected only by polyagglutination, which is insensitive to low levels of Tn expression and requires antigen expression on a large proportion of the RBCs. We therefore used a monoclonal antibody (MoAb) against Tn, developed in our laboratory, for systematic examination of normal and abnormal bone marrow aspirates so as to detect Tn expression on hemopoietic cells.
Tn POLYAGGLUTINATIONor persistent mixed-field agglutination is an acquired condition that was first described by Moreau et al.' in 1957 and that has been reported in association with a number of hematologic abnormalities, including acute leukemia, myelofibrosis, non-Hodgkin's lymphoma, and hemolytic anemia.2-6 Common characteristics of Tn red cells (RBCs) include agglutination by most human adult sera irrespective of blood group, reduced membrane sialic acid and galactose levels, and agglutination by the Tn-specific lectin, Salvia sclarea ,' The precise nature of the Tn antigen on RBCs has been identified as terminal N-acetylgalactosamine linked a-glycosidically to a serine or threonine residue of glycophorin A.8 The defect that leads to the in vivo production of Tn RBCs is due to a somatic mutation that occurs at the level of the pluripotent hemopoietic stem cell and results in a selective deficiency of the enzyme 3-P-~-galactosyltransferase.~ This gives rise to the heterogeneous cell population characterized by both Tnpositive and Tn-negative RBCs in the blood of the same individual. The Tn antigen can also be exposed on the
Case Reports Case 1
F.C., a 63-year-old woman (group B, Rh-positive), was admitted to the hospital in July 1988 with a diagnosis of acute
From the Hematology Department, Flinders Medical Centre, Bedford Park, South Australia, Australia. Rcccivcd for publication Dcccmbcr 16, 1991; rcvision rcceivcd May 10, 1992, and acceptcd May 20, 1992.
myeloblastic leukemia. Following induction and consolidation chemotherapy, the patient remained in remission. In October 1989, a follow-up bone marrow aspirate showed a sideroblastic
834
TRANSFUSION
835
TN ANTIGEN IN HEMATOLOGIC DISEASE
1992-Val. 32. No. 9
/
e f
80
-
40
8 F o
s
,
c
5
I
i
20 0 n
4Nov I S
Jan I 4
Junm (I6
Jan (IB
Mar I 6
Nov I S Jan I 4 Nzv
~~
Jun
Jan I 6
Mar Ma). II
Peripheral Blood (AT)
Bone Marrow Aapirationa (AT)
r--f ~
0.0
(I0
~
~~
NW (I7
0.0
(I9 Juna 90
Bone Marrow Aspirations (FMc)
Peripheral Blood (FMc)
FIG. 1. Percentage of Tn-positive erythroid (0)and myeloid (+) cells in bone marrow and peripheral blood from patients A.T. and F.Mc.
picture with 25-percent ringed sideroblasts present. Two percent of the blasts in this bone marrow aspirate were Tn positive. On the most recent clinical review (August 1991), she remained in remission and was receiving no chemotherapy. Case 2 E.M., a 58-year-old woman (group 0, Rh-positive), presented in December 1984 with diffuse, large-cell non-Hodgkin’s lymphoma with bone marrow involvement. This bone marrow was Tn negative. Remission was induced with combined chemotherapy and radiotherapy. In October 1989, she presented with a small bowel obstruction that was surgically corrected. This bowel obstruction was due to relapse of the lymphoma, which was Tn negative. A bone aspirate was performed in November 1989 and showed no evidence of involvement by lymphoma, but 4 percent of myeloid cells (including blasts and immature forms) were Tn positive. It is interesting that, on retesting of the marrow after chemotherapy, the Tnpositive cells had disappeared.When last seen in February 1991, the patient was in remission. Case 3 D.D., a 33-year-old man (group A, Rh-positive), was diagnosed as having chronic myeloid leukemia in March 1987. He was treated with hydroxyurea and interferon. In March 1988, the leukemia underwent acute myeloid transformation, and the patient died the following March of septicemia. Tn expression was detected in 2 percent of blast cells at the time of transformation.
Case 4 A.T., a 68-year-old man (group 0,Rh-positive), has been described previously.6 Briefly, he presented with acute myelomonocytic leukemia in November 1983 and was treated with standard chemotherapy. His disease relapsed in February 1985 and he died of septicemia in March 1985. Tn polyagglutination of his RBCs was noted 2 days prior to his death. Retrospective examination of archival bone marrow smears by immunohistochemistry showed Tn-positive cells in small numbers in April 1984 (Fig. lA), 8 months before relapse and 11 months before Tn polyagglutination was detected as a blood group anomaly. There was a progressive increase in the number of erythroid cells and, to a lesser extent, of myeloid cells expressing the Tn antigen. Similar immunohistochemical results were seen in his peripheral blood (Fig. 1B). Case 5 F.Mc., a 69-year-old man (group A, Rh-positive), presented in April 1986 with pancytopenia, splenomegaly, and some clinical features suggestive of an autoimmune disorder. Bone marrow examination was inconclusive. A second bone marrow examination was carried out in June 1986, and a diagnosis of myelodysplasia (subtype: refractory anemia with excess blasts) was made. Cytogenetic studies at that time showed a deletion occurring on a long arm of chromosome 11 in the banding region 13 to 23. In August 1986, he underwent splenectomy for persistent refractory thrornbocytopenia. He developed overt acute myeloid leukemia in January 1990 and died of septicemia in June 1990. Treatment involved administration of cyclo-
836
TRANSFUSION
ROXBY ET AL.
phosphamide, vincristine, danazol, and prednisolone and regular transfusions of RBCs and platelets. The original bone marrow sample, screened as part of this study, showed a small population of Tn-positive RBCs and white cells. With time, the percentage of Tn-positive cells within the marrow gradually increased (Fig. 1C). As in the case of A.T., peripheral blood smears from F.Mc. were also screened. The first available peripheral blood smear (December 1986) showed a small population of Tn-positive cells (Fig. 1D). Repeated peripheral blood examinations showed an increase in circulating Tn-positive RBCs over the period from December 1986 to December 1987 (Fig. lD), which was followed by a decrease over the next 12 months. From August 1989, there was a dramatic increase in the number of circulating Tn-positive RBCs; this preceded the transformation to acute leukemia in January 1990. The level of circulating Tnpositive RBCs had reached 95 percent at the time of the patient's death. Immunohistochemical examination of paraffinembedded spleen sections from F.Mc. showed that RBCs in the splenic sinusoids were Tn positive, but that splenic lymphatic tissue was negative. Using peripheral blood samples from F.Mc., we compared the sensitivity of S. sclureu and FBT3 in hemagglutination tests with the immunohistochemical technique for detection of Tn RBCs. As can be seen from Table 1, immunohistochemical staining using FBT3 was far more sensitive than hemagglutination.
Materials and Methods MoAb production The production of the MoAb, FBT3, was as described previously." In brief, FBT3 was cloned by using spleen cells from a male Balb/c mouse immunized with intact Tn-positive RBCs. We confirmed antibody specificity by demonstrating reactivity against seven examples of Tn-positive RBCs and lack of reactivity against Tn-negative RBCs which expressed between them various blood group specificities (ABO, Rh, Ii, Fy, K, MNSs, PI, Lu, Jk, Le, Xg, Wr, En(a-), Ma+, SD+, Sd(a+ +), St(a + ), T). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of Tn-positive and -negative RBC membranes and immunoblotting further confirmed Tn specificity.
Immunohistochemistly After obtaining informed consent, we made smears from bone marrow aspirates collected intraoperatively from volunTable
1.
Date (monthlyear) 12/86 2/87 3/87 4/87 1 1 I87 1 I88 1 1 189 1 190 2/90 5/90
Sensitivity of tests to detect Tn red cells in peripheral blood from F.Mc. Hemagglutination reaction score' S. sclarea FBT3 0 0 0 0 0 0 2 4 7 7
* Reaction score scale from 0 to
lmmunohistochemical stainingt loercentaoe of TN-positive"ce1ls)
0 0 0 1 6
5 12 12 12 12 12.
t This test used monoclonal anti-Tn (FBT3)
.r-
~~
3 4 5 7 23 14 52 47 66 95
Vol. 32, No. 9-1992
teers who were undergoing elective orthopedic or gynecologic surgery. They were presumed to have normal hemopoiesis and not to have cancer. Smears were also made from patients with documented hematologic abnormalities who were undergoing a diagnostic marrow aspirate. We used both fresh and previously stained (Jenner-Giemsa) bone marrow aspirate or peripheral blood smears for immunohistochemical staining. Previously stained smears were briefly destained in acid alcohol without artifactual problems being encountered. Fresh smears were fixed in methanol. We treated all smears with H,O,/methanol for 30 minutes at room temperature to block endogenous peroxidase, incubated them with normal horse serum for 20 minutes to reduce nonspecific binding, and then incubated them with either FBT3 or HO-2-2 for 60 minutes at room temperature. HO-2-2 is an IgM MoAb that is directed against the Lyt-2-2 murine T-lymphocyte differentiation antigen and that acted as a negative control. We then washed the smears three times each for 5 minutes in 0.02M (0.02 mol/L) Trisbuffered saline (TBS) (pH 7.6) and incubated then with biotinylated horse anti-mouse IgM for 60 minutes at room temperature, after which we again washed them in TBS. Avidinbiotin-peroxidase complex (Vector Laboratories, Burlingame. VT) was then applied to the smears for 45 minutes at room temperature. Following further washing in TBS, the smears were incubated in 1 mM (1 mmol/L) diaminobenzidine containing 0.015-percent (vol/vol) H202in TBS (pH 7.6) for 10 minutes. After rinsing them with TBS, we counterstained the smears in chloroform-extracted methyl green and examined them by light microscopy.
Results FBT3 was isotyped as IgM and was shown to react specifically with Tn-positive RBCs by hemagglutination and indirect immunofluorescence. We further confirmed Tn specificity by immunoblotting in which binding to a,6, and dimers of a and S sialoglycoproteins from Tn-positive RBC membranes was observed with FBT3, but none was evident with normal RBC membranes. We examined aspirates from 35 normal patients and 725 patients with hematologic disorders (Table 2). No hemopoietic cells from the 35 normal bone marrow aspirates examined expressed the Tn antigen, whereas membrane and cytoplasmic staining of mature and immature erythroid cells and some myeloid cells was seen in five pathologic bone marrow aspirates (Fig. 2) and in one case each of myelodysplasia, acute myelomonocytic leukemia, acute myeloblastic leukemia, chronic myeloid leukemia in blast phase, and non-Hodgkin's lymphoma. Table 2. Expression of Tn antigen in normal and hematologic disease Diagnosis Positive Total Normal 0 35 Acute lymphoblastic leukemia 0 27 Acute nonlymphoblastic leukemia 2 55 Myeloproliferative disorders 1 84 Myelodysplasia 1 60 Chronic lymphocytic leukemidymphoma 1 206 Others' 0 293 ' Includes autoimmune hemolytic anemia, iron deficiency and megaloblastic anemia, erythroid hypoplasia and hyperplasia, neutropenia, idiopathic thrombocytopenia purpura, and pancytopenia.
TRANSFUSION
837
TN ANTIGEN IN HEMATOLOGIC DISEASE
1992-Vol. 32, No. 9
FIG. 2. Bone marrow aspirate showing Tn-positive (1))and Tn negative (4)red cells.
Discussion
There is now much evidence that blood group-related antigens may be inappropriately expressed or deleted in patients with cancer, although the mechanisms responsible are poorly understood. However, the documentation of altered expression and the evaluation of distribution of these antigens in neoplastic tissues may provide further insights into mechanisms of tumor growth, metastasis, and the immune recognition process of the host that enhance or impair tumor proliferation. The Tn syndrome is a clonal disorder that affects granulocytic, lymphoid, erythroid, and megakaryocytic lineages. The biochemical defect in RBCs is a gross deficiency or absence of 3-P-~-galactosyltransferase,~ and its absence results in the exposure of terminal N-acetylgalactosamine serine or threonine residues that make up the Tn antigen on these cells. Bigbee et al.14 showed that normal individuals appear to have very low levels (c 1 x of circulating Tn-positive RBCs, and they gave several possible explanations for the presence of low numbers of variants in normal individuals, including the loss of both 3-P-~-galactosyltransferase alleles, the elimination of Tn-positive RBCs in vivo by serum antibodies against Tn that are present in most normal individua1s,l2 and a mutation that requires two genetic events to produce the Tn phenotype. However, the true prevalence of Tn expression in hematologic disorders is unclear, as Tn expression has previously been detected only by polyagglutination. Overall, our results suggest that the Tn antigen is not commonly expressed in hematologic malignancies, as compared to other tumors.”J* In a previous study,ll we showed that Tn was expressed in 97 percent of colon tumors, 92 percent of breast tumors, 88 percent of ovarian tumors, 86 percent of gastric tumors, and 56 percent of bladder tumors. Nevertheless, the finding of five cases of Tn expression in 725 bone marrow aspirates and, in
particular, in two patients with new acute leukemia and two patients whose disease terminated in acute leukemia suggests that this phenomenon is more common than has previously been recognized. In Cases 1, 2, and 3, only a small percentage of cells in the marrow showed Tn expression. In these cases, that expression may have represented clonal evolution of the underlying hemopoietic malignancy. At no time were Tn-positive cells detectable in the peripheral blood of Patients 1, 2, or 3. Cases 4 and 5 were the most interesting, in that Tnpositive cells became detectable in peripheral blood in low numbers and slowly increased in both the bone marrow and peripheral blood, until, at the time of the patients’ deaths, nearly the entire population of RBCs was Tn positive, as was a proportion of myeloid cells (Fig. 2). In Cases 4 and 5, there appeared to be a relationship between Tn expression and disease progression, which suggests that this is an epi-phenomenon and that the unmasking of this determinant could be of biologic importance in disease progression and may have provided a proliferative advantage to the leukemic clone. Our data indicate that Tn is expressed in a small proportion of patients with acute leukemia, and it seems possible that, in such patients, monitoring with FBT3 may be of value in the detection of small numbers of leukemic cells and in predicting relapse. Although this study used immunohistochemistry to detect Tn expression, in other studies (Roxby DJ, Burpee My Morley AA. Unpublished observations, 1985) we have found that RBCs expressing Tn can be detected with equally with flow cytometry. high sensitivity, down to References 1. Moreau R, Dausset J, Bernard J. Antmie htmolytique acquist avec polyagglutinabilitt des htmaties par un nouveau facteur prisent dans le serum humain normal (anti-Tn). Bull Soc Mtd Hop Paris 1957;73:569-87. 2. Bird GW, Wingham J, Pippard MJ, Hoult JG, Melikian V. Erythrocyte membrane modification in malignant diseases of myeloid and lymphoreticulartissues. I. Tn-polyagglutinationin acute myelomonocytic leukaemia. Br J Haematol 1976;33:289-94. 3. Bird GW, Wingham J, Chester GH, Kidd P, Payne RW. Erythrocytc membrane modification in malignant diseasc of myeloid and lymphoreticular tissues. 11. Erythrocyte “mosaicism” in acute erythroleukaemia. Br J Haematol 1976;33:295-9. 4. Bird GW, Wingham J, Richardson SG. Myelofibrosis, autoimmune haemolytic anaemia and Tn-polyagglutinability. Haematologia (Budap) 1985;18:99-103. 5. Wallner M, Waldner R. Tn polyagglutinability occurring in a patient with B cell lymphoma. Blut 1985;51:355-60. 6. Roxby DJ, Morley AA, Burpee M. Detection of thc Tn antigen in leukacmia using monoclonal anti-Tn antibody and immunohistochcmistry. Br J Haematol 1987;67:153-6. 7. Bird GW, Wingham J. Haemagglutinins from Sufviu. Vox Sang 1974;26:163-6. 8. Dahr W, Uhlenbruck G, Gunson HH, Van Dcr Hart M. Molecular basis of Tn-polyagglutinability. Vox Sang 1975;29:36-50. 9. Cartron JP, Cartron J, Andreu G, Salmon C, Bird GW. Selective deficicncy of 3-P-D-galactosyltransferase(T-transfcrase) in Tnpolyagglutinable erythrocytes. Lancet 1978;1:856-7.
838
ROXBY ET AL.
10. Ness PM. Garratty G, Morel PA, Perkins HA. Tn polyagglutination preceding acute leukemia. Blood 1979;54:30-4. 11. Roxby DJ. Skinner JM, Morley AA, Weeks S. Burpee M. Expression of a Tn-like epitope by carcinoma cells. Br J Cancer 1987;56:734-7. 12. Springer GF. T and Tn, general carcinoma autoantigens. Science 1984;224: 1198-206. 13. Nishiyama T, Matsumoto Y, Watanabe H, Fujiwara M, Sato S. Detection of Tn antigen with Kciu villosu agglutinin in urinary bladder cancer: its relevance to the patient’s clinical course. J Natl Cancer lnst 1987;78:1113-8. 14. Bigbee WL. Langlois RG, Stanker LH. Vanderlaan M, lensen RH. Flow cytometric analysis of erythrocyte populations in Tn
TRANSFUSION
Vol. 32. No. 9-1941
syndrome blood using monoclonal antibodies to glycophorin A and the Tn antigen. Qtometry 1990;11:261-71. David J. Roxby, MSc, Principal Hospital Scientist, Hematology Department, Flinders Medical Centre, Flinders Drive, Bcdford Park, South Australia 5042. Australia. [Reprint requests] Marie B. Pfeiffer, BSc, Research Assistant, Hematology Department, Flinders Medical Centre. Alec A. Morley. MD, Professor and Head, Hematology Department, Flinders Medical Centre. Mark A. Kirkland, BMedSc, MBBS, PhD, Registrar. Hematology Department, Flinders Medical Centre.
Corrections Goodnough LT, Verbrugge D, Vizmeg K, Riddell J IV. Identifying elective orthopedic surgical patients transfused with amounts of blood in excess of need: the transfusion trigger revisited. TRANSFUSION 1992;32:648-53.An error occurred in the last sentence of the abstract. The last sentence of the abstract should have read: With this method, it was found that the transfusion trigger is different for women and for men as well as for autologous blood donors and those who dld not donate autologous blood undergoing elective orthopedic surgery.
Ramos RR, Curtis BR, Chaplin H. A latex particle assay for platelet-associated IgG. TRANSFUSION 1992;32:235-8.An error occurred on page 236, column 1, lines 12-15.The sentence should have appeared as: Alternatively, the normal platelets were frozen at -80°C in dimethyl sulfoxide ( 5 % ) and PGE, (prostaglandin El) 1 Fg per mL (Sigma).
