Role of Immunotyping in Chronic Lymphocytosis: Review of the Natural History of the Condition in 145 Adult Patients

Transcription

1 Subject Review Role of Immunotyping in Chronic Lymphocytosis: Review of the Natural History of the Condition in 145 Adult Patients AYALEW TEFFERI, M.D., Division of Hematology and Internal Medicine; CHIN-YANG LI, M.D., Department of Laboratory Medicine and Pathology and Division of Hematology and Internal Medicine; ROBERT L. PHYLIKY, M.D., Division of Hematology and Internal Medicine We investigated the clinical value of immunotyping in 145 consecutive adult patients with absolute or relative lymphocytosis: 132 (91%) had B-cell lymphocytosis, 5 (4%) had T-cell lymphocytosis, 2 (1%) had hairy cell leukemia, and 6 (4%) had reactive lymphocytosis. Of the five patients with T-cell lymphocytosis, four were best categorized as having Ty-chronic lymphoproliferative disease and had an indolent clinical course. Of the 132 patients with B-cell lymphocytosis, 121 (92%) had B-cell chronic lymphocytic leukemia (B-CLL), and 11 (8%) had small cleaved ("lymphosarcoma") cell leukemia. Patients with small cleaved cell leukemia had a worse clinical outcome than did those with B-CLL. We further analyzed the surface immunoglobulin (slg) and CD20 (B-l) antigen expression patterns in B-CLL to determine whether any correlation existed with clinical outcome. A subset of patients with B-CLL in whom slg was expressed in less than 20% of their lymphocytes had the best clinical outcome. HLA-DR (la-like) antigen typing helped identify B-CLL cases with minimal or no slg expression. CD20 (B-l) antigen was weak or undetectable in most cases of B-CLL. Patients with B-CLL who had CD20 (B-l) in more than 20% of their lymphocytes did not have a different prognosis. Our data provide the incidence and natural history of the various subsets of CLL in a series of patients at a single institution. The type and extent of immunotyping necessary and practical in the clinical management of patients with CLL are explored. Clinicians are often confronted with patients who have absolute or relative lymphocytosis. Previously, in the absence of clinical symptoms or signs of leukemia, some of these patients were observed for a period before being labeled as having chronic lymphocytic leukemia (CLL). The development of monoclonal antibodies, improvement in the techniques of immunocytochemical staining, and, more recently, the advent of immunoglobulin or T-cell receptor gene rearrangement studies have made it possible to distinguish reactive from clonal lymphoproliferative dis- Address reprint requests to Dr. C.-Y. Li, Department of Laboratory Medicine, Mayo Clinic, Rochester, MN orders. 1 " 6 These techniques have also been instrumental in subclassifying patients with CLL into phenotypically different clinicopathologic entities. 7 Little is known about the relative incidences of these various subsets in a cohort of patients examined at a single institution. Of patients with B-cell CLL (B-CLL), 10 to 20% show minimal or no surface immunoglobulin (slg) expression. 8 The clinical relevance of this observation in patients with B-CLL is unknown. In this study, we (1) report incidence figures and review the natural history of the various subsets in CLL, (2) investigate the correlation of slg and CD20 (B-l) antigen expression patterns in B-CLL with clinical outcome, (3) assess clinical and morphologic criteria in identifying reactive Mayo Clin Proc 63: ,

2 802 IMMUNOTYPING IN CHRONIC LYMPHOCYTOSIS Mayo Clin Proc, August 1988, Vol 63 lymphocytosis, and finally (4) provide guidelines for determining the type and extent of immunotyping needed in the management of patients with CLL. MATERIAL AND METHODS Between August and December 1984, peripheral blood specimens of 145 consecutive adult patients with absolute or relative lymphocytosis were immunophenotyped at our institution. All patients had relative lymphocytosis in the context of normal or elevated leukocyte counts. At the time of immunotyping, CLL had previously been diagnosed in most patients. Air-dried blood smears were used for lymphocyte typing with monoclonal antibodies specific for HLA-DR (la-like), CD3 (Leu-4) (Becton Dickinson Monoclonal Antibody Center, Inc., Mountain View, California), CD4 (OKT4), CD8 (OKT8) (Ortho Diagnostic Systems, Inc., Raritan, New Jersey), and CD20 (B-l) (Coulter Electronics, Inc., Hialeah, Florida) antigens. An indirect immunocytochemical technique with alkaline phosphatase as indicator was used, as previously described. 9 Alkaline phosphataseconjugated goat anti-human κ and λ light chain antisera (Tago, Inc., Burlingame, California) were used in a direct immunoalkaline phosphatase technique for study of slg in suspension, as described elsewhere. 10 Minimal slg expression was defined as either no detectable slg expression or weak slg expression in less than 20% of the circulating lymphocytes. Interpretation of "weak" and "strong" staining was based on previously published criteria. 9 The medical records of all the patients were reviewed retrospectively. The prognosis was evaluated by assessing progression of disease, development of cytopenia (or cytopenias), presence of hypogammaglobulinemia, and frequency of therapeutic interventions. Progression of disease was defined as the development of or an increase in palpable lymph nodes, hepatosplenomegaly, and cytopenia. Cytopenia was defined as a hemoglobin concentration of less than 10 g/dl and a platelet count of less than 100,000/mm 3. Reactive lymphocytosis was defined as an increase of CD8 (OKT8)+ cells with no appreciable changes in the proportion of B- and T-cell subsets. RESULTS Immunophenotypic, morphologic, and clinical characteristics were used to determine the subtype of lymphocytosis in 145 patients (Table 1). Of the 145 patients, 132 (91%) were typed as having a B-cell lymphoproliferative disorder [HLA-DR (Ia-like)+, light chain restricted, CD3 (Leu-4)-], 5 (4%) had a T-cell lymphoproliferative disorder [CD3 (Leu-4)+, CD4 (OKT4) or CD8 (OKT8)+, slg-], 2 (1%) had hairy cell leukemia [(tartrate-resistant acid phosphatase)+, slg+, CD3 (Leu-4)-], and 6 (4%) had reactive lymphocytosis. No symptoms or signs of chronic leukemia developed in any patient with reactive lymphocytosis, and four had substantiated resolution of the lymphocytosis on follow-up examinations. The median leukocyte count of these patients with reactive lymphocytosis was 8 χ lovmm 3 (range, 5 to 15 χ 10 3 /mm 3 ). The underlying diseases in four of these patients were colonic cancer, rheumatoid arthritis, Wegener's granulomatosis, and postsplenectomy state, respectively. Both patients with hairy cell leukemia had strong slg, HLA-DR (la-like), and CD20 (B-l) antigen expression. After a median follow-up of more than 5 years, both patients with hairy cell disease had undergone splenectomy and were receiving cr-interferon therapy at the time of their last clinic examination. Four of the five patients with T-cell lymphoproliferative disorders had the T cytotoxic/suppressor phenotype CD8 (OKT8)+. After a median follow-up of 6 years, none of these patients had progression of disease, and only one required treatment. This latter patient had autoimmune hemolytic anemia, and two other patients had hypogammaglobulinemia. The median leukocyte count of 3 x 10 3 /mm 3, the concomitant granulocytopenia, and the morphologically identifiable large granular lymphocytes in all these patients Table 1. Subtypes of Lymphocytosis in 145 Patients Subtype* B-cell lymphocytosis B-CLL SCCL T-cell lymphocytosis Ty-CLPD Sezary syndrome Hairy cell leukemia Reactive lymphocytosis Patients No % 91 *B-CLL = B-cell chronic lymphocytic leukemia; SCCL = small cleaved ("lymphosarcoma") cell leukemia; Ty-CLPD = Tychronic lymphoproliferative disease

3 Mayo Clin Proc, August 1988, Vol 63 IMMUNOTYPING IN CHRONIC LYMPHOCYTOSIS 803 best classified them as having Ty-chronic lymphoproliferative disease. The fifth patient with T-cell lymphocytosis had circulating Sozary cells phenotyped as CD4 (OKT4)+. Of the 132 patients with B-cell lymphoproliferative disorders, 121 (92%) were typed as having B-CLL [with mature-appearing circulating lymphocytes that were HLA-DR (Ia-like)+, slg+, and CD3 (Leu-4)-], and 11 (8%) were typed as having small cleaved cell leukemia [with circulating small cleaved cells that strongly expressed sig and CD20 (B-l) ]. After a median follow-up of 6 years, patients with small cleaved cell leukemia had a poorer prognosis than did those with B-CLL (73% of the patients had progression of disease and an increased frequency of therapeutic interventions). All patients with small cleaved cell leukemia in whom serum protein electrophoresis was done had hypogammaglobulinemia. The 121 patients with B-CLL were further subcategorized into various groups on the basis of the degree and the extent of their sig and CD20 (B-l) antigen expression (Table 2). No clear-cut morphologic difference was noted in the circulating lymphocytes among the groups. The median duration of follow-up of these groups ranged from 7 to 10 years. Group A [16 patients with weak sig and CD20 (B-l) antigen expression in less than 20% of their lymphocytes] had the best prognosis; only 6 to 7% of the patients had progression of disease or hypogammaglobulinemia. Of these patients, only 12% required treatment. Although HLA-DR (la-like) antigen was always expressed in this group, CD20 (B-l) antigen expression was minimal or absent. The patients in groups B [74 patients with more than 50% of their lymphocytes weakly expressing sig but less than 20% of their lymphocytes expressing CD20 (B-l) ], C [16 patients with CD20 (B-l) expression in more than 20% of their lymphocytes and variable sig expression], and D [15 patients with strong sig expression but lacking circulating small cleaved cells and showing variable CD20 (B-l) expression] had similar prognoses, which were intermediate between that of group A and that of patients with small cleaved cell leukemia. In each of these last three groups (B, C, and D), approximately 30% of patients had progression of disease and an increased need for therapy, and approximately 50% had hypogammaglobulinemia. The difference in progression of disease and development of cytopenia between groups A and B on one hand and groups B and small cleaved cell leukemia on the other was statistically significant (P values of 0.05 and 0.03, respectively; chisquared test with Yates' correction). Among the subgroups of B-CLL, no significant differences were found in the median age of patients and the Rai clinical stages at the time of diagnosis, as summarized in Table 2. The Rai clinical staging at the time of immunotyping is also shown in Table 2. Overall, the incidence of hypogammaglobulinemia in our series of patients with B-CLL was 42% of those in whom serum protein electrophoresis was done. Of the patients with hypogammaglobulinemia, 66% had progression of disease and an increased need for treatment; thus, hypogammaglobulinemia seems to have a negative prognostic value. In the B-lymphocytosis group, 11 patients (8%) had monoclonal gammopathy (group A = 0, B = 4, C = 2, D = 3, and small cleaved cell leukemia = 2). Seven patients had μ heavy chains, and four had y heavy chains. Seven patients had λ light chains, and four had κ light chains. The prognosis of patients with monoclonal gammopathy did not seem to differ from that of patients without monoclonal gammopathy. Four patients had immune thrombocytopenic purpura, and only one had confirmed autoimmune hemolytic anemia in the B-CLL group. DISCUSSION Until recently, proposed classifications of CLL have been confusing and inconsistent. The development of monoclonal antibodies and improvements in the techniques of immunocytochemical staining have contributed to improved diagnosis and classification of CLL. 1-4 Currently, various subsets of the disorder with clinical, morphologic, and phenotypic differences are recognized, 7 but little is known about the relative incidences of these subsets. Our data provide general incidence figures within the various subsets of CLL in a cohort of patients from a large series at a single institution. The incidence of each subgroup in this study, however, may not be representative of the general population, inasmuch as our patient population is subject to unavoidable referral bias. Light chain restriction in sig expression is generally considered as denoting a monoclonal B-cell lymphoproliferative disorder. 11 Such a monoclonal pattern has been indispensable in the distinction of B-cell malignant lesions from reactive

4 804 IMMUNOTYPING IN CHRONIC LYMPHOCYTOSIS Mayo Clin Proc, August 1988, Vol 63 Subgroup* Table 2. Profile of 132 Patients With Various Types of B-Cell Lymphocytosis Rai staging at time Disease Median Median of diagnosis and at progression Chemo- Hypogamma- Patients "*Γ!^Γ" foif^wtud time of immunotyping or cytopenia therapy globulinemia No. % (yr) (yr) ~~Ö Ϊ II III IV No % No! % No~f %~ B-CLL Group A: HLA-DR (Ia-like)+; <20% of lymphocytes weakly expressing slg or CD20 (B-l) antigen Group B: HLA-DR (Ia-like)+; >50% of lymphocytes weakly expressing slg; <20% expressing CD20 (B-l) Group C: HLA-DR (Ia-like)+; >20% of lymphocytes expressing CD20 (B-l); variable slg staining Group D: HLA-DR (Ia-like)+; majority of cells strongly expressing slg; variable expression ofcd20(b-l) SCCL HLA-DR (Ia-like)+; presence of circulating small cleaved cells strongly expressing slg 10 11; 3; 2; 0; 0; of ; 13; 21; 0; 0; of ; 4; 5; 0; 0; ; 2; 4; 0; 0; ; 2; 3; 0; 0; of of of *B-CLL = B-cell chronic lymphocytic leukemia; SCCL = small cleaved ("lymphosarcoma") cell leukemia; slg = surface immunoglobulin. fno. with hypogammaglobulinemia among those in whom serum protein electrophoresis was done. B-cell processes in lymph node, bone marrow, and body fluid examinations. 10,12 In the peripheral blood, however, reactive lymphocytosis is almost always of T-cell lineage, and slg-like markers are not immediately available to determine clonality in T-cell lymphocytosis. Recently, new procedures in molecular biology have helped determine monoclonality in T-cell lineage through T-cell receptor gene rearrangement studies. 6 At present, these techniques are tedious, time consuming, and not widely available. Their use in T-cell lymphocytosis should be restricted to selected cases. Clinical and morphologic criteria alone, or with additional quantitative estimation of the proportion of the different subsets of lymphocytes in equivocal cases, may help in separating T-cell lymphoproliferative disorders from reactive lymphocytosis. We recommend immunotyping in patients with unexplained "relative lymphocytosis." Our data show the usual presence of associated diseases in patients with reactive lymphocytosis and the resolution of lymphocytosis with time. The two subsets that could potentially be confused with reactive lymphocytosis are Ty-chronic lymphoproliferative disease and T-suppressor CLL. The former is usually associated with granulocytopenia and large granular lymphocytes. 13,14 Profound granulocytopenia is seldom seen in reactive lymphocytosis. T-suppressor CLL is usually characterized by significant lymphocytosis (lymphocyte counts that exceed 20 χ lovmm 3 ). 15 This degree of lymphocytosis is unusual in reactive lymphocytosis. The indolent clinical course of Ty-chronic lymphoproliferative disease and T- suppressor CLL allows watchful observation of equivocal cases without resorting to gene rearrangement studies. T-cell lymphoproliferative disorder can be further subtyped into helper/inducer CD4 (OKT4) or cytotoxic/suppressor CD8 (OKT8) type. Our patients with Ty-chronic lymphoproliferative disease had an indolent clinical course, similar to

5 Mayo Clin Proc, August 1988, Vol 63 IMMUNOTYPING IN CHRONIC LYMPHOCYTOSIS 805 those observed in earlier studies. 16 Conversely, patients with T-helper CLL are known to have an aggressive clinical course, with a median duration of survival of less than 2 years. 17 Consequently, distinction between these two subtypes has a definite clinical relevance and should be attempted in all cases. In addition, no case in our series was identified as T-helper CLL, a finding that suggests the incidence of this disease is low probably less than that of Ty-chronic lymphoproliferative disease. The immunologically more differentiated subsets of B-lymphocytic leukemia that is, prolymphocytic leukemia and small cleaved cell leukemia are easily identified by their clinical and morphologic characteristics. Patients in these subsets have a much worse prognosis than do those with B-CLL, as evidenced by the clinical course of our 11 patients with small cleaved cell leukemia and those described in other studies. 18,19 Both of these subsets are usually associated with strong slg expression. The clinical relevance of strong slg expression in the context of B-CLL (that is, in the absence of circulating prolymphocytes or small cleaved cells) has not been described previously. In our group of 121 patients with B-CLL, we identified 23 patients (19%) with strong slg expression, and their clinical outcome did not differ from the outcome of those patients with B-CLL who had weak slg expression. Even though B-cell lymphocytes are thought to express weak slg, the degree and the extent of slg expression vary. Of patients with B-CLL, 10 to 20% are reported to have minimal or no slg expression. 8,20,21 This finding is in agreement with our observed incidence of 13%. Only in our study has the clinical relevance of this observation been investigated, and we found an indolent clinical course in this group of patients. Our data show that CD20 (B-l) antigen staining was not always present on slg- B cells, and when it was present along with slg, it had no prognostic implication. In contrast, HLA-DR (lalike) antigen was universally present in patients with B-CLL and helped identify those patients with minimal or undetectable slg expression. HLA-DR (la-like) antigen is an extremely sensitive marker for B-CLL but lacks specificity because of its presence in activated T cells and monocytes. The latter can easily be distinguished from small lymphocytes by morphologic analysis alone. Combining CD3 with HLA-DR staining should distinguish B lymphocytes (CD3-) from activated T lymphocytes (CD3+). Hypogammaglobulinemia was shown to have a negative prognostic effect in our series. This influence was not due to increased infectious complications alone but to actual deterioration and development of cytopenia and increased chemotherapeutic requirements. The incidence of serologically identified monoclonal gammopathy in our series was 8%, similar to that reported in the literature. 22,23 Most patients have μ heavy chains and λ light chains. The extent of immunotyping necessary for the evaluation of lymphocytosis depends on the practicality and ease of performing the test in a specific patient and the information needed to be obtained. Using a combination of HLA-DR (lalike) and CD3 (Leu-4) is a practical way of distinguishing B-cell from T-cell lymphoproliferative disorders of small lymphocytic subtype. It is extremely sensitive and its specificity is enhanced because of the rarity of benign B-cell lymphocytosis. Determination of light chain restriction is more confirmatory but is also more tedious and less practical. In the context of a T-cell lymphoproliferative disorder, CD4 (OKT4) and CD8 (OKT8) antigen typing should be done, for reasons already discussed. The characteristic strong CD20 (B-l) expression in small cleaved cell leukemia, prolymphocytic leukemia, and hairy cell disease may be helpful in the diagnosis of these disorders. Its role in the diagnosis and prognostication of B-CLL is minimal. If our observation of an indolent clinical course for patients with minimal or undetectable slg expression is supported by findings of other investigators, initial routine slg typing for all patients with B-CLL may be helpful. Finally, development of new antibodies with restricted antigen specificity, such as the antibody against the CLL antigen, may simplify the specific diagnosis of CLL. 24 ACKNOWLEDGMENT We thank Mary Ann Morris for her technical assistance. REFERENCES 1. Foon KA, Schroff RW, Gale RP: Surface markers on leukemia and lymphoma cells: recent advances. Blood 60:1-19, Li C-Y: Immunocytochemical techniques for identifying leukemias. Mayo Clin Proc 59: ,1984

6 806 IMMUNOTYPING IN CHRONIC LYMPHOCYTOSIS Mayo Clin Proc, August 1988, Vol Van der Reijden HJ, van der Gaag R, Pinkster J, Riimke HC, van 't Veer MB, Melief C JM, von dem Borne AEGK: Chronic lymphocytic leukemia: immunologic markers and functional properties of the leukemic cells. Cancer 50: , Romain PL, Schlossman SF: Human T lymphocyte subsets: functional heterogeneity and surface recognition structures. J Clin Invest 74: , Waldmann TA, Korsmeyer SJ, Bakhshi A, Arnold A, Kirsch IR: Molecular genetic analysis of human lymphoid neoplasms: immunoglobulin genes and the c-myc oncogene. Ann Intern Med 102: , Minden MD, Toyonaga B, Ha K, Yanagi Y, Chin B, Gelfand E, Mak T: Somatic rearrangement of T-cell antigen receptor gene in human T-cell malignancies. Proc Natl Acad Sei USA 82: , Gale RP, Foon KA: Biology of chronic lymphocytic leukemia. Semin Hematol 24: , Freedman AS, Boyd AW, Bieber FR, Daley J, Rosen K, Horowitz JC, Levy DN, Nadler LM: Normal cellular counterparts of B cell chronic lymphocytic leukemia. Blood 70: , Li C-Y, Ziesmer SC, Yam LT, English MC, Janckila AJ: Practical immunocytochemical identification of human blood cells. Am J Clin Pathol 81: , Li C-Y, Witzig TE, Phyliky RL, Ziesmer SC, Yam LT: Diagnosis of B-cell non-hodgkin's lymphoma of the central nervous system by immunocytochemical analysis of cerebrospinal fluid lymphocytes. Cancer 57: , Preud'homme JL, Seligmann M: Surface bound immunoglobulins as a cell marker in human lymphoproliferative diseases. Blood 40: , Cossman J, Neckers LM, Hsu S-M, Longo D, Jaffe ES: Low-grade lymphomas: expression of developmentally regulated B-cell antigens. Am J Pathol 115: , Reynolds CW, Foon KA: Ty-lymphoproliferative disease and related disorders in humans and experimental animals: a review of the clinical, cellular, and functional characteristics. Blood 64: , Brouet J-C, Flandrin G, Sasportes M, Preud'Homme J-L, Seligmann M: Chronic lymphocytic leukaemia of T-cell origin: immunological and clinical evaluation in eleven patients. Lancet 2: , Nair KG, Han T, Minowada J: T-cell chronic lymphocytic leukemia: report of a case and review of the literature. Cancer 44: , Phyliky RL, Li C-Y, Yam LT: T-cell chronic lymphocytic leukemia with morphologic and immunologic characteristics of cytotoxic/suppressor phenotype. Mayo Clin Proc 58' Witzig TE! Phyliky RL, Li C-Y, Homburger HA, Dewald GW, Handwerger BS: T-cell chronic lymphocytic leukemia with a helper/inducer membrane phenotype: a distinct clinicopathologic subtype with a poor prognosis. Am J Hematol 21: , Melo JV, Catovsky D, Gregory WM, Galton DAG: The relationship between chronic lymphocytic leukaemia and prolymphocytic leukaemia. IV. Analysis of survival and prognostic features. Br J Haematol 65:23-29, Mintzer DM, Hauptman SP: Lymphosarcoma cell leukemia and other non-hodgkin's lymphomas in leukemic phase. Am J Med 75: , Freedman AS, Nadler LM: B cell development in chronic lymphocytic leukemia. Semin Hematol 24: , Dillman RO, Beauregard JC, Lea JW, Green MR, Sobol RE, Royston I: Chronic lymphocytic leukemia and other chronic lymphoid proliferations: surface marker phenotypes and clinical correlations. J Clin Oncol 1: , Deegan MJ, Abraham JP, Sawdyk M, Van Slyck EJ: High incidence of monoclonal proteins in the serum and urine of chronic lymphocytic leukemia patients. Blood 64: , Azar HA, Hill WT, Osserman EF: Malignant lymphoma and lymphatic leukemia associated with myeloma-type serum proteins. Am J Med 23: , Faguet GB, Agee JF: Monoclonal antibodies against the chronic lymphatic leukemia antigen cclla: characterization and reactivity. Blood 70: ,1987