A unique case of splenic marginal zone-cell lymphoma with synchronous clonal T-cell large granular lymphocyte proliferation: an immunologic, immunohistochemical and genotypic study

1. Introduction 

Splenic marginal zone-cell lymphoma (SMZL) and T-cell large granular lymphocytic (LGL) leukemia are low-grade lymphoproliferative processes which result from clonal expansion of previously stimulated B- and T-cells [1]. MZL cells express surface and/or cytoplasmic monotypic immunoglobulin (Ig) and display rearrangement of the Ig heavy and light chain genes [1]. The cells of most T-LGL leukemias express characteristic T-cell differentiation antigens, such as, CD3 and CD8, and bear rearranged T-cell receptor (TCR) beta, gamma and delta chain genes [2]. In addition, LGLs express the human natural killer (CD57) marker; however, they usually display poor natural killer activity. Both SMZL and T-LGL leukemia may present with splenomegaly and neutropenia or pancytopenia [1], [2].

In the present study, we describe a patient with synchronous occurrence of SMZL and T-LGL leukemia in the spleen, liver, bone marrow (BM) and peripheral blood (PB). We report the results of detailed immunologic and molecular genetic analyses of the neoplastic cells; finally, we discuss the possibility of an eventual relationship between these two rare disorders co-existing in the same patient.

2. Case history 

A 53-year-old female was admitted to our hospital in July 1994 for anemia following a 2-month history of weakness and weight loss. Her past medical history was unremarkable. Physical examination revealed pallor and profound splenomegaly. Laboratory values were as follows: hematocrit: 22.0%; white blood cells 3.1×109l−1 (neutrophils 18%, lymphocytes 73%, monocytes 7%, eosinophils 2%); platelets 125.0×109l−1; erythrocyte sedimentation rate 60mm per first hour; liver function tests, serum glucose, creatinine, electrolytes, proteins and uric acid: normal; anti-nuclear, -mitochondrial and -DNA antibodies (Abs): negative; herpesviruses and human immunodeficiency virus Abs: negative; lactate dehydrogenase 334U/l (normal: <250U/l); serum IgG: 11.6g/l, IgA: 2.25g/l, IgM: 2.4g/l; serum electrophoresis: monoclonal IgM λ; direct antiglobulin test: negative. Chest X-ray and CT of the thorax were normal. U/S and CT scans of the abdomen showed profound splenomegaly (25cm), without enlargement of the abdominal lymph nodes. Upper gastrointestinal tract endoscopy: normal. Examination of PB smear revealed hypochromia and anisopoikilocytosis; the lymphocytes appeared small with well-condensed, coarse chromatin. BM aspiration demonstrated a 30% lymphocytic infiltration, consisting mainly of small lymphocytes; occasional large lymphocytes with prominent nucleoli were also observed. In August 1994, the patient underwent splenectomy; histological and immunohistochemical findings of spleen and liver biopsy specimens are shown in Table 1. Starting from September 1994, she was given eight courses of CHOP combination chemotherapy; the patient eventually attained complete remission and was off therapy by June 1995. However, in October 1995 the lymphoma relapsed; a second complete remission was attained by six courses of the IVPP protocol (idarubicin (10mg/m2 i.v., days 1–3), etoposide (100mg/m2 i.v., days 1–3) and prednisone (100mg p.o., days 1–7)) [3]. Seven years later, the patient remains in complete remission.

Table 1. Histological and immunohistochemical findings

	Morphology	Immunophenotype
	Spleen
	Extensive neoplastic small lymphocytic infiltration (SLI) of the red pulp	CD3+CD45RO+CD8+CD16+CD56−TIA-1+Granzyme B+CD20ass−
	Moderate expansion of the white pulp, suggestive of SMZL	CD20ass+CD79a+SIgM(λ)+CD5−CD23−CD10−CD3−
	Liver
	Extensive SLI of the sinusoids	CD3+CD45RO+CD8+CD16+CD56−TIA-1+Granzyme B+CD20ass−
	Absence of portal tract neoplastic infiltration	Scattered small B (CD20ass+CD3−) cells
	BMBa
	15% SLI and focal and interstitial growth	Focal: CD20+CD79a+CD3−
		Interstitial: CD3+CD57+CD8+CD4−CD20ass−

a Obtained before splenectomy.

3. Study design 

4. Results 

5. Discussion 

T-LGL leukemia has been reported to co-exist with various B-lymphoproliferative disorders [2]. The evidence presented herein suggests that in our patient two separate entities grew simultaneously in the same anatomic sites; their distribution in the tissues examined was immunophenotypically and genotypically discriminate. The morphologic findings in the spleen were indicative but not diagnostic of SMZL; immunohistochemical analysis confirmed the neoplastic nature of the cells of the expanded marginal zone and, furthermore, revealed a minimal B-small lymphocytic infiltration (SLI) of the splenic red pulp and hepatic sinuses as well as a mild B-SLI of the BM with a nodular growth pattern. Nevertheless, evidence for the neoplastic nature of the B-lymphocytic infiltration of the liver and the BMB was only obtained by PCR which identified identical clonal Ig gene rearrangements in the all tissue samples (including PB).

The other cell clone consisted of small to medium sized lymphocytes which extensively infiltrated the splenic red pulp cords and sinuses and the hepatic sinuses. The immunophenotype of the neoplastic cells was compatible with T-LGL leukemia. The clonality of the limited T-SLI detected in the BM could not be determined solely on immunophenotype, but was shown to be monoclonal by the PCR. The exact nature of the T-cell lymphoproliferation observed in our patient could not be ascertained on the basis of PB morphology because cells lacked typical LGL morphology. However, occasionally, clonally expanded lymphocytes with a characteristic CD3+CD57+ phenotype may not have LGL morphology on PB smear [2]. With this in mind, and given the highly characteristic immunophenotypical profile and genotype, the patient was diagnosed as having T-LGL leukemia. T-LGL leukemia has striking clinicopathologic, immunophenotypic and genotypic similarities to S-100+ T-cell lymphoproliferative disorder (T-LPD) and hepatosplenic γδ T-cell lymphoma: S-100+ T-LPD was easily excluded because the neoplastic cells were negative for the β fraction of the S-100 protein. Hepatosplenic γδ T-cell lymphoma was also excluded on both clinical grounds and immunophenotypical findings (in our case the cells were CD8+, CD56−, δTCR-1−).

The low frequency both of SMZL and T-LGL leukemia (1.5% of chronic lymphocytic leukemias) suggests that their simultaneous occurrence might not be fortuitous. The possibility that the T-LGL population detected in our patient might react to the SMZL seems unlikely because of the massive neoplastic infiltation of the spleen and liver by T-LGL and the documented clonality of this population. Alternatively, one could argue that the two entities might share a common malignant precursor or that B-non-Hodgkin’s lymphoma emerged as a secondary phenomenon due to anomalous B-cell function in the setting of T-LGL leukemia; however, the cytotoxic/suppressor phenotype of the neoplastic T-cells of our patient does not support the latter explanation. Normally, cytotoxic/suppressor T-cells down-regulate Ig production by activating B-cells; nevertheless, other studies indicate that, in most cases, unlike LGLs from healthy individuals, the abnormally expanded T-LGL leukemia cells (despite CD8 positivity) instead of suppressing Ig synthesis in vivo [2], [8] function as contrasuppressors promoting Ig production. In this context, one could envisage an uncontrolled Ig production with autoantibody formation that could eventually lead to the development of clonal B-cell disorders, as in the few cases cited in the literature and in our case as well [2], [9]. In conclusion, our case gives further support for the notion that granular CD3+, CD8+, CD57+ lymphocyte proliferation is associated with B-cell dysfunction; the exact pathophysiologic mechanism awaits further clarification.