Persistent Lymphocytosis of Natural Killer Cells After Splenectomy

British Journal of Haematology, 1995, 91, 253-258

Correspondence PERSISTENT LYMPHOCYTOSIS OF NATURAL KILLER CELLS AFTER SPLENECTOMY In their report on persistent lymphocytosis of natural killer

cells following splenectomy in three patients with autoimmune thrombocytopenic purpura (ATP). Garcia-Suarez et a1 (1995) consider the possibility of LGL lymphocytosis being the consequence of the asplenic state itself but lack data to support this assumption. Otherwise healthy asplenic patients are indeed the optimal model in whom to study the haematological consequences of the asplenic state. Large granular lymphocytosis has been found repeatedly in the peripheral blood of splenectomized individuals (Millard & Banerjee, 1979; Airo et al, 1985), but was studied systematically first by Kelemen et a1 (1986) in patients splenectomized because of splenic trauma and/or non-malignant haematological diseases, mainly idiopathic (autoimmune) thrombocytopenic purpura. The increase in the number of large granular lymphocytes (LGLs) was of similar magnitude in both groups of patients (Kelemen et al, 1986). The large granular lymphocytosis accompanied by a normal NK cell function could be confirmed in our study of post-traumatic splenectomy patients (Demeter et al, 1990). Others have found the number of Leu7' and Leull+ leucocytes and the natural killer cell cytotoxicity in the peripheral blood of post-traumatic splenectomy patients to be increased (Ferrante et al, 1986). These data show that the increase in the number of LGLs after splenectomy is not restricted to patients with ATP, but seems to be a more general phenomenon. ACKNOWLEDGMENTS J.D. is a scientific fellow of the Alexander von Humboldt Foundation, on leave from the First Department of Medicine, Semmelweis University Medical School, Budapest, Hungary. Department of Medicine and Haematology, J. DEMETER University of Ulm, Ulm, Germany

REFERENCES Aro, R.. Baldini, P., Ferrari, C.M., Neri, A. & Airo, P., Jr (1985) Persistent asymptomatic lymphocytosis with large granular lymphocyte predominance (NK phenotype). Haematologica. 70, 245-259. Demeter, J., Paloczi, K., Benczur, M. & Lehoczky, D. (1990) Observations on NK cells, K cells and their function a long time following post-traumatic splenectomy. International Archives of Allergy and Applied Immunology, 98, 287-292. Ferrante, A.. Kiroff, G.K. & Drew, P.A. (1986) Elevated natural killer(NK) cytotoxicity of mononuclear leukocytes from splenectomized patients: increase in Leu-7' and Leu-11' leucocytes. Clinical and Experimental Immunology, 64, 173-180.

0 1995 Blackwell Science Ltd

Garcia-Suarez, J., Prieto, A., Reyes, E., Arribalzaga. K.. PerezMachado, A.. Rubio-Lopez, M., Manzano, L. & Alvarez-Mon. M. (1995) Persistent lymphocytosis of natural killer cells in autoimmune thrombocytopenic purpura (ATP) patients after splenectomy. British Journal of Haematology, 89. 653-655. Kelemen, E., Gergely, P., Lehoczky, D., Triska, E., Demeter, J. & Vargha, P. (1986) Permanent large granular lymphocytosis in the peripheral blood of splenectomized individuals without concomitant increase of in vitro natural killer cell cytotoxicity. Clinical and Experimental Immunology, 63. 696-702. Millard, R.E. & Banerjee. D.K. (1979) Changes in T and B blood lymphocytes after splenectomy. Journal of Clinical Pathology, 32, 1405- 1049.

Lymphocyte phenotypical characteristics are actually the main parameter for the recognition of the different cellular subsets found in T cell and natural killer (NK) cell compartments. The morphological appearance of the large granular lymphocytes (LGL) is a nonspecific finding shared by both lymphocytic populations. Thus, an LGL expansion can be ascribed to either T or NK lymphocytes. Furthermore NK cells represent a heterogenous compartment that is presently best defined phenotypically as CD56+CD3lymphocytes. Several antigens such as CD2, CD8, CD11, CD16, CD56. etc., are expressed in both T and NK cells. In agreement with this, some authors have observed an expansion of morphologically defined LGL or of cells expressing the C D l l or CD16 antigens in peripheral blood mononuclear cells (PBMCs) from patients splenectomized because of splenic trauma or non-malignant haematological diseases (Kelemen et al, 1986; Ferrante et al, 1986; Demeter et al, 1990). It is important to emphasize that these phenotypical studies were performed as a single-colour analysis of PBMCs expression of CDll or CD16 antigens. Therefore it is not possible to conclude that splenectomy is associated with an expansion of NK cells in peripheral blood. Our data show that splenectomy in autoimmune thrombocytopenic purpura (ATP) is associated with a marked enhancement of CD56+CD3- NK cells and LGL. In previous work we found that ATP patients tend to accumulate CD56+CD3- NK cells in peripheral blood (Garcia-Suarez et al, 1995). It is possible that the removal of the spleen in ATP patients makes this abnormal NK cell compartment behaviour more apparent. We did not conclude that the NK cell alterations found in splenectomized ATP patients were due to the removal of this organ, as Demeter seems to have concluded. Therefore we stress that NK cell lymphocytosis has not yet been well studied in asplenic patients. Our report is a clear demonstration of the existence of NK cell lymphocytosis after

253

2 54

Correspondence

splenectomy in a non-malignant disease such as ATP. More well-performed immunophenotypic studies are needed in other asplenic states to better characterize the lymphocyte redistribution in peripheral blood after splenectomy. Servicio de Hematologia, J. GARCIA-SUAREZ A. PRIETO Servicio de Enfermedades del M. ALVAREZ-MON Sistema lnmune,

Hospital Universitario Principe de Asturias, Departmento de Medicina, Universidad de Alcala de Henares, Alcala de Henares, Spain Keywords: splenectomy, large granular lymphocytes.

ERYTHROPOIETIN I N THE MYELODYSPLASTIC SYNDROMES: META-ANALYTICAL STUDY We have read with interest the paper by Hellstrom-Lindberg (1995) concerning the efficacy of erythropoietin (EPO) in myelodysplastic syndromes (MDS). We have recently published a similar meta-analytical study (Rodriguez et al, 1994) including some of the series reviewed by HellstromLindberg among others. We accepted the response criteria indicated in each study since differences were slight. The statistical analysis was performed using the Mann-Whitney U-test for quantitative and the Chi-square test for qualitative parameters. Our study covered 115 patients, in 2 7 of whom a response was observed (23.5%). We analysed route of administration (s.c. versus i.v.), posology (daily, twice or three times a week), duration of the study, response according to initial diagnosis, and finally we compared the epidemiological and analytic parameters of responders and non-responders. The overall rate of response in our study (23.5%)was slightly higher than that reported by HellstromLindberg (16.1%).We found no significant difference in the response rate with respect to route of administration, posology or duration of treatment when patients were treated for less or more than 3 months, but this difference was significant when we considered 6 months as the cut-off point. However, this conclusion must be treated cautiously as only a few patients were treated for 6 months. According to initial diagnosis, 50 patients had refractory anaemia (RA), 38 RA with ringed sideroblasts (RARS), 18RA with excess of blasts (RAEB) and eight RAEB in transformation (RAEB-t), one patient had an unclassified MDS. The response rates in these different groups were 22%, 18.4% and 50%, respectively; no response was observed in the RAEB-t group and in the patient with the unclassified MDS no response was observed. Considering only RA, RARS and RAEB we found a significant difference in the response rate, this being significantly higher in the RAEB group than in the other two groups. No significant difference was observed between RA and RARS. Comparing the responder and nonresponder groups we found no significant difference in the response rate with respect to age, sex, dosage employed

(maximum dosage and dosage with which response was obtained), transfusional dependency and degree of transfusional dependency, basal serum EPO, time since diagnosis, transfusional period, haemoglobin level in the transfusiondependent patients and the non-transfusion-dependent patients. Considering as a cut-off point 250U/1 for basal serum EPO, we observed that 50% of responding patients had a level lower than this. In conclusion, we agree with Hellstrom-Lindberg that the number of responses to the treatment with EPO in patients with MDS is low; higher dosages of EPO do not correlate with a higher number of responses and 3 months of treatment seems to be enough to evaluate the possibility of response; no response was observed in the RAEB-t group; of the remaining groups RARS displayed the lowest number of responses (moreover, according to our results, patients with RAEB should be included in groups being treated with EPO). A maximum level of basal serum EPO of approximately 200-25OU/l might serve to identify patients who may respond. However, this factor alone does not ensure response.

Service of Haematology, JUANNICOLASR O D R ~ G L J E Z Hospital ‘JuanRambn Jimknez’ J O S E CARLOSDIEGUEZ 21005 Huelva, Spain D A L M I RPRADOS O REFERENCES Hellstrom-Lindberg, E. (1995) Ef!Ecacy of erythropoietin in the myelodysplasticsyndromes: a meta-analysis of 205 patients from 17 studies. British Iournul of Huematology, 89, 67-71. Rodriguez, J.N., Dieguez, J.C., Muni, R.. Martino, M.L., FernandezJurado. A., Amian, A., Cafiavate, M. & Prados, D. (1994) Eritropoyetina recombinante humana en el tratamiento de la anemia de 10s sindromes mielodisplasicos. Estudio Metaanalitico. Sungre (Barcelonu), 39, 435-439.

Keywords: myelodysplastic syndrome, erythropoietin.

FATAL HAEMOPTYSIS IN FILAMENTOUS MYCOSIS In a retrospective study over a 5-year period Pagano et a1 (1995) report 116 cases of filamentous mycosis (FM) in patients with acute leukaemia. 12 patients died from massive haemoptysis. In 11 of these patients the neutrophil count had recently recovered after a course of induction

chemotherapy. An in vivo diagnosis of FM was made in only four patients. The authors discuss the aetiology of haemoptysis in FM. When the bone marrow recovers, mycetomas which have formed during immunosuppression liquefy due to proteolytic

0 1995 Blackwell Science Ltd. British Journal of Huematology 91: 253-258