Singapore Med J 2000 Vol 41(6) : 279-285 Original Article Bone Marrow and Peripheral Blood Changes in Non-Hodgkin’s Lymphoma E J Lim, S C Peh ABSTRACT INTRODUCTION Background: 47 patients with non-Hodgkin’s lymphoma Changes in the full blood counts, peripheral blood (PB) (NHL) were studied retrospectively to determine differential white cell counts, as well as bone marrow their marrow and blood changes at diagnosis. involvement (MI) frequently occur in non-Hodgkin’s lymphoma (NHL). Examination of these tissues at the Methods: The blood counts, blood films, marrow time of diagnosis often provides useful information on smears, trephine and tissue biopsies of patients at the staging of the disease and its effects on the diagnosis were reviewed. The scheme proposed by haemopoietic system(1-3). Literature search yields limited the International Lymphoma Study Group (REAL) information on these haematological parameters in was utilised for lymphoma subclassification. Asian NHL patients. The purpose of this study is (1) to Results and Conclusion: 21.3% had lymphoblastic ascertain the frequency and pattern of MI, (2) to lymphoma, 21.3% had peripheral T-cell lymphoma determine the other accompanied changes in the (unspecified), 29.8% had diffuse large B-cell NHL, marrow such as the presence of lymphocytosis, 10.6% had Burkitt’s lymphoma and 17% had others. eosinophilia and increase in reticulin, and (3) to The incidences of anaemia, one or more abnormal determine the changes in the PB cell counts and counts, lymphocytopenia, increased marrow reticulin haematological parameters in the different subtypes of and marrow eosinophilia at diagnosis were 66%, NHL, in a group of Malaysian patients. 85.1%, 41.3%, 40.9% and 44.7% respectively. Marrow involvement was present in 46.8% of the patients, METHODS with diffuse infiltration noted in 71.4% of these cases. 47 patients who had been diagnosed to have NHL (based Abnormal counts and anaemia were common in all on tissue biopsies) between January 1993 and July 1995 the NHL subtypes. In lymphoblastic lymphoma, the at University Hospital Kuala Lumpur, and had available common haematological abnormalities were pretreatment tissue biopsies, marrow specimens and peripheral atypical lymphocytes and diffuse marrow full blood counts, were studied retrospectively. involvement. In peripheral T-cell lymphoma The full blood counts (consisting of haemoglobin (unspecified), common features were peripheral level, total white blood cell count and platelet count) of lymphocytopenia, increased marrow reticulin and all 47 patients at presentation were assessed for the eosinophilia. In diffuse large B-cell NHL, peripheral presence of anaemia, leucopenia or leucocytosis, and lymphocytopenia, peripheral myeloid precursors thrombocytopenia or thrombocytosis by comparison and/ or nucleated red cells and marrow involvement with the age-specific normal ranges in the 8th edition were common. In Burkitt’s lymphoma, diffuse of Practical Haematology by Dacie and Lewis. These Unit of Haematology marrow involvement and eosinophilia were parameters had been measured using a Coulter STKS Department of Pathology common. No significant differences were noted counter (Coulter Electronics, Miami, Florida). Faculty of Medicine between most of the haematological parameters of University of Malaya The PB film at diagnosis (retrieved in 46 patients) Kuala Lumpur 50603 B and T-NHLs. In comparison with other reports, we had been stained using the Leishman’s stain. A manual Malaysia recorded higher overall incidences of anaemia and differential count was performed on each patient’s blood E J Lim, MBBS, MPath Lecturer diffuse marrow involvement, and a lower incidence film by counting 200 white blood cells per film. of marrow infiltration in peripheral T-cell lymphoma S C Peh, MBBS, Lymphoma cells, atypical lymphocytes and myeloid FCRPath (unspecified). precursors were included in the differential counts. Associate Professor Examination for the presence of nucleated red cells was Correspondence to: Keywords: Non-Hodgkin’s lymphoma, marrow, E J Lim peripheral blood also carried out. The absolute counts of neutrophils, Tel: (603) 750 2583 lymphocytes, monocytes and eosinophils were calculated Fax: (603) 755 6845 Email: limej@ Singapore Med J 2000 Vol 41(6):279-285 using the manually obtained differential counts, and medicine.med.um.edu.my 280 : 2000 Vol 41(6) Singapore Med J the presence of neutropenia or neutrophilia, were performed on the trephine biopsies when deemed lymphocytopenia or lymphocytosis, monocytosis and necessary, to assist in the detection of MI. A standard eosinophilia were determined by comparing with the three-stage immunoperoxidase technique, following normal ranges outlined in the 8th edition of Practical antigen retrieval using the microwave was used. A panel Haematology by Dacie and Lewis. of antibodies against CD20 antigen (L26), CD30 Adequate marrow smears (retrieved in 42 patients) antigen (Ber-H2), CD61 (RUU-PL7F12, Becton had been obtained by unilateral aspiration of marrow Dickinson, CA) and CD3 antigen (polyclonal CD3) from the posterior superior iliac crest using Klima were utilised. needles. They were fixed using absolute methanol, The diagnostic tissue biopsy material for all the stained with May-Grunwald-Giemsa’s stain and patients studied was retrieved and reviewed by both evaluated by light microscopy for the presence of authors. The lesions were subclassified using the lymphoma cells. A differential count was also performed proposed list by the International Lymphoma Study by counting 500 nucleated cells. Marrow lymphocytosis Group (REAL). The slides were stained using as detected by examination of the marrow smear was haematoxylin and eosin. Immunohistological studies defined as the presence of more than 20% lymphocytes, were performed on 45 cases using the same staining and eosinophilia as the presence of more than 6% technique as mentioned above. A panel of antibodies eosinophils and eosinophilic precursors, in patients against the CD15 antigen (Leu-M1, Becton Dickinson, above 12 years of age(4). For patients who were 12 years CA), CD20 antigen (L26), CD30 antigen (Ber-H2), old or younger, the normal ranges used were as outlined CD43 antigen (DFT1), CD45 antigen (DAKO-LC), by Glaser K et al(5). CD45RO antigen (UCHLl) and CD3 antigen Adequate unilateral trephine biopsies (retrieved in (polyclonal CD3) were used. Unless otherwise stated, 46 patients) had been obtained using Jamshidi needles all antibodies used for the trephine and tissue biopsies from the same anatomical site as the aspirates, fixed in were from Dako, Denmark. 10% buffered formalin, decalcified using 5% formic acid The frequencies of each abnormal feature were and embedded in paraffin. 4 micron-thick sections were tabulated according to the NHL subtype. The data was stained with haematoxylin and eosin as well as silver analysed with non-parametric tests using the SPSS reticulin stain using the Gordon and Sweet method, and 8.0 programme. Analysis of correlation between the evaluated by both the authors. The presence of haematological findings and the different subtypes lymphoma infiltrates, increase in reticulin fibres as of NHL was not performed due to the small numbers well as other abnormalities were recorded. The pattern of patients in each subtype. of trephine involvement by lymphoma was classified according to guidelines given by Bain BJ(6). The presence RESULTS of aggregates of lymphocytic cells were regarded as Of 47 patients studied, 29 had B-NHL, 15 had T-NHL benign in nature if they were small, with cells which and 1 patient had non-B, non-T NHL. Tissue appeared mature, had a well-defined border and were immunophenotyping was not performed in 2 patients randomly distributed(7). Immunohistochemical stains due to the inadequacy of material. 10 (21.3%) had Table I. The changes in the full blood count in each subtype of NHL. NHL Anaemia TWBC* Platelet Abnormal Multiple subtype counts cytoperna Decr# Incr# Decr Incr LB 6/10 1/10 4/10 2/10 4/10 8/10 2/10 (60%) (10%) (40%) (20%) (40%) (80%) (20%) PTCL 7/10 2/10 3/10 4/10 0/10 9/10 4/10 (unspecified) (70%) (20%) (30%) (40%) (0%) (90%) (40%) BLCNHL 9/14 2/14 3/14 4/14 4/14 12/14 4/14 (64.3%) (14.3%) (21.4%) (28.6%) (28.6%) (85.7%) (28.6%) BL 4/5 0/5 1/5 1/5 2/5 5/5 1/5 (80%) (0%) (20%) (20%) (40%) (100%) (20%) Others 5/8 0/8 1/8 1/8 1/8 6/8 0/8 (62.5%) (0%) (12.5%) (12.5%) (12.5%) (75%) (0%) Total 31/47 5/47 12/47 12/47 11/47 40/47 11/47 (66%) (10.6%) (25.5%) (25.5%) (23.4%) (85.1%) (23.4%) *TWBC – total white blood cell count # Incr – increased; Decr - decreased Singapore Med J 2000 Vol 41(6) : 281 lymphoblastic lymphoma (LB), 10 (21.3%) had or younger, while the remaining 36 were more than peripheral T-cell lymphoma (unspecified) (PTCL-u), 12 years old. A significant difference was noted 14 (29.8%) had diffuse large B-cell lymphoma between the mean ages of patients in the different (BLCNHL) and 5 (10.6%) had Burkitt’s lymphoma subtypes of NHL (p=0.001, Kruskal Wallis test). (BL). Of the 10 patients with LB, 3 had B-LB, 5 had BLCNHL affected mainly older patients (mean age: T-LB while the remaining 2 cases were the ones without 53.7 years) while BL and LB affected mainly younger immunohistochemical staining. No significant difference patients. The mean ages of BL patients and LB patients was noted in the frequencies of each of these 4 main were 11.6 years and 19.6 years respectively. LB was subtypes (p=0.243, chi square test). Of the remaining 8 diagnosed in 6 of the 36 patients (16.7%) who were (17%) , 2 had B mantle cell NHL (BNHLMC), 2 had above 12 years of age and 4 of the 11 patients (36.4%) follicular lymphoma (FL), 1 had T-cell rich, B cell who were 12 years old or younger.