Identification of blast cells in peripheral blood through automatic assessment of nuclear density: a new tool for monitoring patients with acute leukaemia

Identification of blast cells in peripheral blood through automatic assessment of nuclear density: a new tool for monitoring patients with acute leukaemia
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  British ]ournal of Haematologn, 1987. 66, 473477 zyxwvusr Identification of blast cells in peripheral blood through automatic assessment of nuclear density: a new tool for monitoring patients with acute leukaemia G. D’ONOFRIO, . MANCINI, . LEONE, B. BIZZI AND G. MANGO ervizio di Ematologia, Zstituto di Semeiotica Medica, Universita’ Cattolica del Sacro Cuore, Roma, Ztaly Received 1 April 1986: acceptedfor publication 28 Ianuary 1987 Summary. The Technicon H*l haematology system is pro- vided with a new method for basophil count, neutrophil lobularity assessment and detection of blasts in peripheral blood through automated measurement of nuclear density. We compared theresults of the H* 1 blast flag with those of the microscope examination in 131 peripheral blood samples from 43 patients with acute leukaemia in different phases of their disease, to determine the degree of sensitivity and specificity of the system in this setting. In zyxwvu ix patients at diagnosis or zyxwvuts n overt relapse, all having large percentages of blasts at the manual differential count, a typical deformation of the profile of the mononuclear cell population on the display was consistently observed, regardless of the morpho- logical subtype of leukaemic cell which was involved. Amongst 34 samples with 4-9 5 morphologically recogniz- able blasts on the peripheral blood film, the sensitivity of the system was 100 . with no false negatives at all. In 43 samples, on the other hand, the H* 1 blast flag was positive in the absence of any morphological evidence of blasts on the smear. These ‘false positives’, however, were always obtained from leucopenic patients who had more than 12 blast infiltration in the bone marrow, compared to the 0-6 value that was found in patients with a negative H*l blast flag. These results suggest that the H*l system is a highly sensitive tool for the detection in peripheral blood of even small concentration of leukaemic cells, which escape morphologi- cal identification. Haematological analysers based on the principle of auto- mated cytochemistry have demonstrated increased precision for detection of blast cells in patients with acute leukaemia (AL) under treatment (Avnstrom et al, 1986: Ross zyxwvu   Bardwell, 1980: Winkel zyxwvuts t al, 1982). The new Technicon H*1 haematological analyser has the capability, according to the manufacturer, of specifically identifying even small numbers of blast cells circulating in the peripheral blood, utilizing a new method for the evaluation of nuclear structure density. We report here the results of a study carried out to assess the sensitivity of the blast flag of the H*1 system in patients with acute leukaemia during the different phases of their disease. PATIENTS AND METHODS Analysis ojperipheraf blood by the Technicon HI1 system The Technicon H*l system is a recently developed automatic haematologic analyser which provides, in 60 zyxwvu , complete blood cell counts and leucocyte differential on 100 p1 Correspondence: Dr Giuseppe d‘onofrio. Servizio di Ematologia, Universita’ Cattolica del Sacro Cuore, Largo Agostiio Gemelli 8, I 00168 Roma. Italy. peripheral blood samples (Technicon H*l System Operators’ Guide, 1985). After red blood cell lysis and cytochemical staining for peroxidase, in particular, leucocytes are classified according to their absorbance and forward scatter properties, measured in a tungsten-based optical system. Thus three populations of cells with increasing peroxidase activity (monocytes, neutrophils and eosinophils) and two popula- tions of unstained cells (lymphocytes and large unstained cells) are differentiated and quantitated. The principle of the H*l peroxidase method is the same that of the other Technicon cytochemical systems (Hemalog D and H6000). the main differences being a shorter time of reaction (33 s instead of 10 min) and computer cluster analysis generating floating thresholds for each blood sample. The basophil count, on the other hand, is carried out by the H*l system in a Merent, laser-based, analytical channel (‘baso/lobularity method’). Here, the addition of a specific reagent (Technicon product no. TO1-1647-56), containing phtalic acid 21.49 mmol and surfactant, causes haemolysis of the erythrocytes, as well as stripping of the cytoplasm from all leucocytes, except basophils. Analysis by a two-angle laser light scatter- ing system (Tycko et aJ. 1985) then provides a leucocyte 473  474 G. d’Onofrio et a2   zyxwvu . .........,......... .........I....... ..I......... L 1: :- C . I ........ l.........~......... r k . ’I ........I.........,.........l.........l......... r z ig 1. The H’1 baso/lobularity display. (A) Normal peripheral blood (bg: basophil granulocytes: mn: mononuclear cells: pmn: polymorphonuclear cells). (B) Acute monoblastic leukaemia zyxwvu ANLGMS), ith 75 blasts in the peripheral blood film bc last cells). (C) Chronic lymphocytic leukaernia, with 82 lymphocytes in the peripheral blood Elm. Note the deformation of the mononuclear population profle in (B), ue to the presence of a large number of blast cells with hypodense nuclel. In chronic lymphocytic leukaemia (C). on the contrary, the mononuclear population is thickened, but not distorted. by the maturelooking leukaemic lymphocytes with dense nuclear structure. classification into three categories: basophils, mononuclear cells and polymorphonuclear cells (Fig 1A). In addition, this ‘baso/lobularity method’ also provides an index of neutrophil lobularity (by the ratio of the signals from the two nuclear classes) and specific flags concerning ‘left shift’, immature granulocytes, atypical lymphocytes, erythroblasts and. finally, blast cells. The blast flag of the sgstem As stated above, the Technicon H*1 ‘baso/lobularity method’ produces selective rupture of cell membranes (except baso- phil) with stripping of most of their cytoplasm, but retaining of their nuclear structure. When analysed in the two-angle laser optic, the loose chromatin structure of their nuclei permits the identification of blast cells, as a population merging on the extreme left of the mononuclear cell popula- tion (Fig 1B). Blast cell quantitation is performed through a hidden threshold, separating them from mononuclear cells, and fixed at channel no. 8. When the percentage of ‘signals’ beyond this threshold exceeds 5.0 n a given sample, the H*1 software generates a flag I+’) for the presence of blast cells. Patients with acute leukaemia Over a 3-week period, we were able to study 131 EIYTA- collected peripheral blood samples obtained from 43 adult patients with AL during different phases of their disease. Four patients (1 samples) were studied at presentation. and two (six samples) were in overt relapse. The remaining ones were observed during induction, reinduction or maintenance treatment (33 patients with 102 samples) with conventional drugs, dosages and schedules. Eight samples, finally, were obtained from four patients who were ‘off therapy’ after long- lasting complete remission. In all cases diagnosis and classifi- cation of zyxwvuts L had been based on morphological examination of blood and bone marrow films with cytochemical confir- mation according to the FAB group suggestions (Bennett et al, 1976). Sample management ll 13 1 peripheral blood samples were both zy un hrough the H* and examined morphologically on May-Griinwald- Giemsa-stained films by three different observers (GdO, SM, GM). For the purpose of this study, each blood film was scanned thoroughly, counting at least 100 cells even in the most leucopenic patients, and classified as positive or nega- tive with regard to the presence of morphologically recogniz- able blast cells. Furthermore, positivity or negativity of the H*l blast flag was recorded. Bone marrow analpis To evaluate the significance of the H*l ‘false positive’ blast flag, we decided to determine the percentages ofblasts in bone marrow of patients in which no blasts were recorded at the microscope examination of the peripheral blood smear. Bone marrow aspiration data, including percentages of bone marrow blast cells, were available from clinical records concerning the same period of our study for all of the 33 patients who had no blasts in the peripheral blood at the morphological examination. In three patients. however, the bone marrow cellularity was too poor for a correct assess- ment of the blast cell percentage. The remaining 30 patients were considered evaluable in terms of marrow residual leukaemic cell percentages. RESULTS A general profile of our 43 patients, subdivided according to the FAB classification of AL and to the particular phase of their clinical course, is given in Tables I and 11 Group A (‘onset/relapse’) We were able to observe four patients at the onset of their disease (one with ANLL-M1, one with ANLGM3, one with ANLGM5 and one with ALLL2). as well as two subjects (one with ANLL-M5 and one with ALL-L2) with overt leukaemic  Automatic Zdent cation of Blast Cells 475 Table I. Classification of patients according to FAB cytological subtyping and phase of the disease. Digits in parentheses indicate the number of peripheral blood samples obtained &om each group of patients. FAB classification zyxwvut  ~ ANLL’. n=27 (81) zyxwv t =15 (SO) Phase of Group thedisease M1 M2 M3 M4 M5 L1 L2 L3 Total zy   Onset/relapse 1 (3) 1(4) 2 (8) 2 (6) 6 (21) B Treatment 3(10) 6(16) 3 (8) 6(21) 3(11) 1 (3) lO(29) 1 (4) 33(102) C Offtherapy 2 (4) 2 (4) 4 (8) Total 4(13) 8(20) 4(12) 6(21) S(19) 1 (3) 14(39) 1 (4) 43(131) * ANLL: Acute non-lymphocytic leukaemias. t ALL: Acute lymphocytic leukaemias. Table 11. Comparison of results of the H*l blast flag versus morphological analysis in terms of sensitivity for blast cell detection in1 3 1 peripheral blood samples obtained from 43 patients with acute leukaemia Detection of blast cells (microscope/H*l blast flag) Phase of Group the disease +/+ /- -/+ -/- Total A Onset/relapse 21 0 0 0 21 B Treatment 13 0 43 46 102 C Offtherapy 0 0 088 Total 34 zyxwvutsrq t 43 54 131 t False negatives. False positives. relapse, with large numbers of leukaemic cells in peripheral blood. Leucocyte counts in peripheral blood samples obtained from patients from this group were generally high, from 12.3 x 10y/l o 126 x 10y/l, he only exception being three samples obtained from a patient with ANLLM3, who was leucopenic (1.85-2.69 x 10y/l), with 50-70 of abnormal promyelocytes on the peripheral blood films. In each of these patients with high circulating blast count, a typical appear- ance was consistently found on the x-y display of the baso/ lobularity channel of the HI1 system, with a clear downward shift to the left of the mononuclear population profile. This pattern arose in every sample from this group, regardless of the different nature of the leukaemic cells (myeloid undiffer- entiated, monoblastic. promyelocytic and lymphoblastic) (Fig 2). In all of the 2 I peripheral blood samples obtained from these patients, the H*l blast flag, obviously, was positive. Group B (‘treatment ) One-hundred-and-two samples were obtained from 3 3 patients receiving aggressive polychemotherapy as induc- tion, reinduction or maintenance treatment for acute leukae- mia. Blast cells were found on peripheral blood films in 13 samples from this group: the H*l blast flag in such cases was also always positive. Forty-three samples, on the other hand, had been flagged by the H*l in the absence of morphologi- cally recognizable blasts on the smear. A common finding shared by these false positives was moderate-to-severe leuco- penia (0.28-3.46 x 109/1, with a mean value of 1.85 x lo9/ 1 . In the remaining 46 samples from this group, finally, both manual and automated examinations were negative regard- ing the presence of circulating blast cells. Group C (‘off therapy’) In the eight samples obtained from four patients in complete remission, in which maintenance chemotherapy had been discontinued for 4-2 7 months, both morphological smear examination and H*l blast flag resulted negative. Bone marrow comparison study Evaluable bone marrow aspiration results, available on clinical records for 30 group B patients, were analysed with particular regard to the number of residual leukaemic cells in the bone marrow (Fig 3). In the 19 patients with a negative H*l blast flag the percentage of leukaemic blasts found in the bone marrow ranged from 0 to 6 ; that is, all of them were under complete remission, or were probably going to attain it. On the contrary, in the 11 patients with discrepant results the percentage of morphologically identifiable blasts in the bone marrow was never below 12 , ranging from 12 to 70 , with a mean value of 35.6 . DISCUSSION The presence of occasional circulating blasts in the blood film of patients with AL in haematological remission has been recently correlated with a high risk of early relapse (Dharma- sena zyxw   Galton, 1986). The purpose of the investigation reported in this paper was to examine the clinical usefulness of the H*l blast flag as a mean for detecting blast cells in peripheral blood ofpatients with acute leukaemia. Our results can be summarized as follows. The presence of large number of leukaemic blasts in peripheral blood samples, in patients at diagnosis or at the  476 G. d Onofrio et a1 .I ................................................. .................... .............................. b 7 ......... .........I ......... ........ ......... i ......... I ......... ......... ......... . ................................................. ;c : 1. .. : I .. i I I l .......... ......... ............................. r zyxw N = 9 ooo ,~oo 0 zyxwvutsrqpo aooo (( - ) % BLAST CELLS IN BONE MARROW Hi ((BLAST)) NEGATIVE N. = 19 MEAN: 3.8 RANGE: 0.6 H1 ccBLAST)) FALSE POSITIVE N. = 11 MEAN: 35.6 RANGE 12-70°o zyxwvuts H I <(BLAST>, LAG (PERIPHERAL BLOOD) tie of relapse, produces a typical deformation downwards and to the left of the profile of the mononuclear cell population on the display of the H*1 baso/lobularity channel, as well as a positive blast flag, regardless of the FA3 subtype of the leukaemic population. Thus, a unique property, shared Pig 2. The H*l baso/lobularity display in four cases of acute leukaemia at diagnosis. (A) ALL z   leucocyte count 72.6 x 109/1. with 74 lymphoblasts at the manual Werential count). (B) ANLGMl leucocyte count 44.1 x 109/1, with 82 poorly Werentiated myeloblasts). (C) ANLLM3 (leucocyte count 2.48 x 109/l. with 66 heavily granulated promyelocytes). (D) ANLGM5 (leucocyte count 12.7 x 109/1, with 59 monoblasts). Note the consistent deformation of the mononuclear population profile. in comparison to the normal pattern Seen in Fig A. In (A) and (B) the upward signal shift across the horizontal basophil threshold reflects a phenomenon of coincidence, due to the very hlgh leucocyte counts. Fig 3. Percentages of blast cells in the bone marrow aspirates from 30 patlents with no morphologically recognizable blasts on the peripheral blood smears, versus positivity or negativity of the W 1 blast flag. by every type of leukaemic cell, probably residing in the particular structure of their nuclear matrlx. is indeed what the system recognizes as 'blastic'. The sensitivity of the H*l blast flag, over a wide range of blood concentrations, was 100 . with no false negatives amongst 34 samples having from 4 to 95 blast cells  Automatic Identification of Blast Cells 477 according to the morphological examination. The H*l blast flag, on the other hand, apparently lacked specificity, because of 32 spuriously positive results in leukaemic patients with no evidence of blast cells on the peripheral blood smear. False positives, however, were especially common with leucopenic blood samples, raising the possibility of an increased sensiti- vity of the H*l system for the detection of very small amounts of circulating blast cells as an explanation for such discrepan- cies. This hypothesis is supported by the results of bone marrow comparison study. All of the 19 patients with available bone marrow data and negative blast flag, in fact, were in complete remission (bone marrow blast from 0 to 6%), whereas in all the 11 patients with zyxwvut n apparently false positive blast flag, heavier marrow infiltration was consistently found (marrow blasts from 12 to 70 ). Thus, the H*l blast flag, in our series, has demonstrated great accuracy and precision in detecting blood samples containing even very small percentages of leukaemic cells, showing a better correlation with the real status of disease, as defined by morphological examination of bone marrow, than the conventional microscope examination of peripheral blood films. ACKNOWLEDGMENTS This work could not have been carried out without the skilled technical assistance of Mr Federico Di Giacomo and Mr Basilio Caruso. and the advice and support of Dr Crescenzio Izzo. REFERENCES Avnstrom, S.. Raltkiar, E.. Winkel, P. zyx   Nissen. N.I. (1985) The relative merit of various cytochemical quantities and manual differential count in predicting remission following chemotherapy of patients with de novo acute myeloblastic leukemia. American Journal zyxwv j Clinical Pathology. 83, 73-76. Bennett, J.M., Catowsky, D.. Daniel, M.T.. Flandrin, G.. Calton, D.A.C., Gralnick. H.R. Sultan. C. (1976) Proposals for the classillcation of the acute leukaemias. British Journal oj Haernato- logy. 33, 451-458. Dharmasena, F. Calton. D.A.G. (1986) Circulating blasts in acute myeloid leukaemia. (Letter). British Journal oj Haernatology, 63, Hinchlie. R.F.. Lilleyman, J.S.. Burrows. N.F. &Swan, H.T. (1981) Use of the Hemalog D automated leukocyte differential counter in the diagnosis and therapy of leukaemia. Acta Haematologica. 65, Ross, D.W. Bardwell, A. (1980) Automated cytochemistry and the white cell differential in leukemia. Blood Cells. 6, 455-470. Technicon H*l System: Operators’ Guide (1985) Technicon Instru- ments Corporation, Tarrytown. New York. Tycko, D.H.. Metz, M.H.. Epstein. E.A. Grinbaum, A. (1985) FIow- cytometric light scattering measurement of red blood cell volume and hemoglobin concentration. Applied Optics. 24, 13 55-1 365. Winkel. P.. Olesen. T. Nissen. N.I. (1982) Automated cyto- chemistry in the prediction of remission following chemotherapy of patients with novo acute myeloblastic leukemia. American Iournal oj Clinical Pathology, 77, 50-53. 21 1-213. 79-84.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks