daunorubicin; ArC = cystosine arabinoside; MOPP: Mustard, Oncovin, Prednisone, Procar- bazine; CYT = cyclophosphamide; ABVD = Adriamycin, Bleomycin, ...
HEMATOPATHOLOGY Original Article
The Prognostic Significance of Proerythroblasts in Acute Erythroleukemia ARETA KOWAL-VERN, M.D.,1 JAMES COTELINGAM, M.D.,2 AND HAROLD R. SCHUMACHER, M.D.3
oxidase staining was negative. The malignant proerythroblasts had increased block and blush periodic acid-Schiff positivity. The most frequent chromosomal abnormalities involved chromosomes 5 and 7. The patients with DiGuglielmo's disease had a median survival time of 2 months (range, 0.06 to 9 months), compared to a median of 16 months (range, 2 to 48 months), in patients with DiGuglielmo's syndrome. The erythroleukemia with the preponderance of proerythroblasts had a worse prognosis because many of the individuals did not survive long enough to respond to the therapy initiated. Erythroleukemia with 30% or more proerythroblasts should be included in the French-American-British classification because it behaves clinically and appears morphologically as an acute leukemia rather than a myelodysplastic syndrome. (Key words: Proerythroblasts; Acute erythroleukemia; Acute leukemia; Di Guiglielmo) Am J Clin Pathol 1992; 98:34-40
Erythroleukemia is a heterogeneous disorder that can have an excess of myeloblasts or proerythroblasts in the setting of dyserythropoiesis. The French-American-British classification, established in 1976 and subsequently revised, allows only for the diagnosis of erythroleukemias whose immature elements are predominated by myeloid blasts, previously described as DiGuglielmo's syndrome. However, there is another form of erythroleukemia, in which the predominant immature elements are proerythroblasts, called DiGuglielmo's disease. To clarify this issue, 23 cases of erythroleukemia were reviewed and classified: 10 with a myeloblast predominance and 13 with a proerythroblastic predominance. These two forms of erythroleukemia can be distinguished on the basis of quantitative and qualitative morphologic features. When there were 30% or more proerythroblasts, calculated by dividing the total erythroid component into the proerythroblasts, there were few to no myeloblasts, no Auer rods, and increased cytoplasmic vacuoles; and myeloper-
originally in 1917, and DiGuglielmo's disease, described in 1926.9"" The blasts in DiGuglielmo disease are abnormal proerythroblasts, in contrast to DiGuglielmo syndrome, in which the blasts are of myeloblasts or monoblastic origin. Since DiGuglielmo's original description of erythroleukemia, many reports have noted that the two conditions may coexist within the spectrum of a malignant erythroid condition. 36 " 8 Dameshek12 and Dameshek and Baldini13 theorized three sequential phases in a spectrum of disease rather than two separate disorders. These were erythroblastic proliferation (erythremia myelosis) progressing to a erythroblastic-myeloblastic phase (erythroleukemia) and terminating as acute myeloblasts leukemia. Many other attempts, reported in the literature, have From the Departments of Pathology, 'Loyola University Medical Cen-unsuccessful in resolving the question of heterogebeen ter. Maywood. J Illinois, 2Nalional Naval Medical Center, Bethesda, neity of the erythroleukemias.14"20 In fact, however, this Maryland, and University of Cincinnati, Cincinnati, Ohio. erythroblastic-myeloblastic leukemia (DiGuglielmo synThe opinions and assertions expressed herein are those of the authors drome) with myeloblasts predominance is the only form and not necessarily those of the United States Navy or Department of of erythroleukemia that can be diagnosed by the current Defense. French-American-British (FAB) criteria.' Because the Received July 24, 1991; received revised manuscript and accepted for FAB classification does not provide a category for Dipublication November 1, 1991. Address reprint requests to Dr. Kowal-Vern: Department of Pathology, Guglielmo disease, a dilemma exists as to whether one Loyola University Medical Center, 2160 South First Avenue, Maywood, should include a fatal acute leukemic disorder in a myIllinois 60153. Erythroleukemia constitutes approximately 3% of all cases of acute leukemia and is very rare in children.1"3 It presents with increased leukemic erythropoiesis, which has both megaloblastoid and megaloblastic features, multinucleation, immature vacuolated erythroid and myeloid precursors, dyserythropoiesis, dysmyelopoiesis, and cytoplasmic budding. These morphologic changes are seen not infrequently in other conditions, such as B, 2 or folate deficiency, and other neoplasms.4"8 In DiGuglielmo's treatise on erythroleukemia (compiled by his sons in 1962 after his death), he documented in detail the entities that have been identified as DiGuglielmo syndrome, described
34
35
KOWAL-VERN, COTELINGHAM, AND SCHUMACHER Prognostic Significance of Proerythroblasts in Acute Erythroleukemia elodysplastic syndrome category; and workshops on leukemia frequently ignore its existence.21 The purpose of this study is to evaluate critically the significance of proerythroblasts in acute erythroleukemia. MATERIALS AND METHODS Selection of Patients Fourteen cases of acute erythroleukemia diagnosed between September 1986 and July 1990 were identified in the files of Loyola University Medical Center, Maywood, Illinois. Nine cases of acute erythroleukemia diagnosed between August 1978 and June 1986 were collected from the files of the Bethesda Naval Medical Center in Bethesda, Maryland. This time frame was selected because the diagnosis was established by one of the authors (H.R.S.) who initially evaluated all cases. One case was previously reported.22 Peripheral Blood and Bone Marrow Evaluation The peripheral blood, bone marrow aspirate smears, and touch preparations were stained with May-GriinwaldGiemsa. The bone marrow clot and biopsy specimens were fixed in B5 and stained with hematoxylin and eosin by standard methods. Five hundred cells were counted on May-Griinwald-Giemsa smears to calculate the percentage of cell types present. Two hundred cells were counted to establish the percentage of ringed sideroblasts on the Prussian blue smears. Results were stated as median value ± 1 standard deviation. Quantitative Calculations Used to Evaluate Proerythroblasts In the current revised FAB classification of October 1985 for M6, the percentage of erythroid cells must be 50% or more of all nucleated cells.2 Also, the percentage
of the nonerythroid blasts calculated by excluding the erythroid population must be 30% or more (the present M6 classification). If the nonerythroid blast cells are less than 30% of the nonerythroid component, the case becomes a myelodysplastic syndrome. These criteria incorporate DiGuglielmo syndrome, but not DiGuglielmo disease, which is a leukemia and not a myelodysplastic syndrome. The erythroid component was evaluated by calculating the percentage of proerythroblasts by exclusively using the erythroid component. For example, a total erythroid population of 60% with 20% proerythroblasts would result in 20 of 60 (33.3%) proerythroblasts. For the purposes of this study, cases with 50% or more erythroid components that have myeloblasts 30% or more of the nonerythroid component are designated M6A, and those cases that have proerythroblasts 30% or more of the erythroid component are designated M6B (Fig. 3). Cytochemical analysis and terminal deoxynucleotidyl transferase. Special cytochemical stains of the bone marrow aspirate included alpha naphthyl acetate esterase (AEST), alpha naphthyl butyrate esterase (B-EST), periodic acid-Schiff (PAS), Leder's chloracetic acetate esterase (CAE), combined alpha-naphthyl esterase (C-EST composed of A-ESt and CAE), Sudan Black B (SBB), and Prussian blue.23 The myeloperoxidase stain (MPEX) was performed according to the new technique of Doe24 or standard benzidine methods. The stains were considered positive if 3% or more of the myeloblasts or monoblasts were stained (MPEX, A-EST, B-EST, CAE, SBB); and 3% or more proerythroblasts were positive for PAS. The terminal deoxynucleotidyl transferase (Molecular Genetic Resources, Tampa, FL) was performed according to the method of Stass and colleagues25 (Figs. 1 and 2). Cytogenetic analysis. Cytogenetic analysis was performed on bone marrow aspirates, biopsy samples, or peripheral blood specimens obtained at the time of diagnosis. Metaphase cells prepared from short-term (24 or 48 hours)
TABLE 1. ACUTE ERYTHROLEUKEMIA CHARACTERISTICS OF PATIENTS WITH THE M6A GROUP Patient 1 2 3 4 5 6 7 8 9 10
Age/Sex
PMH
Therapy
69/M 45/M 63/M 35/M 66/M 25/M 57/F 72/F 63/F 26/F
NC Multiple myeloma NC Hodgkin's RAEB NC NC NC NC RAEB
D; ArC VCR; PRED; Alkeran. D; ArC D; ArC ABVD, RAD; MOPP. Cyt, Bus; ABMT D;ArC D; ArC D;ArC D;ArC D; ArC; 6-TG D;ArC
M6A = 2:30% myeloblasts; M6B = 2:30% proerythroblasts; M = male; F = female; PMH = past medical history; NC = noncontributory; RAEB = refractory anemia with excess blasts; D = daunorubicin; ArC = cystosine arabinoside; MOPP: Mustard, Oncovin, Prednisone, Procar-
Survival (months) 3 24* 24* 4 24* 48 4* 6
— 13
bazine; CYT = cyclophosphamide; ABVD = Adriamycin, Bleomycin, Vincristine, DTIC; Rad = radiation; Bus = Busulfan; ABMT = autologous bone marrow transplant; *alive and well. Age is expressed in years.
Vol. 9 8 - N o . 1
36
HEMATOPATHOLOGY Original Article TABLE 2. ACUTE ERYTHROLEUKEMIA CHARACTERISTICS OF PATIENTS WITH THE M6B GROUP
Patient
Age/Sex
1 2 3 4 5 6 7 8 9 10 11 12 13
65/F 68/M 32/M 79/M 72/M 72/M 26/M 36/F 4.5/F 20/M 64/M 82/M -/F
Survival (months)
PMH
Therapy
NC NC Bloom's NC NC NC NC Hodgkin's ITP RAEB NC RAEB NC
D; ArC D;ArC D; ArC None None D;ArC D;ArC COPP, RAD; none for M6B PRED, VCR, L-asp, 5AZA, ADR, CYT, ArC D;ArC None D;ArC
3 2 1.5 1 2 days 1 4 1 9
— 1 5
—
M6A = £30% myeloblasts; M6B = £30% proerythroblasts: M = male; F = female; PMH = past medical history: NC = noncontributory; RAEB = refractory anemia with excess blasts; D = daunorubicin; ArC = cytosine arabinoside; COPP = Cytoxan. Oncovin, prednisone, procarbazine;
unstimulated cultures of bone marrow or peripheral blood cells were stained using a trypsin-Giemsa banding technique.26 Chromosomal abnormalities were defined according to the International System for Cytogenetic Nomenclature and were analyzed at the University of Chicago in Illinois.27 Immunophenotyping. Surface marker analyses at diagnosis were performed using Ficoll-Hypaque-separated blood and bone marrow cells. These cells were stained by direct and indirect techniques, using monoclonal antibodies directed to erythroid: glycophorin A (Amac, Inc., Westbrook, ME); transferrin receptor (CD71) and CD45 (Becton-Dickinson, Mountain View, CA); and myeloid cells (CD13, CD33, CDllb) (Coulter Immunology, Hialeah, FL); and CD15 (Becton-Dickinson). The cells were
— Rad = radiation: PRED = prednisone. L-asp = L-asparaginase; 5-AZA = Adriamycin; 6-TG: 6-thioguanine; *alive and well. Age is expressed in years.
;
5-azacytidine. ADR
analyzed using aflowcytometer (FACScan, Becton Dickinson). Cell populations with light scatter characteristics of myeloid and erythroid cells were observed, analyzed by one-color fluorescence histograms, and displayed on a logarithmic scale. The results were reported as a percentage of total gated population and were considered positive if 30% or more of the cells stained with the specific surface markers. Gene rearrangement. High-molecular-weight DNA was prepared from mononuclear cells isolated by Ficoll-Hypaque density centrifugation. Methods and probes used in Southern blot analyses to detect clonal heavy and light chain immunoglobulins and T-cell gamma and beta antigen receptor gene rearrangements were performed as previously described.28
TABLE 3. PERIPHERAL BLOOD PARAMETERS Parameters median ±1 SD (range) WBC lO'/L HGB g/L I2
RBC 10 /L HCT I Reticulocytes % Platelets 109/L PB blasts % Nucleated RBC %
M6A
M6B
5.6 ± 7.0 (0.8-7.9) 92 ± 14 (50-100) 2.8 ± 0.6 (2.5-3.0) 0.27 ± 0.04 (0.15-0.31) 1.1 ±0.90 (0.20-2.6) 49.0 ± 27.0 (20.0-96.0) 1.0 ± 12.0 (0-32.0) 10.0 ± 8.0 (1-23.0)
7.2 ± 7.3 (1.9-21.0) 84 ± 3 0 (38-118) 3.2 ±0.9 (1.4-4.2) 0.26 ± 0.08 (0.09-0.35) 0.45 ± 2.4 (0.10-6.8) 34.0 ± 37.0 (11.0-118.0) 0.0 ± 7.0 (0-25.0) 2.0 ± 13.0 (0.0-4.0)
NL (4.5-11.0) M (140-170); F (120-160) M (4.5-6.0); F (3.6-5.5) M (0.40-0.54); F (0.34-0.51) (0.2-1.5) (150.0-400.0)
WBC = white blood count; HGB = hemoglobin: RBC = red blood cells: HCT = hematocrit: PB = peripheral blood: M6A = myeloblasts a 30%: M6B = proertythroblasts z 30%: normal = NL.
A.J.C.P.-July 1992
KOWAL-VERN, COTELINGHAM, AND SCHUMACHER in Acute Erythroleukemia
Prognostic Significance ofProery Statistical analysis. A database for the study was set up with the use of RBASE for disk-operating system 2.11 (Microrim Inc., Redmond, WA). The data were analyzed by a two-sample Student's Mest using Statgraphics 2.6 (STSC Inc., Rockville, MD). RESULTS The 10 patients with M6A (7 male, 3 female) had a mean age of 54 years (range, 25 to 77 years); the M6B group of 13 patients (9 male, 4 female), had a mean age of 50 years (range, 4.5 to 72 years). The median survival was 2 months (range, 0.06 to 9 months) for the M6B group and 16 months (range, 2 to 48 months) for the M6A group. The past medical history, therapy, and survival are listed in more detail in Tables 1 and 2. The patients died as a result of fulminant leukemia or of treatment complications, such as infection and/or hemorrhage. No difference in the findings was apparent in patients with M6A or M6B in relation to age, sex, or past history. However, a significant difference in survival was observed in the two groups. Treatment usually undertaken was daunorubicin and cytosine arabinoside; however, the M6B group, in many instances, did not survive long enough to establish the benefit of chemotherapy. The peripheral blood findings listed in Table 3 were similar, although patients with M6A had a higher median nucleated red blood cell and reticulocyte count percentage in the peripheral blood than did patients with M6B. The comparison of the bone marrow myeloblasts and proerythroblasts are listed in Table 4 as median ± 1 standard deviation. In Table 5, the bone marrow elements are listed, with the myeloid and erythroid components separated from the myeloblasts and proerythroblasts. Ringed sideroblasts were present in 18 of 23 patients (86%). The four patients who did not have ringed sideroblasts had a high percentage of very immature erythroblasts (three were M6A). The fatxell ratio in patients with M6B and M6A ranged from 1:99 to 70:30 and from 1:99 to 20:80, respectively. The myeloid:erythroid ratio varied from 1:1 to 1:20 in the M6B and from 1:2 to 1:5 in the M6A patients.
TABLE 4. PERCENTAGE OF MYELOBLASTS AND PROERYTHROBLASTS IN ACUTE ERYTHROLEUKEMIA
Myeloblasts ANC Myeloblasts NEC Proerythroblasts ANC Proerythroblasts EC Total blasts ANC
M6A
M6B
P Value
19 ± 10 45 ± 2 2 5±4 9±7 27+11
2±3 9 + 20 30 ± 18 47 ± 2 5 39 ± 15
