Volume 104 Number 2
globulin), which act to decrease the negative charge (the zeta potential) on the surface of erythrocytes and thus increase the rate at which they sediment.2.~ As such, these assays serve as simple and nonspecific tests for the presence and intensity of an inflammatory process. Studies in adults have analyzed ZSR resul[s in relation to the clinical status of patients and have concluded that it accurately reflects the presence or absence of inflammation) 7 We evaluated the ZSR in children by comparing it with simultaneous ESR measurements. Our results found good agreement (86%) in samples in the 31% to 40% Hct range. Discrepant results were encountered most frequently in patients with Hct <31%. All discrepant samples in this range had an elevated ESR and a normal ZSR, which probably reflects the tendency for the ESR to be artifactually elevated in anemic patients.8 Five of the six discrepant samples in the >40% Hct range had an elevated ZSR and a "normal ESR, which might reflect the impact that the ZSR fo/mulation has in this range. Without actually measuring the acute-phase reactants that the sedimentation rates indirectly reflect, it is impossible to know with certainty in each instance which test best reflects the presence or absence of inflammation. The ESR has two major advantages over the ZSR. One is physicians' familiarity with a test that has over 50 years of clinical use. The other is that its expanded scale provides a greater range of values when marked inflammation is present, whereas the ZSR has an upper limit of about 70% in most cases. This expanded scale may be of use in assessing disease progression in patients with markedly elevated erythrocyte sedimentation rates.
Clinical attd laboratory observations
The ZSR correlates closely with the ESR in most patients, is probably superior in anemic patients, and has significant practical advantages. Although other modifications of the ESR have been suggested for use in children,9 we believe that the ZSR is the better alternative and have introduced it into routine use in our Clinical Hematology Laboratory. We thank Karen Maxwell and Carol Groves for technical assistance, Carol Morris for secretarial assistance, and Evelyn Saxon and Dr. Kenneth Boyer for reviewingthe manuscript.
REFERENCES 1. Lascari AD: The erythrocyte scdimentation rate. Pediatr Clin North Am 19:1113, 1972. 2. Bull BS, Brailsford JD: The zeta sedimentation ratio. Blood 40:550, 1972. 3. Dawson JB" The ESR in a new dress. Br Mcd J 1:1697, 1960. 4. Bull BS: Clinical and laboratory implications of present ESR methodology. Clin Lab Hacmatol 3:283, 1981. 5. Bucher WC, Gall EP, Woodworth R: Zeta scdimentation ratio in rheumatic disease. Am J Ctin Pathol 64:613, 1975. 6. Morris MW, Pinals RS, Nelson DA: The zeta sedimentation ratio (ZSR) and activity of discasc in rheumatoid arthritis. Am J Clin Pathol 68:760, 1977. 7. Buchcr WC, Gall EP, Beckcr PT: The zcta scdimcntation ratio (ZSR) as the routine monitor of disease activity in a general hospital. Am J Clin Pathol 72:65, 1979. 8. BullBS, Brechcr G: An evaluation of the rclativc merits of the Wintrobc and Westcrgren scdimentation methods, including. hcmatocrit correction. Am J Clin Pathol 62:502, 1974. 9. Adler SM, Denton RL: The erythrocyte scdimcntatlon rate in the newborn period. J PEDIA'rR86:942, 1975.
Transient neonatal leukemoid reactions in mosaic trisomy 21 Nita L. Seibei, M.D., Annemarie Sommer, M.D., and James Miser, M.D. Columbus, Ohio, and Bethesda, Md.
NEONATAl. LEUKEMOID REACTIONS, documented in infants with Down syndrome,' have been described recently in phenotypically normal infants with mosaic trisomy
Front the Department of Pediatrics, Coluntbus Children's llospital, and the Pediatric Oncolog)' Branch. National Cancer Institute. Reprint requests: Nita L. Seibel, M.D., Children's Hospital, 700 Children's Dr., Columbus, OH 43205.
21. 2.4 These normal appearing infants with mosaicism appear to experience spontaneous gradual resolution of both the leukoeytosis and the trisomy 21 karyotype over a period of months to years without residual effects. We describe two cases of neonatal leukemoid reactions associated with mosaic trisomy 21 karyotype. M A T E R I A L S AND M E T H O D S Banded chromosome studies were performed using a method modified from Moorehead et al? and SeabrighP
Clinicaland laboratory observations
The Journalof Pediatrics February1984
Table. Cytogenetic data in two children with leukemoid reaction in infancy
Percentof cells counted Age
Patient 1 2 I 2 3 4 5
d yr yr yr yr yr
Blood Blood Blood Blood Blood Blood
32 52 33 51 50 65
63 0 0 0 0 0
50 90 87 98 94 93
50 10 13 2 6 7
Blood Blood* Blood Skin Bone marrow
65 32 32 32 21
18 18 0
Patient 2 7d 3 mo
9P|IA nonstimulated. on cultures of blood lymphocytes incubated overnight without p h y t o h e m a g g l u t i n i n ( P H A ) and for 72 hours with P H A stimulation. Bone marrow cells in patient 2 were prepared for eytogenetie examination by direct method without culture and analyzed with regular G i e m s a staining. Fibroblasts were grown from a skin biopsy specimen and harvested for cytogenetic analysis after 33 to 51 days in culture. CASE REPORTS Patient 1. This 7-hour-old, 4240 gm infant boy was referred to Columbus Children's Hospital for evaluation of abdominal distention. Ile was the product of a 31-year-old G6P3Abj white woman whose pregnancy, labor, and delivery were uncomplicated. Physical examination revealed a large, plethoric appearing infant in no acute distress. Abnormal physical findings included an enlarged liver, palpable 5 cm below the right costal margin, as well as generalized petechiae. Except for epicanthal folds, no other features suggestive of Down syndrome were present. Laboratory data included hemoglobin 15.7 gm/dl and WBC count 80,000/mm ~ with 3% polymorphonuclear leukocytes, 5% bands, 3% promyelocytes, 2% myelocytes, I% metamyelocytes, 19% lymphocytes, 3% monocytes, 1% basophils, and 63% blast ceils. The platelet count was 155,000/mm 3, and rcticulocyte count 8.5%. Chromosome analysis on peripheral blood revealed 46,XY in 50% of karyotypes and 47,XY,+21 in 50%. Bone marrow aspirate at I day of age revealed a slight decrease in cellularity, with increased crythropoiesis, decreased myelopoiesis, and the presence of 32% blast cells. Because the blood picture was consistent with a leukemoid reaction, chemotherapy was withheld. Gradually over the next month the peripheral blood count returned to normal. By 2 months of age, the organomegaly had resolved; over the following 5 years the patient has continued to have normal growth and development. Chromosome analysis has been performed annually, and the percentage of trisomic cells has gradually decreased (Table). Patient 2. This 2240 gram white infant boy was the product of a
23-year-old G2PzAbl white woman whose pregnancy was complicated by premature labor and vaginal bleeding at 3589 weeks' gestation. On admission to the Ohio State University Neonatal Intensive Care Unit, the infant was grunting and slightly pale. No dysmorphic features, hepatosplenomegaly, or masses were present; the estimated gestational age was 34 weeks. The patient initially required oxygen by hood (50%) for mild hyaline membrane disease, but was easily weaned to room air over 4 days. The anemia noted on admission (hemoglobin 13.7 gm/dl) was thought to be secondary to a combination of a cephalohematoma and traumatic delivery, for which the patient was transfused at 12 hours of age with packed red blood cells. The WBC count was 22,900 cells/mm 3 with 25% PMNs, 13% bands, 1% monocytes, 3% cosinophils, 2% basophils, 23% lymphocytes, 2% promyelocytes, 6% myclocytes, 2% metamyelocytes, and 23% blast cells, with 20 nucleated RBC/100 WBC. The initial platelet count was 96,000/mm 3. Ampicillin and gentamicin were administered intravenously for 14 days for presumptive sepsis, although blood, gastric aspirate, and cerebrospinal fluid cultures were negative. TORCll antibody titers were not elevated. Over the 15 days of hospitalization the WBC count varied from 18,300 to 31,500 cells/ram ~ with blast cells always present; the highest value was noted at 4 days of age with 25% blasts. The platelet count reached a low of 27,000/mm ~ without evidence of disseminated intravascular coagulopathy at 6 days of age, and rose tO 84,000/mm 3 by discharge, without evidence of bleeding problems. A bone marrow aspirate at 5 days of age revealed normal cellularity for age and a myeloid to erythroid ratio of 15:1; differential showed 4% erythroid, 61% myeloid, 6% lymphocytes, 9 I% monocytes, and 28% blast cells. Sudan black and periodic acid-Schiff stains of the bone marrow were negative. Blast cells in the peripheral blood were peroxidase, sudan black, and terminal deoxynucleotidyl transferase (Tdt) negative. The peripheral lymphocytes were characterized by the sheep RBC rosette technique to be I% B cells and 95% T cells. The blast cells were E rosette, surface immunoglobulin, la, and CALLA negative. The ccrebrospinal fluid contained no blast cells. Chromosome analysis on peripheral blood incubated without
Volume 104 Number 2
PHA disclosed 47.XY,+21 on all cells, and with PHA stimulation disclosed 46,XY on all cells. Complete disappearance of the blast cells was noted from the peripheral blood by 39 days of age. Repeat bone marrow aspirate at 3 months of age revealed normal cellularity with a myeloid to erythroid ratio 1.3:1; the differential showed 26% crythroid, 33% myeloid, 34% lymphoid, 1% monocyte, and 6% blast cells. Chromosome analyses on peripheral blood, skin, and bone marrow disclosed 46,XY (Table). The patient's growth and developmenthave continued to be normal at 6- and 9-month examinations. DISCUSSION Transient myeloproliferative syndromes with signs and symptoms indistinguishable from congenital acute leukemia, except that complete clinical and hematologic recovery occurs within weeks to months without specific antileukemic treatment, have been described in at least 30 newborn infants with Down syndrome, t However, there has only been one reported case ofa phenotypically normal newborn infant with a normal karyotype experiencing a transient spontaneous remission of a condition indistinguishable from acute leukemia.7 Two phenotypically normal infants have been reported 'who at birth had circulating blast ceils in the peripheral blood and karyotypes revealing trisomy 21 mosaicism. Brodeur et al.' reported a 3-day-old infant with organomegaly, leukocytosis, and myeloblasts whose karyotype revealed 47,XY,+21 (100%) in blood and bone marrow cells and 4% in skin fibroblasts. Without treatment, the leukocytosis and organomegaly resolved, and serial cytogenetic studies disclosed gradual reduction of trisomic ceils in the blood and bone marrow over 3 months. Weinberg et al? reported a phenotypically normal infant who at birth had hepatosplenomegaly, thrombocytopenia, and leukocytosis with circulating blast cells. No treatment was given, and 9the organomegaly as well as the leukocytosis gradually disappeared. Chromosome analysis at 12 days of age revealed 47,XX,+21 (10% to 14%) in peripheral blood. Repeat chromosome analysis at 1 year of age disclosed complete absence of 47,XX,+21 karyotype from bone marrow and skin; at 31 months of age, 3% trisomic cells remained in the peripheral blood. Findings in our patient ! are similar; however, patient 2 had no physical evidence of leukemia. Except for the presence of circulating blast cells in the peripheral blood, chromosome analysis probably would never have been done. The blood stimulated with PtlA revealed a normal karyotype, which is consistent with PHA stimula"tion of normal thymus-derived lymphocytes; the nonstimulated specimen revealed 100% 47,XY,+21. These findings are in contrast to the results reported by Brodeur et al.'; in their patient the stimulated specimen revealed an abnormal karyotype. The response to PIIA could be a reflection of
Clinical attd laboratory observations
the potential suppression that is exerted by the trisomic cells on the cytogenetically normal cells. All three patients demonstrated complete disappearance of circulating blast ceils by 3 months of age, which preceded the gradual resolution of the trisomic cells in the blood. None of the patients had any stigmata of Down syndrome, and all have had normal growth and development. Because of the persistence of a small clone of cells that have a +21 karyotype long after the disappearance of the blast cells from the blood and bone marrow, it is likely that the cause of this syndrome is a true somatic mosaicism. The significance of trisomy 21 and of the transient leukemoid reaction in newborn infants is not known. Ross et al) proposed that trisomy of chromosome 21 was responsible for an intrinsic intracellular defect in the regulation of leukocyte distribution, release, and maturation. Further evidence to support this concept is suggested by Heaton et al., 3 who described a normal appearing newborn infant with a leukemoid reaction with 58% blast cells and thrombocytopenia that disappeared over I month. Chromosome analyses of blood, bone marrow, and skin revealed mosaicism, with the abnormal cell line consisting of translocation of trisomy 21 [46,XY,t(21;21)(pl l;ql 1)1. Over 6th months this mosaicism gradually decreased. Whether the loss of maturational control, as suggested by Ross, is permanent or temporary could determine the development of true leukemia or a leukemoid reaction) .9 In this patient the chromosomal defect could represent an initial insult that results in an abnormality of the hematopoietic maturation and distribution process, producing a leukemoid reaction. A second insult might, if present, induce transformation to leukemia?.~~ There are now four reported cases of phenotypically normal infants with mosaic trisomy 21 karyotypes associated with a transient leukemoid reaction in the newborn period. All four infants have had resolution of the leukocytoffs and associated findings. Whether these patients are at a higher risk for the development of acute leukemia, as has been described in Down syndrome, is yet to be determined.~ ~.~2 REFERENCES I. Rosncr F, Lee SL: Down's syndrome and acute leukemia:
Myeloblastic or lymphoblastic? Am J Mcd 53:203, 1972. 2. Weinberg AG, Schiller G, Windmiller J: Neonatal leukemoid
reaction. Am J Dis Child 136:310, 1982. 3. Hcaton DC, Fitzgerald PH, Fraser GJ, Abbott GD: Transient leukemoid proliferation of the cytogenetieally unbalanced +21 cell line of a constitutional mosaic boy. Blood 57:883, 1981. 4. Brodeur GM, Dahl GV, Williams DL, Tipton RE, Kalwinsky DK: Transient leukemoid reaction and trisomy 21 mosaicism in a phenotypically normal newborn. Blood 55:691, 1980. 5. Moorhead PS, Nowell PC, Mellman WJ, Battips DM,
Clinical attd laboratory observations
}lungerford DA: Chromosome preparations of leukocytes cultured from human peripheral blood. Exp Cell Res 20:613, 1960. Seabright M: A rapid banding technique for human chromosomes. Lancet 2:971, 1971. Van Eys J, Flexner JM: Transient spontaneous remission in a case of untreated congenital leukemia'. Am J Dis Child 118:507, 1969. Ross JD, Moloney WC, Desforges JF: Ineffective regulation of granulopoiesis masquerading as congenital leukemia in a mongoloid child. J PEDIATR 63:1, 1963. Lazarus KII, Iieerema NA, Palmer CG, Baehner RL:
The .lournal of Pediatrics February 1984
Myeloproliferative reaction in a child with Down's syndrome: Cytological and chromosomal evidence for a transient leukemia. Am J ttematol 11:417, 1981. 10. Rowley JD: Down's syndrome and acute leukemia: Increased risk may be due to trisomy 21. Lancet 2:1020, 1981. I1. Lin tl-P, Menaka II, Lim K-tl, Yong ItS: Congenital leukemoid reaction followed by fatal leukemia: A case with Down's syndrome. Am J Dis Child 134:939, 1980. 12. Chu J-Y, Weldon BC: Acute leukemia in a patient with Down's syndrome and transient congenital leukemia. Am J Dis Child 136:367, 1982.
Cerebral spinal fluid pleocytosis with bone marrow contamination Naomi L. C. Luban, M.D., Roberta M. Alessi, M.T.(ASCP)S.H., Bess G. Gold, M.D., and Sudesh Kapur, M.D. Washington, D.C.
EVALUATION OF TIlE CEREBROSPINAL FLUID by white blood cell count allows for the differential diagnosis of diseases of the central nervous system. W e have recently encountered an unusual complication of l u m b a r puncture t h a t may obscure the C S F cell count a n d m a y be more c o m m o n t h a n is currently realized.
CASE R E P O R T This 15-day-old infant boy was admitted with fever and fussiness. The baby was born after 39 weeks' gestation to a 27-year-old G,P, mother who did not believe in standard medical care and delivered the infant at home without medical assistance. The infant was brought to the emergency room with a temperature of 39~ heart rate 160, respiratory rate 36. tie was vigorous and well hydrated, with a soft anterior fontanelle, mild icterus, and well perfused skin. There was no hepatosplenomegaly, and cardiovascular and neurologie findings were normal. Laboratory assessment revealed hemoglobin 18.4 gm/dL, hematocrit 53.5%, WBC count 15,600/#L with 28% segmented forms, 3% bands, 57% lymphocytes, 12% monocytes, and 340,000/taL platelets. Serum electrolyte and glucose concentrations were normal. Cultures of blood, urine, and CSF were obtained. A lumbar puncture was performed without local anesthesia at the L4-5 interspace. The CSF was clear and xanthochromic, with no red blood cells and 1772 WBC/~L on chamber count. The differential was performed on a cytocentrifuged
From the Departments of Laboratory Medicine, Infectious Disease. attd Pathology, Children's llospital National Medical Center. Reprint requests: Naomi L. C. Luban, M.D., Department of Laboratory Medicine, Children's llospital National Medical Center, 111 Michigan Ave., N.W.. Washington. DC 20010.
specimen and revealed 7% segmented forms, 17% lymphocyteso 9% monocytes, 7% histiocytes, and 29 nucleated red blood cells; 40% of the cells observed were large mononuclear cells with scanty blue-gray cytoplasm and perinuelear halos, resembling pronormoblasts, and 20% were larger ceils with dark blue foamy cytoplasm, resembling reactive lymphocytes (Figure). Mitotic forms were noted in several of these cells. No organisms were seen on Gram stain. The CSF glucose concentration was 42 mg/dL (with a peripheral glucose concentration of 79 mg/dL), and the CSF pt'otein was 129 mg/dL. The infant was given antibiotics because of the unexplained CSF pleocytosis. Titers for TORCH, viral cultures of urine, throat, and rectum, liver function analysis, and bilirubin fractionation were obtained. Blood and CSF cultures revealed no growth of bacteria by day 2, and a repeat lumbar puncture was obtained. The CSF was bloody and xanthochromie with RBC 50,000/#L and WBC 3789/t~L. A differential performed on a cytocentrifuged specimen revealed 14% segmented forms, 61% lymphocytes, and 25% mononuclear forms. No pronormoblasts, nucleated red cells, or bizarre cells were observed. CSF cultures for virus, fungus, tuberculosis, and bacteria were obtained. Antibiotic therapy was discontinued when initial cultures were negative after 72 hours. The infant was discharged against medical advice on the fourth hospital day without further medical evaluation. On return to the clinic 3 ,days later, he was well. DISCUSSION T h e C S F in the infant usually contains no red blood cells and up to 30 white blood cells per microliter of fluid, with a p r e d o m i n a n c e of monocytes (72 _ 22%) a n d lymphocytes (20 ___ 150%). Lymphocyte pleocytosis is associated with viral, fungal, or tuberculous meningitis and other primary neurologic disease. Neutrophilic pleocytosis can be seen