PEDIATRICS Vol. 108 No. 1 July 2001, pp. 181-184

EXPERIENCE AND REASON:
Congenital Autoimmune Neutropenia in Two Premature Neonates

	ABSTRACT

Top Abstract Introduction Discussion References

Autoimmune neutropenia (AIN) has been reported in infants and children, but not in neonates. AIN is caused by antibodies produced by the patient against their own neutrophils; therefore, it differs from the more common alloimmune neonatal neutropenia and the neonatal neutropenia because of a maternal autoimmune disease in which antineutrophil antibodies of maternal origin cross the placenta. We observed 2 cases of congenital AIN in premature neonates. These are the youngest reported cases, and indicate that AIN can have a prenatal onset. Examination of the bone marrow biopsies revealed an increase in B lymphocytes and myeloperoxidase-positive cells with a maturation arrest at the myelocyte stage. Recombinant human granulocyte colony-stimulating factor effectively treated the neutropenia, as it does in infantile AIN. Ten months after the diagnosis, 1 of the patients still requires recombinant human granulocyte colony-stimulating administration.

Autoimmune neutropenia (AIN) is caused by patient-derived antibodies directed against the patient's own neutrophils. AIN can occur in childhood, but we are aware of no congenital cases. The youngest reported case of AIN is 33 days of age.¹ We observed 2 premature neonates who had congenital neutropenia that was determined to be AIN. Both patients developed severe neutropenia, but responded to treatment with recombinant human granulocyte colony-stimulating factor (rhG-CSF).

	CASE REPORTS

Case 1

Dizygotic male twins were born to a 34-year-old, gravida 2, para 1, Hispanic woman, whose first child was healthy and born at term without complications. The twin pregnancy was complicated by a febrile illness that occurred at 34 weeks of gestation. During the illness, parvovirus B19 titers were negative.

The twins were born by spontaneous vaginal delivery at 35 weeks of gestation after spontaneous rupture of membranes 2 hours before the delivery. Twin A's birth weight was 2195 g and Twin B's was 2296 g. The infants seemed to be healthy, and physical examination revealed no abnormalities. On the second day of life, Twin A had hypothermia and tachypnea (respiratory rate 80/min). He was taken to the special care nursery and a complete blood count (CBC) showed a total white blood cell count of 3600/mm³, with an absolute neutrophil count (ANC) of 1296/mm³. Red blood cell count, hemoglobin concentration, and platelet count were normal. Blood cultures were obtained, and ampicillin and gentamicin were initiated for the treatment of potential sepsis. Antibiotics were continued for 7 days despite no bacterial growth from the blood cultures. The neutropenia persisted but was not severe (ANC's between 1160 and 1376/mm³), and the infant remained asymptomatic. Additional evaluation of the neutropenia included samples from the patient and mother for antineutrophil antibody studies. The CBC's of both parents were normal.

Granulocyte typing, granulocyte cross-matching, and assays of serum to detect neutrophil-specific antibodies and immune complexes were performed at the Neutrophil Serology Reference Laboratory of the American Red Cross, St Paul, Minnesota. The granulocyte agglutination assay and the granulocyte immunofluorescence assay were both negative using maternal serum. However, when the infant's serum was examined, the granulocyte immunofluorescence assay was positive. Antibodies to human leukocyte antigens (HLA) can react in these assays, therefore, a monoclonal antibody immobilization of granulocyte antigens assay (MAIGA)² was performed to differentiate between HLA and neutrophil-specific antibodies. The MAIGA detected specificity for the NA1 antigen using a monoclonal antibody to the CD 16 molecule, and a diagnosis of congenital autoimmune neutropenia was made. Granulocyte typing of the mother's and the father's neutrophils revealed both to be NA1 antigen-positive.

The remainder of the infant's hospitalization was unremarkable, and he was discharged on day of life 22. He continued to do well, and subsequent ANC's remained around 1200/mm³. However, at 2 months of age, when he received his first set of immunizations, he developed a fever to 104°F. The ANC was 496/mm³, and he was admitted to the hospital for evaluation of fever and neutropenia. Blood, urine, and cerebrospinal fluid cultures were obtained and antibiotics initiated. The following day, the ANC had increased to 3192/mm³, and the serum G-CSF concentration was 55 pg/mL (G-CSF concentration in healthy adult volunteers is <100pg/mL). He was discharged when cultures were negative. During the next 2 days, the ANC progressively decreased to 438/mm³ with a serum G-CSF concentration of 1124 pg/mL. At this time, a repeat antineutrophil antibody screen was obtained, and a bone marrow biopsy³ was performed. Recombinant human G-CSF therapy was initiated subcutaneously at 10 µg/kg/d every other day for 3 doses, with a subsequent ANC of 2900/mm³. Repeat antineutrophil antibody testing was positive. The bone marrow had a maturation arrest at the myelocyte stage. (Figure 1) Megakaryocyte and erythroid precursors seemed normal. Lymphocytes were present in small aggregates. A myeloperoxidase stain revealed abundant promyelocytes and myelocytes. (Fig 1) T and B cell stains showed a marked predominance of B lymphocytes over T lymphocytes. With the bone marrow findings and the presence of anti-NA1 antibodies in the infant's serum on 2 separate occasions, the diagnosis of congenital autoimmune neutropenia was confirmed, and the neonate was enrolled in the Severe Chronic Neutropenia International Registry.

View larger version (59K): [in this window] [in a new window]   Fig. 1.   Bone Marrow Biopsy. The top panel (A) shows a hematoxylin and eosin-stained section demonstrating left-shifted myeloid maturation (absence of metamyelocytes, bands, and segmented forms). The center panel (B) shows an immunohistochemical stain for myeloperoxidase. The stain identifies the many immature myeloid cells. The bottom panel (C) shows an immunohistochemical stain to detect the B cell antigen CD20. This stain demonstrates clusters of B lymphocytes. (Original magnification 400×).

He was initially maintained on weekly doses of rhG-CSF and then received biweekly doses as an outpatient to maintain his ANC at >1000/mm³ for 5 months. During the past 10 months, he has not been treated prophylactically with antibiotics or had any recognized infections.

Case 2

Monozygotic female twins were born at 31 weeks' gestation to a 40-year-old, gravida 4, para 0, white woman by cesarean section secondary to prolonged premature rupture of the membranes and advanced cervical dilation. The mother's pregnancy was otherwise uncomplicated.

Twin A's birth weight was 1914 g and Twin B's was 1994 g. Apgars were 9¹ and 9⁵ for each twin. Within 1 hour, Twin A developed respiratory distress requiring endotracheal intubation and surfactant administration. She also developed hypotension requiring administration of intravenous normal saline. An initial ANC was 576/mm³, and she received intravenous immune globulin (IVIG) and ampicillin and gentamicin for suspected sepsis. The ANC increased, and she completed a 10-day course of antibiotics despite no bacterial growth from the blood cultures. Because the infant presented with a clinical picture consistent with septic shock, the initial ANC was attributed to the presence of bacterial infection. Two CBCs obtained on days three and four of life had normal ANCs (3484 and 2288/mm³). The neonate continued to improve and was discharged to home 18 days after birth.

At 28 days of age the mother brought the infants to the pediatrician for well-child visits. At that time, delayed separation of the umbilical cord without omphalitis was noted in Twin A. The infant was referred to an immunologist for additional evaluation. The peripheral blood leukocyte subsets (CD3, 4, 8, 16, 10, 11b, 11c, 56, 57, 4/8 ratio, 16/56 ratio, and HLA-DR) were normal as assessed by flow cytometric studies. A CBC revealed a total white blood count of 6700/mm³ and an ANC of 67/mm³. Samples for antineutrophil antibodies (Neutrophil Serology Reference Laboratory of the American Red Cross) were obtained on the mother and the infant, and the neonate underwent a bone marrow biopsy. The marrow showed a maturation arrest at the myelocyte stage with normal megakaryocytes and erythroid cells, and increased B lymphocytes. The CBC's of both parents and her identical twin were normal.

The infant was treated with 10 µg/kg of rhG-CSF subcutaneously every day for 3 doses with a subsequent ANC of 24 738/mm³ 5 days after the initiation of rhG-CSF therapy. The dosing frequency was changed to every other day. Her serum G-CSF concentration before rhG-CSF treatment was 46 pg/mL. Maternal antineutrophil antibody testing was negative. The neonate's serum was positive using the granulocyte immunofluorescence assay. The extended antibody testing in the MAIGA assay did not define a granulocyte antibody specific for NA1, NA2, or NB1 antigens, or the CD18, the common subunit of the -2 integrins.

At 10 months of age, she continues to require biweekly rhG-CSF to maintain an ANC 1000/mm³. She has remained well and has not been treated prophylactically with antibiotics. The patient was enrolled in the Severe Chronic Neutropenia International Registry.

	DISCUSSION

Top Abstract Introduction Discussion References

Neutropenia is relatively common in premature infants,^4,5 and the majority of cases are associated with maternal hypertension or bacterial infection.^5,6 However, the differential diagnosis of neonatal neutropenia is extensive and includes alloimmune neonatal neutropenia (ANN),⁷ cyclic neutropenia,⁴ IVIG-induced neutropenia,⁸ and severe chronic neutropenia, such as Kostmann syndrome.⁹

AIN has not previously been reported in preterm infants. It differs from ANN, which is well-described in this population. ANN is diagnosed by granulocyte-specific antibodies in the maternal serum, directed against fetal neutrophil antigens, which are foreign to the mother, but are inherited from the father.⁷ In contrast, AIN is characterized by the presence of antineutrophil antibodies in the infant's serum that are absent from the maternal serum.¹

In neonates who have been treated with IVIG and subsequently develop neutropenia, the diagnosis of immune neutropenia attributable to antineutrophil antibodies in the IVIG⁸ must be considered. In the single case report of a neonate who developed neutropenia associated with IVIG therapy, the neutropenia developed rapidly after administration and persisted for 4 weeks.⁸ Our case 2 received IVIG on the first day of life and maintained a normal ANC for 6 weeks afterward. Because the clinical course of our neonate did not fit the pattern for development of neutropenia described by Lassiter,⁸ we felt it unlikely that the administration of IVIG produced her neutropenia.

In older children with AIN, mild infections during the period of neutropenia are common, and spontaneous remission of the neutropenia after 6 to 12 months is typical.¹⁰ The incidence of AIN is about 1:100 000,¹¹ and thus it is more common than severe chronic neutropenia (incidence 1:1 000 000).¹² The cause of AIN is not known. Bone marrow examination of patients with AIN have the appearance of maturation arrest in the myeloid series. This finding is not attributable to arrest of maturation, but to the destruction of the more mature myeloid forms (myelocytes, metamyelocytes, bands, and segmented forms) on which the relevant neutrophil antigen is expressed.¹ The expanded early myeloid precursors in our patients, as demonstrated by the myeloperoxidase stain, represent an appropriate compensatory increase in neutrophil production. The large number of B lymphocytes in the marrow is interesting because in normal adult bone marrow samples, T cells generally outnumber B cells.¹³ In autoimmune disorders in adults, lymphocytes and plasma cells are often increased in the marrow.^14,15

Our patients presented with neutropenia during the first days of life, indicating an intrauterine onset. Previously, an association between AIN and Parvovirus B19 infection was suggested.¹⁶ However, Bux and colleagues¹ tested 110 serum samples with AIN and found that only 36 sera had significant Parvovirus B19 titers. They concluded that Parvovirus B19 infection is not an invariant antecedent to AIN. This was apparently the case for our first infant, whose maternal serology for Parvovirus B19 was negative. An association between NA1-specific autoantibodies and HLA-DR2 has been found. However, the occurrence of AIN in only one of monozygotic twins in our second case, and that of others,^9,17 is strong evidence against a strictly hereditary mechanism.

AIN is thought to be exclusively of postnatal onset.¹ However, perhaps some of the infants with AIN reported by Bux actually had congenital neutropenia, which was not recognized until they were several months old when they developed an infection. Indeed, the neutropenia in our first patient was not severe initially, and only became severe over time. On the basis of these 2 cases, it seems likely that AIN of infancy is, at least in some cases, a disorder of prenatal onset and that this diagnosis should be considered in neonates with persistent unexplained neutropenia, even if mild.

Strauss and coworkers¹⁸ demonstrated the presence of detectable anti-leukocyte antibodies in 13% (4 of 30) of multiply-transfused neonates. Of these, 3 had reactivity against HLA class I antigens and 1 had antineutrophil specificity. This unequivocally demonstrates that neonates are capable of producing antibodies to neutrophil antigens.

Because AIN has not previously been reported in preterm neonates, the best treatment option is undefined. Bux and coworkers¹ reviewed 3 treatment regimens for children with AIN: IVIG, corticosteroids, and rhG-CSF. The remissions induced by IVIG and corticosteroids were inconsistent, but rhG-CSF treatment was successful in all patients reported, and in the 2 premature infants in our report. We have previously shown that endogenous G-CSF concentrations in neonates increase during infection, resulting in an increase in the peripheral ANC.¹⁹ This was seen in our first patient during his febrile illness and has been observed in other patients with AIN during infection.¹ Our choice to use rhG-CSF to treat these 2 neonates was based on their severe and prolonged course of neutropenia. However, its use in neutropenic premature neonates requires more comprehensive studies because the potential effects of rhG-CSF on nonhematopoietic tissues in the developing neonate remain to be determined.²⁰

Darlene A. Calhoun, DO^*
Lisa M. Rimsza, MD
David J. Burchfield, MD^*
Melinda Millsaps, ARNP^*
Robert D. Christensen, MD^*
* Department of Pediatrics, Division of Neonatology
 Department of Pathology, Division of Hematopathology
University of Florida College of Medicine
Gainesville, FL 32610-0296

Jyoti Budania, MD
Worthington Pediatrics
Gainesville, FL 32606

Jeffrey McCullough, MD
American Red Cross North Central Blood Services
St Paul, MN 55107

	FOOTNOTES

Received for publication Jul 19, 2000; accepted Oct 25, 2000.

Reprint requests to (D.A.C.) University of Florida, Box 100296, Department of Pediatrics, Division of Neonatology, Gainesville, FL 32610-0296. E-mail: calhoda{at}peds.ufl.edu

	ABBREVIATIONS

AIN, autoimmune neutropenia; rhG-CSF, recombinant human granulocyte colony-stimulating factor; CBC, complete blood count; HLA, human leukocyte antigens; MAIGA, monoclonal antibody immobilization of granulocyte antigens; IVIG, intravenous immune globulin; ANN, alloimmune neonatal neutropenia.

	REFERENCES

Top Abstract Introduction Discussion References

1. Bux J, Behrens G, Jaeger G, Welte K Diagnosis and clinical course of autoimmune neutropenia in infancy: analysis of 240 cases. Blood 1998; 91:181-186 [Abstract/Free Full Text]
2. Bux J, Jung K, Kauth T, Mueller-Eckhardt C Serological and clinical aspects of granulocyte antibodies leading to alloimmune neonatal neutropenia. Transfus Med 1992; 2:143-149 [Medline]
3. Sola M, Rimsza L, Christensen R A bone marrow biopsy technique suitable for use in neonates. Br J Haematol 1999; 107:458-460 [CrossRef][Medline]
4. Al-Mulla Z, Christensen R Neutropenia in the neonate. Clin Perinatol 1995; 22:711-739 [Medline]
5. Calhoun D, Christensen R. Human developmental biology of granulocyte colony-stimulating factor. In: Christensen R, ed. Clinics in Perinatology. Philadelphia, PA: WB Saunders; 2000: 559-76
6. Schibler K. Leukocyte development and disorders during the neonatal period. In: Christensen R, ed. Hematologic Problems of the Neonate. Philadelphia, PA: WB Saunders; 2000: 311-342
7. Felix J, Calhoun D Neonatal alloimmune neutropenia in premature monozygous twins. Pediatrics 2000; 106:340-342 [Abstract/Free Full Text]
8. Lassiter H, Bibb K, Bertolone S, Patel C, Stroncek D Neonatal immune neutropenia following the administration of intravenous immune globulin. Am J Pediatr Hematol Oncol 1993; 15:120-123 [Medline]
9. Calhoun D, Christensen R Kostmann syndrome in preterm neonates. Pediatrics 1997; 99:259-261 [Free Full Text]
10. Conway L, Clay M, Kline W, Ramsay N, Krivit W, McCullough J Natural history of primary autoimmune neutropenia in infancy. Pediatrics 1987; 79:728-733 [Abstract/Free Full Text]
11. Lyall E, Lucas G, Eden O Autoimmune neutropenia of infancy. J Clin Pathol 1992; 45:431-434 [Abstract/Free Full Text]
12. Welte K, Dale D Pathophysiology and treatment of severe chronic neutropenia. Ann Hematol 1996; 72:158-165 [CrossRef][Medline]
13. Foucar K. Bone Marrow Pathology. Chicago, IL: ASCP Press; 1995:259
14. Rosenthal N Bone marrow findings in connective tissue disease. Am J Clin Pathol 1989; 92:650-654 [Medline]
15. Navone R, Valpreda M, Pich A Lymphoid nodules and nodular lymphoid hyperplasia in bone marrow biopsies. Acta Haematol 1985; 74:19-22 [CrossRef][Medline]
16. Cartron J, Bader-Meunier B, Deplanche M, Human Parvovirus B19-associated childhood autoimmune neutropenia. Int J Pediatr Hematol Oncol 1995; 2:471-473
17. Bux J, Mueller-Eckhardt G, Mueller-Eckhardt C Autoimmunization against the neutrophil-specific NA1 is associated with HLA-DR2. Hum Immunol 1991; 30:18-21 [CrossRef][Medline]
18. Strauss R, Johnson K, Cress G, Cordle D Alloimmunization in preterm infants following repeated transfusions of leukocyte-reduced red blood cells from the same donor. Transfusion 2000; 40:1463-1468 [CrossRef][Medline]
19. Calhoun D, Lunøe M, Du Y, Hutson A, Veerman M, Christensen R Granulocyte colony-stimulating factor (G-CSF) serum and urine concentrations in neutropenic neonates before and following the intravenous administration of recombinant G-CSF. Pediatrics 1999; 105:392-397 [Abstract/Free Full Text]
20. Calhoun D, Christensen R The distribution of granulocyte colony-stimulating factor receptor (G-CSF-R) and its messenger RNA expression in the human fetus. Pediatr Res 1999; 43:333-338