Intracranial hemorrhage in utero as a complication ofisoimmune thrombocytopenia

Volume 95 Number 4

The etiology of the ST depression in premature infants with RDS/PDA m a y be related to several factors. One possibility is that the ST depression is related to digoxin therapy. Digoxin may cause ST changes, specifically coving of the ST segment without J-point depression? It is unlikely that digoxin therapy produced the ST depression in these infants because (1) all of our infants had significant J-point depression; (2) ST depression was not necessarily related to the presence of digoxin therapy (6 of 19 received no digoxin at all), and (3) ST depression returned to baseline after surgery whether the patient was continued on digoxin or not. The etiology of ST depression seen in these infants is most likely subendocardial ischemia. Recent radionucleotide myocardial imaging techniques have demonstrated subendocardial or global ischemia in infants with perinatal cardiorespiratory distress) In addition, Achtel et al ~ and Vincent et al, 7 utilizing experimental aortapulmonary communications, have shown that reduced coronary flow in this model can lead to left ventricular subendocardial ischemia and subsequent myocardial decompensation. They proposed that this factor should be considered as a cause of refractory congestive heart failure in naturally occurring aorta-pulmonary steal syndromes, such as PDA. Another important consideration is the effect of mildto-moderate pulmonary hypertension on myocardial perfusion and performance. Brooks et al" have found that with decreased right coronary blood flow and only mild right ventricular hypertension, there is significant alteration of right ventricular function and subsequently left ventricular function. The cardiac pathology of 11 infants who died after prolonged respirator courses was reported by DeSa. 9 All infants had pulmonary hypertension secondary to severe broncho-pulmonary dysplasia. At autopsy all infants had cardiac hypertrophy and areas of myocardial fiber necrosis. The ECG findings in these infants were not reported. Considering the presence of pulmonary hypertension in infants with RDS, TM it is rational that decreased coronary

B r i e f clinical and laboratory observations

6 11

flow secondary to a ductal runoff, and mild right ventricular hypertension secondary to RDS might lead to myocardial compromise. Serial ECGs are an important adjunct in the management of infants with RDS/PDA; the character and magnitude of ST depression strongly suggests subendocardial ischemia. This subendocardial ischemia may be one of the mechanisms of refractory CHF in the RDS/ PDA group; persistent ST depression greater than 2 mm may be another indication for PDA ligation in infants with RDS/PDA.

REFERENCES

1. Cotton RB, Stahlman MT, Lovar I, and Catterton WZ: Medical management of small preterm infants with symptomatic patent ductus arteriosus, J PEDIATR92:467, 1978. 2. Cotton RB, Stahlman MT, Bender HW, Graham TP, Catterton WZ, and Kovar I: Randomized trial of early closure of symptomatic patent ductus arteriosus in small preterm infants, J PEDIATR93:647, 1978. 3. Walsh SZ: Evolution of the electrocardiogram o~"healthy premature infants during the first year of life, Acta Paediatr Scand Suppl 145:26, 1963. 4. Nadas AS, and Fyler DC: Pediatric Cardiology, Philadelphia, 1972, WB Saunders Company, p 72. 5. Finley JP, Howman-Giles RB, Gilday DL, Bloom KR, and Rowe RD: Transient myocardial ischemia of the newborn infant demonstrated by thallium myocardial imaging, J PEDIATR 94:263, 1979. 6. Achtel RA, Utley JR, Todd EP, and Erickson JL: Patent ductus arteriosus as a cause of subendocardial ischemia in newborn lambs, Circulation 53,54:185, 1976 (Suppl lI) (abst). 7. Vincent WR, Sakamoto JR, and Buckberg GD: Mechanism of cardiac failure in experimental patent ductus arteriosus, Am J Cardiol 29:295, 1972 (abstr). 8. Brooks, H, Kirk ES, Vokonas PS, et al: Performance of the right ventricle under stress: relation to right coronary flow, J Clin Invest 50:2176, 1971. 9. DeSa D J: Myocardial changes in immature infants requiring prolonged ventilation, Arch Dis Child 52:138, 1977. 10. Stahlman MT: in Avery GB, editor: Neonatology pathophysiology and management of the newborn, Philadelphia, 1975, JB Lippincott Company, pp 230.

Intracranial hemorrhage in utero as a complication of isoimmune thrombocytopenia Edwin L. Zalneraitis, M.D.,* Richard S. K. Young, M.D., and Kalpathy S. Krishnamoorthy, M.D., Boston, Mass. From the Pediatric Neurology Unit, Massachusetts General Hospital; and Departments of Neurology and Pediatrics, Harvard Medical School *Reprint address: Pediatric Neurology Unit, Massachusetts General Hospital Fruit St., Boston, MA 02114.

0022-3476/79/100611 + 04500.40/0 9 1979 The C. V. Mosby Co.

ISOIMMUNE THROMBOCYTOPENIA is a known cause of bleeding in the newborn infant? Fetal platelets, of a different histocompatibility type from the mother's, cross the placenta. The resultant sensitization of the mother

6 12

Brief clinical and laboratory observations

leads to the p r o d u c t i o n o f IgG antibodies directed against fetal platelets. These m a t e r n a l antibodies cross the placenta and cause destruction o f fetal platelets. A m o n g the h e m o r r h a g i c complications resulting from the thrombocytopenia, those in the central nervous system carry significant morbidity a n d mortality. H e m o r r h a g i c central nervous system insults have been r e p o r t e d with intrapartum and other perinatal stresses. F o r this reason, it has been advocated that infants at risk for i s o i m m u n e t h r o m b o c y t o p e n i a be delivered by cesarean sectionY We describe two infants who h a d intracranial h e m o r r h a g e in utero, and who would not have b e e n helped by operative delivery. CASE R E P O R T S Patient 1. The patient was born weighing 3,200 gm after a 38-week gestation to a 29-year-old, gravida 3, para 2, abortus 0, type O-Rh positive nurse who was not employed while pregnant. The second of her pregnancies ended at 36 weeks, and was remarkable for presumed abruptio placenta and an anemic infant. Neither of the prior two infants had any evidence of a bleeding tendency. Significant in this pregnancy was a weight gain of only 3.5 pounds. Abbreviations used CCT: computerized cranial tomography EEG: electroencephalogram Labor was uncomplicated, with a total duration of four hours; rupture of membranes occurred 40 minutes prior to delivery. The delivery was spontaneous and vaginal with the patient under local anesthesia, without the use of forceps. Apgar score was ten at one and at five minutes. Head circumference was 39 cm ( > ninety-seventh percentile) at birth. During the first 12 hours, increasing petechial rash and brown vomitus after two sterile water feedings caused concern. The infant had split cranial sutures, a bulging fontanelle, liver palpable 4 cm below the costal margin, and spleen palpable 2 cm below the costal margin. The platelet count at 12 hours of age was 7,000/mm 3 with a hematocrit of 35%. The patient was transferred tt~ this hospital. On arrival, the infant was jittery. The hepatosplenomegaly and petechial rash were unchanged from the prior examination. The patient had good spontaneous movement and normal reflexes, Funduscopic examination revealed multiple retinal hemorrhages. Transillumination of the head wa~ negative. Admission hematocrit was 33%, and platelet count 3,000/mmL Lumbar puncture was omitted, but all other routine cultures were obtained prior to the initiation of therapy with ampicillin and gentamicin. A platelet count of 6,000/ram ~ was obtained just prior to the infusion of two units of random donor platelets at.!8 hours of age. The post-transfusion platelet count was 11,000/Nm ~. Two more units of random donor platelets were given, and the platelet count was 3,000/mm 3 two hours later. A two-volu~e exchange transfusion with two more units of random donor platelets increased the platelet count to 60,000/mmL Twelve hours later,

The Journal of Pediatrics October 1979 without further therapy, the platelet count was 25,000/mm ~. A transfusion of one unit of washed maternal platelets, given six hours later at 42 hours of age, raised the platelet count to 600,000/mmL Twelve hours post-transfusion, the platelet count was 481,000/mmL The platelet counts were 318,000 189,000, and 67,000, respectively, on the next three days. At the end of the third day, another unit of washed maternal platelets was given, raising the count from 67,000 to 1,118,000/mm :~. The platelet count remained normal after this second transfusion. Computerized cranial tomography at the time of admission demonstrated bifrontal and left posterior fossa cysts with marked obstructive hydrocephalus (Figure). The right frontal cyst communicated with the lateral ventricle. Several areas of calcification, most prominently tiear the left atrium, were noted. An air ventriculogram on the third day of life confirmed the obstructive hydrocephalus with no air beyond the third ventricle. The large communicating cyst on the right contained clear, xanthochromic fluid with hemosiderin-laden macrophages (protein 60 mg/dl, glucose 31 mg/dl). The loculated cyst on the left contained turbid brown fluid with innumerable RBCs arid hemosiderin-laden macrophages. There was no evidence of a bacterial or viral infection in the fluid obtained. Acute and convalescent serum titers and culture for cytomegalovirus, toxoplasma, rubella, and herpes were unremarkable in the patient and mother. Bacterial cultures produced no growth. Umbilical cord blood IgM level was 23 mg/dl (4 to 24 mg/dl). The patient had two tonic seizures on the second day of life. After treatment with phenobarbital, no subsequent seizures occurred. A ventriculoperitoneal shunt produced decompression of the hydrocephalus and communicating cyst, confirmed by follow-up CCT. The EEG, which was severely abnormal preoperatively, became normal by the time of discharge. Spasticity of the lower extremities, probable cortical blindness, and psychomotor retardation were noted on follow-up examination at 4 months of age. Patient 2. This child was born to a 20-year-old primigravida. The pregnancy was complicated by increasing edema in the third trimester for which hydrochlorothiazide was taken. The intraparturn course was complicated by uterine dystocia. Low forceps were applied to rotate the head. Multiple petechiae and a cephalohematoma were noted at birth. Meconium staining was noted, but the infant was vigff~ous: Although the weight was at the fiftieth percentile, the head circumference of 37.5 cm was above the ninety-seventh percentile. A hematoma appeared at the site of the vitamin K injection. Bilateral conjunctival hemorrhages were noted. The cry was vigorous, and the tone and neonatal reflexes were normal. Initial laboratory values, at 12 hours of life, included a platelet count of less than 3 000/ram ~ a prothrombin time of 12.1 seconds with a control of 11.6 seconds, and a partial thromboplastin time of 34.8 seconds. A lumbar puncture yielded bloody spinal fluid with approximately 35,000 RBC/mm ~in all four tubes. Two units of random donor platelets were infused, and the platelet count increased to 32,000/mm 3 two hours later. The platelet count fell again to less than 10;000/ram ~ within 12 hours. A two-volume exchange transfusion with fresh blood was then performed prior to the administration of two units of random donor platelets, and

Volume 95 Number 4

raised the platelet count to 192,000/mm ~. At the conclusion of the procedure, bradycardia and a tense fontanelle were noted. Left-sided clonic seizures ensued within 18 hours. An EEG showed right-sided spikes. The serum sodium concentration subsequently fell to 112 mEq/1; but returned to normal with fluid restriction. A bone marrow examination on the third day of life showed normal cellularity with normal granulocytic precursors, occasional erythrocyte percursors, and no blasts or megakaryocytes, although no bone spicules were seen. Platelet typing of the parents showed the mother to be PLA-1 negative, and the father to be PLA-1 positive. One unit of washed, maternal platelets was then administered, which raised the platelet count from 6,000/ mm 3 to 375,000/mm 3. The platelet count subsequently remained normal. CCT on the third day of life revealed a large right parieto-occipital hematoma producing a midline shift. Split cranial sutures and bulging fontanelle on the eighth day of life led to a second CCT, which showed the presence of hydrocephalus despite resolution of the hematoma. Ventricular fluid, obtained at that time, was brown and turbid (protein 1,46 gm/dl). Cultures and titers for cytomegalovirus, toxoplasma, rubella, and herpes, as well as bacterial cultures, were negative on this fluid. Despite repeated ventricular drainage of proteinaceous material, follow-up CCT showed a right parieto-occipital porencephalic cyst, with further enlargement of the ventricles. A ventriculoperitoneal shunt was performed to decompress hydrocephalus. Follow-up examination of this patient at 18 months of age revealed significant motor deficits in the form of deformity and weakness of the arms, with hypotonia of the legs and inability to walk. He was able to say four to five words, and did better in psychosocial than in motor activities. DISCUSSION isoimmune thrombocytopenia should be considered in the differential diagnosis when a newborn infant presents with hemorrhagic tendencies. 3 Both of our patients had perinatal hemorrhage in the presence of very low platelet counts, which responded dramatically to transfusion of maternal platelets. This response was obtained after finding a poor response to random donor platelets, and an 9intermediate response to the combination of exchange transfusion plus r a n d o m donor platelets. This pattern of response supports the proposed humoral mechanism for neonatal isoimmune thrombocytopenia. 1'4 Other causes of neonatal thrombocytopenia to be considered in these cases include congenital infection)' 6 neonatal leukemia, 7 and the maternal use of hydrochlorothiazide drugs? In the face of established isoimmune thrombocytopenia, the risk of intracranial hemorrhage is substantial. O f the 55 patients with isoimmune thrombocytopenia reviewed by Pearson et al, 1 there were eight proved and four suspected cases o f intracranial hemorrhage. Vaginal delivery and other perinatal stresses are known to increase the risk of intracranial bleeding in these infants. 2 Neither of our patients was known to be at risk for isoimmune

Brief clinical and laboratory observations

6 13

Figure. Computerized cranial tomography in Patient 1 demonstrating bifrontal cysts and obstructive hydrocephalus. The right frontal cyst communicates with the lateral ventricle.

platelet destruction. Although the vaginal deliveries may have contributed to the morbidity, it is likely that both infants had prior, in utero central nervous system insults. Both infants had enlarged heads with split sutures at birth. In Patient 1, the hemorrhage was demonstrated by C C T obtained at birth, and by the cyst fluid obtained on the third day of life. The large cyst with clear xanthochromic fluid and hemosiderin-laden macrophages may well have occurred as a result of bleeding several months earlier, in the late second trimester? 19 The liquefied, rusty fluid obtained on the eighth day of life in Patient 2 may have resulted from hemorrhage which is estimated pathologically to have occurred one to two weeks prenatally2 Although there has been a report of intracranial hemorrhage in utero in a patient with isoimmune hemolytic disease, 11 we know of no prior report of an in utero intracranial hemorrhage in a patient with isoimmune thrombocytopenia. Although both patients were promptly treated, both had congenital hydrocephalus and required ventriculoperitoneal shunts. Thus, it would seem prudent to continue to deliver infants at known risk for isoimmune thrombocytopenia by cesarean section, as suggested by Sitarz et al,'-' to decrease morbidity and mortality. H o w e v e r , ' i t is clear from our experience that this alone would not eliminate the risk of hemorrhagic central nervous system complications in infants with isoimmune thrombocytopenia. The authors are grateful to Dr. John T. Truman for his valuable assistance with the hematologic aspects of this paper. We thank Mrs. Cynthia A. Roback for assistance in typing this paper.

6 14

Brief clinical and laboratory observations

REFERENCES 1. Pearson HA, Schulman NR, Marder VJ, and Cone TE: Isoimmune neonatal thrombocytopenic purpura-clinical and therapeutic considerations, Blood 23:154, 1964. 2. Sitarz AL, Driscoll JM, and Wolf JA: Management of isoimmune neonatal thrombocytopenia, Am J Obstet Gynecol 124:39, 1976. 3. Hathaway WE: The bleeding newborn, Sem Hematol 12:175, 1975. 4. Adner MM, Fisch GR, Starobin SG, and Aster RH: Use of "compatible" p[atelet transfusions in the treatment of congenital isoimmune thrombocytopenic purpura, N Engl J Med 280:244, 1969. 5. Oski FA, and Naiman JL: Hematologic problems in the newborn, Philadelphia, 1966, WB Saunders Company, pp 27-28.

The Journal of Pediatrics October 1979

6.

Zinkham WH, Medearis DR, and Osborn JE: Blood and bone marrow findings in congenital rubella, J PEDIATR 71:512, 1967. 7. Stransky F, and Sarcia SR: On neonatal leukemia, Ann Pediatr 203:68, 1964. 8. Rodr!guez SU, LeNin SL, Hiller MC: Neonatal thrombocytopenia associated with antepartum administration of thiazide drugs, N Engl J Meal 270"881, 1964. 9. Tedeschi CG, Eckert WG, and Tedeschi LG: Forensic medicine: a study in trauma and environmental hazards, Philadelphia, 1977 WB Saunders Company, p 46. 10. Strassman G:. Hemosiderin and tissue iron in the brain, its relationship, occurrence, and importance: A study of 93 human brains, J Neuropathol Exp Neurol 4:393, 1945. 11. Bose C: Hydrops fetalis and in utero intracranial hemorrhage, J PEDIATR 93:1023, 1978.

Systemic vascular cotton fiber emboli in the neonate Kusuma Bavikatte, M.D., Janet Hilliard, D.O., Richard L. Schreiner, M.D.,* David Mirkin, M.D., Bruce Williams, M.D., James A. Lemons, M.D., and Edwin L. Gresham, M.D., Indianapolis, Ind.

V A R I O U S AUTHORS h a v e n o t e d cotton fiber emboli in organs of adults a n d children'-e~ however, to our knowledge there are n o r e p o r t e d cases in neonates. A review o f the literature suggests t h a t such cotton fiber emboli m i g h t not be innocuous. O u r s t u d y was m o t i v a t e d by finding foreign bodies r e s e m b l i n g cotton fibers d u r i n g routine autopsy o f two neonates. CASE REPORTS Patient 1. A postmature female infant weighing 3,487 gm was transferred to the intensive care nursery with a history of asphyxia and meconium aspiration. Umbilical arterial catheter placement was attempted both at the local hospital and at the intensive care nursery without success. A percutaneous right dorsalis pedis arterial catheter was then inserted. The infant had a total of 67 blood samples drawn from this source.during her ten days of hospitalization. Wet alcohol swabs (Kendall Hospital Products, Boston) were used to wipe the stopcock prior to drawing each blood sample. Fluids were also infused through a scalp vein so that intravenous medications could be administered by means of an in-line three-way stopcock. On the tenth day the infant had a cardiac arrest. At that time an umbilical venous catheter was inserted for the infusion of emergency drugs, but the infant failed to respond to vigorous resuscitative efforts. An

From the Section of Neonatal-Perinatal Medicine, Department of Pediatrics and the Section of.Pediatric Pathology, Department of Pathology, Indiana University School of Medicine, and the Jarhes Whitcomb Riley Hospital for Children. *Reprint address: Department of Pediatrics, Indiana University School of Medicine, 1100 W. Michigan St., Indianapolis, IN 46223.

autopsy revealed severe pneumonia and, quite unexpectedly, refractile foreign bodies in the pulmonary arteries (Fig. 1). Patient 2. This term female infant was transferred to the intensive care nursery with a history of amnionitis, low Apgar scores, and respiratory distress. An umbilical venous catheter had been inserted at the referring hospital but was replaced on admission by an umbilical artery catheter with the tip at the level of the third lumbar vertebra. During the hospital course a minimum of 31 arterial blood samples were withdrawn through the line. An alcohol swab was always used to wipe off the stopcock prior to blood letting. On the second day of life the infant developed seizures and renal failure, and she died on the following day. At autopsy .diffuse pulmonary consolidation and hemorrhage and cerebral softening were found. An unusual observation was the presence of thrombi in small hepatic arteries surrounding birefringent material resembli~ cotton fiber. MATERIALS

AND

METHODS

This study was predicated o n the hypothesis t h a t the foreign bodies f o u n d on autopsy originated from the alcohol swabs used to wipe the stopcocks prior to infusion o f m e d i c a t i o n s or withdrawal o f b l o o d samples. A n i n t r a v e n o u s fluid system identical to that used in the nursery was set up in the laboratory. T h e system consisted of: (1) a n e w bottle (500 ml) o f dextrose 10% in water (Abbott, Chicago); (2) i n t r a v e n o u s fluid t u b i n g ( A b b o t t ) ; (3) t h r e e - w a y stopcock (Medex, Hilliard, Ohio); a n d (4) a 5 - F r e n c h umbilical artery catheter ( S h e r w o o d Medical, St. Louis). Fluid was allowed to r u n t h r o u g h the t u b i n g a n d the catheter. Two milliliters o f this fluid were d r a w n i n t o a syringe a n d injected into a 0.22 m i c r o n filter

0022'3476/79/100614+ 03500.30/0 9 1979 The C. V. Mosby Co.