- Email: [email protected]
Persistent antithrombin III deficiency: Risk factor for thromboembolic complications in neonates small for gestational age
Persistent antithrombin III deficiency: Risk factor for thromboembolic complications in neonates small for gestational age Coagulation and fibrinolytic factors were investigated daily in 24 SGA and 26 AGA neonates. The results were correlated with placental form and structure and with hematologic values. In the SGA infants, a higher incidence of placental infarction (P < 0.01), polycythemia (P < 0.005), and thromboeytopenia (P < 0.001) was present. During the first 9 days, the mean antithrombin III level in the AGA group increased from 0.36 to 0.53 U/ml, whereas in SGA neonates this value was initially significantly lower (0.27 U/ml) and remained at that level for the entire observation period. The same pattern was found for c~e-antiplasmin. The persistent AT-Ill deficiency and the reduced blood flow associated with polycythemia may explain the increased risk of thromboembolic complications in SGA infants described by others. (J PeDZArR 105:310, 1984)
M. Peters, M.D., J. W. ten Cate, M.D., Ph.D., L. H. Koo, and C. Breederveld, M.D., Ph.D. A m s t e r d a m ,
The Netherlands
NEONATES WHO ARE SMALL FOR GESTATIONAL AGE frequently (17% to 50%) have polycythemia~-3 related to development during chronic_hypoxia, which is caused by placental dysfunction: Polycythemia in these infants is the result of shunting of the hematopoietic stem cells in the direction of erythropoiesis at the expense of thrombopoiesis and leukopoiesis)-7 Neonatal polycythemia is the primary cause of hyperviscosity, 1 which results in a reduced blood flow within the microcirculation. Several associated clinical symptoms have been observed in neonates with this syndrome, including cardiorespiratory distress, L2,8 transient renal failure,9, ~0 neurologic disorders, 1,2,7,~0 and necrotizing enterocolitis. H In addition to these clinical symptoms, thrombotic complications have been seen in association with polycythemia:. ~2It is unknown whether these clinical symptoms are the result of hyperviscosity alone or whether fibrin thrombi within the microcirculation contribute to them. From the Department of Hematology, Division of Hemostasis and Thrombosis; and the Department of Pediatrics, University Hospital of Amsterdam, Academic Medical Centre. Supported by Grant 32-80-21 from the Astma Foundation. Submitted for publication Sept. 9, 1983; accepted Feb. 3, 1984. Reprint requests: M. Peters, M.D., Department of Hematology, Division of Hemostasis and Thrombosis, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.
310
The Journal of P E D I A T R I C S
Only a few coagulation studies in SGA neonates have been reported. The results of these investigations suggest that intravascular coagulation does occur in such patients ,3. ,4 although actual fibrin deposition has not been described. In a previous study in preterm infants with idiopathic respiratory distress syndrome, low A T - I l l levels were found to be associated with microthrombi in vital organs, '5 and were also observed in SGA neonates. AT-III FDP/fdp
Antithrombin III Fibrin/fibrinogen degradation products
I
I
I
Because low AT-III levels of either congenital or acquired origin are known to be a risk factor in causing thromboembolic complications,16 we undertook a prospective study of coagulation and fibrinolytic and hematologic measurements in AGA and SGA neonates during the first 9 days of life. In both groups, macroscopic placental form and structure were also studied.
METHODS Patients. Fifty consecutively born infants were included in one of two groups: a study group comprising 24 SGA neonates (14 boys) and a control group consisting of 26 AGA neonates (14 boys). The mean birth weights of the AGA and SGA groups were 2120 gm (range 1460 to 3190 gin) and 1520 gm (range 920 to 2120 gm) (P < 0.01),
Volume 105 Number 2
respectively, and the mean gestational ages of these groups were 33.5 weeks (range 31 tO 37 weeks) and 34.7 weeks (range 31 to 39 weeks) (NS), respectively. SGA was defined as birth weight <10th centile on the intrauterine growth chart.~7 The birth weight of nine S G A neonates was <2.3 centile. A possible explanation of the intrauterine growth retardation could be found in 11 of the 24 infants (preexistent hypertension in two, preeclampsia in nine). The umbilical cord was clamped within 3 minutes after birth in all infants. Birth asPhyxia (1- and 5-minute Apgar scores <7, and cord blood pH <7.25) was present in eight SGA neonates a n d in seven A G A neonates. The mean 1and 5-minute Apgar scores in the nonasphyxiated S G A group were 7.8 and 9.2, respectively, and in the A G A group were 8.5 and 9.7, respectively. For placental and hematologic investigations, all infants were included. To investigate the influence of intrauterine growth retardation on coagulation and fibrinolytic values, only nonasphyxiated neonates were included in the analysis. Excluded from the study were those infants with complications other than birth asphyxia (i.e., idiopathic respiratory distress syndrome, septicemia, ABO or rhesus Sensitization, and chromosomal abnormalities). All infants survived the study period. Major symptoms of thromboembolism, cardiorespiratory distress, neurologic disorders, or necrotizing enterocolitis were not observed in any of the nonasphyxiated SGA neonates. None of the polycythemic infants received either (partial) plasma exchange transfusion or other plasma products. The study was approved by the Medical Ethical Committee of the University Hospital of Amsterdam. Placental morphology. Twenty-five of the 26 placentas in the A G A group and all placentas of the SGA infants were macroscopically examined by the same experienced investigator (B.L.H.), who was unaware of the patients' hematologic and coagulation status. Placental infarction was expressed .as a percentage loss of tissue. Grade 0 (no infarction) and grade 1 (<10%) were considered normal, whereas grade 2 (11% to 25%) and grade 3 (>25%) were defined as abnormalY The weights were expressed in eentiles on a placental growth curve. A centile <10 was defined as abnormal. Laboratory methods. The first blood sample was obtained within 6 hours after birth: subsequent samples were taken along with routine daily blood collection until the ninth day after birth. Blood (0.5 ml) was collected into polypropylene tubes (Greiner, Niirtingen, West Germany) containing solid ks EDTA (1.5 mg/ml), which was usedfor hematologic as well as for coagulation studies. Venous and capillary EDTA blood samples were suitable for hemostatic measurements, as described previous-
Persistent antithrombin I I I deficiency in S G A neonates
3 11
Table I. Clinical data AGA neonates (n = 26)
SGA neonates (n = 24)
3
11
1
9
6
14
2
15
Polycythemia (Ht >0.65 L/L) Leukocytopenia (WBC <4,000 ~1) Normoblastosis (>20 cells/100 WBC) Thrombocytopenia (<100,000 M)
Table II. Coagulation and fibrinolytic parameters on the first day of life in nonasphyxiated newborn infants t U/ml
AT-Ill Factor II Factor X Plasminogen a2-Antiplasmin
AGA neonates (n = 19)
0.36 0.32 0.39 0.38 0.61
_+ 0.07 ___0.07 4- 0.08 ___0.09 -+ 0.19
SGA neonates (n = 16)
0.27 0.32 0.39 0.39 0.52
+ 0.06* _+ 0.07 _ 0.09 _+ 0.10 - 0.25
*Values are mean _+_SD. P < 0.01 as comparedwith AGA neonates.
ly. 19 Microhematocrit levels were determined in duplicate by standard methods, and WBC using a Coulter electronic counter. Normoblasts were counted relative to 100 WBC. Platelet counts were perform~:l according to Feissly and Lfidin3~ Polycythemia was defined as hematocrit value >0.65 L / L , ~ thrombocyt0penia as platelet count <100,000/~1, ~ leUkocytopenia as WBC <4000 #!, and normoblastosis as normoblast count >20/100 WBC. Plasma was prepared by centrifugation at 13,000 x g for 4 minutes a t room temperature. Automated chromogenic determinations of AT-III, factors II and X, plasminogeni and a2=antiplasmin could be performed with 70 #1 plasma. 19 Normal adult ranges of these assays are 0.80 to 1.40 U/ml. Statistical methods. Statistical evaluation of the data was performed using the one-way analysis of variance, the Student t test, and the chi-square test. P values <0.05 were considered significant. RESULTS Placental morphology. Fourteen of 24 SGA placentas and none of the A G A placentas weighed <10th centile. Infarction grades 2 and 3 were present in 11 SGA placentas, and not present in the A G A placentas. Five SGA placentas were diagnosed as noirmal (weight > 10th centile with concurrent infarction grade <2). Hematologic studies (Table I). A significantly higher
3 12
(
Peters et al.
0.80
The Journal o f Pediatrics August 1984
newborn infants (0.36 U / m l ) (P < 0.01). Initial A T - I l l levels <0.25 U / m l were observed in eight of 16 nonasphyxiated SGA neonates and in three of 19 nonasphyxiated A G A infants. Over the course of the study, the mean A T - I l l level in the A G A group increased to 0.53 U/ml, whereas in the SGA group it remained at a lower level (Figure, A) (P < 0.01 on days 1, 3, 4, and 6 to 9). An increased a2-antiplasmin level to almost normal adult values was noted in the A G A infants after day 1. The concentration of this fibrinolysis inhibitor remained at the initial low level in the SGA group (Figure, B) (P < 0.01 on days 6 and 7). No statistically significant differences were detected in the plasma levels of plasminogen and factors II and X in either group of neonates during the observation period (Table II) (P > 0.5).
Antithrombin U/ml
~)0.70
0.60
0.50
0.40
0.30
0.20
0.10
~1
0
I
- - I
2
3
I--
/
I
4
5
6 time (days)
DISCUSSION
Q
1.000.90
Antiplasmin U/m1
0.800.70 0.60 0.50" 0.40. s Q20
time (days)
Figure. Course of A T - H I and ~x2-antip]asmin in nonasphy• A G A ( o ~ o ) and SGA neonates ( H ) . Values are mean _+ SD.
incidence of polycythemia (P < 0.005), leukocytopenia (P < 0.001), normoblastosis (P < 0.02), and thrombocytopenia (P < 0.001) was found in the SGA neonates as compared with the A G A group. Seven of 16 nonasphyxiated SGA infants had hematocrit values >65%. Polycythemia with concurrent thrombocytopenia was the most consistent combination and was present in eight SGA infants. No correlation was found between either macroscopic infarction grade or placental weight and polycythemia (P > 0.1). All SGA neonates had normal hematologic values 5 days after birth. Coagulation investigations (Table II). On the first day the mean A T - I l l level (0.27 U/ml) was significantly lower in the SGA group as compared with that in the A G A
Placental form and structure and hematologic and coagulation measurements were prospectively investigated in consecutively born SGA and A G A neonates. A significantly lower placental weight and a higher incidence of infarction grades 2 and 3 were demonstrated in the SGA group. Scott and Jordan 4 found this phenomenon to be associated with placental insufficiency, and thus a concurrent chronic hypoxic state of the fetus. A study in hypoxic mice revealed that impaired oxygenation of the fetus interferes with intrauterine growth? Bone marrow analyses in these animals showed a decreased number of megakaryocytes during the first 4 days of life. Thrombocytopenia and polycythemia were found in the peripheral blood. Bone marrow investigations in four polycythemic SGA neonates revealed a high normoblast count and reduced megakaryopoiesis and leukopoiesis (Brubakk AM, Ruys JH: Personal communication, 1981). In our study, polycythemia with concurrent thrombocytopenia was present in many SGA neonates but not in any A G A infants. The grade of placental infarction and the occurrence of polycythemia did not correlate significantly (P > 0.5). Rivers 6 described three A G A infants (gestational a g e >~35 weeks) with polycythemia combined with thrombocytopenia and circulating fibrin monomers; the findings were thought to be compatible with low-grade intravascular coagulation. Henriksson 22 investigated a heterogeneous group of 15 A G A and SGA neonates, all having hyperviscosity Syndrome related to polycythemia. No fibrin/ fibrinogen degradation products and normal levels of fibrinogen and factors V and VIII were demonstrated. The mean A T - I l l level, however, was decreased, as compared with adult values. Katz et al. 7 investigated 20 polycythemic newborn infants and found coagulation tests and clotting factor levels appropriate for their age. Perlman and Dvilansky ~3
Volume 105 Number 2
found prolonged prothrombin time, partial thromboplastin time, and thrombin time; diminished platelet counts; and increased F D P / f d p in 15 S G A neonates as compared with a control group. Weissbach et al. TM studied 24 S G A neonates and measured significantly lower fibrinogen and progressive A T - I I I and factor V levels as compared to those in an A G A group; increased FDP was found in some of the S G A infants. Both authors suggested that their findings, which were detected in the first days of life, are indicative of a consumptive coagulopathy. Controversy exists as to whether coagulation abnormalities play an additional role in the hyperviscosity syndrome. The studies have been performed in heterogeneous groups of polycythemic newborn infants or in homogeneous groups of S G A infants, in whom it was unknown whether polycythemia was present. W e compared a homogeneous group of S G A neonates with an appropriate control group of A G A infants. Significantly lower levels of A T - I I I were observed in the S G A group during the entire study period. Concurrently, decreased levels of a2-antiplasmin were present. Thus the inhibitors of coagulation (AT-III) as well as of the fibrinolytic system (az-antiplasmin) were relatively deficient in S G A neonates as compared with the A G A group. The cause of the decreased plasma levels of A T - I I I and of a2-antiplasmin is unknown. Perlman and Dvilansky la speculated that their results were caused by the infusion of placental thromboplastic material in the fetal circulation, resulting from placental infarction. W e could not find a correlation between the grade of infarction and the A T - I I I levels within the S G A group ( r = 0.05) (data not shown). Reduced blood flow resulting from hyperviscosity initiates hypoxia and local acidosis. Both of these phenomena contribute to activation of the coagulation system. 23,24 Venous stasis with concurrent local hypoxia and acidosis causes increased secretion of tissue plasminogen activators, 25 which induces activation of the fibrinolytic system and consumption of its inhibitor through the formation of plasmin-a2-antiplasmin complexes. We could not demonstrate any correlative relationship between the incidence of polycythemia and the decreased levels of either A T - I I I or a2-antiplasmin levels (data not shown). Our investigations did not show any evidence that could contribute to one of the above-mentioned pathophysiologic mechanisms, although more refined tests to prove disseminated intravascular coagulation were not included in this study. Our findings suggest an increased risk of thromboembolic complications in S G A neonates related to the combined mechanisms of polycythemia and relatively insufficient levels of AT-III.
Persistent antithrombin I I I deficiency in S G A neonates
3 13
We thank Prof. W. E. Hathaway for critical review of and valuable comments on the content of thls manuscript; Dr. J. G. Koppe for allowing the study infants in the neonatal intensive care unit; and Mrs. B. L. Huidekoper for examination of the placentas. REFERENCES 1. Gross GP, Hathaway WE, McGaughey HR: Hyperviscosity in the neonate. J PEDIATR82:1004, 1973. 2. Humbert JR, Abelson H, Hathaway WE, Battaglia FC: Polycythemia in small for gestational age infants. J PEDIATR 75:812, 1969. 3. Hakanson DO, Oh W: Hyperviscosity in small for gestational age infants. Biol Neonat e 37:109, 1980. 4. Scott JM, Jordan JM: Placental insufficiency and the small -for-dates baby. Am J Obstet Gynecol 113:823, 1972. 5. Meberg A: Transitory thrombocytopenia in newborn mice after intrauterine hypoxia. Pediatr Res 14:1071, 1980. 6. Rivers RPA: Coagulation changes associated with a high haematocrit in the newborn infant. Acta Pediatr Scand 64:449, 1975. 7. Katz J, Rodriquez E, Mandani G, Branson HE: Normal coagulation findings, thrombocytopenia, and peripheral hemoconcentration in neonatal polycythemia. J PEDIATR 101:99, 1982. 8. Gatti RA, Muster A J, Cole RB, Paul HM: Neonatal polycythemia with transient cyanosis and cardiorespiratory abnormalities. J PEDIATR69:1063, 1966. 9. Aperia A, Bergquist G, Broberger O, Thodenius K, Zetterstrfm R: Renal function in newborn infants with high hematocrit values before and after isovolemic hemodilution. Acta Paediatr Scand 63:878, 1974. 10. Herson VC, Ray JR, Rowe JC, Phillipps AF: Acute renal failure associated with polycythemia in a neonate. J PEDIATR 100:137, 1982. 11. Hakanson DO, Oh W: Necrotizing enterocolitis and hyperviscosity in the newborn infant. J PEDIATR90:458, 1977. 12. Papageorgiou A, Stern L: Polycythemia and gangrene of an extremity in a newborn infant. J PED1ATR81:985, 1972. 13. Perlman M, Dvilansky A: Blood coagulation status of smallfor-dates and post-mature infants. Arch Dis Child 50:424, 1975. 14. Weissbach G, Lenk H, Vogtmann C, Donula H, B6ttcher H: Gerinnungsparameter bei Neugeborenen mit Symptomen der Dysmaturit~it. Wiss Z Karl Marx Univ Leipzig 25:155, 1976. 15. Peters M, ten Cate JW, Breederveld C, de Leeuw R, Emeis J J, Koppe JG: Low antithrombin 111 levels in neonates with IRDS: Poor prognosis. Pediatr Res (In press.) 16. Egeberg O: Inherited antithrombin deficiency causing thrombophilia. Thromb Diath Haemorrh 13:516, 1965. 17. Kloosterman G J: On intrauterine growth: The significance of prenatal care. Int J Gynecol Obstet 8:895, 1970. 18. Kloosterman G J, Huidekoper BL: The significance of the placenta in "Obstetrical mortality": A study of 2000 births. Gynaecologia 138:529, 1954. 19. Peters M, Breederveld C, Kahl6 LH, ten Cate JW: Rapid microanalysis of coagulation parameters by automated chromogenic substrated methods: Application in neonatal patients. Thromb Res 28:773, 1982. 20. Feissly R, Lfidin H: Microscopie par contrastes de phases. Ill. Application ~t l'h6matologie. Rev Hematol 4:481, 1949.
314
Peters et al.
21. Forfar JO, Arneil CC: Textbook of paediatrics, ed 2. London, 1978, Churchill Livingstone, p 177. 22. Henriksson R: Hyperviscosity of the blood and hemostasis in the newborn infant. Acta Paediatr Stand 68i701, 1979. 23. Broersma R J, Bullemer GD, Mammen EF: Acidosis-induced disseminated intravascular microthrombi and its dissolution by streptokinase. Thromb Diath Haemorrh 24:55, 1970.
The Journal of Pediatrics August 1984
24. Kisker CT, Robillard JE, Clarke WR: Blood coagulation changes following hypoxemia in the near-term fetal lamb. Pediatr Res 16:732, 1982. 25. Wiman B, Melbring G, R/mby M: Plasminogen activator release during venous stasis and exercise as determined by a new specific assay. Clin Chim Acta 127:279, 1983.
M. Peters, M.D., J. W. ten Cate, M.D., Ph.D., L. H. Koo, and C. Breederveld, M.D., Ph.D. A m s t e r d a m ,
The Netherlands
NEONATES WHO ARE SMALL FOR GESTATIONAL AGE frequently (17% to 50%) have polycythemia~-3 related to development during chronic_hypoxia, which is caused by placental dysfunction: Polycythemia in these infants is the result of shunting of the hematopoietic stem cells in the direction of erythropoiesis at the expense of thrombopoiesis and leukopoiesis)-7 Neonatal polycythemia is the primary cause of hyperviscosity, 1 which results in a reduced blood flow within the microcirculation. Several associated clinical symptoms have been observed in neonates with this syndrome, including cardiorespiratory distress, L2,8 transient renal failure,9, ~0 neurologic disorders, 1,2,7,~0 and necrotizing enterocolitis. H In addition to these clinical symptoms, thrombotic complications have been seen in association with polycythemia:. ~2It is unknown whether these clinical symptoms are the result of hyperviscosity alone or whether fibrin thrombi within the microcirculation contribute to them. From the Department of Hematology, Division of Hemostasis and Thrombosis; and the Department of Pediatrics, University Hospital of Amsterdam, Academic Medical Centre. Supported by Grant 32-80-21 from the Astma Foundation. Submitted for publication Sept. 9, 1983; accepted Feb. 3, 1984. Reprint requests: M. Peters, M.D., Department of Hematology, Division of Hemostasis and Thrombosis, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.
310
The Journal of P E D I A T R I C S
Only a few coagulation studies in SGA neonates have been reported. The results of these investigations suggest that intravascular coagulation does occur in such patients ,3. ,4 although actual fibrin deposition has not been described. In a previous study in preterm infants with idiopathic respiratory distress syndrome, low A T - I l l levels were found to be associated with microthrombi in vital organs, '5 and were also observed in SGA neonates. AT-III FDP/fdp
Antithrombin III Fibrin/fibrinogen degradation products
I
I
I
Because low AT-III levels of either congenital or acquired origin are known to be a risk factor in causing thromboembolic complications,16 we undertook a prospective study of coagulation and fibrinolytic and hematologic measurements in AGA and SGA neonates during the first 9 days of life. In both groups, macroscopic placental form and structure were also studied.
METHODS Patients. Fifty consecutively born infants were included in one of two groups: a study group comprising 24 SGA neonates (14 boys) and a control group consisting of 26 AGA neonates (14 boys). The mean birth weights of the AGA and SGA groups were 2120 gm (range 1460 to 3190 gin) and 1520 gm (range 920 to 2120 gm) (P < 0.01),
Volume 105 Number 2
respectively, and the mean gestational ages of these groups were 33.5 weeks (range 31 tO 37 weeks) and 34.7 weeks (range 31 to 39 weeks) (NS), respectively. SGA was defined as birth weight <10th centile on the intrauterine growth chart.~7 The birth weight of nine S G A neonates was <2.3 centile. A possible explanation of the intrauterine growth retardation could be found in 11 of the 24 infants (preexistent hypertension in two, preeclampsia in nine). The umbilical cord was clamped within 3 minutes after birth in all infants. Birth asPhyxia (1- and 5-minute Apgar scores <7, and cord blood pH <7.25) was present in eight SGA neonates a n d in seven A G A neonates. The mean 1and 5-minute Apgar scores in the nonasphyxiated S G A group were 7.8 and 9.2, respectively, and in the A G A group were 8.5 and 9.7, respectively. For placental and hematologic investigations, all infants were included. To investigate the influence of intrauterine growth retardation on coagulation and fibrinolytic values, only nonasphyxiated neonates were included in the analysis. Excluded from the study were those infants with complications other than birth asphyxia (i.e., idiopathic respiratory distress syndrome, septicemia, ABO or rhesus Sensitization, and chromosomal abnormalities). All infants survived the study period. Major symptoms of thromboembolism, cardiorespiratory distress, neurologic disorders, or necrotizing enterocolitis were not observed in any of the nonasphyxiated SGA neonates. None of the polycythemic infants received either (partial) plasma exchange transfusion or other plasma products. The study was approved by the Medical Ethical Committee of the University Hospital of Amsterdam. Placental morphology. Twenty-five of the 26 placentas in the A G A group and all placentas of the SGA infants were macroscopically examined by the same experienced investigator (B.L.H.), who was unaware of the patients' hematologic and coagulation status. Placental infarction was expressed .as a percentage loss of tissue. Grade 0 (no infarction) and grade 1 (<10%) were considered normal, whereas grade 2 (11% to 25%) and grade 3 (>25%) were defined as abnormalY The weights were expressed in eentiles on a placental growth curve. A centile <10 was defined as abnormal. Laboratory methods. The first blood sample was obtained within 6 hours after birth: subsequent samples were taken along with routine daily blood collection until the ninth day after birth. Blood (0.5 ml) was collected into polypropylene tubes (Greiner, Niirtingen, West Germany) containing solid ks EDTA (1.5 mg/ml), which was usedfor hematologic as well as for coagulation studies. Venous and capillary EDTA blood samples were suitable for hemostatic measurements, as described previous-
Persistent antithrombin I I I deficiency in S G A neonates
3 11
Table I. Clinical data AGA neonates (n = 26)
SGA neonates (n = 24)
3
11
1
9
6
14
2
15
Polycythemia (Ht >0.65 L/L) Leukocytopenia (WBC <4,000 ~1) Normoblastosis (>20 cells/100 WBC) Thrombocytopenia (<100,000 M)
Table II. Coagulation and fibrinolytic parameters on the first day of life in nonasphyxiated newborn infants t U/ml
AT-Ill Factor II Factor X Plasminogen a2-Antiplasmin
AGA neonates (n = 19)
0.36 0.32 0.39 0.38 0.61
_+ 0.07 ___0.07 4- 0.08 ___0.09 -+ 0.19
SGA neonates (n = 16)
0.27 0.32 0.39 0.39 0.52
+ 0.06* _+ 0.07 _ 0.09 _+ 0.10 - 0.25
*Values are mean _+_SD. P < 0.01 as comparedwith AGA neonates.
ly. 19 Microhematocrit levels were determined in duplicate by standard methods, and WBC using a Coulter electronic counter. Normoblasts were counted relative to 100 WBC. Platelet counts were perform~:l according to Feissly and Lfidin3~ Polycythemia was defined as hematocrit value >0.65 L / L , ~ thrombocyt0penia as platelet count <100,000/~1, ~ leUkocytopenia as WBC <4000 #!, and normoblastosis as normoblast count >20/100 WBC. Plasma was prepared by centrifugation at 13,000 x g for 4 minutes a t room temperature. Automated chromogenic determinations of AT-III, factors II and X, plasminogeni and a2=antiplasmin could be performed with 70 #1 plasma. 19 Normal adult ranges of these assays are 0.80 to 1.40 U/ml. Statistical methods. Statistical evaluation of the data was performed using the one-way analysis of variance, the Student t test, and the chi-square test. P values <0.05 were considered significant. RESULTS Placental morphology. Fourteen of 24 SGA placentas and none of the A G A placentas weighed <10th centile. Infarction grades 2 and 3 were present in 11 SGA placentas, and not present in the A G A placentas. Five SGA placentas were diagnosed as noirmal (weight > 10th centile with concurrent infarction grade <2). Hematologic studies (Table I). A significantly higher
3 12
(
Peters et al.
0.80
The Journal o f Pediatrics August 1984
newborn infants (0.36 U / m l ) (P < 0.01). Initial A T - I l l levels <0.25 U / m l were observed in eight of 16 nonasphyxiated SGA neonates and in three of 19 nonasphyxiated A G A infants. Over the course of the study, the mean A T - I l l level in the A G A group increased to 0.53 U/ml, whereas in the SGA group it remained at a lower level (Figure, A) (P < 0.01 on days 1, 3, 4, and 6 to 9). An increased a2-antiplasmin level to almost normal adult values was noted in the A G A infants after day 1. The concentration of this fibrinolysis inhibitor remained at the initial low level in the SGA group (Figure, B) (P < 0.01 on days 6 and 7). No statistically significant differences were detected in the plasma levels of plasminogen and factors II and X in either group of neonates during the observation period (Table II) (P > 0.5).
Antithrombin U/ml
~)0.70
0.60
0.50
0.40
0.30
0.20
0.10
~1
0
I
- - I
2
3
I--
/
I
4
5
6 time (days)
DISCUSSION
Q
1.000.90
Antiplasmin U/m1
0.800.70 0.60 0.50" 0.40. s Q20
time (days)
Figure. Course of A T - H I and ~x2-antip]asmin in nonasphy• A G A ( o ~ o ) and SGA neonates ( H ) . Values are mean _+ SD.
incidence of polycythemia (P < 0.005), leukocytopenia (P < 0.001), normoblastosis (P < 0.02), and thrombocytopenia (P < 0.001) was found in the SGA neonates as compared with the A G A group. Seven of 16 nonasphyxiated SGA infants had hematocrit values >65%. Polycythemia with concurrent thrombocytopenia was the most consistent combination and was present in eight SGA infants. No correlation was found between either macroscopic infarction grade or placental weight and polycythemia (P > 0.1). All SGA neonates had normal hematologic values 5 days after birth. Coagulation investigations (Table II). On the first day the mean A T - I l l level (0.27 U/ml) was significantly lower in the SGA group as compared with that in the A G A
Placental form and structure and hematologic and coagulation measurements were prospectively investigated in consecutively born SGA and A G A neonates. A significantly lower placental weight and a higher incidence of infarction grades 2 and 3 were demonstrated in the SGA group. Scott and Jordan 4 found this phenomenon to be associated with placental insufficiency, and thus a concurrent chronic hypoxic state of the fetus. A study in hypoxic mice revealed that impaired oxygenation of the fetus interferes with intrauterine growth? Bone marrow analyses in these animals showed a decreased number of megakaryocytes during the first 4 days of life. Thrombocytopenia and polycythemia were found in the peripheral blood. Bone marrow investigations in four polycythemic SGA neonates revealed a high normoblast count and reduced megakaryopoiesis and leukopoiesis (Brubakk AM, Ruys JH: Personal communication, 1981). In our study, polycythemia with concurrent thrombocytopenia was present in many SGA neonates but not in any A G A infants. The grade of placental infarction and the occurrence of polycythemia did not correlate significantly (P > 0.5). Rivers 6 described three A G A infants (gestational a g e >~35 weeks) with polycythemia combined with thrombocytopenia and circulating fibrin monomers; the findings were thought to be compatible with low-grade intravascular coagulation. Henriksson 22 investigated a heterogeneous group of 15 A G A and SGA neonates, all having hyperviscosity Syndrome related to polycythemia. No fibrin/ fibrinogen degradation products and normal levels of fibrinogen and factors V and VIII were demonstrated. The mean A T - I l l level, however, was decreased, as compared with adult values. Katz et al. 7 investigated 20 polycythemic newborn infants and found coagulation tests and clotting factor levels appropriate for their age. Perlman and Dvilansky ~3
Volume 105 Number 2
found prolonged prothrombin time, partial thromboplastin time, and thrombin time; diminished platelet counts; and increased F D P / f d p in 15 S G A neonates as compared with a control group. Weissbach et al. TM studied 24 S G A neonates and measured significantly lower fibrinogen and progressive A T - I I I and factor V levels as compared to those in an A G A group; increased FDP was found in some of the S G A infants. Both authors suggested that their findings, which were detected in the first days of life, are indicative of a consumptive coagulopathy. Controversy exists as to whether coagulation abnormalities play an additional role in the hyperviscosity syndrome. The studies have been performed in heterogeneous groups of polycythemic newborn infants or in homogeneous groups of S G A infants, in whom it was unknown whether polycythemia was present. W e compared a homogeneous group of S G A neonates with an appropriate control group of A G A infants. Significantly lower levels of A T - I I I were observed in the S G A group during the entire study period. Concurrently, decreased levels of a2-antiplasmin were present. Thus the inhibitors of coagulation (AT-III) as well as of the fibrinolytic system (az-antiplasmin) were relatively deficient in S G A neonates as compared with the A G A group. The cause of the decreased plasma levels of A T - I I I and of a2-antiplasmin is unknown. Perlman and Dvilansky la speculated that their results were caused by the infusion of placental thromboplastic material in the fetal circulation, resulting from placental infarction. W e could not find a correlation between the grade of infarction and the A T - I I I levels within the S G A group ( r = 0.05) (data not shown). Reduced blood flow resulting from hyperviscosity initiates hypoxia and local acidosis. Both of these phenomena contribute to activation of the coagulation system. 23,24 Venous stasis with concurrent local hypoxia and acidosis causes increased secretion of tissue plasminogen activators, 25 which induces activation of the fibrinolytic system and consumption of its inhibitor through the formation of plasmin-a2-antiplasmin complexes. We could not demonstrate any correlative relationship between the incidence of polycythemia and the decreased levels of either A T - I I I or a2-antiplasmin levels (data not shown). Our investigations did not show any evidence that could contribute to one of the above-mentioned pathophysiologic mechanisms, although more refined tests to prove disseminated intravascular coagulation were not included in this study. Our findings suggest an increased risk of thromboembolic complications in S G A neonates related to the combined mechanisms of polycythemia and relatively insufficient levels of AT-III.
Persistent antithrombin I I I deficiency in S G A neonates
3 13
We thank Prof. W. E. Hathaway for critical review of and valuable comments on the content of thls manuscript; Dr. J. G. Koppe for allowing the study infants in the neonatal intensive care unit; and Mrs. B. L. Huidekoper for examination of the placentas. REFERENCES 1. Gross GP, Hathaway WE, McGaughey HR: Hyperviscosity in the neonate. J PEDIATR82:1004, 1973. 2. Humbert JR, Abelson H, Hathaway WE, Battaglia FC: Polycythemia in small for gestational age infants. J PEDIATR 75:812, 1969. 3. Hakanson DO, Oh W: Hyperviscosity in small for gestational age infants. Biol Neonat e 37:109, 1980. 4. Scott JM, Jordan JM: Placental insufficiency and the small -for-dates baby. Am J Obstet Gynecol 113:823, 1972. 5. Meberg A: Transitory thrombocytopenia in newborn mice after intrauterine hypoxia. Pediatr Res 14:1071, 1980. 6. Rivers RPA: Coagulation changes associated with a high haematocrit in the newborn infant. Acta Pediatr Scand 64:449, 1975. 7. Katz J, Rodriquez E, Mandani G, Branson HE: Normal coagulation findings, thrombocytopenia, and peripheral hemoconcentration in neonatal polycythemia. J PEDIATR 101:99, 1982. 8. Gatti RA, Muster A J, Cole RB, Paul HM: Neonatal polycythemia with transient cyanosis and cardiorespiratory abnormalities. J PEDIATR69:1063, 1966. 9. Aperia A, Bergquist G, Broberger O, Thodenius K, Zetterstrfm R: Renal function in newborn infants with high hematocrit values before and after isovolemic hemodilution. Acta Paediatr Scand 63:878, 1974. 10. Herson VC, Ray JR, Rowe JC, Phillipps AF: Acute renal failure associated with polycythemia in a neonate. J PEDIATR 100:137, 1982. 11. Hakanson DO, Oh W: Necrotizing enterocolitis and hyperviscosity in the newborn infant. J PEDIATR90:458, 1977. 12. Papageorgiou A, Stern L: Polycythemia and gangrene of an extremity in a newborn infant. J PED1ATR81:985, 1972. 13. Perlman M, Dvilansky A: Blood coagulation status of smallfor-dates and post-mature infants. Arch Dis Child 50:424, 1975. 14. Weissbach G, Lenk H, Vogtmann C, Donula H, B6ttcher H: Gerinnungsparameter bei Neugeborenen mit Symptomen der Dysmaturit~it. Wiss Z Karl Marx Univ Leipzig 25:155, 1976. 15. Peters M, ten Cate JW, Breederveld C, de Leeuw R, Emeis J J, Koppe JG: Low antithrombin 111 levels in neonates with IRDS: Poor prognosis. Pediatr Res (In press.) 16. Egeberg O: Inherited antithrombin deficiency causing thrombophilia. Thromb Diath Haemorrh 13:516, 1965. 17. Kloosterman G J: On intrauterine growth: The significance of prenatal care. Int J Gynecol Obstet 8:895, 1970. 18. Kloosterman G J, Huidekoper BL: The significance of the placenta in "Obstetrical mortality": A study of 2000 births. Gynaecologia 138:529, 1954. 19. Peters M, Breederveld C, Kahl6 LH, ten Cate JW: Rapid microanalysis of coagulation parameters by automated chromogenic substrated methods: Application in neonatal patients. Thromb Res 28:773, 1982. 20. Feissly R, Lfidin H: Microscopie par contrastes de phases. Ill. Application ~t l'h6matologie. Rev Hematol 4:481, 1949.
314
Peters et al.
21. Forfar JO, Arneil CC: Textbook of paediatrics, ed 2. London, 1978, Churchill Livingstone, p 177. 22. Henriksson R: Hyperviscosity of the blood and hemostasis in the newborn infant. Acta Paediatr Stand 68i701, 1979. 23. Broersma R J, Bullemer GD, Mammen EF: Acidosis-induced disseminated intravascular microthrombi and its dissolution by streptokinase. Thromb Diath Haemorrh 24:55, 1970.
The Journal of Pediatrics August 1984
24. Kisker CT, Robillard JE, Clarke WR: Blood coagulation changes following hypoxemia in the near-term fetal lamb. Pediatr Res 16:732, 1982. 25. Wiman B, Melbring G, R/mby M: Plasminogen activator release during venous stasis and exercise as determined by a new specific assay. Clin Chim Acta 127:279, 1983.