Cerebral palsy (CP) is the most common congenital neurologic disorder. The world wide prevalence of cerebral palsy is 2-2.5/1000 live births 1. In the past it was thought to occur as a result of acute intrapartum hypoxic events. Recent epidemiological studies suggest that in only 10% of cases this is true, and 90% of cases are due to chronic events or congenital factors 2-4.
CP is a chronic disability of central nervous system origin characterized by aberrant movement and posture, appearing early in life and associated with defect or lesion of the immature brain.
Neonatal encephalopathy and hypoxic-ischemic encephalopathy (HIE) have been defined in term and near term infants as a group of criteria to collectively include altered consciousness, tone, reflexes, feeding ability respiration, and/or seizures, and may or may not result in permanent neurologic injuries. Previous epidemiologic studies show that only 19% of neonatal encephalopathy met old nonstringent criteria for neonatal intrapartum hypoxia. CP due to HIE must be preceded by neonatal encephalopathy. Incidence of CP due to intrapartum asphyxia is only 1.6/ 10,000 and 70% of neonatal encephalopathy occurs as a result of prenatal events. Additionally intense intrapartum monitoring in recent past has not reduced the incidence of CP. However, in the past, birth asphyxia due to poor obstetric management has been needlessly and without scientific basis implicated medicolegaly.
Spastic quadriplegia, is the only type of CP associated with intrapartum acute hypoxic events. Dyskinetic or ataxic CP associated with learning problems, epilepsy, hyperactivity disorders, attention deficit or mental retardation do not have their origin in acute birth hypoxia.
Our current knowledge of cerebral palsy is highly enhanced by both internationally published scientific evidence and a joint project by the American College of Obstetricans and Gynecologists and the American Academy of Pediatrics (ACOG/AAP). A multidisciplinary International Cerebral Palsy Task force of scientists and clinicians published a consensus statement in 1999, which set out the criteria defining an acute intrapartum hypoxic event necessary to have had occurred prior to the onset of CP 5. ACOG/AAP revised the international template and updated the literature on the pathogenesis and pathophysiology of neonatal encephalopathy and cerebral palsy in 2003 6 (Table 1). Of the nine criteria four when present together help to prove the existence of severe hypoxia at birth. The remaining five when present collectively suggest intrapartal timing but are not specific to asphyxial insults. These efforts help to truly identify intrapartum hypoxic events and other pathologies responsible for CP. Recently, cord blood gases, placental pathologies and neonatal brain imaging allow better identification of antepartum etiologies. Additionally, definition of acute hypoxia and its correlation with neonatal nucleated red blood cells increase in chronic hypoxic events can be used to differentiate between acute and chronic pathologies 7-8.
Neonatal encephalopathy and HIE are related to CP and long term neurologic deficits specifically in term and near term infants in whom neurologic signs can be accurately recognized, classified, and scientifically studied.
Very preterm and very low birth weight infants contribute disporportionately to occurrence of CP eg, 25% of the total number of CP cases come from those with a birth weight of < 1500g. These babies have stormy neontal course, respiratory and other vital problems. However they are hard to study since they do not exhibit typical neonatal encephalopathy pathophysiology and may have a variety of other neurologic deficits besides CP.
Maternal and fetal conditions as risk factors are depicted in Table 2.
Theoretically, vaginal bleeding and shock due to placenta previa and placental abruption can reduce placental perfusion and oxygenation of the fetus leading to neonatal encephalopathy resulting in CP. In reality, third trimester bleeding is rarely associated with HIE 9.
A prospective US study of 42,704 infants weighing > 2500 g from pregnancies complicated by bleeding due to placenta previa had a slightly higher risk of CP (OR of 6; 95% CI, 1,9-18.8)10. This risk was not elevated in those whose mothers had bleeding due to abruptio placenta although the risk of death in first year of life was greater for these infants.
A case control study of 46 normally weighing infants with spastic CP from California Birth Defects Monitoring Program revealed that neither placenta previa nor abruptio placenta were associated with increased risk of CP (OR 14.3 and OR 2.8 with 95% CI) 11.
In two case control studies from Western Austrialia involving babies > 2500 g, vaginal bleeding was associated with slightly increased risk of neonatal encephalopathy (OR 5.95%; CI 1.5-17.3 and OR 3.57; 95% CI 1.30 - 9.85) 12. The second time the results were higher but not associated with hypertension.
Hence CP does not seem to be increased with vaginal bleeding due to placenta previa or abruption. Confounding variables of growth restriction, hypertension, coagulopathy, substance abuse etc, may be operative in genesis of HIE and its consequences.
Studies of CP in preterm infants are rare. However, in a Danish review, neither placenta previa nor placental abruption was found to be responsible for HIE or CP 13.
Underlying long standing and chronic conditions in third trimester vaginal bleeding may lead to fetal neurological injury antedating antepartum bleeding. Symptom of third trimester bleeding is rarely associated with HIE 14,15.
Intrauterine infection is associated with CP in term and near term infants. It is now realized as the most common precursor of adverse neonatal outcomes viz., low apgars culminating in encephalopathy and CP.
Intrauterine infection may lead to hypotension, neonatal seizures, respiratory distress syndrome (RDS), preterm delivery and HIE. Intraventricular hemorrhage, periventricular leukomalasia, bronchopulmonary dysplasia and CP may result from these serious problems 9,16,17.
These and other responses imitate adult systemic inflammatory responses mediated by endothelial injury in adults. Over production of cytokines leads to brain damage in small infants. Chorioamnionitis increases periventricular leukomalasia in preterm infants. An increased risk of CP has been documented in preterm infants exposed to chorioamnionites (RR 1.9) 18. Added exposure to asphyxia due to obstetric complications certainly makes this risk even greater.
Recent evidence suggests that intrauterine infection preceding pregnancy or arising early in pregnancy may persist from one pregnancy to next, and have significant impact on outcome 19.
Maternal and fetal response to intrauterine infection may be mediated by invidiual’s genetically controlled inflammatory process 20. Infectious and inflammatory processes may progress to all manifestations of intrauterine infections. Additionally altered immune recognition, cell damage and cell death from reactive oxygen species and damaging molecules may result from these responses. These mechanisms may persist causing cell damage even after the original processes are corrected 21,22.
Intrauterine infection appears to be the most common cause of preterm birth in USA and possibly world over. Ascending infection from bacterial vaginosis from endogenous microflora, anerobes, mycoplasmas, gardenerella etc, or less often infection by way of maternal blood stream of microbes is often responsible for perinatal infection.
Elevated cytokine concentrations in amniotic fluid and cord blood have been associated with white matter damage and CP 21,23. Severe neonatal morbidity due to neonatal sepsis including multiorgan involvement eg. pneumonia, respiratory distress syndrome (RDS), intraventricular hemorrhage and periventricular leukomalasia have been associated with increased levels of Il-6 in cord blood (78% vs 30%) 23
Ultrasound evidence of perivenricular leukomalasia and clinical markers of infection eg. chorioamnionitis, C-reactive proteins and recovery of causative microbes are closely associated 24. Increased neonatal cytokines and evidence of neonatal thrombophilla are closely associated with CP 25,26.
Clinical and subclinical intrauterine infections are significant causes of reproductive loss and preterm birth. Infection and associated inflammatory cytokines are associated with increased risk of CP in all infants. Thrombophilia is associated with inflammatory activation in children manifesting CP and vice versa.
Interacting associations of intraamniotic infection, white matter lesions, echoluscency, and periventricular leukomalasia are under study. However, multiple studies demonstrate increased CP in chorioamnionitis in term and preterm infants 27-32
Bacterial perinatal infections including intrauterine infections are known to imitate adult systemic inflammatory responses involving cytokines, chemokines, coagulation factors, and other cellular responses causing direct cell necrosis and inducing programed cell death.
A host of organisms have been observed in these infections, including group B streptococci, E. coli, klebsiella and proteus, hemophilus, listeria monocytogenes, salmonella etc.. Anerobes and mycoplasma may be recovered from amniotic fluid and often not cultured from the neonate 22-30.
Proteolytic enzymes released during killing of the microorganisms can perpetuate cell damage long after clinical infection has been controlled. Viral and protozoal perinatal infections are less common, involving < 5% of cases 16. Organisms, causing brain damage due to viral/protozoal infections include, cytomegalovirus, herpes I and II, varicella- zoster, rubella and some enteroviruses. Toxoplasma and congenital malaria infections may be involved in causing brain lesions in the fetus occasionally.
Maternal and fetal coagulopathies and placental thrombosis are important contributors to fetal morbidity and mortality including stillbirth, early neonatal demise and organ thrombosis 31. In addition to fetal coagulopathies, maternal coagulation abnormalities can be transmitted across placenta to the fetus.
Variety of brain injuries may occur prenatally and three major categories were seen during a review of nearly 100 autopsies 32
Evidence of coagulopathies like factor V Leiden mutation, and protein C or protein S abnormalities without cytokines indicating absence of infection were seen in blood samples of 31 children with CP of unknown origin. All blood samples had cytokines indicating relationship between perinatal inflammation and coagulopathy 26.
Spastic CP in near term infants was associated with inflammation and coagulopathies. Both maternal and fetal thrombophilias eg. maternal antibodies to coagulation factors, cardiolipin in systemic lupus erythematosus and phospholipids, may injure early embryo and fetus causing fetal demise 33.
Fetal genome influencing outcome is well established in case of factor V Leiden mutation directly predisposing the fetus to thrombotic conditions and CP 34. Other adverse outcomes in similar conditions may include intrauterine growth restriction, low birth weight, fetal distress, neonatal seizures, and a clinical picture resembling birth asphyxia. Cerebral infarction or intraventricular hemorrhage may be seen on imaging studies.
Coagulation disorder in mother (eg. antithrombin - III deficiency, abnormalities of protein C or S, and factor V Leiden mutation) may contribute to origin of CP. Only severe Maternal infections are associated with CP.
Earlier data from the National Collaborative Perinatal Project Identified maternal mental retardation and seizure disorders as risk factors for CP which was not confirmed later 3,35. Family history of neurologic disorders including seizure disorder is associated with a 2.5 fold increase in HIE in the infant. Antiepileptics are not a risk factor for CP. The link between maternal epilepsy or mental retardation and CP may only be an association.
Maternal hypothyroidism is a risk factor for abnormal neurodevelopment in the child however it being causative in origin of CP needs further research and evidence.
Maternal severe hypothyroidism adversely affects fetal neurodevelopment and may be significantly linked to neonatal encephalopathy 36. Transient hypothyroidism may be associated with abnormal neurodevelopment in preterm babies 37-39.
Maternal consumption of alcohol may result in neurologic, cognitive and behavioral abnormalities. However no relationship has been observed between maternal use of alcohol and CP 40,41. Maternal use of cocaine and exposure to organic mercury may affect neurologic development and the latter has been linked to CP 42. Environmental influence on etiology of CP has not been established to date. Association between CP and lead levels and maternal hyperthernia have not been fully investigated yet.
In a multivariate analysis, the strongest risk factor for neonatal encephalopathy was fetal growth restriction 9. This risk increased with advancing gestational age beyond 38 weeks.
Multiple causal pathways lead to neonatal encephalopathy and CP. However, 70% of cases of neonatal encephalopathy were due to events present prior to onset of labor, and were not amenable to intervention by the provider.
The literature to demonstrate long term relationship between antepartum testing and neurologic outcome is scarce. Randomized prospective trials of antepartum testing have shown no benefit 43,44.
A single randomized trial of fetal movement counting has shown decrease in fetal death but no impact on neurologic outcome 45. In an Australian study, a nonreactive positive contraction stress test was associated with 28% perinatal mortality and 27% neurologic deficit rate 46. Manning et al 47 have shown incidence of CP to be 1.33/1000 in those infants who underwent biophysical profile vs 4.74/10000 among those who did not.
Overall predictive value of antepartum testing for prediction of long term neurologic abnormality is poor, though abnormal results are associated with more abnormal outcome. Antepartum electronic fetal heart rate monitoring has not shown to reduce CP. Antepartum fetal movement monitoring has shown to reduce fetal deaths.
Since 1970’s, National Collaborative Perinatal Project data have concluded that routine auscultatory monitoring of FHR is not helpful.
Advent of electronic fetal monitoring is associated with fewer intrapartum and neonatal deaths, fewer low apgar scores, and less need for resuscitation. Randomized trials of electronic fetal heart rate monitoring vs intensive nursing and auscultations failed to show advantages of electronic monitoring seen in previous trials. In fact increased cesarean deliveries occurred 48-53. The only benefit of electronic monitoring shown was decrease in neonatal seizures 54,55. No benefit was observed in abnormal neurologic outcome. Opportunity for appropriate action to prevent bad outcome was still not available.
Use of scalp blood sampling for acid-base status with abnormal patterns decreased cesarean section rate 49. Fetal pulse oxymetry for patients with nonreassuring tracings decreased cesarean delivery for abnormal tracings but increased cesarean deliveries for dystocia 56. Further studies were recommended.
In an earlier study, intrapartum fetal monitoring analysis in 1,35,000 patients revealed fetal death rate to be 0.54/1000 in monitored group vs 1.76/1000 in nonmonitored group, with a death ratio of 3.26 favoring monitoring 57. As previously stated, both intensive auscultation and electronic monitoring prevent fetal demise.
In a study by Nelson et al 58, fetal heart rate patterns associated with CP were repetitive late decelerations and decerased beat to beat variability. A high false positive rate of prediction was observed.
The National Institute of Child Health and Human Development consensus report has stated definition of reassuring fetal heart rate pattern to be extremely predictive of well oxygenated fetus. Reassuring pattern includes, normal baseline, moderate heart rate variability, presence of accelerations and absence of decelerations. It reiterated patterns indicative of impeding asphyxia and risk of neurologic injury associated patterns with recurrent late or severe variable decelerations or significant bradycardia and absent variability. A sentinel event precipitating these patterns is necessary when previously normal patterns become ominous. A fetus whose mother is admitted with abnormal pattern at the onset of labor may already have neurologic injury which cannot be improved by any means.
Specific fetal heart patterns indicate uteroplacental insufficiency and umbilical cord compression. Profound fetal hypoxia resulting in prolonged deceleration and leading to encephalopathy can occur due to any cause.
Intrapartum fetal monitoring has shown to reduce neonatal seizures. However, it is not associated with reduction in CP when compared with dedicated auscultation. It has high false positive rate of prediction for CP.
Animal data have contributed profoundly to our knowledge of mechanisms of neurologic injury in the fetus. The pattern of neurologic insult following acute catastrophic hypoxic-ischemic insult usually involves the thalami and basal ganglia and is different from the pattern following chronic insult seen in the neonate involving cerebral cortex and the subcortical white matter predominantly 59,60. The former is highly predictive of poor outcome and mimics animal studies of acute hypoxia. Additionally it may not be asscoiated with multiorgan system injury 60,61. In an anlysis of 14 cases of isolated brain injury following acute asphyxia, the average duration of fetal heart decelertion was 32.1 minutes (range 19-51 minutes) 61. A high metabolic rate of subcortical region may account for the damage to this region without involvement of other systems. Conversely, more chronic hypoxia associated with uteroplacental ischemia allows redistribution of blood supply and oxygen from the cerebral cortex and other organ systems to thalami and brainstem resulting in injury to the cortical region and multiorgan dysfunction.
Neurologic outcome and time needed for delivery in cord prolapse depend upon the presentation and station, and severity of uterine contractions (low vertex is associated with more severe outcome). Recent advents in neonatal management may have improved neurologic outcome due to better resuscitation. Fetal mortality and neurologic injury in cord prolapse are affected by the time from diagnosis to delivery, and the extent of the cord prolapse 62-65.
The pathophysiology in shoulder dystocia is similar to that in cord prolapse but more severe. Head to body delivery interval is the determining factor. In a review of 56 cases of shoulder dystocia, 47% had < 4 minutes and 20% had > 10 minutes head to body delivery interval. Twenty one babies had HIE. Autopsy on 18 of the 23 nonsurviving babies revealed acute hypoxic damage 66.
Perinatal mortality and morbidity associated with uterine rupture is dependent upon whether the fetus was extruded from the uterus or not. In a review of 106 cases, extreme neonatal morbidity including perinatal asphyxia was much worse when fetus was partially or completely extruded out of the uterus 67.
When delivery occurred within 17 minutes of onset of prolonged deceleration, no significant perinatal morbidity was seen. However, when prolonged fetal heart rate deceleration was preceded by 36-90 minues of severe late decelerations, perinatal asphyxia occurred within a short time, as early as within 10 minutes of prolonged deceleration. Morbidity was minimal when delivery occurred within 13 minutes from onset of prolonged deceleration 68.
The extent of neonatal asphyxia following maternal cardiopulmonary arrest depends upon numerous factors independent of the time factor including the etiology of the collapse, maternal state and physiologic stability / instability prior to collapse, and gestational age of the fetus. In a review from the national registry of maternal collapse with a live fetus, neurologically intact survival was inversely related to the time from cardiac arrest to delivery 68.
A reivew of published data of neonatal outcome following maternal death between 1900 to 1985, found that the most fetal survivors were delivered within 5 minutes and none survived 10 minutes after maternal demise 69.
AAP / ACOG have emphasized a 30 minute time interval from decision to delivery (cesarean section) 69, 70. Most medical centers in USA provide necessary facilities including presence of qualified obstetricians, pediatricians, operating rooms, anesthesia, and nursing support. However, in the event of catastrophic emergency, extent of fetal hypoxic ischemic state and neurologic damage is virtually unpredictable and often not related to event-delivery time.
Multiple pregnancy and CP
Multiple pregnancy contributes disproportinately to the incidence of CP. In fact 2.4% of multiples contribute to 25% of children with CP. Prevalence of CP in multiples is epidemoiologically quoted to be 7/1000 births, and the probability of at least one child with CP from a multiple pregnancy is 15/1000 for twins, 80/1000 for triplets, and 429/1000 for quadruplets. Association of preterm birth, low birth weight, discordant growth, birth order, monozygocity, monochorionicity, and intrauterine demise of one fetus all predispose these babies to CP.
Preterm births
The overall risk of CP in twins weighing > 2500 g is 3.38 times more than singletons of similar birth weight. In a recent study, relative risk (RR) for CP in a comparison of similar birth weight was 4.5 (95% CI), while RR in comparison of gestational age of > 37 weeks was 6.3 (95% CI). Multiplicity of fetus itself appears to be a risk factor for CP 71-73.
Monochorionicity and demise of one
Monozygotic and monochorionic twins are more likely to result in neurologic abnormalities than others due to vascular anastomoses, fetal growth problems, and cord abnormalities. This risk is higher for the survivor when demise of one occurs in a monochorionic gestation 74,75.
A recent study investigating antenatal origin of neurologic injury by ultrasound assessment of neonatal brain revealed cerebral atrophy and white matter necrosis to occur more often in monochorionic twins than in dichorionic twins (30% vs 3.3%, P<0.001) 76. Polyhydramnios, demise of one, and vascular abnormalities within the placenta all played a role in increasing this risk.
The most outstanding pathologic lesions in the surviving twin in a monochorionic pregnancy complicated by demise of one include renal cortical necrosis and multicystic encephalomalacia. These processes occur due to intraplacental vascular connections. Major placental vascular connections allow twin to twin transfusion (TTS) to occur leading to the syndrome of TTS due to uncompensated vascular anastomoses and imbalanced distribution of blood between the twins. Often extreme degree of transfusion may predispose one or the other twin to severe decompensation ultimately resulting in CP.
Multiple pregnancy is independent risk factor for neurologic deficit and CP. Monochorionicity and demise of one twin predispose twins to even greater risk of neurologic impairment.
Intrauterine growth retardation (IUGR) vs small for gestational age infant (SGA)
These terms refer to a baby with a birth weight less than an average weight for comparable gestational age. IUGR refers to a pathophysiologic state resulting in small size while SGA is simply a small birth weight baby. Birth weight of the baby in question is compared to a standard nomogram of birth weight in the community.
In US, standard nomograms for females and males are available 77,78. A baby falling below the 10th percentile is considered SGA. However, adverse perinatal outcome and consequently risk of encephalopathy and even mortality rises for newborns weighing below the 5th and specially below the 3rd percentile for age 9,79. Relative risk of encephalopathy for birth weights below the 3rd percentile when compared to those between 3rd to 9th percentile is significantly greater
(OR 38.23vs 4.37) and so also when compared to those above the 9th percentile (OR 1.54). Arrest of fetal growth or loss of weight are distinct but imprecise measures of fetal risk.
Numerous other variables may determine the ultimate outcome and risk of encephalopathy eg. perinatal infections, genetic and structural abnormalities, substance abuse, coagulopathies etc.. An IUGR infant born preterm has obviously much greater risk.
National Collaborative Perinatal Project found infants without any hypoxia related risk factors not to be at increased risk for neurologic morbidity or CP. Conversely, IUGR newborns with hypoxia related risk factors were more likely to have neurologic deficits 80 . Oligohydramnios and IUGR combination suggesting chronic fetal hypoxia may also result in fetal cerebral ischemia and white matter lesions 81,82. Adverse developmental outcome with IUGR/SGA in otherwise healthy babies results from nonnutritional causes. Etiologies of IUGR are multiple and determine the ultimate outcome. No interventions other than delivery have any impact on outcome of IUGR.
Postterm birth / Postmaturity syndrome
Perinatal mortality increases after 41 weeks of gestation. A significant increase in RR for encephalopathy beyond 39th week of gestation was found in a large population based study (RR 1.41 at 40, 3.34 at 41, and 13.2 at 42 weeks of gestation) 9. Currently in US, postdate fetuses routinely undergo biweekly heart rate surveillance by NST and are delivered by 42 weeks.
Meconium passage
Passage of meconium was implicated as a marker for new born hypoxia/asphyxia. It has been found to be a poor predictor of HIE. A review of 42,000 term singletons did not find any increased incidence of adverse neonatal findings in newborn with meconium stained fluid 83,84. Vasoconstriction of cord vessels leading to hypoxia due to meconium has been theorized. It has not been confirmed scientifically and 99.6% of term fetuses with meconium stained fluid do not have CP. In fact, Meconium passage is a physiologic event and by itself is not a marker for fetal hypoxia. It may form a nidus for bacterial growth in the amniotic cavity.
Genetic, anatomic, and metabolic etiologies of neonatal encephalopathy
Neonatal altered neurologic manifestations can occur as a result of genetic, metabolic, and anatomic etiology, and may contribute to neonatal encephalopathy 85,86. Many of these disorders result in hypoglycemia, hyperammonemia or other metabolites leading to irreversible metabolic, anatomic or functional effects and adversely affect the brain structure and/or function. Table 3 gives the congenital disorders associated with neonatal encephalopathy / CP.
Neonatal hypoglycemia
Neonatal hypoglycemia may occur as a result of hyperinsulinism, endocrine deficiencies, or metabolic defects and produce neuronal as well as glial injury. Table 4 lists the various conditions causing neonatal hypoglycemia.
