Causal pathways in cerebral palsy
- Declaration of conflict of interest: None declared.
Correspondence: Professor Nadia Badawi, Grace Centre for Newborn Care, The Children's Hospital at Westmead, Locked Bag 4001, Westmead, NSW 2145, Australia. Fax: +61 29845 2251; email: firstname.lastname@example.org
In 1977, the National Health and Medical Research Council (NHMRC), the peak funding body for medical research in Australia, allocated $4000 to a junior researcher from Perth, Western Australia (WA). The brief was to set up a population-based register of cerebral palsy (CP) in a state with a relatively stable and non-mobile population. CP, the most common disability of childhood, is a non-progressive, motor disorder of infancy and childhood, frequently but not invariably linked with other conditions such as epilepsy and intellectual disability.[1, 2] It affects one in 500 liveborn babies and, until recently, was stubbornly resistant to all efforts to reduce its incidence. Currently, nearly 40 000 Australians live with CP. In 1977, with preterm babies beginning to receive neonatal intensive care and neurological follow-up showing a high chance of long-term disability, it seemed possible that we were on the edge of a CP epidemic. Conventional wisdom, then largely unchallenged for 200 years, stated that CP was almost invariably due to ‘birth asphyxia’ (hypoxia and cerebral ischaemia during the birth process); in other words, a failure of obstetric care.[4-7] Around this assumption had grown an industry of punitive litigation, where families in great need of assistance sought to fund the expenses of care for their disabled child through an expensive and personally destructive legal system. The system was slow, unwieldy and confrontational. Moreover, it required a villain to be identified responsible for each victim. It was, in effect, a modification of a criminal code to deal with a situation where it was believed that the outcome was invariably a result of negligence.
Obstetricians largely bought into this story. Watching their professional indemnity premiums soar, they sought to improve their capacity to identify where they were going wrong. Believing that the answer to this seemingly insoluble problem must lie in their practice around labour, they focused almost exclusively on this window and sought to identify a monitoring system where an early warning would allow them to intervene and thus avoid injury. Electronic fetal monitoring was accepted wholesale into care, despite the fact that it was completely unproven.[8-10] Caesarean section rates soared[11, 12] as clinicians responded to abnormal fetal heart rate patterns in labour, the significance of which were, in fact, uncertain. Birth increasingly became identified with risk, and yet, in all of this, CP rates stayed frustratingly constant.[4, 7] The acceptance of this dogma held back research into the wider picture of CP for a generation, and although it is the most common disability of childhood, progress in this area remains well behind other less common conditions such as childhood cancer. To be fair, the pressure on the obstetricians was unreasonable. Any clinician who has looked after a woman through her pregnancy knows how close that bond becomes by the time of birth. The thought that a doctor might be responsible for a child in his/her care's lifelong disability was, and remains, too terrible to face. The knowledge that this was going to be highlighted and reinforced through years of court action only made this feeling of despair more acute. It was no wonder that they became less and less tolerant of any variation of labour and moved more and more quickly to Caesarean section, where no one could say that they had not done everything that was possible.
Entering into this state of dogma, dispute and defensive medicine are Karin Nelson and the Collaborative Perinatal Project. Somewhat unexpectedly, her research demonstrated that many cases of CP had uneventful births and newborn periods. Many strong associations were suggested during pregnancy and even before pregnancy. Conventional wisdom was, it seemed, potentially flawed. Interestingly, Sigmund Freud had previously asked the question why, with 280 days in pregnancy, all neurological injury should be assumed to take place in the last 12 h of pregnancy, often the last minutes. This radical idea had gained little traction, and in general, there was great resistance to the findings of Nelson and her colleagues. Fiona Stanley, who recently returned to WA and embarking on a research career, decided that if we were to make any headway in the field of CP, it was time to stop being informed by expert opinion and to start collecting some hard data. The princely sum of $4000 allocated by the NHMRC must remain one of the best investments ever made by any funding body in the world. Out of that seed fund grew the Western Australian Cerebral Palsy (WA CP) Register, which identified and collected extensive data on all cases of CP in WA. It could later be linked to the Maternal and Child Health Research Data Base and the Birth Defects Register (another innovation of Stanley's team). Cases were notified from multiple sources, and case ascertainment was remarkably robust. This allowed not only data collection on cases but also, unlike so many previous case series, comparison with contemporaneous controls.
Stanley and Blair started to produce data demonstrating that, far from most CP being due to birth injury, at most 10% of CP could be attributed to events in this epoch. Suddenly, because someone was looking, all sorts of new associations were being discovered. Thyroid disease, growth restriction, low socio-economic status, family history of epilepsy and neurological conditions, post-maturity and many others became promising areas for research. Risk factors and associations were identified in the preconceptional, early pregnancy and late pregnancy epochs. Like water brought to the desert, there was a flowering of interest in a field that had been thought barren and unresponsive. Once it became apparent that many cases of CP had their origins well away from birth, the range of possibilities for prevention opened up.
CP is a clinical diagnosis and is a complex condition with multiple subgroups. Classification may be by grade of severity, distribution (hemiplegia, diplegia or quadriplegia) and pattern (spasticity, ataxia, hypotonia or dyskinesia). From a causal-pathway point of view, it is pragmatic to consider CP in two larger subsets: those in term babies and those in preterm infants.[2, 4]
While only 1.4% of live births in Australia are 31 weeks gestation or below, these babies account for 30% of all CP cases. The WA CP Register demonstrated that, following an initial rise in CP in extremely preterm survivors, this rate is now falling steadily, with improvements in perinatal care. Antenatal steroids, the use of antenatal antibiotics, better monitoring in intensive care and closer attention to homeostasis, as well as the post-natal use of caffeine and delivery by Caesarean section in this preterm population, were all independent factors associated with a decrease in the risk of CP.[18, 19] Modifications of in vitro fertilisation protocols to replace only one rather than several embryos significantly decreased the risk of multiple pregnancy and preterm births in this assisted reproductive technology cohort, which now accounts for more than 3.2% of all Australian births.
In recent years, the promising finding that antenatal magnesium sulphate given to women in preterm labour before 30 weeks gestation may decrease the risk of CP by up to 30% has provided a new and potentially powerful preventative strategy. One challenge is getting research findings translated into routine clinical care without undue delay. The example of antenatal steroids is a salutary example where, due to reluctance to accept the findings of international groups, uptake in some countries was particularly slow, leading to higher-than-necessary rates of death and disability for many children. Antenatal steroids have been shown to protect against not only respiratory morbidity but also cerebral haemorrhage. Guidelines from the Royal Australian and New Zealand College of Obstetricians and Gynaecologists and the NHMRC relating to the use of magnesium sulphate should foster a smoother uptake.[23, 24]
Most term infants who have CP seem normal at birth and are delivered without any suggestion of a problem. The diagnosis slowly becomes apparent when parents notice their development is different from other children and seek medical input.
Only 25% of term infants who develop CP have neonatal encephalopathy. These infants, in the past, were uniformly described as having hypoxic-ischaemic encephalopathy. However, the WA Newborn Encephalopathy study showed that, even among term encephalopathic babies, only 29% had identifiable intrapartum risk factors, of whom 24% had both antenatal and intrapartum factors, and only 5% had solely intrapartum risk factors.[26, 27]
Risk factors were identified in the preconceptional period and all epochs of pregnancy, and included maternal thyroid disease, a history of infertility, socio-economic factors, a family history of neurological disease, intra-uterine growth restriction, postmaturity, pregnancy-induced hypertension and the presence of birth defects.[26-28]
The difficulty that faces researchers now is to move beyond the delineation of associations and to focus on the mechanisms and pathways that link these risk factors to CP. With the delineation of these pathways to damage, the number of points of possible intervention becomes greater. The illustration of kernicterus secondary to isoimmunisation gives a good example of this. Intervention at the point of exchange transfusion is expensive and risky. Understanding the causal pathway has allowed the introduction of antenatal and post-natal anti-D injections to rhesus-negative women, and this has virtually eradicated rhesus isoimmunisation as a cause of kernicterus in Western neonates. Small interventions early in the process may have a greater impact than large interventions late in the pathway. Other pathways need to be defined around growth restriction and placental insufficiency, thyroid disease and cerebral dysfunction, and social disadvantage and CP.
While we applaud the progress made with cooling in the infant born with evidence of birth asphyxia, the future probably rests with intervention at more ‘distal’ points along the pathway, preventing rather than ameliorating the injury.
To identify these pathways will take collaborative effort between international teams with wide-ranging skills. Basic scientists, epidemiologists, clinicians and research funders will need to pool their efforts. In order to gather more meaningful data on rarer risk factors and outcomes, it would be sensible to consider pooling the raw data from all studies internationally in a central clearing house similar to a CP register. Currently, for instance, because of small numbers in any one study, risk factors are often grouped inappropriately, which diminishes the ability of researchers to draw clinically meaningful conclusions. Maternal medical disorders in pregnancy are a good example where unrelated pathologies are often linked in a discordant common risk factor, for instance, epilepsy being grouped with cardiac disease. If all raw data from publications were subsequently sent to a clearing house, then very large data sets could be accumulated quickly, where relatively uncommon factors could be examined in sufficiently large numbers.
The year 2008 saw the birth of the Australian Cerebral Palsy Register, which is now the largest such register in the world. This national initiative owes its existence to the original WA register, which spawned other state-based registers. It is hosted by the Cerebral Palsy Alliance and is an invaluable resource to monitor the prevalence and trends in CP, investigate the aetiology and review the impact of various interventions in the perinatal period. It will also help governments to plan the health service needs of people living with CP. This register is now collaborating with other international registries. The potential this has for facilitating research and, thus, allowing results to be achieved much more quickly in larger populations cannot be overstated.
Another benefit of the refocusing of attention on CP research has been the establishment of the Cerebral Palsy Institute. The aim of this body is to bring together international researchers to work on the prevention and, ultimately, the cure of CP. In addition to hypothermia and magnesium sulphate, other strategies currently under investigation include erythropoeitin, melatonin and creatine, to name a few. Research into stem cells is at an early stage, and their role is as yet unproven. However, these therapies have theoretical benefits with tantalising potential for repair of damaged neurons.
Optimising resuscitation, especially in preterm infants, is an important area, as it seems likely that some damage occurs in this context, and other injury may be minimised by improvements in this area.
While research into the causes and prevention of CP remains vital, it is also important to work to improve the lives of people and families living with CP. To this end, the establishment of the National Disability Insurance Scheme is to be applauded. Under this scheme, resources would be allocated on the basis of need, which represents a much more equitable system.
For many years, CP languished behind other areas of paediatric research because of a widely held view that little could be done. Thanks to the efforts of key researchers such as Fiona Stanley, Eve Blair, Linda Watson and Karin Nelson, we are on the move again at last, and now, the vision of prevention and cure no longer seems an unattainable goal. There is much to be done before this common lifelong disabling condition can be relegated to history. However, perhaps in our children's lifetime, we will see CP recede from its current status to become a rare but treatable condition.