Feature

Premature Birth and Infection

Research into the role cytokines play in infection-driven premature delivery may help prevent premature births.

J. Keelan and M.D. Mitchell

Normal human pregnancy usually ends about nine months (40 weeks) after conception with the delivery of a healthy baby. Unfortunately, about 10% of pregnancies fail to progress to term, and the baby is delivered prematurely. Modern neonatal intensive care units presently achieve fairly good survival rates for babies born after 26 weeks gestation (three months premature) but, sadly, survival is frequently accompanied by serious short-term and long-term health problems. In fact, 75% of all neonatal deaths and illnesses are due to premature delivery.

From a health funding perspective, the problem of prematurity is a major financial burden. The cost to the hospital of providing care for a very premature infant can be in excess of $100,000, while long-term health-care costs, such as caring for children with physical or mental handicaps, can be enormous as they may be required for the lifetime of the affected individual.

Current treatment options for women presenting with premature labour are of limited effectiveness, and there have been few improvements over the past 20 years or so. A class of drugs called b-agonists are widely used in hospitals around the world to delay labour for 24-48 hours, but large international studies have failed to show that the use of these drugs results in significant reduction in neonatal deaths or illnesses due to premature delivery. Moreover, they have been linked to a number of serious maternal side effects, as have other different drugs which have been tried in the past.

Part of the reason for our failure to successfully treat premature delivery is that its causes have been poorly understood. However, in the last ten years it has become apparent that a significant proportion of women with preterm labour, perhaps up to 70%, have infections of the placenta or membranes that surround the fetus in the womb. It appears that in many such cases, the infection is not clinically obvious -- the mother does not have a fever or inflammation or tenderness in her womb or vagina. However, biochemically it has been shown that inflammatory reactions are established in the tissues that surround the fetus, and the chemical products of these reactions (cytokines) are thought to be the agents that cause the onset of premature labour.

The Anatomy Involved

To comprehend the process through which bacterial invasion can cause premature labour, one must first have a working knowledge of the tissues and organs that protect and nourish the fetus during pregnancy.

Throughout pregnancy the baby is immersed in a watery bath of amniotic fluid surrounded by a tough semi-transparent membranous sack. This fluid is composed principally of fetal urine, and is repeatedly swallowed and "breathed" by the baby as it grows and develops (distasteful as this may seem). The membranous sack is actually composed of three distinct layers: the inner membrane is called the amnion, the middle layer is called the chorion, and the outer layer is the decidua. This outer tissue is actually of maternal origin, originating from the cells that line the uterus before pregnancy, while the other two are derived from the fetus itself. Only the decidua is supplied with blood; the amnion and chorion are devoid of blood vessels and derive their oxygen and nutrients from the amniotic fluid or from the decidua by diffusion.

The placenta, which also shares its origin with fetal tissue, is the organ responsible for allowing the exchange of nutrients and oxygen from the maternal to fetal blood supply, while maintaining a physical barrier to stop the two circulations mixing. The baby's waste products in turn cross into the maternal blood and are cleared through the mother's urine or faeces.

The placenta usually lies along the wall of the uterus where it cannot obstruct the cervix. This is a tough ring of tissue which is normally tightly closed during pregnancy, and which then softens and dilates during labour to allow delivery of the baby. The placenta also serves as a barrier to prevent infections from crossing from the mother to the baby. Some viral infections appear to be able to penetrate this barrier and can harm the fetus, but such infections do not commonly cause premature labour.

The vagina contains a range of bacteria, the presence of which is normal and is usually harmless. They are prevented from ascending into the amniotic cavity by the cervix, which is not only constricted during pregnancy but is plugged with a thick mucus which is an effective barrier against microbial invasion. However, under some circumstances (most of which are unclear), bacteria appear to gain access to the membranes or amniotic fluid. The normal response of tissues to the presence of bacteria is an inflammatory reaction, and this seems to occur during pregnancy following infection of the amniotic fluid, membranes or placenta.

Response to Infection

All animals have a repertoire of defences -- the immune system -- which can be called upon to fight off an infection. One arm of the body's immunological armory involves the production of special targeting proteins called antibodies. Antibodies are produced by specific cells in the blood following activation by fragments of an invading microorganism or cell. They bind to a unique recognition sequence on the foreign cell, targeting it for destruction.

The other arm involves the release of chemical messengers which alert the body to the presence of an invader and cause the recruitment of special cells which are able to engulf and kill the foreigner with toxic chemicals. These cells, which include macrophages and neutrophils, are present in blood and also to a lesser extent in many tissues of the body, including the placenta and decidua. Since they can be recruited to the site of an infection, macrophages and neutrophils are often found in large numbers in infected tissues.

Infection is usually associated with inflammation, which involves vasodilation or widening of the blood vessels. This results in increased blood flow to the site of infection, swelling (fluid build-up) and tissue destruction. Vasodilation is caused by the local release of a variety of rapid-acting molecules, including lipid-like chemicals called prostaglandins. These inflammatory reactions are coordinated, in part, by a family of signalling proteins called cytokines.

Although we are still trying to understand the very complex actions and interactions of the different cytokines, a clear picture of the role of these messengers in inflammation and cellular immunity is beginning to emerge.

Cytokines are classified according to their actions. The pro-inflammatory cytokines, as their name suggests, act to propagate the inflammatory reaction. Examples of these include interleukin-1 (IL-1), IL-2 and tumour necrosis factor-a (TNFa). Other cytokines (called chemokines) are involved in the recruitment of immune cells, an example of which is IL-8.

There are also immunomodulatory cytokines, such as IL-6, which coordinate systemic responses to an infection, and inhibitory cytokines (transforming growth factor-b(TGFb), IL-4 and IL-10) which dampen the immune response by inhibiting the cytokine "cascade".

Many of the cytokines known to be produced by immune cells have also been detected in the fetal membranes, decidua, or placenta. It appears that the cells which normally compose these tissues are capable of releasing cytokines such as IL-1, IL-4, IL-6, IL-8, IL-10, TGFb and TNFa in response to an infection, and that the production of such cytokines is not dependent on the presence of infiltrating immune cells into the tissue. However, it has also been found that if the membranes are infected, they become considerably enriched with macrophages and neutrophils, and their capacity to secrete cytokines such as IL-1 and TNFa greatly increases.

Tissues obtained from women who have gone into labour at term and delivered normally produce more IL-1, IL-6, IL-8 and TNFa than tissues obtained by caesarean section (ie non-laboured). This has led to the suggestion that cytokines are produced in the uterus in normal, uninfected pregnancies and that they may play a role in the normal birth process. Alternatively, the stresses of delivery, and the opportunity for exposure of the membranes to vaginal bacteria during labour, may be the explanation for these findings.

Infection-driven Preterm Labour

While the role of the pro-inflammatory cytokines in normal pregnancy remains a matter of debate, their role in the mechanism of infection-driven preterm labour is fairly well established. Levels of IL-1, IL-6, IL-8 and TNFa in amniotic fluid are all greatly elevated in infected compared to uninfected pregnancies, either at term or preterm.

In animal studies it has been convincingly shown that inoculation of the amniotic cavity with bacteria or bacterial proteins is followed within a few hours by measurable increases in amniotic fluid cytokine concentrations. After six to twelve hours, strong uterine contractions begin and, several hours later, delivery ensues. Uterine contractions can be induced by inoculation with IL-1 or TNFa, while in mice, bacteria-induced labour can be blocked by the co-administration of an "anti-inflammatory cytokine" such as IL-10.

The pro-inflammatory cytokines have powerful effects on the tissues surrounding the baby. They are capable of stimulating the production of prostaglandins from cells in the amnion, chorion, decidua and uterus; greatly elevated levels of prostaglandins are present in the amniotic fluid of women with infected pregnancies.

It is through this route that cytokines are thought to play such an important role in infection-driven preterm labour, since prostaglandins cause the uterine muscles to contract and commence the labour process. Once started, the membranes and cervix become stretched and inflamed, and this sets up a secondary inflammatory-like response which amplifies the process already in progress. Cytokines and prostaglandins also cause the ripening of the cervix. In fact, obstetricians sometimes administer a prostaglandin gel during labour for this purpose.

As mentioned above, inflammation is often associated with tissue destruction. Cytokines mediate this process through inducing the release of enzymes such as collagenase or elastase which dissolve the connective matrix which holds cells together. TNFa may also kill cells directly though the induction of a process known a apoptosis (programmed cell death).

If the membranes become too thin and weak due to these processes they may rupture prematurely, allowing the protective amniotic fluid to leak out and bacteria to get in. Premature rupture of the membranes is a common occurrence with infected pregnancies, and is a serious complication if it occurs more than a few weeks before term.

Unusual Immunological Responses

While many of the actions of the pro-inflammatory cytokines on the fetal membranes are typical of the effects that these cytokines have on other cells, the actions of some of the "inhibitory" cytokines are more unusual. For example, IL-10, which inhibits prostaglandin and pro-inflammatory cytokine production by macrophages and decidual cells, has little effect on prostaglandin production by the amnion, while it actually stimulates secretion of IL-6 and IL-8 by this tissue. IL-4 also has some stimulatory effects on amnion cells, in contrast to its generally inhibitory actions on production of inflammatory mediators.

Perhaps the most unexpected is the action of a protein called IL-1 receptor antagonist (IL-1ra) on the decidua. This protein is similar in structure to IL-1, but blocks the actions of IL-1 by preventing it from binding to its receptor. However, in the decidua (unlike any other tissue), IL-1ra actually behaves like IL-1, in that it has a pro-inflammatory action.

What is the significance of these findings? We have postulated that in the presence of a virulent bacterial infection of the fetal membranes or amniotic fluid, the fetus is faced with a "fight or flight" crisis. Should the fetus remain in the uterus while the mother's immunological defences attempt to fight off the infection, or should the fetus be delivered and escape the hostile environment of an infected womb?

We believe that evolution has answered that question by significantly attenuating the cytokine regulatory mechanisms that would normally switch off the cytokine-generated inflammatory reaction. Hence, if infection sets in and initiates a significant cytokine response, the production of cytokines and prostaglandins by the fetal membranes accelerates unchecked and leads to levels of these substances in the amniotic fluid and uterine tissues that are high enough to initiate labour resulting in delivery of the baby.

Such an option also has benefits from the maternal point of view, since the infected tissue is expelled from the womb along with the baby, reducing the mother's risk of death from infection and thus allowing her the opportunity of further pregnancies.

Potential Therapy

Preliminary experiments with the anti-inflammatory cytokines IL-10 and transforming growth factor-b (TGFb) have shown that, at least in animal models, the administration of anti-inflammatory cytokines can prevent preterm delivery in response to bacterial infection. These encouraging results must, however, be viewed with considerable caution.

Firstly, while inoculation of bacteria or bacterial proteins into the peritoneal cavity of mice is useful in studying some of the molecular events that take place in infection-driven preterm delivery, this somewhat artificial infection mechanism does not accurately mimic the mode of infection that can occur in women. Secondly, the atypical responses of the fetal membranes to inhibitory cytokines throws considerable doubt on the likely effectiveness of this approach in humans, and also highlights the likely importance of the route of administration.

From a clinical perspective, the danger that delaying delivery of a baby in an infectious environment by administering anti-inflammatory agents might cause more harm than good is very real. Under such circumstances, perhaps the optimal course of action is to administer antibiotics and delay labour for a short time while giving steroids to mature the fetal lungs to minimise the problems that premature babies encounter. Administration of broad-spectrum antibiotics are effective in reaching the infected tissues, either by diffusion from the decidua, or from the fetal side via secretion into the amniotic fluid. For this course of action, anticytokines may have a role to play since current approaches are not particularly effective in these circumstances and have significant side-effects.

Perhaps more encouraging is the prospect of measuring cytokines or cytokine-induced proteins in amniotic or cervico-vaginal fluids to assess the risk of premature delivery. Presently, trials are underway in New Zealand to assess the usefulness of measuring cytokines and fetal fibronectin (a connective tissue protein released from the chorion under the influence of tissue-dissolving enzymes and cytokines) to predict preterm labour. Previous studies overseas have shown that the presence of fetal fibronectin in cervico-vaginal fluid is indicative of delivery within two weeks, while IL-6 is also highly predictive of an intrauterine infection.

It is anticipated that the combination of these and perhaps other measurements will prove to be an effective diagnostic tool in the management of women who are at high risk of delivering prematurely, or who present with preterm labour of unknown cause.

The importance of cytokines in the mechanism of infection-driven premature delivery is now widely accepted by the medical community. We hope that continued research into the biochemical processes that occur in the uterus and placental tissues in infected pregnancies will allow us to devise useful and effective diagnostic tools and therapeutic strategies that will result in significant improvements in the management of women with preterm labour. The unique immunological characteristics of the tissues that surround the fetus make the development of anti-cytokine therapies a challenging goal for future research.

Jeff Keelan is with the Department of Pharmacology and Clinical Pharmacology at the Auckland University School of Medicine.
Professor Murray Mitchell is head of the Department of Pharmacology and Clinical Pharmacology in Auckland University's Faculty of Medicine and Health Science.