Muscle contraction is triggered by a rise in the intracellular concentration of calcium ions. Electrical activity, calcium ion influx and development of myometrial tension are precisely synchronized (López Bernal, 2003). Calcium ions bind to the calcium-binding protein calmodulin (CAM) which regulates the activity of many of the intracellular enzymes generating a cascade of reactions leading to binding of actin and myosin (Wray et al., 2003). Binding of calcium to CAM forms a Ca–CAM complex which activates MLCK. MLCK is inhibited when calcium levels are low. The activated MLCK undergoes a conformational change which exposes the catalytic site so the MLCK can be phosphorylated and will interact with actin, initiating contractions. Removal of calcium results in dephosphorylation of myosin by myosin light-chain phosphatase and causes muscle relaxation. Smooth muscle contraction can therefore be increased either by activating MLCK or by inhibiting myosin phosphatase.

Calcium enters the myometrial cell from the extracellular fluid or is released from intracellular binding sites and organelles including the sarcoplasmic reticulum of the myometrial cells. Calcium released from the sarcoplasmic reticulum can prime contractions that need to be sustained by calcium influx. Voltage-operated calcium channels regulate contractility by allowing calcium entry. Calcium-activated potassium channels set the threshold of activation of the cell membrane. There is a good correlation between intracellular calcium concentration and the muscular force developed with ionized calcium (Ca²⁺) concentration increasing from approximately 100–500 nm during contraction. Uterotonins (substances that stimulate myometrial contractility), such as prostaglandins and oxytocin, increase calcium influx and mobilize intracellular calcium stores, therefore increasing intracellular calcium concentrations and MLCK phosphorylation and increasing myometrial activity. Agents that inhibit myometrial activity, such as progesterone, β-mimetics, relaxin and prostacyclin, decrease intracellular Ca²⁺ by promoting calcium uptake into the intracellular stores such as the sarcoplasmic reticulum so free calcium levels decrease and the uterine muscle relaxes. Calcium channel blockers, such as nifedipine, prevent calcium entry into the cells promoting relaxation of the uterus.

Intracellular calcium levels (and myometrial activity) are controlled by various receptors on the myometrial cell surface. Most hormones that affect myometrial contractility bind to receptor sites that are coupled to one of two G-proteins: G_αs or G_αq (see Chapter 3; Bernal et al., 1995). The G-proteins act as transducers between the receptor and the effector regulating the cellular response by coupling the receptor to different signal-generating enzymes within the cell. These enzymes, in turn, generate an amplifying cascade of second messengers. The G-proteins allow the myometrial tissue to respond to a large number of agonists with a limited number of effects, either relaxation or contraction (Fig. 13.6). One of the G-protein pathways (G_αq) is linked to the inositol phosphate pathway. Binding of agonists (uterotonins) to receptors coupled to this G-protein activates phospholipase C (PLC), generating inositol triphosphate and diacylglycerol. Inositol trisphosphate stimulates the release of calcium ions from the sarcoplasmic reticulum and diacylglycerol activates the enzymes protein kinase C and phospholipase A. The latter releases arachidonic acid from membrane phospholipids; arachidonic acid is the precursor of prostaglandins. The result is a rise in intracellular calcium and therefore smooth muscle contraction. The other G-protein pathway (G_αs) activates adenylate cyclase, an enzyme that generates cyclic adenosine monophosphate (cAMP) and inhibits release of calcium. Agonists that stimulate this pathway, such as β-adrenergic receptor agonists and prostacyclin, will cause decreased release of Ca²⁺ from intracellular Ca²⁺ stores and dephosphorylation of myosin (so MLCK cannot bind Ca–CAM) and thus uterine relaxation (Yuan and López Bernal, 2007).

The onset of regular uterine contractions is gradual. As pregnancy progresses, the resting potential of the myometrial cells falls so that action potentials are generated with greater ease. Initially spontaneous uterine contractility is inhibited but becomes significant by mid-gestation. At about 20–24 weeks, hardening (or ‘tightening’) of the uterus can be felt as muscle contractions start. Initially, uterine activity is mostly of low amplitude and high frequency, with peak activity at night. This circadian rhythm stops at about 3 weeks before delivery, and may be mediated by cortisol (Germain et al., 1993). At first, small groups of myometrial cells contract together causing small fluctuations in intrauterine pressure. As pregnancy progresses, adjacent cells begin to contract synchronously so the contractions become more coordinated. This increased coordination results from cell–cell coupling and is due to the formation of intercellular gap junctions.

Gap junctions are formed from bundles of proteins, called connexins, that align forming pore-like symmetrical channels protruding through adjacent cells, so allowing contact and communication. Gap junctions exist in other tissues that act together in a coordinated fashion, such as cardiac muscle and pancreatic islets. In the open state, gap junctions allow rapid transmission of signals such as electrical stimuli and second messengers such as calcium and inositol trisphosphate (Fig. 13.7). This means that depolarization and smooth muscle contraction in one cell are quickly communicated to adjacent cells so there is a spread of excitation and a synchrony of contractions.

There is an increase in the synthesis of gap junctional protein connexin 43 and the number of gap junctions increases during gestation to about 1000 per myometrial cell. Physiological regulation of connexins and formation of gap junctions is by prostaglandins and steroid hormones; oestrogen and some prostaglandins increase and progesterone and nitric oxide suppress gap junction formation (Garfield et al., 1988). Gap junction density can be determined by measuring electrical resistance of tissue. The resistance of human myometrium at term is about half of that in the non-pregnant uterus and much less than that in other smooth muscle (for instance bladder and stomach), which indicates that the cells of the pregnant myometrium are very well coupled. Gap junction formation increases prior to spontaneous labour, resulting in synchronization and coordination of high-amplitude high-frequency myometrial contractions so increased intrauterine pressure is generated (Neulen and Breckwoldt, 1994). This is why palpation especially near to term, can stimulates uterine tightening; palpation initiates the first stimulation of the uterus which then spreads throughout the myometrium. Often fetal movements will trigger uterine tightening for the same reasons. Suppression of gap junctions may be important at the time of implantation and in early pregnancy (Grummer and Winterhager, 1998).

In practice, the cervix can be assessed by using a simple scoring system, the Bishop’s score (Table 13.1). This is particularly useful prior to induction of labour and for monitoring the changes in the cervix as the induction progresses.

If the cervix is soft, effaced and has started to dilate, induction may be implemented by artificial rupture of the membranes, which augments endogenous prostaglandin production. If the cervix is less favourable, prostaglandin E₂ (PGE₂) is administered into the posterior fornix of the vagina to facilitate effacement. However, PGE₂ should be used with caution in multiparous women because they have a greater sensitivity to it.

The cause of these structural changes in the cervix is not clear. It is thought to be hormonally controlled. Relaxin has been shown to be important in cervical ripening in rodents and has been used clinically to promote cervical ripening in humans. However, the levels of endogenous relaxin in pregnant women seem to be highest at the beginning of the second trimester. Oestrogen affects the synthesis of connective tissue components in vitro but has limited success when used pharmacologically as an induction method. PGE₂ causes cervical softening or ‘ripening’ and is produced naturally by both the cervix and the fetal membranes. PGE₂ appears to act by increasing collagenolytic activity rather than by changing the composition of the ground substance, which probably precedes prostaglandin use in successful induction of labour. Following delivery, glycoproteins that bind strongly to collagen are re-formed so the rigidity of the cervix is re-established; however, it never completely regains its original form (see Chapter 14). Damage to the cervix may have long-term consequences (Box 13.1).

In sheep, maturation of the fetal HPA initiates parturition. Human fetal malformations such as anencephaly (no cerebrum), malformed pituitary glands or hypoplasia of the adrenal glands are associated with an increased range of gestation length (both longer and shorter) but women still undergo spontaneous labour (Steer and Johnson, 1998). In sheep, similar fetal malformations, either accidental or deliberate, result in significantly prolonged gestation which adversely affects the fetus. This implies that the fetal–adrenal axis has a supportive rather than a direct role in parturition, acting to fine tune the gestational length in humans, rather than functioning as the ‘on–off switch’ for the initiation of labour, as seen in sheep.

It is evident that the human fetal anterior pituitary gland undergoes maturational changes in the last weeks of gestation as the profile of hormonal release changes. However, there are no defined changes measurable in the maternal circulation prior to the onset of labour. The raised cortisol levels in the cord blood of infants, who have experienced a spontaneous delivery rather than an induced one, or delivery by caesarean section, are maternally derived and are secondary to maternal pain. Labour itself, whether spontaneous or induced, causes stress and therefore an increased production of cortisol, which can cross the placenta. It is possible to differentiate between cortisol of maternal origin and cortisol of fetal origin by comparing the levels in umbilical cord arterial blood with those in the cord vein. Blood in the vein flows from the placenta to the fetus so if cortisol levels are higher in the vein than in the arteries this suggests the source of cortisol is maternal.

Exogenous corticosteroids (such as dexamethasone and betamethasone) given to a pregnant woman at 24–34 weeks to promote fetal lung maturity (see Chapter 15) do not initiate parturition, although oestrogen and cortisol levels fall. Pharmacological doses of glucocorticoids introduced into the amniotic fluid increase uterine activity and induce labour; however, this effect is not observed with physiological doses. Most researchers have not been able to measure an increase in cortisol prior to the onset of labour, nor an effect on placental hormone production. Therefore, it now seems unlikely that cortisol is important in initiating labour in humans, although it has a vital role in the maturation of fetal lungs and other organs. Several factors may affect physiological responses to cortisol; these include corticosteroid-binding protein (CBP) which would influence the relative concentration of free cortisol, metabolism of cortisol to inactive cortisone by 11β-hydroxysteroid dehydrogenase and modification of corticosteroid receptor expression.

In most animal models, progesterone withdrawal is the trigger for parturition. In the sheep, fetal cortisol affects steroidogenesis so progesterone levels fall and oestrogen levels increase. This is a result of increased expression and activity of placental 17α-hydroxylase (see above). Across almost all species, the fall in plasma progesterone concentration is a common endocrine event leading to parturition. However, progesterone levels do not fall prior to labour in humans. In humans, the placenta does not express inducible 17α-hydroxylase activity and therefore cannot convert progesterone to oestrogen. However, increases in free oestriol levels have been observed in saliva of women before the onset of labour, both preterm and at term (Goodwin, 1999).

High progesterone levels inhibit myometrial activity during implantation and favour uterine quiescence throughout the pregnancy. This suppression of myometrial activity is essential to the maintenance of pregnancy, although parturition begins without a measurable decrease in maternal peripheral progesterone levels. Large doses of progesterone are relatively unsuccessful in inhibiting preterm labour in humans. However, mifepristone, the progesterone receptor antagonist previously known as RU-486, inhibits progesterone so it blocks the effect of progesterone and induces labour (Thong and Baird, 1992). Mifepristone is used clinically as an effective abortifacient. It also has anti-glucocorticoid activity and can cause ripening of the cervix and increase sensitivity to contractile prostaglandins.

It is possible in humans that there could be a functional progesterone withdrawal such as local changes in progesterone concentration or receptor binding which, even in the absence of a fall in progesterone concentration in peripheral blood, may affect myometrial activity. As the uterus enlarges during the pregnancy, the part of the uterine wall distal to the site of implantation, and therefore furthest away from the major source of progesterone production, may re-establish uterine contractions and trigger the onset of labour (Challis et al., 2000). Measurement of absolute progesterone levels may therefore be misleading as the biological effects will be altered by receptor density, levels of binding proteins or postreceptor changes. The fetal membranes and maternal decidua can both metabolize progesterone and produce cortisol. Cortisol is antagonistic to progesterone and can increase the synthesis of CRH (Karalis et al., 1996). Fetal membranes, therefore, offer a mechanism for local control of progesterone concentration. Alternative mechanisms have been suggested for the apparent lack of progesterone withdrawal prior to human parturition. These include sequestration of progesterone, another progesterone-like steroid being involved as a natural antagonist or local inactivation of progesterone, for instance progesterone is converted into an inactive metabolite that competes for receptor sites (Challis et al., 2000). For instance, before labour, the major products of the paracrine steroid hormone synthesis could be progesterone and oestrone (a weak oestrogen) and following the onset of labour the major products could be less active progesterone metabolites and biologically active oestradiol (Mitchell and Wong, 1993). It has also been suggested that in humans there is a regionalization of uterine activity and high progesterone levels are important in promoting relaxation of the lower uterine segment while contractions in the fundal area facilitate the descent of the fetus (Challis et al., 2000). Changes in progesterone receptor expression could potentially act as a functional progesterone block and thus control myometrial progesterone responsiveness (Mesiano, 2004). The altered progesterone receptor profile is postulated to modulate oestrogen effects (and hence oxytocin responses) via expression of oestrogen receptors. The human progesterone receptor exists in several isoforms produced from a single gene. PR-B is the full-length isoform which is the major mediator of progesterone effects whereas PR-A lacks the N-terminal part of the protein that contains one of the functional domains so it tends to have effects that oppose those of PR-B. There are also PR-C and two other truncated isoforms. It seems that the expression of PR-A may increase significantly before the onset of labour at term so that the ratio of the two forms changes during gestation thus affecting the response of the uterine tissue and the ability of progesterone to suppress myometrial activity (Goldman and Shalev, 2007). The increased synthesis of PR-A may be mediated by inflammatory factors and Toll-like receptors (Smith et al., 2007); this pathway may be increased in some cases of preterm labour associated with infection.

The interaction of progesterone with its receptors requires specific co-activators; these co-activators are less abundant towards the onset of labour (Smith, 2007). Progesterone metabolites may also interact with the progesterone receptors in a different way to progesterone (Mitchell and Taggart, 2009). Progesterone bioactivity could also be regulated at the post-receptor level.

Human placental production of oestrogen increases throughout labour and the rate seems to increase in the latter part of gestation. In sheep, increased oestrogen causes increased contraction-associated proteins (CAPs) and up-regulation of prostaglandin synthase and consequent increased synthesis of PGE₂, promoting a positive feedback on myometrial contractility. However, spontaneous labour in humans can occur in the absence of measurable changes in oestrogen concentration. Exogenous oestrogen infusion in humans causes a transient increase in uterine activity and decreases the oxytocin threshold of the uterus but does not induce premature delivery or fetal membrane changes (Challis et al., 2000).

Synthesis of placental oestrogen depends on fetal cooperation in the provision of the precursors, and could thus potentially provide an opportunity for the human fetus to manipulate the progesterone:oestrogen ratio and uterine activity. The fetal adrenal gland in humans and higher primates is a relatively large percentage of body mass compared with the adult and has three zones: the outer adult zone produces mostly aldosterone, the unique fetal zone produces dehydroepiandrosterone sulphate (DHEAS) and the transitional zone produces mostly cortisol. This large fetal zone of the adrenal gland disappears in the neonatal period and may be regulated by hCG levels. DHEAS is the fetal adrenal C19 steroid precursor for placental oestradiol-17α and oestrone synthesis, which are implicated as having a role in labour. DHEAS is converted to oestrogens by placental sulphatase, aromatase and other enzymes (see Chapter 3). Production of DHEAS is controlled by ACTH from the anterior pituitary, which is itself regulated by hypothalamic CRH. CRH of placental origin can also stimulate production of DHEAS from the fetal zone (see below). However, women who have placental sulphatase or aromatase deficiencies and produce very little placental oestrogen can have normal pregnancy and labour (López Bernal, 2003). Thus, the changing ratio of oestrogen to progesterone may facilitate effective uterine contractions but is probably not critical for the induction of labour (Steer and Johnson, 1998).

Hypothalamic CRH controls the function of the pituitary–adrenal axis in response to stress. CRH stimulates the anterior pituitary gland to produce corticotrophin which then stimulates the cortex of the adrenal gland to release cortisol. However, even in the presence of significant stress, levels of CRH are relatively low compared to the levels reached in pregnancy. CRH is expressed abundantly in the syncytiotrophoblast cells of the human placenta (but not in non-primate placentas) and CRH-receptors are expressed in the primate placenta and myometrium. Placental CRH seems to have an important role in the initiation of parturition in higher primates and humans. CRH is synthesized by the placenta predominantly into the maternal circulation though some enters the fetal circulation (Smith, 2007). Levels of CRH steadily increase in the maternal circulation from mid-term (about 90 days before the onset of labour) until about 35 weeks when levels sharply rise and are usually particularly high in pregnancies ending with premature labour (Wolfe et al., 1990) and those complicated by pre-eclampsia (Smith and Nicholson, 2007). CRH activity is attenuated by CRH-binding protein (CRH-BP), which is synthesized by the liver, placenta and brain so most CRH is in the bound (inactive) form during pregnancy. However, towards term, when levels of CRH increase, levels of CRH-BP simultaneously fall and the capacity of the CRH-BPO is saturated so circulating levels of physiologically active, free CRH markedly increase. The maternal adrenal CRH receptors are down-regulated so the ACTH response to CRH is blunted in late pregnancy; this protects the maternal pituitary-adrenal axis from over-stimulation.

The stress hormone, cortisol, normally has a negative feedback effect to inhibit CRH secretion and thus ACTH and cortisol secretion as part of the HPA axis. The opposite situation occurs with placental CRH where cortisol has a positive feedback effect to further increase CRH release by the placenta. This positive feedback is further augmented by increased prostaglandin production (Petraglia et al., 1995), which also increases placental synthesis of CRH. Fetal stresses such as hypoxia and hypoglycaemia result in increased CRH concentrations. CRH is a vasodilator in the placental vascular bed (Clifton et al., 1994) so increased CRH should result in increased blood flow and abrogation of the fetal insult. However, if the insult persists, then CRH would increase fetal ACTH secretion and increase DHEAS production thus increasing oestrogen synthesis which leads to myometrial activation. Thus, CRH appears to be a mechanism by which a compromised fetus can precipitate labour when intrauterine life becomes unfavourable. The pathway by which CRH in the fetal circulation increases pituitary corticotrophin production which subsequently stimulates cortisol secretion by the fetal adrenal gland is important in the maturation of the fetal lungs and other organs particularly the central nervous system and gut (see Chapter 15). The maturing fetal lungs increase their production of surfactant proteins which can then enter the amniotic fluid; these surfactant proteins together with phospholipids and inflammatory cytokines are thought to stimulate inflammation in the fetal membranes and underlying myometrium and contribute to the amplification of the signals leading to labour (Smith, 2007).

It has been suggested that the ratio of placental CRH and CRH-BP acts as a placental ‘clock’ which controls the length of gestation and allows a distressed fetus to ‘wind-on’ the clock to initiate premature labour (Smith, 2007). CRH receptors can couple to different G-proteins in different tissues and at different stages of pregnancy (Karteris et al., 1998). So for most of pregnancy, CRH is coupled to the G-protein (G_αs) involved in increasing cAMP levels. It therefore has a ‘protective’ role and promotes myometrial quiescence via the generation of cAMP and cGMP, and up-regulation of nitric oxide synthase expression. Near term, the expression of CRH receptors changes to the form that couples to the other G-protein pathway (G_αq) so that CRH now activates PLC and mobilizes Ca²⁺ from intracellular calcium stores thus promoting increased myometrial contractility and labour (Hillhouse and Grammatopoulos, 2002). The involvement of a physiological stress pathway in the onset of labour could explain the relationship between maternal stress and reproductive failure (Nakamura et al., 2008) and the interesting possible relationship between maternal periconceptional nutritional status and the timing of parturition (MacLaughlin and McMillen, 2007). Maternal stress would stimulate the maternal pituitary-adrenal axis resulting in increased maternal cortisol production which would stimulate placental CRH release (Smith et al., 2007). Myometrial contractility is also enhanced by up-regulation of oxytocin receptor expression and cross-talk between the oxytocin and CRH receptors.

Prostaglandins (PGs) of the 2-series are known to be important in the feed-forward signal amplification and progression of labour; they promote myometrial contractions, cervical dilatation and membrane rupture. Prostaglandins are present in the circulation in low concentrations and are cleared in the pulmonary circulation so they are difficult to measure as they have paracrine activity and a short half-life. Prostaglandins can be formed as a consequence of tissue trauma (including labour itself and any manipulative or tactile stimuli). In late pregnancy, prostaglandin synthesis is readily stimulated by minor local stimuli, such as coitus, vaginal examination, sweeping the membranes or amniotomy, which are associated with inducing labour. Exogenous prostaglandins can be used therapeutically to ripen the cervix and to induce uterine contractions and labour. Mid-trimester abortion can be induced by procedures, such as intra-amniotic injection of hypertonic saline, that result in the increased synthesis and release of prostaglandins. Increasing DHEAS or oestrogen concentration and decreasing progesterone concentration increase production of prostaglandins (Challis et al., 2000). An increase is myometrial expression of prostaglandins receptors, particularly the receptor of PGF_2α, is implicated in the causes of some preterm labour (Olson et al., 2003).

Prostaglandins are synthesized in the fetal membranes, decidua, myometrium and cervix; levels fall abruptly following placental separation. The rate-limiting step of prostaglandin production may be important in initiating labour. Activity of phospholipase A₂ (PLA₂) may regulate the level of arachidonic acid, which is the precursor of prostaglandins (see Chapter 3). Increased oestrogen (or decreased progesterone) stimulates the release of PLA₂ from decidual lysosomes and therefore increases free arachidonic acid and subsequent prostaglandin synthesis. Arachidonic acid can also be produced indirectly via phospholipase C activity. Alternatively, the activity of cyclooxygenase (COX) may be rate-limiting (Aitken et al., 1990). The chorion produces prostaglandin dehydrogenase (PGDH), the enzyme which inactivates prostaglandins (Smith, 2007). In late pregnancy, chorionic PGDH activity decreases so the levels of PGE₂ rises.

The two most important prostaglandins in labour appear to be PGF_2α and PGE₂. At term, concentrations of PGE₂ and PGF_2α are higher in the decidua and myometrium (but these tissues are subject to tissue trauma so some authors believe raised prostaglandin levels are caused by increased uterine activity). PGE₂ is involved in cervical ripening, by mediating the release of MMP, and is metabolized by the myometrium to produce PGF_2α. However, women with extrauterine pregnancies go into labour at term and experience painful uterine contractions even though there are no fetal membranes in contact with the uterus.

Maintenance of human pregnancy may depend on the synthesis of PGE₂ being inhibited and this inhibition of prostaglandin synthesis being attenuated at the onset of labour. Prostaglandin concentrations in the pregnant uterus are very low (about 200 times lower than at any stage in the menstrual cycle), but increase sharply in the maternal circulation from 36 weeks’ gestation probably in response to the increasing level of CRH (Hertelendy and Zakar, 2004). Arachidonic acid, the precursor, is plentiful but synthesis of prostaglandins is inhibited, even if the pregnancy is extrauterine. Progesterone or the fetus may directly or indirectly moderate synthesis or metabolism of prostaglandins. Endogenous inhibitors of prostaglandin synthesis have been identified in maternal plasma; levels fall towards the end of gestation. Inhibitors of the COX enzymes, such as aspirin and indomethacin, block prostaglandin synthesis and are used therapeutically to treat preterm labour. Prostaglandin synthesis is up-regulated by cortisol (and dexamethasone), oestradiol, CRH and the inflammatory cytokines interleukin 1β (IL-1β) and tumour necrosis factor α (TNFα; Bowen et al., 2002).

Low doses of prostaglandins, particularly PGF_2α, increase myometrial responsiveness to prostaglandins and oxytocin possibly by increasing the formation of gap junctions. In vitro, PGE₂ has a biphasic effect, stimulating at nanomolar concentrations and inhibiting at micromolar concentrations. It also has a dual action: when its effects are mediated by the EP₁ and EP₃ receptors, PGE₂ increases intracellular calcium concentration, but the EP₂ receptor is coupled to adenylate cyclase so PGE₂ acting at this receptor decreases intracellular calcium and favours relaxation. Variations in regional prostaglandin synthesis may occur (Wilmsatt et al., 1995).

Prostacyclin (PGI₂) is synthesized in the myometrium and cervix. It promotes myometrial quiescence. It has an important role in maintaining uterine blood flow in labour; it causes vasodilation of smooth muscle and inhibits platelet aggregation, potentially inhibiting thromboembolic complications. Myometrial relaxation between uterine contractions in labour prevents occlusion of the uterine vessels and hypoxia which could lead to uterine dystocia (Hertelendy and Zakar, 2004). Placental thromboxane (TxA₂) has opposing effects to PGI₂ and is important in the closure of the fetal ductus arteriosus and haemostasis after delivery. In pre-eclampsia, levels of prostacyclin are low and levels of thromboxane are high. This is the rationale for aspirin treatment of pre-eclampsia. Aspirin-like drugs, by inhibiting the COX enzymes and thereby prostaglandin synthesis, restore thromboxane levels in pre-eclampsia. Trials using aspirin caused a slight prolongation of gestation and diminution of uterine contractions, but also increased the risk of premature closure of the ductus arteriosus and postnatal bleeding problems in the mother.

Prostaglandins are metabolized by the enzyme PGDH which is located in the fetal membranes. Deficiency of PGDH is associated with preterm labour. Expression of PGDH is up-regulated by progesterone and IL-10 and suppressed by cortisol (and dexamethasone), oestradiol, CRH, IL-1β and TNFα. Parturition is, therefore, essentially an inflammatory process. The association between infection and preterm labour is thought to be due to the release of phospholipases from bacterial organisms (Romero et al., 2003), which cause an increase in arachidonic acid release and so prostaglandin synthesis. Decidual macrophages respond to bacterial products by releasing pro-inflammatory cytokines. Bacterial endotoxins, such as lipopolysaccharide, can either increase prostaglandin release directly or further stimulate release of cytokines which then increase prostaglandin synthesis. Lipopolysaccharide also contributes to premature rupture of the membranes.

Prostaglandins also have an important role in the establishment of a neonatal circulatory pattern (see Chapter 15). Respiratory distress syndrome is associated with high levels of PGF_2α in the infant’s circulation, and patent ductus arteriosus with high PGE₂ levels. PGE₂ can prevent the ductus arteriosus from closing and inhibitors of prostaglandin synthesis can promote its closure. There is indirect evidence (i.e. fetal breathing movements, FBM, stop) that PGE₂ increases in the fetal circulation 48–72 h before the onset of labour (Thorburn, 1992). It is possible that this is mirrored by a local increase in PGE₂ concentration in uterine tissue, which initiates labour.

Oxytocin is a peptide synthesized by the hypothalamus and released from the posterior pituitary gland (see Chapter 3). A synthetic form (Oxytocinon or Syntocinon) is used extensively for induction and augmentation of human labour and can prevent postpartum bleeding or haemorrhage. Endogenous oxytocin production can also be stimulated, for instance by nipple stimulation, with a favourable outcome. However, it is not certain whether oxytocin is important in the initiation of labour. As oxytocin receptors are generally localized to the uterus, mammary glands and pituitary, oxytocin antagonists and agonists have few systemic effects. Maternal oxytocin levels are very low and do not change very much before labour. Maternal pituitary production of oxytocin dramatically increases in the first stage of labour. However, focusing on circulating levels of oxytocin may be misleading. The concentration of oxytocin receptors in the myometrium and decidua rise dramatically (by 100–200 times) during late pregnancy so the sensitivity of the uterus increases (Fuchs et al., 1984). This means the uterus can be stimulated by low concentrations of maternal oxytocin levels that previously had no effect. Therapeutic doses adequate to augment labour are very variable, which probably reflects individual differences in receptor number. If used to augment labour infusions of Oxytocinon should commenced on low dosage which is gradually increased until regular, strong contractions (around 3 in every 10 min time period) are present to ensure hyperstimulation of the myometrium is reduced. The pattern of oxytocin release changes at the onset of labour, with an increased frequency of pulses (Fuchs et al., 1991). It may be important that maternal oxytocin levels stay low during the pregnancy so the sensitivity of the uterus to oxytocin is maintained. Oxytocin antagonists, such as atosiban, have been used to inhibit uterine contractility in preterm labour (see Table 13.2) but not always effectively. Atosiban is useful in slowing down preterm labour so that steroids can be administered to facilitate fetal lung maturity however it can also cause raised blood glucose levels so should be used with caution with diabetic women (Berkman et al., 2003). Note that all tocolytic medications have side effects some of which can be extreme and potentially life-threatening (Blumenfeld and Lyell, 2009).