CORRECT IDENTIFICATION
OF PRETERM LABOUR &
pPROM
GIAN CARLO Di Renzo, MD, PhD, FACOG, FRCOG, ficogDept of Obstetrics and Gynecology &
Centre of Perinatal and Reproductive Medicine
University of Perugia, Perugia Italy
Prediction Preterm Labour
and pPROM
Gerard H.A. Visser
University Medical Center
Utrecht
The Netherlands
Prediction Preterm Labour and PPROM
• Preventive measures • lifestyle
• Cerclage
• Progesteron
• Pessary
• Prevention unnecessary treatment• Hospitalization
• Antibiotics
• Tocolytics
• corticosteroids
The short cervix at 24 wks
Iams et al NEJM, 1996
Effect of vaginal progesteron on
preterm birth < 33 wks
Do we have to screen all women for a
short cervix?
Do we have to screen all women for a
short cervix?
• 400-600 exams to prevent 1 preterm birth<34
wks (Fonseca 1.7% cervix<15 mm; Hassan 2.3% CL 1-2 cm)
• In low risk patients more exams necessary(Parry & Elovitz Clin O&G, 2014)
Do we have to screen all women for a
short cervix?
• 400-600 exams to prevent 1 preterm birth<34
wks (Fonseca 1.7% cervix<15 mm; Hassan 2.3% CL 1-2 cm)
• In low risk patients more exams necessary(Parry & Elovitz, Clin O&G, 2014)
• Sweden 0.5% CL <2.5 cm (Kuusela et al, Acta OGS, 2015
• Netherlands 0.8% CL <3 cm (van Os et al, Am J Perinatol, 2015
Do we have to screen all women for a
short cervix?
• 400-600 exams to prevent 1 preterm birth<34
wks (Fonseca 1.7% cervix<15 mm; Hassan 2.3% CL 1-2 cm)
• In low risk patients more exams necessary(Parry & Elovitz, Clin O&G, 2014)
• Sweden 0.5% CL <2.5 cm (Kuusela et al, Acta OGS, 2015
• Netherlands 0.8% CL <3 cm (van Os et al, Am J Perinatol, 2015
In these countries we should do at least 3 times
as many exams to prevent 1 preterm birth
Do we have to screen all women for a
short cervix?
• 400-600 exams to prevent 1 preterm birth<34
wks (Fonseca 1.7% cervix<15 mm; Hassan 2.3% CL 1-2 cm)
• In low risk patients more exams necessary(Parry & Elovitz, Clin O&G, 2014)
• Sweden 0.5% CL <2.5 cm (Kuusela et al, Acta OGS, 2015
• Netherlands 0.8% CL <3 cm (van Os et al, Am J Perinatol, 2015
NLs; n=20.234; short cervix PG 2 PTB; control
arm 4 PTB; i.e 5-10.000 exams to present 1 PTB
Standards of quality and reproducibility
1 of 4 ultrasound images submitted
for certification did not meet
published quality criteriaIams et al, AJOG, 2013
Suhag 2015
So, are we ready for nationwide
implementation of CL screening?
• Probably not yet
• Alternatives
- previous PT delivery
- all nullip’s or nullip’s with risk factors
- or
So, are we ready for nationwide
implementation of CL screening?
• Probably not yet
• Alternatives
- previous PT delivery
- all nullip’s or nullip’s with risk factors
- or a ‘risk-based screening’
( previous PTD,Cerv proced, race-ethnicity, smoking. Miller et al O&G, 2015)
Risk-based screening
Shortening of the cervix; do
we have to manage/act?
• YES
• Should we screen every pregnant
woman:probably not yet
• Will depend on risk profile, expertise and
ongoing trials (OPPTIMUM)
Thank you
Prediction Preterm Labour and PPROM
• Preventive measures• lifestyle
• Cerclage
• Progesteron
• Pessary
• Administration of MgSO4
• Prevention unnecessary treatment• Hospitalization
• Antibiotics
• Tocolytics
• corticosteroids
Unnecessary treatment modalities
• Antibiotics may cause fetal brain damage
• Tocolytics: most cause maternal side-effects
without improving perinatal outcome
• Corticosteroids: poison with some positive
side-effects
Antenatal CSs in low to middle income countries(Argentina,Guatemala, India, Kenya, Pakistan, Zambia)
Althabe et al, Lancet Febr 14, 2015
• Implementation program of CS in case of threatened
preterm birth versus standard care (n=98.000)
• Proxi for preterm birth: birthweight< 5th centile (36-
37wks)
• Intervention group CS in 45%, in control group 10%
• What will be neonatal outcome in infants weighing <
5th centile?
• What will be the overall perinatal mortality?
• And what about maternal morbidity?
Antenatal CSs in low to middle income countries(Argentina,Guatemala, India, Kenya, Pakistan, Zambia)
Althabe et al, Lancet Febr 14, 2015
• Implementation program of Cs in case of threatened
preterm birth versus standard care (n=98.000)
• Proxi for preterm birth: birthweight< 5th centile (36-
37wks)
• Intervention group CS in 45%, in control group 10%
• Neonatal mortality (<28d; <5th c group): RR 0.96 (0.87-1.06)
Antenatal CSs in low to middle income countries(Argentina,Guatemala, India, Kenya, Pakistan, Zambia)
Althabe et al, Lancet Febr 14, 2015
• Implementation program of Cs in case of threatened
preterm birth versus standard care (n=98.000)
• Proxi for preterm birth: birthweight< 5th centile (36-
37wks)
• Intervention group CS in 45%, in control group 10%
• Neonatal mortality (<28d; <5th c group): RR 0.96 (0.87-1.06)
• Total mortality : RR 1.12 (1.02-1.22)
• Maternal infections : RR 1.45 (1.33-1.58)
Antenatal CSs in low to middle income countries(Argentina,Guatemala, India, Kenya, Pakistan, Zambia)
Althabe et al, Lancet Febr 14, 2015
• 87% of CS were given to infants weighing> 2000-
2500g, where there is no evidence of its usefulness
• With risks of side-effects such as reduced
fetal/placental growth, apoptosis in the brain, CP and
maternal infection, which may explain the overall
poorer outcome
• These data stress the importance of adequate dating of
the pregnancy and of identifying women at real risk
of preterm birth (Visser & DiRenzo, correspondence Lancet, 2015)
Unnecessary treatment modalities
So, it is essential to restrict threatment to those
women who are less than 34 wks and who are
really going to deliver
Identification of true Preterm Labour
• Treatment may well be reduced by a better identification of
women who really are at increased risk of preterm delivery (CL
measurement, fibronectin)
• 24-34 wks, preterm contractions n=665, PTD 12%; Van Baaren et al, O&G 2014
n PTD<1wk
• CL >30mm 255 2 0.7%
• CL 15-30mm 297 25 8 %• CL 15-30 FN- 139 4 3 %
• CL 15-30 FN+ 148 21 14 %
• CL < 15 mm 113 53 47 %
Cervical length (TVUS)
EMG
Maternal BMI
Previous History
Bacterial vaginosis
IL-6
IL-8
IL1
fetal fibronectin (fFN)ferritin
α-fetoprotein
Placental alpha microglobulin-1
(PAMG-1)
human chorionic gonadotropin
prolactin
C-terminal propeptide of
procollagen
pIGFBP-1
Cervix or vagina Amniotic fluidcalgranulins
defensins
IL-6
IL-8
SerumG-CSF
ferritin
defensins
calgranulins
IGF BP-1 fragment
relaxin
Vitamins and micronutrients
CRP, CD163
Salivaoestriol
Prediction of PTB in Symptomatic Women
Biochemical and Biophysical Methods Overview
PartoSure vs Fetal Fibronectin vs Cervical Length
Nikolova T, Bayev O, Nikolova N, Di Renzo GC. Comparison of a novel test for placental alpha microglobulin-1 with fetal fibronectin and cervical length measurement for the prediction of
imminent spontaneous preterm delivery in patients with threatened preterm labor. J Perinat Med. 2015 Jan 6. [Epub ahead of print]
PartoSure is the single most accurate test when compared to fFN
and CL for prediction of imminent spontaneous delivery in patients
presenting with signs, symptoms, or complaints suggestive of PTL.“
”
Unnecessary Admissions
fFN1
32%29%
71%
68%87%
13%
“Admit”Necessary
Unnecessary
“Discharge”
PAMG-11
18%76%
24%
82%96%
4%
“Admit”Necessary
“Discharge”
+
-
+
-
Fetal
Fibronectin
(fFN)
Unnecessary
(1) Di Renzo et al. JPM 2015. (2) Lucovnik et al. AJOG 2013
Average cost of an
unnecessary admission
estimated at $20,372 USD 2
PartoSure may reduce
unnecessary admissions by
up to 80%1
Key Points
That the PartoSure test was found to be statistically superior to fFN and CL with respect to SP and
PPV (P < 0.01) provides evidence toward being able to significantly enhance current practice to
ultimately reduce the unnecessary administration of potentially harmful therapeutics to patients,
as well as reduce the economic burden associated with unnecessary hospital admissions.
“”
NEW! Study Data (Medway Hospital, NHS Trust, UK)
Test Dates Audited # Pts Observed PPV (%) Patients Admitted % Patients Admitted
Fetal Fibronectin (fFN) Test Jan – April 2015 79 8% 18 23%
PAMG-1 Test May – August 2015 119 50% 16 13%
A retrospective analysis of fetal-fibronectin (fFN) related admissions vs. PartoSure (PAMG-1)
related admissions in a UK birthing center
• 43% reduction in admissions• No pts sent home unnecessarily
Brume D, Harris K, Basu M, & Griffin S. A retrospective cost-benefit analysis: real-world application of the PartoSure test for the prediction of imminent spontaneous delivery in a UK birthing center. Medway Hospital, NHS Trust, UK, 2015. (unpublished)
PTL: diagnostic algorithmClinical signs of preterm labor:
Contractions of at least 30-40 seconds with at least 8 per hour
CL shortening by 80% or dilatation 1-3 cm
Changes in the cervix during 2 hours of observation
CL < 15 мм
Admission
Steroids
Tocolytics
CL 15-35 мм
PartoSure
(PAMG-1)
Positive (+)
Negative (-)
CL > 35 мм
Observation
PROM MANAGEMENT OPTIONSA. Antibiotic therapy
A. Intrapartum prophylaxis against early-onset of neonatal GBS infection
B. Antibiotics to prolong latency
B. Antenatal corticosteroidsA. Proven benefit <32 weeks’
gestationB. Unclear benefit 32-34 weeks
(lack of data)
C. Elective Delivery
D. Tocolytics (?) A. No consistent evidence of
prolonged latency
B. No evidence of improved neonatal outcome
C. Significant maternal side-effects
Mercer Obstet Gynecol 2003
Survival by GA at delivery
COCHRANE, 2010
40-47% of ROM cases are not obvious. These are the cases that diagnostic tests are used on.
Traditional ApproachesTrue Accuracy
Traditional Test SN (%) SP (%)
Ferning 50-63 71-86
Nitrazine (pH) 72-88 64-68
Pooling, AFI, Ferning 85 79
Pooling, AFI, nitrazine 72 98
Traditional Test NPV (%)
Pooling, nitrazine, ferning 55
“Evaluation of ferning, nitrazine, and/or ultrasound has shown that they add little, if anything, to speculum examination alone and that none of them are as accurate as the test based on biochemical markers. Accordingly, we believe that there is little to merit their use in modern practice.“
Di Renzo et al JMFNM 2011.
Traditional ApproachesClinical Practice Guidelines - EAPM
Sosa CG, Herrera E, Restrepo JC, Strauss A, Alonso J. Comparison of placental alpha microglobulin-1 in vaginal fluid with intra-amniotic injection of indigo carmine for the diagnosis of rupture of membranes. J Perinat Med. 2014 Sep;42(5):611-6.
Search for an Accurate TestPAMG-1 Indigo Carmine Study
• A prospective multi-site study incl. women reporting signs, symptoms, or complaints suggestive of ROM 21 - 42 weeks of gestation without obvious leakage of fluid from the cervical os during sterile speculum examination and without confirmation of ROM by traditional methods.
• A total of 140 patients were recruited with a prevalence of ROM of 19.3%.• The PAMG-1 immunoassay in vaginal fluid yielded results that were comparable to those of the
instillation of indigo carmine into the amniotic cavity; therefore, we propose that PAMG-1 is a sensitive and specific test to assess ROM in patients with an equivocal diagnosis based on simple tests.
Search for an Accurate Test – META-ANALYSISAmniSure (PAMG-1) vs IGFBP-1 Tests
Ramsauer B1, Vidaeff AC, Hösli I, Park JS, Strauss A, Khodjaeva Z, de la Cruz ÁA, Martínez-Astorquiza T, Horovitz J, Coatleven F, Helmer H. The diagnosis of rupture of fetal membranes (ROM): a meta-analysis. J Perinat Med. 2013 May;41(3):233-40.
OBJECTIVE: to compare the performance of tests based on the detection of insulin-like growth factor binding protein 1 (IGFBP-1) and placental α-microglobulin-1 (PAMG-1) in diagnosing rupture of fetal membranes (ROM) across different patient populations.
METHODS: A meta-analysis was conducted on 39 prospective observational or cohort studies investigating ROM tests based on the detection of IGFBP-1 and PAMG-1 meeting the following criteria:
1) performance metrics calculated by comparing results to an adequate reference method;
2) sensitivity thresholds of the investigated tests matching those of the currently available tests;
3) study population, as a minimum, included patients between 25 and 37 weeks of gestation.
Ramsauer B, Duwe W, Schlehe B, Pitts R, Wagner D, Wutkewicz K, Chuvashkin D, Abele H, Lachmann R. Effect of blood on ROM diagnosis accuracy of PAMG-1 and IGFBP-1 detecting rapid tests. J Perinat Med. 2014 Nov 8. [Epub ahead of print]
Search for an Accurate Test AmniSure (PAMG-1) vs IGFBP-1 Tests
in Patients w/ Vaginal Bleeding
OBJECTIVE: In up to 30% of patients presenting with signs and symptoms of a rupture of the fetal membranes (ROM), vaginal bleeding may be present. The presence of blood may lead to false positive results with biochemical markers.
METHODS: Herein data is presented from a multi-centric prospective observational clinical study that for the first time systematically evaluates the performance of PAMG-1 and IGFBP-1 detecting tests in 151 women with vaginal bleedings and signs and symptoms indicative of ROM.
RESULTS: Our data showed a better performance for the PAMG-1 detecting tests compared to the IGFBP-1 detecting tests in all quality parameters evaluated. In detail: Sensitivity is 97.8% (91.0%), Specificity is 91.5% (75.0%), PPV is 94.6% (83.5%) and NPV is 96.4% (85.7%) for PAMG-1 tests (and IGFBP-1 tests, respectively). This difference and also those in Specificity and PPV are statistically significant demonstrating superiority of PAMG-1 detecting tests over IGFBP-1 detecting tests.
“The PAMG-1 detecting test is significantly less susceptible to interference by blood than the IGFBP-1 detecting test.”
Biomarker Tests SN(%) SP (%)
PAMG-1 94-100 99-100
IGFBP-1 71-81 71-88
Traditional Tests SN(%) SP (%)
Ferning 50-63 71-86
Nitrazine (pH) 72-88 64-68
Pooling, AFI, Ferning 85 79
Pooling, AFI, nitrazine 72 98
Traditional Tests NPV (%)
Pooling, nitrazine, ferning 55
Search for an Accurate Test - SUMMARYBiomarkers vs. Traditional Tests – True Accuracy
TAKE HOME MESSAGE
• Preterm PROM is one leading cause of preterm birth, and its proper identification is essential
• Biochemical markers are better than the traditional methods, as they are specific to proteins found in amniotic fluid
• The rapid strip test based on PAMG-1 (AmniSure) seems to be a more sensitive bedside test compared with other tests
- 6 comparative studies 2008-2011
- All arrive at the same conclusionthat PAMG-1 is superior to IGFBP-1
for ROM diagnosis
PAMG-1 vs. IGFBP-1
Finally some progress as to preterm labour
• The old: -Corticosteroids
-Antibiotics
-Tocolytic drugs
-Cerclage
• The new: -MgSO4
-Progesteron
-Importance of short cervix:
-Cerclage
-Arabin pessary
-Progesteron
-
Perinatal deaths
Cochrane data base; courtesy of Z.Alfirevic
Preterm births
<37+0 weeks
The Arabin pessary
The Arabin pessary
However, Nicolaides et al: no
beneficial effects of pessary(presented in Marbella, June 2013)
Progesteron and preterm birth
• Past history of preterm birth:
- 6 trials; 1453 women
- results independent of route of administration
- delivery< 34 wks RR 0.31 (0.14-0.69)
- perinatal mortality RR 0.50 (0.33-0.75)
• Short cervix:
- 2 studies; n=438
- delivery<34 wks RR 0.64 (0.45-0.90)
- delivery< 28 wks RR 0.59 (0.37-0.93) (n=1115)
- vaginal progesteron only?
Dodd et al, Cochrane, 2013
Progesteron and preterm birth
• Past history of preterm birth:
- 6 trials; 1453 women
- results independent of route of administration
- delivery< 34 wks RR 0.31 (0.14-0.69)
- perinatal mortality RR 0.50 (0.33-0.75)
• Short cervix:
- 2 studies; n=438
- delivery<34 wks RR 0.64 (0.45-0.90)
- delivery< 28 wks RR 0.59 (0.37-0.93) (n=1115)
- vaginal progesteron only?
Dodd et al, Cochrane, 2013
20% of women with a
preterm delivery
Progesteron and preterm birth
• Past history of preterm birth:
- 6 trials; 1453 women
- results independent of route of administration
- delivery< 34 wks RR 0.31 (0.14-0.69)
- perinatal mortality RR 0.50 (0.33-0.75)
• Short cervix:
- 2 studies; n=438
- delivery<34 wks RR 0.64 (0.45-0.90)
- delivery< 28 wks RR 0.59 (0.37-0.93) (n=1115)
- vaginal progesteron only?
Dodd et al, Cochrane, 2013
20% of women with a
preterm delivery
Do we have to manage/act?
17 alpha hydroxy PG caproate
Cervix< 25 mm + other risk factors
N=105
Discontinued because of futility ( <34 wks 24-30%;
<32 wks 14-20%)
Winer et al AJOG 2015
Short cervix: do we have to manage/act?
YES
Gerard Visser
Poor Prof Antsaklis
Should preterm labour be
stopped at all?
Gerard H.A. Visser
University Medical Center
Utrecht
The Netherlands
Magnesium
- Not an effective tocolytic drug
- Blocks voltage dependent Ca2+ channels
- Glutamate N MDA receptor antagonist
extracellular Magnesium
calcium influx into cells is blocked
neural injury will be blocked
Antenatal magnesium sulfate and
neuroprevention
• 5 RCTs (in 4 neuroprevention was primary aim)
• Meta-analysis:
• RR
• Cerebral Palsy 0.69 (CI 0.54-0.87)
• Motor dysfunction 0.61 (CI 0.44-0.85)
• No effects on mortality or on other neurological
impairments or disabilities in the first years of life
Doyle LW, Current Opinion in Pediatrics 2012;24:154-159
Effect of magnesium sulfate on
cerebral palsy
Magnesium sulphate
To prevent one case of moderate/severe CP:
<28 wks, number needed to treat : 30
<32 wks, number needed to treat : 63
Dosage:
4 g i.v. loading dose; 1-2g/h maintenance for 12-24 h*
Maternal side effects:
Vasodilatation
Neuromuscular blocking agent Prevent overdosing
*Guidelines Australian & Canadian COG
Vogel et al(WHO) Lancet Nov 22, 2014
Use of CSs and tocolytic drugs in 29 low to
middle income countries
Main tocolytic drug: B-mimetics
Toc
21%
CSs
51%
Tocolytic drugs
Treatment Side effects Positive effects
1 Indomethacin ++++ ++++
2 -agonists +++ ++
3 Calcium channel blockers ++ ++
4 Oxytocin antagonists + ++
Mg-SO4, NO donors do not work
*Significant difference compared with atosiban
de Heus R, et al. BMJ 2009
Tocolytic drug N Severe Mild
Nifedipine 543 5 (0.9%)* 8 (1.5%)*
-agonists 158 3 (1.9%)* 4 (2.5%)*
Atosiban 576 0 (0%) 1 (0.2%)
Indomethacin 35 0 (0%) 0 (0%)
Side effects observed after a single course
of a tocolytic drug (n=1.333)
Side effects observed after tocolytics
Treatment Patients (n) Severe* Mild
Single course 1,333 8 (0.6%) 10 (0.8%)
Sequential courses 282 1 (0.4%) 3 (1.1%)
Combined courses 311 5 (1.6%) 1 (0.3%)
*Dyspnoea n=6; hypotension n=4; lung edema n=2; hypoxia n=1; cardiac failure n=1.
de Heus R, et al. BMJ 2009
Tocolytics and complicated pregnancies
Treatment (%) Atosiban Nifedipine -agonists
Overall distribution 43 40 12
Multiple gestations 50 40 10
Pre-existing CV disease 41 42 10
Diabetes 51 30 16
PIH/PE 43 26 24
Placenta praevia 39 27 25
R.De Heus et al, BMJ, 2009
Tocolytic drugs
• So, apparently in the Netherlands and Flanders the
underlying maternal condition is not taken into account,
when deciding which tocolytic drug to use
So…
So…
• Do not use -agonists anymore
• Do not give combined courses
• Consider giving atosiban
So…
• Do not use -agonists anymore
• Do not give combined courses
• Consider to give atosiban
• Especially in cases of multiple gestation, diabetes and
maternal cardiovascular problems
– i.e. take the maternal condition into account when deciding
which drug to use
Reassess the role of prostaglandin inhibitors (but not in MC twins)
Nifedipine versus atosiban; RCT, n=500
• nifedipine atosiban
• Primary outcome (maternal level):
• Prolongation of pregnancy 6 (1-38) 4 (1-30)
(median n days and IQR)
Vliet et al. AJOG 2015 (SMFM; Suppl Jan 2015, S4)
Nifedipine versus atosiban; RCT, n=500
• nifedipine atosiban RR
• Primary outcome (child level):
• Adverse perinatal outcome 12% 12%
• - perinatal death 5% 2% 2.2 (0.86-5.8)
• - BPD 3% 6%
• - sepsis 3% 1%
• - PVH> grade 2 1% 1%
• - NEC 3% 2%
Vliet et al. AJOG 2015 (SMFM; Suppl Jan 2015, S4)
Treatment NMulti
gestCVD Inf Diab
Other
tocol
Myocardial infarction 2 2
Acute pulmonary
edema9 5 1 1 2
Severe dyspnoea 8 6 1 4 3
Hypotension, IUFD 1
Hypotension, em CS 18
Atrial fibrillation 1
Calcium channel blockers: Side effects
Case reports: Oei,1999;Hodges,2004;Verhaert,2004;Vaast,2004;Van Geijn,2005;Nassar,2007;Gatault, 2008;Perbet.2008
Meta analyses on tocolytic drugs
placebo tocolytic
• Birth delay > 48 h 53% 75-93%
• Birth delay> 7 days 39% 61-78%
• With no lengthening of gestation beyond one week
Haas et al, Obstet Gynecol 2009;113:585-594
Meta analyses on tocolytic drugs
placebo tocolytic
• Birth delay > 48 h 53% 75-93%
• Birth delay> 7 days 39% 61-78%
• And no significant difference in RDS or neonatal survival
(in studies in which corticosteroids were given in both
arms)
Haas et al, Obstet Gynecol 2009;113:585-594
Meta analyses on tocolytic drugs
• placebo tocolytic
• Birth delay > 48 h 53% 75-93%
• Birth delay> 7 days 39% 61-78%
• No significant difference in RDS or neonatal survival ( in
studies in which corticosteroids were given in both arms)
Haas et al, Obstet Gynecol 2009;113:585-594
RCOG Greentop Guideline, 2010: no tocolytic
drug has been associated with a reduction in
prenatal or neonatal morbidity
Reason for absence of beneficial effects?
• The majority of preterm labours –with or without intact
membranes- is associated with infections or
inflammation
• And both are related to neurological and respiratory
complications, including PVL and CP
• So, delaying delivery may not prevent neurological
damage, but may even make it worse ( see also Oracle
trial: increased incidence of CP after 7 years in intact
membranes group; Kenyon et al, Lancet 2008)
Reason for absence of beneficial effects?
• The majority of preterm labours –with or without intact
membranes- is associated with infections or
inflammation
• And both are related to neurological and respiratory
complications, including PVL and CP
• So, delaying delivery may not prevent neurological
damage, but may even make it worse ( see also Oracle
trial: increased incidence of CP after 7 years in intact
membranes group; Kenyon et al, Lancet 2008)
So why don’t we only give a ( rescue) course
of corticosteroids and wait and see
Reason for absence of beneficial effects?
• The majority of preterm labours –with or without intact
membranes- is associated with infections or
inflammation
• And both are related to neurological and respiratory
complications, including PVL and CP
• So, delaying delivery may not prevent neurological
damage, but may even make it worse ( see also Oracle
trial: increased incidence of CP after 7 years in intact
membranes group; Kenyon et al, Lancet 2008)
Or corticosteroids and MgSO4
Reason for absence of beneficial effects?
• The majority of preterm labours –with or without intact
membranes- is associated with infections or
inflammation
• And both are related to neurological and respiratory
complications, including PVL and CP
• So, delaying delivery may not prevent neurological
damage, but may even make it worse ( see also Oracle
trial: increased incidence of CP after 7 years in intact
membranes group; Kenyon et al, Lancet 2008)
Or corticosteroids and MgSO4
The more so since MgSO4 works < 2 h*
* See also RCOG opinion paper 29, August 2011
Anyhow,
• 2 days should be more than enough
• Also for the achievement of proper action of
corticosteroids
• And for in utero transfer to a level 3 hospital
And what about maintenance tocolytic therapy?
• Oxytocin antagonists, one trial only
• Oral betamimetics, 13 trials
• Ca channel blockers, one trial only ( until 2102)
No effect on incidence of preterm birth or neonatal
morbiditiy
Cochrane databases: Papatsonis et al, 2009; Dodd et al, 2011; Gaunekaret al, 2010
Nifedipine maintenance, Dutch RCT
• N=406
• Mean age at inclusion: 29.2 wks
• Adverse outcome:
--11.9% placebo
--13.7% nifedipine
Roos et al, JAMA, 2013
Tocolytic maintenance therapy
• Who is still using oral maintenance therapy?
Conclusions
• There is no convincing evidence that tocolytics improve
neonatal outcome
• So, if you want to treat, do it only for a short time ( i.e. in
utero transfer) and with a drug that is safe for the mother
• But you may also consider to give corticosteroids and
MgSO4, instead.
• There is no place for tocolytic maintenance therapy
• And should we restrict tocolytics for cases with a
CL>1cm?
Asymptomatic patients with cervix< 15mm
• Women with intraamniotic inflammation ( AF MMP-
8>23ng/ml have a shorter cervix ( median1.5 mm (0-6.5)
vs 11 mm (4-13)
• And shorter diagnosis-delivery interval
Vaisbuch et al, AJOG 2010; gestational age at diagnosis 19-24 wks
RCT vaginal progesterone ; women with cervical
length 10-20 mm at 19-24 wks
Hassan et al UOG 2011
Anyhow
• There is no convincing evidence that tocolytics improve
neonatal evidence
• There is an indication for acute tocolysis during labour, in
case of overstimulation, for intrauterine resuscitation,
and/or to gain time for a CS
However,
Antenatal corticosteroids
Poison with some positive side effects
Sir Graham Liggins
Serendipity
Lambs born preterm were more
viable after antenatal glucocorticoid
administration
(Liggins, 1969)
Antenatal corticosteroids
1969 – Liggins
1972 – first RCT ( Liggins & Howie)
1994 – 15 RCT
2006 – 21 RCT
So there is a case to give corticosteroids in women at risk of preterm
delivery between 24 weeks and 34 weeks
Betamethasone is more effective than dexamethasone; but be aware of its
effects on FHR variation and movements
(Roberts and Dalziel, Cochrane, 2006)
1970s 1980s 1990s
RDS 0.55 0.71 0.69
PVH 0.50 0.61 0.53
Neonatal death 0.73 0.98 0.50
Antenatal steroids: RCT’s over the decades
Retrospective cohort studies
Corticosteroids also work at the threshold of
viability (22-24 weeks)
-Hayes et al, Obstet Gynecol 2008;111: 921-6; data from 1998-2007
-Abbasi et al, Am J Perinat.2010, 27:61-66; data from 1998-2008
-Mantlelow et al, Arch Child Fetal Neonatal Ed 2010, 95: F95-F98; data 1993-2007
Antenatal corticosteroids
• Reduce inflammation
• Promote surfactant mediated increases in functional
residual capacity
• Induce alveolar thinning
• Increase air special volume
• Increase lung antioxidant capacity
• Accelerate lung fluid reabsorption
Arayama et al 1992, Thibeault et al 1993, Vyas et al 1997, Bolt et al
2001, Halliday et al 2009
So,
Antenatal corticosteroids improve
outcome
In 40% of 420 European Centres >3 courses will be given
(Eurail, 2001)
37%
22%
14%
0
20
40
60
80
100
120
140
160
1 2 3 4 5 6 7 8 9 10 11 12 20
Number of courses, Europe 2000
Crowther and Harding, 5 RCTs
Repeated versus single corticosteroid doses
RR CI
Neonatal lung disease 0.82 0.72–0.93
Severe lung disease 0.60 0.48–0.75
Serious infant morbidity 0.79 0.67–0.93
Cochrane, 2007
Should steroids be repeated?
Author N Reduction severe/comp morbidity
Total group Early
Guinn 01 502 No Yes <27 weeks
Wapner 06 495 No Yes <32 weeks
Crowther 06 982 Yes <32 weeks
MACS trial 08 2304 No No < 32 weeks
MACS trial, results
Murphy et al, Lancet 2008
MACS trial, delivered < 32 weeks
Side effects multiple courses, at birth
Author N Side effect
Guinn 01 502 weight (ns p=0.10)
Wapner 06 495 weight, <10th centile*
Crowther 06 982 weight, head circumference
*Sadawy 07 93 Placental weight >32 weeks sign ↓
Direct side effects
MACS, Lancet December 2008
Decreased birth weight and head circumference
5-year follow up (MACS trial);Asztalos et al AJOG 2013(abstract)
Placebo Repeat
N 855 873
Composite mortality/morbidity 210 217
24.6% 24.9%
2-year follow up (Crowther et al, NEJM, 2007)
Placebo Repeat
N 526 521
Weight/HC/Bayley – –
Absence major disability 81% 84.4%
2-year follow up (Wapner et al, NEJM, 2007)
* 5 of 6 cases >3 courses, 5 >32 weeks of gestation
Placebo Repeat
N 236 248
Weight/HC/Bayley – –
CP 1 (0.5%) 6 (2.9%)*
(data from 19 RCT; Halliday, 2001)
Early neonatal treatment with corticosteroids
• 14 more extubated by 7 days
• 11 less have CLD
• 7 less will die
• 14 avoid late CS treatment
• 6 more have GI bleeding
• 4 more have GI perforation
• 12 have cerebral palsy
• 14 have abnormal
neurological development
at follow-up
For every 100 babies treated…
0.05–0.20 mg/day for 2 days
0.5 mg/kg/day for many days
Fetal versus neonatal dose
Potent drugs may have potent side effects
0.05–0.20 mg/day for 2 days
0.5 mg/kg/day for many days
Fetal versus neonatal dose
Corticosteroids
• Poison with some positive side
effects
(Noorlander et al, Dev Neurobiol, 2008 and PLoSOne 2013)
Apoptosis in fetal hippocampus following
antenatal corticosteroids in mice
0
50
200
E16
Ap
op
toti
c c
ell
s in
CA
SalineDexamethasone
100
150
250
E18 P0 P5 P10 P20 Adult
0
20
80
E16
Ap
op
toti
c c
ell
s in
DG
40
60
140
E18 P0 P5 P10 P20 Adult
100
120
SalineDexamethasone
** *
*
**
0
4,000
14,000
Ki-
67 p
osit
ive c
ell
s in
DG
8,000
10,000
20,000
E18 P0 P5 P10 P20 Adult
*
*
** *
18,000
16,000
12,000
6,000
2,000
(Noorlander et al, Dev Neurobiol, 2008 and PLoSOne 2013)
Cell proliferation following antenatal
corticosteroids in mice
SalineDexamethasone
Apoptosis versus cell proliferation
Noorlander et al, 2013; similar findings pren/neon exposure: Zuloaga et al, 2011; Chun-I Sze et
al, 2013
controls
Decreased number of proliferating cells in hippocampus after dexamethasone
Molecular analysis
0
100
200
300
400
500
600
Saline Dexamethasone
To
tal n
um
be
r K
i-67
po
sit
ive
ce
lls
Morris water maze:Hippocampus dependent
learning and memory task
No difference in swimming speed
Behavioural analysis
02468
1012141618
1 2 3 4 5Time (days)
Velo
cit
y (
cm
/sec)
SalineDexamethasone
Camera
Visual Cue
Light
Hidden
Platform
Morris water maze
*
*
*
Hidden platform
*
*
**
Impaired learning and memory after dexamethasone treatment
0
20
40
60
80
100
1 2 3 4 5
Late
ncy (
se
c)
0
20
40
60
80
100
1 2 3
0
400
800
1,200
1,600
1 2 3 4 5
Time (days)
Dis
tan
ce
sw
um
(cm
)
1 2 3
Visible platform
Behavioural analysis
SalineDexamethasone
SalineDexamethasone
SalineDexamethasone
SalineDexamethasone
Time (days)
Late
ncy (
sec)
0
400
800
1,200
1,600
Dis
tan
ce s
wu
m (
cm
)
Pericentral areas of fibrosis that extends to portal fields
Saline Dexamethasone
Peripheral analysis: Liver
Dexamethasone
Peripheral analysis: Heart
Saline
Myocytes → thin layer of collagen, indicating
degeneration of myocytes replaced by fibrous tissue
Dexamethasone induces precocious
aging and reduced lifespan in mice
Implications for the human…?
Follow-up after one course of
corticosteroids is reassuring
(McArthur et al, 1990; Smolders – de Haas et al, 1990; Schmand et al 1990;
Dessens et al, 2000; Dalziel et al, 2005 (2x),Karemaker 2006)
• no impairment at the age of 6 (maybe some
impaired visual memory)
. normal behaviour and motor function at 7-10 years
• normal physical and psychological development at
the age of 12 and 20 years
• normal cardiovascular and psychological development at
the age of 30 years (apart from increased insulin resistance)
Impact of corticosteroids on the density of
large neurons in the human hippocampus
(22 infants, 25–32 weeks, who died <4 days after delivery; Thijsseling et al, PLoSOne 2013)
Density of neurons Antenatal CS No antenatal CS
High (4) 1 6
Moderate (3) 4 3
Moderate/low (2) 6 2
Low (1) 0 0
Total n of neonates 11 11 (p<0.02)
Again, that was just to remind you, that…
• Corticosteroids are very potent drugs, and…
• Potent drugs may have serious side effects
Should steroids be repeated?
• Multiple courses of antenatal steroids do not increase or decrease the risk of death or developmental difficulties by 5 y of age.
• Because there is no clear benefit, this approach is not recommended for routine use
• Future research may be warranted for a more specified use of repeated courses
MACS-5; Asztalos et al, AJOG 2013 (abstract)
However,……WHO recommendations
• Interval Death RDS CVS haemorrhage
• <24 h RR 0.6 RR 0.87
• <48 h RR 0.59 RR 0.67 RR 0.26
• 1-7 d RR 0.81 RR 0.46
• >7 days RR 1.42 RR 0.82*
• * with a 150 g lower birth weight
A.Shennan Florence Sept 10, 2015
Most importantly
• Use of corticosteroids may well be reduced by a better
identification of women who really are at increased risk
of preterm delivery (CL measurement, fibronectin); Van
Baaren et al O&G 2014
• And by determining fetal lung maturation by
amniocentesis before a planned preterm delivery (CS).
Note: almost 50% of IUGR infants at 32 wks will have
sufficient lung maturation and do not need CSs
• Question: How many of your patients who received
corticosteroids actually delivered preterm? Utrecht area: 34%
delivered < 1 wk; Boesveld et al AJOG, 2014)
(Derks et al, BJOG,1995) ( Mulder et al, Ped Res 2004)
Betamethasone, FHR variation, F movements
Fetal Heart Rate Variation
Body and Breathing Movements
FHR var, Body moves & Breathing
moves day 2 as compared to day 0
Singleton Twins
FHR var -16 -14 %
Body Move’s -49 -42 %
Breathing Ms -86 -88 %
Corticosteroids in twins
• Neon death,2 trials n=236 RR 0.79 (0.39-1.61)
• RDS, 4 trials n=320 RR 0.85 (0.60-1.20)
• CVH, 1 trial n=137 RR 0.39 (0.07-2.06)
Roberts & Dalziel, Cochrane, 2006
Corticosteroids in twins
• Neon death,2 trials n=236 RR 0.79 (0.39-1.61) 0.69
• RDS, 4 trials n=320 RR 0.85 (0.60-1.20) 0.66
• CVH, 1 trial n=137 RR 0.39 (0.07-2.06) 0.28
Roberts & Dalziel, Cochrane, 2006
RRs more or less similar to those of overall analysis, though
small numbers meant the CIs were wide and crossed one
SGA, normal Doppler ua
Torrance et al. 2009
SGA, abnormal Doppler ua
Torrance et al. 2009
100 280 1.18 1.0 1.48
2015
2015
Sir Graham Liggins
And don’t forget:
• Women want to be treated
• And doctors want to treat
• but……………………..
And don’t forget:
• Women want to be treated
• And doctors want to treat
• but……………………..
• Doctors should know better
Thank you
A myth ??
• Kenyon & Peebles 2x
Finally some progress as to preterm labour
• The old:
-Corticosteroids
-Antibiotics?
-Tocolytic drugs?
-Cerclage?
• The new:-Importance of the short cervix
-Progesteron
-Arabin pessary
-MgSO4
Summary of Studies on PAMG-1
Study Year N PPV (%) NPV (%)
Nikolova et al.1 2014 101 78% 97%
Nikolova et al.2 2015 203 76% 96%
Lotfi et al.3 2015 150 75% 98%
Heverhagen et al.4 2015 64 100% 94%
Bolotskih et al.5 2015 49 75% 100%
Van Holsbeke et al.6 2015 35 75% 94%
1. Nikolova et al. J Perinat Med. 2014 Jul;42(4):473-7.
2. Nikolova et al. J Perinat Med. 2015 Jan 6.
3. Lotfi et al. World Congress of Perinatal Medicine (WCPM). 2015 (submitted abstract).
PAMG-1 has demonstrated a high negative predictive value (NPV) and a positive predictive
value (PPV) across studies and geographies
Prediction of del ≤ 7 days in tPTL pts
4. Heverhagen et al. Swiss Society of Obstetrics and Gynecology Annual Congress. 2015 (abstract).
5. Bolotskikh et al. Scientific and Practical Journal of Obs and Gyn Russian Fed. 2015; 2:94-98.
6. Van Holsbeke et al. World Congress of Perinatal Medicine (WCPM). 2015 (submitted abstract).
Because of inaccessibility of the gold standard that is essential to diagnose unknown/questionable PROM cases
IS INDIGO CARMINE STUDY AN UNREALISTICREFERENCE METHOD FOR A PROM STUDY?
Why would anyone conduct a ROM study on aknown/irrelevant population??