The document discusses the assessment and diagnosis of implantation failure, ectopic pregnancy, and miscarriage, highlighting limitations in current diagnostic methods and the importance of using relevant tests. It outlines various predictive modeling parameters, including the role of hormone levels such as human chorionic gonadotrophin (hcg) and progesterone in determining pregnancy outcomes. The author emphasizes the need for a practical and evidence-based approach for management while noting that a significant percentage of pregnancies of unknown location resolve spontaneously or require intervention.

1. Identifying the signs for
Implantation Failure and Miscarriage
By
Roy Farquharson
Liverpool Women's Hospital
UK
Contact: rgfarquharson@yahoo.com
1

2. Declaration of Interests
• Executive Committee member, European Society of
Human Reproduction and Embryology (2011 - 2015)
• NICE Guideline Development Group for ectopic
pregnancy and miscarriage (CG 154; 2010-2013)
• Chair, Association of Early Pregnancy Units
(2006-2011)
• Co-ordinator, ESHRE Special Interest Group for Early
Pregnancy (2007-2010)
• Associate Editor, Human Reproduction Update (20102014)
2

3. Educational Objectives
• Learning Objectives
•
• At the conclusion of this presentation, participants
should be able to:
• Describe the assessment for the diagnosis of
implantation failure, ectopic pregnancy and
miscarriage
• Acknowledge the limitations of available diagnostic
methods
• Develop a practical approach to using relevant tests
and management protocols.
3

4. 4

5. Predictive Modelling for Early
Pregnancy
Best
Area of Interest Diagnostic
Utility
Parameter(s)
Ovulation
Biomarker
D21 Progesterone
Pregnancy of
Unknown
Location (PUL)
Transvaginal
(TVU) Scan
and Biomarker
TVU Scan
+ HCG doubling
time/ratio
+/- Progesterone
Pregnancy of
Scan
Fetal heart action
Uncertain Viability Scan
plus Crown-rump
(PUV)
ExclusivelyScan length
5

6. Practical Advice - PULs
3. Predicting outcome
 Hormones
 Human chorionic gonadotrophin (hCG)
 Progesterone
 Other:




Creatine kinase
CA 125
Activin A
Inhibin A
 Mathematical Models
6

7. HCG changes in normal
pregnancy
•
Mean (SE) serum concentrations of human chorionic gonadotrophin (adapted
from Braunstein et al 1976)
7

8. Haemodynamically stable
Pain free
Expectant management
PUL
Haemodynamically stable
Haemodynamically unstable
Pain
Pain
? Serum hCG
Serum hCG levels
at 0 and 48 hrs +/progesterone
? Intra-uterine Pregnancy
Consider
laparoscopy/laparotomy
Consider laparoscopy
? Ectopic Pregnancy
? Failing PUL
8

9. Practical Advice - PULs
3. Predicting outcome
 Hormones
 Human chorionic gonadotrophin (hCG)
 Progesterone
 Other:




Creatine kinase
CA 125
Activin A
Inhibin A
 Mathematical Models
9

10. Serum hCG Levels
Single Levels
Serial Levels
Discriminatory Zone
10

11. Serum hCG Levels
Single Levels
Serial Levels
Discriminatory Zone
 Developed with respect to transabdominal USS
 Lower levels of hCG used with TVS
 Using a single value of hCG in a PUL population
is of limited value:
 Many ectopic pregnancies have a low hCG
 Clinicians may be falsely reassured
11

12. Serum hCG Levels
Single Levels
Serial Levels
Change over 48hrs
(hCG ratio)
Intrauterine Pregnancies (IUPs)
 Kadar et al. (1981) first to describe the minimal
rate of rise for an IUP to be 66% over 48hrs
 More recently minimal rise reported to be 53%
(Barnhart et al. 2004)
 In clinical practice a more conservative cut-off of
35% has been suggested
12

13. Serum hCG Levels
Single Levels
Serial Levels
Change over 48hrs
(hCG ratio)
Failing PULs
 A decline of 21-35% at 48 hours depending on initial
hCG level ( levels at presentation – rate of
decrease) (Barnhart et al. 2004)
 An hCG decrease of >13% (hCG ratio < 0.87) has
been shown to have a sensitivity of 92.7% and a
specificity of 96.7% for the prediction of a failing PUL
(Condous et al., 2006)
13

14. Serum hCG Levels
Single Levels
Serial Levels
Change over 48hrs
(hCG ratio)
Ectopic Pregnancies (EPs)
 ‘No single way to characterize the pattern of serum
hCG behaviour’ (Silva et al., 2006)
 hCG profile mimicked IUP in 21% and a spontaneous
miscarriage in 8% (Silva et al., 2006)
 Sensitivity of 83% for EP when IUP excluded by hCG
rise < 35% and failing PUL excluded by hCG decrease >
14
21-35% (Seeber et al., 2006)

15. Evidence based management of PULs
Predicting outcome
 Hormones
 Human chorionic gonadotrophin (hCG)
 Progesterone
 Other:




Creatine kinase
CA 125
Activin A
Inhibin A
 Mathematical Models
15

16. Serum Progesterone Levels
Serum
Progesterone
< 20 nmol/L
PPV > 95% to predict
pregnancy failure
(Banerjee et al., 2001)
Viable IUPs reported with
levels < 16nmol/L
> 60 nmol/L
‘Strongly’ associated
with viable
pregnancies
Discriminative capacity
insufficient to diagnose
ectopic pregnancy with
certainty (Mol et al., 1998)
Good at predicting viability but not location
16

17. Pregnancies of Unknown
Location (PULs)
• The majority of PULs fail and resolve
spontaneously (44% – 69%) RCOG green top guideline on
Tubal Pregnancy 2004 sourcing five observational studies
• Of the remainder, ectopic pregnancy was
subsequently diagnosed in 14 to 28%
• Intervention (medical or surgical) was
required in approx 25% cases
17

18. HCG in practice (NICE 2012)
• Clinical symptoms more important than HCG results
• HCG levels do not ‘locate’ the pregnancy nor assess
viability
• 2 levels 48 hours apart are useful for ‘risk
stratification’ and act as best evidence for
subsequent management
• Limitations of prediction should be shared and
acknowledged to patients (eg ectopic pregnancy HCG levels
mimic viable IUP in 21% and EPL in 8%)
• Ectopic pregnancy and miscarriage: diagnosis and initial management in
early pregnancy of ectopic pregnancy and miscarriage. (NICE Clinical
guideline 154; 2012; www.nice.org.uk)
18

19. Sites of ectopic pregnancies
Illustration: John Yanson.Seeber. Suspected Ectopic Pregnancy. Obstet Gynecol 2006.
From: Seeber: Obstet Gynecol, Volume 107(2, Part 1).February 2006.399-413
19

20. Ectopic Pregnancy
• Variable mode of presentation
• ‘Mask of invisibility’
• High index of suspicion and vigilance
eg against diagnosis of complete early
pregnancy loss
• All areas of emergency care provision will
receive cases of undiagnosed ectopic
pregnancy
20

21. Ectopic Pregnancy
presentation
• ACUTE (typical)
• Collapse with lower
abdominal pain, tachycardia
and hypotension
• Pain, amenorrhoea and sign
of pelvic tenderness
• EPU presentation with
positive pregnancy test,
scan showing empty uterus
and adenexal
inhomogeneous mass
• CHRONIC (atypical)
• Symptoms mimicking
gastroenteritis
• Light irregular bleeding
• >1/3rd of all patients have
no risk factors
21

22. HCG studies
•
•
•
•
•
•
•
Review question
What is the diagnostic accuracy of two or more hCG measurements for
determining an ectopic pregnancy in women with pain and bleeding and
pregnancy of unknown location?
Description of included studies
Nine studies were included in this review (Barnhart et al., 2010; Condous et al.,
2004; Condous et al., 2007; Dart et al., 1999; Daus et al., 1989; Hahlin et al., 1991;
Mol et al., 1998; Stewart et al., 1995; Thorburn et al., 1992).
Five prospective cohort studies (Condous et al., 2004; Condous et al., 2007; Hahlin
et al., 1991; Mol et al., 1998; Thorburn et al., 1992)
Four retrospective cohort studies (Barnhart et al., 2010; Dart et al., 1999; Daus et
al., 1989; Stewart et al., 1995).
Conducted in the UK (Condous et al., 2004; Condous et al., 2007), the USA (Dart et
al., 1999; Daus et al., 1989; Stewart et al., 1995), the Netherlands (Mol et al.,
1998) and Sweden (Hahlin et al., 1991; Thorburn et al., 1992). One study (Barnhart
et al., 2010) was conducted in both the UK and USA.
22

23. GRADE system
•
•
GRADE (Grading of Recommendations
Assessment,Development and Evaluation) assesses
evidence on an outcome-by-outcome basis
Quality can vary within a study and is based on 5
factors:
–
Study design
–
Limitations
Inconsistency
Indirectness
Imprecision
–
–
–
23
23
23

24. Summary of findings
Quality assessment
No. of
studies
Design
Limitati
ons
Incon
siste
ncy
Indirect
ness
Imprec
ision
Other
consider
ations
Nu
m
be
r
of
w
o
m
en
Measure of diagnostic accuracy
Sensiti
vity %
(95%
CI)
Specifi
city %
(95%
CI)
Positi
ve
predic
tive
value
%
(95%
CI)
Negative
predictiv
e value
% (95%
CI)
Positiv
e
likelih
ood
ratio
%
(95%
CI)
Negati
ve
likelih
ood
ratio
%
(95%
CI)
Qualit
y%
(95%
CI)
GRADE findings for the diagnosis of ectopic pregnancy using two or more hCG measurements
Model M4
1 study
Condous
et al.,
2007
prospec
tive
study
1 study
Barnhart
et
al.,
2010
retrospe
ctive
study
(2
included
cohorts:
UK and
adjuste
d USA)
serious1,
6
serious4,
6
serious4,
6
no
seriou
s
incons
istenc
y
serious2,
no
seriou
s
incons
istenc
y
serious2,
no
seriou
s
incons
istenc
y
serious1
12
12
no
serious
impreci
sion
none
no
serious
impreci
sion
none
173
431
80.0
(59.8,
100)
80.8
(65.6,
95.9)
88.6
(83.7,
93.6)
88.9
(85.8,
92.0)
40.0
(22.5,
57.5)
31.8
(20.6,
43.1)
97.9
(95.6,
100)
98.6
(97.4,
99.8)
7.02
(4.25,
11.61)
7.27
(5.21,
10.14)
0.23
(0.08
,
0.62)
0.22
(0.10
,
0.48)
LOW
V.
LOW
3
no
serious
impreci
sion
none
544
54.8
(45.2,
64.4)
87.7
(84.7,
90.8)
51.4
(42.1,
60.7)
89.2
(86.2,
92.1)
4.47
(3.29,
6.06)
0.52
(0.46
,
0.64)
V.
LOW
24

25. Treatment Options for EcP and PUL
•
Laparoscopic surgery ESEP RCT 2013 (NL)
– Salpingectomy versus Salpingostomy
•
Systemic Methotrexate (MTX)
- DEMETER RCT 2013 (Fr)
•
–
Expectant management
- METEX RCT 2013 (NL)
25

26. When can expectant management be
employed?
•
•
•
•
•
•
•
Clinically stable
Minimal symptoms
Discriminatory HCG zone: 1000-2000iu/l
Weekly USS
Twice weekly HCG until <20 iu/l
Compliance with follow up
Immediate access to hospital
26

27. Take home messages
• Laparoscopic surgery is cornerstone of
treatment intervention with ectopic
pregnancy
• Results of RCT’s (DEMETER) improve evidence
level and inform practice eg less MTX use
• Advantages of centralisation of care in EPU
• CMACH report on awareness of failed medical
management
27

28. It’s all about Quality of Care
28

29. The signs of Miscarriage
• Exclusively ultrasound based
• Updated CRL measurements
• Revised crown rump length criteria for
confirmed diagnosis of early pregnancy loss
(>7mm; NICE GDG & RCOG 2012)
• Acknowledgement of inherent, wide biological
variation of embryo growth velocities
• Specificity of viability assessment is 99.9%
29

30. Comparison of the CRL curve (solid line) with the Robinson
curve (dashdotted) and the Hadlock curve (dotted)
90
80
70
CRL (in mm)
60
50
40
30
20
10
0
40
50
60
70
GA (in days)
80
90
100
30

31. Updated Gestational Age
Measurement in early pregnancy
• Total number of pregnancies: 6666 (2002-2008)
• No. Excluded = 2956 (uncertain dates, redated, infertility treatment,
miscarriage, stillbirth, genetic or congenital abnormalities)
• No. Included = 3710 normal singleton pregnancies dated according to
known and recorded last menstrual period (LMP) with confirmed viability
at the time of the nuchal scan
• Predominantly transvaginal ultrasound below 10 weeks by contrast with
Robinson transabdominal derived CRL curve (BMJ, 1972)
• The gestational age (GA) ranged between 35 and 98 days
• Linear mixed-effects model in order to account for possible codependency of multiple CRL measurements in the same patient
Reference: Bottomley C ,Bourne T. Dating and
growth in the first trimester. Best Practice and
research Clin Obstet Gynaecol 2009 ; ESHRE
precongress course, Roma, 2010
31

32. TV Ultrasound
Fetal loss with CRL =7mm
32

33. Embryoscopy – the
close-up
H=head/heart prominence, Y=yolk sac, B=bubble
33

34. TVU – small embryonic structure
in disproportionately large sac
34

35. Embryoscopy
– short body stalk with 6mm CRL
- cytogenetics = 47XY+7
35

36. Fetal loss at 7 weeks
CRL = 19mm
36

37. Cytogenetics = 47XY+15
Small head compared to CRL, dysplastic face, partial encephalocele
37

38. Is Treatment Failure in RM a valid
concept?
- Cytogenetic Analysis of Pregnancy Loss in RM
38

39. Opportunityisnowhere
39

40. microarrays
• technique
high resolution WHOLE genome scan
cytogenetics
FISH
arrays
40

41. • Microarray
Advantages
- SINGLE test vs Karyotype + 5 FISH tests
- DNA extraction directly vs cell culture
- detect low level fetal cells vs maternal cell contamination
- higher resolution vs lower resolution
Disadvantages
- CANNOT detect ‘balanced’ rearrangements
- confirmatory follow up studies
41

42. • Trisomy 10 - TR
Karyotype = Normal Female
Array = Abnormal MALE result +10
FISH = confirmed +10 (70% MCC)
42

43. • 14q deletion - JS
Karyotype = Normal
Female
Array = Abnormal Female –
deletion 14q
FISH = confirm deletion in 11%
of cells (89% MCC)
43

44. RM – Evaluation of Array CGH v Conventional Cytogenetics
(McNamee et al, British Journal of Hospital Medicine, 2013, 74, 36-40 )
Array CGH and
conventional
cytogenetics
N=50
Triploidy on
FISH
N=4
Normal result
N=23(46%)
Abnormal result
N=27 (54%)
Diagnosed with
conventional
cytogenetics
Missed with
N=14
N=9
NUMERICAL
+16 x3
+15 x2
+21 x2
+13 x2
+22
+10
+14
-X x2
conventional
cytogenetics
NUMERICAL
+22
+10
+15
+8
+16
STRUCTURAL
>dup(22)(q11.2q11.2) ,
>del(14q)(q31.1) ,t(1:q1
6)mat
>del(13q)12.3-q34
44

45. Pregnancy Success Prediction Matrix
Following idiopathic RM, the predicted probability (%) of successful pregnancy is determined by age
and previous miscarriage history ( 95% confidence interval <20% in bold).
_____________________________________________________________________________
Age
Number of Previous Miscarriages
(yrs)
2
3
4
5
_____________________________________________________________________________
20
92
90
88
85
25
89
86
82
79
30
84
80
76
71
35
77
73
68
62
40
69
64
58
52
45
60
54
48
42
_______________________________________________________________________
Brigham et al, Hum Rep, 1999, 14, 2868-2871; PROMISE Trial 2010 MRC/HTA funded
45

46. Previous single miscarriage
Risk of preterm delivery <37 weeks
OR 1.1-1.4
ONE MISCARRIAGE
Basso '98
1.333/ 21.166
Buchmayer '04 1.293/ 21.631
Martius '98
1.069/ 13.461
Pickering '91
?/ 8.589
Smith '06
673/ 9.215
Hammoud '07
369/ 5.973
Pickering '85
?/ 3.927
Thom '92
174/ 2.146
Lang '96
?/?
Bhattacharya '08 128/ 1.404
Lekea '90
117/ 1.291
El-Bastawissi '03
69/ 143
Schoenbaum '80
17/189
Nguyen '04
16/164
0.1
1
10
46

47. Previous two or more miscarriages
Risk of preterm delivery <37 weeks
I
OR 1.1-1.4
II
OR 1.6-2.1
III
OR 1.5-3.0
TWO MISCARRIAGES
Basso '98
432/ 5.268
Martius '98
309/ 2.788
Smith '06
178/ 1.792
Buchmayer '04
146/ 1.742
Pickering '91
?/ 1.524
Hammoud '07
88/ 908
Lang '96
?/ ?
Pickering '85
?/ 689
Lekea '90
73/ 439
El-Bastawissi '03
31/ 57
Nguyen '04
8/ 33
RECURRENT MISCARRIAGE
Martius '98
151/ 639
Thom '92
63/ 638
Lang '96
?/ ?
Hammoud '07
36/ 225
Lekea '90
?/ ?
Hughes '91
11/88
Jivraj '01
7 /61
0.1
1
10
47

48. Previous miscarriage(s)
Risk of very preterm delivery <34 weeks
I
II
OR 2.2-3.4
III
ONE MISCARRIAGE
Basso '98
466/ 21.166
Buchmayer '04 219/ 21.631
Martius '98
195/ 13.461
Smith '06
138/ 9.215
Hammoud '07
92/ 5.973
Bhattacharya '08 39/ 1.404
Thom '92
26/ 2.146
El-Bastawissi '03
16/ 90
OR 1.5-1.7
NS 2.4-6.7
TWO MISCARRIAGES
Basso '98
158/ 5.268
Martius '98
71/ 2.788
Smith '06
56/1.792
Buchmayer '04
44/ 1.742
Hammoud '07
6/ 908
El-Bastawissi '03
6/ 32
RECURRENT MISCARRIAGE
Martius '98
52/ 639
Thom '92
27/ 638
Hammoud '07
5/ 225
0.1
1
10
48

49. Previous miscarriage(s)
Risk of small for gestational age
I
NS
II
OR 1.4
III
NS (?)
ONE MISCARRIAGE
Basso '98
1.291/ 21.166
Pickering '85
?/ 3.927
Lang '96
?/ ?
Thom '92
94/ 2.146
Parazzini '07
96/ 439
TWO MISCARRIAGES
Basso '98
395/ 5.268
Pickering '85
?/ 689
Lang '96
?/ ?
RECURRENT MISCARRIAGE
Thom '92
41/ 638
Lang '96
?/ ?
Jivraj '01
5/ 61
Hughes '91
3/ 88
0.1
1
10
49

50. Risk of adverse outcome in subsequent
pregnancy
OR/
*RR
Termination of
pregnancy
Miscarriage
1
≥2
≥3
1
≥2
1.0-3.31.2
1.0-1.54
-
ns
ns
Placental abruption
ns
1.54
-
ns
ns
Placenta previa
ns
1.74
*6.04
ns
ns
Preterm <37
1.1-1.43,5
1.6-2.13.5
*1.5-3.01,6
1.1-1.36,8
1.6-2.36,8
Preterm <34
1.5-1.73,5
2.2-3.43,5
*2.4-6.71,6
1.3-1.57,8
1.8-2.97,8
SGA p<10
ns
1.45
?1
ns
ns
LBW <2500
ns
?4,5
*2.04
ns
ns
LBW <1500
ns
ns
-
?9,10
?9
Cong. Malformation
ns
ns
*1.84
ns
ns
Low AS
ns
ns
ns
ns
ns
1.92
ns
ns
ns
ns
Number
Preeclampsia
Intrauterine Fetal death
1 Thom et al. 1992; 2 Bhattacharya et al., 2008; 3 Buchmayer et al., 2004; 4 Sheiner et al., 2005; 5 Basso et al.,
50
1998; 6 Martius et al., 1998; 7 Moreau et al 2005; 8 Ancel et al., 2004; 9 Lumley 1985; 10 Reime et al 2008

51. Vanishing Twin phenomenon
• Spontaneous reduction of a multiple pregnancy
• IVF-population (~5%)
• Incidence 10-30%1-3
• Studies: IVF population
• Vanishing twin IVF pregnancies, which were
spontaneous reduced from twin to single
pregnancies, were compared to single IVF
pregnancies
1 Dickey et al., 2002; 2 Landy and Keith 1998; 3 Pinborg et al., 2005
51

52. Vanishing Twin: Risk of
Preeclampsia and SGA
PREECLAMPSIA
*Pinborg 2007
Chasen 2006
X
SGA
Shebl 2007
*Pinborg 2007
La Sala 2004
Dickey 2002
Chasen 2006
0
1
2
3
4
5
6
Preeclampsia and SGA
52

53. Vanishing Twin;
SGA-LBW
Low birht weight <2500g
Cas
e
Control
O
R
95%CI
%Case
2004
Retrospective
62
437
Pinborg et al
2007
Retrospective
642
5.237
1,7
1,1-2,7
S
187
424
2,8
1,1-7,1
S
2,0
1,5-2,6
2005
Retrospective
642
5.237
Shebl et al
2007
Retrospective
46
8,9%
Signi
La Sala et al
Pinborg et al
9,7%
%Control
NS
vanishing twin > 8 wks vs <8 wks
11,7%
6,3%
0,001
92
26,1%
12,0%
0,036
3,2%
2,7%
4,1%
1,5%
Very low birth weight <1500g
La Sala et al
2004
Retrospective
62
437
Pinborg et al
2005
Retrospective
642
5.237
3,0
1,9-4,7
NS
0,001
Small for gestational age p<10
Chasen et al
2006
Retrospective
55
168
14,5%
9,6%
NS
Dickey et al
2002
Retrospective
140
4.683
15,7%
4,5%
NS
La Sala et al
2004
Retrospective
62
437
9,7%
15,6%
NS
Pinborg et al
2007
Retrospective
642
5.237
1,6
1,1-2,3
S
187
424
2,1
0,99-4,4
NS
46
92
Shebl et al
2007
Retrospective
32,6%
16,3%
Vanishing twin > 8 wks vs. <8 wks
0,029
53

54. The Sound of Life - Greetings
From Liverpool!!
54

55. Sites of ectopic pregnancies
Illustration: John Yanson.Seeber. Suspected Ectopic Pregnancy. Obstet Gynecol 2006.
From: Seeber: Obstet Gynecol, Volume 107(2, Part 1).February 2006.399-413
55