Ovulation Induction

for Anovulatory Infertility

(PCOS)
Adam Balen
Department of Reproductive Medicine
Leeds Teaching Hospitals, UK

ESHRE Campus, Kiev, May 2010

Ovulation Induction for PCOS

Learning Objectives

1. Options for OI
2. Weight reduction
3. Oral agents (Clomiphene Citrate, Aromatase Inhibitors)
4. Gonadotrophin therapy
5. Laparoscopic ovarian diathermy
6. Metformin
The Rotterdam ESHRE/ASRM Consensus Group
Revised 2003 Diagnostic Criteria for PCOS
2 out of 3 criteria required
Oligo- and/or anovulation
Hyperandrogenism (clinical and/or biochemical)
Polycystic ovaries
Exclusion of other causes of menstrual disturbance
and hyperandrogenism

Human Reproduction 2004; 19: 41-47. Fertility & Sterility, 2004; 81: 19-25.

Ultrasound Assessment of the Polycystic Ovary:

International Consensus Definitions
The polycystic ovary contains 12 or more
follicles measuring 2-9 mm in diameter

and/or

increased ovarian volume (>10 cm3)

Balen, Laven, Tan & Dewailly; Hum Reprod Update 2003; 9: 505
ESHRE/ASRM Consensus 2003

Elevated Luteinising Hormone:

- not mandatory for diagnosis

- most likely to be elevated in slim women
- may help predict outcome of fertility therapy:
- Worse outcome after CC if elevated day 8
- Better prognosis for response to ovarian drilling
1741 Women with PCOS

symptoms CENTRAL
OBESITY
hormones

ultra-
WEIGHT soundd
LOSS

INSULIN

Balen et al Hum Reprod 1995; 10: 2107

PCOS: Investigations
1. Testosterone (SHBG)
2. FSH, LH (E2)
3. AMH?
4. Prolactin / TFTs
5. Ultrasound scan
6. GTT, lipid profile
7. Semen analysis
8. Tubal patency assessment
First line therapy for anovulatory PCOS
• Weight loss

• Clomiphene citrate

?
• Aromatase inhibitors
• Gonadotrophins
G d t hi
• In vitro maturation of oocytes
• Ovarian surgery

• Insulin sensitisers???
sensitisers???

1 2 3 4
Preconcepti Clomifene Gonado-
Gonado-
onal care tropins IVF
(no (single ET)
- diet Metformin or
- exercise Aromatase
- other inhibitors))
inhibitors LOD

Should there be a cut off weight / BMI

before any treatment?

• Reduced chance conception

• Increased risk miscarriage

• Increased rate of congenital anomalies

• Obstetrical problems
(Gest DM, PET, delivery ….)
Balen, Dresner
Balen, Dresner,, Scott & Drife
BMJ 2006;332;434-
2006;332;434-435
Weight loss in PCOS

• 5-10% reduction in weight can achieve 30% reduction

in visceral fat

• Metabolic & endocrine profile improve significantly

• Improvement reproductive function and outcomes

Kiddy et al Clin Endo 1992 36:105

Clark et al Hum Rep 1995 10:2705
Jakubowicz & Nestler JCEM 1997 82:556

Weight Reduction:
RCOG Guidelines, 2007

No evidence for one type of diet

Strategies may include pharmacotherapy

(Orlistat,
Orlistat, not sibutramine or rimonabant)
rimonabant)

Bariatric surgery

Avoid pregnancy during rapid weight loss

BFS Guidelines, 2007

“Treatment should be deferred

until BMI < 35 kg/m2
although in those with more time
(under 37y, normal ovarian reserve)
a weight reduction to < 30 kg/m2 is preferable”

Balen & Anderson, Human Fertility 2007; 10: 195-

195-206
Weight loss and exercise
BMI > 30, > 2y anovulatory infertility, CC resistance

13/18 completed 6 month study:

weight loss improved endocrinology
12 - lower insulin
insulin, testosterone
all ovulated
11 conceived (5 naturally)

Clark et al H. Rep 1995 10:2705

Clomifene Citrate
n = 5268 patients
Ovulation - 3858 (73%)
Pregnancies - 1909 (36%)

Miscarriage - 20%
Multiple pregnancy rate - 10%

Single live-
live-birth rate – 25%

Homburg, Hum Reprod

Reprod,, 2005

To give hCG in CC cycles?

“ Routine addition of hCG at mid
mid--cycle
does not improve conception rates”

………but helps in timing of intercourse

or IUI

Agrawal & Buyalos,

Buyalos, 1995
 Should we monitor clomiphene
cycles with ultrasound?

3 cycles of CC

 G
Group 1:
1 N=105,
N 105
with U/S monitoring + hCG
 Group 2: N=150,
no U/S monitoring, no hCG

Konig,, Homburg et al, ESHRE, 2009

Konig

With U/S + hCG No U/S or hCG

48% Cumulative conception rate 34.7%

35.6%
35 6% Deliveries 26.7%
26 7%

0 Multiple pregnancies 1

Clomiphene Citrate
Starting…
 on day 2,3,4 or 5 makes no difference (Wu, 1989
1989))
 dose 50 mg/day, rising by 50mg if no ovulation
 even without withdrawal bleeding (Farhi,
Farhi, 2009)
2009)

Stopping…
Stopping…
 when 6 ovulatory cycles fail to yield a pregnancy
 when no ovulation with 150mg/day
 if endometrial thickness <7mm at ovulation
 Non-Response to Clomiphene

Failure to ovulate

 FAI
 BMI
 LH
 Insulin

Reasons for Clomiphene Failure

Ovulation but no conception

 Anti-estrogen effects
Anti-
- cervical mucus
- endometrium

 High LH

CC ovulation induction in WHO 2

- outcomes, predictors, and nomogram

Outcomes

Nomogram

Predictors
 Modern use of clomiphene citrate
in induction of ovulation

80
70
60
50
% 40
30
20
10
0
cycles
0 1 2 3 4 5 6 >6

Kousta, White & Franks. Hum Reprod Update 1997; 3: 359

Aromatase Inhibitors
- Theoretical Advantages
Letrozole (2.5 mg)
Do not block estrogen receptors –

 No detrimental effect on endometrium

or cervical mucus

 Negative feedback mechanism not

turned off – less chance of multiple
follicular development

Aromatase inhibitors
-questions
 Do they work?

 Better than CC for first

first--line treatment?

 Useful in CC resistance?

 Letrozole or anastrozole
anastrozole?
?

 Safety?
 Aromatase inhibitors for PCOS –
RCT’s vs CC
 Superiority or equivalence,
CC (100mg vs letrozole 2.5mg)
2.5mg)
Atay et al, 2006; Bayar et al, 2006

 CC, 100mg vs Letrozole

CC, Letrozole,, 5mg
 n=438 (1063 cycles)
 Pregnancy/cycle – CC 17.9%,
- letrozole 15.1% (NS)
Badawy et al, 2007

Aromatase inhibitors vs CC

 Meta-analysis,
Meta- analysis, 4 RCT’s
 Clear superiority of aromatase
inhibitors in pregnancy rates (OR 2
2.0)
0)
and deliveries (OR 2.4).

Atay 2006; Bayar 2006; Sohrabvand 2008; Sipe 2006

Meta--analysis: Polyzos et al, Fertil Steril
Meta Steril,, 2008

Letrozole induction of ovulation in women with

CC––resistant PCOS…
CC

• Ovulation
Ovulation-- 24/
24/44
44 cases (54.6%)

• Clinical pregnancy-
pregnancy- 6/44 cases
(25%of ovulators)
ovulators)

Elnashar et al, 2006

 Anastrozole
Anastozole (1mg/day) vs CC (100mg/day)
Anastrozole produced fewer follicles,
thicker endometrium.
endometrium. May be used
successfully for ovulation induction
Wu et al,
al 2007,
2007 n = 33

Anastrozole (1mg) vs Letrozole (2.5mg)

Letrozole superior in ovulation and
pregnancy rates
Al--Omari et al, 2004, n = 40
Al

Outcome – Letrozole vs CC

n=911 newborns in 5 centers

CC Letrozole
Pregnancies 397 514

Congenital
malformations
+ 19 (4.8%) 14 (2.4%)
Chromosomal
abnormalities Tulandi et al, 2006

Outcome – Letrozole vs CC

CC Letrozole
Pregnancies 397 514

M j malformations
Major lf ti 12 (3%) 6 (1.2%)
(1 2%)

VSD 4 (1.0%) 1 (0.2%)

Total cardiac anomalies 1.8% 0.2%

Tulandi et al, 2006

 Gonadotropin Regimens

 Step up

 Low dose step up

 Step down

Low dose rec-FSH

100--150 IU
100
75--112.5 IU
75
50--75 IU
50

14 7 7
Days

Reduction of multiple pregnancy rate

• Low dose regimens

• Careful monitoring

• Strict criteria for hCG:

1 follicle > 17mm
< 2 folls. in total > 14mm
 Gonadotrophin therapy in 103 women with PCOS

90
80
70
60
50
%
40
30
20
10
0
cycles
0 1 2 3 4 5 6 7 8 9 10 11 12

Balen et al Human Reproduction 1994; 9:1563

Low Dose Gonadotropins

Summary of Results
Patients - 1040, Cycles 2472
Pregnancies 411 (40%)
Fecundity/ov cycle
Fecundity/ov.cycle 23%
Uniovulation 71%
OHSS 0.14%
Multiple pregs. 5.1%

Updated from Homburg & Howles

Howles,, 1999

Results with low-dose gonadotrophins

-100 women

No response 15
Ovulate 85

No conception 40
Conception 45

Miscarriage 9
Multiple pregs 2
Singleton live birth 34
 Low-dose gonadotrophins
-questions
 Step--up or step-
Step step-down?

 Starting
g dose?

 Incremental dose rise?

 Use as first
first--line treatment?

21
14 days
1
100
75
50 Step--up
Step

If <10mm
10mm
10mm

Step--down
Step
100
75
50
3 days hCG

Conclusions

 Step-up safer and more efficient than

Step-
step--down
step
- Lower rate of overstimulation
- Higher
Hi h rate off monofollicular
f lli l cycles
l
- Higher ovulation rate

Christin--Maitre & Hugues

Christin Hugues,, 2003
Comparison of 2 starting doses
(37.5 vs 50 IU) r-hFSH for 14 days
N= 22: Mean Age 30.4 yrs: BMI 24.6 Increase after 14 days (37.5 & 50)

 Use of 37.5IU FSH as a starting dose resulted in similar

outcome but with less IUs FSH vs 50 IU

37.5IU 50IU

(Balasch et al 2000)

Only minimal dose increment needed

(Orvieto & Homburg, 2008)
 Incremental dose rise of 8.3 IU each week

64.6
64 6 IU
58.3 IU
50 IU
7 14 21 Days

 N=25, PCOS, CC failures, 69 cycles

Only minimal dose increment

needed (Orvieto & Homburg, 2008)
 Treatment days –
10.8 +
+\\- 4.3 (range 5-
5-25)

 Total dose of FSH (IU) –

622 +/-
+/- 286 (208-
(208-1641)

 Cycle cancellation – 1/69

 Ovulation rate – 98.5% of started cycles

 Only minimal dose increment
needed (Orvieto & Homburg, 2008)
 1 follicle only > 16mm 82.6%
> 14mm 62.3%
 Clinical ppregnancies
g 20 (29%
( cycles)
y )
Miscarriages 4
 Live births 16 / 25 patients
 Twins 1
 OHSS 0

Predictors FSH ovulation induction

- response dose, mono-follicle development, ongoing pregnancy

FSH response
p dose Mono-follicle
Mono- Ongoing
d l
development pegnancy

Conventional OI (CC + FSH)

singleton live birth

predicted chance
 Pregnancy rates of 240 WHO anovulatory
infertile women (CC, FSH, IVF)

Initial screening characteristics predicting

treatment outcome in WHO 2 anovulatory infertility

The way forward towards

patient tailored OI protocols
CC or rec-FSH for first-line treatment?

PCOS
Randomization

CC rec-
rec-FSH
3 cycles
 Randomized
N=268
CC FSH

Allocated Allocated
N=130 N=138

Drop-outs
Drop- Drop-outs
Drop-
N=18 N=24

Analysed Per--protocol
Per Analysed
N=112 N=114

Results
CC FSH P
Patients per protocol 112 114
Cycles 287 249

P
Pregnancies
i 46 (41%) 64 (56%) 0
0.02
02
Miscarriage rates 15% 12.5%
Multiple pregnancies 0 2 (3%)
Pregnancies/cycle 16% 26% 0.006
Live births 40 (35.7%) 56 (49%) 0.03

Cumulative live-birth rates

70%
60%
50%
40% CC
30% FSH

20%
10%
0%

Cycle 1 2 3
After 3 cycles - CC 42%, FSH 62%
 Summary

Clear superiority of low-

low-dose FSH over CC for first
line treatment of anovulatory PCOS

Absolute difference -
 off 24% iin CCR over 3 cycles
l
 of 10% in pregnancy rates/cycle
 of 20% in cumulative live birth rates

 More than x2 chance of conception in 1st cycle

 Shorter treatment to pregnancy time

Laparoscopic ovarian surgery

40W, monopolar,
monopolar, 4 points, 4 seconds

LOD versus rFSH - RCT

LOD rFSH
n 83 85
ovulatory 63% 64%
pregnant 34% 67% CCR 6 cycles
RR 0.54 ((95%CI 0.39-
0.39-0.76))
12m CCR 67% 67%
RR 1.01 (95%CI 0.81-
0.81-1.24)
miscarriage 9% 13%

After 8w 45 received addition of CC  49% CCR

and 21 then received rFSH  67% CCR at 12m
Bayram et al, BMJ 2004; 328:192
LOD versus rFSH - RCT
LOD rFSH
n 83 85
ovulatory 63% 64%
pregnant 34% 67% CCR 6 cycles
RR 0.54 ((95%CI 0.39-
0.39-0.76))
12m CCR 67% 67%
RR 1.01 (95%CI 0.81-
0.81-1.24)
miscarriage 9% 13%

After 8w 45 received addition of CC  49% CCR

and 21 then received rFSH  67% CCR at 12m
Bayram et al, BMJ 2004; 328:192

LOD versus rFSH - RCT

LOD rFSH
n 83 85
ovulatory 63% 64%
pregnant 34% 67% CCR 6 cycles
RR 0.54 ((95%CI 0.39-
0.39-0.76))
12m CCR 67% 67%
RR 1.01 (95%CI 0.81-
0.81-1.24)
miscarriage 9% 13%

After 8w 45 received addition of CC  49% CCR

and 21 then received rFSH  67% CCR at 12m
Bayram et al, BMJ 2004; 328:192

LOD versus rFSH - RCT

LOD rFSH
n 83 85
ovulatory 63% 64%
pregnant 34% 67% CCR 6 cycles
RR 0.54 ((95%CI 0.39-
0.39-0.76))
12m CCR 67% 67%
RR 1.01 (95%CI 0.81-
0.81-1.24)
miscarriage 9% 13%

After 8w 45 received addition of CC  49% CCR

and 21 then received rFSH  67% CCR at 12m
Bayram et al, BMJ 2004; 328:192
LOD versus rFSH - RCT
LOD rFSH
n 83 85
ovulatory 63% 64%
pregnant 34% 67% CCR 6 cycles
RR 0.54 ((95%CI 0.39-
0.39-0.76))
12m CCR 67% 67%
RR 1.01 (95%CI 0.81-
0.81-1.24)
miscarriage 9% 13%

After 8w 45 received addition of CC  49% CCR

and 21 then received rFSH  67% CCR at 12m
Bayram et al, BMJ 2004; 328:192

Laparoscopic “drilling” by diathermy or laser in

anovulatory PCOS
 studies small
 main outcomes ovulation & pregnancy

 6 month ppregnancy
g y rate vs 6 cycles
y gonadotrophin
g p
therapy: OR 0.48, 95% CI 0.28 – 0.81
 12 month pooled OR 1.27, 95% CI 0.77 – 2.09

Farquhar et al, Cochrane database 2002

Laparoscopic “drilling” by diathermy or laser in

anovulatory PCOS

 miscarriage rates - similar

 multiple pregnancy rates - lower
(OR 0.16, 95% CI 0.03 – 0.98)

Farquhar et al, Cochrane database 2002

 Effects of metformin on PCOS
 Improve reproductive function
 Improve response to both clomifene and
gonadotropin induced ovulation

Lord et al,
al, 2003, Cochrane Review & BMJ
Costello et al,
al, 2003, Human Reproduction

A multi-centre randomised, placebo-controlled ,

double-blind study, of combined life-style modification
& metformin in obese patients with PCOS

 8 centres U.K., co-ordinated by Leeds

 Placebo controlled, double blind RCT
 6 months metformin 850mg b.d.
 143 women randomised, with BMI > 30 kgm-2
mean BMI 38 kgm-2
power 0.90 for significance 0.05, requires 55 per arm of study)

Tang et al, Human Reproduction 2006; 21: 80-

80-89.

Randomised
143

Metformin Placebo
69 74

Withdrew Withdrew
13 8

Completed Completed
56 66
 Metformin vs Placebo

Significant increase in number of cycles,

and fall in BMI and waist circumference in both groups

No difference in ovulation rate between the groups

Improvements seen in those who lost weight in either group

Tang et al, Human Reproduction 2006; 21: 80

80--89.

A randomised double blind clinical trial comparing

clomifene citrate plus metformin with clomifene citrate
plus placebo in newly diagnosed PCOS

228 women with PCOS

Randomly allocated to receive either metformin 2000 mg/d

or p
placebo for 1 month

Then clomifene citrate 50 up to 150 mg for 6 ovulations or

until CC-resistance

BMI ~ 28 kg/m2

Moll et al, BMJ; 332: 1485

Ovulation per dosage clomifene citrate

CC + metformin CC + placebo P

CC 50mg 49/80 (61%) 50/92 (54%) 0.36

CC 100mg 27/44 (61%) 35/53 (66%) 0.63

CC 150mg 8/17 (47%) 13/23 (57%) 0.55

Moll et al BMJ 2006; 332: 1485

Ovulation, pregnancy and spontaneous abortion rates

CC CC Relative Risk
+ metformin + placebo (95% CI)

n=111 n=114

Ovulation 71 (64%) 82 (72%) 0.89 (0.7 - 1.1)

Ongoing
Pregnancy 44 (40%) 52 (46%) 0.87 (0.6 - 1.2)

Spontaneous
Abortion 13 (12%) 12 (11%) 1.11 (0.5 - 2.3)

Moll et al BMJ 2006; 332: 1485

Discontinuation due to side effects:

16% versus 5% (95% CI 5 - 16%)

Moll et al BMJ 2006; 332: 1485

CC and/or metformin alone or in combination

626 anovulatory PCOS

Metformin vs Placebo 2000 mg / day

Clomiphene or Placebo 50 – 150 mg for 5d

6 cycles or 30 weeks

Mean BMI ~ 35 kg/m2

Legro et al, NEJM 2007, 356:551

CC and/or metformin alone or in combination

CC M CC + M
Conception 39.5% 8.4% 46.0%
/ovulation
Miscarriage 8 3%
8.3% 20 8%
20.8% 9 2%
9.2%

Live birth 22.5% 7.2% 26.8%

(47/209) (15/208) (56/209)

CC superior to metformin and combination confers no

advantage in achieving live birth
Legro et al, NEJM 2007, 356:551
 Clomiphene with Metformin or Placebo

Two very large RCTs have failed to show any

benefit from metformin

Clomiphene alone results in highest livebirth rate

Moll et al, BMJ 2006; 332:1485

Legro et al, NEJM 2007; 356:551

Revised Cochrane Meta-analysis

Tommy Tang, Rob Norman, Adam Balen

2009

Metformin vs placebo or no treatment: Body weight

Metformin Control Mean Difference Mean Difference

Study or Subgroup Mean SD Total Mean SD Total Weight IV, Fixed, 95% CI IV, Fixed, 95% CI
Baillargeon 2004 61.4 1.6 28 61.4 1.6 30 96.3% 0.00 [-0.82, 0.82]
Kelly 2002 91 24 10 94 31 10 0.1% -3.00 [-27.30, 21.30]
Lord 2006 94.7 27.1 16 94.9 15.5 15 0.3% -0.20 [-15.62, 15.22]
Pasquali 2000 94 17 10 97 18 8 0.2% -3.00 [-19.33, 13.33]
Tang 2006 99 15 56 99.2 17.3 66 2.0% -0.20 [-5.93, 5.53]
Trolle2007 92.9 19 42 96.1 21.1 45 0.9% -3.20 [-11.63, 5.23]
Vandermolen 2001 96.9 26.53 11 106.9 23.2 14 0.2% -10.00 [-29.84, 9.84]

Total (95% CI) 173 188 100.0% -0.06 [-0.87, 0.75]

Heterogeneity: Chi² = 1.70, df = 6 (P = 0.94); I² = 0%
-10 -5 0 5 10
Test for overall effect: Z = 0.15 (P = 0.88) Favours metformin Favours control
Live birth rate: OR 1.00 95% CI 0.16, 6.39
OR -0.06 95% CI -0.87, 0.75
 Metformin versus placebo or no treatment: Live birth rate

Metformin Control Odds Ratio Odds Ratio

Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Ng 2001 1 9 2 9 79.6% 0.44 [0.03, 5.93]
Yarali 2002 1 16 0 16 20.4% 3.19 [0.12, 84.43]

Total (95% CI) 25 25 100.0% 1.00 [0.16, 6.39]

Total events 2 2
Heterogeneity:
g y Chi² = 0.87,, dfor=Subgroup
Outcome 1 (P
( = 0.35);
); I² = 0% Participa
Studies Statistical Method Effect Estimate
nts 0.001
0 001 01
0.1 1 10 1000
Test for overall effect: Z = 0.00 (P = 1.00) Favours control Favours metformin

Live birth rate: OR 1.00 95% CI 0.16, 6.39

Metformin versus Clomifene Citrate: Live birth Rate

Metformin Clomifene Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Legro 2007 15 208 47 209 87.2% 0.27 [0.14, 0.50]
Palomba 2005 26 50 9 50 0.0% 4.94 [1.99, 12.26]
Zain 2008 4 42 7 41 12.8% 0.51 [0.14, 1.90]

Total (95% CI) 300 300 100.0% 0.30 [0.17, 0.52]

Total events 45 63
Heterogeneity: Chi² = 0.76, df = 1 (P = 0.38); I² = 0%
0.01 0.1 1 10 100
Test for overall effect: Z = 4.25 (P < 0.0001)
Favours Clomifene Favours Metformin

OR 0.30 95% CI 0.17, 0.52

Live birth rate: OR 1.00 95% CI 0.16, 6.39

Metformin plus ovulation induction agent

vs ovulation induction agent alone: Ovulation Rate
Treatment Control Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Random, 95% CI M-H, Random, 95% CI
3.3.1 PCOS and clomifene sensitive
Jakubowicz 2001 26 28 22 28 5.5% 3.55 [0.65, 19.37]
Subtotal (95% CI) 28 28 5.5% 3.55 [0.65, 19.37]
Total events 26 22
Heterogeneity: Not applicable
Test for overall effect: Z = 1.46 (P = 0.14)

3.3.2 PCOS and clomifene resistant

Hwu 2005 17 40 5 40 8.2% 5.17 [1.68, 15.98]
Kocak 2002 21 27 4 28 6.8% 21.00 [5.21, 84.66]
Malkawi 2002 11 16 3 12 5.6% 6.60 [1.23, 35.44]
Ng 2001 4 9 1 9 3.4% 6.40 [0.55, 74.89]
Sturrock 2002 5 12 4 14 5.8% 1.79 [0.35, 9.13]
Vandermolen 2001 9 12 4 15 5.4% 8.25 [1.45, 46.86]
Subtotal (95% CI) 116 118 35 2%
35.2% 6 55 [3
6.55 [3.40,
40 12
12.63]
63]
Total events 67 21
Heterogeneity: Tau² = 0.05; Chi² = 5.36, df = 5 (P = 0.37); I² = 7%

CC resistant Test for overall effect: Z = 5.61 (P < 0.00001)

3.3.3 PCOS and clomifene sensitivity not defined

OR 6.55 95% CI 3.40, 12.63
El-Biely 2001 35 45 29 45 9.3% 1.93 [0.76, 4.90]
Khorram 2006 7 16 1 15 3.9% 10.89 [1.14, 103.98]
Legro 2007 582 964 462 942 12.9% 1.58 [1.32, 1.90]

Live birth rate: OR 1.00 95% CI 0.16, 6.39

Moll 2006
Nestler 1998
84
19
141
21
98
2
168
25
11.9%
4.4%
1.05 [0.67, 1.66]
109.25 [14.04, 850.33]
Sahin 2004 38 51 34 55 9.8% 1.81 [0.79, 4.15]
Zain 2008 38 41 24 41 7.1% 8.97 [2.37, 33.91]
Subtotal (95% CI) 1279 1291 59.2% 2.75 [1.48, 5.11]
Total events 803 650
Heterogeneity: Tau² = 0.42; Chi² = 29.18, df = 6 (P < 0.0001); I² = 79%

CC sensitive Test for overall effect: Z = 3.20 (P = 0.001)

OR 2.75 95% CI 1.48, 5.11
Total (95% CI) 1423 1437 100.0% 3.93 [2.32, 6.65]
Total events 896 693
All Test for overall effect: Z = 5.08 (P < 0.00001)
OR 3.93 95% CI 2.32, 6.65
Heterogeneity: Tau² = 0.55; Chi² = 53.17, df = 13 (P < 0.00001); I² = 76%
0.01 0.1 1 10 100
Favours control Favours treatment
 Metformin plus ovulation induction agent
vs ovulation induction agent alone: Live Birth Rate
Treatment Control Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Legro 2007 56 209 47 209 50.7% 1.26 [0.81, 1.97]
Moll 2006 21 111 31 114 36.5% 0.62 [0.33, 1.17]
Sahin 2004 3 11 3 10 3.4% 0.88 [0.13, 5.82]
Vandermolen 2001 4 12 1 15 0.9% 7.00 [0.66, 73.93]
Zain 2008 7 41 7 41 8.6% 1.00 [0.32, 3.16]

Total (95% CI) 384 389 100.0% 1.04 [0.75, 1.46]

Total events 91 89
Heterogeneity: Chi² = 5.79, df = 4 (P = 0.22); I² = 31%
0.01 0.1 1 10 100
Test for overall effect: Z = 0.25 (P = 0.80) Favours control Favours metformin

OR 1.04 95% CI 0.75, 1.46

Live birth rate: OR 1.00 95% CI 0.16, 6.39

Insulin sensitising agents in PCOS:

ESHRE/ASRM Consensus, 2007

• No clear role of metformin in management

anovulatory
y infertility
y either alone or in combination

• No evidence of improvement in pregnancy outcome

Human Reproduction 2008; 23:462

Fertility & Sterility 2008; 89: 505

RCOG Scientific Advisory Committee Guideline, 2008

Ovulation Induction for PCOS

Learning Objectives

1. Options for OI
2. Weight reduction
3. Oral agents (Clomiphene Citrate, Aromatase Inhibitors)
4. Gonadotrophin therapy
5. Laparoscopic ovarian diathermy
6. Metformin
 Metformin vs placebo or no treatment: Ovulation rates
Metformin Control Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
1.3.1 All patients
Baillargeon 2004 27 32 1 32 0.3% 167.40 [18.40, 1523.16]
Fleming 2002 37 45 30 47 11.5% 2.62 [0.99, 6.90]
Hoeger 2004a 4 9 3 9 3.7% 1.60 [0.24, 10.81]
Hoeger 2004b 3 9 6 11 7.9% 0.42 [0.07, 2.58]
Jakubowicz 2001 8 28 0 28 0.8% 23.63 [1.29, 433.02]
Lord 2006 9 22 9 22 11.7% 1.00 [0.30, 3.33]
Nestler 1996 5 11 1 13 1.1% 10.00 [0.94, 105.92]
Nestler 1998 12 35 1 26 1.7% 13.04 [1.57, 108.36]
Ng 2001 3 9 3 9 4.4% 1.00 [0.14, 7.10]
Onalan 2005a 17 153 20 150 39.6% 0.81 [0.41, 1.62]
Onalan 2005b 5 63 5 51 11.2% 0.79 [0.22, 2.91]
Sturrock 2002 0 12 1 14 3.0% 0.36 [0.01, 9.68]
Vandermolen 2001 1 12 1 15 1.8% 1.27 [0.07, 22.72]
Yarali 2002 6 16 1 16 1.4% 9.00 [0.94, 86.52]
Subtotal (95% CI) 456 443 100.0% 2.21 [1.57, 3.10]

All Total events 137

OR 2.21 95% CI 1.57, 3.10
82
Heterogeneity: Chi² = 40.58, df = 13 (P = 0.0001); I² = 68%
Test for overall effect: Z = 4.57 (P < 0.00001)

1.3.2 Patients with BMI < 30kg/m2

Baillargeon 2004 27 32 1 32 0.8% 167.40 [18.40, 1523.16]
Ng 2001 3 9 3 9 9.6% 1.00 [0.14, 7.10]
Onalan 2005a 17 153 20 150 86.6% 0.81 [0.41, 1.62]
Yarali 2002 6 16 1 16 3.0% 9.00 [0.94, 86.52]
Subtotal (95% CI) 210 207 100.0% 2.33 [1.43, 3.81]
Total events 53 25
Heterogeneity: Chi² = 25.45, df = 3 (P < 0.0001); I² = 88%

BMI < 30 OR 2.33 95% CI 1.43, 3.81

Test for overall effect: Z = 3.39 (P = 0.0007)

1.3.3 Patients with BMI > 30Kg/m2

Live birth rate: OR 1.00 95% CI 0.16, 6.39

Fleming 2002
Hoeger 2004a
37
4
45
9
30
3
47 21.2%
9 6.8%
2.62 [0.99, 6.90]
1.60 [0.24, 10.81]
Hoeger 2004b 3 9 6 11 14.6% 0.42 [0.07, 2.58]
Jakubowicz 2001 8 28 0 28 1.4% 23.63 [1.29, 433.02]
Lord 2006 9 22 9 22 21.6% 1.00 [0.30, 3.33]
Nestler 1996 5 11 1 13 2.0% 10.00 [0.94, 105.92]
Nestler 1998 12 35 1 26 3.1% 13.04 [1.57, 108.36]
Onalan 2005b 5 63 5 51 20.6% 0.79 [0.22, 2.91]
Sturrock 2002 0 12 1 14 5.4% 0.36 [0.01, 9.68]
Vandermolen 2001 1 12 1 15 3.3% 1.27 [0.07, 22.72]
Subtotal (95% CI) 246 236 100.0% 2.11 [1.31, 3.37]
Total events 84 57

BMI > 30 OR 2.11 95% CI 1.31, 3.37

Heterogeneity: Chi² = 15.35, df = 9 (P = 0.08); I² = 41%
Test for overall effect: Z = 3.09 (P = 0.002)

0.001 0.1 1 10 1000

Favours control Favours metformin
Metformin vs placebo or no treatment: Clinical pregnancy rates
Metformin Control Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
1.2.1 All patients
Nestler 1996 1 11 0 13 2.9% 3.86 [0.14, 104.65]
Yarali 2002 2 16 0 16 3.0% 5.69 [0.25, 128.50]
Kocak 2002 1 28 0 28 3.4% 3.11 [0.12, 79.64]
Fleming 2002 4 23 1 19 6.4% 3.79 [0.39, 37.20]
Lord 2006 3 22 2 22 12.3% 1.58 [0.24, 10.52]
Tang 2006 6 69 2 74 12.5% 3.43 [0.67, 17.60]
Ng 2001 1 9 2 9 12.6% 0.44 [0.03, 5.93]
Karimzadeh 2007 40 100 11 100 46.9% 5.39 [2.57, 11.34]
Subtotal (95% CI) 278 281 100.0% 3.84 [2.21, 6.68]
Total events 58 18

All Heterogeneity: Chi² = 4.41, df = 7 (P = 0.73); I² = 0%

Test for overall effect: Z = 4.76 (P < 0.00001)
OR 3.84 95% CI 2.21, 6.68
1.2.2 Patients with BMI < 30kg/m2
Yarali 2002 2 16 0 16 4.8% 5.69 [0.25, 128.50]
Ng 2001 1 9 2 9 20.2% 0.44 [0.03, 5.93]
Karimzadeh 2007 40 100 11 100 75.0% 5.39 [2.57, 11.34]
Subtotal (95% CI) 125 125 100.0% 4.41 [2.24, 8.66]
Total events 43 13
Heterogeneity: Chi² = 3.33, df = 2 (P = 0.19); I² = 40%
BMI < 30 Test for overall effect: Z = 4.31 (P < 0.0001) OR 4.42 95% CI 2.24, 8.66
1.2.3 Patients with BMI > 30kg/m2

Live birth rate: OR 1.00 95% CI 0.16, 6.39

Nestler 1996
Kocak 2002
1
1
11
28
0
0
13
28
7.7%
9.0%
3.86 [0.14, 104.65]
3.11 [0.12, 79.64]
Fleming 2002 4 23 1 19 17.2% 3.79 [0.39, 37.20]
Lord 2006 3 22 2 22 32.8% 1.58 [0.24, 10.52]
Tang 2006 6 69 2 74 33.4% 3.43 [0.67, 17.60]
Subtotal (95% CI) 153 156 100.0% 2.89 [1.09, 7.64]
Total events 15 5
Heterogeneity: Chi² = 0.52, df = 4 (P = 0.97); I² = 0%

BMI > 30 Test for overall effect: Z = 2.14 (P = 0.03)

OR 2.89 95% CI 1.09, 7.64
0.001 0.1 1 10 1000
Favours control Favours metformin

Metformin plus ovulation induction agent

vs ovulation induction agent alone: Clinical Pregnancy Rate
Treatment Control Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
3.2.1 All patients
El-Biely 2001 13 45 4 45 3.4% 4.16 [1.24, 14.00]
Hwu 2005 6 40 0 40 0.5% 15.26 [0.83, 280.72]
Khorram 2006 5 16 0 15 0.4% 14.83 [0.74, 295.97]
Kocak 2002 3 27 0 28 0.5% 8.14 [0.40, 165.53]
Legro 2007 80 209 62 209 45.2% 1.47 [0.98, 2.21]
Malkawi 2002 9 16 2 12 1.2% 6.43 [1.05, 39.33]
Moll 2006 57 111 64 114 36.3% 0.82 [0.49, 1.39]
Sahin 2004 5 11 3 10 2.0% 1.94 [0.32, 11.76]
Sturrock 2002 3 12 4 14 3.3% 0.83 [0.15, 4.78]
Vandermolen 2001 6 12 1 15 0.5% 14.00 [1.37, 142.89]
Zain 2008 8 41 7 41 6.7% 1.18 [0.38, 3.62]
Subtotal (95% CI) 540 543 100.0% 1.58 [1.20, 2.07]
Total events 195 147
Heterogeneity: Chi² = 20.60, df = 10 (P = 0.02); I² = 51%

All OR 1.58 95% CI 1.20, 2.07

Test for overall effect: Z = 3.31 (P = 0.0009)

3.2.2 Patients with BMI < 30kg/m2

Hwu 2005 6 40 0 40 1.2% 15.26 [0.83, 280.72]
Malkawi 2002 9 16 2 12 3.0% 6.43 [1.05, 39.33]
Moll 2006 57 111 64 114 90.7% 0.82 [0.49, 1.39]
Sahin 2004 5 11 3 10 5.1% 1.94 [0.32, 11.76]
Subtotal (95% CI) 178 176 100.0% 1.23 [0.78, 1.93]

BMI < 30 OR 1.23 95% CI 0.78, 1.93

Total events 77 69
Heterogeneity: Chi² = 8.54, df = 3 (P = 0.04); I² = 65%
Test for overall effect: Z = 0.88 (P = 0.38)

Live birth rate: OR 1.00 95% CI 0.16, 6.39

3.2.3 Patients with BMI > 30kg/m2
El-Biely 2001 13 45 4 45 5.6% 4.16 [1.24, 14.00]
Khorram 2006 7 16 1 15 1.1% 10.89 [1.14, 103.98]
Kocak 2002 3 27 0 28 0.8% 8.14 [0.40, 165.53]
Legro 2007 80 209 62 209 75.1% 1.47 [0.98, 2.21]
Sturrock 2002 3 12 4 14 5.4% 0.83 [0.15, 4.78]
Vandermolen 2001 6 12 1 15 0.9% 14.00 [1.37, 142.89]
Zain 2008 8 41 7 41 11.1% 1.18 [0.38, 3.62]
Subtotal (95% CI) 362 367 100.0% 1.83 [1.31, 2.56]
Total events 120 79
Heterogeneity: Chi² = 10.53, df = 6 (P = 0.10); I² = 43%

BMI > 30 OR 1.88 95% CI 1.31, 2.56

Test for overall effect: Z = 3.52 (P = 0.0004)

0.01 0.1 1 10 100

Favours control Favours treatment