Background Women following unilateral oophorectomy (UO)
are occasionally encountered during assisted reproduction
treatment.
Objective To explore the impact of UO on ovarian reserve in
assisted reproduction.
Search strategy An electronic database search was performed
using PubMed, EBSCO, ISI, Trip, ClinicalTrial.gov and the
Cochrane library followed by a manual search to identify
published research between January 1978 and December 2015.
Selection criteria Controlled studies that compared infertile
women following UO undergoing IVF-ET treatment with women
with two intact ovaries.
Data collection and analysis Two reviewers independently
extracted the data concerning the impact of UO on ovarian
reserve tests, ovarian response to controlled ovarian
hyperstimulation and clinical pregnancy rate. Meta-analysis was
performed using these measures.
Main results Twenty-one studies were eligible for quantitative
analysis. They included 1045 and 18 172 IVF cycles in women
with one and two intact ovaries, respectively. Basal FSH
weighted mean difference (WMD) was significant (2.01 IU/l;
95% CI: 0.24–3.79, P = 0.026). Similarly, the WMD of serum
E2 level on the day of hCG administration was significant
(WMD: 431 pg/ml; 95% CI: 616 to 246, P < 0.001).
However, the weighted overall odds ratio (OR) of clinical
pregnancy between women with a single ovary and women with
two ovaries was comparable (overall OR: 0.76; 95% CI: 0.57–
1.00, P = 0.054). All eligible studies were retrospectively
conducted and the heterogeneity among ovarian response
measures was high.
Conclusions Available pooled data supports an adverse effect of
UO on ovarian reserve involving the quantity but not the quality
of the ovarian pool.
Keywords Assisted reproduction, in vitro fertilisation, metaanalysis,
ovarian reserve, systematic review, unilateral
oophorectomy.

Introduction
During reproductive years, unilateral oophorectomy (UO)
could be an undesirable outcome of a gynaecological emergency
or a necessary intervention for a wide range of conditions,
including ovarian torsion, rupture, uncontrollable
bleeding, pregnancy, endometrioma, abscess and tumour.
However, the literature is limited on the effect of complete
unilateral oophorectomy on women’s fertility, or short and
long-term wellbeing. Infertile women following UO are
occasionally encountered in the assisted reproduction technology
(ART) setting when undergoing in vitro fertilisation
and embryo transfer (IVF-ET) treatment.
In a recent, well-designed cohort study investigating UO
in the general premenopausal population, the incidence of
this occurrence was 56–125 per 100 000 person years, with
temporal trends over the six decades of the study.1
Although the exact prevalence of single ovary in women
undergoing ART treatment is not known, there is evidence
to show that more ovarian surgeries are performed on women over 30 years of age compared with younger
women.1 It is therefore likely that the risk of undergoing
UO as the result of an ovarian emergency or disease in
women delaying their first child may be on the rise.
With women intentionally delaying childbirth2,3 and an
increasing demand for ART treatment,4,5 ovarian ageing
has become a dominant issue of infertility. As such, the
European Society of Hum Reproduction and Embryology
(ESHRE) published the Bologna criteria in 2011 seeking to
standardise the definition of poor ovarian response,6
although risk factors for poor response in young infertile
women were not explicitly defined.7 One potential risk
factor is a previous UO.7
The effect of UO on ovarian reserve and its association
to early reproductive ageing is not clearly defined. Reproductive
ageing is characterised by a gradual decline of the
quantity and quality of the oocyte-follicle pool. A fixed
temporal relationship, with large inter-individual variation,
is thought to be present among various reproductive events
including the most favourable fertility period, commencement
of sub-fertility, sterility, cycle irregularity and eventually
menopause.8 As such, several proxies of ovarian ageing,
representing different facets of ovarian reserve assessment,
could be involved. Consequently, UO may have an impact
on ovarian reserve testing, ovarian response to controlled
ovarian hyperstimulation (COH), IVF-ET outcomes, age at
natural menopause, and the development of premature
ovarian failure or early menopause.
Our aim in this systematic review and meta-analysis was
to explore the impact of UO during the reproductive years
on ovarian reserve testing, ovarian response to conventional
COH, IVF-ET outcomes and pregnancy rates, in an OR in vitro fertilisation OR in vitro fertilisation. We have combined these terms by AND term to complete the search.

Interpretation
The studies included in this systematic review were conducted
over three decades, and only three were carried out
in the last 10 years. One reason could be a more conservative
approach to treating ovarian pathology, or an ovariansparing
policy within gynaecological surgery. Indeed, today
any ovarian emergency or disease during the reproductive
years is seldom treated with total ovary removal. Indeed,
the vast majority of studies performed today focus on the
impact of ovarian cystectomy on ovarian reserve.52–54 Furthermore,
the release of the Women’s Health Initiative
study results in 2002 and the ACOG recommendation in
2008 to avoid elective salpingo-oohphorectomy,55,56 may
have contributed considerably to the reduction in the total
number of oophorectomies performed today in premenstrual
women.57,58
The results of our systematic review and meta-analysis
should be interpreted with caution. Ovarian reserve has
been shown adversely to affect the quantity but not the
quality of the ovarian follicular pool and this may be true
in the general infertile population. There is evidence to
show that UO may increase the risk of premature ovarian
failure and early menopause39,41–45 and decrease the age at
menopause;37–40 it may therefore have a negative impact
on women’s ovarian reserve and fertility. The same has
been shown in women undergoing bilateral ovarian surgery
for endometrioma resection.59 In light of these findings, it
may be detrimental for women with a priori low ovarian
reserve to undergo UO. The impact on their future fertility
may be significant. Ovarian reserve evaluation and fertility
preservation should be discussed before UO in cases when
time is not clinically critical.
Conversely, although UO crudely signifies acute reduction
of ovarian reserve by half, two high-quality studies
have shown that age at menopause is reduced only by
1.2–1.4 years, suggesting compensatory mechanisms for
ovarian ageing.39,40 Two recent reports in the ART setting
support this view, showing that the remaining ovary
appears to compensate in follicular yield following after
UO in women.31,32 The nature and magnitude of
reproductive ageing mechanisms of compensation, by
intra-, inter, or extra-ovarian signalling, are still to be
determined.32
Conclusion
Available data in the literature support the notion that UO
in women of the reproductive age adversely affects their
ovarian reserve. This is demonstrated in the ART setting,
with higher basal FSH level and inferior ovarian response
outcome measures in women following UO as compared
with controls. Nevertheless, available data do not support
an adverse effect on fertilisation rate and clinical pregnancy
rate, suggesting that the reduced ovarian reserve involves
the quantity but not the quality of ovarian pool. Our
results should interpreted with caution as they rely entirely
on retrospective studies with high heterogeneity in ovarian
response outcome measures. The role of compensatory
mechanisms of the remaining ovary to reproductive ageing
should be further investigated; future research should also
assess AFC and AMH, the more reliable biomarkers of
ovarian reserve.