Key words opportunistic salpingectomy - prophylactic salpingectomy - ovarian cancer - p53 signature
- laparoscopy - hysterectomy
Schlüsselwörter opportunistische Salpingektomie - prophylaktische Salpingektomie - Eierstockkrebs
- p53-Signatur - Laparoskopie - Hysterektomie
Introduction
The recent discovery and its supporting arguments that precancerous lesions of the
fallopian tube precede high trade serous cancer of the ovary have given rise to change
in paradigm [1 ]. Indeed, recent histopathological analyses of prophylactic salpingo-oophorectomies
for genetic risk have revealed new occult precancerous lesions named Serous Tubal
Intraepithelial Carcinoma or STIC. These lesions consist of nonciliated cells and
are defined by the following features: epithelial stratification, nuclear atypias
with an increase in the nucleocytoplasmic ratio, loss of nuclear polarity, nuclear
pleiomorphism, loss of ciliated cells, intense and diffuse immunohistochemical expression
of p53. Other earlier tubal lesions have also been described: the p53 signature, defined
by a succession of at least 12 secretory cells with intense nuclear p53 staining.
Finally, an ultra earlier lesion termed SCOUT (Secretory Cell Outgrowths), characterised
by a succession of at least 30 pseudostratified secretory epithelial cells with a
low expression of PAX2, PTEN and Ki67, and no p53 mutation has also been revealed.
It is now presumed that the SCOUT lesions could develop into p53 signature and then
into STIC. The STIC could metastasize in the ovary and peritoneum. Several series
of non-genetic sporadic high-grade serous ovarian cancers were re-analysed and revealed
the presence of the same tubal carcinogenic sequence. All these arguments would be
in favour of the tubal origin of ovarian carcinogenesis and could lead to preventive
surgical strategies as exclusive prophylactic salpingectomy without ovariectomy [1 ], [2 ].
Taking advantage of benign hysterectomy to offer prophylactic salpingectomy with conservation
of the ovary would on the one hand avoid provoking menopause and on the other could
be an efficient means of preventing ovarian cancer [2 ].
In a very recent meta-analysis specifically addressing the protective effects of prophylactic
salpingectomy in the general population, the risk of ovarian cancer was significantly
reduced, with an OR = 0.51 (95% CI 0.35 – 0.75) [3 ]; however, these results must be interpreted with some prudence: the meta-analysis
included only 3 studies and did not take the various histologic types of ovarian cancer
into account [3 ]. Moreover, while hysterectomy alone or salpingectomy alone are both related with
a reduced incidence of ovarian cancer, the degree of protection obtained by associating
hysterectomy and salpingectomy was not assessed. In terms of ovarian cancer prevention,
would opportunistic salpingectomy with ovarian conservation during perimenopausal
benign hysterectomy result in any advantages [4 ], [5 ]?
In all Western countries we are witnessing an exponential increase in opportunistic
salpingectomy during benign hysterectomy: exclusive salpingectomy without ovariectomy
is recommended by several societies in Austria (the Austrian Society of Obstetricians
and Gynaecologists OEGGG) [6 ], USA (the Society of Gynecologic Oncology and the American Congress of Obstetricians
and Gynecologists) [7 ], [8 ], Canada (the Society of Gynecologic Oncology of Canada and the Ovarian Cancer Research
Program of British Columbia) [9 ], [10 ], United Kingdom (the Royal College of Obstetricians and Gynaecologists RCOG) [11 ], Australia/New Zealand (Royal Australian and New Zealand College of Obstetrics and
Gynaecology RANZCOG) [12 ] and in Germany (the German Kommission Ovar of the Arbeitsgemeinschaft Gynäkologische
Onkologie AGO) [13 ].
We propose here to assess this new practice, focusing on the prevalence of tubal abnormalities
(STIC or p53 lesion) and the prevalence of peri- and postoperative complications related
with the salpingectomy procedure during laparoscopic hysterectomy.
Materials and Methods
Study design
We carried out a continuous prospective study between January 1, 2014 and December
31, 2015 at the Hôpital Femme Mère Enfant in Lyon, France. Bearing in mind the new
tubal hypothesis of ovarian carcinogenesis [2 ], our surgical protocol for benign laparoscopic hysterectomy was modified to add
systematic bilateral salpingectomy (or unilateral in case of history of previous salpingectomy).
In the event of subtotal hysterectomy, the tubes were never morcellated and were always
removed inside a bag via the suprapubic trocar immediately after salpingectomy. After
prior informed consent from the patient with ethics board approval, the decision to
carry out complementary ovariectomy was made by the surgeon depending on her age (postmenopausal
status) and/or the existence of ovarian abnormalities found preoperatively by ultrasound,
or discovered macroscopically during surgery. The study was approved by the local
ethics committee.
The exclusion criteria were:
Hysterectomy in a context of gynecologic cancer
Hysterectomy not carried out using laparoscopy
Fimbriated end missing or damaged or impossible for pathologist to analyze
No consent obtained from the patient, salpingectomy refused, inability to understand
the information given
Surgical technique
The salpingectomy took place at the beginning of surgery, by section of the tuboovarian
ligament, then section of the mesosalpinx closer to the fallopian tube than the ovary,
and finally section of the fallopian tube from the uterus closer to the uterine horn.
Hemostasis was performed by electrosurgical bipolar vessel sealing.
Histological analysis
Histopathological analysis of the tubes was carried out by two experienced pathologists,
using a specific and validated protocol, the Sectioning and Extensively Examining
of the Fimbriated end protocol (SEE-FIM protocol) [1 ]. Briefly, the tube was embedded completely in buffered formalin, then the fimbriated
end and infundibulum were sampled using longitudinal sections 2 to 3 mm thick, with
cross sections for the remainder of the tube. Each sample taken was embedded in paraffin;
a 3 micron spread was stained by HPS (hematoxylin, phloxine, saffron) for morphologic
analysis. There was also a systematic immuno-histochemical examination with amplification
(Ultraview, Ventana). The slides were first deparaffinized; after heat-induced antigen
retrieval using EDTA pH 9, anti-p53 antibody (Dako France, clone DO-7, dilution 1/400)
was applied and the reaction revealed with DAB (diaminobenzydine).
The definition of a STIC (serous tubal intraepithelial carcinoma) lesion was based
on the following criteria [1 ]: epithelial stratification, nuclear atypias with an increase in the nucleocytoplasmic
ratio, loss of nuclear polarity, nuclear pleiomorphism, loss of ciliated cells, intense
and diffuse immunohistochemical expression of p53. The p53 signature was defined by
at least 12 consecutive secretory cells that were p53-positive by immunohistochemistry
[1 ].
A standard protocol was used for histopathologic analysis of the ovariectomy and hysterectomy
samples.
Data collections
We collected the following data for each patient: age, BMI, gestational, parity and
menopause status, personal and/or family past history of gynecologic cancer, surgical
indication, type of surgery, peri- and postoperative complications, uterine (cervix,
myometrium, endometrium), ovarian and tube histology. The duration of salpingectomy
was timed and postsalpingectomy blood loss was measured.
Statistical analysis
The main criterion assessed was prevalence of histopathologic tube anomalies (STIC
or p53 lesion). The secondary criterion assessed was prevalence of peri- and postoperative
complications related with achievement of the salpingectomy. In this descriptive study,
the results are expressed in terms of percentage of the total number of patients.
Results
One hundred patients were included in this study.
Demographic data
The mean age of the population was 49.3 years old with no personal or family past
history of gynecologic cancer. The type of laparoscopic hysterectomies was total in
72% and subtotal in 28%. It should be noted that laparoscopic salpingectomy was always
possible without any peri- or postoperative complication attibutable to the salpingectomy
itself. The mean duration of salpingectomy was 428 seconds (354 – 596) and the blood
loss during salpingectomy was 9 cm3 (2 – 15). The complications attributable to the hysterectomy were rare: bleeding
at the uerine pedicle (1%), ureteral injury (1%), vaginal cuff bleeding (2%) and urinary
tract infection (2%) ([Table 1 ]).
Table 1 Demographic data, per and postoperative (%). N = 100. The mean age of the population
was 49.3 years. The type of laparoscopic hysterectomies was total in 72% and subtotal
in 28%. Laparoscopic salpingectomy was always possible without any peri- or postoperative
complication attibutable to the salpingectomy itself. The mean duration of salpingectomy
was 428 seconds (354 – 596) and the blood loss during salpingectomy was 9 cm3 (2 – 15).
Age (years)
49.3
BMI (kg/m2)
26.3
Gestational status
2.9
Parity
2.35
Personal or family past history of gynecologic cancer
0
Type of laparoscopic hysterectomy
28 subtotal hysterectomies (28%)
72 total hysterectomies (72%)
Preoperative complications during salpingectomy
0
Postoperative complications attributable to salpingectomy
0
Postoperative complications attributable to the hysterectomy
6 (6%)
1 (1%)
1 (1%)
2 (2%)
2 (2%)
Duration of salpingectomy (seconds)
428 (354 – 596)
Postsalpingectomy blood loss (cm3 )
9 (2 – 15)
Clinical indications for surgery
Surgical indications were mainly related to symptomatic leiomyomas in 50% followed
by menorrhagia in 27% of the population, uterine prolapse in 11% of the cases, pelvic
pain, clinically benign ovarian cyst and cervival dysplasia in respectively 5, 4 and
3% of the cohort. None of the patients had a history of gynecologic cancer ([Table 2 ]).
Table 2 Clinical indications for surgery (%). Surgical indications were mainly related to
symptomatic leiomyomas in 50% followed by menorrhagia in 27% of the population. None
of the patients had a history of gynecologic cancer (n = 100).
Symptomatic leiomyomas
50 (50%)
Menorrhagia
27 (27%)
Uterine prolapse
11 (11%)
Pelvic pain
5 (5%)
Clinically benign ovarian cyst
4 (4%)
Cervical dysplasia
3 (3%)
Pathologic diagnoses in uterus, ovary and fallopian tubes
Histopathological analysis was always benign. Benign leiomyomas were found in 50%
of the population. In the other cases, histopathological results showed adenomyosis
(30%), benign mucosal polyps (10%), benign endometrial hyperplasia (1%) and chronic
endometritis (1%). Ovariectomy was performed in 36% of cases: histopathological assessment
was always benign with functional cysts (19.4%), benign dermoid cyst (2.7%), serous
cystadenoma (2.7%) and ovarian adenofibroma (2.7%) ([Table 3 ]).
Table 3 Main pathologic diagnosis in uterus (n = 100) and ovary (n = 36). Histopathological
analysis was always benign.
In uterus
100
30 (30%)
50 (50%)
10 (10%)
1 (1%)
1 (1%)
In ovary
36
26 (72.2%)
7 (19.4%)
1 (2.7%)
1 (2.7%)
1 (2.7%)
During hysterectomy for benign indications, 99 bilateral salpingectomies and one unilateral
salpingectomy (past history of salpingectomy for ectopic pregnancy) were carried out
by laparoscopy. So the histopathologic analysis concerned 199 tubes.
There were 11 tubes with the p53 signature ([Figs. 1 ] and [2 ]) in 9 patients, indicating a prevalence of 5.52% (11/199):
Fig. 1 p53 signature, hematoxylin phloxine saffron, original magnification × 400.
Fig. 2 p53 signature, immunohistochemistry, original magnification × 400.
7 (3.57%) unilateral p53 signatures,
2 (1%) bilateral p53 signatures.
No STIC and no associated cancer were observed.
The other associated pathological findings were: paratubal cysts (2.5%), tubal papilloma
(0.5%), hydrosalpinx (0.5%), tubal endometriosis (0.5%) and paratubal hemangioma (0.5%)
([Table 4 ]).
Table 4 Main pathologic diagnosis in fallopian tubes (n = 199). There were 11 tubes with
the p53 signature in 9 patients, indicating a prevalence of 5.52% (11/199) with the
following distribution: 7 (3.57%) unilateral p53 signatures and 2 (1%) bilateral p53
signatures. No STIC and no associated cancer were observed.
STIC
0
p53 signatures
11 (5.52%)
2 (1%)
7 (3.57%)
No pathologic abnormality
179 (89.9%)
Benign paratubal cysts
5 (2.5%)
Benign tubal papilloma
1 (0.5%)
Hydrosalpinx
1 (0.5%)
Benign tubal endometriosis
1 (0.5%)
Benign paratubal hemangioma
1 (0.5%)
Discussion
With the recent discovery of a tubal origin for high-grade serous cancers of the ovary
not only in the population at genetic risk (BRCA mutation) but also the general population
[1 ], [2 ], the concept of so-called opportunistic prophylactic salpingectomy has spread quickly
and is carried out increasingly during hysterectomy for a benign indication during
the perimenopause [2 ], [4 ], [14 ]. Up until the menopause it is essential to conserve the ovaries in order to avoid
cardiovascular and osteoporosis morbidity and mortality [15 ]. Certain mathematical models even suggest conserving the ovaries beyond physiological
menopause (up until age 65) due to the residual hormonal secretions [16 ]. Indeed the ovary produces androgens (androstenedione and testosterone) significantly
with an aromatization in fat tissue into estrone after the menopause [17 ]. Inversely conservation of the tubes during hysterectomy does not appear to present
any advantage: associated salpingectomy reduces postoperative infectious morbidity
in both univariate (OR = 7.2; 95% CI 1.6 – 32.1) and multivariate analysis (OR = 4.9;
95% CI 1.1 – 22.9) [18 ]. Conservation of the tubes during hysterectomy could also result in nearly 35.5%
hydrosalpinx and thus possibly repeat surgery [19 ].
If opportunistic salpingectomy is to be recommended, three questions must first be
raised:
Is salpingectomy always technically feasible? What is the morbidity?
Does salpingectomy during hysterectomy result in ovarian insufficiency or even premature
menopause? In other words, what impact does salpingectomy have on the ovary?
Is opportunistic salpingectomy an efficient means of preventing ovarian cancer in
the general population (that is, in the absence of any associated genetic risk)?
Concerning the first question, all depends on the approach used. Laparoscopy does
not appear to involve any particular difficulties: in our series, it was possible
for all the patients to undergo salpingectomy, which is confirmed by other studies
[18 ], [20 ], [21 ]. Perioperative complications can be found in each series, including ours, but there
are no complications at all that are attributable exclusively to salpingectomy. All
the complications are related with the main indication for surgery, i.e. the hysterectomy
[18 ], [20 ], [21 ].
However, a vaginal approach to the tubes may be more difficult and the success rate
for salpingectomy varies between 73.9 and 88% for experienced teams [22 ], [23 ]. The complication rate attributable to the salpingectomy via the vaginal route could
be estimated at up to 3.8% [23 ].
Concerning the second question, the salpingectomy surgical procedure must not affect
ovarian vascularization. First the surgeon must coagulate and section the tuboovarian
ligament without coagulating the infundibulopelvic ligament. Then the mesosalpinx
must be coagulated and sectioned, closer to the fallopian tube (in order to protect
the ovarian artery) [17 ]. For other authors, the surgeon should perform a wide excision of soft tissues adjacent
to the ovary and fallopian tube because the mesosalpinx structure could give rise
to carcinoma [24 ]. Finally, the surgeon has to section and coagulate the tube level with the uterine
horn [17 ]. The intrauterine portion of the fallopian tube should be removed because STIC lesions
have been described in this location despite previous salpingectomy [25 ]. In another study about 522 women at low risk of ovarian cancer and from whom systematic
prophylactic salpingectomy was performed during benign surgery, 4 STICs were identified:
3 STICs were located in the fimbriae whereas one STIC was in the nonfimbriated portion
of the tube [26 ]. Several studies thus have analyzed the impact of salpingectomy on the ovarian reserve
by comparing hormone levels (in particular the anti-mullerian hormone) and Doppler
flow data in the ovary before and after salpingectomy [24 ], [27 ], [28 ]: a recent meta-analysis based on 13 studies concluded that “the salpingectomy does
not appear to affect ovarian function. It may, however, impair the ovarian reserve
in the long run” [29 ]. Despite there is no data in the literature concerning when menopause commences
after salpingectomy, it seems that the addition of opportunistic salpingectomy during
laparoscopic hysterectomy does not cause any long-term effects (3 to 5 years later)
on ovarian function [30 ].
Finally, with respect to the last question, several precancerous lesions have been
identified in the tube: in the carcinogenic sequence, the p53 signatures come before
those of STIC lesions. The latter then metastasize to the ovary and neighboring peritoneum
[1 ], [2 ]. However we still do not know which p53 lesions will evolve towards STIC and which
p53 lesions will never evolve or may even disappear. Various molecular biology techniques
such as next generation sequencing have been able to discriminate between different
STIC lesions, some of which gave rise to ovarian cancer while others were micrometastases
of uterine endometrioid carcinoma [31 ]. It is probable that this kind of molecular technology will be able to discriminate
between different p53 signatures.
Our results also raise the question of how to share information with the patients
(possibly causing anxiety [32 ]) and of monitoring, if any: the clinical significance of STIC in the general population
with low risk is unknown. Only 10% of STIC diagnosed in the high risk population with
BRCA mutation could subsequently develop into ovarian carcinoma. We could therefore
suppose that this prevalence may be lower in the general population, and this has
been demonstrated in a previous study with 522 patients at low risk and only 4 STICs
(4/522 = 0.76%) [26 ]. The risk of p53 lesions evolving could therefore be very low.
On the other hand, it is important to consider the evolutionary timeline of ovarian
cancer development: a very recent molecular study about whole-exome sequence and copy
number analyses of laser capture microdissected precancerous tubal lesions has demonstrated
that p53 signatures and STICs are precursors of ovarian cancers (clonal relationship
between them with the same molecular mutations) and that the timing of the progression
from these precancerous lesions to invasive ovarian cancers was on average 6.5 years
[33 ]. Until now, the histopathological analysis of fallopian tube in the general population
(i.e. no genetic mutation) was only a macroscopic examination with a single representative
tissue section per tube. We and others really think that there should be a change
in tubal tissue histopathological sampling protocol not only for the genetic population
with BRCA mutations but also in the general population when an opportunistic salpingectomy
is performed [26 ].
To conclude, there is practically no doubt that laparoscopic salpingectomy is both
feasible and innocuous. Whether it is useful in terms of preventing ovarian cancer
when combined with benign hysterectomy in the general population has not yet been
established [1 ], [2 ]. The only way to find the answer to this tricky question is to set up registers
and long-term prospective studies (as is the case in the province of British Columbia
in Canada) [10 ].
Funding source: None.
