Keywords
ovarian torsion - pelvic pain - peritoneal defect - surgery
Introduction
Surgical evaluation for presentation of acute pelvic pain is often necessary to explore
its etiology as well as its risk factors in females of reproductive age.[1] Laparoscopy plays a role in diagnosis and treatment in emergency gynecology cases
as it confirms ovarian torsion and could show the possible cause and risk of twisted
ovary in women.[2] Some underlying causes are rare and unexpected, such as a peritoneal hernia or a
defect in which the ovary and adnexa can fall and become incarcerated along with the
torsion.[1]
Some of the congenital causes are Mullerian duct anomalies such as Mayer–Rokitansky–Kuster–Hauser
(MRKH) syndrome. In patient diagnosed with MRKH syndrome, congenital absence of the
upper two-thirds of the vagina, the cervix, and hypoplasia, or complete aplasia of
the uterus.[3] In such patients, the ovaries lack the support of the utero-ovarian ligament, and
are only supported by the infundibulopelvic ligament, making the ovaries highly mobile
and more suspectable to ovarian torsion.[4] Due to the absence of the uterus, the ovaries are only supported by the infundibulopelvic
ligament, which provides little fixation to the pelvic side wall, making ovarian torsion
more common in such cases. Ovarian torsion is more common in patients with anatomical
variation, including those with ovarian cyst or anatomical agenesis, such as patients
with MRKH in which the torsion might include the fallopian tubes with or without the
ovary.[5]
In general, the presence of incarceration with ovarian torsion plays a role in enhancing
the ischemic changes in ovarian tissue by reducing the blood supply and leading to
venous congestion, ischemia, and eventually necrosis, making prompt diagnosis crucial.
Incarcerated adnexa is rare and there are only two cases reported to date.[1]
[2] We present a case of an incarcerated ovary combined with ovarian torsion and ischemic
change. This case was followed up by a laparoscopic second-look re-exploration procedure
and management of the identified peritoneal defect.
Case Report
A 27-year-old, single, nulliparous female, not known to have any medical conditions,
with a surgical history of a gastric sleeve, presented to the emergency room with
severe right lower quadrant pain for 1 day. The pain was associated with nausea, however,
without any episodes of vomiting. The patient denies any fevers, abnormal vaginal
discharges, dysuria, or changes in bowel habits. Her last menstrual cycle was 2 weeks
prior to presentation. She had no history of chronic pelvic pain or secondary dysmenorrhea.
Upon examination, the patient was visibly in distress and crying out in pain. Laboratory
tests showed a negative urine pregnancy test, a white blood cell count of 6.34 × 109 cells/L, with 71% neutrophils, a hemoglobin level of 106 g/L, and a platelet count
of 284 × 109 cells/L. The lactic acid level was 2.32. A formal gynecological transabdominal ultrasound
was performed, which showed an enlarged right ovary in size with absence of blood
flow.
The patient was consented and taken to the operating room and a diagnostic laparoscopy
was started. Upon visualization, the right ovary and the fallopian tube were noted
to be torsed four times ([Fig. 1]). The right ovary was found to be impacted inside a pelvic peritoneal pouch above
the right uterosacral ligament ([Fig. 2]). The ovary was noted to be enlarged and bluish in color. We started by removing
the ovary from the pouch gently using atraumatic graspers, gradually pulling the ovary,
which was very fragile ([Fig. 3]). Once the ovary was completely out, we began detorsion of the right adnexa, ovary,
and tube. The inspected peritoneal pouch was found to be 8 × 6 cm, rounded in shape,
with visible uterine vessels inside in the medial pouch wall ([Figs. 4] and [5]). There were no other visible structures or any abnormalities like an endometriotic
patch. The left adnexa tube and ovary were inspected and appeared normal with no peritoneal
defect in the left side. After inspecting the left adnexa, our gaze was shifted back
to the right ovary, which began to regain its normal color after few minutes of detorsion.
No ovariopexy was performed. We ended the procedure with a plan for a second laparoscopic
look for closure of the identified defect as the pouch tissue was edematous and fragile
at the time and would make the suturing difficult. The patient tolerated the surgery,
and she was discharged in stable condition.
Fig. 1 Right adnexa torsion with ovarian ischemic changes.
Fig. 2 Right ovary incarcerated inside the peritoneal pouch.
Fig. 3 Right ovary completely extracted.
Fig. 4 Appearance of uterine vessel's peritoneal pouch.
Fig. 5 Pelvic retroperitoneal pouch.
Four weeks post-op, the patient was taken for a second look laparoscopic surgery for
a better closure of the identified defect. Intraoperatively, both ovaries and tubes
looked normal. The right tube and ovary had regained their normal structure, size,
and color ([Fig. 6]). The defect seen previously in the right side of the posterior pelvic peritoneum
was smaller than before, measuring around 5 × 6 cm. The defect was closed by approximating
the edges and suturing using an interpreted 0 Vicryl stitch interrupted after refreshing
the edges around. The closure was started from medial to the lateral, 1 to 2 cm left
over as opening in the lateral side. The procedure ended with no complications. The
patient tolerated the second surgery and was discharged from the hospital on post-op
day 1 with no complications. The patient was then seen in the clinic 3 weeks postoperation
for a follow-up, which was reassuring. The patient claimed to have no more pain and
that her menstrual cycle had since resumed regularly.
Fig. 6 Closing peritoneal pouch with interrupted suture.
Discussion
The diagnosis of ovarian torsion can be challenging due to its vague symptoms upon
presentation. Abdominal pain is usually reported to be the most common symptoms followed
by nausea and vomiting, which present in 85 to 90% of the cases.[6]
[7] Ultrasound is the first imaging modality that is used for differential diagnosis
in patients with acute abdominal or pelvic pain. B-mode ultrasonography is used to
assess the ovarian morphology while color Doppler ultrasonography (CDU) is used to
evaluate the blood flow to the ovaries.[8] It has been proposed that the decrease of blood flow to the ovaries is highly suggestive
of ovarian torsion. However, recent studies have argued about the reliability of color
Doppler in the diagnosis of ovarian torsion due to its low sensitivity. It was reported
in a study that the CDU sensitivity in the diagnosis of ovarian torsion reaches 43%
with a specificity of 91.7%, while the negative predictive value reached 71% and the
positive predictive value reached a 78 to 100%.[9]
[10] Such findings have implied that it is not recommended to rule out the diagnosis
of ovarian torsion based on normal CDU. Furthermore, the American College of Obstetricians
and Gynecologists suggested not to use CDU for the diagnosis of ovarian torsion and
rely more on the clinical judgment.[11] The pathophysiology of ovarian torsion is still not clear. A recent published article
by Carugno stated that an association was noted between leukocytosis and absent blood
flow in patients with ovarian torsion. It is believed that the etiology behind having
a high inflammatory marker is a response to necrosis of the tissue during the ovarian
torsion. Therefore, it is hypothesized that the presence of no blood flow and high
inflammatory markers is an indicator that ischemia started to occur and ovarian torsion
is more.[12]
Torsion affects the right side of the body more frequently than the left, with a 3:2
ratio, as explained earlier.[13] The first step during surgical approach is to detorse the adnexa after the laparoscopic
ports have been inserted. Untwisting the torsed adnexa was originally thought to cause
vascular emboli, therefore most torsion was treated by removing the adnexa without
untwisting it.[14] However, this has been proven to be false.[15] Even if the ovary seems blue and dusky at first glance, after only 6 weeks, most
ovaries, ∼90%, show normal follicular development on ultrasound, normal Doppler flow,
and normal gross appearance.[16]
Pelvic peritoneal defect is a rare occurrence, and can be congenital or iatrogenic.[2] Dysfunctional embryonic uterine development is believed to be the cause of congenital
peritoneal defect. It is believed that during embryonic development, the remnant of
the Mesonephric or Mullerian duct forms a cyst in the broad ligament which later can
rupture and lead to a defect in the peritoneum. Iatrogenic pelvic peritoneal defects
are typically seen in endometriosis, prior surgery, intraperitoneal inflammation,
obstetric or abdominal trauma, and iatrogenic operations as the most common predisposing
factors of peritoneal defect.[17] Another etiology is Allen–Masters syndrome, a syndrome which is associated with
laceration to the uterine-supportive ligaments following traumatic delivery.[18]
This differential makes the defect in this case more likely to be congenital in nature
due to the absence of a history of pelvic surgery, nulliparity, the absence of prior
sexual intercourse, and any history of sexually transmitted diseases, along with the
lack of signs of endometriosis. Together, these findings suggest that our patient's
peritoneal defect was likely congenital. However, the etiology of peritoneal defect
is idiopathic. In 0.5% of regular biopsy specimens, congenital defects were discovered.[19] Ovarian incarceration into a pelvic peritoneal defect is a rare phenomenon, and
there are only few published case reports to guide management of the defect.[1]
Literature review of the pelvic peritoneal defect with incarcerated ovary showed two
cases published describing ovarian incarceration into a posterior pelvic peritoneal
defect like our case, by Kataoka et al in 2009 and by Jackson et al in 2015.[1]
[2] This is a third case describing an incarcerated ovary. The previous two published
cases were on the right adnexa pelvic side with similar clinical presentations and
normal ovarian tissue on laparoscopy with no evidence of ischemic changes such as
abnormal discoloration or evidence of ovarian torsion. Our case was associated with
ovarian torsion four times, and there were ischemic changes of the affected ovary
which made the removal of the incarcerated ovary from the defect more challenging
considering the fragility of the ovarian tissue.
In this case, we chose to close the peritoneal defect compared with the two published
studies, which left the defect with no suturing. Surgical repair of the peritoneal
defect is still controversial, with literature review revealing lack of consensus.[1]
[20]
[21] We closed the defect because we believed it may prevent the recurrence of an incarcerated
ovary that would otherwise increase the risk of ovarian ischemia with ovarian torsion.
On the other hand, most reports of a peritoneal defect support not closing the peritoneum
to prevent peritoneal tension and the risk of vascular injury.[1] Other methods for closing peritoneal defects include using an omental flap, mesh,
or a space-occupying compound.[5]
[22] More cases are required to be collected and reported to support or go against the
defect closure, weighing the benefits and risks of such intervention and the risk
of re-occurrence.
Conclusion
In conclusion, pelvic peritoneal defect is considered to be a rare phenomenon. More
case reports and long-term sequels are required to identify whether the defect is
associated with enhancing the ovarian ischemia with torsion. Closure of the peritoneal
defect is still controversial and requires more studies for an expert consensus on
the best approach to those findings.