Study design
This single centre based retrospective study was conducted at Aster Medcity Kochi, a 670 bedded NABH accredited quaternary care centre in Kerala, India, from the period of 2015 to 2021.
All obstetric and gynecological cases in which interventional radiology techniques were used during the period of January 2015 to December 2021 were included. None of the cases were excluded. For all elective cases, a multidisciplinary team meeting consisting of Obstetrician and Gynecologist, Interventional Radiologist and Anesthesiologist was conducted prior to the procedure. Informed consent was taken from every patient prior to the procedure.
The interventional radiology procedures were performed in ‘PHILIPS Hybrid cath lab’ by two interventional radiologists, with one of them having a minimum five years of experience. All patients with arteriovenous malformations (AVM), underwent a contrast CT angiography for assessing the feeding the vessels and draining veins. Patients with Placenta Accreta Spectrum (PAS) disorders were subjected to a non-contrast MRI of abdomen in the late third trimester for maximum possible characterization of abnormal placentation. Nine women of the PAS disorder underwent prophylactic balloon placement in bilateral internal iliac arteries under local anesthesia. Common femoral arteries (CFA)were used as access sites and through 6F sheath, 5mm/6mm of 40 mm length balloons were placed in internal iliac arteries distal to common iliac bifurcation. The sheaths and catheters were anchored to the thigh with sutures and ‘Tegaderm’. Patients were shifted to the operation theatre (OT) for anesthesia and surgery. The interventional radiologists were called to the OT for inflating balloons immediately after the delivery of baby. The balloons were kept inflated to their optimum pressures till the placental extraction/hysterectomy was completed. Then the balloons and sheaths were removed in the theatre and hemostasis achieved by manual compression or using closure devices.
For uterine fibroid embolization (UFE), all patients underwent non-contrast MRI as a baseline imaging for future comparison. Right femoral artery access was chosen for access in all patients; in rare circumstances, left femoral access was also taken. The anterior division of internal iliac arteries was cannulated using 4 or 5F catheters and microcatheter was used for selective access into the uterine artery distal to the branches to cervix. After achieving stable position, the arteries were embolized using PVA particles ranging from 250 to 500micron sizes without causing significant reflux of particles. Good stasis of contrast material in the proximal uterine arteries was taken as the endpoint and confirmed adequate embolization of arteries.
Pelvic AVM cases were rare, and often difficult to differentiate between retained products of conception with significant vascularity in the setting recent pregnancy with occasional coexistence. So, all patients were extensively evaluated and followed up. Recent contrast CT angiogram was taken to identify the feeding arteries, size of the nidus and draining veins. Access was planned based on the CT findings. Arterial route embolization was preferred often, however one of our cases needed both arterial and venous route embolization.
Patients were followed up for 72 hours as in patients for any recurrence of symptoms and complications. Continuous monitoring of vitals (heart rate, blood pressure ,temperature, saturation ) was done. Clinical success was defined in terms of recurrence and or need for further surgery. Both obstetric and gynaecology cases were followed up at the time of postsurgical review as per protocol.
Statistical analysis: No statistical analysis of data was required as it was a single centre retrospective analysis of cases.
Results:
We had a total of 35 cases, both obstetrics and gynecological, who underwent various interventional radiological procedures including embolization of uterine artery, peripheral angiography and embolization, internal iliac artery balloon placement and ovarian vein embolization and coil insertion.
Observations have been outlined below in Table-1 and Table-2, for gynaecologic and obstetric cases respectively.
Our case series included nine cases of placenta accreta spectrum (PAS)disorders, four each of fibroid complicating pregnancy and postpartum hemorrhage and the latter underwent obstetric or cesarean hysterectomy for primary or secondary hemorrhage, two with secondary hemorrhage following hysterectomy for gynecological indications, ten cases of Fibroid and six cases of arteriovenous malformations (AVM). The AVM cases included one uterine AVM and five cases of pelvic AVM. One patient of pelvic AVM had rectal ischemia post embolisation and underwent pelvic exenteration surgery. In another case of pelvic congestion syndrome planned for bilateral ovarian vein embolisation, only left side embolisation could be performed due to difficult anatomy.
We now highlight the clinical scenarios in our hospital where interventional radiology and obstetrics and gynaecology collaborated for the management of the patients.
Postpartum hemorrhage :
A 32 year old G3P1L1A1 at 36+ weeks gestation with 2 previous LSCS referred with MRI suggestive of Placenta previa with accreta/increta with placenta in the left and inferior wall of uterus, completely covering internal os. There was no evidence of extra uterine extension of placenta, bladder infiltration or bladder tenting/ureteric compression/hydronephrosis.
An elective Caesarean hysterectomy with classical uterine incision with bilateral uterine artery occlusion under general anaesthesia intraoperatively showed placenta completely covering the lower segment, adherent-partially separated. Adnexa were normal. Hysterectomy with placenta insitu was performed. Estimated blood loss was around 500 ml. (Figure1).
A 37 year old referred for postpartum haemorrhage following an emergency LSCS was taken up for laparotomy and exploration with a preoperative emergency uterine artery embolization by interventional radiologist. Intraoperative findings were suggestive of a longitudinal lower posterior uterine wall rupture of 5x1 cm which was repaired. B lynch sutures were applied for atonicity along with other measures to control the haemorrhage. Post operative collapse in ICU led to relaparotomy with total abdominal hysterectomy which showed Couvelaire uterus of 24 weeks size and hemoperitoneum of 1.3-1.5 litres managed by component replacement.