Abstract
Background
Transcatheter arterial embolization (TAE) is not common for hemorrhagic complications after gynecologic hysterectomy.
Purpose
To evaluate the effectiveness and safety of TAE for hemorrhage after hysterectomy for gynecologic diseases.
Material and Methods
This is a retrospective, multicenter study, which investigated 11 patients (median age = 45 years) who underwent TAE for hemorrhage after gynecologic hysterectomy between 2004 and 2020.
Results
The median interval between surgery and angiography was one day (range = 0–82 days). Hemodynamic instability and massive transfusion were present in 6 (54.5%) and 4 (36.4%) patients, respectively. CT scans (n = 7) showed contrast extravasation (n = 5), pseudoaneurysm (n = 1), or both (n = 1). On angiography, the bleeding arteries were the anterior division branches of the internal iliac artery (IIA) (n = 6), posterior division branch (lateral sacral artery, n = 1), and inferior epigastric artery (n = 1) in eight patients with active bleeding. In the remaining three patients, angiographic staining without active bleeding foci was observed at the vaginal stump, and the feeders for staining were all anterior division branches of the IIA. Technical and clinical success rates were 100% and 90.9% (10/11), respectively. In one patient, active bleeding focus was successfully embolized on angiography, but surgical hemostasis was performed for suspected bleeding on exploratory laparotomy. Postembolization syndrome occurred in one patient.
Conclusions
TAE is effective and safe for hemorrhage after hysterectomy for gynecologic diseases. Angiographic findings are primarily active bleeding, but angiographic staining is not uncommon. A bleeding focus is possible in any branch of the IIA, as well as the arteries supplying the abdominal wall.
Introduction
Hemorrhage after gynecologic abdominal, vaginal, or laparoscopic surgery is a life-threatening complication that requires immediate management (1,2). Although surgery is a common choice of treatment for these postoperative complications, it is challenging for the surgical and anesthetic team when the patient is hemodynamically unstable (1,2).
Transcatheter arterial embolization (TAE) is well known as a common minimally invasive non-surgical treatment for postpartum hemorrhage (3–6), but it has not been commonly used for hemorrhagic complications after gynecologic surgery (1,2,7,8). We retrospectively evaluated the effectiveness and safety of TAE for the treatment of hemorrhage after abdominal, vaginal, or laparoscopic hysterectomy for gynecologic diseases.
Material and Methods
Patients
Eleven consecutive patients (median age = 45 years; age range = 34–80 years) who underwent TAE for hemorrhage after hysterectomy for gynecologic diseases from January 2004 to November 2020 were included. Four academic tertiary referral centers participated in this study. The study was approved by the institutional review board at each participating hospital, and written informed consent was waived owing to the retrospective nature of the study.
Angiography and TAE techniques
All angiography and subsequent TAE procedures were performed by one of five experienced interventional radiologists. A 5-Fr introducer sheath (Terumo, Tokyo, Japan) was inserted through the common femoral artery under local anesthesia. Arteriography of the bilateral internal iliac arteries (IIAs) was performed to identify the bleeding focus using a 5-Fr diagnostic catheter (Cobra; Cook Medical, Bloomington, IN, USA). The Waltman loop technique was used to select the ipsilateral IIA. Super-selective angiography of the bleeding artery was performed using a 1.9- to 2.2-Fr microcatheter (Renegade; Boston Scientific, Natick, MA, USA or Progreat; Terumo, Tokyo, Japan). When the bleeding focus was recognized on computed tomography (CT), initial angiography was performed in the relevant arteries. Contrast media extravasation, a pseudoaneurysm, or abrupt arterial cutoff were considered as positive angiographic findings for bleeding. The embolization endpoint was when no further active bleeding focus was shown with complete stasis of contrast medium on completion angiography.
Absorbable gelatin sponge particles (Spongostan; Johnson & Johnson, Gauteng, South Africa), microcoils 3–7 mm in diameter (Nester; Cook Medical, Bloomington, IN, USA), and/or N-butyl-cyanoacrylate (NBCA) (Histoacryl; B Braun, Sempach, Switzerland) were used as embolic agents. Microcoils or NBCA were used for active bleeding foci at the discretion of the attending interventional radiologist. NBCA was mixed with ethiodized oil (Lipiodol; Andre Guerbet, Aulnay-Sous-Bois, France) at a ratio of 2:1–6:1 (oil:NBCA).
Definitions and analysis
Hemodynamic instability was defined as hypotension with systolic blood pressure < 90 mmHg despite the use of vasopressors (9). Massive transfusion was defined as transfusion of 10 units of packed red blood cells within a 24-h period. Patients were identified as having coagulopathy if they met one of the following criteria: international normalized ratio (INR) > 1.5, or thrombocytopenia with a platelet count < 50,000/μL (9). Technical success was defined as the non-visualization of the bleeding focus seen on post-TAE angiography. Clinical success was defined as the cessation of bleeding after TAE and without the need for repeated TAE or additional hemostatic surgery during the patient’s hospital stay. Surgical hematoma evacuation was not regarded as hemostatic surgery.
The electronic medical records and picture archiving and communication system database of all patients were reviewed to collect data on demographics, clinical presentation, hemodynamic stability, coagulopathy, angiographic findings, embolization details, technical and clinical success, and complications.
Results
Patient characteristics
The clinical characteristics of the 11 patients are summarized in Table 1. The gynecologic diseases for hysterectomy were leiomyoma (n = 2), adenomyosis (n = 2), combined leiomyoma and adenomyosis (n = 1), ovarian cancer (n = 1), cervical cancer (n = 1), endometrial cancer (n = 1), retroperitoneal cyst with an inflammatory adhesion (n = 1), uterine prolapse (n = 1), and leiomyosarcoma (n = 1). Transabdominal hysterectomy (n = 8) was the most common method for hysterectomy, followed by laparoscopically assisted vaginal hysterectomy (n = 2) and laparoscopic radical hysterectomy (n = 1). The median interval between surgery and angiography was one day (range = 0–82 days). The most common clinical presentations were bloody Jackson–Pratt (JP) drainage (n = 5) and vaginal bleeding (n = 4).
Patient clinical characteristics and embolization details.
*On laparotomy, the suspicious bleeding focus was at the embolized artery and successful surgical hemostasis followed.
ant., anterior; BP, blood pressure; br., branch; BSO, bilateral salpingo-oophorectomy; CE, contrast extravasation; CM, clinical manifestation; CV, cervicovaginal; GSP, gelatin sponge particles; Hb, hemoglobin; IIA, internal iliac artery; JP, Jackson–Pratt; lat., lateral; LAVH, laparoscopically assisted vaginal hysterectomy; LRH, laparoscopic radical hysterectomy; NA, not available; NBCA, N-butyl cyanoacrylate; PES, postembolization syndrome; PNLD, pelvic lymph node dissection; PSA, pseudoaneurysm; pud., pudendal; Rt, right; TAH, transabdominal hysterectomy.
Hemodynamic instability was noted in six patients (54.5%; patients 3, 4, 7, 9, 10, and 11), of whom four (36.4%; patients 4, 7, 9, and 11) required massive transfusion. Coagulopathy was noted in two patients (patients 1 and 11). CT scans were available in seven patients, and the findings were contrast extravasation (n = 5), pseudoaneurysm (n = 1), and both (n = 1) (Fig. 1).
A 34-year-old woman (patient 7) who underwent transabdominal hysterectomy for endometrial cancer two days previously presented with bloody Jackson–Pratt drainage and blood pressure drop. (a) Axial and (b) coronal enhanced computed tomography scans show a pseudoaneurysm (arrow) with contrast extravasation (arrowheads) from an iliac branch (black arrows) of the right obturator artery (black arrowhead). Note the large amount of hemoperitoneum (asterisks). (c) Right internal iliac angiogram shows contrast extravasation (arrow). (d) Selective right obturator angiogram shows contrast extravasation (asterisk) from the iliac branch (arrows) of the obturator artery (arrowheads). Right anterior oblique view with microcatheter super-selection shows contrast extravasation (asterisk) from the iliac branch (arrows) (inset). N-butyl cyanoacrylate embolization was performed with successful hemostasis (not shown).
Characteristics and outcome of TAE
The embolization details are summarized in Table 1. Bleeding foci were observed on angiography in eight patients (72.7%), including contrast extravasation (n = 6) (Fig. 1) and both pseudoaneurysm and contrast extravasation (n = 2) (Fig. 2). The other three patients (27.3%) showed angiographic staining (Fig. 3) at the vaginal stump.
A 56-year-old woman (patient 1) who underwent transabdominal hysterectomy for ovarian cancer one day previously presented with bloody Jackson–Pratt drainage. (a) Right common iliac angiogram (left anterior oblique view) shows a pseudoaneurysm (arrow) from the lateral sacral artery (arrowheads). Delayed angiogram shows a pseudoaneurysm (arrow) and contrast extravasation (asterisks) (inset). (b) Postembolization angiogram shows compact coil packing (arrows) at the lateral sacral artery with no visualization of the bleeding focus.
A 45-year-old woman (patient 5) who underwent laparoscopically assisted vaginal hysterectomy for adenomyosis 82 days previously presented with vaginal bleeding. (a) Left internal iliac angiogram show angiographic staining (arrowheads) from the cervicovaginal branch and vaginal artery (arrows) at the vaginal stump. (b) Delayed angiogram shows more prominent angiographic staining (arrowheads). Gelatin sponge particle embolization was performed with successful hemostasis (not shown). The right internal iliac angiogram showed similar findings (not shown).
In eight patients with active bleeding, the bleeding artery was mostly the anterior division branches of the IIA (n = 6; patients 3, 4, 7, 9, 10, and 11) (Fig. 1), but there were also posterior division branches (lateral sacral artery, n = 1; patient 1) (Fig. 2) and the inferior epigastric artery (n = 1; patient 2). In the remaining three patients, angiographic staining was observed without active bleeding foci at the vaginal stump, and the feeders for staining were all anterior division branches of the IIA (patients 5, 6, and 8) (Fig. 3). Embolic materials were NBCA (n = 4) (Fig. 1), coils (n = 1) (Fig. 2), NBCA/coils (n = 1), coils/GSP (n = 1), and NBCA/GSP (n = 1) in eight patients with active bleeding, and were gelatin sponge particles (n = 2) (Fig. 3) and coils (n = 1) in three patients with angiographic staining.
In all patients with active bleeding or angiographic staining, TAE was technically successful. Clinical success was achieved in 90.9% (10/11) of patients. In one patient (patient 4), active bleeding from the anterior division of the right IIA was successfully embolized with coils and NBCA, and bleeding ceased on post-TAE angiography; however, a small amount of bleeding was suspected at the embolized anterior division of the right IIA on exploratory laparotomy, and hemostasis was performed surgically. Hematoma evacuation was performed surgically immediately after TAE in four patients (patients 1, 2, 4, and 11), and in one of them, the hemostasis described above was also performed.
The mean level of hemoglobin increased from 7.9 to 10.2 g/dL in 10 patients whose hemoglobin levels were available before and after the procedure. Postembolization syndrome occurred in one patient (patient 4), and her fever and elevated C-reactive protein level improved after three days with antibiotics. All patients were discharged from the hospital uneventfully at a median of eight days (mean = 9.9 days; range = 2–35 days) after the TAE.
Discussion
Hemorrhagic complications after hysterectomy can occur owing to improper vascular ligation and/or insufficient hemostasis of an anastomotic site or cut surface, despite advanced surgical techniques (1). Given that the incidences of hemodynamic instability and massive transfusion were 55% and 36%, respectively, hemorrhage can be regarded as life-threatening, so prompt management is essential. If there are any indications of bleeding after surgery, such as bloody JP drainage and vaginal bleeding, CT or angiography should be considered immediately. CT is very sensitive for the detection/localization of active bleeding, as all seven patients in this study who underwent CT examination had active bleeding. Angiography showed active bleeding in 73% of patients, especially in patients with hemodynamic instability. Therefore, it is important to promptly diagnose and treat patients with clinically suspected bleeding through CT and angiography.
When analyzing the eight bleeding arteries on angiography, the origin from the anterior division of the IIA was the most common and included its branches (the internal pudendal artery, obturator artery, and cervicovaginal branch), as well as the anterior division of the IIA itself (1,2). It is reasonable that anterior division branches of the IIA are the most common bleeding arteries, but the posterior division branches of the IIA or inferior epigastric artery that supplies the abdominal wall may also be responsible. Posterior division branches of the IIA, such as the lateral sacral artery in this study and a previous study (10) and the gluteal artery in another report (1), can also be damaged when removing an enlarged uterus. Although for hemorrhage after obstetric hysterectomy, the remnant uterine artery was reported as the most common bleeding artery (10), there were various bleeding arteries associated with hemorrhage after gynecologic hysterectomy.
Technically, the approach seems less difficult than TAE for postpartum hemorrhage because the uterine artery is absent or remains as a stump, and the active bleeding rate is high mostly in the anterior division branches of the IIA. Clinically, TAE of direct bleeding foci or suspicious blood vessels even without foci has good hemostatic effects, such as vital sign recovery and hemoglobin elevation. If there is a large intraperitoneal hematoma, hematoma evacuation is necessary regardless of clinical improvement after TAE. In this study, identification of suspicious bleeding foci and additional surgical hemostasis were possible, as shown in one patient.
Regarding embolic materials, permanent materials, such as NBCA and coils, were primarily used for active bleeding in this study. The use of permanent embolic agents, especially NBCA, is advantageous for more effective hemostasis, even in patients with coagulopathy. The mechanism of NBCA embolization involves rapid polymerization and hardening of NBCA in anionic fluid, and it provides effective embolization without abundant clotting factor (6). On the other hand, gelatin sponge particles involve the hemostatic mechanisms of blocking blood flow and inducing thrombus formation, requiring good coagulation capacity, and they show good embolization effect for angiographic staining.
In this study, the postembolization syndrome in one patient was tolerable and there were no other complications, so TAE as a whole is very safe. Postembolization syndrome presents in the first three days and is characterized by nausea, vomiting, pain, fever, and/or flu-like symptoms. Symptoms are self-limiting and treatment is supportive. TAE can minimize disadvantages, such as general anesthesia accompanying surgery and surgical complications, including bleeding, infection, and ureter injury. In addition, because the pelvic region is supplied with an extensive anastomotic vascular network, hemostatic surgery aimed at the identification and ligation of bleeding vessels is rather limited in its probability of success owing to minimal or unidentifiable bleeding sites, anatomical inaccessibility of the bleeding arteries, or associated pelvic adhesions (1,2).
The present study has several limitations. First, it involves a retrospective design based on a review of medical records. Second, the number of patients in this study was small and were collected from four hospitals. The hysterectomy method in the four hospitals was not the same, and the diagnostic criteria and management for hemorrhage after hysterectomy were not uniform. Thus, it was difficult to determine the frequency and role of TAE for hemorrhage after hysterectomy.
In conclusion, TAE is effective and safe for hemorrhage after hysterectomy for gynecologic diseases. Angiographic findings are primarily active bleeding, but angiographic staining without active bleeding focus is not uncommon. A bleeding focus is possible in any branch of the IIA, as well as the arteries supplying the abdominal wall.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received the following financial support for the research, authorship, and/or publication of this article: This study was supported by Natural Science Foundation of Henan Province (Grant No. 202300410463).
