Stereotactic Body Radiotherapy (SBRT) for an Extracranial Arteriovenous Malformation of the Pelvis.

Extracranial arteriovenous malformations (AVMs) are rare pathological, benign conditions that are characterized by aberrantly connected arteries and veins without normal intervening capillary vasculature. Although stereotactic radiosurgery is an established, efficacious, safe treatment for intracranial AVMs, there is no known published data on the use of stereotactic body radiotherapy (SBRT) for the treatment of abdominopelvic AVMs. One patient with an extracranial AVM in the pelvis that was only partially responsive to embolization was treated with SBRT to a dose of 21 Gy, delivered in three fractions over six calendar days. At presentation, the patient was non-ambulatory due to neuropathic pain from a sciatic impingement of the AVM. The patient underwent two prior catheter-based embolization procedures that had achieved partial obliteration, but with the persistence of neuropathic pain and symptoms. After SBRT to the pelvic AVM, the patient had marked improvement in pain over 10 months and was able to ambulate again. Follow-up angiography and CT demonstrated the obliteration of previously visualized AVM. We describe the first known report of pelvic AVM successfully treated with a combination of embolization and SBRT. Three-fraction SBRT to a total dose of 21 Gy appears to be safe and effective for extracranial AVMs arising in the pelvis. This strategy may be considered for patients with pelvic AVMs that are refractory to standard interventional therapies. However, these findings should be validated in larger cohorts.

Initial Canadian Experience of Prostate Artery Embolization for Symptomatic Benign Prostatic Hyperplasia: Midterm Outcomes.

INTRODUCTION: This study aimed to assess the midterm outcomes and safety of prostate artery embolization (PAE) for the treatment of benign prostatic hyperplasia (BPH). METHODS: A single-center, retrospective review of PAE performed for BPH was performed. Validated International Prostate Symptom Score (IPSS), quality of life (QoL) index, and International Index for Erectile Function (IIEF-5) questionnaires were completed at baseline and at least 12 months post-procedure. Prostate imaging was performed preprocedure as well as 3 months and 12 months post-procedure to assess prostate volume (PV). Uroflowmetry was also performed at baseline and 12 months post-procedure to assess urine flow rate (Qmax) and post-void residual (PVR) volume. Adverse events were graded according to Society of Interventional Radiology (SIR) guidelines. RESULTS: Eighty male patients underwent the PAE procedure (mean age 69 years). Prostate volume decreased significantly from a mean volume 156 to 107 mL after 12 months post-procedure, commensurate with a mean reduction of 27.5% (P < .05). Significant improvements were seen in IPSS (21.8 vs 10.5) and QoL (4.5 vs 2.0) from baseline to 12 months post-procedure (P < .05). There was no significant change in IIEF-5 score. There was a significant reduction in PVR (202 vs 105 mL) and improvement in Qmax (5.9 vs 10.0 mL/s) between baseline and 12 months post-procedure (P < .05). No major complications occurred; 4 minor complications occurred (SIR grade A or B). CONCLUSION: Prostate artery embolization achieved a clinically and statistically significant prostate volume reduction, symptom and QoL improvement, and enhanced uroflowmetry parameters in patients with BPH.

Correction to: Minimally Invasive Treatment for Benign Prostatic Hyperplasia: Economic Evaluation from a Standardized Hospital Case Costing System.

The original version of this article published previously contained an error.

Minimally Invasive Treatment for Benign Prostatic Hyperplasia: Economic Evaluation from a Standardized Hospital Case Costing System.

PURPOSE: Minimally invasive alternatives to transurethral resection of the prostate (TURP) such as prostate arterial embolization (PAE) and photoselective vaporization of the prostate (PVP) are being explored as adjuncts in the care of patients with benign prostatic hyperplasia. However, there are conflicting reports of the costs of these procedures. The purpose of this study was to compare the direct and indirect hospital costs of TURP, PAE and PVP. MATERIALS AND METHODS: A chart review was performed in patients who underwent TURP, PVP and PAE from April 2015 to March 2017. All hospital costs were collected in accordance with the Ontario Case Costing Initiative, a standardized medical case costing system. Costs were characterized as direct or indirect and fixed or variable. Probabilistic sensitivity analysis was conducted to study cost uncertainty. RESULTS: During the study period, a total of 209 men underwent TURP, 28 PVP and 21 PAE. Mean age (years) was as follows: TURP 71.43; PVP 73.66; PAE 70.77 (p = 0.366). Mean length of stay (days) was as follows: TURP 1.63; PVP 1.55; PAE 1 (p = 0.076). Total costs of the PAE group ($3829, SD $1582) were less than both PVP ($5719, SD $1515) and TURP groups ($5034, SD $1997, p < 0.001). There was no significant difference in direct costs between the groups. Monte Carlo simulation demonstrated that PAE was the least costly alternative majority of the time. CONCLUSIONS: The total hospital costs of PAE at our institution are significantly lower than those of PVP and TURP.

Successful balloon assisted percutaneous thrombin injection of right subclavian artery pseudoaneurysm.

PURPOSE: To report the use of a balloon catheter to facilitate percutaneous ultrasound-guided thrombin injection of a subclavian artery pseudoaneurysm. CASE REPORT: A 36-year-old man presented with hoarseness after the insertion of a right temporary internal jugular line. Arteriography showed a short neck right subclavian artery pseudoaneurysm. A 6 mm × 40 mm balloon catheter was positioned across the neck of the pseudoaneurysm to permit percutaneous ultrasound-guided thrombin injection. Immediate post-treatment angiogram demonstrated no filling of the pseudoaneurysm. Seven-month clinical and imaging follow-up confirmed resolution of the pseudoaneurysm and hoarseness. CONCLUSIONS: This clinical vignette highlights the potential of balloon catheter-assisted percutaneous ultrasound-guided thrombin injection as an alternative to open or endovascular repair of pseudoaneurysms of the subclavian arteries.

Intraarterial Lidocaine for Pain Control in Uterine Artery Embolization: A Prospective, Randomized Study.

PURPOSE: To assess efficacy of two different techniques of lidocaine injection in the uterine arteries to reduce pain following uterine artery embolization (UAE) for leiomyomas. MATERIALS AND METHODS: This prospective randomized single-blinded study was performed with 60 patients enrolled between November 2014 and December 2015 equally randomized to 3 arms. Group A received 10 mL lidocaine 1% (100 mg) mixed with polyvinyl alcohol particles (355-500 µm). Group B received the same dose of lidocaine injected after embolization. Group C was a control group. Pain was assessed on a 100-point visual analog scale at 4, 7, and 24 hours after the procedure. Narcotic agent dose to 24 hours was recorded. Outcomes were examined by analysis of variance and pairwise comparison. Leiomyoma infarction was assessed with magnetic resonance imaging 3 months after the procedure. RESULTS: Technical success rate of UAE was 100%. Mean pain score at 4 hours was significantly lower in the lidocaine groups (group A, 28.6; group B, 35.8) compared with the control group (59.4; P = .001). Pain scores at 7 and 24 hours were not statistically different among the 3 arms. The mean in-hospital narcotic agent dose was significantly lower in both lidocaine groups than in the control group (group A, 8.5 mg [P = .002]; group B, 11.1 mg [P = .03]; group C, 17.4 mg). There were no adverse events related to the use of lidocaine. The number of patients with complete infarction of leiomyomas at 3 months was significantly lower in group A at 38.9% (group B, 77.8%; group C, 75%; P = .0451). CONCLUSIONS: Lidocaine injected in the uterine arteries reduced postprocedural pain and narcotic agent dose after UAE. There were more cases of incomplete necrosis when lidocaine was mixed with the particles.

Safety and Effectiveness of Percutaneously Inserted Peritoneal Ports Compared to Surgically Inserted Ports in a Retrospective Study of 87 Patients with Ovarian Carcinoma over a 10-Year Period.

PURPOSE: Placement of peritoneal ports has become a favorable technique for direct chemotherapy infusion in treating peritoneal metastases from ovarian cancer. We aim to outline an approach to the percutaneous insertion of peritoneal ports and to characterize success and complication rates compared to surgically inserted ports. MATERIALS AND METHODS: Retrospective analysis was collected from 87 patients who had peritoneal port insertion (28 inserted surgically and 59 percutaneously) for treatment of peritoneal metastases from ovarian cancer from July 2004 to July 2014. Complications were classified according to the SIR Clinical Practice Guidelines as major or minor. RESULTS: Technical success rates for surgically and percutaneously inserted ports were 100 and 96.7 %, respectively (p = 0.44), with the two percutaneous failures successful at a later date. There were no major complications in either group. Minor complication rates for surgically versus percutaneously inserted ports were 46.4 versus 22.0 %, respectively (p = 0.02). The infection rate for surgically inserted versus percutaneously inserted ports was 14.3 and 0 %, respectively (p = 0.002). The relative risk of developing a complication from percutaneous peritoneal port insertion without ascites was 3.4 (p = 0.04). For percutaneously inserted ports, the mean in-room procedure time was 81 ± 1.3 min and mean fluoroscopy time was 5.0 ± 4.5 min. CONCLUSION: Percutaneously inserted peritoneal ports are a safe alternative to surgically inserted ports, demonstrating similar technical success and lower complication rates.

Does Technical Success of Angioplasty in Dysfunctional Hemodialysis Accesses Correlate with Access Patency?

PURPOSE: To study if <30 % residual stenosis post angioplasty (PTA) correlates with primary access circuit patency, and if any variables predict technical success. MATERIALS AND METHODS: A prospective observational study was performed between January 2009 and December 2012, wherein 76 patients underwent 154 PTA events in 56 prosthetic grafts (AVG) and 98 autogenous fistulas (AVF). Data collected included patient age, gender, lesion location and laterality, access type and location, number of prior interventions, and transonic flow rates pre- and postintervention. Impact of technical outcome on access patency was assessed. Univariate logistic regression was used to assess the impact of variables on technical success with significant factors assessed with a multiple variable model. RESULTS: Technical success rates of PTA in AVFs and AVGs were 79.6 and 76.7 %, respectively. Technical failures of PTA were associated with an increased risk of patency loss among circuits with AVFs (p < 0.05), but not with AVGs (p = 0.7). In AVFs, primary access patency rates between technical successes and failures at three and 6 months were 74.4 versus 61.9 % (p = 0.3) and 53.8 versus 23.8 % (p < 0.05), respectively. In AVGs, primary access patency rates between technical successes and failures at three and six months were 72.1 versus 53.9 % (p = 0.5) and 33.6 versus 38.5 % (p = 0.8), respectively. Transonic flow rates did not significantly differ among technically successful or failed outcomes at one or three months. CONCLUSION: Technical failures of PTA had a significant impact on access patency among AVFs with a trend toward poorer access patency within AVGs.

Elastic Recoil after Balloon Angioplasty in Hemodialysis Accesses: Does It Actually Occur and Is It Clinically Relevant?

Purpose To qualify and quantify elastic recoil and determine its effect on access patency. Materials and Methods Research ethics board approval was obtained and all patients signed an informed consent form. This was a prospective, nonrandomized study of mature accesses that underwent balloon percutaneous transluminal angioplasty (PTA) between January 2009 and December 2012. After PTA, completion fistulography was performed at 0-, 5-, 10-, and 15-minute intervals. From Digital Imaging and Communications in Medicine images, percentage of lesion stenosis before and after PTA was measured at each time point. A total of 76 patients (44 men, 32 women; mean age, 59.6 years) were enrolled and underwent 154 PTAs in 56 grafts and 98 fistulas. Venous elastic recoil was defined as recurrent luminal narrowing greater than 50% within 15 minutes after full effacement of the stenosis by the angioplasty balloon. Data collected included sex, age, access type and location, lesion location, length, and time to next intervention. Access patency was estimated by using Kaplan-Meier survival method, association of variables with the risk of loss of patency was assessed by using a Cox proportional hazards model, and a multiple variable model was examined by considering all variables. Results Technical success of PTA with less than 30% residual stenosis was 78%. By 15 minutes, 15.6% (24 of 154) of treated lesions recurrently narrowed by more than 50%, with a majority observed at 5 minutes (15 of 24). Technical failure of PTA was predictive of elastic recoil (P < .001), as was cephalic arch stenosis in fistulas (P = .047) and autogenous fistulas (P = .04). Elastic recoil, when it did occur, did not influence patency. Six-month primary patency was 34.8% in grafts and 47.1% in fistulas. Conclusion Venous elastic recoil after PTA of stenoses in hemodialysis access circuits is common, but its occurrence does not influence access primary patency after PTA. (©) RSNA, 2015.

Randomized Trial Comparing Radiologic Pigtail Gastrostomy and Peroral Image-Guided Gastrostomy: Intra- and Postprocedural Pain, Radiation Exposure, Complications, and Quality of Life.

PURPOSE: To prospectively compare radiologically created pigtail gastrostomy (PG), in which the tube is inserted directly through the abdominal wall, versus peroral image-guided gastrostomy (POG), in which the tube is inserted through the mouth. Pain profiles (primary outcome measure), fluoroscopy times, total room times, technical success, complications, and quality of life (QOL) were measured. MATERIALS AND METHODS: Sixty patients were prospectively randomized to receive 14-F PG or 20-F POG tubes. All patients received prophylactically created gastrostomies before radiation therapy for head and neck squamous-cell carcinoma. Patients receiving palliative treatment were excluded, as were those with established pharyngeal obstruction. Pain was measured by numeric rating scale (NRS) scores for 6 weeks after the procedure and by intraprocedural fentanyl and midazolam doses and postprocedural 24-h morphine doses. Fluoroscopy times, total room times, technical success, complications up to 6 months, and gastrostomy-related QOL (using the Functional Assessment of Cancer Therapy-Enteral Feeding questionnaire) were determined. RESULTS: Fifty-six patients underwent the randomized procedure. The POG group required significantly higher intraprocedural midazolam and fentanyl doses (mean, 1.2 mg and 67 µg, respectively, for PG vs 1.9 mg and 105 µg for POG; P < .001) and had significantly longer fluoroscopy times (mean, 1.3 min for PG vs 4.8 min for POG; P < .0001). NRS scores, morphine doses, total room times, technical success, complication rates, and QOL did not differ significantly between groups. The one major complication, a misplaced PG in the peritoneal cavity, followed a technical failure of POG creation. CONCLUSIONS: Despite the differences in insertion technique and tube caliber, the measured outcomes of POG and PG are comparable.

Efficacy of Ovarian Artery Embolization for Uterine Fibroids: Clinical and Magnetic Resonance Imaging Evaluations.

PURPOSE: The objective of the study was to assess the efficacy of ovarian artery embolization (OAE) treatment for symptomatic uterine leiomyomas. METHODS: A retrospective review of 17 patients who underwent OAE in conjunction with uterine artery embolization in a 6-year period (2006-2012) was performed. Ten patients had previous failed embolization, while 7 had not received any embolization therapy before. Percent uterine volume change, percent dominant fibroid volume change, and percent dominant fibroid infarction were assessed with magnetic resonance (MR) imaging. Resolution of menorrhagia, dysmenorrhea/pain, and bulk and/or pressure symptoms including urinary frequency were evaluated clinically. Change in menopausal state was also an outcome of interest. RESULTS: Mean MR imaging follow-up was performed 3 months post-OAE. MR images showed complete infarction in the majority of cases (64.7%; n = 11), with infarction rates of 90%-100% in 3 cases, 1 case with 30%-50% infarction, and 2 cases with 0%-10% infarction. Average uterine size reduction on MR was 32.3% (95% confidence interval [CI]: 22.5%-42.2%; P < .001). The average size reduction for the dominant fibroid was 42.4% (95% CI: 27.7%-57.0%; P = .01). The mean time to final follow-up visit was 11 months. At this point complete symptom resolution (menorrhagia, dysmenorrhea and bulk-related) was achieved in 82.4% (n = 14) of cases. At the final follow-up 11.8% (n = 2) of cases reported menopause. CONCLUSIONS: We observed OAE to be an effective and safe adjunct to uterine artery embolization when hypertrophic ovarian artery(ies) require intervention. However, incomplete fibroid infarction of 23% remains a concern with a potential for long-term treatment failure. In addition, long-term effect on ovarian function is uncertain.

Venous aneurysms in autogenous hemodialysis fistulas: is there an association with venous outflow stenosis.

PURPOSE: To determine whether patients with venous aneurysms in their arteriovenous fistulas (AVFs) have associated venous outflow stenoses. MATERIALS AND METHODS: A retrospective study was performed, which included all patients presenting with dysfunction and had venous aneurysms in their AVFs. Patient's medical records and imaging studies were examined and data collected including access characteristics, patient demographics and imaging findings. Data were analyzed using Fisher's exact test. RESULTS: A total of 89 patients (58 men, 31 women; mean age 60) presented for intervention related to access dysfunction with incident venous aneurysms over the study period. Of the 89 patients with venous aneurysms (mean diameter 2.3 cm) of their AVF's, 69 (78%) patients had an associated venous outflow stenosis. The stenoses were present most commonly in the outflow cephalic vein (57%), followed by the cephalic arch (20%), brachiocephalic vein (10%) and subclavian vein (6%). Outflow stenoses in AVFs with venous aneurysms were observed in 87% of brachiocephalic AVFs, 60% of radiocephalic AVFs and 80% of brachiobasilic AVFs. Brachiocephalic AVFs with venous aneurysms were significantly more likely to have an associated outflow stenosis than radiocephalic AVFs with venous aneurysms (P=0.007). AVFs with outflow stenosis were on average 1502 days old while AVFs without outflow stenosis were on average 2351 days old, which was a statistically significant difference (P=0.031). No statistically significant differences were observed for sex and side of the fistula. CONCLUSIONS: Outflow stenosis was observed to be associated with venous aneurysms in AVFs with a more statistically significant association in brachiocephalic AVFs compared to other AVFs.

Concealed postpartum hemorrhage treated with transcatheter arterial embolization.

BACKGROUND: Postpartum hemorrhage is an obstetric emergency and is a major preventable cause of maternal morbidity and mortality. An arteriovenous fistula is a rare cause of concealed postpartum hemorrhage. CASE: A 20-year-old woman spontaneously delivered at 40 0/7 weeks of gestation. Twelve hours after delivery, she became hemodynamically unstable, developing a large left vaginal hematoma. An angiogram revealed extravasation originating from the right pudendal artery with early filling of a draining vein, consistent with a traumatic arteriovenous fistula. Complete occlusion of the fistula was achieved by embolizing a branch of the right pudendal artery. The postprocedure course was uneventful. CONCLUSION: A vaginal arteriovenous fistula should be considered in cases of concealed postpartum hemorrhage; transcatheter arterial embolization is an effective treatment for these cases.

Central venous stenosis associated with pacemaker leads: short-term results of endovascular interventions.

PURPOSE: To determine the safety and outcomes of performing endovascular interventions across pacemaker (PM) lead-related central vein stenosis/occlusion. MATERIALS AND METHODS: Clinical and imaging records from April 2002 to August 2010 were reviewed for patients presenting with clinically significant central venous stenosis or central venous obstruction with indwelling PM leads and were evaluated to determine if any lead dysfunction or complications occurred following intervention. Thirteen patients were identified who underwent 14 interventions. Outcomes were assessed based on recurrent clinical symptoms and repeat venograms, and patency was determined with Kaplan-Meier analysis. RESULTS: All procedures were technically successful (100%) with no procedural complications. All patients underwent balloon angioplasty; two patients also underwent stent placement for persistent flow-limiting stenosis within the superior vena cava. No incidence of PM or lead dysfunction, infection, or need for PM intervention was identified in any patient at the time of the index procedure and at 30 or 90 days. In the two patients who received stents across PM leads, no manifestations of PM or lead dysfunction or infection, or need for intervention, were identified at 1 year after stent placement in one patient and at 3 years in the other patient. For eight patients, sufficient follow-up studies were available, with 30-, 60-, and 120-day primary patency rates of 87.5%, 75%, and 50%, respectively. CONCLUSIONS: Endovascular interventions across PM leads resulted in no adverse clinical cardiac or PM sequelae, but with limited primary patency compared with historical outcomes for central venous interventions in the absence of PM leads.

Patency of femoral tunneled hemodialysis catheters and factors predictive of patency failure.

PURPOSE: To determine the patency rates of and factors associated with increased risk of patency failure in patients with femoral vein tunneled hemodialysis catheters. METHODS: All femoral tunneled catheter insertions from 1996 to 2006 were reviewed, during which time 123 catheters were inserted. Of these, 66 were exchanges. Patients with femoral catheter failure versus those with femoral catheter patency were compared. Confounding factors, such as demographic and procedural factors, were incorporated and assessed using univariate and multivariable Cox proportional hazards regression analyses. RESULTS: Mean catheter primary patency failure time was 96.3 days (SE 17.9 days). Primary patency at 30, 60, 90, and 180 days was 53.8%, 45.4%, 32.1%, and 27.1% respectively. Crude rates of risk of catheter failure did not suggest a benefit for patients receiving catheters introduced from one side versus the other, but more cephalad location of catheter tip was associated with improved patency. Multivariate analysis showed that patients whose catheters were on the left side (p = 0.009), were of increasing age at the time of insertion (p = 0.002) and that those who had diabetes (p = 0.001) were at significantly greater risk of catheter failure. The catheter infection rate was 1.4/1000 catheter days. CONCLUSION: Patients who were of a more advanced age and had diabetes were at greater risk of femoral catheter failure, whereas those who received femoral catheters from the right side were less at risk of catheter failure.

Exaggerated increase in cardiac output during exercise in patients with peripheral high-flow arteriovenous malformations.

PURPOSE: To test the hypothesis that cardiac output (Q) in patients with arteriovenous malformations (AVMs) is normal at rest and increases disproportionately during exercise. MATERIALS AND METHODS: Q was measured in eight patients with large peripheral AVMs and in nine healthy subjects using a noninvasive carbon dioxide (CO2)-based differential Fick method. Subjects were tested while seated at rest and during mild exercise (repeated leg straightening while sitting). Oxygen consumption (VO2) was monitored as an index of the degree of exercise. RESULTS: Average resting Q was similar between AVM patients and healthy subjects (7.40 L/min ± 3.29 vs 6.13 L/min ± 0.94, respectively, P = .29). During exercise, AVM patients showed a smaller increment in VO2 (0.50 L/min ± 0.11 vs 0.78 L/min ± 0.26, P = .012) but with more apparent effort and shortness of breath compared with healthy subjects. The change in Q per unit change in VO2 (ΔQ/ΔVO2) was greater in AVM patients than in healthy subjects (16.00 L/min ± 6.50 vs 9.79 L/min ± 5.33, P < .045). CONCLUSIONS: Exercise intolerance in AVM patients may be due to an imbalance in ΔQ/ΔVO2 resulting from increased shunting through the AVM. Exercise provocation may increase the sensitivity of Q in the clinical evaluation of AVM patients.

Effectiveness of stent-graft placement for salvage of dysfunctional arteriovenous hemodialysis fistulas.

PURPOSE: To determine the clinical effectiveness of stent-graft placement in dysfunctional autogenous hemodialysis fistulas. MATERIALS AND METHODS: Between September 2006 and June 2008, 24 Fluency Plus stent-grafts were inserted in 17 patients with failing autogenous hemodialysis fistulas (two radiocephalic, 12 brachiocephalic, and three brachiobasilic). Six fistulas were thrombosed at presentation. Indications for stent-graft insertion included 10 residual stenoses, six pseudoaneurysms, and one fistuloplasty-induced rupture. Fistula function before and after intervention was assessed with ultrasound dilution technique. RESULTS: The technical success rate was 100%, as were the anatomic and clinical success rates. Via Kaplan-Meier method, the primary access patency rates were 94.1% at 3 months (95% CI, 0.461-0.995) and 88.2% at 6 and 12 months (95% CI, 0.468-0.980). The primary lesion patency rate was 94.1% (95% CI, 0.650-0.992) at 3, 6, and 12 months. The secondary access patency rate had a point estimate of 100% at 3 months, and was 93.8% (95% CI, 0.632-0.991) at 6 and 12 months. Univariate analysis suggested that male sex and left-sided accesses were associated with a longer primary access patency rate (P = .02). The binary restenosis rate at 6 months was 5.9%. Mean transonic flow rates were 667.5 mL/min before intervention and 2,076.0 mL/min after intervention (P = .03, Wilcoxon signed-rank test). No patients were lost to follow-up. CONCLUSIONS: Stent-graft placement in dysfunctional autogenous hemodialysis fistulas is technically feasible and effective in preserving function and preventing access abandonment, with patency rates that exceed historical patency rates with angioplasty and/or uncovered stents.

Arterial embolization for primary postpartum hemorrhage.

PURPOSE: To evaluate the efficacy and safety of arterial embolization (AE) for treatment of primary postpartum hemorrhage (PPH), and the factors associated with clinical success. MATERIAL AND METHODS: A retrospective analysis of all patients undergoing AE for primary PPH at three institutions (N = 43) from 1996 through 2007 was conducted. Patients with an antepartum diagnosis of invasive placenta were excluded from the study. Mean patient age was 31 years +/- 5 (range, 21-40 y). Eighteen women (42%) were primiparous. Delivery details, transfusion requirements, hematology and coagulation results, embolization details, and clinical outcomes were collected. Clinical success was defined as cessation of bleeding without the need for repeat embolization, laparotomy, or hysterectomy after embolization; or death. The Fisher exact test was used to analyze nonparametric data. RESULTS: The clinical success rate was 79% (n = 34). Four patients underwent successful repeat embolization. Two of 35 patients who had not undergone hysterectomy before embolization underwent hysterectomy for continued bleeding (without repeat embolization). One underwent hysterectomy 2 weeks after AE for uterine necrosis. One of eight patients who had undergone hysterectomy before AE required a laparotomy for a large retroperitoneal hematoma, and one patient died from cerebral anoxia secondary to hypotension despite repeat embolization. Clinical success was not related to mode of delivery, cause of PPH, transfusion requirements, time from delivery to embolization, or hysterectomy before AE (P > .05). Patients with active extravasation visualized angiographically were more likely to require repeat embolization (five of 13 [38%] vs 0 of 30 without extravasation; P < .01). CONCLUSIONS: AE for primary PPH is safe and effective. Repeat embolization may be necessary in patients with active extravasation on angiography.