Faculty Bibliography

We sought to delineate factors of inferior vena cava filter placement associated with increased radiation and cost and difficult subsequent retrieval. In total, 299 procedures from August 2013 to December 2014, 252 in a fluoroscopy suite (FS) and 47 in the operating room (OR), were reviewed for radiation exposure, fluoroscopy time, filter type, and angulation. The number of retrieval devices and fluoroscopy time needed for retrieval were assessed. Multiple linear regression assessed the impact of filter type, procedure location, and patient and procedural variables on radiation dose, fluoroscopy time, and filter angulation. Logistic regression assessed the impact of filter angulation, type, and filtration duration on retrieval difficulty. Access site and filter type had no impact on radiation exposure. However, placement in the OR, compared to the FS, entailed more radiation (156.3 vs 71.4 mGy; P = 0.001), fluoroscopy time (6.1 vs 2.8 min; P < 0.001), and filter angulation (4.8 degrees vs 2.6 degrees ; P < 0.001). Angulation was primarily dependent on filter type (P = 0.02), with VenaTech and Denali filters associated with decreased angulation (2.2 degrees , 2.4 degrees ) and Option filters associated with greater angulation (4.2 degrees ). Filter angulation, but not filter type or filtration duration, predicted cases requiring >1 retrieval device (P < 0.001) and >30 min fluoroscopy time (P = 0.02). Cost savings for placement in the FS vs OR were estimated at $444.50 per case. In conclusion, increased radiation and cost were associated with placement in the OR. Filter angulation independently predicted difficult filter retrieval; angulation was determined by filter type. Performing filter placement in the FS using specific filters may reduce radiation and cost while enabling future retrieval.

Pulmonary embolism (PE), a potentially life-threatening entity, can be treated medically, surgically, and percutaneously. In patients with right ventricular dysfunction (RVD), anticoagulation alone may be insufficient to restore cardiac function. Because of the morbidity and mortality associated with surgical embolectomy, clinical interest in catheter-directed interventions (CDI) has resurged. We describe specific catheter-directed techniques and the evidence supporting percutaneous treatments.

OBJECTIVE: To estimate the incidence rates of kidney-related clinical outcomes among patients with tuberous sclerosis complex (TSC)-related angiomyolipoma compared to an age-matched control cohort in the United States. METHODS: This was a retrospective, observational study. Administrative data from the MarketScan(R) Research Databases were used to select patients with TSC and renal angiomyolipoma. An age-matched group with no TSC or renal angiomyolipoma was identified for comparison. Outcomes were incidence rates per 100 patient-years and number of months to development of hematuria, chronic kidney disease (CKD), renal hemorrhage, kidney failure, and inpatient death. RESULTS: Among the commercially insured TSC-renal angiomyolipoma patients (N=605) and matched controls (N=1,815), 37.2% were <18 years old. Among Medicaid TSC-renal angiomyolipoma patients (N=246) and matched controls (N=738), 38.6% were aged <18. In the commercial sample, in both age groups, the incidence rate of each clinical outcome measured was higher in the TSC-renal angiomyolipoma cohort than in the control cohort, with several differences reaching statistical significance. Compared with younger patients, older TSC-renal angiomyolipoma patients had higher incidence rates of clinical outcomes (hematuria: 20.4 vs. 8.7; CKD: 9.6 vs. 3.5; renal hemorrhage 2.7 vs. 0.7; kidney failure: 1.9 vs. 0.4) and took more time on average to develop each clinical outcome. A similar pattern of results was observed among patients with Medicaid insurance. CONCLUSION: TSC-renal angiomyolipoma patients are at significantly higher risk for renal morbidity relative to the general population.

Learning Objectives: 1. Describe anatomic and physiologic aspects of creating left hepatic vein (LHV) to left portal vein (LPV) transjugular intrahepatic portosystemic shunts (TIPS). 2. Illustrate through a series of 15 cases the technical aspects unique to left-sided TIPS. 3. Understand outcomes of left-to-left TIPS creation for refractory ascites and variceal bleeding, compared with right and middle hepatic vein and right portal vein TIPS Background: TIPS creation is a primary treatment for complications of portal hypertension. Classically, the shunt is created between the right hepatic and right portal veins, owing to the relatively larger size of the right hepatic lobe that would theoretically increase safety and the potential for accessing one of multiple portal venous branches. LHV to LPV TIPS is much less frequently done, though preliminary studies have suggested possibly increased safety. In addition, there may be specific indications for left to left TIPS, most of which are anatomic in nature Clinical Findings/Procedure Details: Fifteen patients underwent successful LHV to LPV TIPS creation at a single institution from 11/2011 to 12/2014. Through a series of examples, the anatomy of left-to-left TIPS creation will be reviewed: left-sided TIPS tend to be shorter and less angulated compared with right-sided TIPS. Given the shorter trajectory, use of a pediatric needle for access may be considered. Patient outcomes regarding control of ascites and bleeding will be reviewed. Only 2 of 15 patients (13%) required a revision with angioplasty or further stent placement over 6 months. All of the patients were alive at 1 month post-TIPS. Conclusions: Left hepatic to left portal vein TIPS creation is safe and provides comparable outcomes compared with right-sided TIPS. Future controlled trials in centers frequently performing TIPS creation could clarify which approach is optimal. However, familiarity with the technical aspects of left-sided TIPS creation is essential as many patients have anatomy unfavorable for right-sided TIPS

Purpose: To evaluate factors predicting contrast-induced nephropathy (CIN) after conventional angiography for symptomatic bleeding. Materials: All angiograms for intra-abdominal bleeding at one academic center (1/2013-6/2015) were reviewed retrospectively. Studies included 66 angiograms in 60 patients (20 women; 40 men) with a mean age of 63 yr (range 21-92). Indications included gastrointestinal (38), urinary tract (7), uterine (5), hepatic (3), splenic (2), extremity (1), or other abdominal (6) bleeding, and hemoptysis (4). Variables assessed included age, sex, history of congestive heart failure (CHF) or diabetes (DM), injected contrast volume or type, and the presence of active extravasation on angiography. Pulse, mean arterial pressure (MAP), and last glomerular filtration rate (GFR) were recorded immediately preceding the examination. The percent change in hematocrit (DELTAHct) and units of packed red blood cells (pRBC) were measured over the 24 hr prior to each angiogram. Studies resulting in CIN (25% increase in creatinine from baseline or 0.5 mg/dL increase in absolute value 48-72 hr from contrast administration) were compared to those not resulting in CIN with parametric and nonparametric tests, and logistic regression. Results: Of 66 angiograms, 10 (15%) resulted in CIN; 56 (85%) did not. Age, sex, CHF, DM, contrast volume/type, pulse, MAP, presence of active extravasation, and the pRBCs or DELTAHct over the 24 hours prior to angiogram did not significantly differ between groups. GFR was significantly lower in patients who experienced CIN (57 vs. >60; p=0.02). On logistic regression, GFR was the only independent predictor for CIN (p=0.03): 67% of patients with severe renal impairment (GFR < 30), 29% of patients with mild-moderate impairment (GFR 30-60), and 8% of patients with normal renal function (GFR > 60) developed CIN. Conclusions: CIN is relatively common following conventional angiography for symptomatic hemorrhage, in light of low rates reported for patients without hemorrhage after contrast-enhanced CT. Pre-existing renal impairment, the only independent predictor of CIN, can help triage candidates for angiographic investigation and therapy and identify patients in need of renal protection

Purpose: To compare placement of inferior vena cava filters (IVCF) in the fluoroscopy suite (FS) and in the operating room (OR) with regard to radiation exposure, filter position, and cost. Materials: All IVCF placements (8/13-12/14) at a single tertiary hospital center were identified through a picture archiving and communication system (PACS) engine search. The administered medications, procedure time, anesthesia services, radiation exposure, and filter type and angulation were recorded. Cost was estimated using information available in the hospital cost allocation system including time of procedure, procedure/recovery location, typically administered medications, and type of filter typically used in each area; the cost of anesthesiology consultation was excluded as it was unavailable in the allocation system. Results: In total, 254 IVCFs were placed in the FS and 47 filters in the OR. IVCF placements performed in the OR entailed radiation doses (156.3 mGy vs. 72.2 mGy, p=0.001) and fluoroscopy times (6.1 min vs. 2.8 min, p<0.0001) that were twice as high as those performed in the FS. Angulation of deployed filters from the central axis of the IVC was significantly less for those placed in the FS than for in the OR (2.6degree vs 4.8degree p<0.0001). There was a significant difference in the type of anesthesia used, with general anesthesia or monitored anesthesia care preferred in the OR (84%), whereas an anesthesiologist was involved in only 5% of FS cases. Most (92%) FS cases involved only a local anesthetic (p< 0.0001) and physiologic monitoring. Cost savings of FS cases compared with OR cases were estimated to be more than $488.77 per case. Conclusions: IVCF placement in the FS, compared to the OR, resulted in less exposure to radiation, more precise filter placement, and reduced cost

Purpose: To compare the antegrade trans-oral and the retrograde trans-abdominal approaches for fluoroscopy-guided percutaneous gastrostomy tube (G-tube) placement. Materials: Following institutional review board approval, all G-tubes placed by Interventional Radiology at two academic hospitals between January 2014 and May 2015 were reviewed retrospectively. Chart review determined type of anesthesia used during placement, dose of radiation used, fluoroscopy time, total procedure time, medical history, and complications. Results: A total of 149 patients (64 women, 85 men; mean age, 64 years, range 18-100 y) underwent G-tube placement, including 93 placed via the retrograde trans-abdominal approach and 56 placed via the antegrade trans-oral approach. Retrograde placement entailed fewer anesthesiology consultations (37% vs. 98%, p <0.0001), less overall procedure time (28 minutes vs. 36 minutes, p=0.02), and less fluoroscopy time (2.1 minutes vs. 8.7 minutes, p <0.0001). A comparison of approaches for placement within the same hospital demonstrated that the retrograde approach led to significantly reduced radiation dose (4.9 mGy vs. 28.1 mGy, p=0.02). There were no differences in complication rates between the two techniques, with minor and major complication rates of 13-19% and 6-7%, respectively, for both approaches. Conclusions: G-tube placement using the retrograde trans-abdominal approach is associated with less fluoroscopy and procedure times, radiation exposure, and need for anesthe-siology consultation with similar safety profile compared with the antegrade trans-oral approach. Additionally, it is proposed that decreased procedure time and anesthesiology consultation using the trans-oral approach are likely associated with reduced cost

Purpose: To identify factors predicting active extravasation on mesenteric and visceral angiography performed for intra-abdominal bleeding. Materials: All conventional angiograms performed (1/20136/2015) for intra-abdominal bleeding at one academic center were reviewed retrospectively for active contrast extravasation. A total 75 angiograms (46, gastrointestinal; 6, splenic; 4 hepatic; 12, uterine; and 7 other sources of intra-abdominal hemorrhages) were performed in 70 patients (26 women; 44 men) with mean age of 59 (range 21-92) yr. Variables were assessed over the 24 hr prior to angiography: percent change in hematocrit (ahct); urine output; vasopressor administration; units of packed red blood cells (pRBC) and volume of intravenous fluids (IVF) given; and mean arterial pressure (MAP) and pulse immediately preceding the study. Results: Of 75 exams, 20 (27%) showed contrast extravasation consistent with active bleeding and 50 (73%) did not. Multiple logistic regression showed ahct was the only independent predictor of contrast extravasation (p=0.017), whereas MAP, pulse, treatment with pressors, number of units of pRBC transfusion, and volume of IVF administered were not predictive. Patients with active extravasation had larger ahct compared to those without extravasation (-17% vs -1%; p=0.01). Findings were similar for the subset of 46 studies performed for gastrointestinal hemorrhage: ahct was the only independent predictor of extravasation (-21% vs -1%; p=0.049) (Table). Neither CT angiography nor endoscopy accurately predicted which studies had contrast extravasation (p>0.3). Conclusions: Greater decrease in hct over 24 hr prior to conventional angiography predicts active extravasation in patients with intra-abdominal and gastrointestinal bleeding. Patients with large hct declines should be triaged for rapid angiography to increase the likelihood of detecting a treatable abnormal vascular focus. (Table Presented)