Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print June 21, 2002, 10.1148/radiol.2242010847
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2242010847v1
224/2/436    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pelage, J.-P.
Right arrow Articles by Merland, J.-J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pelage, J.-P.
Right arrow Articles by Merland, J.-J.
(Radiology 2002;224:436-445.)
© RSNA, 2002

Experimental Studies

Uterine Artery Embolization in Sheep: Comparison of Acute Effects with Polyvinyl Alcohol Particles and Calibrated Microspheres1

Jean-Pierre Pelage, MD, Alexandre Laurent, MD, PhD, Michel Wassef, MD, Michel Bonneau, Delphine Germain, PhD, Roland Rymer, MD, Patrice Flaud, PhD, Jacques Martal, VMD and Jean-Jacques Merland, MD

1 From the LNAT (Centre de Recherche de l’Association Claude Bernard) rattaché au Laboratoire de Biorhéologie et Hydrodynamique Physiologique ESA-CNRS 7057, University of Paris, France (J.P.P., A.L., P.F., J.J.M.); Center for Research in Interventional Radiology (Cr2i, AP-HP/INRA), Jouy-en-Josas, France (J.P.P., A.L., M.B., D.G., J.J.M.); Departments of Body and Vascular Imaging (J.P.P., R.R.), Neuroradiology (A.L., J.J.M.), and Pathology (M.W.), Hôpital Lariboisière, AP-HP, 2 rue Ambroise Paré, 75475 Paris Cedex 10, France; and Department of Animal Physiology, INRA, Jouy-en-Josas, France (J.M.). Received April 26, 2001; revision requested June 15; final revision received December 17; accepted January 22, 2002. Supported in part by a grant from the Société Française de Radiologie (SFR). Address correspondence to J.P.P. (e-mail: jean-pierre.pelage@lrb.ap-hop-paris.fr).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
PURPOSE: To compare the effects on the myometrium of polyvinyl alcohol (PVA) particles and calibrated microspheres (MS) in embolization of the uterine arteries in sheep.

MATERIALS AND METHODS: Superselective and bilateral embolization of the uterine arteries was performed with PVA particles and calibrated MS within 24 hours after artificial ovulation in 26 adult nonpregnant sheep. PVA particles of four diameters, 150–250, 250–400, 400–600, and 600–1,000 µm, were compared with calibrated MS of similar diameters, 100–300, 300–500, 500–700, and 700–900 µm, in eight groups of sheep. Evaluation was based on histopathologic study of uterus, ovaries, and vascular pedicles after sacrifice 5 days after embolization. The scores of necrosis, the diameter of occluded arteries, and the number of particles were determined. The scores of uterine necrosis were compared by using nonparametric tests (Mann-Whitney U and Kruskal-Wallis). Spearman rank test was used for correlations.

RESULTS: PVA particles clumped more readily than did MS. Small particles had a higher score (P = .02) of uterine necrosis than did large particles. PVA particles produced more necrosis than did MS. Size of MS and diameter of occluded arteries showed significant correlation ({rho} = 0.762, P < .001). Size of PVA particles and diameter of occluded arteries showed no correlation. PVA particles occluded vessels of a wider range of size than did calibrated MS.

CONCLUSION: PVA particles are associated with intense uterine necrosis and extensive arterial occlusion regardless of size. Calibrated MS, which are associated with less uterine necrosis, permit a segmental arterial occlusion correlated with size.

© RSNA, 2002

Index terms: Animals • Arteries, chemotherapeutic embolization, 854.1264, 964.1264 • Arteries, uterine, 854.1264, 964.1264 • Experimental study • Interventional procedures, 854.1264, 964.1264 • Uterine neoplasms, therapeutic radiology, 854.1264, 964.1264


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Uterine fibroids represent the most common pelvic tumors in women. Usual medical treatment includes administration of progesterone or gonadotropin-releasing hormone agonists (1,2). Surgical resection, including myomectomy and hysterectomy, is often needed (3–5). In 1995, a group of radiologists and gynecologists (6) at Lariboisière Hospital in Paris, France, reported the successful preoperative use of uterine artery embolization to reduce bleeding during surgery. The procedure also led to shrinkage of the fibroids themselves. The technique was then performed as an alternative to surgery in the treatment of symptomatic fibroids, with encouraging midterm results (7). Since publication of this initial report, uterine fibroid embolization (UFE) with polyvinyl alcohol (PVA) particles has been performed worldwide (8–11). Clinical success rates and tumor size reduction are encouraging (8–11).

Despite its effectiveness, UFE is associated with a low rate of major ischemic complications. Uterine infection leading to hysterectomy, and even the death of a woman, has been reported (8,11,12) in 1%–2% of the procedures. Uterine necrosis has also been reported (9) in 2% of the procedures in a series of 53 patients. Transient or permanent amenorrhea has been reported (11) in up to 4% of the procedures. Nontargeted embolization, which can potentially result in uterine necrosis and ovarian infarction, also has been reported (11,13). Potential explanations for these complications are that the fibroids and surrounding myometrium have a common arterial supply and that there is a functional anastomosis between the ovaries and the uterus. These complications also may be related to the noncalibration of embolic particles that leads to uncontrolled devascularization (14). Therefore, there is a great need to improve the procedure with regard to the type and size of the embolic particles. The ideal embolic particle should achieve fibroid shrinkage without adverse effects on normal myometrium and without ischemic damage to the ovaries.

To evaluate the effects of embolic particles on the uterus and the ovaries, we used an animal model for uterine artery embolization. The ideal animal model in which to perform experimental uterine embolization should replicate the anatomy of the human female reproductive system.

During a feasibility study, we verified that the arterial blood supply to the uterus in a ewe was similar to that in a woman (15,16). From an anatomic point of view, the main blood supply to the uterus was provided by the uterine arteries (17). The uterine arteries arose from the anterior division of the internal iliac artery in all animals. Anastomosis between the uterine and the ovarian arteries was observed in one (14%) of seven animals (18). This frequency was similar to that reported in women (19). The diameter of the anastomosis was comparable to that of a woman (usually < 500 µm) in all animals. The cervicovaginal branches supplied the cervical part of the uterus as well as the vagina.

From a physiologic point of view, we found the same variations in uterine artery size as in clinical practice. Thus, an artificially controlled hormonal cycle was necessary to perform embolization of enlarged uterine arteries. This hormonal control allowed us to obtain an arterial size similar to that of a woman with uterine fibroids in all animals. In our study, arterial embolization was performed by using the same sheath, catheters, and guide wires as in clinical procedures (8,11).

The purpose of our study was to compare the effects on the myometrium of reference PVA particles and calibrated microspheres (MS) in experimental embolization of the uterine artery in sheep.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Study Protocol
All experiments were approved by the institutional animal care and use committee of the Center for Research in Interventional Radiology, Jouy-en-Josas, France, and were conducted according to European community rules of animal care (20). PVA particles (UltraIvalon; Nycomed, Paris, France) of four granulometric ranges, 150–250, 250–400, 400–600, and 600–1,000 µm, were compared with calibrated MS (Embosphere; Biosphere Medical, Louvres, France) of similar sizes, 100–300, 300–500, 500–700, and 700–900 µm.

Animal Model
The study was performed in 26 nonpregnant adult Pre-Alpes sheep (Cr2i-APHP/INRA; Center of Research in Interventional Radiology, Jouy-en-Josas, France) that had a mean weight of 59.6 kg ± 7.3 (SD) (range, 46–72 kg). There was no difference in weight between the groups of animals in which embolization was performed with PVA particles and those in which it was performed with MS (P = .590, Mann-Whitney U test).

The number of animals in which embolization was performed with each type and size of particles is summarized in Table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Type and Diameter of Embolic Particles and Diameter of Uterine Arteries According to Animal

 
As previously stated, hormonal control of the menstrual cycle was performed before the procedure. A sponge impregnated with 40 mg of a progestational agent (Gonolone; Laboratoire Intervet, Angers, France) was placed inside the vaginas of the sheep. Fourteen days later, the sponge was removed, and an intramuscular injection of 500 IU of pregnant-mare serum gonadotropin (Folligon; Laboratoire Intervet, Angers, France) was administered to achieve ovulation. Pelvic angiography and embolization of the uterine arteries was then performed within 24 hours after the injection of pregnant-mare serum gonadotropin.

Anesthesia
The sheep were not fed for 24 hours before the procedure. Anesthesia was induced by means of intramuscular injection of 15 mg per kilogram of body weight of thiopental sodium (Nesdonal; Specia Rhone-Poulenc, Paris, France). Each animal was placed in the supine position, intubated, anesthetized with a mixture of 1.5% halothane (Trofiels, Zug, Switzerland) and 98.5% oxygen (CFPO, Paris, France), and ventilated with a unit (Logic 0.5; Ohmeda, Steeton, England). End-tidal CO2 levels were measured continuously and maintained between 26 and 36 mm Hg with a monitor (N1000; Nellcor, Pleasanton, Calif). Peripheral arterial oxygen saturation, maintained at a level higher than 95%, was monitored with a probe applied to the ear. An electrocardiogram was used to continuously monitor each animal during the procedure.

Angiographic Procedure
Arterial embolization was performed by using the same sheath, catheters, and guide wires as were used in clinical procedures (8,11). A 4- or 5-F vascular sheath (Radifocus Introducer; Terumo, Tokyo, Japan) was placed into the common femoral artery by means of the standard Seldinger technique with sterile conditions. A 4-F pigtail catheter (Optitorque Radifocus; Terumo) was placed at the level of the iliac bifurcation to perform digital substraction aortography and to identify the ovarian and uterine arteries. Selective catheterization of the contralateral internal iliac artery and superselective catheterization of the uterine artery were performed by using a 4-F cobra-shaped catheter (Radifocus Angiographic Catheter; Terumo, Tokyo). Free-flow superselective embolization was then performed by one interventionalist (J.P.P.) by using the selected embolic agent.

All animals that had uterine arteries too small to be cannulated with a 4-F catheter were excluded from the study, and no embolization was performed in them. Catheterization of the ipsilateral internal iliac artery and superselective study of the uterine artery were performed before embolization of the second uterine artery after a Waltman loop was formed by using a previously described procedure (21,22). The diameter of the uterine artery was evaluated by means of visual observation with a three-grade scale as follows: grade 0, thin diameter; grade 1, intermediate diameter; and grade 2, enlarged diameter (Fig 1). When the diameter of the uterine artery was similar to the size of the ipsilateral internal iliac artery, the width was considered enlarged (grade 2).



View larger version (195K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. Angiograms in a ewe. (a) Aortogram demonstrates two enlarged (grade 3) uterine arteries (arrowheads). The left utero-ovarian anastomosis (arrow) is also seen. (b) Superselective angiogram of the left uterine artery.

 


View larger version (189K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. Angiograms in a ewe. (a) Aortogram demonstrates two enlarged (grade 3) uterine arteries (arrowheads). The left utero-ovarian anastomosis (arrow) is also seen. (b) Superselective angiogram of the left uterine artery.

 
Embolization Protocol
After superselective catheterization of the uterine artery, embolization was performed. The type or size of the particle was randomly selected before angiographic study. Embolization was performed in free flow, with the catheter placed 4 cm distally into the uterine artery at the level of the uterine arch (19). Each vial of PVA foam particles and MS was diluted in a mixture of 40 mL of 50% iodinated contrast medium (meglumine ioxitalamate, Telebrix 35; Laboratoire Guerbet, Aulnay-sous-Bois, France) and 50% saline. This mixture was diluted to a weaker concentration than that used in clinical practice (23).

Prior to each injection, the mixture was agitated to maintain homogeneous particle suspension. The injection of the mixture was performed slowly with fluoroscopic control by using 2.5-mL syringes. After each injection of particles, the catheter was purged with 2.5 mL of saline. Embolization was stopped when proximal arterial flow was reduced at angiographic evaluation. The same end point of embolization was used for the two types of particles. The number of vials used was noted on a data sheet. The quantity of embolic agent for each procedure was evaluated. Each vial of PVA contained 1.0 mL of particles, whereas each vial of MS contained 0.6 mL of particles.

The same procedure was performed for both uterine arteries. The injectability of the particles through the catheter was evaluated during the procedure. The degree of resistance at injection was subjectively evaluated by the interventional radiologist (J.P.P.) with the following three-grade scale: grade 0, saline-like injection; grade 1, resistance; and grade 2, catheter occlusion.

Postembolization Follow-up
Veterinary follow-up evaluation was performed for each ewe before sacrifice. In accordance with the protocol, if the animal exhibited any abnormal behavior or was in pain, it would be sacrificed immediately.

Sacrifice
Five days after embolization, laparotomy along the alba linea and pelvic exploration were performed. Without knowledge of the particle type or size, the clinician visually inspected the genital tract, the bladder, and the gastrointestinal tract to search for necrosis. Removal of the uterus, the ovaries, and the afferent arterial pedicles (ovarian, uterine, and vaginal arteries) was then performed before the animals were sacrificed with an overdose of 60 mg/kg of pentobarbital.

Magnetic Resonance Imaging Studies
Magnetic resonance (MR) examinations were performed with a 0.23-T imager (Outlook; Marconi-Picker, Cleveland, Ohio) and a surface coil by two radiologists (J.P.P., D.G.). The gross specimens removed at hysterectomy and ovariectomy were examined by using transverse and coronal T2-weighted spin-echo images (2,500/100 [repetition time msec/echo time msec], 5–8-mm section thickness, 244 x 325 field of view). Acquired images were compared with gross specimens removed at hysterectomy and with transverse sections of the uterus. Evaluation was based on the signal intensity of the myometrium observed during our feasibility study, and imaging findings were compared with pathologic findings (15,16). Normal myometrium manifested an intermediate signal intensity, whereas necrosis was associated with a heterogeneous (patchy) or homogeneous high signal intensity. A visual evaluation of the extent of necrosis was made with the following three-grade scale: grade 0, grade 1, and grade 2 corresponded to absence of necrosis, moderate necrosis (which affected less than 50% of the surface), and intense necrosis (which affected more than 50% of the surface), respectively.

On MR images, ovarian necrosis (ie, diffuse high signal intensity of the ovary), which was observed as different from the localized high signal intensity of the physiologic follicle, was evaluated. Each MR image was reviewed without knowledge of the type or diameter of embolization particle used.

Histopathologic Study
The evaluation of gross specimens removed at hysterectomy and ovariectomy was based on an abnormal brown area, which corresponded to an area of necrosis, at external inspection. The extent of necrosis was based on a three-grade scale: grade 0, grade 1, and grade 2, which respectively corresponded to absent; moderate, which affected less than 50% of the surface; and extensive, which affected more than 50% of the surface, necrosis. Transverse 10-mm-thick uterine sections were then prepared. In the normal uterus, the inner mucosal part manifests a yellow and homogeneous color. A clear delineation of the serosa and of the mucosal parts with the myometrium was visible. The normal myometrium has a white and homogeneous color. External serosa is thin and homogeneous. During the histologic feasibility study, we verified that the brown color of the uterus corresponded to necrosis (15,16). Analysis was performed by two observers (J.P.P., M.W.) who worked on a consensus basis, and they had no knowledge of the particle type.

Each transverse section was obtained for each uterine region (ie, the body, the cervix, and the horns) (Fig 2). The number of sections was dependent on the uterine size. Each section was then divided radially into eight parts. The mucosal part was separated from the myometrium, and this separation permitted the individualization of 16 areas. For each area, presence or absence of necrosis was assigned grade 1 or 0, respectively. For each transverse section, the score was the total number of areas of necrosis assigned grade 0 to 1 for the 16 areas. Three uterine necrosis scores were used: (a) Total score of necrosis was the percentage of necrosis corresponding to the total number of areas of necrosis of all the transverse sections for the total number of areas for the entire uterus. (b) Transmural score of necrosis represented necrosis involving the entire uterine wall from the cavity to the serosa. (c) Score of necrosis for each anatomic region was also evaluated for the horns, the cervix, and the body of the uterus with regard to arterial anastomoses between uterine, ovarian, and vaginal arteries.



View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Gross specimen removed at hysterectomy and ovariectomy performed 5 days after embolization in a ewe demonstrates the appearance of the uterine regions. B = body, C = cervix, H = horns, O = ovaries.

 
All the resected tissues were routinely fixed in formaldehyde and processed by means of paraffin embedding. Standard hematoxylin-eosin–saffron staining was then performed, and histopathologic analysis was performed by the same pathologist (M.W.) and the same radiologists (J.P.P., A.L.), and they worked on a consensus basis without knowledge of the embolic particle used.

Histologic analysis was performed in each of the previously described anatomic regions (ie, the cervix, the body, and the horns). The presence or absence of necrosis was noted.

In each slice obtained, each artery was studied. The following elements were evaluated: presence and size of embolic particles, diameter of the occluded artery, and number of particles in the arterial lumen. Measurement of the size of the artery and of the particles was precisely determined by using a calibrated ocular microscope with graduations of 10 µm. The distribution of embolic particles according to uterine layer (ie, the perimyometrium, the myometrium, and the endometrium) was also recorded.

Statistical Analysis
Summary descriptive statistics were used, and values were the means and SDs. In relevant cases, the median and range (minimum to maximum) were determined. Scores of uterine necrosis for the particle types and sizes were compared by using nonparametric tests. The Mann-Whitney U test was used to compare two groups with a continuous variable (eg, PVA particles versus MS), and the Kruskal-Wallis test was used for comparison among more than three groups (eg, scores of uterine necrosis according to the four sizes of embolic agents).

We used a Spearman rank correlation test to determine whether the degree of arterial occlusion (ie, diameter of occluded arteries) was related to size of the particle for each type of embolic agent. We compared the scores of uterine necrosis of small particles (ie, PVA particles < 400 µm and MS < 500 µm in diameter) and large particles (ie, PVA particles > 400 µm and MS > 500 µm in diameter) by using the Mann-Whitney U test. The difference in diameter between small and large particles was based on granulometries provided by the manufacturing companies. Statistical analysis was carried out by using software (Statview 512; SAS Institute, Cary, NC).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Embolization
Superselective and bilateral catheterization and free-flow embolization of the uterine arteries were achieved in all animals. No difference in arterial diameter was observed in the group of animals in which embolization was performed with MS compared with the group in which embolization was performed with PVA particles, and no difference was observed between the left and the right uterine arteries (Table 1).

The total number of vials necessary to obtain arterial occlusion with MS (2.8 ± 0.5; range, 2.0–4.0) was different from that which was necessary to obtain arterial occlusion with PVA particles (1.6 ± 0.5; range, 1.0–2.5) (P < .001, Mann-Whitney U test), but the total quantity of embolic agent was not different with both PVA particles and MS.

No catheter occlusion was observed with the calibrated MS. During embolization with MS of 700–900 µm, grade 1 injection was noticed in one animal (Table 2).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Ease of Injection for Each Type of Embolic Particle

 
During embolization with PVA particles, catheter occlusion (grade 2) was observed in two of four, two of three, one of three, and two of three animals with particles of 150–250, 250–400, 400–600, and 600–1,000 µm, respectively (Table 2).

Veterinary Evaluation
No abnormal behavior or pain was noticed after embolization during the follow-up period.

Surgical Findings
The bladder and the gastrointestinal tract appeared to be normal in all animals.

Uterine Necrosis
Total score of necrosis.—The total score of necrosis was higher with PVA particles than with MS (P < .004) (Fig 3). The results are shown in Figure 4. The uterine necrosis obtained with PVA particles and MS considered in the aggregate was dependent on the size of the particles (P < .003, Kruskal-Wallis test). The uterine necrosis obtained with PVA particles also depended on the size of the particles (P < .013, Kruskal-Wallis test). The uterine necrosis obtained with MS differed according to the diameter of the particle (P < .02, Kruskal-Wallis test).



View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Transverse section of the uterus obtained in a ewe after embolization performed with 150-250-µm PVA particles. Each section is divided into eight inner (mucosal area) and eight outer (myometrial area) parts. Total score of uterine necrosis (brown areas) for this section is 13 (81%) of 16 parts and transmural score of necrosis is five (62%) of eight parts.

 


View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Total score of uterine necrosis (mean ± SD) in sheep. Graphs show comparison of sizes (Kruskal-Wallis test) and of PVA particles ({square}) and MS ({circ}) (Mann-Whitney U test). The total score of uterine necrosis depended on the particle size and was higher with PVA particles than with MS.

 
Uterine necrosis was related to the size of the MS. The uterine necrosis was less than 5% of the uterus with large MS (ie, with a diameter > 700 µm), whereas more necrosis was observed with small MS. A similar score of uterine necrosis was observed with 600–1,000-µm PVA particles and with 300–500-µm MS (Fig 4).

Small particles (ie, MS of < 500 µm and PVA particles of < 400 µm) were associated with more necrosis than were large particles (P > .001, Mann-Whitney U test). Small MS were associated with more necrosis than were large MS (P < .003, Mann-Whitney U test). The same findings were observed with small versus large PVA particles (P < .003, Mann-Whitney U test). These results are shown in Figure 5. Small MS were associated with less necrosis than were small PVA particles (P > .018). The same findings were observed with large particles (P < .004) (Figs 5–7).



View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Total score of uterine necrosis (mean ± SD) in sheep. Graph shows comparison of small (ie, PVA particles [gray bars] < 400 µm and MS [white bars] < 500 µm) and large (ie, PVA particles > 400 µm and MS > 500 µm) particles and comparison of PVA particles and MS (Kruskal-Wallis test). The total score of uterine necrosis was higher with PVA particles than with MS for small and large particles.

 


View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6a. Results of embolization performed with 150-250-µm PVA particles in a ewe. (a) Gross specimen removed at hysterectomy demonstrates reddish heterogeneous area that corresponds to area of necrosis. (b) Coronal T2-weighted fast spin-echo MR image (2,000/80; flip angle, 90°) of the gross specimen demonstrates the patchy high signal intensity of the myometrium, which corresponds to an area of necrosis. Localized high signal intensity of ovarian cyst (arrowhead) is also identified. (c) Transverse sections of the uterus show extensive brown areas that correspond to areas of necrosis. Foci (arrowheads) of transmural necrosis are identified.

 


View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6b. Results of embolization performed with 150-250-µm PVA particles in a ewe. (a) Gross specimen removed at hysterectomy demonstrates reddish heterogeneous area that corresponds to area of necrosis. (b) Coronal T2-weighted fast spin-echo MR image (2,000/80; flip angle, 90°) of the gross specimen demonstrates the patchy high signal intensity of the myometrium, which corresponds to an area of necrosis. Localized high signal intensity of ovarian cyst (arrowhead) is also identified. (c) Transverse sections of the uterus show extensive brown areas that correspond to areas of necrosis. Foci (arrowheads) of transmural necrosis are identified.

 


View larger version (94K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6c. Results of embolization performed with 150-250-µm PVA particles in a ewe. (a) Gross specimen removed at hysterectomy demonstrates reddish heterogeneous area that corresponds to area of necrosis. (b) Coronal T2-weighted fast spin-echo MR image (2,000/80; flip angle, 90°) of the gross specimen demonstrates the patchy high signal intensity of the myometrium, which corresponds to an area of necrosis. Localized high signal intensity of ovarian cyst (arrowhead) is also identified. (c) Transverse sections of the uterus show extensive brown areas that correspond to areas of necrosis. Foci (arrowheads) of transmural necrosis are identified.

 


View larger version (153K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7a. Results of embolization performed with 100-300-µm MS in a ewe. (a) Gross specimen removed at hysterectomy demonstrates few reddish areas (arrowheads), which correspond to areas of necrosis. (b) Coronal T2-weighted fast spin-echo MR image (2,000/80; flip angle, 90°) of the gross specimen demonstrates heterogeneous signal intensity of the myometrium, which corresponds to an area of necrosis. (c) Transverse sections of the uterus show numerous brown areas that correspond to areas of necrosis.

 


View larger version (191K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7b. Results of embolization performed with 100-300-µm MS in a ewe. (a) Gross specimen removed at hysterectomy demonstrates few reddish areas (arrowheads), which correspond to areas of necrosis. (b) Coronal T2-weighted fast spin-echo MR image (2,000/80; flip angle, 90°) of the gross specimen demonstrates heterogeneous signal intensity of the myometrium, which corresponds to an area of necrosis. (c) Transverse sections of the uterus show numerous brown areas that correspond to areas of necrosis.

 


View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7c. Results of embolization performed with 100-300-µm MS in a ewe. (a) Gross specimen removed at hysterectomy demonstrates few reddish areas (arrowheads), which correspond to areas of necrosis. (b) Coronal T2-weighted fast spin-echo MR image (2,000/80; flip angle, 90°) of the gross specimen demonstrates heterogeneous signal intensity of the myometrium, which corresponds to an area of necrosis. (c) Transverse sections of the uterus show numerous brown areas that correspond to areas of necrosis.

 
Transmural score of necrosis.—The score of necrosis was higher with PVA particles compared with the score with MS (P < .001, Mann-Whitney U test) (Fig 3).

The score of necrosis was higher with small particles (PVA particles or MS) than it was with large particles (P < .004, Mann-Whitney U test). More necrosis was observed with small MS than with large MS (P > .038, Mann-Whitney U test). More necrosis was observed with small PVA particles than with large PVA particles (P > .032, Mann-Whitney U test). Small MS were associated with less necrosis than were small PVA particles (P = .006). The same findings were observed for large particles (P > .016).

Uterine necrosis in each anatomic region.—More necrosis was observed in the groups in which embolization was performed with PVA particles than in those in which embolization was performed with MS for each size of particles. Small particles were associated with more necrosis than were large particles in the cervix, the body, and the horns, with P values of less than .072, less than .001, and equal to .035, respectively (Mann-Whitney U test).

Correlation between scores of necrosis.—The total score of necrosis and the transmural score of necrosis correlated well for the two types of particles ({rho} = 0.819, P < .001, Spearman rank test). The correlation was higher with MS ({rho} = 0.831, P = .004) than with PVA particles ({rho} = 0.698, P < .016).

The total uterine necrosis was well correlated with the necrosis of the body for the two types of particles ({rho} = 0.947, P < .001, Spearman rank test).

Effects on the Ovaries
No ovarian necrosis was observed with PVA particles or with MS. In one ovary of a ewe in which embolization was performed with MS of 100–300 µm, three MS were observed (Fig 8). In one ovary of a ewe in which embolization was performed with MS of 300–500 µm, one particle was observed.



View larger version (189K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8. Results of embolization performed with 100-300-µm MS in a ewe. Photomicrograph of the histologic aspect of the left ovary. Three MS (arrowheads) are identified. No ovarian necrosis is observed. (Hematoxylin-eosin-saffron stain; original magnification, x6.)

 
MR Imaging Studies
More myometrial necrosis was observed with PVA particles than was observed with MS (P > .047, contingency table). The signal intensity of the ovaries was normal in all animals.

Location of Particles in Arterial Network
The total number of occluded vessels was 790 (ie, n = 600 for those in which MS were used and n = 190 for those in which PVA particles were used). Statistical values regarding diameter of occluded arteries are shown in Table 3.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Diameter of Occluded Arteries

 
The size of the arteries in which PVA particles and MS were located inside the arterial network differed (P < .001, Kruskal-Wallis test) (Fig 9). The diameter of occluded arteries was dependent on the particle size of MS (P < .001). Small MS were located more distally even into the endometrium (vessel diameter range, 100–500 µm), whereas large MS were located more proximally.



View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9. Location of the particles in the arterial network in sheep. Graph shows comparison of PVA particles (gray bars) and MS (white bars) (Mann-Whitney U test) and comparison of sizes (Kruskal-Wallis test). The diameter of occluded arteries depended on the particle size for MS but not for PVA.

 
Conversely, no effect of size on location was observed with PVA particles. All sizes of PVA particles were located both proximally in aggregates and distally (vessel diameter range, 300–1,000 µm). There was a wider range of occluded vessels with PVA particles compared with the range with MS. For instance, 150–250-µm PVA particles were observed in vessels ranging from 50 to 1,300 µm in diameter, whereas 100–300-µm MS were observed in arteries 50–700 µm in diameter (Figs 10, 11).



View larger version (162K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10a. Results of embolization performed with 150-250-µm PVA particles in a ewe. (a) Photomicrograph of uterine section. Small fragments (arrow) of PVA particles were identified in distal arteries inside the endometrium. Diffuse areas of necrosis (arrowheads) are seen. (Hematoxylin-eosin-saffron stain; original magnification, x4.) (b) Photomicrograph of uterine section obtained in the same animal. PVA particles (arrows) are observed proximally in aggregates inside perimyometrial arteries. (Hematoxylin-eosin-saffron stain; original magnification, x4.)

 


View larger version (154K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10b. Results of embolization performed with 150-250-µm PVA particles in a ewe. (a) Photomicrograph of uterine section. Small fragments (arrow) of PVA particles were identified in distal arteries inside the endometrium. Diffuse areas of necrosis (arrowheads) are seen. (Hematoxylin-eosin-saffron stain; original magnification, x4.) (b) Photomicrograph of uterine section obtained in the same animal. PVA particles (arrows) are observed proximally in aggregates inside perimyometrial arteries. (Hematoxylin-eosin-saffron stain; original magnification, x4.)

 


View larger version (144K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 11. Results of embolization performed with 100-300-µm MS. Photomicrograph of uterine section. MS (arrowheads) are identified in distal arteries inside the endometrium. * = uterine cavity. (Hematoxylin-eosin-saffron stain; original magnification, x4.)

 
Number of Particles in Occluded Arteries
The total number of particles observed in occluded arteries was 1,349 for those in which MS were used and 2,311 for those in which PVA particles were used for embolization. The mean number of PVA particles in each occluded artery was higher than that of MS (P < .001, Mann-Whitney U test). The mean number of MS observed in occluded arteries was 2.2, whereas that of PVA particles was 12.2. The number of particles in occluded arteries varied more with PVA particles (range, 1–100; median, 8) than with MS (range, 1–43; median, 1) (Fig 12).



View larger version (27K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 12. Number of particles in occluded arteries (mean ± SD). Graph shows comparison of PVA particles (gray bars) and MS (white bars) (Mann-Whitney U test) and comparison of sizes (Kruskal-Wallis test). The number of particles in occluded arteries was higher with use of PVA particles (median, eight particles) than with use of MS (median, one particle).

 
Quantity of Particles according to Uterine Layer
The distribution of PVA particles and MS demonstrated a statistically significant difference. PVA particles were observed more frequently inside the perimyometrial (P < .001) and myometrial (P < .001) arteries than were MS. The quantity of embolic particles differed according to the layer within the uterus.

Correlation between Particle Size and Arterial Diameter
For MS, there was a significant correlation between the size of the particle and the arterial diameter ({rho} = .842, P <.001, Spearman rank test). For PVA particles, no correlation was observed.

Correlation between Number of Particles and Arterial Diameter
There was a significant correlation between the number of particles and the arterial diameter for PVA particles ({rho} = 0.688, P < .001) and a low correlation for MS ({rho} = 0.485, P < .001).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
UFE is a therapeutic option in the management of symptomatic fibroids. UFE is an effective procedure, but several investigators have reported two types of severe ischemic complications: (a) extensive uterine necrosis and infection, which leads to hysterectomy, and (b) amenorrhea after the procedure (8,9,11,12). Although PVA particles are the most common embolic agent used to perform UFE, to the best of our knowledge, no published report describes experimental embolization of the uterus with PVA particles or other embolic agents (8–12). PVA particles have been widely used as an embolic agent in various territories for 20 years (24).

Despite recent improvement of the calibration, the irregular shape of the material is associated with a large granulometric size range of the particles (14,25). In addition, there is a risk of injection of particles smaller than the theoretic size in conjunction with aggregates associated with obstruction of the catheter (25–27). The potential adverse effects of PVA particles have been reported (14). Even if ischemic complications of UFE seem to be observed in the case of large-diameter fibroids, in a submucosal or subserosal location, it may be hypothesized that the size of PVA particles used may account for these complications (11).

In our study, the degree of penetration of the particles into the vascular system was different for PVA and MS. PVA particles were associated with an extensive arterial occlusion involving distal and proximal arteries simultaneously. In addition, PVA particles tended to clump and form aggregates within the catheters and within the vessels. Arterial occlusion may occur with all sizes of particles, regardless of the dilution of the particles. We believe this is a hazardous arterial occlusion, since the range of occluded arteries is the same with all diameters of PVA particles.

In our study, the level of arterial occlusion was not correlated with the size of the PVA particles. The advertised size of a PVA particle is primarily a function of its intermediate axis. It is the reason for a wide granulometric size range of the particles, despite the theoretic size of the particles (25). Even if one type of PVA particle was used in the present study, there would be only minor differences between the particles supplied by the manufacturers because of recent changes in the manufacturing process. These findings already have been reported in women treated with preoperative embolization (28,29).

Conversely, calibrated MS are associated with a more controlled arterial occlusion (30,31). A significant correlation between the level of arterial occlusion and the diameter of the particles was observed for all sizes of MS. Thus, proximal occlusion can be achieved precisely with large-diameter (>700 µm) MS, whereas distal occlusion can be achieved with small (<500 µm) MS. A small number (less than three) of MS were found in embolized arteries, and this finding accounted for accurate segmental occlusion. A deeper penetration of embolic particles can be advantageous and can lead to a more effective tumoral devascularization of many hypervascularized lesions, including meningiomas (32,33). In other situations, the size of the particles should be carefully selected according to the level of occlusion to be achieved. This is of major importance in clinical practice.

During UFE, there is a potentially dangerous anastomosis between the uterine and the ovarian arteries, and the presence of this anastomosis compels the clinician to focus attention on choosing the correct particle size (19,34). Ovarian infarction and subsequent amenorrhea may be related to nontargeted embolization of the ovaries. From our experimental model, we know that particles smaller than 500 µm can pass through a patent utero-ovarian anastomosis. From clinical reports, we also know that small fragments of PVA particles can be observed inside the ovaries (35). With all types of embolic particles, there is a potential risk of ovarian embolization.

In our study, small (MS or PVA) particles were associated with more uterine necrosis than were large particles. It has been hypothesized that the lower rate of pelvic pain reported by the U.S. groups of investigators was related to the large diameter of PVA particles compared with that used by the French investigators (8). The same hypothesis was used to explain ischemic complications reported by the researchers from Paris who were using small PVA particles (7,11). In fact, several cases of extensive uterine necrosis after embolization with larger PVA particles were reported by Vashisht et al (12) and by an author of this investigation (J.P.P., unpublished data, 1999).

We observed significant differences in the extent of uterine necrosis that occurred in embolization performed with calibrated MS and PVA particles of similar size. There was a significant correlation between the diameter of MS and the extent of uterine necrosis. Small MS were associated with intense uterine necrosis, whereas large MS were associated with less necrosis. Conversely, more intense uterine necrosis was observed with PVA particles. Even medium-sized, that is, 250–400- and 400–600-µm, PVA particles were associated with more than 40% and 25% of uterine necrosis, respectively. The length of arterial occlusion may account for intense necrosis with large PVA particles. Additional proximal occlusion may have additional ischemic effect in the state of widespread distal embolization.

In conclusion, our experimental model provides an assessment of the physiopathologic uterine response to embolization. PVA particles are associated with more intense uterine necrosis than are calibrated MS. Calibrated MS permit more segmental arterial occlusion than do PVA particles. To prove that calibrated MS are advantageous in uterine artery embolization, the ideal range of vessel diameters to be occluded (ie, one that is safe for normal myometrium and ovaries and that will cause effective fibroid shrinkage) should exist. Preliminary clinical trials conducted in various countries are encouraging, but further clinical studies should be conducted to find the optimal balance, in terms of particle size and type, between safety and efficacy.

Practical application: Choosing a size of calibrated MS solely on the basis of experience with PVA particles could result in marked complications. Since the level of arterial occlusion and the extent of uterine necrosis correlate well with MS, the investigator should carefully choose the optimal size for a specific procedure. Our preliminary clinical results of UFE with calibrated MS larger than 500 µm are encouraging (36,37).

Dilution of calibrated MS is of paramount importance to precisely fill the arterial network and derive the benefit from the precise calibration of the embolic agent (38,39).

With regard to improvement of the safety of the UFE procedure for the ovaries and the control of uterine necrosis, the use of small MS is potentially dangerous because of the risks of uterine necrosis and ovarian ischemia. MS larger than 500 µm can permit the occlusion of the perifibroid vascular plexus and, therefore, should reasonably be associated with reduced nontargeted embolization.


   ACKNOWLEDGMENTS
 
The authors thank Jean-Pierre Albert, Melanie Bazire, Nathalie Blanck, Christian Bourgeois, Valerie Scotto, and Natacha Trofleau for their helpful assistance.


   FOOTNOTES
 
Abbreviations: MS = microspheres, PVA = polyvinyl alcohol, UFE = uterine fibroid embolization

Author contributions: Guarantors of integrity of entire study, J.P.P., A.L.; study concepts and design, J.P.P., A.L.; literature research, J.P.P., J.J.M.; experimental studies, J.P.P., M.W., M.B.; data acquisition, J.P.P., D.G., M.B.; data analysis/interpretation, J.P.P., A.L., P.F., M.W.; statistical analysis, A.L.; manuscript preparation, J.P.P., A.L.; manuscript definition of intellectual content, J.J.M., R.R.; manuscript editing, J.P.P., R.R.; manuscript revision/review, A.L., J.J.M.; manuscript final version approval, J.P.P., A.L., J.J.M.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 

  1. Vollenhoven BJ, Lawrence AS, Healy DL. Uterine fibroids: a clinical review. Br J Obstet Gynecol 1990; 97:285-298.[Medline]
  2. Healy DL, Vollenhoven BJ. The role of GnRH agonists in the treatment of uterine fibroids. Br J Obstet Gynaecol 1992; 99:23-26.
  3. Wallach EE. Myomectomy. In: Thompson JD, Rock CA, eds. Te Linde’s operative gynecology. 7th ed. Philadelphia, Pa: Lippincott, 1992; 647-662.
  4. Carlson KJ, Nichols DH, Schiff I. Indications for hysterectomy. N Engl J Med 1993; 328:856-860.[Free Full Text]
  5. Iverson RE, Chelmow D, Strohbehn K, Waldman L, Evantash EG. Relative morbidity and mortality of abdominal hysterectomy and myomectomy for management of uterine leiomyomas. Obstet Gynecol 1996; 88:415-419.[CrossRef][Medline]
  6. Ravina JH, Bouret JM, Fried D, et al. Intérêt de l’embolisation pré-opératoire des fibromes utérins: à propos d’une série multicentrique de 31 cas. Contracept Fertil Sex 1995; 23:45-49.[Medline]
  7. Ravina JH, Herbreteau D, Ciraru-Vigneron N, et al. Arterial embolisation to treat myomata. Lancet 1995; 346:671- 672.[CrossRef][Medline]
  8. Goodwin SC, Vedantham S, Mc Lucas B, Forno Perella R. Preliminary experience with uterine artery embolization for uterine fibroids. J Vasc Interv Radiol 1997; 8:517-526.[Medline]
  9. Worthinghton-Kirsch RL, Popky GL, Hutchins SL, Jr. Uterine arterial embolization for the management of fibroids: quality-of-life assessment and clinical response. Radiology 1998; 208:625-629.[Abstract/Free Full Text]
  10. Spies JB, Scialli AR, Jha RC, et al. Initial results from uterine fibroid embolization for symptomatic leiomyomata. J Vasc Interv Radiol 1999; 10:1149-1157.[Medline]
  11. Pelage JP, Le Dref O, Soyer P, et al. Fibroid-related menorrhagia: treatment with superselective embolization of the uterine arteries and mid-term follow-up. Radiology 2000; 215:428-431.[Abstract/Free Full Text]
  12. Vashisht A, Studd J, Carey A, Burn P. Fatal septicaemia after fibroid embolisation. Lancet 1999; 354:307-308.[Medline]
  13. Holmgren B. Some observations on the blood vessels of the uterus under normal conditions and in myoma. Acta Obstet Gynecol Scand 1938; 18:192-213.
  14. Repa I, Moradian GP, Dehrer LP, et al. Mortalities associated with use of a commercial suspension of polyvinyl alcohol. Radiology 1989; 170:395-399.[Abstract/Free Full Text]
  15. Pelage JP, Laurent A, Wassef M, Bonneau M, Rymer R, Merland JJ. Uterine artery embolization: choice of an embolic particle (abstr). J Vasc Interv Radiol 2000; 11:189.[Medline]
  16. Farrer-Brown G, Beilby JO, Tarbit MH. The blood supply of the uterus: arterial vasculature. J Obstet Gynaecol Br Com 1970; 77:673-681.[Medline]
  17. Barone R. Anatomie comparée des mammifères domestiques Paris, Vigot: Splanchnologie, 1976; 4.
  18. Janson PO, Williams D, Petrucco OM, Amato F, Seamark RF, Findlay JK. Blood flow in the ovary and adjacent structures of the non-pregnant sheep. Acta Endocrinol (Copenh) 1983; 103:259-265.
  19. Pelage JP, Le Dref O, Soyer P, et al. Arterial anatomy of the female genital tract: variations and relevance to transcatheter embolization of the uterus. AJR Am J Roentgenol 1999; 172:989-994.[Free Full Text]
  20. Directive CEE 86/609. J Officiel Communautés Européennes Dec 18 1986; L358:1-28.
  21. Waltman AC, Courey WR, Athanasoulis C, Baum S. Technique for left gastric artery catheterization. Radiology 1973; 109:732-734.[Medline]
  22. Pelage JP, Soyer P, Le Dref O, et al. Uterine arteries: bilateral catheterization with a single femoral approach and a single 5-F catheter. Radiology 1999; 210:573-575.[Abstract/Free Full Text]
  23. Goodwin SC, Lai AC. Uterine embolization technique (abstr). J Vasc Interv Radiol 2000; 11:26-32.
  24. Kerber CW, Bank WO, Horton JA. Polyvinyl alcohol foam: prepackaged emboli for therapeutic embolization. AJR Am J Roentgenol 1978; 130:1193-1194.[Medline]
  25. Derdeyn CP, Moran CJ, Cross DTAI, Dietrich HH, Dacey RG, Jr. Polyvinyl alcohol particle size and suspension characteristics. AJNR Am J Neuroradiol 1995; 16:1031-1036.[Abstract]
  26. Barr JD, Lemley TJ, Petrochko CN. Polyvinyl alcohol foam particle sizes and concentrations injectable through microcatheters. J Vasc Interv Radiol 1998; 9:113-118.[Medline]
  27. Siskin G, Eglander M, Stainken BF, Ahn J, Dowling K, Dolen EG. Embolic agents used for uterine fibroid embolization. AJR Am J Roentgenol 2000; 175:767-773.[Free Full Text]
  28. Aziz A, Petrucco OM, Makinoda S, et al. Transarterial embolization of the uterine arteries: patient reactions and effects on uterine vasculature. Acta Obstet Gynecol 1998; 105:235-240.
  29. Kardache M, Soyer P, Le Dref O, Pelage JP, Abitbol M, Rymer R. Uterine leiomyoma: histopathological features after superselective transcatheter embolization of the uterine arteries (abstr). AJR Am J Roentgenol 1998; 170(suppl):40.
  30. Derdeyn CP, Graves VB, Salamat MS, Rappe A. Collagen-coated acrylic microspheres for embolotherapy: in vivo and in vitro characteristics. AJNR Am J Neuroradiol 1997; 18:647-653.[Abstract/Free Full Text]
  31. Laurent A, Beaujeux R, Wassef M, Rüfenacht D, Boschetti E, Merland JJ. Trisacryl gelatin microspheres for therapeutic embolization. I. Development and in vitro evaluation. AJNR Am J Neuroradiol 1996; 17:533-540.
  32. Bendszus M, Klein R, Burger R, Warmuth-Metz M, Hofmann E, Solymosi L. Efficacy of trisacryl gelatin microspheres and polyvinyl alcohol particles in the preoperative embolization of meningiomas. AJNR Am J Neuroradiol 2000; 21:255-261.[Abstract/Free Full Text]
  33. Beaujeux R, Laurent A, Wassef M, et al. Trisacryl gelatin microspheres for therapeutic embolization. II. Clinical evaluation in tumors and arteriovenous malformations. AJNR Am J Neuroradiol 1996; 17:541-548.
  34. Siddle N, Sarrel P, Whitehead M. The effect of hysterectomy on the age at ovarian failure: identification of a subgroup of women with premature loss of ovarian function and literature review. Fertil Steril 1987; 47:94-100.[Medline]
  35. Walker WJ, Green A, Sutton C. Bilateral uterine artery embolisation for myomata: results, complications and failures. Minim Invasive Ther Allied Technol 1999; 8:449-454.
  36. Pelage JP, Beregi JP, Nonent M, et al. Uterine fibroid embolization using calibrated microspheres (abstr). Cardiovasc Intervent Radiol 2000; 23S:104.
  37. Spies JB, Benenati JB, Worthington-Kirsch R, Pelage JP. Initial experience using trisacryl gelatin microspheres for uterine artery embolization for leiomyomata. J Vasc Interv Radiol 2001; 12:1059-1063.[Medline]
  38. Dion JE. Dextran microspheres embolization: experimental and clinical experience with pathologic correlation. Radiology 1986; 160:717-721.[Abstract/Free Full Text]
  39. Kai Y, Hamada JI, Morioka M, Todaka T, Hasegawa S, Ushio Y. The utility of microcrystalline cellulose sphere as a particulate embolic agent: an experimental study. AJNR Am J Neuroradiol 2000; 21:1160-1163.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2242010847v1
224/2/436    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pelage, J.-P.
Right arrow Articles by Merland, J.-J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pelage, J.-P.
Right arrow Articles by Merland, J.-J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE