| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
1 From the Division of Abdominal Imaging and Intervention, Massachusetts General Hospital, 55 Fruit St, White 270, Boston, MA 02114. From the 2001 RSNA scientific assembly. Received August 1, 2002; revision requested October 1; revision received December 15; accepted January 27, 2003. Address correspondence to M.G.H. (e-mail: mharisinghani@partners.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The medical records of 140 patients who underwent percutaneous CT-guided transgluteal drainage of 154 deep pelvic abscesses were reviewed to determine the origins, location, and size of the abscesses; volume of initial aspirate; organisms identified in fluid culture; duration of catheter drainage; incidence of catheter-related pain and procedure-related complications; and short- and long-term outcomes. The resultant data were analyzed with a Fisher exact test for difference in the incidence of postprocedural catheter-site pain between transpiriformis and infrapiriformis approaches.
RESULTS: The origins of the pelvic abscesses included postoperative fluid collection (n = 115), perforating appendicitis (n = 6), diverticulitis (n = 16), tubo-ovarian inflammation (n = 5), Crohn disease (n = 10), and internal bowel fistula due to irradiation (n = 2). The abscesses were 4–12 cm in diameter. The volume of the aspirate was 5–310 mL. Laboratory cultures of the aspirate grew mixed flora, but the organism most frequently isolated was Escherichia coli. Catheters were removed after a mean of 8 days. In 134 (96%) of 140 patients, there was complete resolution of the abscess following transgluteal drainage, without subsequent surgery. In six of 140 (4%) patients, incomplete resolution necessitated subsequent surgery for postoperative fluid collection (n = 3), diverticulitis (n = 2), or perforating appendicitis (n = 1). Complications of transgluteal drainage were rare and included hemorrhage in three (2%) of the 140 patients. There was no procedure-related mortality. A transpiriformis approach was significantly more likely to be associated with postprocedural pain (P < .001) than was an infrapiriformis approach.
CONCLUSION: Percutaneous CT-guided transgluteal drainage is a safe and effective alternative to surgery for deep pelvic abscesses. Major complications are rare.
© RSNA, 2003
Index terms: Abscess, percutaneous drainage, 80.126 • Pelvic organs, abscess, 80.211 • Pelvic organs, interventional procedures, 80.126
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Each patient’s medical record was reviewed separately by one of four radiologists (M.G.H., D.A.G., M.M.M., J.V.) to determine the origins, location, and size of the abscess or abscesses, volume of aspirate, presence of organisms in fluid culture, duration of drainage, incidence of catheter-related pain and procedure-related complications, and outcome of drainage. Catheter size and insertion technique were recorded. One of two radiologists (P.F.H., P.R.M.) with more than 30 years of experience in performing interventional procedures reviewed the images of the drainage procedure to determine the anatomic location and direction of catheter insertion with respect to the piriformis muscle. Butch et al (4) had reported that the insertion of the catheter through the piriformis muscle predicts catheter-related pain, and they had suggested that insertion below the piriformis might minimize this pain. The medical record reviewers worked independently from the image reviewers, and each group was blinded to the findings of the other during data acquisition.
Abscess Drainage
All transgluteal drainage procedures were performed with CT guidance and with intravenous conscious sedation of the patient. Because sedation increases the risk of patient aspiration of orally administered contrast material, no oral contrast agent was administered. Sedation was achieved with the intravenous administration of 1–4 mg (mean, 2 mg) of midazolam hydrochloride (Versed; Roche, Nutley, NJ) and either 25–50 mg (mean, 35 mg) of meperidine hydrochloride (Demerol; Sanofi Winthrop, New York, NY) or 20–100 µg (mean, 68 µg) of fentanyl citrate (Sublimaze; Taylor, Decatur, Ill). In addition, six patients received 12.5–25.0 mg of intravenous promethazine hydrochloride (Phenergan; Wyeth-Ayerst, Philadelphia, Pa). The intravenous sedatives were administered in keeping with institutional guidelines, with continuous monitoring of heart rate and oxygen saturation and with recording of heart rate, blood pressure, and oxygen saturation every 5 minutes. Intravenous sedatives were titrated to achieve adequate pain control for the procedure.
Patients who tolerated the prone position were placed prone (n = 102), whereas those with recent surgical incision (n = 24), respiratory distress (n = 8), or ostomy (n = 6) were placed in a prone oblique or lateral decubitus position. Intravenous antibiotics (1 g ampicillin, 80 mg gentamycin sulfate, and 500 mg metronidazole) were administered to all patients immediately prior to the procedure.
All procedures were performed by both an attending interventional radiologist and an interventional radiology fellow. Self-retaining locking pigtail catheters with distal hydrophilic tips (Cook, Bloomington, Ind) were used in all patients. Catheters ranged in size from 8 to 14 F, and the choice of catheter size for each procedure was made by the attending radiologist performing that procedure. Catheters were placed by using either a tandem trocar technique (in 134 abscesses) or the Seldinger technique (in 20 abscesses) (1,3). The tandem trocar technique involved initial CT-guided placement of an 18–22-gauge spinal needle into the abscess. This needle provided an externally visible guide, enabling catheter insertion at the correct angle, perfectly parallel to the needle. The Seldinger technique involved the insertion of an 18-gauge sheathed needle (Long dwell needle; Baxter, Round Lake, Ill) into the abscess. The sheath allowed the needle to be exchanged for a 0.035-inch Amplatz guide wire (Cook) over which dilators were inserted and serial dilation was performed. The catheter was then inserted over the wire to a depth determined on the basis of the initial transverse CT scan. The need for a second catheter was determined by the attending radiologist on the basis of CT images depicting more than one collection of fluid in the same patient. A single catheter was placed in each of 126 patients, and two catheters were placed in each of 14 patients.
The catheter was inserted below the piriformis muscle when possible (119 procedures), but an approach through the piriformis (35 procedures) was used if necessary to accomplish successful drainage. A medial approach was used to avoid injuring the bowel and the sciatic nerve. The sciatic nerve courses through the lateral and inferior aspect of the sciatic foramen. In the transgluteal approach through the greater sciatic foramen, careful attention must be given to anatomic landmarks in order to avoid injury to the sciatic nerve. At the superior aspect of the foramen lies the piriformis muscle, with the sacral plexus and the inferior gluteal vessels lying anterior to this muscle (Fig 1). The inferior aspect of the greater sciatic foramen contains the sacrospinous ligament, and all major vascular and neural structures are located cephalad to this ligament. The preferred approach for transgluteal access is the infrapiriformis approach, in which the catheter is inserted as close to the sacrum as possible, at the level of the sacrospinous ligament, below the piriformis muscle (Figs 2, 3). At this level the sciatic nerve is situated laterally and can be avoided easily. The location of the abscess or of the bowel, however, sometimes requires a transpiriformis approach—insertion of the catheter through the piriformis muscle.
|
|
|
|
Catheter Management
Immediately after catheter placement, a syringe was inserted and the fluid was aspirated until the flow ceased; after aspiration, the catheter was left in place to allow gravity drainage. Postdrainage images of the area of interest were obtained to verify the adequacy of drainage. The interventional radiology staff, including an attending interventional radiologist and a fellow in interventional radiology, and the surgical staff jointly performed patient follow-up. To maintain catheter patency, the catheters were flushed every 8 hours with 5 mL of 0.9% saline solution. The success of percutaneous drainage was assessed postprocedurally with radiologic imaging and was verified by observation of improvement in the patient’s general clinical condition as indicated by defervescence, diminishing catheter output, and diminishing leukocytosis.
Statistical Analysis
A Fisher exact test was used to assess for difference in the incidence of postprocedural catheter-site pain between transpiriformis and infrapiriformis approaches. A P value of less than .05 was considered to indicate a significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In 134 (96%) of 140 patients, there was a complete resolution of the abscess following transgluteal drainage, with no need for further surgical intervention. Incomplete resolution of abscesses arising from postsurgical fluid collection (n = 3), diverticulitis (n = 2), or perforating appendicitis (n = 1) necessitated subsequent surgery in six (4%) of the 140 patients. The need for surgery was established on the basis of imaging and clinical findings. In four of the six patients, although images showed diminished abscess size, there was persistent high catheter output and little clinical improvement. Fluoroscopic images obtained with catheter injection in these four patients depicted communication of the abscess with the bowel, and subsequent surgical drainage combined with fistula repair was performed. In the other two patients, the abscess size as depicted at CT was unchanged while catheter output persisted, accompanied by persistent fever and worsening leukocytosis. All six patients underwent surgical drainage of the residual abscess cavity, with catheter removal at the time of surgery.
Intraprocedural Pain
During the percutaneous drainage procedure, four (3%) of the 140 patients experienced buttock pain radiating to the leg. CT scans in all four patients showed the localizing needle in the vicinity of the sciatic nerve. After the localizing needle was withdrawn and repositioned more medially, there was immediate, complete resolution of pain. In these four patients, there was no need for additional analgesia, beyond initial sedation, during or after the procedure.
Postprocedural Pain
There were no cases of postprocedural radiating leg pain. However, 27 (19%) of the 140 patients had moderate localized pain at the catheter site immediately after the procedure. In all 27 of these patients, the pain resolved completely within 24 hours after the procedure. Oral analgesics were adequate to control the pain in the first 24 hours. In 20 (74%) of these 27 patients the catheter was introduced through the piriformis muscle, whereas in the remaining seven patients it was inserted below the piriformis. In six other patients severe postprocedural catheter-site pain required patient-controlled intravenous administration of analgesics. In these six patients the catheter had been inserted through the piriformis muscle; complete and immediate resolution of pain followed catheter removal. A transpiriformis approach was significantly more likely to be associated with postprocedural pain (P < .001).
Complications
The only major complication was hemorrhage, which occurred in three (2%) of the 140 patients; in all three, a transpiriformis approach had been used. In two of these patients, hemorrhage occurred through the catheter tract after removal of the catheter; in both patients, arteriography revealed a pseudoaneurysm of the inferior gluteal artery. After the pseudoaneurysms were embolized with coils and gelatin sponge (Gelfoam; Upjohn, Kalamazoo, Mich), no further bleeding occurred in these patients. The third patient developed a large pelvic hematoma 1 day after catheter placement. Arteriography showed no arterial injury, and there was spontaneous resolution of the hematoma with conservative management.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Several aspects of our study results merit further comment. First, the clinical effectiveness of this approach is confirmed by the fact that 134 (96%) of the 140 patients in our study required no further surgery. Percutaneous abscess drainage is usually the preferred treatment in the absence of other indications for surgery. On the basis of the pooled results from several reports on the use of percutaneous drainage to treat abdominal and pelvic abscesses of diverse origins, the American College of Radiology and the Society of Interventional Radiology have developed and published standards for percutaneous abscess drainage (10). The thresholds established with these standards are complete success in 80% of patients and partial success in another 10%. Our 96% (134 of 140 patients) success rate for resolution of abscess following transgluteal drainage without further surgery exceeds these established thresholds. Furthermore, four of the six failures of the procedure occurred in patients who had a fistula in the bowel. Previous investigators have reported that fistulas pose great challenges in percutaneous drainage and may necessitate a longer period of drainage (11).
The second important point emerging from our study is the significantly decreased incidence of catheter-site pain with use of an infrapiriformis approach as opposed to a transpiriformis approach. Nonradiating pain at the catheter site likely results from irritation caused by the catheter to the muscle. Butch et al (4) noted that pain seemed to occur more commonly in patients with the catheter inserted through the piriformis muscle than in patients with the catheter inserted through the gluteal muscles and below the piriformis. The size of our patient group allowed us to quantify this observation by Butch et al and test the data for statistical significance. Results of statistical testing confirmed a significant difference in the incidence of pain between the infrapiriformis approach and the transpiriformis approach, with the latter significantly more likely to be accompanied by postprocedural pain. The overall 24% incidence of catheter-site pain in our patients is comparable to the 20% incidence reported by Butch et al. Based on the results of these two studies, the expected incidence of catheter-site pain with use of an infrapiriformis approach is 20%–24%. Most important, however, in 27 (82%) of 33 patients with catheter-site pain, the pain resolved within 24 hours and was readily treated with oral analgesics in the first few hours.
For percutaneous drainage of abscesses that are not accessible with a transabdominal approach, the choice is usually among the transvaginal, transrectal, and transgluteal approaches in women and between the transgluteal and transrectal approaches in men. For drainage of pelvic collections of fluid near the vagina and rectum, excellent results have been reported with transvaginal (12) and transrectal approaches (13), respectively. However, the transgluteal approach has advantages over these approaches. First, the transvaginal approach cannot be employed to drain presacral abscesses effectively and is impossible in men (6). If long-term catheter drainage is indicated, catheter fixation is achieved much more easily with a transgluteal approach (14). This approach, however, also has drawbacks. Because patients must lie prone on the transgluteally placed catheter for proper gravity drainage, they may experience discomfort, and catheter kinking may occur.
When more than one of these three approaches is technically feasible and clinically acceptable, the choice of approach will depend on the radiologist who performs the procedure. The choice likely will be based on personal experience with the various approaches and on the relative availability of CT (for guidance during a transgluteal approach) compared with that of ultrasonography (for guidance during a transrectal or transvaginal approach). At our institution, in patients in whom a transgluteal approach is possible, this approach usually is preferred over the transrectal and transvaginal approaches. The transrectal and transvaginal approaches usually are reserved for small fluid collections close to the rectum or vagina, particularly in women with pelvic inflammatory disease, in whom short-term drainage is indicated.
The most common complication with the transgluteal approach is bleeding, and this was the only complication that occurred in our study, affecting only three (2%) of our 140 patients. Conventional arteriography was performed in these three patients for evaluation of bleeding and was followed by arterial embolization in two of the three. Bleeding after transgluteal drainage of a pelvic abscess was reported previously by Malden and Picus (14), who also treated this complication successfully with percutaneous arterial embolization. Our experience suggests that the vessel at highest risk of injury is the inferior gluteal artery and that an infrapiriformis approach may diminish the risk of injury to this artery.
Our results confirm that an infrapiriformis approach to transgluteal drainage is preferable when possible, because this approach is associated with a significantly decreased incidence of catheter-site pain and a probable decreased risk of hemorrhagic complications. Caudocephalic gantry angulation may allow infrapiriformis access to relatively cephalic abscesses. If infrapiriformis drainage is not readily accomplished, the transpiriformis approach is an option, especially for drainage of high presacral abscesses, but the radiologist performing the procedure should anticipate a higher incidence of catheter-site pain.
There were no cases of sciatic nerve injury during catheter placement in the 154 drainage procedures included in this study. It is likely that such injury was avoided by our careful monitoring of intraprocedural radiating leg pain. Four of our patients reported radiating leg pain during initial needle placement. Review of the images for these four patients showed that the needle was in close proximity to the sciatic nerve—not the planned location. The needle most likely had deviated from the planned trajectory because the patient had tensed the gluteal muscles in response to insertion of the needle. We stress, therefore, that after the localizing needle has been placed and before the catheter is inserted, the needle placement should be evaluated for lateral deviation of the needle tip. If imaging or other evidence suggests sciatic nerve irritation due to the proximity of the localizing needle to the sciatic nerve, it is probably safest to reposition the localizing needle to avoid sciatic nerve damage. Image review following initial needle placement is especially important if a surgeon is performing abscess drainage in a patient who is receiving general anesthesia or in an otherwise minimally responsive patient, because the images may provide the only evidence that the needle is approaching the location of the sciatic nerve. In all four patients in our study who experienced intraprocedural radiating leg pain, repositioning of the needle resulted in immediate resolution of pain without sequelae.
In conclusion, a CT-guided transgluteal approach is effective for percutaneous drainage of deep pelvic abscesses that are inaccessible with an anterior approach. Catheter-site postprocedural pain occurs in few patients, is readily controlled by oral analgesics, and resolves within 24 hours following the procedure. An infrapiriformis approach significantly decreases the incidence of pain and also may decrease the risk of bleeding complications, the most common of which is hemorrhage. Hemorrhage occurred in only three (2%) of the 140 patients in this study.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. D. Beland, D. A. Gervais, D. A. Levis, P. F. Hahn, R. S. Arellano, and P. R. Mueller Complex Abdominal and Pelvic Abscesses: Efficacy of Adjunctive Tissue-Type Plasminogen Activator for Drainage Radiology, May 1, 2008; 247(2): 567 - 573. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. O. Jacobs Diverticulitis N. Engl. J. Med., November 15, 2007; 357(20): 2057 - 2066. [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Gannon, J. S. Zager, G. J. Chang, B. W. Feig, C. G. Wood, J. M. Skibber, and M. A. Rodriguez-Bigas Pelvic Exenteration Affords Safe and Durable Treatment for Locally Advanced Rectal Carcinoma Ann. Surg. Oncol., June 1, 2007; 14(6): 1870 - 1877. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Cahill, K. M. Baskin, R. D. Kaye, C. R. Fitz, and R. B. Towbin Transgluteal Approach for Draining Pelvic Fluid Collections in Pediatric Patients Radiology, March 1, 2005; 234(3): 893 - 898. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. H. Woo, S. F. Millward, and M. Harisinghani Transgluteal Approach for Percutaneous Drainage of Deep Pelvic Abscesses: How to Avoid Injury to Vital Structures * Dr Harisinghani responds: Radiology, October 1, 2004; 233(1): 300 - 302. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
