| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Department of Radiology, Division of Abdominal Imaging and Intervention, Massachusetts General Hospital, 55 Fruit St., White 270, Boston, MA 02114.
Received June 18, 2003;
accepted after revision August 15, 2003.
Address correspondence to D. A. Gervais.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Imaging studies from patients who underwent multiple drainages were reviewed to define a cohort of patients who underwent secondary percutaneous abscess drainage after successful initial percutaneous abscess drainage of the same abscess. Medical records of these patients were then reviewed to assess the results of secondary percutaneous abscess drainage.
RESULTS. Forty-five abscesses in 43 patients required secondary percutaneous abscess drainage. Twenty-four of the 43 patients avoided surgery. Secondary percutaneous abscess drainage was successful in evacuating the abscess cavity in 39 (91%) of 43 patients. Duration of drainage and time until recurrence were not significant predictors for avoiding surgery. Mean duration of secondary percutaneous abscess drainage was significantly longer than mean duration of primary percutaneous abscess drainage, but duration of secondary percutaneous abscess drainage (25 vs 14 days, respectively; p = 0.007) did not differ significantly between patients who ultimately required surgery and those who did not (17 vs 11 days, respectively; p = 0.10). Time to recurrence ranged from 2 days to 1 year (mean, 51 days).
CONCLUSION. After successful primary percutaneous abscess drainage, secondary percutaneous abscess drainage of recurrent abscesses succeeded in evacuating the abscess cavity in most patients, and surgery was avoided by slightly more than half. Patients with postoperative abscesses were significantly more likely to avoid surgery (p = 0.008), whereas patients with pancreatic abscesses were significantly more likely to require it (p = 0.03).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We undertook this study to assess the results of repeated percutaneous drainage in patients with recurrent abscesses after successful initial drainage. In particular, we sought to assess whether the abscesses had been completely drained at the initial procedure and whether patients subsequently required surgery for definitive treatment of the underlying disease. In addition, we sought to assess the factors that might predict which recurrent abscesses could be successfully treated with repeated percutaneous abscess drainage, such as duration of initial drainage, time until recurrence, and cause.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
For patients in whom imaging review confirmed that an abscess had undergone secondary percutaneous drainage after a successful primary percutaneous drainage, medical and surgical records were reviewed for duration of the initial and repeated drainages, time until recurrence, underlying cause of abscess, documentation of a fistula, and outcome of the second percutaneous drainage. The minimum follow-up period was 4 years.
The overall distribution of underlying causes and abscess locations was determined for all 785 patients to determine whether certain underlying causes or locations might be associated with a higher incidence of repeated drainage. Locations were categorized as abdominopelvic, retroperitoneal, peripancreatic, or confined within a parenchymal organ such as the liver or spleen. Abscesses related to underlying pancreatitis were categorized as peripancreatic, retroperitoneal, or abdominopelvic, depending on location, not on cause.
Drainage Procedure and Patient Treatment
Abscess drainage was performed in all cases by both an interventional
radiology fellow and a staff interventional radiologist. The use of CT or
sonography for imaging guidance was at the discretion of the staff
radiologist. Eight- to 14-French locking pigtail catheters were placed using
the trocar or Seldinger technique. Catheter size and insertion technique were
also at the discretion of the staff interventional radiologist (Fig.
1A,
1B). Immediate imaging after
the procedure confirmed adequate catheter placement in all cases. Catheters
were left to gravity drainage and flushed with 0.9% saline solution every
8–12 hr to maintain catheter patency. Catheters were managed jointly by
the interventional radiology service and the referring clinical service. The
interventional radiology service performed daily rounds to ensure that saline
flushes were performed and that the catheter remained in correct position.
Parameters of clinical improvement that were monitored were fever,
leukocytosis, and overall well-being. Catheters were removed when outputs
diminished to less than 10–20 mL per day.
|
|
CT was performed before removing the catheters to ensure evacuation of the abscess contents. Successful primary abscess drainage was defined as improvement in the clinical parameters listed previously, negligible catheter output, and complete evacuation of the abscess contents visible on CT. In patients with clinical evidence of a fistula (persistent high catheter outputs or drainage that resembled bowel contents or bile), a fistulogram was obtained. Patients with documented fistulas underwent prolonged catheter drainage until the fistula closed or until they were treated surgically.
Statistics
The Student's t test was used to evaluate differences between
group means with respect to duration of first drainage, duration of second
drainage, and time until recurrence. The chi-square test was used to assess
the difference in fistula formation between groups and to determine whether
the predominant causes predisposed patients for successful secondary
percutaneous abscess drainage and avoidance of surgery. A p value of
less than 0.05 was considered significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Mean duration of primary percutaneous abscess drainage was 14 days. Mean duration of secondary percutaneous abscess drainage was significantly longer at 25 days (p = 0.007). Time to recurrence ranged from 2 to 365 days (mean, 51 days). Among the patients who avoided surgery, time until recurrence ranged from 2 to 365 days (mean, 44 days.) Among the patients who ultimately required surgery, time until recurrence ranged from 15 to 191 days (mean, 60 days). This difference in means was not significant (p = 0.47). The duration of primary and secondary percutaneous abscess drainage did not differ significantly between the patients who required surgery and those who did not. The mean primary percutaneous abscess drainage times were 17 and 11 days (p = 0.10), respectively; the mean secondary percutaneous abscess drainage times were 27 and 23 days (p = 0.56), respectively.
Among the 19 patients who underwent surgery, the abscess cavity had been completely evacuated by primary abscess drainage in 15 (79%). Thus, in combination with the 24 patients who did not require surgery, secondary abscess drainage succeeded in evacuating the abscess contents in a total of 39 (91%) of the 43 patients.
Causes and Outcomes
Most abscesses that required secondary drainage were postoperative. A total
of 459 postoperative abscesses in 374 patients were drained during this
period. Among the patients who underwent primary and secondary percutaneous
abscess drainage, 22 (51%) of 43 underwent drainage of 23 postoperative
abscesses. Thus, the repeated percutaneous abscess drainage rate for
postoperative abscesses was 23 (5%) of 436.
Abscesses developed after surgery directly involving bowel, with the creation of anastomoses or oversewn stumps in eight patients. Seven of these eight patients underwent repeated percutaneous abscess drainage without requiring further procedures, and one required surgery for anastomotic revision and abscess drainage. Four patients had previously undergone hepatectomy and developed abscesses in the postoperative bed. Three of these four patients underwent successful repeated percutaneous abscess drainage, and one required surgical drainage. Three patients presented after appendectomy for appendicitis with perforation, and all three underwent successful repeated percutaneous abscess drainage without requiring further surgery. Two patients had previously undergone cholecystectomy and developed abscesses as a complication of cystic duct stump leaks. One of these two patients underwent successful repeated drainage, and the other required surgery for drainage of an enlarging abscess. Two patients developed abscesses caused by bowel fistulas that developed after thoracoabdominal aneurysm repair. Both ultimately required surgery—for fistula excision in one patient and diverting gastrostomy in the other.
The remaining three postoperative patients did not require additional procedures after repeated drainage of abscesses, which developed after bile duct surgery in two patients and after a gastrectomy and a hernia repair in two patients. Thus, in total, 17 (77%) of 22 patients with postoperative abscesses who underwent repeated percutaneous drainage were treated successfully without needing additional surgery. Patients with postoperative abscesses were significantly more likely to avoid surgery (p = 0.008).
The second most common underlying cause for secondary percutaneous abscess drainage among these 43 patients was pancreatitis or pancreatic duct injury, which involved 13 patients and 14 abscesses. One hundred seventy patients underwent drainage of 217 abscesses related to pancreatitis or pancreatic duct leaks, so the rate of repeated drainage for this disease process was 14 (7%) of 203. Four of these 13 patients (31%) required no further surgery, but patients in whom pancreatitis was the underlying cause were significantly more likely to require surgery (p = 0.03). One patient presented with an infected pseudocyst and underwent percutaneous abscess drainage twice, with complete evacuation of the contents. However, the abscess recurred, and the patient underwent surgery. The remaining eight patients required pancreatic débridement, six without requiring additional drainage of the index abscess.
The underlying causes in the remaining eight of the 43 patients varied widely. In five of these eight patients, the diagnosis was known at the time of primary and secondary percutaneous abscess drainage. Tuberculous psoas abscesses were drained successfully and recurred in two patients, both of whom underwent repeated percutaneous abscess drainage without subsequently needing surgery. One patient with Crohn's disease ultimately required small-bowel resection after the second percutaneous drainage. A perirenal abscess in one patient thought to be related to urolithiasis was successfully drained twice without needing further procedures. The fifth patient developed sacral osteomyelitis and abscess formation as a result of trauma. The abscess was drained completely twice, and the patient then underwent sacral débridement. The abscess recurred and was successfully drained percutaneously a third time.
The three remaining patients required surgery before a definitive diagnosis could be reached. Two were diagnosed with tumors that were not suspected initially. A 72-year-old man presented with a pelvic abscess 1 year after an abdominoperineal resection for rectal cancer. He underwent percutaneous abscess drainage and secondary percutaneous abscess drainage 3 months later. Two months after the secondary percutaneous abscess drainage, he underwent surgical drainage for recurrence. The abscess recurred and at the third surgical drainage, curettage confirmed recurrent rectal cancer. A 46-year-old man presented with a large pelvic abscess with associated inflammatory changes in adjacent sigmoid colon and was diagnosed as having diverticulitis with abscess formation. The abscess was successfully drained but recurred 7 days later. At surgery, adenocarcinoma of the sigmoid colon was diagnosed.
Another elusive diagnosis occurred in a 39-year-old woman who underwent pelvic abscess drainage after appendectomy. Bacteroides species were recovered, and the abscess was completely evacuated at percutaneous abscess drainage on three separate occasions in 13 months. At 20 months, the collection was removed surgically, and an infected endometrioma was diagnosed.
Our sample contained 11 documented fistulas: stomach, n = 3; cystic duct, n = 2; small bowel, n = 3; colon, n = 1; duodenum, n = 1; and bile duct, n = 1. Three fistulas occurred among the 24 patients who did not require surgery, and eight occurred among the 19 patients who did require surgery. This difference was significant (p = 0.03).
Repeated Drainage by Abscess Location
Most abscesses undergoing repeated percutaneous abscess drainage were in
the abdomen or pelvis, with 37 (5.6%) of 659 abscesses needing repeated
drainage. Three (3.7%) of 82 retroperitoneal abscesses required repeated
percutaneous abscess drainage. Repeated percutaneous abscess drainage rates in
peripancreatic and liver abscesses were three (4.7%) of 64 and two (1.9%) of
102, respectively. Four splenic abscesses were drained during this period, and
none required repeated percutaneous drainage.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although abscesses had diverse causes, postoperative abscesses and abscesses related to pancreatitis were the most common, and we have confirmed that for secondary percutaneous abscess drainage, results are similar to reported results for primary drainage. Our significantly higher success rate in avoiding surgery in 17 (77%) of 22 patients with postoperative abscesses compared with the significantly lower success rate in four (31%) of 13 patients with pancreatitis is consistent with prior reports [8–12] that indicate higher success rates for primary drainage of postoperative abscesses and lower success rates for primary drainage of abscesses related to pancreatitis. In a prospective study of 96 patients, Cinat et al. [9] have recently shown that a postoperative process is a significant (p = 0.04) independent predictor of success and that a pancreatic process is a significant (p = 0.02) predictor of failure.
Our results confirm the experience of others [1, 13] who had less success with percutaneous abscess drainage in the presence of a fistula. The 11 patients with fistulas in our series were significantly more likely to undergo surgery. Nevertheless, successful repeated percutaneous abscess drainage was possible in some patients with fistulas. Furthermore, we speculate that small undetectable leaks or fistulas may have contributed to recurrence in other patients in whom we did not identify a fistula. For example, in postoperative cases, a persistent or recurrent leak from the bowel or the biliary system might have resulted in abscess recurrence. Likewise, with pancreatitis, persistent peripancreatic inflammation or pancreatic duct leak could have resulted in abscess recurrence. Multiple abscess recurrences after successful evacuation of the abscess contents should prompt consideration of unsuspected diagnoses. Cancer was the most common in our series.
Our study is limited by its retrospective nature. The 4.9% secondary percutaneous abscess drainage rate is not the same as that of abscess recurrence after percutaneous abscess drainage, nor does it reflect the total failure rate for percutaneous abscess drainage. Other reasons for failure exist besides recurrence after a successful drainage.
The abscess recurrence rate is probably higher than the rate we found in this sample because our methods identified only those patients who returned for secondary percutaneous abscess drainage; we missed any with recurring symptoms who went directly to surgical drainage. For those patients, we cannot know how secondary percutaneous abscess drainage might have altered their clinical course. However, our secondary percutaneous abscess drainage rate establishes a minimum possible value of the actual recurrence rate for these 785 patients and 956 abscesses during 4 years of follow-up. Moreover, our institution relies heavily on percutaneous abscess drainage even for abscess recurrences, so the number of patients with abscess recurrences taken directly to surgery was probably small.
Reported rates of recurrence in other smaller series and shorter follow-up periods place our minimum well within expected ranges. In a seminal article, Gerzof et al. [3] reported only one (1%) recurrence of 71 abscesses in a study with follow-up periods ranging from 1 month to 5 years. However, subsequent reports have established higher recurrence rates. VanSonnenberg et al. [1] reported 20 (8%) recurrences of 250 abscess drainages. Lambiase et al. [4] reported seven (2.1%) recurrences of 355 abscess drainages during 1 year of follow-up in all patients, with all recurrences occurring within 3 months.
In this series, 4.9% of abscesses required secondary percutaneous abscess drainage. Success of secondary percutaneous abscess drainage in evacuating the abscess contents was 91%, and secondary percutaneous abscess drainage was successful in obviating surgery in 56% of patients. Duration of primary drainage and time until recurrence were not significant predictors of the results of secondary percutaneous abscess drainage. However, after secondary abscess drainage, patients with postoperative abscesses were significantly more likely to avoid surgery, and patients with pancreatic abscesses were significantly more likely to need it.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. J. Kim, J. K. Han, J. M. Lee, S. H. Kim, K. H. Lee, S. H. Park, S. K. An, J. Y. Lee, and B. I. Choi Percutaneous Drainage of Postoperative Abdominal Abscess with Limited Accessibility: Preexisting Surgical Drains as Alternative Access Route Radiology, May 1, 2006; 239(2): 591 - 598. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
