Gastrointestinal Imaging Original Research

Marin et al. Abscess Drainage for Acute Appendicitis Gastrointestinal Imaging Original Research

Percutaneous Abscess Drainage in Patients With Perforated Acute Appendicitis: Effectiveness, Safety, and Prediction of Outcome
Daniele Marin1 Lisa M. Ho1 Huiman Barnhart 2 Amy M. Neville1 Rebekah R. White 3 Erik K. Paulson1
Marin D, Ho LM, Barnhart H, Neville AM, White RR, Paulson EK

OBJECTIVE. The purposes of this study were to retrospectively investigate the effectiveness and safety of CT-guided percutaneous drainage in the treatment of patients with acute appendicitis complicated by perforation and to identify CT findings and procedure-related factors predictive of clinical and procedure outcome. MATERIALS AND METHODS. From March 2005 through December 2008, 41 consecutively registered patients (24 men, 17 women; age range, 1875 years) underwent CTguided percutaneous drainage for the management of acute appendicitis complicated by perforation and abscess. Three board-certified radiologists independently reviewed preprocedure CT images. Patients were assigned to one of three risk categories on the basis of the CT findings. Success and failure of percutaneous drainage were defined on a per-patient (i.e., clinical outcome) and per-procedure (i.e., technical outcome) basis. Immediate, periprocedure, and delayed complications were recorded. The association between candidate predictive variables, including demographic characteristics, preprocedure CT findings, and procedure-related factors and clinical or technical outcome was assessed with logistic regression models. RESULTS. Fifty-two CT-guided procedures were performed on 41 patients. Percutaneous drainage had clinical and technical success rates of 90% (37 of 41 patients, 47 of 52 procedures) with no procedure-related complications. In seven patients (19%) clinical success required repeated drainage procedures. A large, poorly defined periappendiceal abscess and an extraluminal appendicolith on preprocedure CT images were independent predictors of clinical failure of percutaneous drainage. CONCLUSION. CT-guided percutaneous drainage is both effective and safe in the treatment of patients with acute appendicitis complicated by perforation and abscess. The clinical and technical success rates are high. cute appendicitis is a common clinical problem with an incidence of approximately 1 case per 1,000 persons per year [1]. Although immediate appendectomy is the treatment of choice of patients with uncomplicated acute appendicitis, there is no consensus on the optimal treatment of the approximately 26% of patients whose condition becomes manifest at a later stage with appendiceal perforation with or without appendiceal abscess [24]. Imaging-guided percutaneous drainage in combination with broad-spectrum IV antibiotics is an effective, minimally invasive treatment of patients with acute appendicitis complicated by perforation and abscess [58]. This approach manages the initial inflammatory process and is followed by either elective interval appendectomy or conservative nonoperative management in selected

Keywords: abdominal abscess, effectiveness, percutaneous drainage, perforated acute appendicitis DOI:10.2214/AJR.09.3098 Received May 26, 2009; accepted after revision July 30, 2009.
1 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710. Address correspondence to Erik K. Paulson (pauls003@mc.duke.edu). 2 Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC. 3 Department of Surgery, Duke University Medical Center, Durham, NC.

AJR 2010; 194:422429 0361803X/10/1942422 American Roentgen Ray Society

cases. For reasons that remain unclear, however, a substantial minority ( 1025%) of patients with perforated acute appendicitis may not respond to initial nonsurgical treatment with percutaneous drainage, leading to prolonged hospitalization, repeated percutaneous procedures, multiple follow-up CT examinations, and in some cases, urgent appendectomy [9, 10]. The results of the few studies [6, 9, 10] conducted in attempts to identify factors predictive of the outcome of percutaneous drainage in patients with perforated acute appendicitis have been conflicting. This lack of agreement generates uncertainty in the clinical care of these patients, leading to wide variations in clinical practice among surgeons at different institutions and even among surgeons at the same institution. The purposes of this study were to retrospectively investigate the effectiveness and safety

422

AJR:194, February 2010

Abscess Drainage for Acute Appendicitis of CT-guided percutaneous drainage in the care of patients with acute appendicitis complicated by perforation and to identify CT findings and procedure-related factors predictive of clinical and procedure outcome. Materials and Methods
This retrospective single-center HIPAA-compliant study was approved by our institutional review board. The requirement for informed consent was waived. traluminal appendicolith. In 24 patients who underwent interval appendectomy a mean of 74 days (range, 0210 days) after the primary drainage procedure, a final diagnosis of appendiceal perforation was confirmed when appendiceal perforation was macroscopically evident (n = 5), when transmural inflammatory cell infiltrate with necrosis was found at pathologic examination (n = 12), or both (n = 7). In two of these patients, pathologic examination showed perforated acute appendicitis was associated with mucocele secondary to mucinous cystadenoma. procedure. All the procedures were performed or closely supervised by one of 12 attending radiologists with 420 years of experience in imagingguided percutaneous drainage and interventional procedures. A senior resident or fellow assisted with the procedure. Before each procedure, diagnostic CT examinations were reviewed for planning of an appropriate route. Catheter size (Flexima APD or APDL, Boston Scientific) was determined by the attending radiologist. Abscess drainage was performed with CT or CT fluoroscopic guidance (CT/i equipped with SmartView, GE Healthcare) and Seldinger technique [13]. After catheter placement, the collections were aspirated as completely as possible, and samples were sent for microbiologic analysis. The catheters were attached to Jackson-Pratt bulb drains (Medi-Vac, Cardinal Healthcare), which generate 3050 mm Hg of suction. The inpatient nursing service ensured catheter patency by flushing the catheter lumen with 1015 mL of 0.9% sterile saline three times per day. The decision for catheter removal was based on the following criteria: clinical improvement (normal body temperature and WBC count, no clinical symptoms), drainage output of 10 mL/d or less, and CT findings of complete resolution of the target fluid collection.

Patient Selection
We reviewed the interventional procedure log for CT-guided percutaneous abscess drainage of the abdomen or pelvis performed from March 2005 through December 2008. Among 843 procedures, 59 consecutive procedures on 48 patients were reported as being performed for acute appendicitis complicated by perforation and abscess (Fig. 1). For each of these patients, we reviewed medical records (radiology, surgery, pathology, and discharge summary) to confirm the diagnosis of acute appendicitis with perforation. Seven of the 48 patients were excluded because of a history of Crohns disease (five patients) or concomitant tuboovarian abscess (two patients). The other 41 patients (mean age, 38 years; range, 1875 years) composed our study cohort, which included 24 men (mean age, 40 years; range, 1875 years) and 17 women (mean age, 29 years; range, 2050 years). The final diagnosis of perforated appendicitis was based on a clinical history of fever, leukocytosis, and right lower quadrant abdominal pain corroborated by at least one of the following CT findings [11, 12]: focal defect in an enhancing appendiceal wall, periappendiceal abscess, periappendiceal phlegmon, extraluminal air, and ex-

Preprocedure CT and Drainage Procedure

Diagnostic CT was performed with an MDCT scanner (LightSpeed 16, GE Healthcare) with the following parameters: detector configuration, 16 0.625 mm; effective section thickness, 5 mm; reconstruction interval, 5 mm; gantry rotation time, 0.5 second; beam pitch, 1.75; 100350 mA depending on the patients body habitus; 140 kVp. Patients ingested 450 mL of a 2% barium sulfate suspension (Readi-Cat 2, E-Z-EM) 12 hours before scanning. After IV administration of 150 mL of nonionic contrast medium (iopamidol, Isovue 300, Bracco), scanning was performed from the dome of the diaphragm through the pubic symphysis during the portal venous phase as determined with bolus tracking and automated triggering technology. In addition to the transverse source images, a set of coronal images of the abdomen and pelvis (effective section thickness, 3 mm; reconstruction interval, 3 mm) were reconstructed by the technologist at the operators console. After referral from the surgical team, each patient gave written informed consent before the drainage procedure. All patients were treated with broad-spectrum antibiotics before the drainage

Data Collection
Preprocedure CT findings and risk catego rizationThree board-certified radiologists with 18, 10, and 3 years of experience in abdominal imaging independently reviewed the preprocedure CT images of each patient on a PACS workstation (Centricity 2.1, GE Healthcare). Readers were aware that the patients had been referred for known or suspected perforated appendicitis, but they were unaware of the clinical data and final

Fig. 1Flowchart shows study enrollment.

AJR:194, February 2010

423

Marin et al.

A
Fig. 2Three-category CT scale of perforated acute appendicitis. A, 32-year-old man with right lower quadrant abdominal pain, tenderness, and mild leukocytosis (category 1, periappendiceal phlegmon or abscess smaller than 3 cm). Transverse CT image shows well-circumscribed, 2.5-cm abscess with thickened wall (straight arrows) in right lower quadrant and 0.5-cm extruded appendicolith (curved arrow). Examination of specimen from elective laparoscopic appendectomy 150 days after drainage procedure showed chronic appendicitis. B, 32-year-old man with acute onset of lower abdominal pain radiating to periumbilical region for 12 hours (category 2, well-circumscribed periappendiceal abscess larger than 3 cm). Transverse CT image shows well-circumscribed 5-cm pelvic abscess (arrows) with airfluid level (asterisk). Patient was treated successfully with percutaneous drainage and antibiotic therapy only. C, 26-year-old woman with 6 hours of generalized abdominal pain, tenderness, and leukocytosis (category 3, large, poorly defined periappendiceal abscesses extending to distant locations). Transverse CT image shows large, poorly-defined pelvic abscess (arrows) extending from periappendiceal region to pouch of Douglas. After initial attempt at percutaneous drainage, patient underwent urgent open appendectomy because follow-up CT showed interval increase in abscess.

C
outcome (see later, Effectiveness). Because we were not attempting to determine the diagnostic accuracy of CT, disagreement was resolved by consensus. Readers assessed contrast-enhanced CT images of each patient for visualization of the appendix and the presence of a periappendiceal abscess (defined as a fluid collection adjacent to the appendix with attenuation of 020 HU), periappendiceal phlegmon (defined as areas of 20-HU or greater attenuation in the fat tissue surrounding the appendix), extraluminal gas or appendicolith, and small-bowel obstruction. If an abscess was identified, readers also documented the size (defined as the single largest transverse diameter) and margins (either well-circumscribed or poorly defined) of the fluid collection. In an attempt at stratification according to severity of inflammatory disease and size and complexity of abscess, patients were assigned to one of three risk categories on the basis of the CT findings. Based on a classification system described by Jeffrey et al. [6], the categories were 1, periappendiceal phlegmon or abscess smaller than 3 cm (n = 17) (Fig. 2A); 2, well-circumscribed periappendiceal abscess larger than 3 cm (n = 10) (Fig. 2B); and 3, large, poorly defined periappendiceal abscesses extending to distant locations, such as the pelvic cul-de-sac, the interloop spaces, or beyond the peritoneal cavity (n = 14) (Fig. 2C). Procedure detailsDetails of each drainage procedure were recorded by one abdominal imaging research fellow who retrospectively reviewed the interventional radiology data sheets, which were prospectively completed by the attending radiologist after the procedure: procedure reports and intraprocedure CT images. Data collected included the approach for catheter placement, number and diameter of catheters, volume and character of aspirate (purulent or not purulent), and results of microbiologic culture. For patients who underwent repeated procedures, technical details were recorded individually for each procedure. Documentation of the duration of catheter placement was not reported because of inconsistent data from patients who were discharged after the drain placement and underwent outpatient follow-up. EffectivenessThe same abdominal radiology research fellow who recorded the procedure details assessed the outcome of drainage therapy by retrospectively reviewing electronic medical records (radiology, surgery, pathology, and discharge summary) for each patient. Success and failure were defined per patient (i.e., clinical outcome) and per procedure (i.e., technical outcome). Clinical success was defined as patient recovery after single or multiple procedures with or without interval elective appendectomy. Clinical failure was defined as progressive deterioration with worsening clinical signs and symptoms of infection after single or multiple drainage procedures that ultimately necessitated urgent appendectomy. Technical success was defined as complete resolution of an abscess as determined at follow-up CT along with negligible catheter output. A procedure was considered a failure if the operator was unable to place a drain, if no fluid was aspirated

424

AJR:194, February 2010

Abscess Drainage for Acute Appendicitis

B
Fig. 318-year-old woman with right lower quadrant pain and fever. A, Transverse contrast-enhanced CT scan shows 5-cm well-circumscribed abscess (straight arrows) in right lower quadrant, inflammatory changes in adjacent fat tissue (category 2), and 1-cm extruded appendicolith (curved arrow). B, CT fluoroscopic image (140 kV, 10 mA, 5-mm section thickness) shows 18-gauge needle (curved arrow) in pelvic abscess (straight arrows). After stepwise dilation, 14-French pigtail catheter (not shown) was placed in abscess. C, Follow-up CT scan 15 days after drainage procedure shows successful abscess drainage (arrow) with no residual fluid. No interval appendectomy was performed.

C
after successful catheter placement, or if followup imaging more than 1 day after the procedure showed enlargement of the abscess that necessitated either secondary drainage or urgent appendectomy. Repeated drainage procedures because of development of a new abscess in a different location were not deemed technical failure and were evaluated separately. Abscesses that became evident after surgical appendectomy were excluded from our analysis. Immediate, periprocedure, and delayed complications were recorded per treatment and were classified in accordance with suggested reporting criteria [14]. Predictive variablesCandidate predictive variables selected included demographic characteristics (age and sex), preprocedure CT findings (risk category and presence of extraluminal gas or appendicolith and small-bowel obstruction), and factors related to the first procedure (approach for catheter placement, number and diameter of catheters, volume and character of aspirate, and results of microbiologic culture). The association between these variables and clinical or technical outcome was assessed with logistic regression models. Because of the small number of patients who underwent multiple procedures, only the first procedure was considered in the association analysis. Variables in the univariate analysis with p < 0.20 were entered into multivariate logistic regression analysis in a search for independent factors predictive of outcome. Backward-forward and stepwise procedures were used for model selection with entry and stay level of 0.10. Because of the exploratory nature of the analyses with a small sample size, p 0.1 was considered to indicate statistical significance. All statistical analyses were performed with statistical software (SAS version 9.1.3, SAS Institute).

Results Preprocedure CT Findings, Risk Category, and Procedure Details Table 1 summarizes the demographic characteristics, preprocedure CT findings, and procedure details for patients in different risk categories. In 39 of the 41 patients (95%), perforated acute appendicitis became manifest as a periappendiceal abscess (mean size, 4.1 cm; range, 0.810.5 cm) at the initial

CT examination. The abscess was associated with an extraluminal appendicolith in 16 of the patients (39%). In no patient were multiple abscesses present throughout the abdomen or pelvis. Unequivocal identification of the appendix was possible in 24 of the 41 patients (59%). In 37 of the 41 patients (90%), percutaneous drainage was preferentially performed through a direct transabdominal approach with a single catheter greater than 10-French. Except for two patients in whom percutaneous drainage was performed despite the absence of fluid collections at preprocedure CT, 5350 mL of fluid was aspirated during the drainage procedure. Aspiration revealed purulent fluid in most of the patients (34 of 41, 83%). Effectiveness Fifty-two CT-guided procedures were performed on 41 patients, including a single procedure on 33 patients, two procedures on six patients, three procedures on one patient, and four procedures on one patient. Percutaneous

AJR:194, February 2010

425

Marin et al. drainage was clinically successful for 37 of the 41 patients (90%; 95% CI, 8199%), including 16 of the 17 patients (94%) in category 1, all 10 patients in category 2, and 11 of 14 patients (79%) in category 3 (Fig. 3). In 30 of the 37 patients (81%), success was achieved after a single drainage procedure; seven patients (19%) needed repeated drainage procedures because of follow-up imaging findings of a new abscess in a different location (six patients) or enlargement of a periappendiceal abscess (one patient). In four of the 41 patients (10%), percutaneous drainage was deemed a clinical failure, and urgent appendectomy was performed. In two of these patients (both category 3), including one patient for whom catheter placement required transgression of the ascending colon, follow-up CT (2 and 4 days after procedure) showed enlargement (from 4 to 6 cm and from 5 to 8 cm) of a periappendiceal abscess despite successful catheter placement during the initial drainage procedure (Fig. 4). Although we do not advocate transcolonic percutaneous drainage, this approach was discussed with both the patient and the surgeon before the procedure. In one patient (category 3) the operator was unable to advance the tip of the catheter into the target fluid collection using a direct transabdominal approach. In the other patient (category 1), urgent appendectomy was performed because of the development of small-bowel obstruction 2 days after the drainage procedure. In this patient, an adhesion was removed at the point of transition during surgery. For three of four patients with clinical failure, catheters were still in place when the patients condition deteriorated. Technical success was achieved in 47 of 52 drainage procedures (90%), including 18 of 19 procedures (95%) on patients in category 1, all 11 procedures on patients in category 2, and 18 of 22 procedures (81%) on patients in category 3. The technical success rate for the first procedure was 88% (36 of 41 patients; 95% CI, 7898%). Although five procedures (10%) in five patients were deemed technical failures, clinical success was achieved in three patients after a course of antibiotic therapy either alone (two patients) or in combination with secondary percutaneous drainage (one patient). Causes of technical failure included inability to place a drain into the target fluid collection despite multiple attempts in two patients and increased size of an abscess as documented at follow-up CT in three patients. No complications were observed during or immediately after any drainage procedure. Predictive Variables The results of the univariate analysis for comparison of the predictive variables (demographic characteristics, preprocedure CT findings, procedure-related factors) with clinical and technical outcome are summarized in Table 2. The multivariate analysis showed that risk category 3 (odds ratio, 0.07; 90% CI, 0.010.67; p = 0.05) and extraluminal appendicolith (odds ratio, 0.09; 90% CI, 0.090.83; p = 0.07) were independent predictors of clinical failure of percutaneous drainage (Table 3) (Fig. 5). Use of a direct transabdominal approach for catheter placement was an independent predictor of technical success of percutaneous drainage (odds ratio, 14.73; 90% CI, 1.545140.48; p = 0.05), and female sex was associated with a

TABLE 1: Demographic Characteristics, Preprocedure CT Findings, and Procedure Details on Patients Stratified to Risk Categories
Risk Category Characteristic Sex (no.) Men Women Age (y) Mean SD Range Abscess size (cm) Mean SD Range Phlegmon (no.) Extraluminal gas (no.) Extraluminal appendicolith (no.) Small-bowel obstruction (no.) No. of procedures Single Multiple Technical approach (no.) Transabdominal Transgluteal Transcolic No. of catheters 0 (aspiration) 1 2 Catheter size (no.) > 10 French 10 French Volume of aspirate (mL) Mean SD Range 36.6 50.6 0200 67 100.8 15350 33.9 24.8 070 43.1 60.8 0350 10 (59) 7 (41) 9 (90) 1 (10) 7 (50) 7 (50) 26 (63) 15 (37) 2 (12) 14 (82) 1 (6) 0 10 (100) 0 0 13 (93) 1 (7) 2 (5) 37 (90) 2 (5) 16 (94) 1 (6) 0 10 (100) 0 0 11 (79) 2 (14) 1 (7) 37 (90) 3 (7) 1 (3) 14 (82) 3 (18) 9 (90) 1 (10) 10 (71) 4 (29) 33 (80) 8 (20) 2.5 0.4 0.83 8 (47) 10 (59) 7 (41) 1 (6) 4.5 1.6 3.57 3 (30) 5 (50) 5 (50) 0 5.3 2.1 410.5 7 (50) 7 (50) 4 (29) 1 (7) 4.1 2.0 0.810.5 18 (44) 22 (54) 16 (39) 2 (5) 42.0 15.0 2375 42.2 20.4 1573 29.9 12.1 1657 37.9 16.3 1875 13 (76) 4 (24) 7 (70) 3 (30) 4 (29) 10 (71) 24 (59) 17 (41) 1 (n = 17) 2 (n = 10) 3 (n = 14) Total (n = 41)

NoteValues in parentheses are percentages calculated with numerators in the rows and denominators in the column headings.

426

AJR:194, February 2010

Abscess Drainage for Acute Appendicitis

Fig. 426-year-old woman with right lower quadrant pain. A, Transverse contrast-enhanced CT scan shows 3.5-cm well-circumscribed abscess (arrows) immediately posterior to cecum (C) with inflammatory changes in adjacent fat tissue (category 2). Abscess would have been difficult to approach percutaneously because of interposed intestine, pelvic bones, and adnexa. B, After discussion with referring surgeon, abscess was drained by intentional transgression of ascending colon. CT fluoroscopic image (140 kV, 10 mA, 5-mm section thickness) shows 18-gauge needle (curved arrow) in pelvic abscess (straight arrows). Despite excellent position of drainage catheter (not shown), follow-up CT showed enlargement of abscess that necessitated urgent appendectomy.

C
AJR:194, February 2010

Fig. 519-year-old woman with right lower quadrant pain and fever. A, Transverse contrast-enhanced CT scan shows 5.8-cm poorly defined abscess (black straight arrows) in right lower quadrant of abdomen with inflammatory changes in adjacent fat tissue (category 3), 1-cm-diameter extruded appendicolith (curved arrow), and thickened and enhanced wall of adjacent sigmoid colon (white straight arrows). B, CT fluoroscopic image (140 kV, 10 mA, 5-mm section thickness) shows 0.38 guidewire (curved arrow) coursing through 18-gauge needle (straight arrow) in pelvic abscess. After stepwise dilation, 10-French pigtail catheter (not shown) was placed in abscess. C, Follow-up CT scan 5 days after procedure shows newly developed 4-cm abscess (arrows) in right paracolic gutter immediately lateral to ascending colon (C). After continued clinical deterioration, open appendectomy was performed.

427

Marin et al. lower rate of technical success (odds ratio, 0.12; 90% CI, 0.010.99; p = 0.1) (Table 3). Discussion Our results show that percutaneous drainage is effective and safe in the treatment of patients with acute appendicitis complicated by perforation and abscess. Both clinical and technical success rates were 90% (37 of 41 patients and 47 of 52 procedures), and no procedure-related complications occurred. In patients with a well-circumscribed periappendiceal abscess (categories 1 and 2), the clinical and technical success rates of percutaneous drainage increased to 92% and 96%. Our data compare favorably with the results of previously published studies [58] and confirm the effectiveness of percutaneous drainage in combination with broad-spectrum antibiotics in the treatment of patients with acute appendicitis complicated by perforation and abscess. There is compelling evidence that among these patients, immediate appendectomy is associated with substantially higher risk of complications, including hemorrhage, fistula formation, wound infection, prolonged ileus, and adhesions [4]. Another clinically important finding of our study is that besides the high clinical success rate (81%, 30 of 37 patients), which was achieved after a single drainage procedure, percutaneous drainage was clinically successful in seven of eight patients (88%) who underwent repeated procedures, most commonly because of the development of a new abscess at a location distant from the primary site of infection. These results, which are consistent with the 92% clinical success reported by McCann and colleagues [15] in a study with patients with acute appendicitis complicated by multiple abdominal abscesses, emphasize the importance of secondary drainage in the treatment of patients with perforated acute appendicitis in whom new intraabdominal abscesses develop after the first drainage procedure. At the same time, our data highlight the need for close clinical and CT follow-up after percutaneous drainage for early detection and prompt management of persistent or newly developed abscesses. In our study, initial nonsurgical management with percutaneous drainage failed to control the acute inflammatory process in four of 41 patients, resulting in a 10% clinical failure rate. In accordance with results of previous analyses [6, 9], we found that a large, poorly defined periappendiceal abscess (category 3) and extraluminal appendicolith were the two most specific predictors of unfavorable clinical outcome of percutaneous drainage. This finding, which remained significant after adjustments for other potential prognostic factors, has two important clinical implications. First, it reinforces the current clinical practice of performing urgent appendectomy on patients with perforated acute appendicitis that becomes evident at a later stage with more generalized, potentially life-threatening signs and symptoms of peritoneal infection. Second, it corroborates the hypothesis that patients with an

TABLE 2: Bivariate Association Between Patient Demographics, Preprocedure CT Findings, and Procedure-Related Factors and Clinical and Technical Outcome
Clinical Outcome Variables Sex (no.) Men Women Age (y) Mean SD Risk category (no.) 12 3 Extraluminal gas (no.) Extraluminal appendicolith (no.) Small bowel obstruction (no.) No. of procedures Single Multiple Technical approach (no.) Standard Nonstandard No. of catheters (no.) 1 2 Catheter size (no.) > 10 French 10 French Volume of aspirate (no.) > 50 mL 50 mL Character of aspirate (no.)a Purulent Nonpurulent Microbiologic culture (no.)a Polymicrobial Monomicrobial 34 (92) 2 (6) 3 (75) 0 (0) 32 (86) 4 (11) 2 (50) 1 (25) NA 34 (90) 2 (6) 3 (75) 0 (0) 7 (19) 30 (81) 1 (25) 3 (75) 0.31 30 (83) 5 (14) 2 (50) 1 (25) NA 24 (65) 13 (35) 2 (50) 2 (50) 0.77 8 (22) 28 (78) 0 (0) 5 (100) NA 33 (90) 2 (5) 4 (100) 0 (0) 0.56 24 (67) 12 (33) 2 (60) 3 (40) NA 35 (95) 2 (5) 2 (50) 2 (50) NA 34 (94) 2 (6) 5 (100) 0 (0) 0.26 30 (81) 7 (19) 3 (75) 1 (25) 0.02 34 (94) 2 (6) 3 (60) 2 (40) NA 0.04 26 (70) 11 (30) 21 (57) 13 (35) 2 (6) 1 (25) 3 (75) 1 (25) 3 (75) 0 (0) 0.25 0.16 NA 0.77 38.9 16.7 28.3 7.9 0.11 26 (72) 10 (28) 20 (56) 13 (36) 2 (6) 1 (20) 4 (80) 2 (40) 3 (60) 0 (0) 0.52 0.32 NA 23 (62) 14 (38) 1 (25) 3 (75) 0.24 39.1 17.1 29.6 3.6 0.05 Success (n = 37) Failure (n = 4) p 0.18 23 (64) 13 (36) 1 (20) 4 (80) 0.24 Outcome of First Procedure Success (n = 36) Failure (n = 5) p 0.09

NoteValues in parentheses are percentages calculated with numerators in the rows and denominators in the column headings. NA = not applicable for data with zero frequency. aNo fluid was in two patients.

428

AJR:194, February 2010

Abscess Drainage for Acute Appendicitis TABLE 3: Multivariate Association Between Candidate Predictive Variables and Clinical and Technical Outcomes
Clinical Outcome Variable Sex Risk category (3 vs 1 or 2) Extraluminal appendicolith Technical approach Odds Ratio NI 0.07 0.09 NI 0.010.67 0.090.83 0.05 0.07 90% CI p Outcome of First Procedure Odds Ratio 0.12 NI NI 14.7 1.55140.48 0.05 90% CI 0.010.99 p 0.10
5. Jamieson DH, Chait PG, Filler R. Interventional drainage of appendiceal abscesses in children. AJR 1997; 169:16191622 6. Jeffrey RB Jr, Federle MP, Tolentino CS. Periappendiceal inflammatory masses: CT-directed management and clinical outcome in 70 patients. Radiology 1988; 167:1316 7. Roach JP, Partrick DA, Bruny JL, Allshouse MJ, Karrer FM, Ziegler MM. Complicated appendicitis in children: a clear role for drainage and delayed appendectomy. Am J Surg 2007; 194:769 772 8. Lasson A, Lundagrds J, Lorn I, Nilsson PE. Appendiceal abscesses: primary percutaneous drainage and selective interval appendicectomy. Eur J Surg 2002; 168:264269 9. Aprahamian CJ, Barnhart DC, Bledsoe SE, Vaid Y, Harmon CM. Failure in the nonoperative management of pediatric ruptured appendicitis: predictors and consequences. J Pediatr Surg 2007; 42:934938 10. Nadler EP, Reblock KK, Vaughan KG, Meza MP, Ford HR, Gaines BA. Predictors of outcome for children with perforated appendicitis initially treated with non-operative management. Surg Infect (Larchmt) 2004; 5:349356 11. Horrow MM, White DS, Horrow JC. Differentiation of perforated from nonperforated appendicitis at CT. Radiology 2003; 227:4651 12. Tsuboi M, Takase K, Kaneda I, et al. Perforated and nonperforated appendicitis: defect in enhancing appendiceal walldepiction with multi-detector row CT. Radiology 2008; 246:142147 13. Paulson EK, Sheafor DH, Enterline DS, et al. CT fluoroscopy-guided interventional procedures: techniques and radiation dose to radiologists. Ra diology 2001; 220:161167 14. Cancer Therapy Evaluation Program. Common terminology criteria for adverse events, version 3.0. National Cancer Institute Website. ctep.cancer.gov/reporting/ctc.html. Published December 12, 2003. Accessed May 1, 2009 15. McCann JW, Maroo S, Wales P, et al. Imageguided drainage of multiple intraabdominal abscesses in children with perforated appendicitis: an alternative to laparotomy. Pediatr Radiol 2008; 38:661668 16. Singh AK, Hahn PF, Gervais D, Vijayraghavan G, Mueller PR. Dropped appendicolith: CT findings and implications for management. AJR 2008; 190:707711 17. Buckley O, Geoghegan T, Ridgeway P, Colhoun E, Snow A, Torreggiani WC. The usefulness of CT guided drainage of abscesses caused by retained appendicoliths. Eur J Radiol 2006; 60:8083 18. Maher MM, Gervais DA, Kalra MK, et al. The inaccessible or undrainable abscess: how to drain it. RadioGraphics 2004; 24:717735

NoteNI = not included and not significant at 0.10 level.

extraluminal appendicolith after appendiceal perforation have a poorer prognosis, including increased risk of recurrent abscess and other complications [16]. The presence of an appendicolith can act as a nidus for continuous infection and abscess formation. This notion justifies the need for elective interval removal of a dropped appendicolith with either surgery or CT-guided percutaneous extraction and stone basket catheters [16, 17]. In our study, percutaneous abscess drainage through a direct transabdominal approach was associated with a significantly higher probability of procedure success. In a minority of patients, however, an alternative approach, including a transgluteal route in three patients and intentional transgression of the ascending colon in one patient, was necessary because of the presence of deep-seated, less accessible abscesses. These approaches were not associated with major periprocedure complications, such as hemorrhage or injury to the sciatic nerve but resulted in both clinical and procedure failure in two of four cases (50%). Although this finding did not reach statistical significance, it needs to be emphasized that many authorities have discouraged transgression of the small or large bowel for drainage of deep-seated abscesses [18]. In our patient who needed colonic transgression, the risks of this approach were discussed with both the patient and the referring surgeon before the procedure. Besides its retrospective nature, potential limitations of our study merit consideration. First, the relatively small sample size probably limited the statistical power to detect associations between CT findings and technique-related factors and the outcome variables. Second, we restricted our analysis to patients with conclusive diagnostic criteria for perforated acute appendicitis at CT [11, 12]. This approach might have introduced selection bias because it is possible that some patients with atypical clinical or CT manifestations of perforated acute appendicitis might have been excluded

from the study. In addition, because we included only patients with perforated acute appendicitis who were referred to our interventional radiology service for a percutaneous drainage procedure, we cannot compare the effectiveness of percutaneous drainage with that of other treatments, such as immediate appendectomy and antibiotic therapy alone. Our results also reflect the experience of a single tertiary referral center with a high volume of percutaneous drainage procedures. It remains to be determined whether our results can be generalized to smaller community hospitals. Finally, our work was focused on adults; we believe similar results can be achieved in the treatment of children with perforated appendicitis. CT-guided percutaneous drainage is both effective and safe in the care of patients with acute appendicitis complicated by perforation and abscess, having clinical and technical success rates of 90%. In patients with preprocedure CT findings of large, poorly defined periappendiceal abscesses or extraluminal appendicoliths, percutaneous drainage is associated with a less favorable clinical outcome. References
1. National Center for Health Statistics. Ambulatory and inpatient procedures in the United States, 1996. National Center for Health Statistics Series 13, no. 139. Atlanta, GA: National Center for Health Statistics 2. Brown CV, Abrishami M, Muller M, Velmahos GC. Appendiceal abscess: immediate operation or percutaneous drainage? Am Surg 2003; 69:829832 3. Goh BK, Chui CH, Yap TL, et al. Is early laparoscopic appendectomy feasible in children with acute appendicitis presenting with an appendiceal mass? A prospective study. J Pediatr Surg 2005; 40:11341137 4. Gibeily GJ, Ross MN, Manning DB, Wherry DC, Kao TC. Late-presenting appendicitis: a laparoscopic approach to a complicated problem. Surg Endosc 2003; 17:725729

AJR:194, February 2010

429