World J. Surg. 25, 362–372, 2001 DOI: 10.1007/s002680020386

WORLD Journal of

SURGERY © 2001 by the Socie´te´ Internationale de Chirurgie

Percutaneous Abscess Drainage: Update Eric vanSonnenberg, M.D.,1 Gerhard R. Wittich, M.D.,2 Brian W. Goodacre, M.D.,3 Giovanna Casola, M.D.,4 Horacio B. D’Agostino, M.D.5 1

Department of Radiology, MA 02115, USA 2 Department of Radiology, 3 Department of Radiology, 4 Department of Radiology, 5 Department of Radiology,

Dana Farber Cancer Institute & Brigham and Women’s Hospital, Harvard Medical School, 44 Binney St., Boston, The University of Texas Medical Branch, 301 University Road, Galveston, Texas 77555, USA School of Medicine, University of Washington, School of Medicine, Seattle, Washington, USA University of California, San Diego Medical Center, 200 W. Arbor Drive, San Diego, California 92103-8756, USA School of Medicine, Louisiana State University, 1501 Kings Highway, Shreveport, Louisiana 71130-3932, USA

Published Online: April 11, 2001

Abstract. During the approximately 20 years that percutaneous abscess drainage (PAD) has been an extant procedure and as the millennium begins, PAD has become, by consensus, the treatment of choice for abscesses. Indications for PAD continue to expand, and currently almost all abscesses are considered amenable. On occasion, PAD is an adjunctive procedure that provides a beneficial temporizing effect for the surgeon who eventually must operate for a coexisting problem such as a bowel leak. Simple unilocular abscesses are cured almost uniformly by PAD; more complicated abscesses, such as those with enteric fistulas (e.g., diverticular abscess) or pancreatic abscesses, have cure rates ranging from 65% to 90%. Various catheters and insertion techniques have proven effective. Ultrasonography, computed tomography, and fluoroscopy are the staple modalities that guide PAD. PAD is the prototype interventional radiology procedure, providing detection of the abscess by imaging, needling for diagnosis, and catheterization for therapy.

Percutaneous abscess drainage (PAD) has become the procedure of choice for treating most intraabdominal abscesses [1, 2]. There is a natural progression from detection of the abscess by ultrasonography (US) or computed tomography (CT), to needle aspiration for diagnosis, to catheter drainage for treatment. In-hospital follow-up is essential after catheter placement, as adjustments in catheter position and number are occasionally required to ensure clinical improvement in the patient’s condition. Repeat imaging with CT or US is indicated in patients who do not respond in a timely fashion. A host of intraabdominal sites in which abscesses accumulate are amenable to PAD. Abscesses in peritoneal compartments, liver, pancreas, and pelvis are among the most common that undergo PAD. Several fundamentals are standard with PAD: concomitant administration of intravenous antibiotics, aggressive drainage and irrigation, and follow-up by the radiologist. Particularly with complex abscesses, a cooperative strategy between the surgeon and interventional radiologist optimizes treatment and outcomes. Correspondence to: E. vanSonnenberg, M.D.

Imaging Prior to PAD In most centers, CT is the preferred modality for imaging abscesses. Once an abscess is identified by CT, US may be perfectly suited to guide catheter insertion (Fig. 1). Thus superficial abscesses, peripheral liver abscesses, and those in peritoneal spaces that abut the abdominal wall are well accessed with US. CT is chosen for guidance of pancreatic, interloop, and complex abscesses. Magnetic resonance imaging has no clinical role in PAD. The CT scan has the imaging advantage of being able to visualize the entire abdomen despite distended bowel or overlying bandages that otherwise might detract from, or preclude, US examination. CT depicts the entire abdomen and retroperitoneum. The characteristics of the abscess also may be defined by CT, including gas in the abscess, trapped bubbles, associated tumor, and contrast leakage that suggests an abnormal communication. Not infrequently patients have more than one abscess. CT is valuable for its ability to detect these multiple collections. These multiple abscesses may be juxtaposed or remote from each other. For example, a right lower quadrant or pelvic periappendiceal abscess may have a second collection in the subhepatic or even subphrenic space. If the patient is doing well, follow-up US or CT imaging is unnecessary. If there is a possible communication of the abscess to the gastrointestinal (GI) tract, pancreatic duct, bile duct, or renal collecting system, a catheter sinogram or abscessogram should be performed for initial documentation and as a baseline for followup.

Diagnostic Aspiration Under either US or CT guidance, a fine needle is inserted into the abscess or fluid collection for diagnosis. The needle provides guidance for catheter placement by defining a safe access route. If

vanSonnenberg et al.: Percutaneous Abscess Drainage

363

the needle is misplaced, a second, even finer, needle can be inserted until the appropriate route is determined. The fine needle also permits aspiration of a small amount of fluid, which is sent to the laboratory for evaluation. If pure pus is extracted, under proper clinical conditions the catheter is placed immediately. However, if there is a question as to the origin and etiology of the contents, a Gram stain is performed. A common finding with the Gram stain is the presence of white blood cells but no bacteria, a so-called sterile abscess. This is typical in patients who have been hospitalized and have had extensive antibiotic administration. Extracted fluid may be characteristic for collections other than abscesses. Creatinine is elevated in urinomas; the presence of lymphocytes and fat globules signifies lymphoceles; amylase characterizes pseudocysts; and bilirubin is diagnostic for bile collections. Seromas typically yield clear yellowish fluid. Specific Abscess Sites Liver Abscesses Pyogenic liver abscesses are a common indication for PAD. Solitary abscesses may be drained with ultrasound guidance. Those that are more complex and multiloculated and those that require the use of more than one catheter are best drained with CT guidance. A particularly challenging liver abscess is one that is located cephalad, near the dome of the diaphragm. The optimal guidance modality under these circumstances may be combined US and fluoroscopy to best achieve the appropriate angulation. CT with an angled gantry also can be used to access these collections. Clinically, virtually all pyogenic liver abscesses require drainage [3]. The adage that “the sun should never set on an abscess” is germane, as these patients frequently are quite toxic. Typical causes for pyogenic liver abscesses include extension of gallbladder and gallstone disease, prior trauma or surgery, and seeding from sepsis, (e.g., diverticulitis). Although amebic abscesses are preferably treated with antibiotics such as metronidazole, certain situations prompt percutaneous drainage [4 – 6] including poor response to antibiotic therapy, uncertainty as to the diagnosis, imminent or even actual leakage or rupture, size larger than 6 to 8 cm in diameter [7], and left lobe abscesses that have a proclivity to rupture into the thorax. These abscesses usually are unilocular and can be drained with ultrasound guidance alone [8]. Catheters need not remain for extended periods; 3 to 5 days is usually ample time for cure (along with concurrent antibiotics) [3]. Bile collections from the liver are usually secondary to trauma or to a recent operation. Percutaneous drainage may be indicated because of pain or secondary infection. A catheter sinogram after evacuating the fluid often results in a visible communication to the bile ducts. With time and external drainage, these communica-

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ Fig. 1. Percutaneous drainage of a hepatic abscess. A. A 42-year-old man complained of right upper quadrant pain and fever. Sagittal sonogram of the right upper quadrant demonstrates a large complex mass in the posterior aspect of the right lobe of the liver. B. A 12F sump catheter (arrows) was inserted into the abscess by an anterior approach. C. After evacuation of 55 cc of purulent material, drainage largely ceased. The catheter was removed 4 days later. Note that the abscess is completely evacuated.

364

World J. Surg. Vol. 25, No. 3, March 2001

Fig. 2. Percutaneous drainage of liquefied pancreatic necrosis in a 32year-old man with gallstone pancreatitis. A. Computed tomography (CT) scan demonstrates abnormal density and gas in the pancreatic bed. His temperature was 104°C, and the white blood cell (WBC) count was 31,000/ml. B. Fine-needle aspiration and localization. Gas and a small amount of brownish fluid were extracted. C. Tandem trocar technique for

catheter drainage. Note the fine needle and the adjacent 14F sump catheter. Multiple gram-negative organisms were cultured. D. Catheter enters the pancreatic bed from a flank approach. Brownish fluid (95 cc) with debris was extracted. The patient recovered and was discharged from the hospital after 8 days.

tions usually seal, and the catheter then can be removed without sequelae. Echinococcal abscesses are endemic in certain areas worldwide, but are extremely uncommon in the United States. Although there is concern about the development of anaphylaxis, substantial experience has accumulated in the Middle East suggesting that these abscesses usually can be treated safely by percutaneous means [9].

tissue and therefore completely undrainable by percutaneous means. If treated, surgical necrosectomy is preferred. Conversely, liquefaction of necrosis can occur, and in this state the collection may be amenable to percutaneous drainage (Fig. 2). Large-bore catheters are used for liquefied necrosis (20 –30F), thereby permitting evacuation of typical brownish fluid that commonly contains debris. The radiologist is occasionally asked to perform diagnostic needle aspiration alone to determine if there is infected necrosis, the most severe complication of pancreatitis; a positive Gram stain or culture often triggers surgery for débridement and necrosectomy. Pancreatic abscesses are not seen until well into the course of pancreatitis, typically longer than 3 to 4 weeks. Multiple, complex collections often are present and require an adequate number and size of catheters. Another cause of pancreatic abscess is after operation on the pancreas itself, such as with a partial pancreatectomy, or from operations on adjacent organs during which the pancreas is inadvertently injured. An occasional source of confusion with pancreatic collections is differentiation of a pancreatic pseudocyst from a cystic pancreatic tumor. The usual procedure of choice for these cystic tumors is

Pancreatic Abscesses Pancreatic collections are diverse, and percutaneous options mirror the complexities of these collections. Thus infected pancreatic pseudocysts and pancreatic abscesses are generally drained routinely and urgently [10 –12]. On the other hand, noninfected pancreatic pseudocysts may be simply observed in a substantial number of cases. Indications for drainage of noninfected pseudocysts include pain, imminent leak or rupture, and obstruction of the biliary or GI tract by the pseudocyst. Perhaps the major challenge with pancreatitis is the diagnosis and management of pancreatic necrosis. The necrosis can be solid

vanSonnenberg et al.: Percutaneous Abscess Drainage

365

surgical extirpation, although on occasion needle aspiration is performed because of an uncertain diagnosis or because of the compromised clinical status of the patient, that may preclude operation. Fluid from a cystic malignant pancreatic tumor is dark and may contain elevated carcinoembryonic antigen (CEA). Renal and Perirenal Abscesses Renal and perirenal abscesses often are associated with nephrolithiasis, and there may be hydronephrosis as well. Septicemia is another cause of renal abscesses; in this situation the abscesses occasionally are multiple, bilateral, or both. Renal abscesses may rupture through the renal capsule into the perirenal space. These abscesses often are quite large. As with intrarenal abscesses, perirenal abscesses usually are amenable to PAD. If renal obstruction is associated with a renal or perirenal abscess, percutaneous nephrostomy may be needed for access to stone removal or to relieve a malignant or benign obstruction. Enteric Abscesses Percutaneous aspiration and catheter drainage permit numerous options for the surgeon and interventional radiologist when caring for patients with enteric abscesses (periappendiceal, peridiverticular) (Fig. 3). In particular, periappendiceal and peridiverticular abscesses can be drained preoperatively; alternatively, PAD may be applied in select cases to obviate all operations or may be utilized postoperatively for a persistent or new abscess. Preoperative drainage permits the surgeon to remove the appendix electively at a later date or remove a portion of the sigmoid colon (or other portion) with diverticulitis when the patient is stabilized and nonseptic. In patients who are poor operative candidates, draining the abscess may allow all symptoms to abate and not to recur; this is particularly helpful in elderly patients with severe cardiopulmonary disease in whom surgery is undesirable [13–18]. A pitfall with presumed periappendiceal and peridiverticular abscesses is overlooking an underlying necrotic carcinoma that has become infected. The infected necrotic tumor may be drained successfully. A subsequent barium study or colonoscopy may be performed to determine if an underlying tumor is present. Postoperative periappendiceal and peridiverticular abscesses generally are straightforward in requiring percutaneous drainage. A preoperative CT scan that demonstrates a periappendiceal or peridiverticular abscess may predictably mean that surgery will be more difficult and thus imply that preoperative drainage would be of benefit. Crohn’s-related abscesses may be treated with PAD [15, 16]. CT guidance is used to avoid transgressing bowel with the catheter. Catheters should remain in place long enough to allow any enteric communications to close, although this does not occur in all cases. Anastomotic Abscesses Leakage of enteric contents at an anastomosis may lead to walling off and development of an abscess. When this occurs, PAD is indicated. As these anatomic situations are usually complex, CT guidance typically is preferred [14]. The strategy for treating these abscesses is to catheterize the collection and evacuate the contents, obtain a catheter sinogram or abscessogram to demonstrate

Fig. 3. Periappendiceal abscess drainage. A. CT scan of a 23-year-old man reveals a large inhomogeneous collection at the base of the cecum. The patient had a fever of 102°F, WBC count of 17,000/ml, and right lower quadrant pain and tenderness. B. Percutaneous 12F catheter drainage of the periappendiceal abscess. Escherichia coli grew from the abscess contents. The patient recovered and was discharged from the hospital 2 days later. He subsequently underwent interval elective appendectomy 3 weeks later.

communication to the GI tract, and allow time for the anastomotic leak to both heal and seal. In some cases the latter does occur, whereas in others the surgeon must reoperate to repair the anastomotic site. During drainage, enteric decompression and diversion of bowel contents are accomplished concomitantly with nasogastric or nasoenteric tubes. Subphrenic Abscesses Subphrenic abscesses offer special challenges in regard to access and occasionally determining the origin of the abscess [19]. The relatively high and cephalad nature of the collections suggest considerations similar to those with abscesses in the dome of the liver. Angulation of the CT gantry and use of US combined with fluoroscopy are valuable for gaining safe access. Occasionally the pleura is transgressed. Usually this is not of clinical concern, but an empyema or even a pneumothorax may develop as a complication. This occurrence may dictate the need for a pleural drainage catheter. As the subphrenic space is in communication potentially with

366

all other peritoneal compartments, abscesses that accumulate in the subphrenic spaces may have remote origins. For example, there may be secondary sites (e.g., subphrenic space) of pus accumulation from primary periappendiceal, postoperative, pelvic, or even pericholecystic abscesses. An important tenet is that each defined abscess requires drainage and needs its own specific catheter. Splenic Abscesses Albeit uncommon, splenic abscesses have a host of interesting etiologies [20]. Perhaps the most frequent nowadays are microabscesses, often from Candida; they are seen typically in immunocompromised patients. Posttraumatic superinfection of a hematoma is another cause of splenic abscess. Similarly, infected infarcts are seen in septicemic patients. Splenic cysts may become infected, and even the uncommon intrasplenic pancreatic pseudocyst may be infected. Considerations similar to those with other abscesses are useful and generally effective when treating splenic abscesses. Pelvic Abscesses Aside from enteric and postoperative abscesses, tuboovarian abscess (TOA) is a frequently encountered lesion that can be treated successfully by PAD [21]. Probably the main issue with pelvic abscesses is the route for PAD [22]. The options are many: directly through the anterior abdominal wall, posteriorly via a transgluteal approach [13], transrectally as with a surgical approach [23], and transvaginally [24]. The latter is an approach well-suited to US guidance. For deep interloop abscesses that are completely encircled by bowel, transenteric needle aspiration alone without catheter placement may be the only percutaneous option [25]. Prostatic abscesses usually are small and may need only needle aspiration combined with antibiotics [26]. Thoracic Drainage A host of collections in the chest is amenable to PAD, including free-flowing pleural effusions, loculated effusions, empyema, infected bronchopleural fistula, mediastinal abscess, lung abscess, pericardial empyema, pneumothorax, and cystic collections. Ultrasonography is the preferred guidance modality for the more superficial and discrete collections. CT is opted with the deeper, loculated, multiloculated, and more complex collections [27–31]. A common scenario is that radiologic drainage is requested after some form of ward therapy that has been inadequate. Examples are an attempted thoracentesis that yields no fluid, a lung abscess that has been treated with antibiotics and respiratory stimulation, or a mediastinal collection in which supportive therapy and antibiotics are unlikely to be successful in an extremely ill patient. Ultrasound- or CT-guided thoracentesis is indicated for failed ward taps, for the patient who has chronic obstructive pulmonary disease (COPD) and little reserve and in whom a pneumothorax would be quite troublesome, and for ventilator-dependent patients. Drainage of empyemas is prompted by standard indications, such as infection and compromise of respiratory status; 12F to 20F catheters are used. For empyemas that are quite thick and do not drain well with conventional catheters, lytic agents are

World J. Surg. Vol. 25, No. 3, March 2001

instilled. These agents presumably break disulfide bonds, allowing lysis of the otherwise thick material. The PAD technique often dramatically improves the condition of patients stricken with a lung abscess (Fig. 4). These patients are often chronically ill and have been undergoing antibiotic and respiratory therapy. Bronchoscopy frequently is performed to remove mucous plugs or a foreign body. Persistent sepsis can be rapidly eliminated by catheter drainage and evacuation of the pus. Cardinal technical rules of PAD are not to violate normal lung en route to the abscess (unless no other route exists) and not to place the patient such that the normal lung is dependent, thereby preventing contralateral spill of pus. Mediastinal abscesses are generally seen in postoperative esophageal or cardiac patients, who are frequently quite ill. As the abscesses often are deep, CT placement of multiple catheters is typical. Catheters are placed on suction, and the abscesses are aggressively evacuated. Pneumothorax is routinely treated percutaneously by radiologic guidance. Probably the most frequent use of radiologic drainage is during or after percutaneous lung or mediastinal biopsy. It is indicated in patients who are dyspneic, for those in whom the pneumothorax is enlarging on sequential chest radiographs, or to continue the biopsy. Small catheters (6 –10 F) suffice for drainage. The catheters may be attached to a water seal, suction drainage, or a Heimlich valve. Other Abscesses Abscesses in the thigh [32], psoas muscle [33], and breast [34] are all accessible to PAD. US guidance suffices, as these lesions are usually located superficially. Not infrequently an attempt at drainage on the ward, in an office, or in an emergency room by clinicians has been unsuccessful because complex or solid inflamed areas have been visualized or palpated that do not yield fluid on needle aspiration. US depicts the most liquid portion of the collection, thereby optimizing success for needle aspiration. Psoas abscesses may have remote etiologies, such as the GI tract, spine, or kidney. Epidural abscesses and spondylodiskitis [35–37] may be relieved by catheter drainage. Unlike other abscesses, transcatheter infusion of antibiotics may be valuable in these conditions. Results and Complications Abscesses can be classified into three groups, organized by complexity and reflected in success and failure rates. The unilocular and discrete abscess is simplest to drain and is cured in more than 90% of cases. The medium-complexity abscess (i.e., abscesses that have a communication to the GI tract) are cured in 80% to 90% of cases. An associated GI communication may require an operation, but if surgery is performed after PAD, the surgeon’s task is simplified because of the noninfected bed. The most complicated collections are, for example, intermixed pancreatic abscess/necrosis, infected tumor, and tenacious organized empyemas. The cure rates in these latter situations are no higher than around 80% and may be as low as 30% to 50%. Table 1 summarizes the results and complications for reported series of abscess drainage. Complications reported with PAD related to bleeding include oozing, frank hemorrhage, and liver and splenic laceration. Enteric complications such as transcatheter puncture of the colon

vanSonnenberg et al.: Percutaneous Abscess Drainage

367

Table 1. Representative series on percutaneous abscess drainage. Study

Fig. 4. A 29-year-old woman had persistent fever due to pneumonia and a lung abscess despite 2 weeks of antibiotic therapy. A. Scout image from a CT scan demonstrates residual right upper lobe pneumonia and a lung abscess. B. Tandem trocar approach via an anterior route demonstrates the 10F catheter in the lung abscess. C. After evacuation of 25 cc of pus, the catheter is coiled in the residual cavity with adjacent parenchymal consolidation. Culture was negative, presumably because of long-term antibiotics. The patient responded well, and the catheter was removed 5 days later.

with a resultant fistula are noted with PAD, similar to percutaneous nephrostomy access. Peritonitis from spilled bowel contents or spilled pus and generalized sepsis have been occurred, albeit rarely.

General Wroblicka [25] Bufalari [38] Shuler [39] Goletti [40] Brolin [41] McNicholas [19] Jang [42] Voros [43] Lang [44] Gerzof [45] Liver Rajak [3] Ch [46] Tazawa [47] Chu [48] Hashimoto [49] Filice [9] Seeto [50] Pancreas Spivak [51] Chen [52] vanSonnenberg [12] Mithofer (A) [10] Mithofer (B) [10] Himal [53] vanSonnenberg [11] Pelvis Nelson [54] vanSonnenberg [24] Casola [21] Tumor Civardi [55] Mueller [56] Iliopsoas Kang [57] Gupta [33] Retroperitoneum Paley [58] Siegel [59] Enteric Jamieson [60] Bernini [14] Sahai [15] Ayuk [61] vanSonnenberg [17] Casola [16] Neff [62] Stabile [18] Thoracic vanSonnenberg [29] Other Staatz [37] Garg [34] Barozzi [26]

Year

No. of cases

Success

1998 1996 1996 1993 1984 1995 1995 1996 1986 1985

97 27 18 200 119 25 10 185 136 125

87/97 (90%) 23/27 (85%) 12/18 (67%) (88%) (76%) 22/25 (88%) 9/10 (90%) 166/185 (92%) 105/136 (77%) 92/125 (73.6%)

4 1 19 7 11

1998 1997 1997 1996 1995 1990 1996

25 101 48 27 39 31 49

25/25 (100%) (96.8%) 46/48 (95%) 24/27 (88.9%) 27/39 (69.2%) 31/31 (100%) 90%

0 3 2 — 4 — —

1998 1997 1997 1997 1997 1991 1989

27 17 59 31 14 14 101

1994 1991 1992

31 14 16

26/31 (84%) 12/14 (86%) 15/16 (94%)

8 0 3

1998 1989

14 16

12/14 (86%) 12/16 (75%)

0 0

1998 1996

14 51

10/14 (71%) 35/51 (68.6%)

0 0

1996 1996

12 33

12/12 (100%) 25/33 (75.7%)

0

1997 1997 1997 1996 1987 1987 1987 1990

46 82 27 14 21 15 16 19

42/46 (91%) 62/82 (76%) 15/27 (56%) 6/14 (42.8%) 19/21 (90.4%) 15/15 (100%) 16/16 (100%) 14/19 (74%)

3 0 4 0 0 0 0 0

1991

19

69/19 (84%)

1

1998 1997 1998

33 25 5

31/33 (94%) (84%) 5/5 (100%)

0 0 0

17/27 (63%) 16/17 (94%) 51/59 (86%) 9/29 (31%) 14/14 (100%) 12/14 (85.7%) 91/101 (90.1%)

Complications 2 3 2

0 0 3

1 10

Situations more likely to result in failure of PAD include those associated with tumor, phlegmon, and necrotic tissue. It has been noted [14] that a high APACHE II score indicates a poor likelihood of success with abscess drainage. In most series, the complication rate is within 5%.

368

Conclusions As the millennium gets under way, the treatment of abscesses has become largely image-guided with percutaneous catheters. Most abscesses may be cured or can undergo a beneficial temporizing effect that, if surgery is necessary, aids the surgeon by relieving sepsis and affording a healthier operative bed [63]. Results vary according to the complexity of the abscess; unilocular abscesses are cured almost uniformly, whereas draining the most complex abscesses more realistically temporizes the situation as a goal rather than providing cure. Although complications are relatively uncommon, they may be serious and even life-threatening. Close communication and strategy between the surgeon and interventional radiologist provide synergy that ultimately benefits the patient. Résumé Dans les 20 dernières années, le drainage percutané (DP) a pris sa place dans l’arsenal thérapeutique. A l’approche du nouveau millénium, il est devenu, par consensus, le traitement de choix pour le traitement des abcès. Les indications du DP continue de s’étendre; à présent, presque tous les abcès peuvent être traités ainsi. De temps à autre, le DP peut être utilisé comme un procédé adjuvant temporaire lorsque, par exemple, le chirurgien doit s’occuper d’un problème annexe tel une fistule anastomotique. On arrive à guérir presque tous les abcès simples uniloculaires uniquement par le DP; pour les abcès plus compliqués, associés à des fistules (par ex un abcès en rapport avec une perforation diverticulaire) ou des abcès d’origine pancréatique, le taux de guérison est de 65–90%. Plusieurs systèmes de cathéters et différentes techniques d’insertion se sont montrés efficaces. La sonographie, la tomodensitométrie et la radio (fluoro)scopie, seules ou combinées, sont toutes efficaces pour guider le DP. Le DP est le prototype de procédé radiologique interventionnel: il permet la détection de l’abcès par l’imagerie, la ponction pour le diagnostique, et l’insertion du cathéter pour le traitement. Resumen Hace aproximadamente 20 años que se inició la punción percutánea como tratamiento de los abscesos (PAD); a medida que se acerca el nuevo milenio esta técnica se ha convertido en el procedimiento electivo para el tratamiento de los abscesos. Las indicaciones de la PAD han ido aumentando y, en la actualidad, todo absceso se considera subsidiario de una PAD. A veces, la PAD no es más que un tratamiento paliativo, temporalmente beneficioso, que no evita el tratamiento quirúrgico p. ej. cuando existe una fistula intestinal. Casi todos los abscesos simples, uniloculares se curan mediante la PAD; en los abscesos complicados producidos por fistulas entéricas (p.ej. abscesos diverticulares) o por pancreatitis, la tasa de curación con la PAD oscila entre el 65–90%. Han sido eficazmente empleadas diversas técnicas de punción y diferentes tipos de catéteres. La punción guiada por ecografía, TAC (CT), radioscopia o combinando varias técnicas de imaginería son las modalidades más empleadas en la PAD. La PAD constituye el procedimiento más frecuente de la radiología intervencionista. Detección del absceso mediante técnicas de imaginería, punción que confirma el diagnóstico y

World J. Surg. Vol. 25, No. 3, March 2001

colocación de un catéter como gesto terapéutico constituyen el tratamiento actual de los abscesos. Acknowledgments Our appreciation goes to Ms. Vicki McDowell and Ms. Faith McDaniel for preparation of the manuscript and to Ms. Linda Aldridge for technical assistance. References 1. Boland, G.W., Mueller, P.R.: Update on abscess drainage. Semin. Interv. Radiol. 13:27, 1996 2. vanSonnenberg, E., D’Agostino, H.B., Casola, G., Halasz, N.A., Sanchez, R.B., Goodacre, B.W.: Percutaneous abscess drainage: current concepts. Radiology 181:617, 1991 3. Rajak, C.L., Gupta, S., Jain, S., Chawla, Y., Gulati, M., Suri, S.: Percutaneous treatment of liver abscesses: needle aspiration versus catheter drainage. A.J.R. 170:1035, 1998 4. Shandera, W.X., Bollam, P., Hashmey, R.H., Athey, P.A., Greenberg, S.B., White, A.C., Jr.: Hepatic amebiasis among patients in a public teaching hospital. South. Med. J. 91:829, 1998 5. vanSonnenberg, E., Mueller, P.R., Schiffman, H.R., Ferrucci, J.T., Jr., Casola, G., Simeone, J.F., Cabrera, O.A., Gosink, B.B.: Intrahepatic amebic abscesses: indications and results of percutaneous catheter drainage. Radiology 156:631, 1985 6. Ken, J.G., vanSonnenberg, E., Casola, G., Christensen, R., Polansky, A.M.: Perforated amebic liver abscesses: successful percutaneous treatment. Radiology 170:195, 1989 7. Tandon, A., Jain, A.K., Dixit, V.K., Agarwal, A.K., Gupta, J.P.: Needle aspiration in large amoebic liver abscess. Trop. Gastroenterol. 18:19, 1997 8. Moazam, F., Nazir, Z.: Amebic liver abscess: spare the knife but save the child. J. Pediatr. Surg. 33:119, 1998 9. Filice, C., Pirola, F., Brunetti, E., Dughetti, S., Strosselli, M., Foglieni, C.S.: A new therapeutic approach for hydatid liver cysts: aspiration and alcohol injection under sonographic guidance. Gastroenterology 98:1366, 1990 10. Mithofer, K., Mueller, P.K., Warshaw, A.L.: Interventional and surgical treatment of pancreatic abscess. World J. Surg. 21:162, 1997 11. vanSonnenberg, E., Wittich, G.R., Casola, G., Brannigan, T.C., Karnel, F., Stabile, B.E., Varney, R.R., Christensen, R.: Percutaneous drainage of infected and noninfected pancreatic pseudocysts: experience in 101 cases. Radiology 170:757, 1989 12. vanSonnenberg, E., Wittich, G.R., Chon, K.S., D’Agostino, H.B., Casola, G., Easter, D., Morgan, R.G., Walser, E.M., Nealon, W.H., Goodacre, B., Stabile, B.E.: Percutaneous radiologic drainage of pancreatic abscesses. A.J.R. 168:979, 1997 13. Ryan, J.M., Murphy, B.L., Boland, G.W., Mueller, P.R.: Use of the transgluteal route for percutaneous abscess drainage in acute diverticulitis to facilitate delayed surgical repair. A.J.R. 170:1189, 1998 14. Bernini, A., Spencer, M.P., Wong, W.D., Rothenberger, D.A., Madoff, R.D.: Computed tomography-guided percutaneous abscess drainage in intestinal disease: factors associated with outcome. Dis. Colon Rectum 40:1009, 1997 15. Sahai, A., Belair, M., Gianfelice, D., Cote, S., Gratton, J., Lahaie, R.: Percutaneous drainage of intra-abdominal abscesses in Crohn’s disease: short- and long-term outcome. Am. J. Gastroenterol. 92:275, 1997 16. Casola, G., vanSonnenberg, E., Neff, C.C., Saba, R.M., Withers, C., Emarine, C.W.: Abscesses in Crohn’s disease: percutaneous drainage. Radiology 163:19, 1987 17. vanSonnenberg, E., Wittich, G.R., Casola, G., Neff, C.C., Hoyt, D.B., Polansky, A.D., Keightley, A.: Periappendiceal abscesses: percutaneous drainage. Radiology 163:23, 1987 18. Stabile, B.E., Puccio, E., vanSonnenberg, E., Neff, C.C.: Preoperative percutaneous drainage of diverticular abscesses. Am. J. Surg. 159:99, 1990 19. McNicholas, M.M.J., Mueller, P.R., Lee, M.J., Echeverri, J., Gazelle, G.S., Boland, G.W., Dawson, S.L.: Percutaneous drainage of

vanSonnenberg et al.: Percutaneous Abscess Drainage

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

subphrenic fluid collections that occur after splenectomy: efficacy and safety of transpleural versus extrapleural approach. A.J.R. 165:355, 1995 Kapur, A., Vasudeva, R., Howden, C.W.: Candida splenic abscess in the absence of obvious immunodeficiency. Am. J. Gastroenterol. 92:509, 1997 Casola, G., vanSonnenberg, E., D’Agostino, H., Harker, C., Varney, R.R., Smith, D.: Percutaneous drainage of tuboovarian abscesses. Radiology 182:399, 1992 Kastan, D.J., Nelsen, K.M., Shetty, P.C., Burke, M.W., Sharma, R.P.: Combined transrectal sonographic and fluoroscopic guidance for deep pelvic abscess drainage. J. Ultrasound Med. 15:235, 1996 Pereira, J.K., Chait, P.G., Miller, S.F.: Deep pelvic abscesses in children: transrectal drainage under radiologic guidance. Radiology 198:393, 1996 vanSonnenberg, E., D’Agostino, H., Casola, G., Goodacre, B., Sanchez, R., Taylor, B.: US-guided transvaginal drainage of pelvic abscesses and fluid collections. Radiology 181:53, 1991 Wroblicka, J.T., Kuligowska, E.: One-step needle aspiration and lavage for the treatment of abdominal and pelvic abscesses. A.J.R. 170:1197, 1998 Barozzi, L., Pavlica, P., Menchi, I., De Matteis, M., Canepari, M.: Prostatic abscess: diagnosis and treatment. A.J.R. 170:753, 1998 Ball, W.S., Jr., Bisset, G.S., III, Towbin, R.B.: Percutaneous drainage of chest abscesses in children. Radiology 171:431, 1989 Westcott, J.L.: Percutaneous catheter drainage of pleural effusion and empyema. A.J.R. 144:1189, 1985 vanSonnenberg, E., D’Agostino, H.B., Casola, G., Wittich, G.R., Varney, R.R., Harker, C.: Lung abscess: CT-guided drainage. Radiology 178:347, 1991 Klein, J.S., Schultz, S., Heffner, J.E.: Interventional radiology of the chest: image-guided percutaneous drainage of pleural effusions, lung abscess, and pneumothorax. A.J.R. 164:581, 1995 vanSonnenberg, E., Wittich, G.R., Goodacre, B.W., Zwischenberger, J.B.: Percutaneous drainage of thoracic collections. J. Thorac. Imaging 13:74, 1998 vanSonnenberg, E., Wittich, G.R., Casola, G., Cabrera, O.A., Gosink, B.B., Resnick, D.L.: Sonography of thigh abscesses: detection, diagnosis, and drainage. A.J.R. 149:769, 1987 Gupta, S., Suri, S., Gulati, M., Singh, P.: Ilio-psoas abscesses: percutaneous drainage under image guidance. Clin. Radiol. 52:704, 1997 Garg, P., Rathee, S.K., Lal, A.: Ultrasonically guided percutaneous drainage of breast abscess. J. Indian Med. Assoc. 95:584, 1997 Hori, K., Kano, T., Fukushige, T., Sano, T.: Successful treatment of epidural abscess with a percutaneously introduced 4-French catheter for drainage. Anesth. Analg. :1384,1386 1997 Borowski, A.M., Crow, W.N., Hadjipavlou, A.G., Chaljub, G., Mader, J., Cesani, F., vanSonnenberg, E.: Percutaneous management of pyogenic spondylodiskitis. A.J.R. 170:1587, 1998 Staatz, G., Adam, G.B., Keulers, P., Vorwerk, D., Gunther, R.W.: Spondylodiskitic abscesses: CT-guided percutaneous catheter drainage. Radiology 208:363, 1998 Bufalari, A., Giustozzi, G., Moggi, L.: Postoperative intraabdominal abscesses: percutaneous versus surgical treatment. Acta Chir. Belg. 96:197, 1996 Shuler, F.W., Newman, C.N., Angood, P.B., Tucker, J.G., Lucas, G.W.:Nonoperative management for intraabdominal abscesses. Am. Surg. 62:218, 1996 Goletti, O., Lipplois, P.V., Chiarugi, M., Ghiselli, G., De Negri, F., Conte, M., Ceragioli, T., Cavina, E.: Percutaneous US-guided drainage of intra-abdominal abscesses. Br. J. Surg. 80:336, 1993 Brolin, R.E., Nosher, J.L., Leiman, S., Lee, W.S., Greco, R.S.: Percutaneous catheter versus open surgical drainage in the treatment of abdominal abscesses. Am. Surg. 50:102, 1984

369

42. Jang, T.N., Fung, C.P.: Treatment of pyogenic splenic abscess: nonsurgical procedures. J. Formos. Med. Assoc. 94:309, 1995 43. Voros, D., Gouliamos, A., Kotoulas, G., Kouloheri, D., Saloum, G., Kalovidouris, A.: Percutaneous drainage of intraabdominal abscess using large lumen tubes under computed tomographic control. Eur. J. Surg. 162:895, 1996 44. Lang, E.K., Springer, R.M., Glorioso, L.W., III, Cammarata, C.A., : Abdominal abscess drainage under radiologic guidance: causes of failure. Radiology 159:329, 1986 45. Gerzof, S.G., Johnson, W.C., Robbins, A.H., Nabseth, D.C.: Expanded criteria for percutaneous abscess drainage. Arch. Surg. 120:227, 1985 46. Ch, Y.S., Hg, L.R., Kan, P.S., Metreweli, C.: Pyogenic liver abscess: treatment with needle aspiration. Clin. Radiol. 52:912, 1997 47. Tazawa, J., Sakai, Y., Maekawa, S., Ishida, Y., Maeda, M., Marumo, F., Sato, C.: Solitary and multiple pyogenic liver abscesses: characteristics of the patients and efficacy of percutaneous drainage. Am. J. Gastroenterol. 92:271, 1997 48. Chu, K.M., Fan, S.T., Lai, E.C.S., Lo, C.M., Wong, J.: Pyogenic liver abscess. Arch. Surg. 131:148, 1996 49. Hashimoto, L., Hermann, R., Broniatowski, S.: Pyogenic hepatic abscess: results of current management. Am. Surg. 61:407, 1995 50. Seeto, R.K., Rockey, D.C.: Pyogenic liver abscess. Medicine 75:99, 1996 51. Spivak, H., Galloway, J.R., Amerson, J.R., Fink, A.S., Branum, G.D., Redvanly, R.D., Richardson, W.S., Mauren, S.J., Waring, J.P., Hunter, J.G.: Management of pancreatic pseudocysts. J. Am. Coll. Surg. 186:507-511, 1998 52. Chen, C., Chen, P.J., Yang, P.M., Huang, G.T., Lai, M.Y., Tsang, Y.M., Chen, D.S.: Clinical and microbiological features of liver abscess after transarterial embolization for hepatocellular carcinoma. Am. J. Gastroenterol. 92:2257, 1997 53. Himal, H.S., Cusimano, R.J.: Pancreatic pseudocysts: the role of percutaneous catheter drainage. Cay J. Surg. 35:403, 1992 54. Nelson, A.L., Sinow, R.M., Renslo, R., Renslo, J., Atamdede, F.: Endovaginal ultrasonographically guided transvaginal drainage for treatment of pelvic abscesses. Am. J. Obstet. Gynecol. 172:1926, 1995 55. Civardi, G., Filice, C., Caremani, M., Giorgio, A., Vallisa, D., Berte, R., Cavanna, L.: Clinical efficacy of ultrasound guided percutaneous drainage of abscesses in patients with leukaemia and lymphoma. Eur. J. Cancer 34:580, 1998 56. Mueller, P.R., White, E.M., Glass-Royal, M., Keman, R.K., Saini, S., Silverman, S.G., Hahn, P.F., Simeone, J.F.: Infected abdominal tumors: percutaneous catheter drainage. Radiology 173:627, 1989 57. Kang, M., Gupta, S., Gulati, M., Suri, S.: Ilio-psoas abscess in the paediatric population: treatment by US-guided percutaneous drainage. Pediatr. Radiol. 28:478, 1998 58. Paley, M., Sidhu, P.S., Evans, R.A., Karani, J.B.: Retroperitoneal collections—aetiology and radiological implications. Clin. Radiol. 52:290, 1997 59. Siegel, J.F., Smith, A., Moldwin, R.: Minimally invasive treatment of renal abscess. J. Urol. 155:52, 1996 60. Jamieson, D.H., Chait, P.G., Filler, R.: Interventional drainage of appendiceal abscesses in children. A.J.R. 169:1619, 1997 61. Ayuk, P., Williams, N., Scott, N.A., Nicholson, D.A., Irving, M.H.: Management of intraabdominal abscesses in Crohn’s disease. Ann. R. Coll. Surg. Engl. 78:5, 1996 62. Neff, C.C., vanSonnenberg, E., Casola, G., Wittich, G.R., Hoyt, D.B., Halasz, N.A., Martini, D.J.: Diverticular abscesses: percutaneous drainage. Radiology 163:15, 1987 63. vanSonnenberg, E., Wing, V.W., Casola, G., Coons, H.G., Nakamoto, S.K., Mueller, P.R., Ferrucci, J.T., Halasz, N.A., Simeone, J.F.: Temporizing effect of percutaneous drainage of complicated abscesses in critically ill patients. A.J.R. 142:821, 1984

370

Invited Commentary DOI: 10.1007/s002680020387

Benno Stinner, M.D. Department of General Surgery, Philipps University, Marburg, Germany Published Online: April 11, 2001

This recent survey of vanSonnenberg and coworkers delineates the development of percutaneous abscess drainage (PAD) over the years from a more restrictive view with regard to the statement that an attempt should be made to drain almost any abscess by PAD first. With the percutaneous route introduced, more has been learned about surgical treatment of abscesses, but care must be taken regarding the level of evidence guiding PAD recommendations. Most series simply reflect the experience of an expert, some report on uncontrolled studies, and some use ill-matched controls or even historical comparison rather than the gold standard of evidence-based medicine, the prospective randomized trial. Indications for PAD were kept restrictive during the early 1980s and were predominantly reserved for the single, easily accessible collection. Absolute contraindications for surgeons [1] and relative contraindications for radiologists were complex multilocular collections, those with a difficult access such as postoperative abscesses, phlegmonous lesions, and collections containing large amounts of sequester and debris (i.e., pancreatic abscesses) [2]. Although each professional group favored its own type of solution, endpoints for treatment success must be clearly defined and held comparable regardless of the technique applied. Definitive success must comprise resolution of the septic process and the principle of doing the least harm to the patient. Initial successful removal of a fluid collection alone does not mean success of inherent value. Indications for PAD were early expanded to treat complicated abscesses as a temporizing measure. In an observational study of 15 patients with complicated abscesses, the authors tried to identify predictive factors. Neither size, depth, drainage route, nor etiology of the abscess (spontaneous versus postoperative) could be correlated with success, favoring vanSonnenberg et al.’s attitude [3]. In another study no difference in success could be documented regarding surgical or interventional therapy; both methods came out to be only the final executing step of the whole treatment process [4]. A retrospective analysis of 91 patients undergoing drainage of postoperative intraabdominal abscesses by PAD or surgically claimed to identify a high APACHE II score as the predictor of failure. Methodologic restrictions (treatment allocation, not randomized) limits evidence-based conclusions, but these results do not favor the attractive opinion that the less invasive PAD is clearly best for the elderly and frail patient [5, 6]. This is especially true if it is to be followed by surgery, as is often the case with complex abscess formations [7]. Once PAD fails, the mortality rate may be as high as 75% [8]. Even in the view of intestinal surgeons, inaccessibility of an abscess (unless traversing pleura or abdominal viscus) is a strict contraindication. With this widely explorative policy, the failure rate (urgent or emergency surgery) was 21%

World J. Surg. Vol. 25, No. 3, March 2001

(4/19) in those with diverticulitis, 31% (9/29) in those with anastomotic leakage, 40% (2/5) in patients with Crohn’s disease, and 75% (3/4) in those with nonoperative perforation. About 42% (8/19) of the diverticulitis patients and 20% (1/5) of the Crohn’s patients could be converted to elective surgery, which does not prove overall benefit from the procedure. Only 34% of severely ill patients (APACHE II . 15) with complex abscesses underwent successful drainage. An exact morphologic definition of the problem before intervention is recommended, contrary to the myth that surgical exploration gives the definitive insight into the abdominal cavity [9]. There are numerous reports comparing PAD and surgical drainage; but according to MEDLINE search criteria there are no prospective randomized trials on this topic. Most of the studies date to the 1980s, and comparisons were made on the level: “27 patients treated over the last 5 years,” with 43 patients treated by operative interventions over the last 10 years,” and the result that “PAD is at least as efficacious as operative drainage” [10]. More experience with interventional radiology changed the conclusion to the more defensive: “surgical drainage for intraabdominal abscesses is as successful as PAD,” suggesting that PAD would have been the preferable measure. The actual message of this report was that frequent clinical reassessment is mandatory to change the procedure in cases of clinical failure [11]. Reviewing the pros and cons of the various therapeutic approaches, better preoperative localization and avoidance of detrimental physiologic derangement prior to therapy were considered the success factors for treating abdominal abscesses. This is well in line with the documented experience that high APACHE II scores correlate with high morbidity regardless of the treatment method chosen [6, 12]. Positively, the general view is increasing that the two methods are complementary rather than competitive, and clinical judgment should not to be governed by imaging procedures alone. Three examples of vanSonnenberg et al.’s review require some additional “surgical” comment. The definitive role of percutaneous abscess drainage for periappendiceal abscesses remains uncertain because treatment alternatives of the underlying disease are still undefined. In most European countries and many AngloAmerican centers, surgical drainage and removal of the appendix together with the abscess is claimed to be “standard,” although this operation is associated with a higher complication rate than with resection of the nonperforated appendix (18% vs. 10%) [13]. PAD has been suggested to be one treatment option leading to avoidance of subsequent appendectomy in some cases, although the hospital stay of 3 to 23 days was well within the range for open surgery [14]. Recently, PAD has been advocated for periappendiceal abscesses in children, with subsequent surgery required in only 4 of 46 cases [15]. Keeping in mind the low morbidity and short hospital stay (mean 4.7 days) for patients with open surgery [16], the benefit is dubious. If the hospital stay and morbidity are not really different, it is a matter of debate whether, for otherwise healthy children, it is of any additional benefit to be catheterized under local anesthesia percutaneously or via a transrectal route, rather be operated on once. The peridiverticular abscess is a typical example how “straightforward” assessment of a new method enters surgical practice without final proof of a prospective randomized trial. What appears to be clear in the management of complicated diverticular disease is that the CT scan provides valuable information for

vanSonnenberg et al.: Percutaneous Abscess Drainage

371

Fig. 1. CT scan of a 76-year-old septic patient with perforated diverticular disease and spread of a gas-forming infection in the retroperitoneum to the distal thigh. A. Gas-containing abscess ventral to left-sided os ileum. B, C. Extension of the gas spread to the ventral thigh down to the knee. D. Puncture of the proximal part of the infected area for guiding percu-

taneous abscess drainage (PAD) insertion. There was no clinical improvement after PAD. Salvage surgery (Hartmann procedure, débridement) was performed 12 hours later. Obviously there was no indication for PAD in this case.

guiding further treatment and such early CT may even reduce costs [17]. The feasibility of draining peridiverticular abscesses has also been proven in several studies [18], which claimed that downstaging the disease permitted more one-stage operations (resection and primary anastomosis). However, one-stage sigmoid resection is possible as well without preoperative drainage in many cases [19]. This, then, poses the question of whether the temporizing effect of PAD is not simply delaying definitive treatment. Care must be taken with regard to the follow-up of the patient: The disease recurred after the initial resolution in 37 of 43 patients with nonresected abscesses [20]. In a series reported by Freeny and coworkers [21], one of the last surgical “bastions” and the least suitable indication for PAD—the infected pancreatic necrosis—was postulated to be conquered. Notably, Freeny et al. were not talking about pancreatic abscesses, which are somewhat mature infected areas of necrosis and are easily drained by PAD [22]; they were reporting on infected nonperfused tissue material. The authors attributed their relative success to a vigorous interventional approach with multiple and repeated catheter placements, multiple CT scans, and manual irrigation through a large-bore catheter, what they

call a “percutaneous necrosectomy.” Facing 146 catheter changes and a drainage duration of 25 to 152 days with multiple appointments at the radiology department by these 34 patients, the question may be justified whether these multiple interventions are really preferable to one visit to the operating room [23]. The development of CT scanning and interventional radiology over the past decades has been a major contribution to the treatment of abscesses in various locations. It is not just that surgical intervention can be guided by the exact topographic localization of the problem and by visualizing sites not accessible to the operating surgeon (i.e., intraparenchymal abscesses), percutaneous abscess drainage may also offer the definitive treatment. There has been a clear development from the early indications—single safely accessible collection—to more complex problems up to the principal statement that almost any abscess, regardless of its origin and localization, can be drained with few limitations. Surgeons have to confess that their more skeptical view of the indication must be revised, as even “forbidden” indications, such as hepatic hydatid cysts, have been treated successfully and safely by PAD first [24]. However, one must be cautious not to become too enamored with this clean “minimally invasive”

372

procedure (as surgeons claim radiologists do). Great temptation arises with any fluid collection that can be reached; but looking more closely at the need for drainage or for a “definitive” solution, the literature differs considerably, as was demonstrated above. So I am not misunderstood: The introduction of PAD represents clear progress in abscess treatment. However, an awareness is advised that many indications are based on little evidence and that there are essentially no prospective randomized trials comparing PAD to open surgery or further treatment modalities, even regarding such simple endpoints as survival, hospital stay, or definitive cure. This does not take into account the more complex issues such as quality of life. If we accept “obvious” indications, we also should disregard PAD in the opposite case, as is shown in Figure 1 in this Invited Commentary. The CT scan of a patient with perforated diverticular disease revealed a gas-containing infection that had already spread from the retroperitoneum to the thigh. It could not be expected in this septic patient that introduction of a small catheter into the huge ill-defined infected area would improve his condition (the patient underwent emergency surgery some 12 hours later). In general, surgeons may need a little more patience and radiologists a little more self-reflection with their beloved interventional “baby” to form a strong and effective team for the benefit of their common patient, and so far we are completely in line with vanSonnenberg’s closing remarks. However, the time for feasibility is over and the time for creating a higher level of evidence for the use of PAD has arrived, including more complex endpoint evaluation.

References 1. Dahnert, W., Gunther, R.W., Borner, N., Braun, B., Gamstatter, G., Rothmund, M.: Percutaneous drainage of abdominal abscesses. II. Value in comparison with septic surgery. Chirurg 56:584, 1985 2. Halasz, N.H., vanSonnenberg, E.: Drainage of intraabdominal abscesses: tactics and choices. Am. J. Surg. 146:112, 1983 3. VanSonnenberg, E., Wing, V.W., Casola, G., Coons, H.G., Nakamoto, S.K., Mueller, P.R., Ferrucci, J.T., Halasz, N.A., Simeone, J.F.: Temporizing effect of percutaneous drainage of complicated abscesses in critically ill patients. A.J.R. 142:821, 1984 4. Deveney, C.W., Lurie, K., Deveney, K.E.: Improved treatment of intra-abdominal abscess: a result of improved localization, drainage, and patient care, not technique. Arch. Surg. 123:1126, 1988 5. Levison, M.A., Zeigler, D.: Correlation of APACHE II score, drainage technique and outcome in postoperative intra-abdominal abscess. Surg. Gynecol. Obstet. 172:89, 1991

World J. Surg. Vol. 25, No. 3, March 2001

6. Levison, M.A.: Percutaneous versus open operative drainage of intraabdominal abscesses. Infect. Dis. Clin. North Am. 6:525, 1992 7. Malangoni, M.A., Shumate, C.R., Thomas, H.A., Richardson, J.D.: Factors influencing the treatment of intra-abdominal abscesses. Am. J. Surg. 159:167, 1990 8. Brolin, R.E., Flancbaum, L., Ercoli, F.R., Milgrim, L.M., Bocage, J.P., Blum, A., Needell, G.S., Nosher, J.L.: Limitations of percutaneous catheter drainage of abdominal absceses. Surg. Gynecol. Obstet. 173:203, 1991 9. Bernini, A., Spencer, M.P., Wong, W.D., Rothenberger, D.A., Madoff, R.D.: Computed tomography-guided percutaneous abscess drainage in intestinal disease. Dis. Colon Rectum 40:1009, 1997 10. Johnson, W.C., Gerzof, S.G., Robbins, A.H., Nabseth, D.C.: Treatment of abdominal abscesses: comparative evaluation of operative drainage versus percutaneous catheter drainage guided by computed tomography or ultrasound. Ann. Surg. 194:510, 1981 11. Olak, J., Christou, N.V., Stein, L.A., Casola, G., Meakins, J.L.: Operative vs percutaneous drainage of intra-abdominal abscesses: comparison of morbidity and mortality. Arch. Surg. 121:141, 1986 12. Hemming, A., Davis, N.L., Robins, R.E.: Surgical versus percutaneous drainage of intraabdominal abscesses. Am. J. Surg. 161:593, 1991 13. Styrud, J., Eriksson, S., Granstrom, L.: Treatment of perforated appendicitis: an analysis of 362 patients treated during 8 years. Dig. Surg. 15:683, 1998 14. Bagi, P., Dueholm, S., Karstrup, S.: Percutaneous drainage of appendiceal abscess: an alternative to conventional treatment. Dis. Colon Rectum 30:532, 1987 15. Jamieson, D.H., Chait, P.G., Filler, R.: Interventional drainage of appendiceal abscesses in children. A.J.R. 169:1619, 1997 16. Warner, B.W., Kulick, R.M., Stoops, M.M., Mehta, S., Stephan, M., Kotagal, U.R.: An evidenced-based clinical pathway for acute appendicitis decreases hospital duration and cost. J. Pediatr. Surg. 33:1371, 1998 17. Brengmann, M.L., Otchy, D.P.: Timing of computed tomography in acute diverticulitis. Dis. Colon Rectum 41:1023, 1998 18. Mueller, P.R., Saini, S., Wittenburg, J., Simeone, J., Hahn, P.F., Steiner, E., Dawson, S.L., Butch, R.J., Stark, D.D., Ottinger, L.W.: Sigmoid diverticular abscesses: percutaneous drainage as an adjunct to surgical resection in 24 cases. Radiology 164:321, 1987 19. Meyer, C., Rohr, S., Iderne, A., Tiberio, G., Bourtoul, C.: The value of peroperative colonic lavage in urgent colonic surgery: apropos of 54 patients. J. Chir. (Paris) 134:271, 1997 20. Farmakis, N., Tudor, R.G., Keighley, M.R.B.: The 5-year natural history of complicated diverticular disease. Br. J. Surg. 81:733, 1994 21. Freeny, P.C., Hauptmann, E., Althaus, S.J., Traverso, L.W., Sinanan, M.: Percutaneous CT-guided catheter drainage of infected acute necrotizing pancreatitis: techniques and results. A.J.R. 170:969, 1998 22. VanSonnenberg, E., Wittich, G.R., Chon, K.S., D’Agostino, H.B., Casola, G., Easter, D., Morgan, R.G., Walser, E.M., Nealon, W.H., Goodacre, B., Stabile, B.E.: Percutaneous radiologic drainage of pancreatic abscesses. A.J.R. 168:979, 1997 23. Mueller, P.R.: Percutaneous drainage of pancreatic necrosis: is it ecstasy or agony? A.J.R. 170:976, 1998 24. Khuroo, M.S., Wani, N.A., Javid, G., Khan, B.A., Yattoo, G.N., Shah, A.H., Jeelani, S.G.: Percutaneous drainage compared with surgery for hepatic hydatid cysts. N. Engl. J. Med. 337:881, 1997