Percutaneous Abscess and Fluid Drainage

Patient Preparation

Any correctable abnormalities such as coagulopathies and fluid electrolyte imbalances should be corrected before abscess drainage. The patient should receive prophylactic intravenous antibiotics as determined by blood culture results. If blood cultures are negative, then an appropriate broad-spectrum antibiotic regimen such as gentamicin, ampicillin, and metronidazole (or other appropriate broad-spectrum coverage recommended by local infectious disease personnel) should be used.

Detection and Localization

Without doubt, computed tomography (CT) is the most appropriate modality for the detection and localization of intraabdominal fluid collections. Sonography may be helpful in detecting upper abdominal collections such as subdiaphragmatic collections, paracolic collections, or collections in solid viscera such as the liver and spleen. However, ultrasound suffers from its inability to penetrate gaseous interfaces. This is a particular problem in patients with intraabdominal abscesses, in that many have an associated ileus, which is particularly problematic in the postoperative patient. CT is therefore the preferred imaging modality for the identification of intraabdominal abscesses. Furthermore, because all adjacent organs can be visualized, an appropriate access route can be planned.

One of the disadvantages of CT is that loculation may be difficult to visualize, because often the septa are of the same density as the adjacent fluid and cannot be distinguished. Septation and loculation are much more easily identified by sonography. It is important also for patients to have appropriate bowel opacification with Gastrografin, when possible, because unopacified bowel may be difficult to differentiate from an abdominal abscess. Additionally, appropriate bowel opacification is necessary for planning the access route to ensure that small or large bowel is not traversed with a catheter when draining the abscess.

Neither sonography nor CT can determine whether a collection is infected or not (unless air is present). Gram stain and culture must be obtained to make this determination ( Box 18-1 ).

For pleural space collections, plain films and sonography are often sufficient to demonstrate the entire fluid collection. With multiloculated empyemas, mediastinal abscesses, and lung abscesses, CT is necessary for full delineation of the abscess cavity.

Technique

Catheter Types

The various catheters available for drainage include sump designs and nonsump designs. Sump catheters have double lumens and are particularly suited for intraabdominal abscesses. The outer lumen in the sump catheter is designed to prevent side holes from becoming blocked when the catheter is adjacent to the wall of an abscess cavity. Twelve- to 14-French sump catheters are suitable for most intraabdominal abscesses (Boston Scientific, Natick, Mass.) ( Fig. 18-1 ). However, sump catheters are not really necessary. Most pigtail catheters with reasonably large side holes work well in conjunction with adequate flushing. Larger (16- to 28-French) catheters are required in specific circumstances such as for pancreatic abscesses, hematomas, or when the abscess cavity contents are extremely viscous (see Fig. 18-1 ). Nonsump catheters are used in the chest. Generally these catheters have large side holes to permit appropriate drainage. Catheters inserted in the chest also tend to be larger (16- to 30-French) because kinking occurs commonly with smaller catheters because of respiratory excursion, which compresses the catheter against adjacent ribs. There is a vogue to place smaller pigtail catheters in the chest, which the author does occasionally for simple fluid collections, but for empyemas, the author prefers to place larger catheters.

Locking pigtail catheters (8- to 10-French) are used in specific circumstances such as when draining lymphoceles and seromas, or when draining deep pelvic abscesses transrectally or transvaginally. It is important to use locking catheters when using the transvaginal or transrectal route because any abdominal straining may dislodge a nonlocking catheter.

Image Guidance

The decision whether to drain an abscess under ultrasound or CT guidance is based largely on the location of the abscess, the size of the abscess, and operator preference. Most pleural fluid collections or empyemas can be drained under ultrasound guidance, as can hepatic abscesses, subphrenic abscesses, paracolic abscesses, and some of the larger, more central intraabdominal collections. However, from a practical point of view many abdominal abscesses are detected by CT scanning and therefore it is often easier to drain the abscess under CT guidance at the time of diagnosis. In addition, some abscesses absolutely require CT guidance, such as retroperitoneal and iliopsoas abscesses, deeply located abscesses, small abscesses, or abscesses that are not visible by ultrasound.

Diagnostic Fluid Aspiration

Diagnostic fluid aspiration is often requested to determine whether a fluid collection detected by either CT or sonography is infected or uninfected. It is important to plan the access route carefully so that bowel is not transgressed en route to the collection. This is to ensure that a potentially sterile collection is not contaminated by a diagnostic aspiration. Generally a 20-gauge needle is used for diagnostic aspiration. This can be performed under ultrasound or CT guidance, provided a safe access route is visible. Two to 3 mL of fluid are aspirated and specimens sent for Gram stain and culture. If fluid cannot be obtained with a 20-gauge needle, an 18-gauge needle is placed in tandem to the 20-gauge needle into the fluid collection. Failure to aspirate fluid through this 18-guage needle usually means that the cavity contents are very viscous. Fluid can usually be aspirated in small amounts if rapid to-and-fro motions with the 18-gauge needle are performed. Alternatively, 1-2 mL of sterile saline can be injected into the cavity and reaspirated for the purpose of Gram stain and culture.

If the sample obtained is pus, a catheter should be placed straight away. If the specimen obtained is not pus and it is unclear whether it is infected or not, either wait for the result of the Gram stain or place a catheter. Some interventional radiologists prefer to wait for the result of the Gram stain. It is the practice in the author’s unit to place a drain in the vast majority of abdominal collections, particularly if the patient is sick and has a high temperature. One can then await the result of the Gram stain and culture. If these are negative, the catheter can be removed after 48 hours ( Box 18-2 ).

It is important for the interventional radiologist to be able to interpret Gram stain results because the result may directly affect decision making. A Gram stain that has abundant bacteria and white cells indicates an abscess. A stain that yields bacteria without white cells may be consistent with colonic contents. The CT scan should be reviewed to confirm that the suspected abnormality does represent an abscess and not unopacified colon and that the aspiration needle did not traverse the colon. Alternatively, bacteria without white cells may mean that the patient is immunocompromised and cannot mount a leukocyte response. It is not uncommon, with the modern use of antibiotics, that a Gram stain may show white cells without bacteria, indicating a so-called sterile abscess. These collections should, however, be drained ( Box 18-3 ).

Drainage Procedure

It is now routine to perform the drainage procedure under either ultrasound or CT guidance at the initial time of localization of the intraabdominal fluid collection. An appropriate access route is chosen that allows a clear route to the collection without passing through adjacent structures.

There are two basic methods of draining an abscess or fluid collection: the Seldinger technique and the trocar technique.

In the Seldinger technique, an 18-gauge long-dwell sheath is placed in the cavity and a 0.038-inch guidewire is coiled within the cavity. Alternatively, a one-stick system using a 22-gauge needle and 0.018-inch guidewire can be used (Neff set, Cook, Bloomington, Ind.). The track is dilated with fascial dilators to two French sizes larger than the catheter to be placed. The catheter is then inserted over a stiff guidewire into the collection. It is important to coil the catheter within the collection so that all of the side holes are within the collection. Initially, when using the Seldinger technique the needle, long-dwell sheath, and wire were placed into the abscess cavity under CT or ultrasound guidance and then the patient was moved to fluoroscopy to complete the procedure. When experience is gained, it is possible to perform the entire procedure under CT or ultrasound guidance. However, the Seldinger technique can be a relatively blind procedure without using fluoroscopy, and for this reason the author prefers to use the trocar technique where possible.

The trocar technique ( Fig. 18-2 ) consists of placing a reference needle into the abscess cavity. A catheter with a sharp stylet is inserted alongside the localizing needle into the collection in a single stab. It is important to leave the reference needle in situ, because the catheter can be directed along the exact trajectory of the localizing needle. Adequate dissection of the skin and subcutaneous tissues with a standard surgical forceps is necessary for this procedure. A “give” is usually felt when the cavity is entered. Once the catheter is felt to be in place, the central stylet is removed and the catheter aspirated to confirm that the catheter is in the cavity. Once pus or fluid is aspirated, the catheter can be coiled in the cavity by disengaging and withdrawing the trocar and pushing the catheter forward ( Fig. 18-3 ).

When the catheter is secure within the cavity, the cavity contents are completely aspirated. This is best performed using a closed system with a three-way stopcock and drainage bag. In this way the cavity contents can be completely aspirated and drained into the drainage bag. When the cavity is completely aspirated, the cavity is irrigated with sterile saline until the aspirate returns clear to ensure that most of the debris and viscous contents, if present, are drained.

There are many methods for securing the catheter to the skin, ranging from simply suturing the catheter to the skin to using commercially available catheter fixation devices. The author’s unit uses the “drain-fix” device (Unomedical, Birkerod, Denmark) ( Fig. 18-4 ). A piece of tape is placed round the catheter and the tape is then sutured to the ostomy disk. This system works quite well for catheter fixation; the ostomy disk usually protects the surrounding skin if there is any pericatheter leakage. If ostomy disks are not available, the tape placed around the catheter can be sutured directly to the patient’s skin or a fixation device can be used.

It is imperative to repeat imaging after evacuation of the cavity contents to ensure that the cavity is completely evacuated and that there is no loculation. If there is an undrained area, placement of a second or more catheters to completely drain the abscess cavity is required because the patient will not defervesce if pus is left behind ( Box 18-4 ).

Aftercare

It is vitally important that the interventional radiologist be actively involved in patient management when a drainage catheter is placed. It is not acceptable to place a catheter and abdicate on the clinical responsibility of looking after the catheter and the patient’s abscess. Respect by clinical colleagues is also gained by this approach and increased referrals to the interventional radiology service usually ensues. Interventional radiologists who have inserted the catheter know the abscess type, size, consistency of the fluid content, and loculation. It is mandatory that daily ward rounds be made on each patient with an indwelling catheter. During these ward rounds, the skin site, catheter and connections, the amount of drainage, clinical well-being, changes in white cell count, and fever are assessed. With daily ward rounds and careful observation, the interventional radiologist can decide whether follow-up imaging or intervention is required, and when the catheter should be removed.

Virtually all catheters are left to gravity drainage on the ward and fluid output is recorded. It is important that the catheter be irrigated three to four times daily with 10-mL aliquots of sterile saline to prevent clogging. This is usually performed by the nursing staff, but if there is any question of catheter patency, the catheter should be irrigated by the interventional radiologist on ward rounds to ensure patency.

The endpoint of catheter drainage is dependent on a number of factors. Primarily these are clinical factors such as clinical well-being, defervescence, reduction in white cell count, and decreased catheter drainage to less than 10-15 mL/day. It is not necessary to perform follow-up imaging on simple collections, particularly when the patient is recovering. Imaging endpoints include disappearance of the collection on repeat imaging and/or a reduction in size of the cavity on a contrast abscessogram. In general, abscessograms are rarely performed unless the possibility of a fistulous communication exists. Resumption of appetite is another good clinical criterion for successful drainage. When some or all of these criteria are met, the catheter is withdrawn. Usually for simple collections that drain quickly and successfully, the catheter can be simply removed. For more complicated abscesses or those that take longer to resolve, the catheter is best withdrawn over a number of days, as with surgical drains, which is preferable to needing to redrain the abscess ( Box 18-5 ).

Specific Abscess Drainages

Enteric Abscesses

Abscesses complicating appendicitis, diverticulitis, or Crohn disease are referred to as enteric abscesses . A complicating abscess in these conditions makes immediate surgery extremely difficult and may make multistage surgery, with its associated cost and discomfort, a reality for these patients. In these circumstances, percutaneous drainage, in combination with appropriate antibiotic therapy, can be used to effectively drain the abscess and resolve sepsis. Elective one-stage surgery can then be performed at an appropriate interval after resolution of sepsis.

Drainage of diverticular abscesses can often avoid two- or three-stage surgery and convert the surgical procedure to an elective one-stage operation. The three-stage operation was in use before the general availability of antibiotics. The three-stage operation consisted of initial surgical abscess drainage and colostomy, a resection of the diseased colon and reanastomosis, and lastly a revision of the colostomy. In general, surgeons resect the diseased segment and any small associated abscess (less than 5 cm in diameter) and do a primary anastomosis. PAD is used for draining the larger abscesses to permit elective one-stage surgery ( Fig. 18-5 ). Success rates between 80% and 90% have been quoted for PAD of diverticular abscesses, permitting single-stage surgery.

Periappendiceal abscesses result from a walled-off appendiceal perforation. Drainage of the periappendiceal abscess and appropriate antibiotic therapy usually permits elective appendicectomy in 4-6 weeks. Indeed, there is some debate in the surgical literature regarding whether interval appendicectomy is necessary. Success rates of 90%-100% have been quoted for PAD in periappendiceal abscesses.

Abscesses complicating Crohn disease occur in approximately 12%-25% of patients at some point in the disease course. Crohn abscesses are difficult to manage and PAD is useful in temporizing patients with enteric communication before definitive surgery ( Fig. 18-6 ). Alternatively, PAD can be curative if there is no enteric communication. Enterocutaneous fistulas resulting from percutaneous drainage in patients with Crohn abscesses have not been reported to date. Success rates for abscess drainage in Crohn disease range from 70% to 90%.

It is important to use CT as both the diagnostic and therapeutic guiding modality in these patients with enteric abscesses. Good bowel opacification is necessary for secure diagnosis and for planning the access route for drainage. These abscesses occur in close proximity to bowel loops, and CT is mandatory to ensure that small or large bowel is not traversed by the catheter during drainage ( Box 18-6 ).

Box 18-6

Enteric Abscess Drainage

• Abscesses associated with Crohn disease, diverticulitis, and appendicitis fall in this group

• Abscess drainage generally allows elective one-stage surgery

• Computed tomography best for abscess localization and access route planning

• Good bowel opacification is mandatory

Abscess-Fistula Complex

Fistulization to collections can occur from various structures including the pancreatic duct, bile duct, urinary system, and bowel. The most common abscess fistula complex is the enteric abscess with fistulous communication to the small or large bowel. Principles of treatment are the same for all abscesses associated with fistulas. Enteric abscesses associated with fistulous communication are discussed here representing the typical abscess-fistula complex.

The index of suspicion for fistulous communication should be high when managing enteric abscesses. Persistent high outputs (>100 mL/day) or an increase in output after 3-4 days of drainage indicates the presence of a probable fistula. This can be confirmed with an abscessogram. Fistulas are designated as high-output when drainage is greater than 200 mL/day. In these high-output abscess-fistula complexes, the communication is usually with small bowel. Management principles include draining the abscess, proximal diversion of bowel contents, and bowel rest. Abscesses associated with fistulas also take longer to heal (often 3-6 weeks for high-output fistulas), which should be communicated to the referring physician and to the patient. Proximal diversion of bowel contents can be achieved by nasogastric suction and by placing a catheter through the fistulous track into the bowel ( Figs. 18-7 and 18-8 ). The catheter placed in the fistulous track is left in situ for approximately 10-14 days to allow a mature fibrous track to form. When catheter output recedes to less than 30-40 mL/day the catheter can be slowly withdrawn. With high-output fistulas it is important to monitor and correct electrolyte and fluid losses from the small bowel to speed fistula healing. Patients with high-output fistulas are usually fed parenterally.

In patients with low-output fistulas from the colon, drainage of the associated abscess with bowel rest is often sufficient for complete healing of the fistula. As might be expected, low-output abscess-fistula complexes usually heal successfully with percutaneous drainage. High-output fistulas do less well. It is useful to clamp the abscess catheter before removing it for 2-3 days in patients with high-output fistulas. If a CT scan after 2-3 days of catheter clamping shows no evidence of recurrence, the catheter can be removed.

Other factors that influence successful drainage and fistula healing include the presence of distal obstruction, the health of the bowel at the fistula site, and the immune status of the patient. In the presence of distal obstruction, fistulas will not heal. Similarly, if the bowel at the fistula site is diseased (e.g., affected by Crohn disease or malignancy), the fistula is unlikely to heal. Additionally, if the patient is immunocompromised, fistula healing will be delayed. Quoted success rates for successful resolution of abscesses associated with fistulas vary from 66% to 82% ( Box 18-7 ).

Box 18-7

Abscess-Fistula Complex

• Diagnosed by catheter outputs < 100 mL/day

• High-output fistula > 200 mL/day

• Managed by abscess drainage, proximal bowel diversion, and bowel rest

• Fistula healing influenced by distal obstruction, integrity of bowel at fistula site, and immune status

Subphrenic Abscess

The vast majority of subphrenic abscesses are postoperative, often resulting from pancreatic, gastric, or biliary surgery. Anatomically, they are located in a difficult position with the pleural attachment often making an extrapleural access route a technical challenge. The pleura is attached at the 12th rib posteriorly, 10th rib laterally, and eighth rib anteriorly. Traditionally, these abscesses were drained using a subpleural or extrapleural approach, which involved angling an 18-gauge sheath needle or 22-gauge single-stick needle up under the rib cage and into the collection under ultrasound guidance and using fluoroscopic guidance to dilate a track over a stiff wire to place the catheter ( Fig. 18-9 ).

It has become apparent that an intercostal approach can be used in selected cases without a major increase in the complication rate, and the author’s unit now uses this route for the vast majority of subphrenic abscesses. It is likely that the two pleural surfaces are firmly adhesed by the time of drainage because of the adjacent abscess, making pneumothorax or empyema unlikely with an intercostal approach. However, it is prudent when draining these abscesses intercostally to go through the lowest intercostal space possible that gives access to the abscess ( Fig. 18-10 ). Quoted success rates for PAD of subphrenic abscesses are between 80% and 90%.

Hepatic Abscess

Pyogenic hepatic abscess has become rare since antibiotic coverage of patients with abdominal sepsis has improved. In earlier days, most hepatic abscesses occurred secondary to bowel infections such as diverticulitis and appendicitis. Now, most hepatic abscesses are secondary to liver or biliary surgery. PAD of hepatic abscess is very successful and should be curative in more than 90% of cases. Many hepatic abscesses at presentation appear loculated with multiple septations; portions may even appear solid on imaging. However, it is worthwhile placing a catheter in all hepatic abscesses, in that almost all such abscesses respond dramatically to PAD ( Fig. 18-11 ). Access can be intercostal or subcostal depending on the location of the abscess. Some interventionalists advocate needle aspiration alone for hepatic abscesses. The author prefers to place catheters for larger abscesses and use needle aspiration for smaller abscesses.

Retroperitoneal Abscess

Retroperitoneal abscesses usually locate in the iliopsoas compartment and can have varied etiologies ranging from acute spinal osteomyelitis to Crohn disease or hematogenous spread. These abscesses require CT guidance for drainage because of their deep location. If the abscess involves the psoas muscle in the abdomen and the iliacus in the pelvis, it is often sufficient to place a catheter in the iliacus muscle because there is extensive communication between the iliacus and psoas muscles.

A catheter is first placed in the iliacus muscle and pus aspirated. If on the postprocedure CT scan the psoas component has also resolved, another catheter may not be necessary. If the psoas component has not fully resolved, another catheter will be necessary ( Fig. 18-12 ). This is best done under fluoroscopic guidance, using the same puncture site as that used for the catheter in the iliacus abscess. The existing catheter is removed over a guidewire and a second guidewire inserted and manipulated up into the psoas muscle. Twelve- to 14-French catheters are then placed over each guidewire, one catheter in the iliacus abscess and the second in the psoas abscess.

Between 80% and 90% success rates have been reported for PAD of iliopsoas abscesses.