Eric vanSonnenberg, Gerhard R. Wittich,

Lung

MD MD2

Horacio Robert

#{149} #{149}

Abscess:

B. D’Agostino, R. Varney, MD

MD

CT-guided

Lung abscesses were drained by means of catheters guided by computed tomography (CT) in 19 patients who still had sepsis despite standard medical therapy; all patients had received antibiotics for at least 5 days, and 11 of the 19 patients had undergone bronchoscopy. The abscess was cured (by clinical and radiographic criteria) in all 19 patients (100%), and surgery was avoided in 16 of the 19 patients (84%). Three patients underwent surgery for removal of organized tissue or decortication after the lung abscess was evacuated. Complications included a hemothorax that required a chest tube in one patient and three minor complications (a clogged catheter in two patients and transient elevation of intracerebral pressure in one patient). The hemothorax occurred in one of two patients in whom the catheter traversed normal lung. The percutaneous drainage catheters traversed juxtaposed abnormal pleura on route to the abscess in 17 of the patients. CT-guided drainage of lung abscess is an effective method to treat lung abscesses that are refractory to conventional therapy; the procedure should obviate major operation in most patients. A catheter route through abscess-pleural syndesis is preferable, and CT is useful for planning this route.

S

Giovanna Harker,

Casola, MD

#{149}

#{149} Colleen

MD

#{149}

Drainage’

treatment

TANDARD

of lung

ab-

tients

had

idiopathic

purpura and splenic sequestration. The white blood cell count in 1 1 patients ranged from 11,000 to 20,000/mm3 (11-20 109/L). Six patients had a white blood cell count greater than 20,000/mm3 (20 X 109/L). Eleven of 19 patients underwent bronchoscopy before catheter insertion; no unexpected tumors or foreign bodies were discovered. The causes of the abscesses were pneumonia (n 9) (Fig 1), trauma (n = 4) (Fig 2), tumor (n 2), postoperative esophagectomy (n 1), pulmonary infarct (n 1), infected sequestration (n = 1), and Streptococcal fasciitis (n =

experience

present

with

drainage

computed tomography tients who had a lung fined as a pulmonary mal collection of pus, this report.

PATIENTS The women, was scess

patients aged

despite age of

AND 8-74

14 men

years.

Each

(up

antibiotics to 22 days

(n

of the abscesses 7), right middle

=

right lower lobe (n = 2), and left lower

scesses

were

from

3 to

cm). in

five

patient absepsis,

for an averprior to

drainage). blood cell 32,800/mm3

At the time of drainage, white counts varied from 1,300 to (1.3-32.8 X 109/L). Two pa-

tients had 10,000/mm3

a white blood (10 X 109/L);

cell count below one of those pa-

Adjacent

I

From

the

G.C.,

Departments

cine (E.V.). University Medical Center, 225

terms:

Computed

Abscess,

percutaneous

tomographic

(CT)

ventional

procedures

Lung,

Lung,

60.1211

CT,

guidance abscess,

drainage #{149} Inter-

60.216

1991;

C.H.)

of California, Dickinson St.

the 1989 RSNA June 18 1990;

quested

revision

July

accepted

20;

October

San San

scientific revision

received

4. Address

(E.V.,

and

17 patients,

tubes

had

All

2;

Current Center, RSNA,

address: Stanford, 1991

Stanford Calif.

University

Medi-

(average,

lung

pleural

(n

6.6

was

fluid

seen

was

thoracostomy

inserted

into

the

pleura

prior to percutaneous abscesses.

patients

2060

underwent

obtained GE 9800

and

(GE

in

drain-

chest

radio-

at 1-cm intervals (GE Medical Systems,

Technicare

Medical

2020

Systems)

CT

and

scanners

were used. The catheter was inserted by means of trocar technique in 14 patients (Fig 3), Seldinger technique in three patients, and one-step drainage 12-F chest drain catheter ton Scientific, Watertown, in

15 patients

(Fig

was

tient.

Routes

in

catheter

one

the catheter

in two patients immediately

since adjacent

Malecot

A used cath-

patient,

was

of drainage abutted

two patients. (Medi-tech/BosMass) was

Spencer, md) was a 9-F Sacks catheter

abnormal

abscess

tients;

used

in

4), a 20-F

eter (Cook Urological, used in two patients,

the

requests

1), lobe

4). All absize ranged

Abscess

and

upper

(n

graphic and CT examinations; all catheters were inserted under CT guidance.

Diego, Diego,

October

reprint

right upper

in diameter

been

10 patients age of the

contiguously

asre-

5), left lobe (n

in 14. Surgical

Medi-

to E.V. 2

178:347-351

R.R.V.,

CA 92103. From sembly. Received

cal

Radiology

of Radiology

G.R.W.,

were lobe

consolidated

10-F pigtail H.B.D.,

=

solitary.

15 cm

Milwaukee)

and

of a lung fever and

1). Sites

lobe

Scans were routinely.

METHODS

for drainage of ongoing

receiving 12 days

by

(CT) in 19 paabscess, deintrapanenchyis described in

included

referred because

guided

x

(Elecath)

Index

thrombocytopenic

scess assumes patency of the draining bronchi, along with an intact cough mechanism, and is aided by postural drainage and antibiotics. Lung abscesses that are refractory to this treatment may be improved by bronchoscopic clearing of mucus or removal of an obstructing foreign body. Surgical resection of the affected lobe traditionally has provided definitive treatment should these regimens fail (1-9). However, mortality rates from lung abscess-even with surgery-continue to be substantial, ranging from 9.3% to 28% (14,7,10-12). Prior reports described the feasibility of radiographically or fluonoscopically guided drainage of lung abscess due to pneumonia (13-22). Our

the

and

used

in one

were

through

a

pa-

pleura

or where

pleura

in

traversed

17 pa-

normal

the pleura was to the abscess.

lung not The

patients or oblique

were scanned positions

in supine, prone, (six, nine, and four

patients, Once

respectively). the catheter

was

abscess

was

evacuated,

in place gentle

and

the

irrigation

347

a.

b.

c.

Figure 1. CT-guided drainage of postpneumococcal pneumonia lung abscess. Patient still had sepsis despite receiving antibiotics for 2’/2 weeks. (a) Chest radiograph demonstrates large left-lower-lobe lung abscess. A surgical chest tube is in the pleura. (b) CT scan shows prone posterior percutaneous catheter drainage through contiguously abnormal pleura on route to the abscess. (c) Radiograph obtained 1 week after drainage. The catheter has been removed. The abscess had resolved almost completely at radiography, and the patient recovered uneventfully.

was

performed

tnieved

fluid

with

saline

was clear.

until

Irrigation

the

re-

was

continued when the patient was on the ward (5-15 mL every 8 hours) with recovery of the irrigant and pus. The catheter was connected to pleurevac suction (Dernatel, Fall River, Mass). All catheter connections were taped to avoid inadvertent

loosening

and pneumothorax.

Chest

ra-

diographs were obtained until the abscess cleared in all patients. All patients received antibiotics for at least the duration of catheterization. Contrast medium was injected through the catheter in five patients (propyliodone [Dionosil, Picker Health Care Products, Euclid, Ohio]) in three patients; nonionic agents in two patients) to determine bronchial communications; communication was observed in three patients. All patients had follow-up chart review for at least a 6-month interval after drainage.

the

percutaneous catheters (Fig 5). One patient developed a hemothorax after drainage of the abscess; a chest tube was inserted to drain the blood (Fig 6). In that patient, the abscess drainage route was transparenchymal, through normal lung. This patient had mild dyscrasia (prothnombin time 4 seconds over control) that was not completely connected with fresh frozen plasma. Other complications included clogging of the catheter that required catheter exchange (to 20 F) in two patients, and one episode of transient elevation of intracerebral pressure in a patient with a prior head injury (1 hour after tube placement). Pneumothorax on bronchopleural fistula did not occur in any patient as a result of percutaneous drainage.

RESULTS In all 19 patients (100%), the abscess was evacuated and sepsis was relieved. Eighteen patients showed clinical improvement (ie, fever abatement, restoration of normal vital signs, and reduction in leukocytosis) within 48 hours of drainage. Cathetens were removed when drainage ceased and when follow-up radiographs demonstnated resolution of the abscess cavity; pleural fluid and adjacent consolidation typically cleared in concert with the abscess. The duration of drainage averaged 9.8 days and ranged from 4 to 38 days. Sixteen of the patients (84%) were spared an operation; surgery was performed in three patients because of adjacent organized pleural tissue that could not be drained via

348

Radiology

#{149}

catheter

well-formed

drainage

can

be used effectively to treat most lung abscesses that are resistant to the standard therapy of antibiotics and postural drainage. A good response to percutaneous drainage-that is, abatement of fever, normalization of white blood cell count, and clearing of the abnormality seen on radiographs-can be dramatic. Catheter drainage should be completely cunative, and, in most cases, surgery can be avoided. Our study comprises only those patients who did not respond to conventional medical treatment after an average of 12 days. These patients had persistent sepsis and toxicity at the time of percutaneous drainage. The effectiveness of

abscess

cavity

are

un-

likely to be cured by means of the percutaneous approach. Failure of percutaneous drainage is reported to occur with multiloculation, poor definition of a cavity, or a thick wall that does not collapse (13). Thus, for the appropriately selected patient with a lung abscess, the data suggest CT-guided drainage should

sidered

DISCUSSION CT-guided

percutaneous drainage was evident; 18 of the 19 patients experienced abatement of fever and were relieved of sepsis within 48 hours. The most common previous indication for surgery with lung abscesspersistent sepsis and toxicity from an abscess over 4 cm in diameter (4,16)-may now be treated effectively by means of catheter drainage. Only those patients with nonfluid thick, organized tissue on without a

as a frontline

method

that con-

be

of then-

apy. Not all patients with a lung abscess require percutaneous catheter drainage. A patient with an intact cough mechanism and patent bronchi should be cured by means of bnonchial toilet and antibiotics. Reports of large surgical series describe the need for surgery in 1 1%-21% of patients with lung abscess (3,4,7). Abscess size is considered an indication for surgery or percutaneous tube drainage; surgical drainage of abscesses larger than 4-8 cm has been advocated by numerous authors (20,23,24). Bronchial obstruction usually is the cause of abscess enlargement. Our data suggest that even abscesses as large as 15 cm can be cured by means of percutaneous drainage.

February

1991

b.

c.

4 Figure

2. CT-guided catheter drainage of posttraumatic lung abscess. Posteroanterior (a) and lateral (b) chest radiographs demonstrate large right-lower-lobe lung abscess with surrounding consolidation. A large he-

mothorax

had been

evacuated

previously

thoracostomy tube. (c) CTthrough abnormal pleura yielded bloody pus. A small amount of normal lung may have been traversed; however, the patient suffered no adverse effects. (d) Follow-up chest radiograph obtained 3 days after drainage with percutaneous catheter shows improvement. (e) Radiograph obtained 8 days after drainage. The catheter has been removed, and the patient was well. with a surgical guided drainage

abscess

e. i4 Figure

3. CT-guided percutaneous drainage of traumatic pneumatocele. Patient had multiple

from

thoracic

a motor

3 weeks tion and

earlier. a right

the left lateral his

and

vehicle

abdominal

accident

There is bilateral pneumothorax.

decubitus

position

injuries

that

occurred

consolidaPatient is in

because

of

injuries.

has

pleura

Another neous

indication approach

patients who mechanism; a developing

Volume

178

for the is drainage

percutaof pus

in

lack an adequate cough radiographic evidence of contralateral pneumo#{149} Number

2

reported

as 11%-

on route

to the

abscess

when-

ever possible. Normal lung is avoided to prevent development of an infected bronchopleural fistula or pyopneumothorax and to avoid

bleeding. nia in a patient with a lung abscess is suggestive of this problem. Surgical treatment of lung abscess may result in bronchopleunal fistula, empyema, bleeding, on spillage of pus into the tracheobronchial tree (3,4,9,13,15,20,25,26). Postoperative empyema has been reported to complicate resection of lung abscess in 10%-29% of patients (3,5,6). Several surgical studies noted that complications were greater after surgery than after tube drainage, even though patients in the latter group typically were more ill than those undergoing surgery (3,5,6,20,24,26,27). Postoperative mortality after surgery for lung

been

16% (1,3). A recent study of 184 patients with lung abscess showed an increasing, rather than decreasing, mortality rate from the 1960s to the 1980s (22% to 28%) (4). A guideline that we used for drainage was that the catheter should traverse contiguous abnormal lung and

One

of two

patients

whom normal lung was traversed by the catheter hemothonax (the patient

in

unavoidably developed also had

a a

mild clotting disorder). Although avoiding normal pulmonary panenchyma appears conservative and prudent, it is controversial. Not all investigators have insisted on this restriction for drainage; the precise abscess-pleura relationship will not always be recognized by using only

fluoroscopic

guidance.

displays the help provide

relevant accurate

CT accurately anatomy catheter

ment. Although lung abscesses drained while the patient

may is on

Radiology

to placebe the #{149} 349

a.

b.

Figure 5. Lung abscess in a 66-year-old man with postoperative pneumonia who developed a right-lower-lobe lung abscess. Large-bore catheters were used to treat the abscess. (a) CT scan of patient in prone position. Drainage with 12-F catheter evacuated 60 mL of pus. (b) Radiograph obtained with use of contrast material. In an effort to better remove thick tenacious material and tissue (note contrast agent injection), two 20-F Malecot catheters were inserted into the abscess. The larger catheters were marginally beneficial; the patient eventually underwent surgical debridement and decortication.

Figure

4. Standard chest catheter (8-10 F) used for percutaneous drainage of lung abscess. This catheter is 12 F, has three wide side holes, and is made of Percuflex (Meditech/Boston Scientific).

ward (20,24,27) roscopic guidance

offers

distinct

or by means (13,17,20-22),

advantages.

of fluoCT

CT dem-

onstrates both the abscess and the surrounding structures. The relationship of the abscess to the adjacent lung and pleura helps determine the optimal catheter pathway (28). The presence and quantity of fluid in the abscess (29), other contents of the abscess (such as a fungus ball) (30), and wall thickness are best displayed by means of CT (15,24). CT helps ascertam whether a collection is an abscess or an empyema (31), although either should be amenable to percutaneous catheter drainage by means of comparable percutaneous management (32). CT guidance permits direct trocar catheter insertion after fine-needle localization. Repeat CT examination after evacuation of pus

yields

immediate

information

on the

adequacy of drainage; this dictates whether more than one catheter will be necessary. Patient positioning is extremely important for successful and safe drainage. Percutaneous drainage may be performed with the patient in the supine, prone, decubitus, or oblique

position. 350

Whatever Radiology

the

position,

the

a. Figure

b. 6.

CT

scans

show

catheter

route

through

normal

lung

parenchyma

for

drainage

of

lung abscess. A hemothorax resulted. (a) Anterior transparenchymal catheter pathway into deep lung abscess. Pus was extracted, and the patient’s fever abated. (b) Thirty-six hours after drainage, the patient’s hematocrit had dropped 10%. CT scan revealed a new pleural collection. Two liters of blood were drained from the pleura.

contralatenal lung must not be dependent. This avoids aspiration of pus into the normal lung, which can cause bilateral disease and overwhelming sepsis. In one series, 22% of 33 deaths from lung abscess were due to aspiration and spread of pus from the lung abscess itself to uninvolved areas of the ipsilateral lung or to the other lung (10). The differentiation of lung abscess from related disorders may be difficult or impossible in certain cases. An infected traumatic or postinfectious

pneumatocele, bronchopleural fistula, bulla, sequestration, empyema, or necrotic malignancy may simulate an abscess juxtaposed to the pleura (18, 21,29,33). In neonates, cystic malformation of the lung may mimic a lung abscess (9). The key therapeutic point with all these disorders is that any patient who still has sepsis despite administration of antibiotics and who has an infected thoracic collection in the periphery adjacent to the pleura should undergo drainage expeditiously (14,15,28).

February

1991

It is well established catheters can routinely ly drain pus. Needle

that 10-14-F and effectiveor sheath aspira-

tion alone has been used previously to drain lung abscess (14) and was effective in two of our patients. In one prior report, catheter exchange from 7 to 10 F was necessary in two pa-

tients

to augment

drainage

and

This study establishes that CTguided percutaneous drainage of lung abscesses is safe and effective for most patients. The caveat of Estrena et al (3) that “external drainage is limited and probably should be avoided” may not be applicable to

current niques.

CT and Further

interventional techstudies are needed to

evaluate whether traversing normal lung by a catheter is deleterious; the hemothorax that developed in one patient with a transparenchymal catheter supports the contention that catheters should avoid normal panenchyma whenever possible. U

drain as our preferred catheters provided age of pus. Catheters

Acknowledgments: Friedman, MD, assistance.

These drainF may

be necessary in patients who have extremely tenacious, viscous material (18,20). Catheters were connected to pleunevac for suction and were inngated with 5-15 mL of saline peniodically. Gentle irrigation may facilitate and expedite percutaneous drainage (20,31). Postdrainage contrast smography theoretically offers information on cavity closure (16); however, we did not use it routinely. Plain chest nadiognaphs normally suffice for follow-up. Most patients were ambulatory during, at least, the latter portion of the drainage procedure. How long percutaneous drainage

should

be delayed

while

antibiotics

dency

on mechanical

One advantage drainage is that ic improvement

Delarue per JD.

2.

Bernhard

tions.

thus the that can tive and decreased.

potential for complications occur with more conservaprolonged treatment may

Volume

178

Estrera

AS,

5.

Platt

1980;

Hagan

JA,

MR,

Wylie

Mills

U,

22.

JD.

of

Lung

184

Ann

DL.

LH.

Mayer

T, Matlak

L.

ME,

ings

of neonatal

Child

1982;

136:39-41.

Siegel

HD,

McCracken

J

Am

Fox trid

V. Shasha

abscess.

GH Dis

J

Am

Jr.

Child

PB.

JR. Hughes lung abscess.

Fatal

lung

experience.

FA,

Sutliff

WD.

J Thorac

Surg

Gittens ment

berc 13.

Kosloske

SA,

Mihaly

of acute

1954;

lung

JP.

Medical

abscess.

Am

Ball

WS

Jr.

Butler

1985;

TB.

Bacterial

infections

of the

lung. In: Shields TW, ed. General thoracic surgery. 2nd ed. Philadelphia: Lea & Febiger, 1983; 575-590. Rice TW, Ginsberg RJ, Todd TRJ. Tube drainage of lung abscesses. Ann Thorac Surg 1987; 44:356-359. T, Vane

lung

DW,

in

disease.

Vainrub

B, Musher

Krishna

G, et al.

the

treatment

Chest

1985;

DM,

Utility

of

87:731-734.

Guinn

GA,

EJ, Septimus EJ, Travis LT. Percudrainage of lung abscess. Am Rev 1978;

117:153-160.

J, vanSonnenberg

Stavas GR.

E, Casola

Percutaneous

G, Wit-

drainage

of infect-

noninfected

J Thorac

MH,

Albelda

SM,

Mayock

McLean GK. Severe hemoptysis ed with pulmonary aspergilloma: neous intracavitary treatment.

RL,

associatpercutaChest 1988;

94:1225-1231.

abscesses:

Med

Non-pu-

1954;

32.

manageTu-

C, Muse-

meche CA. Drainage of pediatric lung abscess by cough, catheter, or complete resection. J Pediatr Surg 1986; 21:596-600. Lorenzo RL, Bradford BF, Black J,Smith JD. Lung abscesses in children: diagnostic and therapeutic needle aspiration. Radiology

pressure

Shapiro

Stark DD, Federle MP, Goodman PC, Podrasky AB, Webb WR. Differentiating lung abscess and empyema: radiology and computed tomography. AJR 1983; 141:163-167. vanSonnenberg E, Nakamoto 5K, Mueller PR, et al. CT- and ultrasound-guided catheter drainage of empyemas after chest-tube

69:673-680. AM,

positive

30.

Dis

133:

Rev

receiving

MS. Perculung bulla

29.

I,

South

MG, Stulbarg of an infected

thoracic fluid collecImaging 1987; 2:80-87. Rice TW, Ginsberg RJ, Todd TRJ. Tube drainage of lung abscesses. Ann Thorac Surg 1987; 44:356-359.

27:255-260.

12.

Ferguson

tions.

31. P. Terry

by

90:186-194.

RespirDis

Neonatal 1979;

treated

Cardiovasc

of cavi-

28.

find-

abscess

ventilation. Chest 1987; 91:928-930. Parker LA, Melton JW, Delany DJ, Yankaskas BC. Percutaneous small bore catheter drainage in the management of lung abscesses. Chest 1987; 92:213-218.

Young taneous

tomographic

lung

abscess.

947-949. Harber

Condon

Computed

JI.

pneumonotomy

68:168-172.

review of 11 years’ J 1981; 74:281-283. 1 1.

1985;

ed and

lung 10.

Surg

lung

J Thorac

Dean NC, Stein taneous drainage

Chidi CC, Mendelsohn HJ. Lung abscess: a study of the result of treatment based on 90 consecutive cases. J Thorac Cardiovasc 1974;

Lee

26.

27.

21:370-

RG,

using Surg

treated

1951;

Giant

thoracostomy.

tich

Glasgow

9.

Surg

Jost

Neonatal lung abscess: resection one-lung anesthesia. Ann Thorac 1983; 36:464-466. Snow N, Lucas A, Horrigan TP.

lung.

of surgically

J Thorac

L.

tube

376.

Surg 8.

A review

resec-

55,

Weber

1983;

Pulmonary

for chronic abscess of the Thorac Surg 1948; 17:514-522. abscess.

revis-

Surg

Sagel

25. RR.

abscess

cases.

Mengoli

tary

197:755-762. Shaw RR, Paulson

lung

to

79:275-282.

a survey

Bosher

due

Cardiovasc

tion 6.

24.

RH.

Shaw

J Thorac

abscess.

JL, Hardy

ited:

15.

be

Malcolm

lung

Surg

14.

nadiolograpid;

WF,

DJ,

of

Percutaneous drainage of lung abscess. AJR 1979; 132:282-283. Yellin A, Yellin EO, Lieberman Y. Percutaneous tube drainage: the treatment of choice for refractory lung abscess. Ann Thorac Surg 1985; 39:265-270. Keller FS, Rosch J, Barker AF, Dotter CT. Percutaneous interventional catheter therapy for lesions of the chest and lungs. Chest 1982; 81:407-412.

in a patient

NC, Pearson FG, Nelems JM, CooLung abscess: surgical implicaCan J Surg 1980; 23:297-302.

Primary

4.

19.

21.

to Paul for their

Lung abscess: a study of 148 cases aspiration. Chest 1963; 43:620-630. 3.

Aronberg

23.

1.

ventilation.

of percutaneous clinical and usually are

appreciation Chambers

17.

20.

References

7.

are administered was not determined from our data. Healing of a lung abscess with conservative medical treatment can be protracted; one study showed that 13% of cavities resolved in 2 weeks, 44% in 4 weeks, 59% in 6 weeks, and 70% by 3 months after medical therapy (34). A recent article (29) suggested that the following cnitenia should lead to aggressive interventional therapy in a patient with a lung abscess: (a) 5-7 days of unremitting sepsis while receiving antibiotics, (b) abscess larger than 4 cm and causing mediastinal shift, (c) enlarging total abscess size or fluid portion of the abscess, and (d) patient depen-

Our Peggy

and

Lawrence GH, Rubin SL. Management giant lung abscess. Am J Surg 1978; 136:134-139.

18.

achieve cure (22). Tubes as large as 28 F have been used (20); the use of this tube size appears unnecessary for most cases of lung abscess. In addition, although the use of large-bone catheters in the pleura is safe, it may cause undesirable trauma to the lung. We used a 12-F single-lumen chest catheter. complete of 20-30

16.

33.

cavity ment 1977;

34.

failure.

151:349-353. MacArthur

Weiss mary, Respir

suction

AM,

Radiology

Fountain and

W. Cavity nonspecific Dis 1973;

SW.

drainage

of emphysematous 32:668-672.

1984;

in

Intrathe

bullae.

treatThorax

behavior in acute, lung abscess. Am 108:1273-1275.

priRev

157:79-80.

Cuestas RA, Kienzle GD, Armstrong JD. Percutaneous drainage of lung abscesses in infants. Pediatr Infect Dis J 1989; 8:390392.

Number

#{149}

2

Radiology

#{149} 351