INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Both primary and metastatic lung tumors can present with cavitation. The cavitation arises from tumor necrosis in the majority of cases (1), but can also occur as a result of bacterial colonization or growth in the necrotic center of tumors (2).

Fever is common in patients with primary lung cancer, with a reported incidence of 21.1% (3), and is often caused by pulmonary infections such as obstructive pneumonitis (4) or abscess formation. Because obstructive pneumonitis is known to be a radiographic opacification resulting from complete or partial obstruction of a bronchus by a pulmonary neoplasm, it can be clearly recognized on clinical grounds. However, currently available imaging tools are usually unreliable in differentiating in situ infection from simple necrosis in cavitating lung tumors. Lung abscesses usually require prolonged antimicrobial treatment, and failure to recognize and treat lung abscesses is associated with poor clinical outcome. Thus, defining the bacterial growth in a cavitating lung tumor and applying the most effective treatment are clinically important.

Ultrasound (US)-guided transthoracic aspiration is a safe and useful method for collecting specimens for accurate bacteriologic diagnosis of lung abscess (5) and obstructive pneumonitis (4). However, there have been no reports on the use of US-guided needle aspiration for bacteriologic diagnosis in patients with cavitating lung tumors. In this study, we examined the safety and effectiveness of US-guided transthoracic aspiration for investigating the bacteriology of infected cavitating lung tumors, and evaluated its impact on modification of antimicrobial regimens.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

For this study, we prospectively enrolled all patients with a diagnosis of cavitating lung tumor treated at National Taiwan University Hospital from January 1996 through October 1998. The diagnosis of cavitating lung tumor required chest radiographic and computed tomographic (CT) evidence of cavitating lung tumor, with or without air- fluid level; cavitating lung tumors were defined as those in which the thickest part of the cavitary wall was greater than 5 mm thick, with an irregular contour of the inner surface (6). In addition, the cavitating lung tumors had to be proven to be either primary or metastatic carcinoma, by cytologic or histologic examination of biopsy specimens taken from the cavitating lung tumor; specimens were collected via US- or CT-guided transthoracic aspiration, or bronchoscopic biopsy. Patients were excluded if (1) the platelet count was less than 100,000/µl or there was prolonged prothrombin time (more than 3 s prolonged as compared with control); (2) there was evidence of systemic or local infection, such as obstructive pneumonia or intra-abdominal or urinary tract infection; or (3) there was no "ultrasound window" (5) for the cavitating lung tumor (7).

All patients provided informed written consent after receiving a detailed explanation of the biopsy procedure and the possible complications. They were examined with real-time, linear-array, convex sector US units equipped with 3.75- and 5.0-MHz transducers (Aloka SSD 630 and Toshiba 100A; Tokyo, Japan). US-guided transthoracic aspiration was performed with a special puncture probe (Aloka UST-507BP). We used a 22-gauge needle with an outer sheath and an inner stylet for aspiration of the cavitating lung tumor. A 20-ml syringe was adapted for use as an aspirator. The needle was inserted through a guiding channel and was advanced to the necrotic cavity with real-time US guidance (4). The transthoracic aspirates, collected in airtight, oxygen-free syringes, were immediately sent to the laboratory for aerobic and anaerobic microbial cultures.

The demographic characteristics (age and sex) and clinical manifestations (fever, leukocytosis, purulent sputum) of patients, cell types of malignancies, tumor location, diagnostic procedures for underlying malignancies, cancer stage, and the antibiotic treatment and management of the underlying malignancies were all recorded.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

During the study period, 26 patients were treated for cavitating lung tumors. Of these, four were excluded from the analysis because of clinical and imaging evidence of obstructive pneumonitis. Thus, 22 patients were included in the study; there were 18 men and four women, with a mean age of 64.9 (range, 54 to 80) yr.

The clinical data of the 22 patients with cavitating lung tumors are summarized in Table 1. In all cases, the US-guided transthoracic needle aspiration samples for microbiological study were taken from areas that were previously proven to be malignant. The cavitating lung tumors were proved to be primary or metastatic carcinomas by cytologic examination of US-guided or CT-guided transthoracic aspiration samples in 15 of the 22 patients; the remaining seven cavitating lung tumors were proved to be malignant by histologic examination of bronchoscopic biopsy samples. All 22 cavitating lung tumors had thick walls and irregular contours of the inner surface, and the imaging features and clinical courses were consistent with a diagnosis of tumor with central necrosis and cavity formation. Figure 1 demonstrates the imaging studies of Case No. 2 with a cavitary lung tumor. The bacterial culture of needle aspirate from necrotic cavity grew Haemophilus influenzae.

View larger version (131K): [in this window] [in a new window]   View larger version (112K): [in this window] [in a new window]   View larger version (140K): [in this window] [in a new window]   Figure 1.   (A) Chest radiograph of a 76-yr-old man with squamous cell   carcinoma of the lung shows a tumor in the left lower lobe (arrowhead ). (B) Chest computed tomography scan reveals a tumor mass with central necrosis and cavitation. (C ) Chest ultrasound shows a cavitating mass at the subpleural area. Arrowheads indicate the thick wall of the cavitating mass. The linear hyperechoic densities indicated by arrows represent air within the cavitating mass with abscess formation. Cultures of samples of the necrotic center obtained with US-guided transthoracic needle aspiration yielded Haemophilus influenzae.

We divided the patients into febrile and nonfebrile groups on the basis of body temperature at the time of US-guided transthoracic aspiration biopsy. The febrile group comprised seven patients with a body temperature greater than 37.5° C at presentation; the nonfebrile group comprised 15 patients with a body temperature less than 37.5° C. Six of the seven patients in the febrile group had microorganisms isolated from the lung aspirates. Empiric antibiotic therapy had been administered for 1 to 14 d (mean, 5 d) before needle aspiration in four of these six patients. One patient in the nonfebrile group had a microorganism isolated from the lung aspirate. Those in the nonfebrile group had not received antimicrobial therapy before lung aspiration.

Bacteriology of Lung Aspirates

The clinical data and bacteriology of the lung aspirates of the seven patients with positive culture results are summarized in Table 2. Nine pathogens were isolated from the seven patients; two patients had two pathogens isolated from the lung aspirates, whereas 1 pathogen was isolated from the aspirate in each of the other five patients. Six of the pathogens were aerobic gram negative bacilli, and the others were an anaerobic gram positive bacillus (Bifidobacterium), an aerobic gram-positive coccus (Enterococcus faecium), and a Mycobacterium tuberculosis.

Clinical Courses

The seven patients in the febrile group were treated with empiric antibiotics, because of clinical suspicion of infection, until the culture results were available. The one patient in the febrile group who had negative results on the lung aspirate cultures had persistent fever despite a course of empiric antimicrobial treatment; he became afebrile after 2 wk of irradiation for the lung carcinoma. The nonfebrile patient with a positive lung aspirate culture received a course of prophylactic antimicrobial therapy to prevent infection after irradiation for lung carcinoma.

The antimicrobial regimen was adjusted according to the culture results in five of the six febrile patients with positive culture results. The patient whose lung aspirate culture yielded Mycobacterium tuberculosis was treated with empirical antibiotics and antipyretic agents initially and became afebrile without antipyretic agents after 2 wk of antituberculous treatment. She received 6 mo of antituberculous therapy and was lost to follow up thereafter. The cavitating lung tumor of this patient was increased in size over the period of antituberculous therapy. The other four patients had persistent fever for a period ranging from 3 to 21 d (mean, 12.5 d) before the adjustment of antibiotics. They became afebrile 3, 4, 5, and 37 d, respectively, after modification of the antimicrobial regimens.

Complications and Outcomes

Only one patient developed pneumothorax after needle aspiration, and his condition improved after a chest tube was inserted. This was the only aspiration-related complication. None of the patients developed hemoptysis.

One patient in the nonfebrile group, whose lung aspirate cultures were negative for bacterial growth, developed obstructive pneumonitis during a course of radiotherapy. The other nonfebrile patients had uncomplicated courses after subsequent management, including resection (one patient), chemotherapy (six patients), and irradiation (four patients). Three patients received conservative treatment only. In addition to parenteral antimicrobial treatment in the febrile group, we performed resection in one patient, chemotherapy in two patients, irradiation in three, and conservative treatment in one. One patient in the febrile group, who had positive lung aspirate culture results, developed life-threatening upper gastrointestinal tract bleeding requiring surgical intervention during hospitalization; he also had prolonged fever and hospitalization, and received only conservative treatment for the underlying malignancy. Two additional patients in the febrile group had prolonged hospitalization owing to postoperative complication and progressive underlying malignancy, respectively. None of the patients died during hospitalization, and there was no morbidity related to cavitating lung tumor infections. The mean duration of hospitalization was 51.4 d for the febrile group and 24.9 d for the nonfebrile group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our study has documented that cavitating lung tumors are commonly associated with in situ infection (6 of 22, 27%), especially in patients who are febrile at presentation (6 of 7, 86%). US-guided transthoracic needle aspiration of cavitating lung tumors with suspected infection is useful for bacteriologic diagnosis and infection control, and may be helpful for determining whether the fever is due to tumor necrosis or infection.

Fever is frequently seen as a consequence of tumor necrosis in patients with pulmonary carcinoma, but is often a result of lung infection (3). Distinguishing simple tumor necrosis from an infected cavity on the basis of clinical and imaging findings in febrile patients can be difficult. Thus, bacteriologic diagnosis of cavitating lung carcinoma is important for investigating the possibility of infection in the cavity and optimizing the antimicrobial regimen. In our study, bacteria were isolated from the lung aspirates of seven of the 22 patients, six of whom were febrile at presentation. In five of the six febrile patients who had positive aspirate culture results, the initial empiric antimicrobial regimens were changed in accordance with the aspirate culture results and the spectrum of antimicrobial susceptibility; all five improved after adjustment of the antimicrobial regimen. On the other hand, one febrile patient who had negative culture results had no improvement after antimicrobial treatment. This patient subsequently became afebrile after the antimicrobial treatment was stopped and irradiation was administered for the lung carcinoma. The cause of fever in this case could be attributed to tumor fever rather than pulmonary infection.

US-guided transthoracic aspiration is a very efficient tool for etiology diagnosis of peripheral pulmonary lesions, lung abscesses, and obstructive pneumonitis (4, 5, 8). In patients with lung abscesses (5), transthoracic needle aspiration is the most direct way to obtain reliable, uncontaminated material for bacterial culture. Although fluoroscopic and CT guidance can also achieve this purpose, lateral fluoroscopy does not adequately demonstrate extremely peripheral and very small lesions, and CT-guided aspiration lacks real-time monitoring of the needle position and carries a relatively high risk of large pneumothorax (9). US-guided transthoracic aspiration, on the other hand, can be performed at the bedside, allows real-time monitoring, costs less than the other two methods (5, 8), and minimizes the risk of hemoptysis and hemothorax. In this study, only one of the 22 patients developed an aspiration-related complication (pneumothorax).

The bacteriologic results in this study differed markedly from those of previous investigations, which reported a predominance of anaerobic pathogens in lung abscesses (5, 10, 11). In an earlier report (5), 41 of 65 isolates (63%) from lung abscesses were anaerobes, and only 12 of the 65 (19%) isolates were gram-negative bacilli. Grinan and coworkers (10) and Bartlett and coworkers (11) had similar results. In the present study, six isolates were aerobic gram negative rods, one was an aerobic gram-positive coccus, and one was Mycobacterium; there was only one anaerobe isolated.

Differences in patient selection and the pathogenesis of abscess formation between our study and previous reports may account for the apparent discrepancies in bacteriologic results. First, the majority of patients in the previous reports had primary lung abscesses, while only a small proportion had underlying lung carcinoma. Second, the infections were within the necrotic centers of lung tumors in the present study, but were in necrotic lung parenchyma in primary lung abscesses in the other reports. Furthermore, four of our six febrile patients who had positive results on lung aspirate cultures had received antimicrobial treatment for suspected lung abscesses, with coverage of anaerobic pathogens, before transthoracic aspiration; this may have led to the low yield of anaerobes from the aspirate cultures. Finally, faulty culture technique, in terms of collection, transportation, and handling of specimens, may also have contributed to the low yield of anaerobes. However, faulty anaerobic culture technique may not be the sole factor responsible for the low yield of anaerobes. In our previous studies (4, 5) by US-guided transthoracic aspiration, we have found that 41 of 65 isolates (63%) from lung abscesses, and eight of 16 isolates (50%) from obstructive pneumonitis were anaerobes. Different diseases with different pathophysiologies may account for the difference in anaerobic isolation rate.

The mean hospitalization length is long in our patients with infected cavitating lung tumor. However, the long stays were due to the complicated clinical courses rather than infection. Two patients who were hospitalized for more than 60 d had life-threatening upper gastrointestinal tract bleeding and postoperative complications, respectively. One patient was hospitalized for 91 d because of multiple metastases of breast cancer that involved several major organs; she underwent several courses of systemic chemotherapy in the hospital.

Our study had several limitations. First, our case number was small, so our results are suggestive rather than conclusive. Second, lung tumors that do not attach to the chest wall are invisible on ultrasonographic examinations, so centrally located cavitating lung tumors were not included in our study. Third, the prior use of antibiotics may have influenced the yield of microorganisms from the infected cavitating lung tumors. Fourth, because our study was not a controlled study, we could not prove that modification of antibiotics on the basis of transthoracic needle aspirate culture results is associated with a good outcome.

In conclusion, in situ cavitary infection is common among febrile patients with cavitating lung tumors. US-guided transthoracic aspiration is helpful for differentiating infection from simple necrosis in cavitating lung tumors in febrile patients. Lastly, the results of lung aspirate cultures are useful for optimizing antimicrobial management. Further controlled trials comparing empiric antibiotic therapy versus empiric antibiotic therapy followed by adjustment according to the aspirate culture results may clarify whether routine bacteriology investigation is needed in this patient group.