Percutaneous biopsy in evaluation of lung nodules.

Management of an indeterminate pulmonary nodule is a diagnostic challenge that commonly confronts primary care physicians and specialists. Patients often present with this radiographic finding in the course of an unrelated medical evaluation. We examined our institution's experience with percutaneous biopsy of lung nodules to determine the impact of this procedure on overall patient care. Although significant complications are uncommon, the expedience of percutaneous lung biopsy often supplants a surgical opinion prior to initiation of therapy without added diagnostic benefit or cost-savings. Hence, we caution practitioners to use this technique as an adjunct to diagnosis and not a substitute for multidisciplinary care.

INTRODUCTION

Discovery of an indeterminate pulmonary nodule mandates prompt, expedient evaluation by the primary care physician or specialist. Unless the nodule is known to be a long-standing, stable finding from previous radiographic studies, accurate tissue diagnosis is needed to guide the patient's further therapy. Diagnosis of a benign lesion, such as a pulmonary hamartoma or focal infarct, might save the patient from a formal surgical procedure. Similarly, histological diagnosis of a lung malignancy in light of extrathoracic disease would preclude lung resection.

Computed tomographic-guided (CT-guided) percutaneous needle biopsy is a common diagnostic procedure in evaluation of pulmonary nodules. The reported safety and sensitivity (>90%) of the procedure has made it a well-established tool in the armamentarium of physicians.[1, 2] Limitations of the technique include inaccessible or small lesions,[3] pneumothorax, inadequate tissue sample, and the low but finite risk of tumor implantation in the chest wall.[4] Patients having a particularly suspicious-appearing lesion often require surgical resection if a biopsy result fails to show an expected malignancy.

To better understand the impact of CT-guided biopsy at our institution, we conducted a retrospective review of patients undergoing percutaneous needle biopsy to evaluate its role in the diagnosis and management of indeterminate lung nodules.

METHODS

We retrospectively examined 48 consecutive CT-guided percutaneous lung biopsies at Columbus Hospital, Chicago, Illinois, between January 1997 through March 1999. A total of 27 women and 21 men comprised the study group. The mean patient age was 64.5 years (range, 45 to 98). Tobacco use (current or prior) was noted in 35 of 44 patients.

A history of occupational exposure, presenting symptoms, radiographic findings (chest radiograph and CT scan), flexible bronchoscopy brushing/biopsy results, percutaneous biopsy results, and surgical pathology findings (if available) were reviewed. Complications of percutaneous biopsy were graded as none, minor (small pneumothorax–no intervention required), intermediate (pneumothorax requiring thoracic vent drainage), and major (hemopneumothorax requiring tube thoracostomy).

RESULTS

Forty-eight needle biopsies yielded 38 malignancies, 5 specific benign diagnoses, 4 nonspecific diagnoses, and 1 result that could not be determined from a review of the records. Nonspecific diagnoses resulted from hypocellular samples, lack of specific benign or malignant cells, specimens indistinguishable from normal tissue, or inability to arrive at a definitive diagnosis. Specific benign diagnoses included peripheral nerve sheath tumor, caseating granuloma, benign mesothelioma, sarcoidosis, and lipoma.

Of the 38 patients who had a malignant diagnosis by CT-guided biopsy, 10 of these patients underwent surgical resection/biopsy. Lung tissue was obtained by lobectomy in 7 patients, video-assisted thoracoscopic biopsy in 2 patients, and mediastinoscopy in 1 patient. Pathologic review of these specimens showed that all were positive for lung cancer. Of the 4 nonspecific diagnoses provided by CT-guided biopsy, one of these patients underwent surgical resection. This specimen was negative for malignancy.

Eighteen of 48 patients underwent bronchoscopy. Seventeen of these patients had nonmalignant results (94%). All of these 17 patients went on to CT-guided biopsy. Thirteen patients in this group of 17 had a positive result for malignancy, while 4 patients' specimens were returned as negative for malignancy. In this subgroup of patients who had a negative bronchoscopy (n=17), only 1 patient who had a negative CT-guided biopsy went on to surgery to confirm this negative diagnosis. The 1 patient (1/18) who had a bronchoscopy that was positive for malignancy had this finding corroborated by CT-guided biopsy. This sole patient did not however undergo surgical resection.

Determination of the accuracy of CT-guided biopsy was achieved by comparing these results with the tissue diagnosis obtained at surgery. All malignant lung cancers were considered positive, with benign and nonspecific results being considered negative. Using this system, CT-guided biopsy produced 9 true positives and 1 true negative. No false negative and no false positive results were obtained.

Using results obtained by surgical resection, bronchoscopy yielded 1 true negative, 4 false negatives, and 1 true positive. When comparing bronchoscopy with CT-guided biopsy, bronchoscopy yielded 13 false negatives, 4 true negatives, and 1 true positive.

The principle complication of the CT-guided biopsy technique was pneumothorax. Data were obtained in 44 patients who underwent a postprocedure chest radiograph. Twenty patients had no complications, and 24 had a pneumothorax (55%). Of these 24, 7 had large pneumothoraces (16%) defined as requiring corrective intervention. Five of these patients underwent thoracostomy tube placement, and 2 required a thoracic vent.

DISCUSSION

Percutaneous lung biopsy has become firmly ensconced in the evaluation of lung nodules,[5,6] although debate is ongoing about the utility of fine needle aspirate versus use of a cutting needle to obtain a tissue sample.[7, 8] Nonspecific diagnoses were obtained in 8.3% of our study patients, and definitive negative benign diagnoses were obtained in 10.4% of patients. It is interesting to note that at our institution percutaneous biopsy was in all cases performed prior to surgical consultation.

CT-guided biopsy has several disadvantages. Although this technique is a sensitive method for diagnosing malignant intrathoracic tumors, it offers no data for tumor staging. Determination of surgical operability or entry to a regimen of neo-adjuvant chemotherapy/radiation mandates accurate histological diagnosis and staging. The reported rate of pneumothorax in CT-guided biopsy is 12% to 62%.[1-3,9-11] Forty-five percent of our patients had pneumothoraces. Within this group, 16% of these patients required additional treatment with either a chest tube or thoracic vent.

The rate of obtaining a specific benign diagnosis via percutaneous biopsy varies from 12% to 68% and can be as low as 2%.[10] In our study, the rate was 12.5%. Although a specific benign result may avert surgery, 2 out of 5 patients in our group (1 with benign mesothelioma and the other with a lipoma) did require surgical resection. Larscheid et al[10] showed that 100% of their “nonspecific” diagnoses were actually malignant. Charig et al,[12] although demonstrating the safety of outpatient percutaneous lung biopsy in their study, reported a negative predictive value as low as 65% for nonspecific results. Our study yielded a negative predictive value of 100% for CT-guided biopsy, although three fourths of our patients were never afforded a definitive surgical diagnosis. We readily acknowledge that our small study group and lack of uniform medical evaluation for each patient might result in a misleadingly high accuracy rate.

Odell and Reid[11] in their series of 113 patients undergoing 117 percutaneous lung biopsies found that 26.5% of their biopsy results were undiagnostic or negative for malignancy. In this subgroup, almost one half of these patients at surgical resection were found to have a malignancy. Almost 30% of their patients had a pneumothorax that required postprocedural hospitalization. They concluded that transthoracic needle biopsy in patients who are surgical candidates adds little to the course of their overall management. Wallace et al[9] stratified patients undergoing percutaneous fine needle aspiration of lung nodules into those with lesions of 0.5 cm to 0.7 cm versus those with lesions of 0.8 cm to 1.0 cm. They found a higher diagnostic sensitivity and accuracy in patients with larger lesions that were subpleural. Sixty-two percent of their patients had pneumothoraces after the procedure with one-third requiring tube thoracostomy.

We also evaluated bronchoscopy at our institution as a diagnostic tool for evaluating suspicious lung masses. This technique was associated with a low negative predictive value (20%) in patients who underwent surgery to confirm the tissue diagnosis. It must again be stressed that this number might seem inordinately high in that only 4 of the 17 patients underwent surgical resection. In the 17 patients who had a negative bronchoscopy and who also underwent CT-guided biopsy, 13 patients had a CT-guided biopsy result that was positive for malignancy, yielding a negative predictive value of bronchoscopy against CT-guided biopsy of only 24%.

Taken in the context of overall patient management, surgical biopsy is often the most cost-effective approach in terms of accuracy and staging in determining the malignant or benign nature of a lung lesion in patients who are surgical candidates. Goldberg-Kahn and colleagues[13] created a hypothetical situation to compare costs of various techniques used to diagnose a lung malignancy. Using the model of a patient with a 2.8-cm lung lesion who was a surgical candidate, they set parameters including a prevalence of malignancy = 0.67, and diagnostic sensitivities for the following modalities: fine needle aspirate (0.95), bronchoscopic biopsy (0.80), and sputum analysis (0.51). Projected costs per correct diagnoses were open biopsy, $12,888; sputum examination, $63,424; fine needle aspirate, $21,543; and bronchoscopy, $16,615. They concluded that open biopsy of a lung nodule in a “surgically fit” patient is the optimal approach to diagnosis and management.

CONCLUSION

CT-guided biopsy, although a valuable tool in the diagnostic armamentarium of the physician, is not without complications and is often an intermediate additional study. We believe a multidisciplinary evaluation is the most appropriate course of action in patients found to have new lung nodules. The goal of treatment in good performance status patients with bronchogenic carcinoma is accurate staging and curative resection. A nonspecific diagnosis mandates repeat CT-guided biopsy, VATS, or open thoracotomy for definitive diagnosis. It is uncommon that diagnosis of a specific benign disease process on percutaneous biopsy of a suspicious pulmonary nodule will abrogate the need for surgical resection.