Should Endoscopic Ultrasound Be the First Test for Lung Cancer?

TEMPLE, TEXAS—It may be time for a major change in the way lung cancer is diagnosed and staged, the results of a recent prospective study suggest.¹ While it is common practice to perform multiple imaging tests after initial thoracic and upper abdominal CT, the study indicated that endoscopic ultrasound–guided fine-needle aspiration (EUS-FNA) “may be considered the next test to establish the diagnosis and stage of lung cancer” because of its high diagnostic yield and level of accuracy.

Among 113 patients with suspected lung cancer, the use of EUS-FNA as a first test after CT led to a diagnosis in 82% of cases and had an overall diagnostic yield of 70%. This level of accuracy is comparable to that obtained with bronchoscopy, with the additional advantage that thoracic and extrathoracic staging can be performed simultaneously, the authors said.

Furthermore, EUS-FNA was similar in accuracy to PET for finding distant metastases to the celiac nodes, liver, and adrenal glands. However, the ability of EUS-FNA to acquire tissue and accurately stage lung tumors makes it preferable to PET as well, asserted the authors.

They therefore recommended that after initial CT assessment, EUS-FNA be considered the next test to establish the diagnosis and stage of lung cancer, particularly when CT shows mediastinal and distant metastases. Principal author Pankaj Singh, MD, is from the Division of Gastroenterology at the Central Texas Veterans Health Care System in Temple.

To be eligible for the study, the patients had to have a newly detected lung mass suspicious of cancer or a recent tissue diagnosis of non–small cell lung cancer (NSCLC). The exclusion criteria were advanced heart or lung disease that precluded moderate sedation, current chemo­therapy or radiotherapy for lung cancer, and an established tissue diagnosis of small cell lung cancer.

In addition to EUS-FNA, the patients had other tests for diagnosis and staging when indicated. They were typically diagnosed with bronchoscopy and transthoracic needle aspiration and staged with mediastinoscopy/mediastinotomy and PET.

Lung cancer diagnoses and metastases to the mediastinal lymph nodes were established by fine-needle aspiration or through the detection of malignant cells in a lung tissue sample obtained during surgical resection. Distant metastases were defined as the presence of cytologically positive malignant cells or as radiologic disease progression on follow-up imaging studies.

After initial CT, EUS-FNA was the first test performed and produced a diagnosis in 93 (82%) of the patients. In 17 cases (15%), EUS-FNA was used for staging, since a lung cancer diagnosis had already been made.

The sensitivity of EUS-FNA for distant metastases, small cell lung cancer, and mediastinal invasion was 100%, 94%, and 92%, respectively. For enlarged mediastinal lymph nodes, NSCLC, and a normal mediastinum, the sensitivity was 79%, 65%, and 57%, respectively.

“Metastases to the [mediastinal lymph nodes] were established in 49 cases,” the authors related. “EUS, CT scan, and PET scan detected metastases to the mediastinum with accuracies of 93%, 81%, and 83%, respectively.”

Compared to CT, EUS-FNA had a greater diagnostic yield for the detection of celiac lymph node metastases, was more specific for liver metastases, and displayed greater specificity and accuracy for distant metastases. EUS-FNA was more accurate than PET for locating liver metastases and just as accurate for finding celiac node and adrenal gland metastases. However, the authors noted, there was no difference between the three tests regarding sensitivity for distant metastases, and no correlation was found between celiac metastases and adrenal or liver metastases.

In a Cox proportional hazards model, metastases to the celiac axis and liver independently predicted poor survival (hazard ratios, 2.4 and 3.5, respectively). “The median survival time for subjects with and without EUS evidence of metastasis to the [lymph nodes at the celiac axis] was 142.5 days and 330 days,” pointed out the authors. “The difference was statistically significant.”

For patients with NSCLC metastases to the celiac lymph nodes, median survival was five months, versus 10.8 months for those without NSCLC metastases; one-year survival in these two groups was 0% and 42%, respectively. The median survival of patients with metastases to the liver or adrenal gland was 7.8 and 4.4 months, respectively, versus about 11 months for those without liver or adrenal gland metastases.

Among 44 patients with resectable tumors on CT, EUS-FNA changed the management of eight cases (18%) by upgrading the tumor stage. “By using EUS-FNA, thoracotomy was avoided in 13.6% of patients with CT findings of a resectable tumor,” the authors observed. “One patient developed chest pain after EUS-guided FNA of the pleural effusion. There were no other immediate or late complications due to EUS-FNA.”

The high incidence of metastases to the celiac axis (11%) was unexpected and important because of its association with reduced survival, the authors remarked. They suggested that bronchoscopy be reserved for cases of suspected lung cancer in which EUS-FNA is nondiagnostic and for operable cases, since those cases require accurate endobronchial tumor staging. The main limitation of EUS, its lack of ability to reach the pretracheal lymph nodes, remains a concern, they noted.

“In skilled hands, EUS offers a genuine advance in the potential for diagnosis and staging, but perhaps it should be seen as complementary to [endobronchial ultrasound–guided fine-needle aspiration],” suggested Sam M. Janes, MD, and Stephen G. Spiro, MD, in an accompanying editorial.² In a previous study, comparison of the two procedures found similar diagnostic yields, although the latter was better for right-sided nodes, they explained. Thus, the most appropriate test at any given time will depend on the situation, they said.

“There is a steep learning curve to these generally safe techniques and expert training is mandatory, but the time to start is now,” Drs. Janes and Spiro added.

—Timothy Begany

Reference
1. Singh P, Camazine B, Jadhav Y, et al. Endoscopic ultrasound as a first test for diagnosis and staging of lung cancer: a prospective study. Am J Respir Crit Care Med. 2007;175:
345-354.
2. Janes SM, Spiro SG. Esophageal endoscopic ultrasound/endobronchial ultrasound–guided fine needle aspiration: a new dawn for the respiratory physician? Am J Respir Crit Care Med. 2007;175:297-299.

Return to table of contents