Lung Cancer Staging: Mediastinoscopy vs Endosonography

Andrew F. Shorr, MD, MPH


December 16, 2010

This feature requires the newest version of Flash. You can download it here.

This is Andy Shorr from the Washington Hospital Center with a Pulmonary and Critical Care literature update. I'd like to talk to you about an article by Annema and colleagues from the Netherlands, which was recently published in JAMA.[1] These investigators focused on patients with suspected mediastinal involvement from lung cancer, and examined the question of how best to approach the issue of staging.

With the advent of PET scanning in the last 5 years, we've faced a number of issues in deciding how to evaluate patients who are thought to have lymph node involvement and who may have N2 or N3 disease, which essentially makes them no longer resectable under our current approach to lung cancer. One parallel advancement in our approach to lung cancer staging has been the development of endobronchial ultrasound (EBUS), along with esophageal approaches to staging the mediastinal nodal involvement.

A number of studies, thus far, have demonstrated that you can access these nodes by both EBUS and esophageal ultrasound, and they've also shown that by using these approaches, you can decrease the number of surgical procedures that might be needed. In this study, Annema and colleagues took patients (n = 241) and in a novel fashion, randomly assigned them to an initially noninvasive approach using EBUS and looking at the esophagus vs going right to mediastinoscopy, which would have been our historical control.

The patients assigned to the less invasive approach first had follow-up mediastinoscopy, if it was necessary, and all patients who went to thoracotomy had lymph node dissection. The investigators specifically focused on the sensitivity and specificity of the procedures, looking at N2 and N3 nodes. Because this study was randomized, it allows us to compare the alternate strategies. With a large observational study, we can look at what happens to patients and then calculate the sensitivity and specificity, but we also want to know how many thoracotomies might be prevented by using one approach vs another, but without a randomized controlled trial approach, we can only tangentially answer that question.

The randomized controlled trial, which was the strategy used in this study, also allows us to look at complication rates in a more systematic fashion, because the issue isn't only "what do you get?", it's also "what price do you pay to get it?"

Overall, the 2 strategies (the initially noninvasive/less invasive approach followed by mediastinoscopy if necessary, if all of lymph node sampling was negative vs going right to mediastinoscopy) had very similar sensitivities, in the 80%-85% range. If you combine the noninvasive approach followed by the surgical approach, and still no lymph nodes were found, but there was a concern about it, the sensitivity went up to about 94%, which is pretty good.

The patients who went right to mediastinoscopy were able to avoid some thoracotomies, but they were also able to avoid a number of thoracotomies that became unnecessary because the patient had been staged by the noninvasive approach. In the end, complication rates between the 2 strategies, in terms of doing the procedures (airway and esophagus first, followed by surgery if needed, staging the lymph nodes, followed by thoracotomy, if still indicated vs going right to mediastinoscopy and then thoracotomy) were similar.

This study was really unique because of its randomized approach to this question and it provides good confirmatory evidence that these noninvasive approaches are really ready for primetime, as long as they are performed by operators who know how to use them, and as long as we understand that even if the findings are negative, and we still suspect lymph node involvement in such a way that would preclude a curative lung cancer surgery, we still need to definitively answer that question.

This doesn't obviate the need for every mediastinoscopy, but hopefully obviates the need for a number of mediastinoscopies, and as pulmonologists, we really need to become comfortable with EBUS technology and these other approaches so we can better stage the lymph nodes and better help our patients and improve our patients' chances of avoiding a trip to the OR that they may not need.

Again, this was recently in JAMA, and it deals with what is one of the most common problems we face as pulmonologists -- staging the mediastinum, and I urge you to look at this study carefully. This is Andy Shorr from the Washington Hospital Center.