Lymph nodal staging for non-small-cell lung cancer

Abstract

We read with interest Matsuguma et al.’s recent paper in the journal [1] on the optimum method of lymph node (LN) staging for non-small-cell lung cancer (NSCLC). We congratulate the authors on their elegant analysis of the data to conclude that the LN ratio had the best discriminatory power, followed by the LN number, both of which were superior to the current nodal staging system. Several other authors [2–5] have come up with similar results and have espoused the use of either the LN ratio or positive LN number for the pathological staging of NSCLC. In an ideal world, with uniform standards of surgery, pathology grossing and reporting, there is no doubt that these methods would be superior to the conventional lymph nodal staging system. We, however, believe that there are several points to be considered before recommending that these be used to modify the existing tumour node metastases (TNM) staging system. First, it is often difficult to get an accurate count of the number of LNs harvested in the mediastinum. Frequently, this depends not only on the expertise of the operating surgeon, but also the diligence of the pathologist grossing the specimen. Moreover, anthracotic LNs in the mediastinum often fragment, thereby giving an erroneous count. Secondly, the ratio of metastatic LNs would be widely variant between surgeons, thoracic units and institutions depending on their surgical philosophy, expertise and again, the pathologist. Surgeons performing mediastinal LN sampling would have very different LN ratios from those who perform a radical systematic mediastinal LN dissection. Paradoxical results could appear due to an inordinately high pick-up of normal LNs, reducing the LN ratio. Thirdly, preoperative LN staging, already inaccurate, would become even more inaccurate as it would be virtually impossible to get a metastatic LN ratio or even an LN number based on radiological investigations or even a mediastinoscopy. Finally, these methods would increase the complexity of staging, making them less likely to be complied with by routine pathology reporting. While we agree wholeheartedly with the authors that the LN ratio or number would be superior to the existing TNM staging system, we would be reluctant to advocate the adoption of these methods in the TNM staging system.