Quantifying the number of lymph nodes for examination in breast cancer

Pathologic node-negative lung cancer: Adequacy of lymph node yield and a tool to assess the risk of occult nodal disease

INTRODUCTION

Accurate lymph node (LN) staging is crucial for prognostication in NSCLC. Diagnosis of pN0 disease is based on the absence of positive LNs, irrespective of the number of LNs excised, and is thus susceptible to sampling error. Tumors that are assumed to be pN0 may in fact be understaged. We developed a tool to quantify the risk of occult nodal disease (OND) among patients with pN0 NSCLC in terms of the number of LNs examined.

METHODS

Patients treated surgically for stage I-III primary NSCLC between 2004 and 2014 (n = 49,356) were extracted from the Surveillance, Epidemiology, and End Results database. The probability of missing a positive node in terms of the number of LNs examined was modeled using a beta-binomial model. A mathematical tool was then used to calculate the negative predictive value (NPV) corresponding to the number of LNs examined. Ranging from 0 to 100%, higher NPV reflects greater confidence in the pN0 diagnosis and a lower probability of OND.

RESULTS

The median number of LNs examined was 7 for N0, 10 for N1/N2, and 8 for N3 disease. The probability of missing a positive node decreased as LNs examined increased. Additionally, higher T stage required more LNs to confirm an N0 diagnosis. After accounting for false-negative diagnoses, the prevalence of node-positive disease was readjusted from 16% to 22% among patients with T1 disease. According to our tool, with 10 LNs examined, the NPV was 85% (15% probability of OND) for a patient with T3 disease, compared with 95% (5% probability of OND) for a patient with T1 disease.

CONCLUSIONS

Accurate pN0 diagnosis depends on the number of LNs examined. The proposed tool offers the ability to quantify, in a patient-specific manner, the confidence in a diagnosis of node-negative disease on the basis of the number of LNs examined.