Christopher Caruso

MD, MPH

University of Pennsylvania

Pulmonary and Critical Care Fellow

NLST (Learn more about this study)

NLST-1245

May 16, 2024

Comparing Non-Malignancy Rate Definitions Utilizing the NLST Cohort

Rationale: Harms from lung cancer screening (LCS) include high false positive rate of low dose computed tomography (LDCT) and potential downstream procedural risks patients may experience during subsequent evaluation. To monitor these downstream harms of LCS one proposed quality metric is the non-malignancy rate (NMR) of thoracic surgery. To be a reliable quality metric NMR must be measured consistently across practice settings. However, to date, the definition of NMR has been inconsistent across publications. Seemingly straightforward, the NMR is the number of surgical lung biopsy (SLB) procedures that identify a benign etiology divided by the total number of SLB procedures performed in a given cohort. However, the denominator (total number of biopsies performed) will differ if SLBs in patients with an established diagnosis prior to surgery are included. SLBs after a diagnosis of lung cancer occur as part of the surgical resection used to treat early stage lung cancer. Understanding the NMR with the exclusion of these biopsies is important because this rate will be more representative of the chances of a benign resection for patients enrolled in LCS. The National Lung Screening Trial (NLST) NMR of 24% has been used as the benchmark for current NMR literature but the definition used to determine that rate is unknown.

Objectives: The primary objective of this study is to analyze the NLST NMR using two different definitions: 1) The total number of SLB procedures with a benign diagnosis divided by the total number of diagnostic SLBs, excluding all individuals that have an established diagnosis of lung cancer before surgery (therapeutic biopsy/resection), and 2) The total number of SLB procedures with a benign diagnosis divided by the total number of surgical lung biopsy/resection procedures performed in NLST. The secondary objective of this study is to compare the NMR of NLST to the NMR reported in current literature to better understand if there has been a change in NMR since the publication of NLST.

Methods: This study will include a cross-sectional evaluation of NLST patients who underwent a surgical lung biopsy (thoracoscopy, thoracotomy, or mediastinoscopy) of a nodule identified with LDCT. Diagnosis dates and pathology from SLB would be evaluated based on the Diagnostic Procedures dataset form. Cases where a pathologic diagnosis of lung cancer from a procedure (bronchoscopy, transthoracic needle aspiration, etc.) completed prior to SLB will be defined “therapeutic biopsy”. Patients with a diagnosis of metastatic cancer non-lung primary and patients in the x-ray arm will be excluded. Primary analysis will be calculating the NMR using the two definitions discussed and comparison using tests of proportions. Secondary analysis will include comparison of the NLST rates to NMR rates in LCS populations identified in recent publications (2014-2024).

Aim 1: To evaluate the non-malignant resection rate (NMR) of surgical lung biopsy for the diagnosis of a pulmonary nodule in lung cancer screening patients in the National Lung Screening Trial (NLST) cohort using two proposed definitions that include or exclude therapeutic biopsy/resection.
Hypothesis 1: The NMR of surgical lung biopsy in the NLST cohort will be different based on definition used.

Aim 2: To compare the variability in NMR in the NLST cohort to NMR in current clinical practice based on recent publications.
Hypothesis: The NMR in the NLST cohort will be higher than the NMR in current clinical practice.