Lung Cancer

The NLST confirmed diagnoses of lung cancer through medical record abstraction (MRA) of participants suspected by the trial to have lung cancer. MRA was triggered by:

• a self-report of lung cancer on a study update form (annual for LSS, semi-annual for ACRIN)
• positive CT or X-ray screening exam
• death certificate indicating lung cancer
• a relative informing screening center of participant's cancer (this was infrequent)
• suspicion of non-lung cancer that was determined (by MRA) to be lung cancer (e.g. metastasis to brain from a lung primary)

Note that if the MRA process did not find records indicating a lung cancer diagnosis, the lung cancer was not considered confirmed, even if a source such as a death certificate indicated lung cancer.

Data collection was pursued for all lung cancers diagnosed at any time during NLST. This included additional primary lung cancers diagnosed after the first, not just the first lung cancer per participant. The focus of data collection was on characteristics of the cancer at the time of diagnosis.

Cancer stage was originally classified according to the AJCC 6 manual, but AJCC 7 stage data are also available from a re-review of medical records after study closeout.

The Participant dataset contains sufficient information on lung cancer diagnosis to be useful for many purposes. Additional information is available in a separate Lung Cancer dataset, which contains one record per primary lung cancer.

Other (Non-Lung) Cancer

In addition to lung cancer, the NLST confirmed diagnoses of non-lung cancer. This was mostly to ensure accurate ascertainment of lung cancer status and cause of death, but the data on non-lung cancer are useful in their own right. Less detail was collected about non-lung cancers than lung cancers.

The NLST confirmed diagnoses of non-lung cancer through medical record abstraction (MRA) of participants suspected by the trial to have non-lung cancer. MRA was triggered by:

• a self-report of non-lung cancer on a study update form (annual for LSS, semi-annual for ACRIN)
• death certificate indicating non-lung cancer
• a relative informing screening center of participant’s cancer (this was infrequent)
• suspicion of lung cancer that was determined (by MRA) to be non-lung cancer (e.g. metastasis to lung)

Note that if the MRA process did not find records indicating a diagnosis of non-lung cancer, the cancer was not considered confirmed, even if a source such as a death certificate indicated non-lung cancer.

Data collection was pursued for all non-lung cancers diagnosed at any time during NLST (except for basal and squamous cell skin cancers). This included additional primary cancers of a non-lung site that were diagnosed after the first cancer at that site.

ACRIN and LSS collected different information on non-lung cancer. ACRIN collected data on each pathology specimen obtained during diagnostic work-up of suspected cancer, while LSS recorded data about each tumor, not each pathology specimen. The ACRIN and LSS data were then harmonized to provide trial-wide, tumor-specific data. This entailed summarizing the ACRIN pathology specimen data to tumor-level through a review of the data. The trial-wide data include the ICD-O-3 code and date of diagnosis for each tumor.

The Participant dataset contains the non-lung cancer data. Data on all cancers (lung and non-lung) are stored in arrays of four elements, one for each confirmed cancer diagnosed during NLST, sorted by date of diagnosis. For instance, the variables confirmed_icd_morph1-4 contain the ICD-O-3 morphology code for the first through fourth confirmed cancers (of any site) that a participant was diagnosed with during NLST. Lung cancers are included in the arrays to provide the sequence of all cancers diagnosed during the trial, but better information on lung cancer is available in other variables in the participant dataset. No participant had more than four cancers diagnosed during NLST.