Studying Tumor Tissue Samples From Patients With Early-Stage Non-Small Cell Lung Cancer

PRIMARY OBJECTIVES:

I. To develop and validate a clinically useful molecular prognostic signature based on
ribonucleic acid (RNA)-expression arrays to accurately distinguish between good and poor
outcome in patients with stage 1B non-small cell lung cancer (NSCLC) (T2aN0M0) by predicting
the risk of cancer recurrence after surgery.

SECONDARY OBJECTIVES:

I. To develop and validate a clinically useful prognostic signature for stage 1A NSCLC
(T1a-bN0M0) that accurately predicts the risk of post-operative cancer recurrence and is part
of a prognostic classifier, which also includes histological, pathological, and demographic
parameters as in the primary objective.

II. To determine the effects on test accuracy of the histological subtype (adenocarcinoma
[AC] vs squamous cell [SQA]) for each of the stage specific prognostic classifiers.

III. To determine the relevant statistical features, including accuracy of predicting overall
survival, cancer specific survival, and disease free survival at 3 and 5 years, for each of
the prognostic classifiers validated.

IV. To utilize the validated prognostic signatures and classifiers in stage 1B NSCLC to
determine whether these can reliably predict good outcome among stage II NSCLC patients.

OUTLINE:

RNA extracted from archived tumor tissue samples are analyzed for whole-genome expression by
Affymetrix microarrays (Gene Profiling Array cGMP U133 P2) and reverse
transcriptase-polymerase chain reaction (RT-PCR).

Inclusion Criteria:

- Registration to CALGB-140202

- The subject population to be studied in this protocol includes patients selected from
CALGB-140202; all such patients have signed a written informed consent document
meeting all federal, state and institutional guidelines as part of entry into that
trial; the CALGB does not require that a separate consent form be signed for this
study

- All samples to be studied were obtained and stored as part of CALGB-140202; the
material and data obtained from the patient's protocol record will be used to obtain
appropriate clinical information; in no instance will the patient be contacted
directly

- Tumor specimens need to have 40% tumor content to be included