Genomic Prognostic Models in Early-Stage Lung Cancer

Johannes R. Kratz; David M. Jablons


Clin Lung Cancer. 2009;10(3):151-157. 

In This Article

Abstract and Introduction


Patients with early-stage lung cancer demonstrate significant recurrence rates and lower-than-expected survival rates after surgical resection, indicating that our current staging methods do not adequately predict outcome. Since the last revision of the TNM staging system, a number of genomic models have been proposed which more accurately predict prognosis in patients with early-stage lung cancer. A variety of prognostic genomic models based on gene-expression profiling and quantitative polymerase chain reaction (PCR) are able to stratify patients with early-stage lung cancer into high- and low-risk groups with respect to disease-free and overall survival. In the future, clinical application of these models may ultimately dictate both the use of adjuvant therapy as well as the choice of surgical procedure in patients with early-stage lung cancer. An effort to develop a robust genomic model for use in the clinical setting should be prompted by encouraging results obtained by the use of a quantitative PCR-based genomic signature in the field of breast oncology.


Lung cancer is the leading cause of cancer-related mortality in the United States.[1] It is estimated that 215,020 new cases of lung cancer are diagnosed annually, leading to approximately 161,840 cancer deaths per year. Despite knowledge that the majority of these deaths are directly related to smoking and thus preventable, the incidence and cancer death rates for lung cancer remain high among men and continue to increase among women. Remarkably, although many breakthroughs in our understanding, diagnosis, and treatment of lung cancer have been made, the 5-year survival rate of lung cancer has changed little over the past 30 years. In 1975, the 5-year survival rate for all lung cancers was 13%; today it is estimated to be 15%. This is commonly attributed to the fact that over 75% of lung cancers demonstrate either regional or distant spread at the time of diagnosis.

The lung cancer 5-year survival rate is dependent on a number of factors, including anatomic factors (eg, histology, tumor size, local obstruction, nodal involvement, presence of distant metastasis, etc.) and a variety of patient, hematologic, and molecular factors.[2] Clinically, the most powerful method of predicting the survival rates of patients with lung cancer is the tumor-node-metastasis (TNM) staging system.[3] This system, currently under revision,[4] uses tumor size, local invasion, and presence of nodal and distant metastases to categorize patients in 1 of 7 clinical stages (IA, IB, IIA, IIB, IIIA, IIIB, IV).[3] Five-year survival rates range from 67% for pathologically diagnosed localized stage IA lung cancer to 1% for metastatic stage IV lung cancers. Surgical resection represents the mainstay of treatment for patients with early-stage (stage I and II) lung cancer and is considered curative in patients with stage I lung cancer.[5,6] Despite this, the 5-year survival rate of even the earliest stage IA lung cancers after complete surgical resection is still only approximately 70%, indicating that TNM staging alone does not adequately predict outcome.[7,8]

Patients with completely resected stage I lung cancer who have no evidence of distant or regional disease spread develop disease recurrence because of the presence of subclinical micrometastatic disease. The recurrence rate of completely resected stage I lung cancers has been reported to be as high as 15.7% in patients with a mean follow-up of 3 years.[9] In addition, the only randomized prospective trial to date of patients with stage I lung cancer undergoing lobectomy versus limited (segment or wedge) resection has shown that patients undergoing lobectomy have less locoregional recurrence.[10] This is thought to be because lobectomy more completely excises occult micrometastases at the time of resection. One of the biggest challenges that remains, therefore, is being able to more accurately predict which patients with early-stage disease are likely to have subclinical metastases undetectable by our current imaging and pathologic modalities.

The ability to identify patients with early-stage lung cancer who are at high risk for disease-specific mortality would be beneficial in a number of ways. Currently, controversy exists with respect to the benefit of adjuvant therapy following resection of limited-stage lung tumors.[11,12,13,14,15,16] Identifying high-risk patients with early-stage disease may reveal a survival benefit from adjuvant therapies post surgical resection. In addition, although the randomized prospective trial by the Lung Cancer Study Group referenced above demonstrated increased survival for those patients undergoing lobectomy versus limited resections for stage I disease,[10] other recent retrospective trials have challenged this view.[17,18,19,20,21] Better prognostic risk-stratification among patients with early-stage disease may allow identification of low-risk patients who achieve adequate survival with less radical lung cancer operations such as wedge resection or segmentectomy.

Many clinical features, pathologic characteristics, and biomarkers have been studied in an attempt to improve risk stratification over the TNM staging system.[2] In particular, much effort has focused on whether prognostic indicators exist on a molecular level. Early efforts in this area led to several proteomic models of risk stratification based on profiling of oncoproteins and other protein products[22,23,24] (for an excellent review of these models, see D'amico[25]). In the past few years, however, attention has shifted to genomic models that provide additional prognostic information beyond that of the TNM staging system. This review focuses on the major microarray and quantitative polymerase chain reaction (PCR)-based models that have been proposed to help us better risk-stratify patients with early-stage lung cancer. Although studies investigating the prognostic significance of other genomic markers such as tumor DNA copy numbers, mutations, methylation changes, microenvironmental analysis, cancer stem cell marker analysis, tumor miRNAs, and germ line DNA polymorphisms have been published and/or are being pursued, they are beyond the scope of this review.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: