Curtis Harris

NCI

Chief

NLST (Learn more about this study)

NLST-165

Sep 25, 2015

Circulating inflammatory markers and urinary metabolites in combination with low dose computed tomography screening (LDCT).

We intend to identify, validate and further test to the highest level of evidence, novel and powerful biomarkers to complement successful screening protocols. Our approach is to use easily accessible biospecimens such as blood and urine. Indeed, having such easily available biospecimens is a great asset to a screening program. Our group uses a multi-factor approach to the generation of biomarker signatures for disease classification, with the hypothesis being that the measurement of disease signals in several biological fluids can impart both complementary and mutually exclusive information regarding disease classification. Cancer cells have distinguishable inflammatory and metabolic fingerprints compared to normal cells, creating the possibility that these profiles can detect cancer in its early stages.
Numerous studies have shown that chronic inflammation is associated with lung carcinogenesis. Using the National Cancer Institute-Maryland [NCI-MD] case control study, we previously found that increased serum levels of IL-8, IL-6 and CRP were associated with an increased risk of lung cancer and that IL-6 was associated with an increased risk of lung cancer mortality. These inflammatory markers were initially chosen as part of a larger spectrum of inflammatory cytokines, and in combination with other findings from the literature, have shown the strongest associations with lung cancer. Conscious that utility is enhanced if the biomarker is associated with a lesion/tumor prior to clinical detection by radiographical screening, we extended our analysis to the prospective Prostate, Lung, Colorectal and Ovarian (PLCO) cohort where we found that levels of these biomarkers, before diagnosis, were predictive of lung cancer. This suggested that they could have additional utility in a setting where a validated screening tool is used. Thus, to test this assumption explicitly, we are now submitting an application for NLST plasma samples.
Furthermore, urinary metabolites may be able to capture subtle shifts in multiple independent metabolic pathways and specifically capture a metabolic fingerprint of lung cancer. Urine sampling is also non-invasive and requires minimal sample preparation. We recently developed an unbiased mass spectrometry-based metabolomics approach using urine samples from the NCI-MD case control study (469 lung cancer cases and 536 population controls) (Mathe E, Patterson A, Haznadar M et.al., Can Res). This is the largest untargeted metabolomics study to date in lung cancer.

Specific Aim 1: Determine the ability of circulating inflammatory markers and urinary metabolites to distinguish between benign and malignant LDCT-detected pulmonary nodules. The success of a screening program at the population level depends on multiple factors; the disease must be important and detectable in a preclinical phase, an effective treatment, one that works better if applied early, must be available, and the test must be accurate, acceptable and carry a low risk. LDCT has an enhanced sensitivity compared to previous methods such as chest X-ray and sputum cytology, but the rate of false positives and the detection of benign lung lesions have increased. In some studies, up to 70% of individuals have an undefined nodule (3-5), while up to 30% of suspicious lesions are benign at time of surgery (3,4,6). Therefore, biomarkers that can discriminate between benign and malignant lesions will have a significant utility for screening programs. With that goal in mind, the aims of this application are:
Specific Aim 1A: Validate the ability of plasma inflammatory profiles and/or urine metabolites to predict lung cancer in samples obtained at lung cancer diagnosis, as well as examine whether they can enhance accuracy and distinguish between benign and malignant nodules, both alone and in combination with LDCT.

Specific Aim 2: To investigate if plasma inflammatory and urine metabolomic profiles are associated with overall and lung cancer-specific mortality. There are few, if any, validated prognostic biomarkers of clinically detected lung cancer. Moreover, the natural history of sub-centimetric LDCT-detected lung cancer is not fully understood. For example, lesions discovered by LCDT, rather than clinical diagnosis, are not just smaller, but may also have lower rates of progression and a better prognosis. By annotating which patients will have a poor prognosis, biomarkers that can predict both overall and lung cancer-specific mortality in patients with LDCT-detected lesions will also be advantageous to screening programs and could potentially aid in the selection of downstream therapeutic protocols.
Specific Aim 2A: Test the ability of plasma inflammatory markers and urine metabolites in samples obtained at diagnosis to predict overall and lung cancer specific mortality, as well as their ability to predict aggressive vs. indolent stage 1 lung cancers detected by LDCT.