1. Metabolic PHENOTYPE by NMR spectroscopy: A biomarker for lung CANCER
E Louis1, K Baeten1, K Vandeurzen2, K Darquennes3, K Vanhove4, L Mesotten5, M Thomeer5, E de Jonge5, K van der Speeten5, J Mebis6, P Adriaensens7
1Biomedical Research Institute, Hasselt University, Diepenbeek, Belgium, 2Mariaziekenhuis Noord-Limburg, Overpelt, Belgium, 3Ziekenhuis Maas en Kempen, Maaseik, Belgium, 4AZ Vesalius, Tongeren, Belgium, 5Ziekenhuis Oost-Limburg, Genk, Belgium, 6Jessa ziekenhuis, Hasselt, Belgium, 7Institute of Material Research, Hasselt University, Diepenbeek, Belgium
INTRODUCTION
Globally, lung cancer is the 2nd most common cancer in men and the 3rd in women. Moreover, it is the leading cause of cancer death worldwide.
Screening for lung cancer remains a challenge. Consequently, it is often only diagnosed at a metastatic stage.
Computed tomography (CT) allows the early detection of lung cancer. However, it is not suitable as a general screening method due to its high cost, the related exposure to radiation and the presence of lead-time bias and overdiagnosis. Therefore, screening tools with an improved specificity and sensitivity are urgently needed!
Recent studies indicate that metabolic fingerprinting of a blood sample by means of nuclear magnetic resonance (NMR) spectroscopy has the potential for early disease detection. For several diseases, changes in the metabolic fingerprint have been shown to correlate with the presence of a certain pathology. Identifying which changes in the metabolic fingerprint correlate with the presence of lung cancer could allow us on the long-term to easily detect the presence of lung cancer in a simple blood sample.
At present, positron emission tomography (PET) is used for the staging of lung cancer. This technique gives information about the metabolic alterations at the level of the tumor. Because we both study metabolic changes in the blood and at the tumor site, it is possible to investigate if there is a correlation between these two.
Intravenous administration of the radioactive tracer 18F-FDG.
Execution of a whole body PET/CT scan one hour after this (Gemini TF Big Bore PET/CT, Philips).
Measurement of the degree of tumoral 18F-FDG uptake with a semi-automatic delineation tool (Frings, Boellaard 2010). In particular, calculation of 3 predefined threshold volume of interests (VOIs) consisting of 41%, 50% and 70% of the maximum voxel value, with correction for local background (A41%, A50% en A70%).
Figure 3: Calculation of threshold VOIs corrected for local background.
PRELIMINARY DATA
Subject characteristics
Table 1: Subject characteristics.
Discrimination of lung cancer patients from healthy controls based on their metabolic fingerprint
Currently, the metabolic fingerprint of 29 lung cancer patients ( ) and 29 healthy controls ( ) was analyzed by NMR spectroscopy.
By means of OPLS-DA analyses on a well-defined and selected panel of metabolites present, a high degree of discrimination between control subjects and patients with lung cancer was achieved with a specificity of 96% and a sensitivity of 93%.
Figure 4: OPLS-DA analyses of plasma samples distinguish between lung
cancer patients and healthy controls.
FUTURE RESEARCH
The metabolic fingerprint of all collected blood samples will be determined by means of NMR spectroscopy.
Statistical analyses will be performed to verify whether there is a difference in the metabolic fingerprint of lung cancer patients as compared to healthy controls, patients with inflammatory lung diseases and patients whose lesion did not show uptake on PET. Furthermore, we will investigate whether the observed metabolic differences can discriminate lung cancer patients and whether these metabolic differences can become a potential tool for the detection of lung cancer.
The predefined threshold VOIs corrected for local background (A41%, A50% and A70%) will be calculated for all patients. Furthermore, regression analysis between these VOIs and all the metabolites present in the metabolic fingerprint will be performed to determine if there is a correlation between the metabolic changes at the tumor site and the metabolic changes in the blood.
SUV max lesion = 7,4
SUV mean background = 2,06
A41% VOI = 3,88
A50% VOI = 4,73
A70% VOI = 6,6
HYPOTHESES
Metabolic fingerprint analyses by NMR spectroscopy allow the early detection of lung cancer.
Metabolic changes, detected in the blood of lung cancer patients, correlate with metabolic alterations at the tumor site.
SUBJECTS AND METHODS
Study population
Patients with a new lesion in the lung on CT, referred for a PET/CT in the Limburgs PET center in Hasselt between March 2011 and August 2011, were included. These patients came from different hospitals, namely Ziekenhuis Oost-Limburg (Genk), AZ Vesalius (Tongeren), Mariaziekenhuis Noord-Limburg (Overpelt) and Ziekenhuis Maas en Kempen (Maaseik). After confirmation of diagnosis by anatomopathologic research, the group of patients turned out to consist of lung cancer patients, patients with inflammatory lung diseases and patients whose lesion did not show uptake on PET (schwannoma, sequel of a passed infection, …). In a second cohort, healthy volunteers without any indication of any pulmonary disease were recruited. These healthy controls did not undergo a CT or a PET. The study protocol was approved by the ethical comité of the Ziekenhuis Oost-Limburg.
NMR spectroscopy
Figure 1: Determination of the metabolic fingerprint of a blood sample by means of NMR spectroscopy.
Collection of blood samples.
Measurement of NMR spectra by means of a 400 MHz Varian spectrometer.
Calculation of metabolic differences between the different groups by student T-tests and supervised orthogonal partial least squares-discriminant analyses (OPLS-DA).
18F-FDG PET/CT
Figure 2: Obtaining an image with 18F-FDG PET/CT.
Specificity: 96%
Sensitivity: 93%