Lung cancer cells are well-documented to rewire their metabolism and energy production networks to support rapid survival and proliferation. This metabolic reorganization has been recognized as a hallmark of cancer. The increased uptake of glucose and the increased activity of the glycolytic pathway have been extensively described. However, over the past years, increasing evidence has shown that lung cancer cells also require glutamine to fulfill their metabolic needs. As a nitrogen source, glutamine contributes directly (or indirectly upon conversion to glutamate) to many anabolic processes in cancer, such as the biosynthesis of amino acids, nucleobases, and hexosamines. It plays also an important role in the redox homeostasis, and last but not least, upon conversion to α-ketoglutarate, glutamine is an energy and anaplerotic carbon source that replenishes tricarboxylic acid cycle intermediates. The latter is generally indicated as glutaminolysis. In this review, we explore the role of glutamine metabolism in lung cancer. Because lung cancer is the leading cause of cancer death with limited curative treatment options, we focus on the potential therapeutic approaches targeting the glutamine metabolism in cancer.

Background Accumulating evidence has shown that cancer cell metabolism differs from that of normal cells. However, up to now it is not clear whether different cancer types are characterized by a specific metabolite profile. Therefore, this study aims to evaluate whether the plasma metabolic phenotype allows to discriminate between lung and breast cancer. Patients and methods The proton nuclear magnetic resonance spectrum of plasma is divided into 110 integration regions, representing the metabolic phenotype. These integration regions reflect the relative metabolite concentrations and were used to train a classification model in discriminating between 80 female breast cancer patients and 54 female lung cancer patients, all with an adenocarcinoma. The validity of the model was examined by permutation testing and by classifying an independent validation cohort of 60 female breast cancer patients and 81 male lung cancer patients, all with an adenocarcinoma. Results The model allows to classify 99% of the breast cancer patients and 93% of the lung cancer patients correctly with an area under the curve (AUC) of 0.96 and can be validated in the independent cohort with a sensitivity of 89%, a specificity of 82% and an AUC of 0.94. Decreased levels of sphingomyelin and phosphatidylcholine (phospholipids with choline head group) and phospholipids with short, unsaturated fatty acid chains next to increased levels of phospholipids with long, saturated fatty acid chains seem to indicate that cell membranes of lung tumors are more rigid and less sensitive to lipid peroxidation. The other discriminating metabolites are pointing to a more pronounced response of the body to the Warburg effect for lung cancer. Conclusion Metabolic phenotyping of plasma allows to discriminate between lung and breast cancer, indicating that the metabolite profile reflects more than a general cancer marker. Clinical trial registration number NCT02362776.

Lung cancer is the leading cause of cancer-related mortality worldwide, with five-year survival rates varying from 3–62%. Screening aims at early detection, but half of the patients are diagnosed in advanced stages, limiting therapeutic possibilities. Positron emission tomography-computed tomography (PET-CT) is an essential technique in lung cancer detection and staging, with a sensitivity reaching 96%. However, since elevated 18F-fluorodeoxyglucose (¹⁸F-FDG) uptake is not cancer-specific, PET-CT often fails to discriminate between malignant and non-malignant PET-positive hypermetabolic lesions, with a specificity of only 23%. Furthermore, discrimination between lung cancer types is still impossible without invasive procedures. High mortality and morbidity, low survival rates, and difficulties in early detection, staging, and typing of lung cancer motivate the search for biomarkers to improve the diagnostic process and life expectancy. Metabolomics has emerged as a valuable technique for these pitfalls. Over 150 metabolites have been associated with lung cancer, and several are consistent in their findings of alterations in specific metabolite concentrations. However, there is still more variability than consistency due to the lack of standardized patient cohorts and measurement protocols. This review summarizes the identified metabolic biomarkers for early diagnosis, staging, and typing and reinforces the need for biomarkers to predict disease progression and survival and to support treatment follow-up.

Pulmonary imaging often identifies suspicious abnormalities resulting in supplementary diagnostic procedures. This study aims to investigate whether the metabolic fingerprint of plasma allows to discriminate between patients with lung inflammation and patients with lung cancer. Metabolic profiles of plasma from 347 controls, 269 cancer patients and 108 patients with inflammation were obtained by 1H-NMR spectroscopy. Models to discriminate between groups were trained by PLS-LDA. A test set was used for independent validation. A ROC curve was built to evaluate the diagnostic performance of potential biomarkers. Sensitivity, specificity, PPV and NPV of PET-CT to diagnose cancer are 96, 23, 76 and 71%. Metabolic profiles differentiate between cancer and inflammation with a sensitivity of 89%, a specificity of 87% and a MCE of 12%. Removal of the glutamate metabolite results in an increase of MCE (38%) and a decrease of both sensitivity and specificity (62%), demonstrating the importance of glutamate for discrimination. At the cut-off point 0.31 on the ROC curve, the relative glutamate concentration discriminates between cancer and inflammation with a sensitivity of 85%, a specificity of 81%, and an AUC of 0.88. PPV and NPV are 92 and 69%. In PET-positive patients with a relative glutamate level ≤ 0.31 the sensitivity to diagnose cancer reaches 100% with a PPV of 94%. In PET-negative patients, a relative glutamate level > 0.31 increases the specificity of PET from 23% to 58% and results in a high NPV of 100%. In case of discrepancy between SUVmax and the glutamate concentration, lung cancer is missed in 19% of the cases. This study indicates that the 1H-NMR-derived relative plasma concentration of glutamate allows discrimination between lung cancer and lung inflammation. A glutamate level ≤ 0.31 in PET-positive patients corresponds to the diagnosis of lung cancer with a higher specificity and PPV than PET-CT. Glutamate levels > 0.31 in patients with PET negative lung lesions is likely to correspond with inflammation. Caution is needed for patients with conflicting SUVmax values and glutamate concentrations. Confirmation is needed in a prospective study with external validation and by another analytical technique such as HPLC-MS.

Metabolite profiling of blood plasma, by proton nuclear magnetic resonance (¹H-NMR) spectroscopy, offers great potential for early cancer diagnosis and unraveling disruptions in cancer metabolism. Despite the essential attempts to standardize pre-analytical and external conditions, such as pH or temperature, the donor-intrinsic plasma protein concentration is highly overlooked. However, this is of utmost importance, since several metabolites bind to these proteins, resulting in an underestimation of signal intensities. This paper describes a novel ¹H-NMR approach to avoid metabolite binding by adding 4 mM trimethylsilyl-2,2,3,3-tetradeuteropropionic acid (TSP) as a strong binding competitor. In addition, it is demonstrated, for the first time, that maleic acid is a reliable internal standard to quantify the human plasma metabolites without the need for protein precipitation. Metabolite spiking is further used to identify the peaks of 62 plasma metabolites and to divide the ¹H-NMR spectrum into 237 well-defined integration regions, representing these 62 metabolites. A supervised multivariate classification model, trained using the intensities of these integration regions (areas under the peaks), was able to differentiate between lung cancer patients and healthy controls in a large patient cohort (n = 160), with a specificity, sensitivity, and area under the curve of 93%, 85%, and 0.95, respectively. The robustness of the classification model is shown by validation in an independent patient cohort (n = 72).