In the present study, we verified the usefulness of PFAA profiling for lung cancer detection using new independent samples that had never been used for previous analysis and a derivation of multivariate discriminating function(s) that could distinguish lung cancer patients from control subjects. The results were highly reproducible for the change in PFAA profiles in lung cancer patients and highly discriminatory for lung cancer patients, including those with early stage cancer. Therefore, the results strongly suggest that our method is robust enough for clinical use. Moreover, because our method is a relatively simple plasma assay and imposes minimal physical burden on subjects, our findings suggest that PFAA profiling has great potential for improving the early detection of lung cancer.
Among the three discriminating functions, several amino acids were used in more than one function. His and Orn were incorporated into all of the functions, and Ser, Gln, Ala, Val, Ile, and Trp were incorporated into two of the three functions (Table 3) [32, 34, 35]. According to a comparison between the study and test data sets, plasma concentrations of Pro, Ile and Orn were higher in each data set, while the concentrations of Gln, His and Trp were lower (Figure 1 and Table 2). Among these amino acids, changes in the plasma concentrations of four amino acids (Pro, Ile, His, and Orn) were identical to the changes in amino acids in lung cancer patients in previous studies. Maeda et al. have reported that plasma concentrations of Ser, Pro, Gly, Ala, Met, Ile, Leu, Tyr, Phe, Orn, and Lys are increased and His is decreased in lung cancer patients . Miyagi et al. have also reported that the plasma concentrations of Ser, Pro, Gly, Ile, and Orn are increased, whereas Gln, Cit, His, and Trp are decreased in lung cancer patients . Therefore, the results strongly suggest the robustness of these three discriminating functions for the detection of lung cancer.
Moreover, Miyagi et al. have also reported that plasma levels of Gln, Trp, His, Pro, and Orn are commonly altered in cancer patients with five types of cancer (lung cancer, gastric cancer, colorectal cancer, breast cancer, and prostate cancer) . Therefore, the data also strongly suggest that the changes in plasma concentrations of Pro, His, and Orn are essentially associated with carcinogenesis and cancer progression regardless of the location of the tumor.
Although tumor markers have been used extensively to detect lung cancer and estimate clinical condition, the markers are not always useful due to low specificity and insufficient sensitivity. Therefore, combinatorial use of two or more independent tumor markers is necessary for clinical utility . Our results suggest that a PFAA-based diagnostic method would be a novel index to improve the insufficient clinical performance of the tumor markers. Combinatorial use of the tumor markers with Discriminant- 3 showed higher sensitivities than any of the tumor markers generally used for lung cancer patients. Additionally, only a low correlation was observed between the discriminating function scores and the tumor marker levels, suggesting the independence of the PFAA profiles from the existing tumor markers. Miyagi et al. have suggested that the change in the PFAA profile in cancer patients reflects two aspects: metabolic changes common to many cancers and metabolic characteristics specific to each cancer . Indeed, although the results were preliminary, the same study demonstrated the possibility of discriminating the cancer type. To clarify this hypothesis, testing the behavior of the discriminating function scores in lung cancer patients after surgery and chemotherapy and in those with recurrence would be necessary.
Because this study was designed as a case–control study, the results cannot be directly applied to further observation or prediction. Therefore, additional validation using a larger sample size is necessary to establish the clinical utility of our approach. Nonetheless, we believe that our results strongly suggest the clinical usefulness of the PFAA-based diagnostic method for the detection of lung cancer.