African Journal of Internal Medicine ISSN: 2326-7283 Vol. 9 (8), pp. 001-008, August, 2021. © International Scholars Journals

Full Length Research Paper

Diagnostic modeling with differences in plasma amino acid profiles between non-cachectic colorectal/breast cancer patients and healthy individuals

Naoyuki Okamoto1*, Yohei Miyagi2, Akihiko Chiba3, Makoto Akaike4, Manabu Shiozawa4, Akira Imaizumi5, Hiroshi Yamamoto6, Toshihiko Ando5, Minoru Yamakado7 and Osamu Tochikubo8

1Cancer Prevention and Cancer Control Division, Kanagawa Cancer Center, 1-1-2, Nakao, Asahi-ku, Yokohama-city, 241-0815 Japan.

2Moleculer Pathology and Genetics Division, Kanagawa Cancer Center, 1-1-2, Nakao, Asahi-ku, Yokohama-city, 241- 0815 Japan.

3Department of Breast Surgery, Kanagawa Cancer Center, 1-1-2, Nakao, Asahi-ku, Yokohama-city, 241-0815 Japan.

4Department of Gastrointestinal Surgery, Kanagawa Cancer Center, 1-1-2, Nakao, Asahi-ku, Yokohama-city, 241-0815 Japan.

5Institute of Life Sciences, Ajinomoto CO., Inc., Japan.

6HI Department, Ajinomoto CO., Inc. , Japan.

7Center for Multiphasic Health Testing and Services, Mitsui Memorial Hospital, Japan.

8Department of Public Health, Yokohama City University, Graduate School of Medicine, Japan.

Accepted 22 December, 2020

Abstract

Metabolic changes in patients with cancer lead to alterations in their amino-acid balances. Thus, amino-acid profiles may be useful as biomarkers of cancers. This study was conducted to analyze amino-acid profiles in plasma by multivariate analysis, in order to elucidate differences between cancer patients and controls. Venous blood samples were taken from colorectal and breast cancer patients, and healthy controls. Plasma free amino acids were measured by liquid chromatography/mass spectrometry. No weight loss was observed in any of the cancer patients. Multiple logistic regression models were used to discriminate between cancer patients and controls. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve for each discriminant score was calculated as 0.860 (95% confidence interval: 0.784 to 0.937) for colorectal cancer and 0.906 (95% CI: 0.845 to 0.967) for breast cancer. The performance of these discriminants was independent of the cancer stage. This study revealed significant differences in plasma amino acid profiles between cancer patients and controls. The development of a cancer alters plasma amino-acid profiles without cachexia or weight loss, and the pattern of change differs between two cancers. Plasma amino-acid profiling might therefore be useful for the early detection of cancer.

Key words: amino acid profiles, plasma, screening, cancer, multivariate analysis