Predictive prognostic biomarkers in patients with COVID‑19 infection
Shunsuke Fukui, Kohei Ikeda, Mayu Kobayashi, Keigo Nishida, Keita Yamada, Shotaro Horie, Yasuaki Shimada, Hiroto Miki, Hiroki Goto, Koken Hayashi, Yuuichi Nakazawa, Hiroki Mizutani, Toshitaka Kamon, Yusuke Tanigaito, Shuji Kodama, Takashi Kato, Yuuki Nishiura, Daisuke Suga, Toshikazu Terashima, Yuhuko Ichikawa, Isao Moritani, Akitaka Yamamoto, Kei Takaba, Kouji Yasumoto, Hideo Wada, Katsuya Shiraki
Abstract
The present study aimed to identify useful biomarkers to predict deterioration in patients with coronavirus disease 2019 (COVID‑19). A total of 201 COVID‑19 patients were classified according to their disease severity into non‑severe (n=125) and severe (n=76) groups, and the behavior of laboratory biomarkers was examined according to the prognosis. Neutrophil count, aspartate aminotransferase (AST), alanine aminotransferase, lactate dehydrogenase (LDH), C‑reactive protein (CRP), sialylated carbohydrate antigen KL‑6 (KL‑6), procalcitonin (PCT), presepsin (PSP) and D‑dimer levels were significantly higher, and lymphocyte count and platelet count were significantly lower in the non‑severe group compared with the severe group. In the non‑severe group, ROC analysis demonstrated that only four biomarkers, CRP, PSP, AST and LDH were useful for differentiating the prognosis between improvement and deterioration subgroups. No strong correlation was revealed for any of the markers. Multivariate analysis identified CRP as a significant prognostic factor in non‑severe cases (odds ratio, 41.45; 95% confidence interval, 4.91‑349.24; P<0.001). However, there were no blood biomarkers that could predict the outcome of patients in the severe group. Overall, several blood markers changed significantly according to disease severity in the course of COVID‑19 infection. Among them, CRP, PSP, LDH and AST were the most reliable markers for predicting the patient's prognosis in non‑severe COVID‑19 cases.