Advertisement

Serum cystatin C and CRP are early predictive biomarkers for emergence of hypoxia in COVID-19

      Abstract

      Background

      In Japan, during the coronavirus disease 2019 (COVID-19) pandemic, patients with non-hypoxia are recommended to recuperate at home or in pre-hospital facilities. However, it was observed that unexpected hypoxia may occur and become severe subsequently in patients whose symptoms were initially expected to improve naturally. The aim of this study is to validate biomarkers that can predict at an early stage the emergence of hypoxia in COVID-19 patients without hypoxia.

      Methods

      We retrospectively enrolled 193 patients with COVID-19, excluding patients with hypoxia and severe disease from the onset. Participants were classified into two groups according to the emergence of hypoxia during the clinical course, and the laboratory data were compared to identify biomarkers that could predict early the emergence of hypoxia.

      Results

      The areas under the curve for serum cystatin C (CysC) and C-reactive protein (CRP) levels for the emergence of hypoxia during the clinical course were higher than those for other biomarkers (CysC, 0.84 and CRP, 0.83). Multivariate analysis showed that high serum CysC and CRP levels were associated with the emergence of hypoxia during the clinical course.

      Conclusions

      Elevated serum CysC and CRP levels were associated with the emergence of hypoxia during the clinical course in COVID-19 patients without hypoxia. These findings may help determine the need for hospitalization in initially non-hypoxic COVID-19 patients.

      Key Indexing Terms

      Introduction

      Coronavirus disease 2019 (COVID-19) cases are increasing exponentially. As of April 22, 2021, over 140 million people have been infected and over 3 million people have died worldwide.
      Johns Hopkins University and Medicine
      COVID-19 map.
      Japan experienced three major waves of a surge in infection between 2020 and 2021, creating social problems.
      • Karako K
      • Song P
      • Chen Y
      • et al.
      Overview of the characteristics of and responses to the three waves of COVID-19 in Japan during 2020-2021.
      In Japan, recuperation at home or in a hotel is recommended for patients with low disease severity during surges in infection. However, some patients who were convalescing at home or in pre-hospital facilities because they did not need oxygen treatment experienced rapid hypoxia during the clinical course, requiring hospitalization or even resulting in death.
      • Nakamura M
      • Hitosugi M.
      Emerging COVID-19-related deaths during home recuperation in Japan.
      Therefore, it is necessary to elucidate the factors that can predict the emergence of hypoxia, even in non-hypoxic COVID-19 patients.
      Various factors that predict disease severity, such as oxygen demand and in-hospital mortality, have been reported. A Japanese study involving 345 COVID-19 patients reported that oxygen demand is associated with age, shortness of breath, and general fatigue, and death is associated with age, hyperuricemia, and chronic kidney disease.
      • Ishii M
      • Terai H
      • Kabata H
      • et al.
      Clinical characteristics of 345 patients with coronavirus disease 2019 in Japan: a multicenter retrospective study.
      In addition, data from patients with acute hypoxic respiratory failure in New York showed that age, chronic cardiac disease, chronic pulmonary disease, interleukin (IL)-6 and D-dimer levels are associated with in-hospital mortality.
      • Cummings MJ
      • Baldwin MR
      • Abrams D
      • et al.
      Epidemiology, clinical course, and outcomes of critically ill adults with COVID-19 in New York City: a prospective cohort study.
      However, there are still few reports on biomarkers that can predict the emergence of hypoxia, validated only in COVID-19 patients without hypoxia. An investigation of these biomarkers may identify the criteria for recommending hospitalization, even in patients who are not hypoxic and whose symptoms are expected to improve naturally. The number of hospital beds is limited universally, and hence, it is important to identify those patients who should be recommended hospitalization during drastic surges in infection. Therefore, in this study, we validated the biomarkers that may predict the emergence of hypoxia during the COVID-19 clinical course in patients who did not require oxygen treatment.

      Methods

      Study design and patients

      In this double-center retrospective study, we investigated the biomarkers that could predict the emergence of hypoxia early during the clinical course in COVID-19 patients without hypoxia. We enrolled 281 COVID-19 patients admitted to the Showa University Hospital and Showa University East Hospital from March 28, 2020 to April 22, 2021. Of the 88 excluded patients, 28 were transferred from or to another hospital during the disease course, 23 required oxygen treatment at admission, 11 were administered COVID-19 therapeutic agents (remdesivir, favipiravir, tocilizumab, baricitinib, and/or steroids) and discharged without oxygen treatment, 14 did not undergo chest computed tomography (CT), five were suspected of having sleep apnea and were administered oxygen during sleep, six had incomplete data, and one patient had initiated home oxygen therapy before admission (Figure 1).
      Figure 1
      Figure 1Consolidated Standards of Reporting Trials flow diagram. COVID-19: Coronavirus disease 2019.
      We analyzed 193 patients who satisfied the inclusion criteria. First, we categorized the patients into two groups: one with patients that were discharged without hypoxia during the clinical course (“non-hypoxia group”) and one in which oxygen treatment was administered due to hypoxia (“hypoxia group”). Hypoxia was defined as the condition at which oxygen treatment was started with percutaneous oxygen saturation <90%. Second, patient characteristics, comorbidities, symptoms, and laboratory data were compared between the two groups. All laboratory data were obtained within 24 hours before and after admission with a diagnosis of COVID-19. We also compared the threshold cycle (Ct) values in real-time reverse transcription-polymerase chain reaction (RT-PCR) for COVID-19 between the groups. This study was approved by the ethics committee of the Showa University (Approval number: 3438).

      Serum CysC and CRP measurements

      Serum cystatin C (CysC) levels were measured using latex agglutination turbidimetric immunoassay by BML (Kawagoe, Japan). Serum C-reactive protein (CRP) levels were measured using latex agglutination turbidimetric immunoassay at the hospital laboratory of Showa University.

      RT-PCR measurements

      We used the reagent of the SARS-CoV-2 Detection Kit (TOYOBO, OSAKA, Japan) and performed RT-PCR on nasopharyngeal swab samples at the Showa University Hospital PCR center for COVID-19. The RT-PCR assays were conducted in a QuantStudio 3 real-time PCR system (Thermo Fisher Scientific, Waltham, MA, USA). A total of 45 cycles were set with the following cycling conditions: 42°C for 5 min for reverse transcription, 95°C for 10 s for pre-denaturation, 95°C for 1 s for denaturation, 50°C for 10 s for annealing, and 55°C for 10 s for extension. All Ct values obtained in this study were measured using a N2 primer-probe set.

      Statistical analysis

      The analyses were performed using JMP Pro ® 15 (SAS Institute Inc., Cary, NC, USA). The Anderson-Darling test showed that no continuous variables were normally distributed. Continuous variables are expressed as median (interquartile range [IQR]), and statistical significance was set at p < 0.05 (two-sided). Differences between the groups were analyzed with χ2 test for categorical variables and unpaired Mann-Whitney U-test for continuous variables.
      Moreover, we considered the emergence of hypoxia during the clinical course as the dependent variable, created receiver operating characteristic (ROC) curves, calculated areas under the curves (AUCs), and measured the cutoff values for white blood cell (WBC), D-dimer, creatinine, lactate dehydrogenase (LDH), ferritin, CRP, and CysC levels. Optimal cutoff values were determined using the Youden index.
      Additionally, we performed multivariate logistic regression analysis with the emergence of hypoxia during the clinical course as the dependent variable and age, gender, pneumonia, CRP, and CysC as independent variables.

      Results

      Patient characteristics

      The median observation period in the hospital was 10 (IQR, 8-12) days. Table 1 shows the patient characteristics. The study population comprised 100 men (51.8%) and 93 women (48.2%), and the median age was 47 (IQR, 30-60) years. The median body mass index (BMI) was 22.7 kg/m2, and 18 (9.3%) patients were obese (BMI ≥ 30 kg/m2). Hypertension was the most common comorbidity (20.2%), followed by hyperlipidemia (13.5%), diabetes (6.7%), and emphysema (6.7%). The most common symptoms were fever (79.3%), followed by cough (45.1%) and olfactory disorder (22.8%). Additionally, 67 (34.7%) patients reported a history of smoking and 117 (60.6%) exhibited pneumonia on CT.
      Table 1Baseline clinical characteristics and laboratory data of patients included in the study.
      Total (N = 193)Non-hypoxia (N = 150)Hypoxia (N = 43)P- value
      Age (years)
       Median47 (30-60)41 (27-55)65 (53-76)<0.0001
       10-193 (1.6%)3 (2.0%)0 (0.0%)
       20-2943 (22.3%)43 (28.6%)0 (0.0%)
       30-3930 (15.5%)28 (18.7%)2 (4.7%)
       40-4928 (14.5%)23 (15.3%)5 (11.6%)
       50-5940 (20.7%)30 (20.0%)10 (23.3%)
       60-6915 (7.8%)7 (4.7%)8 (18.6%)
       70-7925 (13.0%)12 (8.0%)13 (30.2%)
       80-898 (4.1%)4 (2.7%)4 (9.3%)
       90-991 (0.5%)0 (0.0%)1 (2.3%)
      Sex
       Female93 (48.2%)80 (53.3%)13 (30.2%)0.0092
       Male100 (51.8%)70 (46.7%)30 (69.8%)
      Race
       Japanese183 (94.8%)143 (95.3%)40 (93.0%)0.6955
       Non-Japanese10 (5.2%)7 (4.7%)3 (7.0%)
      BMI (kg/m2)22.7 (20.0-25.1)21.8 (19.8-24.6)24.0 (22.8-27.0)0.0002
       Obesity18 (9.3%)11 (7.3%)7 (16.3%)0.1316
      Smoking history
       Current or former smoker67 (34.7%)43 (28.7%)24 (55.8%)0.0018
      Finding by CT
       Pneumonia117 (60.6%)76 (50.7%)41 (95.3%)<0.0001
      Respiratory support
       Nasal cannula oxygen therapy37 (19.2%)037 (86.0%)<0.0001
       Face mask oxygen therapy3 (1.6%)03 (7.0%)
       IMV3 (1.6%)03 (7.0%)
      Symptoms
       Fever153 (79.3%)120 (80.0%)33 (76.7%)0.6715
       Cough87 (45.1%)69 (46.0%)18 (41.9%)0.7287
       Sputum16 (8.3%)14 (9.3%)2 (4.7%)0.5308
       Sore throat49 (25.4%)44 (29.3%)5 (11.6%)0.0178
       Rhinorrhea20 (10.4%)15 (10.0%)5 (11.6%)0.7783
       Dyspnea25 (13.1%)22 (14.7%)3 (7.0%)0.3011
       Taste disorder43 (22.3%)37 (24.7%)6 (14.0%)0.1518
       Olfactory disorder44 (22.8%)36 (24.0%)8 (18.6%)0.54
       Nausea8 (4.1%)6 (4.0%)2 (4.7%)1.0000
       Diarrhea11 (5.7%)7 (4.7%)4 (9.3%)0.2672
      Comorbidities
       Hypertension39 (20.2%)20 (13.3%)19 (44.2%)<0.0001
       Diabetes13 (6.7%)8 (5.3%)5 (11.6%)0.1688
       Hyperlipidemia26 (13.5%)16 (10.7%)10 (23.3%)0.0430
       Hyperuricemia10 (5.2%)4 (2.7%)6 (14.0%)0.0092
       Bronchial asthma11 (5.7%)8 (5.3%)3 (7.0%)0.7112
       Emphysema13 (6.7%)3 (2.0%)10 (23.3%)<0.0001
       COPD2 (1.0%)0 (0%)2 (4.7%)0.0487
       Sleep apnea syndrome5 (2.6%)2 (1.3%)3 (7.0%)0.0743
      Laboratory data
      White blood cell count (/ μl)4500 (3700-5550)4350 (3600-5300)5100 (4200-6200)0.0020
      Lymphocytes count (/ ml)1040 (780-1355)1080 (790-1370)950 (770-1250)0.2959
      D-dimmer (mg/ml)0.74 (0.59-0.98)0.70 (0.56-0.88)0.96 (0.78-1.17)<0.0001
      Creatinine (mmol/L)0.7 (0.57-0.85)0.65 (0.55-0.82)0.82 (0.70-0.96)<0.0001
      Lactate dehydrogenase (U/L)189 (161-233)176 (155-215)241 (214-301)<0.0001
      Creatine kinase (U/L)74 (51-119)74 (50-117)75 (55-149)0.3520
      C-reactive protein (mg/dl)0.89 (0.24-2.78)0.51 (0.19-2.05)4.20 (1.44-7.83)<0.0001
      Ferritin (ng/mL)207 (87-505)171 (66-361)598 (229-893)<0.0001
      Cystatin C (mg/L)
      Cystatin C was measured in a total of 152 patients, 116 in the non-hypoxia group and 36 in the hypoxia group.
      0.87 (0.75-0.99)0.83 (0.73-0.92)1.04 (0.93-1.31)<0.0001
      Data are expressed as median (interquartile range) or N (%).
      BMI, body mass index; COPD, Chronic obstructive pulmonary disease; IMV, Invasive mechanical ventilation
      low asterisk Cystatin C was measured in a total of 152 patients, 116 in the non-hypoxia group and 36 in the hypoxia group.

      Comparison between hypoxia and non-hypoxia groups

      Table 1 shows the results of comparison between the groups. Among patient characteristics, there were significant differences in age, sex, BMI, and smoking history between the two groups. Moreover, significant differences were evident in findings suggestive of pneumonia on chest CT and in the symptom of sore throat. Among comorbidities, there were significant differences in the incidence of hypertension, diabetes, hyperlipidemia, hyperuricemia, emphysema, and chronic obstructive pulmonary disease (COPD). Furthermore, the laboratory data showed significant differences in WBC, D-dimer, creatinine, LDH, CRP, ferritin, and CysC levels.

      Early predictive factors for need of oxygen treatment

      The results of multivariate logistic regression analysis showed that older age (hazards ratio [HR], 1.60 [95% CI 1.08-2.46] per 10-year increase), presence of pneumonia (HR, 10.40 [95% CI 1.63-186.13]), high CRP (HR 1.39 [95% CI 1.10-1.85]), and high CysC (HR 1.36 [95% CI 1.09-1.82] per 0.1 mg/L increase) were associated with the emergence of hypoxia during the clinical course (Table 2).
      Table 2Logistic regression analysis for the emergence of hypoxia in COVID-19 patients without hypoxia.
      Risk factorsOdds ratio95% CIP- value
      Age (per 10-year increase)1.601.08-2.460.0219
      Sex (Male)3.050.90-11.660.0835
      Pneumonia10.401.63-186.130.0397
      C-reactive protein1.391.10-1.850.0145
      Cystatin C (per 0.1mg/L increase)1.361.09-1.820.0146
      95% Cl; 95% confidence interval
      Supplementary Figure 1 shows the Ct values in RT-PCR for all patients. The median Ct value for the entire study sample was 28.6 (IQR 23.1-32.4) cycles, and the median time from disease onset to data collection was 6 (IQR 2-10) days. As shown in Supplementary Figure 1A, 132 patients who underwent RT-PCR at least once during hospitalization and in whom Ct values were measured were categorized into 25 and 107 patients in the hypoxia and non-hypoxia groups, respectively. Moreover, the Ct value in RT-PCR for COVID-19 is thought to attain a peak within one week of onset.
      • Sethuraman N
      • Jeremiah SS
      • Ryo A.
      Interpreting diagnostic tests for SARS-CoV-2.
      With reference to this report, 78 patients who underwent RT-PCR at our hospital within one week of onset were categorized into 13 and 65 patients in the hypoxia and non-hypoxia groups, respectively (Supplementary Figure 1B). The two groups were compared using the first Ct values measured at our hospital, and no significant differences were observed.

      AUC and cutoff value for each biomarker

      Figure 2 and Table 3 show the ROC curve and AUC for each biomarker. The AUCs for CysC, CRP, LDH, ferritin, D-dimer, creatinine, and WBC levels were 0.84 (95% confidence interval [CI] 0.74-0.89), 0.83 (95% CI 0.74-0.89), 0.82 (95% CI 0.73-0.87), 0.77 (95% CI 0.69-0.85), 0.75 (95% CI 0.66-0.82), 0.74 (95% CI 0.65-0.81) and 0.65 (95% CI 0.54-0.73), respectively. Additionally, the cutoff values for CysC, CRP, LDH, ferritin, D-dimer, creatinine, and WBC levels were 0.97 mg/L, 2.75 mg/dL, 196 U/L, 482 ng/mL, 0.76 mg/ml, 0.68 mmol/L, and 5700 /μl, respectively.
      Figure 2
      Figure 2Receiver operating characteristic curve for each biomarker pertaining to the emergence of hypoxia during the clinical course.
      Table 3Cutoff, area under the curve, and P values for each biomarker.
      Cutoff valueAUC (CI)P- value
      Cystatin C (mg/L)0.970.84 (0.75-0.90)<0.0001
      C-reactive protein (mg/dl)2.750.83 (0.75-0.89)<0.0001
      Lactate dehydrogenase (U/L)1960.82 (0.74-0.88)<0.0001
      Ferritin (ng/mL)4820.77 (0.69-0.84)<0.0001
      D-dimmer (mg/ml)0.760.75 (0.67-0.82)<0.0001
      Creatinine (mmol/L)0.680.74 (0.65-0.81)<0.0001
      White blood cell count (/μl)57000.65 (0.55-0.74)0.0063
      The sensitivity, specificity, positive predictive value, and negative predictive value of CysC were 0.69, 0.86, 0.58, and 0.90, respectively, while those for CRP were 0.75, 0.82, 0.57, and 0.91, respectively. Moreover, the sensitivity, specificity, positive predictive value, and negative predictive value when both CRP and CysC levels were considered concomitantly were calculated for 152 patients. When CRP ≥2.57 mg/dL and CysC level ≥0.97 mg/L, these values were 0.55, 0.95, 0.8, and 0.87, respectively, and when CRP <2.57 mg/dL and CysC <0.97 mg/L, they were 0.86, 0.74, 0.50, and 0.94, respectively. In other words, 80% cases exceeding the cutoff values for both biomarkers were in the hypoxia group, and 94% cases not exceeding the cutoff values were in the non-hypoxia group (Figure 3).
      Figure 3
      Figure 3Frequency of patients with hypoxia of COVID-19 according to serum CRP and CysC levels. Hypoxia group: the group in which oxygen treatment was required in patients because of hypoxia during the clinical course Non-hypoxia group: the group in which patients were discharged without hypoxia during the clinical course. COVID-19: Coronavirus disease 2019; CRP: C-reactive protein; CysC: Cystatin C
      Furthermore, we investigated the relationship between CRP and CysC levels and the number of days from disease onset to laboratory investigations. We categorized 39 patients in the hypoxia group with known dates of disease onset (for asymptomatic patients, the PCR positive date was considered the onset date) into 26 patients with CRP ≥2.75 mg/dL and 13 patients with CRP <2.75 mg/dL, and compared the number of days from onset to data collection. For CysC, we categorized 34 patients in hypoxia group with known onset dates into 25 patients with CysC ≥0.97 mg/L and nine patients with CysC <0.97 mg/L, and compared the number of days from onset to data collection (Supplementary Table 1). The median number of days was six for patients with CRP ≥2.75 mg/dL and three for patients with CRP <2.75 mg/dL; the difference was statistically significant (p = 0.0003). However, the median number of days was five for both patients with CysC ≥0.97 mg/L and patients with CysC <0.97 mg/L, and the difference was not statistically significant (p = 0.8586).

      Discussion

      Our study showed two important findings. First, Serum CysC and CRP levels in COVID-19 patients without hypoxia were associated with the emergence of hypoxia during the clinical course. Second, the cut-off values for serum cystatin C and CRP were 0.97 mg/L and 2.75 mg/dL, respectively. The strength of this study is that it excluded severe cases who required oxygen treatment and only validated the study in a group of patients without hypoxia. This may have clarified the criteria for recommending hospitalization, even for initially patients who are not hypoxic and whose symptoms are expected to improve naturally.
      Serum CysC has recently been determined a marker for evaluating renal function decline. It has high sensitivity and is not usually affected by age, sex, body weight, and inflammation.
      • Delanaye P
      • Cavalier E
      • Morel J
      • et al.
      Detection of decreased glomerular filtration rate in intensive care units: serum cystatin C versus serum creatinine.
      It is also considered a predictor of death from heart failure and cardiovascular disease.
      • Garcia CR
      • Vigil ML
      • Barquero PO
      • et al.
      Cystatin C as a predictor of cardiovascular outcomes in a hypertensive population.
      Serum CysC is a 13 kDa protein consisting of 122 amino acids and is a cysteine proteinase inhibitor.
      • Simonson O
      • Grubb A
      • Thysell H.
      The blood serum concentration of cystatin C (γ-trace) as a measure of the glomerular filtration rate.
      It suppresses the activation of cysteine protease, an enzyme used by the invading pathogens for proliferation, extracellularly.
      • Turk B.
      • Turk D.
      • Salvesen G.
      Regulating cysteine protease activity: essential role of protease inhibitors as guardians and regulators.
      An increase in serum CysC is also observed in human immunodeficiency virus infection,
      • Odden MC
      • Scherzer R
      • Bacchetti P
      • et al.
      Cystatin C level as a marker of kidney function in human immunodeficiency virus infection: the FRAM study.
      which may reflect the in-vivo reactions associated with viral infections. Furthermore, serum CysC level is significantly correlated with inflammatory cytokines IL-6 and Tumor necrosis factor (TNF)-α levels,
      • Okura T
      • Jotoku M
      • Irita J
      • et al.
      Association between cystatin C and inflammation in patients with essential hypertension.
      and cytokines from inflammatory cells regulate the expression of serum CysC.
      • Zi M
      • Xu Y.
      Involvement of cystatin C in immunity and apoptosis.
      COVID-19 is often exacerbated by the release of excess cytokines (“cytokine storm”), which can cause excessive inflammation and multiple organ failure.
      • Lucas C
      • Wong P
      • Klein J
      • et al.
      Longitudinal analyses reveal immunological misfiring in severe COVID-19.
      Therefore, inflammatory cytokines that cause cytokine storm are involved in the expression of serum CysC, as reflected in our research.
      In this study, elevated serum CRP levels were associated with the emergence of hypoxia in COVID-19 patients without hypoxia. Serum CRP levels have also been reported to be significantly associated with the increased possibility of ICU hospitalization, intubation, and mortality.
      • Ayanian S
      • Reyes J
      • Lynn L
      • Teufel K.
      The association between biomarkers and clinical outcomes in novel coronavirus pneumonia in a US cohort.
      Serum CRP level is largely reflected in the exacerbation of COVID-19, which may be due to its association with IL-6. It has been considered that the activation of IL-6 amplifier by COVID-19 specific angiotensin 2-angiotensin receptor type 1, which is a mechanism of IL-6 overproduction for cytokine storm induction, induces inflammatory cytokines.
      • Hirano T
      • Murakami M.
      COVID-19: A New Virus, but a Familiar Receptor and Cytokine Release Syndrome.
      Reflecting this, IL-6 is elevated in patients with severe COVID-19 symptoms,
      • Hadjadj J
      • Yatim N
      • Barnabei L
      • et al.
      Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients.
      and hence, tocilizumab, a neutralizing antibody against the IL-6 receptor, is now being used as treatment. Elevated IL-6 is significantly associated with the severity of COVID-19. Serum CRP is produced by the action of inflammatory cytokines such as TNF-α, IL-1, and IL-6 produced by macrophages, monocytes, and T cells by lipopolysaccharide stimulation due to infection.
      • Pepys MB
      • Hirschfield GM.
      C-reactive protein: a critical update.
      Therefore, an elevated serum CRP may mean an elevated IL-6 acting on it. In fact, CRP has been normalized in rheumatoid arthritis and Castleman's disease by treatment that inhibits IL-6.
      • Nishimoto N
      • Yoshizaki K
      • Miyasaka N
      • et al.
      Treatment of rheumatoid arthritis with humanized anti-interleukin-6 receptor antibody: a multicenter, double-blind, placebo-controlled trial.
      ,
      • Nishimoto N
      • Kanakura Y
      • Aozasa K
      • et al.
      Humanized anti-interleukin-6 receptor antibody treatment of multicentric Castleman disease.
      The greatest benefit of serum CRP over IL-6 is that it is a biomarker which can be easily and quickly measured at hospital laboratories. However, the problem is with CRP that it is not elevated in the early stages of infection, as also seen in our study (Supplementary Table 1), meaning that if laboratory investigations are performed exceedingly early in the course of the disease, hypoxia may be emerged, even if serum CRP is below the cutoff value.
      Unlike previous reports,
      • Williamson EJ
      • Walker AJ
      • Bhaskaran K
      • et al.
      Factors associated with COVID-19-related death using Open SAFELY.
      ,
      • Docherty AB
      • Harrison EM
      • Green CA
      • et al.
      Features of 20133 UK patients in hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study.
      the multivariate analysis showed no significant difference in need for oxygen treatment during the clinical course in men (p = 0.0835), which may be due to COPD and the exclusion of severe disease.
      Previous reports have shown that smoking itself is associated with death and hospitalization in COVID-19.
      • Neira DP
      • Watts A
      • Seashore J
      • et al.
      Smoking and risk of COVID-19 hospitalization.
      Moreover, COPD is associated with an increased risk of death from COVID-19 in men, and a five-fold increased risk of severe COVID-19 symptoms compared to those without any underlying disease.
      • Lippi G
      • Henry BM.
      Chronic obstructive pulmonary disease is associated with severe coronavirus disease 2019 (COVID-19).
      ,
      • Zhao Q
      • Meng M
      • Kumar R
      • et al.
      The impact of COPD and smoking history on the severity of COVID-19: a systemic review and meta-analysis.
      However, studies that exclude patients with severe symptoms, such as this study, may exclude COPD patients, which could result in lesser impact of sex. In fact, the proportion of men and COPD patients in this study was even lower than in studies including patients with severe symptoms.
      • Cummings MJ
      • Baldwin MR
      • Abrams D
      • et al.
      Epidemiology, clinical course, and outcomes of critically ill adults with COVID-19 in New York City: a prospective cohort study.
      Unfortunately, there was no significant difference in Ct values between the two groups in our study. Several previous reports have shown a correlation between Ct values in RT-PCR of COVID-19 and mortality.
      • Huang JT
      • Ran RX
      • Lv ZH
      • et al.
      Chronological changes of viral shedding in adult inpatients with COVID-19 in Wuhan, China.
      ,
      • Liu Y
      • Yan LM
      • Wan L
      • et al.
      Viral dynamics in mild and severe cases of COVID-19.
      ,
      • Shi F
      • Wu T
      • Zhu X
      • et al.
      Association of viral load with serum biomakers among COVID-19 cases.
      However, the association of the Ct value with disease severity is controversial, with variable opinions. The lack of significant differences in this study may be attributed to the fact that there were few patients with severe symptoms, such as those who required oxygen significantly and those who died. Moreover, the number of patients who underwent RT-PCR within one week after the onset was small, necessitating further research into the utilization of Ct value.
      Our study has several limitations. First, the severity of COVID-19 is affected by several host factors, such as gene mutations and race.
      • Bhala N
      • Curry G
      • Martineau AR
      • et al.
      Sharpening the global focus on ethnicity and race in the time of COVID-19.
      ,
      • Yamamoto N
      • Bauer G.
      Apparent difference in fatalities between Central Europe and East Asia due to SARS-COV-2 and COVID-19: Four hypotheses for possible explanation.
      As approximately 5% of patients in this study population were of non-Japanese descent and ethnic differences were not considered in the analysis, further research is required. Second, a variant of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) from the UK was reported in December 2020,
      • Kirby T.
      New variant of SARS-CoV-2 in UK causes surge of COVID-19.
      which may have been partially included in this study. The recent SARS-CoV-2 variant is associated with a 1.64 times higher risk of death than the original strain.
      • Challen R
      • Brooks-Pollock E
      • Read JM
      • et al.
      Risk of mortality in patients infected with SARS-CoV-2 variant of concern 202012/1: matched cohort study.
      Therefore, it is necessary to evaluate variant-specific effects in the future. Third, serum CysC and CRP levels were included in the multivariate logistic regression analysis to identify biomarkers, but changes in concentration over time were not evaluated. The levels of these markers can change significantly over time, and their effects should be further investigated.

      Conclusions

      In conclusion, our findings showed that high serum CysC and CRP in COVID-19 patients without hypoxia was associated with the emergence of hypoxia during the clinical course. The cutoff values were 0.97 mg/L for serum CysC and 2.57 mg/dL for serum CRP. These findings may help in determining the need for hospitalization of COVID-19 patients without hypoxia.

      Authors Contributions statement

      Y.M. designed the study, wrote and revised the paper; H I wrote and revised the paper; K.H. collected and analysed the data; F.I., S.O., H.S., K.M., T.E., H.M., T.K., Y.F., Y.K., T.H., H.O., S.K., M.Y., S.S., Y.U., A.T., K.I. and Y.K. collected data; H.S. supervised and reviewed the manuscript; all authors provided critical review of the manuscript and approved the final draft for publication.

      Source of Funding Statement

      This research was supported by the GSK Japan Research Grant 2021 to HI.

      Conflict of Interest Statement

      The authors declare no conflicts of interest associated with this manuscript.

      Acknowledgments

      We would like to thank all the members who participated in this study.

      References

        • Johns Hopkins University and Medicine
        COVID-19 map.
        Johns Hopkins Coronavirus Resource Centre, 2021 (accessed April 22)
        • Karako K
        • Song P
        • Chen Y
        • et al.
        Overview of the characteristics of and responses to the three waves of COVID-19 in Japan during 2020-2021.
        Biosci Trends. 2021; 15: 1-8
        • Nakamura M
        • Hitosugi M.
        Emerging COVID-19-related deaths during home recuperation in Japan.
        Geriatr Gerontol Int. 2021; https://doi.org/10.1111/ggi.14143
        • Ishii M
        • Terai H
        • Kabata H
        • et al.
        Clinical characteristics of 345 patients with coronavirus disease 2019 in Japan: a multicenter retrospective study.
        J Infect. 2020; 81: e3-e5
        • Cummings MJ
        • Baldwin MR
        • Abrams D
        • et al.
        Epidemiology, clinical course, and outcomes of critically ill adults with COVID-19 in New York City: a prospective cohort study.
        Lancet. 2020; 395: 1763-1770
        • Sethuraman N
        • Jeremiah SS
        • Ryo A.
        Interpreting diagnostic tests for SARS-CoV-2.
        JAMA. 2020; 323: 2249-2251
        • Delanaye P
        • Cavalier E
        • Morel J
        • et al.
        Detection of decreased glomerular filtration rate in intensive care units: serum cystatin C versus serum creatinine.
        BMC Nephrol. 2014; 15: 9
        • Garcia CR
        • Vigil ML
        • Barquero PO
        • et al.
        Cystatin C as a predictor of cardiovascular outcomes in a hypertensive population.
        J Hum Hypertens. 2017; 31: 801-807
        • Simonson O
        • Grubb A
        • Thysell H.
        The blood serum concentration of cystatin C (γ-trace) as a measure of the glomerular filtration rate.
        Scand J Clin Invest. 1985; 45: 97-101
        • Turk B.
        • Turk D.
        • Salvesen G.
        Regulating cysteine protease activity: essential role of protease inhibitors as guardians and regulators.
        Curr Pharm Des. 2002; 8: 1623-1637
        • Odden MC
        • Scherzer R
        • Bacchetti P
        • et al.
        Cystatin C level as a marker of kidney function in human immunodeficiency virus infection: the FRAM study.
        Arch Intern Med. 2007; 167: 2213-2219
        • Okura T
        • Jotoku M
        • Irita J
        • et al.
        Association between cystatin C and inflammation in patients with essential hypertension.
        Clin Exp Nephrol. 2010; 14: 584-588
        • Zi M
        • Xu Y.
        Involvement of cystatin C in immunity and apoptosis.
        Immunol Lett. 2018; 196: 80-90
        • Lucas C
        • Wong P
        • Klein J
        • et al.
        Longitudinal analyses reveal immunological misfiring in severe COVID-19.
        Nature. 2020; 584: 463-469
        • Ayanian S
        • Reyes J
        • Lynn L
        • Teufel K.
        The association between biomarkers and clinical outcomes in novel coronavirus pneumonia in a US cohort.
        Biomark Med. 2020; 14: 1091-1097
        • Hirano T
        • Murakami M.
        COVID-19: A New Virus, but a Familiar Receptor and Cytokine Release Syndrome.
        Immunity. 2020; 52: 731-733
        • Hadjadj J
        • Yatim N
        • Barnabei L
        • et al.
        Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients.
        Science. 2020; 369: 718-724
        • Pepys MB
        • Hirschfield GM.
        C-reactive protein: a critical update.
        J Clin Invest. 2003; 111: 1805-1812
        • Nishimoto N
        • Yoshizaki K
        • Miyasaka N
        • et al.
        Treatment of rheumatoid arthritis with humanized anti-interleukin-6 receptor antibody: a multicenter, double-blind, placebo-controlled trial.
        Arthritis Rheum. 2004; 50: 1761-1769
        • Nishimoto N
        • Kanakura Y
        • Aozasa K
        • et al.
        Humanized anti-interleukin-6 receptor antibody treatment of multicentric Castleman disease.
        Blood. 2005; 106: 2627-2632
        • Williamson EJ
        • Walker AJ
        • Bhaskaran K
        • et al.
        Factors associated with COVID-19-related death using Open SAFELY.
        Nature. 2020; 584: 430-436
        • Docherty AB
        • Harrison EM
        • Green CA
        • et al.
        Features of 20133 UK patients in hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study.
        BMJ. 2020; 369: m1985
        • Neira DP
        • Watts A
        • Seashore J
        • et al.
        Smoking and risk of COVID-19 hospitalization.
        Respir Med. 2021; 182106414
        • Lippi G
        • Henry BM.
        Chronic obstructive pulmonary disease is associated with severe coronavirus disease 2019 (COVID-19).
        Respir Med. 2020; 167105941
        • Zhao Q
        • Meng M
        • Kumar R
        • et al.
        The impact of COPD and smoking history on the severity of COVID-19: a systemic review and meta-analysis.
        J Med Virol. 2020; 92: 1915-1921
        • Huang JT
        • Ran RX
        • Lv ZH
        • et al.
        Chronological changes of viral shedding in adult inpatients with COVID-19 in Wuhan, China.
        Clin Infect Dis. 2020; 71: 2158-2166
        • Liu Y
        • Yan LM
        • Wan L
        • et al.
        Viral dynamics in mild and severe cases of COVID-19.
        Lancet Infect Dis. 2020; 20: 656-657
        • Shi F
        • Wu T
        • Zhu X
        • et al.
        Association of viral load with serum biomakers among COVID-19 cases.
        Virology. 2020; 546: 122-126
        • Bhala N
        • Curry G
        • Martineau AR
        • et al.
        Sharpening the global focus on ethnicity and race in the time of COVID-19.
        Lancet. 2020; 395: 1673-1676
        • Yamamoto N
        • Bauer G.
        Apparent difference in fatalities between Central Europe and East Asia due to SARS-COV-2 and COVID-19: Four hypotheses for possible explanation.
        Med Hypotheses. 2020; 144110160
        • Kirby T.
        New variant of SARS-CoV-2 in UK causes surge of COVID-19.
        Lancet Respir Med. 2021; 9: e20-e21
        • Challen R
        • Brooks-Pollock E
        • Read JM
        • et al.
        Risk of mortality in patients infected with SARS-CoV-2 variant of concern 202012/1: matched cohort study.
        BMJ. 2021; 372: n579