RESEARCH ARTICLE


https://doi.org/10.5005/jp-journals-10028-1621
Journal of Postgraduate Medicine, Education and Research
Volume 57 | Issue 3 | Year 2023

Predictive Model for the Assessment of COVID-19 Severity based on Acute Phase Proteins: An Analysis from a Clinical Laboratory in North India


Ram K Saini1, Neha Saini2, Sant Ram3, Ankita Goyal4, Shiv Soni5, Vikas Suri6, Ravjit Jassal7, Arnab Pal8, Deepy Zohmangaihi9https://orcid.org/0000-0002-5054-4001

1-4,7-9Department of Biochemistry, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

5Department of Anaesthesiology and Intensive Care, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

6Department of Internal Medicine, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

Corresponding Author: Deepy Zohmangaihi, Department of Biochemistry, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India, Phones: +91 8414863196, +91 7087008274, e-mail: drdeepyz14@gmail.com

Received on: 30 November 2022; Accepted on: 12 April 2023; Published on: 26 July 2023

ABSTRACT

Aim: Pestilence owing to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus in recent times and its ongoing resurgence poses an urgent need for the identification of a predictive marker for determining coronavirus disease 2019 (COVID-19) severity. Considering inflammation as the fundamental cause of morbidity following COVID-19 infections, our study targeted acute phase proteins (APPs), potential markers reflecting the inflammatory state of the body.

Materials and methods: We retrospectively analyzed the serum levels of seven APPs, namely, α-1 antitrypsin (AAT), α-1 acid glycoprotein (AGP), ceruloplasmin (Cp) and C-reactive protein (CRP), transferrin (Tf), transthyretin (TTR), and human serum albumin (HSA) in 788 COVID-19 patients. Multivariate regression analysis and receiver operating characteristic (ROC) analysis were performed to develop a mathematical model for predicting COVID-19 severity.

Results: Out of the seven APPs, levels of CRP (p < 0.0001) and HSA (p < 0.0001) were found to have significant associations with the incidence of intensive care unit (ICU) admission in COVID-19 patients, as analyzed primarily in 219 patients. A mathematical model was designed with ROC displaying sensitivity and specificity of 88.89 and 66.67% with 0.802 [95% confidence of interval (CI), 0.4408–0.6508] area under the curve (AUC). The logistic regression equation thus developed was validated in a different cohort of 569 COVID-19 patients where probabilities for ICU admission calculated on the 1st day of admission were higher (p < 0.0001) in patients actually admitted to ICU at any time point during their hospital stay.

Conclusion: The model appears fit for predicting the severity of COVID-19 patients at its early stages. Utilization of this model is feasible and practical as severity is decided based on a few biochemical parameters.

How to cite this article: Saini RK, Saini N, Ram S, et al. Predictive Model for the Assessment of COVID-19 Severity based on Acute Phase Proteins: An Analysis from a Clinical Laboratory in North India. J Postgrad Med Edu Res 2023;57(3):117-123.

Source of support: Nil

Conflict of interest: Dr Vikas Suri is associated as the National Editorial Board member of this journal and this manuscript was subjected to this journal’s standard review procedures, with this peer review handled independently of this editorial board member and his research group.

Keywords: Acute phase proteins, Coronavirus disease 2019, Model, Prediction, Severity

INTRODUCTION

The COVID-19 pandemic, triggered by SARS-CoV-2, has severely impacted millions of people worldwide. According to World Health Organization, nearly 637 million people have been infected, with approximately 6 million deaths as of November 28, 2022.1 India had recorded 44 million confirmed cases with a mortality rate of 1.2% as of November 2022,2 which has severely affected the healthcare system and economy of the country. It was initially observed that among the people infected with COVID-19, few required hospitalization, while the rest recovered with mild symptoms. Elderly individuals, patients with obesity, and previous medical history of diabetes mellitus, hypertension, cancer, respiratory disorders, and other chronic diseases were more prone to develop severe manifestations of the disease, leading to intensive care treatment and mortality.3 In spite of the recoveries in the majority of the COVID-19 patients and the mortality rate being relatively low, the recurring disease occurrence urges a dire need for the prediction of SARS-CoV-2 severity at an early stage. Consequently, some predictive models have been designed to assess the severity of COVID-19 in recent times. Some of these are based upon the biochemical and hematogical parameters of the patients with severe and nonsevere consequences,4,5 while others include the findings of chest computed tomography (CT) scans to predict the outcomes of the disease.6 Employment of machine learning algorithms makes clinical data more conclusive, especially for novel diseases like COVID-19, where outcomes are unpredictable. Applications of these models in hospital settings rely upon the achieved accuracy along with a rapid prediction for early clinical decision-making.

As stated earlier, in COVID-19-positive patients, no severe clinical manifestations are seen at early stages, with many presenting with only mild fever, cough, or muscle soreness. However, clinical studies have shown that in critical patients, with the host immune response to the SARS-CoV-2 virus being hyperactive, there is an excessive and aggressive inflammatory response with the release of a large number of proinflammatory cytokines in an event known as “cytokine storm.” Unfavorable prognosis of severe COVID-19, acute respiratory distress syndrome, and multi-organ failure resulting in death within a short time was found to be correlated directly with this “cytokine storm,”7,8 thereby proving its important role in the process of disease aggravation.

In the same context, proteins that are expressed in the acute phase are potential biomarkers for the diagnosis of inflammatory disease. APPs are primarily synthesized in hepatocytes. The acute phase response is a spontaneous reaction triggered by disrupted homeostasis resulting from environmental disturbances.9 Acute phase reactions (APRs) usually stabilize quickly after recovering from a disruption to homeostasis within a few days to weeks; however, APPs expression levels often remain elevated in long-lasting infections and chronic disease states,10,11 thereby making them potential markers of inflammatory conditions.

There are eight proteins that are over-expressed in APRs denoted as “positive” APPs, such as haptoglobin, serum amyloid, fibrinogen, Cp, AGP, AAT, lactoferrin, and CRP. Similarly, there are a number of “negative” APPs, the expression levels of which are reduced, which include albumin (ALB), Tf, and TTR.12 The APPs are elicited by cytokines possessing either proinflammatory or anti-inflammatory activity. Acute-phase proteins are, therefore, reactive proteins, which are directly manifested as part of the body’s inflammatory response.

Currently, there is no systematic analysis regarding the characteristics of APPs in COVID-19. In this paper, we evaluated and compared the levels of positive APPs, namely, AAT, AGP, Cp, and CRP, and negative acute phase reactants, Tf, TTR, and HSA in the serum of the severe COVID-19-positive patients requiring ICU treatment and those not severely affected and not admitted in the ICU (non-ICU). The aim of our study was the prompt recognition of high-risk COVID-19 patients at early stages through a model based on the levels of these APPs. Our findings, we speculate, may provide references for the prognosis and prediction of severity and, subsequently, effective treatment and control of COVID-19. In the present scenario, with the resurgence of infections and identification of mutant strains, a model for rapidly assessing the severity of COVID-19 is critical and paramount.

MATERIALS AND METHODS

Study Design

This study included the data of 788 reverse transcription polymerase chain reaction (RT-PCR) confirmed COVID-19 patients admitted to the Nehru Extension Block, the designated COVID-19 Care Center at the Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India. Of these 788 patients, the development of the predictive model was based on the data obtained from 219 patients, while data from another 569 patients were employed for the validation of the predicted model. Since the disease is caused by a novel SARS-CoV-2 virus, displaying different statistics in subsequent peaks, its true prevalence is not known for the estimation of the sample size for the study. We, therefore, have included all the adult patients of both genders admitted to the hospital between May and December 2020. Samples were collected from the patients as part of routine care and were not obtained solely for the purpose of research. The study plan was approved by the Institutional Ethics Committee, Postgraduate Institute of Medical Education and Research, Chandigarh, India (Ref No-NK/6384/study/231). Since it is a retrospective analysis, the Ethics Committee granted permission to publish the retrospectively collected data and also allowed a waiver for the consent of the patients participating in the study.

Sample collection and laboratory analysis of APPs—venous blood (3 mL) from the patients drawn in red top vacutainers under aseptic conditions wearing personal protective equipment was transferred to the clinical biochemistry laboratory as a routine patient care practice. After centrifugation at 3000 g for 10 minutes, separated serum was used for the analysis of the levels of APPs, namely, AAT, AGP, Cp, Tf, and TTR through “nephelometry” in IMMAGE 800, immunochemistry system, Beckman Coulter autoanalyzer. CRP and HSA were analyzed through “immunoturbidimetry” and “colorimetry” methods, respectively, in Roche Cobas 8000 autoanalyzer. Normal limits for the different APPs were defined as AAT, 88–174 mg/dL; AAG, 51–117 mg/dL; Cp, 20–58 mg/dL; CRP, 0–5 mg/L, Tf, 202–336 mg/dL; TTR, 18–38 mg/dL; HSA, 3.4–4.8 gm/dL.

Development of Model

For the construction of a predictive model for COVID-19 severity, we included the data of serum APPs levels in 219 patients with RT-PCR-confirmed COVID-19 status admitted between May and July 2020. Multivariate regression analysis was used to design the mathematical model. The accuracy of the logistic regression equation was confirmed through ROC curve analysis.

Validation of the Model

Further validation of the predictive model was carried out using the data of serum APPs levels in a different cohort of 569 COVID-19 patients admitted between September and December 2020, where probabilities for ICU and non-ICU admission were compared using the developed model.

Statistical Analysis

Trial versions of Prism 8.0 and MedCalc-19.6.1 were used for statistical analysis. The normality of the proteins was analyzed through “KolmogorovSmirnov” test. Comparison between the levels of APPs in different groups was made through the Mann–Whitney test. Multivariate regression analysis was done through enter, backward, forward, and stepwise methods using MedCalc. Coefficients of the regression equation thus obtained were used to design the mathematical model and to calculate the logit (p) values.

RESULTS

Descriptive analysis of APPs levels in the study population

Initially, serum analysis of APPs was carried out in 219 recruited patients, 88 (40%) of which required ICU treatment at any of the time points during their hospitalization. Among the positive APPs, the levels of AAT, AGP, and CRP were found to be significantly elevated (p < 0.0001) in ICU patients in comparison to the non-ICU group (Fig. S1A), with median values for CRP, 67.97 mg/L, and 14 mg/L respectively (Table 1). However, no significant difference was observed in Cp levels in these two groups (Fig. S1A). Among the negative APPs, the levels of Tf, HSA, and TTR were found to be significantly lower (p < 0.0001) in ICU patients in comparison to non-ICU patients (Fig. S1B) with substantial differences in median values (Table 1). Abnormal levels of APPs, that is, above the upper limit of normal for the positive and below the lower limit of normal for negative APPs, were compared in ICU and non-ICU patients (Figs 4A to D), displaying the significant difference in the levels of AAT (p = 0.004), AGP (p = 0.0008), CRP (p < 0.0001), and TTR (p = 0.0123). Out of the 219 patients primarily recruited in the study, 132 were males, and 87 were females. Levels of APPs did not display any significant difference in males and females except for AGP (p = 0.0112), CRP (p < 0.0001), and Tf (p < 0.0001) (Figs 1A and B).

Table 1: Descriptive analysis of the levels of APPs in the study population (n = 219)
Minimum 25% percentile Median 75% percentile Maximum Range
AATT (mg/dL) 55 155 200 297 593 538
AATnon-ICU 55 149.5 177.5 246.8 562 507
AATICU 82 182.3 259.5 341.3 593 511
AGPT (mg/dL) 11 99 143 226 442 431
AGPnon-ICU 11 91.50 120 183 402 391
AGPICU 43 123 200 248.8 442 399
CpT (mg/dL) 12 35 41 53.50 334 322
Cpnon-ICU 14 35 40 49 133 119
CpICU 12 35 46 55 334 322
CRPT (mg/L) 0.20 4 28.10 79.67 392.30 392.10
CRPnon-ICU 0.20 2.16 14 47.61 235.10 234.90
CRPICU 0.48 30 67.97 138.2 392.3 391.80
TfT (mg/dL) 23.90 144.3 204 275.3 877 853.10
Tfnon-ICU 23.90 175 249.50 304 877 853.10
TfICU 42 129 163 209 492 450
TTRT (mg/dL) 2.97 11 17 22.25 193 190
TTRnon-ICU 2.90 16 20 24 79 76.03
TTRICU 3 8.07 12 16 193 190
HSAT (gm/dL) 1.60 3.30 3.87 4.44 5.62 4.02
HSAnon-ICU 1.60 3.70 4.30 4.60 5.62 4.02
HSAICU 1.70 2.80 3.40 3.70 4.86 3.16

T, Non-ICU, ICU representing the values for total patients, patients with non-ICU and ICU outcomes, respectively

Figs 1A and B: (A) Figure showing the comparison between the levels of positive and negative; (B) APPs in males and females in the study population (n = 219). (M and F indicate the levels in males and females, respectively)

Fig. 2: ROC curve for multivariate regression analysis (stepwise) for the built mathematical model (n = 219)

Figs 3A and B: (A) Histogram showing the distribution of COVID-19 patients admitted in ICU and non-ICU compared to the percent probability for ICU admission calculated using the regression model on 1st day of their hospital admission; (B) comparison between the probability of ICU admission using regression model in COVID-19 patients actually admitted in ICU and non-ICU (n = 569)

Figs 4A to D: Figures showing the total levels (A and B) and abnormal levels; (C and D) of positive and negative APPs in non-ICU and ICU patients

Logistic Regression Analysis

Stepwise logistic regression analysis incorporating the data for all APPs displays significant p-values only for CRP (0.0045) and HSA (0.0002) (Table 2), based on which the following mathematical model for predicting ICU requirement in COVID-19 patients was designed:

Table 2: Multivariate regression statistics
Overall model fit
Null model—2 log-likelihood 271.030
Full—2 log-likelihood 219.351
Chi-squared 51.679
DF 2
Significance level p < 0.0001
Cox and Snell R2 0.2267
Nagelkerke R2 0.3062
Regression coefficients and Wald test along with p-values
Variable Coefficient Standard Error Wald p
Albumin −0.85965 0.22840 14.1659 0.0002
CRP 0.0081705 0.0028730 8.0876 0.0045
Constant 2.28878 0.93439 6.0000 0.0143
Odds Ratios and 95% CI
Variable Odds ratio 95% CI
Albumin 0.4233 0.2705–0.6623
CRP 1.0082 1.0025–1.0139
Hosmer–Lemeshow test
Chi-squared 23.3483
DF 8
Significance level p = 0.0029

DF, Degree of freedom

Logit (p) = 2.28878 + 0.00817 (CRP)−0.85965 (HSA)

The estimated model forms a ROC curve with an area under curve 0.802 (p < 0.0001, Fig. 2). Youden index was 0.556 (95% CI, 0.4408–0.6508) with sensitivity and specificity of 88.89 and 66.67%, respectively. This model was found to be the best among the other models obtained through enter, forward, and backward regression analysis (Table S1).

Table S1: A comparative analysis of different regression models incorporating the APPs data of the study population (n = 219)
Enter Forward Backward Stepwise
AUC 0.806 0.802 0.806 0.802
95% CI 0.746–0.866 0.740–0.864 0.746–0.865 0.740–0.864
Youden index 0.5324 0.5556 0.5025 0.556
Specificity 79.17 66.67 70.00 66.67
Sensitivity 74.07 88.89 80.25 88.89
Significance in validations (p) 0.0007 <0.0001 0.0003 <0.0001

Validation of Mathematical ModelFurther validation of the mathematical model was done using the data of serum APPs levels in a different cohort of 569 COVID-19 patients. Using the above model, the probabilities for ICU admission were significantly higher (p < 0.0001) in the patients actually admitted in ICU (Figs 3A and B) as compared to the patients without ICU requirement indicating prodigious accuracy of the designed model.

DISCUSSION

The ongoing COVID-19 pandemic has beleaguered the worldwide healthcare system, with more than 6 million people succumbing to it,1 and it is still continuing to affect people globally. Supportive therapies, besides others, include the use of dexamethasone and other corticosteroids, as the severity of the disease has been reported to be linked with inflammatory responses.7 Proinflammatory cytokines released during infections induce the secretion of APPs that may display different functions, such as protease inhibitors (α1-antitrypsin), antioxidants (Cp and Tf), or scavengers (CRP) to eliminate the foreign bodies. As a result of the enormous production of these positive APPs, the circulatory concentration of negative APPs decreases in order to maintain the osmotic pressure.13

The defensive role of APPs has been used as a diagnostic tool for many diseases in recent times, ranging from bacterial infections to carcinoma.14,15 Although the alterations in their levels upon COVID-19 infections have been reported in a few studies,16,17 the knowledge regarding precise levels of these APPs and their association with the severity of the disease is lacking.

In our study carried out in the tertiary care center of PGIMER, Chandigarh, India, we primarily included the data of serum APPs levels in 219 patients with RT-PCR-confirmed COVID-19 status. A total of 88 (40%) of them required ICU treatment at any time point since their admission. Taking into account that inflammations are the key indicators of COVID-19 severity, we analyzed the levels of both positive and negative APPs in the serum samples of these patients. We found significantly raised levels of AAT, CRP, and AGP in ICU patients in comparison to the patients with non-ICU status. CRP is a nonspecific acute phase reactant secreted primarily by hepatocytes as a host response toward infections and inflammations. In concomitance with the recent studies showing the raised levels of CRP in COVID-19 patients,16,17 this study shows a nearly fivefold increase in the median values of CRP in ICU patients as compared to non-ICU. This finding is quite obvious as raised levels of proinflammatory cytokines, predominantly interleukin (IL-6) in severe COVID-19 patients, are the major inducer for CRP secretion along with IL-1 and tumor necrosis factor-α.20 Significantly raised levels of CRP in ICU patients (p < 0.0001) in spite of its rapid decline rate with a half-life of 19 hours21 makes it, although not specific but a suitable marker for the prognosis of COVID-19 and determining the severity in terms of ICU requirement. Significantly increased levels (nearly twofold, p < 0.0001) of AGP in severe COVID-19 patients can be correlated with studies showing a two to sixfold rise in the levels of AGP upon infections and inflammations.22

α-1 antitrypsin (AAT), being a protease inhibitor, is responsible for maintaining the levels of neutrophil elastase in the lung parenchyma. It has been observed recently that the deficiency of AAT protein is more commonly found in Europe and America in comparison to some Asian populations.23 In our study also, significantly increased levels of AAT (p < 0.0001) were found in the ICU patients. Despite the anti-inflammatory and antioxidant roles,24 Cp levels were not significantly increased either in ICU or non-ICU patients in our study. Although increased ferritin levels have been reported in COVID-19 patients and Cp has a role in iron trafficking.25

Among the negative APPs, significantly reduced values of HSA have been reported as an independent risk factor for COVID-19-induced morbidity.26 We have also found significantly reduced levels of HSA in ICU patients (p < 0.0001) in comparison to non-ICU patients. Such conditions of hypoalbuminemia in severe COVID-19 patients may be attributed to the reduced synthesis of ALB due to inflammations, liver damage, and reduced food intake or in order to maintain the osmotic pressure in response to the elevated levels of positive APPs. The levels of Tf and TTR were also found to be significantly reduced (p < 0.0001) in the ICU patients of our study, which may be attributed to multiple factors as reported in other studies.27

Multivariate logistic regression analysis indicated CRP and HSA as potential indicators for determining the severity of COVID-19 and ICU requirements. Values for these parameters can be used to determine the probabilities for ICU admission through the mathematical equation designed in this study. This mathematical model was further validated through ROC analysis displaying key parameters within the acceptable range. The accuracy of this model was further confirmed using the data on serum APPs levels in a different cohort of 569 patients.

Despite the ample sample size and significant findings, our study has a few limitations. Firstly, this study is a single center-based study, which has been carried out in the tertiary care center of PGIMER, Chandigarh, India, that serves as a referral center for the majority of north Indian states. Secondly, among the several existing APPs, we could analyze the levels of only seven, yet others might also have a role in determining the severity that could have improvised the mathematical model. Also, the clinical and medication history of the patients was not available.

Since the development of our predictive model was based on the retrospective data obtained from the COVID-19 patients, it might have certain limitations, including the absence of potential confounders, thus making it difficult to identify the appropriate comparison groups, leading to outcome bias. Also, the associations between disease outcomes and risk factors often change with time. But, considering COVID-19, a novel and fatal disease, follow-up of the critically ill patients was not possible. Therefore, we found this observational retrospective approach suitable for the prediction of severity, thus leading to the better management of COVID-19-positive patients.

CONCLUSION

We have designed and validated a mathematical equation for defining the ICU requirement for COVID-19 patients at the early stages. This equation could also be applicable to other disease conditions requiring ICU admissions, as both CRP and HSA are not entirely COVID-19-specific parameters. Further studies are required in this direction to examine the efficacy of this mathematical model in other health conditions as well as design a superior model including other potential APPs.

Clinical Significance

  • Predictive biomarkers for assessment of COVID-19 severity at an early stage may help in better management of the patients.

  • Increased levels of inflammatory cytokines have been the hallmarks of morbidity and mortality in COVID-19 patients.

  • Alterations in the level of APPs were observed in COVID-19 patients admitted to ICU.

  • Regression analysis revealed significance in the levels of ALB and CRP. A mathematical model with the acceptable AUC, specificity, and sensitivity was developed for predicting ICU admission.

  • Furthermore, the accuracy of the model was validated in a different cohort of COVID-19-positive patients and found fit.

  • Application of this model is feasible and practical as the prediction of severity is based on the levels of routine biochemical parameters such as ALB and CRP.

AUTHORS’ CONTRIBUTION

Ram Krishan Saini and Neha Saini did data analysis, statistical analysis, and graphs preparation and drafted the manuscript. Sant Ram and Arnab Pal participated in the revision and editing of the manuscript. Shiv Soni and Vikas Suri provided the clinical samples for the study. Ankita Goyal and Ravjit Jassal participated in data acquisition. Deepy Zohmangaihi conceived, designed, and analyzed the study and created the final version of the manuscript. All authors approved the final manuscript.

ORCID

Deepy Zohmangaihi https://orcid.org/0000-0002-5054-4001

ACKNOWLEDGMENT

We thank Professor Sadhna Sharma, Professor Indu Verma and the entire faculty and technical staff of the Biochemistry Department, PGIMER, Chandigarh, India, for their contributions.

REFERENCES

1. World Health Organization, 2022. Corona. Coronavirus disease (COVID-19) pandemic. Available from: https://covid19.who.int/ [Accessed 28 November 2022].

2. Hopkins John, 2022. Coronavirus Resource Center. Mortality Analysis. Available form: https://coronavirus.jhu.edu/data/mortality [Accessed 28 November 2022].

3. Liu D, Cui P, Zeng S, et al. Risk factors for developing into critical COVID-19 patients in Wuhan, China: a multicenter, retrospective, cohort study. EClinicalMedicine 2020;25:100471. DOI: 10.1016/j.eclinm.2020.100471

4. Meng Z, Wang M, Zhao Z, et al. Development and validation of a predictive model for severe COVID-19: a case-control study in China. Front Med (Lausanne) 2021;8:663145. DOI: 10.3389/fmed.2021.663145

5. Zeng F, Deng G, Cui Y, et al. A predictive model for the severity of COVID-19 in elderly patients. Aging (Albany NY) 2020;12(21):20982–20996. DOI: 10.18632/aging.103980

6. Kitamura FC, Prado GF, Kuriki PEA. Machine learning model for predicting severity prognosis in patients infected with COVID-19: study protocol from COVID-AI Brasil. PLoS One 2021;16(2):e0245384. DOI: 10.1371/journal.pone.0245384

7. Huang C, Wang Y, Li X, et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet 2020;395(10223):497–506. DOI: 10.1016/S0140-6736(20)30183-5

8. Ruan Q, Yang K, Wang W, et al. Clinical predictors of mortality due to COVID-19 based on an analysis of data of 150 patients from Wuhan, China. Intensive Care Med 2020;46(5):846–848. DOI: 10.1007/s00134-020-05991-x

9. Cray C, Zaias J, Altman NH. Acute phase response in animals: a review. Comp Med 2009;59(6):517–526.

10. Mackiewicz A. Acute phase proteins and transformed cells. Int Rev Cytol 1997;170:225–300. DOI: 10.1016/s0074-7696(08)61623-x

11. Koj A. Termination of acute–phase response: role of some cytokines and anti–inflammatory drugs. General Pharmacol 1998;31(1):9–18. DOI: 10.1016/s0306-3623(97)00435-7

12. Carrasco L, Salguero FJ, Barranco I, et al. Acute phase proteins as biomarkers in animal health and welfare. Acute phase proteins as early non-specific biomarkers of human and veterinary diseases 2011;10:259–280. DOI: 10.5772/19166

13. Lakota K, Buzan O, Tomšič M, et al. Acute Phase Proteins in Prototype Rheumatic Inflammatory Diseases. Acute Phase Proteins: Regulation and Functions of Acute Phase Proteins. IntechOpen, 303. DOI: 10.5772/19166

14. Charlie-Silva I, Klein A, Gomes JMM, et al. Acute-phase proteins during inflammatory reaction by bacterial infection: Fish-model. Sci Rep 2019;9(1):4776. DOI: 10.1038/s41598-019-41312-z

15. Chen Y, Lim BK, Hashim OH. Different altered stage correlative expression of high abundance acute-phase proteins in sera of patients with epithelial ovarian carcinoma. J Hematol Oncol 2009;2:37. DOI: 10.1186/1756-8722-2-37

16. Skevaki C, Fragkou PC, Cheng C, et al. Laboratory characteristics of patients infected with the novel SARS-CoV-2 virus. J Infect 2020;81(2):205–212. DOI: 10.1016/j.jinf.2020.06.039

17. Li L, Chen C. Contribution of acute-phase reaction proteins to the diagnosis and treatment of 2019 novel coronavirus disease (COVID-19). Epidemiol Infect 2020;148:e164. DOI: 10.1017/S095026882000165X

18. Saini RK, Saini N, Ram S, et al. COVID-19 associated variations in liver function parameters: a retrospective study. Postgrad Med J 2022;98(1156):91–97. DOI: 10.1136/postgradmedj-2020-138930

19. Sharifpour M, Rangaraju S, Liu M, et al. C-Reactive protein as a prognostic indicator in hospitalized patients with COVID-19. PLoS One 2020;15(11):e0242400. DOI: 10.1371/journal.pone.0242400

20. Zhang D, Sun M, Samols D, et al. STAT3 participates in transcriptional activation of the C-reactive protein gene by interleukin-6. J Biol Chem 1996;271(16):9503–9509. DOI: 10.1074/jbc.271.16.9503

21. Vigushin DM, Pepys MB, Hawkins PN. Metabolic and scintigraphic studies of radioiodinated human C-reactive protein in health and disease. J Clin Invest 1993;91(4):1351–1357. DOI: 10.1172/JCI116336

22. Kremer JM, Wilting J, Janssen LH. Drug binding to human alpha-1-acid glycoprotein in health and disease. Pharmacol Rev 1988;40(1):1–47.

23. Bhattacharyya C, Das C, Ghosh A, et al. SARS-CoV-2 mutation 614G creates an elastase cleavage site enhancing its spread in high AAT-deficient regions. Infect Genet Evol 2021;90:104760. DOI: 10.1016/j.meegid.2021.104760

24. Bakhautdin B, Febbraio M, Goksoy E, et al. Protective role of macrophage-derived ceruloplasmin in inflammatory bowel disease. Gut 2013;62(2):209–219. DOI: 10.1136/gutjnl-2011-300694

25. Shoenfeld Y. Corona (COVID-19) time musings: our involvement in COVID-19 pathogenesis, diagnosis, treatment and vaccine planning. Autoimmun Rev 2020;19(6):102538. DOI: 10.1016/j.autrev.2020.102538

26. Huang J, Cheng A, Kumar R, et al. Hypoalbuminemia predicts the outcome of COVID-19 independent of age and co-morbidity. J Med Virol 2020;92(10):2152–2158. DOI: 10.1002/jmv.26003

27. Zinellu A, Mangoni AA. Serum prealbumin concentrations, COVID-19 severity, and mortality: a systematic review and meta-analysis. Front Med (Laussane) 2021;8:638529. DOI: 10.3389/fmed.2021.638529

________________________
© The Author(s). 2023 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted use, distribution, and non-commercial reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.