C-Reactive Protein-to-Albumin Ratio as a Prognostic Inflammatory Marker in COVID-19

Content of the Study and Participants

This retrospective cohort study was performed at Medical Faculty Research and Application Hospital between March 11, 2020, and June 01, 2020, following the MoH Diagnosis and Treatment Guidelines.^[13,14] A total of 125 patients initially diagnosed with COVID-19 and 30 healthy individuals constituted the study universe. Among the 125 COVID-19 patients, 68 were positive for the RT-PCR test (LC) and 57 were negative for the RT-PCR test (LUC). The LUC group had negative PCR tests and respiratory cultures, but their computed tomography (CT) findings were compatible with COVID-19 and could not be attributed to any other cause. A total of 23 individuals in both groups were excluded because they were under 18 or pregnant or required intensive care. Nasopharyngeal samples of the remaining 56 patients with positive RT-PCR tests (LC) and 46 patients with negative (LUC) were included in the study. Additional nasopharyngeal samples from 30 age-matched healthy participants, with negative RT-PCR tests, were evaluated as the control group (►Fig. 1).

Close

Fig. 1

Study participants.

Export to PPT

Data

Demographic information, epidemiological history, clinical symptoms, comorbid diseases, imaging features, laboratory data, and length of stay were collected through an electronic medical record system. CBC data were determined using an ABX Pentra DX 120 (Horiba Medical, Montpellier, France) hematology analyzer. Biochemical tests were performed with Roche's Cobas 8000 c502 Analyzer (Roche diagnostics; Geneva, Switzerland). Coagulation tests were performed with the Sysmex CS-2500 System coagulation analyzer (Siemens Healthcare Diagnostics, Erlangen, Germany).

COVID-19 PCR Test

RNA isolation from the nasopharyngeal sample was conducted with Speedy vNAT Bio-viral extraction kit (Bioeksen, Turkey). From the obtained mixture, the real-time qPCR RdRp gene was identified with Bio-speedy COVID-19 RT-qPCR kit (Bioeksen, Turkey). The PCR process was performed by programming the thermal cycler device for 20 minutes at 450°C (once), 5 minutes at 950°C (once), at 950°C for 15 seconds + at 550°C for 40 seconds (45 times). PCR results were evaluated as positive/negative considering the cycle threshold (ct) values according to the amplification curves.

Diagnosis and Follow-Up

COVID-19 was diagnosed according to Diagnosis and Treatment Guidelines.^[14] Patients with acute respiratory tract infection developed in the last 14 days who required hospitalization due to fever, cough, and dyspnea, tachypnea, hypoxemia, hypotension, had diffuse radiological findings on lung imaging, change in consciousness, and who were found to be SARS-CoV-2 positive by molecular methods were considered to have COVID-19.

In this study, we classified COVID-19 patients into three groups based on the severity of their illness (mild, severe, and critical). Patients were assigned to these groups based on their length of hospitalization, symptoms, accompanying diseases, radiological findings, and treatments. Mild patients were those patients who had mild or no signs of pneumonia. Patients with dyspnea, hypoxia, or opacities over 50% in lung imaging within 24 to 48 hours of hospitalization were categorized in the severe group, and respiratory failure, shock, or multiple organ failure in the critical group.^[15] Patients who were admitted to the intensive care unit were not evaluated further. The end of the follow-up period was June 1, 2020.

Statistical Evaluation

In ►Table 1, we report the frequency and percentage values for categorical variables, and mean, standard deviation, and minimum and maximum values for continuous variables (►Table 1).

Parametric/nonparametric distribution was determined by performing the Kolmogorov–Smirnov test for all groups. Student's t-test was used for variables with parametric distribution and the Mann–Whitney U-test for variables with the nonparametric distribution. A cut-off level for biomarkers was determined using the ROC (receiver operator characteristics) curve and the median value. All statistical analyses were performed with the SPSS 22.0 program and the p-values for significant findings were less than 0.05.

Complete Blood Count and Immature Parameters

Similar to other publications, in our study absolute numbers of WBC#, LY#, and EOS# were found to be lower than the control group, and NEU# absolute counts were found to be higher. Among the platelet fraction parameters, PLT# absolute count and PCT% were low, and MPV and PDW% were found to be significantly higher in LC and LUC patients compared with the control group. Sun et al reported that CBC parameters of COVID-19 patients were significantly abnormal with hemoglobin, platelets, and lymphocyte count significantly lower than those in the control group.^[18] Xu et al also reported infiltration of mononuclear cells, especially lymphocytes, in the interstitial lung in COVID-19 patients.^[19,20] Li et al found that some deaths from COVID-19 had overlapping bacterial pneumonia.^[21] Yang et al reported lymphopenia in 80% of adult COVID-19 patients with critical illnesses, while Chen et al reported lymphopenia in only 25% of patients with mild COVID-19 infection.^[22,23] Lymphocyte counts reflect the effect of acute physiological stress and there is reason to expect that lymphopenia may be related to the severity of the SARS-CoV-2 infection.^[24]

In our study, immature cell parameters IMG, IMM, IML, LIC, and ALY, in absolute numbers, and platecrit (PCT%) were also found to be lower in confirmed cases than in the control group. However, there is limited literature on this subject regarding COVID-19. It would not be correct to make a comparison since different devices such as IG%^[25] and LUC%^[26] have different units such as the measured parameters and percentage.^[27-29] LY did not differ significantly in LUC compared with the LC group, while PLT, WBC, NEU, EO, IMM, LIC, and ALY had significantly lower values. Additionally, while LY was significantly lower in LUC cases compared with the healthy control group, NEU numbers of unconfirmed cases were higher than healthy control and confirmed groups. This indicates that the number of NEU may be a good indicator in unconfirmed patients with positive CT findings but negative PCR tests. We found no significant difference between the confirmed and unconfirmed groups in other coagulation and biochemical parameters. Therefore, inflammatory markers are very valuable in terms of diagnostic and prognostic follow-up of unconfirmed patients.

Inflammatory Markers

Higher numbers of proinflammatory cytokines in the serum of COVID-19 patients and the use of anti-inflammatory agents for treatment highlight the critical role of inflammation in the progression of the disease. Zeng et al concluded that inflammatory markers, especially CRP, PCT, interleukin (IL)-6, and ESR, had a positive correlation with the severity of COVID-19.^[30] In our study, we examined CRP, NLR, PLR, SII, and CAR tests in detail as markers of inflammation. CRP, SII, CAR, PLR, and NLR were significantly higher in confirmed cases compared with the control group. CRP, SII, CAR, PLR, and NLR were higher than the control group also in unconfirmed cases. Compared with unconfirmed patients, SII, CRP, and CAR were significantly lower in confirmed patients.

The literature only points to relationships between NLR, PLR, and COVID-19.^[10,31] Mazza et al have also found a direct correlation between obsessive-compulsive disorders and MLR, NLR, and SII, suggesting that higher initial inflammation may be associated with later obsessive-compulsive disorder symptoms.^[32,33] Our findings of CAR are unique and an important addition to the literature in this rapidly evolving field.

Disease Severity

We also evaluated LC patients according to the severity of the disease. CRP, CAR, PLR, and NLR indices showed increasing values as disease severity increased. In our study, the fact that NLR was significantly higher in cases with severe disease supports the literature.^[10,34] However, although the optimal cut-off determined for NLR is 1.82 and LR⁺ is 1.21, this offers a low diagnostic accuracy (54.65%) in making a differential diagnosis between severe and nonsevere patients. In the literature, different NLR cut-off values have been reported in different populations (with or without cancer) with different methods. No universal cut-off values are available yet. It would be reasonable to interpret such a parameter carefully in a clinical context, even if based on data from a comparable population.^[35] Our study also revealed that CRP level was positively correlated with the severity of COVID-19, similarly to other studies.^{[19,23,36,37]}

Low albumin levels are associated with poor prognosis and mortality for COVID-19 patients in the literature and serum albumin-mediated strategies were shown to be useful in treating COVID-19 patients.^[38] One of the noninvasive markers for albumin, CAR, stands out as the most appropriate potential marker in our study. CBC parameters used in other markers are affected by any inflammation, infection, physiological stress, changes in significant serum parameters, or biological variation, such as age, gender, and race.^[39]

Consisting of CRP and ALB, CAR has the potential to show inflammation and infection processes more clearly. CAR has also been recognized as an independent factor useful in predicting short-term survival within 2 weeks for COVID-19 patients.^[40] The fact that CRP and ALB are less affected by changes such as biological variation compared with other blood count parameters increases the diagnostic and prognostic potential of CAR.

In our study, ALB values at hospitalization were found to be significantly lower compared with the control group. Also, there was a significant decrease in ALB levels in the group with severe disease compared with the mild group. The CAR index, in which albumin and CRP values were evaluated together, was significantly higher in LC compared with the control group, and lower in LUC compared with LC. In the examination of disease severity, CAR was found to be significantly higher in patients with severe illness compared with mild patients. Among the parameters we studied as an inflammatory marker, CAR had the highest AUC of 93% and PPV of 92.5%. It was also 6.59 times better at distinguishing between patients with severe and mild disease progression with a high positive predictability rate (LR⁺). Its diagnostic accuracy was 89.29%.

Inflammatory markers show activation of systemic inflammatory processes but are not specific to COVID-19. Although there are evident abnormalities in these inflammatory markers in COVID-19 patients, they are important only for diagnostic and prognostic predictability. The use of markers alone for systemic inflammation can also be misleading. Until the pathogenesis of COVID-19 is fully clarified and specific diagnosis and treatment methods are developed, the use of markers together with other data (history, symptoms, comorbidity, radiological findings, and other laboratory parameters) will be more meaningful.

Nonetheless, CAR, which has the highest diagnostic accuracy compared with other inflammatory markers, is a useful parameter in the diagnosis and estimation of disease severity for COVID-19. Of course, it would be appropriate to support our findings with data from larger samples.

Author Contribution

Substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data: A.Ç., E.Ç.G., M.D., B.E., A.Y.

Drafting the article or revising it critically for important intellectual content: A.Ç., E.Ç.G., A.Y., B.E., M.D.

Final approval of the version to be published: A.Ç., E.Ç.G., M.D., B.E., A.Y.

Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved: A.Ç., E.Ç.G., M.D., B.E., A.Y.

Guarantor: AÇ.

Ethical Approval

This study was approved by the Ministry of Health (2020–06–15T15_01_35) and the Faculty of Medicine Non-Interventional Clinical Research Ethics Committee on June 18, 2020 with the number of 2020.167.06.29.

Statement on Liability

The responsibility for the correctness of data and statements made in the manuscript rests entirely with the author.

Copyright Statement

The submitted manuscript represents original research, not previously published nor being considered for publication elsewhere.