From: Blood biomarkers differentiating viral versus bacterial pneumonia aetiology: a literature review
Author (REF) | Date | Country | Age range | Type of the study | No. of CAP cases | No. of viral and bacterial cases | Markers tested | Conclusion and notes |
---|---|---|---|---|---|---|---|---|
Bhuiyan et al. [14] | 2019 | Australia | ≤17 years | Prospective cohort | n = 230 | Bacterial n = 30 | CRP, WBC, absolute neutrophil count (ANC) | CRP, WBC, and ANC were higher in definite bacterial pneumonia. The median blood CRP concentration was more than 6 times higher in definite bacterial cases than viral. |
Viral n = 118 | ||||||||
Other n = 82 | ||||||||
Huang et al. [15] | 2019 | China | 1 to 13 years old | Prospective cohort | n = 40 | HAdV pneumonia n = 20 | miRNAs | miRNAs biomarkers for HAdV pneumonia, at least in the cohort experiment. Importantly, neither pair of miRNAs could independently distinguish HAdV-infected patients from healthy children, highlighting the requirement for combining the two miRNA pairs identified in the present study. |
Healthy n = 20 | ||||||||
Yang et al. [16] | 2018 | China | 0.8 to 9.6 years | Retroperspective cohort | n = 321 | Healthy controls n = 50 | YKL-40, IL-6, IL-10, TNF-α, CRP | No significant difference between the levels of YKL-40 in the serum in all 3 pneumonia subgroups. The levels of YKL-40 in the BALF specimens of patients with bacterial pneumonia were significantly higher than with viral pneumonia. The levels of IL-6, TNF-α, and C-reactive protein were positively correlated with the serum levels of YKL-40. IL-10 levels were negatively correlated with YKL-40 levels in all pneumonia subgroups. |
Viral = 104 | ||||||||
Bacterial = 110 | ||||||||
Co-infection = 66 | ||||||||
Wallihan et al. [17] | 2018 | United States (Ohio) | 2 months to 18 years old | Prospective cohort | n = 152 | Healthy controls n = 39 | WBC, CRP, PCT | CRP, and PCT values were highter in patients with pyogenic bacteria than those with viral pneumonia or M. pneumonia. |
Pyogenic bacteria = 16 | ||||||||
M.pneumoniae = 41 | ||||||||
Viral n = 78 | ||||||||
Undetermined n = 14 | ||||||||
Yang et al. [18] | 2018 | China | ≤10 years | Prospective cohort | n = 82 | Healthy controls = 21 | HPR | HPR is higher in the mycoplasma pneumonia and the viral pneumonia than in the bacterial pneumonia. |
Bacterial = 21 | ||||||||
M.pneumoniae = 21 | HPR is a potential biomarker differentiate bacterial pneumonia and non-bacterial pneumonia. | |||||||
Viral = 19 | ||||||||
Higdon et al. [19] | 2017 | Bangladesh, Gambia, Kenya, Mali, South Africa, Thailand, Zambia | <5 years | Case-control study | n = 3981 | HIV-negative tested controls n = 601 | CRP | CRP was highter with bacterial pneumonia and negatively associated with RSV pneumonia. CRP may be useful for distinguishing bacterial from RSV-associated pneumonia, although its role in discriminating against other respiratory viral pneumonia needs further study. |
HIV-negative with comfirmed bacterial pneumonia n = 119 | ||||||||
HIV-negative with comfirmed viral pneumonia n = 556 | ||||||||
Esposito et al. [20] | 2016 | Italy | < 14 years old | Prospective cohort | 110 radiologically confirmed CAP | n = 20 no aetiology identified | Lcn2, SYN4, CRP, WBC | Lcn2 and SYN4 cannot predict aetiology. CRP together with WBC and clinical data, when combined, is the best predictor. |
Bacterial n = 74; | ||||||||
Viral n = 16; | ||||||||
Confirmed with PCR from nasopharyngeal swab positive | ||||||||
Valim et al. [21] | 2016 | Mozambique | < 10 years | Prospective cohort | n = 117 | Bacterial n = 23; | Fifty-six markers in a multiplex immunoassay. Hap, TNF receptor 2 or IL-10. The full list of markers was beyond the scope of this review. | Combination of three proteins (Hap), TNF receptor 2 or IL-10, and tissue inhibitor of metalloproteinases 1 provided the best tool to differentiate bacteria from the virus. |
Viral n = 30; | ||||||||
Bacteria isolated from blood or pleural fluid; | ||||||||
Viruses identified with PCR. | ||||||||
Esposito et al. [22] | 2016 | Italy | 4 months-14 years | Prospective cohort Multicentre | 433 radiologically confirmed CAP | Bacterial n = 235; | CRP and PCT, MR-proANP, MR-proADM, WBC, neutrophil percentages | CRP and PCT are better at predicting viral and bacterial aetiology than others. |
One or more viruses n = 111; | ||||||||
Unknown n = 87; real-time PCR tests on blood samples and nasopharyngeal swabs were used to identify agents. | PCT and MR-proANP helped to identify severe cases. | |||||||
Naydenova et al. [23] | 2016 | Gambia | 2–59 months | Retrospective Case-control | 780 clinically and radiologically confirmed CAP | Only in 84 cases, the aetiology was identified using blood culture. | CRP, Lcn2, Hap and CD163 protein | Aetiology can be determined using three vital signs (RR, HR, and SaO2) and a newly proposed biomarker (lipocalin-2) (81.8% sensitivity and 90.6% specificity). Lcn2 values below 200 ng ml are suggestive of a viral cause. Note: cases with no bacterial growth in blood culture were diagnosed as viral pneumonia. This is not reliable, as blood cultures are not always positive in bacteria CAP. This can lead to diagnostic bias. |
22 bacterial and 62 viral. | ||||||||
Zhu et al. [24] | 2016 | China | 10 months to 7 years | Prospective cohort | 65 based on criteria provided by IDSA and American Thoracic Society | Bacterial n = 34; | PCT | Bacterial pneumonia had far greater levels of PCT than non-bacterial. Also, statically significant changes in PCT level were tested before, and after treatment. Therefore, PCT is an important marker. |
Non-bacterial n = 32 | ||||||||
Note: The type of detection used to isolate the agents are not disclosed in this study. | ||||||||
Moreover, there is a mismatch between the sample size given in abstract, methods and results. | ||||||||
Engelmann et al. [25] | 2015 | France | 0–16 years | Prospective cohort Multicentre | n = 41 | Viral clinically* diagnosed n = 4. | MxA1 | Over 200 ng/ml of MxA has very high sensitivity and specificity in diagnosing a viral infection. |
CRP | ||||||||
Viral microbiologically confirmed n = 6 | NOTE: Not all cases were microbiologically confirmed as the study design did not accommodate this. The request for confirmation was based on the decision of the treating physician. | |||||||
Bacterial clinically* diagnosed n = 16 | ||||||||
Bacterial microbiologically confirmed n = 6 | ||||||||
No data n = 9 | ||||||||
*clinical diagnosis included signs and symptoms, and routine laboratory workup such as CRP | ||||||||
Hoshina et al. [26] | 2014 | Japan | < 15 years old | Retrospective cohort | n = 31 | Bacterial n = 21 | WBC, neutrophil counts, CRP, PCT | PCT was a very useful marker to differentiate bacterial pneumonia. Neutrophil count helped to discriminate bacterial bronchitis. |
Viral n = 10 | ||||||||
Bacterial CAP was confirmed by sputum culture. | ||||||||
Viral CAP was confirmed by nasopharyngeal aspirate, sputum or throat swab | ||||||||
Elemraid et al. [27] | 2014 and 2013 | United Kingdom | ≤16 years | Two prospective aetiological studies | n = 241 (in 2002) | 2002: | CRP, total WBC, and absolute neutrophil count | CRP and WBC/ neutrophil count can indicate aetiology to a great extent when combined |
Elemraid et al. [28] | n = 160 (in 2011) based on signs and symptoms and radiographic findings | Viral n = 47 | ||||||
Bacterial n = 58 | ||||||||
Mixed n = 12 | ||||||||
2011: | ||||||||
Viral n = 50 | ||||||||
Bacterial n = 28 | ||||||||
Mixed n = 20 | ||||||||
Zhou et al. [29] | 2011 | China | 0.11–2.57 years | Prospective cohort | n = 78 | Bacteria n = 27 | CRP, WBC, IgA, IgG, IgM, percent of T, Tc, Th, B, NK, CD23+, and CD25+cells and degree of expression of HMGB1 mRNA | HMGB1 and WBC can differentiate between bacterial, viral and co-infected cases of bronchial pneumonia |
Viruses n = 25 | ||||||||
Bacteria and viruses n = 26 |