Eosinophilia: A Diagnostic Clue for Nonbacterial Diseases in Patients With Systemic Inflammatory Response Syndrome*

Eosinophil count (EC) as a diagnostic and prognostic marker for infection in the internal medicine department setting

INTRODUCTION

Eosinopenia has been previously investigated as a marker to differentiate infectious from non-infectious diagnoses and as a prognostic marker. Most previous studies were conducted in intensive care unit patients. Our study focuses on the value of eosinopenia in patients admitted to the Internal Medicine department.

METHODS

We retrospectively analyzed 271 consecutive patients with infection and 31 patients presenting with fever or inflammatory syndrome and a non-infectious diagnosis. We evaluated and compared the following markers for differentiating infectious from non-infectious diagnoses: eosinophil count [EC], CRP, WBC and neutrophil-to-lymphocyte count ratio [NLCR]. We also evaluated the value of eosinopenia as a monitoring parameter in patients with infections.

RESULTS

Eosinopenia at admission was found in 71% of patients with infection compared to 32% in the non-infection group. EC and NLCR were moderate markers for discriminating infection from non-infection, with an area under the receiver operating characteristic curve of 0.789 and 0.718 respectively. Significant eosinopenia (≤ 10/µL) had a high specificity (90%) for diagnosing infections. High EC at admission (> 400/µL) was rare in the infection group (1.5%), but not uncommon in the non-infection group (25.8%). Persistent eosinopenia was noted in non-survivors, compared to the rapid normalization of EC in survivors.

CONCLUSIONS

Among patients presenting with fever and/or high inflammatory markers a low EC is supportive of infection, while a high EC may suggest non-infectious diagnoses. The persistence/ resolution of eosinopenia may be a useful monitoring parameter to predict response to therapy.

The thymus and activation-regulated chemokine (TARC) level in serum at an early stage of a drug eruption is a prognostic biomarker of severity of systemic inflammation

BACKGROUND

In severe drug eruptions, precise evaluation of disease severity at an early stage is needed to start appropriate treatment. It is not always easy to diagnose these conditions at their early stage. In addition, there are no reported prognostic biomarkers of disease severity in drug eruptions. The aim of this study was to test whether the thymus and activation-regulated chemokine (TARC) level in serum at an early stage of a drug eruption can serve as a prognostic biomarker of systemic inflammation.

METHODS

Study participants included 76 patients who received a diagnosis of a drug eruption, one of the following: drug rash with eosinophilia and systemic symptoms/drug-induced hypersensitivity syndrome, maculopapular exanthema, and erythema multiforme. Stevens-Johnson syndrome (SJS)/toxic epidermal necrolysis (TEN) was eliminated in this study because scoring system for evaluating the severity was established. Correlation coefficients between serum TARC levels and indicators of systemic inflammation, including the neutrophil-to-lymphocyte ratio, Glasgow prognostic score, modified systemic inflammatory response syndrome (mSIRS) score, and C-reactive protein in serum were evaluated.

RESULTS

Serum TARC levels positively correlated with the neutrophil-to-lymphocyte ratio, Glasgow prognostic score, mSIRS score, C-reactive protein, albumin, white blood cell count, body temperature, and pulse rate. TARC levels negatively correlated with systolic blood pressure. Among these parameters, the mSIRS score showed strong correlation (correlation coefficient: 0.68).

CONCLUSIONS

Serum TARC levels correlate well with indicators of systemic inflammation and of disease severity among patients with a drug eruption except SJS/TEN. Serum TARC may be a prognostic biomarker of severity of inflammation in drug eruptions.