Discriminant Value of Volume, Conductivity and Scatter Properties of Leucocytes (VCS Technology) for Rapid and Reliable Diagnosis of Malaria and Dengue Fever

Malaria and dengue are highly prevalent mosquito-borne tropical diseases that share several clinical and haematological features: a peak incidence in the rainy season, presentations with malaise, body ache and high fever, almost invariable thrombocytopenia and a propensity to develop DIC with bleeding and adverse outcomes if neglected or misdiagnosed. They are challenging to differentiate clinically from each other and also from the more common viral infections. Current laboratory tests are relatively time consuming and/or require specialized expertise or equipment. Expected findings of leukopenia and thrombocytopenia may be absent, or the parasites missed, thus delaying the ordering of confirmatory testing and definitive therapy.

Beckman Coulter analyzers incorporate VCS technology to quantify morphological characteristics of neutrophils, lymphocytes, monocytes and eosinophils: cell volume by voltage impedance (V); cytoplasmic/nuclear ratio by radiofrequency conductivity (C); and cytoplasmic granularity/nuclear complexity by laser light scatter (S). All these measurements (Mean and Standard Deviations (SD)) are reported as numerical values, called Cell Population Data (CPD) for every sample.VCS has helped improve the diagnosis of malaria and septicaemia in adults and neonates.

We compared the VCS parameters in dengue fever, malaria and other febrile illnesses with a view to generate automated factors for the identification of these illnesses.

We studied CBC parameters and VCS indices from 115 malaria patients (diagnosed by blood smear examination and immunochromatographic strip test for the plasmodium LDH and histidine rich protein-2) along with 105 patients with dengue fever diagnosed on positive NS1 antigen or IgM ELISA assays along with 105 control patients referred for testing for febrile illnesses who were negative for both malaria and dengue by all the above tests.

We used the originally developed program “EMMA” [Sukhachev DV, Zefirov NS. 10th European Symposium on Structure-Activity relationships: QSAR and molecular modeling. Barcelona, 4–9. 1994:A104] to generate discriminant functions to differentiate between the groups. The program uses combinatorial algorithms of selected parameters for regression equations by a modified stepwise procedure. It allows computation of a number of “best” regression equations with different parameter combinations. Only parameters statistically significantly different between the groups (checked with Mann-Whitney U test) are included in the discriminant functions. The diagnostic performances of various functions thus generated were assessed by ROC curve analysis with calculation of areas-under-the-curves (AUC).

The following equations were generated:

• To discriminate Malaria from Controls:

• Improved Malaria Factor =-0.473-0.00163*PLT+0.0524*LySDV+0.0302*LySDC

• To discriminate Dengue from Controls:

• Dengue Factor =0.3-0.00183*PLT+0.00619*LY%+0.0335*LySDC

• To discriminate Malaria versus Dengue: =5.51+0.0579*MCHC+0.00549*NE%+0.0138*LyMV+0.00956MoMV +0.027*MoSDV

• The performance characteristics of these 3 factors are shown in Table 1.

The parasitic/virological stimuli elicit strong immune responses resulting in leukocyte abnormalities in malaria and dengue that permit their distinction from other causes of fever. Our discriminant functions are easily calculable by LIS and flags thus generated are likely to have a high sensitivity and specificity for dengue and malaria. Our Improved Malaria Factor represents an enhancement over the original result by Briggs et al (In the ROC curves analysis Malaria Factor has AUC = 0.864, cut-off >4.25, Sens 86%, Spec 78%). Since the VCS data are obtained automatically as part of any CBC-diff, these results can greatly improve the detection of serious febrile illnesses in a timely and cost effective manner.

No relevant conflicts of interest to declare.