Evaluating High Fluorescence Lymphocyte Count as a Predictor of Severe Dengue Infection

ABSTRACT

Background

Dengue infection (DI) is a significant global health concern, with severe cases leading to plasma leakage, organ failure, and shock. Identifying reliable biomarkers for early risk stratification is crucial for improving patient management.

Methods

This descriptive cross‐sectional study included 268 DI patients at 103 Military Hospital, Vietnam, from July 2022 to October 2023. Patients were classified into Non‐Severe DI and Severe DI groups per the 2009 WHO dengue guideline. High Fluorescent Lymphocyte Count (HFLC), a parameter from the Sysmex hematology analyzer, was analyzed. HFLC% represents the proportion of high fluorescent lymphocytes among total WBCs, while HFLC# is these cells' absolute count (G/L). Statistical methods included Mann–Whitney U, Kruskal‐Wallis, Chi‐square tests, ROC curve analysis, and binary logistic regression.

Results

HFLC% and HFLC# were significantly elevated in Severe DI compared to Non‐Severe DI (p < 0.001). HFLC% negatively correlated with platelet count and positively with liver enzymes (AST, ALT), suggesting an association with severe complications. ROC analysis showed that HFLC# (AUC = 0.913, p < 0.001, cut‐off = 1.00 G/L) had 70.6% sensitivity and 90.8% specificity, while HFLC% (AUC = 0.833, p < 0.001, cut‐off = 13.15%) had 70.6% sensitivity and 80.5% specificity for predicting severe DI.

Conclusion

Elevated HFLC at admission is strongly associated with severe DI, with HFLC# demonstrating excellent predictive accuracy. These findings suggest that HFLC is a promising biomarker for early identification of high‐risk dengue patients. Further large‐scale validation is required to confirm its clinical utility.

A descriptive cross‐sectional study was performed on 268 DI patients at 103 Military Hospital, Vietnam, from July 2022 to October 2023. The results revealed that HFLC% and HFLC# were significantly elevated in Severe DI compared to Non‐Severe DI. HFLC% negatively correlated with platelet count and positively with liver enzymes (AST, ALT), suggesting an association with severe complications. ROC analysis showed that HFLC# (AUC = 0.913, p < 0.001, cut‐off = 1.00 G/L) had 70.6% sensitivity and 90.8% specificity, while HFLC% (AUC = 0.833, p < 0.001, cut‐off = 13.15%) had 70.6% sensitivity and 80.5% specificity for predicting severe DI.

1. Introduction

Dengue infection (DI) is the most common arthropod‐borne viral disease, with over 6.5 million cases and 6800 deaths reported globally in 2023 [1]. It is caused by four antigenically distinct dengue virus (DENV) serotypes transmitted by infected Aedes mosquitoes. DI manifests with a wide clinical spectrum, ranging from mild fever to severe complications such as hemorrhage, organ failure, or hypovolemic shock. Severe dengue infection (DI) happens in about 5% of cases [2]. Mild cases can sometimes progress to severe dengue [3]. Finding reliable ways to predict which patients might develop complications is important.

One major risk factor for severe DI is a second infection with a different dengue virus strain, which significantly increases the chances of complications [4]. After initial infection, patients develop lifelong immunity to that specific virus type but only short‐term protection (2–3 months) against other types [5]. This temporary immunity makes them more vulnerable to reinfection, which is linked to severe disease. Studies suggest that secondary infections are 4–6 times more likely to lead to severe dengue compared to primary infections [6, 7].

Given the pivotal role of immune response in DI severity, identifying reliable immunological markers may help predict disease progression. One promising marker is the high fluorescence lymphocyte count (HFLC), a value obtained from a routine complete blood count (CBC) test. The HFLC is measured using the SYSMEX hematology automated analyzer, which identifies different types of white blood cells based on how they scatter light and fluoresce. HFLC represents a group of cells with low side scatter and high fluorescence, indicating they contain a lot of RNA [8, 9]. Studies have shown that HFLC is strongly linked to plasmablasts and plasma cells (PBs/PCs) [10], which are responsible for producing antibodies [11]. Previous research found that PB/PC levels are higher in DI than in other febrile illnesses (OFI), suggesting they play a crucial role in the body's immune defense. Because of this connection, HFLC may serve as an automated indicator of PBs/PCs and could help assess disease severity. While several studies have confirmed that HFLC is useful for diagnosing DI [8, 12, 13], its link to severe DI has not been studied. Therefore, this study aims to evaluate whether HFLC at hospital admission can serve as an early warning sign for severe DI.

2. Patients and Methods

2.1. Study Design

The study was conducted at the Infectious Diseases Department, 103 Military Hospital, Vietnam. This descriptive cross‐sectional study included a total of 268 patients who were diagnosed with DI according to the World Health Organization Guideline for Dengue (2009 edition) between July 2022 and October 2023.

2.2. Study Population

Patients with DI confirmed by clinical symptoms with a positive NS1 antigen and/or positive dengue immunoglobulin M (dengue IgM) using Rightsign Biotest were included in this study. We excluded patients with hematological or autoimmune disorders, who used immunosuppressive drugs, or who had known infections with other viruses.

2.3. Data Collection

Laboratory indices, including complete blood count (CBC), biochemical tests, and serology tests, were gathered on the first day of arrival. Manual platelet counts were performed in cases where automated results showed very low platelet levels (usually < 50 G/L) or when platelet clumping was suspected. A peripheral blood smear was prepared, stained with Wright‐Giemsa, and examined under a microscope. Platelets were counted in 10 high‐power fields, and the average was used to estimate the total count. This method helped confirm platelet levels in critical cases, ensuring accuracy for clinical decision‐making in dengue patients. The absolute count (HFLC#) and percentage (HFLC%) of high‐fluorescence lymphocytes (HFLC) were obtained from the CBC test conducted using the Sysmex XN1000 automated analyzer (Sysmex, Kobe, Japan) on the day of admission. Biochemical tests were performed by Beckman Coulter AU5800 chemistry analyzer (Beckman Coulter, USA). Serological tests were done using Rightsign Dengue rapid test. Clinical parameters such as presenting symptoms, warning signs, and severe complications were documented by a study team member.

According to the 2009 WHO guideline for DI, we classified patients into three categories: dengue without warning signs, dengue with warning signs (including abdominal pain, persistent vomiting, fluid accumulation, mucosal bleeding, lethargy, liver enlargement, and an increase in hematocrit with a decrease in platelets), and severe dengue (characterized by severe plasma leakage, severe bleeding, or organ failure) [3]. We then merged the first two groups into the Non‐Severe DI group, while the last group was classified as the Severe DI group.

2.4. Data Analysis

Statistical calculations were performed using SPSS (version 20.0; SPSS Chicago, IL, USA). In order to compare between groups, the Mann–Whitney U test and the Kruskal–Wallis test were applied for continuous variables. The chi‐square test was used to compare categorical variables. The receiver operating characteristic (ROC) curve was generated, and Youden's index was calculated to obtain the predictive cut‐off value of HFLC% to suspect severe DI. The Spearman test was used to examine correlations. Furthermore, we also created a suitable model for predicting Severe DI using binary logistic regression (BLR). A p < 0.05 was determined to be statistically significant.

3. Results

Table 1 summarizes the demographic characteristics of patients in the Non‐Severe and Severe DI groups. The two groups had no significant difference in age distribution or the number of fever days upon admission. However, a statistically significant association was found in gender distribution (p = 0.017), with a higher percentage of females in the Severe DI group compared to the Non‐Severe DI group. Severe complications of DI were observed in only 6.3% of cases in our study.

TABLE 1.

Comparison of demographic characteristics between Severe DI and Non‐Severe DI groups.

Variable	Total (n = 268)	Non‐Severe DI (n = 251)	Severe DI (n = 17)	p
Age, median (IQR)	43 [32–61]	43 [32–62]	42 [31–59]	0.662
Female; n (%)	130 (48.5)	117 (46.6)	13 (76.5)	0.017
Male; n (%)	138 (51.5)	134 (53.4)	4 (23.5)
Days of fever on admission, median (IQR)	4 [3–5]	4 [3–5]	5 [4–5]	0.449

Note: Significance of bold indicates statistical significance of p value when P < 0.05.

Table 2 presents serological, complete blood count (CBC), and biochemical parameters of patients in both groups. Serological analysis revealed a statistically significant difference in IgG positivity between the groups, with a higher percentage of IgG‐positive cases in the Severe DI group (93.3%).

TABLE 2.

Comparison of laboratory results between Severe DI and Non‐severe DI group.

Variable	Total (n = 268)	Non‐Severe DI (n = 251)	Severe DI (n = 17)	p
Result of serology test
NS1; n (%)	233 (94.0)	218 (94.4)	15 (88.2)	0.306
IgM; n (%)	104 (39.5)	93 (37.5)	11 (73.3)	0.06
IgG; n (%)	173 (66)	159 (64.4)	14 (93.3)	0.021
Result of complete blood count test
WBC (G/l), median (IQR)	4.09 [2.05–6.01]	3.93 [2.78–5.60]	8.55 [5.39–13.88]	< 0.001
NEU (G/l), median (IQR)	2.03 [1.31–3.08]	2.00 [1.32–2.94]	3.16 [1.17–7.62]	0.137
LYMP (G/l), median (IQR)	1.08 [0.68–1.91]	1.01 [0.66–1.80]	2.92 [2.09–5.70]	< 0.001
MONO (G/l), median (IQR)	0.46 [0.28–0.70]	0.44 [0.27–0.67]	0.80 [0.51–1.06]	0.002
HGB (g/l), median (IQR)	142.00 [132.00–158.00]	143.00 [132.00–159.00]	133.00 [113.50–146.00]	0.016
HCT (L/L), median (IQR)	0.42 [0.39–0.46]	0.42 [0.39–0.46]	0.39 [0.33–0.43]	0.008
PLT (G/l), median (IQR)	42.50 [17.00–116.00]	45.00 [17.00–122.00]	21.00 [15.00–52.00]	0.044
Result of biochemical indices
GLU, median (IQR)	6.61 [5.65–7.85]	6.61 [5.70–7.85]	5.97 [5.03–8.02]	0.264
URE, median (IQR)	4.23 [3.20–5.47]	4.22 [3.20–5.47]	4.32 [2.69–10.24]	0.484
CREA, median (IQR)	85.17 [73.62–99.55]	85.10 [73.76–99.19]	88.67 [62.72–132.74]	0.877
AST, median (IQR)	99.91 [59.51–189.60]	95.75 [56.48–173.74]	1000.80 [151.03–1374.99]	< 0.001
ALT, median (IQR)	55.91 [32.33–108.76]	51.55 [30.93–98.40]	343.73 [70.09–643.11]	< 0.001
CRP, median (IQR)	7.59 [3.43–13.83]	7.12 [3.44–12.71]	13.18 [2.92–33.67]	0.075

Note: Bold value: significant differences.

Abbreviations: ALT, alanine aminotransaminase; AST, aspartate aminotransaminase; CREA, creatinine; CRP, C‐reactive protein; GLU, glucose; HCT, hematocrit; HGB, hemoglobin; IgG, immunoglobulin G; IgM, immunoglobulin M; LYMP, lymphocyte; MONO, monocyte; NEU, neutrophil; NS1, nonstructural protein 1; PLT, platelet count; URE, urea; WBC, white blood cell count.

Significant variations were observed in CBC parameters. White blood cell indices showed that the median WBC, lymphocyte (LYMP), and monocyte (MONO) counts were markedly lower in the Non‐Severe DI group compared to the Severe DI group. Conversely, red blood cell indices revealed significantly reduced hemoglobin and hematocrit levels in the Severe DI group. Additionally, platelet counts were significantly lower in the Severe DI group than in the Non‐Severe DI group.

Regarding biochemical indices, the median levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were significantly higher in the Severe DI group than in the Non‐Severe DI group.

The result in Table 3 shows that both HFLC absolute count (HFLC#) and HFLC percentage (HFLC%) were significantly elevated in the Severe DI group compared to the Non‐Severe DI group (p < 0.001).

TABLE 3.

Comparison of high fluorescent lymphocyte count and percentage between Severe DI and Non‐Severe DI groups.

	Total (n = 268)	Non‐Severe DI (n = 251)	Severe DI (n = 17)	p
HFLC#, median (IQR)	0.11 [0.02–0.52]	0.09 [0.02–0.40]	1.24 [0.65–2.17]	< 0.001
HFLC%, median (IQR)	3.55 [0.50–12.20]	15.70 [11.30–24.65]	2.70 [0.40–11.10]	< 0.001

Note: Bold value: significant differences.

Abbreviations: HFLC#, high fluorescence lymphocyte count; HFLC%: high fluorescence lymphocyte percentage.

Spearman correlation analysis in Table 4 indicates significant negative correlations between platelet count and HFLC# (r = −0.625, p < 0.001) as well as HFLC% (r = −0.638, p < 0.001). Increased HFLC# and HFLC% were also associated with elevated liver enzyme levels, including GOT and GPT.

TABLE 4.

Correlations of PLT and liver enzyme with HFLC.

Variables	HFLC#		HFLC%
	ρ	p	ρ	p
PLT	−0.625	< 0.001	−0.638	< 0.001
AST	0.526	< 0.001	0.511	< 0.001
ALT	0.446	< 0.001	0.415	< 0.001

Note: Bold value: significant differences.

Abbreviations: ALT, alanine aminotransaminase; AST, aspartate aminotransaminase; HFLC#, high fluorescence lymphocyte count; HFLC%, high fluorescence lymphocyte percentage; PLT, platelet.

Regarding the predictive value of HFLC# and HFLC% for severe dengue infection (DI), ROC analysis presented in Figure 1 indicates that both parameters can serve as useful predictors. For HFLC#, the area under the curve (AUC) was 0.913 (p < 0.001), with a cut‐off value of 1.00 G/L, yielding a sensitivity of 70.6% and a specificity of 90.8%. For HFLC%, the AUC was 0.833 (p < 0.001), with a cut‐off value of 13.15%, showing a sensitivity of 70.6% and a specificity of 80.5%.

FIGURE 1.

ROC of High Fluorescent Lymphocyte Count (green line) and percentage (blue line) in predicting severe DI patients. HFLC#: AUC = 0.913, p < 0,001, cut‐off value = 1.00 G/L, Sensitivity = 70.6%, specificity = 90.8%. HFLC%: AUC = 0.833, p < 0,001, cut‐off value = 13.15%, Sensitivity = 70.6%, specificity = 80.5%.

Binary logistic regression (BLR) analysis using the enter method (Table 5) identified HFLC# as an independent laboratory marker for predicting severe DI.

TABLE 5.

Binary logistic analysis prognosis factor for Severe DI.

Variables	OR	95% CI	p
WBC	1.361	1.015–1.825	0.039
LYMPH	0.551	0.283–1.074	0.08
MONO	0.800	0.248–2.580	0.708
HGB	0.966	0.944–0.990	0.005
PLT	0.995	0.983–1.007	0.394
HFLC#	22.881	1.303–401.745	0.032
HFLC%	0.981	0.846–1.139	0.805

Note: Bold value: significant differences.

Abbreviations: HFLC#, high fluorescence lymphocyte count; HFLC%, high fluorescence lymphocyte percentage; HGB, hemoglobin; LYMPH, absolute lymphocyte count; MONO, absolute monocyte count; PLT, platelet; WBC, white blood cell count.

4. Discussion

Our study demonstrates that High Fluorescent Lymphocyte Count (HFLC) is significantly elevated in severe dengue Infection (DI) compared to non‐severe cases. This finding suggests a potential role for HFLC in disease progression and severity assessment.

HFLC reflects the antibody secreting cells (ASC) population, which increases in dengue infection, especially in secondary infections caused by a different serotype. Linssen et al. [10] demonstrated that HFLC corresponds to ASC, including short‐lived plasmablasts producing low‐affinity antibodies and long‐lived plasma cells generating high‐affinity antibodies. Secondary dengue infections are more likely to be severe due to enhanced plasmablast responses. Garcia‐Bates et al. [14] reported a significantly higher proportion of plasmablasts among the total number of B lymphocytes in severe secondary dengue cases and secondary infections regardless of severity. Our study supports these findings with 93.3% of severe cases exhibiting higher IgG positivity than non‐severe cases. Since plasmablasts dominate secondary immune responses [15], their increased presence in secondary DI likely results from memory B cell activation, leading to a more robust but often suboptimal immune response.

Despite the increased HFLC, patients with secondary infections frequently experience severe complications due to the phenomenon of “Original Antigenic Sin” (OAS). In secondary dengue infections, the immune response is more rapid and intense due to the activation of memory B cells and robust plasmablast expansion. However, cross‐reactivity between antibodies from the primary infection and the new serotype results in incomplete viral neutralization, increasing disease severity [16, 17, 18]. Differences in viral surface proteins (epitopes) among serotypes enable immune evasion, leading to ineffective binding by pre‐existing antibodies. Additionally, cross‐reactive antibodies suppress the activation of naive B cells that could generate more effective, serotype‐specific responses. This dysregulated immune response may prolong viral clearance and exacerbate disease severity [19].

Furthermore, our study highlights correlations between HFLC, platelet count, and liver enzymes, suggesting a link between HFLC elevation and severe complications. Using the Spearman correlation, there were significant negative correlations between platelet count, HFLC#, and HFLC% (Table 4). Abeysuriya et al. have similar findings that there was an inverse relationship between absolute as well as percentage of atypical lymphocytes (AL) and platelet count with the correlation coefficient of −0.130 and −0.116, respectively [20]. According to Raharjo et al. [21], in severe dengue hemorrhagic fever, immune complexes are formed when circulating dengue virus binds to specific IgG antibodies. These complexes often attach to platelet surfaces and are subsequently cleared by the reticuloendothelial system, contributing to thrombocytopenia. The observed increase in HFLC, reflecting elevated antibody‐secreting lymphocytes, may explain the inverse correlation with platelet count. Whereas AL reflects activated T and B lymphocytes, HFLC reflects only those cells involved in antibody secretion that have been studied to be elevated in the peripheral blood of DI patients. Therefore, chances are, HFLC is more closely related to thrombocytopenia—one of the main causes of severe bleeding in DI patients. We also found that as HFLC# and HFLC% increased, liver enzymes such as AST and ALT also elevated. Since platelet count (PLT) and liver enzymes are key laboratory markers for bleeding risk and liver failure, two life‐threatening complications used to diagnose severe DI, the strong correlation between HFLC and these markers suggests that HFLC may have potential value in predicting severe DI. Moreover, increased vascular permeability in severe dengue cases may be partially attributed to plasmablast‐derived cytokines, such as IL‐6, which contribute to endothelial dysfunction, plasma leakage, hypovolemia, and, in extreme cases, dengue shock syndrome (DSS) [22]. The interplay between virus‐specific plasmablast expansion, complement activation, and platelet dysfunction may trigger vasculopathy, ultimately leading to plasma leakage and severe disease outcomes. Virus‐specific antibodies produced by plasmablasts contribute to immune complex formation, activating the complement cascade and releasing vasoactive anaphylatoxins [23].

Given its strong correlation with disease severity, HFLC may serve as a potential biomarker for early identification of severe dengue cases. Monitoring HFLC levels at hospital admission could facilitate risk stratification, allowing clinicians to implement timely interventions to prevent severe complications. Future research should focus on establishing HFLC threshold values that can reliably predict disease progression, thereby enhancing patient management and reducing dengue‐related morbidity and mortality.

This study has several limitations. First, it was conducted at a single center with a relatively small sample size, which may limit the generalizability of the findings to broader populations. Second, while elevated HFLC levels were observed, the study did not assess how this parameter was utilized in clinical decision‐making or patient management. Third, all serological tests were performed using rapid card assays without ELISA‐ or PCR‐based confirmation, potentially affecting the accuracy of infection classification. Fourth, reinfection history was not evaluated, and the study did not report the number of patients with primary versus secondary dengue infections, limiting the interpretation of IgG‐related findings. Finally, the number of patients with severe dengue infection was relatively low, which may have reduced the statistical power and stability of the regression model.

5. Conclusion

Our study suggests that patients with severe dengue infection (DI) exhibit significantly higher HFLC compared to those with dengue with or without warning signs. We propose that an admission HFLC% > 13.15% and HFLC# > 1.00 G/L could serve as novel positive predictive factors for severe DI. While these findings are promising, further validation is needed using the Sysmex automated hematology system and other analyzers. Future larger prospective longitudinal studies are essential to confirm whether admission HFLC levels can predict the development of complications in the later stages of dengue infection.

Author Contributions

Conceptualization: A.V.H. and H.T.V.; Methodology: H.T.V. and K.N.T.; Formal analysis and investigation: A.V.H., H.N.T.H., K.N.T., and H.T.V.; Writing – original draft preparation: A.V.H.; Writing – review and editing: H.N.T.H. and K.N.T.; Supervision: H.T.V.

Ethics Statement

Animals did not participate in this research. All human research procedures followed were in accordance with the ethical standards of the committee responsible for human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008. This study was approved by the Ethical Committee of Vietnam Military Medical University (No: 89/HĐĐĐ).

Consent

Informed consent was obtained from all subjects involved in the study.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.