Comparison of severity signs between patients with primary and secondary dengue infection during the 2016 Burkina Faso (West Africa) outbreak

Background The factors that expose the severity of dengue are still controversial, particularly the relationship between severe dengue and secondary dengue. More importantly, the severity of dengue infection remains poorly studied in Africa. The objective of this study was to compare severity signs between patients with primary and secondary dengue infection during the 2016 dengue outbreak in Burkina Faso. Methods This was a cross-sectional study through a retrospective examination of patient medical records managed in Ouagadougou for dengue fever from 1 January 2015 to 31 December 2017. All health facilities with the capacity to perform dengue diagnosis in Ouagadougou were considered in the survey. Primary dengue was dened as the presence of AgNS1 and/or IgM and secondary dengue as the presence of IgG associated with one of these two markers. Patients with only IgG were excluded. Univariate and multivariable analyzes were performed using a logistic regression with dengue infection (primary or secondary dengue) as the binary dependent variable. The statistical signicant level was set at 0.05.


Introduction
Dengue is the most common arboviruses in the tropics. It is a potentially serious disease with about 500,000 cases of severe dengue each year worldwide and 2.5% of deaths (1). While this endemic disease is known in Africa, several sources of data indicate a lack of knowledge on its epidemiology background and severity in this region (2,3). Several West African countries have experienced dengue outbreaks during the past decade. Meanwhile the medical and clinical environment is characterized by the de ciency of the diagnostic capacity and the low knowledge on the infection and clinical signs by local health workers. In addition, until now the disease has been very infrequent and poorly taught and the surveillance system in place is inadequate (2,4). Moreover, the context of malaria prevalence contributes to a signi cant underestimation of the disease and its fatality rate in the region (5).
Severe dengue is clinically de ned by the presence of severity signs and symptoms such as plasma leakage, severe bleeding and organ failure. However, severe dengue fever remains poorly described in Africa (6,7). The most frequently reported sign of severity is hemorrhage, however shock syndrome or organ failure have been also reported in very few studies in Africa (8,9). In addition, the pathogenesis of severe dengue fever is not fully understood (10). The theory of Antibody-dependent enhancement (ADE) is supported by several publications to explain plasma leakage with shock syndrome. According to the ADE-theory, the risk of dengue hemorrhagic fever (DHF) is increased during re-infection with heterologous serotypes (sequential theory). The presence of a number of biological and in ammatory phenomena may lead to an increase of vascular permeability and leakage plasma. However, this hypothesis is subject to many controversies and could not explain all the biological disturbances and organ failures (encephalitis, hepatitis, myocarditis ...) observed during severe forms (11).
Although complications are most often reported during dengue re-infection, it should be noted that some publications report cases of severe dengue fever during primary dengue infection (12)(13)(14). For instance, in India about 10 to 30% of patients hospitalized for dengue hemorrhagic fever had primary dengue infection (15). Some viral serotypes could expose to a higher frequency of DHF.
The pathogenicity of severe dengue fever is often described by a combination of factors involving the patient's pre-existence of clinical conditions and medical history, the serotype virulence, viral load, vectorial skills which reinforces the controversy over the theory of ADE (16)(17)(18).
The rst cases of dengue fevers were reported in 1925 in Burkina Faso (19). Other outbreaks subsequently appeared in 1980 and 2013. Since 2013, cases of severe dengue fever have been regularly reported in Burkina Faso. On 18 November 2016, WHO recognized dengue fever outbreak in Burkina Faso, with a total of 1061 suspected cases noti ed between August and November 2016 mainly in the capital city, Ouagadougou (20). This disease represents one of the most important causes of infectious diseases in Burkina Faso and has a signi cant economic impact for the country (21). Meanwhile, the prevalence of severity signs as well as case fatality rate remains poorly studied. Indeed, so far, only one study conducted on 98 patients reports a prevalence of 18.4% and 11.2%, for DHF and dengue shock syndrome (DSS) respectively (22).
Entomological surveillance data indicate the circulation of Aedes aegypti and data from the National Reference Laboratory for Viral Haemorrhagic Fevers con rmed the presence of the 3 serotypes in the country (DENV-1, DENV-2 and DENV-3) (23). Routine surveillance of dengue fever has recently been initiated in Burkina Faso but is still restricted to seven sentinel sites in four cities (Ouagadougou, Bobo-Dioulasso, Pouytenga and Dori).
Our objective in this study was to compare severity signs between patients with primary dengue infection and secondary dengue infection during the 2016 outbreak in Burkina Faso.

Study design and study population
This was a cross-sectional study through the retrospective examination of patient medical records managed in Ouagadougou (capital city of Burkina Faso) for dengue fever from 1 January 2015 to 31 December 2017. The city of Ouagadougou is situated on the central plateau (12.4°N 1.5°W), in the Kadiogo province. The total population is estimated at about 2.700.000 with 48% of men and 52% of women. We considered all the health facilities that had the capacity to perform dengue testing in the city of Ouagadougou. We identi ed in total 15 health structures divided as follows: 4 tertiary health facilities, 3 secondary health facilities, 8 private clinics.
We included in this study all patients managed for dengue diagnosed on the basis of antigenic and/or serological tests (NS1 antigen positive and/or IgM positive). Patients who were tested positive for NS1 antigen and/or IgM were classi ed as primary dengue. When these two markers were isolated or simultaneously associated with IgG positivity, the patient was classi ed as secondary dengue. Patients who had only anti-dengue IgG and patients with dengue suspicion (not documented by an antigenic and serological test) were not included in the study.
The biological test used was the SD Bioline dengue duo which allows the detection of AgNS1, IgM and IgG.

Operational de nitions
For the dengue classi cation into primary or secondary, the following parameters were used: Primary dengue was de ned as the presence of AgNS1 and/or IgM, whereas secondary dengue was de ned as the presence of IgG associated with AgNS1 or IgM.
In this study, the following dengue clinical de nitions were used according to the WHO 2009: 2.3 Severe dengue fever: probable dengue fever with at least one of the following severity signs: Plasma leakage with hemoconcentration de ned by a hematocrit> 45%; Severe bleeding: de ned as diffuse and / or abundant bleeding with deglobulization requiring blood transfusion.
Hypovolemic shock: plasma leakage with shock syndrome or respiratory distress.
Severe thrombocytopenia: de ned by a platelet count of less than 20 × 10 3 per μ 2.4 Collection of data and statistical analysis Data were extracted from patient medical le and laboratory records. When clinical data were not available, patients were contacted by telephone to obtain more information about their signs and care pathways.
The questionnaire included socio-demographic characteristics (age, sex, educational level, and occupational activities), history and co-morbidities (asthma, diabetes, hypertension), clinical characteristics, biological results (thick blood smear or malaria RDT, blood cells count, biochemistry), the WHO dengue classi cation 2009, the status of patient evolution (complications, deaths).
For the investigation of severity signs, we compared, respectively, patients with primary dengue fever and patients with secondary dengue fever. The data analysis was performed using Epi-Info version 7 software. The statistical tests used for the statistical comparisons are the Chi 2 test or the Fisher's exact test (if Chi2 conditions were not met). Univariate and multivariable analyzes were performed using a logistic regression with dengue infection as the dependent variable. This dependent variable was a binary variable considering the value 1 if it is a primary infection and 0 if it is a secondary infection. The signi cant level was set at 0.05.

Socio-demographic characteristics of the study participants
Of the 811 patients managed for dengue fever during the study period, 418 (51.5%) were male. The median age was 30.5 years (Min = 1; max= 83 years). The peak prevalence (38.5% of patients) of registered cases was observed in November 2016. The number of hospitalized patients was 793 (97.8%). The majority of patients (477; 58.8%) were treated in private clinics. The distribution of patients according to their level of education and socio-professional level shows that respectively 464 patients (60.0%) had a secondary and higher education and 263 patients (35.4%) were public or private sector employees (Table  1). Severe thrombocytopenia was observed and con rmed in 37 patients (4.6%). (Figure 1).

Relationship between the severity signs of dengue fever and primary or secondary dengue infection
In univariate analysis, severe bleeding was associated with primary dengue infection (OR = 2.65, 95%IC: 1.16 -6.03), p = 0.01). However, in multivariable analysis, none of the signs of severity are statistically associated with primary or secondary dengue fever ( Table 2).

Discussion
In our study, about 1/3 of patients experienced severe dengue episode. The most common severity signs observed were renal failure followed by severe bleeding, and plasma leakage. Severe bleeding was statistically more common in primary dengue fever.
According to the WHO, nearly 500,000 cases of severe dengue (DHF and DSS) occur each year with nearly 20,000 deaths in the world. In the Americas, the incidence of severe dengue has been estimated at more than 200 cases per 100 000 persons in 2010 (24). Prothapregada et al. reported 37.4% of severe dengue in hospitalized children in India which is quite similar to the 33,5% of prevalence reported in our study (25).
Severe dengue fever is underestimated, particularly in Africa, where many cases probably remain undiagnosed due to the high frequency of other febrile non-malarial diseases and limited diagnostic capabilities (8,26).
The high prevalence of severe dengue in this study may be related both to the WHO classi cation used for organ failure and to the de nition of severity signs that vary from one study to another (27). Indeed, these severity signs are not precisely de ned by WHO, and most studies published so far have used the 1997 de nition. The new WHO de nition has improved sensitivity for earlier identi cation of severe cases and better screening of patients (28).
Renal failure appears in this study as the most common severity sign (13.1%). In the literature, the frequency of renal failure during severe dengue fever varies widely (0.2 to 36%) (29)(30)(31). In addition, the thresholds de ning renal failure during dengue fever are also variable according to these studies. The high frequency of renal failure in our study could be explained by the choice of creatinine threshold at 120 μmol/l. There are few published studies on renal failure during dengue fever in Africa. However, renal failure is common in this region, particularly because of malaria or other tropical infections. During dengue infection, several mechanisms may cause renal failure: renal hypoperfusion in the setting of sepsis, immunological damage and other unsolved mechanisms. Renal failure may also be a result of severe bleeding and plasma leakage. Sepsis, advanced age, use of nephrotoxic drugs, a history of high blood pressure or diabetes are also risk factors for renal failure during dengue fever (17,32).
The frequency of severe bleeding varies from one study to another. This can be explained by a lack of homogeneity in the de nitions of severe bleeding between studies. In our study, we considered severe bleeding patients with extensive bleeding and signi cant deglobulization that required blood transfusion. This strict de nition may explain the lower frequency of severe bleeding in our study compared to those reported in several studies (33). However, in West Africa, although the majority of the published work is case studies or small series, bleeding is the most common clinical severity sign that usually leads to the biological diagnosis of dengue fever and patients hospitalization (8,33). This could be explained by the fact that, in the context of limited resources, hemorrhage is a sign of severity that is easier to identify than other signs for which laboratory diagnosis remains essential. It is also one of the symptoms that is increasingly leading to the use of rapid diagnostic tests.
Plasma leakage was observed in 6.9% of study's patients and only 1.4% of the patients experienced hypovolemic shock.
In India, the frequency of DSS was 39% in children and was higher than the frequency of DHF (34). A frequency of 11.1% of DSS was reported by Abdallah et al. in East Sudan (35). Plasma leakage is suspected being underestimated complication in West-Africa due to lack of knowledge about the disease, the lack of systematic measurement of vital parameters and the unavailability of hematocrit test (4,36).
Regular monitoring of these clinical and biological parameters during the disease evolution is rarely done in the West-African context and more especially in Burkina Faso. As previously described, plasma leakage occurs during the critical phase of the disease, which varies according to the individual from 3 to 7 days, thus after the onset of symptoms and corresponds to the moment of thermal decline. A standardized surveillance study would better assess the frequency and importance of plasma leakage, its risk factors in the West-African context and therefore take appropriate measures to prevent patient deaths.
In our study, the rate of hemorrhage was signi cantly higher in primary dengue. This is in contrast with the majority of studies reporting that severe bleeding complications occur during secondary dengue infection which is consistent with ADE theory. Similar to our ndings other authors have reported, a higher prevalence of DHF in primary dengue infection (12,37). Furthermore, Soo et al. in a meta-analysis showed that the severity of dengue fever during primary infection was related to the virulence of the viral serotype (38). Tee et al. showed that severe bleeding occurred more frequently in elderly patients with primary dengue fever (39). An important pathogenic role of pro-in ammatory cytokines, including a signi cant elevation of IFN-γ, IL-12, TNF-α and IL-6 during primary dengue with serotype 2 could explain the severity of the disease during primary infection (40). Data from the national reference laboratory during the same outbreak in 2016 showed a predominant circulation of DENV-2 in Burkina Faso (41).
The difference between primary dengue and secondary dengue fever is often di cult and requires more advanced diagnostic techniques, including the measurement of antibody titration and for this, WHO also recommended haemagglutination inhibition technique which is limited in Burkina Faso (42). Therefore, the results of our study need to be interpreted with caution, since only 4% of our patients were con rmed and tested by PCR primary dengue infection. Thanachartwet et al. reported similar results with 3.5% of patients with primary dengue fever (43). Previous data from studies of patients who developed fever in community showed a seroprevalence of 66% IgG and con rm the endemic circulation of the dengue virus in the country (21). The proportion of severe dengue in these two groups may in uence a difference in favor of the lower strength group.

Conclusion
This study showed that severe dengue fever is a reality in West Africa and that its prevalence is high. The most common severity signs observed, were renal failure, severe bleeding and plasma leakage. As demonstrated, primary dengue is a source of hemorrhagic complications. This study con rms the need to conduct further studies to investigate on the severity factors of dengue fever. In addition, environmental sanitation and better knowledge of dengue fever by health workers are urgently needed to reduce the social and demographic consequences of dengue fever. However, this requires the commitment of all socio-political and health stakeholders in order to reach successful actions.

Declarations
Ethics approval and consent to participate An authorization from the Ministry of Health was obtained to carry out the data collection and the study in all health structures (2016/16-474/MS/SG). The data were reported to the Information and Freedoms Commission of Burkina Faso. The para clinical investigation was done as part of the routine management of the disease and therefore no additional cost was generated by the study for the patient.
The data were anonymous collected on a paper questionnaire to ensure con dentiality according to the ICHGCP standards.
Hospitalization of the patient was dependent on the severity level and/or tolerance of the disease by the patient. Burkina Faso government health subsidy scheme, which was launched in October 2016, allows free treatment for cases of severe dengue in tertiary level structures.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.