1. Introduction

Viral meningitis occurs at any age with young children being the most commonly affected. Unfortunately, little is known regarding the incidence of viral meningitis (VM) in Sudan and other African countries unlike that of Europe and US. In a large cohort (n = 12,000) of children in Finland, the annual incidence of presumed VM was 219 per 100,000 in infants under 1 year and 27.8 per 100,000 overall in children under 14 years.[1] The incidence of aseptic meningitis in people aged 16 years and above was lower at 7.6 per 100,000 in another smaller study from the same country.[2] Viral meningitis is a notifiable disease in England and Wales, but many cases go unreported[1,3]; in 2005 to 2006, a total of 2898 people were admitted with a diagnosis of VM, 10 times the number of cases notified over the same period.[3] In the United States, VM accounts for approximately 26,000 to 42,000 hospitalizations each year, affecting mainly infants under 1 year of age and children 5 to 10 years.[4]

In Africa, some studies reported the involvement of potentially curable causes of central nervous system (CNS) infections; 2 (0.99%) children in Burkina Faso were found positive for HSV-1 among aseptic meningitis suspected patients whereas none for HSV-2.[5] In Uganda, cerebrospinal fluid (CSF) polymerase chain reaction (PCR) testing for herpes simplex virus type-1/2 (HSV-1/2), varicella zoster virus (VZV), human parechoviruses (HPeV), human enteroviruses (HEV), mumps and rubella viruses were all found negative in a suspected young child.[6] Herpes encephalitis was confirmed in 1 case (0.96%) among children presented with fever and coma in 3 main hospitals in Khartoum, Sudan, during April to November of 2011.[7] Some other studies reported more aggressive and potentially fatal neuroinvasive newly discovered viruses (i.e., Ntwetwe virus)[6] or previously known viruses (i.e., West Nile, Zika, Chikungunya and Ebola viruses)[8] in sub-Saharan Africa as major causes of CNS infections in children.

HSV-1 is responsible for more than 85% of all cases with sporadic necrotizing encephalitis and is infrequently associated with aseptic meningitis while neonatal infections account for 25% of all cases of neonatal HSV infection, with an overall incidence ranging from 1 case in 2000 to 5000 deliveries per year.[9,10] Aseptic meningitis is usually caused by HSV-2 primary infection; it occurs mainly in adults, since infections are mainly acquired by sexual contact, but can also occur in young children. HSV-2 accounts for 75% of neonatal infections by HSV.[10–12] In temperate climates, most non-polio HEVs infections occur in outbreaks during the warmest months, however, sporadic cases can develop throughout the year. In the United States, infections are seen more frequently from June to October. In tropical and subtropical regions, the incidence remains stable throughout the year.[4,13] HPeVs have been identified in Japan, Netherlands and United States. In a retrospective study from the Netherlands, CSF PCR testing from 716 children less than 5 years of age identified HPeVs in 4.6%. The yearly prevalence varied from 0.4% to 8.2%.[14,15] Since the introduction of varicella vaccine, the incidence of VZV infection (i.e., chickenpox, herpes zoster) in children has declined substantially and the overall incidence of the neurologic complications in healthy children, including aseptic meningitis, is reduced to less than 1%.[16] Excluding the single case report mentioned above,[7] we couldn’t find other studies on HSV1/2, VZV, non-polio HEVs and HPeVs infections in children in Sudan.

Viral meningitis is difficult to be differentiated clinically from bacterial meningitis or from other non-CNS related febrile infections that are more common in Sudan. In fact, the diagnosis of VM is uncommonly considered or reported in pediatric hospital practice in Sudan; unlike the highly inflated over-diagnosis of bacterial meningitis that was discussed and reported in our previous publication (Abdelrahim et al).[17] Besides, there is no well-established local viral etiological map. We, therefore, intended in this study to differentiate VM among suspected febrile attendees of a major referral children’s hospital in Khartoum, Sudan, and to explore viruses causing neurological infections amongst those that are mostly reported in children worldwide, namely; HSV-1, HSV-2, VZV, non-polio HEVs and HPeVs.

2. Methods 2.1. Materials

This is a cross-sectional hospital-based study. All children age 0 to 15 years attending Omdurman Hospital for Children in Khartoum, Sudan during December to August of 2010 and presented with fever (>37oC), seizures and a suspicion of neuroinfections were included (i.e., complete coverage) (n = 503). Demographic (i.e., age, sex), clinical (i.e., symptoms of meningeal irritation) and conventional laboratory data (i.e., total white blood cell count and type, glucose and protein concentrations, evidence of bacterial etiology) were obtained from hospital records. An aliquot of CSF was frozen in −80oC for nucleic acid analysis at the department of Clinical Microbiology, Umeå University, Umeå, Sweden. These included; viral deoxyribo nucleic acid (DNA) and RNA extraction using QIAamp® UltraSens Virus Technology, estimation of nucleic acids concentrations using NanoDrop™ spectrophotometery, preparation of complementary DNA (cDNA) from viral RNA using GoScriptTM reverse transcription system (for HEVs and HPeVs), quantitative real time PCR using TaqMan® method (for HSV-1/2, VZV, and HEVs) and conventional PCR followed by agarose gel electrophoresis (for HPeVs).

2.2. Secondary analysis using published clinical decision rules

Data collected throughout this project were used to identify infectious meningitis and differentiate aseptic/VM according to the criteria listed in the Clinical Case Definition for Aseptic Meningitis (2004) developed by the Brighton Collaboration Aseptic Meningitis Working Group,[18] these are; presentation with meningeal irritation, CSF pleocytosis and negative CSF bacterial culture with or without identified microbial etiology by nucleic acids test, or normal CSF cellular count but with evidence of microbial origin. Following the criteria (Table 1), cases were classified as aseptic meningitis based on clinical evidence of acute meningitis, CSF pleocytosis and negative CSF Gram stain and routine bacterial culture. Cases showing CSF pleocytosis (≥5 cells/mm3) were considered as Infectious Meningitis, those fulfilling all criteria of the Clinical Case Definition for Aseptic Meningitis were considered as Aseptic Meningitis, and cases with confirmed viral etiology were considered as Proven Viral Meningitis (Table 2).

Table 1 - Clinical case definition for aseptic meningitis and number of cases fulfilling the criterias. # Criteria for aseptic meningitis
Level (1) of diagnostic certainty Number of cases fulfilling only one criterion Number of cases fulfilling all 4 criteria 1 Clinical evidence of acute meningitis: fever, headache, vomiting, bulging fontanel, nuchal rigidity or other signs of meningeal irritation 361 out of 361 (142 MD) 18 (4.5%; 18/404) AND 2 CSF pleocytosis: >5 leukocytes/µL if patient is ≥ 2 mo, or > 15 leukocytes/µL in infants < 2 mo 23 out of 404 (97 MD) AND 3 Absence of any microorganism on Gram stain of CSF 497 out of 503 AND 4 Negative routine bacterial culture of CSF in the absence of antibiotic treatment before obtaining the first CSF sample 497 out of 503 Case definition developed by the Brighton Collaboration Aseptic Meningitis Working Group.[18]

MD = missing data 1.

Table 2 - Classification of Sudanese pediatric cases based on clinical and nucleic acids findings. ♯ Defined groups Cases out of 503* % Out of 40† % Frequency 1 Proven infectious meningitis‡ 8% 40† 100% 2 Proven aseptic meningitis 6.8% 34§ 85% 3 Proven viral meningitis 3.2% 16‖ 40% 4 HSV-1 meningitis 0.4% 2 5% 5 Non-polio HEV meningitis 0.4% 2 5%

*Total number of suspected febrile patients who attended the hospital during the study period and were subjected to lumber puncture.

†This number is the sum of 17 cases with positive microbial origin -but with normal cellular count-along with 23 cases with CSF pleocytosis with or without positive microbial etiology (this is the overall project findings described in 4 articles -including this one[20,21,22]).‡This information is presented in detail in our previous publication.[20]§This number is the sum of 18 cases with level-1 of diagnostic certainty for aseptic meningitis along with 16 cases with laboratory evidence of non-cultivable microbial etiology (further information is presented in our previous publications.[21,22]‖Four cases with confirmed HSV-1 and HEV DNA/RNA and 12 cases with other confirmed viral nucleic acids (this information is presented in our previous publication.[21]
2.3. Viral DNA and RNA extraction

Viral DNA and RNA were extracted using QIAamp UltraSens virus kit (Qiagen®, Hilden/Germany) following their recommended standard protocol. One mL specimen was used to concentrate viral DNA/RNA. Two aliquots were prepared (30 µL each), one was preserved at −80oC and was used for detecting viral DNA, the second was preserved at −20oC and was used for the preparation of cDNA for viral RNA testing.

2.4. Synthesis of cDNA

The aliquot stored at −20oC was used for converting viral RNA to cDNA using RNasin® Plus RNase Inhibitor product (Promega®, Madison, WI). The enzyme is supplied in a buffer composed of 20 mM HEPES-KOH (pH 7.6), 50 mM KCL, 8 mM DTT and 50% v/v glycerol. The procedure is used to convert up to 5 μg of total RNA or up to 500 ng of poly-(A) RNA into the first strand of cDNA. In brief, per each reaction, 1 μL of random hexameric primer (0.5 μg/reaction) was added into a sterile thin-walled reaction tube containing 4 μL experimental RNA (the template) to reach a volume of 5 μL. The mixture was vortexed using Eppendorf ThermoMixer Compact® (Hamburg, Germany) adjusted at 70oC and 400 rpm for 5 minutes. Immediately, the tube was chilled in ice for at least 5 minutes, centrifuged for 10 seconds in a microcentrifuge using Eppendorf Centrifuge 5415D® (Hamburg, Germany) and 15 μL reverse transcription reaction mix was added to reach total volume of 20 μL per each cDNA reaction. The reverse transcription reaction mix was prepared, freshly for each group of cDNA reactions, by adding the following components into a clean, sterile and RNase-free 1.5 mL Eppendorf tube; 4 μL nuclease-free water, 4 μL GoScriptTM (Promega®, Madison, WI) 5x reaction buffer, 4 μL MgCl2 (final concentration 4 mM), 1 μL PCR nucleotide mix (final concentration 0.5 mM each dNTP), 1 μL Recombinant RNasin® (Promega®, Madison, WI) Ribonuclease Inhibitor (20 units), and 1 μL GoScriptTM Reverse Transcriptase (Promega®, Madison, WI). The 20 μL RNA/RT reaction mix was then placed in the thermo-cycler machine PTC-200 Peltier DNA Engine (MJ Research®, Deltona, FL) and programmed as follows; annealing at 25oC for 5 minutes and extension at 42oC for up to 1 hour. The newly synthesized cDNA was then stored at −20oC. Before proceeding with qPCR, the reverse transcriptase was inactivated by incubating at 70oC for 15 minutes.

2.5. Quantitative real time PCR

Viral genomic DNA (gDNA) and cDNA amplification and detection was performed by real-time analysis using Applied Biosystems® 7900HT Fast Real-Time PCR (Foster City, CA). TaqMan®, (Eurogentec®, Seraing, Liège, Belgium) probe (reporter dye FAMTM on 5′ end and quencher dye TAMRATM on 3′ end) and 2 sets of primers, forward and reverse, were used for each of the 4 target DNA templates (3 gDNAs and 1 cDNA). Primers and probes (DNA Technology®, Aarhus, Denmark) were diluted to reach final working concentration of 25 μM, which is the validated and standardized oligonucleotide concentration used by the department of Clinical Microbiology at Umeå University. Commercially provided oligonucleotide products were diluted to the suitable working solutions, as shown on Tables 3 and 4. Recommendations of (QuantiTect®, Hilden, Germany) QPCR protocols were followed: 3 out of 4 different PCR master mix solutions were prepared similarly for Rune (HSV-1), Sune (HSV-2), and Valle (VZV) primers (Table 3). Per 1 PCR reaction, the following components were added in clean, sterile and RNase-free Eppendorf tubes: 12.5 μL of 2x QuantiTectTM Probe PCR Master Mix (Qiagen®, Hilden, Germany), 1.0 μL of each of forward and reverse primers (final concentration of 25 μM), 0.5 μL probe (25 μM final concentration) and 7.5 μL RNase-free water (Ambion©, Waltham, MA). The mixture was pulse-vortexed and centrifuged briefly (ie, 3000–5000xg). Then the 22.5 μL single reaction mix was transferred into a well of a MicroAmpTM Optical 96-Well Reaction Plate (Applied Biosystems®, Foster City, CA) to which 2.5 μL template gDNA was added to reach total volume of 25 μL per 1 reaction. Experiments were conducted in groups of 10 to 96 reactions; volumes were multiplied by the total number of reactions per PCR run.

Table 3 - Primers and probes specifications for herpesviruses. ♯ Title Specification (A) Herpes simplex virus type-1 F1 Forward primer name Sune-1 Fw  2 Oligo scale 40 nmol  3 No. of couplings 26 bases  4 Sequence 5′ AGA TAT CCT CTT TAT CAT CAG CAC CA 3′ R1 Reverse primer name Sune-2 Rw  2 Oligo scale 40 nmol  3 No. of couplings 17 bases  4 Sequence 5′ TTG TGC TGC CAA GGC GA 3′ Pr1 Probe name Sune-3 Prb  2 Conc. per amount 100 pmol/μL  3 Molecular weight 7471  4 Reporter dye VIC  5 Quencher dye TAMRA  6 No. of couplings 20 bases  7 Sequence 5′ CGG CGG CGT TCG TTT GTC TG 3′ (B) Herpes simplex virus type-2 F1 Forward primer name Rune-1 Fw  2 Oligo scale 40 nmol  3 No. of couplings 18 bases  4 Sequence 5′ GGC CTG GCT ATC CGC AGA 3′ R1 Reverse primer name Rune-2 Rv  2 Oligo scale 40 nmol  3 No. of couplings 17 bases  4 Sequence 5′ GCG CAG AGA CAT CGC GA 3′ Pr1 Probe name Rune-3 Prb  2 Conc. per amount 100 pmol/μL  3 Molecular weight 8794  4 Reporter dye FAM  5 Quencher dye TAMRA  6 No. of couplings 25 bases  7 Sequence 5′ CAG CAC ACG ACT TGG CGT TCT GTG T 3′ (C) Varicella-zoster virus F1 Forward primer name Valle-1 Fw  2 Oligo scale 40 nmol  3 No. of couplings 22 bases  4 Sequence 5′ TCG TGG AAT ATT GAA AGG ATC C 3′ R1 Reverse primer name Valle-2 Rv  2 Oligo scale 40 nmol  3 No. of couplings 21 bases  4 Sequence 5′ CCG TAG GTT CGG CAA ATC TAA 3′ Pr1 Probe name Valle-3 Prb  2 Conc. per amount 100 pmol/μL  3 Molecular weight 7801  4 Reporter dye FAM  5 Quencher dye TAMRA  6 No. of couplings 22 bases  7 Sequence 5′ CCC GTC AAC AGC CGG TCT GTC A 3′
Table 4 - Primers and probes specifications for enteroviruses. ♯ Title Specification (A) Non-polio human enteroviruses F1 Forward primer name NMF1  2 Oligo scale 0.2 μmol  3 No. of couplings 15 bases  4 Sequence 5′ GCC CCT GAA TGC GGC 3′ R1 Reverse primer name NMR1  2 Oligo scale 0.2 μmol  3 No. of couplings 19 bases  4 Sequence 5′ AAT TGT CAC CAT AAG CAG C 3′ Pr1 Probe name MP  2 Conc. per amount 100 pmol/μL  3 Molecular weight 9123  4 Reporter dye FAM  5 Quencher dye TAMRA  6 No. of couplings 26 bases  7 Sequence 5′ CGG AAC CGA CTA CTT TGG GTG TCC GT 3′ (B) Human parechovirus F1 Forward primer name AN353  2 Oligo scale 40 nmol  3 No. of couplings 29 bases  4 Sequence 5′ GAC AAT AGT TTT GAA ATN ACN ATH CCN TA 3′ R1 Reverse primer name AN355  2 Oligo scale 0.2 µmol  3 No. of couplings 37 bases  4 Sequence 5′ CTC CAA CTA TAA TGC CAT ART GYT TRT ARA ANC CGY K 3′

PCR master mix solution for NM (HEVs) primers (Table 4A) was prepared as follows: 12.5 μL QuantiTectTM, (Qiagen®, Hilden, Germany) Probe PCR Master Mix (2x), 1.0 μL forward primer (25 μM), 1.0 μL reverse primer (25 μM), 0.8 μL probe (20 μM) and 7.2 μL RNase-free water were added into 2.5 μL template cDNA to reach a final working solution of 25 μL per single PCR reaction. PCR reaction plates were covered by MicroAmpTM Optical Adhesive Film (Applied Biosystems®, Foster City, CA) and concentrated at the bottom of the plate wells by spinning at low speed in a cold centrifuge (ie, 1200xg) using AllegraTM X-12R Centrifuge (Beckman Coulter©, Brea, CA). Each PCR reaction plate was designed for carrying 8 standard dilutions, 1 negative reverse transcriptase preparation, 1 non-template control containing ddH2O and duplicates or triplicates of each experimental DNA template.

The standard (obtained from the department of Clinical Microbiology, Umeå University) was prepared with the concentrations; 5E6, 5E5, 5E4, 5E3, 5E2, 5E1, 5 and 1 virus copies. Negative reverse transcriptase preparation is a mixture containing cDNA PCR component with the exception of RNA template (replaced by RNase-free ddH2O) and was subjected to the same cDNA PCR conditions. Non-template control is a qPCR preparation in which template DNA is replaced by RNase-free ddH2O. Real-time cycler thermal conditions were similar for all 4 PCRs; heating at 50°C for 2 minutes (for carryover prevention since UNG will eliminate any dUMP-containing PCR products resulting from carryover contamination), followed by PCR initial activation step necessary for HotStarTaq® DNA Polymerase (Qiagen®, Hilden, Germany) at 95°C for 10 minutes, then denaturation at 94°C for 15 seconds and combined annealing and extension at 60°C for 60 seconds. This cycle was repeated 45 times. The duration of final PCR cycles was approximately 90 minutes including generation of the amplification curve. Resulting data were analyzed using software provided by Applied Biosystems® (Foster City, CA) 7900HT Fast Real-Time PCR system (ABI 7900HT SDS Plate Utility Version 2.4).

2.6. Conventional HPeVs PCR

HPeV nucleic acids were investigated using a validated and optimized method reported by W. Allan Nix et al.[19] In brief, 5 μL template cDNA was added to 20 μL reaction mix (12.5 μL of 2x QuantiTectTM Probe PCR Master Mix [Qiagen®, Hilden, Germany], 1 μL of 50 pmol AN353 forward primer, 1 μL of 50 pmol AN355 [DNA Technology®, Aarhus, Denmark] reverse primers [Table 4B], 5.5 μL nuclease-free ddH2O [Ambion©, Waltham, MA]) to reach total volume of 25 μL per 1 reaction. The PCR amplification protocol included 35 cycles; starting at 95oC for 30 seconds, followed by 42oC for 40 seconds and 72oC for 60 seconds, with 0.4oC/s ramp from 42oC to 72oC. A representative clinical isolate for HPeVs (i.e., Echovirus 23) was used as positive control in the PCR procedures. The same recommended PCR amplification protocol was followed using DNA Engine PTC-200 Peltier Thermal Cycler (MJ Research©, Deltona, FL). PCR amplicons were kept at −20oC until processed by agarose gel electrophoresis.

2.7. Agarose gel electrophoresis

1% normal grade melting temperature agarose gel adequate for separating 0.1 to 10 kb DNA fragments was prepared as follows; 1 g agarose (Agarose Standard provided by Saveen Werner ABTM, Malmö, Skåne län, Sweden) was dissolved in 100 mL of 0.5x Tris Borate EDTA (TBE). The 0.5x TBE was prepared as follow; stock solution of 5x TBE concentrate buffer was firstly prepared by dissolving 54 g Tris Base and 27.5 g Boric Acid in 900 mL of deionized water, 20 mL of 0.5 M (pH 8.0) EDTA was added and the overall volume was adjusted to 1 L by adding deionized water. The working 0.5x TBE buffer solution was then prepared by adding 1 part of stock 5x TBE to 9 parts of deionized water. The solution was heated until the agarose was completely dissolved in a microwave (Electrolux®, Stockholm, Sweden) and was then allowed to cool in a water bath set at 50oC for 10 minutes. Gel casting tray was adjusted with 2 combs. After the gel solution was cooled, 5 μL of 10 mg/mL Ethidium Bromide (Bio-Rad®, Hercules, CA) was added and mixed well, then poured into the gel tray and allowed to cool and solidify at room temperature for 30 minutes. Once completely solidified, combs were removed and the gel tray was placed in an electrophoresis chamber (Separation System Inc., OWL®, Gulf Breeze, FL) and covered with 0.5 x TBE buffer. For preparing samples; 2.5 μL of 6x MassRulerTM DNA Loading Dye (Fermentas Life Sciences®, Vilnius, Lithuania) was added to 25 μL PCR amplicons. The 27.5 μL mixture was transferred into the gel wells placed in the electrophoresis chamber. Equal volumes of 2 types of standard DNA Ladders (FastRulerTM Low and Middle Range DNA Ladders [Fermentas Life Sciences®, Vilnius, Lithuania]) were loaded in each run. The agarose gel was electrophoresed at 80 Volts for 75 minutes. DNA bands were visualized by ethidium bromide staining and ultra-violet light box using Ultraviolet Transilluminator device (Spectroline®, Melville, NY). Images were transferred into a monitor (Ikegami®, Tokyo, Japan) and printed in a K65HM-CE High Density Paper (Mitsubishi Electric®, Tokyo, Japan).

2.8. Statistical analysis

We used the statistical package program IBM/SPSS version 21 (Chicago, IL) to analyze all data where numerical variables were organized into categories and expressed in frequencies and percentages. Numerical variables were described using measures of central tendency and of dispersion. Cross-tabulations to compare categorical variables and tests for detecting statistically significant differences (P ≤ .05) and associations (P ≤ .05, r ≥ +/- 0.5) were conducted using Pearson chi-square or Fisher’s Exact along with Phi and Cramer’s V Correlation and its 95% confidence intervals.

2.9. Ethics approval and consent to participate

The ethical clearance for conducting this study was obtained from the Ethical Review Board of the National Center for Neurological Sciences on September of 2009. Patients were not contacted directly; data were obtained from hospital files and were kept anonymous at all stages of the study. Patient consent was considered unnecessary and was waived. Excess specimens were obtained from the hospital main laboratory after all officially requested tests were applied. Permission to collect data and specimens was granted from hospital authorities verbally (from hospital general director) and in writing (from Laboratories Administration, Federal Ministry of Health).

3. Results 3.1. Secondary analysis using standardized clinical codes

The criteria (Table 1) listed in the Clinical Case Definition for Aseptic Meningitis[18] were used to perform secondary-analysis using data collected throughout this project. When all criteria were met, level-1 of diagnostic certainty for aseptic meningitis was designated (Table 1). When viral nucleic acids were detected by PCR, confirmed VM was designated (Table 2). The overall clinical, nucleic acids and conventional microbial findings are illustrated in Figure 1 which describes findings from this study alongside findings reported in our other publications.[17,20,21]

Figure 1.:

Categorization of all cases based on clinical case definition for aseptic meningitis, bacterial meningitis score and molecular and conventional microbial findings.

3.2. General nucleic acids findings

PCR testing for HSV-1 DNA in CSF revealed 2 (0.4%) positive results out of total 503. Average Ct for both HSV-1 positive specimens was 38.67 + 0.96 SD and median virus copy was 2.5 × 102/PCR run (1 × 105 virus copies/mL CSF). PCR testing for non-polio HEVs cDNA in CSF revealed another 2 (0.4%) positive results. Average Ct for both HEVs positive specimens was 38 + 0.4 SD and median virus copy was 3.5 × 102/PCR run (1.4 × 105 virus copies/mL CSF). No genetic materials were detected for HSV-2, VZV and HPeV. On TaqMan-based real-time PCR procedures, standard dilutions with viral copies of 5E6 to 5E2 pr