Epstein-Barr virus infections in patients with acute febrile illness in Kenya
Abstract
Epstein Barr Virus (EBV) has emerged as an important etiologic agent of infectious
mononucleosis and still-growing number of lymphoproliferative disorders. It is also a potential
cause of acute febrile illness (AFI). At present, the viral-prevalence of EBV and its association
with acute febrile illness have not been adequately studied in Africa. This study sought to
catalogue the prevalence and viral loads of EBV in patients with acute febrile illness in Kenya.
A voluntary hospital based cross-sectional study of EBV infection was conducted between
September 2008 and April 2013 on 2107 patients presenting with AFI at eight out-patient clinics
across Kenya. EBV was detected in whole blood by a TaqMan-based real-time PCR for the
BALF5 region of EBV genome. The primers and probes used were obtained from commercial
sources (Applied Biosystems, Foster City, CA, USA), and the positive control was purchased
from Vircell Microbiologists, Granada, Spain. Prevalence rates and viral load level were
determined and examined according to demographic characteristics, region of residence, malaria
endemicity and malaria infection. EBV was detected in 29.1% (614/2,107) of the recruited AFI
patients. Their viral load ranged from 52 to 7.19 x106 copies/mL (mean 35,699 copies/mL;
median 4,699 copies/mL). Viral prevalence was highest in patients aged below 5 years but
reduced with increase in age. Of the 614 positive patients, 443 (72.1%) had viral load above the
2,000 copies/mL cutoff that is considered clinically important. Again the <5 year olds constituted
the majority in this group (p<0.05). Regionally, the Lake Victoria and the coastal regions were
found to have higher prevalence rates compared to the highland of Kisii and semi-arid areas (p=
0.1636). Of the following clinical features that were commonly associated with AFI, chills,
cough, sore throat, headache, runny nose, abdominal pain, joint aches and muscle pain, EBV
viral load significantly associated with runny nose (p=0.0006). Regarding malaria endemicity,
incidences of EBV infection were higher (35.7%) in regions considered holoendemic for malaria
(L. Victoria basin and coast) compared to (22.9%) for hypoendemic regions (highland of Kisii
and semi-arid areas) (p<0.05). Finally, comparison of viral loads between AFI patients coinfected
with EBV and malaria and those with EBV infections alone revealed higher viral loads
in the co-infected group (5,329 copies/mL compared to 3,655 copies/mL, p= 0.004), and this
difference was much more pronounced in the 5-9 year olds. The study demonstrates that EBV is
a common infection among patients with AFI; hence the need to create more awareness among
healthcare workers that EBV can be an important cause of AFI. The age, malaria endemicity, and
infection with malaria influence EBV prevalence and viral load levels. Like other AFI, clinical
features are not helpful in arriving at EBV diagnosis. The findings further suggest the need to
investigate whether increased EBV viraemia in EBV-malaria co-infections among the 5-9 year
olds is a precursor to the development of Burkitt lymphoma.
Key words: Epstein-Barr virus, acute febrile illness, viral load and Malaria
Publisher
University of Nairobi
Description
Thesis