Page 417 - HIVMED_v21_i1.indb

Page 4 of 6 Original Research

diagnosed with HIV encephalopathy by the Centers for Most children (76%) with epilepsy were classified as stage
Disease Control and Prevention (CDC) criteria, where 3 or 4 according to the WHO staging system for HIV/AIDS
25
neurotoxic effect of HIV was considered as the cause of at diagnosis of HIV. Thus, most of the children with
epilepsy. In one child with spastic quadriplegic cerebral epilepsy were diagnosed at an advanced stage of HIV
palsy, it was assumed that intra-partum hypoxic ischaemic infection. In contrast, 80% of the children in the group
encephalopathy caused the epilepsy as the child developed without epilepsy were assessed as stage 1 at diagnosis. (see
seizures right after birth. Two children presenting with Table 1 and Figure 1)
cerebrovascular accidents, both caused by persistent severe
thrombocytopenia, developed epilepsy thereafter. The In about half of the patients with epilepsy (48%), treatment
average duration between CNS infection and diagnosis of for HIV infection was initiated at the time of diagnosis.
epilepsy was 11 months (range: 1–24 months). In only one Twenty-eight per cent had treatment initiation months, and
case, the diagnosis of epilepsy was made prior to CNS sometimes years after HIV diagnosis, probably because of
infection. For more than one-third of children (37%), no cause the guidelines for that specific time period. For the rest, the
could be found. exact date of diagnosis of HIV infection was not available,
which is another limitation of this study.
Features of neurodevelopmental delay, as assessed by the
medical officer in the antiretroviral therapy (ARV) clinic, None of the children with epilepsy and stage 1 HIV infection
were present in 17 (34%) of the epileptic children. Several had prior CNS infection, while 81% of the children with
children were referred either for assessment in the epilepsy and stage 4 HIV infection had prior CNS infection.
neurodevelopmental clinic or to the occupational, physio- or Thirty-seven (75%) of the children were diagnosed with
speech therapist. One of the eight children seen in the epilepsy before or at initiation of ART. Retrospectively, it was
neurodevelopmental clinic was diagnosed with hemiplegic not possible to differentiate in how many children the
cerebral palsy post-stroke, one with quadriplegic cerebral complaint of seizures was the cause of further investigations,
palsy because of hypoxic ischemic encephalopathy, two with followed by the discovery of HIV infection. These findings
might nevertheless indicate not only that many children were
speech and cognitive impairment of unknown cause and only diagnosed as HIV-positive when presenting at the
four with HIV encephalopathy.
hospital with seizures or CNS infection, but also that some of
them did not yet qualify for ART according to the specific
School failure, school problems, reports from the educational HIV treatment guidelines in place at the time. This led to
psychologist of intellectual impairment or the notice of progression of HIV infection in these children and made
attendance of a special school was noted for 27 (55%) of the them vulnerable to opportunistic infections and increased
epileptic children (70% of whom were boys). the risk of developing epilepsy. The above findings also
19
support the findings of a prior study, where early initiation
Most children were diagnosed with epilepsy before or at of ART was assumed to be protective for epilepsy and
the time of diagnosis of HIV infection (35/49). Almost all seizures.
(48 of 49) children with epilepsy were treated with sodium
valproate; 18 received other antiepileptic drugs either before The most common specific aetiology for epilepsy was a prior
sodium valproate or as a dual- or multi-drug regime, CNS infection, with meningitis caused by Mycobacterium
including phenobarbitone (n = 16), carbamazepine (n = 2), tuberculosis being the most frequent. Again, CNS infection
clonazepam (n = 2), lamotrigine (n = 1) and/or ethosuximide was probably because of late diagnosis of HIV infection,
(n = 1). The two children treated with carbamazepine were and/or late initiation of ART, making the children vulnerable
referred from other health facilities on carbamazepine to opportunistic infections. Human immunodeficiency virus
and were changed to sodium valproate in our clinic. Half neurotoxicity was the second most common suspected cause.
(24/49) of the children became seizure-free with the use Those data correlate well to those of Bearden et al. in
19
of antiepileptic drugs and eight (16%) had a significant Botswana. Two children had infection-independent strokes
(50% – 75%) reduction in seizure frequency.
Discussion Stage 1 Cohort Epilep c pa ents

In this retrospective survey, we were able to show a 2.5% Stage 2
prevalence of epilepsy in children with HIV infection, which WHO stages
is similar to other areas in South Africa and sub-Saharan Stage 3
Africa, as well as in India. 19,20,21 This number is about two to Stage 4
three times greater compared to the overall prevalence of 0 10 20 30 40 50 60 70 80 90
epilepsy in Africa and South Africa, which is estimated % for epilep c parents and cohort at
between 0.73% and 1.2%. 13,28,29 However, this prevalence diagnosis of HIV infec on
might still be underestimated as the data were retrospectively WHO, World Health Organization; HIV, human immunodeficiency virus.
collected from case notes, which is a key limitation of FIGURE 1: World Health Organization stages in percentage of total number of
this study. patients (blue) and in percentage of total number of epileptic patients (green).

http://www.sajhivmed.org.za 409 Open Access