Cognitive impairments in Polish children and adolescents with perinatal HIV infection

Summary Aim. The aim of the study was to assess the presence of cognitive impairments in children and adolescents with vertically transmitted HIV infection and to determine possible relationships with clinical and sociodemographic variables. Method. The study included: 50 children with perinatal HIV infection aged 6–18 years (PHIV+), 24 healthy children perinatally HIV-exposed but uninfected (PHEU) and 43 healthy children of uninfected parents (HIV-nA). CANTAB Research Suite was used to assess cognitive impairments. Results. In comparison with the HIV-nA group, the PHIV+ group scored significantly lower in shifting of attention, cognitive flexibility, use of feedback, and working memory. In comparison with the PHEU group, the PHIV+ group had significantly longer planning time while performing the task. The analysis of results for the 12–18 year-old age group revealed deterioration of cognitive functions in all tests of the PHIV+ children in comparison with the HIV-nA group. A higher logarithm of viral load at the start of the ARV treatment was associated with lower results in the use of feedback, shifting of attention, cognitive flexibility, and worse information processing. Conclusions. Results of the research indicate deterioration of executive functioning in the PHIV+ group associated with longer duration of HIV neuroinfection and severity of infection before treatment.


Introduction
In children and adolescents with perinatal HIV (human immunodeficiency virus) infection, cognitive functioning may be affected by fetal factors (HIV infection, medications taken by the mother, drugs, alcohol, malnutrition, and other infections and diseases), HIV course-related factors and environmental factors (cognitive and educational levels of the parents, socioeconomic status, the presence of illness, the experience of loss of a parent, etc.) [1]. In the majority of studies on perinatally HIV-infected (PHIV+) subjects, general intellectual functioning tests have typically been used as indicators of neurodevelopmental cognitive deficits. The general result of the examination of the intellectual level within population norm levels does not fully reflect the cognitive functioning [2,3] mainly due to the fact that the tools used for this purpose are not adapted to the assessment of subtle abnormalities, e.g. related to the functioning of the prefrontal cortex (e.g., responsible for executive functions) [1].
Importantly, executive functions deficits in PHIV+ children with no symptoms of neuroinfection were found to be independent of infection stage [2,8]. Some authors claim that the presence of executive functions deficits applies to all PHIV+ children and that they serve as early markers of the CNS HIV infection, despite the absence of clinical symptoms, found in most of the respondents. This seems to be a consequence of the implementation of HAART therapy, owing to which encephalopathy (diagnosed according to the Centers for Disease Control and Prevention [19] definition) ceases to be the expected consequence of HIV neuroinfection, and executive function disorders become such a consequence [2,6,8,18].
Cognitive functioning is primarily affected by the infection's course. Performance on tests of cognitive functions in children and adolescents with AIDS (acquired immune deficiency syndrome) acquired before the age of five is much worse despite improvement in clinical status later in life [7,14,18,20]. Problems regarding memory and visuospatial functions were most often observed in these groups [8] as opposed to children with no class C infection, who had normal results, similar to HIV-exposed but uninfected children (PHEU) [7].
The results of a study by Koekkoek et al. [2] testing possible relationships between infection-related clinical factors and executive function deficits showed better results regarding attention and working memory domains for children with higher CD4 lymphocytes level at the beginning of treatment and longer ARV (antiretroviral) treatment time. The results of other studies have confirmed that high viral load and low CD4 lymphocytes level are predictors of neurocognitive disorders. This has been explained by worse CNS renewal potential compared to other systems but also less effectiveness of HAART treatment in the CNS domain [6,12,14,18,21,22]. Although the consequences of ARV therapy on the developing brain are yet to be fully recognized, the current guidelines recommend starting ARV therapy right after diagnosis of the infection [23,24]. It seems that early introduction of treatment that has a positive impact on a child's immunological status may be beneficial for cognitive development [25]; however, not all studies have confirmed this observation [12,26].
The expression of cognitive impairments in PHIV+ children is also influenced by environmental factors [6,14,27,28], yet their interaction with clinical factors is not entirely clear. Llorente et al. [29], analyzing the cognitive functioning of PHIV + children, showed that the differences ceased to be significant after taking into account the influence of many environmental factors (such as race, maternal education, family income, prematurity, low birth weight, alcohol and drug use during pregnancy). The formulation of ultimate conclusions about the cognitive functioning of PHIV+ children is limited by the small number of studies, group heterogeneity and the diversity of assessed cognitive functions, which justifies further research.
This study aimed to assess the presence of cognitive impairments in children and adolescents with vertically transmitted HIV infection and to determine possible relationships with clinical and sociodemographic variables.

Material
The study group included 50 patients with perinatal HIV infection, aged 6-18 yyears, from the Department of Children's Infectious Diseases at the Medical University of Warsaw. The exclusion criteria included other routes of HIV infection, the presence of a chronic somatic condition other than HIV infection, the presence of any problem that could limit cooperation (inability to communicate in the Polish language), and the lack of child or caregiver consent. Two age -and sex-matched control groups were formed. The first reference group consisted of 24 healthy children and adolescents with HIV-infected mothers (perinatally HIV-exposed but uninfected; PHEU), who were recruited as ex-patients of the Department of Children's Infectious Diseases at the Medical University of Warsaw and clients of non-governmental organizations helping people infected with HIV. The second reference group consisted of 43 healthy children and adolescents with uninfected parents or without a history of HIV infection in family members (HIV-non-Affected; HIV-nA), who were recruited through social networking websites and in Warsaw schools.

Methods
CANTAB Research Suite (Cambridge Neuropsychological Test Automated Battery; CANTAB) [30] was used in the assessment of cognitive functions. CANTAB is language-independent, culturally neutral, non-invasive, and require no technical knowledge or prior familiarity with computers. The CANTAB battery has never been used before in studies with children and adolescents with HIV infection (it was used in HIV-infected adults). However, it is a tool commonly used in other pediatric populations (e.g., in studies in children with ADHD).
Due to the lack of other studies on the Polish population and small samples, no comparisons with population norms were made in the current study. Untransformed scores were used in the statistical analyses.
The following tests were used: 1. MOT (Motor Screening Tasks), to introduce the study procedure. Its outcome measures include (1) the latency -speed of response between the presentation of the stimuli and the response; and (2) the mean error -which refers to the accuracy of touch, measured as a distance between the touch point and the centers of the stimuli. smaller in cases when the task is not completed; in such a case, 25 errors are added to make comparisons with participants who completed the task; and (6) total trials (adjusted) -the number of deals for each level completed; with 50 deals added for each uncompleted level. Additionally, some analyses use parameters from level 9 (errors, latency, trials) as a measure of extradimensional reversal learning. 5. SWM (Spatial Working Memory) -a measure of the ability to update, refresh, and maintain spatial location information and maintain this data in working memory. The outcome measures include (1) between error -defined as selecting boxes that have already been found to contain a token in the ongoing task; and (2) strategythe assessment of strategy based on the number of new box-scanning sequences on the most difficult level (six and eight boxes), with a high score (multiple sequences starting with different squares) reflecting weak strategy use. During the initial interview, using an original survey filled in by a legal guardian, we collected basic information about the participant and the family, as well as information about environmental and sociodemographic factors. Participation in the study was voluntary. Informed consent forms were signed by both a child and a legal guardian prior to the study. The study was approved by the Ethics Committee of Medical University of Warsaw (approval no: KB 4/2012).
Descriptive statistics include medians and interquartile ranges (IQR). Due to the limited sample size and significant differences in this aspect between the compared groups, as well as the strong skewness of a number of outcome measures and the presence of outliers, the Mann-Whitney U test was used for all intergroup comparisons for continuous variables. Relationships between continuous variables were assessed using the non-parametric Spearman's rho correlation coefficient. The Chi-square or Fisher's exact test were used to assess the significance of differences in the distributions of dichotomous variables. p-values below 0.05 were considered statistically significant.

CANTAB
Further analyses were performed for the whole sample, as well as separately for the 6-11 and 12-18 age groups. The decision was motivated by the wide age range in the experimental and reference groups and significant differences in median age between the PHIV+ and PHEU groups. Data were obtained for n = 50 people from the PHIV+ group, n = 24 from the PHEU group and n = 43 from the HIV-nA group. The CANTAB test results and intergroup comparisons are presented in Table 4. Significant group differences were found both in the analysis for the whole sample and in the 12-18 year-old subgroup.

Clinical and sociodemographic factors
Due to the small subgroup sizes, analyses of relationships between the CANTAB results and the clinical and sociodemographic factors were carried out on a 6-18 year-old group. Significant differences in CANTAB results were found in groups differentiated by sex, awareness of diagnosis and ARV treatment initiation (Table 5).  The following statistically significant relationships were found between the results obtained in the CANTAB tests and the child's age at the time of the study: (5) CD4% on ARV treatment initiation and MOT mean error (rho = 0.59; p <0.001); IED completed stage errors (rho = -0.30; p = 0.040); errors block 9/total errors stage 9 (rho = -0.33; p = 0.023); total latency stage 9 (rho = -0.34; p = 0.021) and total trials stage 9 (rho = -0.34; p = 0.019); (6) CD4n/l at the time of the study and IED stages completed (rho = -0.32; p = 0.024); errors block 9/total errors stage 9 (rho = -0.38; p = 0.007) and total latency stage 9/Total trials stage 9 (rho = -0.39; p = 0.006).

Discussion
Our study was the first in Poland to evaluate the majority of members of the HIVinfected population aged 6-18 years. The CANTAB results were analyzed for two domains: (1) information processing speed (MOT and RTI tests); and (2) executive functions including working memory, visuospatial memory, planning, control, cognitive flexibility, abstract reasoning, reversal learning, goal-directed activity, and modulation of impulsive reactions (IED, SOC, SWM tests).
Impairments in cognitive functioning were found in all tests in the PHIV+ group: the results were significantly worse for working memory and movement initiation time compared with the HIV-nA group, as well as longer in-task planning time in comparison with the PHEU group. In comparison with both the HIV-nA and PHEU groups, the PHIV+ group had significantly worse results in attentional switching, cognitive flexibility and reversal learning. The results in the group aged 12-18 years indicated worse speed processing and executive functioning in the PHIV+ group in comparison with the HIV-nA group, in all tests. As executive functions develop and mature with age [31], it is therefore important to emphasize the results obtained by adolescents. This issue is also pointed out by Nichols et al. [14,32] and Hunter et al. [14,32], who emphasize that executive function disorders are especially visible in adolescence, hence the correct results obtained in younger age groups cannot be interpreted without reservations as evidence of the absence of cognitive deficits. Our observations are in line with the results of a study conducted by Cohen et al. [13], where lower processing speed was observed in PHIV+ children over 12 years of age in comparison with healthy subjects.
The negative impact of HIV infection on processing speed may be observed irrespective of the patient's immunological condition [2,14,32]. This is linked with the occurrence of HIV encephalopathy as well as the time-point at which the lowest level of CD4 lymphocytes coincided with the highest viral load. Worse cognitive functioning of adolescents from our experimental group in comparison with healthy subjects suggests that infection duration adversely affects executive functions. On the other hand, the conclusion can be drawn that, despite the negative influence of HIV, there was a developmental potential in patients with congenital infection (as evidenced by the positive correlation between age and CANTAB results in our sample), such that the quality of performance in cognitive tests was better in older participants.
The results of the current study did not confirm significant differences between children diagnosed and undiagnosed with encephalopathy observed in other studies [5,9,14]. Additionally, no significant differences or relationships were found according to HIV clinical class, including class C. The observation is not in line with the findings of other authors who reported cognitive deficits in children with AIDS, irrespective of improved health condition [7,8,14,18,20,22,34].
Contrary to other studies [25], no definite protective role of early (before the age of 3 or 12 months) ARV treatment initiation was observed in the current results with regard to cognitive functions. Our results showed worse reversal learning, attention shifting and cognitive flexibility in children provided with ARV treatment before three months of age in comparison with a group who initiated treatment after that time. We assume that it could have been a group of children whose early treatment was due to the rapid expression of HIV symptoms immediately after delivery (or symptoms observed in pregnant mothers), which suggests greater severity of infection and, thus, a more negative effect on the development of the CNS. Furthermore, an analysis of the relationships between CANTAB test results and clinical factors showed that higher viral load values at ARV treatment initiation correlated with slower information processing, weaker planning, reversal learning, and cognitive flexibility. In view of the foregoing, it can be concluded that cognitive functioning is influenced not only by the current health condition but also the course of HIV infection. Similarly, Koekkoek et al. [2] emphasized the importance of the child's immunological condition before initiating ARV treatment and its duration as variables which determined the quality of, among others, working memory functioning. Observations suggest the decrease in CNS regeneration potential, despite successful ARV treatment [6,12,14,18,21]. On the other hand, Martin et al. [5] demonstrated that processing speed is better in children in better actual immunological condition.
The results of studies on sex-related speed processing are inconclusive -some results suggest higher processing speed in younger girls [35] and some in boys. Significant sex differences are observed in adolescence [31]. In our study, girls from the PHIV+ experimental group had worse movement times, which influenced processing speed in the RTI test. However, no negative impact on other cognitive tests was observed.
There are several important limitations of the study. The first is the lack of selection of control groups according to socio-economic status, which should be accounted for in the context of observations of Cohen et al. [13]. Other limitations which precluded extending the analyses beyond one-factor relationships was the small size of experimental and reference groups. Because more complex analyses could not be performed, the current study should be regarded as descriptive. The size and age of the PHEU group, which was significantly younger than the experimental group, is another important limitation. Furthermore, the mothers of children who were exposed to the infection received ARV treatment when pregnant, which raises questions about the influence of this procedure on future cognitive functioning. Moreover, the CANTAB test battery has never before been used in the assessment of children and adolescents with congenital HIV infection. The tools are also not commonly used in Poland, which limits comparisons to the results of other studies or norms. An unsuccessful attempt (related to the sample size and sex distribution) was made to compare the results to population norms delivered by CANTAB software. For the above reasons, only intergroup comparisons were performed. Comorbid HCV infection diagnosed in seven children in the experimental group was a potential factor that could have influenced the cognitive functioning.

The cognitive functioning of adolescents with vertically transmitted HIV infection
was worse compared to healthy subjects with uninfected parents. 2. A more severe course of HIV infection and its severity before starting treatment is associated with a greater severity of deficits in executive functions.