**3. Results**

### *3.1. Original Studies Evaluating Cognitive Assessment Tools for Pediatric MS*

The workflow of this scoping review is shown in Figure 1. We identified eleven original papers evaluating the performance of cognitive assessment tools in the context of pediatric MS, with the earliest one published in 2009 [13]. A summary of these articles is provided in Table 1. These studies were largely cross-sectional in design. They were either single- or multi-centric, and all were carried out in North America and Europe. The focus of these studies di ffered from one another. Some studies tried to establish normative data using regression-based approach [14–16], in which age-squared variable could be incorporated to better model the nonlinear quality of cognitive development [15,16]. Others investigated the performance of various cognitive assessment tools through comparisons between patients and healthy controls [7,13,15–21]. Among these, two studies aimed to construct batteries by picking up three to four tests with better discriminating abilities, and evaluated the performance of these batteries as screening tools [13,21]. Participants' satisfaction with the test was quantitatively reported in a study [17]. One study examined the interrelationships between tests tapping di fferent cognitive domains [8]. A recurring finding yielded by these studies was the significant role of age and educational level in cognitive task performance in pediatric populations [14,17,21]. For instance, the symbol digit modalities test (SDMT) performance steadily improves with age in healthy children (8–17 years) [17]. On the other hand, older age predicted poorer SDMT performance in POMS after adjustment for disease severity (i.e., the expanded disability status scale, EDSS) [18]. Together these sugges<sup>t</sup> divergent cognitive trajectories between normal children and pediatric patients with MS.

**Figure 1.** Diagram showing the workflow of literature search and selection process.


*Children* **2020**,

 *7*, 183





children—fourth edition; WLG, word list generation; WM, working memory.

### *3.2. Original Studies Evaluating Cognitive Rehabilitation for POMS*

We identified three original studies evaluating the e ffects of cognitive rehabilitation for POMS [22–24]. Their study design was summarized in Table 2. All three studies comprised interventions using home-based, computerized cognitive training. The targeted cognitive domains were working memory [22,23] and attention [24], respectively. The duration of a single training session was similar across these studies (45 min to 1 h), while the intensity varies from twice to five times per week. These studies were all pilot and exploratory in nature. The sample size was small (5–16 patients) and was not preplanned based on power analysis. No healthy control group was included in these studies. The study on attention retraining was a double-blind randomized clinical trial, using nonspecific training as the comparator arm. There was no comparator arm in the other two studies, hence the role of practice e ffect in neuropsychological evaluation cannot be clarified. The outcome measures included not only targeted cognitive function but also more extensive neuropsychological performance [23,24], and aspects of feasibility (adherence and tolerance to the training program) were evaluated as the main outcome in one study [23].

The results of these studies were summarized in Table 3. The study on attention retraining showed not only positive e ffects on attention and related cognitive domains, but also far transfer e ffect on visuospatial memory [24]. On the other hand, the other two studies, both focusing on working memory training, showed only modest e ffect on objective working memory measures, at group level. The far transfer e ffect was either inconspicuous [23] or not assessed [22] in these two studies. Hubacher et al. demonstrated that the training e ffect was sustained for nine months in both responders [22]. The sustainability of training e ffect was not assessed in the other two studies. Reported factors associated with outcomes of cognitive rehabilitation include various measures of disease burden, normalized brain volume, and general intelligence [22,23]. These are generally in line with the theory of brain reserve and cognitive reserve [25].

### *3.3. Registered Clinical Trials Primarily Focusing on Cognitive Issues in POMS*

We searched on ClinicalTrials.gov, European Union Clinical Trials Register, and Open Science Framework database for pertinent trials or projects on 4 March 2020. Only three trials were considered most relevant. One trial aimed to explore the electrophysiological mechanisms underlying cognitive dysfunction in pediatric MS. The other was a randomized controlled trial assessing the e fficacy of a home-based computerized program for retraining attention in pediatric patients with MS. These two trials were completed, and the results of the latter one has been published and included in the present review [24]. The third one is an ongoing randomized clinical trial aiming to assess the cognitive impact of a virtual reality videogame exercise program. More information about these trials is summarized in Table 4.




**Table 3.** Original studies evaluating cognitive rehabilitation for POMS: Summary of results.


**Table 4.** Registered clinical trials primarily focusing on cognitive issues in POMS.
