Cognitive function and risperidone long-acting injection vs. paliperidone palmitate in schizophrenia: a 6-month, open-label, randomized, pilot trial

To the best of our knowledge, the present randomized study is the first attempt to evaluate changes in cognitive function in patients with schizophrenia during RLAI or PP treatment. The most clinically relevant finding obtained by this preliminary study is that patients who switch from RLAI to PP might have higher improvement in attention and processing speed than those who continue treatment with RLAL. Although the differences in effects on cognitive function observed between the two groups were very small, our results might have some implications for the treatment of schizophrenia.

To date, a number of studies have investigated the influence of pharmacological therapy on cognitive function in patients with schizophrenia. Many reports have demonstrated a slightly to moderately favorable influence of antipsychotic treatment on the cognitive function of patients with schizophrenia [30]. However, some studies have reported that cognitive function is negatively affected by standard- and high-dose antipsychotic treatment [31–34], antipsychotic polypharmacy [27], concomitant use of anti-Parkinson’s disease drugs [35], or concomitant use of benzodiazepines [36]. Thus, particular attention to drug doses and combinations is the first measure to avoid cognitive dysfunction. As a further step, the effects of different antipsychotics on cognitive function have been examined by several studies. Differences between the impact of SGAs and first-generation antipsychotics (FGAs) on cognition have frequently been reported [37]. However, studies focused on the possible different effects of different SGAs found both negative [38, 39] and positive results [13, 40, 41]. Given that individual SGAs show different pharmacological profiles and that cognitive function consists of diverse domains, it seems likely that the effects on cognitive function may differ (even if slightly) among drugs [42].

Regarding the present study, the different pharmacokinetic and pharmacodynamic properties between RLAIs and PPs should be considered. The greatest pharmacokinetic difference between the two compounds pertains to the PAL/RIS drug level ratio in the blood. PP works exactly as PAL, while the PAL/RIS blood ratio during RLAI treatment (both 25, 37.5, and 50 mg/14 days) was reported to be 2.4–3.0 [43] and much higher during Oral-RIS treatment (2 mg/day: 13.3, 4 mg/day: 11.5, 6 mg/day: 13.3). One possible reason for such a low PAL/RIS for RLAI is the lack of a first pass effect in the liver. This suggests that RLAI is a drug that more strongly reflects the influence of RIS compared to PP and Oral-RIS.

Under the pharmacodynamic point of view, PAL and RIS show many similarities in their affinities for receptors. However, their affinities for the alpha 2-adrenergic receptor are different, since PAL has a higher affinity than RIS for this receptor [12]. Animal studies have shown that the blockage of alpha 2-adrenergic receptors greatly affects the release of dopamine and noradrenaline from the medial prefrontal cortex [44]. Clinical studies have demonstrated that treatment with antidepressants with strong alpha 2-adrenergic receptor antagonistic activity (e.g. mirtazapine [45] or mianserin [46]) may improve cognitive function. Furthermore, many studies in humans, monkeys, and rodents have demonstrated that the noradrenergic nervous system, particularly through the alpha 2-adrenergic receptor, contributes to the regulation of attention [47]. Given that attention is a markedly disturbed cognitive domain in schizophrenia, the results of the present study deserve to be investigated by future studies. The results of a previous study [13] that compared Oral-RIS and Oral-PAL are poorly comparable with the present ones, since the reported pharmacokinetic differences among RLAIs, PPs, Oral-RIS, and Oral-PAL. In addition to these differences, the following issues should be considered: (1) the range of fluctuations in blood levels of Oral-RIS and Oral-PAL differ substantially (peak-to-trough fluctuation: 1.47 for Oral-PAL, 3.30 for Oral-RIS) [48]; (2) the frequency of sedation is higher with Oral-RIS than with Oral-PAL [49]; and (3) Oral-RIS may adversely affect working memory, a cognitive functional domain related to short-term memory [50].

In the present study, there were no differences between the two groups in the change in the total scores of the PANSS and DIEPSS. These results suggest similar efficacy and tolerability of RLAI and PP, consistently with previous studies [51–53].

The total score of the SWNS did not differ between the two groups, and there were no correlations between the changes in the total SWNS score and the BACS items. Thus, the subjective well-being of patients with schizophrenia did not differ between RLAI and PP, despite the higher improvement in some cognitive domains observed in the PP group. However, a correlation between emotional regulation (a SWNS subscale) and attention/processing speed improvement during treatment was found. Gross et al. assumed a model of emotion regulation that includes five phases and hypothesized that attention plays an important role in this process [54]. The event-related potentials measured by an electroencephalographic study and a functional magnetic resonance imaging study demonstrated a relationship between attention and emotional regulation [55, 56]. In addition, numerous reports have examined the dysfunctional interactions between attention and unpleasant stimuli, which are particularly interesting in regard to schizophrenia [57]. However, the current evidence appears still insufficient to state that improvements in attention during the pharmacological treatment of schizophrenia can result in improvements of emotional regulation [57]. In addition, the SWNS subscales may be greatly affected by the total score and they may have low reliability [58]. Thus, further studies focused on this issue are needed.

This study has several limitations. First, the small sample size increases the risk of false negative findings. The lack of multiple-testing correction (e.g. Bonferroni correction) may also result in type I errors, but given the pilot nature of the present study, the results should be considered preliminary. Therefore, replication studies are needed, possibly with larger samples and through a randomized, double blind design. Second, the open-label design might have affected the results, since the expectations of patients or raters might have affected the assessments. However, the similar effects of the two drugs on symptomatology scores do not support this bias and the randomized design represents a point of strength. Third, we did not consider the drug blood level at the time of the cognitive assessment or of the other assessments. The drug blood level may have affected these evaluations [31], despite drug level fluctuations for LAIs are considerably smaller than for oral antipsychotics [59]. Finally, we did not strictly limit the use of anticholinergic agents, which may have an impact on cognitive function. However, only one patient in the RLAI group used anticholinergic agents, with probably no impact on the present results. Future studies should fix the timing of cognitive assessments and strictly control the use of concomitant medications.