In the current subgroup analysis of a randomized clinical trial among Chinese patients with OSA-EDS, solriamfetol treatment displayed robust efficacy in managing OSA-EDS, as evidenced by the consistent prolonged MWT sleep latency and decreased ESS scores among three subgroups stratified by baseline AHI. Moreover, similar safety profiles were detected among subgroups after solriamfetol administration, which supports the conclusion that solriamfetol demonstrates reliable efficacy and a well-tolerated safety profile in patients across different baseline AHI levels.

Of note, among the available wake-promoting agents for addressing residual EDS in patients with OSA, a network meta-analysis based on 18 studies showed that solriamfetol achieved the greatest improvement in ESS scores and sleep latency compared with pitolisant, modafinil, and armodafinil [18]. Findings from the study by Cheng et al. further indicated good efficacy and acceptable safety profile among Chinese patients with OSA and EDS [20]. Our study group recently conducted a subgroup analysis based on different adherence levels to primary OSA therapy, showing that solriamfetol efficacy and safety was not affected by usage of primary OSA therapy [21], which is quite important considering the real-world complexity of OSA management. Further, our current study provides evidence to solve another clinical question regarding the complexity of disease severity.

The baseline AHI-based stratification in this study may not fully reflect the complex pathophysiology and intrinsic severity of OSA [24], owing to confounding factors including CPAP adherence, nocturnal variability, and scoring bias. Nonetheless, it retains practical reference value for guiding the clinical identification of treatment responses across distinct patient subgroups. Besides, findings from another post hoc analysis of the solriamfetol clinical trial showed consistent AHI levels across the 12-week study duration [25], which, to some extent, indicates that stratification by baseline AHI adequately captures the variability in disease severity during the randomized controlled trial, thus supporting the validity of subgroup analyses assessing treatment efficacy across different OSA severity levels.

Most of the patients with baseline AHI < 5 had well-controlled OSA symptoms with CPAP yet still experienced residual sleepiness, which reflects the wide range of real-world patients with residual EDS. A recently conducted prospective study reported that as high as 28.2% patients with OSA still had EDS even after a medium duration of 5 months treatment with CPAP [26]. Consistent with previous studies [27, 28], solriamfetol demonstrated efficacy in reducing EDS in this subgroup cohort by significantly improving MWT sleep latency and decreasing ESS scores, suggesting its potential as an adjunctive therapy to CPAP and highlighting its prominent value in managing patients with post-treatment residual symptoms.

Of note, patients with OSA and AHI levels ranging from 5 to 30 are highly prevalent and account for about 5.4–45.7% of the general OSA population [29]. In the parent randomized controlled trial, a notable proportion of patients with mmOSA (N = 69, 34.5% of total cohort) were included for efficacy and safety evaluation, of which around 60% patients had poor or no adherence on any primary OSA therapy. This is consistent with findings that CPAP is not universally recommended, well accepted, or adherent among patients with OSA of milder severity. Notably, although mmOSA is reported to be associated with increased risk of incident hypertension in a previous study [5], it was found that conventional therapy such as CPAP is not always successful with mmOSA, especially in the presence of non-anatomical traits for patients [30]. The absence of compelling evidence for cardiovascular morbidity and mortality improvement by CPAP in patients with mild OSA [13] also contributed to the poor adherence of CPAP primary therapy. It was once reported that up to 47% of the CPAP-treated patients with OSA quit treatment by 3 years [31]. Solriamfetol was found to be effective and well tolerated for patients with OSA and baseline 5 ≤ AHI < 30 in the current analysis, as evidenced by a significant prolongation in MWT duration and a reduction in ESS scores. Given the poor adherence to CPAP therapy among patients with milder severity of OSA, wake-promoting agents such as solriamfetol might also serve as a possible supportive therapeutic option for sleepy patients with OSA and low AHI levels to improve their quality of life [24].

The current analysis included around 23% of patients who had baseline AHI ≥ 30, most of whom were not on any primary OSA therapy. Such patients were included to better reflect real-world practice in the management of OSA [20]. Similar trends in efficacy and safety of solriamfetol were also found in this subgroup. Specifically, small numerically increases in heart rate and blood pressure compared to placebo were observed in this subgroup, which is also consistent with previous results from other subgroups. Taken together, these findings suggest that solriamfetol may serve as a symptomatic supportive treatment for OSA-EDS with different baseline AHI levels. It is important to note that solriamfetol should not be considered as a substitute for therapy targeting airway problems. Especially for patients had AHI higher than 30, the use of primary OSA therapy should be highly recommended. Physicians should accordingly reinforce the critical role of primary OSA interventions and encourage consistent patient adherence.

Similar to prior subgroup analyses of other factors [19, 21, 22], the efficacy and safety profile of solriamfetol was also consistently observed in subgroups of OSA severity stratified by baseline AHI. The improvement in MWT sleep latency exceeded the clinically meaningful threshold of 4 min in all subgroups [32]. Meanwhile, the magnitude of improvement in ESS scores with solriamfetol vs. placebo was relatively smaller and statistically non-significant in the subgroup with an AHI ≥ 30. This finding may be attributable to the limited sample size and a stronger placebo effect observed for this subjective outcome measure compared with objective indices. However, caution should be exercised regarding potential cardiovascular risks (e.g., mild pulse rate elevation), and routine monitoring of blood pressure and heart rate is recommended for patients with severe OSA-EDS receiving solriamfetol. Notably, solriamfetol represents a valuable adjunct to CPAP therapy in this patient population, thereby optimizing the comprehensive management of OSA-EDS.

A key strength of this study lies in its simulation of real-world clinical scenarios commonly encountered by clinicians in the management of patients with OSA-EDS, including patient subgroups stratified by primary therapy receipt, treatment adherence status, and the extent of therapeutic improvement achieved. In addition to AHI, a recent study indicated that elevated arousal threshold may contribute to EDS by prolonging respiratory events, exacerbating intermittent hypoxia, increasing sleep pressure, and disrupting architecture [33]. One potential future research direction would be stratified analysis of solriamfetol based on the baseline arousal index.

This study has several limitations. First, the baseline AHI levels of the current study may not reflect the original disease severity because of confounding by different adherence to primary OSA therapy. The mild (5 ≤ AHI < 15) and moderate (15 ≤ AHI < 30) OSA subgroups contained inadequate sample sizes for independent testing; we therefore merged these two strata for pooled statistical analyses in the present study. Further comparison is warranted in future large-scale studies. In addition, as a result of the small sample size in each subgroup, a further sensitivity analysis by adherence to primary therapy was difficult to perform and is therefore highly encouraged in future studies. Second, the exploratory nature of this post hoc analysis and enrollment of a relatively small sample size in each subgroup, all of whom were of Chinese ethnicity, limited the robustness of the study findings to some extent. Further validation is warranted in larger cohorts and across diverse ethnic populations in real-world settings. Besides, the accuracy of ESS for measuring daytime sleepiness may not be really relevant for Asian population since some questions are not suitable for the Asian population who usually do nap after lunch and use a motorcycle but not a car. Finally, the behavioral impacts of solriamfetol on patients remain unexplored, specifically its potential effects on occupational injury and motor vehicle crash risk in OSA-EDS cohorts. The real-world clinical impact of solriamfetol thus merits further investigation.