Narcolepsy type 2: phenotype is fundamental

In a detailed analysis of narcolepsy type 2 (NT2), Barateau and colleagues assess their Narcolepsy Severity Scale, modified to remove cataplexy (called the NSS-2), and Idiopathic Hypersomnia Severity Scale (IHSS) [1]. This work emphasizes the symptom phenotype of this challenging diagnosis and probes the common framing that NT2 exists on a symptom spectrum with narcolepsy type 1 (NT1) on one end, IH on the other, and NT2 in the middle [2]. In support of such a spectrum, prior work indicates that classic narcolepsy symptoms, such as sleep paralysis and sleep-related hallucinations, are most common in people with NT1, intermediate in people with NT2, and least (but not absent) in people with IH [3]. Similarly, for classic IH features, such as long sleep durations and sleep inertia, patients with IH and NT1 segregate fairly cleanly into opposite poles, whereas available data suggest patients with NT2 may fall in the middle [4, 5].

To some extent, this apparent spectrum may reflect diagnostic misclassification [2]. Occult hypocretin deficiency in people with narcolepsy without cataplexy likely results in some people with NT1 being classified as NT2, a risk that may vary by population [6]. However, this risk was mitigated substantially by Barateau et al., with cerebrospinal fluid confirmation of hypocretin levels in 69% of their participants [1]. More challenging, perhaps even impossible with current tools, is to definitively distinguish those with NT2 from those with IH, given the relatively poor test–retest reliability of the MSLT for these disorders [7–10].

The Barateau work broadly supports the symptom spectrum construct, with a notably smaller proportion of NT2 participants with typical narcolepsy symptoms than with typical IH symptoms. Classic narcolepsy symptoms were seen in fewer than half, including 40% with sleep-related hallucinations and 28% with sleep paralysis [1], less than in people with NT1 [11]. Classic IH symptoms were much more common, including long sleep (76%) and sleep inertia (83%), but less than in IH patients [12]. Unrefreshing naps, a typical symptom of IH [13], were seen in most (63%) of the NT2 participants.

Where the findings from Barateau et al. are particularly intriguing, perhaps challenging the symptom spectrum framework, is the high proportion of NT2 participants with disrupted nocturnal sleep. Over half (58%) of NT2 participants had disrupted sleep based on their responses to the question, “Currently, how disturbed is your nighttime sleep?.” Among those NT2 patients with classic IH symptoms of long sleep or sleep inertia, 63% endorsed disrupted sleep. This nocturnal feature is a well-described symptom and polysomnographic finding in people with NT1 [14], but is less studied among people with NT2. In meta-analyses, objectively measured disrupted sleep is clearly more common in the combined narcolepsy group than in healthy controls [15]. However, much of this is driven by those with NT1, who demonstrate multiple polysomnographic markers of more severely disrupted sleep than people with NT2, including significantly lower sleep efficiency, more wake after sleep onset, and more N1 [15]. In contrast, polysomnographic sleep in NT2 is similar to that in IH, apart from REM features [16]. Disrupted sleep is a multifaceted construct that is likely to be driven by several factors, including features unique to NT1, such as state dyscontrol caused by hypocretin deficiency, and those not exclusive to NT1, such as sleep and mood comorbidities and treatment side effects. As with NT1 [17], a single question about disrupted sleep may not provide enough detail to fully understand this phenomenon in people with NT2.

The Barateau study also reveals a discordance in how the sleepiness of NT2 was captured on each questionnaire. Untreated patients with NT2 universally endorsed sleepiness on the NSS-2, and all but one (4%) endorsed sleepiness on the IHSS. However, 16% of NT2 participants were not considered to be sleepy based on the Epworth [1]. This highlights the ongoing challenge in sleep medicine with defining the multifaceted construct of sleepiness. There are well-described differences between objective measures of sleepiness and self-reported measures [18], but the work by Barateau et al. demonstrates that self-reported measures are also influenced by differences in the specific construct being assessed. In this context, the recent text revision of the International Classification of Sleep Disorders [13] offers an interesting amendment to the definition of sleepiness in the central disorders of hypersomnolence. The third edition defined sleepiness as “periods of irrepressible need to sleep or daytime lapses into sleep” [19]. In the text revision, the second half of this definition has reverted to that of the second edition, “. . . daytime lapses into drowsiness or sleep” (emphasis added). This change encompasses the range of sleepiness manifestations across different hypersomnia disorders, not always occurring as sudden sleep attacks [2]. It is presently unclear which scale may be optimal to measure this revised definition of sleepiness, given that the Epworth specifically requires a lapse into sleep (i.e. to doze) [20]. The NSS-2 and IHSS both offer multiple definitions of sleepiness, for narcolepsy focusing mostly on irresistible need for sleep [21] and for IH focusing on need for and duration of naps and the struggle to stay awake [12], suggesting a combination of these tools may be better aligned with this broader definition of sleepiness.

Barateau and colleagues define the minimal clinically important difference (MCID) for the NSS-2 via standard distribution-based methods, both Cohen and empirical rule effect sizes [1]. Anchor-based approaches, in which changes are compared to an anchor scale that defines meaningful improvement, such as the patient's global impression of change [22], are also commonly used [23]. In the work by Barateau et al, although an anchor-based definition of MCID was not used, the inclusion of the European QoL-5 Dimensions quality of life scale as an external reference in a subset of participants clarifies how well the NSS-2 relates to the primary clinical goal of improving quality of life. Correlations with quality of life were modest in treated patients, absent in untreated patients. This, along with relatively low proportions of NT2 patients achieving MCIDs despite expert treatment (53% on the IHSS, 62% on the NSS-2 and Epworth) [1], highlights the urgent need for additional, more effective, and comprehensive treatments for this disorder.

Finally, the symptom overlap across hypersomnia disorders and the potentially fluid nature of IH and NT2 support the need for a single, validated scale for all these non-episodic disorders. Such a disease-independent scale could improve clinical utility and allow a better understanding of symptoms across diagnostic boundaries. In pursuit of such a transdiagnostic scale, some well-validated constructs from the NSS and IHSS could be used directly, while others, such as disrupted sleep and quality of life, might be expanded or added to allow more granular assessment. It would be critical to consider problematic symptoms not included in the NSS or the IHSS. As Barateau et al. note, neither scale assesses fatigue, attention/cognition, brain fog, or automatic behaviors [1], all common and problematic in hypersomnia disorders [24–26]. A transdiagnostic scale might also incorporate symptoms of common comorbidities, such as dysautonomia and mood disorders [27–29]. Such scale development should partner with patients, to ensure the lived experience of people with hypersomnolence is fully captured.

Funding

This work was supported by the National Institutes of Health (R01 NS111280 to Dr. Trotti).

Disclosure Statement

Financial disclosure: Drs. Trotti & Nichols have no financial conflicts to disclose. Nonfinancial disclosure: Dr. Trotti is a member of the Board of Directors of the American Academy of Sleep Medicine. Any opinions expressed are those of the authors, and do not necessarily reflect that of the AASM or its board.