Craniopharyngioma is a rare intracranial tumor located in the hypothalamic and pituitary region (e.g., Halac and Zimmerman, 2005; Müller, 2014). Approximately 6 to 10% of children with brain tumors have craniopharyngioma, with peak pediatric onset between 5 and 14 years of age. The typical treatment approach is a combination of surgical resection and radiation therapy (Müller et al., 2019). Survival rates for craniopharyngioma are very high, with a recent treatment study follow-up demonstrating an overall survival rate of 96% at 10 years (Edmonston et al., 2022). Despite high survival rates, quality of life may be substantially affected due to significant morbidity, including headaches, visual impairment, endocrine dysfunction, cognitive late effects, and sleep disturbances (e.g., Halac and Zimmerman, 2005; Müller, 2008; Khan et al., 2012; Fournier-Goodnight et al., 2017; Mandrell et al., 2020; Merchant et al., 2022).

Related to sleep disturbance, rates of narcolepsy and hypersomnia due to medical condition in pediatric craniopharyngioma are as high as 80%, despite the low prevalence of these conditions in the general population (e.g., Plazzi et al., 2018; Mandrell et al., 2020). Children and adolescents with craniopharyngioma who are overweight or obese are more likely to manifest hypersomnia or narcolepsy, with higher grade of hypothalamic tumor involvement associated with greater likelihood of a narcolepsy diagnosis (Mandrell et al., 2020). Children and adolescents with hypersomnia or narcolepsy are at risk for poor psychosocial functioning (Plazzi et al., 2018; Graef et al., 2020; Ingram et al., 2021), which further compounds poorer quality of life outcomes among youth with craniopharyngioma. High rates of sleep disorders are likely related to effects of the tumor location and treatment on nearby brain structures (i.e., optic chiasm, pituitary gland, hypothalamus) critical for sleep-wake regulation (España and Scammell, 2011).

Damage to the hypothalamic region also suggests likely disruptions to circadian rhythms, which are regulated by the suprachiasmatic nucleus (SCN) in the hypothalamus (Dibner et al., 2010). An examination of potential circadian rhythm disruption in patients diagnosed with craniopharyngioma is important to improve our understanding of sleep disturbance.

A theoretical overview of potential mechanisms of circadian rhythm disruption is outlined in Figure 1 and details both static and dynamic vulnerabilities to circadian disruption. Static vulnerabilities, or those that may be unchanging consequences of the tumor and/or surgical intervention, may include damage to the optic pathway, damage to the SCN, endocrine disruption, and changes to MAPK/ERK signaling pathways (Halac and Zimmerman, 2005; Müller et al., 2019; Hengartner et al., 2020). Damage to the optic pathway may cause changes in light sensitivity with downstream effects on circadian entrainment (e.g., Abbott et al., 2021), while damage to the SCN may directly affect timing of melatonin secretion (Dibner et al., 2010). Endocrine disruption, such as adrenal insufficiency, can alter the rhythm of cortisol secretion and other glucocorticoids with the potential for downstream effects on signaling and synchronization between the SCN and peripheral clocks (e.g., Chan and Debono, 2010). The potential for overactivity in the MAPK/ERK pathway is proposed in craniopharyngioma, and, if present, may lend vulnerability to circadian phase shifts (Kalkman, 2012; Hengartner et al., 2020).

Dynamic vulnerabilities may interact with and be further exacerbated by resulting circadian disruption. Behavioral changes, such as increased napping, decreased physical activity, and increased caloric intake/changes to meal timing may lead to disruption and misalignment between circadian timing and behavior. This may further contribute to weight gain and poor sleep health (e.g., Carskadon and Acebo, 2002; Culnan et al., 2021). Therefore, there is the theoretical potential for circadian disruption to arise through various mechanisms—although the effects of vulnerabilities on circadian function are yet to be demonstrated.

Thus far, a few studies have investigated the potential for circadian rhythm disruption in craniopharyngioma by measuring overall levels of melatonin, a hormone that helps establish circadian rhythmicity (Benloucif et al., 2008). Patients with craniopharyngioma have been found to have lower levels of melatonin than healthy controls, providing initial evidence that the rhythm of melatonin production may be disrupted (Lipton et al., 2009; Pickering et al., 2014). Furthermore, low melatonin levels are associated with higher self-reported sleepiness and additional sleep problems (as measured by actigraphy) such as decreased nighttime sleep duration and reduced sleep efficiency (Müller et al., 2002; Pickering et al., 2014).

While these findings are suggestive of potential circadian rhythm disruption, these studies have typically been small and/or included a wide age range from children to adults. Further study in a large sample of children and adolescents with craniopharyngioma would be beneficial for more precise characterization of circadian health in pediatric craniopharyngioma. In addition to measuring overall melatonin levels, the timing of melatonin secretion provides key information about circadian phase. Dim light melatonin onset (DLMO) reflects the timing of the rapid increase in melatonin secretion that occurs just prior to habitual bedtime (Crowley et al., 2006). If circadian phase is shifted (i.e., advanced or delayed), this could have implications for the understanding of sleep disturbance in craniopharyngioma and the development of effective treatments.

The present study is derived from an exploratory aim of a treatment protocol for pediatric craniopharyngioma. Details of the larger treatment study are provided below. The purpose of the exploratory aim was to examine circadian factors in pediatric craniopharyngioma. Within this aim, we had two study goals:

Prior research on circadian factors and craniopharyngioma has been focused on melatonin levels, rather than timing of melatonin secretion. Therefore, our study hypotheses were largely exploratory. We hypothesized that DLMO timing would correlate with the presence of hypersomnia due to medical disorder or narcolepsy, sleep duration, and sleep efficiency, consistent with reported associations between low melatonin and other sleep parameters (self-reported daytime sleepiness, decreased nighttime sleep duration, reduced sleep efficiency; Müller et al., 2002; Pickering et al., 2014). We also predicted that DLMO timing would correlate with mid-sleep time, which reflects a behavioral estimate of circadian timing (Terman et al., 2001; Roenneberg et al., 2003). We hypothesized that DLMO timing would correlate with BMI based on prior reported findings of an association between melatonin levels and BMI in pediatric craniopharyngioma (Müller et al., 2002). Finally, we expected DLMO to be later with more mature Tanner stage, consistent with the typical circadian shift in adolescence (Carskadon, 2011; Crowley et al., 2018).

In this study, we examined circadian rhythms among children and adolescents with craniopharyngioma by estimating DLMO. DLMO was missed for almost half our sample (n = 25), with most patients in the missed DLMO group above the threshold at the time of first collection. For these patients, it is possible that they had an earlier circadian phase. There is some evidence to suggest that at least a small subset of individuals with craniopharyngioma (2 of 14 patients in one study) show a phase advanced rhythm (Pickering et al., 2014). Circadian phase advances are also consistent with our clinical observations that patients often describe an early morning preference. Anecdotally, even adolescents with craniopharyngioma commonly report a morning preference, which is atypical for this age group.

The likelihood of missed DLMO was not associated with any examined demographic variables. However, it was associated with greater disease severity, as defined by hypothalamic involvement and presence of diabetes insipidus. Diabetes insipidus, unless present prior to surgery, reflects extensive surgical resection (Edmonston et al., 2022). This suggests that capturing DLMO within the sampling window was less likely for those with more severe encroachment on brain structures from the tumor and subsequent surgical intervention(s) (Halac and Zimmerman, 2005; Müller et al., 2019).

We also examined whether missed DLMO may have been related to sleep health variables. Surprisingly, missed DLMO did not relate to the presence of a central disorder of hypersomnolence. This was an unexpected finding, although it could have been due to the very high rates (83%) of these conditions in our sample, limiting variability. We found that patients in the missed DLMO subgroup took longer to fall asleep at night and had a more variable sleep onset latency, as measured by actigraphy, than those in the DLMO subgroup. It is not clear why patients in the missed DLMO subgroup had longer and more variable SOL. Many of these patients showed melatonin values that were above threshold at the beginning of sampling, which suggests that their DLMO may have occurred earlier (phase advanced). However, we do not know for certain if this is the case without more data to confirm a phase advance. It is also possible that those in the missed DLMO subgroup have circadian dysregulation overall, which could be directly or indirectly (e.g., napping behavior) contributing to lengthened sleep onset latency. Unfortunately, we did not gather data on napping behavior.

For those with a valid DLMO estimate, our hypotheses were partially supported. Although DLMO timing was not associated with the presence or absence of hypersomnia or narcolepsy, later DLMO timing was associated with increased sleepiness severity on the MSLT (reflected by a shorter average sleep onset latency). That we observed significant differences when data were looked at continuously, but not categorically, is not entirely surprising given the very high rates of disorders of hypersomnolence in our sample. In contrast to our hypotheses, we did not find associations between DLMO timing and sleep duration or efficiency. However, later DLMO timing was associated, as one might expect, with later midsleep and wake times. Later DLMO timing was also associated with decreased variability in actigraphy sleep onset time. Consistent with our hypothesis that DLMO timing and BMI would be associated, we found that patients with higher BMIs had later DLMO timing. Finally, we did not find an association between DLMO timing and age or Tanner stage, which is inconsistent with well-established findings of circadian shift in adolescence (Carskadon, 2011; Crowley et al., 2018). This may reflect disrupted sleep/wake regulation in this sample of children and adolescents with craniopharyngioma; however, a lack of association between DLMO timing and Tanner stage may simply reflect a power issue as the majority of participants with valid DLMO estimates were in Tanner stage 1.

Overall, our findings suggest possible circadian phase advance in a subset of children and adolescents with pediatric craniopharyngioma and that circadian timing is related to additional health factors. It will be important to further study circadian rhythms in these patients in order to inform clinical practice. In clinical settings, DLMO is not routinely used, but actigraphy may be used as a proxy to identify patients with advanced phase or other forms of circadian disruption. Note, however, that validation of using actigraphy to estimate circadian phase has not been systematically tested in this patient group. If indicated, targeted interventions, such as bright light therapy and melatonin administration, may be beneficial (Abbott et al., 2020). Circadian health is important for broader sleep and overall health, and consideration of circadian health may improve health and daytime function for patients.

Our study findings suggest that a subset of children and adolescents with craniopharyngioma may have phase advance and that this relates to poorer prognostic indicators (e.g., hypothalamic involvement of the tumor). For those with valid DLMO estimates, timing correlated with sleepiness and BMI. Further study of circadian health in pediatric craniopharyngioma is needed to refine treatments and improve care for patients.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by St. Jude IRB. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

TM is the principal investigator for the prospective treatment protocol from which circadian data was collected. TM, MW, BM, and DH completed sleep and health assessments and collected all data for the project. Specific project conceptualization was completed by BM, VC, DK, and SC. SC provided circadian expertise and consultation. Data management and analysis was completed by DK, YL, HD, and JS. All authors contributed to manuscript development and editing. All authors contributed to the article and approved the submitted version.