Anesthesia outcomes in lysosomal disorders: CLN3 and GM1 gangliosidosis

Abstract

Natural history studies of pediatric rare neurometabolic diseases are important to understand disease pathophysiology and to inform clinical trial outcome measures. Some data collections require sedation given participants’ age and neurocognitive impairment. To evaluate the safety of sedation for research procedures, we reviewed medical records between April 2017–October 2019 from a natural history study for CLN3 (NCT03307304) and one for GM1 gangliosidosis (NCT00029965). Twenty-two CLN3 individuals underwent 28 anesthetic events (age median 11.0, IQR 8.4–15.3 years). Fifteen GM1 individuals had 19 anesthetic events (9.8, 7.1–14.7). All participants had American Society of Anesthesiology classification of II (8/47) or III (39/47). Mean sedation durations were 186 (SD=54; CLN3) and 291 (SD=33; GM1) minutes. Individuals with GM1 (6/19, 31%) were more frequently prospectively intubated for sedation (CLN3 3/28, 11%). Minor adverse events associated with sedation occurred in 8/28 (28%, CLN3) and 6/19 (32%, GM1) individuals, frequencies within previously reported ranges. No major adverse clinical outcomes occurred in 47 anesthetic events in pediatric participants with either CLN3 or GM1 gangliosidosis undergoing research procedures. Sedation of pediatric individuals with rare neurometabolic diseases for research procedures is safe and allows for collection of data integral to furthering their understanding and treatment.

1. INTRODUCTION

Advances in molecular sequencing have spurred improvements in the diagnosis of and therapeutic development for rare inherited genetic conditions. For CLN3 and GM1 gangliosidosis, multiple interventions involving small molecule or gene replacement therapies are at various clinical trial development stages (Nicoli et al., 2021; Specchio et al., 2021). CLN3 [Juvenile NCL, Batten, OMIM 204200] is an autosomal recessive, neurodegenerative, lysosomal disorder with a classic presentation of blindness, seizures, and behavior and motor dysfunction. GM1 gangliosidosis [Type I, OMIM 230500; Type II, OMIM 230600; Type III, OMIM 230650) involves ophthalmologic findings with blindness in the infantile presentation, neurodevelopmental delay and regression, seizures, hearing loss, and visceral and musculoskeletal pathologies.

As lack of sufficient number of participants to power studies and of clinically meaningful disease biomarkers are recognized challenges in rare disorder research, broadly designed natural history studies that surveil multiple organ systems and collect relevant biosamples are essential for longitudinal deep phenotyping, and biomarker and hypothesis generation (Dickson et al., 2011; Shapiro et al., 2016). The neurologic involvement of CLN3 and GM1 gangliosidosis calls for neuroimaging modalities with high sensitivity and collection of biospecimen proximal to the brain, e.g. cerebral spinal fluid. Both procedures require comprehension and cooperation from the study participants, abilities most often limited in the pediatric population and these conditions, and thus necessitating sedation.

Information on anesthetic management and outcomes in CLN3 and GM1 gangliosidosis is sparse. Association of these disorders to inherited metabolic conditions perpetuates word-of-mouth, unspecified concerns of worsening disease outcome with certain anesthetics. The few cases of anesthetic outcome in individuals with CLN3 noted no significant adverse effect (Pereira et al., 2006; Yamada et al., 2002; Yamaguchi et al., 2020). Management recommendations for sedation in individuals with gangliosidoses often address the potential airway obstruction from soft tissue substrate deposits, and cervical instability from skeletal dysplasia (Kloesel & Holzman, 2017; Lerman et al., 2016). While these issues may be seen in GM1-gangliosidosis, both features are more in line with the mucopolysaccharidosis conditions.

In this report, we provide data on the anesthetic experience at a single center from natural history studies of CLN3 and GM1 gangliosidosis. The study cohorts include individuals of similar age range, similar length of anesthesia exposure within a specified time span so that they can serve as comparator group for one another. We hypothesize that while the underlying pathophysiology may contribute to different anesthesia events, both groups have outcomes that were within the expected range for individuals with American Society of Anesthesiology (ASA) classifications II/III.

2. PARTICIPANTS and METHODS

2.1. Editorial Policies and Ethical Considerations

The observational studies of CLN3 (NCT03307304) and glycosphingolipid disorders including GM1 gangliosidosis (NCT00029965) were approved by the NIH Institutional Review Boards. At enrollment, we obtained consents from guardians or participants 18 years and older, and assents from participants 7 years and older if deemed cognitively appropriate by the investigators.

2.2. Observational Natural History Studies

Participants of all ages, genders, demographic, and geographic locations were included in the studies. For either condition, eligible participants were diagnosed based on molecular findings resulted from a CLIA-certified laboratory or combinations of molecular, laboratory and clinical findings consistent with the disease. Individuals who were unable to travel or could not comply with the protocol, were not included in the studies.

CLN3 participants underwent sedated procedures at yearly intervals. Evaluations under anesthesia included MR brain imaging, ophthalmologic evaluations, auditory brain stem response, lumbar puncture, and skin biopsy. GM1 participants underwent sedated procedures every 6 months to 1 year that included MR brain, ophthalmologic evaluations, auditory brain stem response, and lumbar puncture.

2.3. Anesthesia

Prior to any procedure under anesthesia, the anesthesiology team evaluated and obtained procedural consents from participants or guardians. Participants were instructed to fast overnight for sedated procedures that started between 7–8 AM. Most of the participants had intravenous access when presenting for anesthesia on the morning of anesthetic. A team including a board-certified Anesthesiologist who was, at times, accompanied by a certified registered Nurse Anesthetist, administered anesthesia and monitored participants per standard clinical guidelines.

Anesthesia was administered to participants in multiple ways, the most common being with total intravenous anesthesia (i.e., propofol infusion). The objective of the propofol infusion was for the participant to breathe spontaneously, but to otherwise refrain from excessive movement. Most participants appear to experience this type of anesthesia as complete unconsciousness.

In general, propofol sedation with the participants spontaneously breathing was selected when it was concluded by the anesthesiologist, based on medical history and physical exam, that the participants would be able to maintain a patent airway for the duration of the case and was at a very low probability for pulmonary aspiration of gastric contents. The participant would remain under sedation for the duration of the case as long as he or she was able to maintain spontaneous ventilation, adequate oxygenation, and a patent airway. After propofol sedation was initiated, in participants who needed assistance maintaining a patent airway, an oropharyngeal or nasopharyngeal airway was placed to maintain adequate oxygenation and ventilation.

Airway management with an endotracheal tube was selected if, based on medical history and physical exam, the anesthesiologist determined it would be safer to have the participants under general endotracheal anesthesia than under propofol sedation without a secured airway for the duration of the procedures. Among the reasons that a participant may be intubated include risk of pulmonary aspiration of gastric contents if the participant did not have a secure airway (i.e., history of gastroesophageal reflux or vomiting) or difficulty maintaining a patent airway or spontaneous respirations during the case (i.e., history of obstructive sleep apnea or severe neuromuscular dysfunction). None of the participants had a laryngeal mask airway applied or a difficult intubation.

A retrospective chart review of anesthetic events at the NIH Clinical Center for the CLN3-Batten and GM1 natural history protocols between April 2017 – October 2019 was conducted. Data from the individual participant’s medical records for the following parameters were extracted or reviewed by an Anesthesiologist on the study protocol: age and weight at the time of the anesthetic, American Society of Anesthesiologist classification, type of anesthesia, airway management, anesthesia duration [time from first sedating drug until the participant went to the post-anesthesia care unit (PACU)], temperature during anesthetics, perioperative adverse events (from start of anesthetic to 48 hours after recovery), and medications given during anesthetic period.

2.4. Statistics

Descriptive and non-parametric statistics are provided with significance set at p < 0.05.

3. RESULTS

3.1. Participant and Evaluation Characteristics

Between April 2017 – October 2019 twenty-two and fifteen study participants with CLN3 and GM1, respectively, received anesthesia for research evaluations per study protocol. Some individuals had more than one anesthetic events, thus the number of anesthetic events reviewed were 28 for CLN3 and 19 for GM1. Demographic information for the individuals at each anesthetic events showed similar median age and sex distributions. Participants with GM1 weighed less than those with CLN3 [median (min,max) in kg: GM1 26 (15,50), CLN3 44.5 (27.5, 54.0)]. Participants in both conditions had ASA classifications of either II or III in comparable distributions. (Table 1)

Table 1.

Demographic and clinical characteristics of the CLN3 and GM1 cohorts.

	CLN3	GM1	p-value*
Individuals, n	22	15
Anesthetic events, n	28	19
Parameters at individual anesthetic event
Age, years			0.50
Median	11.0	9.8
Quartile (1st; 3rd)	8.4; 15.3	7.1; 14.7
Sex			0.14
Female, n (%)	13 (46)	13 (68)
Male, n	15	6
Weight, kg			0.004
Median	44.5	26.0
Quartile (1st; 3rd)	27.5; 54.0	22.5; 35.0
Disease severity †
Median score	9	n/a
Quartile (1st; 3rd)	4; 21.5	n/a
Seizure
Present, n (%)	16 (57)	10 (53)
Antiepileptics, n	16 (57)	5 (26)
ASA Classification			0.85
II, n (%)	5 (18)	3 (16)
III, n (%)	23 (82)	16 (84)
IV, n (%)	0 (0)	0 (0)

^*Mann-Whitney (0.05 significance, 2-tailed): Age, Weight. Chi-square (0.05 significance): Sex, ASA Classification.

^†CLN3: Unified Batten Disease Rating Scale Physical score. ASA – American Society of Anesthesiologists. n/a – not applicable.

3.2. Peri-Anesthetic Parameters and Events

The average sedation durations were 186 (SD=54) and 291 (SD=33) minutes for CLN3 and GM1 participants, respectively. Evaluations and procedures under sedation are listed in Table 2. For participants with CLN3, those most frequently performed included lumbar puncture, neuroimaging, and ophthalmologic evaluation (including optical coherence tomography). Skin biopsy was frequently done immediately following lumbar puncture for participants’ comfort but was not the primary or only indication for anesthesia. For participants with GM1, frequently performed procedures were neuroimaging, ophthalmologic and audiologic evaluations, and lumbar puncture.

Table 2.

Procedures and evaluations completed under sedation for the CLN3 and GM1 cohorts.

	CLN3	GM1
Anesthetic events, n	28	19
Neuroimaging
Magnetic resonance imaging, n	24	19
Magnetic resonance spectroscopy, n	24	15
Ophthalmologic evaluation
Exam under anesthesia, n	14	17
Optical coherence tomography, n	12	0
Electroretinogram, n	7	0
Audiologic evaluations
Auditory brainstem response, n	5	17
Cerumen removal, n	2	0
Procedures
Lumbar puncture, n	27	15
Skin biopsy, n	22	7
Electromyogram/Nerve conduction velocity, n	0	9
Dental exam, n	0	1

Propofol was the most used anesthetic agent in participants with either condition (Figure 1). Other gaseous and IV agents were also used, including dexmedetomidine in three events for participants with CLN3 and parental report of irritable mood or previous agitation following anesthesia. Participants with CLN3 also received > 2 doses of pressors during 8 of 28 anesthetic events. Other medications administered in the peri-anesthesia period included fentanyl, midazolam, and glycopyrrolate.

Figure 1.

Medications administered peri-anesthetic events for the CLN3 and GM1 cohorts. Gray: CLN3. Black: GM1. Dashed lines: total number of anesthetic events. Paralytics: cisatracurium, rocuronium, succinylcholine. Pressors: > 2 doses of ephedrine or phenylephrine.

A higher percentage of events in the CLN3 cohort were classified as deep sedation than in the GM1 cohort (86% versus 53%), and vice versa for general anesthesia (14% versus 47%) (Table 3). Similarly, the GM1 cohort had higher level of respiratory support as reflected in more events with planned endotracheal intubation. However, the chi-square test of independence for relation between GM1 and need for any respiratory support was not significant at p < 0.05.

Table 3.

Peri-anesthesia parameters for the CLN3 and GM1 cohorts.

	CLN3	GM1
Anesthetic events, n	28	19
Type of anesthesia
Deep sedation, n	24	10
General anesthesia, n	4	9
Airway management
None, n	18	11
Oral/Nasopharyngeal airway, n	7	2
Endotracheal intubation, n	3	6

Most anesthetic events in either cohort had no complication [CLN3: 20/28 (71%); GM1: 13/19 (68%); Figure 2]. Most complications can be classified as mild to moderate severity (Common Terminology Criteria for Adverse Events version 4.0) and resolved with minimal interventions in an outpatient setting. Two adverse events in CLN3 participants classified as severe involved fasting-related syncope and prolonged post-sedation recovery. These resolved within 24 hours with monitoring and maintenance intravenous fluid. Post-anesthesia headache and agitation upon emergence were observed only in the CLN3 cohort. One of the two post-anesthetic headaches was related to post-lumbar puncture procedure based on presentation and timeline, whereas the other was of brief duration and not likely related. The three emergence agitation events occurred in CLN3 female participants who were 10, 15, and 16 years old at the time of anesthesia. The primary anesthetic agent in these three events was propofol. Of note, the 10- and 15-year-old participants had a respectively preceding and subsequent anesthetic event, where propofol and sevoflurane/nitrous oxide were primary anesthetic agents, that were not associated with emergence agitation. The three anesthetic events where dexmedetomidine was administered were not associated with emergence agitation.

Figure 2.

Peri-anesthetic event complications for the CLN3 and GM1 cohorts. An anesthetic event can be associated with more than one complication. Gray: CLN3. Black: GM1. Dashed lines: total number of anesthetic events. Others: aspiration/chemical pneumonitis, suspected hypoglycemia (CLN3); dehydration, hypothermia, shivering (GM1).

4. DISCUSSIONS

As approaches for therapeutic management of rare pediatric heritable diseases expand, the need for comparative natural history data becomes more evident (Goring et al., 2019; Jahanshahi et al., 2021; Mulberg et al., 2019). Concerns for adverse events and worsening of disease presentation are reasons parents of children with CLN3 cited for not participating in observational natural history studies where anesthesia may be used. In this retrospective review of 22 participants with CLN3 in comparison with 15 participants with GM1 who received anesthesia, the majority had no complications. Observed complications did not require prolonged or escalated care. Emergence agitation is observed in participants with CLN3 and may be pre-emptively managed with dexmedetomidine. The data presented here will be useful towards informing risk determination for sedated procedures in rare pediatric neurological diseases by Institutional Review Boards.

The frequencies of peri-anesthesia adverse events in both the CLN3 and GM1 cohorts are higher than the calculated 534 per 10 000 rate reported for a pediatric research cohort (Kiringoda et al., 2010). However, the Kiringoda et al. (2010) study did not report several adverse event categories such as headache, poor oral intake, or fever. The frequencies of adverse events observed in our cohorts are similar to previously reported numbers of ~30% in a mostly healthy pediatric cohort (Cohen et al., 1990) and a cohort of individuals with various types of neuronal ceroid lipofuscinosis (Yamaguchi et al., 2020).

Effects of anesthesia on cerebral spinal fluid marker levels for inflammation or injury have been reported for older adult cohorts without known neurocognitive diseases. Inflammatory cytokine IL-6 and IL-8 levels are the most evaluated and consistently reported to be elevated in CSF during 2–48 hours post-anesthesia (Pikwer et al., 2017; Reinsfelt et al., 2012; Tang et al., 2011). Changes in the CSF levels of markers for neuronal injury following anesthesia are variable. CSF neurofilament light chain did not change (Danielson et al., 2021; Reinsfelt et al., 2012), whereas glial fibrillary acidic protein (Anckarsater et al., 2014; Reinsfelt et al., 2012) appear to increase following anesthesia. CSF total tau (Danielson et al., 2021; Palotas et al., 2010; Pikwer et al., 2017; Reinsfelt et al., 2012; Tang et al., 2011; Zhang et al., 2013) and amyloid beta peptide 42 (Danielson et al., 2021; Palotas et al., 2010; Pikwer et al., 2017; Tang et al., 2011; Zhang et al., 2013) were variably reported as increased or unchanged following anesthesia. Heterogeneous study designs contribute to the lack of overwhelming evidence to support a clear effect of anesthesia on inducing an inflammatory or lasting injurious effects in the studied adults. Notably, a few of these studies (Danielson et al., 2021; Palotas et al., 2010) attempted to correlate the biochemical findings with neurocognitive outcomes and reported none or not clinically significant findings. We were unable to identify similar biomarker studies for the pediatric population.

The US Food and Drug Administration published a warning of potential adverse effects on neurodevelopment in children under 3 years of age who undergo repeated or prolonged sedation (FDA). The recommendation cited animal studies showing neuronal toxic effects and pediatric studies reporting conflicting results of the effect of anesthesia in children younger than 3 years of age on neurocognitive development, learning abilities, and behavioral issues. More recent or larger pediatric study cohorts continue to leave uncertainties about the effects of anesthesia on long-term neurodevelopment (Walkden et al., 2020; Warner et al., 2018; Yang et al., 2021; Zaccariello et al., 2019). In addition to the focus of these pediatric studies on the effect of anesthesia exposure before 3 years of age, it was not feasible based on the study design to examine the effects of anesthesia on an existing, underlying neurodevelopmental or neurodegenerative disorders where neurocognitive and behavioral dysfunctions are major components of the disease phenotype.

In the CLN3 and GM1 cohorts evaluated here, behavioral dysfunctions during emergence from anesthesia were observed only in the former. Behavior problems associated with CLN3 have been described as “diverse and complex psychiatric symptoms” including symptoms that are “social, thought, attention problems, somatic complaints and aggressive behavior” (Backman et al., 2005), with anxiety a frequent feature. A survey of 42 juvenile neuronal ceroid lipofuscinosis (CLN3) patients in Finland identified fear in 45% of the CLN3 patients studied (Santavuori et al., 1993). In the pediatric population, certain psychosocial factors that contribute to emergence delirium include patient and parental pre-operative/pre-anesthetic anxiety and decreased sociability (Kain et al., 2004; Kain et al., 2007). Dexmedetomidine is one of many medications studied and evaluated for its efficacy in managing emergence delirium (Rao et al., 2020). A study of 100 children ages 2 to 8-years-old showed decreased numbers of emergence delirium events in those receiving dexmedetomidine or propofol versus saline near the end of the elective surgical intervention (Makkar et al., 2016). In the anesthetic events described in our cohort, three CLN3 events involved aggression or irritability during recovery. Dexmedetomidine was not administered during any of these 3 anesthetic events. In contrast, in 3 CLN3 anesthetic events where dexmedetomidine was administered based on reported history or observations of behavioral dysfunctions did not demonstrate evidence of behavior problems during recovery.

In summary, we report the anesthetic experience of pediatric participants with CLN3 or GM1 undergoing research sedated procedures at the NIH. No major adverse events were observed. The types and frequencies of non-major adverse events were within expectations for children with ASA classification II/III. Pre-anesthesia evaluations is essential to determine existing clinical features (e.g., behavioral dysfunction) that may inform improved post-anesthesia care. Sedation for research procedures by experienced anesthesiology teams is safe and vital for the study of and therapeutic development for these orphaned diseases.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.