Comparison of Antiemetics in the Management of Pediatric Cannabinoid Hyperemesis Syndrome

Abstract

OBJECTIVE

As a result of recent legislative changes allowing for increased access to marijuana products, there have been increasing rates of cannabis abuse among adolescents and subsequent diagnoses of cannabinoid hyperemesis syndrome (CHS). Most available literature on this syndrome exists within the adult population and describes benzodiazepines, haloperidol, and topical capsaicin as potentially efficacious in the management of CHS. The objectives of this study were to identify antiemetics and compare their efficacy and safety in the management of pediatric CHS.

METHODS

A retrospective review of Penn State Children's Hospital electronic health record was performed to identify patients 18 years or younger who had an emergency department or inpatient encounter, a cannabis hyperemesis–related diagnosis code, and met diagnostic criteria for CHS. Antiemetic efficacy was determined using subjective patient perception of nausea and objective documentation of vomiting. Benzodiazepines, haloperidol, and topical capsaicin were classified as nontraditional antiemetics, whereas all other antiemetics were classified as traditional.

RESULTS

Nontraditional antiemetic medications appeared to be more effective in resolving patient symptoms compared with traditional antiemetics. Analysis of all ordered antiemetics demonstrated a gap in partial or full symptom resolution between nontraditional and traditional agents. Reported adverse effects were minimal.

CONCLUSIONS

Cannabinoid hyperemesis syndrome is an underrecognized and underdiagnosed condition characterized by cyclic vomiting related to chronic cannabis use. Abstinence from cannabis remains the most effective approach to mitigating morbidity associated with CHS. Medications such as lorazepam or droperidol may have benefit in managing toxidrome symptoms. Traditional antiemetic prescribing remains a key barrier to effective management of pediatric CHS.

Introduction

Cannabis is the most widely used illicit drug in the United States, with the highest prevalence of use in individuals ages 18 to 25 years.¹ As of November 29, 2021, a total of 18 states have legalized recreational marijuana and 37 states have regulations allowing for medical marijuana use with a valid prescription.² As a result of legislative changes, there have been increasing rates of cannabis abuse and subsequent diagnoses of cannabinoid hyperemesis syndrome (CHS).³

Cannabinoid hyperemesis syndrome is characterized by episodes of cyclic nausea and vomiting as a result of chronic cannabis use. It is an acute on chronic condition that is often refractory to standard antiemetic medications.⁴ It is a diagnosis of exclusion requiring the elimination of other potential etiologies prior to effective treatment for the condition.⁴ The diagnosis is further complicated by the perceived social stigma and repercussions associated with cannabis.⁵ These and other factors cause patients to be admitted to the hospital for extensive medical testing and symptomatic management with ineffective initial therapies.⁶

In pediatric patients, the use of marijuana products has significantly increased since 2014, when Colorado became the first state to legalize recreational marijuana. Following these legislative changes, the cases of emergency department visits for adolescent CHS doubled during 1.5 years.^4,7 Moreover, prevalence results from the Monitoring the Future study show statistically significant increases in marijuana use in US students in the 8th, 10th, and 12th grades.^4,8 As of 2021, the lifetime prevalence of marijuana use for students in these grades was 10.2%, 22.0%, and 38.6%, respectively.⁸

Most studies evaluating the effective management of CHS have been performed in adult patients. These studies have found traditional antiemetic therapies, including serotonin (5-HT₃) antagonists and phenothiazines, to be ineffective for the management of this condition because they do not target the cannabinoid receptors 1 and 2, and transient receptor potential vanilloid 1 receptors implicated in CHS.¹ This has led to the use of nontraditional antiemetic medications as first-line therapy for symptomatic management, such as benzodiazepines, haloperidol, and topical capsaicin.⁶

The literature assessing antiemetics used to treat pediatric patients with CHS is limited, with only a handful of studies describing effective treatment modalities. The available studies have described the use of benzodiazepines, haloperidol, and topical capsaicin as effective treatment modalities in pediatric patients with diagnosed CHS.⁴ The primary objective of this study was to identify antiemetics used for the management of pediatric CHS. The secondary objectives of this study were to compare the efficacy and safety of these antiemetics for the treatment of pediatric CHS.

Materials and Methods

As part of the standard Penn State Health (PSH) Institutional Review Board protocol, an external company provided a list of potential individuals to be reviewed for the study using all ICD-10 codes involving cannabis from October 1, 2018, to September 30, 2021, inclusive. Potential patients collected from the electronic health record (EHR) at PSH Children's Hospital were retrospectively evaluated for inclusion in the study based on the following criteria: age ≤18 years; admitted as observation or inpatient status in either the emergency department (ED) or pediatric units; and had a clinical diagnosis of CHS, defined as a minimum frequency of weekly use of marijuana plus at least 1 of the following symptoms: 1) stereotypical episodic vomiting resembling cyclic vomiting syndrome in terms of onset, duration, and frequency; 2) resolution of symptoms with cessation of tetrahydrocannabinol products; 3) relief with external thermoregulation (hot showers); or 4) epigastric or periumbilical pain (Rome IV criteria).⁹ Exclusion criteria included patients admitted to an adult service (i.e., not cared for by a pediatrician), patients who were erroneously included in the report (i.e., did not have an encounter involving cannabis), patients who had a cannabis ICD-10 code but did not receive a formal diagnosis of CHS (i.e., cannabis use disorder), and patients who had an outpatient encounter outside of the ED (e.g., health care visit with institution provider at outpatient clinic).

Patient data extracted from the PSH EHR included patient demographics (age, sex, ethnicity, height, weight), type of admission (ED or inpatient pediatric unit), and the presence of nausea and/or vomiting. Additional data points collected included antiemetic medication(s) received, patient documented response to antiemetic therapy, scheduled versus as-needed prescribed antiemetic therapy, described medication adverse effects, and length of stay. Patient documented response to antiemetic therapy was elucidated from the EHR through chart documentation as either a reduction in subjective nausea or objective recorded vomiting. Efficacy was described as either a partial or full reduction in nausea and/or vomiting symptoms. If multiple antiemetics were ordered for a patient, they were recorded in order of administration time to best capture response to prescribed therapy. Benzodiazepines, butyrophenones, and topical capsaicin have been previously described in literature as effective treatment for pediatric CHS⁴ and were classified in this study as nontraditional antiemetic therapy. All other antiemetics were classified as traditional therapy. One investigator reviewed each patient record to determine eligibility for inclusion in the study, whereas 2 investigators reviewed 5% of the patient charts as a quality assurance marker of data collection accuracy.

Statistical analysis was performed using SPSS v28 (IBM Corp, Armonk, NY). Median values were calculated for demographic data. Frequencies and percentages were also calculated for sex, ethnicity, admission type, and length of stay. Kruskal-Wallis and χ² tests were performed to compare partial or full symptom resolution with traditional and nontraditional antiemetic therapies. A p value of <0.05 was determined to be statistically significant.

Results

From October 1, 2018, to September 30, 2021, a total of 432 patient encounters were screened for inclusion in the study, of which 19 met inclusion criteria (Figure 1). Participants had a median age of 16.8 years, and 68% were female. There were 19 initial patient encounters with 33 repeat encounters involving 9 patients. Demographic information can be found in Table 1.

Figure 1.

Patient selection based on exclusion criteria.

Table 1.

Patient and Admission Characteristics (N = 19)

Parameter	Value
Age, median (range), yr	16.8 (12–18)
Sex, n (%)
Male	6 (31.6)
Female	13 (68.4)
Height, median (range), cm	167 (150.0–182.9)
Weight, median (range), kg	62.3 (36.5–94.3)
Ethnicity, n (%)
White	9 (47.4)
Black or African American	3 (15.8)
Hispanic or Latino	4 (21.1)
Not documented or other	3 (15.8)
Admission type, n (%)
Emergency department	5 (26.3)
Inpatient	14 (73.7)
Length of stay, n (%), days
1	3 (15.8)
2	7 (36.8)
3	2 (10.5)
5	4 (21.1)
6	1 (4.3)
7	2 (8.7)

Traditional antiemetics were ordered 77% of the time, whereas nontraditional antiemetic medications were ordered 23% of the time during the initial patient encounter. Ondansetron was the most frequently ordered traditional antiemetic, whereas lorazepam was the most frequently ordered nontraditional antiemetic. Other traditional antiemetics ordered for patients included phenothiazines (promethazine, prochlorperazine), antihistamines (diphenhydramine, hydroxyzine, cyproheptadine, scopolamine), metoclopramide, and trimethobenzamide. Haloperidol and droperidol were the other nontraditional antiemetics ordered. Capsaicin was briefly ordered for 1 patient during a repeat admission but was immediately discontinued because of a burning sensation and was not included in the efficacy analysis. A list of antiemetic medications ordered by class can be found in Table 2.

Table 2.

Frequency of First Prescribed Medication by Drug Class and Classification of Antiemetics Received

Parameter	Value, n (%)
Antiemetic–Drug Class
First prescribed
5-HT3 antagonist	5 (26.3)
Butyrophenone	2 (10.5)
Phenothiazine	1 (5.3)
Antihistamine	4 (21.1)
Benzodiazepine	5 (26.3)
D2 antagonist	2 (10.5)
Second prescribed
5-HT3 antagonist	11 (78.6)
Benzodiazepine	2 (14.3)
D2 antagonist	1 (7.1)
Third prescribed
Phenothiazine	1 (25)
Antihistamine	2 (50)
D2 antagonist	1 (25)
Fourth prescribed
5-HT3 antagonist	3 (75)
Benzodiazepine	1 (25)
Fifth prescribed
5-HT3 antagonist	1 (50)
Phenothiazine	1 (50)
Sixth prescribed
Phenothiazine	1 (100)
Antiemetic Classification
First
Nontraditional	7 (36.8)
Traditional	12 (63.2)
Second
Nontraditional	2 (14.3)
Traditional	12 (85.7)
Third
Nontraditional	0 (0)
Traditional	4 (100)
Fourth
Nontraditional	1 (25)
Traditional	3 (75)
Fifth
Nontraditional	0 (0)
Traditional	2 (100)
Sixth
Nontraditional	0 (0)
Traditional	1 (100)

5-HT₃, serotonin receptor subtype 3 D₂, dopamine-2 receptor

Clinically significant differences were found between traditional and nontraditional antiemetic agents and response in nausea and vomiting symptoms during the initial encounter (Figure 2). Nontraditional antiemetic medications had a greater reduction for nausea and vomiting symptoms compared with traditional anti-emetics. Patients who received traditional antiemetics continued to have nausea and vomiting symptoms after medication administration more often than those who received nontraditional antiemetics (16 of 35 vs 2 of 10, p = 0.08). Across both initial and repeat encounters, partial or full symptom resolution occurred 59% of the time with traditional antiemetic therapy and 72% of the time with nontraditional antiemetic therapy (Figure 3). No differences were noted between traditional and nontraditional antiemetic prescribing on sex, ethnicity, or length of stay.

Figure 2.

Antiemetic agent prescribed and symptom outcomes after medication administration for initial encounter.

Figure 3.

Percent improvement of symptom resolution (partial or full) following traditional and nontraditional antiemetic administration.

In general, antiemetic medications were well tolerated among patients. One patient reported extrapyramidal effects following administration of metoclopramide, which resolved upon discontinuation of the medication. The patient was subsequently switched to trimethobenzamide and had no further adverse effects. One patient reported burning with topical capsaicin immediately after application and was quickly discontinued after application.

Discussion

To date, this study represents the largest analysis of antiemetic use in pediatric patients with CHS. Patient demographics of this study were consistent with those in other published pediatric CHS studies. Traditional antiemetics were more frequently ordered compared with nontraditional antiemetics. Although statistical significance was not observed, clinically significant improvements in nausea and vomiting symptoms were noted with nontraditional antiemetics compared with traditional antiemetics. Adverse effects were generally reported as mild and well tolerated.

Previous literature has identified more traditional antiemetic therapies, such as ondansetron and promethazine, as being ineffective in the treatment of CHS.^4,6 Prescribers ordered traditional antiemetics almost 77% of the time in this study. However, these agents collectively reduced nausea and vomiting symptoms 59% of the time when ordered. Ondansetron and promethazine were the most prescribed antiemetics and demonstrated partial or full resolution of nausea and vomiting symptoms best of the traditional antiemetic agents (33 of 52 [63.5%] and 12 of 19 [63%], respectively). Diphenhydramine and metoclopramide had less of an overall effect on nausea and vomiting symptoms, with partial or full symptom resolution occurring infrequently (5 of 15 [33.3%] and 5 of 13 [38.5%], respectively).

Conversely, literature-classified effective (nontraditional) antiemetics for the treatment of pediatric CHS were ordered approximately 23% of the time. When assessed collectively, greater improvement in nausea and vomiting symptoms was observed, with partial or full symptom resolution occurring 72% of the time. Lorazepam was the most frequently prescribed appropriate antiemetic and demonstrated improvement of nausea and vomiting symptoms in 14 of 18 orders (78%). This supports described success with benzodiazepines in previous studies.^10–12 Interestingly, haloperidol resolved symptoms in 5 of 9 patients (56%), although droperidol improved nausea and vomiting symptoms in 4 of 4 patients for whom it was prescribed. Like haloperidol, droperidol is a butyrophenone that is thought to provide antiemetic activity through antagonism of dopamine-2 receptors.^13,14 Compared with haloperidol, it is associated with a quicker onset of action and a greater incidence of QTc prolongation, warranting a black box warning. Although droperidol demonstrated improvements in nausea and vomiting in this study, previous studies in adult patients suggest that the antiemetic efficacy of droperidol may not be as favorable as that of haloperidol.¹⁵ At present, droperidol may be useful in the treatment of pediatric CHS and deserves additional assessment.

Several antiemetic medications were identified in this study that have not been studied in the treatment of pediatric CHS to date. Cyproheptadine (5 of 6 [83%]), hydroxyzine (1 of 1 [100%]), scopolamine (2 of 2 [100%]), and trimethobenzamide (1 of 1 [100%]) demonstrated good reduction of nausea and vomiting symptoms. However, the small number of orders for these medications, in addition to droperidol, makes it difficult to draw meaningful conclusions. The observed efficacy of hydroxyzine and scopolamine in this study is inconsistent with previous adult literature.^16,17 Moreover, the use of cyproheptadine and trimethobenzamide has not previously been described in pediatric or adult literature. Due to the very small sample size for these agents, further studies are warranted to describe these agents and their role in the management of pediatric CHS.

All antiemetics were well tolerated with minimal reported adverse effects. While the various antiemetics used were generally safe, nontraditional antiemetics for CHS can have significant adverse effects. Pediatric patients exposed to benzodiazepines may be at a higher risk of experiencing paradoxical effects including hyperactivity, aggressive behavior, and respiratory distress.¹⁸ Haloperidol and droperidol are known to cause QTc prolongation, hypotension, and extrapyramidal symptoms including tardive dyskinesia to varying extents.^13,14,19 Literature suggests the proarrhythmic potential of droperidol may be more pronounced compared to haloperidol.¹⁵ Topical capsaicin can elucidate a burning or tingling sensation when applied to the skin that may be intolerable to some patients,²⁰ as was observed in this study. Appropriate consideration and monitoring of adverse effects should factor into therapy selection for pediatric patients.

This study has several limitations that merit discussion. First, it is a retrospective study that relies on documentation of multiple healthcare providers during a patient encounter. It is possible that inadequate or inaccurate documentation of nausea and vomiting may have led to underreporting of symptoms and overestimation of perceived benefit. Second, the use of ICD-10 codes for identification of patients in this study may not have captured all cases of pediatric CHS within our hospital. All ICD-10 codes involving cannabis were included in the search to control for this limitation as much as possible. Third, despite being the largest study of pediatric CHS to date, the sample size is small (n=19), and study power could not be assessed. Given the nature of this condition, multicenter studies should yield more robust outcomes. Fourth, there is no validated nausea assessment tool for pediatric CHS. Clinical assessment of nausea can be variable and inconsistent from patient to patient and by healthcare providers. Fifth, the criteria for diagnosing CHS in pediatric patients is extrapolated from adult criteria using Rome IV classification. This could affect identification of pediatric CHS cases. The development of pediatric-specific diagnostic criteria is an area for future consideration. Finally, there were instances in which patients had symptoms consistent with diagnoses of CHS, but were excluded from the study as they were not admitted to the institution. These patient encounters may have provided valuable insight into the outpatient antiemetic management of pediatric CHS.

Conclusion

Cannabinoid hyperemesis syndrome is an underrecognized and underdiagnosed condition characterized by cyclic vomiting related to chronic cannabis use. Abstinence from cannabis remains the most effective approach to mitigating morbidity associated with CHS, although medications such as lorazepam or droperidol may have some benefit in managing toxidrome symptoms. Traditional antiemetic prescribing remains a key barrier to effective management of pediatric CHS.

ABBREVIATIONS

CHS

cannabinoid hyperemesis syndrome;

ED

emergency department;

EHR

electronic health record;

PSH

Penn State Health