Evaluation of the Development of Tolerance Following Frequent Administration of Propofol in Children Requiring Successive Sedation for Separate Procedures

The delivery of sedation by non-anesthesiologists is continuing to emerge globally. Propofol (2,6-diisopropylphenol) is one of the widely used sedative-hypnotic medications for short procedures that require an uninterrupted motionless activity for pediatric patients. Its quick recovery time and rapid induction with an easily titratable depth of sedation make it a preferred medication for outpatient procedures. Propofol sedation by non-anesthesiologists for outpatient procedures has been shown to be safe and cost efficient. In an extensive database review of 49,836 propofol sedations from 37 locations, Cravero et al²  found an overall complication rate of 5.3%. Similar reports²  of sedations performed by anesthesiologists have shown complication rates beginning at 1.3%. However, there is variation in the definition of serious adverse events between sedations in different studies. Despite having an excellent safety profile for procedural sedation, at larger doses and longer durations propofol is not a benign drug, particularly in small children. In those patients, it is associated with the development of metabolic acidosis, myocardial failure, rhabdomyolysis, and death (propofol-related infusion syndrome),³  so it is essential that clinicians be mindful of escalating doses or total cumulative doses of propofol within a short period. The FDA has a “black box” warning on the use of propofol for continuous sedation in the setting of the pediatric ICU.⁴  Even for outpatient sedation, total propofol dose was found²  to be independently associated with the risk of complications in a large retrospective study.

Drug tolerance is a progressively diminished response to a drug with repeated exposure. Although the development of tolerance to drugs that depress the central nervous system is common, tolerance to the effects of propofol is not well recognized. The effect of propofol is mediated by the gamma-aminobutyric acid receptor. This is also an important mechanism related to abusive substances (e.g., alcohol, barbiturates, benzodiazepines) that are also associated with the development of tolerance following repeated usage.⁵  Similar to opioids and benzodiazepines, propofol also acts through specific cell surface receptors and alters chloride conductance in the central nervous system, potentially causing tolerance on repeated use. Sudden discontinuation of propofol after use for more than 1 day has been shown⁶  to cause an acute withdrawal response. Because withdrawal phenomena are closely associated with agents that also result in tolerance with long-term use, propofol can be expected to display the same pattern.⁷  Propofol is used frequently for outpatient sedation in children; however, the development of tolerance to the sedating effects of propofol is not well characterized. There have been reports suggesting the development of tolerance on repeated usage of propofol both in human and animal models^8–12  and some recent reports arguing against the development of tolerance.^2,5 

Our institution almost exclusively uses propofol for outpatient sedation for non–pain-inducing procedures like MRI or radiation therapy sedation in children. Radiation therapy patients typically require multiple rounds of radiation in a short period and do not require other adjuvant medications; thus, they form an ideal group of the population in which to analyze the impact of multiple procedures under propofol sedation. The propofol-based sedations in our practice are highly protocolized, thereby providing minimal provider-related variation in terms of the dosing.

This study was designed with a specific aim of identifying if frequent administration of propofol in children undergoing outpatient sedation by non-anesthesiologists leads to the development of tolerance, as identified by the requirement of successive higher doses of propofol.

OSF Children's Hospital of Illinois–OSF HealthCare is an affiliate clinic to the St Jude Children's Research Hospital, and most of the patient data were obtained from this population. These data were extracted from the electronic medical record system from January 2011 through April 2019.

All patients who underwent 3 or more sedations during the study period were included. In our institution, all the patient procedural information during each sedation is collected by sedation nurses on a real-time data collection sheet that is then scanned into the patient's chart.

All patients who were either admitted or were scheduled as an outpatient for radiation planning, radiation chemotherapy, or wound dressing change were pre-screened for deep sedation by either a pediatric intensivist or a hospitalist. The most common procedures that were selected as meeting the inclusion criteria were radiation chemotherapy, wound dressing change, and CT planning for radiation therapy. Before the first encounter, a plan was established for each patient undergoing sedation that would enable administration of the smallest dose that could achieve the desired level of sedation. Sedation was induced by a loading dose of propofol ranging from 2 to 4 mg/kg followed by a continuous infusion ranging from 1 to 7 mg/kg/hr. The dose was titrated up or down to maintain deep sedation, as defined by American Society of Anesthesiologists practice guidelines.¹³  Once the initial sedation was successful for the patient, a similar protocol was carried out for subsequent sedations unless larger doses were needed to achieve the desired effect. Regardless of prescriber, this protocol remained the same for different sedation events.

The sedation procedures were mainly performed on a daily or sometimes on an alternate-day basis. The procedures generally lasted from 5 to 70 minutes, and the duration of each session was similar for that specific patient each time. Patients were monitored using the 2002 practice guidelines from the American Society of Anesthesiologists¹³  throughout the procedure until they displayed recovery from sedation. Some children undergoing sedation also had underlying anxiety or pain and were prone to emesis as an adverse effect of radiation treatments; hence, based on underlying conditions, each patient was provided additional medications as needed in similar doses across each sedation event.

The patient's demographic characteristics included patient age, weight, and sex. Other data included diagnosis, type of procedure, American Society of Anesthesiologists class, induction, and total dose with an interval time between sedations. After being abstracted from the electronic medical records the data were compiled in Excel. The induction and total dose and total time of sedation were recorded. Each sedation procedure for each patient was considered a separate event. Statistical analysis was conducted using JMP Pro version 14.0 and SAS version 9.4 (SAS Institute, Cary, NC). The average propofol dose for the serial number of sedations (i.e., serial No. 1 is the first sedation event of all patients; serial No. 2 is the second sedation event for all patients, and so on) and average propofol dose for all sedations by the patient were calculated and displayed as mean ± SD. A mixed model with a random intercept was used to account for correlation within subjects. Age, sex, additional medications given, and the time interval between sedations were included in the model for multivariate analysis. A p value of <0.05 was considered statistically significant.

We reviewed 255 separate sedation events (radiation chemotherapy [n = 220, 86.2%], CT/radiation planning [n = 6, 2.3%], and wound dressing [n = 29, 11.35]) in 24 different patients who received procedural sedation between January 2011 and April 2019. Age ranged from 0.8 to 20 years (median = 4 years [IQR: 2.7. 6.0]). The median weight of the patient was 16 kg (IQR: 13.8–22; range 10–108.8 kg). There were 16 males (66.6%) and 8 females (33.3%). The largest number of patients had diagnoses of Wilms tumor (45.8%) and neuroblastoma (25%). Other diagnoses included Ewing's sarcoma (n = 1), rhabdomyosarcoma (n = 1), congenital heart block (n = 1), pacemaker infection (n = 1), trauma to leg (n = 1), small bowel perforation with fistula (n = 1), left heel trauma (n = 1), right popliteal artery injury (n = 1), and right knee dislocation (n = 1).

The mean ± SD time for sedation administration for procedural events was 36.8 ± 16.1 minutes (range: 5–73 minutes). Out of all of the sedation events, fentanyl (Hospira, Lake Forest, IL) was used for pain in 16 patients (6.2%), and ondansetron was given to 12 patients (4.7%) for nausea and vomiting. When necessary, adjuvant medications were used in the same doses for each subsequent sedation event.

Patient-specific description of age, sex, the interval between first and last sedation, total sedation, and mean total dose and mean induction dose are shown in Table 1. The median interval between the first and last sedation events of patients was 16 days (IQR: 8.5, 22.5), with a range from 6 to 126 days. The median number of sedations was 10.5 (IQR: 6.2, 12) and ranged from 3 to 28. The mean total dose of propofol for all sedations was 0.19 ± 0.14 mg/kg/min, and the mean induction dose was 3.2 ± 0.97 mg/kg. The mean time for all sedation procedures was 36.8 ± 16.1 minutes and ranged from 5 to 73 minutes (Table 1).

The mean total propofol rate (mg/kg/min) and induction dose (mg/kg) for each serial number of sedation and the total number of sedations for each serial number are shown in Table 2. There were a total of 24 patients; however, data on time for 2 patients for sedation serial No. 1 and 1 patient for sedation serial No. 2 were not available; thus, data for 21 patients for sedation serial No. 1 and 23 patients for sedation serial No. 2 were analyzed. The total number of sedations per serial number decreased as expected, with only one patient requiring more than 20 sedations (Table 2).

Mixed-model analysis of the total propofol dose (p = 0.089) and induction doses of propofol (p = 0.180) for serial numbers including variables of age, sex, additional medication, and time interval did not show a significant trend of increase in propofol dose by serial number (Figure). There was, however, a statistically significant trend of decrease in propofol dose per unit of time between 2 sedations. The cumulative propofol dose decreased by 0.003 mg/kg/min for every one more day between 2 sedations (parameter estimate, −0.003; CI: −0.0045 to −0.0017; p value < 0.0001).

In this article, we have described our experience with 24 patients who required 3 or more outpatient procedures with propofol sedation and the impact of multiple sedations on the development of propofol tolerance. Our results indicate that the frequent administration of propofol does not lead to a development of tolerance, as defined by an increased requirement of propofol for induction or maintenance of sedation.

Cohen et al⁹  evaluated 30 adult patients undergoing propofol sedation for 5 days of successive electroconvulsive therapy at an outpatient psychiatric clinic. They observed that 13 of 30 patients required more than a 100% increase in the dose of propofol to achieve the same sedation starting from their third treatment onward and concluded that the repeat injections of propofol under deep sedation could induce a tolerance-like reaction to the drug. More recent reports on children, however, have not shown this behavior from propofol.

Anghelescu et al²  evaluated the propofol sedation for radiotherapy sessions at St Jude Children's Research Hospital from 2004 to 2006. They did not find any evidence of the development of tolerance to propofol with subsequent sedations. They found a negative correlation between the total propofol dose per session and the number of previous propofol-based anesthesia sessions. They believe that this difference was reflective of a “gradual dose-finding” clinical approach by anesthesia providers. In a recent, similarly sized study from South Korea,⁵  investigators evaluated the propofol dosing on all patients undergoing proton radiation therapy. They identified 58 children with an average of 19 sedations per patient. They found that the repeated prolonged deep sedations for multiple radiation therapy over several weeks did not lead to the development of tolerance in the majority of children. However, in their study, 15 of 58 (26%) patients showed an escalating requirement of propofol with multiple sedations.

For their overall patient population, the odds ratio for the risk of tolerance based on the number of treatments was 0.99 (CI: 0.90–1.07; p = 0.89). They also did a case-control evaluation of the 2 groups. They did not find any significant difference between the children who developed tolerance and the children who did not develop tolerance. Both of these studies included sedations performed by anesthesia providers. While it is possible that there are differences in methodology, monitoring, and dosing in sedations performed by anesthesia and non-anesthesia providers, our results correlate closely with their conclusions.

Larsson and Wahlström¹⁰  evaluated the impact of propofol on the development of tolerance in rats of different ages (20 ‘young' and 20 ‘old' rats). They observed that maintenance of propofol anesthesia alters the sensitivity to propofol (development of tolerance) in older animals, but young animals did not develop acute tolerance to propofol. This finding has not been replicated in other studies, including clinical studies by Kang et al⁵  and our study.

We observed a slight decrease in the total mean propofol rate requirement by the duration of the interval between the 2 doses. In an experiment on the effect of propofol on rats, Fassoulaki et al⁸  observed a decrease in sleeping times with a single dose of propofol at 24 and 48 hours, but not at 72 hours, after the first dose. This effect has not been replicated in other laboratory and clinical studies to date.

The mean time of propofol sedation in our study was 36.8 minutes; this may not be long enough for the development of tolerance. In a cost-benefit analysis of propofol versus midazolam sedation in the ICU, investigators¹⁴  found the development of tolerance after 144 hours of infusion. There are currently no large studies on the development of propofol tolerance when the drug is administered as a continuous infusion in the ICU. A small number of patients who required fentanyl may have influenced the level of sedation. However, this effect was mitigated by maintaining similar doses of fentanyl throughout the subsequent sedations for an individual patient while titrating propofol to achieve the desired level of sedation.

Limitations of the present study include its retrospective nature; the fact that it represents single-center experience, with a small patient population; and use of a non-standardized database with some variables that were challenging to objectify. Because we relied on the procedure nurses to record data, some data points were not recorded in the chart and were missing. The amount of missing data, however, was very minimal and compared favorably to that seen in other retrospective chart reviews. There was considerable variation in duration between successive sedations, and different sedations were performed by different providers. However, the use of a sedation protocol negates the providers' related variations to some extent.

Repetitive use of propofol in children for outpatient sedation by non-anesthesiologists does not lead to the development of tolerance to propofol. There is a small statistically significant impact of time on successive doses of propofol. The difference in propofol dosing per day was not significant enough to be clinically relevant for outpatient sedation; however, it may be significant in inpatient settings, such as those involving sedation in the PICU.

CT

computed tomography;

FDA

US Food and Drug Administration;

ICU

intensive care unit;

MRI

magnetic resonance imaging