Ketamine infusion for sedation in a patient on extracorporeal membrane oxygenation (ECMO)

Abstract

Introduction: we sought to explore the efficacy of ketamine in a patient on extracorporeal membrane oxygenation (ECMO) receiving ketamine for sedation by investigating the utility of plasma ketamine concentrations.

Case report: retrospective chart review of one critically ill patient on ECMO

Discussion: This was a descriptive review of serial plasma ketamine concentrations in an ECMO patient. Although no reference plasma concentrations exist in ECMO patients, ketamine levels appeared to be lower than those seen in surgical patients not on ECMO.

Conclusion: At this point, no reference plasma concentrations exist for ketamine in ECMO patients, further research may help understand the effects of ECMO on ketamine disposition and that lower ketamine concentrations may be used for effective analgesia or sedation.

Introduction

Extracorporeal membrane oxygenation (ECMO) is a life support modality for patients with refractory cardiac and respiratory failure who cannot be managed with conventional medical therapies. ¹ Patients receiving ECMO often require analgesics and sedatives to help facilitate ventilator synchrony and to reduce discomfort and oxygen consumption. ² Extracorporeal Life Support Organization (ELSO) guidelines recommend light sedation during cannulation for the first 12–24 h of management but they do not provide any specific agents or dosing strategies. ³

ECMO circuits are known to sequester drugs given the large surface area of the membranes and tubing which can increase the volume of distribution of drugs. ⁴ Additionally, over time, the circuit may become saturated, which can result in increased serum drug levels as no more drug can be sequestered. Importantly, the ECMO circuit may continue to release the drug into the circulation after discontinuation of the drug resulting in unpredictable pharmacological effects. ⁵ Highly protein bound and lipophilic medications are more prone to sequestration in ECMO circuits. ⁵ In this case study, we introduce challenges in providing optimal sedation in a critically ill patient receiving venovenous ECMO (VV-ECMO) support for respiratory failure. Ketamine was added to improve sedation and decrease agitation after other sedatives including propofol, fentanyl, and dexmedetomidine continuous infusions were already initiated. Ketamine plasma levels were measured to determine therapeutic values and degree of sequestration in the ECMO circuit. This case report was approved by local Institutional Review Board (IRB).

Case report

A 63-year-old male with a history of smoking, chronic obstructive pulmonary disease with mild obstruction on pulmonary function testing, major depression, and new diagnosis of lung cancer underwent right thoracotomy in our hospital (BW: 77 kg, BMI: 24.9, normal renal and hepatic function upon admission). His post-operative course was complicated by prolonged intubation due to hypoxemic hypercarbic respiratory failure, pneumonia, and acute respiratory distress syndrome (ARDS). To achieve a target Richmond Agitation Sedation Score (RASS) goal of 0 to −2, he was started on propofol, fentanyl, and dexmedetomidine infusions. He subsequently required intravenous vecuronium for ventilation synchronization. Due to increased oxygen requirement and persistent hypercarbia he was cannulated for VV-ECMO support. Despite dose escalation of multiple sedative agents, he remained agitated and intravenous haloperidol doses were given as needed. After 6 days of ECMO support, intravenous ketamine infusion was added for persistent agitation. He was initiated on 0.1 mg/kg/hour, which was increased to 0.15 mg/kg/hour after 24 h. He remained on this dose for 1 day before this infusion was stopped due to improvement of agitation and vent synchrony. Three days later, ketamine was restarted due to worsening of agitation. Ketamine was started at a dose of 0.2 mg/kg/hour but was quickly up-titrated to 0.5 mg/kg/hour in 24 h for RASS goal of 0 to −2. A day later, this dose was further up-titrated to 0.625 mg/kg/hour. At this rate, ketamine provided adequate sedation. Serial plasma ketamine levels were obtained while on this dose (Table 1). Mean plasma ketamine level was reported to be 402 ng/mL (range: 330–490 ng/mL: reference range in non-ECMO: by Quest Diagnostics Nichols Institute: 500–6500 ng/mL). Ketamine was discontinued after 8 days and the plasma level was reported as undetected approximately 24 h after discontinuation. Ketamine was well tolerated. Patient remained on dexmedetomidine infusion (dose range: 0.2–1.5 mcg/kg/hour) during this time but propofol and fentanyl infusions were tapered off. His respiratory status improved and ECMO was discontinued 4 days after ketamine infusion was stopped.

Table 1.

Sedative doses and plasma levels while on VV ECMO ^a .

Days post ECMO	6	7	8	9	10	11	12	13	14	15	16	17	18	19
Fentanyl infusion range/day (mcg/hr)				50–100	100–300	50
Propofol infusion range (mcg/kg/min)	50	25–50	50	25–50	10–40
Dexmedetomidine (mcg/kg/hr)	1.5	1.5	1.3	1.4	1.0	1.3	1.5	1.2	1	1.5	1.5	1.5	1.5	1.3
Ketamine dose (mg/kg/h)	0.1	0.15				0.2	0.5	0.625	0.625	0.625	0.625	0.625	0.625	0
Ketamine level (ng/mL)	—	—	—	—	—	—	—	—	440	330	360	390	490	ND*

^aVV ECMO information: pump flow rate: 3.98 L/min, rotation per minute (RPM) 3055 L/min, FiO₂: 100%, Sweep flow rate: 3.75 L/min.

*ND: not detected.

Discussion

Pharmacokinetic and pharmacodynamic properties of ketamine make this agent an attractive sedative in ECMO. Compared to commonly used sedatives such as fentanyl, propofol, and midazolam, ketamine has lower lipophilicity and protein binding, making this agent less prone to be removed by the ECMO circuit. ⁶ Despite this, clinical evidence for its use in ECMO patients is limited. Tellor and colleagues reported their experience with ketamine in 26 ECMO patients retrospectively. Median starting infusion rate was 50 mg/hr (range: 6–150 mg) for a median duration of 9 days (range: 0.2–21 days) and other sedative and opioid requirements reduced in 9/26 patients (35%). ⁷ Similarly, Floroff et al. reported dose reduction of midazolam infusion from 16 mg/hour to 8 mg/hour 8 hours after initiation of ketamine in a an ECMO patient. Ketamine was initiated at a rate of 4.2 mg/hour which was titrated to 42 mg/hour in this case. ⁸ On the other hand, a prospective trial of 20 ECMO patients randomized to ketamine infusion plus standard of care versus standard of care alone did not result in dose reduction of other sedatives. In fact, in this study, median cumulative amount of fentanyl and midazolam were higher in the ketamine group. Ketamine was given as an intravenous bolus dose of 40 mg followed by a continuous infusion of 5 mcg/kg/min. ⁹ In our case, ketamine at 0.625 mg/kg/hour seemed to provide desired level of sedation (RASS target of 0 to −2). Although the normal laboratory reference range in our institution was reported to be 500–6500 ng/mL in non-ECMO patients, a review article by Mion et al. described plasma concentrations as low as 70 ng/mL as adequate for memory alteration and analgesic effects at steady state concentrations of 100–260 ng/mL. ¹⁰ This may explain our findings despite lower than laboratory reference range levels seen in our patient.

He remained on dexmedetomidine during this time but other sedatives were tapered off. To our knowledge this is the first study to report plasma ketamine levels in a patient receiving ECMO.

Our study has a few limitations. First, although ketamine levels were collected, these levels resulted few days after collection and real time intervention was not possible. Second, given the descriptive nature of the study and sample size of one, we cannot make recommendations regarding optimal dose of ketamine in ECMO. Third, only total ketamine levels (and not free levels) were monitored and therefore we are unable to assess how changes in protein binding in acute illness and ECMO could have impacted our outcomes. Finally, although ketamine helped with the tapering of other sedatives, other modalities such as antibiotic therapy, resolution of pneumonia, and general respiratory care combined with dexmedetomidine could have contributed to this outcome.

Conclusion

Pharmacokinetic properties of ketamine make this agent an attractive sedative in patients receiving ECMO. At this point, no reference plasma concentrations exist for ketamine in ECMO patients, further research may help understand the effects of ECMO on ketamine disposition and that lower ketamine concentrations may be used for effective analgesia or sedation.