Serotonin Syndrome in the Intensive Care Unit: Clinical Presentations and Precipitating Medications

Transcription

1 Neurocrit Care (2014) 21: DOI /s ORIGINAL ARTICLE Serotonin Syndrome in the Intensive Care Unit: Clinical Presentations and Precipitating Medications Swetha Pedavally Jennifer E. Fugate Alejandro A. Rabinstein Published online: 20 September 2013 Ó Springer Science+Business Media New York 2013 Abstract Background Serotonin syndrome (SS) is becoming a more frequent diagnosis in the intensive care unit (ICU). We sought to determine the clinical presentation, drug exposures, and outcomes of SS in critically ill patients. Methods A retrospective study of 33 consecutive ICU patients with SS between March 2007 and March 2012 in ICUs in a large teaching hospital. SS was defined using the Hunter Serotonin Toxicity Criteria. Results Seventeen patients (52 %) were admitted for mental status changes, including seven patients (21 %) with drug overdose and four cases (12 %) in which SS was considered the primary admission diagnosis. In 13 patients (39 %) the features of SS developed only after a mean of 6.8 ± 9 days of hospitalization. Most received multiple serotonergic drugs upon diagnosis (median three drugs, range 1 5). Antidepressants were the serotonergic medications most often used before admission, and opioids (principally fentanyl) and antiemetics were the most frequently prescribed new serotonin-enhancing medications. Altered mental status was present in all patients and myoclonus, rigidity, and hyperreflexia were the most prevalent examination signs. All but one patient had documented recovery. The mean time to neurological improvement was 56 ± 5 h, but ranged from 8 to 288 h. There were no cases of renal failure related to rhabdomyolysis, or death or persistent disability caused by SS. Conclusion SS in the ICU occurs most often because of exposure to multiple serotonergic agents. Continuation of S. Pedavally J. E. Fugate A. A. Rabinstein (&) Division of Neurocritical Care, Department of Neurology, Mayo Clinic, 200 First Street SW, Mayo W8B, Rochester, MN 55905, USA rabinstein.alejandro@mayo.edu antidepressants plus the addition of opioids and antiemetics during hospitalization are most commonly responsible for this complication. Keywords Serotonin syndrome Neurotoxicity Antidepressants Opioids Intensive care unit Introduction Serotonin syndrome (SS) is an adverse drug reaction caused by excessive activation of postsynaptic serotonin receptors [1 3]. It primarily affects the central nervous system and is characterized by altered mental status, signs of neuromuscular irritability, and autonomic instability [1, 4]. Although it can occur because of intentional or accidental overdose with a single serotonergic agent, most severe cases result from the interaction of two or more drugs that enhance serotonin transmission [1, 4 7]. The range of clinical severity can vary from mild to lifethreatening [5, 8, 9], and resolution requires discontinuation of all serotonergic agents. Thus, delayed recognition can have serious consequences [10]. Patients with SS can present to the intensive care unit (ICU) with a diagnosis of unexplained encephalopathy or with a toxidrome. However, SS can also develop in patients who are in the ICU for a different diagnosis and receive serotonergic medications. The rapidly growing use of serotoninergic medications for various indications (depression, chronic pain, emesis, migraine, among others) and their frequent use as part of a polypharmacy may explain why we are encountering this adverse reaction more often among patients in the ICU. Literature on SS in the ICU is very scarce and mostly limited to very small case series and reviews [10 13]. As a

2 Neurocrit Care (2014) 21: consequence, there is insufficient information on the characteristics of SS among critically ill patients, the drugs implicated in the pathogenesis of SS in the ICU, and the outcomes of the syndrome in this specific population. Patients in the ICU can be exposed to medications and drug interactions that differ from those typically found in the outpatient setting [14]. They can also be at greater risk of secondary complications which may worsen the outcome. In this study, we sought to improve our knowledge of SS in the ICU by analyzing the clinical presentation, evolution, and outcome of a consecutive series of critically ill patients with SS. Patients and Methods We used the Mayo Clinic Medical Records Linkage system to identify consecutive cases of possible SS by searching for the term serotonin syndrome in the records of patients hospitalized at one of the Mayo Clinic, Rochester hospitals between March 2007 to March With this approach, any case where the examining physician mentioned SS was identified (n = 148). These medical records were then manually reviewed to determine whether the patient met inclusion criteria for SS while in an ICU. We relied on the Hunter s toxicity criteria for the definition of SS (Table 1) [4]. Patients suspected of having SS in the ICU but not meeting these criteria (n = 9), SS cases attended in the general wards (n = 24) and patients in which SS was considered as a differential diagnosis but then dismissed (n = 82) were excluded from the analysis. Consequently, our final study population consisted of 33 critically ill patients with SS admitted to one of the ICUs of a tertiary referral center. We collected information on patients age, gender, admission diagnosis, APACHE-3 scores upon ICU admission, vital signs and physical examination findings on the day of SS diagnosis, laboratory values (serum concentrations of lactic acid, CPK, creatinine, AST, ALT, and white blood cell count) both on the day of SS diagnosis and when at their peak during the course of SS, requirement of mechanical ventilation, treatments prescribed for SS, secondary complications during the hospitalization, and functional outcome Table 1 The hunter serotonin toxicity criteria Exposure to a serotonin agent plus one of the following Spontaneous clonus Inducible or ocular clonus and agitation or diaphoresis Inducible or ocular clonus and increased muscle tone and temperature >38 C Tremor and hyperreflexia at discharge and at 3-month follow-up. We gathered all the drugs with reported serotonin enhancing activity received by the patients before the diagnosis of SS. We defined tachycardia as a heart rate >100 beats per minute, hypertension as blood pressure >140/90 mmhg, tachypnea as respiratory rate >20 breaths per minute, and fever as body temperature >38 C (to be consistent with the Hunter criteria) [4]. Presence of agitation, clonus, muscle hypertonicity, tremor, hyperreflexia, and diaphoresis were extracted from detailed review of the medical and nursing notes. We also noted if the patient had fluctuating or depressed level of consciousness, inattention, or disorganized thinking; presence of any of these features or agitation were considered manifestations of abnormal mental status. Time to improvement was defined as the interval between diagnosis and recovery of ability to volitional interaction. Functional outcome was noted at discharge and 3 months later using the modified Rankin scale. We used descriptive statistics as pertinent to present the data. A Fisher exact test was used to assess for a statistical difference in functional outcome between patients presenting with SS versus those who developed it in the ICU. The study was approved by the Mayo clinic Institutional Review Board Rochester, Minnesota. Results Median age in our cohort was 42 years (18 71 years) and 24 (73 %) were women. The various reasons for admission to the hospital are shown in Table 2. Seventeen patients (52 %) were admitted primarily for alterations in the level or content of consciousness, including seven patients (21 %) with drug overdose and four cases (12 %) in whom SS was considered the most likely diagnosis upon ICU admission. Table 2 Reasons for admission Reason for hospital admission n (%) Alteration in consciousness 17 (52) a Sepsis syndrome 4 (12) Resuscitation after cardiac arrest 3 (9) Severe pneumonia 2 (6) Tumor-related complications 2 (6) Emergency surgery 2 (6) Graft versus host disease 1 (3) Liver failure 1 (3) Trauma 1 (3) a Including seven patients with drug overdose and four patients in whom the diagnosis of SS was deemed the most likely explanation upon ICU admission

3 110 Neurocrit Care (2014) 21: Among the seven cases presenting after drug overdose, a single serotonergic agent had been ingested by four and a combination of two agents by three. The unit harboring most of these patients was the medical ICU (22 patients, 67 %). Mean admission APACHE-3 score was 67 ± 30 (median 58.5, range ). Eighteen patients (55 %) required mechanical ventilation. Six patients (18 %) had undergone surgery before the diagnosis of SS. Thirteen patients (39 %) developed the features of SS after hospitalization. The mean duration of hospitalization prior to the documentation of features of SS was 6.8 ± 9 days (median 4, range 1 35). Overall, SS was diagnosed within 24 h of first documentation of clinical features fulfilling the diagnostic definition in 31 patients (94 %) and within 48 h in the other two patients. The serotonergic drugs implicated are shown in Table 3. Twenty-two different drugs with serotonergic activity were identified. Patients were exposed to a median of three serotonergic drugs (range 1 5). Twenty-three patients (70 %) received one or more new serotonergic drugs once in the hospital. Antidepressants (selective serotonin-reuptake inhibitors and serotonin norepinephrine reuptake inhibitors) were the medications most commonly used by the patients before hospitalization. Opioids and antiemetics were the most commonly added serotonergic medications in the hospital. Twenty-two patients (67 %) were exposed to a combination of an antidepressant and an opioid or an antiemetic. Serotonergic medications that were part of the outpatient pharmacological regimen were stopped upon hospitalization in 20 patients (61 %); in the remainder those medications were only discontinued after SS diagnosis. Among the 13 patients who developed SS once in the ICU, 11 received opioids before the diagnosis (sometimes along with tramadol or antiemetics), including all patients who developed SS more than 48 h after ICU admission. The symptoms and signs present upon diagnosis of SS are summarized in Table 4. All patients had abnormal level of content of consciousness including 42 % who were agitated. Fever was present in less than half and hypertension was uncommon. Clonus (inducible or spontaneous), hyperreflexia, and muscle hypertonicity were very prevalent. Predominant rigidity in the legs was noted in 24 %. At the time of diagnosis of SS, 7/18 patients in whom serum lactic acid was measured had lactic acidosis (median 3.3 mmol/l, range ) and elevated CPK was documented in 12/26 patients with available measurements (median 1,524 U/L, range ,085). Leukocytosis was present in 14 cases (42 %) (median /L, range /L /L). Abnormal serum creatinine concentration was observed in eight patients (24 %) and abnormal liver transaminases in were noted in 11 patients (33 %). All laboratory abnormalities peaked within 48 h of Table 3 Serotonergic medications implicated in our 33 cases of serotonin syndrome in the ICU Serotonergic drug Before admission Added after admission SSRI 25 (76 %) 1 (3 %) Citalopram 7 (21 %) Fluoxetine 6 (18 %) 1 (3 %) Sertraline 5 (15%) Escitalopram 4 (12 %) Paroxetine 1 (3 %) SNRI 10 (30 %) Venlafaxine 8 (24 %) Duloxetine 2 (6 %) Other antidepressants 9 (27 %) 1 (3 %) Trazodone 5 (15 %) 1 (3 %) Mirtazapine 3 (9 %) Buspirone 1 (3 %) Clomipramine 1 (3 %) Antiemetics 4 (12 %) 7 (21 %) Ondansetron 4 (12 %) 7 (21 %) Metoclopramide 1 (3 %) 1 (3 %) Granisetron 1 (3 %) Opiates 3 (9 %) 21 (64 %) Fentanyl 1 (3 %) 19 (58 %) Tramadol 1 (3 %) 3 (9 %) Meperidine 2 (6 %) Methadone 1 (3 %) Others Sumatriptan 3 (9 %) Lithium 2 (6 %) Amphetamine 1 (3 %) Methylene blue 1 (3 %) SSRI selective serotonin-reuptake inhibitors, SNRI serotonin-norepinephrine reuptake inhibitors SS diagnosis. There were no cases of renal failure induced by rhabdomyolysis. Treatment consisted of removal of all serotonergic agents, supportive measures and administration of benzodiazepines in all cases. Cyproheptadine was prescribed to 13 patients (39 %) for a mean duration of 3.5 days with an average maximum dose of 21.7 mg daily (range 6 25 mg daily). Chlorpromazine was used in only one case. Eight patients (24 %) were treated with non-depolarizing neuromuscular blocking agents. The most common complication during the ICU stay was hospital acquired pneumonia, which occurred in ten cases (30 %). Mean time to improvement was 56 ± 5h (median 24, range h) and improvement occurred in all but one patient who died from end-stage heart failure without regaining his ability to interact. Mean length of stay in the ICU was 6.5 ± 8.8 days. There were four in-

4 Neurocrit Care (2014) 21: Table 4 Symptoms and signs of serotonin syndrome upon diagnosis Symptom/sign n (%) Altered metal status 33 (100) Agitation 14 (42) Tachypnea 23 (70) Tachycardia 20 (61) Fever 14 (42) Hypertension 4 (12) Clonus (inducible or spontaneous) 29 (88) Ocular clonus 3 (9) Hyperreflexia 24 (72) Rigidity 26 (79) Predominant in the legs 8 (24) Mydriasis 14 (42) Tremor 12 (36) Diaphoresis 7 (21) hospital deaths, all unrelated to SS or to direct complications from it. Functional recovery among survivors was generally but not uniformly favorable. At 3 months, full functional recovery was noted in 18 patients (55 %), mild to moderate symptoms with independent function in four (12 %), moderate disability in six (18 %), and delayed death due to respiratory failure in one (3 %). In all patients with persistent disability, the functional impairment was caused by underlying primary disease and comorbidities rather than by complications related to SS. Functional outcome did not differ between patients admitted with features of SS versus those who developed during the ICU stay (P = 0.16). All but one of the survivors had been restarted on a serotonergic agent (typically an antidepressant) upon last follow-up. The mean interval from diagnosis of SS to first reintroduction of a serotonergic medication was 21.6 days. No episodes of recurrence of SS were documented. Discussion We present the first sizable series of patients with SS requiring ICU care. It includes patients admitted for the manifestations of SS as well as patients admitted for other primary diagnosis who developed SS while in the ICU. Most patients were exposed to multiple serotonergic medications, often including some started in the hospital. Antidepressants (especially selective serotonin-reuptake inhibitors, venlafaxine, and trazodone) were the predominantly implicated serotonergic agents in the pharmacological regimes preceding the hospitalization, and opioids (particularly fentanyl) and antiemetics were the most common serotoninenhancing drugs added in the ICU. In fact, the great majority of patients who did not present with a drug overdose were exposed to a combination of an antidepressant and an opioid or an antiemetic. Almost all patients improved with supportive therapy thanks to prompt recognition and discontinuation of all serotonergic medications, but recovery was sometimes slow. In reviewing these cases, we were impressed by the variety of initial presentations. Some patients presented after an overdose with a single or multiple drugs. Others presented with encephalopathy and signs of neuromuscular irritability which led to the correct diagnosis upon first evaluation or, more frequently, within the following day. Yet, a subgroup of our patients were admitted for an unrelated disease and developed SS in the ICU after being kept on their outpatient serotonergic medications and receiving additional serotonergic agents in the hospital. These potentially preventable cases deserve special attention because transient discontinuation of antidepressants and chronic analgesics, and selective and cautious use of opioids and antiemetics in patients with previous exposure to serotonergic drugs can avert this complication [1, 10]. Using opioids without known serotonergic properties (such as morphine) could be a reasonable option in patients with severe pain [15, 16]. When ICU patients develop mental status changes accompanied by signs of neuromuscular disinhibition (clonus, hyperreflexia, rigidity, tremors), it is crucial that clinicians think of SS and promptly check the medication list to stop all serotonergic drugs without delay [5, 13]. In our cohort we did not find any cases related to the use of monoamine oxidase inhibitors, the first agents associated with SS in the medical literature [17]. This is because the use of this class of antidepressants has declined markedly in recent years. Also, we did not find cases related to the use of linezolid, which also inhibits the enzyme monoamine oxidase [18]. Local practices restricting and carefully monitoring the use of this antibiotic may explain the absence of linezolid-related cases of SS. We identified one case of SS triggered by the administration of methylene blue, a known monoamine oxidase inhibitor [19], for treatment of ifosfamide toxicity. There has been debate in the literature regarding the best criteria to use for the definition of SS. We employed the Hunter Serotonin Toxicity Criteria, which has undergone the most solid validation and is most often preferred [4, 10]. We found that the majority of patients with a clinical diagnosis of SS documented in the medical records fulfilled these formal criteria and could be included in this study. Abnormal mental status was uniformly present upon diagnosis, but agitation was not very prevalent. Clonus, rigidity, and hyperreflexia were the predominant neurological signs. Fever was present in less than half of the patients and hypertension was uncommon at the time of SS

5 112 Neurocrit Care (2014) 21: diagnosis. Thus, the absence of fever should not deter clinicians from making the diagnosis of SS. Although there are no pathognomonic laboratory test abnormalities in SS we collected information to evaluate for the presence of lactic acidosis, rhabdomyolysis, leukocytosis, and renal and liver dysfunction because critically ill patients might be at greater risk of developing complications from SS. Abnormalities were noted in a minority of patients, peaked within 48 h of diagnosis, and were not consequential. Specifically, we did not observe any cases of acute renal failure caused by severe rhabdomyolysis. Our patients were managed with supportive measures, sedation with benzodiazepines, and, when necessary, neuromuscular paralysis with non-depolarizing blockers to control the rigidity and hyperthermia and facilitate adequate ventilation. Cyproheptadine, a serotonin antagonist, has been proposed to alleviate the manifestations of SS [20, 21], but to our knowledge it has not been proven to improve patient outcomes. We used cyproheptadine in selected cases based on our practice to start this agent when SS is deemed particularly severe. This selection bias impeded us to determine the value of cyproheptadine (or other anti-serotonergic medications) for the treatment of SS. Nearly all patients had resolution of the manifestations of SS, except for one patient who never regained awareness and was transitioned to palliative care because of heart failure deemed intractable. Yet, time to improvement (defined by recovery of volitional interaction) varied widely. While most patients improved within 2 3 days, some improved much more slowly. Therefore, time to recovery from SS in critically ill patients can be prolonged [13] and this information should be taken into account when estimating neurological recovery in these patients. Eventually, patients admitted for SS had good functional outcomes and in patients with other primary diagnoses the functional outcome depended on the evolution of the primary disease. We acknowledge that our study has limitations. We required that cases meet the Hunter Serotonin Toxicity Criteria and relied on the medical and nursing notes to determine if the cardinal signs were present. Therefore, it is possible that we may have under-recognized cases of SS during the study period. This is even more likely when considering that physician awareness of SS is suboptimal [22]. Some signs of SS may be less commonly noted or documented, such as ocular clonus, tremor, or diaphoresis, and this may have affected the rates of these features presented in our results. For instance, the predominance of rigidity in the legs more than in the arms is a useful sign to identify cases of SS [11, 23]. However, this gradient in the severity of rigidity may not be consistently documented in the records. The Hunter s Criteria were not specifically developed for use in the ICU and their specificity in ICU patients has never been tested. However, we are confident that the patients included in this series had definite SS based on their clinical evolution after discontinuation of the implicated drugs. The lack of a well validated scale to grade the severity of SS made it impossible for us to evaluate the response to a specific treatment. Other diagnosis (such as sepsis) can mimic SS in critically ill patients. However, strict reliance on the Hunter s criteria to define the diagnosis of SS and knowing the evolution of the patients after discontinuation of serotonergic drugs makes misclassification unlikely. Intensivists should be cognizant of SS because of its preventable nature and favorable prognosis when timely recognition is followed by immediate discontinuation of all serotonergic drugs. Avoiding the unnecessary continuation of outpatient polypharmacy (including antidepressants) and exercising caution when prescribing opioids (particularly fentanyl) and antiemetics to patients previously exposed to serotonergic medications are simple measures that can be taken to reduce the burden of this complication in the critical care population. Conflict of interest Swetha Pedavally, Jennifer Fugate, and Alejandro Rabinstein declare that they have no conflict of interest. References 1. Sternbach H. The serotonin syndrome. Am J Psychiatry. 1991;148: Boyer EW, Shannon M. The serotonin syndrome. N Engl J Med. 2005;352: Isbister GK, Buckley NA. The pathophysiology of serotonin toxicity in animals and humans: implications for diagnosis and treatment. Clin Neuropharmacol. 2005;28: Dunkley EJ, Isbister GK, Sibbritt D, et al. The Hunter Serotonin Toxicity Criteria: simple and accurate diagnostic decision rules for serotonin toxicity. QJM. 2003;96: Mills KC. Serotonin syndrome: a clinical update. Crit Care Clin. 1997;13: Mitchell PB. Drug interactions of clinical significance with selective serotonin reuptake inhibitors. Drug Saf. 1997;17: Keegan MT, Brown DR, Rabinstein AA. Serotonin syndrome from the interaction of cyclobenzaprine with other serotonergic drugs. Anesth Analg. 2006;103: Neuvonen PJ, Pohjola-Sintonen S, Tacke U, Vuori E. Five fatal cases of serotonin syndrome after moclobemide citalopram or moclobemide clomipramine overdoses. Lancet. 1993;342: Bronstein AC, Spyker DA, Cantilena LR Jr, et al Annual report of the American Association of Poison Control Centers National Poison Data System (NPDS): 29th Annual Report. Clin Toxicol (Phila). 2012;50: Bienvenu JO, Neufeld KJ, Needham DM. Treatment of four psychiatric emergencies in the intensive care unit. Crit Care Med. 2012;40: Nijhawan PK, Katz G, Winter S. Psychiatric illness and the serotonin syndrome: an emerging adverse drug effect leading to intensive care unit admission. Crit Care Med. 1996;24: Chechani V. Serotonin syndrome presenting as hypotonic coma and apnea: potentially fatal complications of selective serotonin receptor inhibitor therapy. Crit Care Med. 2002;30:473 6.

6 Neurocrit Care (2014) 21: Torre LE, Menon R, Power BM. Prolonged serotonin toxicity with proserotonergic drugs in the intensive care unit. Crit Care Resusc. 2009;11: Gollapudy S, Kumar V, Dhamee MS. A case of serotonin syndrome precipitated by fentanyl and ondansetron in a patient receiving paroxetine, duloxetine, and bupropion. J Clin Anesth. 2012;24: Gillman PK. Monoamine oxidase inhibitors, opioid analgesics and serotonin toxicity. Br J Anaesth. 2005;95: Rastogi R, Swarm RA, Patel TA. Case scenario: opioid association with serotonin syndrome: implications to the practitioners. Anesthesiology. 2011;115: Denton PH, Borrelli VM, Edwards NV. Dangers of monoamine oxidase inhibitors. BMJ. 1962;2: Lawrence KR, Adra M, Gillman PK. Serotonin toxicity associated with the use of linezolid: a review of postmarketing data. Clin Infect Dis. 2006;42: Ramsay RR, Dunford C, Gillman PK. Methylene blue and serotonin toxicity: inhibition of monoamine oxidase A (MAO A) confirms a theoretical prediction. Br J Pharmacol. 2007;152: Graudins A, Stearman A, Chan B. Treatment of the serotonin syndrome with cyproheptadine. J Emerg Med. 1998;16: McDaniel WW. Serotonin syndrome: early management with cyproheptadine. Ann Pharmacother. 2001;35: Mackay FJ, Dunn NR, Mann RD. Antidepressants and the serotonin syndrome in general practice. Br J Gen Pract. 1999;49: Bodner RA, Lynch T, Lewis L, et al. Serotonin syndrome. Neurology. 1995;45: