The job of the pharmacist rarely involves evaluating diagnostic impressions or making definitive diagnoses. However, the pharmacist may often be called upon to give his or her assessment of a clinical picture when drug-related or drug-induced factors may be at play. In addition, pharmacists may be called upon to give therapy recommendations, frequently evaluating appropriate dosing, drug interactions and even therapeutic endpoints of medications. As patient and drug-regimen complexity increase, the ability of the pharmacist to give valuable input to the diagnostic impression or assessment of drug-related factors regarding a patient's clinical presentation becomes paramount. This may be particularly true in the context of mental health patients treated with more than one psychotropic agent.
A phenomenon that the pharmacist may encounter in this work is that the literature regarding drug initiation and combination is often more robust and informative than is the literature guiding drug discontinuation. For example, there are reasonably good data to support the addition of a second-generation antipsychotic to a conventional antidepressant agent when there is not an optimal response of depressive symptoms with antidepressant agents alone. However, the duration of therapy and/or the effects of discontinuation of the antipsychotic therapy in this scenario are not well studied or defined. Most antidepressant-antipsychotic combination therapy trials have only been eight weeks in duration and are not followed up with analysis of the effects of antipsychotic maintenance versus discontinuation.1 The well-known metabolic side effects of many of the atypical antipsychotic agents may serve as a stimulus for reducing or eliminating these medications once stability of symptoms is realized. The degree to which depression or anxiety relapse or other adverse outcomes may occur is unknown. Recent reports describing a case of serotonin syndrome (SS) due to duloxetine after discontinuation of olanzapine,2 and a case of neuroleptic malignant syndrome (NMS) occurring after the withdrawal of amisulpiride,3 indicate that there is a potential for unexpected effects to occur.
Here, the author reports the case of a severe neurotoxic syndrome in a patient in whom risperidone had recently been discontinued and in whom venlafaxine and sertraline combination therapy were continued. The case, diagnosed in the clinical setting as SS, is instructive regarding the nuances of SS and NMS diagnoses as well as in the consideration, evaluation and monitoring of psychotropic therapy discontinuations.
Ms. S is a 68-year-old female nursing home resident with medical conditions including hypertension, hypothyroidism, major depressive disorder, anxiety disorder, not otherwise specified, syringomyelia, and gastroesophageal reflux disease. At the time of presentation, the patient's scheduled nursing home medication list included: venlafaxine, sertraline, trazodone, lorazepam, hydrocodone/acetaminophen, baclofen, conjugated estrogens/medroxyprogesterone, omeprazole, lisinopril, furosemide, potassium chloride, levothyroxine, docusate sodium, and calcium carbonate/vitamin D. The nursing home medication administration record indicated that risperidone 0.5 mg twice daily had been in place for the previous sixteen months, but was discontinued four days prior to presentation for unknown reasons. Approximately thirteen months prior to presentation, sertraline was added to venlafaxine for better control of panic attacks, resulting in a stable regimen of venlafaxine extended release 300 mg daily, sertraline 200 mg daily, and risperidone 0.5 mg twice daily for more than one year. On the date of presentation, Ms. S was transferred from the nursing home to the emergency room due to confusion, fever, and “uncontrolled jerking”. Family members reported the patient had been feeling sick for the past couple of days. At presentation she was delirious, confused, hyperthermic (105.9ºF), tachycardic (168 bpm), but normotensive (117/62 mmHg) dehydrated with hyperkalemia (potassium 6.0 mg/dL) and hypernatremia (sodium 152 mg/dL) and in mild acute renal failure (SCr 1.2 mg/dL, BUN 30 mg/dL) which was presumed to be prerenal due to dehydration. Her white blood cell count (18,600 u/L), creatinine kinase (CK,1479 IU/L) and aspartate amino transferase (AST, 82 U/L) were elevated. All other laboratory studies were unremarkable. She was documented as having significant diffuse tremulousness, muscle spasms, dramatic nystagmus and diarrhea. The patient also had documented muscle rigidity of the lower extremities which was known to be chronic (thought to be related to syringomyelia), and was overlapped with frequent muscle spasms. Serotonin syndrome, neuroleptic malignant syndrome and septicemia were in the initial differential diagnosis. All psychotropic medications were discontinued and supportive care and fluid resuscitation were initiated. Regular insulin and 50% dextrose were given to correct hyperkalemia. Vancomycin and piperacillin/tazobactam were initiated, and she was admitted to the intensive care unit. Blood cultures, influenza antigen and urinalysis were unremarkable and infection was subsequently ruled out. Electrolyte abnormalities improved after dehydration was corrected and temperature slowly normalized, returning to baseline within twenty-four hours. Mental status began to improve significantly on day three. The patient was diagnosed with probable serotonin syndrome. She showed complete resolution of symptoms on day seven and she was discharged from the hospital with bupropion as the only psychotropic therapy prescribed.
The diagnosis of SS in the intensive care unit in this case was one of exclusion. Based on the significantly elevated temperature, neurologic symptoms and relative normalcy of the patient's vital signs and oxygenation as well as negative cultures, sepsis and other infectious causes were quickly ruled out. Due to alteration in the patient's mental status, a head CT scan was also done and was unremarkable. Acute baclofen withdrawal may also present with symptoms that are somewhat similar to what was seen in this case;5 however, this was ruled out in this patient due to documented consistent baclofen administration prior to admission. At this point, SS and NMS were the remaining considerations in the differential diagnosis. The patient's discontinuation of risperidone four days prior was thought to essentially eliminate NMS from further consideration. In addition, the patient's continuation of high dose combination serotonergic antidepressants coupled with dehydration was thought to contribute to elevation of antidepressant levels and was presumed to be the major impetus for the observed neurologic and physical symptoms. Thus, diagnosis for serotonin syndrome was made.
The differential diagnosis between SS and NMS can be difficult and the medical literature contains several commentaries discussing the differentiation of the two.5 Several different criteria for the diagnosis of SS have been applied in the medical literature over time, the first of which was the Sternbach criteria.6 However, the Hunter Serotonin Toxicity Criteria7 are supported as the most specific and sensitive of the available criteria for diagnosis of SS and are the most widely accepted in the field.5,8 These criteria are shown below in Figure 1.
When applying the Hunter Serotonin Toxicity Criteria (hereafter referred to as Hunter criteria)7 to the patient case presented here, one notes the absence of documented clonus. However, the medical record for this patient describes “frequent muscle spasms” and the family reported the patient was having “uncontrolled jerking.” It is difficult to determine whether these descriptions represent spontaneous clonus or not since that specific terminology was not used. Also, ocular clonus was not specified in this case, yet the documenting physician described “resting nystagmus that is pretty dramatic,” leading one to wonder whether this was purely nystagmus or whether ocular clonus terminology might have applied. In the absence of definitive description of clonus (spontaneous, inducible, ocular, or otherwise), the next step in applying the Hunter Criteria is to evaluate for the presence of tremor coupled with hyperreflexia. Tremor was indeed present (described as significant diffuse tremor), however, there is no mention in the chart documentation regarding the presence or absence of reflexes in this patient. Hence, the diagnosis of SS, according to the Hunter criteria, cannot definitively be applied. However, according to the Sternbach criteria (Figure 1),6 this patient meets the diagnostic threshold for SS (mental status changes, tremor, and fever were all present) and perhaps these were applied to the case as the diagnosis was made.
The likelihood that SS was present based on pharmacology is yet another topic to consider. Serotonin syndrome has been described as a concentration-dependent phenomenon,5,7 whereby the severe, life-threatening clinical effects of SS, such as rigidity and hyperthermia, are mediated by excessive stimulation at serotonin 2A (5-HT2A) receptors.4 In this patient, it stands to reason that SS could have resulted from the combined inhibition of serotonin reuptake induced by sertraline, venlafaxine and its active metabolite, desmethylvenlafaxine, leading to excessive 5-HT2A receptor stimulation. It should be noted the potential for trazodone to have contributed to SS in this case is highly unlikely. Though there are reports of SS occurring in association with combinations of trazodone and a serotonergic antidepressant, according to Gillman, trazodone is unlikely to contribute to SS because of its mechanism of action, primarily 5-HT2A antagonism and evidence that it does not appreciably elevate serotonin levels.9–11 Recalling the fact that this patient was taking the combination of sertraline and venlafaxine along with risperidone for more than one year without manifestations of severe serotonin toxicity being reported, might lead one to question whether discontinuation of risperidone four days prior could have brought on such symptoms. It is possible to surmise that risperidone's 5-HT2A antagonism may have been preventing onset of SS. The concomitant administration of risperidone with sertraline and venlafaxine may have limited the impact of high serotonin concentrations on the 5-HT2A receptor, possibly allowing no or only minor symptoms of SS (chronic mild rigidity) to be clinically evident. This is consistent with published pharmacology data regarding risperidone, indicating that it has a very strong binding affinity at the 5-HT2A receptor (Ki of 0.19nM) such that it displaced ketanserin (a compound with strong affinity for the 5-HT2A receptor) equally strongly in all three brain regions studied.12 Further, risperidone reportedly confers nearly 80% occupancy of 5-HT2A receptors in doses as low 1mg,13 and in animal models of SS, risperidone attenuated the increase of extracellular glutamate and nitric oxide concentrations that were thought to be implicated in the development of SS.14 Taking these data together, one could hypothesize that perhaps only when risperidone was discontinued and the 5-HT2A receptor was left unopposed, did severe SS develop. Recent cases implicating risperidone in the occurrence of SS,3,15 and the above described lack of consistency of symptoms with the Hunter criteria, lend some doubt to this line of thinking. Similarly, the continuation of the 5-HT2 antagonist, trazodone, is puzzling. The binding affinity of trazodone at the 5-HT2A receptor is unknown and therefore, the strength with which it would be a barrier to the binding of excessive concentrations of serotonin at that receptor cannot be determined.
It wasn't until much later that the diagnosis of NMS was more thoroughly considered in regard to this patient's case. The diagnostic criteria for NMS as described in the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM 5) are listed in Table 1.16 The very first descriptive criterion for NMS involves exposure to a dopamine antagonist within 72 hours prior to symptom development. It is challenging to interpret this criterion in the context of this patient since risperidone was discontinued 4 days prior to the patient's presentation, yet the family reported the patient had been “feeling sick for the past couple of days.” Perhaps there was overlap between the onset of her symptoms and risperidone ingestion in this patient, but that is virtually impossible to determine. Additional closer inspection and application of the remaining NMS diagnostic criteria is enlightening. The presence of elevated temperature (105.9ºF on first measurement, and 105.0ºF on second measurement) in the clinical presentation of this patient is certainly consistent with NMS. This patient also had rigidity in the lower extremities, but interestingly, this was described as chronic in nature, without mention of severity or “lead-pipe” quality, making the application of this diagnostic criterion for NMS somewhat challenging. Further, the patient had tremor, myoclonus, changes in consciousness, tachycardia, leukocytosis and elevated CK, all of which are consistent with the diagnosis of NMS. There was no mention of diaphoresis (described as a distinguishing feature of NMS)16 and the patient's respiratory status remained stable throughout the entire course. Serum iron concentrations were not available in the medical record.
Despite some of the similarities of this patient's case to typical NMS presentation, as mentioned previously, the fact that risperidone had been discontinued in this patient was very prominent in the discarding of the NMS diagnosis from the differential. Yet, a review of the NMS medical literature reveals several cases of NMS that were associated with discontinuation of an antipsychotic medication. Ball and colleagues report a fatal case of withdrawal NMS that developed seven days after discontinuation of amisulpiride, lithium and clomipramine.3 Margetic and Margetic describe a case of NMS that developed ten days after abrupt discontinuation of clozapine and haloperidol.17 In all cases of withdrawal NMS reported, the clinical presentation involved the common features of elevated temperature, rigidity, elevation of CPK, and leukocytosis and, therefore, was indistinguishable from that of the more traditional NMS that has been known to occur with antipsychotic initiation or dosage increase.3,18 These cases of withdrawal NMS developed within one to ten days after antipsychotic discontinuation and two were fatal.3,18 Three cases describe the presence of mild muscular rigidity prior to antipsychotic discontinuation, similar to that described in our patient, but all cases describe the index episode of NMS-like symptoms as distinct and separate from the clinical features of the patient's illness and/or presentation while they were taking the antipsychotic agent in question.18–20
Thought to be quite rare, the emergence of withdrawal NMS is possibly related to an “imbalance” in the dopaminergic system, as opposed to excessive dopaminergic blockade.18 Leading hypotheses regarding the pathophysiology of traditional NMS have centered on 1) dopaminergic hypofunction or 2) autonomic sympathetic nervous system hyperactivity.5 It seems that withdrawal NMS may be consistent with the latter of these. It has been proposed that withdrawal NMS may occur as a result of rebound cholinergic hyperactivity, through a complex cascade of antipsychotic-induced effects, revolving around induction of a supersensitized cholinergic network and resultant dopaminergic supersensitivity. Amore further theorizes that long term antipsychotic treatment is associated with an up-regulation of acetylcholine receptors; on antipsychotic withdrawal, dopamine blockade is reduced, along with a reduction in the anticholinergic “load” so that a cholinergic hyperactivity can occur, resulting in a state of reduced dopaminergic transmission that seems to be a crucial point in precipitating NMS.18 This theory is consistent with the observation that with the exception of the case presented here, in all NMS cases associated with antipsychotic withdrawal reported thus far, either the antipsychotic agent that was discontinued or another medication that was also simultaneously discontinued immediately prior to symptom onset, possessed significant anticholinergic properties and could have contributed to the onset of NMS because of rebound cholinergic hyperactivity. However, that is not the case in the patient presented here, in which only risperidone was discontinued. Risperidone has very low affinity for anticholinergic receptors (Ki value of >10,000) and very little appreciable anticholinergic action when used clinically.13 Therefore, the discontinuation of risperidone would be expected to carry very little risk of cholinergic rebound or dopaminergic hypofunction. As a result, it seems neither of the hypothesized NMS pathophysiologic mechanisms applies well to the case presented here. Authors acknowledge that while cholinergic overdrive after antipsychotic discontinuation may be a risk factor for development of withdrawal NMS, no proposed mechanism fully explains the complexity of the pathophysiology behind withdrawal NMS.3,18
While the clinical features of the case presented here are likely more in line with that of NMS as compared with SS, the pharmacology and proposed pathophysiology underlying each of these conditions may suggest SS as the more plausible explanation. Unfortunately, the distinction cannot definitively be made on the information available. Interestingly, a similar conundrum has been described in a patient taking risperidone and fluvoxamine combination therapy and in whom the symptoms resolved, not began, as these therapies were discontinued.21 Regardless of which diagnosis is the most correct to apply in this case, it is important for the mental health clinician to be cognizant of the potential for severe reactions such as SS and NMS to occur, not only when psychotropic therapy is initiated and titrated, but also upon its discontinuation.
The clinical features of SS and NMS can be difficult to distinguish in some patients who are taking combinations of psychotropic medications. It is plausible that either of these conditions could develop after one or more psychotropic medications are discontinued. Pharmacists and other clinicians should be aware of the potential for SS and NMS to occur in this scenario and keep it in mind as they consider and monitor therapeutic endpoints with antidepressant and antipsychotic medications.