Hoigne Syndrome Secondary to Intravenous Lidocaine in a Woman with Metastatic Ewing Sarcoma

Introduction

Parenteral lidocaine infusions are increasingly used in the management of refractory neuropathic cancer pain. ¹ Hoigne syndrome (HS) is a rarely described psychiatric reaction that has been reported after administration of numerous medications, but most commonly subcutaneous or intravenous local anesthetics. HS does not have well-delineated diagnostic criteria but is commonly described as an acute nonallergic reaction to a medication (typically a local anesthetic), characterized by visual hallucinations, anxiety surrounding one's own death, and mood changes, usually without a change in sensorium. These changes are typically self-limited, resolve in hours to days, and do not have a lasting effect.^2–7 Although HS has been described from parenteral lidocaine in treating chronic noncancer pain ⁷ and postcardiac arrest, ⁴ it has not yet been described in the setting of a patient with advanced cancer. Ethics approval for this case report was not deemed necessary.

Case Report

A 29-year-old South Asian woman with Ewing sarcoma of the left scapula developed cancer recurrence one year after initial resection, with diffuse lung and bone metastases. After initiation of palliative chemotherapy with nab-paclitaxel, her course was complicated by rapid disease progression with increasingly refractory pain. She transferred care from another institution for a possible salvage therapy and was admitted to our Solid Oncology ward for an acute pain crisis. She described her pain as “all over her body” although largely concentrated in her ribs, low back, and pelvis. Her pain was characterized by a deep throbbing ache that would shoot down her legs bilaterally, thought to have a predominant neuropathic element.

During a prior hospitalization, the patient had been started on a bolus-only hydromorphone patient-controlled analgesia (PCA), which had been continued as an outpatient through a Continuous Ambulatory Delivery Device. At the time of this hospital admission, she was also on a regimen of methadone 20/20/30 mg PO TID, acetaminophen 1000 mg PO q8 hours, duloxetine 60 mg PO qAM, and methocarbamol 750 mg PO QID. Despite uptitration of her hydromorphone PCA to a 2.5 mg IV bolus q10 minutes prn (no basal rate) and eventual transition to a fentanyl PCA up to a 150 mcg IV bolus q10 minutes prn with a 25 mcg/hour basal rate, the patient's pain remained poorly controlled.

An interdisciplinary approach involving the palliative care and interventional pain teams was used to guide and discuss further options. A ketamine infusion was initiated and eventually titrated up to 0.3 mg/kg/hour but did not result in adequate pain relief. An intrathecal drug delivery system was implanted on hospital day (HD) six with constant infusion of intrathecal hydromorphone (9.8 mg qday) and bupivacaine (6.5 mg qday) without ziconotide. Despite this regimen, new and enlarging spinal metastases contributed to severe radicular pain in her upper and lower extremities. The patient inquired about an intravenous lidocaine infusion, which, during prior hospitalizations, had resulted in significant relief that persisted even after the infusion was stopped.

An intravenous lidocaine infusion was initiated on HD 16 with a loading dose of 2 mg/kg followed by a continuous infusion of 1 mg/kg/hour, based on ideal body weight. The patient reported significant relief after the bolus dose and was able to achieve continuous relief with ongoing titration of her lidocaine infusion. Ketamine infusion was discontinued one hour after initiation of parenteral lidocaine. The lidocaine infusion rate was titrated to a maximum of 1.6 mg/kg/hour on hospital day 17, at which time the patient reported perioral numbness but denied tinnitus, tremor, or myoclonus. At that point plasma levels of lidocaine were elevated to 7.2 mcg/mL. As the lidocaine infusion was decreased to 1 mg/kg/hour, plasma levels normalized to consistently <5 mcg/mL with continued relief of pain.

On HD 19, the patient developed worsening bilateral lower extremity weakness and numbness suggestive of progressive spinal cord compression. Dexamethasone 10 mg IV qAM was started on HD 20 to provide relief of compression and pain as the patient was not a surgical candidate. On HD 25, dexamethasone was increased to 10 mg IV q8 hours in the hopes of reducing bony pain and compression. On HD 27, the patient began to have increasingly frequent hallucinations related to her death, beginning with seeing the spirits of dead family members. She was able to identify these as hallucinations and was not distressed by them; rather, she found these hallucinations helpful in processing her mortality and providing closure to her family. She maintained a clear sensorium throughout this period, and was aware that her hallucinations were not real, with good attention during conversation, making steroid induced delirium less likely. Owing to the nature of her hallucinations (related to her death), while maintaining a clear sensorium, these changes were thought to be most consistent with HS secondary to her lidocaine infusion. Given that these hallucinations were helpful in her processing her mortality and that her primary goal was pain relief, a shared decision was made to continue parenteral lidocaine.

In the following days, however, these hallucinations became increasingly persecutory and centered around death. The patient described “talking to God” who was “angry with [her] for using the PCA.” She shared being “trapped in Purgatory and waiting to be taken to Hell.” At this time, she began wearing a traditional religious head covering and praying fervently while confessing her sins. According to her family, these behaviors were extremely uncharacteristic, as she had never professed a strong religious identity in the past.

Given the improvement in her neuropathic pain, her lidocaine infusion was continued while attempts were made to control the symptoms of HS. The distress caused by her hallucinations were unresponsive to nightly olanzapine but did improve with doses of oral and intravenous lorazepam. As dexamethasone was reduced to once daily dosing and lidocaine was weaned, these death-focused hallucinations stopped. The patient reported greater clarity, although she was somnolent due to increased opioids. The patient was transitioned to hospice on a reduced dose continuous lidocaine infusion and died at home.

Discussion

HS was first noted in 1951⁸ and was later given its eponym in 1959 by Drs. Hoigne and Schoch ⁹ who noted that a minority of patients who received intramuscular procaine penicillin developed rapid onset anxiety about death, dissipating within hours without lasting consequences.

Typically, patients with HS have acute psychiatric changes that develop within minutes of subcutaneous or intramuscular medication administration and can occur days after an offending infusion is initiated. ⁴ Reactions from subcutaneous or intramuscular injections are thought to be related to inadvertent intravenous injections. ⁵ The most common symptom reported is anxiety about impending death, commonly called “doom anxiety” or “death anxiety.^4,5” Patients also can develop agitation, mood changes, and auditory or visual hallucinations. Although some case reports describe patients becoming acutely confused, typically patients maintain normal orientation and can discern that their hallucinations are not real, differentiating this from delirium. These psychiatric effects resolve within hours of removal of the offending medication, with anxiety and sleep disturbances at times lasting days to weeks.^2,3

The pathophysiology initially was thought to be mediated by procaine penicillin crystal microembolization going from the venous system to the brain. ⁹ This theory has fallen out of favor with the absence of pulmonary symptoms which one would expect as microemboli move through the pulmonary vasculature. Furthermore, microemboli would likely cause more heterogeneous deficits from scattered emboli, whereas the psychiatric changes in HS are relatively consistent. ¹⁰

More recently, it has been hypothesized that the symptoms of HS may be related to limbic kindling. ¹⁰ Kindling is the heightened physiological response to a stimulus after multiple initial exposures do not have a response. Pathophysiological kindling is supported by a case series showing that HS typically only occurs after multiple prior doses of procaine penicillin, and that the anxiety and perceptual changes seen in HS resemble temporal lobe epilepsy. ¹⁰ Our patient had previously received lidocaine infusions for intractable neuropathic pain at another institution, yet per family's report, she did not have symptoms characteristic of HS during those infusions. Prior case series have demonstrated a theoretical link between anxiety disorders and HS, with high rates of panic disorder found in a cohort of patients after an acute episode of HS. With many of these patients there was concern for an anxiety disorder preceding their acute episode of HS, raising the possibility that anxiety disorders may predispose patients to HS, although this remains unclear. In our patient she did not have any previously diagnosed anxiety disorders. ³

Although HS from lidocaine or procaine does not appear to represent an overdose of local anesthetics, ² it is unclear if HS is dose responsive, as opposed to the more common and predictable side effects of tinnitus and oral numbness that roughly correlate to plasma lidocaine levels. In our case, the patient developed HS while having plasma levels <5 mcg/mL, although symptoms did improve with dose reduction of her infusion.

The most common reported cases of HS have been from local injections of procaine penicillin, although studies have varied as to whether procaine or penicillin is the identified as culprit.^5,9 Since the initial reports were first documented by Hoigne and Batchelor, case series have documented intravenous lidocaine causing death-related anxiety and hallucinations in patients postcardiac arrest. ⁴ Case reports have also documented HS from antimicrobials, including clarithromycin, ceftriaxone, and meglumine.^6,11,12 One case report also documented HS secondary to a nail bed prednisolone injection. ¹³

The primary treatment for HS is discontinuing the offending agent, as symptoms typically resolve within minutes to hours. Multiple case reports have also documented improvement in agitation from short-term intravenous benzodiazepine use; both lorazepam (1 mg IV) and midazolam (4 mg IV) have been reported as being effective in acutely managing the anxiety and agitation from HS.^2,7

Since our patient was near the end of life, our case posed some unique challenges when managing HS. First, the patient initially found therapeutic benefit from the HS-associated hallucinations in processing her mortality. Second, the patient had refractory cancer pain that was only well controlled with parenteral lidocaine. Through shared decision making with the patient and her family, the decision was made to continue the infusion, acknowledging that the hallucinations could become increasingly distressing over time. When the hallucinations became more persecutory, completely stopping the lidocaine was not thought to be realistic given the benefit in controlling her previously refractory pain. Again, with shared decision making, we decreased the dose of her infusion and added adjunctive benzodiazepines to minimize the distressing nature of her hallucinations while maintaining good analgesia. ¹

The concurrent use of dexamethasone raises the possibility that some of her symptoms could have been attributed to steroid-induced delirium. Moreover, as her symptoms improved with reduction of both dexamethasone and lidocaine simultaneously, it is difficult to differentiate the sole source of her symptoms. However, because she was aware that her hallucinations were not real, maintained a clear sensorium, with persecutory hallucinations centered around her death, we felt that her clinical syndrome was more closely aligned with HS.

Conclusion

With the increasing use of parenteral lidocaine for refractory cancer pain, palliative care providers should be aware of HS as a possible complication that can occur with repeated or prolonged use. Although typically represented as a distressing psychiatric syndrome, this case demonstrates that when patients are near the end of life, hallucinations centered around one's own death are not necessarily harmful and may be helpful in processing their mortality. In addition, when patients are near the end of life and lidocaine has been useful for refractory pain, stopping it may not be realistic. In our case adding adjunctive benzodiazepines and decreasing the infusion were both helpful in reducing the distressing nature of the hallucinations and anxiety secondary to HS while maintaining adequate analgesia.