Lithium Toxicity — Clinical Case

DCE Long Case

Patient brief (provided to trainee)

Patient: Mrs Margaret Chen, 72 years old, retired pharmacist. [1]

Presenting complaint: Progressive confusion, unsteadiness, tremor, and nausea over 5 days, brought to the emergency department by her daughter. [1]

History of the presenting complaint: Mrs Chen's daughter became concerned when her mother, who normally manages her own affairs independently, began calling her at unusual times, repeating herself in conversation, and forgetting recent events. Over the preceding 5 days the family noted a worsening coarse shaking of her hands that made it difficult for her to hold a cup, slurred speech, unsteadiness when walking (she nearly fell in the bathroom), and intermittent nausea with one episode of vomiting. She has been drinking less than usual because of the nausea. Her daughter brought her to the emergency department when she became unable to dress herself this morning due to the confusion and the tremor. [1]

Past history:

• Bipolar I disorder (diagnosed at age 47, 25 years ago) — managed with lithium carbonate 1000 mg at night for the last 18 years, with a stable level of 0.75 mmol per litre. Her last psychiatric admission was 8 years ago for a depressive episode. She has not had a manic episode for 12 years.
• Hypothyroidism (attributed to lithium) — on levothyroxine 75 micrograms daily for 6 years.
• Hypertension (diagnosed 3 weeks ago at a GP visit).
• Osteoarthritis of the knees.
• Acute myocardial infarction 3 weeks ago, treated with PCI and a drug-eluting stent to the left anterior descending artery. Echocardiography showed a left ventricular ejection fraction of 42 percent (heart failure with reduced ejection fraction). [1]

Current medications:

1. Lithium carbonate 1000 mg at night (bipolar disorder)
2. Levothyroxine 75 micrograms daily (hypothyroidism)
3. Aspirin 100 mg daily (post-MI, stent)
4. Ticagrelor 90 mg twice daily (post-MI, stent — dual antiplatelet therapy for 12 months)
5. Atorvastatin 80 mg daily (post-MI)
6. Bisoprolol 5 mg daily (post-MI, heart failure)
7. Indapamide 1.5 mg daily (started 2 weeks ago by GP for hypertension)
8. Ramipril 2.5 mg daily (started 2 weeks ago by cardiologist for heart failure post-MI)
9. Ibuprofen 400 mg three times daily as needed (started 2 weeks ago by GP for osteoarthritic knee pain — she has been taking it regularly)
10. Glyceryl trinitrate spray as needed (not used recently) [1]

Social history: She lives alone in her own single-storey home since her husband died 3 years ago. Her daughter visits daily. She is independent in personal care and shopping. She does not smoke and does not drink alcohol. She is a retired pharmacist and is meticulous about her medications. [1]

Examination findings (trainee elicits):

• Alert but drowsy, Glasgow Coma Scale 13 (E4 V4 M5). Disoriented to time (thinks it is March when it is July) but oriented to place and person.
• Coarse tremor of both hands at approximately 6 to 8 Hz, present at rest and on action, and a coarse tremor of the jaw.
• Speech is slurred (dysarthric) but fluent and not dysphasic.
• Ataxic on finger-to-nose and heel-to-shin testing bilaterally.
• Hyperreflexia in all four limbs. Occasional myoclonic jerks of the hands. No fasciculations visible in the tongue. Plantar responses downgoing.
• Cranial nerves intact. No nystagmus. No papilloedema on fundoscopy.
• Cardiovascular: pulse 88 in sinus rhythm, blood pressure 105 over 62, JVP not elevated, no peripheral oedema, dual heart sounds with no murmurs.
• Respiratory: clear, oxygen saturation 97 percent on room air.
• Abdomen: soft, non-tender.
• Temperature 37.2 degrees Celsius. [1]

Investigations:

• Serum lithium (drawn approximately 14 hours after the last dose): 3.0 mmol per litre (therapeutic range 0.6 to 1.0).
• Creatinine: 152 micromoles per litre (baseline 92 from 6 months ago). eGFR approximately 32.
• Sodium: 146 mmol per litre. Potassium: 4.8 mmol per litre. Chloride: 108. Bicarbonate: 22. Urea: 9.2.
• Glucose: 5.8 mmol per litre.
• Full blood count: haemoglobin 128, white cell count 9.2 (mild leucocytosis, consistent with lithium effect), platelets 240.
• TSH: 7.8 mU per litre (mildly elevated, consistent with her known lithium-related hypothyroidism). Free T4: 12 pmol per litre (low-normal).
• Corrected calcium: 2.62 mmol per litre (mildly elevated). Albumin: 42 g/L.
• Troponin: not elevated (no acute coronary event).
• ECG: sinus rhythm, rate 88, QTc 480 milliseconds, T-wave flattening in the lateral leads. No acute ischaemic changes.
• CT brain (performed to exclude an intracranial event given her confusion): no acute infarct, no haemorrhage, no mass. Mild age-related atrophy.
• Urine: specific gravity 1.008 (inappropriately dilute for her degree of dehydration), no blood, no protein, no leucocytes, no nitrites. [1]

Candidate's long-case presentation (SASPOP)

"Mrs Margaret Chen is a 72-year-old retired pharmacist presenting with a 5-day history of progressive confusion, a worsening coarse tremor, slurred speech, unsteadiness, and nausea. She has a 25-year history of bipolar I disorder, maintained on lithium carbonate 1000 mg at night for the last 18 years with a stable level of 0.75 millimoles per litre, and she has hypothyroidism, hypertension, osteoarthritis, and a recent myocardial infarction with a drug-eluting stent and mild heart failure with reduced ejection fraction. [1]

The critical history is that two weeks ago her GP and cardiologist started three new medications that all increase serum lithium levels: indapamide 1.5 mg daily (a thiazide-like diuretic) for her hypertension, ibuprofen 400 mg three times daily (an NSAID) for her osteoarthritic knee pain, and ramipril 2.5 mg daily (an ACE inhibitor) for her heart failure. Her lithium was not dose-adjusted and was not checked after these changes. The combination of these three interacting drugs — each of which increases proximal tubular lithium reabsorption or reduces glomerular filtration of lithium — precipitated acute-on-chronic lithium toxicity. [1]

On examination she is drowsy with a Glasgow Coma Scale of 13, disoriented, and has a coarse tremor of her hands and jaw, slurred speech, ataxia on coordination testing, hyperreflexia, and myoclonic jerks. Her serum lithium level is 3.0 millimoles per litre. Her creatinine has risen from a baseline of 92 to 152, consistent with an acute kidney injury on a background of probable chronic interstitial nephritis from her 18-year lithium exposure. Her ECG shows a QTc of 480 milliseconds with T-wave flattening. Her TSH is mildly elevated at 7.8, consistent with her known lithium-related hypothyroidism, and her corrected calcium is mildly elevated at 2.62, raising the possibility of lithium-induced hyperparathyroidism. [1]

Her main problems are:

1. Acute-on-chronic lithium toxicity at 3.0 millimoles per litre with significant neurological toxicity (GCS 13, ataxia, coarse tremor, myoclonus) and acute kidney injury — she meets the EXTRIP criteria for consideration of haemodialysis.
2. Acute kidney injury, contributing to the reduced lithium clearance, on a background of probable lithium-induced chronic interstitial nephritis.
3. QT prolongation on the ECG, a cardiac conduction effect of lithium toxicity.
4. The three precipitating drug interactions — indapamide, ibuprofen, and ramipril — all of which must be stopped.
5. Chronic lithium end-organ effects — hypothyroidism, mild hypercalcaemia, probable chronic interstitial nephritis, and a concentrating defect (inappropriately dilute urine).
6. The underlying bipolar I disorder — the indication for lithium must be revisited with her psychiatrist.
7. Her recent myocardial infarction with a drug-eluting stent — she must continue her dual antiplatelet therapy (aspirin and ticagrelor), which is a consideration if she requires haemodialysis or any procedure. [1]

My immediate management plan is to stop the lithium and all three interacting drugs (indapamide, ibuprofen, ramipril), assess and resuscitate with ABCDE, and give aggressive intravenous normal saline to restore her volume and enhance renal lithium excretion. I will not give loop diuretics, which paradoxically raise the lithium level. Given her level of 3.0 with decreased consciousness and acute kidney injury, I am discussing urgent intermittent haemodialysis with the nephrology and toxicology teams — she meets the EXTRIP criteria for extracorporeal treatment on the basis of her decreased level of consciousness (a recommended indication irrespective of the level) and her impaired renal function. I will monitor her lithium level every 2 hours, and after any dialysis I will check a post-dialysis rebound level at 4 to 6 hours. I will monitor her renal function, her electrolytes, her ECG, and her neurological state continuously. [1]

Once she has recovered, I will work with her psychiatrist and her GP to decide whether to restart lithium at a reduced dose with a lower target and very close monitoring, or to switch to an alternative mood stabiliser such as valproate, lamotrigine, or an atypical antipsychotic, balancing her excellent long-term response to lithium against her renal trajectory and her cardiovascular comorbidity. I will ensure her antihypertensive and analgesic regimens are changed to agents that do not interact with lithium (amlodipine for blood pressure, paracetamol for pain). I will check a parathyroid hormone level and a 24-hour urinary calcium to investigate the mild hypercalcaemia. And I will provide written education to Mrs Chen, her daughter, her GP, and her cardiologist about the lithium-interaction risk of thiazides, NSAIDs, and ACE inhibitors, to prevent recurrence." [1]

Examiner discussion questions and model answers

Q1: "You plan to dialyse this patient. She is on dual antiplatelet therapy for a recent stent. Does that change your plan?" [1]

"The dual antiplatelet therapy is an important consideration but it does not preclude haemodialysis. Intermittent haemodialysis can be performed through a temporary (non-tunnelled) central venous catheter, which avoids the need for a tunnelled catheter or an arteriovenous fistula creation that would carry a higher bleeding risk on dual antiplatelets. The access is a temporary internal jugular or femoral catheter placed under strict aseptic technique, with appropriate local measures to manage the bleeding risk (compression, careful technique). The bigger issue is that she absolutely must not stop her dual antiplatelet therapy — she had a drug-eluting stent placed 3 weeks ago, and stopping aspirin or ticagrelor within the first 12 months carries a very high risk of acute stent thrombosis and a fatal myocardial infarction. So I would proceed with the dialysis via a temporary catheter, continue the dual antiplatelet therapy, and manage the access-site bleeding risk with local measures. I would involve the cardiology team to confirm the stent timeline and the antiplatelet plan, and the haematology team if there is concern about the bleeding risk. The principle is that the lithium toxicity is immediately life-threatening (it is causing a decreased level of consciousness and seizures are a risk), and it must be treated; the dual antiplatelet therapy is managed around the dialysis, not used as a reason to withhold it." [1]

Q2: "How would you explain to Mrs Chen's daughter what happened, in a way she will understand?" [1]

"I would say: your mother has been on a medication called lithium for her bipolar disorder for many years, and it has been keeping her well. Lithium is an excellent medication, but it has a very narrow safe range in the body, and the body gets rid of it through the kidneys. In the last two weeks, your mother was started on three new medications — one for her blood pressure, one for her heart, and one for her knee pain. These are all safe medications for most people, but each of them interferes with the way the body gets rid of lithium, and together they caused the lithium to build up in her body to a toxic level. That build-up is what caused her confusion, the shaking, the slurred speech, and the unsteadiness. We have stopped the lithium and the three interacting medications, we are giving her intravenous fluids to help her kidneys flush out the lithium, and we are considering a dialysis treatment to remove the lithium quickly from her blood. I want to reassure you that we expect her to recover, though it may take some days, and once she is better we will work with her psychiatrist to decide the best plan for her bipolar medication going forward. I would also say that this was preventable — the three medications should have been started with a reduction of the lithium dose and a blood test to check the level — and we will make sure that the communication between her doctors is better going forward so that this does not happen again." [1]

Q3: "She has been stable on lithium for 18 years with no psychiatric admissions for 8 years. How does that factor into your longer-term plan?" [1]

"Her excellent and durable response to lithium is a very strong point in favour of restarting it, if we can do so safely. Lithium is the gold-standard mood stabiliser for bipolar I disorder — it has unmatched efficacy for mania prophylaxis, bipolar depression prophylaxis, and crucially for suicide prevention, which is the leading cause of premature death in bipolar disorder. A patient who has been stable for 8 years on lithium has clearly responded well, and switching to an alternative agent carries a real risk of relapse. However, the contraindications to simple restart are her progressive renal impairment (the baseline eGFR of 52 suggests early chronic interstitial nephritis, and now she has an acute kidney injury on top), her age (72), and the cardiovascular comorbidity that will require medications that interact with lithium. So my plan, made jointly with her psychiatrist, would be along these lines: once she has recovered fully from the acute toxicity and her renal function has stabilised, I would favour restarting lithium at a reduced dose — perhaps 500 mg at night — targeting a lower trough level of 0.5 to 0.6 millimoles per litre, which is adequate for maintenance prophylaxis (the higher levels are needed for acute mania, which is not her current situation). I would ensure her antihypertensive is changed to amlodipine (no lithium interaction), her analgesic is paracetamol (no lithium interaction), and that any future medication change is accompanied by a lithium dose review and a level check. I would arrange close monitoring — a lithium level every 2 weeks initially, then every 3 months, with renal and thyroid function at each visit. If her renal function continues to decline despite the dose reduction, I would revisit the decision with her psychiatrist and consider a switch to valproate or an atypical antipsychotic. The decision is shared, it is ongoing, and it is anchored to her goals — stability of her mood, preservation of her renal function, and avoidance of further toxicity." [1]

Q4: "What is the single most important system-level change to prevent this happening to another patient?" [1]

"The single most important change is a structured lithium prescribing and monitoring system that flags the drug-interaction risk at the point of prescribing. This patient became toxic because three medications, each known to increase lithium levels, were started by two different doctors (the GP and the cardiologist) without a lithium dose adjustment or a level check. The systemic fix is a decision-support tool in the electronic prescribing system that alerts the prescriber whenever a thiazide, an NSAID, an ACE inhibitor, or an ARB is prescribed to a patient on lithium, prompting a dose review and a level check. Alongside the technical fix, the patient education is critical — every patient on lithium should carry a card or a written list of the medications to avoid, and should be instructed to tell any doctor or pharmacist that they are on lithium whenever a new medication is prescribed. And the communication between the psychiatrist, the GP, the physician, and the cardiologist must be explicit about the lithium and the monitoring plan at every transition of care. The episode was preventable at multiple points — the GP starting the indapamide and the ibuprofen, the cardiologist starting the ramipril, the absence of a lithium level check after any of these — and a layered system of decision support, patient education, and structured communication would have caught it." [1]

Q5: "Her QTc is 480. Is this from the lithium, and what will you do about it?" [1]

"The QT prolongation is very likely from the lithium toxicity. Lithium can cause cardiac conduction effects — QT prolongation, T-wave inversion or flattening, sinus node dysfunction, and occasionally bradycardia. These are features of tissue-saturated toxicity, and they carry a risk of torsades de pointes, particularly in a patient with other QT-prolonging factors (she is on a stable dose of atorvastatin and bisoprolol, neither of which significantly prolongs the QT, but her electrolytes should be checked — hypokalaemia and hypomagnesaemia would compound the risk). The management is to treat the underlying lithium toxicity — stop the lithium, correct any electrolyte abnormalities, and remove the lithium with intravenous fluids and, if indicated, haemodialysis. As the lithium level falls, the QT should normalise. I would monitor her on continuous cardiac telemetry, check and correct her potassium and magnesium, and avoid any additional QT-prolonging drugs (macrolides, fluoroquinolones, ondansetron, antipsychotics) during the admission. If she develops torsades de pointes, I would treat it with intravenous magnesium sulfate and, if needed, overdrive pacing or isoprenaline, but the primary strategy is the rapid removal of the lithium." [1]

DCE Short Case — Focused Neurological Examination of the Lithium-Toxic Patient

Examination instruction

Examiner: "This 68-year-old man, who is on lithium for bipolar disorder, has been brought in with confusion and unsteadiness. Please perform a focused neurological examination and present your findings." [1]

Key findings this patient demonstrates

• Mental state: drowsy, disoriented to time, but cooperative with examination.
• Speech: slurred (dysarthric), fluent, not dysphasic.
• Tremor: coarse, 6 to 8 Hz, affecting both hands and the jaw, present at rest and on action.
• Cranial nerves: intact, no nystagmus, no papilloedema.
• Motor system: normal tone, normal power, brisk reflexes symmetrically, downgoing plantar responses.
• Involuntary movements: occasional myoclonic jerks of the hands.
• Coordination: dysmetria on finger-to-nose testing bilaterally, impaired heel-to-shin testing.
• Gait: ataxic, wide-based, unable to tandem walk.
• Sensation: intact. [1]

Presentation template

"I examined the neurological system of this 68-year-old man on lithium for bipolar disorder who presents with confusion and unsteadiness. He is drowsy and disoriented to time but cooperative. His speech is slurred but not dysphasic. He has a coarse tremor at 6 to 8 hertz affecting both hands and his jaw, present at rest and on action. His cranial nerves are intact with no nystagmus. In the limbs, the tone and power are normal, but the reflexes are brisk symmetrically, and there are occasional myoclonic jerks of the hands. His coordination is impaired, with dysmetria on finger-to-nose testing bilaterally. His gait is ataxic and wide-based. His sensation is intact. My findings are of a diffuse encephalopathy with cerebellar involvement, neuromuscular irritability, and a coarse tremor — a picture consistent with lithium neurotoxicity. I would now check a serum lithium level, a renal function, and an ECG, and proceed with the toxicological management." [1]

Discussion

Q: "What is the differential diagnosis of this neurological picture?" [1]

"The differential of a coarse tremor with ataxia, confusion, hyperreflexia, and myoclonus in a patient on lithium includes: lithium toxicity itself (the leading diagnosis given the medication history and the classic picture); serotonin syndrome (if the patient is also on a serotonergic agent such as an SSRI, SNRI, or tramadol — the distinguishing features are clonus, which is highly specific for serotonin syndrome, and a more rapid onset); neuroleptic malignant syndrome (if the patient is on an antipsychotic — the distinguishing features are lead-pipe rigidity, profound hyperthermia, and a markedly raised creatine kinase); Wernicke encephalopathy (particularly in a patient with alcohol use or malnutrition — give thiamine empirically); hepatic encephalopathy (check the liver function); a meningoencephalitis (fever, meningism, and a lumbar puncture); and a cerebellar stroke (sudden onset, focal signs, confirmed on imaging — but the CT brain here showed no acute event). The serum lithium level and the clinical context usually resolve the differential, and this patient's level of 3.0 millimoles per litre with the classic picture confirms lithium toxicity. The key point is to treat the lithium toxicity while keeping the other possibilities in mind and investigating as appropriate." [1]

Q: "Why is the tremor of lithium toxicity described as 'coarse,' and why does that matter?" [1]

"The distinction between a fine and a coarse tremor is clinically important because it distinguishes the benign, therapeutic-level tremor from the toxic tremor. Many patients on lithium develop a fine, action tremor at therapeutic levels — this is common, it is benign, and it is managed with reassurance, a reduction in caffeine, and occasionally a beta-blocker. It is an exaggeration of the physiological tremor. The coarse tremor of toxicity is different: it is lower-frequency (6 to 8 Hz versus 8 to 12 Hz for the fine tremor), it is higher-amplitude, and critically, it is present at rest as well as on action. The appearance of a coarse tremor, or the worsening of a previously fine tremor into a coarse one, is a red flag for toxicity and demands an immediate lithium level. The tremor is one of the earliest signs of toxicity — it often precedes the confusion and the ataxia — and recognising it can catch toxicity before it becomes severe. This is why every patient on lithium should be asked about a new or worsening tremor at every review." [1]