Cytotoxic Lesion of the Corpus Callosum in a Patient with Acute-on-chronic Lithium Toxicity

Abstract

A 30-year-old woman with history of bipolar disorder presented to the emergency department with impaired consciousness and myoclonus following regular lithium treatment. Brain magnetic resonance imaging revealed a hyperintense lesion in the splenium of the corpus callosum, consistent with cytotoxic lesions of the corpus callosum (CLOCCs). Serum lithium levels were markedly elevated (3.54 mmol/L), and acute-on-chronic lithium toxicity was diagnosed. Hemodialysis and supportive care resulted in clinical and radiological recovery. This case highlights that CLOCCs may serve as important neuroimaging evidence of lithium toxicity and may support an early diagnosis and treatment, especially in patients with long-term lithium use.

Introduction

Lithium is widely used in the treatment of bipolar disorder with a narrow therapeutic window (1). Lithium toxicity can cause central nervous system symptoms (e.g., tremors, ataxia, impaired consciousness), particularly in the presence of dehydration or renal dysfunction. However, abnormalities on brain magnetic resonance imaging (MRI) have rarely been reported in cases of lithium toxicity (2-6).

In clinical practice, the diagnosis of lithium toxicity is often complicated by the poor correlation between serum blood levels of lithium with symptom severity and the delay in obtaining lithium levels. Therefore, neuroimaging may play a valuable role in its early diagnosis, particularly in patients with atypical clinical presentations.

Cytotoxic lesions of the corpus callosum (CLOCCs) are secondary intracranial lesions characterized by transient cytotoxic edema in the splenium of the corpus callosum. These lesions have diverse etiologies and exhibit distinct imaging features. CLOCCs are believed to arise from cytotoxic edema caused by various systemic or neurological conditions, including infections, metabolic disturbances, epilepsy, and certain drug toxicities (7,8). Although generally reversible, CLOCCs require prompt identification and management of their underlying causes.

We report a rare case of CLOCCs due to acute-on-chronic lithium toxicity in a patient presenting with impaired consciousness, in which early MRI findings facilitated a prompt diagnosis and management.

Case Report

A 30-year-old woman with bipolar disorder was being treated by her primary care physician with daily doses of mirtazapine (45 mg), duloxetine (60 mg), quetiapine (300 mg), lithium carbonate (400 mg), triazolam (0.25 mg), brotizolam (0.5 mg), and lorazepam (3 mg). Seven days before admission, the patient overdosed on lithium carbonate (amount unknown) due to insomnia, resulting in diarrhea, vomiting, and poor oral intake. Five days before admission, the patient discontinued all medications. The next day, she developed tremors and dysphagia. One day before admission, she visited a psychiatrist at another hospital and was admitted with suspected dehydration and benzodiazepine withdrawal; however, owing to impaired consciousness and fever, she was transferred to our hospital for further physical evaluation and management.

On admission, the primary survey revealed: patent airway, respiratory rate was 23 breaths/min with oxygen saturation of 97% on room air, heart rate 114 beats/min, blood pressure 127/78 mmHg, impaired consciousness (Glasgow Coma Scale score of 6, E4V1M1), body temperature 37.4°C, height 152 cm, and weight 53.3 kg. Initial Sequential Organ Failure Assessment yielded a score of 5. A neurological examination revealed nystagmus and generalized myoclonus.

The laboratory tests (Table 1) demonstrated leukocytosis, elevated inflammatory markers, mild hepatic dysfunction, severe renal impairment [estimated glomerular filtration rate (eGFR), 14.7 mL/min/1.73 m²] indicating acute kidney injury, and mild hyponatremia. Cerebrospinal fluid analysis revealed no pleocytosis or hypoglycorrhachia and culture results were unremarkable. Electroencephalography showed generalized slow-wave activity, however, no findings suggestive of status epilepticus were observed.

Table 1.

Laboratory Data.

On admission
Hematology		Biochemistry
WBC (4,000-9,000 /μL)	28,000	TP (6.7-8.3 g/dL)	7.2	Ca (8.6-10.1 mg/dL)	9.7
Differential count		ALB (3.8-5.3 g/dL)	4.1	IP (2.7-4.5 mg/dL)	2.8
Neutrophils (45-55%)	88.0	AST (8-38 U/L)	46	Mg (1.8-2.3 mg/dL)	2.9
Eosinophils (1-5%)	0.0	ALT (4-44 U/L)	66	CRP (<0.1 mg/dL)	1.76
Basophils (0-1%)	0.5	LDH (124-222 U/L)	372	Procalcitonin (0.0-0.49 ng/mL)	0.29
Monocytes (4-7%)	5.0	ALP (38-113 U/L)	220	NH3 (12-66 μg/dL)	57
Lymphocytes (25-45%)	6.5	γGTP (8-38 U/L)	286	Cerebrospinal fluid
RBC (376-500×104/μL)	454	T-bil (0.3-1.2 mg/dL)	0.6	TP (10-40 mg/dL)	46
Hb (12-16 g/dL)	13.4	CK (42-150 U/L)	56	ALB (mg/dL)	29
Ht (33.5-45.0%)	40.1	UN (8-20 mg/dL)	65.2	IgG (mg/dL)	3.1
MCV (80-100 fL)	88.3	Cre (0.47-0.79 mg/dL)	3.29	IgG index (0-0.73)	0.43
MCH (28-32 pg)	29.5	UA (4-7 mg/dL)	24.5	Glucose (50-75 mg/dL)	105
MCHC (31-35 g/dL)	33.4	Na (135-147 mEq/L)	132	Cell count (/mm3)
Plt (15-35×104/μL)	30.1	K (3.6-5.0 mEq/L)	3.7	Mononuclear cell	1
		Cl (98-108 mEq/L)	92	Polymorphonuclear cell	0

The range of normal values are given in parenthesis.

WBC: white blood cell, RBC: red blood cell, Hb: hemoglobin, Ht: hematocrit, MCV: mean corpusclular volume, MCH: mean corpusclular hemoglobin, MCHC: mean corpuscular hemoglobin concentration, Plt: platelet, TP: total protein, ALB: albumin, AST: aspartate aminotransferase, ALT: alanine aminotransferase, LDH: lactate dehydrogenase, ALP: alkaline phosphatase, γGTP: gamma-glutamyltransferase, T-bil: total bilirubin, CK: creatine kinase, BUN: blood urea nitrogen, Cre: creatinine, UA: uric acid CRP: C-reactive protein

Brain MRI was promptly performed on admission to identify potential structural or toxic-metabolic causes. A hyperintense ovoid lesion was revealed in the splenium of the corpus callosum on fluid-attenuated inversion recovery and diffusion-weighted imaging, with corresponding hypointensity on the apparent diffusion coefficient map. These findings were consistent with CLOCCs (Fig. 1a-c).

Figure 1.

Brain magnetic resonance imaging (MRI) findings. Brain MRI findings on admission (a-c) and 5 months later (d-f). On admission, fluid-attenuated inversion recovery (a) and diffusion-weighted imaging (b) show a hyperintense lesion in the splenium of the corpus callosum (arrowheads), with corresponding hypointensity on the apparent diffusion coefficient map (c). The lesion completely resolved on follow-up imaging (d-f). Note: Slice angles differ between the two scans because they were obtained at different facilities.

Based on the clinical presentation, severe lithium toxicity with central nervous system involvement was suspected, and hemodialysis was initiated on the day of admission and continued for 7 days. The following day, serum lithium levels were confirmed to be markedly elevated (3.54 mmol/L), leading to a diagnosis of acute-on-chronic lithium toxicity. The patient's serum creatinine and lithium levels declined steadily, and eGFR showed a corresponding improvement to normalization with supportive care and hemodialysis (Fig. 2). By hospital day 8, the myoclonus had resolved and the patient regained the ability to communicate. On hospital day 11, she was transferred to the care of her previous physician. She began speaking the day after being transferred and oral intake was resumed on the fifth day. Three months later, the patient was discharged and admitted to a group home, although mild dysarthria persisted. Five months after discharge from our hospital, follow-up MRI scans at another facility showed complete resolution of the lesion in the corpus callosum, with no new abnormalities (Fig. 1d-f).

Figure 2.

Clinical course showing serum lithium concentration, estimated glomerular filtration rate, and level of consciousness. The serum lithium levels and estimated glomerular filtration rate (eGFR) are plotted over the course of hospitalization. The consciousness level is indicated by the Glasgow Coma Scale (GCS). Following the early initiation of hemodialysis, the patient’s lithium levels decreased, renal function improved, and neurological status gradually improved.

Discussion

The diagnosis of lithium toxicity can be challenging because blood lithium levels do not always correlate with the severity of neurological symptoms and results may be delayed for several days. In the present case, brain MRI revealed findings consistent with CLOCCs in a patient with impaired consciousness and myoclonus who had been regularly taking lithium carbonate. This case highlights the importance of considering lithium toxicity as a possible cause of CLOCCs.

The concept of reversible splenial lesion syndrome was proposed in 2011 to describe the transient splenial lesions associated with antiepileptic drug withdrawal, infection, high-altitude sickness, and metabolic or electrolyte abnormalities (9). Since then, CLOCCs has been introduced as a broader term encompassing the reversible splenial lesion syndrome and other transient lesions of the corpus callosum, unified by a common pathophysiological mechanism. These lesions are typically caused by infections, metabolic disturbances, and drug-related factors, and are characterized by transient cytotoxic edema in the splenium that generally resolves by treating the underlying condition (7,8). Therefore, prompt identification and management of the causes of CLOCCs are cardinal.

The splenium consists of densely packed glial cells and a combination of thin, late-myelinating fibers anteriorly and thick, early myelinating fibers posteriorly, supporting high-bandwidth interhemispheric transfer of visual, sensory, and cognitive information. These densely arranged axonal fibers and glial cells mean that any change in osmotic gradients or cellular water content is more likely to produce localized swelling and cytotoxic injury in the splenium than in other parts of the corpus callosum. Therefore, making the splenium especially susceptible to cytotoxic edema (7,10). On the other hand, lithium competitively disrupts the sodium-lithium counter-transport and interferes with normal cellular sodium gradients, which can lead to changes in neuronal excitability and water transport, predisposing to cytotoxic edema (11). Lithium also affects inositol monophosphatase signaling, glutamate transport, and mitochondrial function (11), which may collectively impair astrocytic osmoregulation and contribute to reversible intramyelinic edema. These anatomical features and pharmacological mechanisms may be associated with lithium-induced CLOCCs.

CLOCCs are associated with various etiologies including infection, metabolic disturbances, epilepsy, and drug toxicity. In this case, the absence of pleocytosis or hypoglycorrhachia in the cerebrospinal fluid excluded infectious causes, while metabolic abnormalities were excluded by normal serum glucose and electrolytes. Withdrawal-related neurological instability was deemed unlikely because electroencephalography demonstrated diffuse slowing without seizure activity as well as no autonomic instability. No evidence suggested drug interactions, as the patient had discontinued all psychotropic medications five days prior, and none of the remaining agents are known to cause isolated splenial lesions. Therefore, lithium toxicity was considered the most likely cause.

Our review of the relevant literature identified six previously reported cases of CLOCCs associated with lithium toxicity (2-6) (Table 2). Six patients, including ours, experienced either chronic or acute-on-chronic lithium toxicity during regular treatment. The dosages of lithium carbonate and serum lithium concentrations at symptom onset varied across the cases, with our patient exhibiting the highest lithium concentration recorded.

Table 2.

Clinical Features of Reported Cases of CLOCCs Caused by Lithium Toxicity.

Reference	Age	Sex	Diagnosis/Symptoms	Lithium dose (mg/day)	Serum lithium concentration at onset (mmol/L)	Treatment	Outcome
(2)	55	Male	Acute lithium toxicity/ Bilateral tremor, dysarthria	400	0.64	Discontinuation of lithium	CR
(3)	28	Female	Chronic lithium toxicity/ Generalized tremor, rigidity, dysarthria, fever, tachycardia	1,200	1.24	Discontinuation of lithium	CR
(3)	59	Female	Chronic lithium toxicity/ Fever, generalized tremor	1,200	0.31	Discontinuation of lithium	CR
(4)	37	Male	Acute-on-chronic lithium toxicity/ Fever and altered mentation	N/A	1.94	Discontinuation of lithium	CR
(5)	40s	Female	Chronic lithium toxicity/ Fever, disturbance of consciousness, dysarthria	800	2.31	Discontinuation of lithium	CR
(6)	35	Female	Acute-on-chronic lithium toxicity/ An apathetic expression, refused food intake	1,750	1.46	Discontinuation of lithium	CR
Our case	30	Female	Acute-on-chronic lithium toxicity/ Fever, disturbance of consciousness, nystagmus	400	3.54	Discontinuation of lithium, hemodialysis	PR

N/A: not available, CR: complete recovery, PR: partial recovery

Acute-on-chronic or chronic lithium toxicity may be associated with more severe neurological manifestations than acute toxicity alone due to cumulative neuronal vulnerability. This may explain the presence of CLOCCs in our patient despite a relatively low daily dose of lithium carbonate.

In this case, identifying CLOCCs through early MRI led to a prompt diagnosis of lithium toxicity and thus timely hemodialysis treatment, even before serum concentration results were available. These actions enabled early recovery despite the patient having the highest plasma lithium levels reported (Table 2). In comparison to previous reports, our case illustrates the diagnostic value of early brain MRI in recognizing CLOCCs, serving as a radiological pointer for lithium neurotoxicity in long-term lithium users; hence, facilitating the timely initiation of renal replacement therapy and prompt recovery.

Conclusion

This case underscores the diagnostic utility of brain MRI in identifying CLOCCs as a potential neuroimaging marker of lithium toxicity, particularly in patients with long-term lithium use. Early recognition of CLOCCs can facilitate prompt intervention, such as hemodialysis, and improve clinical outcomes. Clinicians should consider lithium toxicity in the differential diagnosis of CLOCCs, especially when neurological symptoms are present without other apparent causes. Further research is warranted to elucidate the mechanisms linking lithium toxicity to CLOCCs and to validate their prognostic significance.

The authors state that they have no Conflict of Interest (COI).