Research report: Case Report: Hyperactive delirium with psychotic symptoms associated with severe hypothyroidism during pembrolizumab therapy. Complete remission after levothyroxine and no recurrence after pembrolizumab rechallenge

A 78-year-old woman with metastatic colorectal cancer developed progressive confusion over about three weeks, followed by marked agitation and psychotic symptoms, roughly three months after starting the immunotherapy drug pembrolizumab, and clinicians in Bern, Switzerland, traced the psychiatric deterioration to a cause that is easy to overlook: profound hypothyroidism. Her thyroid hormones had collapsed to undetectable levels, and once she received thyroid hormone replacement, the hallucinations, persecutory delusions, and agitation disappeared completely. The case, published in Frontiers in Psychiatry on 27 October 2026, also documents that the same drug was restarted afterward with no return of the neuropsychiatric illness.

Immune checkpoint inhibitors such as pembrolizumab have transformed the treatment of several advanced cancers, but they work by unleashing the immune system, and that unleashed immune system sometimes attacks the patient’s own organs. These immune-related adverse events can affect multiple organ systems, and endocrine toxicities are among the most frequent. Thyroid dysfunction is particularly common during therapy with programmed cell death protein 1 inhibitors, the class to which pembrolizumab belongs. In most patients, the thyroid problem is picked up through routine laboratory monitoring or announces itself with nonspecific complaints such as fatigue, cold intolerance, or weight gain. Severe hypothyroidism that presents primarily as an acute neuropsychiatric emergency, with hyperactive delirium, hallucinations, and persecutory delusions, is rare and can be extremely difficult to distinguish from other neurological, metabolic, or psychiatric causes.

The woman in this report had no previous psychiatric history and had been living independently. She began palliative pembrolizumab on July 4, 2025, at which point her baseline thyroid-stimulating hormone was entirely normal at 1.99 mU/L. Her relevant medical history included chronic kidney disease of moderate severity and mild aortic stenosis. About three months later, after four administrations of the drug, she developed progressive confusion and a functional cognitive decline that unfolded over roughly three weeks. This was followed by marked agitation and psychotic symptoms, including persecutory and poisoning delusions. On October 3, 2025, she was brought to the emergency department in a state of severe disorganization and agitation: she was disoriented in all spheres, her responses were incoherent, and her neurological examination showed no focal deficits.

The clinical picture during hospitalization fluctuated dramatically. She was intermittently more accessible, but at other times she remained disoriented, motorically restless, and increasingly agitated. She experienced vivid visual hallucinations, persecutory and poisoning delusions, and ideas of external control. Her agitation worsened in the evening and at night, and she barricaded herself and became verbally and physically aggressive toward staff. The diagnosis of hyperactive delirium was made clinically, based on the acute disturbance of consciousness and cognition, disorientation, incoherent thought and speech, psychomotor agitation, hallucinations and delusions, sleep-wake disturbance, and the fluctuating course in the context of acute medical abnormalities. On examination, her thyroid was not enlarged or tender, her cardiovascular findings were normal, and her neurological examination showed normal strength and tone with no pathological reflexes or ataxia.

The laboratory findings were striking. During interval monitoring on September 8, 2025, her thyroid-stimulating hormone had already climbed to 41.20 mU/L. On the day of the emergency presentation, it measured 37.90 mU/L, while her free T3 was below the assay’s lower limit of detection at less than 0.60 pmol/L, against a reference range of 3.74 to 6.11, and her free T4 was likewise below the lower limit at less than 0.5 pmol/L, against a reference range of 11.5 to 19.6. The authors noted that the relatively modest TSH elevation despite very low free thyroid hormone levels was interpreted cautiously: a rapidly evolving immune-related thyroiditis with incomplete TSH equilibration was considered possible, and a partial central component could not be fully excluded from the available data. Concomitant abnormalities included moderate hypoosmolar hyponatremia, with sodium falling from 135 mmol/L at baseline to 124 to 125 mmol/L at presentation, along with mild hypokalemia.

The diagnostic workup illustrates how complicated acute mental status changes can be in patients receiving checkpoint inhibitors. The differential diagnosis included immune checkpoint inhibitor-associated encephalitis, acute structural brain pathology, adrenal insufficiency or hypophysitis, metabolic encephalopathy related to the hyponatremia, and a primary psychiatric disorder. Neuroimaging did not reveal an acute explanatory lesion: a cranial CT showed a small punctate hyperdensity interpreted as possible hemorrhage versus calcification, and a subsequent MRI described an occipital cavernoma without acute hemorrhage. Available cortisol values did not support clinically overt adrenal insufficiency, although adrenocorticotropic hormone values and a comprehensive assessment of other pituitary axes were not available. Lumbar puncture and electroencephalography were considered but never performed, first because of severe agitation and later because of rapid clinical improvement, a gap the authors acknowledge limits diagnostic certainty, since immune-mediated encephalitis and subclinical epileptic activity could not be fully excluded.

Treatment proceeded on two fronts. Pembrolizumab was paused, and levothyroxine was initiated on October 3, starting at 50 micrograms per day. Severe agitation during the medical admission required intravenous midazolam, and she was transferred to inpatient psychiatry on October 12. The levothyroxine dose was increased to 100 micrograms per day on October 23 and to 150 micrograms per day on November 6 because of persistent biochemical hypothyroidism. The clinical recovery tracked the biochemical one. On October 21, she still reported persecutory ideas. By October 27, staff documented fewer psychotic statements and increasing independence. By October 31, she said she no longer felt persecuted, describing the previous experience as similar to a nightmare. By November 6 and 11, she was clear in consciousness, calm, coherent, and free of hallucinations and delusional content. At discharge on November 12, 2025, she was fully oriented, independent in daily activities, and without positive psychotic symptoms.

The temporal sequence was clinically suggestive, the authors argue: thyroid function was normal before pembrolizumab initiation, severe hypothyroidism was documented at the time of acute neuropsychiatric deterioration, and symptoms resolved after thyroid hormone replacement and correction of metabolic abnormalities. This presentation overlaps phenomenologically with myxedema psychosis, a classic description of psychosis in severe thyroid failure, but the prominent disturbance of consciousness and orientation, the fluctuating cognitive status, the motor agitation, and the medical precipitants favored a diagnosis of delirium with psychotic symptoms rather than a primary psychotic syndrome. Still, the authors are careful about causality. The presentation was probably multifactorial, with hyponatremia, sleep disruption, psychosocial stress, and possible medication-related factors potentially contributing. Thyroid antibodies were not measured during the acute episode; delayed testing in August 2026 showed mildly elevated thyroperoxidase antibodies with negative TSH receptor and thyroglobulin antibodies, suggesting thyroid autoimmunity but unable to determine whether the antibodies were present at onset or emerged later. As a single case report, the findings are not generalizable.

The rechallenge is perhaps the most practically interesting element. After clinical stabilization and adequate thyroid hormone replacement, pembrolizumab was restarted at three-week intervals with close laboratory monitoring, and no recurrence of delirium or psychotic symptoms followed. At follow-up in April 2026, the patient remained on levothyroxine 150 micrograms per day, was living independently, had no recurrence of delirium or psychosis, and her tumor staging showed stable disease. Thyroid laboratory values, including TSH and cortisol, are monitored every six weeks. Although conclusions from a single case must remain cautious, the authors suggest that rechallenge may be feasible after stabilization of endocrine immune-related adverse events, provided careful interdisciplinary follow-up and regular thyroid monitoring are ensured.

The practical message for clinicians is straightforward. New-onset agitation, hallucinations, psychotic symptoms, or delirium in a patient receiving immune checkpoint inhibitors should prompt rapid assessment for potentially reversible immune-related endocrine toxicities, including thyroid dysfunction and adrenal insufficiency, before the symptoms are attributed to a primary psychiatric disorder. This matters especially because such presentations can initially resemble primary psychiatric illness when psychotic symptoms dominate the picture. Early recognition and treatment, the authors argue, may avoid prolonged morbidity, unnecessary escalation of psychiatric treatment, and premature discontinuation of effective cancer therapy. The case also underscores that antibody negativity does not exclude immune checkpoint inhibitor-associated thyroid dysfunction, and that a normal baseline thyroid value offers no protection against a later collapse. For a drug class now given to vast numbers of patients with advanced cancer, a reversible cause of acute psychosis hiding in a routine blood test is a finding worth remembering.

Subject of Research: A case report of hyperactive delirium with psychotic symptoms linked to severe hypothyroidism during pembrolizumab therapy, with remission after levothyroxine and no recurrence after rechallenge.

Article Title: Case Report: Hyperactive delirium with psychotic symptoms associated with severe hypothyroidism during pembrolizumab therapy. Complete remission after levothyroxine and no recurrence after pembrolizumab rechallenge

Article References: Vieira Fregolente, E., Bühler, S., & Lahr, J. (2026). Case Report: Hyperactive delirium with psychotic symptoms associated with severe hypothyroidism during pembrolizumab therapy. Complete remission after levothyroxine and no recurrence after pembrolizumab rechallenge. Frontiers in Psychiatry, 17, 1905725. https://doi.org/10.3389/fpsyt.2026.1905725

Image Credits: AI Generated

DOI: 10.3389/fpsyt.2026.1905725

Keywords: pembrolizumab, hypothyroidism, delirium, psychosis, immune checkpoint inhibitors, immune-related adverse events, levothyroxine, hyponatremia, thyroid dysfunction, case report, oncology, neuropsychiatry