Acute Mental Status Changes in Two Patients Taking Capecitabine: A Case Report
Case Presentations
Patient #1
A 79-year-old Caucasian male was referred to hospice because of encephalopathy after a witnessed generalized tonic-clonic seizure at home. He had been in his usual state of health until the day of admission when he experienced at least one witnessed seizure. He was brought unconscious to the hospital Emergency Department and never fully woke up.
His past medical history was significant for colon cancer, hypertension, and hyperlipidemia. His outpatient medications were capecitabine, atenolol, simvastatin, and lisinopril. His social history was negative for alcohol or illicit drug use.
He had been working out in the sun and exercising quite vigorously prior to his seizure. According to his son, the patient was drinking iced tea to keep himself hydrated. A few days earlier, he had had a near syncopal episode, for which he did not seek medical help.
The initial neurological examination failed to reveal any specific deficits. Computed tomography scan of the brain, without contrast, showed only atrophy. He was started on phenytoin and IV fluids. Admission blood work included a metabolic profile, a complete blood count (CBC), and a drug screen. The metabolic profile was normal except for a mildly elevated glucose of 150 mg/dL. Liver function tests were also within normal range except for a total protein of 6.0 g/dL (normal = 6.2-8.6 g/dL) and an albumin of 2.9 g/dL (normal = 3.2-5.5 g/dL). The toxicology screen was negative. The CBC showed a white blood cell (WBC) count of 5.5/µL, hemoglobin of 13.4 g/dL, hematocrit of 38.0%, mean corpuscular volume (MCV) of 103.4 µm3, mean corpuscular hemoglobin (MCH) of 36.4 pg/cell, and platelet count of 157 x 103/µL.
The day after admission, the patient briefly regained consciousness but was lethargic most of the time. Phenytoin levels were within the therapeutic range. The neurologist thought it may be subclinical seizures and started the patient on divalproex sodium. However, the patient remained unresponsive.
A magnetic resonance imaging (MRI) scan done with gadolinium showed only some mild periventricular white matter disease and atrophy. An electroencephalogram failed to show any epileptiform activity. Chest x-ray was negative for acute infiltrates, and blood and urine cultures were negative, as was a spinal tap. Thyroid function tests and ammonia levels were within normal limits. Five days into his admission, when the patient failed to awaken, hospice was called on consult at the request of the family. As part of the evaluation, hospice consulted a PharmD regarding the possible relationship between elevated MCV and MCH and capecitabine. Reports of thiamine deficiency were found in patients treated with 5-fluorouracil (5-FU), of which capecitabine is a prodrug. That led hospice to recommend a trial of intravenous thiamine, using standard doses. The patient woke up after 2 days on this therapy. At that point, the family decided not to keep him in hospice, and he was discharged to a rehabilitation facility.
Patient #2
A 75-year-old Caucasian female was referred to hospice because of stage IV squamous cell carcinoma of the vulva that was not responding to therapy with capecitabine. She was initially admitted to an inpatient unit because of electrolyte abnormalities (low phosphorus and low magnesium levels), weakness, and a left lower extremity cellulitis. She was treated with oral replacement of her magnesium deficiency and ciprofloxacin for her cellulitis.
At the time of admission, she was alert and oriented to person, place, time, and situation. The CBC showed a WBC count of 14.3/µL, hemoglobin of 10.3 g/dL, hematocrit of 30.4%, MCV of 102.3 µm3, MCH of 34.5 pg/cell, and platelet count of 331 x 103/µL. Her metabolic profile was normal except for a calcium level of 7.7 mg/dL, and her liver function tests were also normal except for an albumin of 2.5 g/dL. The corrected calcium was 9.2 mg/dL, which is normal (8.4-10.2 mg/dL). Vitamin B12 levels were 492 pg/mL (normal = 212-911 pg/mL) and folate levels were 3.80 ng/mL (normal = 1.1-20.0 ng/mL).
The patient complained of lower extremity weakness that had been bothering her for several weeks prior to admission and anorexia. During her stay at the inpatient unit she developed diarrhea. It was thought to be due to the oral magnesium so this electrolyte was replaced parenterally. The patient was given a loading dose of magnesium in 500 ml of D5W, to run over- night. Within 24 hours, she became restless, confused, disoriented with confabulation and visual hallucinations. When her husband was interviewed to update him and to get more history on his wife’s habits, he disclosed that she had been drinking, mostly wine, quite heavily for several years.
The patient was immediately started on IV thiamine. Within several hours of the first dose, she started to recognize her husband and her surroundings. By the second day, she was back to baseline. She was continued on IV thiamine for 1 more day and was discharged to a nursing home on oral thiamine and multiple vitamins.
Discussion
Capecitabine is an oral chemotherapeutic agent that is used to treat solid gastrointestinal tumors. It is a prodrug that is enzymatically converted to 5-FU in the tumor.1 5-FU is an antimetabolite, which works as a thymidylate synthetase inhibitor.2 It has been associated with a depletion of thiamine stores in both experimental3,4 and human subjects.5-7 It has also been associated with neurotoxicity by other mechanisms.7,8
Thiamine is a member of the B vitamins. It is water-soluble. Its chemical structure contains a pyrimidine ring, and it plays an important role in carbohydrate metabolism.9 Deficiencies of this vitamin can manifest as megaloblastic anemia, neurodegenerative changes (eg, Wernicke-Korsakoff syndrome; the MRI findings in the first patient were not consistent with this diagnosis), and cardiac dysfunction (beriberi).10 Caffeine and tea can inhibit absorption of thiamine.10,11 Alcohol is the best documented cause of 
thiamine deficiency.12-15 Other substances and treatments that can interfere with thiamine absorption or metabolism are outlined in Table I.23-25 The classic presentation of acute thiamine deficiency is summarized in Table II.12 However, it is important to note that, more often than not, delirium or some alteration of mental status can be the actual presentation, which can lead to underdiagnosing.16-19 Several studies show that thiamine deficiency may be more prevalent than previously thought. In fact, in one study of necropsies of 36 community-dwelling elderly people without dementia, 17% were found to be thiamine-deficient.17
Capecitabine neurotoxicity has been previously reported.20-22 In the first patient’s case, it was felt that the excessive consumption of caffeine-containing fluids plus therapy with this prodrug may have contributed to his encephalopathy and seizures. In the second patient, it was believed that the combination of alcohol consumption and capecitabine left the patient in a deficient state that was made worse by the IV glucose load used to replace her magnesium.
Unfortunately, thiamine levels were not drawn in either patient before starting treatment. Therefore, it is entirely possible that the response to treatment was entirely coincidental. However, even in the absence of rigorous scientific methodology, given the presentation, one should consider the possibility of a cause-and-effect phenomenon.
There are several messages to be gleaned from these cases. The first is to consider lifestyle and complementary medicine usage in patients who present with acute medical illnesses not explainable by the current therapeutic regime alone. Sometimes, interactions between the prescribed drugs and nonprescription supplements may be responsible for the otherwise unexplainable clinical presentation. For a review of possible drug interactions with natural health products, please refer to http://ncp.sagepub.com/cgi/content/abstract/20/1/33.
Another, no less important, message is that close cooperation with other members of the interdisciplinary team—in this case, the pharmacist—might shed light and facilitate solutions that may not be readily apparent. And, last but not least, we must remember that older, debilitated patients are more likely to suffer adverse effects of medications, which may result in hospitalization. Furthermore, cognitive impairment in the elderly should not be taken as an expected consequence of aging without a search for contributing causes, as has been reported elsewhere.26
The authors report no relevant financial relationships.
Dr. Gonzalez is a Palliative Medicine Specialist, Hospice of Palm Beach County, and is a part of the teaching program at that institution’s Palliative Medicine Fellowship, West Palm Beach, FL; and Dr. Gates is a Pharmacist, Hospice of Palm Beach County.