Introduction
Delirium is one of the most common neuropsychiatric complications among patients with cancer, especially during the advanced and terminal stages. It is associated with higher mortality and causes severe distress for patients, families, and caregivers (Breitbart et al., Reference Breitbart, Gibson and Tremblay2002; Hui et al., Reference Hui, dos Santos and Reddy2015; Morita et al., Reference Morita, Tei and Tsunoda2001). Understanding the various causative factors of delirium among patients with cancer is important to relieve symptoms (Inouye et al., Reference Inouye, Marcantonio and Metzger2014; Meagher, Reference Meagher2001).
Wernicke's encephalopathy (WE), caused by a thiamine deficiency, is one cause of delirium in patients with cancer (Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Hsu and Hatzoglou2014b; Onishi et al., Reference Onishi, Kawanishi and Onose2004, Reference Onishi, Sugimasa and Kawanishi2005, Reference Onishi, Ishida and Toyama2015; Sechi & Serra, Reference Sechi and Serra2007; Yae et al., Reference Yae, Okuno and Onishi2005). This disorder is potentially reversible if properly diagnosed and treated with thiamine; however, it causes severe, irreversible, and fatal brain damage if left untreated (Korsakoff syndrome). The estimated mortality rate is 17% (Victor et al., Reference Victor, Adams and Collins1971); therefore, early detection and immediate thiamine administration is critical.
This disorder often remains unrecognized and overlooked (Harper, Reference Harper1979; Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Kutner and Nicolson2012; Ogershok et al., Reference Ogershok, Rahman and Nestor2002) in part because of the clinical diversity of symptoms (Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Kutner and Nicolson2012). Only 16% of patients exhibit the typical clinical symptoms of WE—namely confusion, ocular palsy, and ataxia—and 19% of patients are asymptomatic (Harper et al., Reference Harper, Giles and Finlay-Jones1986).
Several clinical approaches have been applied to diagnose WE. Typical magnetic resonance imaging findings are not evident in all patients, computed tomography scans are inadequate (Antunez et al., Reference Antunez, Estruch and Cardenal1998), and measuring blood thiamine levels takes several days (Onishi et al., Reference Onishi, Ishida and Toyama2015); thus, a clinical diagnosis of WE remains the standard (Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Kutner and Nicolson2012). The clinical symptoms and situations in which WE appear need to be clarified.
To date, most reports of WE arising in patients with cancer describe inpatient or emergency settings (Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Alici and Hatzoglou2016; Onishi et al., Reference Onishi, Kawanishi and Onose2004, Reference Onishi, Sugimasa and Kawanishi2005, Reference Onishi, Ishida and Toyama2015); the prevalence of WE in regular outpatient clinics is unknown.
Here, we describe two outpatients with advanced cancer who developed delirium that was discovered during regular attendance at a psycho-oncology clinic. These patients did not exhibit the typical symptoms of WE, but a thiamine deficiency as a cause of delirium was suspected from detailed interviews with family members. Thiamine administration resulted in complete recovery from delirium and prevented irreversible brain damage.
Case reports
Case 1
A 76-year-old woman with pancreatic cancer and liver metastasis was referred by her physician to a psycho-oncology outpatient clinic because of depression and insomnia. Major depressive disorder (MDD) was diagnosed and she was treated with paroxetine. She achieved clinical remission after 2 months of psychotherapy and pharmacotherapy and attended a psycho-oncology outpatient clinic monthly for maintenance MDD therapy.
Although MDD did not recur, she weakened because of progressive pancreatic cancer and received palliative care at home but no psychiatric treatment. Thirteen months after initially attending the psycho-oncology clinic as an outpatient, she presented with drowsiness and ataxia that her family confirmed had persisted for several days. They also noted that she had been absent-minded all that day and the day before, and had frequently asked her family about the next consultation date at the clinic.
Her symptoms fulfilled the Diagnostic and Statistical Manual of Mental Disorders-IV criteria for delirium, but the findings of a neurological examination were unremarkable except for ataxia. A detailed interview with her family revealed that she had lost her appetite about 8 weeks previously.
Ingested thiamine is stored in the body for about 18 days (MacLean et al., Reference MacLean, Rhode and Shizgal1983) and the appetite loss in this patient had persisted for about 8 weeks, indicating that a thiamine deficiency had caused the delirium. She was administered oral thiamine (75 mg/day), which resolved the delirium within a few days. Five days later, her serum thiamine level, measured by high-performance liquid chromatography, was 16 ng/mL (reference range: 24-66 ng/mL), confirming the diagnosis of WE.
Case 2
A physician referred a 79-year-old man with advanced stomach cancer and depression to the psycho-oncology clinic as an outpatient. He had been diagnosed with stomach cancer accompanied by bone and lymph node metastasis 1 month previously and was under treatment with 100 mg TS-1 combination (tegafur, gimeracil, oteracil potassium) and 10 mg oxycodone. Hypocalcemia developed after 2 weeks and TS-1 therapy had been suspended.
An appointment psycho-oncology outpatient clinic was made 2 weeks before his first consultation at the request of family members because he had become depressed after learning that he had cancer.
When he initially presented at the psycho-oncology clinic, he was unable to describe the date or the name of hospital. He was unable to concentrate, and family members noted that he had been absentminded for most of the day and could not answer questions. His symptoms fulfilled the Diagnostic and Statistical Manual of Mental Disorders-IV criteria for delirium, but a neurological examination revealed no motor weakness, ataxia, or nystagmus and head computed tomography scans revealed no significant findings.
Laboratory tests revealed a serum magnesium level of 2.6 mg/dL (reference range, 1.8–2.5 mg/dL), but this was not regarded as the cause of the delirium.
A detailed interview with his family revealed that his appetite had been about 70% normal for several weeks. We could not define the cause of delirium; a thiamine deficiency was suspected for the reasons described previously. High-performance liquid chromatography showed that his serum thiamine level was 22 ng/mL (reference range, 24–66 ng/mL), which confirmed a diagnosis of WE.
He was intravenously administered 100 mg of thiamine, which resolved the symptoms of delirium within a few days. One week later, he provided adequate answers to questions posed by a psycho-oncologist at the outpatient clinic. Disorientation disappeared, his speech had become comprehensible, and he was able to communicate with his family.
Discussion
We described two outpatients with cancer who developed WE that did not involve the complete typical triad of symptoms. Both of them regularly attended a psycho-oncology clinic as outpatients. The detection of WE and immediate thiamine administration resulted in recovery and prevented irreversible brain damage due to Wernicke-Korsakoff syndrome.
Routine laboratory or neuroimaging findings cannot identify causes of delirium, a condition that is rare among outpatients who attend psycho-oncology clinics (Akechi et al., Reference Akechi, Nakano and Okamura2001; Kissane & Smith, Reference Kissane and Smith1996; Massie & Holland, Reference Massie and Holland1987; Tada et al., Reference Tada, Matsubara and Kawada2012). However, a careful clinical examination led to a diagnosis of WE in our two patients.
That both patients did not exhibit the classical triad of WE is not surprising because they are found in only 16% of autopsy samples and 11% of consultation samples (Harper et al., Reference Harper, Giles and Finlay-Jones1986; Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Alici and Hatzoglou2016). The critical indicator of WE in both patients described here was appetite loss that had persisted for several weeks. The human body cannot synthesize thiamine; thus, it has to be ingested orally or delivered intravenously. Thiamine is stored mainly in the liver for about 18 days (MacLean et al., Reference MacLean, Rhode and Shizgal1983).
The main reason for the thiamine deficiency in our patients was reduced intake from the persistent appetite loss. Other factors that could have also been associated with the development of WE is the liver metastasis in case 1, which might have reduced the hepatic storage capacity (Onishi et al., Reference Onishi, Sugimasa and Kawanishi2005) and the 5-fluorouracil administered to case 2 to treat gastric cancer. Clinical and experimental findings indicate that 5-fluorouracil could be associated with a thiamine deficiency via either increased thiamine utilization or catabolism (Aksoy et al., Reference Aksoy, Basu and Brient1980; Basu et al., Reference Basu, Aksoy and Dickerson1979).
We identified appetite loss in our patients only by interviewing family members, because both patients were delirious and could not answer questions posed by a psycho-oncologist. We previously emphasized that appetite loss persisting for >18 days might lead to a thiamine deficiency (Onishi et al., Reference Onishi, Ishida and Toyama2015) because humans cannot synthesize thiamine and the liver stores thiamine for only 18 days (MacLean et al., Reference MacLean, Rhode and Shizgal1983). However, the exact amount of thiamine ingested by the patients was not investigated because the amount consumed food was not monitored. When outpatients with cancer present with delirium, family members should be questioned about the appetite of the patient over the past few weeks. If appetite loss has persisted for longer than 2 weeks, WE should be considered as a cause of delirium, and thiamine should be immediately administered because it is safe for such patients (Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Chabon and Nicolson2014a).
Alternative diagnostic criteria that are more sensitive than the classical triad for a diagnosis of WE have been developed (Caine et al., Reference Caine, Halliday and Kril1997). These criteria compose eye signs, cerebellar signs, malnutrition, and mild memory impairment or confusion. Any two of these signs indicate a diagnosis of WE, and both of our patients fulfilled these diagnostic criteria for WE. Although these criteria were developed based on findings from patients with alcoholism, recent studies have indicated that they are more effective than the classic triad for diagnosing WE in patients with cancer (Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Alici and Hatzoglou2016).
In case 1, we administered thiamine orally. Usually, the intravenous administration of thiamine is recommended for the treatment of WE (Galvin et al., Reference Galvin, Brathen and Ivashynka2010; Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Kutner and Nicolson2012). In elderly patients, oral thiamine (10 mg/day) was reported to have improved the quality of life compared with a placebo (Wilkinson et al., Reference Wilkinson, Hanger and Elmslie1997). We think that the ability to absorb thiamine via the small intestine remained intact in this patient. The suspension of TS-1, which is associated with thiamine deficiency via either increased thiamine utilization or catabolism, might affect this.
A thiamine deficiency was not evident in our patients before they developed WE. This suggests that patients with cancer who often experience appetite loss are at risk of developing WE (Hollen et al., Reference Hollen, Gralla and Kris1993). Such patients should be screened and administered thiamine to prevent delirium developing from a thiamine deficiency. However, the symptoms of WE are broad and nonspecific (Isenberg-Grzeda et al., Reference Isenberg-Grzeda, Kutner and Nicolson2012), and the clinical symptoms and laboratory findings of such patients should be established to facilitate diagnostic accuracy.
In conclusion, awareness of WE in the outpatient setting is important when patients do not exhibit the classical triad. Appetite loss might be the key to a diagnosis of WE in the absence of other causes of delirium.
