APOS Clinical E-Mail Update #3
22 October 2003

In this Update:

When patients refuse to eat, what should be considered?

Ganzini L, Goy ER, Miller LL, Harvath TA, Jackson A, Delorit MA. Nurses’ experiences with hospice patients who refuse food and fluids to hasten death. New Engl J Med 2003;349:349-365.

This is a provocative study placed prominently. Because there are few reports describing the quality of death of patients who voluntarily refuse food and drink, these researchers sent a survey to 429 hospice nurses to ask about patients who refused intake with the intention of dying. One third of the 307 nurses who filled out the survey had in the previous 4 years cared for a patient who refused intake. For the great majority of these patients described in nurses’ surveys, the two weeks before death seemed to have low levels of pain and suffering, and the death was rated as good. According to hospice nurses, patients chose this method of dying because they were ready to die, saw continuing to live as pointless, and had a sense of poor quality of life. In the nurses’ view, 12% of these patients had depression that influenced the decision to stop eating.

The nurses’ perception of hospice patients who deliberately hastened death by refusing food and fluids were compared with their perceptions of hospice patients who sought legalized physician-assisted suicide to hasten death during the same period. Compared to the nurses’ description of patients who had sought physician-assisted suicide, these patients were not as apt to want to control the circumstances of their death. They were more prepared to die, less likely to fear loss of dignity, and more likely to lack social support. Compared to those who chose physician-assisted suicide, these patients were more apt to be older, less likely to have cancer, and more likely to have a terminal neurological disease.

The setting is the 50 Medicare-certified home hospice programs in Oregon and two additional programs that serve Oregon patients. Nurses were asked to comment by survey on the patient’s overall peacefulness and suffering in the two weeks before death and the quality of the dying. They were asked to compare the family members with other hospice patients’ families. A group of 55 nurses were asked the same questions about patients who sought physician-assisted suicide.

The patient’s decision not to eat was reviewed at an interdisciplinary hospice conference in 74% of the 126 cases; 62% of cases were evaluated by a hospice social worker, and 9% by a mental health professional. Of these patients, 16 reversed their decision because of family encouragement to eat (5), discomfort (4), amelioration of depression (one), and alleviation of other concerns (one).

This article is prominently placed in the New England Journal of Medicine. It sheds light on the minority of hospice patients who refuse food in order to die. However, the light is shed from a distance, by voluntary survey reports of hospice nurses who had an investment in the care delivered. They report through the lens of their own values and biases. Many questions remain unanswered and are worthy, as the authors suggest, of discussion.

What is the patients’ story? What is their emotional pain? Those of us interested in the psychological care of patients with cancer or patients at the end of life might wonder what the patients would say if we spoke with them directly about their decision and their inner distress. While 45% of patients seeking physician-assisted suicide were seen by a mental health professional, ONLY 9% of these patients who chose to stop eating in order to die were evaluated by a mental health professional.

We really don’t know the differential diagnosis of the patients. Not eating may be the last bit of control that they have, or not eating may be related to anorexia and eating difficulties. Since clinical depression is not always a straightforward diagnosis, the key question is how many of these patients did have clinical depression and associated anhedonia, loss of appetite, and suicidal thoughts. Since more of these patients had neurological disease, we wonder also whether difficulty swallowing made intake more difficult. Since these were patients with less social support than those who sought physician help with suicide, we might wonder what more respect and attention might have been offered to these patients.-DBG

     [back to top]

What are novel treatments for itching related to cancer?

Davis MP, Frandsen JL, Walsh D, Andresen S, Taylor S. Mirtazapine for pruritis. J Pain Symptom Manage 2003;25:288-291.

Itching is an unusual symptom among cancer patients, but it can be a clue to lymphoma or biliary obstruction. It is also a sensation associated with neuropathy. These authors review the pathophysiology and report benefit from mirtazapine in three cases.

Pruritis is often treated with antihistamines, diphenhydramine and doxepin. Paroxetine has also been helpful for the itching from lymphoma, opioid, or cholestasis.1  Ondansetron has also relieved itching in cholestasis, opioids, and renal failure.2  Stimulation of unmyelinated C-fibers in the dermal-epidermal junction lead to pruritis, and H1 receptors, 5HT2 and 5HT3 receptors have been implicated. A scratch stimulates myelinated A delta sensory fibers to block the itch.

Mirtazepine is a H1, 5HT2, and 5HT3 receptor blocker. Mirtazepine 15 mg at night relieved a man with metastatic adenocarcinoma to the liver in spite of persistent jaundice. Mirtazepine 30 mg worked for a woman with Hodgkin’s disease as did 15 mg for a man with chronic lymphocytic leukemia. - DBG

  1. Zylicz Z, Smits C, Krajnik M. Paroxetine for pruritis in advanced cancer. J Pain Symptom Manage 1998, 16:121-124.
  2. Wilde MI, Markham A. Ondansetron: a review of its pharmacology and preliminary clinical findings in novel applications. Drugs 1996;52:773-794.

     [back to top]

What is the best strategy to allay anxiety due to tumor marker reports?

Dahele M, Camidge R. Tumour marker reference ranges and patients’ anxiety.[comment]. Lancet 2003; 361:882.

Sundar S, Symonds P. Cancer patients’ anxiety. Lancet 2003;361:1746.

Two comments in Lancet highlight the regular problem for oncologists as they report tumor markers to patients who worry. Dahele and Camidge had suggested that the reference ranges be expressed in smaller units. Sundar and Symonds take issue and report that they offer patients a detailed explanation about the tests. The explanations include context: for instance, a prostatic specific antigen of 0.2 micrograms/liter may seem to be present in a patient without a prostate as it can represent the lowest number reported. The peritoneum also can produce CA 125 in a patient without ovaries.

They emphasized the importance of kinetics rather than actual serum values. If the value drops after treatment and then doubles, the change may be clinically significant even in the normal range. - DBG

     [back to top]

What if the cancer patient is scared of needles?

Fernandes PP. Rapid densensitization for needle phobia. Psychosomatics 2003;44:253-254.

Needle phobias, associated with bradycardia and vasovagal syncope, occur in 3-4% of the population. Fernandes reports a rapid desensitization technique for a patient who needed dialysis. This patient was frightened not by the sight of the needle but by the prick in the skin. The therapy exposed the patient over three hour-long sessions to a hierarchy of stimuli: to the smell of alcohol pads, the stroke of the needle cap, written descriptions of needle insertion spoken by the therapist, and active imagination by the patient of needle placement. These sessions with the therapist over the course of a week were interrupted by practice of counter-responses to vasovagal stimulations. Homework included self-assisted exposure twice a day. - DBG

     [back to top]

Is there risk of antidepressants causing cancer?

Theoharides TC, Konstantinidou A. Antidepressants and risk of cancer: a case of misguided associations and priorities. J Clin Psychopharmacol 2003:23:1-4.

In an editorial, Theoharides and Konstantinidou put in perspective an article that appeared in the journal Blood in 2002 about fluoxetine, paroxetine and citalopram preventing apoptosis in Burkitt lymphoma cells.1  The authors of the laboratory article concluded that serotonin by the serotonin transporter influenced apoptosis. Theoharides and Konstantinidou argue that these in vitro results required serotonin concentrations that are not attainable systemically unless the tumor secretes serotonin.

In an epidemiological study paroxetine was associated with a 7-fold increased risk of breast cancer; this was later reduced to one-tenth the risk.2,3  Reports of laboratory growth of tumors in mice as a result of exposure to antihistamines, amitriptyline, and fluoxetine had been picked up by general news media in the past.4

They conclude that the confusion comes from laboratory evidence taken out of context and contradictory epidemiological findings without any clear mechanism proposed. Unfortunately, these isolated reports were picked up by the media and led to alarm of patients. - DBG

  1. Serfeim A, Grafton G, Chamba A et al. 5-hydroxytryptamine drives apoptosis in biopsylike Burkitt lymphoma cells: reversal by selective serotonin reuptake inhibitors. Blood 2002;99:2545-2553.
  2. Cotterchio M, Kreiger N, Darlington G, et al. Antidepressant medication use and breast cancer risk. Am J Epidemiol 2000;151:951-957.
  3. Traub M. Antidepressants (SSRIs) may increase breast cancer risk. Proc. 35th Annual Meeting Society for Epidemiologic Research, Seattle, WA 2000.
  4. Brandes LJ, Warrington RC, Arron RJ, et al. Enhanced cancer growth in mice administrered daily human-equivalent doses of some H1-antihistamines: predictive in vitro correlates. J Natl Cancer Inst 1994;86:770-775.

     [back to top]


Donna B. Greenberg, MD, Associate Professor of Psychiatry at Harvard Medical School and Psychiatric Consultant in the Massachusetts General Hospital Cancer Center, Dana Farber Partners Cancer Care