|
|
||
 |
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 22, No 1 (January 1), 2004: pp. 204-a-205 © 2004 American Society of Clinical Oncology. DOI: 10.1200/JCO.2004.99.206
Fluoxetine Versus Placebo in Advanced Cancer OutpatientsHematology/Oncology Division, General Hospital of Annemasse, Annemasse, France To the Editor: I read with interest the article by Fisch et al [1] in the May 15, 2003 issue. In a double-blinded randomized study, the authors compared fluoxetine versus placebo in patients with various advanced neoplasms and depressive symptoms. Although overall quality of life and depression were improved in the fluoxetine group, the clinical significance was uncertain because of equivalent best-change scores in the placebo group. Previous randomized trials comparing antidepressant drugs and placebo have also shown a modest benefit of drug therapy, leading to this study [2-7]. In addition, two randomized trials comparing paroxetine versus amitriptyline and fluoxetine versus desipramine, respectively, also showed equivalent improvements in each study [8,9]. The authors stated that conclusions cannot be drawn from previous studies because of a wide predominance of gynecologic cancers and frequent patient dropout. Despite investigators' motivation, sophisticated statistical analysis, and recruitment of a variety of cancer sites, this study does not resolve the problem of generalizability of the results. Low compliance and difficulty in demonstrating a significant improvement raise some hypotheses. The variety of tumor sites reflects neither the epidemiology of cancer nor the morbidity of depression, which may be particularly serious in cancers of the breast, pancreas, or head and neck, and more generally, in rapidly progressing diseases [10-12]. From a methodologic point of view, patients opposed to psychotropic drugs probably are excluded from this type of study, introducing an accrual bias. Recruitment has been difficult. The double-blinded procedure might be questionable because logically, more patients experienced emesis in the fluoxetine group. Other side effects of fluoxetine, such as neuropsychiatric symptoms, dizziness, or diarrhea, also might have had an impact. Patients in each arm are not strictly comparable because family history of depression or psychologic counseling was more frequent in the placebo group, suggesting a higher risk of depression but also possibly recourse to psychological support. The dropout rate remained elevated and approximately 25% of the patients were not assessable. This phenomenon, which depends on multiple factors such as age, comorbidity, education, or social support, is inherent in the recruitment of a large population. The impact of fluoxetine might have been leveled off by the placebo effect but also by the follow-up, which might represent a psychological support. The results confirm that patients with no familial history have had less depression. Fluoxetine has been widely used with success in various settings such as stroke, HIV infection, or diabetes mellitus [13], and the toxicity profile is improved in comparison with that of tricyclics. Nevertheless, fluoxetine may interact with several drugs frequently administered to cancer patients, such as phenytoin, carbamazepine, nonsteroidal anti-inflammatory drugs, antivitamin K, or 5-HT3 antagonists [14,15]. Fluoxetine may cause hypoglycemia in diabetics or seizures in the patient at risk. The treatment should be tailored according to the type and severity of depression, but also to the psychological profile and lifestyle. Conceptually, a sole therapy cannot be generalized to a heterogeneous population with such a multifactorial syndrome. In particular, comparison of any given therapy to placebo seems no longer justified. Author's Disclosures of Potential Conflicts of Interest The author indicated no potential conflicts of interest. REFERENCES
Related Article
Related Correspondence
Related Reply
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
