Stimulation of Malignant Growth in Rodents by Antidepressant Drugs at Clinically Relevant Doses¹

Tricyclic antidepressants, such as amitriptyline (Elavil), and the non-tricyclic agent, fluoxetine (Prozac), bind to growth-regulatory intracellular histamine receptors, associated with anti-estrogen binding sites in microsomes and nuclei. The prototype anti-estrogen binding site/intracellular histamine receptor ligand, N,N-diethyl-2-[4-(phenylmethyl)phenoxy]ethanamine HCl, inhibits normal cell proliferation in vitro but stimulates tumor growth in vivo. Because of their structural similarity to N,N-diethyl-2-[4-(phenylmethyl)phenoxy]ethanamine HCl, we carried out studies to determine whether amitriptyline and fluoxetine stimulate tumor growth and/or development in rodents at concentrations relevant to the treatment of human depression (equivalent human dose range, ∼100–150 mg/day for amitriptyline and ∼20–80 mg/day for fluoxetine). All experiments were performed blinded. In studies of growth stimulation of transplantable syngeneic tumors, groups of mice were inoculated s.c. with C-3 fibrosarcoma cells or given i.v. or s.c. injections of B16f10 melanoma cells, followed 24 h later by daily i.p. injections of saline, amitriptyline, or fluoxetine. Tumor latency (fibrosarcoma), aggregate tumor weight (s.c. injected melanoma), or time to death from pulmonary metastasis (i.v. injected melanoma) was determined; drug-induced stimulation of DNA synthesis in C-3 fibrosarcoma cells in vitro was correlated with tumor growth acceleration in vivo. In a mammary carcinogenesis model, the effects of chronic saline, amitriptyline, or fluoxetine administration on the rate and frequency of development of mammary tumors in rats fed dimethylbenzanthracene (DMBA) were compared. Eight of 20 amitriptyline- or fluoxetine-treated mice developed fibrosarcoma tumors by day 5, as compared to none of 20 saline controls (P < 0.002). Similarly, 20 of 21 DMBA-treated rats receiving the antidepressant drugs developed 33 mammary tumors by week 15 as compared to 5 tumors in 4 of 7 DMBA-treated rats receiving saline (P < 0.001). For both models, tumor latency decreased 30–40% and, in the DMBA model, tumor frequency increased > 2-fold in the antidepressant-treated rats as compared to controls. Stimulation of fibrosarcoma growth in vivo correlated with a corresponding bell-shaped drug-induced increase in DNA synthesis in vitro. While the median time to death from pulmonary metastases did not differ among groups given i.v. injections of melanoma cells, a significant (P < 0.01) stimulation of growth of s.c. injected melanoma was observed in mice receiving the antidepressants. On the basis of these findings, we suggest that: (a) properly designed epidemiological studies be carried out to determine whether antidepressant drugs adversely affect the development or course of cancer in humans; and (b) preclinical drug screening procedures should include studies of tumor promotion, in addition to carcinogenesis.