Risk of cancer associated with the use of antidepressants

Abstracts

INTRODUÇÃO: Alguns estudos sugerem que o uso de antidepressivos poderia aumentar o risco de câncer. Este estudo visa realizar uma revisão sobre o tema. MÉTODO: Foi feita uma busca nas bases de dados MEDLINE e LILACS, utilizando como palavras de busca antidepressant, cancer e nomes das diferentes drogas antidepressivas. RESULTADOS: Onze artigos foram selecionados. Foram encontrados seis artigos sugerindo uma associação positiva fraca entre o uso de antidepressivos e o crescimento tumoral e cinco artigos que não sugeriam a associação. Discussão: Os resultados dos estudos com relação ao risco de câncer associado ao uso de antidepressivos são ainda conflitantes. Na maioria dos estudos, a análise multivariada não mostra associação positiva em uso de antidepressivos e câncer, a não ser em casos específicos, como linfoma de Hodgkin.

Antidepressivos; câncer; risco

REVIEW ARTICLE

Risk of cancer associated with the use of antidepressants

Camila Silva Bôaventura^I; Andréia Nunes Guimarães^I; Guilherme Rebello Soares^I; Anna Milena Barreto Ferreira Fraga^I; Flávia Branco Cerqueira Serra Neves^I; Milena Pereira Pondé^II

^IMedical students, Escola Bahiana de Medicina e Saúde Pública (EBMSP), Salvador, BA, Brazil

^IIPsychiatrist. Associate professor, Pharmacology, EBMSP, Salvador, BA, Brazil

^{Correspondence}

ABSTRACT

INTRODUCTION: Some studies suggest that the use of antidepressants could increase the risk of cancer. This study aims at performing a review on this subject.

METHOD: A search was performed in the MEDLINE and LILACS databases using the keywords antidepressant, cancer and names of varied antidepressant drugs.

RESULTS: Eleven articles were selected. Six articles were found suggesting a positive weak association between use of antidepressants and tumoral growth. In five articles this association was not found.

DISCUSSION: The results of studies on increased risk of cancer associated with antidepressants are still conflicting. In most studies the multivariate analysis did not show positive association between use of antidepressants and cancer, unless in specific cases, such as Hodgkin's lymphoma.

Keywords: Antidepressants, cancer, risk.

Introduction

There are controversial reports in the literature concerning the use of antidepressants (AD) and their association with cancer (CA). Some authors suggest an association between use of AD and increased risk of CA.^1–10 CA is a potentially fatal disease, and use of AD has increased over the years,^1,2 which makes this possible association an important theme for discussion. It is also worth stressing that AD help fight depression in patients with CA.^11–14

The authors suggesting that use of AD reduces risk of CA^4,11–18 are based on different arguments. Firstly, there is the idea that depression causes depletion of natural killer (NK) cells,¹⁹ and of anti–neoplastic surveillance cells,²⁰ increasing risk of CA development. By treating depression, NK cells would return to normal standards, reducing risk of CA. Another mechanism that supports this idea is the dose–reversible and voltage–dependent block of hEag1 channels caused by imipramine and astemizole. These channels were detected in many lineages of tumor cells, influencing their proliferation rate.^11,13 Another channel blocked by imipramine and desipramine is the SK3 present in TE671 cells of the human medulloblastoma, which also influences division of cancer cells.¹² There is a study indicating use of clomipramine as adjuvant in CA chemotherapy, since it enhances the potential of vincristine effect on leukemia refractory to this drug.¹⁶

Other authors claim that AD may increase risk of CA,^6,21 using as arguments the confirmed increase in incidence of colon CA in rats treated with desipramine, after intoxication by azoxymethane, in relation to controls;²¹ reversible reduction in NK cell activity after treatment with desmethylimipramine,^22,23 which, as previously mentioned, is an anti–tumor surveillance cell; and structural similarity between tricyclic AD (TAD), such as amitriptyline, and selective serotonin reuptake inhibitors (SSRI), such as fluoxetine, with growth regulator histamine receptor ligand, N, N–diethyl–2–[4–(phenylmethyl) phenoxy] ethanamide HCl, which stimulates tumor growth in vivo.²⁴

Due to the controversies found in the literature, this study aims at performing a review on the association of AD and stimulation of CA appearance and/or growth.

Method

A survey was carried out in the following databases: MEDLINE, Cochrane Database of Systematic Reviews and Controlled Trials Register, Update Software and LILACS. Search words were: antidepressant, fluoxetine, paroxetine, imipramine, sertraline, nortriptyline, venlafaxine, citalopram, escitalopram, fluvoxamine, mirtazapine, trazadone, nefazodone, milnacipran, duloxetine, maprotiline, amitriptyline, clomipramine and cancer. Only studies using primary data were included: cohort, case–control, cohort nested case–control and case reports using humans. Exclusion criteria were literature reviews, meta–analyses, studies using animals and those that approached AD in the treatment of depression resulting from CA.

Results

Considering the inclusion and exclusion criteria, 11 studies published between 1998 and 2005 were selected. Of selected studies, there was one with case reports, eight case–control studies and two cohort studies.

Three cases of male breast cancer²⁵ were associated with use of SSRI. SSRI used were fluoxetine at 20 mg/day for 1 to 3 years in two cases, and paroxetine at 20 mg/day for more than 5 years in one case. The authors suggest that assessing the activity of AD on male breast CA is easier to be interpreted than women's CA, due to lower number of risk factors that male gender presents in relation to breast CA. In addition, male breast suffers fewer physiological changes throughout life. In cases under investigation, however, the patients were exposed to other variables, probably implied in CA development, such as use of other drugs (for example, finasteride), excessive alcohol consumption and family history of CA.

Eight case–control studies were included; in six of them, there was no evidence of association between use of AD and CA development, and in two studies such association was suggested in specific situations.

Two studies analyzed risk of ovarian CA and use of AD. The first study analyzed risk of ovarian CA according to use and type of use (regular or irregular) of TAD, SSRI, phenothiazine drugs and benzodiazepines.⁸ Regular use was considered in 4 days a week for at least 4 weeks. Questionnaires responded by patients receiving care for 22 years in four health centers in important American metropolitan regions were analyzed, representing a sample of 748 cases of women with ovarian epithelial CA. For each case, two controls diagnosed with some type of CA and who were not using drugs and two controls with diagnoses other than CA and who were not using drugs either were selected. What stands out in the results is the value of odds ratios (OR) = 3.00 (CI95% 0.80–10.50) for regular use of SSRI in patients with CA, compared with controls with no diagnosis of CA. When correction for other variables, such as menarche, use of oral contraceptives, alcohol and tobacco consumption, was performed, OR fell to 0.70 (CI95% 0.10–1.10). Considering the OR obtained both for regular and irregular use of both AD classes, compared with both control groups, the OR is not different from 1.00, with lowest confidence interval of 0.20 and highest of 3.30. Therefore, results do not suggest association between AD and ovarian CA.

The second study was based on the analysis of a database maintained by a health surveillance agency over a 16–year period (1981–1997).³ A total of 314 cases of women with ovarian epithelial CA were included. They were paired with 790 controls, women between 35 and 79 years with sociodemographic characteristics similar to cases and who did not have CA. The use of varied AD classes was analyzed considering their prescription for a 6–month period and continuous use or not for more than 6 months. Data were obtained from an electronic database of drugstores linked to the health agency, which record information concerning number of prescriptions, dosages and posology of each drug acquired by the patient. For cases that had only one prescription for AD, over a 6–month period, OR was 1.1 (0.69–1.90), and for those with two prescriptions in 6 months, OR was 0.71 (0.47–1.10). In non–continuous users, for more than 6 months, OR was 0.94 (0.64–1.4), and for continuous users, for a period longer than 6 months, OR was 0.64 (0.36–1.10). The most used AD were doxepine, amitriptyline and imipramine. In this study, that fact that SSRI started being widely used only after the 1980's compromised the analysis of their long–term use.

One case–control study⁶ indicated positive association, but with no statistical significance, between AD and breast CA, with OR = 2.20 (0.8–6.30) for use of SSRI for 36 months or more. This study was based on a database, previously setup, of a case–control study on breast CA in North Carolina, USA, and the information about use of AD was provided by the patients. A total of 1,445 cases of women with diagnosis of breast CA and who were making continuous use of AD for 3 months or more before the diagnosis were selected, and then paired with 1,226 controls, women with breast CA and who were not using AD before the diagnosis. Continuous use of AD was reported in 1/5 of patients in both cases and controls (after diagnosis of breast CA), and there was no association between increased risk of CA and continuous use of AD (OR = 1.00; CI95% 0.7–1.20). When separated by AD class (TAD, SSRI, atypical, lithium or multiple types), all OR values were lower than 1.10, and CI ranged between 0.50 and 1.90.

Another study involving use of AD and breast CA, carried out in Ontario, Canada, analyzed cases of CA included in the Ontario Cancer Registry and use of AD, considering sociodemographic, reproductive, anthropometric, clinical and psychiatric history variables, lifestyle and use of other drugs.⁷ A total of 3,133 cases (women with primary diagnosis of breast CA, aged between 25–74 years at diagnosis) and 3,062 controls (randomly selected among a female population from Ontario) were selected. Analyzing use of all AD classes and total prevalence of CA, there was weak association, with OR = 1.17 (1.01–1.36), but when this value was corrected considering other risk factors for CA, the result was not statistically significant. Use of sertraline and paroxetine alone suggests a weak increase in risk of breast CA, with OR = 1.58 (1.03–2.41) and OR = 1.55 (1.00–2.40), respectively. However, in multivariable analysis, this result loses its significance, with OR = 1.45 (0.88–2.40) for sertraline and OR = 1.60 (0.93–2.77) for paroxetine. These results show the multi–influence existing in development of breast CA.

In opposition to the findings of previous articles, a study carried out including women aged 35 years or more, in Saskatchewan, also in Canada, suggested positive association between prolonged use of TAD (11 to 15 years) and risk of breast CA.²⁶ From the databases of the Saskatchewan Cancer Agency and Saskatchewan Prescription Drug Plan, 5,882 cases of women with CA and 23,517 controls were selected, randomly chosen among the population with no history of CA, with mean of four controls for each case. From the same database, information was obtained about AD prescription, class of prescribed drug, number of pills and drug dosage in each pill, but there was no information concerning daily dose and treatment duration in each prescription. Taking into account all TAD, the study found increased risk of breast CA development in patients taking those drugs for periods between 11 and 15 years, with relative risk (RR) = 2.02 (1.34–3.04). Based on previous studies on TAD genotoxicity in Drosophila melanogaster, the author divided TAD into two groups: genotoxic (amoxapine, clomipramine, desipramine, doxepine, imipramine and trimipramine) and non–genotoxic (amitriptyline, maprotiline, nortriptyline and protriptyline), to assess which drugs were associated with risk of breast CA development. Over a period of use between 11 and 15 years, genotoxic TAD caused an increase in risk of CA development, with RR = 2.47 (1.38–4.40), whereas the group of non–genotoxic was not related to increase in this risk, with RR = 1.92 (0.93–3.95).

A study carried out by the Food and Drug Administration (USA) aimed at determining the risk of tumor recurrence and appearance of a second primary tumor in patients taking AD.¹⁰ The sample was taken from a database that records diagnosed cases of tumor, its classification, stage, treatment and follow–up. Patients included in that database have benefits as to prescription of drugs, so that the authors used the information from the drugstore of this health program to select those that used AD. A total of 1,467 patients diagnosed with breast and colon CA or malignant melanoma were selected. Such neoplasias were selected due to their prevalence, patients' survival and possibility of recurrence. For each case, a control included in the database and with the same characteristics of the cases was selected. Of all patients with CA, 206 used AD, and only 20 (4.2%) had recurrence and 17 (6.5%) had a second primary CA, a relatively small sample when compared with the initial number of patients with CA. Based on OR values found – relation between AD and CA recurrences, with OR = 0.97 (0.52–1.78), and use of AD and occurrence of a second primary CA, with OR = 0.94 (0.51–1.77) –, the study suggests that there is no relationship between AD and risk of recurrence and appearance of a second primary tumor over a period shorter than 6 months.

In the study carried out by Cancer Care Ontario, women aged between 25 and 74 years with diagnosis of primary breast CA were selected between 1995 and 1996 from the Ontario Cancer Registry.²⁷ Controls were randomly chosen among women aged between 25 and 74 years from the database of the Ontario Ministry of Finance. Self–applicable questionnaires were sent to cases and controls, assessing: 1) sociodemographic data; 2) duration, dosage and type of AD; and 3) potential confounding factors. Women taking AD for 2 weeks or less or who were taking AD during the 6 months before diagnosis of CA were excluded from the survey. There was a high proportion of family history and fat consumption in cases, compared with controls. Use of TAD, only related to age, had weak association with risk of breast CA (OR = 1.2; CI95% 0.8–1.8), but with no statistical significance. Relating age and OR of varied factors, estimated for use of TAD for 25 months or more, there was increased risk of breast CA (OR = 2.5; CI95% 1.2–5.1; multivariate odds ratio estimate (MVOR) = 2.1; CI95% 0.9–5.0). Paroxetine was associated with increased risk of breast CA after clinical control of depression (MVOR = 7.2; CI95% 0.9–58.3). Use of other SSRI–type AD, such as amitriptyline and fluoxetine, for more than 2 years, was not associated with increased risk of breast CA (MVOR = 0.7; CI95% 0.2–2.2).

An important retrospective cohort study, carried out in a region in Denmark including 9% of the population, used information obtained from a database of the Danish Health Service, which gathers epidemiological information on pharmaceutical prescriptions crossed with information from the National Cancer Center.⁹ A population of 30,807 individuals taking AD was selected, with frequency of CA cases, compared with expected frequency according to variables such as gender, age and time of year. The dose/response relationship in the association of AD and CA was also assessed; for that, the sample was divided into groups classified as occasional, periodic or regular users. During the study, there were 766 cases of CA, compared with 746 expected cases, resulting in standardized incidence ratio (SIR) = 1.00 (1.00–1.10) and refuting the hypothesis of a relationship between AD and risk of CA. However, there was positive association between regular use of TAD (five or more prescriptions) and occurrence of non–Hodgkin's lymphoma, with SIR = 2.50 (1.40–4.20). Therefore, the authors suspected that use of TAD could be related to occurrence of non–Hodgkin's lymphoma or, based on the literature, that depression could be associated with immunosuppression and consequent development of CA.

Haque et al.⁵ analyzed 635 patients with CA and users of AD with the aim of establishing the risk of tumor growth by using paroxetine and other AD. Results indicate that use of paroxetine in relation to non–users of AD is responsible for an increase that is not statistically significant of risk of breast CA in women, with RR = 1.12 (0.96–1.31), but there was no association with duration of use. Exclusive use of paroxetine in relation to other AD revealed RR = 1.16 (0.98–1.38), presenting weak association that does not reach statistical significance. The same study did not find association between combined use of paroxetine and other AD, RR = 1.01 (0.67–1.51).

A cohort nested case–control study¹ analyzed the association between AD and CA, specifically TAD and SSRI and breast CA. The search was performed using the General Practice Research Database. It is an electronic database containing information about patients (socioeconomic data, clinic visits, hospitalizations, medical records, examinations and prescriptions), collected by physicians from the whole UK, and which is especially used to perform scientific papers. A total of 3,708 patients were found as sample. As a result, there was no relationship between use of AD (both TAD and SSRI) and risk of breast CA, with OR = 0.98 (0.81–1.19) and OR = 0.86 (0.73–1.00), respectively.

Discussion

Use of AD and risk of CA development comprehend some important aspects. One of them is the association of these drugs with increased risk of CA, since it has clinical implications, due to extensive use of AD and to lack of public knowledge concerning this risk. Which AD are involved, types of CA, most affected age group and time of drug exposure needed to CA development should be also considered. It is important to stress the need of searching for concrete evidence of the absence of other risk factors associated with neoplasia in individuals taking AD drugs, so that their role in promoting CA can be inferred.

The case report by Wallace et al.²⁵ suggests the association between male breast CA and use of SSRI. Such hypothesis is based on the idea that this type of effect is easier to be interpreted in men than in women, as male breast suffers fewer complex physiological changes than female breast and, for that reason, it presents fewer risk factors for CA development. However, of all reported cases, there was presence of other risk factors, such as use of finasteride, alcohol and family history. The fact that there are few studies on use of SSRI for long periods justifies the selection of that article. Nevertheless, the analysis of only three cases reduces the possibility of generalizing these data.

In a case–control study,¹⁰ there was no positive association between use of AD drugs and increased incidence of recurrent CA or secondary tumors. That study, however, had some limitations, such as small number of patients in studied groups, lack of correlation between cases of CA and a specific AD and arbitrary definition to establish the time that distinguishes a recurrent tumor from a treatment–resistant primary tumor.

Cotterchio et al.²⁷ obtained results that do not corroborate the hypothesis that use of any AD is always associated with risk of breast CA. That study suggests that use of TAD and paroxetine for more than 2 years could be associated with increased risk of CA development, but results did not reach statistical significance in the multivariate analysis. The authors believe that there has been no evidence so far to support a shift in frequent use of AD, but warn to the fact that there might be unknown effects with prolonged use.

Coogan et al.⁸ concluded that there is no increased risk of ovarian epithelial CA after using AD, TAD and SSRI. That study, however, was carried out using a small number of regular AD users, thus reducing the possibility of assessing risk according to duration, time of treatment onset and aging with drug use. In addition, there were few SSRI users, since this class of drugs was launched in the market only in the 1980's, being widely used after the 1990's. Another limitation is that information about patients, previous to diagnosis, concerning use of AD was obtained from questionnaires applied by nurses. Therefore, the information obtained as to type of drug used, time of use and dosage is directly dependent on patients' memory, and may create a risk of bias.

Dublin et al.³ suggest that there is no consistent evidence supporting the association between increased risk of ovarian epithelial CA in women and use of some types of AD. The fact that all information collected on the characteristics of using AD was obtained from drugstore data does not dismiss the possibility of patients having obtained the drugs in other drugstores that were not registered and does not assure that patients adhered to the treatment or are appropriately taking the drugs.

Sharpe et al.²⁶ associate the use of genotoxic TAD (amoxapine, clomipramine, desipramine, doxepine, imipramine and trimipramine) with increased risk of CA, which is not suggested with use of non–genotoxic TAD (amitriptyline, maprotiline, nortriptyline and protriptyline). This idea is supported after a period of use longer than 10 years. The study limitation, however, consists of the absence of control of other possible carcinogenic agents other than AD.

The case–control study by Moorman et al.⁶ did not indicate increased risk of breast CA development in AD users. The only result found with OR higher than unit was not statistically significant. However, the study presented an unexpected result, even for the authors, as to in situ breast carcinoma. Women taking AD had less chance of developing this type of CA, when compared with the control group. In this case, drugs would have probably acted as a protection factor against CA. Its advantage is the large sample of 1,445 cases, divided into in situ carcinoma and breast invasive.

Streingart et al.⁷ studied the association of breast CA with use of paroxetine and sertraline using a model adjusted for age. Results were not statistically significant in multivariate analyses.

The findings in the study by Gónzalez–Pérez et al.¹ do not corroborate the hypothesis that use of AD increases risk of development of breast CA. This result was found even when analyzing different variables, such as duration of drug use, daily dose or specific type of drug. Since each physician performed their own data collection, some information could have been omitted. The data were not collected for the study, since they were already in medical records.

In the cohort study carried out by Dalton et al.,⁹ there was correlation between non–Hodgkin lymphoma and use of AD. That study used data from an integrated system in Denmark, in which each individual has a national identification number, and all drugstores in the region under investigation are connected to the same database, reducing losses. Another relevant factor is the use of all available AD classes. It is believed that non–Hodgkin's lymphoma is strongly associated with immunosuppressed transplanted patients and with patients with AIDS, but none of these cases can be applied to that study, since immunosuppression drugs were not administered for the individuals. The duration/response relationship with AD for non–Hodgkin's lymphoma was consistent in two interpretations: in the first one, the fact that depression is associated with immunodepression is important, and patients with depression and severely ill were given too many prescriptions for AD; the second interpretation involves individuals with long–term treatment using TAD and with positive etiology for CA, but who do not present depression. However, restricting interpretation of results, that study does not have information about the daily AD dosage used by each individual neither why patients used these drugs.

The study recently carried out by Haque et al.⁵ suggests that there is no difference between increased risk of breast CA in women who used paroxetine and those who used other AD. In addition, women who used paroxetine for more than 2 years did not present association with CA, when compared with those who used that drug for a shorter period of time. However, this analysis was not adjusted by an analytical comparison with non–users of AD or with information associated with other risk factors for breast CA, such as pregnancy and family history, neither considered daily dosage used by each patient, which could lead to reduced OR values.

Most analyzed studies did not find positive association between use of AD and CA, or the positive association did not reach statistical significance. Case reports suggest an association between male breast CA and use of AD. Most case–control studies did not find association between use of AD and occurrence of varied types of CA. Cohort studies were divergent concerning results about the association between varied types of CA and use of AD, since one of them found relationship between non–Hodgkin's lymphoma and TAD, whereas breast CA is based on the same result obtained by case–control studies, which did not demonstrate such association.

In conclusion, this review of the literature indicates conflicting results as to the effects of AD in promoting, causing or inhibiting the development of varied types of CA. It is possible that AD have an adverse effect on the growth of cancerous tumors only in the presence of other carcinogenic factors. New studies in this area are needed to better define the action of each AD as a risk or protection factor in relation to varied forms of CA.

References

Received January 16, 2007.

Accepted March 21, 2007.

• ¹
  Gonzalez–Perez A, Garcia–Rodriguez LA. Breast cancer risk among users of antidepressant medications. Epidemiology. 2005;16(1):101–5.
• ²
  Lawlor DA, Juni P, Ebrahim S, Egger M. Systematic review of the epidemiologic and trial evidence of an association between antidepressant medication and breast cancer. J Clin Epidemiol. 2003;56(2):155–63.
• ³
  Dublin S, Rossing MA, Heckbert S, Goff BA, Weiss NS. Risk of epithelial ovarian cancer in relation to use of antidepressants, benzodiazepines, and other centrally acting medications. Cancer Causes Control. 2002;13(1):35–45.
• ⁴
  Abdul M, Logothetis CJ, Hoosein NM. Growth–inhibitory effects of serotonin uptake inhibitors on human prostate carcinoma cell lines. J Urol. 1995;154(1):247–50.
• ⁵
  Haque R, Enger SM, Chen W, Petitti DB. Breast cancer risk in a large cohort of female antidepressant medication users. Cancer Lett. 2005;221(1):61–5.
• ⁶
  Moorman PG, Grubber JM, Millikan RC, Newman B. Antidepressant medications and their association with invasive breast cancer and carcinoma in situ of the breast. Epidemiology. 2003;14(3):307–14.
• ⁷
  Steingart A, Cotterchio M, Kreiger N, Sloan M. Antidepressant medication and breast cancer risk: a case control study. Int J Epidemiol. 2003;32(6):961–6.
• ⁸
  Coogan PF, Rosenberg L, Palmer JR, Strom BL, Stolley PD, Zauber AG, et al. Risk of ovarian cancer according to use of antidepressants, phenothiazines and benzodiazepines (United States). Cancer Causes Control. 2000;11(9):839–45.
• ⁹
  Dalton S, Johansen C, Mellemkjær L, Sørensen H, McLaughlin J, Olsen J, et al. Antidepressant medications and risk for cancer. Epidemiology. 2000;11(2):171–6.
• ¹⁰
  Weiss SR, McFarland BH, Burkhart GA, Ho PT. Cancer recurrences and secondary primary cancers after use of antihistamines or antidepressants. Clin Pharmacol Ther. 1998;63(5):594–9.
• ¹¹
  Gavrilova–Ruch O, Schonherr K, Gessner G, Schonherr R, Klapperstuck T, Wohlrab W, et al. Effects of imipramine on ion channels and proliferation of IGR1 melanoma cells. J Membr Biol. 2002;188(2):137–49.
• ¹²
  Carignani C, Corsi M. Inhibition of SK3 channels in the TE671 human medulloblastoma cell line by desipramine and imipramine. Eur J Pharmacol. 2002;448(2–3):139–42.
• ¹³
  Garcia–Ferreiro RE, Kerschensteiner D, Major F, Monje F, Stuhmer W, Pardo LA. Mechanism of block of hEag1 K+ channels by imipramine and astemizole. J Gen Physiol. 2004;124(4):301–17.
• ¹⁴
  Daley E, Wilkie D, Loesch A, Hargreaves IP, Kendall DA, Pilkington GJ, et al. Chlorimipramine: a novel anticancer agent with a mitochondrial target. Biochem Biophys Res Commun. 2005;328(2):623–32.
• ¹⁵
  Xia Z, Bergstrand A, DePierre JW, Nassberger L. The antidepressants imipramine, clomipramine, and citalopram induce apoptosis in human acute myeloid leukemia HL–60 cells via caspase–3 activation. J Biochem Mol Toxicol. 1999;13(6):338–47.
• ¹⁶
  Tsuruo T, Iida H, Nojiri M, Tsukagoshi S, Sakurai Y. Potentiation of chemotherapeutic effect of vincristine in vincristine resistant tumor bearing mice by calmodulin inhibitor clomipramine. J Pharmacobiodyn. 1983;6(2):145–7.
• ¹⁷
  Sauter C. Cytostatic activity of commonly used tricyclic antidepressants. Oncology. 1989;46(3):155–7.
• ¹⁸
  Tutton PJ, Barkla DH. Influence of inhibitors of serotonin uptake on intestinal epithelium and colorectal carcinomas. Br J Cancer. 1982;46(2):260–5.
• ¹⁹
  Frank MG, Hendricks SE, Burke WJ, Johnson DR. Clinical response augments NK cell activity independent of treatment modality: a randomized double–blind placebo controlled antidepressant trial. Psychol Med. 2004;34(3):491–8.
• ²⁰
  Whiteside TL, Herberman RB. The role of natural killer cells in immune surveillance of cancer. Curr Opin Immunol. 1995;7(5) 704–10.
• ²¹
  Iishi H, Tatsuta M, Baba M, Taniguchi H. Enhancement by the tricyclic antidepressant, desipramine, of experimental carcinogenesis in rat colon induced by azoxymethane. Carcinogenesis. 1993;14(9):1837–40.
• ²²
  Miller AH, Asnis GM, van Praag HM, Norin AJ. Influence of desmethylimipramine on natural killer cell activity. Psychiatry Res. 1986;19(1):9–15.
• ²³
  Eisen JN, Irwin J, Quay J, Livnat S. The effect of antidepressants on immune function in mice. Biol Psychiatry. 1989;26(8):805–17.
• ²⁴
  Brandes LJ, Arron RJ, Bogdanovic RP, Tong J, Zaborniak CL, Hogg GR, et al. Stimulation of malignant growth in rodents by antidepressant drugs at clinically relevant doses. Cancer Res. 1992;52(13):3796–800.
• ²⁵
  Wallace WA, Balsitis M, Harrison BJ. Male breast neoplasia in association with selective serotonin re–uptake inhibitor therapy: a report of three cases. Eur J Surg Oncol. 2001;27(4):429–31.
• ²⁶
  Sharpe CR, Collet JP, Belzile E, Hanley JA, Boivin JF. The effects of tricyclic antidepressants on breast cancer risk. Br J Cancer. 2002;86(1):92–7.
• ²⁷
  Cotterchio M, Kreiger N, Darlington G, Steingart A. Antidepressant medication use and breast cancer risk. Am J Epidemiol. 2000;151(10):951–7.

Publication Dates

• Publication in this collection
  06 Sept 2007
• Date of issue
  Apr 2007

History

• Accepted
  21 Mar 2007
• Received
  16 Jan 2007