|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 26, No 14 (May 10), 2008: pp. 2305-2310 © 2008 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.15.9681 Use of Antidepressants and Risk of Cancer in Individuals Infected With HIV
From the Imperial College School of Medicine, Departments of Oncology, Imperial Healthcare National Health Service Trust; St Bartholomew's Hospital, Department of Oncology; St Stephen's Centre, Chelsea and Westminster Hospital; Imperial College School of Medicine, Department of Oncology and HIV Medicine, Chelsea and Westminster Hospital, London, United Kingdom Corresponding author: J. Stebbing, MA, MRCP, MRCPath, PhD, Imperial College, Imperial Healthcare NHS Trust, Charing Cross Hospital, Dept Medical Oncology, 1st Floor, E Wing, Fulham Palace Rd, London W6 8RF, UK; e-mail: j.stebbing{at}ic.ac.uk
Purpose Preclinical and cohort studies suggest that certain antidepressants are associated with a predisposition to cancer whereas others decrease the risk. We aimed to assess whether different classes of antidepressants were associated with changes in cancer incidence in a population of HIV-1 infected individuals, based on duration of exposure. Methods Antidepressant exposure was measured from date of first prescription of the antidepressant until the date of last follow-up or cancer diagnosis. Univariate and multivariate analyses were performed to establish the risk of AIDS-related cancers and non–AIDS-related cancers according to whether patients were receiving selective serotonin reuptake inhibitors, tricyclic antidepressants, or other medicines for depression. We analyzed data for time exposed to antidepressants, before and during the era of highly active antiretroviral therapy (HAART). Results From a cohort of 10,997 patients representing 52,656 years of follow-up attending a large HIV center during the pre-HAART and HAART eras, a total of 2,004 (18%) were prescribed antidepressants representing 15,850 years exposed. A total of 1,607 (15%) individuals were diagnosed with cancer. There were no significant associations between any class of antidepressant and any type of cancer (P = .19), in either the pre-HAART or HAART era (P = .23), and use of serotonin reuptake inhibitors did not alter the risk of Burkitt lymphoma. Conclusion Antidepressants, irrespective of their class, do not affect cancer risk in HIV-infected individuals.
Data regarding the possible carcinogenic effect of antidepressants remain inconclusive. Some epidemiologic studies have recorded an increased risk of cancer with use of antidepressants,1-3 whereas others have shown no association.4-10 Animal studies have suggested that tricyclic antidepressants (TCAs) enhance proliferation in the intestinal epithelium,11,12 and cause genotoxicity during Drosophila wing development.13 One retrospective paper has shown an increased incidence ratio for breast cancer with use of TCAs,3 data that were not supported in subsequent studies of either breast or colorectal cancer.14 Serotonin reuptake inhibitors (SSRIs), now the most frequently prescribed antidepressants,15 have been shown in one study to have malignant growth promoting effects,16 and in others they slow the growth of colorectal carcinoma xenografts and suppress carcinogenesis in chemically induced colon tumors.17-19 These data are consistent with findings that serotonin promotes cellular proliferation20 and the SSRI drug fluoxetine decreased the viability of human colon cells, possibly through a direct cytotoxic effect.21 Importantly for this study, the serotonin transporter (SERT) was found to be an unexpected therapeutic candidate in Burkitt lymphoma.22 Burkitt lymphoma cell lines carry immunoreactive SERT and transport serotonin with first-order kinetics.23 Moreover, a capacity for serotonin to drive rapid and extensive apoptosis in biopsy-like Burkitt lymphoma cells was reversed by SSRIs that act by blocking serotonin uptake into SERT-carrying cells. Most surprising was that SSRIs themselves provoked apoptotic death in biopsy-like Burkitt lymphoma cells.24 Supporting the potential antitumor effects of SSRIs, a study has recently shown that SSRIs increase natural killer–mediated innate cytolytic immunity in samples derived from HIV-infected individuals.25 We therefore investigated this further in a large clinical cohort infected with HIV in which medication history, virologic parameters, and other illnesses were prospectively recorded. Despite extensive data demonstrating that HIV infection and associated immunosuppression predisposes individuals to a wide range of cancers26-28 (including non–AIDS-related malignancies29), no studies have specifically investigated the association between antidepressant use, length of antidepressant exposure, and the development of both AIDS-related and non–AIDS-related cancers in the highly active antiretroviral therapy (HAART) and pre-HAART eras.
The Chelsea and Westminster HIV cohort in London, United Kingdom, is one of the largest single center cohorts in Europe and we prospectively collect routine data on the individuals who attend which is entered into computerised databases. HIV positive patients are seen at regular intervals for clinical assessment, trial follow-up, and immunologic assessments. All HIV patients who have attended the Chelsea and Westminster since routine prospective data collection commenced in 1986 were identified, and we have defined HAART as therapy consisting of at least three antiretroviral drugs in accordance with published guidelines (dual nucleoside analogs alone are not considered HAART, though they are considered backbones of therapy). This study compares cancer incidence in the pre-HAART and the HAART era, which we have defined as commencing on January 1, 1996, when HAART became routinely available at our institution and many others. We also examined risk of AIDS-related cancers (Kaposi sarcoma, high grade B-cell non-Hodgkin's lymphoma including Burkitt lymphoma, and invasive cervical cancer) and non–AIDS-related cancers (all other tumors). Appropriate ethical approval was obtained specifically for this study (REC number 07/Q0411/3). We obtained exposure to antidepressants using our prospectively recorded computerised cohort data collected through the hospital pharmacy. Individual antidepressants prescribed were studied separately and in groups as defined (Table 1). Each occasion a prescription was given to a patient, this information was entered into the database. First exposure to antidepressants was defined as the first prescription of an antidepressant drug after HIV positivity was diagnosed. We have analyzed both exposure and response variables, in those who were exposed to different classes of antidepressants.
Person-days of follow-up were converted to person-years at risk (PYAR); this was not categorised. PYAR was estimated from entry into the cohort to the last available observation, the development of cancer, first exposure to an antidepressant or death. PYAR for antidepressant exposed individuals started from the first date of antidepressant prescription while time in the cohort before antidepressant prescription for these individuals was included in the unexposed PYAR (the none category in the tables). In order to keep the coefficient of the PYAR constant, this was log transformed and used as the offset in the Poisson regression (in a log linear model as an extension of Cox's model). These data were analyzed using the Genmod procedure in SAS (Version 9.1, 2004; SAS Institute, Cary, NC) with loge link and Poisson error distributions. This fits generalized linear models allowing time dependent measures of probability, in order to incorporate time exposed to antidepressants. Case control analyses were not used as the primary consideration; we utilized duration of exposure. Median and interquartile ranges were used to create unbiased categoric data. A separate category was created for all variables with missing data which ensured no degrees of freedom were lost when building multivariate models. Log linear models with single variables were initially used to estimate rate ratio of cancers intrinsically corrected for the length of time exposed to antidepressants. All variables found to approach significance (P < .2) in univariate log linear model were then used to build a multivariate model which allowed the risk of a particular prognostic variable to be assessed while controlling for the others in the model. The final multivariate model presented was tested for its distributional assumptions using Cox Snell residual plots and adjusted for possible confounding or residual effects. As the missing categories comprised less than three percent of the data and therefore had wide confidence intervals the presented graphs exclude point estimates of missing data categories. All P values presented are two-sided.
We investigated a cohort of 10,997 patients attending a large HIV center in the pre-HAART and HAART eras representing 52,656 years of follow-up data. The antidepressants and doses of medications in each category (SSRI, TCA, and other), are presented in Table 1. Similar numbers of patients were exposed to the SSRIs (n = 952) and TCAs (n = 919), which avoids statistical instability. For this reason, the ‘other' group was not further broken down into its representative components (low-dose thioxanthenes,  -2 antagonists, and TCA-related drugs). For SSRIs, person-days at risk measured 2,577,785 (7,062 years); for TCAs, person-days at risk were 2,803,010 (7,679 years); and the other category, exposure was 404,458 (1,108 years) days at risk. The total follow-up for the cohort not exposed to antidepressants measured 13,434,022 days (36,806 years). A total of 144 cases (9%) were prescribed an antidepressant at least 3 months before the cancer diagnosis. We wished to study which antidepressants were more likely associated with a particular cancer based on time exposed. Table 2 demonstrates the univariate log linear regression model showing the rate ratio of HIV-related cancer incidence in patients who were exposed to an antidepressant at any time since 1986. We have included the CD4 count data in this table as an internal control for these data, as we (and others) have shown previously using this cohort that lower CD4 counts are associated with an increased cancer incidence. This is shown graphically in Figure 1, demonstrating the rate ratios using SSRI exposure as the reference category. For both Table 2 and Figure 1 the lower panel shows cancer incidence since January first 1996, the HAART era, with each antidepressant class. For the entire cohort, there were no statistically significant univariate associations with antidepressant use (P = .191); data that were maintained when the HAART era was examined in isolation (P = .234). There were also no significant effects when individual antidepressants were examined, when specific TCAs or SSRIs were investigated or when concomitant HAART was protease inhibitor–or nonnucleoside reverse transcriptase inhibitor–based. As expected, when these variables were entered into a multivariate model that adjusted for all other factors, only a lower CD4 count was found to predispose individuals to an increased cancer risk.
To investigate this further, we specifically evaluated the effect of SSRIs on high grade B-cell non-Hodgkin's lymphoma, an AIDS-related cancer. A total of 265 cases of non-Hodgkin's lymphoma have been observed, and of these, 34 (13%) were exposed to an antidepressant before diagnosis. Similarly, 1,177 cases of Kaposi sarcoma, the most common AIDS-related cancer, have been diagnosed, and of these, 78 (7%) were exposed to an antidepressant. Table 3 demonstrates a univariate analysis for the entire cohort and the HAART era, examining the effect of antidepressant exposure on risk of non-Hodgkin's lymphoma, Kaposi sarcoma, and non–AIDS-related cancers.
Next, we examined primary CNS lymphomas (n = 65; 16 diagnosed post-antidepressant exposure) and Burkitt lymphoma (n = 29; 3 diagnosed post-antidepressant exposure) and again no significant associations were observed (P = .21 and P = unstable due to small numbers). In multivariate analyses only the CD4 count remained significant.
To our knowledge, this is the first study that assesses the effects of antidepressants (SSRI, TCA, and other) on cancer risk in HIV-positive individuals with a clear a priori hypothesis based on duration of exposure. We found that in HIV-infected individuals who are predisposed to a wide range of cancers, different antidepressants are not associated with a change in cancer risk, in either the pre-HAART or during the HAART eras. Despite extensive in vitro and animal data with different antidepressant classes, SSRIs were not associated with a decreased risk of any cancer including Burkitt lymphoma, and TCA use did not increase the hazard ratio for any cancer. Furthermore, we noticed no significant relationship between different antidepressant classes, AIDS-related versus non–AIDS-related cancers, or whether the patients were followed up in the pre-HAART and/or HAART era, based on the duration of exposure to different antidepressants. Selection bias and recall bias were unlikely in this study as outcomes and length of exposure were prospectively recorded. Detection bias could have occurred because individuals who take antidepressants have more opportunities to see a doctor, and thus have more potential to have cancer detected. However, this possibility would lead to an underestimation of a true SSRI antitumour effect and overestimation of a true carcinogenic effect for TCAs. Other limitations include other possible causes of cancer including lifestyle, obesity, diet, and smoking, which are not prospectively recorded, and antidepressant use itself may be associated with such factors. Confounding by such factors would have had different effects for SSRI and TCA analyses. As duration of exposure was fundamental to our analysis here, we were unable to compare exposure versus no exposure because this would have precluded calculations of incidence based on duration exposed. However, the ‘other' group of antidepressants is comprised of medications known to treat depression but without any preclinical or clinical data of an anti- or pro-tumorigenic effect. Thus, the ‘other' category can be considered as a control for the SSRI and TCA groups. As well as improving symptoms of depression, two studies demonstrate that use of antidepressants in HIV positive individuals is associated with improved antiretroviral adherence.30,31 In both studies, after antidepressant therapy for 6 months, adherence to antiretroviral regimes was statistically higher in HIV-depressed patients on treatment than in HIV-depressed patients not treated with antidepressants. Thus, we suggest that use of antidepressants represent an additional strategy to manage nonadherence to HAART, although longer term and larger studies are warranted. We have previously described a small uncontrolled data set in which 32 cancers occurred following a prescription for the sleeping tablet zopiclone in 606 HIV-1 infected individuals.32 The label for eszopiclone contains notable warnings regarding carcinogenicity and mutagenesis, and this medication was approved by the US Food and Drug Administration following the withdrawal of rofecoxib (Vioxx) and antegren (Tysabri) from the market in North America.33,34 We believe that such potential carcinogenesis assumes increased importance when the drug is chronically used in the primary care setting (thus potentially exposing millions of individuals), and when patients at increased risk of cancer are exposed. Our data do not lend support to the hypothesis that TCAs increase risk of certain cancers and SSRIs decrease this risk. Future studies should have a more complete assessment of confounders including smoking and alcohol consumption, physical activity and recreational drug use. We suggest that cohorts such as this represent an ideal setting for clinico-epidemiologic studies of potential carcinogenicity.
The author(s) indicated no potential conflicts of interest.
Conception and design: Justin Stebbing Financial support: Justin Stebbing, Tom Powles, Sundhiya Mandalia, Mark Bower Administrative support: Justin Stebbing, Sundhiya Mandalia, Mark Nelson, Brian Gazzard Provision of study materials or patients: Justin Stebbing, Tom Powles, Mark Nelson, Brian Gazzard Collection and assembly of data: Justin Stebbing, Tom Powles, Sundhiya Mandalia, Mark Nelson, Brian Gazzard, Mark Bower Data analysis and interpretation: Justin Stebbing, Mark Bower Manuscript writing: Justin Stebbing, Sundhiya Mandalia, Mark Nelson, Brian Gazzard Final approval of manuscript: Justin Stebbing, Mark Bower
We thank Deborah K. Zeitin for help with categorization of antidepressants and Rak Patel for help with statistics.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Dalton SO, Johansen C, Mellemkjaer L, et al: Antidepressant medications and risk for cancer. Epidemiology 11:171-176, 2000[CrossRef][Medline] 2. Harlow BL, Cramer DW, Baron JA, et al: Psychotropic medication use and risk of epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev 7:697-702, 1998[Abstract] 3. Sharpe CR, Collet JP, Belzile E, et al: The effects of tricyclic antidepressants on breast cancer risk. Br J Cancer 86:92-97, 2002[CrossRef][Medline] 4. Moorman PG, Berchuck A, Calingaert B, et al: Antidepressant medication use [corrected] and risk of ovarian cancer. Obstet Gynecol 105:725-730, 2005[Medline] 5. Moorman PG, Grubber JM, Millikan RC, et al: Antidepressant medications and their association with invasive breast cancer and carcinoma in situ of the breast. Epidemiology 14:307-314, 2003[CrossRef][Medline] 6. Dublin S, Rossing MA, Heckbert SR, et al: Risk of epithelial ovarian cancer in relation to use of antidepressants, benzodiazepines, and other centrally acting medications. Cancer Causes Control 13:35-45, 2002[CrossRef][Medline] 7. Coogan PF, Rosenberg L, Palmer JR, et al: Risk of ovarian cancer according to use of antidepressants, phenothiazines, and benzodiazepines (United States). Cancer Causes Control 11:839-845, 2000[CrossRef] 8. Wang PS, Walker AM, Tsuang MT, et al: Antidepressant use and the risk of breast cancer: A non-association. J Clin Epidemiol 54:728-734, 2001[CrossRef][Medline] 9. Cotterchio M, Kreiger N, Darlington G, et al: Antidepressant medication use and breast cancer risk. Am J Epidemiol 151:951-957, 2000 10. Kelly JP, Rosenberg L, Palmer JR, et al: Risk of breast cancer according to use of antidepressants, phenothiazines, and antihistamines. Am J Epidemiol 150:861-868, 1999 11. Tutton PJ, Barkla DH: Effect of an inhibitor of noradrenaline uptake, desipramine, on cell proliferation in the intestinal crypt epithelium. Virchows Arch B Cell Pathol Incl Mol Pathol 57:349-352, 1989[Medline] 12. Iishi H, Tatsuta M, Baba M, et al: Enhancement by the tricyclic antidepressant, desipramine, of experimental carcinogenesis in rat colon induced by azoxymethane. Carcinogenesis 14:1837-1840, 1993 13. van Schaik N, Graf U: Structure-activity relationships of tricyclic antidepressants and related compounds in the wing somatic mutation and recombination test of Drosophila melanogaster. Mutat Res 286:155-163, 1993[Medline] 14. Xu W, Tamim H, Shapiro S, et al: Use of antidepressants and risk of colorectal cancer: A nested case-control study. Lancet Oncol 7:301-308, 2006[CrossRef][Medline] 15. Fisch M: Treatment of depression in cancer. J Natl Cancer Inst Monogr 2004:105-111, 2004 16. Brandes LJ, Arron RJ, Bogdanovic RP, et al: Stimulation of malignant growth in rodents by antidepressant drugs at clinically relevant doses. Cancer Res 52:3796-3800, 1992 17. Tutton PJ, Barkla DH: Influence of inhibitors of serotonin uptake on intestinal epithelium and colorectal carcinomas. Br J Cancer 46:260-265, 1982[Medline] 18. Tutton PJ, Steel GG: Influence of biogenic amines on the growth of xenografted human colorectal carcinomas. Br J Cancer 40:743-749, 1979[Medline] 19. Tutton PJ, Barkla DH: The influence of serotonin on the mitotic rate in the colonic crypt epithelium and in colonic adenocarcinoma in rats. Clin Exp Pharmacol Physiol 5:91-94, 1978[CrossRef][Medline] 20. Seretis E, Gavrill A, Agnantis N, et al: Comparative study of serotonin and bombesin in adenocarcinomas and neuroendocrine tumors of the colon. Ultrastruct Pathol 25:445-454, 2001[CrossRef][Medline] 21. Yue CT, Liu YL: Fluoxetine increases extracellular levels of 3-methoxy-4-hydroxyphenylglycol in cultured COLO320 DM cells. Cell Biochem Funct 23:109-114, 2005[CrossRef][Medline] 22. Serafeim A, Grafton G, Chamba A, et al: 5-Hydroxytryptamine drives apoptosis in biopsylike Burkitt lymphoma cells: Reversal by selective serotonin reuptake inhibitors. Blood 99:2545-2553, 2002 23. Meredith EJ, Holder MJ, Chamba A, et al: The serotonin transporter (SLC6A4) is present in B-cell clones of diverse malignant origin: Probing a potential anti-tumor target for psychotropics. Faseb J 19:1187-1189, 2005 24. Serafeim A, Holder MJ, Grafton G, et al: Selective serotonin reuptake inhibitors directly signal for apoptosis in biopsy-like Burkitt lymphoma cells. Blood 101:3212-3219, 2003 25. Evans DL, Lynch KG, Benton T, et al: Selective serotonin reuptake inhibitor and substance p antagonist enhancement of natural killer cell innate immunity in human immunodeficiency virus/ acquired mmunodeficiency syndrome. Biol Psychiatry, [epub ahead of print on October 16, 2007] 26. Bower M, Gazzard B, Mandalia S, et al: A prognostic index for systemic AIDS-related non-Hodgkin lymphoma treated in the era of highly active antiretroviral therapy. Ann Intern Med 143:265-273, 2005 27. Stebbing J, Sanitt A, Nelson M, et al: A prognostic index for AIDS-associated Kaposi's sarcoma in the era of highly active antiretroviral therapy. Lancet 367:1495-1502, 2006[CrossRef][Medline] 28. Bower M, Powles T, Nelson M, et al: Highly active antiretroviral therapy and human immunodeficiency virus-associated primary cerebral lymphoma. J Natl Cancer Inst 98:1088-1091, 2006 29. Silverberg MJ, Neuhaus J, Bower M, et al: Risk of cancers during interrupted antiretroviral therapy in the SMART study. Aids 21:1957-1963, 2007[CrossRef][Medline] 30. Yun LW, Maravi M, Kobayashi JS, et al: Antidepressant treatment improves adherence to antiretroviral therapy among depressed HIV-infected patients. J Acquir Immune Defic Syndr 38:432-438, 2005[CrossRef][Medline] 31. Dalessandro M, Conti CM, Gambi F, et al: Antidepressant therapy can improve adherence to antiretroviral regimens among HIV-infected and depressed patients. J Clin Psychopharmacol 27:58-61, 2007[CrossRef][Medline] 32. Stebbing J, Waters L, Davies L, et al: Incidence of cancer in individuals receiving chronic zopiclone or eszopiclone requires prospective study. J Clin Oncol 23:8134-8136, 2005 33. Psaty BM, Furberg CD: COX-2 inhibitors–lessons in drug safety. N Engl J Med 352:1133-1135, 2005 34. Okie S: What ails the FDA? N Engl J Med 352:1063-1066, 2005 Submitted December 24, 2007; accepted January 22, 2008.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
