Serotonergic antidepressants and increased bleeding risk in patients undergoing breast biopsy

Abstract

Objectives:

Recent investigations have shown that serotonergic antidepressant (SAd) use may increase the risk of peri-operative bleeding events. Our objective was to evaluate the possibility of a similar association in patients undergoing radiologic breast biopsies.

Methods:

We used data from 3890 patients undergoing 6300 biopsy procedures between January 2011 and October 2014 in the Breast Clinic of McGill University Health Centre, Montreal, Canada. In this case-control study, cases were patients reported to have abnormal bleeding during their biopsy by board-certified radiologists. A control group of nonbleeders was selected using matching according to age and type of biopsy. The correlation between abnormal bleeding and SAd use was assessed using bivariate and multivariate statistical analyses.

Results:

There were 97 patients with abnormal bleeding and 137 matched controls; 10 bleeders (cases) were on SAds (7 citalopram, 3 paroxetine) while only 1 nonbleeder (control group) was on a SAd (low-dose sertraline, 25 mg/day). SAds were significantly associated with increased bleeding risk (10.3% versus 0.7%, Fisher’s Exact p = 0.001). Moreover, after adjusting for confounding factors (age, type of biopsy, size of biopsy, needle caliber, pathology result and nonsteroidal anti-inflammatory drug use, multivariate logistic regression confirmed that SAds were associated with elevated bleeding risk (16.2, 95% confidence interval 1.87–140.1, p = 0.01).

Conclusions:

This is the first study demonstrating increased bleeding events in breast biopsy patients using SAds. Clinicians should be aware that SAds may be associated with peri-operative bleeding risk, even in relatively minor procedures such as breast biopsies.

Keywords

bleeding breast biopsy serotonergic antidepressants

Introduction

Serotonergic antidepressants (SAd) are among the most common type of prescribed medications with almost 11% of adults in the US being on antidepressants [Pratt and Broady, 2014]. They can be administered for many psychiatric illnesses including major depressive disorder, anxiety disorders and nicotine dependence [Sayadipour et al. 2012]. SAds consist of selective serotonergic reuptake inhibitors (SSRIs), serotonin–norepinephrine reuptake inhibitors (SNRIs) and some tricyclic antidepressants (TCAs) with high affinity to serotonin receptors [Mahdanian et al. 2014]. Though generally considered safe with high tolerability, SAd use can sometimes be associated with adverse events [Andrade et al. 2010]. These effects include gastrointestinal symptoms, sexual dysfunction, hyponatremia, osteoporosis, diabetes [Abitbol et al. 2015], falls, decreased bone density and increased bleeding risk [Kikuchi et al. 2013].

Breast cancer is one of the most frequent cancers in women. In order to improve early detection and reduce mortality, image-guided biopsies for suspicious lesions have been introduced. Needle biopsies can cause several complications including breast bruising, bleeding and hematoma formation [Chou et al. 2012]. Controlling an active hemorrhage is a major challenge as a complication of biopsy procedures and therefore clinicians try to take necessary precautions to prevent this complication [Sun et al. 2014].

SAd-related increased bleeding risk is an emerging concern since it may contribute to the morbidity and mortality of hemorrhage in surgical settings [Tseng et al. 2013; Singh et al. 2015]. This may be particularly important when SAds are used concurrently with aspirin, nonsteroidal anti-inflammatory drugs (NSAIDs) and anticoagulants [Seitz et al. 2013]. There is a small, but growing literature highlighting the increased peri-operative bleeding risk with SAd use [Mahdanian et al. 2014; Singh et al. 2015]. To the best of the authors’ knowledge there have only been a couple of reports examining the bleeding risk in breast surgery procedures [Basile et al. 2013; Akbulut et al. 2014]. However, no one has looked at relatively minor procedures such as breast biopsies. In this paper, we hypothesized that compared with nonbleeders, patients who bleed during their breast biopsies have higher rates of SAd exposure.

Materials and methods

Study design and setting

The current case-control study utilized retrospective data collected from 3890 patients undergoing 6300 biopsy procedures between January 2011 and October 2014 in the outpatient Breast Clinic of the McGill University Health Centre (MUHC), Montreal, Canada. Patients were only included if they were female, aged ⩾18, had complete medical records and were not using Coumadin (warfarin). Data from patients’ first biopsy procedure between 2011 and 2014 were used. Ethical approval for the study was obtained from the research ethics board of the MUHC. As this was a retrospective chart review study, no consent was obtained from the patients.

Cases

In this study, cases were patients who had abnormal bleeding during their image-guided biopsy procedure. Bleeding events were reported by the board-certified radiologists who performed the biopsy procedure. The radiologists were not aware a priori of a study examining the association between SAds and bleeding. All procedures performed by radiologists were systematically documented, including any kind of complications and particularly any abnormal bleeding right at the time of the biopsy, in a checklist, with one item ‘was there abnormal bleeding during the biopsy’. A survey from the radiologists of our center revealed that none of them were aware of this hypothesis that SAds may be associated with increased bleeding during surgical interventions.

Controls

Potential controls were patients undergoing biopsy procedures during the same time frame without any radiologist-reported bleeding. Using this pool of nonbleeders, up to two controls were matched with every bleeding case according to exact age and type of biopsy. With infrequently performed biopsy procedures, e.g. stereo or magnetic resonance imaging (MRI) guided vacuum biopsy, it was not always possible match two controls to each case, in which instance one control was used. This led us to have 97 patients in the case group and 137 patients in the control group.

Clinical covariates

In our study sample of female breast biopsy patients, we had matched cases and controls for important clinical covariates: age (continuous variable) and type of biopsy. Type of biopsy was categorized into three groups: ultrasound (US) guided core needle, US-guided fine needle aspiration cytology (FNAC) and stereo- or MRI-guided vacuum. Other variables associated with increased bleeding risk were adjusted for in our statistical analyses. The size of the largest biopsy sample (mm) and needle caliber (gauge) were recorded as continuous variables [Al-Harethee et al. 2013]. In terms of patient medications, we looked at current SAd use, as well as pre biopsy use of medications associated with increased bleeding risk (of which only NSAIDs and aspirin had been used). It should be noted that all patients with NSAID and aspirin exposure had been strongly advised by their treating physician to stop the medication at least 1 week before the procedure. We also assessed whether biopsy pathology results were positive, since cancer can be associated with hemostatic dysfunction (either thromboembolism or hemorrhage) [Burbury, 2013]. Radiologists had used similar local anesthesia technique and agents in all patients, so these were not included amongst covariates.

Statistical analysis

Bleeders and nonbleeders were compared for type of biopsy, biopsy pathology results and use of SAds and NSAIDs using the Chi-square test. Groups were compared for age, size of biopsy and needle caliber using Student’s t-test. A multivariable logistic regression was performed to determine whether SAd use was associated with bleeding, independently of age, type of biopsy, size of biopsy, pathology results, needle caliber, or NSAIDs and aspirin usage.

Results

A total of 3890 patients underwent a breast biopsy procedure. Of these, 97 (2.5%) had reports of abnormal bleeding during the procedure. These were age- and type of biopsy- matched to a control group of nonbleeders of 137 patients.

Mean age was 57.9 and 57.7 in bleeders and controls, respectively [t(232) = 0.10, p = 0.92]. All patients were female. There was a trend towards higher NSAID use amongst bleeders (12.4% versus 5.8%, χ² = 3.10, p = 0.078) though patients were supposed to stop those medications at least a week ahead of the breast biopsy. There was no significant difference between two groups in term of type of biopsy, size of biopsy, needle caliber and pathology results. Table 1 shows the baseline demographic and clinical characteristics of the cases and controls.

Table 1.

Baseline demographic and clinical characteristics.

	Bleeders (n = 97)(% or mean ± SD)	Non-bleeders (n = 137)(% or mean ± SD)	Statistics
Demographics
Mean age (years)	57.9 (±13.5)	57.7 (±12.8)	t(232) = 0.10, p = 0.92
Female gender	100% (n = 97)	100% (n = 137)	χ²=0, p = 1.0
Biopsy variables
Type of biopsy			χ²=0.090, p = 0.96
US-guided core needle	53.6% (n = 52)	53.3% (n = 73)	–
US-guided FNAC	5.2% (n = 5)	4.4% (n = 6)	–
Stereo- or MRI-guided vaccuum	41.2% (n = 40)	42.3% (n = 58)	–
Size of biopsy (mm)	1.70 (±0.95)	1.73 (±1.10)	t(232) = 0.26, p = 0.79
Pathology results positive	44.3% (n = 43)	49.6% (n = 68)	χ²=0.64, p = 0.42
Needle caliber (guage)	12.21 (±4.6)	12.40 (±4.3)	t(224) = 0.31, p = 0.76
Medication variables
SAd	10.3% (n = 10)	0.7% (n = 1)	Fisher’s Exact p = 0.001
Antiplatelet or anticoagulant medication	12.4% (n = 12)	5.8% (n = 8)	χ²=3.10, p = 0.078

NB Patients in the ‘antiplatelet/anticoagulant medication category’ had only been exposed to NSAIDs or aspirin.

FNAC, fine needle aspiration cytology; MRI, magnetic resonance imaging; NSAID, nonsteroidal anti-inflammatory drug; SAd, serotonergic antidepressant; SD, standard deviation; US, ultrasound.

SAds were significantly associated with increased bleeding risk (10.3% versus 0.7%, Fisher’s Exact p = 0.001). A total of 10 bleeding cases were exposed to SAds (7 citalopram, 3 paroxetine), whereas only one nonbleeder in the control group had used a SAd (low-dose sertraline, 25mg/day), amounting to an incidence rate of bleeding of 10/11 = 90.9% amongst SAd users in our sample. Other than SAds, none of the bleeders or nonbleeders had exposure to other types of antidepressants.

After adjusting for potentially confounding factors (age, type of biopsy, size of biopsy, pathology result, needle caliber and NSAID use), multivariate logistic regression confirmed that SAds were independently associated with elevated bleeding risk [16.2, 95% confidence interval (CI) 1.87–140.1; p = 0.01) (Table 2). Of note, SAds were not intercorrelated with age, type of biopsy, size of biopsy, pathology result, needle caliber and NSAIDs use suggesting that there was no collinearity between the variables. None of the SAd users currently claimed to be taking NSAIDs or aspirin.

Table 2.

Logistic regression: independent predictors of bleeding during breast biopsy.

Predictor variable	Univariate OR (95% CI)	Multivariate OR (95% CI)	Wald χ²	p value
Age (years)	1.001 (0.98–1.02)	1.00 (0.98–1.02)	0.068	0.79
Needle calibre (gauge)	1.001 (0.93–1.06)	0.94 (0.85–1.03)	1.80	0.18
Type of biopsy:
• FNAC	1.19 (0.35–4.00)	1.32 (0.18–9.70)	0.074	0.78
• Core	1.01 (0.60–1.71)	1.36 (0.52–3.56)	0.38	0.54
• Vacuum	0.96 (0.56–1.62)	1.02 (0.59–1.74)	0.003	0.95
Size of biopsy sample (mm)	0.96 (0.75–1.24)	1.00 (0.68–1.45)	0.001	0.98
Abnormal biopsy pathology results	0.8 (0.48–1.36)	0.71 (0.40–1.26)	1.39	0.24
NSAIDs or aspirin	2.3 (0.89–5.80)	2.72 (0.98–7.6)	3.67	0.055
SAds	15.6 (1.97–124.3)	16.2 (1.87–140.1)	6.38	0.01

CI, confidence interval; FNAC, fine needle aspiration cytology; NSAID, nonsteroidal anti-inflammatory drug; OR, odds ratio; SAd, serotonergic antidepressant.

Discussion

To our knowledge, this is the first study to demonstrate that SAd use increases the risk of abnormal bleeding during breast biopsy procedures. Our novel use of radiological confirmation of bleeding, as opposed to using blood loss or red blood cell transfusions as a primary outcome measure, may have increased our sensitivity for detecting bleeding and facilitated finding this effect in this population. After adjusting for possible confounders, SAds were significantly associated with increased bleeding events in breast biopsy procedures (16.2, 95% CI 1.87–140.1; p = 0.01).

Most studies have found a significant association between SAd use and bleeding risk during operative procedures (OR range 1.21–4.14) [Tavakoli et al. 2012; Mahdanian et al. 2014], with the possible exception of coronary artery bypass grafting [Kim et al. 2009; Xiong et al. 2010]. Similarly, the 2 previous studies on breast cancer surgery [relative risk (RR) = 2.3, 95% CI 1.4–3.9] [Gartner et al. 2010] and breast cosmetic surgery [odds ratio (OR) = 4, 95%CI 1.90–9.04] [Basile et al. 2013] also found significantly increased bleeding associated with SAd use. Our finding is in line with the current theory stating that SAds may increase bleeding risk during breast interventions, but extends this to relatively minor and commonly performed interventions such as breast biopsies.

Bleeding still remains one of the most common complications of breast biopsies. The rate of bleeding in our biopsy procedures was estimated at 2.5%, which was consistent with rates of 0.76% to 7.5% in different studies [Allen et al. 2011; Zagouri et al. 2011]. Several drugs such as platelet inhibitors, vitamin K antagonists, oral anticoagulants, NSAIDs and aspirin are known to increase bleeding and therefore are routinely stopped well ahead of a biopsy procedure. In contrast, SAds are rarely, if ever, stopped because of concerns about bleeding. There are little data addressing the risks and benefits of stopping SAds and its timing. SAds are believed to increase bleeding by decreasing platelets’ binding affinity due to reducing its serotonin levels [Harirchian et al. 2012]. This is important because platelet serotonin concentration decreases by more than 80% after 2 weeks of using a SAd [Hergovich et al. 2000]. One study by Richter and colleagues found that taking sertraline 24 hours before gastrostomy had a higher impact of bleeding (adjusted OR = 3.87, 95% CI 1.06–11.88), but not if it was used more than 48 hours before the procedure (OR = 1.0, 95% CI 0.0–11.79) [Richter et al. 2011]. Antiplatelet agents are usually withheld 7–10 days prior to biopsy procedures to allow platelet renewal [Douketis et al. 2012].

Taking all this together, it can be inferred that SAd-related bleeding maybe less even after 2–10 days of stopping SAd. Such approaches will require testing in future clinical trials, particularly given the mental health risks of even brief SAd discontinuation [Howland, 2010]. In the meantime, clinicians should more vigilant for this possible complication in the population on SAds, particularly in patients on other agents associated with bleeding (such as NSAIDs), as they may have synergistic effects with SAds on bleeding [Douketis et al. 2012]. We should emphasize that we used a sensitive radiological assessment of bleeding, which may be a useful research probe for examining medications’ potential for bleeding in moderate-sized samples (n = 100–1000), but may not be clinically important in individual breast biopsy patients. Therefore at this time, we would suggest that clinicians be aware of the potential for bleeding and not routinely discontinue SAds prior to breast biopsies.

Limitations

Our study has some limitations. First, we used retrospective data from a clinical database. As a case-control study it was more susceptible to selection bias, although we tried to minimize this by using logistic regression analyses. Our use of an age- and biopsy type-matched control group may not necessarily have controlled for confounding, but we tried to adjust for these factors in the analyses. Also, because of the low exposure rates to SAds, particularly in the control group (n = 1), we had wide 95% CIs for our effect size estimates. Another limitation was that we did not assess depression, so that the patients suffering from it might have had a less healthy lifestyle which could possibly have played some role in their bleeding tendency. However, these possible lifestyle confounders are present in almost all studies examining SAds’ relationship with bleeding [Palmsten et al. 2013]. Finally, bleeding was ascertained according to radiologists’ clinical reports following biopsies. A quantitative estimate of bleeding or a follow-up visit to check for bruising/hematomas may have been helpful. Still, our study was the first to use subjective radiological confirmation of bleeding, an outcome measure that was likely more sensitive than some quantitative measures used (e.g. amount of blood loss), although perhaps less rigorous. Nonetheless our method was comparable with measures used in some studies in this field, in which bleeding has been defined using physician diagnostic billing codes and subjective clinical reports [Arnason et al. 200; Borzecki et al. 2011; Napenas et al. 2011; Auerbach et al. 2013; Nouraei et al. 2015].

Conclusion

Clinicians should be aware that SAds may be associated with increased periprocedural bleeding, even in relatively minor procedures like breast biopsy. Therefore, clinicians caring for patients undergoing breast biopsies should be particularly vigilant about bleeding risk in patients using SAds. However, the decision to stop SAds before the procedure should be made with a complete discussion of the risks and benefits with other physicians involved in the care of the patient. Future research can investigate potential strategies to diminish SAd-related bleeding risk.

Clinical practice points

SAds are associated with an increased risk of bleeding during breast biopsies. From a clinical perspective, this potential risk needs to be carefully weighed against the mental health benefits of SAd use. Clinicians caring for patients undergoing breast biopsies should be particularly vigilant about bleeding risk in patients using SAds.

Footnotes

Acknowledgements

We are grateful to Cedars Breast Clinic of McGill University Health Centre, especially to Dr Armen Parsyan, Anita Svadzian and Kim Martire who were actively involved in the development of the database.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. A.M. receives research assistantship funding from the Cedars Breast Centre and Master’s Award from the Fonds de Recherche du Quebec-Santé (FRQS). S.B. holds New Investigator salary awards from Canadian Institutes of Health Research (CIHR) and FRQS. S.R. has received Master’s Awards from CIHR and FRQS for separate projects.

Conflict of interest statement

The authors declare no conflicts of interest in preparing this article.

References

Abitbol

Rej

Segal

Looper

(2015) Diabetes mellitus onset in geriatric patients: does long-term atypical antipsychotic exposure increase risk? Psychogeriatrics 15: 43–50.

Akbulut

Yagmur

Gumus

Babur

(2014) Breast ecchymosis: unusual complication of an antidepressant agent. Int J Surg Case Rep 5: 129–130.

Al-Harethee

Theodoropoulos

Filippakis

Papapanagiotou

Matiatou

Georgiou

. (2013) Complications of percutaneous stereotactic vacuum assisted breast biopsy system utilizing radio frequency. Eur J Radiol 82: 623–626.

Allen

Nerurkar

Della Rovere

(2011) The breast lesion excision system (BLES): a novel technique in the diagnostic and therapeutic management of small indeterminate breast lesions? Eur J Radiol 21: 919–924.

Andrade

Sandarsh

Chethan

Nagesh

(2010) Serotonin reuptake inhibitor antidepressants and abnormal bleeding: a review for clinicians and a reconsideration of mechanisms. J Clin Psychiatry 71: 1565–1575.

Arnason

Wells

Van Walraven

Forster

(2006) Accuracy of coding for possible warfarin complications in hospital discharge abstracts. Thromb Res 118: 253–262.

Auerbach

Vittinghoff

Maselli

Pekow

Young

Lindenauer

(2013) Perioperative use of selective serotonin reuptake inhibitors and risks for adverse outcomes of surgery. J Am Med Assoc Intern Med 173: 1075–1081.

Basile

(2013) Use of selective serotonin reuptake inhibitors antidepressants and bleeding risk in breast cosmetic surgery. Aesthetic Plast Surg 37: 561–566.

Borzecki

Kaafarani

Cevasco

Hickson

Macdonald

Shin

. (2011) How valid is the AHRQ Patient Safety Indicator ‘postoperative hemorrhage or hematoma’? J Am Coll Surg 212: 946–953.

10.

Burbury

(2013) Haemostatic challenges in the cancer patient: focus on the perioperative period. Best Pract Res Clin Anaesthesiol 27: 493–511.

11.

Chou

Wang

Chang

Liu

Kuo

(2012) Evaluation of the effects of chitosan hemostasis dressings on hemorrhage caused by breast biopsy. Breast Care (Basel) 7: 220–224.

12.

Douketis

Spyropoulos

Spencer

Mayr

Jaffer

Eckman

. (2012) Perioperative Management of Antithrombotic Therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 141(2 Suppl.): e326S–e350S.

13.

Gartner

Cronin-Fenton

Hundborg

Pedersen

Lash

Sorensen

. (2010) Use of selective serotonin reuptake inhibitors and risk of re-operation due to post-surgical bleeding in breast cancer patients: a Danish population-based cohort study. BMC Surg 10: 3

14.

Harirchian

Zoumalan

Rosenberg

(2012) Antidepressants and bleeding risk after face-lift surgery. Arch Facial Plast Surg 14: 248–252.

15.

Hergovich

Aigner

Eichler

Entlicher

Drucker

Jilma

(2000) Paroxetine decreases platelet serotonin storage and platelet function in human beings. Clin Pharmacol Ther 68: 435–442.

16.

Howland

(2010) Potential adverse effects of discontinuing psychotropic drugs: Part 2: antidepressant drugs. J Psychosoc Nurs Ment Health Serv 48: 9–12.

17.

Kikuchi

Suzuki

Uchida

Watanabe

Mimura

(2013) Association between antidepressant side effects and functional impairment in patients with major depressive disorders. Psychiatry Res 210: 127–133.

18.

Kim

Daskalakis

Whellan

Whitman

Hohmann

Medvedev

. (2009) Safety of selective serotonin reuptake inhibitor in adults undergoing coronary artery bypass grafting. Am J Cardiol 103: 1391–1395.

19.

Mahdanian

Rej

Bacon

Ozdin

Lavoie

Looper

(2014) Serotonergic antidepressants and perioperative bleeding risk: a systematic review. Expert Opin Drug Saf 13: 695–704.

20.

Napenas

Hong

Kempter

Brennan

Furney

Lockhart

(2011) Selective serotonin reuptake inhibitors and oral bleeding complications after invasive dental treatment. Oral Surg Oral Med O 112: 463–467.

21.

Nouraei

Hudovsky

Frampton

Mufti

White

Wathen

. (2015) A study of clinical coding accuracy in surgery: implications for the use of administrative big data for outcomes management. Ann Surg 261: 1096–1107.

22.

Palmsten

Hernandez-Diaz

Huybrechts

Williams

Michels

Achtyes

. (2013) Use of antidepressants near delivery and risk of postpartum hemorrhage: cohort study of low income women in the United States. Br Med J 347: f4877.

23.

Pratt

Brody

(2014) Depression and obesity in the U.S. adult household population, 2005–2010. NCHS Data Brief: 1–8.

24.

Richter

Patrie

Richter

Henry

Pop

Regan

. (2011) Bleeding after percutaneous endoscopic gastrostomy is linked to serotonin reuptake inhibitors, not aspirin or clopidogrel. Gastrointest Endosc 74: 22–34.

25.

Sayadipour

Mago

Kepler

Chambliss

Certa

Vaccaro

. (2012) Antidepressants and the risk of abnormal bleeding during spinal surgery: a case-control study. Eur Spine J 21: 2070–2078.

26.

Seitz

Bell

Gill

Reimer

Herrmann

Anderson

. (2013) Risk of perioperative blood transfusions and postoperative complications associated with serotonergic antidepressants in older adults undergoing hip fracture surgery. J Clin Psychopharmacol 33: 790–798.

27.

Singh

Achuthan

Chakrabarti

Rajagopalan

Srinivasan

Hota

(2015) Influence of pre-operative use of serotonergic antidepressants (SADs) on the risk of bleeding in patients undergoing different surgical interventions: a meta-analysis. Pharmacoepidemiol Drug Saf 24: 237–245.

28.

Sun

Hennessey

Kam Nakch

Alsharif

Meterissian

Mesurolle

(2014) Compression-refractory breast hematoma secondary to pseudoaneurysm after stereotactically guided vacuum-assisted biopsy: the critical role of urgent surgical evacuation. J Clin Ultrasound 42: 492–494.

29.

Tavakoli

Demaio

Wingert

Rieg

Cohn

Balmer

. (2012) Serotonin reuptake inhibitors and bleeding risks in major orthopedic procedures. Psychosomatics 53: 559–565.

30.

Tseng

Chiang

Lane

Lai

(2013) Selective serotonin reuptake inhibitors reduce P2Y12 receptor-mediated amplification of platelet aggregation. Thromb Res 131: 325–332.

31.

Xiong

Jiang

Clare

Shaw

Smith

O’Connor

. (2010) Safety of selective serotonin reuptake inhibitor use prior to coronary artery bypass grafting. Clin Cardiol 33: E94–98.

32.

Zagouri

Gounaris

Liakou

Chrysikos

Flessas

Bletsa

. (2011) Vacuum-assisted breast biopsy: more cores, more hematomas? In Vivo 25: 703–705.