Left ventricular assist devices (LVADs) provide therapy to advanced heart failure patients in a realm of medicine where few therapeutic options exist. Although significant advancement in LVAD technology has improved survival and reduced complications, bleeding remains a common source of morbidity and mortality in this patient population. Bleeding after LVAD implantation affects 32–44% of patients, and is a multifactorial problem that remains poorly elucidated.1–4 Although directly or indirectly related to hematologic changes after LVAD implantation or the necessary anticoagulation/antiplatelet therapies, bleeding events extend well beyond the LVAD operative period and tend to be systemic in impact. Bleeding complications encountered by LVAD patients include mediastinal bleeding, gastrointestinal (GI) bleeding, recurrent epistaxis, and intracerebral bleeding.
The impact of sex on LVAD patient bleeding risk remains understudied,5–8 a deficit that can be largely be attributed to the lower frequency of LVAD support in females compared with males. Females comprise only 20–25% of LVAD clinical studies and ~20% of the 19,013 patients within the Interagency Registry for Assisted Circulatory Support (INTERMACS). In a 2012 analysis of n = 1,936 patients supported with pulsatile and continuous-flow (CF) LVADs in INTERMACS, sex was not associated with mortality or bleeding.9 Likewise, a multicenter analysis of 147 females and 577 males found no difference in survival or GI bleeding events in men versus women on CF-LVAD support.10 However, a detailed analysis of bleeding risk was not undertaken and other studies have shown a higher rate of hemorrhagic stroke and major bleeding events in females compared with males on LVAD support.4,11–13
In modern medicine, there is an effort to personalize medical and device therapy based on individual risk factors and response. We hypothesized that due to a combination of the aforementioned factors, female LVAD recipients suffer a higher rate of bleeding complications. We thus sought to examine the incidence and prevalence of all clinically significant bleeding, characterizing the type of bleeds by sex and time period after LVAD implantation.
This was a single-center, retrospective review of consecutive patients (n = 375) implanted with Food and Drug Administration approved CF-LVADs from 2008 to 2017. Patients with a prior first-generation device, HeartMate 3 (Abbott/St. Jude, St. Paul, MN), or total artificial heart were excluded. Anticoagulation strategy did not differ by sex and all patients were treated with Aspirin 325 mg daily as well as warfarin with an international normalized ratio (INR) goal of 2–3. All patients are enrolled into the institutional anticoagulation monitoring program, which follows a protocol adjusted of warfarin dosing and regular INR monitoring. Bleeding events were categorized as follows: mediastinal/thoracic (requiring reoperation or >3 units blood occurring within 7 days of implant or transfusion >1 unit after 7 days), naso-oropharyngeal (requiring emergency room visit or hospitalization with transfusion or intervention by an ear, nose, throat [ENT] specialist), GI (hemoglobin drop ≥2 g with positive stool blood occult test and no other obvious sources), gynecologic ([GYN], requiring emergency room visit or hospitalization with intervention or transfusion), intracranial hemorrhage ([ICH], ischemic strokes with secondary hemorrhage were excluded), and “other” (required admission with surgical intervention or transfusion of >1 unit of blood). The time to index bleeding event and time to event for all bleeding categories were tallied.
SPSS (Chicago, IL) version 24 statistical software was used for data analysis. Continuous data were evaluated for normality and are presented as mean ± standard error or median [25th, 75th] as appropriate. Student’s t-tests or Mann–Whitney U tests were used for group comparisons based on data normality. Kaplan–Meier survival estimates were generated and compared with log-rank testing. Patients were censored at the time of transplant or explant for recovery. Cox regression was then used to generate hazard ratios (HRs, [95% CI]) for bleeding. Adjusted Cox analyses were performed and included variables with a p < 0.10 on univariable analysis (age, sex, destination therapy [DT], diabetes, device type). Incident and prevalent event rates were compared after calculation of confidence intervals.14 For all analyses except the multivariable model entry criterion, a p ≤ 0.05 was considered statistically significant.
Internal review board approval was obtained for data collection and analysis.
There were 375 CF-LVADs implanted with a median support duration of 399 days [141,762] (mean 534 days). Of these, 84 CF-LVADs (22%) were placed in females and 291 (78%) were placed in males. Table 1 shows the baseline characteristics and demographics of the cohort. In total, 33% of patients were INTERMACS profile 1–2 and 49% had ischemic cardiomyopathy.
There were 189 patients (50%) with an aggregate of 406 incident bleeding events prompting transfusion or admission for monitoring. As demonstrated in Figure 1, there were 226 GI bleeding events, 54 naso-oropharyngeal, 53 mediastinal, 37 ICH, and 12 GYN bleeding events. Table 2 shows the frequency or rate of events by bleeding category. Mediastinal bleeding occurred almost exclusively in the operative period, whereas all other bleeding events tended to occur after 30 days but within 1 year of VAD implant. There were 54 naso-oropharyngeal bleeds, including 41 epistaxis episodes (5 requiring embolization) and 3 major bleeds after dental extractions (1 requiring intensive care unit admission). Other (non-categorized) bleeds included tracheal bleeds (n = 3), spontaneous rectus sheath hematomas (n = 4), retroperitoneal bleeds (n = 3), implantable defibrillator hematoma after device revision/placement (n = 2), severe hematuria (n = 1), bleeding after cholecystectomy (n = 1) and hip surgery (n = 2), chest wall bleed after thoracentesis (n = 2), spontaneous gluteal/thigh bleed (n = 2), and splenic hematoma/spontaneous rupture (n = 4).
Predictors of Bleeding in the Entire Cohort
Freedom from a clinically significant bleeding event in the cohort at 30 days and 2 years was 77 ± 2.2% and 43 ± 3.2%, respectively. On unadjusted analysis, patients with bleeding events were more likely to be female and older, and were receiving an LVAD for D with more time on LVAD support (Table 1). Diabetic patients and patients implanted with axial flow devices showed a nonsignificant trend toward increased bleeding. On multivariable analysis, female sex (adjusted HR 1.6 [1.1–2.2], p = 0.006) and age (adjusted HR 1.2 [1.1–1.3 per 10 years], p = 0.002) were associated with at least one episode of clinically significant postoperative bleeding. DT intent (p = 0.63), diabetes (p = 0.23), and axial flow device type (p = 0.44) were not associated with bleeding on adjusted analysis.
Bleeding Events in Males Versus Females
Table 3 shows the baseline characteristics of males versus females in this cohort. Women were younger and had lower preoperative serum creatinine and hemoglobin values than men. There was no significant difference in duration of device support or LVAD type implanted. Figure 2 visualizes the freedom from bleeding over time by sex. In females, 30-day and 2-year freedom from bleeding was 75 ± 4.8% and 33 ± 6.2% compared with 78 ± 2.5% and 46 ± 3.7% in men, respectively (p = 0.027).
Overall there were 293 clinically significant bleeding events in males and 113 in females. Table 2 shows the bleeding event rates categorized by patient sex. Although there was no difference in the frequency of mediastinal bleeding (p = 0.35 at 30 days) or rates of GI bleeding (p = 0.93), incident event rates for overall bleeding (eppy: 0.52 males and 0.80 females) and nasopharyngeal bleeding (eppy: 0.07 males and 0.13 females) were higher in females (p < 0.05). Nonsignificant trends were noted for increased rates of ICH in females versus males (p = 0.14). The burden of gynecological (GYN) bleeding was also not low in females. Of the 84 females (mean 52 ± 1.4 years), 8 (10%) experienced 12 incident GYN bleeding events requiring transfusion or surgical intervention. GYN bleeding events included nonspecific vaginal bleeding or uterine/endometrial bleeding (n = 10) and one large ovarian hemorrhage. Of these, two had a hysterectomy, two required endometrial ablation, one required oophorectomy, and the remaining received hormonal therapy.
Overall survivals at 30 days and 2 years were 94 ± 1.2% and 65 ± 3.0%. There was no significant difference in 2-year survival between sexes with survival in females and males of 65 ± 6.8% and 65 ± 3.4% (p = 0.36), respectively.
Thirty-day and 2-year survivals in those with (30-day = 97 ± 1.2%; 2Y = 63 ± 4.0%) and without (30-day = 91 ± 2.1%; 2Y = 67 ± 4.5%) bleeding events were similar (p = 0.31). However, in patients who suffered intracranial hemorrhage, survival at 2 years was 38 ± 8.6% vs. 69 ± 3.0% in those without ICH (p < 0.001). There was no difference in overall survival for patients with or without naso-oropharyngeal bleeding, mediastinal bleeding, or GI bleeding (data not shown).
In this single-center cohort study of 375 patients on CF-LVAD support, we found that 57% of patients experienced at least one clinically significant bleeding event within 2 years of LVAD implant. Mucosal bleeding, such as epistaxis, dental, and vaginal bleeding, is typically benign in non-LVAD patients and does not usually necessitate admission. Conversely, mucosal bleeding events in LVAD patients, especially females, were a source of marked morbidity in this cohort. These events tended to be recurrent, often leading to need for transfusions or invasive procedures. Risk factors for the occurrence of a clinically significant bleeding included female sex and older age. Compared with males, we observed higher rates of both overall and naso-oropharyngeal bleeding in females, with significant morbidity encountered from gynecologic bleeding.
Although females represent the minority of patients on LVAD support, there are ~4,000 females to date enrolled into INTERMACS,8 and the prevalence of females on LVAD support can be expected to increase as device size and implant complexity decrease. In order to individualize care and better assess or mitigate LVAD candidate risk, an understanding of sex-specific differences in outcomes after LVAD implantation is essential. In this cohort study, we did not find a difference in overall survival between females and males on LVAD support. However, we did find that females on LVAD support had a 60% higher adjusted risk of bleeding compared with males. The higher overall bleeding rate in females was largely due to non-GI mucosal bleeding (oral–nasopharyngeal and vaginal), with trends noted for an increase in the rate of increased ICH in women compared with men as well. A single-center study of 139 HeartMate II patients found similar cohort rates (1.4 eppy vs. 1.2 eppy herein) of overall major bleeding, but the authors did not find an association between female sex and bleeding risk.15 The findings of increased bleeding risk in this analysis, however, have been supported by other cohort studies,10,11 secondary analysis of clinical trials,12 and other large database analyses.13,16 A study by Boyle et al.4 of more than 900 HeartMate II clinical trial patients also identified female sex (HR 1.5) and age >65 years (HR 1.3) as correlates of increased bleeding risk after index LVAD hospital discharge. Likewise, a study of the European Registry for Patients with Mechanical Circulatory Support (EUROMACS) showed that females (0.3 eppy) had higher rates of major bleeding than males (0.14 eppy).13 In the Boyle analysis and a study by Morris, an increased risk of hemorrhagic stroke was also found in females compared with males supported with LVADs.4,11 In the analysis herein, we found a trend toward increased incident rates of ICH in females (eppy 0.10) versus males (eppy 0.10) and identified ICH as the only bleeding event that was associated with inferior survival. Clearly more studies with larger numbers of female LVAD patients examining sex-specific outcomes will be required to get a more accurate and precise estimate of differential outcomes of females versus males on LVAD support.
Bleeding events in CF-LVAD recipients are likely multifactorial in origin, with contributions from clotting factor insufficiencies, deregulation of angiopoietin-2, cytokines such as tumor necrosis factor-α, requirements for anticoagulation and antiplatelet agents, and development of angiodysplasias or vascular wall fragility during nonpulsatile support.1,2,17–19 The etiology for increased bleeding events in females compared with males on LVAD support cannot be clearly elucidated from this analysis. Clearly, females have a uniquely increased bleeding risk from gynecological causes on anticoagulation. One could also hypothesize that bleeding risk is increased in females due to increased operative complexity secondary to smaller body sizes. However, we did not find an increased frequency of mediastinal bleeding in females compared with males, and the other bleeding events occurred largely after the operative risk period. The main difference in bleeding rates in this analysis was driven by mucosal bleeding.
Insight into increased bleeding events in female LVAD patients may be gleaned from other patient populations. In acute coronary syndromes, increased mortality observed in females has been attributed by some to be due to sex-specific differences in platelet activity.20 Female sex has also been found to be a risk factor for bleeding in transcatheter aortic valve implantation.21 This increased hazard of bleeding has been theorized to be due to smaller body surface area, increased frailty, and sex hormone differences.21 However, data sufficient for identifying a causal etiology for the increased bleeding in females versus males in general are lacking. As such, future studies in the LVAD patient population are needed to examine sex-specific differences in the hematologic cascades that may contribute to bleeding risk in females. In addition, well-powered studies are needed to determine whether alternative anticoagulation or antiplatelet management strategies should be employed based on sex-specific characteristics in LVAD patients.
Although the presence of a bleeding event did not confer an increased mortality in this analysis, bleeding is a source of morbidity for LVAD patients. Seemingly benign events, such as epistaxis and dental-related bleeding, led to the admission and intervention in 14% of patients. Rates of naso-oropharyngeal bleeding were almost double in females. Others have also identified an increased risk of bleeding in LVAD patients during dental surgery22 and routine noncardiac surgeries23 compared with the general population. Because the definitions used in this analysis required admission or intervention, it is likely that the impact of mucosal bleeding in this patient cohort on patient morbidity is underestimated. Quality-of-life metrics, although not employed in this analysis, would be an important addition to future assessments of bleeding in males and females on LVAD support.
This study has limitations inherent to its retrospective, single-center design, including that of statistical power. Small patient numbers limited the number of covariates entered into the multivariable regression analysis. In addition, although our center follows a standardized anticoagulation and antiplatelet protocol, the analysis lacked data granularity on patient drug regimens, time in therapeutic range, INRs, platelet counts, and measure of platelet function at the time of bleeding events. Further, we did not track which female patients were postmenopausal or surgically sterilized.
Clinically significant bleeding impacted 57% of LVAD patients in this cohort within 2 years of device implant. Females on LVAD support had a 60% adjusted higher risk of bleeding than their male counterparts with significant morbidity encountered from mucosal (GYN and oral–nasopharyngeal) bleeding. Future analyses aiming to identify an optimal anticoagulation or antiplatelet regimen for females on LVAD support are warranted. Premenopausal women should work closely with their gynecologist before LVAD implant to determine a plan for menses control. Finally, although mortality on LVAD support is similar in males and females, several studies have now demonstrated differences in bleeding risk. Thus, more studies with a larger number of female patients are needed to better understand the sex-specific outcomes after LVAD implantation and the impact thereof on quality of life during long-term LVAD support.
1. Suarez J, Patel CB, Felker GM, Becker R, Hernandez AF, Rogers JG. Mechanisms of bleeding
and approach to patients with axial-flow left ventricular assist devices. Circ Heart Fail 2011.4: 779784.
2. Tabit CE, Coplan MJ, Chen P, et al. Tumor necrosis factor-α levels and non-surgical bleeding
in continuous-flow left ventricular assist devices. J Heart Lung Transplant 20172498: 31834X.
3. Uriel N, Pak SW, Jorde UP, et al. Acquired von Willebrand syndrome after continuous-flow mechanical device support contributes to a high prevalence of bleeding
during long-term support and at the time of transplantation. J Am Coll Cardiol 2010.56: 12071213.
4. Boyle AJ, Jorde UP, Sun B, et al.; HeartMate II Clinical Investigators: Pre-operative risk factors of bleeding
and stroke during left ventricular assist device
support: an analysis of more than 900 HeartMate II outpatients. J Am Coll Cardiol 2014.63: 880888.
5. Mehra MR, Naka Y, Uriel N, et al. MOMENTUM 3 Investigators. A fully magnetically levitated circulatory pump for advanced heart failure
. N Engl J Med 2017.376: 440450.
6. Pagani FD, Miller LW, Russell SD, et al.; HeartMate II Investigators: Extended mechanical circulatory support with a continuous-flow rotary left ventricular assist device
. J Am Coll Cardiol 2009.54: 312321.
7. Rogers JG, Pagani FD, Tatooles AJ, et al. Intrapericardial left ventricular assist device
for advanced heart failure
. N Engl J Med 2017.376: 451460.
8. INTERMACS Quarterly Statistical Report 2016 Q4. The Data and Clinical Coordinating Center University of Alabama at Birmingham. Available at: <https://www.uab.edu/medicine/intermacs/images/Federal_Quarterly_Report/Federal_Partners_Report_2016_Q4.pdf+>
. Accessed March 27, 2017.
9. Hsich EM, Naftel DC, Myers SL, et al. Should women receive left ventricular assist device
support?: Findings from INTERMACS. Circ Heart Fail 2012.5: 234240.
10. Meeteren JV, Maltais S, Dunlay SM, et al. A multi-institutional outcome analysis of patients undergoing left ventricular assist device
implantation stratified by sex and race. J Heart Lung Transplant 2017.36: 6470.
11. Morris AA, Pekarek A, Wittersheim K, et al. Gender differences in the risk of stroke during support with continuous-flow left ventricular assist device
. J Heart Lung Transplant 2015.34: 15701577.
12. Bogaev RC, Pamboukian SV, Moore SA, et al.; HeartMate II Clinical Investigators: Comparison of outcomes in women versus men using a continuous-flow left ventricular assist device
as a bridge to transplantation. J Heart Lung Transplant 2011.30: 515522.
13. Magnussen C, Bernhardt AM, Ojeda FM, et al. Gender differences and outcomes in left ventricular assist device
support: The European Registry for patients with mechanical circulatory support. J Heart Lung Transplant 2017.S1053–S2498: 31882X.
14. Sahai H, Khurshid A. Statistics in Epidemiology: Methods, Techniques, and Applications. 1996. Boca Raton: CRC, Print.
15. Bunte MC, Blackstone EH, Thuita L, et al. Major bleeding
during HeartMate II support. J Am Coll Cardiol 2013.62: 21882196.
16. Kirklin JK, Naftel DC, Pagani FD, et al. Seventh INTERMACS annual report: 15,000 patients and counting. J Heart Lung Transplant 2015.34: 14951504.
17. Tabit CE, Chen P, Kim GH, et al. Elevated angiopoietin-2 level in patients with continuous-flow left ventricular assist devices leads to altered angiogenesis and is associated with higher nonsurgical bleeding
. Circulation 2016.134: 141152.
18. Marsano J, Desai J, Chang S, Chau M, Pochapin M, Gurvits GE. Characteristics of gastrointestinal bleeding
after placement of continuous-flow left ventricular assist device
: A case series. Dig Dis Sci 2015.60: 18591867.
19. Skouri H, Shurrab M, Zahnan J, et al. Hemorrhoids screening and treatment prior to LVAD: Is it a necessity? J Cardiothorac Surg 2016.11: 58.
20. Di Giosia P, Passacquale G, Petrarca M, Giorgini P, Marra AM, Ferro A. Gender differences in cardiovascular prophylaxis: Focus on antiplatelet treatment. Pharmacol Res 2017.119: 3647.
21. Chandrasekhar J, Dangas G, Yu J, et al. STS/ACC TVT registry. Sex-based differences in outcomes with transcatheter aortic valve therapy: TVT registry from 2011 to 2014. J Am Coll Cardiol 2016.68: 27332744.
22. Morimoto Y, Nakatani T, Yokoe C, Kudo C, Hanamoto H, Niwa H. Haemostatic management for oral surgery in patients supported with left ventricular assist device
–a preliminary retrospective study. Br J Oral Maxillofac Surg 2015.53: 991995.
23. Taghavi S, Jayarajan SN, Ambur V, et al. Noncardiac surgical procedures after left ventricular assist device
implantation. ASAIO J 2016.62: 370374.