What are medications that have data to prevent LVAD related GI bleeds?

Comment by InpharmD Researcher

Available data suggest that several medications, including digoxin, angiotensin II inhibitors, long-acting octreotide, and omega-3 fatty acids, may be associated with significant reductions in gastrointestinal bleeding (GIB) in patients with left ventricular assist devices (LVADs). However, the studies are limited by their retrospective, observational nature, and the variability in patient response underscores the need for further research to establish reliable prophylaxis protocols for GI bleeds in LVAD patients. In the absence of comparative randomized controlled trials, prophylaxis strategies are recommended to be individualized based on patient history, bleeding pattern, and tolerance of pharmacologic therapies.

Background

A 2020 review article comprehensively examined the treatment and prophylactic strategies for managing gastrointestinal bleeding (GIB) in patients with left ventricular assist devices (LVADs). Initial management of GIB involves supportive care with blood transfusions and withholding anticoagulation, while endoscopic evaluation remains a cornerstone for both diagnosis and hemostatic therapy when feasible. However, due to the often diffuse or recurrent nature of bleeding, especially from small bowel arteriovenous malformations (AVMs), prophylactic pharmacologic interventions have gained increasing attention. Prophylactic medications commonly assessed in literature include octreotide, thalidomide, danazol, and estrogen-based therapies. Octreotide, a somatostatin analogue, has shown promise in decreasing splanchnic blood flow and inhibiting angiogenesis, with some observational data supporting reduced bleeding recurrence. Thalidomide, with its anti-angiogenic properties via suppression of vascular endothelial growth factor (VEGF), demonstrated efficacy in small case series, though its neurotoxicity and thrombotic risks limit broader use. Danazol, a synthetic androgen, may improve mucosal integrity and has been associated with fewer bleeding episodes, though data remain limited. Estrogen-based therapies, while previously considered for angiodysplasia-associated GIB, have shown inconsistent results and are generally less favored in current practice due to side effects and lack of robust efficacy. The authors concluded that no single agent has been definitively proven to prevent GIB in all LVAD patients, and clinical decisions are often individualized based on patient history, bleeding pattern, and tolerance of pharmacologic therapies. Furthermore, there remains a need for randomized controlled trials to better define the efficacy and safety of prophylactic medications. Overall, a combination of endoscopic surveillance, tailored anticoagulation strategies, and selective use of prophylactic agents appears to offer the most pragmatic approach in reducing the burden of GIB in this population. [1]

A 2020 meta-analysis evaluated the association between angiotensin II antagonists, specifically angiotensin-converting enzyme inhibitors/angiotensin II receptor blockers (ACEi/ARB), and the incidence of GIB in patients implanted with continuous-flow left ventricular assist devices (CF-LVADs). The investigation pooled data from three retrospective cohort studies encompassing a total of 619 patients, 467 of whom received ACEI or ARB therapy. Eligible studies included in the analysis were required to report on the incidence of GIB and/or AVM-related GIB in adult CF-LVAD patients who received and did not receive ACEI/ARB therapy. The meta-analysis demonstrated a statistically significant association between ACEi/ARB use and reduced overall GIB in CF-LVAD recipients, with a pooled OR of 0.35 (95% CI, 0.22–0.56; p < 0.001; I² = 0%). However, the relationship between ACEI/ARB therapy and AVM-related GIB was not statistically significant (pooled odds ratio [OR], 0.46; 95% CI, 0.19–1.07; p = 0.07; I² = 51%). Sensitivity analyses suggested that excluding data from the largest contributing study revealed statistical significance for AVM-related bleeding, indicating potential heterogeneity in methods or diagnostic sensitivity across included studies. Dosing effects were inconclusive—though doses equivalent to >5 mg daily of lisinopril were linked to fewer major GIB events, lower doses were paradoxically associated with reduced AVM-related GIB in at least one included report. Duration of exposure was inconsistently defined across studies, and was limited by retrospective design, making it difficult to conclude causality or optimal therapeutic parameters. No differences in outcomes between ACEIs and ARBs were delineated. [2], [3]

A 2023 narrative review synthesized clinical knowledge surrounding GIB related to CF-LVAD. Based on registry analyses and small retrospective studies, the review highlighted several pharmacological interventions targeting both primary and secondary prophylaxis. Utilizing ACEi/ARB was associated with reduced GIB risk by downregulating angiogenic mediators such as TGF-β, VEGF, and angiopoietin-2. A 2019 retrospective cohort analysis demonstrated a 57% reduction in GIB with ACEi/ARB use and a dose-response threshold, although findings remain inconsistent across studies. In secondary prevention, long-acting octreotide showed favorable outcomes in multicenter retrospective series, with reductions in bleeding recurrence and need for transfusions. Other studied agents included digoxin, shown to reduce GIB incidence via HIF-1α inhibition, and omega-3 fatty acids, which in a 2018 observational cohort resulted in a 97% freedom from GIB at one year. Despite these promising pharmacologic strategies, no randomized controlled trials have been conducted, and current evidence is derived exclusively from non-randomized, registry-based, or observational datasets. The review further underscored that newer-generation devices like the HeartMate 3, while reducing thromboembolic events and possibly AVM formation due to restored pulsatility and shear stress mitigation, still confer a considerable residual GIB risk, warranting continued investigation into integrated prophylactic and therapeutic algorithms. [4]

Lastly, a 2021 network meta-analysis (NMA) assessed the efficacy of various pharmacologic interventions in preventing recurrent GIB in patients with LVAD. The analysis included 13 observational studies that compared six strategies, with four focusing on primary prevention and nine on secondary prevention of GIB. The findings revealed that thalidomide, omega-3 fatty acids, octreotide, and danazol were associated with a reduced risk of recurrent GIB. Specifically, thalidomide showed a hazard ratio (HR) of 0.016 (Credible Interval [CrI] 0.00053 to 0.12), omega-3 fatty acids had an HR of 0.088 (CrI 0.026 to 0.77), octreotide had an HR of 0.17 (CrI 0.0589 to 0.41), and danazol also had an HR of 0.17 (CrI 0.059 to 0.41). On the other hand, ACEi/ARB and digoxin did not show any significant reduction in GIB risk. Overall, the findings suggested that thalidomide, danazol, and octreotide were effective in reducing recurrent GIB, while omega-3 fatty acids were particularly helpful in lowering the risk of the primary episode of GIB in the LVAD population. However, it is important to note that the observational nature of included studies may increase the heterogeneity, and the severity of GIB can not be specified without the detailed information of endoscopy and the number of units of blood transfusion. [5]

References:

[1] Vedachalam S, Balasubramanian G, Haas GJ, Krishna SG. Treatment of gastrointestinal bleeding in left ventricular assist devices: A comprehensive review. World J Gastroenterol. 2020;26(20):2550-2558. doi:10.3748/wjg.v26.i20.2550
[2] Kittipibul V, Vutthikraivit W, Kewcharoen J, et al. Angiotensin II antagonists and gastrointestinal bleeding in left ventricular assist devices: A systematic review and meta-analysis. Int J Artif Organs. 2021;44(3):215-220. doi:10.1177/0391398820951811
[3] Converse MP, Sobhanian M, Taber DJ, Houston BA, Meadows HB, Uber WE. Effect of Angiotensin II Inhibitors on Gastrointestinal Bleeding in Patients With Left Ventricular Assist Devices. J Am Coll Cardiol. 2019;73(14):1769-1778. doi:10.1016/j.jacc.2019.01.051
[4] Hammer Y, Bitar A, Aaronson KD. Gastrointestinal bleeding on continuous-flow left ventricular assist device therapy. ESC Heart Fail. 2023;10(4):2214-2224. doi:10.1002/ehf2.14433
[5] Rai D, Tariq R, Tahir MW, et al. Primary and Secondary Prevention Strategies for Gastrointestinal Bleeding in Patients with Left Ventricular Assist Device: A Systematic Review and Network Meta-analysis. Curr Probl Cardiol. 2021;46(11):100835. doi:10.1016/j.cpcardiol.2021.100835

Literature Review

A search of the published medical literature revealed 6 studies investigating the researchable question:

What are medications that have data to prevent LVAD related GI bleeds?

Level of evidence

C - Multiple studies with limitations or conflicting results  Read more→



Please see Tables 1-6 for your response.


 

Association between digoxin use and gastrointestinal bleeding in contemporary continuous flow left ventricular assist device support

Design

Non-randomized, retrospective cohort study 

N= 649

Objective

To the association between digoxin use and gastrointestinal bleeding (GIB) in a multicenter continuous flow left ventricular assist device (LVAD) cohort

Study Groups

Digoxin (n= 213)

No digoxin (n= 436)

Inclusion Criteria

Adult patients who underwent first-time continuous flow LVAD implementation 

Exclusion Criteria

Patients with incomplete data regarding hospitalization for GIB over the follow-up period

Methods

Patients' demographic information and laboratory values were collected from their electronic medical records. At both studied institutions, all LVAD recipients had a target international normalized ratio (INR) of 2 to 3. HeartMate 2 and 3 patients receive aspirin 81 mg daily, and HeartWare patients receive 325 mg daily. Patients were considered to have received digoxin if it was listed on the medical chart between time of discharge to 3 months after LVAD implantation.

Negative binomial regression analyses were used to evaluate the association between digoxin use and the number of GIB events on LVAD support. 

Duration

June 2005 to July 2019

Follow-up: up to 2 years

Outcome Measures

Incidence of rate of GIB over 2 years 

Documented GIB: hospital admission for hematemesis, melena, or hematochezia requiring the transfusion of ≥ one unit of blood

Likely arteriovenous malformations (AVM) bleeding: AVMs were visualized or no other identifiable source of bleeding, or if patients experienced > 2 GIB events on LVAD support 

Baseline Characteristics

 

Digoxin (n= 213)

No digoxin (n= 436)

p-value

Age, years

57 ± 14 58 ± 113 0.482

Male

171 (80%) 349 (80%) 0.944

White

130 (61%) 306 (70%) 0.02

Underlying disease 

Ischemic cardiomyopathy

Diabetes

 

107 (50%)

97 (46%)

 

252 (58%)

175 (40%)

 

0.069

0.19

Device type 

HeartMate 2

HAVD

HeartMate 3

 

169 (79%)

18 (8%)

27 (12%)

 

340 (78%)

40 (9%)

56 (13%)

0.921

--

--

--

Device allocation

Bridge to transplant 

Destination therapy 

 

71 (33%)

142 (67%)

 

222 (51%)

213 (49%)

< 0.001

--

--

INTERMACS profile

1

2-3

4+

 

36 (17%)

121 (57%)

56 (26%)

 

55 (13%)

206 (47%)

175 (40%)

0.002

--

--

--

GI bleed

57 (27%)

123 (28%) 0.698

HVAD, HeartWare ventricular assist device; INTERMACS, Interagency Registry for Mechanically Assisted Circulatory Support

Results

Endpoint

Digoxin (n= 213)

No digoxin (n= 436)

Incidence rate ratio (95% confidence interval [CI]); p-value

Reduction in the incidence of rate of GIB over 2 years

Unadjusted 

Adjusted 

 


32% 

34%

 


--

--

 


0.68 (95% CI 0.46 to 0.99); p= 0.049

0.67 (95% CI 0.45 to 0.99); p= 0.048

Patients with likely arteriovenous malformation (AVM) associated GIB

Unadjusted 

Adjusted

 


52%

53%

 


--

--

 


0.48 (95 % CI 0.26 to 0.89); p= 0.02

0.47 (95% CI 0.25 to 0.9); p= 0.022

Patients without AVM GIB events

29% -- 0.71 (95% CI 0.48 to 1.13); p= 0.153

Adverse Events

N/A

Study Author Conclusions

In this multicenter study, inclusive of contemporary devices, digoxin use was associated with reduced GIB events. Prospective data will be required to confirm this association.

InpharmD Researcher Critique

Even though this study showed an association between digoxin use and decreased GIB incidence, prospective randomized trials need to be conducted to validate the correlation and potential causation. The timing of digoxin initiation/discontinuation and specific doses were also not well-defined in this study. 

References:

El Rafei A, Trachtenberg BH, Schultz J, et al. Association between digoxin use and gastrointestinal bleeding in contemporary continuous flow left ventricular assist device support. J Heart Lung Transplant. 2021;40(7):671-676. doi:10.1016/j.healun.2021.03.005

 

Digoxin Is Associated With a Decreased Incidence of Angiodysplasia-Related Gastrointestinal Bleeding in Patients With Continuous-Flow Left Ventricular Assist Devices

Design

Single-center retrospective study 

N= 199

Objective

To evaluate if digoxin use reduces the incidence of gastrointestinal angiodysplasia (GIAD) and gastrointestinal bleeding (GIB) in patients with continuous-flow left ventricular assist devices (CF-LVAD)

Study Groups

GIB (n= 54)

No GIB (n= 145)

Inclusion Criteria

Patients aged ≥18 years implanted with a CF-LVAD (HeartMate II [Thoratec, Pleasanton, CA] or Heartware VAD [HeartWare, Framingham, MA]) 

Exclusion Criteria

Not specified

Methods

A retrospective chart review with available electronic medical records was conducted to identify eligible patients. Patientt baseline characteristics were collected from the first outpatient follow-up visit after the implant hospitalization. Patients were categorized as digoxin group if they had received digoxin for at least 7 days after LVAD implant and continued at the time of discharge. Per institutional lVAD protocols, all patients received warfarin with goal INR 2–3 and aspirin 81–325 mg daily for HeartMate II or Heartware VAD. 

A backward variable selection method of the Cox proportional hazards model was used to control for potential confounders including demographic characteristics, underlying disease states, and use of concurrent medications. 

Duration

From February 2006 and February 2017

Outcome Measures

Primary outcome: time from LVAD insertion to nonsurgical bleeding

Other outcomes: frequency of GIB in each group and risk factors associated with GIB

GIB: clinical evidence of bleeding (guaiac-positive stool, melena, hematochezia, hematemesis, or the presence of blood in the GI tract on endoscopic evaluation, video capsule endoscopy, or bleeding scan) with a decrease in hemoglobin of ≥ 1 g/dL

GIB caused by GIAD: visualization of a GIAD on an endoscopic procedure or video capsule endoscopy without other known source 

GIB due to likely GIAD: GIB episodes without an identified source

Baseline Characteristics

 

GIB (n= 54)

No GIB (n= 145)

p-value

Mean age, years

60.6 ± 9.7

52.1 ± 13.8

0.0001

Male

39 (72%)

135 (24%)

0.46

White

26 (48%)

35 (24%) 0.001

HM II

45 (83%)

125 (86%)

0.61

Cardiovascular function 

Ischemic cardiomyopathy 

Ejection fraction pre-LVAD, %

Moderate RV failure post-LVAD

Severe RV failure post-LVAD

Acute severe RV failure post-LVAD

 

31 (57%)

21 ± 6

14 (26%)

13 (24%)

3 (6%)

 

49 (34%)

19 ± 6

57 (39%)

27 (19%)

4 (3%)

 

0.003

0.07 

0.18

--

--

Chronic disease

Diabetes mellitus 

CKD

Atrial fibrillation/flutter 

 

27 (50%)

9 (23%)

30 (52%)

 

50 (35%)

31 (77%)

68 (48%)

 

0.046

0.43

0.55

INR, unit

2.04 ± 0.41

2.02 ± 0.33

0.67

Pharmacological agents 

Digoxin

Aspirin

Warfarin

 

10 (16)

48 (89)

51 (94)

 

54 (36)

130 (90)

3 (2)

 

0.012

0.88

0.06

Median blood concentration of digoxin, ng/ mL (interquartile range [IQR]) 

0.80 (0.60 to 1.16)

--

CKD, chronic kidney disease; INR, international normalized ratio; and RV, right ventricular.

Results

Endpoint

GIB (n= 54)

No GIB (n= 145)

p-value

Meidan time to first GIB, days (interquartile range [IQR])

51 (31 to 63) 63 (22 to 164) --

GIB

GIAD/likely GIAD

Non-GIAD

GIAD-induced 

 

30%

7%

3%

 

78%

11%

13%

 

--

0.41

0.039

Univariate Cox Regression Analysis for Overall GIB: Variable

HR (95% CI); p-value

Digoxin

0.43 (0.22 to 0.86); 0.017

Multivariable-adjusted Cox Regression Analysis for Overall GIB: Variable

HR (95% CI); p-value

Digoxin

0.49 (0.24 to 0.98); 0.045

Age (per year)

1.03 (1.01 to 1.06); 0.019

Multivariable-adjusted Cox Regression Analysis for GIAD/Likely GIAD

HR (95% CI); p-value

Digoxin

0.18 (0.06 to 0.60); 0.005

Survival free of GIB was significantly higher (p= 0.022) in the digoxin group than non-digoxin group (specific number not reported in the figure). 

There was no significant difference in pump speed between the digoxin and the non-digoxin groups.

Propensity score analysis showed similar results in reduced GIB associated with digoxin HR 0.49 (95% CI: 0.22 to 1.08; p= 0.08) for overall GIB and HR 0.17 (95% CI: 0.05 to 0.58; p= 0.005) for GIAD/likely GIAD bleeding. 

Adverse Events

Common Adverse Events: incidence of thrombosis digoxin vs. non-digoxin group (14% vs. 17%; p= 0.31)

Study Author Conclusions

Use of digoxin was associated with a significant reduction in GIAD-related GIB in patients with CF-LVAD.

InpharmD Researcher Critique

The non-randomized observational nature of the study limits its use in clinical practice, and the results should only be considered hypothesis-generating. Whether the serum concentration of digoxin in this study is a safe and effective target in the general CF-LVAD populations needs to be further evaluated in prospective, randomized studies. 

 

 

References:

Vukelic S, Vlismas PP, Patel SR, et al. Digoxin is associated with a decreased incidence of angiodysplasia-related gastrointestinal bleeding in patients with continuous-flow left ventricular assist devices. Circulation: Heart Failure. 2018;11(8):e004899.

 

Impact of Heart Failure Drug Therapy on Rates of Gastrointestinal Bleeding In LVAD Recipients: An INTERMACS Analysis

Design

Pooled analysis

N= 13,732

Objective

To evaluate the effect of guideline-directed medical therapies (GDMT) for heart failure on rates of gastrointestinal bleeding (GIB) through analysis of the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) database

Study Groups

GI bleed (n= 2,674)

No GI bleed (n= 11,058)

Inclusion Criteria

Patients who were implanted with a left ventricular assist device (LVAD) between 2008 to 2017 who were alive at 3 months post-implant and receiving both warfarin and antiplatelet therapy

Exclusion Criteria

Pulsatile device

Methods

The INTERMACS was used to select patients implanted with an LVAD. Use of GDMT was determined based on prescription at 3 months after LVAD surgery.

Data were censored at death, transplantation, or device removal after myocardial recovery. Data was also censured at the time of the first GIB event. Univariable and multivariable Cox regression were used to control for confounding of each drug class on each other and additional variables including age, gender, renal function, heart failure etiology, and device strategy.

Duration

Follow-up: 2 years after device implantation

Outcome Measures

 GIB during follow-up (suspected or internal or external bleeding resulting in death, re-operation, hospitalization, or transfusion of red blood cells), effect of GDMT use on rates of GIB

Baseline Characteristics

 

GI bleed (n= 2,674)

No GI bleed (n= 11,058) p-value

Age, years

61.1 ± 10.4 54.5 ± 13.3 < 0.001

Female

541 (20.2%) 2,309 (20.9%) 0.73

Race

White

Ethnicity

Hispanic



1,804 (67.5%)

 

142 (5.3%)



7,409 (67%)

 

748 (6.8%)



0.647

 

0.008

Device type

LVAD

LVAD + RVAD



2,620 (98%)

54 (2%)



10,782 (97.5%)

276 (2.5%)

0.149

 

LVAD type

Centrifugal

Axial



396 (14.8%)

2,278 (85.2%)



2,590 (23.4%)

8,468 (76.6%)

< 0.001

 

 

INTERMACS profile

Profile 1

Profile 2

Profile 3

Profile 4-7



346 (13%)

964 (36.1%)

893 (33.5%)

465 (17.4%)



1,734 (15.7%)

3,927 (35.7%)

3,579 (32.5%)

1,770 (16.1%)

0.003

 

 

 

 

New York Heart Association class IV

2,068 (77.3%)

8,421 (76.2%)

0.198

Hemoglobin, g/dL

11.4 ± 2.15

11.1 ± 2.07

< 0.001

Medication use

ACE inhibitor or ARB

Beta-blocker

Mineralocorticoid receptor antagonist

Loop diuretic

Phosphodiesterase-5 inhibitor

Digoxin

Amiodarone



47.8%

64.7%

34.3%

74%

25.5%

18.4%

46.6%



52.7%

65.4%

39%

69.7%

22.7%

18.9%

43.1%

N/A

RVAD: right ventricular assist device; ACE: angiotensin-converting enzyme; ARB: angiotensin-receptor blocker

Results

Endpoint

Hazard ratio

95% confidence interval

p-value

Univariable analysis for association of medications with GIB on LVAD support

ACE inhibitor or ARB

Beta-blocker

Mineralocorticoid receptor antagonist

Loop diuretic

Phosphodiesterase-5 inhibitor

Digoxin

Amiodarone




0.81

0.88

0.85

1.19

1.19

0.97

1.19




0.75 to 0.87

0.81 to 0.95

0.79 to 0.92

1.09 to 1.3

1.08 to 1.31

0.87 to 1.07

1.1 to 1.28




< 0.001

< 0.001

< 0.001

< 0.001

0.001

0.552

< 0.001

Multivariable analysis for association of medications with GIB on LVAD support

ACE inhibitor or ARB

Beta-blocker

Mineralocorticoid receptor antagonist

Loop diuretic

Phosphodiesterase-5 inhibitor

Digoxin

Amiodarone




1

0.87

0.98

1.28

1.25

-

1.09




0.91 to 1.11

0.79 to 0.97

0.88 to 1.08

1.14 to 1.42

1.12 to 1.39

-

0.99 to 1.2




0.947

0.008

0.647

< 0.001

0.001

-

0.095

The overall incidence of GIB was 19.5% at 2 years of support, which included 44.5%, 27.3%, 24.3%, and 3.5% as upper, lower, occult, and multisite bleeding, respectively.

Adverse Events

N/A

Study Author Conclusions

This analysis suggests that receipt of most GDMT is not associated with the rate of GIB in LVAD with the possible exceptions for beta-blockers, loop diuretics, and phosphodiesterase-5 inhibitors. Additional study is warranted before any GDMT can be advocated to reduce the risk of GIB in patients on durable device support. Lastly, the results of the present analysis may form the basis for future clinical trials to further investigate the effects of these medications on rates of GIB after LVAD.

InpharmD Researcher Critique

Treatment assignment was not randomized; no causality between medication use and rates of GIB can be inferred. Additionally, confounding due to variation in indications for medication may be present, but regression analysis was used to control for certain confounding factors. Additionally, medication adherence was not accounted for, and treatment effects may have been confounded by polypharmacy.

Digoxin was not associated with GIB in univariate analysis; however, only the univariable analysis contained results for digoxin, which does not take into account relationships between the different variables.

 

References:

Jennings DL, Truby LK, Littlefield AJ, et al. Impact of heart failure drug therapy on rates of gastrointestinal bleeding in LVAD recipients: An INTERMACS analysis. Int J Artif Organs. 2021;44(12):965-971. doi:10.1177/03913988211013366

 

Omega-3 Therapy Is Associated With Reduced Gastrointestinal Bleeding in Patients With Continuous-Flow Left Ventricular Assist Device

Design

Retrospective

N=166 patients

Objective

To assess the prophylactic efficacy of treatment with omega-3 on the incidence of gastrointestinal bleeding (GIB) in patients with left ventricular assist devices (LVAD)

Study Groups

No omega-3 (n= 136)

Omega-3 (n= 30)

Methods

Utilized patients implanted with LVAD at a single study center from April 2014 and May 2017.

All patients received guideline-directed medical therapy, including antiplatelet therapy, oral proton pump inhibitor and warfarin with a target INR between 2.0 and 3.0 for HeartWare LVAD and between 2.0 and 2.5 for HeartMate II LVAD, unless recurrent GIB occurred.

Duration

1 year

Outcome Measures

Primary endpoint: Rates of GIB during the 1-year observational period 

Other important endpoints:freedom from GI bleeds and association of omega-3 with GI bleeds 

Baseline Characteristics

Baseline Characteristic

No Omega-3 (n= 136)

Omega-3 (n= 30)

     Age, y

58

49

     Male

111 (82%)

26 (87%)

     Body mass index

29.9

32.8

     LVAD duration before the    enrollment

20

25

     Destination therapy

119 (88%)

26 (87%)

Device

     Axial LVAD

83 (61%)

17 (57%)

     Centrifugal LVAD

53 (39%)

13 (43%)

     Temporary RVAD use

10 (7%)

2 (7%)

Comorbidity

     Ischemic Cause

50 (37%)

7 (23%)

     Hypertension

77 (57%)

18 (60%)

     Diabetes mellitus

50 (37%)

8 (27%)

     Atrial fibrillation

50 (37%)

10 (33%)

     History of stroke

25 (18%)

3 (10%)

     History of ventricular tachyarrhythmia

51 (38%)

10 (33%)

     Hemoglobin, g/dL

9.50±1.23

9.80±1.35

Results

 

None

Omega-3

p-value

1-year freedome from GIB

73%

93%

0.019

Frequency of GIB

0.08 ± 0.42 events/yr

0.37 ± 0.93 events/yr

0.010

# death

20 (17%)

0

N/A

Association between omega-3 and GI Bleeds Hazrad ratio (HR), 0.13 (95% CI, 0.02-0.98; p= 0.047)

Adverse Events

Common Adverse Events: none

Serious Adverse Events: none

Percentage that Discontinued due to Adverse Events: 0%

Study Author Conclusions

Omega-3 therapy was associated with significantly less GIB, as well as better clinical outcomes during LVAD support. A randomized placebo-controlled study and a biomarker study to approach the mechanism of omega-3 are warranted to support the current results.

InpharmD Researcher Conclusions

Association studies, such as this one, can describe a correlation between variables, but do not effectively elucidate a causative effect or the extent to which a therapy may be beneficial. Still, the use of omega-3 in this setting appears safe, but the extent of its benefit in this setting should be further evaluated in randomized, controlled trials.

References:

Imamura T, Nguyen A, Rodgers D, et al. Omega-3 Therapy Is Associated With Reduced Gastrointestinal Bleeding in Patients With Continuous-Flow Left Ventricular Assist Device [published correction appears in Circ Heart Fail. 2018 Nov;11(11):e000030. doi: 10.1161/HHF.0000000000000030.]. Circ Heart Fail. 2018;11(10):e005082. doi:10.1161/CIRCHEARTFAILURE.118.005082

 

Acid Suppression to Prevent Gastrointestinal Bleeding in Patients With Ventricular Assist Devices

Design

Observational cohort study

N= 138

Objective

To study the relationship between acid suppression regimens (PPI, H2RA, or neither) and gastrointestinal bleeding (GIB) after ventricular assist device (VAD) implantation

Study Groups

PPI (n= 48)

H2RA (n= 51)

No acid suppressive therapy (n= 39)

Inclusion Criteria

Patients who underwent VAD implantation between May 2010 and May 2014 and survived at least to the end of the 30-day postoperative study period

Exclusion Criteria

Age< 18 years, any solid organ transplant, use of direct oral anticoagulants before VAD implantation, temporary VAD duration of less than 7 days, or steady state use of both a PPI and H2RA during the study period

Methods

Patients were retrospectively identified via admission diagnosis or procedure code for VAD placement. Treatment exposure was defined as starting either a PPI or H2RA on or before postoperative day #4 and having this therapy continued for at least four doses to reach steady state. Multivariable logistic regression was used to compare treatment cohorts to GIB

Duration

May 2010 to May 2014

Outcome Measures

Primary: Incidence of GIB within 30 days after VAD implantation

Secondary: Timing, etiology, and location of GIB

Baseline Characteristics  

None (n= 39)

H2RA (n= 51) PPI (n= 48)

Age, years

56 ± 13 60 ± 12 59 ± 14

Male 

32 (82%) 43 (84%) 43 (90%)
BMI, median (IQR) 27 (23, 33) 27 (24, 30)

27 (24, 31)

APACHE II, median (IQR)

14 (11, 16) 18 (14, 22) 15 (13, 19)

Pre-existing GIB risk

10 (26%) 28 (55%) 42 (88%)
Results

Endpoint

None (n= 39) H2RA (n= 51) PPI (n= 48) p-value

GIB within 30 days

7 (18%) 7 (14%) 5 (10%) 0.6

Abbreviations: APACHE II= Acute Physiology and Chronic Health Evaluation II; BMI= body mass index; IQR= interquartile range

The study noted that GIB after VAD implantation varied in timing, location, and cause. However, in 37% of cases, endoscopic evaluation failed to identify the source or cause of bleeding. Among the cases where ulcers were identified as the cause, all were located beyond the stomach, indicating a distal origin.

Adverse Events

See results

Study Author Conclusions

Using PPI postoperatively in patients with new VAD was associated with a reduced incidence of GIB. Clinicians should consider the use of acid suppressive therapy for primary prevention.

Critique

The study was limited by its retrospective nature and the small number of GIB events, which restricted the number of covariates in the multivariate model. Despite these limitations, the study provides valuable insights into the potential benefits of PPI therapy in reducing GIB risk in VAD patients. 

References:

Hickman AW, Lonardo NW, Mone MC, et al. Acid Suppression to Prevent Gastrointestinal Bleeding in Patients With Ventricular Assist Devices. J Surg Res. 2019;234:96-102. doi:10.1016/j.jss.2018.08.023

 

Thalidomide Use Reduces Risk of Refractory Gastrointestinal Bleeding in Patients with Continuous Flow Left Ventricular Assist Devices
Design

Single-center, retrospective review

N= 17

Objective

To evaluate the efficacy of thalidomide in reducing the risk of recurrent GI bleeding from arteriovenous malformations (AVMs) in CF-LVAD patients

Study Groups

All patients (N= 17)

Inclusion Criteria

Patients who underwent HeartMate II or HeartWare HVAD CF-LVAD implantation at Stanford Hospital and Clinics and experienced GI bleeding events, with a trial both off and on thalidomide

Exclusion Criteria

Patients with insufficient follow-up after the initial bleeding event (<30 days) and those without an observation period off thalidomide before initiation

Methods

Retrospective review of CF-LVAD patients with GI bleeding from AVMs, treated with thalidomide. Thalidomide was initiated at 50 mg once or twice per day, with dose adjustments based on bleeding persistence and side effects. Octreotide was used as a bridge to thalidomide. 

Duration

January 1, 2010 through June 1, 2018

Outcome Measures

Primary: Time to first GI rebleeding event

Secondary: GI bleeding events per year, PRBC transfusion requirements per year, side effects/tolerability of thalidomide

Baseline Characteristics Characteristic

All patients (N= 17)

Age, years

65.1 ± 10.2

Male 

82.4%

White

52.9%

HeartMate II LVAD

82.4%

HeartWare LVAD

17.6%

LVAD as destination therapy

64.7%

Ischemic cardiomyopathy

52.9%

Melena

70.6%

Days after LVAD implantation until initial GI bleed (IQR)

46 (30-88)

Abbrevation: IQR= interquartile range

Results

Endpoint

Pre-thalidomide Post-thalidomide p-value
1-year bleed-free rate 

5.9%

57.7% 0.022
Median number of GI bleeds per year

4.6

0.4 0.0008
Median number of PRBC transfusions per year

36.1

0.9 0.004
Adverse Events

Dizziness (17.6%), peripheral neuropathy (11.8%), leukopenia (5.9%), hemolysis (17.6%), ischemic stroke (5.9%)

Study Author Conclusions

Thalidomide reduced the risk of AVM-associated GI rebleeding, number of bleeding events, and PRBC requirements in CF-LVAD patients. When initiating therapy, potential side effects and overall clinical context should be considered.

Critique

The study provides valuable insights into the use of thalidomide for reducing GI bleeding in CF-LVAD patients. However, its retrospective design and small sample size limit the generalizability of the findings. The use of octreotide as a bridge to thalidomide introduces a potential confounder, though it reflects a real-world approach to treatment. Further randomized controlled trials are needed to confirm these findings.

References:

Namdaran P, Zikos TA, Pan JY, Banerjee D. Thalidomide Use Reduces Risk of Refractory Gastrointestinal Bleeding in Patients with Continuous Flow Left Ventricular Assist Devices. ASAIO J. 2020;66(6):645-651. doi:10.1097/MAT.0000000000001054