Society: AGA

Background: The 2021 ACG Guidelines suggested an infusion of intravenous (IV) erythromycin before endoscopy in patients with upper gastrointestinal bleeding (UGIB) as it improved gastric visualization during endoscopy and reduced in need for repeat endoscopy. However, IV erythromycin is not widely available. Evidence assessing IV metoclopramide, which is more accessible drug, is scant especially in patient with “active” UGIB. This study aimed to evaluate the efficacy of metoclopramide for gastric visualization in patients with active UGIB.
Methods: This double-blind, double center, randomized controlled trial study was conducted between April 2021 and October 2022. Patients with active UGIB (defined by either hematemesis or the presence of blood in nasogastric tube) were enrolled. We excluded patients with previous diagnosis of esophageal, duodenal, or gastric cancer or who had a gastric lavage of more than 50 mL. The included patients were randomly assigned to either metoclopramide or placebo in a 1:1 concealed allocation. The primary outcome was clear stomach which defined by endoscopic visualized gastric scores (EVS) 6 or more (total 8). The secondary outcomes included mean difference of EVS, duration of EGD, immediate hemostasis, need for a second look EGD, units of blood transfusion, length of hospital stay, and 30-day rebleeding rate.
Results: Sixty-eight eligible patients with active UGIB were randomized equally. Then, 4 patients were excluded due to protocol violation. Finally, 64 patients (metoclopramide 31, placebo 33) were analyzed. The baseline characteristics were not different between 2 groups. The Percentage of patient with clear stomach in the metoclopramide and placebo group was 77.4 % and 63.6% (p=0.22), respectively. The mean EVS was not different between two groups (6.6 vs 6.2, mean difference -0.4; p=0.37). The other secondary outcomes were not different between two assigned groups. In non-esophageal lesions subgroup analysis, the percentage of patient with clear stomach was significantly higher in metoclopramide compared to placebo group (88.9 % vs 57.9%, p=0.038)
Conclusion: Metoclopramide did not improve endoscopic gastric visualization in overall active UGIB lesions but significantly increased the clear stomach rate of those patients with non-esophageal lesions.

<b>Table1 Characteristics of patients at admission</b>

Table1 Characteristics of patients at admission

<b>Table2 Primary and secondary outcomes</b>

Table2 Primary and secondary outcomes

Introduction
Among patients with upper gastrointestinal bleeding (UGIB) who take aspirin, management of aspirin at the time of UGIB poses challenges. In a small randomized controlled trial, aspirin continuation in patients with UGIB increased recurrent UGIB risk but decreased mortality vs placebo. Overall, data related to aspirin management in the setting of UGIB are sparse. In a retrospective study, we characterized clinical practice patterns of aspirin management at the time of UGIB, and ascertained the risks of recurrent UGIB, major adverse cardiovascular events (MACE) [myocardial infarction (MI), stroke, transient ischemic attack (TIA)] or death depending on aspirin management.

Methods
We identified patients admitted for non-variceal, non-cirrhotic UGIB who were taking aspirin at our academic center from 2008-2022. We explored patterns of aspirin resumption (0-3d, 4-14d, 15-60d, and >60d/never). We contrasted aspirin management patterns among patients on aspirin for primary vs secondary prevention (defined by past MI, stroke or TIA). Outcomes were recurrent UGIB, MACE, or death within 60d. Pairwise comparisons between groups were performed using high-dimensional propensity score matching (hdPSM) and logistic regression.

Results
Among 1210 patients, 34% restarted aspirin within 0-3d, 7% within 4-14d, 6% within 15-60d, and 53% >60d/never (Table 1). Patients resuming aspirin within 0-3d were more likely to have history of MI (p=0.01) or CVA (p=0.02) (i.e., secondary prevention), and less likely to have malignancy (p=0.01) vs patients restarting aspirin >60d/never. Among patients taking aspirin for primary prevention, 73% had their aspirin restarted at >60d/never. Among patients taking secondary prevention aspirin, 36% had their aspirin restarted within 0-3d and 50% at >60d/never. The odds for recurrent UGIB were lower when aspirin was resumed >60d/never vs 4-14d (OR 0.43 [0.22-0.82]). Risk for MACE was lower with aspirin resumption at >60d/never vs 0-3d (OR 0.62 [0.44-0.86]) or 4-14d (OR 0.43 [0.23-0.79]) (Table 2). There were no differences in rates of TIA or death (Table 2) as a function of aspirin resumption timing.

Discussion
Among patients with UGIB taking aspirin for secondary cardiovascular prevention, only 1/3 resumed aspirin promptly, and 50% did not resume aspirin within 60d. This practice pattern deviates from recently published guidelines regarding secondary cardiovascular prevention during GIB. Our observed outcomes are not consistent with those of prior studies and were counterintuitive. These results suggest that despite hdPSM, unmeasured confounders probably exist. Baseline differences in history of MI suggest aspirin management may have been guided by clinical judgment regarding individual patient risk for GIB vs. CV complications. Future prospective research is needed to personalize aspirin management at the time of UGIB.

<b>Table 1</b><br /> P-values calculated using chi-squared test. CHF, congestive heart failure; MACE, major adverse cardiovascular events; MI, myocardial infarction; PUD, peptic ulcer disease; PVD, peripheral vascular disease; UGIB, upper gastrointestinal bleed.

Table 1
P-values calculated using chi-squared test. CHF, congestive heart failure; MACE, major adverse cardiovascular events; MI, myocardial infarction; PUD, peptic ulcer disease; PVD, peripheral vascular disease; UGIB, upper gastrointestinal bleed.

<b>Table 2</b><br /> Pairwise comparisons between various aspirin resumption times. Odds ratios calculated using high-dimensional propensity score matching and logistic regression. MACE, major adverse cardiovascular events; MI, myocardial infarction; TIA, transient ischemic attack; UGIB, upper gastrointestinal bleed.

Table 2
Pairwise comparisons between various aspirin resumption times. Odds ratios calculated using high-dimensional propensity score matching and logistic regression. MACE, major adverse cardiovascular events; MI, myocardial infarction; TIA, transient ischemic attack; UGIB, upper gastrointestinal bleed.

OBJECTIVE: We have previously published data on the use of IV bevacizumab for the management of refractory GI bleeding and severe anemia secondary to GAVE and/or small bowel angioectasia (SBA). We now present our initial experience with the use of oral low-dose Pazopanib in the same patient population. Pazopanib is a multiple tyrosine inhibitor with known anti-angiogenic properties.

METHODS: Oral low-dose pazopanib was prescribed to 27 patients from March 2021 to Dec 2022. We present data on 10 patients for whom blood transfusion, iron infusion, endoscopy and laboratory data was available for at least a 6 month duration pre and post pazopanib initiation.

RESULTS: Ten patients (50% female) with a median age of 72 (54-90) years were identified (table 1). Five (50%) patients had been previously treated with other antiangiogenic approaches and were switched to pazopanib (bevacizumab in 4, doxycycline in 1 and Bevacizumab + doxycycline in 1). Four patients received pazopanib at a dose of 200 mg daily throughout the follow up period. Three were started on 200 mg every other day of whom the dose was further reduced to 100 mg in two patients. Two patients were started on 100 mg daily and one patient was started on 200 mg five days/week. There were 4 (40%) patients on antiplatelet therapy and 5 (50%) patients on anticoagulation. Total transfusions in the 6 months before and after pazopanib appeared to decrease (18.5 to 14.7) with similar decreases in median endoscopy numbers (1.4 to 0.8). Median hemoglobin levels increased from 8.4 to 10.6 and ferritin levels increased from 139.7 to 185.2.

CONCLUSIONS: We report our initial results with the use of oral low-dose pazopanib for the management of refractory GI bleeding and severe anemia secondary to GAVE and/or SBA. Oral pazopanib appears to be a promising treatment alternative for these patients, but further follow up in a larger cohort of patients will be necessary to confirm the initial treatment benefit noted in this cohort.

Background
Dual anti-platelet therapy (DAPT) increases the risk of gastrointestinal bleeding (GIB) after coronary stenting. How often a GIB causes changes in P2Y12 inhibitor continuation remains unclear. Therefore, we sought to determine the changes to P2Y12 inhibitor management after coronary stenting following a GIB. Moreover, we aimed to derive a model to estimate the probability of a GIB at 180 days.

Methods
We identified patients who underwent coronary catheterization with stent placement and initiation of DAPT at a tertiary care academic center between January 2015 and June 2021. Demographic, laboratory, endoscopic and catheterization data were collected retrospectively. The primary outcome was hospitalization for GIB at 180 days. Multivariable logistic regression was performed for the primary outcome to derive a risk score. Accuracy of the model was evaluated by a calibration evaluation and the area under the receiver operating characteristic curve (AUROCC) using bootstrap resampling. Likelihood ratio tests (LRT) were performed to evaluate for a statistical difference among the multivariable models.

Results
529 patients were included. 168 (33.2%) had a GIB after coronary stenting at 180 days. Of these patients, 166 (98.8%) were on DAPT at the time of GIB and only 14 (8.4%) had their P2Y12 inhibitor discontinued at discharge. On univariable analysis, obesity, moderate alcohol use, acute coronary syndrome (ACS), proton pump inhibitor (PPI), and hemoglobin < 10 g/dL at catheterization were significantly associated with a GIB at 180 days. Figure 1. On multivariable analysis, a model of 6 predictors with the best discrimination was derived. The model, named POEMA-10 score, had a moderate accuracy to predict a GIB at 180 days, AUROCC: 0.664 (95% CI: 0.614-0.714). Addition of male gender, age>65 years, and pre-catheterization endoscopy did not change the predictive accuracy of the model, (LRT: p > 0.05). At the lowest sextile of the predicted probability of GIB, the observed rate of GIB was 17.6% compared to 50.5% at the highest sextile, (RR: 2.88, p<0.001). Stenting for ACS was the strongest predictor of a GIB, (p<0.001), and was associated with 95% increased odds of a GIB. After coronary stenting, 25% of patients had a GIB at 36, 41 or 159 days for NSTEMI, STEMI, or stable CAD, respectively, (p=0.0015). Figure 2.

Conclusion
We found most patients are continued on their P2Y12 inhibitor after a GIB at 180 days following coronary stenting. We derived the POEMA-10 score, which consists of 6 variables: [P]PI, [O]besity, [E]thanol, prior [M]ajor bleed, [A]CS, and hemoglobin < [10] g/dL at catheterization. This score had a moderate accuracy to predict a GIB at 180 days. Most importantly, our study suggests that coronary stenting for stable CAD before the manifestations of ACS could prevent the incidence of GIB.

Figure 1: Univariable Analysis and Cumulative Incidence of GIB at 180 days After Coronary Stenting

Figure 1: Univariable Analysis and Cumulative Incidence of GIB at 180 days After Coronary Stenting

Figure 2: Multivariable Logistic Regression Models and Calibration Evaluation of POEMA-10 Score

Figure 2: Multivariable Logistic Regression Models and Calibration Evaluation of POEMA-10 Score

Jennifer Zeng

Appearances