Influence of sarcopenia on clinical outcomes in peptic ulcer bleeding: a retrospective single-center analysis in Korea
Article information
Abstract
Background/Aims
Peptic ulcer bleeding (PUB) is a major cause of nonvariceal gastrointestinal bleeding. As the older population increases, the prevalence of sarcopenia is also growing. Although sarcopenia influences the prognosis of various diseases, its association with clinical outcomes of PUB remains unknown. Therefore, this study aimed to explore this correlation.
Methods
We retrospectively analyzed 2,050 patients who underwent esophagogastroduodenoscopy for suspected gastrointestinal bleeding between January 2014 and December 2021. Patients who underwent computed tomography scans were included for sarcopenia evaluation based on the psoas muscle index, defined as the total psoas area normalized by the square of the height of the patient. Sarcopenia was defined using specific cutoffs: ≤7.3 cm2/m2 and ≤5.1 cm2/m2 for men and women, respectively. The primary outcome measured was the 30-day mortality rate.
Results
Out of 358 patients, 149 were diagnosed with sarcopenia. The 30-day mortality rate was significantly higher in patients with sarcopenia than in those without. Multivariate regression analysis revealed significant associations between sarcopenia, a high age, blood tests, and comorbidity score, and administration of inotropic agents with 30-day mortality.
Conclusions
Our study showed that the presence of sarcopenia, elevated comorbidity scores, and use of inotropes were associated with higher 30-day mortality rates. Considering that sarcopenia may influence the clinical outcomes in patients with PUB, it is crucial to manage patients with sarcopenia with particular care.
INTRODUCTION
Peptic ulcer bleeding (PUB) is a leading cause of nonvariceal upper gastrointestinal (GI) bleeding and can result in poor prognosis if not properly managed.1 Recent studies have shown a decline in the incidence of PUB owing to increased Helicobacter pylori eradication and proton pump inhibitor (PPI) use. However, the mortality rate remains high at 6% to 14%,2-4 which is attributed to the increased use of aspirin, nonsteroidal anti-inflammatory drugs, and antiplatelet agents, along with the aging population and rise in musculoskeletal and cardiovascular diseases.5-8
Due to the recent increase in life expectancy, the proportion of older patients with sarcopenia is also increasing. Sarcopenia is a progressive and generalized skeletal muscle disorder common among older individuals and is associated with multiple comorbidities.9,10 Sarcopenia is associated with various disease outcomes, including frailty and inflammatory bowel, cardiovascular, and liver diseases.11-18 Studies have shown that sarcopenia may lead to unfavorable clinical outcomes in various diseases.19-22 A recent study revealed that sarcopenia is associated with a higher prevalence of peptic ulcer disease.23 However, few studies have investigated the relationship between sarcopenia and the clinical outcomes of PUB.
Several factors and scoring systems have been used to predict clinical outcomes in patients with PUB. The Glasgow-Blatchford bleeding score (GBS) is designed to determine the need for clinical interventions such as transfusion, emergency endoscopy, or hospitalization and predict mortality in patients with acute upper GI bleeding.24 Based on the GBS, patients with a score of 0 can be safely managed as outpatients.25 The AIMS65 score (albumin level, international normalized ratio, mental status, systolic blood pressure, and age older than 65 years) is a useful tool for assessing the prognosis of upper GI bleeding. According to the AIMS65 criteria, patients with a score of 2 or more are considered high risk.26 The Rockall score is the most widely used scoring system and consists of age, shock, comorbidities, and evidence of bleeding. It effectively predicted the risk of rebleeding and mortality.27-29 Recently, age, blood tests, and comorbidity (ABC) score, designed to predict mortality in individuals with upper and lower GI bleeding, showed superior performance in predicting patient clinical outcomes compared with other scoring systems.30,31
Few studies have analyzed the relationship between sarcopenia and the clinical outcomes of PUB; however, none have evaluated the role or predictive accuracy of bleeding scoring systems in patients with sarcopenia. Therefore, this study aimed to investigate the correlation between sarcopenia and PUB outcomes and assess the efficacy of bleeding scores in these patients.
METHODS
Patients
We conducted a retrospective review of 2,050 patients who underwent esophagogastroduodenoscopy (EGD) for suspected GI bleeding at the emergency department of the Bundang Jesaeng General Hospital between January 2014 and December 2021. Patients without evidence of GI bleeding (n=661), small bowel bleeding (n=30), or lower GI bleeding (n=323) were excluded. Furthermore, we excluded patients with variceal bleeding (n=190), Mallory-Weiss syndrome (n=74), postprocedural bleeding (n=19), gastric angiodysplasia (n=13), marginal ulcer bleeding (n=9), cancer bleeding (n=85), or other bleeding foci (n=90). Other bleeding foci include bleeding from Zenker’s diverticulum, feeding tube-induced esophageal injury, esophageal tuberculosis, esophageal ulcer, Boerhaave syndrome, and hemorrhagic gastritis. Our center routinely performs enhanced computed tomography (CT) in patients with suspected bleeding to check for contrast extravasation in the GI tract. Among these patients, a CT scan was not performed if there were concerns regarding cost, kidney disease, and contrast allergy, or if the patient declined the examination. We excluded patients with PUB who did not undergo CT at presentation to the emergency department (n=198). Finally, our study included 358 patients with PUB who underwent CT (Fig. 1).
Flowchart of the study. EGD, esophagogastroduodenoscopy; GI, gastrointestinal; CT, computed tomography. a)Zenker’s diverticulum, feeding tube-induced mucosal injury, esophageal tuberculosis, esophageal ulcer bleeding, Boerhaave syndrome, and hemorrhagic gastritis.
We also reviewed the medical history of the patients, including medication history, underlying diseases, laboratory findings, and endoscopic findings such as the location (stomach and duodenum) and size of the peptic ulcer, Forrest classification, and hemostasis methods from electronic medical records.
Definition
An emergent EGD was performed based on the judgment of the on-call gastroenterologist responsible for endoscopic treatment within 24 hours, following established guidelines for upper GI bleeding. The decision considered clinical signs, such as melena, hematochezia, hematemesis, new onset or worsening anemia, and hemodynamic instability. All patients with suspected PUB received high-dose acid suppression therapy (80 mg esomeprazole or pantoprazole via intravenous bolus, followed by a continuous infusion at a rate of 8 mL/h for 72 hours).32 Treatment was based on physician judgment and tailored to the clinical condition of the patient. Bleeding was identified based on the signs of active bleeding, and endoscopic hemostasis was performed in patients with a Forrest classification of Ia, Ib, or IIa during endoscopy. The main bleeding focus site was anatomically determined to be the stomach and duodenum. Successful hemostasis was defined as effective bleeding control achieved at the end of the endoscopic procedure. In cases of hemostasis failure, we investigated the implementation of salvage therapies, such as transarterial embolization (TAE) and surgery. Rebleeding was defined as the recurrence of active bleeding with clinical manifestations such as melena, hematochezia, hematemesis, new onset or worsened anemia (a drop of >2 g/dL), and hemodynamic instability following the initial endoscopy. Rebleeding events were categorized based on their occurrence within 7 days or 30 days after the initial bleeding event. Mortality, measured as all-cause mortality, was classified into occurrence within 7 and 30 days after the initial bleeding event. The physical status of patients was assessed using the American Society of Anesthesiologists (ASA) classification system. The ASA classification includes ASA I (a normal healthy patient), ASA II (a patient with mild systemic disease), ASA III (a patient with severe systemic disease), and ASA IV (a patient with severe systemic disease that poses a constant threat to life).
Endoscopic findings
Ulcer size was classified into two groups: larger than 1 cm and smaller. Endoscopic hemostasis methods included hemoclips, argon plasma coagulation, heat probes, band ligation, injection, and combination therapy. Combination therapy was defined as using two or more endoscopic hemostasis methods.
After the initial endoscopy, patients were closely monitored and treated with PPI. Discharge criteria were based on the absence of rebleeding signs. If signs of rebleeding were observed, a second-look endoscopy was performed. If rebleeding was identified, the endoscopist applied appropriate endoscopic hemostasis techniques similar to those used initially. In the Forrest classification, Ia is defined as spurting bleeding, Ib as oozing bleeding, IIa as a non-bleeding visible vessel, IIb as an adherent clot, IIc as hematin on the ulcer base, and III as a clean ulcer base.
Bleeding scores
This study used several bleeding scores to assess the severity of PUB. Four bleeding scoring systems (GBS, AIMS65, Rockall, and ABC) were used. GBS assesses blood urea, hemoglobin, systolic blood pressure, pulse, melena, syncope, hepatic disease, and cardiac failure to determine the urgency of endoscopic therapy.26 The AIMS65 score includes albumin level, international normalized ratio (INR), systolic blood pressure, altered mental status, and age, effectively predicting mortality. The Rockall score consisted of age, systolic blood pressure, heart rate, comorbidities, endoscopic findings, and stigmata of recent hemorrhage. This study utilized the post-Rockall score, which integrates endoscopic findings.26 The ABC score includes the following factors: age, blood urea, albumin, creatinine, mental status, presence of liver cirrhosis, presence of disseminated malignancy, and an ASA score.30 Each score was defined and used based on established criteria.
Measurement of the psoas muscle index for definition of sarcopenia
We used abdominal CT imaging to measure the psoas muscle index (PMI) and defined sarcopenia according to the methods established in a previous study.33 An experienced radiologist (S.K.J.) retrospectively analyzed all CT scans on a standard desktop computer screen using INFINITT PACS (INFINITT Healthcare). In each study, we measured the cross-sectional areas of the right and left psoas muscles in the L4 vertebral body (Fig. 2). We then calculated the total psoas area by adding the measurements of both sides and normalizing them using the height of each patient, resulting in a PMI of cm2/m2.34 The definition of sarcopenia was based on previously validated criteria, with specific cutoffs of ≤7.3 cm²/m² and ≤5.1 cm²/m² for men and women, respectively.35
Psoas muscle area—yellow color delineates the right and left psoas muscle with the respective areas. (A, B) Patient without sarcopenia (C, D) Patient with sarcopenia.
Statistical analysis
Data were analyzed using R ver. 4.3.1 (The R Foundation for Statistical Computing). Continuous values are presented as mean and standard deviation. The area under the receiver operating characteristic (ROC) curve (AUC) was used to assess the predictive accuracy of bleeding scores. Multiple regression analysis was performed to identify independent predictors of 30-day mortality. Based on previous studies, variables considered in our analysis included mental changes; INR prolongation; comorbidities; high GBS, Rockall, AIMS65, and ABC scores; Forrest classification; incidence of rebleeding within 30 days, presence of sarcopenia, and use of inotropic agents.
IRB approval
This study was approved by the Institutional Review Board of the Bundang Jesaeng General Hospital (approval number: 2023-12-006).
RESULTS
Patients
Of 358 patients, 149 had sarcopenia. Regarding baseline characteristics, patients with sarcopenia were significantly older than those without (72.6 years vs. 62.6 years, p<0.001). Additionally, the use of inotropic medications or vasopressors was more common in patients with sarcopenia than in those without (9.4% vs. 2.9%, p=0.02). In terms of laboratory findings, patients with sarcopenia exhibited significantly lower hemoglobin levels (8.2 g/dL vs. 9.0 g/dL, p=0.01) and albumin (3.2 g/dL vs. 3.5 g/dL, p<0.001). They were more likely to have a higher ASA classification (III and IV) than those without sarcopenia (III: 70.5% vs. 59.8%, IV: 6.0% vs. 2.4%, p=0.02) (Table 1).
Baseline characteristics of patients with peptic ulcer bleeding categorized by the presence of sarcopenia
Endoscopic findings
Patients with sarcopenia had a significantly higher prevalence of gastric ulcers than those without (80.5% vs. 61.2%, p<0.001). Patients with sarcopenia had lesions in the gastric body more frequently than those without (47.7% vs. 28.7%, p<0.001). No significant differences were observed between the groups regarding Forrest classification, ulcer size, or hemostasis method. Although the hemostasis success rate was lower in the sarcopenia group, the difference was not statistically significant. (87.1% vs. 91.7%, p=0.47) (Table 2).
Endoscopic finding and methods of hemostasis in patients with peptic ulcer bleeding categorized by the presence of sarcopenia
Clinical outcomes
Bleeding rates at 7 and 30 days were lower in patients with sarcopenia than in those without; however, these differences were not statistically significant (3.4% vs. 7.2%, p=0.19; 7.4% vs. 10.5%, p=0.41). When categorizing the patients who experienced rebleeding within 30 days, patients with sarcopenia (n=149) had the following percentages of Forrest classifications: Ia, Ib, IIa, IIb, IIc, and III were 1 (14%), 2 (12%), 6 (14%), 1 (5%), 0 (0%), and 1 (3%), respectively. In contrast, patients without sarcopenia (n=209) showed percentages of Forrest classifications as follows: Ia (33%), Ib (17%), IIa (25%), IIb (8%), IIc (0%), and III (5%) (Fig. 3).
Number of patients experiencing rebleeding within 30 days, categorized based on Forrest classification and the presence of sarcopenia.
The 30-day mortality rate was significantly higher in patients with sarcopenia than in those without (4.0% vs. 0.0%, p=0.01). No significant difference in the rate of TAE as salvage therapy was observed between patients with and without sarcopenia (6.0% vs. 6.2%, p>0.99). Surgery was only performed in one patient with sarcopenia (Table 3).
Bleeding scores and clinical outcomes of patients diagnosed with peptic ulcer bleeding according to the presence of sarcopenia
Bleeding scores
In a study of 358 patients with PUB, those with sarcopenia (n=149) had significantly higher Rockall, GBS, and ABC scores than those without (n=209) (pre-Rockall, 4.3±1.6 vs. 3.4±1.8, p<0.001; post-Rockall, 5.4±2.0 vs. 4.7±2.0, p<0.001; GBS, 12.1±4.0 vs. 10.7±4.6, p=0.003; and ABC, 5.0±2.7 vs. 3.7 ± 2.6, p<0.001). Moreover, AIMS65 scores were higher in patients with sarcopenia than in those without; however, this difference was not statistically significant (2.4 vs. 2.3, p=0.199).
Among patients with sarcopenia, the mortality rate was 36.3% in those with an ABC score >10 (the cutoff point), compared with 1.4% in those with a score <10. Of the 92 patients with a GBS >12, 6.5% experienced mortality. Similarly, a mortality rate of 5.6% was observed in patients with an AIMS65 score >3. Seventy-three patients had a Rockall score >6 (the cutoff point) with a mortality rate of 6.8% (Fig. 4).
No. of patients with sarcopenia experiencing mortality within 30 days based on the bleeding score with optimal cut-off value. ABC, age, blood test, and comorbidities; GBS, Glasgow-Blatchford bleeding score; AIMS65, albumin level, international normalized ratio, mental status, systolic blood pressure, and age older than 65 years.
ROCs for bleeding scores
The ROCs for the bleeding scores (ABC, AIMS65, Rockall, and GBS scores) were analyzed to predict the 30-day mortality in the sarcopenia group. GBS ≥1, AIMS65 ≥2, Rockall score ≥5, and ABC score ≥8 predicted high-risk poor clinical outcomes (Supplementary Table 1).36 In patients with sarcopenia, the AUCs for the ABC, AIMS65, Rockall, and GBS scores were 0.82, 0.63, 0.72, and 0.82, respectively. The optimal cutoff value for the ABC score in predicting bleeding events was 10, demonstrating the highest sensitivity (66.7%) and specificity (96.6%) among the four bleeding scores. The AIMS65 score had a cutoff value of 3, yielding a sensitivity and specificity of 66.7% and 60.2%, respectively. The Rockall score, with a cutoff value of 6, showed sensitivity and specificity of 83.3% and 60.8%, respectively. The GBS, with a cutoff value of 12, demonstrated 100% specificity and an AUC of 0.74 (Fig. 5).
receiver operating characteristic curve (AUC) according to the bleeding scores. ABC, age, blood test, and comorbidities; AIMS65, albumin level, international normalized ratio, mental status, systolic blood pressure, and age older than 65 years; GBS, Glasgow-Blatchford bleeding score; PPV, positive predictive value; NPV, negative predictive value.
Multivariate regression analysis
Multivariate regression analysis revealed that a high ABC score (Beta=0.34, t=6.554, p<0.001), administration of inotropic medications or vasopressors (Beta=0.12, t=2.249, p=0.03), and the presence of sarcopenia (Beta=0.10, t=2.005, p=0.046) were significantly associated with an increased 30-day mortality risk (Table 4).
Multivariate regression analysis of the factors associated with 30 days mortality
DISCUSSION
Our study is the first to analyze the association between sarcopenia and PUB clinical outcomes using various scoring systems, including the Rockall, GBS, AIMS65, and ABC scores. This study revealed that sarcopenia was associated with an increased 30-day mortality rate in patients with PUB. Additionally, the ABC score significantly predicted 30-day mortality in the PUB group.
A previous study reported that survival was lower in patients with sarcopenia after endoscopic submucosal dissection for early-stage gastric cancer.37 Other studies have indicated that patients with liver cirrhosis experience higher rates of complications, including GI bleeding, and lower survival rates in the group with sarcopenia than in those without.16,38 In our study, mortality in patients with PUB was significantly higher in the sarcopenia group than in those without sarcopenia. This is likely because patients with sarcopenia had significantly higher ASA scores at baseline, were more frequently administered inotropic medications or vasopressors upon emergency department admission, and had more severe systemic conditions with lower hemoglobin levels. Further multicenter studies with larger patient cohorts are needed to assess the influence of sarcopenia on PUB.
Endoscopic hemostatic treatment is the primary therapeutic option for acute upper GI bleeding.39 Endoscopic hemostasis rates for PUB are typically approximately 90%, consistent with the findings of this study, which reported a rate of 89.9%.40 Previous studies showed that the skeletal muscle mass was significantly lower in patients with sarcopenia.41-43 This may be attributed to the presumed lower transmural thickness of the stomach muscle layer in the sarcopenia group, potentially contributing to the higher severity of PUB. In our study, we observed more patients with sarcopenia in the Forrest Ia group. This suggests that the mucosal damage from peptic ulcers may be more severe in patients with sarcopenia, potentially complicating adequate endoscopic hemostasis.
Notably, we observed six cases of 30-day mortality, all of which occurred in the sarcopenia group. GI bleeding was the direct cause of death in three cases, while in the remaining three cases, it indirectly exacerbated the underlying conditions, leading to a fatal outcome (Supplementary Table 2). Our findings suggest that all-cause mortality occurred exclusively in the sarcopenic group. It is noteworthy that mortality occurred only in patients with sarcopenia, and further prospective studies are needed to investigate this further.
The ABC score has recently emerged as a powerful prognostic tool, providing a more appropriate assessment of upper GI bleeding than established scoring systems such as AIMS65, the Rockall score, and GBS.30,31 In our study, the sarcopenia group exhibited a higher ABC score with an AUC of 0.82 and multivariate analysis showed a statistically significant association between an ABC score of ≥10 and increased 30-day mortality, with the ABC score being the only factor significantly affecting 30-day mortality. These findings indicate that an ABC score cutoff of 10 may be more effective in predicting outcomes in patients with sarcopenia than other bleeding scores. Although the optimal cutoff value for the ABC score was previously set to 8 in previous studies, our study identified it as 10 owing to the older age and higher prevalence of comorbidities in the sarcopenia group.30
Sarcopenia is defined as low muscle strength, quantity, quality, and physical performance.44 Studies showed that various methods can assess sarcopenia. First, muscle strength was assessed using simple and cost-effective techniques, such as grip strength and a calibrated handheld dynamometer.45-47 Another method for determining muscle strength is the chair stand test, which measures the time required for patients to rise five times from a seated position without using their arms.48,49 Muscle quantity or mass can be reported as total body skeletal muscle mass, appendicular skeletal muscle mass, or muscle cross-sectional area of specific muscle groups or body locations.44 Ultrasonography is another widely used technique for measuring muscle quantities. It demonstrates accuracy in elderly individuals with strong intra-and interobserver reliability.50 Dual-energy X-ray absorptiometry and bioelectrical impedance analysis have also been used for noninvasive muscle mass measurement.51,52 However, these methods are not easily applicable in the clinical setting for patients with PUB. Therefore, this study used CT to assess sarcopenia.
This study had some limitations. First, five endoscopists performed the endoscopic examinations, potentially introducing variations in techniques and skills that might have influenced the results. However, the gastroenterologists who conducted the endoscopy were all experienced endoscopists with over 10 years of expertise and adhered to the established procedural guidelines. Second, diagnosing sarcopenia requires evaluation of the physical performance of the patient. However, we defined sarcopenia solely based on CT quantification of the muscle. Therefore, future studies should incorporate diverse definitions of sarcopenia to explore this relationship further. Third, while we showed a significant association between sarcopenia and 30-day mortality, it is essential to recognize that the sarcopenia group was generally older and had more comorbidities, which could also influence clinical outcomes. Although we adjusted for several confounding factors, including age and comorbidities, residual confounding factors may remain. Age is a powerful predictor of poor outcomes in patients with PUB, and older patients may experience worse outcomes regardless of their sarcopenic status. Therefore, while performing propensity score matching based on the presence of sarcopenia would have been appropriate, this was not possible owing to the limited sample size of our study. Consequently, future multicenter studies with larger patient populations are warranted. Finally, as this was a single-center retrospective analysis, validation in a larger multicenter prospective study is required.
In conclusion, sarcopenia, high ABC scores, and inotropic administration were associated with increased 30-day mortality. Therefore, because sarcopenia may affect the clinical outcomes of patients with PUB, careful management is essential for patients with sarcopenia.
Supplementary Material
Supplementary Table 1. Abbreviation of bleeding score in gastrointestinal bleeding.
Supplementary Table 2. Causes of death in patients with sarcopenia.
Supplementary materials related to this article can be found online at https://doi.org/10.5946/ce.2024.209.
Notes
Conflicts of Interest
The authors have no potential conflicts of interest.
Funding
None.
Author Contributions
Conceptualization: JS; Data curation: JHY, HTL, JS; Formal analysis: JHY, JS; Investigation: JHY, HTL, JS; Methodology: JS; Project administration: JS; Resources: JS, SKJ; Supervision: JS; Validation: JS, AYL; Visualization: JHY, JS; Writing–original draft: JHY, HTL, JS; Writing–review & editing: all authors.