Low skeletal muscle mass predicts early mortality in cirrhotic patients with acute variceal bleeding

Accepted Manuscript Low skeletal muscle mass predicts early mortality in cirrhotic patients with acute variceal bleeding Yoji Ishizu, MD, PhD, Masatoshi Ishigami, MD, PhD, Teiji Kuzuya, MD, PhD, Takashi Honda, MD, PhD, Kazuhiko Hayashi, MD, PhD, Tetsuya Ishikawa, MD, PhD, Yoshiki Hirooka, MD, PhD, Hidemi Goto, MD, PhD PII:

S0899-9007(17)30115-6

DOI:

10.1016/j.nut.2017.06.004

Reference:

NUT 9975

To appear in:

Nutrition

Received Date: 27 February 2017 Revised Date:

4 May 2017

Accepted Date: 17 June 2017

Please cite this article as: Ishizu Y, Ishigami M, Kuzuya T, Honda T, Hayashi K, Ishikawa T, Hirooka Y, Goto H, Low skeletal muscle mass predicts early mortality in cirrhotic patients with acute variceal bleeding, Nutrition (2017), doi: 10.1016/j.nut.2017.06.004. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Low skeletal muscle mass predicts early mortality in cirrhotic patients with

2

acute variceal bleeding

4

Running head

5

Muscle mass and mortality in variceal bleeding patients

6

RI PT

3

Yoji Ishizu, MD, PhD 1, Masatoshi Ishigami, MD, PhD 1, Teiji Kuzuya, MD, PhD 1,

8

Takashi Honda, MD, PhD 1, Kazuhiko Hayashi, MD, PhD 1, Tetsuya Ishikawa,

9

MD, PhD 1, Yoshiki Hirooka, MD, PhD 1, Hidemi Goto, MD, PhD 1

SC

7

·

11

1. Department of Gastroenterology and Hepatology, Nagoya University

M AN U

10

12

Graduate School of Medicine, 65 Tsuruma-cho, Showa-ku, Nagoya,

13

466-8550 Japan

14

Role of each author in the work:

16

YI and MI were involved in the conception and design of the study. Generation,

17

collection, assembly, analysis and interpretation of data were handled by YI, MI,

18

TK, TH, KH, TI, YH and HG. Drafting or revision of the manuscript was the

19

responsibility of YI and MI. All the authors read and approved the final version of

20

the manuscript.

EP

21

TE D

15

A word count: 4160 words

23

Conflict of interest: None

24

Financial support: None

25

AC C

22

26

Corresponding Author:

27

Masatoshi Ishigami, MD, PhD

28

Department of Gastroenterology and Hepatology, Nagoya University Graduate

29

School of Medicine

30

65 Tsuruma-cho, Showa-ku, Nagoya 466-8550, Japan

ACCEPTED MANUSCRIPT 31

Tel:

+81-52-744-2169;

32

[email protected]

Fax:

+81-52-744-2178;

E-mail:

AC C

EP

TE D

M AN U

SC

RI PT

33

ACCEPTED MANUSCRIPT Abstract

35

Background & Aims: Low skeletal muscle mass adversely affects outcomes in

36

cirrhotic patients; however, its impact in patients with acute variceal bleeding

37

remains unknown. This study aimed to evaluate the impact of low skeletal

38

muscle mass on outcomes in cirrhotic patients with acute variceal bleeding.

39

Methods: A total of 122 patients were evaluated to identify factors associated

40

with two outcomes: failure to control bleeding, defined as either rebleeding or

41

death within 5 days, and 6-week mortality. Skeletal muscle mass was estimated

42

by calculating the psoas muscle area at the third lumbar vertebra on computed

43

tomographic images. Results: Forty-two patients had low skeletal muscle mass.

44

Fifteen patients had failure to control bleeding and 32 patients died within 6

45

weeks. Six of the patients with low skeletal muscle mass and nine of the patients

46

without low skeletal muscle mass had failure to control bleeding; these

47

proportions did not differ significantly (p=0.628). On the other hand, 15 (35.7%)

48

of the patients with low skeletal muscle mass died within 6 weeks; this proportion

49

was marginally higher than 17 (21.3%) of the patients without low skeletal

50

muscle mass who died within 6 weeks (p=0.084). On multivariate analysis,

51

presence of low skeletal muscle mass (odds ratio [OR] 4.69, p=0.024),

52

non-alcoholic etiology (OR 10.3, p=0.024), higher international normalized ratio

53

of prothrombin time (OR 41.4, p<0.001), and rebleeding within 6 weeks (OR

54

27.0, p<0.001) were associated with 6-week mortality. Conclusions: Low skeletal

55

muscle mass is an independent predictor of 6-week mortality in cirrhotic patients

56

with acute variceal bleeding.

57 58

AC C

EP

TE D

M AN U

SC

RI PT

34

ACCEPTED MANUSCRIPT 59 60

Key words: Low skeletal muscle mass, acute variceal bleeding, cirrhosis, early

61

mortality, portal hypertension

63

Highlights

64




SC




The 6-week mortality was significantly higher in patients with low muscle




M AN U

mass.

67 68

The 5-days rebleeding and mortality was unrelated to presence of low muscle mass.

65 66

RI PT

62

Liver failure was the main cause of death in patients with low muscle mass.

69

Abbreviations:

71

AVB, acute variceal bleeding; CT, computed tomography; EBL, endoscopic band

72

ligation; HCC, hepatocellular carcinoma; TIPS, transjugular intrahepatic

73

portosystemic shunt

EP AC C

74

TE D

70

ACCEPTED MANUSCRIPT Introduction

76

Acute variceal bleeding (AVB) is a fatal complication in patients with cirrhosis.

77

Although the introduction of endoscopic band ligation (EBL) has improved the

78

outcomes of such patients, the mortality rate is still about 20%.1 A further

79

reduction in mortality would require the identification and more aggressive

80

treatment of patients at high risk for early mortality. 2

81

In patients with cirrhosis, malnutrition, which is caused by poor dietary intake,

82

malabsorption, low protein synthesis and hypermetabolism, is often seen3 and

83

represented by decreased skeletal muscle mass.4 Recently, low skeletal muscle

84

mass, seen in an approximately half of all patients with cirrhosis,5 was shown to

85

be a predictor of mortality in cirrhotic patients under several conditions, such as

86

waiting for liver transplantation, post-liver transplantation,5 and bearing

87

hepatocellular carcinoma,6 however, the influence of low skeletal muscle mass

88

on cirrhotic patients with AVB remains unclear.

89

We studied the impact of low skeletal muscle mass on the outcomes of cirrhotic

90

patients with AVB treated by endoscopic treatment to attempt to identify high-risk

91

patients.

94

SC

M AN U

TE D

EP

AC C

92 93

RI PT

75

Methods

95

Patients

96

A total of 169 cirrhotic patients who underwent EBL for acute gastroesophageal

97

variceal bleeding in our department between April 2003 and April 2016 were

98

evaluated. We excluded 47 of these patients because they dropped out within 6

99

weeks after endoscopic treatment, had incomplete laboratory test data, did not

ACCEPTED MANUSCRIPT undergo computed tomographic (CT) examinations, or had a history of ABV

101

within the past 6 weeks. The remaining 122 patients were studied. Baseline

102

characteristics are shown in Table 1. The causes of cirrhosis were as follows:

103

heavy alcohol intake in 27 patients, hepatitis C virus infection in 56 patients,

104

hepatitis B virus infection in 14 patients, autoimmune hepatitis in four patients,

105

primary biliary cholangitis in two patients and miscellaneous in 19 patients.

106

Emergency endoscopy was performed as soon as possible, and if esophageal

107

varices were established to be the source of bleeding, EBL was performed using

108

an elastic band ligating device (Pneumoactivated EVL device, Sumitomo

109

Bakelite, Tokyo, Japan) at the same time. If the patient’s condition was good,

110

additional endoscopic treatment was performed every week until the varices

111

were eradicated. No patient received vasoconstrictive drugs, non-selective

112

β-blockers, or transjugular intrahepatic portosystemic shunt (TIPS) procedures

113

because they are not approved for the treatment of AVB in Japan. Although the

114

prophylactic use of antibiotics is suggested for cirrhotic patients with GI bleeding

115

in

116

recommendation is indicated as weak. 7 Therefore, prophylactic antibiotics were

117

not used routinely and antibiotic therapy was started when patients showed any

118

signs or symptoms suggesting infection. Immediate administration of anti-acid

119

therapy with a proton pump inhibitor or histamine H2-receptor antagonist after

120

EBL in all but one patient at the discretion of the physician in charge because

121

antacid treatment after endoscopic therapy for esophagogastric varices is

122

recommended in the Japanese guidelines.

123

transplantation within 6 weeks. Rebleeding was treated by additional EBL. The

124

diagnosis of liver cirrhosis was based on clinical, laboratory, radiological, and

Japanese

guidelines for

liver

cirrhosis,

the

strength

of

this

AC C

EP

the

TE D

M AN U

SC

RI PT

100

8

No patient underwent liver

ACCEPTED MANUSCRIPT histological data. Hepatocellular carcinoma (HCC) was diagnosed on the basis

126

of typical findings of arterial enhancement with subsequent washout on portal or

127

delayed phase on multidetector row CT scanning or magnetic resonance

128

imaging performed within 2 months before or after admission. This study was

129

approved by the Nagoya University Hospital Ethics Committee and was

130

conducted in accordance with the principles of the 1975 Declaration of Helsinki.

RI PT

125

SC

131

Data collection

133

Clinical and laboratory data were collected retrospectively. The severity of

134

cirrhosis was classified according to the MELD score. Skeletal muscle mass was

135

assessed on the basis of the cross-sectional areas of the right and left psoas

136

muscles at the middle of the third lumbar vertebra (L3). The areas were

137

measured by manual tracing on CT imaging obtained within 2 months before or

138

after treatment, which was performed to evaluate the development of

139

portosystemic collateral pathways, the presence of portal vein thrombosis, and

140

hepatocellular carcinoma. The muscle area (cm2) was divided by the body height

141

squared (m2) to normalize for height. We used 4.24 cm2/m2 for males and 2.50

142

cm2/m2 for females as the cut-off values for low skeletal muscle mass, which

143

were calculated by subtracting two standard deviations from the mean psoas

144

muscle area at the middle of the L3 level in healthy Japanese subjects younger

145

than 55 years, as reported in a previous study.9 The time to endoscopy was the

146

interval from initial assessment at the hospital to the time of initial endoscopy.

147

Hypovolemic shock was defined as the maintenance of a systolic blood pressure

148

of less than 90 mmHg and a heart rate of more than 100 beats per minute. To

149

evaluate the potential effects of the presence of HCC, the patients were divided

AC C

EP

TE D

M AN U

132

ACCEPTED MANUSCRIPT into two groups according to their prognosis:10 patients who had no HCC or were

151

within the Milan criteria (single tumor ≤5 cm, or maximum of three total tumors

152

with none >3 cm) and patients who exceeded the Milan criteria. Failure to control

153

bleeding was defined as meeting at least one of the following four criteria, in

154

accordance with the Baveno consensus workshops: any further clinically

155

significant bleeding after therapeutic endoscopy, a 3-g/dL drop in hemoglobin

156

levels within 24 hours if no transfusion was administered, development of

157

hypovolemic shock, and death within the first 5 days after initial endoscopic

158

treatment. Early mortality was defined as death occurring within 6 weeks after

159

endoscopic treatment.2,

160

presence of further bleeding or a 3-g/dL drop in hemoglobin level within 24 hours

161

during the first 6 weeks after initial endoscopic treatment.

11

M AN U

SC

RI PT

150

We defined rebleeding within 6 weeks as either

162

Statistical analysis

164

Quantitative variables were compared with the use of Student’s t-test or the

165

non-parametric Mann-Whitney U test, and the distributions of qualitative

166

variables were compared using the chi-square test or Fisher’s exact test, as

167

appropriate. We used one-way analysis of variance and the Tukey post hoc test

168

or Kruskal-Wallis test to evaluate differences in the proportions of low skeletal

169

muscle mass among the following three groups: patients with a MELD score of

170

less than 10, 10 to 14, or 15 and higher. Multiple logistic regression analysis was

171

performed to determine the factors that contributed significantly to 6-week

172

mortality. Variables with p values of less than 0.1 on univariate analysis were

173

included in the model.12 To avoid collinearity, the MELD score was removed from

174

the analysis, and its individual components (i.e., serum bilirubin level, serum

AC C

EP

TE D

163

ACCEPTED MANUSCRIPT creatinine level, and international normalized ratio of prothrombin time) were

176

incorporated into multivariate analysis. P values of less than 0.05 were

177

considered to indicate statistical significance. Statistical analyses were

178

performed using SPSS software (SPSS, Inc., Chicago, IL, USA).

RI PT

175

179

Results

181

Prevalence of low skeletal muscle mass in cirrhotic patients with AVB

182

The prevalence of low skeletal muscle mass was 34.4% (42 of 122 patients) in

183

cirrhotic patients with AVB. The prevalence of low skeletal muscle mass

184

according to the MELD score was 29.2% (7 of 24 patients) in patients with a

185

score of less than 10, 35.6% (16 of 45 patients) in patients with a score of 10 to

186

14, and 35.8% (19 of 53 patients) in patients with a score of 15 or higher. These

187

prevalences did not differ significantly (p=0.832). Normalized skeletal muscle

188

area (cm2/m2) also did not differ according to the MELD score (p=0.416).

M AN U

TE D

189

SC

180

Factors associated with failure to control bleeding

191

Fifteen patients met the criteria for failure to control bleeding. Nine of these

192

patients had further clinically significant bleeding, and six patients died within 5

193

days. The causes of death were liver failure in five patients and hypovolemic

194

shock in one patient. Six (14.3%) of the 42 patients with low skeletal muscle

195

mass and nine (11.3%) of the 80 patients without low skeletal muscle mass had

196

failure to control bleeding. These proportions did not differ significantly (p=0.628).

197

Because the limited number of patients with failure to control bleeding precluded

198

a meaningful multivariate analysis, only univariate analysis was performed.

199

Three factors were related to failure to control bleeding: higher serum bilirubin

AC C

EP

190

ACCEPTED MANUSCRIPT 200

level (p=0.033), lower serum sodium level (p<0.001), and presence of an

201

advanced stage of HCC (p=0001) (Table 2).

202

Factors associated with mortality within 6 weeks

204

Thirty-two patients died within 6 weeks. Of the 42 patients with low skeletal

205

muscle mass, 15 (35.7%) died within 6 weeks; this proportion was marginally

206

higher than 17 (21.3%) of the 80 patients without low skeletal muscle mass

207

(p=0.084) who died within 6 weeks (Table 3). Univariate analysis identified

208

another seven factors that significantly or marginally influenced 6-week mortality:

209

non-alcoholic etiology (p=0.043), higher serum bilirubin level (p<0.001), lower

210

serum albumin level (p<0.001), higher international normalized ratio of

211

prothrombin time (p<0.001), higher serum creatinine level (p<0.001), lower

212

serum sodium level (p<0.001), presence of an advanced stage of HCC

213

(p=0.001) and occurrence of rebleeding within 6 weeks (p<0.001). On

214

multivariate analysis, non-alcoholic etiology (OR 10.3, 95% CI 1.36-76.9,

215

p=0.024), higher international normalized ratio of prothrombin time (OR 41.4,

216

95% CI 6.79-253, p<0.001), occurrence of rebleeding within 6 weeks (OR 27.0,

217

95% CI 5.54-131, p<0.001), and presence of low skeletal muscle mass (OR 4.69,

218

95% CI 1.23-17.9, p=0.024) were identified as factors significantly related to

219

increased 6-week mortality (Table 3).

SC

M AN U

TE D

EP

AC C

220

RI PT

203

221

Causes of death within 6 weeks in patients with or without low skeletal

222

muscle mass

223

Of the 32 patients who died within 6 weeks, 15 had low skeletal muscle mass.

224

Two-thirds of the deaths were caused by liver failure, and one patient died of

ACCEPTED MANUSCRIPT infection associated with aspiration pneumonia. Seventeen patients without low

226

skeletal muscle mass died within 6 weeks, and of them, 11 patients died of liver

227

failure, five died of hypovolemic shock, and one died of cytomegalovirus

228

infection, which had already occurred before variceal bleeding (Table 4).

RI PT

225

229

Discussion

231

This study showed that low skeletal muscle mass is one of the significant

232

independent predictors of 6-week mortality in cirrhotic patients who have AVB

233

after emergency ligation therapy.

M AN U

SC

230

234

Despite recent advances in treatment for AVB enabled by the development of

236

pharmacologic and band ligation therapy, approximately one in five patients died

237

within 6 weeks. More aggressive therapy, such as early TIPS, might improve the

238

outcomes of these patients;13 however, the implementation of such therapy in all

239

hospitals around the world be difficult. Therefore, it is necessary to elucidate risk

240

factors that could be used to identify high-risk patients who need to be

241

transferred to facilities where aggressive medical care can be provided. Previous

242

studies identified several factors that influence early morality in cirrhotic patients

243

with ABV: worsening liver function, 14, 15, 16, 17, 18, 19, 20 impaired renal function,18, 19

244

non-alcoholic etiology,17, 18 rebleeding,18 the presence of HCC,19 and infection.18,

245

19

246

function;18, 20 however, there is still room for improvement. Over a period of about

247

10 years, many studies have shown that low skeletal muscle mass is associated

248

with mortality in cirrhotic patients;5 however, the impact of low skeletal muscle

249

mass in cirrhotic patients with ABV has not been adequately investigated. Our

AC C

EP

TE D

235

Some authors have proposed good prognostic models based on liver

ACCEPTED MANUSCRIPT study showed that presence of low skeletal muscle mass is a significant

251

independent predictor of 6-week mortality, along with three other previously

252

reported factors, including non-alcoholic etiology, worsening liver function, and

253

occurrence of rebleeding within 6 weeks. Our results suggest that evaluation of

254

both liver function and presence or absence of low skeletal muscle mass might

255

allow risk in individual patients to be classified more accurately than previously

256

reported.

SC

RI PT

250

257

On the other hand, presence of low skeletal muscle did not influence

259

development of failure to control bleeding. Since failure to control bleeding

260

includes two outcomes: death and rebleeding, we further analyzed the impact of

261

low skeletal muscle mass on each of them, however, presence of low skeletal

262

muscle mass is not associated with both death and rebleeding within 5 days

263

(p=0.339, p=0.943, respectively). Whereas, low sodium levels have been

264

identified as a factor associated with failure to control bleeding in addition to

265

previously proven factors, such as worsening liver function14,

266

presence of HCC.5 Decreased serum sodium level are well known to negatively

267

affect outcomes in patients with severe cirrhosis,21 and we obtained consistent

268

results at a very early phase in patients with AVB.

TE D

and

EP

15, 16, 17

AC C

269

M AN U

258

270

The reason why low skeletal muscle mass negatively affected outcomes

271

remains unclear. Low skeletal muscle mass can be caused by several factors:

272

malnutrition due to inadequate dietary intake and malabsorption, playing major

273

roles in the development of skeletal muscle, metabolic disturbances, systemic

274

inflammation, and inadequate hormone production.22 Based on the hypothesis

ACCEPTED MANUSCRIPT that patients with low skeletal muscle mass are prone to infections because of

276

an impaired immune system due to malnutrition,23 several studies have

277

investigated the causes of death in such patients; however, the results were

278

controversial. In postsurgical patients, presence of a low skeletal muscle mass

279

was associated with a higher incidence of fatal infection irrespective of the

280

presence or absence of cirrhosis.24,

281

cirrhosis who did not undergo surgical procedures, low skeletal muscle mass

282

was not related to death due to infection.6, 26 In our study, although patients with

283

AVB had a high risk of fatal bacterial infection,27 the main cause of death in

284

patients with a low skeletal muscle mass was liver failure, and only one patient

285

died of infection due to aspiration pneumonia.

On the other hand, in patients with

M AN U

SC

25

RI PT

275

286

Our study had two major limitations. First, it was a retrospective study performed

288

in a single institution. Second, vasoactive therapy, which is recommended for

289

cirrhotic patients with AVB by several guidelines,2, 28 was not performed in this

290

study because its use is not approved for the treatment of AVB in Japan.

291

However, the rate of rebleeding and mortality in our study was nearly the same

292

as that in a previous study in which patients received standard vasoactive

293

therapy.29 These results are most likely attributed to two factors: very early

294

endoscopy and the administration of anti-acid drugs. Early endoscopy, defined

295

as endoscopy performed within 12 hours or less, was associated with decreased

296

rebleeding and mortality within 6 weeks in a previous study.30 In our study, the

297

mean time to endoscopy was only about 2 hours. Another study reported that

298

patients who received combination therapy with anti-acid drugs and EBL had

299

similar 6-week mortality and rebleeding rates to those who received standard

AC C

EP

TE D

287

ACCEPTED MANUSCRIPT 300

therapy with vasoactive drugs and EBL.31 To confirm the results of our study, a

301

prospective multicenter study including a control group given standard therapy is

302

needed.

RI PT

303

In conclusion, the presence of a low skeletal muscle mass is significantly and

305

independently related to increased 6-week mortality among cirrhotic patients

306

with ABV treated by EBL. Further prospective studies are needed to confirm our

307

results.

SC

304

AC C

EP

TE D

M AN U

308

ACCEPTED MANUSCRIPT 309

References [1] Augustin S, Altamirano J, González A, et al. Effectiveness of combined

311 312

pharmacologic and ligation therapy in high-risk patients with acute esophageal variceal bleeding. Am J Gastroenterol. 2011; 106: 1787-95.

313

[2] de Franchis R, Faculty BV. Expanding consensus in portal hypertension:

314

Report of the Baveno VI Consensus Workshop: Stratifying risk and individualizing care for portal hypertension. J Hepatol. 2015; 63: 743-52.

315

RI PT

310

318

[4] Moriwaki H, Miwa Y, Tajika M, Kato M, Fukushima H, Shiraki M.

319

Branched-chain amino acids as a protein- and energy-source in liver cirrhosis. Biochem Biophys Res Commun. 2004; 313: 405-9.

M AN U

320

SC

317

[3] O'Brien A, Williams R. Nutrition in end-stage liver disease: principles and practice. Gastroenterology. 2008; 134: 1729-40.

316

323

[6] Meza-Junco J, Montano-Loza AJ, Baracos VE, et al. Sarcopenia as a

324

prognostic index of nutritional status in concurrent cirrhosis and hepatocellular carcinoma. J Clin Gastroenterol. 2013; 47: 861-70.

325 326

TE D

322

[5] Kim HY, Jang JW. Sarcopenia in the prognosis of cirrhosis: Going beyond the MELD score. World J Gastroenterol. 2015; 21: 7637-47.

321

[7] Fukui H, Saito H, Ueno Y, et al. Evidence-based clinical practice guidelines

328

for liver cirrhosis 2015. J. Gastroenterol. 2016, 51, 629–650.

330 331 332 333

[8] The Japanese Society of Gastroenterology. Evidenced-based Clinical Practice Guidelines for Liver Cirrhosis 2015, 2nd ed. Tokyo: Nankodo; 2015, p.77 (written in Japanese).

AC C

329

EP

327

334

[9] Hiraoka A, Aibiki T, Okudaira T, et al. Muscle atrophy as pre-sarcopenia in

335

Japanese patients with chronic liver disease: computed tomography is useful for evaluation. J Gastroenterol. 2015; 50: 1206-13.

336 337 338 339

[10] Sangiovanni A, Colombo M. Treatment of hepatocellular carcinoma: beyond international guidelines. Liver Int. 2016; 36 Suppl 1: 124-9.

ACCEPTED MANUSCRIPT 340

[11] de Franchis R, Faculty BV. Revising consensus in portal hypertension:

341

report of the Baveno V consensus workshop on methodology of diagnosis and therapy in portal hypertension. J Hepatol. 2010; 53: 762-8.

342

345

[13] García-Pagán JC, Caca K, Bureau C, et al. Early use of TIPS in patients

346

with cirrhosis and variceal bleeding. N Engl J Med. 2010; 362: 2370-9.

347

[14] Vanbiervliet G, Giudicelli-Bornard S, Piche T, et al. Predictive factors of

348 349

bleeding related to post-banding ulcer following endoscopic variceal ligation in cirrhotic patients: a case-control study. Aliment Pharmacol Ther. 2010; 32:

350

225-32.

351

353

[15] Bambha K, Kim WR, Pedersen R, Bida JP, Kremers WK, Kamath PS. Predictors of early re-bleeding and mortality after acute variceal haemorrhage in patients with cirrhosis. Gut. 2008; 57: 814-20.

354

[16] de Souza AR, La Mura V, Berzigotti A, García-Pagán JC, Abraldes JG,

355 356

Bosch J. Prognosis of acute variceal bleeding: Is being on beta-blockers an aggravating factor? A short-term survival analysis. Hepatology. 2015; 62:

357

1840-6.

358

[17] Abraldes JG, Villanueva C, Bañares R, et al. Hepatic venous pressure

359 360

gradient and prognosis in patients with acute variceal bleeding treated with pharmacologic and endoscopic therapy. J Hepatol. 2008; 48: 229-36.

361

[18] Augustin S, Muntaner L, Altamirano JT, et al. Predicting early mortality

362 363

after acute variceal hemorrhage based on classification and regression tree analysis. Clin Gastroenterol Hepatol. 2009; 7: 1347-54.

364

[19] Hung TH, Tsai CC, Tseng CW, Tseng KC, Hsieh YH. No difference in

365 366

mortality between terlipressin and somatostatin treatments in cirrhotic patients with esophageal variceal bleeding and renal functional impairment. Eur J

367

Gastroenterol Hepatol. 2016; 28: 1275-9.

368

[20] Reverter E, Tandon P, Augustin S, et al. A MELD-based model to

369

determine risk of mortality among patients with acute variceal bleeding.

SC

M AN U

TE D

EP

AC C

352

RI PT

344

[12] Lang T. Documenting research in scientific articles: Guidelines for authors: 3. Reporting multivariate analyses. Chest. 2007; 131: 628-32.

343

ACCEPTED MANUSCRIPT Gastroenterology. 2014; 146: 412-19.e3.

371 372

[21] Kim WR, Biggins SW, Kremers WK, et al. Hyponatremia and mortality among patients on the liver-transplant waiting list. N Engl J Med. 2008; 359:

373

1018-26.

374

[22] Sinclair M, Gow PJ, Grossmann M, Angus PW. Review article: sarcopenia

375 376

in cirrhosis-aetiology, implications and potential therapeutic interventions. Aliment Pharmacol Ther. 2016; 43: 765-77.

377

[23] Scrimshaw NS, SanGiovanni JP. Synergism of nutrition, infection, and

378

immunity: an overview. Am J Clin Nutr. 1997; 66: 464S-77S.

379 380

[24] Masuda T, Shirabe K, Ikegami T, et al. Sarcopenia is a prognostic factor in living donor liver transplantation. Liver Transpl. 2014; 20: 401-7.

381

[25] Lieffers JR, Bathe OF, Fassbender K, Winget M, Baracos VE. Sarcopenia

382 383

is associated with postoperative infection and delayed recovery from colorectal cancer resection surgery. Br J Cancer. 2012; 107: 931-6.

384

[26] Hanai T, Shiraki M, Nishimura K, et al. Sarcopenia impairs prognosis of

385

patients with liver cirrhosis. Nutrition. 2015; 31: 193-9.

386 387

[27] Lee YY, Tee HP, Mahadeva S. Role of prophylactic antibiotics in cirrhotic patients with variceal bleeding. World J Gastroenterol. 2014; 20: 1790-6.

388

[28] Garcia-Tsao G, Sanyal AJ, Grace ND, Carey W. Prevention and

389 390

management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology. 2007; 46: 922-38.

391

[29] Amitrano L, Guardascione MA, Manguso F, et al. The effectiveness of

392 393

current acute variceal bleed treatments in unselected cirrhotic patients: refining short-term prognosis and risk factors. Am J Gastroenterol. 2012; 107: 1872-8.

394

[30] Chen PH, Chen WC, Hou MC, et al. Delayed endoscopy increases

395 396

re-bleeding and mortality in patients with hematemesis and active esophageal variceal bleeding: a cohort study. J Hepatol. 2012; 57: 1207-13.

397

[31] Lo GH, Perng DS, Chang CY, Tai CM, Wang HM, Lin HC. Controlled trial

AC C

EP

TE D

M AN U

SC

RI PT

370

ACCEPTED MANUSCRIPT 398 399

of ligation plus vasoconstrictor versus proton pump inhibitor in the control of acute esophageal variceal bleeding. J Gastroenterol Hepatol. 2013; 28: 684-9.

AC C

EP

TE D

M AN U

SC

RI PT

400

ACCEPTED MANUSCRIPT 401

Table 1. Baseline characteristics

402

61.6±12.8 92/30 27/95 60/62

Serum bilirubin (mg/dL) Serum albumin (g/dL) Prothrombin time (INR) Serum creatinine (mg/dL)

3.00±4.04 2.59±0.584 1.50±0.416 1.16±1.01

Serum sodium (mEq/L) Platelet count (×109/L) Low skeletal muscle mass (present/absent)

136±4.64 119±63.8 42/80

SC 121±110 70/52 12/110 80/42

M AN U

Time to endoscopy † (min) Active bleeding (present/absent) Hypovolemic shock ‡ (present/absent) HCC (absent or early stage §/advanced stage)

RI PT

Age (years) Gender (male/female) Etiology of alcohol/others Past history of bleeding (with/without)

403

408

‡ Systolic blood pressure <90 mmHg and heart rate >100 beat per minute

409

§ Single <5 cm or ≤3 nodules <3 cm

410

EP

406

AC C

405

TE D

407

Data are presented as means ± standard deviation. INR, international normalized ratio; HCC, hepatocellular carcinoma † The interval from initial assessment at the hospital to the time of initial endoscopy

404

ACCEPTED MANUSCRIPT

RI PT

Table 2. Factors associated with failure to control bleeding Bleeders (n=15)

Non-bleeders ( n =107)

Age (years) Gender (male/female) Etiology of alcohol/others Past history of bleeding (with/without) Serum bilirubin (mg/dl)

64.3±10.6 12/3 1/14 8/7 3.70 (1.55-5.50)

61.2±13.1 80/27 26/81 52/55 1.50 (0.950–2.65)

0.387 0.469 0.123 0.731 0.033

Serum albumin (g/dl) Prothrombin time (INR) Serum creatinine (mg/dl) Serum sodium (mEq/l)

2.36±0.549 1.44 (1.35-2.02) 0.990 (0.755-1.42) 131 (129-135)

2.62±0.584 1.40 (1.21-1.56) 0.85 (0.645-1.27) 137 (134-139)

0.106 0.137 0.364 <0.001

Platelet count (×109/L) Active bleeding (present/absent) Low skeletal muscle mass (present/absent)

136 (88.5-193) 11/4

96.0 (78.5-139) 59/48

0.053 0.182

6/9

36/71

0.628

Hypovolemic shock † (present/absent) HCC (absent or early stage ‡ / advanced stage)

1/14

11/96

0.549

76/31

0.001

M AN U

TE D

EP

AC C 4/11

SC

Variable

p

Data are presented as means ± standard deviation or medians (interquartile range). INR, international normalized ratio; HCC, hepatocellular carcinoma † Systolic blood pressure <90 mmHg and heart rate >100 beat per minute

value

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

‡ Single <5 cm or ≤3 nodules <3cm

ACCEPTED MANUSCRIPT

RI PT

Table 3. Factors associated with death within 6 weeks Univariate analysis The survivors ( n =90)

Age (years) Gender (male/female) Etiology of alcohol/others

63.0±9.22 26/6 3/29

61.1±13.9 66/24 24/66

p value 0.380 0.372 0.043

13/19

47/43

0.260

3.00 (1.55-7.23) 2.24±0.436

1.30 (0.900-2.40) 2.71±0.582

<0.001 <0.001

Prothrombin time (INR) Serum creatinine (mg/dl) Serum sodium (mEq/l) Platelet count (×109/L)

1.65 (1.44-2.05) 1.31 (0.832-1.89) 133 (129-135) 127 (78.5-149)

1.34 (1.20-1.46) 0.800 (0.620-1.04) 137 (134-139) 94.0 (79.0-137)

<0.001 <0.001 <0.001 0.236

20/7

50/30

0.495

27/63

0.084

7/83

0.173

67/23

0.001

AC C

EP

TE D

Past history of bleeding (with/without) Serum bilirubin (mg/dl) Serum albumin (g/dl)

Active bleeding (present/absent) Low skeletal muscle mass (present/absent)

15/17

Hypovolemic shock † 5/27 (present/absent) HCC (absent or early stage ‡/ 13/19

OR

95%CI

p value

10.3

1.36-76.9

0.024

41.4

6.79-253

<0.001

4.69

1.23-17.9

0.024

SC

The deceased ( n =32)

M AN U

Variable

Multivariate analysis

ACCEPTED MANUSCRIPT

19/13

10/80

<0.001

27.0

RI PT

advanced stage) Rebleeding within 6 weeks (present/absent)

5.54-131

SC

Data are presented as means ± standard deviation or medians (interquartile range). INR, international normalized ratio; HCC, hepatocellular carcinoma; OR, odds ratio; CI, confidence interval. † Systolic blood pressure <90 mmHg and heart rate >100 beat per minute

AC C

EP

TE D

M AN U

‡ Single <5 cm or ≤3 nodules <3 cm

<0.001

ACCEPTED MANUSCRIPT

Table 4. Causes of death within 6 weeks in patients with or without low skeletal muscle mass

Patients without low skeletal muscle mass (n =17)

n

Cause of death

n

Liver failure Hypovolemic shock Infection

10 1 1

Liver failure Hypovolemic shock Infection

11 5 1

Hyperkalemia Rupture of HCC Tumor lysis syndrome

1 1 1

M AN U

AC C

EP

TE D

HCC, hepatocellular carcinoma

SC

Cause of death

RI PT

Patients with low skeletal muscle mass (n =15)

24

ACCEPTED MANUSCRIPT

Highlights


The 5-days rebleeding and mortality was unrelated to presence of low muscle mass. The 6-week mortality was significantly higher in patients with low muscle

RI PT




mass.

EP

TE D

M AN U

SC

Liver failure was the main cause of death in patients with low muscle mass.

AC C