+Model

CLINRE-898; No. of Pages 7 ARTICLE IN PRESS

Clinics and Research in Hepatology and Gastroenterology (2016) xxx, xxx—xxx

Available online at

ScienceDirect
www.sciencedirect.com

CASE REPORT

The predictors of mortality and secular

changes in management strategies in
emphysematous gastritis
Andrew Watson , Vadim Bul , Jonas Staudacher ,
Robert Carroll , Cemal Yazici ∗

University of Illinois at Chicago, Department of Medicine, Division of Gastroenterology,

840 South Wood Street, Suite 718E-M/C 716, Chicago, IL 60612, USA

Summary
Background and aims: Emphysematous gastritis (EG) is caused by invasion of the gastric wall
by gas-producing organisms and carries mortality rate up to 60%. Our investigation aimed to
determine the predictors of survival and the secular trends in survival rates of subjects with
EG.
Methods: PubMed search was completed to identify previous cases of EG. In addition, we
included a recent case from our center. Statistical analysis was completed with two-sided Chi2
tests for categorical data and t-tests for continuous variables using SPSS v. 22.0 (SPSS Inc,
Chicago, IL).
Results: Study cohort included 59 adults. Mean age was 55.5 years; mean LOS was 28.6 days,
and 44.1% of subjects were female. Subjects who had EG before 2000 had significantly higher
rates of exploratory laparotomy compared to subjects who had EG after 2000 (62.5% vs. 22.2%,
P = 0.002). In contrast, subjects with EG after 2000 had significantly higher rates of EGD (55.6%
vs. 18.8%, P = 0.003) and lower rates of mortality (33.3% vs. 59.4%, P = 0.046) compared to
subjects with EG on or before 2000. In multivariate logistic regression analysis, the only inde-
pendent predictor of mortality was length of stay (P = 0.047).
Conclusion: We showed that previously reported 60% mortality rate of EG has been reduced
to 33.3% for cases reported after 2000. EGD has been utilized more often while surgical inter-
ventions are used only in carefully selected cases. Our data suggests that early endoscopic
evaluation and optimal medical management can perhaps continue to improve survival in
subjects with EG.
© 2016 Elsevier Masson SAS. All rights reserved.

∗ Corresponding author.
E-mail address: [email protected] (C. Yazici).

http://dx.doi.org/10.1016/j.clinre.2016.02.011
2210-7401/© 2016 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Watson A, et al. The predictors of mortality and secular changes in management strate-
gies in emphysematous gastritis. Clin Res Hepatol Gastroenterol (2016), http://dx.doi.org/10.1016/j.clinre.2016.02.011
+Model
CLINRE-898; No. of Pages 7 ARTICLE IN PRESS
2 A. Watson et al.

Introduction acute renal failure with renal replacement therapy and van-
comycin and piperacillin-tazobactam for presumed sepsis.
Emphysematous gastritis (EG) is defined as gastric wall CT with intravenous contrast of the chest, abdomen,
inflammation with the presence of intramural gas and sys- and pelvis was notable for bilateral pleural effusions, gas-
temic toxicity [1]. EG is a rare condition but its mortality tric wall and diffuse small bowel thickening with air fluid
rate reaches up to 61% [2]. The condition was first described levels concerning for pneumatosis, a small amount of air
by Fraenkel [3] in 1889 in a 35-year-old male who developed along the posterior gastric wall (Fig. 1A), cirrhosis with
hematemesis and abdominal pain after undergoing debride- splenomegaly and esophageal varices, as well as diffuse
ment of an injured finger. Fraenkel’s patient presented with mesenteric edema. An EGD was performed showing grade III
several of the classic symptoms: abdominal pain, diarrhea, varices in the middle and distal esophagus and one large non-
vomiting and hematemesis. The patient rapidly deteriorated bleeding cratered ulcer suggestive of phlegmonous gastritis
and died after only three days. The autopsy showed ‘‘a great along the greater curvature of the stomach (Fig. 1B). Biopsy
number of partly confluent over hempseed sized mucosal results of this ulcer showed gastric mucosa with ulceration,
detachments’’ in the stomach and pars superior duodeni necrosis and inflammatory debris with bacterial colonies
due to trapped sub mucosal air. Rod-shaped bacteria were later determined to be gram-negative rods (Fig. 1C). Cul-
found between the mucosa and submucosa and the authors tures of the biopsy grew Candida glabrata and Candida
hypothesized that this infectious source was responsible for dubliniensis. The patient’s mental status gradually improved
the formation of air. Since this original landmark case, 68 and he was weaned of vasopressors and extubated on the
cases have been reported with 59 of those cases occurring seventh day of his hospitalization. He was able to tolerate
in subjects older than 18 years. an oral diet and a subsequent EGD showed grade II varices
Over the past 15 years, there have been advancements in in the middle of the esophagus and hematin in the gastric
endoscopic and medical management of EG. However, the antrum with resolution of the previously seen ulcer (Fig. 1D).
effects of such advancements and management strategies on Accordingly, a repeat CT scan showed resolved pneumatosis
mortality of the disease have not been investigated yet. In with improved small bowel and gastric thickening. Unfortu-
this manuscript, we report an interesting case of EG from our nately, despite continued aggressive medical management,
center and perform a retrospective analysis of 59 adult cases the patient developed fungemia with C. glabrata and multi-
of EG to determine the predictors of survival and analyze the organ failure. Withdrawal of care was performed on the 20th
secular trends in management strategies and their impact on day of hospitalization.
survival.

Methods
Case report
A PubMed search was performed to identify previous cases
A 46-year-old male with a history of alcoholic cirrhosis of EG using the search terms ‘‘gastric emphysema or emphy-
(complicated by grade III esophageal varices, portal hyper- sematous gastritis.’’ All reports prior to June 2014 were
tensive gastropathy, and hepatic encephalopathy), chronic identified and reviewed. The cases that met the inclu-
alcoholic pancreatitis, insulin dependent diabetes mellitus sion criteria (the presence of gastric wall inflammation,
and anemia presented to an outside hospital with multiple intramural gas and systemic toxicity) were included. A
episodes of hematochezia and hematemesis. On initial pre- systematic analysis was completed using these confirmed
sentation his hemoglobin was 9.2 g/dL, INR was 2.3, and reports and a recent case from our center. Data on age,
platelets 52 × 103 /␮L. He was placed on continuous pro- gender, report date, existing co-morbidities (diabetes mel-
ton pump inhibitor and octreotide infusions and received litus, gastroenteritis, prior surgery, pancreatitis, corrosive
transfusions of red blood cells and platelets. Esophagogas- ingestion, end-stage renal disease, chronic liver disease, and
troduodenoscopy (EGD) revealed multiple, small purple malignancy), drug exposure (NSAIDs and immunosuppres-
vesicles in the distal half of the esophagus and the proximal sives), alcohol use, imaging modalities used for diagnosis,
half of the stomach was found to be circumferentially ulcer- serum leukocyte count, microbiology results, use of endo-
ated and friable. The patient was hypotensive and anuric scopic and surgical interventions, timing and findings of
requiring vasopressor support and was subsequently trans- esophagogastroduodenoscopy, mortality and length of stay
ferred to our center for further care. (LOS) were collected.
Upon transfer, his physical exam was notable for scleral The severity of mucosal injury was categorized into four
icterus and a protuberant abdomen without fluid wave. An groups based on endoscopic appearance: group 1 (no injury),
arterial blood gas showed a pH of 7.17, paCO2 23, PaO2 group 2 (superficial injury), group 3 (deep injury) and group
180, bicarbonate 8 mEq/L, and lactic acid was 8 mEq/L. 4 (healing lesions). In order to analyze the secular changes in
Chemistry panel was notable for sodium of 129 mmol/L, BUN management strategies, the cases were categorized into two
of 31 mg/dL and creatinine of 2.14 mg/dL. Liver function groups in regards to their report date: those reported prior
tests showed a total bilirubin of 8.1 mg/dL, direct bilirubin to 2000 vs. after 2000. Statistical analysis was completed
of 5.1 mg/dL, aspartate aminotransferase (AST) 94 units/L, with two-sided Chi2 or Fisher exact tests for categorical
alanine aminotransferase (ALT) 25 units/L and albumin of data and t-tests for continuous variables using SPSS v. 22.0
1.9 g/dL. Hemoglobin was 7.4 g/dL, platelets 62 × 103 /␮L (SPSS Inc, Chicago, IL). Subsequently, a multivariate logistic
and INR was 2.5. Urine toxicology screen was positive for regression analysis was completed to determine the predic-
cocaine and benzodiazepines. The patient was treated for tors of survival.

Please cite this article in press as: Watson A, et al. The predictors of mortality and secular changes in management strate-
gies in emphysematous gastritis. Clin Res Hepatol Gastroenterol (2016), http://dx.doi.org/10.1016/j.clinre.2016.02.011
+Model
CLINRE-898; No. of Pages 7 ARTICLE IN PRESS
The predictors of mortality in emphysematous gastritis 3

Figure 1 A. Cross-sectional computed tomography (CT) imaging demonstrated air along gastric wall. B. Esophagogastroduo-
denoscopy (EGD) image showed diffuse mucosal injury suggestive of phlegmonous gastritis. C. Gastric biopsy revealed cystic dilation
of the gastric ducts, ulceration, diffuse inflammation as well as focal hemorrhage and necrosis. D. Repeat EGD showed resolution
of mucosal injury.

Results cases were reported after the year 2000 and 32 cases prior
to the year 2000. Subjects who had EG in the earlier group
Literature search had significantly higher rates of exploratory laparotomy
compared to those in the latter group (62.5% vs. 22.2%,
respectively; P = 0.002). In contrast, subjects with EG after
Sixty-seven cases were identified through the PubMed search
2000 had significantly higher rates of early endoscopic evalu-
and with the addition of our case the study size reached 68
ation (within 14 days of presentation) compared to subjects
subjects. Fifty-nine of these were adults aged 18 years or
with EG in or before 2000 (55.6% vs. 18.8%, respectively;
older. The remaining 9 pediatric cases were excluded from
P = 0.003). Subjects who developed EG after 2000 had signifi-
our final analysis. Fig. 2 shows the flow diagram of published
cantly lower rates of mortality compared to those diagnosed
work search.
with EG in or before 2000 (33.3% vs. 59.4%, respectively;
P = 0.046).
Demographic and clinical data

Table 1 summarizes clinical data. Among adult subjects, The predictors of survival
mean age was 55.5 ± 17.5 years. Twenty-six of 59 (44.1%)
subjects were female. The mean LOS was 28.6 ± 34.5 days. In univariate analysis, patients who recovered from EG were
No causative organism was isolated in 25 of 59 (42.4%) more likely to have been diagnosed after the year 2000
cases. When an organism was isolated, a polymicrobial infec- (58.1% vs. 41.9%, P = 0.046). Other predictors of survival for
tion was most common, being present in 12/34 (35.3%) of patients with EG were increased frequency of repeat endo-
reports. Clostridium species was the most common single scopic evaluation (29% vs. 3.6%, P = 0.009), longer LOS (43.6
isolate, identified in 6/34 (17.6%) cases. Organisms that vs. 18.5 days, P = 0.005) and higher rates of gastric ulcer
were identified are listed in Table 2. (35.5% vs. 10.7%, P = 0.026). No single organism was found
to be an independent predictor of mortality in univariate
Secular changes in management strategies analysis. Age, gender, degree of mucosal injury, underlying
co-morbidities, drug exposure (NSAIDs and immunosuppress-
Table 3 shows the secular changes in management strate- ants), alcohol use, and serum leukocyte count did not have
gies among EG patients before and after 2000. Twenty-seven a statistically significant impact on survival (Table 4).

Please cite this article in press as: Watson A, et al. The predictors of mortality and secular changes in management strate-
gies in emphysematous gastritis. Clin Res Hepatol Gastroenterol (2016), http://dx.doi.org/10.1016/j.clinre.2016.02.011
+Model
CLINRE-898; No. of Pages 7 ARTICLE IN PRESS
4 A. Watson et al.

Figure 2 Flow diagram of literature search.

Table 1 Summary of clinical data. Table 2 Organisms identified.

Number of cases reviewed 59 Organisms n (%)

Mean age (years) 55.5 ± 17.5
No organisms identified 25 (42.4)
Sex M/F (% female) 33/26 (44.1)
Polymicrobial 12 (20.3)
Mean white blood cell count (WBC) 19.2 ± 17.3
Clostridium Sp. 6 (10.2)
Length of stay (days) 28.6 ± 34.5
Candida Sp. 5 (8.5)
Diagnostic modality n (%)
Streptococcus Sp. 2 (3.4)
Computed tomography (CT) 30 (50.8)
Klebsiella Sp. 2 (3.4)
X-ray 21 (35.6)
Strongyloides stercoralis 1 (1.7)
Upper gastrointestinal series 5 (8.5)
Staphylococcus Sp. 1 (1.7)
Autopsy 2 (3.4)
Bacillus Sp. 1 (1.7)
Ultrasound 1 (1.7)
Enterobacter colacae 1 (1.7)
Organism identified 34 (57.6)
Enterococcus Sp. 1 (1.7)
Organism unidentified 25 (42.8)
Mucor Sp. 1 (1.7)
Organisms (% of those identified)
Escherichia coli 1 (1.7)
Polymicrobial 12 (35.3)
Gram positive 11 (32.4)
Fungi 6 (17.6)
Gram negative 4 (11.8) A multivariate logistic regression analysis was completed
Parasite 1 (2.9) using univariate variables, which were significantly different
Esophagogastroduodenoscopy (EGD) 21 (35.6) between those who survived vs. died as well as key clinical
performed (% performed) variables such as age and gender. The only independent pre-
EGD findings (% of endoscopies) dictor of survival on this multivariate analysis was length of
Group 1 (no injury) 0 (0) stay (P = 0.047).
Group 2 (superficial injury) 5 (23.8)
Group 3 (deep injury) 15 (71.4) Discussion
Group 4 (healing injury) 1 (4.8)
Exploratory laparoscopy performed (% 26 (44.1) Patients with EG usually present approximately one week
performed) after the inciting event although the time to presenta-
Gastric resection performed (% 11 (18.6) tion can range up to six weeks [1]. Common triggers
performed) include the ingestion of corrosives [4,5], alcohol abuse
Mortality (%) 28 (47.5) [6], recent surgery, gastroenteritis, and non-steroidal anti-
inflammatory drug (NSAID) usage [6—8]. Symptoms include

Please cite this article in press as: Watson A, et al. The predictors of mortality and secular changes in management strate-
gies in emphysematous gastritis. Clin Res Hepatol Gastroenterol (2016), http://dx.doi.org/10.1016/j.clinre.2016.02.011
+Model
CLINRE-898; No. of Pages 7 ARTICLE IN PRESS
The predictors of mortality in emphysematous gastritis 5

Table 3 Differences in management strategies and outcomes before and after 2000.

Variable Before 2000 (n = 32) After 2000 (n = 27) P-value

Exploratory laparotomy 20 (62.5%) 6 (22.2%) 0.002

Early EGD (within 14 days) 6 (18.8%) 15 (55.6%) 0.003
LOS (days) 43.6 ± 36.3 18.0 ± 30.0 0.005
Mortality 19 (59.4%) 9 (33.3%) 0.046
EGD: esophagogastroduodenoscopy.

Table 4 Predictors of survival in emphysematous gastritis in univariate analysis.

Variable Survived (n = 31) Died (n = 28) P-value

Mean age (years) 53.4 ± 18.7 57.9 ± 16.1 0.324

Gender n (%) n (%) 0.219
Female 16 (51.6) 10 (35.7)
Male 15 (48.4) 18 (64.3)
Diabetes mellitus 11 (35.5) 4 (14.3) 0.062
Length of stay (days) 43.6 ± 36.3 18.5 ± 30.0 0.005
EG diagnosis after 2000 18 (58.1) 9 (32.1) 0.046
Second look EGD 9 (29.0) 1 (3.6) 0.009
Gastric ulcer 11 (35.5) 3 (10.7) 0.026
EGD: esophagogastroduodenoscopy.

epigastric abdominal pain, nausea, vomiting, and diarrhea While the methods of diagnosing EG have evolved since
with signs of hemodynamic instability and leukocytosis. the first case was reported, there is no general consensus
Some patients also develop melena and hematemesis. The on the preferred imaging modality [22,23]. Abdominal plain
pathognomonic finding of emphysematous gastritis is emesis films of EG may show gas bubbles in an irregular pattern
of a necrotic mucosal clot in the shape of the gastric wall contouring to the wall of the stomach. Unlike simple intra-
caused by dissection of the muscularis mucosa by bacterial luminal air or fluid, gas from EG will often stay constant
organisms [1,6,9,10]. Patients with an immunocompromised regardless of position. X-ray can also be used to help dif-
state such as those with diabetes mellitus, cirrhosis or ferentiate between EG and gastric emphysema as the latter
end-stage renal disease (ESRD) can have less dramatic pre- tends to have a more uniform linear pattern of bubbles [24].
sentations of the illness [11—13]. Computed tomography (CT) is more sensitive and specific for
The exact pathophysiology of this severe illness is not yet detecting intramural gas and is better able to delineate its
entirely understood. Normally, the gastric mucosa protects precise anatomical involvement [25]. CT findings classically
the deeper layers of the stomach through close connec- include cystic pockets or streaks of air within the gastric
tions between mucosal cells, the presence of an acidic wall, thickened mucosal folds, pneumoperitoneum [26,12]
pH and through its rich bloody supply which leads to ade- and portal venous gas [27—31]. Endoscopic findings greatly
quate oxygenation and tissue repair [1]. Possible etiologies vary between cases but typically reveal erythematous, ero-
for mucosal injuries include liquefaction necrosis by alkali sive lesions with occasional cobblestoning and necrosis of
ingestion, coagulation necrosis from acid ingestion and the gastric mucosa. Previous publications have noted that
alteration in the mucosal barrier in patients with chronic X-ray in conjunction with clinical suspicion is enough to
alcohol abuse [14,15]. It is still unclear whether bacteria make the diagnosis of EG but a CT scan is more sensitive
enter the gastric wall through mucosal ulceration or through and better able to detect small quantities of air compared
hematogenous spread to the stomach, but it is likely that to abdominal radiography alone [2,32]. Regardless of how
both mechanisms may exist. EG is diagnosed, early treatment and intervention is essen-
When a patient is found to have air in the stomach tial. The most commonly documented causative organisms
wall, it is important to differentiate between EG, gastric in the literature are Streptococci species, Escherichia coli,
emphysema and phlegmonous gastritis. Patients with gas- Enterobacter species, Clostridium welchii and Staphylococ-
tric emphysema are usually non-toxic-appearing and the air cus aureus [2]. Multiple cases of fungal etiologies have also
enters the stomach wall through either trauma, a pulmonary been reported in the literature, usually Candida species or
etiology such as a ruptured pulmonary bleb [16,17], or by Mucormycosis [1,2,26,13,33]. These organisms have been
gastrointestinal obstruction leading to increased intragas- identified by gastric aspirate culture or tissue biopsy and
tric air and cystic pneumatosis [18,19]. The prognosis for culture.
patients with gastric emphysema is good as long as the cause Treatment for EG consists of intravenous fluid resuscita-
of intramural air is corrected [20]. Phlegmonous gastritis is tion, antimicrobials, blood product replacement and other
inflammation of the gastric wall due to infection but in the supportive care. Due to the wide range of organisms impli-
absence of intramural air [21]. cated in EG, starting patients on broad-spectrum antibiotics

Please cite this article in press as: Watson A, et al. The predictors of mortality and secular changes in management strate-
gies in emphysematous gastritis. Clin Res Hepatol Gastroenterol (2016), http://dx.doi.org/10.1016/j.clinre.2016.02.011
+Model
CLINRE-898; No. of Pages 7 ARTICLE IN PRESS
6 A. Watson et al.

is recommended. Surgery for EG is not indicated during survival is not unexpected, as patients who survive this
acute infection unless there is evidence of perforation, devastating disease will be hospitalized longer than those
ischemia and necrosis or if conservative management fails who succumb early to the illness. In fact, on multivariate
[1,34,35]. Limiting surgical intervention is essential as there analysis, only LOS was found to be an independent predic-
is a high risk of anastomotic leak and occurrence of fistu- tor of mortality. Co-morbidities such as pancreatitis, ESRD,
las and strictures postoperatively in patients who undergo and diabetes mellitus were not associated with increased
surgery during an active EG infection [22]. The mortality mortality in our analysis. These conditions are associated
rate of EG has been reported as high as 61% [2]. Of patients with chronic inflammation and immunosuppression that typ-
who do survive, approximately 20—25% will develop gastric ically predisposes to worse outcomes from severe infections,
strictures that may require dilation [20,36]. therefore it would have been expected that they would also
While two reviews of EG have been published previ- portend a worse prognosis in patients with EG. The lack of
ously, no study has yet performed a systematic analysis association between these co-morbid conditions and mor-
to determine predictors of survival and to identify the tality rates is likely due to small sample size; only four
secular trends in regards to endoscopic evaluation of EG. patients were diagnosed with pancreatitis, and just seven
Using the largest dataset in the literature (n = 59, adult had ESRD. Another possibility is that the definitions of these
subjects), we determined that patients diagnosed with EG medical co-morbidities have evolved over time, resulting in
before 2000 were about three times more likely to undergo the under- or over-reporting of co-morbid conditions.
surgical intervention (62.5% vs. 22.2%, P = 0.002) whereas There are several limitations in our study. First, despite
those diagnosed after 2000 were three times as likely to our review being the most comprehensive to date, the study
undergo EGD (55.6% vs. 18.8%, P = 0.003). This is probably size remains small and this may lead to limit the detection
due to more prompt recognition of EG by advanced imaging of variables, which could effect the survival but requires
techniques and enhanced supportive care with intravenous a larger power to detect the statistical significances. Sec-
fluids, vasopressors, and broad-spectrum antibiotics, and a ondly, while we have performed the most comprehensive
better understanding of the overall disease process lead- analysis to determine the predictors of mortality, there may
ing to decreased need for surgery. Also likely contributing be other factors that we could not analyze due to the ret-
to the lower rate of surgical intervention and higher rate rospective design of our study and lack of data in previously
of endoscopic evaluation is the advancement in biomedi- published case reports.
cal technology, which has led to increased availability and In conclusion, EG is a very rare condition of gastric
utilization of high-definition, low-cost endoscopic tools. Pre- wall inflammation with resultant intramural gas formation
vious reviews have suggested that surgery for EG is not and systemic toxicity that is associated with high rates of
indicated during acute infection unless there is evidence of morbidity and mortality. Through our comprehensive ret-
perforation, ischemia and necrosis or if conservative man- rospective analysis, we confirmed that management of EG
agement fails [1,34,35]. However, our review demonstrated over the years has shifted from early surgical intervention
that surgical intervention does not confer any statistically towards more conservative measures including advanced
significant benefit on mortality. It is likely that the factors supportive care and earlier endoscopic evaluation. This has
above which contributed to lower rates of surgical interven- coincided with a significant decrease in mortality rate in
tion also contributed to the significant decrease in mortality cases reported after the year 2000. Our systematic anal-
from EG after 2000 in our study. ysis of previously published cases did not identify any
In univariate analysis, patients who were found to have co-morbid conditions or specific organisms that are asso-
a gastric ulcer, were diagnosed with EG after 2000, had ciated increased mortality in EG. The only independent
more than 1 endoscopic evaluation or greater LOS were predictor of survival was LOS, which correlates well with
more likely to survive. No single organism was found to be the notion of advanced supportive care and earlier utiliza-
an independent predictor of mortality. Patients who had tion of endoscopic therapy, which improves survival. EG is
gastric ulcers may have had localized disease which could an important diagnosis to consider in patients with abdom-
be both endoscopically and medically treated with rela- inal pain that are toxic-appearing and have evidence of
tively high efficacy. This can explain the survival benefit gastric wall emphysema on imaging as early endoscopic
of this patient group. Accordingly, patients who develop evaluation in such cases can be crucial for prolonged
EG from hematogenous dissemination may be more likely survival.
to have multi-organ dysfunction and higher mortality rates
than those with EG with isolated gastric ulcers. Unfortu-
nately, the majority of published cases did not have data Disclosure of interest
on bacteremia and we could not perform further analysis on
this subject. The authors declare that they have no competing interest.
The association of multiple endoscopic evaluations with
improved survival is likely correlational and secondary to
the fact that patients who survive the initial insult of EG References
will have an increased LOS and undergo follow-up endoscopy
during their hospital stay. However, it is also a possibility [1] Moosvi AR, Saravolatz LD, Wong DH, Simms SM. Emphyse-
that patients undergoing multiple endoscopic interventions matous gastritis: case report and review. Rev Infect Dis
are also receiving therapeutic benefit from endoscopic 1990;12(5):848—55.
treatment, which could contribute to improved survival. A [2] Yalamanchili M, Cady W. Emphysematous gastritis in a
positive correlation between increased LOS and improved hemodialysis patient. South Med J 2003;96(1):84—8.

Please cite this article in press as: Watson A, et al. The predictors of mortality and secular changes in management strate-
gies in emphysematous gastritis. Clin Res Hepatol Gastroenterol (2016), http://dx.doi.org/10.1016/j.clinre.2016.02.011
+Model
CLINRE-898; No. of Pages 7 ARTICLE IN PRESS
The predictors of mortality in emphysematous gastritis 7

[3] Fraenkel E. Ueber einen Fall von Gastritis acuta emphyse- [19] Kussin SZ, Henry C, Navarro C, Stenson W, Clain DJ. Gas within
matosa wahrscheinlich mykotishcen Ursprungs. Virchows Arch the wall of the stomach report of a case and review of the
[A] 1889;118:526—35. literature. Dig Dis Sci 1982;27(10):949—54.
[4] Berry WB, Hall RA, Jordan GL. Necrosis of the entire stom- [20] Grayson DE, Abbott RM, Levy AD, Sherman PM. Emphysema-
ach secondary to ingestion of a corrosive avid: report of a tous infections of the abdomen and pelvis: a pictorial review.
patient successfully treated by total gastrectomy. Am J Surg Radiographics 2002;22(3):543—61.
1965;109:652—5. [21] Kim GY, Ward J, Henessey B, et al. Phlegmonous gastritis: case
[5] Clearfield HR, Shin YH, Schreibman BK. Emphysematous gastri- report and review. Gastrointest Endosc 2005;61(1):168—74.
tis secondary to lye ingestion. Report of a case. Am J Dig Dis [22] Meyers HI, Parker JJ. Emphysematous gastritis. Radiology
1969;14(3):195—9. 1967;89(3):426—31.
[6] Smith TJ. Emphysematous gastritis associated with ade- [23] Stein LB, Greenberg RE, Ilardi CF, Kurtz L, Bank S. Acute necro-
nocarcinoma of the stomach. Am J Dig Dis 1966;11(5): tizing gastritis in a patient with peptic ulcer disease. Am J
341—5. Gastroenterol 1989;84(12):1552—4.
[7] Van mook WN, Van der geest S, Goessens ML, Schoon EJ, Ramsay [24] Tuck JS, Boobis LH. Interstitial emphysema of the stomach due
G. Gas within the wall of the stomach due to emphysematous to perforated appendicitis. Clin Radiol 1987;38(3):315—7.
gastritis: case report and review. Eur J Gastroenterol Hepatol [25] Loi TH, See JY, Diddapur RK, Issac JR. Emphysematous gastri-
2002;14(10):1155—60. tis: a case report and a review of literature. Ann Acad Med
[8] Badrinath K, Jairam N, Ravi HR, Augustine T. Gastric Singapore 2007;36(1):72—3.
emphysema following drug-induced gastritis. Br J Surg [26] Jung JH, Choi HJ, Yoo J, Kang SJ, Lee KY. Emphysematous gas-
1992;79(4):320. tritis associated with invasive gastric mucormycosis: a case
[9] Mckelvie PA, Fink MA. A fatal case of emphysematous gastritis report. J Korean Med Sci 2007;22(5):923—7.
and esophagitis. Pathology 1994;26(4):490—2. [27] Iqbal M, Ishtiaq O, Akhtar A, Chaudery Y, Khizar B. Empysema-
[10] Gonzalez LL, Schowengerdt C, Skinner HH, Lynch P. Emphyse- tous gastritis — a case report with literature review. J Pak Med
matous gastritis. Surg Gynecol Obstet 1963;116:79—87. Assoc 2006;56(12):605—7.
[11] Fidvi SA, Klein SA. Radiological case of the month—–clinical [28] Ho DC, Moore LA, Downey DB. Residents’ corner. Answer to case
quiz. Appl Radiol 2002;31:33—6. of the month #86. Emphysematous gastritis. Can Assoc Radiol
[12] Van allan RJ, Krebs TL, Henzie GM, Fermelia DE, Fichman MP. J 2002;53(5):309—11.
Emphysematous gastritis in an immunocompromised host. Clin [29] Shipman PJ, Drury P. Emphysematous gastritis: case report and
Imaging 1995;19(2):129—30. literature review. Australas Radiol 2001;45(1):64—6.
[13] Cherney CL, Chutuape A, Fikrig MK. Fatal invasive gas- [30] Hadas-halpren I, Hiller N, Guberman D. Emphysematous gastri-
tric mucormycosis occurring with emphysematous gastritis: tis secondary to ingestion of large amounts of Coca Cola. Am J
case report and literature review. Am J Gastroenterol Gastroenterol 1993;88(1):127.
1999;94(1):252—6. [31] Binmoeller KF, Benner KG. Emphysematous gastritis secondary
[14] Loeb PM, Eisenstein AM. Caustic injury to the upper gas- to gastric infarction. Am J Gastroenterol 1992;87(4):526—9.
trointestinal tract. In: Slesinger M, Fordtran J, editors. [32] Caudill JL, Rose BS. The role of computed tomography in the
Gastrointestinal disease: pathophysiology, diagnosis, manage- evaluation of pneumatosis intestinalis. J Clin Gastroenterol
ment. 3rd ed. Philadelphia: W.B. Saunders; 1983. p. 148—9. 1987;145:51e2.
[15] Ivey K, Roth J. Drug and chemical-induced injuries of the stom- [33] Williamson MR, Shah HR, Harper RR, Angtuaco TL. CT
ach. In: Berk JE, editor. Bockus gastroenterology, Vol. 2, 4th ed. of emphysematous gastritis. Comput Med Imaging Graph
Philadelphia: W.B. Saunders; 1985. p. 975—1003. 1989;13(2):175—7.
[16] Keyting WS, Mccarver RR, Kovarik JL, Daywitt AL. Pneumatosis [34] Bashour CA, Popovich MJ, Irefin SA, et al. Emphysematous gas-
intestinalis: a new concept. Radiology 1961;76:733—41. tritis. Surgery 1998;123(6):716—8.
[17] Naggar CZ. Pneumatosis intestinalis following common upper- [35] Ocepek A, Skok P, Virag M, Kamenik B, Horvat M. Emphyse-
respiratory-tract infection. JAMA 1976;235(20):2221—2. matous gastritis — case report and review of the literature. Z
[18] Caruana V, Swyane LC, Salaski JS. Indium-111 WBC detec- Gastroenterol 2004;42(8):735—8.
tion of emphysematous gastritis in pancreatitis. J Nucl Med [36] Makhoul E, Mana F, Vandermeulen L, Basile N. Emphysematous
1990;31:112—4. gastritis. Acta Gastroenterol Belg 2013;76(4):445—6.

Please cite this article in press as: Watson A, et al. The predictors of mortality and secular changes in management strate-
gies in emphysematous gastritis. Clin Res Hepatol Gastroenterol (2016), http://dx.doi.org/10.1016/j.clinre.2016.02.011