THE AMERICAN JOURNAL OF GASTROENTEROLOGY
1, 2000
© 2000 by Am. Coll. of Gastroenterology
Published by Elsevier Science Inc.
Right Colonic Involvement Is
Associated With Severe Forms
of Ischemic Colitis and Occurs
Frequently in Patients With Chronic
Renal Failure Requiring Hemodialysis
Christelle Flobert, M.D., Christophe Cellier, M.D., Ph.D., Anne Berger, M.D., Alain Ngo, M.D.,
Emmanuel Cuillerier, M.D., Bruno Landi, M.D., Philippe Marteau, M.D., Ph.D., Paul-Henri Cugnenc, M.D.,
and Jean-Philippe Barbier, M.D.,
Digestive Disease Department, Laennec and Georges Pompidou European Hospital, AP-HP, University René
Descartes, Paris, France
OBJECTIVE: The aim of this study was to identify factors
associated with severe outcome in patients with ischemic
colitis.
METHODS: The files of 60 consecutive inpatients (34
women, 26 men, mean age 67 yr) with ischemic colitis were
reviewed. The following data were analyzed: age, sex,
smoking, medications, history of cardiovascular disease,
metabolic disease, chronic renal failure and hemodialysis,
the time elapsed between the first symptoms and the diag-
nosis, and the site and extension of their colonic involve-
ment. Patients were divided into two groups according to
outcome: those with severe disease, including those who
died from ischemic colitis (n ⫽ 3) or who required surgical
resection (n ⫽ 21); and those with mild forms of colitis who
were treated successfully without surgery (n ⫽ 36). The two
groups were compared by means of univariate and multi-
variate analysis to identify factors associated with unfavor-
able outcomes. Only patients who had a complete exami-
nation of the colon (n ⫽ 51) were entered into the statistical
analysis.
RESULTS: By univariate analysis, chronic renal failure (p ⫽
0.03), hemodialysis (p ⫽ 0.01), short delay between symp-
toms and diagnosis (p ⫽ 0.01), and right colonic involve-
ment (p ⫽ 0.002) were significantly more common in the
patients with severe colitis. By logistic regression, right
colonic involvement was the only factor independently as-
sociated with severity (p ⫽ 0.01). Right-sided lesions were
present in 82% of patients on dialysis but in only 26% of
patients not on dialysis (p ⫽ 0.0005).
CONCLUSIONS: Right colonic involvement is associated
with severe forms of ischemic colitis and occurs frequently
in patients with chronic renal failure requiring hemodialysis.
(Am J Gastroenterol 2000;95:195–198. © 2000 by Am.
Coll. of Gastroenterology)
Vol. 95, No.
ISSN 0002-9270/00/$20.00
PII S0002-9270(99)00703-0
INTRODUCTION
The severity of ischemic colitis may range from mild to
life-threatening. In the mild form of ischemia, which is the
most frequent, only the mucosa and submucosa are affected.
Colitis in these patients will most often heal if the cause of
ischemia is corrected, although colonic stricture sometimes
develops during recovery. The severe forms of ischemic
colitis are characterized by transmural gangrenous necrosis
of the colon, rapidly leading to peritonitis and death (1).
Management depends on the severity of the colitis. The
decision to intervene surgically is based on the presence of
peritoneal signs, failure to respond to medical management,
or persistence of clinical signs of subacute sepsis (1). There
have been few studies analyzing the factors associated with
poor prognosis in ischemic colitis (2– 6), and most factors
studied have proved to be unrelated to severity. Only right
colonic involvement has been implicated, and that in only
one study (2).
MATERIALS AND METHODS
Patients
The files of 60 consecutive patients with ischemic colitis
admitted to our department between 1985 and 1997 were
reviewed. The diagnosis of ischemic colitis was made from
the sum of the clinical, radiological, endoscopic, and patho-
logical findings. Patients with a previous history of ulcer-
ative colitis or Crohn’s disease, or with pathogenic bacteria
in their stools, or who had received antibiotics in the pre-
ceding week were excluded.
Data on the following variables were obtained for all the
subjects: age, sex, smoking habit, history of cardiovascular
disease, metabolic disease including diabetes and dyslipide-
mia, chronic renal failure with or without hemodialysis (no
patients had hemodialysis for acute renal failure), medica-
196 Flobert et al.
Table 1. Risks Factors in the Overall Population
n %
Smoking 23 39%
Hypertension 36 60%
Metabolic disease 13 22%
Cardiovascular disease 34 57%
Medications 8 14%
Chronic renal failure 26 43%
Hemodialysis 20 33%
tions (especially neuroleptics, cardiac glycosides, NSAIDs,
or estrogen or progestogen preparations), and the delay
elapsed between the first symptoms and the diagnosis. Of 60
patients, 51 (85%) had complete examination of the colon,
either by complete colonoscopy (n ⫽ 28), by partial
colonoscopy completed by colon enema (n ⫽ 3), or only
during surgery (n ⫽ 20). The colon was divided into three
segments (left, transverse, and right) to describe the site of
colonic involvement. An extensive colitis was defined as a
form involving more than one segment.
Diagnostic angiography was not employed and early an-
giography and vasodilator therapy was not used. An opaque-
enema CT scan carried out preoperatively in 10 patients was
strongly suggestive of ischemic colitis in seven, showing
thickening of the bowel wall with fatty infiltration and
parietal pneumatosis.
Statistical analysis
Patients were divided retrospectively into two groups: those
with “severe” and those with “mild” colitis. The “severe”
group included the patients who died from colonic ischemia
soon after admission and those who required colonic resec-
tion for peritonitis, severe sepsis, or failure of medical
management. The “mild” group included all the other pa-
tients, who all recovered without surgery.
The two groups were compared by means of univariate
analysis, and logistic regression was performed to identify
independent factors associated with poor outcome. Only
patients with the whole length of the colon explored where
entered into the statistical analysis. Qualitative variables
were compared by using the ␹2 test, and quantitative vari-
ables by Student’s t test and Wilcoxon’s test. A p value
⬍0.05 was considered as statistically significant.
RESULTS
Patient and Group Characteristics
A total of 34 women and 26 men with a mean age of 67 yr
(range, 23–97) were included in the study. Presenting symp-
toms and signs were abdominal pain in 49 patients (82%),
blood in the stools in 29 (48%), diarrhea in 23 (38%),
vomiting in 16 (27%), and fever in 14 (23%). Associated
conditions are shown in Table 1. The mean duration of
symptoms before the diagnosis of ischemic colitis was 14
days (range 0.5–90). The distribution of lesions, among the
AJG – Vol. 95, No. 1, 2000
Table 2. Localization of the Colonic Involvement
Site of With With With
Colonic Right Transverse Left With
Involvement Alone Colon Colon Colon Pancolitis Total
Right 15 1 3 4 23 (45%)
Transverse 2 1 3 4 10 (20%)
Left 24 3 3 4 34 (67%)
51 patients with the whole length of the colon examined, is
shown in Table 2.
The group of severe ischemic colitis included 24 patients.
Three of them died before surgery, and right colonic trans-
mural necrosis was found at autopsy in the three cases.
Twenty-one patients were operated on for peritonitis in 11
cases, sepsis with hemodynamic instability in eight cases,
failure of medical management in one case (5%), and pneu-
moperitoneum in one case. Among the 21 patients operated
on, the time elapsed between the onset of symptoms and the
surgical procedure was ⱕ24 h in 17 cases, 2 days in two
cases, and 9 days and 48 days in one case each, respectively.
Nineteen (90%) of the 21 patients who underwent surgery
were found to have transmural necrosis of the colon. One
patient was operated on for severe sepsis with hemodynamic
instability and had nontransmural necrosis of the whole
colon associated with ileal infarction. The other patient had
stenosis of the right colon and was operated on for failure of
medical management 48 days after the diagnosis. Colonic
ischemia was found during surgery to be nonocclusive in
75% of these cases. A second-look laparotomy was neces-
sary in three patients (14%), because of intra-abdominal
hemorrhage, sepsis, or postoperative ileal infarction. Four
patients died in the postoperative period.
The “mild” group comprised 36 patients, all of whom
were managed conservatively (withdrawal of the medication
causing mesenteric vasoconstriction, nasogastric suction to
rest the bowel in patients with ileus, intravenous rehydration,
and broad-spectrum antibiotics). None of these patients died.
Comparison of Mild and
Severe Ischemic Colitis (Table 3).
The two groups did not differ significantly in age, sex,
smoking, hypertension, cardiovascular, metabolic disease,
and medications. In univariate analysis, short delay between
onset of symptoms and diagnosis (p ⫽ 0.01), chronic renal
failure (p ⫽ 0.03), hemodialysis (p ⫽ 0.01), and right
colonic involvement (p ⫽ 0.002) were significantly more
frequent in the severe group. Right colonic involvement was
the only factor independently related to severity by multi-
variate analysis (p ⫽ 0.01).
Severe colitis was more common in patients on dialysis
(71%) than in those not on dialysis (32%), whether or not
they had chronic renal failure or not (p ⫽ 0.009). Right-
sided lesions were found in 82% of patients on dialysis (the
right colon was visualized in 17 patients and was involved
in 14) versus 26% of those not on dialysis (p ⫽ 0.0005).
AJG – January, 2000
Table 3. Comparison of Severe and Mild Forms of Ischemic Colitis*
Severe IC
(n ⫽ 23)
Age 65 (39–92)
Sex (M/F) 7/16
Smoking 6 (27%)
Hypertension 14 (61%)
Metabolic disease 6 (26%)
Cardiovascular disease 15 (65%)
Drugs 2 (9%)
Delay before diagnosis† 3 days (0.5–30)
Right-sided lesions 17 (74%)
Extensive colitis 4 (17%)
Chronic renal failure 14 (61%)
Hemodialysis 12 (52%)
* Only the patients with the whole length of the colon visualized have been included.
† Time elapsed between symptoms and diagnosis.
IC ⫽ ischemic colitis; M/F ⫽ male to female ratio.
DISCUSSION
The management of patients with ischemic colitis depends
on the severity of the disease (1). Mortality ranges between
10% and 50%, and there is a need to identify factors asso-
ciated with poor prognosis. In this study univariate analysis
identified four factors associated with an unfavorable out-
come: a short delay between onset of symptoms and diag-
nosis, right colonic involvement, chronic renal failure, and
hemodialysis; however, right colonic involvement was the
only factor independently associated with severity in
logistic regression. The high rate of severe colitis with
right colonic involvement observed in this study may not
reflect the spectrum of ischemic colitis elsewhere, in as
much as only inpatients were included, and outpatients with
mild colitis (who generally have a good outcome) were not
studied (7).
The severity of the ischemic colitis was judged retrospec-
tively on the patient outcome, e.g.: death or surgical proce-
dure for peritonitis, sepsis with hemodynamic instability,
and failure of medical management. A scoring system of
severity such as APACHE II would have been more appro-
priate to identify severe forms at diagnosis, but this score
could not be calculated in our retrospective study. A total of
92% (22 out of 24) of the patients classified in the severe
group had severe colonic lesions with transmural necrosis.
Among the two patients operated on who did not have
transmural necrosis, one patient had severe sepsis with he-
modynamic instability, and the surgical procedure showed
severe nontransmural necrotic lesions of the whole colon
with an ileal infarction. The other patient was operated on
because of occlusion and failure of conservative manage-
ment, and had a right colonic stenosis also associated with
an ileal infarction. Our retrospective classification according
to patient outcome may have reflected more fairly the se-
verity of the disease at admission.
Several studies have compared severe ischemic colitis
(either fatal, or requiring surgery) with the mild form that
can be treated conservatively, but most failed to identify
Severe Ischemic Colitis and Chronic Renal Failure 197
Mild IC Unifactorial Logistic
(n ⫽ 28) Analysis Regression
67 (23–97) NS
13/15 NS
12 (43%) NS
16 (57%) NS
6 (21%) NS
13 (46%) NS
5 (18%) NS
19 days (1–90) p ⫽ 0.01 NS
6 (21%) p ⫽ 0.002 p ⫽ 0.01
10 (36%) NS
7 (25%) p ⫽ 0.03 NS
5 (18%) p ⫽ 0.01 NS
factors predictive of severity (3–5). Our study confirms that
sex, age, and associated cardiovascular or metabolic disease
have no prognostic value. Longo et al. studied 47 patients
and suggested that the duration of symptoms before diag-
nosis, the number of concomitant medical conditions, and
initial hemodynamic instability were associated with severe
forms of ischemic colitis (6). In a study of 34 patients,
Robert et al. (2) suggested that right colonic involvement
might be associated with more severe forms, as they found
involvement of the right colon in 62% of their patients
requiring surgery, compared to only 19% of those managed
conservatively. Our study confirms that right colonic in-
volvement (present in 74% of our group of severe cases)
may be one of the major factor that is associated with a poor
outcome. The extension of colonic lesions was not associ-
ated with severe outcome, and extensive forms of ischemic
colitis were found even more frequently in patients with a
good outcome. A potential bias of our study may be that
nine of 36 (25%) patients in the mild group did not have
the right colon evaluated. Therefore, only patients with
the whole colon evaluated were entered into the statistical
analysis. A right colonic involvement would still have
been significantly associated with severe outcome, as the
only independent factor if the nine patients not evaluated
had had right colonic involvement (p ⫽ 0.007, data not
shown).
Ischemic colitis may be a significant cause of morbidity
and mortality in chronic hemodialysis patients (8, 9). The
preferential involvement of the right colon in our hemodi-
alysis population has been noted in previous reports (6,
10 –12). In our study, the right colon was involved in 82%
of patients on long term dialysis, but in only 26% of those
not on dialysis, and the colitis was also severe in 71% of the
former group but in only 32% of the latter. Hence, even mild
symptoms suggestive of ischemic colitis in patients on he-
modialysis should be taken as an indication for early vizual-
isation of the whole length of the colon to confirm or refute
the diagnosis, either by complete colonoscopy or by CT
198 Flobert et al.
scan. The latter is less invasive and, in our experience, has
considerable diagnostic value (13).
The pathogenesis of right colonic involvement, as well its
high incidence in patients on hemodialysis, remains to be
explained. Hemodialysis tends to cause repeated hypoten-
sive episodes, and may thus induce vasoconstriction of the
vasa recta in the right colon, leading to colonic ischemia,
probably in association with other classical predisposing
factors. Most of our patients with severe disease had
nonocclusive ischemic colitis. The right colon appears to
be particularly sensitive to non occlusive ischemia, as
there is little natural collateral circulation to meet tissue
needs in the event of loss of the primary arterial blood
supply (14). Furthermore, the vasa recta of the right colon
are longer and originate further away from the bowel than
those on the left side; this may increase their resistance to
reperfusion after an episode of hypotension (15).
To summarize, right colonic involvement was signifi-
cantly associated with severe colonic ischemia in this series;
it occurred frequently in patients with chronic renal failure
on hemodialysis. This suggests that patients with ischemic
colitis involving the right colon who do not require imme-
diate surgery should be carefully monitored in a special-
ized medicosurgical unit. Clinicians should be aware of
the frequency and severity of right colonic involvement
in patients on hemodialysis, and should prescribe a CT
scan or total colonoscopy if in doubt. Methods of pre-
venting ischemic colitis need to be found, especially in
patients on hemodialysis, in whom the main goal should
be to avoid or minimize hypotensive episodes during
dialysis.
Reprint requests and correspondence: Christophe Cellier, M.D.,
Service d’Hépatogastroentérologie, Hôpital Laennec and Hôpital
Européen Georges Pompidou, 42 rue de Sèvres 75007 Paris,
France.
Received Sep. 17, 1998; accepted July 6, 1999.
AJG – Vol. 95, No. 1, 2000
REFERENCES
1. Toursarkissian B, Thompson RW. Ischemic colitis. Surg Clin
North Am 1997;77:461–70.
2. Robert JH, Mentha G, Rohner A. Ischaemic colitis: Two
distinct pattern of severity. Gut 1993;34:4 – 6.
3. West BR, Ray JE, Gathright JB. Comparison of transient
ischemic colitis with that requiring surgical treatment. Surg
Gynecol Obstet 1980;151:366 – 8.
4. Abel ME, Russel TR. Ischemic colitis: Comparison of surgical
and nonoperative management. Dis Colon Rectum 1983;26:
113–5.
5. Guttormson NL, Bubrick MD. Mortality from ischemic colitis.
Dis Colon Rectum 1989;32:469 –72.
6. Longo WE, Ballantyne GH, Gusberg RJ. Ischemic colitis:
patterns and prognosis. Dis Colon Rectum 1992;35:726 –30.
7. Petit A, Guedon C, Duhamel C, et al. Colites ischémiques
“ambulatoires”. Gastroenterol Clin Biol 1990;14:739 – 43.
8. Diamond SM, Emmett M, Henrich WL. Bowel infarction as a
cause of death in dialysis patients. JAMA 1986;256:2545–7.
9. Zeier M, Wiesel M, Rambausek M, et al. Non-occlusive mes-
enteric infarction in hemodialysis patients. The importance of
prevention and early intervention. Nephrol Dial Transplant
1995;10:771–3.
10. Dahlberg PJ, Kisken WA, Newcomer KL, et al. Mesenteric
ischemia in chronic dialysis patients. Am J Nephrol 1985;5:
327–32.
11. Jablonski M, Putzki H, Heymann H. Necrosis of the ascending
colon in hemodialysis patients. Dis Colon Rectum 1987;30:
623–5.
12. Charra B, Cuche J, Ruffet M. Segmental necrosis of ascending
colon in hemodialysis patients. Nephrol Dial Transplant 1995;
10:2281–5.
13. Berger A, Mamzer-Bruneel MF, Wind P, et al. Opaque enema
CT scan allows early diagnosis of non-occlusive right colonic
ischemia in dialysis patients. Nephrol Dial Transplant 1997;
12:2179 – 81.
14. Landreneau RJ, Fry WJ. The right colon as a target organ of
nonocclusive mesenteric ischemia. Arch Surg 1990;125:
591– 4.
15. Stewart JA, Rankin FW. Blood supply of the large intestine.
Arch Surg 1933;26:843–91.