Digestive and Liver Disease 36 (2004) 461466

Alimentary Tract

Prognosis of severe attacks in ulcerative colitis:

effect of intensive medical treatment
L. Benazzato a , R. DInc a , F. Grigoletto b , E. Perissinotto b ,
V. Medici a , I. Angriman a , G.C. Sturniolo a,
a

Department of Surgical and Gastroenterological Sciences, University of Padua, Via Giustiniani 2, 35100 Padova, Italy
b Department of Environmental Medicine and Public Health, University of Padua, Padova, Italy
Received 21 May 2003; accepted 4 December 2003
Available online 30 April 2004
See related commentary on pages 448449

Abstract
Background. Severe attacks of ulcerative colitis have a high risk of colectomy.
Aims. To evaluate the effects of standard medical management and to identify the clinical and laboratory variables capable of predicting
the clinical outcome.
Materials and methods. Prospective study monitoring the clinical and laboratory variables in 67 patients with severe colitis. Therapy
consisted of prednisone, cyclosporin if no response, and azathioprine for maintenance. End-points were colectomy or remission. Logistic
regression analysis was applied for statistical evaluation.
Results. Fourteen (20%) patients required colectomy, 34 (50%) patients achieved remission with steroids, 25 (37%) patients received
cyclosporin, 19 (76%) with benefit. Increased body temperature, pulse rate, sedimentation rate and C-reactive protein levels on admission
were significantly associated with colectomy. Sedimentation rate greater than 75 mm/h and body temperature exceeding 38 C at admission
had 4.6- and 8.8-fold increased risk of colectomy. Less than 40% reduction in the bowel movements within 5 days predicted no response to
steroids. Azathioprine maintained remission in 70% of the patients.
Conclusions. Elevated sedimentation rate and fever at day 1 best predict colectomy in severe colitis. Less than 40% reduction in the bowel
movements at day 5 predicts no response to steroids. Cyclosporin has a high rate of success in acute attacks and azathioprine in maintaining
remission.
2004 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
Keywords: Azathioprine; Cyclosporin; Severe colitis

1. Introduction
Severe ulcerative colitis is a potentially life-threatening
condition with significant morbidity and even mortality. Despite intensive medical treatment, 2530% of patients presenting with a severe episode still need urgent colectomy
[1].
Attempts were made to identify simple clinical and laboratory criteria that would predict the outcome and assist in
the decision of the most appropriate interval before resorting
Corresponding author. Tel.: +39-049-821-2893;
fax: +39-049-876-0820.
E-mail address: gc.sturniolo@unipd.it (G.C. Sturniolo).

to surgical intervention, possibly avoiding early colectomy.

Retrospective studies identified tachycardia, fever, hypoalbuminemia and mucosal islands or colonic dilation as associated with colectomy during the same admission [2]. Further
prospective studies have identified C-reactive protein (CRP)
>45 mg/l and stool frequency >8 per day after 3 days intensive treatment or severe endoscopic lesions, Truelove and
Witts criteria and attacks lasting more than 6 weeks as useful prognostic parameters [3,4]. Furthermore, first episode
and extensive disease are associated with a poor medical response [5,6].
Many patients not responding to steroids have now
been treated with cyclosporin with the aim of avoiding/delaying urgent surgical procedures. A good response

1590-8658/$30 2004 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.dld.2003.12.017

462

L. Benazzato et al. / Digestive and Liver Disease 36 (2004) 461466

to the drug in the acute phase, however, drops significantly after discontinuing cyclosporin. The addition of
6-mercaptopurine or azathioprine has proven efficacy in
maintaining cyclosporin-induced remission [7,8].
The aims of our study were: (1) to evaluate if simple
clinical and laboratory variables at admission were able to
predict prognosis of acute attacks of severe colitis in patients
treated according to conventional guidelines; (2) to identify
the clinical or laboratory variables capable of predicting the
response to corticosteroids.

admission, three stool samples were obtained for culture.

Plain abdominal X-ray was taken at admission and every
other day unless there was a change in the clinical conditions.
Measurements of full blood cell count, erythrocyte sedimentation rate (ESR), CRP, 1 acid glycoprotein (1GPA),
potassium, creatinine and albumin concentrations were obtained at day 1, 3 and 5. Sigmoidoscopy was performed to
assess severity within 3 days of admission unless clinically
contraindicated. Severe endoscopic lesions were defined as
the presence, in any segment of the examined colon, of extensive deep ulcerations, large mucosal abrasions and spontaneous mucosal bleeding.

2. Methods
2.4. Outcome
2.1. Patients
All patients admitted to the Department of Surgical and
Gastroenterological Sciences of the University of Padua
from January 1996 to December 2001 with severe colitis
were evaluated. The diagnosis was made following the
clinical, radiological and pathological criteria and severity
was defined by Lichtiger criteria [9]. Patients with a clinical activity index (CAI) score >12 were included in the
study. Each patient recorded in a diary the number of bowel
movements, the presence of blood in each stool, abdominal
pain and the number of evacuations at night every day after
admission and up to discharge or colectomy.
2.2. Management
All patients received standard intensive medical therapy with intravenous methyl-prednisolone 1 mg/kg per day
following Truelove and Witts guidelines [10]. They also
received treatment for fluid, electrolyte and haemoglobin
deficiencies as well as for malnutrition.
Patients with no response as defined by the persistence
of >6 bowel movements per day and/or elevated inflammatory indexes at day 7, were treated with cyclosporin intravenously at the initial dose of 24 mg/kg per day for 7 days
and then orally. Blood cyclosporin levels were measured
daily by enzyme immunoassay (plasma levels between 150
and 400 g/l) to adapt the doses. Patients who responded
to oral cyclosporin had azathioprine 1.52 mg/kg per day
added within 3 months, unless contraindicated.
Patients deteriorating (severe diarrhoea, fever or abdominal pain, perforation or toxic megacolon) at any time during
intensive treatment were referred for colectomy.

Short-term outcome was assessed at day 5. Responders,

defined as patients having a reduction of at least 6 points in
clinical disease activity index, entered remission either with
steroids or with cyclosporin, non-responders were patients
who failed to respond or whose conditions deteriorated and
were operated on.
2.5. Statistical analysis
Quantitative data are reported as means S.D. and qualitative data as frequencies.
In order to determine which variables were important in
predicting successful outcome, i.e. response to medical therapy, responders and non-responders were compared with
reference to age, gender, number of years since initial diagnosis, number of bowel movements, pulse rate, body temperature, white blood cell count, haemoglobin levels, ESR,
CRP, 1GPA, platelet count and albumin levels. Groups
were compared by means of the Students t-test or 2 -test
(with Yates correction) when appropriate. For each statistically significant variable, positive predictive value (PPV),
negative predictive value (NPV), odds ratio (OR) and confidence intervals (CI) were calculated considering quantiles
as cut-off. We eventually performed the logistic regression
analysis, considering colectomy or response to medical treatment as dependent variables, and variables with statistically
significant OR as independent variables. In order to identify
the variables capable of predicting the response to steroids,
responders and non-responders were compared with reference to the difference of the value of each variable between
admission and day 3 and 5.
Estimation of the time to colectomy from admission was
based on the KaplanMeier method.

2.3. Measurements
Demographic (age and gender) and disease characteristics
(duration and extent) were recorded as well as the following
variables: CAI [9] (number of daily bowel movements, entity of abdominal pain and tenderness, use of antidiarrhoics,
blood in stools, general well-being, faecal incontinence,
nocturnal diarrhoea), pulse rate and body temperature. At

3. Results
In our experience patients with severe colitis had a 20%
chance of colectomy during the same admission, urgent
colectomy was performed in 14% of the patients with a
cumulative rate of colectomy at 1 year exceeding 35%.

L. Benazzato et al. / Digestive and Liver Disease 36 (2004) 461466

to colonic perforation (one patient) or toxic megacolon (one

patient). The remaining 25 patients received cyclosporin,
however, six of them did not respond to cyclosporin and had
colectomy performed within 1 month.

1,2

1,0

Cumulative survival

463

,8

3.1. Predictors of colectomy at admission

,6

,4

,2

0,0
0

200

400

600

800

1000

days
Fig. 1. KaplanMeier estimated time to colectomy in patients with severe
attacks of ulcerative colitis. Day 0 denotes admission to hospital.

Estimation of the time to colectomy from admission is

shown in Fig. 1.
Sixty-seven patients (40 males, 27 females, median age
45 years) were treated for a total of 67 episodes. In 34 patients the disease settled promptly and they were discharged
from the hospital after a mean of 7 1.8 days. Three patients needed colectomy during the follow up due to a second episode of severe colitis not responding to conventional
treatment (after 16, 21 and 26 months). The other patients
were kept on steroids, which were tapered over the following 2 months, while 5-ASA was maintained at a dosage of
2.4 g/day. Four patients started azathioprine because they had
experienced more than two relapses in the preceding year.
Eight patients needed colectomy within 3 weeks (13.8 7.6
days) due to clinical worsening, two of them were urgent due

All patients were experiencing their first episode of severe colitis. Five patients did not have a previous diagnosis
of ulcerative colitis and seven patients were referrals from
nearby hospitals. Patients with Crohns disease were not included in the study.
The demographic and disease specific characteristics of
the two groups of patients who responded to medical treatment (53) and non-responders requiring colectomy (14) are
given in Table 1. A slight prevalence of patients with more
extensive disease (P = 0.06) was observed in the group
which did not respond to therapy.
Table 2 summarises the clinical and laboratory variables
at study entry in responders and non-responders. The colectomy group had significantly increased CRP levels (P =
0.02), ESR (P = 0.003), body temperature (P = 0.001) and
pulse rate (P = 0.045). Logistic regression analysis identified ESR >75 mm/h and temperature >38 C on admission
capable of predicting medical treatment failure (OR = 4.9,
CI 1.220.5 and OR = 8.6, CI 2.135.8, respectively).
Sigmoidoscopy was performed in 60 patients and
documented severe endoscopic lesions in 53 patients.
Non-responders had significantly more large and deep ulcers (90%) and pseudopolypoid islands (73%) than responders (33 and 41%, respectively) (P < 0.05). However, the
severity of the endoscopic lesions or the extent of the disease beyond the sigmoid colon did not influence prognosis
when regression analysis was applied.
All patients initially received methyl-prednisolone
1 mg/kg per day i.v. for at least 1 week (7 1.8 days).

Table 1
Baseline demographic and clinical characteristics of 67 patients with severe ulcerative colitis

Responders (n = 53)
Non-responders (n = 14)
a
b

Mean age
(range)

M/F

45 19
44 14

36/17
8/6

No previous
UC diagnosis

Mean duration of
disease (years)

7.6 7
5.0 3

Extent of diseasea
Left sided

Extensive

Pancolitis

25
10

17
2

10
3

Mean CAIb at
admission
12.8 0.8
13.2 0.7

Maximal extent during the course of the disease.

CAI: clinical activity index.

Table 2
Clinical and laboratory characteristics at study entry
Hb (g/dl)

WBC
(109 per l)

ESR
(mm/h)

CRP (mg/l)

PLT
Albumin
(109 per l) (g/l)

Pulse
rate

Bowel
movements

Responders
11.6 2.1 11,496 5662 52 29 48.3 47.5 399 132 30.1 7.2 84 13 10 4
(n = 53)
Non-responders 10.8 1.9 11,587 4522 80 32 97.3 110
467 173 26.4 6.2 93 14
11 3
(n = 14)

P < 0.005;

< 0.05.

Temperature
( C)

Follow up

37.0 0.8

22 months
(255)
16.5 days
(831)

37.7 1.0

464

L. Benazzato et al. / Digestive and Liver Disease 36 (2004) 461466

Table 3
Clinical and laboratory characteristics of patients responding and not responding to steroids at day 5

Responders
(n = 34)
Non-responders
(n = 19)

Hb (g/dl)

ESR (mm/h)

CRP (mg/l)

PLT (109 per l)

Pulse rate

Bowel
movements

Temperature ( C)

Follow up

11.1 1.7

55 27

29 36

405 167

76 9

42

36.7 0.3

20 months (250)

11.3 1.7

44 29

39 40

380 125

81 10

7 3

36.7 0.3

29 months (448)

P = 0.001.

Fourteen patients (21%) required colectomy within 4 weeks,

34 (54%) obtained clinical response with steroids and in 25
patients cyclosporin was added to steroids: six patients were
referred to surgery in the short term despite cyclosporin
while the remaining 19 patients initially responded to cyclosporin. Eleven patients remained long-term responders
(mean 25.9 months, range 246 months) and maintained
remission with azathioprine, while eight patients experienced another severe relapse 8.3 2.2 months after the
first episode despite azathioprine, and all of them were
subsequently operated on.
3.2. Predictors of response to corticosteroid therapy
Clinical and laboratory variables of responding (34) and
non-responding (19) patients are shown in Table 3. The effect of glucocorticoid treatment on bowel frequency in responding and non-responding patients is shown in Fig. 2.
Clinical and laboratory variables were graded between admission and the 5th day: bowel movements and haemoglobin
levels were statistically different between responders and
non-responders (P = 0.001 and P = 0.03, respectively).
Logistic regression analysis identified the change in the number of bowel movements from admission (mean 10.5, median 10, range 420) to day 5 (mean 3.6, median 3, range
18) capable of predicting the response to steroid therapy

Bowel movements (median and CI 95%)

resp

non resp

14
12
10
8
6
4
2
0
1

days

Fig. 2. Effect of glucocorticoid treatment on bowel frequency (median

and CI 95%) in patients responding and not responding to corticosteroids
therapy during the first 5 days of admission.

with 63% sensitivity, 88% specificity, 75% PPV and 81%

NPV.
None of the clinical or biochemical variables examined
were able to predict the likelihood of cyclosporin success.
No mortality was observed during the study.
4. Discussion
We identified significant differences on admission between patients needing surgery and patients improving with
intensive medical therapy. In detail, ESR and temperature
on admission were the variables capable of predicting the
response to medical therapy.
Severe colitis has a significant rate of morbidity and therefore early recognition of the severity would be helpful in
predicting which patients have good chances of a positive
response to medical therapy and which should undergo early
colectomy. A few studies have already addressed this issue in
the past, most of them retrospectively. Lennard-Jones found
an association between the need for surgical treatment if mucosal islands and toxic dilation were present at the X-ray, if
pulse rate >100 and temperature >38 C were present in the
first 24 h, or if >9 bowel movements per day and albumin
<30 g/l persisted at day 4 [2]. Chakravarty characterised albumin levels and number of bowel movements at admission
as predictive of response. However, data were only analysed at day 1, and modalities and duration of therapy are
not given in detail [11]. Lindgren et al. retrospectively analysed 97 episodes of moderatesevere colitis needing hospitalisation and treated with betamethasone and found that
CRP 25 mg/l and >4 bowel movements per day at day 3
independently predicted a high risk for colectomy [12]. We
analysed data on day 3 and 5 applying Lindgrens criteria:
we obtained 73% efficiency when considering the response
to steroids at day 5 and 79% efficiency when considering the
reduction in the number of bowel movements. A novel scoring system has been recently developed in order to stratify
the risk of any given patient admitted with severe colitis, but
the data need further validation in a prospective cohort [13].
We found that the clinical parameters regarding disease
extension and severity during the acute attack did not have
a significant impact on prognosis.
Carbonnel et al. recently retrospectively analysed the
outcome of 85 patients admitted to a tertiary care referral
centre [4]. The multivariate analysis found the presence

L. Benazzato et al. / Digestive and Liver Disease 36 (2004) 461466

of Truelove and Witts criteria (P = 0.0018), an attack

that had lasted more than 6 weeks (P = 0.001) and severe endoscopic lesions (P = 0.0007) as associated with
a failure of medical therapy by 85%. In our experience
the presence of severe endoscopic lesions was associated
with the risk of urgent colectomy, while the parameter was
not discriminatory in the long term if an initial therapeutic
response occurred. Travis et al. prospectively followed 51
consecutive episodes of severe ulcerative colitis [3], identified CRP and the number of daily movements as predictors
of response to intensive medical care. We also observed
statistically different values of CRP in operated versus
unoperated patients, but logistic regression discarded this
parameter.
Cyclosporin is generally started after 1 week on steroids
in severe attacks of ulcerative colitis, although a recent paper
reported approximately the same results with cyclosporin
used as first line therapy [14]. In our experience, the first line
treatment with steroids has a good rate of response (51%)
and we identified candidates for cyclosporin in patients with
less than 40% reduction in bowel movements between admission and day 5. Moreover, cyclosporin induced remission in the short term in a high proportion of patients, but in
the long term the response was not as good as described in
previous studies. Overall, 56% of the patients treated with
cyclosporin needed surgery in the long term.
These results are in agreement with the retrospective study
by Rowe et al. [15], who observed that 70% of patients
treated with cyclosporin had a satisfactory immediate response, but overall 66% needed colectomy during the follow
up. The study by Rowe et al. was able to characterise prognosis by looking at the presence of bands of neutrophils on
admission. We were unable to identify any clinical or biochemical marker capable of predicting a good response to
cyclosporin. In another retrospective study, AZA/6MP entered the routine management for patients who entered remission after cyclosporin [7,8]. Fernndez-Banres et al. in
1996 [8] using azathioprine in 10 patients with cyclosporin
induced-remission observed one disease relapse during a
mean follow up of 16.3 months. Cohen et al. [16] in 1999
administered cyclosporin i.v. to 42 patients with severe UC,
and AZA/6MP to 25 of the 36 responding patients. In a 18
month follow up, 20% of the patients required colectomy,
versus 45% of those not receiving AZA/6MP.
In our clinical practice, steroids are tapered once remission has been achieved and azathioprine is added after 23
months on cyclosporin. The duration of cyclosporin treatment is prolonged up to 6 months. The management is, however, quite variable from institution to institution and recent
data suggest starting azathioprine after 7 days of cyclosporin
i.v. [17].
The possibility of contracting serious infections has limited the use of cyclosporin for non-transplant patients. Although we did not observe any major episode of infection in
the current series of patients, we recently described the case
of Aspergillus fumigatus-associated pneumonitis occurring

465

in a man with ulcerative colitis treated with cyclosporin

[18]. However, patients who have undergone colectomy after having tried cyclosporin do not seem to suffer from an
increased risk of post-surgical complications [19]. No major clinical complications occurred in our patients after the
operation.
Several side-effects were observed in patients treated
with cyclosporin: hirsutism (five patients), tremors (five patients), mild elevation of creatinin (eight patients). All the
side-effects subsided once cyclosporin was interrupted.
In conclusion, severe attacks of UC need to be promptly
recognised and objectively assessed with clinical and biochemical parameters. These simple clinical measures should
be performed daily in order to identify as early as possible
those patients who will require colectomy. The number of
bowel movements may select patients needing cyclosporin.
Additional parameters such as endoscopic and radiological
examinations may be informative but do not consistently
evaluate prognosis. We confirm efficacy and safety of cyclosporin for steroid-refractory patients: the drug in this experience had a 50% chance of saving the colon in the long
term and enabled to avoid urgent colectomy in 80% of patients. The strategy of adding azathioprine for maintaining
cyclosporin-induced remission was successful, on the whole,
except for patients resistant/intolerant to azathioprine where
the indication for long-term cyclosporin is still under debate
and deserves further investigation.
Conflict of interest statement
None declared.

References
[1] Hyde GM, Jewell DP. Review article: the management of severe
ulcerative colitis. Aliment Pharmacol Ther 1997;11:41924.
[2] Lennard-Jones JE, Ritchie JK, Hilder W, Spicer CC. Assessment of
severity in colitis: a preliminary study. Gut 1975;16:57984.
[3] Travis SPL, Farrant JM, Ricketts C, Nolan DJ, Mortsen NM, Kettlewell MGWK, et al. Predicting outcome in severe ulcerative colitis.
Gut 1996;38:90510.
[4] Carbonnel F, Gargouri D, Lemann M, Beaugerie L, Cattan S, Cosnes
J, et al. Predictive factors of outcome of intensive intravenous
treatment for attacks of ulcerative colitis. Aliment Pharmacol Ther
2000;14:2739.
[5] Jarnerot G, Rolny P, Sandberg-Gertzen H. Intensive intravenous treatment of ulcerative colitis. Gastroenterology 1985;89:100513.
[6] Meyers S, Level PK, Feuer EJ, Johnson JW, Janowitz HD. Predicting
the outcome of corticoid therapy for acute ulcerative colitis. Results
of a prospective, double blind trial. J Clin Gastroenterol 1987;9:
504.
[7] George J, Present DJ, Pou R, Bodian C, Rubin PH. Long term
outcome of ulcerative colitis treated with 6-mercaptopurine. Am J
Gastroenterol 1996;91:17114.
[8] Fernndez-Banres F, Bertrn X, Esteve-Comas M, Cabr E, Menacho M, Humbert P, et al. Azathioprine is useful in maintaining long-term remission induced by intravenous cyclosporin
in steroid-refractory severe ulcerative colitis. Am J Gastroenterol
1996;91:24989.

466

L. Benazzato et al. / Digestive and Liver Disease 36 (2004) 461466

[9] Lichtiger S, Present DH, Kornbluth A, Gelernt I, Bauer J, Galler

G, et al. Cyclosporin in severe ulcerative colitis refractory to steroid
therapy. New Engl J Med 1994;330:18415.
[10] Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Br Med J
1955;2:10418.
[11] Chakravarty BJ. Predictors and the rate of medical treatment failure
in ulcerative colitis. Am J Gastroenterol 1993;88:8525.
[12] Lindgren SC, Flood LM, Kilander AF, Lofberg R, Persson TB,
Sjodahl RI. Early predictors of glucocoticosteroids treatment failure
in severe and moderately severe attacks of ulcerative colitis. Eur J
Gastroenterol Hepatol 1998;10:8315.
[13] Ho GT, Mowat C, Goddard C, Fennel J, Shah N, Prescott R, et al.
Predicting outcome and risk assessment in severe ulcerative colitis.
Gastroenterology 2003;124:A2.
[14] Cohen RD. Intravenous cyclosporin in severe ulcerative colitis: ready
to stand alone? Gastroenterology 2001;120:15413.
[15] Rowe FA, Walker JH, Karp LC, Vasiliauskas EA, Plevy SE, Targan
SR. Factors predictive of response to cyclosporin treatment for severe,

[16]

[17]

[18]

[19]

steroid-resistant ulcerative colitis. Am J Gastroenterol 2000;95:2000

8.
Cohen RD, Stein R, Hanauer SB. Intravenous cyclosporin in ulcerative colitis: a five year experience. Am J Gastroenterol 1999;94:1587
92.
Domenech E, Garcia-Planella E, Bernal I, Rosinach M, Cabre E,
Fluvia L, et al. Azathioprine without oral cyclosporin in the long-term
maintenance of remission induced by intravenous cyclosporin in
severe, steroid-refractory ulcerative colitis. Aliment Pharmacol Ther
2002;16:20615.
Caroli A, Fregonese D, Di Falco G, DInc R. Aspergillus fumigatus
pneumonia during cyclosporin treatment for ulcerative colitis. Am J
Gastroenterol 2000;95:30167.
Hyde GM, Jewell DP, Kettlewell MG, Mortensen NJ. Cyclosporin for severe ulcerative colitis does not increase the rate
of perioperative complications. Dis Colon Rectum 2001;44:1436
40.