Digestive Diseases and Sciences, VoL 39, No. 2 (February 1994), pp.

374-380

Efficacy of Cyclosporine in Treatment of

Fistula of Crohn's Disease
D A N I E L H. PRESENT, MD, and SIMON LICHTIGER, MD

Sixteen Crohn "s disease patients with active fistula who had failed standard medical
therapy were treated with intravenous cyclosporine. Ten patients had perirectal disease,
four had enterocutaneous fistula, and two had rectovaginal fistula. Patients were initialty
treated with intravenous cyclosporine, 4 mg/kg/day, and then switched to oral cyclospor-
ine, 6-8 mg/kg/day. Improvement was graded using the Present-Korelitz criteria, and
success was defined as moderate to total closure o f the fistula. Fourteen o f 16 patients
(88%) responded in the acute phase to parenteral cyclosporine. Closure o f fistula occurred
in seven (44%) with moderate improvement in the remaining seven (44%). Subsequently,
five patients (36%) relapsed to some degree on oral cyclosporine (three severe and two
mild relapses). Nine (64%) patients maintained their improvement in the chronic phase.
Chronic steroids could be discontinued in 6/8 (75%) of patients. Mild side effects were
common [paresthesias (75%) and hirsuitism (19%)]. A single patient had severe pares-
thesias requiring discontinuation o f therapy. Mild hypertension was noted in four (25%)
and one patient (6%) had to be withdrawn because o f nephrotoxicity, which reversed after
stopping cyclosporine. We conclude that intravenous cyclosporine is effective therapy for
perianal, rectovaginal, and enterocutaneous fistula in Crohn's disease. Its future role
awaits controlled trials as well as determination o f the risk-benefit ratio.

KEY WORDS: cyclosporine; immunosuppression; treatment; fistula; Crohn's disease.

Crohn's disease is a chronic inflammatory bowel
disease of unknown etiology. Whether as a primary
or secondary factor, the immune system is impli-
cated in the persistent inflammatory response. Sev-
eral immunosuppressive drugs, including 6-mercap-
topurine (6-MP) (1), azathioprine (2), steroids (3),
and recently methotrexate (4) have shown efficacy
in the treatment of Crohn's disease. However, peri-
rectal disease and enterocutaneous and rectovagi-
nal fistula remain a stubborn complication of the
disease and refractory to standard steroid and sul-
fasalazine therapy (5). The effective use of metro-
nidazole for rectal fistula has been reported only in
uncontrolled studies (6), and thus far only 6-mer-
Manuscript received September 8, 1992; revised manuscript
received May 5, 1993; accepted May 20, 1993.
From the Mount Sinai School of Medicine, One Gustave L.
Levy Place, New York, New York.
Address for reprint requests: Dr. Daniel H. Present, 12 East 86
Street, New York, New York 10028.
captopurine has been proven to be effective in re-
fractory fistulous patients in a controlled trial (1).
Cyclosporine, a rapidly acting and potent immu-
nosuppressive agent has multiple effects on the im-
mune system as it inhibits formation of interleukin-2
(7). Several open-label studies (8, 9) and one double
blinded study (10) have shown its effectiveness in
improving refractory Crohn's disease. In these tri-
als there has been little mention of the response of
fistula to cyclosporine. We now report the results of
a pilot study using cyclosporine in Crohn's patients
with refractory perirectal, rectovaginal, and entero-
cutaneous fistula.
MATERIALS AND METHODS
From January 1987 through January 1991, 16 patients
with fistula were selected for the trial (Table 1). All
patients had failed standard medical therapy, including
the combinations of antibiotics, steroids, and sulfasala-

374 Digestive Diseases and Sciences, Vol. 39, No. 2 (February 1994)
0163-211619410200-0374507.00/0~ 1994PlenumPublishingCorporation
C Y C L O S P O R I N E IN T R E A T M E N T O F C R O H N ' S F I S T U L A

TABLE 1. PATIENTDATA
Response Steroids
Fistula Disease Prior Initial Relapse on Well on Well off Start End of Subsequent
Patient~sex site* sitet 6-MP response$ oral CSA CSA CSA Surgery of trial trial drugs
MR (f) Cut I Yes¶ +3 Yes Yes Yes No
MC (m) P C Yes¶ +2 No Yes No No No MTX
GF (m) P IC Yes¶ +2 No Yes No Yes Yes 6-MP
SR (f) P C No¶ +2 No Yes No Yes No
SC (0 Cut IC No +3 No Yes Yes*§ No No
AU (m) P C No +3 +2 Yes Yes No Yes No
MG (m) P IC Yes¶ +1 No Yes No Yes No
NN (m) Cut I No¶ Fail No Yes No No
CM (f) P C Yes +2 No Yes No No No
WC (O P C Yes +2 No Yes No Yes No MTX
MC (f) R-V IC Yes¶ +3 Yes Yes No No
PM (f) R-V/P C Yes¶ +2 Yes No No No MTX
DR (m) P C Yes¶ +3 No Yes No No No
SM (f) P C Yes¶ +3 +2 Yes Yes No Yes No
HS (f) P C No +3 No Yes No Yes Yes
JR (f) Cut/P IC Allergy +2 No Yes No No No
*Cut = cutaneous fistula, R-V = Rectovaginal fistula, P = perianal fistula.
tI = small bowel, IC = small bowel and colon, C = colon alone.
:~Grades: +1, +2, +3, fail.
§Patient well on CSA, but developed toxicity and was discontinued. Underwent surgery.
¶Patient on Flagyl at start of trial.

zinc, and 10 of the 16 had failed an adequate course of
immunosuppressive therapy with 6-MP or azathioprine.
One patient was allergic to 6-MP. Patients were selected
on a consecutive basis but only if they had failed standard
therapy. In the month prior to the initiation of cyclospor-
ine therapy, there was no change in the regimen of med-
ications, and 6-MP was stopped at least two weeks prior
to initiating cyclosporine therapy. All patients were main-
tained on antibiotics (our standard regimen for Crohn's
patients with fistulization), seven of whom were placed
on parenteral antibiotics upon hospitalization. The dose
was kept constant throughout the hospital course. Met-
ronidazole was not added to the regimen prior to initiation
of cyclosporine but was maintained if the patient was
already receiving this antibiotic. No patient had an in-
crease in their steroid dosage, and none were placed on
parenteral steroids. No patient was placed on total paren-
teral nutrition, and all were allowed to eat a low-residue
or regular diet.
The diagnosis of Crohn's disease and extent of bowel
involvement were documented by clinical and radio-
graphic criteria; endoscopic and pathologic data were
reviewed in all cases. Each patient had an evaluation for
extent of disease within one y e a r of the current study. Of
16 patients, 10 were female and six male. The mean age
was 32.6 years (range 22-55 years), seven of the 16
patients had prior resections. Two patients had ileitis
alone, five had ileocolitis, and nine had Crohn's colitis.
Two of the 16 had rectovaginal fistula (one also had a
perirectal fistula), four had enterocutaneous fistula (one
also had a perirectal fistula), and the remaining 10 had
only perirectal fistula. All patients had normal renal func-
tion as defined by a serum creatinine of 1.0 or less and a
creatinine clearance of greater than 90 cc/min. None of
the patients were hypertensive and none had a history of
neoplasia. All patients gave informed consent before en-
tering into the trial.
Cyelosporine Treatment. On entering the study, pa-
tients were started on parenteral cyclosporine in a dose of
4 mg/kg/day by continuous infusion. Therapeutic levels of
500-700 ng/ml (polyclonal TDX assay, normal 200-800
ng/ml) were achieved in all patients. Cyclosporine levels
were evaluated every other day along with serum BUN,
creatinine, magnesium, and serum electrolytes. Changes
in dosage of cyclosporine were based on drug levels,
adverse effects, and renal function. An increase of 33% in
serum creatinine was immediately followed by a lowering
of cyclosporine dosage by 50%. After patients had clini-
cally improved, they were switched to oral cyclosporine
with an initial dose of 6 - 8 mg/kg/day. On the day prior to
discharge, a pharmacokinetic profile was established by
obtaining cyclosporine levels every 2 hr during a 10-hr
period after an oral dose. This technique aided in estab-
lishing an exact dosing schedule for outpatients. Cyclo-
sporine trough levels, CBC, and SMA-18 were monitored
once weekly for the first month, biweekly during the
second month, and monthly thereafter. No patient was
maintained on parenteral or oral hyperalimentation, and
no new medications were added during the chronic phase.
Antibiotics were occasionally rotated during the chronic
phase.
Criteria for Response. The degree of improvement or
worsening was based on the Present-Korelitz criteria (1).
Each patient was graded on a scale ranging from - 3 to
+3. Excellent (+3) improvement was defined as complete
closure of the fistula, moderate (+2) was defined as sub-
stantial improvement (decreased drainage, decreased dis-
comfort), whereas mild (+1) was defined as slight im-
provement. Severe ( - 3 ) reflected development of a new
fistula or reactivation of an old fistula, whereas - 2 rep-

Digestive Diseases and Sciences, Vol. 39, No. 2 (February 1994) 375