prospective randomized open-label multicenter phase i/ii dose escalation trial of visilizumab...

$: Prospective randomized open-label multicenter phase I/II dose escalation trial of visilizumab (HuM291) in severe steroid-refractory ulcerative colitis†$
ORIGINAL ARTICLE

Prospective Randomized Open-label Multicenter Phase I/IIDose Escalation Trial of Visilizumab (HuM291) in SevereSteroid-refractory Ulcerative Colitis

Daniel C. Baumgart, MD,* Stephan R. Targan, MD,† Axel U. Dignass, MD,‡ Lloyd Mayer, MD,§

Gert van Assche, MD,¶ Daan W. Hommes, MD,k Stephen B. Hanauer, MD,** Uma Mahadevan, MD,††

Walter Reinisch, MD,‡‡ Scott E. Plevy, MD,§§ Bruce A. Salzberg, MD,¶¶ Alan L. Buchman, MD,kk

Grigor M. Mechkov, MD,*** Zahariy A. Krastev, MD,††† James N. Lowder, MD,‡‡‡

Matthew B. Frankel, MD,‡‡‡ and William J. Sandborn, MD§§§

Background: Visilizumab is a humanized IgG2 monoclonal

anti-CD3 antibody. We evaluated its safety and dose response in

severe intravenous steroid-refractory ulcerative colitis (UC).

Methods: In all, 104 patients were treated. In Stage I, 73

patients were randomly assigned to receive intravenous visilizu-

mab 5, 7.5, 10, or 12.5 lg/kg/day for 2 consecutive days. In Stage

II, 33 patients received visilizumab at the optimal clinical dose

(OCD) of 5 lg/kg/day for 2 days. Symptomatic response and

remission were defined by the modified Truelove–Witts severity

index. Clinical response and remission were defined by the Mayo

score.

Results: The rates of symptomatic response at day 15 in the 5,

7.5, 10, or 12.5 lg/kg dose groups were 71%, 70%, 50%, and

61%, respectively, in Stage I and in 54% in Stage II. The sympto-

matic remission rates were 35%, 5%, 22%, and 11% in Stage I

and 18% in Stage II. The rates of clinical response at day 30 in

the 5, 7.5, 10, or 12.5 lg/kg dose groups were 71%, 65%, 50%,

and 67%, respectively, in Stage I and 55% in Stage II. The clini-

cal remission rates were 6%, 5%, 0%, and 11% in Stage I and

6% in Stage II. All patients experienced adverse events. Serious

adverse events included abdominal abscess, cytomegalovirus

infection, atrial fibrillation, herpes zoster, and esophageal

candidiasis.

Conclusions: Treatment with visilizumab induced symptomatic

response and clinical response. Results with 5 lg/kg/day were

similar to those observed with higher doses (NCT00267306 at

www.clinicaltrials.gov).

(Inflamm Bowel Dis 2010;16:620–629)

Key Words: ulcerative colitis, anti-CD3, monoclonal antibody

U lcerative colitis (UC) is an inflammatory disease of the

colon.1,2 Patients with severely active UC require hos-

pitalization for intravenous steroids.1,3 Patients who fail in-

travenous steroids may respond to cyclosporine or tacroli-

mus.4,5 Patients who fail salvage therapy undergo

colectomy.3,6 One potential novel treatment strategy is to

target T-cells (blockade of T-cell cytokines, T-cell migra-

tion, and T-cell surface receptors.7–11

Visilizumab is a humanized IgG2 monoclonal anti-

body to the CD3 complex of the T-cell receptor.12 In vitro,

visilizumab causes activated CD3 cells to undergo apopto-

sis, while leaving quiescent T-cells unchanged.12–14 Visili-

zumab has been investigated in Phase I and/or Phase II

Additional Supporting Information may be found in the online version

of this article.

See online supplement for full investigator list.

Received for publication July 12, 2009; Accepted July 15, 2009.

From the Departments of Medicine, Divisions of Gastroenterology and

Hepatology, *Charite Medical School of the Humboldt-University of Berlin,

Germany, †Cedars Sinai Medical Center, University of California, Los

Angeles, CA, USA, ‡Markuskrankenhaus, Frankfurt, Germany, §Mount

Sinai School of Medicine, New York, NY, USA, ¶Universitaire

Ziekenhuizen Gasthuisberg, Leuven, Belgium, kUniversity of Leiden,

Leiden, The Netherlands, **University of Chicago, Chicago, IL, USA,††University of California San Francisco, CA, USA, ‡‡Medizinische

Universitat Wien, Vienna, Austria, §§University of Pittsburgh, Pittsburgh,

PA, USA, ¶¶Atlanta Gastroenterology Associates, Atlanta, GA, USA,kkFeinberg School of Medicine, Northwestern University, Chicago, IL,

USA, ***Military Medical Academy, Sofia, Bulgaria, †††University of Sofia,

Sofia, Bulgaria, ‡‡‡Facet Biotech Corp. (formerly Protein Design Labs and

then PDL BioPharma), Redwood City, CA, §§§Mayo Clinic, Rochester, MN,

USA.

Disclosure/Grant Support/Writing Assistance: The clinical trials of

visilizumab were supported through research grants from Facet Biotech

Corp. D.C.B., S.R.T., B.A.S., L.M., D.H., S.H., U.M., W.R., S.E., A.U.D.,

G.V.A., A.B., G.M., Z.K., and W.J.S. were investigators, consultants for,

and received research support from Facet Biotech Corp. J.N.L. and M.B.F.

are or were employees of Facet Biotech Corp.

Reprints: Dr. Daniel C. Baumgart, MD, PhD, Charite Medical Center,

Virchow Hospital, Medical School of the Humboldt-University, Department

of Medicine, Division of Gastroenterology and Hepatology, Augustenburger

Platz 1, 13353 Berlin, Germany (e-mail: [email protected])

Copyright VC 2009 Crohn’s & Colitis Foundation of America, Inc.

DOI 10.1002/ibd.21084

Published online 27 August 2009 in Wiley InterScience (www.

interscience.wiley.com).

Inflamm Bowel Dis � Volume 16, Number 4, April 2010620

trials in renal transplantation recipients, bone marrow trans-

plant recipients with graft-versus-host disease (GvHD), and

patients with UC.10,15,16 A previous uncontrolled Phase I

trial of visilizumab at doses of 10 and 15 lg/day adminis-

tered for 2 consecutive days demonstrated that visilizumab

might be effective for the treatment of severely active in-

travenous steroid-refractory UC and that dose-limiting tox-

icity from cytokine release syndrome (CRS) occurred at

the 15 lg/kg dose.11 In the Phase I trial patients with

Epstein–Barr virus (EBV) DNA above the limit of quanti-

tation were excluded because treatment with another anti-

CD3 antibody, muromonab (OKT3), had been associated

with reactivation of latent EBV and EBV-associated post-

transplant lymphoproliferative disease (PTDL). Here we

report the results of a randomized Phase I/II trial designed

to determine the optimal dose of visilizumab for the treat-

ment of severely active intravenous steroid refractory UC

and to determine the safety and tolerability of administer-

ing visilizumab in patients with whole blood EBV DNA

present at concentrations up to 30,000 copies/mL.

MATERIALS AND METHODS

PatientsThis randomized, open-label, Phase I/II dose escala-

tion multicenter trial was conducted at 21 sites in Europe

and North America. Eligible patients had severely active

UC with a Modified Truelove–Witts Severity Index

(MTWSI) score of 11 to 21 despite treatment with intrave-

nous steroids for a minimum of 5 days within 60 days prior

to study entry.17,18 For patients who had received intrave-

nous steroids, failed to respond, and were either receiving

oral steroids or had discontinued steroids, intravenous ste-

roids were resumed at the time of enrollment in the study.

The MTWSI score (also known as the Lichtiger score) is a

21-point symptomatic index that measures diarrhea, noctur-

nal diarrhea, visible blood in stool, fecal incontinence, ab-

dominal pain or cramping, general well-being, abdominal

tenderness, and the need for antidiarrheal drugs.4,17,18 The

diagnosis of UC was confirmed by colonoscopy performed

within 36 months prior to study entry. Males and females

between the ages of 16 and 70 years were included.

Patients were excluded if any of the following criteria were

present: UC requiring immediate surgical, endoscopic, or

radiologic interventions, including massive hemorrhage,

perforation and sepsis, suppurative complications (intraab-

dominal or perianal abscesses), or toxic megacolon; history

of total proctocolectomy, or subtotal colectomy with ileor-

ectal anastomosis; presence of ileostomy; total leukocyte

cell count less than 2.5 � 103/lL, platelet count less than

150 � 103/lL, or hemoglobin less than 8 g/dL; those with

serious infections including any incidence of opportunistic

infections within the past year; those with a positive stool

test for Clostridium difficile within 10 days prior to treat-

ment with visilizumab; those who had received a live vac-

cine within 6 weeks prior to study entry; those with a his-

tory of thrombophlebitis or pulmonary embolus; those with

significant organ dysfunction, including cardiac, renal,

liver, central nervous system, pulmonary, vascular, GI, en-

docrine, or metabolic dysfunction; those with a history of

coronary artery disease within 6 months prior to study

entry; those with a history of PTLD or malignancy other

than nonmelanoma skin cancer or carcinoma in situ of the

cervix that had been adequately treated; pregnant women

or nursing mothers; those with seropositivity for infection

with human immunodeficiency virus-1, hepatitis B virus

surface antigen, or hepatitis C virus antibody; those with

an EBV DNA load �5000 copies/mL in Stage I and

�30,000 copies/mL in Stage II (see below for details on

Stages I and II); treatment with any investigational drugs

or therapies within 60 days prior to study entry; treatment

with any monoclonal antibody therapy within 60 days prior

to study entry; treatment with cyclosporine or tacrolimus

within 3 months prior to study entry; and those with a his-

tory of seizures, a history of both chronic and current treat-

ment with anticonvulsant medication, and no documenta-

tion of therapeutic blood level(s) of anticonvulsant

medication within 7 days before study enrollment.

Study DesignThis was a Phase I/II, open-label study conducted in

2 stages. In Stage I, patients were stratified according to

the presence and load of detectable EBV DNA in whole-

blood samples (EBV status). Patients with EBV loads

�5000 copies/mL were excluded from Stage I of the trial.

Patients with EBV loads below the limit of detection by a

relevant assay were randomized equally to treatment with

intravenous visilizumab at doses of 5, 7.5, 10, or 12.5 lg/kg (Suppl. Tables 1 and 2).

An optimum clinical dose (OCD) was determined at

the end of Stage I. The OCD was selected based on the

balance of clinical response among the dose groups with

acceptable safety profiles in the EBV-negative, EBV-posi-

tive, and combined patient populations. A true symptomatic

response rate (based on the MTWSI criteria, see below)

�60% was proposed for the selection criteria. If there were

multiple dose groups with symptomatic responses �60%

then the dose group with the largest proportion of sympto-

matic responses, or the one with the lowest dose level/best

safety profile was selected for the Stage II dosing. If the

observed number of symptomatic responses were �40%

for all of the dose groups, then no Stage II dosing was to

be conducted. In Stage II, additional patients were enrolled

at the OCD and the exclusion criteria for EBV load was

increased to �30,000 copies/mL.

Inflamm Bowel Dis � Volume 16, Number 4, April 2010 Visilizumab in Severe UC

621

Study Drug and Pre- and PosttreatmentMedications

Patients received visilizumab at doses of 5, 7.5, 10,

or 12.5 lg/kg/day administered as a rapid intravenous

bolus on 2 consecutive days.11 As previously reported,

ondansetron, acetaminophen, and diphenhydramine were

administered within 1 hour prior to each dose of study

drug and as needed after dosing for treatment of cytokine

release syndrome symptoms.11 Patients were also hydrated

with at least 1 L of intravenous fluids before receiving

study drug and hydration was continued after dosing. Me-

peridine or morphine sulfate were used for treatment of

chills.

Concomitant Medical TherapiesPatients were permitted to continue stable doses of

oral or rectal 5-aminosalicylate (5-ASA). Patients receiving

intravenous steroids at the time of enrollment into the study

continued intravenous steroids. For patients who had

received intravenous steroids within the last 60 days, failed

to respond, and were either receiving oral steroids or had

discontinued steroids, intravenous steroids were resumed at

the time of enrollment. On day 3 of the study, all patients

were switched to a dose of oral steroid equivalent to their

intravenous dose and could continue on this dose for up to

5 days. Thereafter, the dose was to be tapered by 5 mg/

week in order to achieve a dose of 20–40 mg/day by day

30. The taper was to be continued as tolerated until treat-

ment with steroids was discontinued. Azathioprine and 6-

mercaptopurine were discontinued at study entry. Azathiop-

rine 2–2.5 mg/day or 6-mercaptopurine 1–1.5 mg/day could

be resumed or started after day 42 if CD3þCD4þ T cells

had recovered to a level of 200 cells/lL.

Patient Schedule and Efficacy/Safety EvaluationsPatients were assessed at baseline and days 1, 2, 3, 5,

8, 15, 22, 30, 60, 90, 180, 270, and 360. The MTWSI4 score

was determined at baseline and at days 1, 8, 15, 22, 30, 60,

90, 180, 270, and 360. Symptomatic response was defined

as an MTWSI score <10 points. The Mayo19 score was

determined at baseline and days 30 and 180. Clinical

response was defined as a decrease from baseline in the

total Mayo score by at least 3 points. Clinical remission

was defined as a total Mayo score of 2 points or lower with

a score of 0 on rectal bleeding and sigmoidoscopy compo-

nents. Symptomatic remission was defined as an MTWSI

score �3 points. At each visit, adverse events (AEs), the

occurrence of colectomy, and concomitant medications

were recorded; blood samples were collected for laboratory

evaluations at selected visits. AEs were classified according

to the medical dictionary for regulatory activities

(MeDRA).20 The severity of AEs was graded based on the

common toxicity criteria of the National Cancer Institute

(Bethesda, MD).21 Safety evaluations included vital signs,

physical examinations, hematology, serum chemistry, EBV

and cytomegalovirus DNA, and flow cytometry for

CD3þCD4þ T cells. Samples for determination of visilizu-

mab serum levels were collected on days 1, 2, and 3. Sam-

ples for determination of antibodies to visilizumab were

collected at days 1, 3, 8, 15, 30, 60, 90, and 270.

The following events were considered dose-limiting

toxicity (DLT). 1) If CD3þCD4þ T cells had not recovered

by day 60. T-cell recovery was defined as a return to �200

cells/lL or, if baseline CD3þCD4þ T cells were <200

cells/lL (either at screening or on day 1 at 15 minutes

before dosing), a return to the baseline value. 2) Grade 3

infusion-related toxicities observed beyond 24 hours after

the second infusion. 3) Opportunistic infections, PTLD, or

other malignancies. Opportunistic infections included Pneu-mocystis jirovecii pneumonia; reactivation of latent viral

infections; serious systemic infections, such as sepsis and

meningitis; and infections requiring IV antibiotics. 4) Any

grade 4 toxicity, including infusion reactions and CRS. 5)

If whole-blood EBV DNA load had not returned to a base-

line level by day 60.

Statistical MethodsThis study was an open-label dose-escalation study

designed to determine the optimal clinical dose and to

determine the safety and tolerability of visilizumab in

patients with baseline elevation of EBV viral load. The

results of this study were primarily intended to guide the

design of a subsequent Phase III trial. As such, no formal

statistical hypothesis testing was performed for this study.

Descriptive statistics and 95% confidence intervals (CI)

were employed where appropriate for an intention-to-treat

data analysis.

Ethical ConsiderationsThe Institutional Review Board at each site approved

the protocol. All patients gave written informed consent.

RESULTS

Disposition of the PatientsA total of 187 patients were screened for the study

and 110 patients met the criteria for enrollment and were

randomized (Fig. 1). Four randomized patients did not

receive visilizumab. Of the remaining 106 patients 73 were

dosed with visilizumab in Stage I of the trial and 33 in

Stage II (Suppl. Table 3).

Overall, 45 patients completed the study to 1 year.

For the 61 patients who discontinued, the reasons for dis-

continuation included: disease relapse or no response (n ¼40); patient decision to withdraw (n ¼ 6); adverse events

Inflamm Bowel Dis � Volume 16, Number 4, April 2010Baumgart et al

622

(atrial fibrillation n ¼ 1, grade IV hypotension n ¼ 1, co-

lon perforation n ¼ 1); investigator judgment (n ¼ 1); lost

to follow-up (n ¼ 5); and other reasons (n ¼ 6) (Fig. 1).

Characteristics of the PatientsThe baseline characteristics of the study population

were consistent with severe UC18 (Table 1).

Stage I

EfficacyThe rates of symptomatic response and symptomatic

remission are shown in Table 2. The distribution of flexible

sigmoidoscopy subscores of the Mayo score are shown in

Supplemental Table 4. Overall, the proportion of patients

with a Mayo score in the severe category dropped from

78% (57/73) at baseline to 17% (10/58) at day 30 and 3%

(1/33) at 6 months. The 5 lg/kg dose group had the best

mucosal response to treatment. At baseline, 88% of the

5 lg/kg cohort had severe disease, but only 6% by day 30.

Disease progression with time after therapy occurred

in 58% of patients; 36% underwent colectomy and 22%

required rescue medical therapy. There was no clear rela-

tionship between the rate of disease progression and dose

level. The lowest rate of colectomy was observed in the

10 lg/kg dose group, and the highest rate of colectomy

was observed in the 12.5 lg/kg dose group. The 7.5 lg/kgdose group had the longest median time to disease progres-

sion (340 days). Figure 2 shows the median time to disease

progression in each dose group. The proportions (95% CI)

of patients who were able to taper steroids to 5 mg day in

the 5, 7.5, 10, or 12.5 lg/kg dose groups were 35% (14%–

61%), 30% (11%–54%), 44% (21%–69%), and 27% (9%–

53%), respectively.

SafetyAn overview of the adverse events reported in Stage

I is reported in Table 3. In all, 514 AEs and 9 serious

adverse event (SAEs) were reported. Every patient reported

at least 1 AE, and 96% of patients reported at least 1 AE

FIGURE 1. Efficacy and safety evaluations included 106 patients.


623

TABLE 1. Patient Demographic and Baseline Data

Characteristics

Stage I Stage II

Dose Group lg/kg

5n ¼ 17

7.5n ¼ 20

10n ¼ 18

12.5n ¼ 18

Totaln ¼ 73

Totaln ¼ 33

Gender: n (%)

Male 9 (53) 14 (70) 10 (56) 9 (50) 42 (58) 19 (58)

Female 8 (47) 6 (30) 8 (44) 9 (50) 31 (42) 14 (42)

Age (yr): mean (�SD) 43.5 (11.4) 32.7 (8.0) 39.7 (13.6) 40.1 (10.2) 38.8 (11.4) 38.4 (13.3)

Ethnicity: n (%)

Asian 1 (6) 0 (0) 0 (0) 1 (6) 2 (3)

Black of African heritage 1 (6) 0 (0) 0 (0) 1 (6) 2 (3) 1 (3)

Caucasian 14 (82) 18 (90) 17 (94) 15 (83) 64 (88) 32 (97)

Multi-racial 0 (0) 0 (0) 1 (6) 0 (0) 1 (1)

Other 1 (6) 2 (10) 0 (0) 1 (6) 4 (6)

Weight (kg): mean (�SD) 74.2 (17.2) 74.9 (17.6) 74.7 (12.3) 68.9 (24.1) 73.2 (18) 73.6 (17.6)

Height (cm): mean (�SD) 170.4 (9.2) 174.1 (8.9) 172.7 (9.9) 170.7 (11.1) 172.1 (9.7) 173 (10.5)

Disease extent n (%)

Less than left flexure 14 (28) 4 (20) 5 (28) 4 (22) 27 (26) 14 (42)

Less than right flexure 6 (12) 2 (10) 2 (11) 2 (11) 12 (11)

Pancolitis 29 (58) 14 (70) 11 (61) 12 (67) 66 (62) 19 (58)

Unknown 1 (2) 0 (0) 0 (0) 0 (0) 1 (1) 0 (0)

Disease duration (yr)

Mean (�SD) 8.5 (10.1) 6.6 (7.67) 2.5 (3.47) 4.6 (4.54) 5.5 (7.13) 4.9 (4.5)

Median (min, max) 4.4 (0.3, 36.1) 3.3 (0.9, 32.9) 1.0 (0, 13.9) 3.3 (0.1, 14.4) 2.7 (0, 36.1) 3.1 (0.9, 17.1)

MTWSI score

Mean (�SD)a 13.9 (1.7) 13.8 (2.0) 13.1 (2.1) 13.8 (2.3) 13.6 (2.0) 14 (2)

Median (min, max) 14 (11, 17) 13 (11, 18) 12.5 (10, 17) 14 (11, 18) 13 (10, 19) 13.0 (11, 18)

Mayo mucosal subscore

Mean (�SD)b 2.9 (0.33) 2.7 (0.57) 2.6 (0.50) 2.9 (0.24) 2.8 (0.45) 2.8 (0.4)

Mayo score

Mean (�SD)c 10.8 (1.6) 10.6 (1.3) 9.7 (1.3) 10.4 (1.8) 10.3 (1.5) 10.3 (1.3)

Median (min, max) 11 (6, 12) 11 (7, 12) 10 (7, 12) 11 (5, 12) 11 (5, 12) 11.0 (7, 12)

EBV viral loadd (copies/mL)e

Mean (�SD) 369 (313) 577 (700) 1068 (1859) 2353 (6996) 1088 (3622) 10635 (40992)

Median (min, max) 250 (200,1337)

250 (200,2648)

250 (200,6815)

250 (200,30131)

250 (200,30131)

250 (200,212066) 28

Current therapies

5-ASA 10 (59%) 12 (60%) 10 (56%) 16 (89%) 48 (66%) 26 (79%)

Systemic steroids 15 (88%) 18 (90%) 18 (100%) 16 (89%) 67 (92%) 32 (97%)

Rectal steroids 1 (6%) 2 (10%) 1 (6%) 2 (11%) 6 (8%) 3 (9%)

CD4þ T-cell count (copies/lL)Mean (�SD) 429 (264) 439 (347) 555 (330) 541 (422) 489 (344) 424 (376)

Current smokers, n (%) 4 (24) 3 (15) 1 (6) 1 (6) 9 (12) 2 (6)

aScale: 0 to 21, where �3 ¼ remission, <10 ¼ positive response, 11 to 21 ¼ active disease.bScale: 0 (normal mucosa) to 3 (very active disease).cScale: 0 to 12, where: 0 to 2 ¼ remission or minimally active disease, 3 to 5 ¼ mildly active disease, 6 to 10 ¼ moderately active disease, 11 to 12 ¼moderately or severely active diseasedEight patients had protocol violations consisting of entry into the study even although they had an EBV titer higher than the allowable limit for that stageof the study.eRounded to the nearest whole copy.


624

related to visilizumab. Overall, 63% of AEs were related to

visilizumab. The most frequently reported AEs were chills

(53%), headache (49%), and pyrexia (41%). Most of the

AEs were reported as mild (grade 1) or moderate (grade

2). No AE category significantly increased in frequency as

the dose level increased. One grade 4 AE (atrial fibrilla-

tion) and 60 grade 3 AEs occurred in 30 patients. The low-

est frequency of grade 3 or 4 AEs was observed in the 5

lg/kg dose group (25%). Eight patients had 9 SAEs (ab-

dominal abscess, gastroenteritis, hypotension, cytomegalo-

virus infection, psychotic disorder, atrial fibrillation, esoph-

ageal candidiasis, anemia, hemorrhoids), 5 of which were

assessed as related to study drug (abdominal abscess, hypo-

tension, cytomegalovirus infection, atrial fibrillation, esoph-

ageal candidiasis). The lowest dose group was associated

with the most SAEs. The proportion of overall AEs was

higher in the patients who were stratified as EBV (þ) at

screening (Suppl. Table 5). The proportions of patients

with grade 3 and 4 AEs was higher in the EBV (þ) cohort

than in the EBV (�) cohort (56% versus 35%). Although

the number of DLTs did not increase linearly with dose

level, the highest dose level (12.5 lg/kg) corresponded to

the highest proportion of patients reporting DLTs (39%)

and the lowest dose level (5 lg/kg) corresponded to the

lowest proportion of DLTs (17%). Thirty-six infections

were reported in 24 patients, and 10 were assessed as

related to visilizumab (candidiasis, nasopharyngitis, bron-

chitis, Clostridium difficile, Herpes simplex, urinary tract

infection). The proportion of infections did not increase

with increasing dose level.

The proportion of patients who experienced cytokine

release syndrome symptoms in Stage I is reported in Sup-

plemental Table 6.

The CD4þ-T-cell levels were reduced to a median

minimum value of 4 cells/lL (range 0–49). The median

time to recovery was 15 days (range 2–122). Overall, 92%

of subjects had recovered by the day 60 visit. When CD4þ-T-cell recovery was examined in relation to dose level, a

reduced rate of recovery was observed at the higher doses.

The whole-blood EBV DNA levels became transi-

ently positive in most patients whose EBV copy number

was negative (i.e., <200–250–copies/mL whole blood) at

baseline, and EBV load transiently increased (by 7–17-

TABLE 2. Summary of Symptomatic Response and Remission and Clinical Response and Remission Over Time, Stage I

Timepoint Symptomatic Response n (%) Symptomatic Remission n (%)

Dose Level (lg/kg)5

n ¼ 177.5

n ¼ 2010

n ¼ 1812.5

n ¼ 185

n ¼ 177.5

n ¼ 2010

n ¼ 1812.5

n ¼ 18

MTWSI

Day 8 11 (65) 14 (70) 10 (56) 10 (56) 2 (12) 1 (5) 2 (11) 4 (11)

Day 15 12 (71) 14 (70) 9 (50) 11 (61) 6 (35) 1 (5) 4 (22) 4 (11)

Day 22 12 (71) 11 (55) 11 (61) 11 (61) 5 (29) 2 (10) 5 (28) 4 (22)

Day 30 12 (71) 14 (70) 11 (61) 12 (67) 5 (29) 5 (25) 4 (22) 5 (28)

Day 60 10 (59) 12 (60) 8 (44) 7 (39) 5 (29) 4 (20) 4 (22) 3 (17)

Day 90 10 (59) 14 (70) 7 (39) 8 (44) 4 (24) 5 (25) 4 (22) 3 (17)

6 months 9 (53) 10 (50) 5 (28) 7 (39) 8 (47) 4 (20) 4 (22) 4 (22)

9 months 8 (47) 8 (40) 5 (28) 4 (22) 5 (29) 6 (30) 3 (17) 3 (17)

1 year 8 (47) 7 (35) 5 (28) 4 (22) 6 (35) 3 (15) 3 (17) 3 (17)

Mayo score Clinical Response n (%) Clinical Remission n (%)

Day 30 12 (71) 13 (65) 9 (50) 12 (67) 1 (6) 1 (5) 0 (0) 2 (11)

Day 180 9 (53) 10 (50) 11 (60) 7 (39) 4 (23) 3 (15) 2 (11) 0 (0)

FIGURE 2. Time to retreatment, colectomy, or salvage ther-apy in 106 patients. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]


625

fold) in the subjects who were EBV (þ) at baseline (Fig.

3A,B). The highest median EBV values were observed at

days 8 and 15. The EBV copies in blood generally

increased after dosing at the same time that the

CD3þCD4þ T-cells decreased. The EBV copy number then

gradually decreased in subjects with an overall median

time to recovery to within ½log of baseline of 10 days.

Overall, 92% of patients had EBV recovery to a level of

200 cells/lL or baseline value by day 60. The proportion

of patients who were recovered by day 60 decreased with

increasing dose. The EBV (�) cohort had a slightly lower

proportion of patients recovered by day 60 (90% versus

96%), and shorter median time to recovery (9 days versus

18 days).

TABLE 3. Adverse Events Stage I

Adverse Event5 lg/kg

n ¼ 17 (%)7.5 lg/kgn ¼ 20 (%)

10 lg/kgn ¼ 18 (%)

12.5 lg/kgn ¼ 18 (%)

Totaln ¼ 73 (%)

Subjects reporting AEs 17 (100) 20 (100%) 18 (100) 18 (100) 73 (100)

Chills 11 (65) 9 (45%) 10 (56) 9 (50) 39 (53)

Headache 6 (35) 15 (75%) 8 (44) 7 (39) 36 (49)

Pyrexia 4 (24) 7 (35%) 10 (56) 9 (50) 30 (41)

Ulcerative colitis 5 (29) 4 (20%) 4 (22) 7 (39) 20 (27)

Fatigue 6 (35) 6 (30%) 4 (22) 4 (22) 20 (27)

Nausea 2 (12) 6 (30%) 6 (33) 5 (28) 19 (26)

Arthralgia 1 (6) 6 (30%) 3 (17) 4 (22) 14 (19)

Back Pain 2 (12) 5 (25%) 3 (17) 3 (17) 13 (18)

Dizziness 4 (24) 2 (10%) 2 (11) 4 (22) 12 (16)

Tachycardia 2 (12) 4 (20%) 0 (0) 5 (28) 11 (15)

Abdominal Pain 2 (12) 5 (25%) 0 (0) 2 (11) 9 (12)

Dyspepsia 1 (6) 2 (10%) 2 (11) 4 (22) 9 (12)

Hemorrhoids 3 (18) 1 (5%) 2 (11) 2 (11) 8 (11)

Musculoskeletal discomfort 1 (6) 5 (25) 0 1 (6) 7 (10)

Myalgia 1 (6) 2 (10) 2 (11%) 2 (11) 7 (10)

Peripheral edema 2 (12) 1 (5) 1 (6%) 3 (17) 7 (10)

Pain 1 (6) 3 (15) 3 (17%) 0 (0) 7 (10)

Tremor 0 (0) 2 (10) 3 (17%) 2 (11) 7 (10)

Vomiting 2 (12) 0 (0) 3 (17%) 2 (11) 7 (10)

Adverse events �10% regardless of relationship.

FIGURE 3. T-cell and EBV recovery in patients who were initially EBV (�) (A) and EBV (þ) (B). [Color figure can be viewed inthe online issue, which is available at www.interscience.wiley.com.]


626

OCDEfficacy following initial treatment with visilizumab

was comparable across the 4 dose regimens tested. How-

ever, in the 5 lg/kg dose cohort there was a trend toward

fewer grade 3 and 4 CRS events, and fewer DLTs due to

AEs, delayed T-cell recovery, and delayed restoration of

baseline EBV values. Thus, 5 lg/kg/day for 2 consecutive

days was chosen as the OCD and was used in Stage II of

this trial. The 5 lg/kg dose group had the greatest numeric

response to treatment across endpoints and was selected as

the OCD.

Stage II

EfficacyThe rates of symptomatic response and symptomatic

remission for the OCD of 5 lg/kg are shown in Table 4.

The distribution of flexible sigmoidoscopy subscores of the

Mayo score are shown in Supplemental Table 7. Overall,

the proportion of patients with a mucosal score in the

severe category dropped from 76% (25/33) of patients at

baseline to 25% (6/24) at day 30 and 17% (2/12) at 6

months. Disease progression over time occurred in 45% of

patients; 33% underwent colectomy and 12% required res-

cue medical therapy. Figure 2 shows the median time to

disease progression. The proportion (95% CI) of patients

who were able to taper steroids to 5 mg/day was 39%

(22%–57%).

SafetyAn overview of the adverse events reported in Stage

II are reported in Table 5 and Supplemental Table 8. Over-

all, 213 AEs and 4 SAEs were reported. All patients

reported at least 1 AE, and 97% of patients reported at

least 1 AE related to visilizumab. Overall, 64% of all AEs

were assessed as related to visilizumab. The most frequent

AEs were chills (54%), headache (48%), and pyrexia

(39%). Most of the AEs were reported as mild (grade 1) or

moderate (grade 2). However, 11 patients (33%) reported

12 AEs that were grade 3 (UC, arthralgia, back pain, chills,

hypertensive crisis, muscle spasm, pyrexia, urinary tract

infection) or grade 4 (colonic perforation). One life-threat-

ening SAE (colonic perforation) occurred on day 2 and

was unrelated to visilizumab. One related SAE was a grade

3 urinary tract infection occurring on day 32. Two other

SAEs (nausea and vomiting), both occurred in 1 patient on

TABLE 4. Summary of Symptomatic Response and Remission and Clinical Response and Remission Over Time, Stage II

TimepointNumber ofSubjects (N)

MTWSI Median Changefrom Baseline (Range)

No. Subjects

Symptomatic Response n (%) Symptomatic Remission n (%)

Day 8 32 �4 (�12, 0) 18 (54) 0 (0)

Day 15 31 �5 (�12, 2) 18 (54) 6 (18)

Day 22 25 �5 (�11, 1) 20 (61) 4 (12)

Day 30 25 �8 (�12, 1) 21 (64) 8 (24)

Day 60 21 �8 (�13, 2) 19 (58) 9 (27)

Day 90 18 �7 (�16, 2) 14 (42) 6 (18)

6 Months 16 �9 (�16, �1) 14 (42) 7 (21)

9 Months 13 �8 (�16, �3) 13 (39) 7 (21)

1 year 14 �9 (�15, �4) 14 (42) 7 (21)

Mayo Median Changefrom Baseline (range) Clinical Response n (%) Clinical Remission n (%)

Day 30 24 �5 (�10–0) 18 (55) 2 (6)

Day 180 12 �7 (�11–2.0) 10 (30) 4 (12)

TABLE 5. Adverse Events, Stage II

Adverse Event Total n ¼ 33 (%)

Chills 18 (55)

Headache 16 (49)

Pyrexia 13 (39)

Arthralgia 10 (30)

Fatigue 7 (21)

Dizziness 6 (18)

Nausea 6 (18)

Back pain 5 (15)

Nasopharyngitis 5 (15)

Ulcerative colitis 4 (12)

Influenza like illness 4 (12)

Rash 4 (12)

Adverse events �10% regardless of relationship.


627

day 39 and were unrelated to study drug. Twenty-four per-

cent of patients developed DLTs, which was below the

allowable limit of DLTs in Stage II. Twelve percent of

patients had delayed CD4þ-T-cell recovery. Nineteen infec-

tions developed in 17 patients, and 8 were assessed as

related to visilizumab (nasopharyngitis, gastroenteritis, oral

candidiasis, sinusitis, urinary tract infection). The propor-

tion of patients who experienced cytokine release syndrome

symptoms in Stage II is reported in Supplemental Table 9.

Overall, the CD4þ-T-cell levels were reduced to a

median minimum value of 5 cells/lL (range: 1–26). The T-

cell levels generally rebounded by day 15. In Stage II,

88% of patients had recovered by day 60 (�15 days). The

median time to EBV recovery for Stage II patients was 9

days. No patients had delayed EBV recovery.

ImmunogenicityThe proportion of patients who had measurable anti-

drug antibodies (ADAb) after the first treatment was 26%.

Neutralizing antibody response was positive in 92% of

ADAb (þ) patients. Similar rates of immunogenicity were

observed in all 4 dosing groups. In general, neutralizing

antibody (NAb) responses were first detectable near day 15

or day 30. Responses in a few patients first appeared at

later timepoints and were of lower magnitude. Approxi-

mately half of NAb-positive patients were positive 6

months after the first treatment (Suppl. Table 10). It should

be noted that all patients experienced a marked reduction

in CD4þ-T-cell levels following the first dose, indicating

that immunogenicity did not affect the pharmacodynamic

effect of the first 2 doses of visilizumab.

DISCUSSIONThe results from this open-label uncontrolled trial

demonstrated that visilizumab at doses of 5–12.5 lg/kg/dayfor 2 days might be of clinical benefit in patients with in-

travenous steroid-refractory UC. Potential clinical benefits

included symptomatic response and clinical response, a

reduction in the mucosal subscore, steroid-sparing, and a

reduction in the rate of colectomy. However, the rates of

symptomatic remission and clinical remission were low.

These data should be interpreted with caution since this

study was not designed to determine efficacy, was not suf-

ficiently powered to do so, did not have a placebo control,

and did not include formal hypothesis testing. AEs

occurred frequently and were generally greatest in the

higher-dose groups. The OCD was 5 lg/kg/day and

patients with whole-blood EBV DNA present at concentra-

tions up to 30,000 copies/mL tolerated visilizumab.

The high rates of symptomatic response, as measured

by the MTWSI score, raise questions regarding the utility

of this instrument in patients with severe intravenous ste-

roid-refractory UC. The definitions of symptomatic

response and symptomatic remission used in this trial were

similar to those used in previous clinical trials of intrave-

nous cyclosporine for the treatment of severe UC. It should

be noted that this instrument was never validated, and the

minimum important difference and the cutpoint for remis-

sion were never determined.

All patients in all stages and dose groups of this trial

experienced at least 1 AE and more than 90% of them

experienced a visilizumab-related AE. Even though no AE

category increased in frequency as the dose level increase,

there was a correlation of higher-grade (3 and 4) AEs in

the higher-dose groups. Most AEs were related to CRS.

Rising EBV titers and posttransplant lymphoprolifera-

tive disease PTLD have previously been reported with visi-

lizumab use for graft-versus-host disease (GvHD).22 PTLD

occurs frequently in the setting of EBV infection and T-

cell depletion.23,24 None of the patients in our study devel-

oped lymphoma or PTLD. The data from this trial suggest

that visilizumab may be tolerated in patients with EBV

viral loads up to 30,000 copies/mL.

After this trial to determine the OCD of visilizumab

in severe intravenous steroid-refractory UC was completed,

a placebo-controlled trial was initiated in the same patient

population, with a planned recruitment of 150 patients. The

placebo-controlled trial was discontinued prematurely by

the Data Safety Monitoring Board (DSMB) after 127

patients had been randomized when an interim analysis

demonstrated lack of efficacy, and trends toward greater

rates of colectomy and vascular disorder AEs in the visili-

zumab-treated patients.25 The placebo response rate for

symptomatic response in this study was 47% at day 45,25

again raising questions regarding the utility of the MTWSI

instrument in this patient population. Treatment with visili-

zumab induced symptomatic response and clinical response

in patients with severe steroid-refractory UC, but sympto-

matic remission and clinical remission rates were low.

Results with 5 lg/kg/day were similar to those observed

with higher doses. All patients experienced AEs.

REFERENCES1. Baumgart DC, Sandborn WJ. Inflammatory bowel disease: clinical

aspects and established and evolving therapies. Lancet. 2007;369:1641–1657.

2. Baumgart DC, Carding SR. Inflammatory bowel disease: cause andimmunobiology. Lancet. 2007;369:1627–1640.

3. Travis SPL, Stange EF, Lemann M, et al. European evidence-basedconsensus on the management of ulcerative colitis: current manage-ment. J Crohn’s Colitis. 2008;2:24–62.

4. Lichtiger S, Present DH, Kornbluth A, et al. Cyclosporine in severeulcerative colitis refractory to steroid therapy. N Engl J Med. 1994;330:1841–1845.

5. Baumgart DC, Macdonald JK, Feagan BG. Tacrolimus (FK506) forinduction of remission in refractory ulcerative colitis. Cochrane Data-base Syst Rev. 2008;CD007216.

6. Biancone L, Michetti P, Travis S, et al. European evidence-basedConsensus on the management of ulcerative colitis: special situations.J Crohn’s Colitis. 2008;2:63–92.


628

7. Rutgeerts P, Sandborn WJ, Feagan BG, et al. Infliximab for inductionand maintenance therapy for ulcerative colitis. N Engl J Med. 2005;353:2462–2476.

8. Feagan BG, Greenberg GR, Wild G, et al. Treatment of ulcerative co-litis with a humanized antibody to the alpha4beta7 integrin. N Engl JMed. 2005;352:2499–2507.

9. Creed TJ, Probert CS, Norman MN, et al. Basiliximab for the treat-ment of steroid-resistant ulcerative colitis: further experience inmoderate and severe disease. Aliment Pharmacol Ther. 2006;23:1435–1442.

10. Baumgart DC, Hommes DW, Reinisch W, et al. The phase I/II visili-zumab study. A report of safety and efficacy of treatment and retreat-ment in ulcerative colitis patients refractory to treatment with i.v.steroids (IVSR-UC). Gut. 2005;54:A57.

11. Plevy S, Salzberg B, Van AG, et al. A phase I study of visilizumab, ahumanized anti-CD3 monoclonal antibody, in severe steroid-refractoryulcerative colitis. Gastroenterology. 2007;133:1414–1422.

12. Cole MS, Stellrecht KE, Shi JD, et al. HuM291, a humanized anti-CD3 antibody, is immunosuppressive to T cells while exhibitingreduced mitogenicity in vitro. Transplantation. 1999;68:563–571.

13. Hsu DH, Shi JD, Homola M, et al. A humanized anti-CD3 antibody,HuM291, with low mitogenic activity, mediates complete and reversi-ble T-cell depletion in chimpanzees. Transplantation. 1999;68:545–554.

14. Chau LA, Tso JY, Melrose J, et al. HuM291 (Nuvion), a humanizedFc receptor-nonbinding antibody against CD3, anergizes peripheralblood T cells as partial agonist of the T cell receptor. Transplantation.2001;71:941–950.

15. Carpenter PA, Appelbaum FR, Corey L, et al. A humanized non-FcR-binding anti-CD3 antibody, visilizumab, for treatment of steroid-re-fractory acute graft-versus-host disease. Blood. 2002;99:2712–2719.

16. Norman DJ, Vincenti F, de Mattos AM, et al. Phase I trial ofHuM291, a humanized anti-CD3 antibody, in patients receivingrenal allografts from living donors. Transplantation. 2000;70:1707–1712.

17. D’Haens G, Sandborn WJ, Feagan BG, et al. A review of activityindices and efficacy end points for clinical trials of medical therapy inadults with ulcerative colitis. Gastroenterology. 2007;132:763–786.

18. Stange EF, Travis SPL, Vermeire S, et al. European evidence-basedConsensus on the diagnosis and management of ulcerative colitis: defi-nitions and diagnosis. J Crohn’s Colitis. 2008;2:1–23.

19. Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral 5-aminosali-cylic acid therapy for mildly to moderately active ulcerative colitis. Arandomized study. N Engl J Med. 1987;317:1625–1629.

20. Medical Dictionary for Regulatory Activities (MedDRA). NorthropGrumman Corp. http://www.meddramsso.com/MSSOWeb/index.htm

21. Common Toxicity Criteria. https://webapps.ctep.nci.nih.gov/webobjs/ctc/webhelp/welcome_to_ctcae.htm

22. Carpenter PA, Lowder J, Johnston L, et al. A phase II multicenterstudy of visilizumab, humanized anti-CD3 antibody, to treat steroid-refractory acute graft-versus-host disease. Biol Blood Marrow Trans-plant. 2005;11:465–471.

23. Cohen JI. Epstein-Barr virus lymphoproliferative disease associatedwith acquired immunodeficiency. Medicine (Baltimore). 1991;70:137–160.

24. Stevens SJ, Verschuuren EA, Pronk I, et al. Frequent monitoring ofEpstein-Barr virus DNA load in unfractionated whole blood is essen-tial for early detection of posttransplant lymphoproliferative disease inhigh-risk patients. Blood. 2001;97:1165–1171.

25. Sandborn WJ, Colombel JF, Frankel M, et al. A placebo-controlledtrial of visilizumab in patients with intravenous (IV) steroid refractoryulcerative colitis (UC). Gastroenterology. 2009;136(suppl 1):A-64.


629

prospective randomized open-label multicenter phase i/ii dose escalation trial of visilizumab...

Documents