Presented in part at the 110th Annual Meeting of the American Gastroenterological Association Institute, Chicago IL.
Natalizumab for moderate to severe Crohn's disease in clinical practice: The Mayo Clinic Rochester experience†
Article first published online: 14 MAR 2012
Copyright © 2012 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 18, Issue 12, pages 2203–2208, December 2012
How to Cite
Kane, S.V., Horst, S., Sandborn, W.J., Becker, B., Neis, B., Moscandrew, M., Hanson, K.A., Tremaine, W.J., Bruining, D.H., Faubion, W.A., Pardi, D.S., Harmsen, W.S., Zinsmeister, A.R. and Loftus, E.V. (2012), Natalizumab for moderate to severe Crohn's disease in clinical practice: The Mayo Clinic Rochester experience. Inflamm Bowel Dis, 18: 2203–2208. doi: 10.1002/ibd.22943
- Issue published online: 15 NOV 2012
- Article first published online: 14 MAR 2012
- Manuscript Accepted: 15 FEB 2012
- Manuscript Received: 14 FEB 2012
- Crohn's disease;
Not all patients with Crohn's disease (CD) respond or maintain response to anti-tumor necrosis factor (TNF) agents and alternative treatment is necessary. Natalizumab, a monoclonal antibody to alpha-4 integrin approved for CD, has demonstrated efficacy in randomized clinical trials. We describe our experience with natalizumab in clinical practice at Mayo Clinic Rochester.
Consecutive patients prescribed natalizumab for active CD were invited to participate and were followed prospectively. Incidence of infection, hospitalization, neoplasm, or other adverse events were recorded. Clinical activity was assessed using the Harvey–Bradshaw Index at each 30-day infusion visit.
Between April 2008 and September 2010, 36 patients were prescribed natalizumab and 30 (83.3%) agreed to participate. Median disease duration was 9 years (range, 3–43). Twenty-three patients had prior exposure to two anti-TNF agents, seven to one agent. All patients experienced at least one adverse event; none of the 13 patients in whom natalizumab was stopped (43%) discontinued due to adverse events. Five patients had infusions held for infection. No patient developed progressive multifocal leukoencephalopathy (PML). Fourteen patients (46%) had clinical response. The cumulative probability of achieving complete response within 1 year was 56% (28%–73%). Four of seven patients were weaned off corticosteroids.
In our experience with natalizumab in clinical practice, adverse events were manageable and did not result in treatment cessation. No PML cases were seen and clinical response was similar to that in clinical trials. Natalizumab results in clinical benefit in patients who have active disease and have failed anti-TNF therapy (Inflamm Bowel Dis 2012;)
Crohn's disease (CD) is a chronic inflammatory condition of the gastrointestinal tract. Patients with moderate to severe disease refractory to other treatments have benefited from the introduction of biologic therapy with monoclonal antibodies to tumor necrosis factor-alpha (TNF-α). Anti-TNF-α antibodies were the first biologic therapies to be developed for CD and infliximab was approved for use in 1998.1 However, some patients with CD do not respond or lose response to anti-TNF agents. In such patients with moderate to severe disease who are refractory to anti-TNF therapy, other options are clearly needed. One such therapy is natalizumab, a monoclonal antibody to alpha-4 integrin, a cell-surface glycoprotein that plays a central a role in the migration of leukocytes across the vascular endothelium, facilitating homing to areas of gut inflammation, where they initiate and maintain this inflammation.2, 3 Natalizumab is administered as an intravenous infusion once every 4 weeks and is effective in inducing and maintaining remission in active luminal CD.4–8
Natalizumab was approved by the U.S. Food and Drug Administration (FDA) in 2008 for use as monotherapy in the treatment of patients with moderate to severe CD with evidence of inflammation and who have had inadequate response to or are unable to tolerate conventional therapies, including anti-TNF agents. In clinical trials, natalizumab was well tolerated in patients with multiple sclerosis (MS) and CD.4–8 It was initially marketed for MS in 2005. However, in early 2005 three cases of progressive multifocal leukoencephalopathy (PML), an opportunistic infection caused by the JC virus that typically occurs in patients who are immunocompromised, were described in two patients with MS9, 10 and one with CD.11 As a result, natalizumab was temporarily removed from the market. The Tysabri Outreach: Unified Commitment to Health (TOUCH) program was subsequently developed to monitor adverse events in patients receiving natalizumab. Natalizumab was then reintroduced to the market for the treatment of MS and later CD. Since its reapproval, natalizumab has been prescribed primarily in patients with MS, with less than 2% of natalizumab use in patients with CD in the U.S.12 As of October 2011, one other case of PML has been described in CD in the postmarketing setting (unpubl. commun., Elan Pharmaceuticals).
To date, the data regarding clinical efficacy and safety of natalizumab in CD comes from updates of patients enrolled in clinical trials and from the TOUCH program in abstract form. The purpose of the current study was to assess the safety and effectiveness of natalizumab for CD in an outpatient, “real-world” tertiary care clinical practice setting.
MATERIALS AND METHODS
This was a prospectively identified cohort of patients being treated with natalizumab for active CD followed at the Inflammatory Bowel Disease Clinic at Mayo Clinic in Rochester, Minnesota. The primary objective was the assessment of the safety of natalizumab—events of interest included adverse events (AE) including infections, hospitalizations, surgeries, neoplasms, and deaths. Secondary endpoints included clinical response and steroid sparing.
All patients with CD prescribed natalizumab at our center from April 2008 to September 2010 were eligible for enrollment in the study. Data from patients included in this registry with infusions from April to October 2008 were collected retrospectively, as Institutional Review Board approval was obtained in September 2008. Patients started on natalizumab after October 2008 were followed prospectively. Inclusion criteria included age ≥18 years, ability to give consent to participate in the study, and clinical evidence of active (symptomatic) CD based on clinical history and radiologic or endoscopic findings. Exclusion criteria included patients who did not give consent or who did not receive at least one infusion of natalizumab.
Informed consent was obtained for all patients and an initial assessment was performed. Baseline characteristics, prior anti-TNF history, prior use of natalizumab, location of CD, prior surgical history, and current corticosteroid use were obtained. Patients were then interviewed at each infusion of natalizumab for the assessment of AEs and clinical efficacy. For those patients receiving infusions at locations other than Mayo Clinic, telephone interviews 3–7 days prior to their scheduled infusion were performed.
Clinical history was evaluated, including the development of any AEs including infection, neoplasm, or drug reaction. Hospitalizations and surgeries were recorded as well as the indication and outcome. Where appropriate, medical records from office visits, hospitalizations, and surgeries were reviewed for additional details of each event. Information regarding corticosteroid dosing, if applicable, was collected.
Clinical response was evaluated at every natalizumab infusion using the Harvey–Bradshaw Index (HBI).13 Partial clinical response was defined as HBI change of greater than or equal to three points and complete clinical response was defined as HBI less than four. Since the infusions were scheduled ≈1 month apart, “time” was measured with respect to infusion number. Kaplan–Meier estimates were generated to evaluate “time from initial infusion” to complete clinical response. Proportional hazards regression models were used to evaluate the association of specific covariates with “time to complete clinical response.” Hazard ratios (HRs) with 95% confidence intervals (CIs) calculated from model estimated coefficients were used to summarize the chance of achieving a complete clinical response.
In total, 36 patients were prescribed natalizumab during the study time period. Four did not consent and two patients never received any therapy. The remaining 30 patients agreed to participate in the study. The median number of natalizumab infusions during the study period was 9 (range, 1–30). Twenty-one patients (70%) were female. The median age at study was 35 years (range, 20–63 years), and a median disease duration of 9 years (3–43 years). All patients had prior anti-TNF therapy. Twenty-three patients had been on two prior anti-TNF medications and seven on one prior anti-TNF medication. One patient had prior natalizumab treatment as part of an investigational study. Four patients were current smokers, and 12 had previously smoked. Seven patients were on corticosteroids at the time of natalizumab initiation. Thirteen patients had no previous surgical resection. Five patients had one surgery, seven patients had two previous surgeries, three patients had three previous surgeries, and one patient had five previous surgeries. Seven patients had permanent ostomies. Other patient characteristics are summarized in Table 1.
|Gender (% female)||21 (70%)|
|Age in years (median, range)||35 (20, 63)|
|Prior anti-TNF therapy||30 (100%)|
|Prior natalizumab treatment||1 (3%)|
|Small bowel disease||3 (10%)|
|Colonic disease||4 (13%)|
|Small bowel and colonic disease||21 (70%)|
|Upper GI and small bowel disease||1 (3%)|
|Limited perianal disease||1 (3%)|
|Perianal disease present||5 (17%)|
|Cutaneous disease present||1 (3%)|
|Corticosteroid at natalizumab initiation||7 (23%)|
All patients experienced at least one AE. AEs are summarized in Table 2. Natalizumab was stopped during the study period in 13 patients (43%). In 11 patients the drug was stopped due to lack of improvement in symptoms, one patient due to an infusion reaction, and one patient stopped due to physician decision. Table 3 outlines more specifically the neurological events that were reported. Headache was the most commonly reported symptom in 88% of patients. These were all self-limited and not related to time of infusion. Headaches that occurred frequently were followed up with a physical exam at the discretion of the treating physician. No patient required initiation of any new chronic therapy for headache. Five head computed tomography (CT) or magnetic resonance imagings (MRIs) were performed in four patients and all were negative or showed stable unrelated disease.
|AE type||Subjects with AEs, n (%)||Total Number of AEs||Total Number of SAEs||SAE Probably Related to Medication|
|Systemic||15 (50%)||41||5||0 (0%)|
|GI||25 (83%)||111||4||0 (0%)|
|Infectious||11 (37%)||31||9||9 (100%)|
|Dermatologic||12 (40%)||20||1||1 (100%)|
|Infusion reaction||6 (20%)||7|
|Neurologic Event||N (%)|
|Temporary forgetfulness||2 (11%)|
Natalizumab was held but subsequently continued in 13 patients. Five patients had infections that required the medication to be held, and one subject had natalizumab held twice for two separate infections. Table 4 summarizes the infectious AEs in more detail. Six patients had medication held secondary to financial reasons, one patient had the medication held secondary to intravenous access issues, and one had the medication held per patient request (prolonged vacation abroad). No patients were diagnosed with PML during the study period.
|Type of Infection||Natalizumab Held?||When Restarted|
|Herpes zoster/abdominal wall abscess||No|
|Clostridium difficile||Yes||2 weeks|
|Gram negative bacillus catheter related bacteremia||No|
|Post op wound chronic infection||Yes||6 weeks|
|Urinary Tract infection/Kidney infection||No|
|Upper respiratory and ear infection with fever||No|
|Coagulase-negative staphyloccal bacteremia||Yes||6 weeks|
|Pneumonia/urinary tract infection||Yes||3 weeks|
|Abscess, location not specified||Yes||6 weeks|
|Epstein-Barr virus||Yes||8 weeks|
During natalizumab treatment, 19 serious AEs were documented; 15 were hospitalizations only, and four were hospitalizations secondary to surgery for CD. Of those who required surgery, one patient required total colectomy with ileostomy formation, one had small bowel resection and partial colectomy, one had a partial colectomy, and one had small bowel resection. Indications for hospitalization other than surgery included volume depletion, acute nonspecific urticaria, and infection. Of the 19 serious AEs, 10 were considered probably related to the medication (nine infections and one acute urticaria). The remainder were considered unlikely related and were secondary to increased symptoms from CD leading to hospitalization for dehydration or need for surgical intervention.
Fourteen patients obtained complete clinical response, 12 obtained a partial clinical response, and four patients had no response (all of those who failed to respond received four or fewer natalizumab treatments). Six had ileocolonic disease, four isolated colonic disease, three small bowel and perianal disease, and one with small bowel disease. The cumulative probability of achieving a complete response within 1 year was 56% (95% CI 28%–73%); 10% achieved a clinical response after the first dose and 50% of patients (seven) had complete response by 4 months. Four of seven patients were able to wean off corticosteroids during natalizumab treatment. The three patients who remained on corticosteroids received four or fewer natalizumab treatments. Three patients required short courses of corticosteroids (less than 1 month) during treatment.
Thirteen endoscopies were performed on nine patients. Four patients had improved disease, with one patient having two subsequent endoscopies showing improved disease. One showed complete endoscopic healing, and two patients had one endoscopy showing improved inflammation but a subsequent one showing new inflammation. One patient's endoscopy showed no change in inflammation, and one patient had high-grade dysplasia on biopsy (this patient received only one dose of natalizumab). Eight abdominal/pelvic CT scans were performed on seven patients. Three showed no improvement (two in the same patient), one showed improved disease, and four showed complete healing.
Cox proportional hazards modeling identified longer duration of disease (≥1 year) to be associated with a decreased chance of achieving clinical response (HR 0.3, 95% CI 0.1–0.8, P < 0.05). Age, gender, and smoking status were not associated with decreased risk of achieving complete clinical response.
Adverse events were common in our patient population; however, therapy was rarely held or stopped secondary to these events. The most common reason for holding natalizumab was for financial reasons, but five patients (16%) had natalizumab held due to a serious infection. Less than half of the patients stopped natalizumab during the study period, and most because of a lack of efficacy.
In this cohort of patients with CD treated with open-label natalizumab, response and remission rates were similar to what has been reported previously in clinical trials.4–8 Forty-three percent of patients achieved complete clinical response. There was a low likelihood of complete clinical response within 30 days (10%); however, by 1 year the probability of achieving complete clinical response was 56%.
A total of 1563 natalizumab-treated CD subjects have been evaluated in clinical trials for both short- and long-term dosing. Thirty-three percent (n = 518) received at least 1 year of treatment and 19% (n = 297) received at least 2 years of treatment.14 In placebo-controlled studies in active CD, 87.4% of natalizumab-treated and 85.6% of placebo-treated subjects experienced at least one AE. The most frequently reported AEs (≥10%) included headache, fatigue, upper respiratory infections, and nausea. Headache was the most commonly reported neurological event, as in our population. Our incidence of headache was higher than in controlled trials, but were self-limited in nature, did not lead to cessation of therapy, or result in the initiation of any new medications. Serious AEs in CD patients treated with natalizumab have included intestinal obstruction or stenosis (2% vs. 1% in placebo), acute hypersensitivity reactions (0.5% vs. 0%), abdominal adhesions (0.3% vs. 0%), and cholelithiasis (0.3% vs. 0%). Overall, natalizumab was well tolerated in controlled trials. In the short-term placebo-controlled treatment studies of active CD, 108 (9.1%) of 1043 natalizumab-treated subjects compared with 57 (11.3%) of 506 placebo-treated subjects experienced an event that led to discontinuation of study drug. A similar percentage of subjects who received placebo and natalizumab discontinued the study due to an AE.4, 5, 14 Furthermore, in a long-term, open-label extension study, ENABLE (Evaluation of Natalizumab Antibody for Long-Term Efficacy), natalizumab was evaluated for long-term maintenance therapy for an additional 12 months in patients in remission at the end of ENACT-2 (Efficacy of Natalizumab as Active Crohn's Therapy), with a total exposure of greater than 2 years of natalizumab. Long-term use was found to be safe and well tolerated.15 The most frequently reported AE that resulted in discontinuation of the medication were exacerbations of CD and acute hypersensitivity reaction.
None of the subjects in this present study were diagnosed with PML during the study period. As of August 31 2011, 88,100 patients have received natalizumab in the postmarketing setting worldwide for either CD or MS (unpubl. commun., Elan Pharmaceuticals). There have been a total of 145 confirmed cases of PML with an incidence rate of 1.62 (1.37–1.91) per 1000. In the postmarketing setting as of October 2011, there have been two confirmed cases of PML among natalizumab-treated CD patients.11 The first reported patient had received treatment with immunosuppressants prior to receiving natalizumab and received 35 doses of natalizumab prior to PML diagnosis. The second case also received immunosuppressants prior to natalizumab and greater than six infusions. The incidence appears to vary with duration of exposure, with the highest rate in those patients receiving 25–36 infusions. Some are now advocating drug holidays, to allow saturation of alpha-4 integrin to decrease and thus decrease the risk of PML and possibly other infectious complications.16 Regardless, extreme vigilance for signs and symptoms of PML is warranted, given there is no known treatment for PML. If any sign or symptom suggesting PML presents in a patient, withholding natalizumab and evaluation for PML should be undertaken, including MRI of the brain along with analysis of the cerebrospinal fluid for JC viral DNA.14
Five randomized controlled trials have evaluated natalizumab for inducing remission in active luminal CD.4–8 A recent systematic review found the relative risk of achieving remission was increased by natalizumab over placebo. In addition, data from the ENACT-2 trial demonstrated that natalizumab was effective in the maintenance of remission in CD.4 The open-label extension study (ENABLE) of those subjects who were in remission at the end of ENACT-2 found the majority maintained remission after an additional 12 months of treatment.15 In our study of using natalizumab in “real-world” clinical practice, there was a high probability of achieving clinical response by 1 year. In addition, more than half of the patients in our study who had endoscopy or radiography after starting natalizumab improved or had complete healing.
In summary, our clinical experience with natalizumab shows that overall patients tolerate natalizumab. The most common reason for medication cessation was lack of improvement in symptoms. This could be because of the complexity of disease and longer duration in these otherwise refractory patients. A number of patients had medication held secondary to infectious complications but all were able to restart therapy. This is significant, given that all of these patients had already failed at least one anti-TNF therapy, many had prior operations for CD complications, and had significant and often refractory disease. Based on our experience, we recommend holding subsequent doses of natalizumab until documented infections are treated. AEs are common, similar to what was seen in clinical trials, but therapy was rarely held or stopped secondary to a noninfectious AE. Natalizumab appears to result in clinical benefit in patients who have failed other medical therapies, with a careful discussion of the risks and benefits. Further studies are warranted regarding longer-term use of natalizumab.
Disclosures: Dr. Kane: Consultant, Elan Pharmaceuticals, Millennium Pharmaceuticals (Takeda), research support: Elan; Dr. Sandborn: Consultant, Elan Pharmaceuticals, Genentech (Roche), Pfizer, Millennium Pharmaceuticals (Takeda), Amgen, research support: Genentech (Roche), Pfizer, Millennium Pharmaceuticals (Takeda); Dr. Loftus: Consultant, Pfizer, research support: Millennium (Takeda), Genentech (Roche), Pfizer; no disclosures for the other authors.
- 15Natalizumab maintains remission for 2 years in patients with moderately to severely active Crohn's disease and in those with prior infliximab exposure: results from an open-label extension study. Gut. 2006; OPG-180 [Abstract]., , , et al.