Efficacy and Safety of Simultaneous Treatment With Two Biologic Medications in Refractory Crohn's Disease

Edward Yang; Nicola Panaccione; Natalie Whitmire; Parambir S. Dulai; Niels Vande Casteele; Siddharth Singh; Brigid S. Boland; Angelina Collins; William J. Sandborn; Remo Panaccione; Robert Battat


Aliment Pharmacol Ther. 2020;51(11):1031-1038. 

In This Article


Baseline Characteristics

A total of 24 therapeutic trials of DBT among 22 patients with Crohn's disease were identified. The median age was 35 (IQR 31–43), 95% of patients were diagnosed before the age of 40 and 32% were diagnosed before the age of 16 (Table 1). The majority of patients had either a stricturing (59%, n = 13) or penetrating (36%, n = 8) phenotype. Fifty-five per cent (n = 12) of the total patients had any prior history of perianal fistulas (Table 1). The median number of prior failed biologics was 4. Of the 94 total failed single biologic trials among 22 patients, the most common reasons for failure were secondary nonresponse without immunogenicity (47%, n = 44), followed by primary nonresponse (43%, n = 40), immunogenicity (7%, n = 7) and adverse drug event (3%, n = 3; Table 2).

Prior to initiating DBT, 79% (n = 19) of trials were receiving immunomodulators, 33% (n = 8) receiving corticosteroids and 33% (n = 8) receiving antibiotics. Additionally, 91% (n = 20) of patients had undergone prior inflammatory bowel disease-related surgery, and 50% of trials had ongoing perianal fistulas prior to initiating DBT (Table 1). Prior to initiating DBT, all patients were receiving biologic therapy with vedolizumab (n = 15), ustekinumab (n = 8) or infliximab (n = 1; Table 2). No patients were in clinical remission; the median PRO-2 score prior to DBT was 24.1 (IQR 20.0–27.0), with 91% (n = 20) having at least moderate disease activity. PRO-2 scores could not be calculated in two patients due to the presence of an ostomy. No patients were in endoscopic remission at baseline. At baseline, median SES-CD was 14.0 (IQR 12.0–17.5) and all patients had at least moderate disease activity on endoscopy. Baseline C-reactive protein (CRP) was elevated in 79% (n = 19), with a median CRP of 17.0 mg/L (IQR 11.0–24.0 mg/L; Table 3).

The most common combinations of individual biologics were vedolizumab and ustekinumab (33%, n = 8). Vedolizumab was combined with a TNF-antagonist in 13 trials (54%) and ustekinumab was combined with a TNF-antagonist in 3 trials (12.5%, Table 2). A total of 79% of all DBT regimens included at least one biologic that had induced a prior initial response with subsequent loss of response (secondary nonresponse) and 29% of DBT trials utilised a biologic that had not been previously used. In therapeutic trials achieving an endoscopic response (n = 10), seven trials utilised at least one biologic that demonstrated prior secondary nonresponse and two trials utilised a biologic that had not been previously used. All baseline biologics and 63% (n = 15) of the second biologic agents in each therapeutic trial were administered as a dose-escalated regimen.

Efficacy of Dual Biologics

Endoscopic improvement occurred in 43% of therapeutic trials, and endoscopic remission was achieved in 26% of trials. The median SES-CD significantly reduced from 14.0 (IQR 12.0–17.5) to 6.0 (IQR 2.5–8.0; P = 0.0005; Figure 1). The median time to endoscopic assessment was 225 days from initiation of DBT.

Figure 1.

Endpoints of dual biologic therapy (DBT). A, SES-CD, simplified endoscopic score—Crohn's disease. B, PRO-2, Crohn's disease—patient reported outcome-2 score. C, CRP, C-reactive protein

Clinical response occurred in 50% of therapeutic trials, clinical remission was achieved in 41% and steroid-free clinical remission was achieved in 36%. The median PRO-2 score significantly reduced from 24.1 (IQR 20.3–27.0) at baseline to 13.4 (IQR 4.6–21.8) after treatment (P = 0.002; Figure 1), and the median CRP concentration declined from 17.0 mg/L (IQR 11.0–24.0) to 9.0 mg/L (IQR 4.0–14.0; P = 0.02; Table 3). Presence of perianal fistulas declined from 50% at baseline to 33% post-treatment (Table 3). Surgery was required after 33% (n = 8) of therapeutic trials and these were considered treatment failures. At the patient level, endoscopic improvement, endoscopic remission, clinical response and clinical remission rates were 43%, 29%, 50% and 40% respectively.

Patients were treated with DBT for a median of 274 days (IQR 191–365) and followed for up to 1 year. At 1 year, the proportion of patients remaining on DBT was 38%, all of which were successful and ongoing trials of DBT (Figure S1). The most common reasons for ending a trial of DBT were nonresponse or worsening of disease activity requiring surgery. At the end of DBT, 75% (n = 18) of trials were additionally receiving immunomodulators, 17% (n = 4) receiving corticosteroids and 25% (n = 6) receiving antibiotics.

The most commonly utilised drug class in DBT was TNF antagonists, which is also the only drug class for Crohn's disease with multiple approved medications. Reasons for prior TNF-antagonist failure that were assessed in this study included mechanistic failure (primary nonresponse and secondary nonresponse despite adequate drug concentrations), secondary nonresponse due to immunogenicity and adverse event (Figure 2). In both trials of DBT that achieved or did not achieve endoscopic improvement utilising a TNF-antagonists (n = 15), all patients had prior failure to both infliximab and adalimumab as separate treatment regimens—five achieved the primary endpoint of endoscopic improvement and 10 did not. The most common reason for prior failure of either infliximab or adalimumab was secondary nonresponse despite adequate drug concentrations, and there was no signal for differences between those with and without improvement. Although specific drug concentration measurements were not available, documentation by expert inflammatory bowel disease specialists that drug concentrations were adequate at the time of treatment failure was utilised.

Figure 2.

Flowchart of reasons for prior TNF-antagonist failure in dual biologic therapy regimens that utilized a TNF-antagonist. TNF, tumor necrosis factor-α antagonist

Reasons for prior biologic failure for the entire cohort were assessed (Table S1). Among all therapeutic trials that achieved endoscopic response (n = 10), nine trials utilised at least one biologic in the regimen that either had demonstrated a prior response (secondary nonresponse) or had never been exposed to (naive) (Table S1). Among therapeutic trials that did not achieve an endoscopic response (n = 13), 10 trials utilised at least one biologic in the regimen that either had prior secondary nonresponse or the patient was naïve to (Table S1).

Adverse Events

Adverse events occurred in three trials (13%). One trial ended due to drug-induced lupus attributed to adalimumab. One patient developed pneumonia and another had reported basal cell skin cancer, recurrent Clostridium Difficile infection and Acinetobacter bacteremia in a patient with a recurrent history of all three diseases prior to initiation of DBT.

Among the combinations used as DBT, vedolizumab combined with ustekinumab had numerically higher rates of endoscopic improvement, but similar endoscopic remission and adverse event rates (Table 4).