Aliment Pharmacol Ther 31, 1296–1309
Background In the randomized, double-blind, placebo-controlled CHARM trial, adalimumab was more effective than placebo in maintaining clinical remission for patients with moderate-to-severe Crohn’s disease (CD) through 56 weeks.
Aim To substantiate the long-term safety and clinical benefits of adalimumab through 2 years of therapy in CHARM and its open-label extension (ADHERE).
Methods Patients entering ADHERE on blinded therapy received adalimumab 40 mg every other week (eow). Patients who had already moved to open-label adalimumab eow or weekly in CHARM continued their regimens. Data were analysed by originally randomized treatment group at CHARM baseline (adalimumab 40 mg eow, adalimumab 40 mg weekly, or placebo), regardless of whether patients entered ADHERE or received open-label adalimumab (eow or weekly).
Results After up to 2 years of therapy, 37.6%, 41.9% and 49.8% of patients originally randomized to placebo, adalimumab eow and adalimumab weekly, respectively, were in clinical remission. All groups experienced sustained improvements on the Inflammatory Bowel Disease Questionnaire. Decreasing hazard rates for both all-cause and CD-related hospitalizations were observed over time. Over a 2-year period, the rates of serious adverse events and malignancies (33.3 and 1.1 events/100-patient-years respectively) were similar to those observed during the overall adalimumab CD clinical development programme.
Conclusions Adalimumab demonstrated sustained maintenance of clinical remission, improvements in quality of life and reductions in hospitalization during long-term treatment for CD, with no new safety concerns identified.
Crohn’s disease (CD) is a chronic, progressive, inflammatory condition of the gastrointestinal tract that frequently progresses to serious complications – fistulas, abscesses and strictures.1 Such complications often require hospitalization and surgery for management, particularly for patients with moderate-to-severe disease. The tumour necrosis factor (TNF) antagonists, adalimumab and infliximab, have proven to be effective therapies for inducing and maintaining remission for up to 1 year in patients with CD.2–6 Certolizumab pegol was recently approved in the United States for reducing signs and symptoms and maintaining clinical response in patients with CD based on 26-week clinical trial data.7, 8 As CD is a chronic disease and evidence for the use of biological therapies earlier in the course of treatment is growing,9, 10 data on the long-term efficacy and safety of TNF antagonists are of interest to prescribing physicians.
Long-term treatment goals in CD include maintaining remission; decreasing CD-related hospitalization and surgery; preventing complications, including fistulizing disease, extraintestinal manifestations and adverse effects of concomitant medications and improving health-related quality of life (HRQOL) and productivity. Patients with sustained remission have lesser risks of complications that require hospitalization and/or surgery for medical management11–13 and the cost benefit of sustained remission has been established.14, 15 A recent publication reported the fistula-healing efficacy and safety of adalimumab through 2 years of therapy in patients with fistulizing CD;16 however, data on the overall CD response and remission beyond 1 year of therapy with TNF antagonists are limited.
Adalimumab is a fully human, anti-TNF monoclonal antibody effective for inducing and maintaining clinical response and remission in both anti-TNF-naïve and infliximab-experienced patients with moderate-to-severe CD.4–6, 17 To substantiate further the published results of efficacy, safety and positive impact on hospitalization/surgery risk and quality of life through 1 year of double-blind maintenance therapy,6, 12, 18, 19 we aimed to evaluate the longer term effects of adalimumab maintenance therapy during the CHARM (Crohn’s Trial of the Fully Human Antibody Adalimumab for Remission Maintenance) trial and its open-label extension (ADHERE, Additional Long-Term Dosing With HUMIRA to Evaluate Sustained Remission and Efficacy in CD). This analysis includes patients with moderate-to-severe CD who received up to 2 years of therapy with adalimumab and is an extension of previously published reports of the first year of treatment.6, 18
The CHARM trial. CHARM was a 56-week, multicentre (92 sites in Europe, the United States, and Canada), Phase III, double-blind, randomized, placebo-controlled trial (http://www.clinicaltrials.gov, NCT00077779) to assess the long-term efficacy and safety of adalimumab every other week (eow) and weekly vs. placebo in maintaining clinical remission in patients with moderately to severely active CD. The study method and entry criteria have been described.6 Briefly, patients were between 18 and 75 years old, had CD for more than 4 months (diagnosis confirmed by endoscopic and radiologic evaluation), had a Crohn’s Disease Activity Index (CDAI) score between 220 and 450, and were to have discontinued any previous anti-TNF therapy at least 12 weeks prior to study enrolment. All patients received open-label induction therapy with adalimumab 80/40 mg at weeks 0/2. At week 4, patients were randomized to receive adalimumab eow, adalimumab weekly, or placebo. After week 12, patients experiencing a disease flare or sustained nonresponse could be switched to open-label treatment with adalimumab 40 mg eow and subsequently to open-label adalimumab 40 mg weekly for continued nonresponse or recurrent flare. At week 8, patients receiving corticosteroids who experienced a significant improvement in CD symptoms could begin to taper their corticosteroid dosages. If a patient subsequently lost clinical response, the corticosteroid dosage could be increased back to the original dosage.
Open-label extension (ADHERE). All patients still blinded at the end of CHARM could enter the open-label extension (ADHERE; http://www.clinicaltrials.gov, NCT00195715) to receive adalimumab eow. Patients who had already moved to open-label adalimumab eow or weekly continued their regimens upon entry in ADHERE. During ADHERE, patients could move from eow to weekly dosing for flares (increase in CDAI ≥70 points compared with week 4 of CHARM and a CDAI score >220) or nonresponse (failure to attain CDAI decrease ≥70 points compared with baseline of CHARM. Patients receiving corticosteroids during CHARM were allowed to continue that regimen upon entry into ADHERE. Patients not previously receiving corticosteroids were allowed to initiate steroids after 3 months of treatment with open-label adalimumab (during either CHARM or ADHERE). Beginning at the ADHERE baseline visit, patients experiencing a significant improvement in their CD symptoms (defined as a CDAI decrease of ≥70 points compared with baseline of CHARM) were able to have their corticosteroid dosages tapered at the investigators’ discretion.
Patients who experienced a flare after they reduced or completely tapered their dosages of steroids in this study or in the previous study could have their corticosteroid dosages increased to the dosage prior to the start of the taper. Adjustments (increases or decreases) in other CD-related concomitant treatments, including initiation of a treatment a patient was not taking previously, were allowed following at least 3 months of open-label adalimumab exposure (during either CHARM or ADHERE).
Data and patient sample
The primary efficacy and safety results of the 56-week CHARM trial (randomized responder population) have been published,6 as have supplemental analyses of treatment regimen.18 The primary CHARM manuscript also summarized the ability of adalimumab to sustain steroid-free remission during the first year of therapy.6
For this analysis, patients were followed through 2 years of treatment (1 year in CHARM and up to one additional year in ADHERE). Patients who were randomized to adalimumab in CHARM were followed up for up to 2 years of adalimumab therapy. Patients who were randomized to placebo in CHARM were followed up from their first doses of blinded study medication, regardless of whether they completed CHARM on placebo or moved to open-label adalimumab therapy during CHARM. Upon entry in ADHERE, all patients randomized to placebo in CHARM and still receiving blinded placebo received open-label adalimumab therapy (40 mg eow).
For patients receiving blinded therapy prior to entering ADHERE, study visits occurred at week 2, week 4, week 8 and week 12 and at 12-week intervals thereafter. For patients entering on open-label therapy, study visits occurred at week 12 and at 12-week intervals thereafter. A standard visit-window method was used to calculate the scores at scheduled visits.
CDAI scores were collected at each visit and clinical remission was defined as CDAI <150. Clinical response, defined as decrease in CDAI of ≥70 points (CR-70) or ≥100 points (CR-100) from CHARM baseline, was also assessed. The Inflammatory Bowel Disease Questionnaire (IBDQ), a valid and disease-specific measure of HRQOL in patients with inflammatory bowel disease,20 was assessed during both CHARM and ADHERE. The IBDQ was assessed at 24-week intervals in ADHERE, starting at ADHERE week 12; assessments during the first 12 weeks were more frequent for patients entering ADHERE on blinded therapy. Hospitalization and surgery events were assessed via review of serious adverse events recorded on the case report forms. Adverse event data were collected throughout the duration of both trials.
Using an intention-to-treat approach, the post-hoc efficacy and safety assessments for the duration of participation in CHARM and ADHERE were analysed based on each patient’s randomized treatment arm in CHARM.
Clinical remission and response. The median CDAI scores were summarized at weeks 26 and 56 in CHARM and weeks 36 and 60 in ADHERE using last-observation-carried-forward (LOCF) method. Week 56 represents the end of CHARM, and weeks 36 and 60 in ADHERE represent 92 and 116 weeks from CHARM baseline respectively. Rates of clinical remission, CR-70 and CR-100 were summarized by randomized treatment group at each time point using LOCF method. Maintenance of remission in patients randomized to adalimumab who were in clinical remission at week 56 of CHARM and rolled over to ADHERE was evaluated using LOCF method. In addition, a subgroup analysis of remission was performed for patients who were in remission on double-blind weekly therapy at the end of CHARM and received only eow therapy for up to 1 year during ADHERE.
Health-related quality of life. The median IBDQ scores were summarized at weeks 26 and 56 in CHARM and weeks 36 and 60 in ADHERE using LOCF method. The percentages of patients with IBDQ ≥170, which correlates with clinical remission20, were also summarized.
Hospitalization and surgery risk. Similar to the methods described by Feagan et al.,12 CD-related hospitalizations were identified based on a list of the Medical Dictionary for Regulatory Activities codes related to CD and confirmed by the review of serious adverse event reports. CD-related hospitalizations and CD-related surgeries were identified by three of the authors through review of patient narratives for serious adverse events (RP, AMR, and JC). A survival analysis of hospitalization was conducted, and Kaplan–Meier curves were plotted by originally assigned treatment group. The Weibull model, a parametric model used in survival analysis, was used to assess the trend in the risk of hospitalization over time. In this model, the ‘shape parameter’ indicates whether the hazard of hospitalization is decreasing (values <1) or increasing (values >1).
Safety. All patients who received at least one injection of adalimumab were included in the adverse event summaries. Treatment-emergent deaths, serious adverse events, adverse events leading to premature discontinuation, adverse events of interest (e.g. serious infections, malignancy, opportunistic infections, tuberculosis), and adverse events reported in >5% of patients were summarized. Adverse events were summarized as number and percentage of patients experiencing each adverse event and as events per 100-patient-years (PYs).
Exposure to adalimumab and sample characteristics
Of the 854 patients enrolled in CHARM, 778 were randomized (261 into placebo, 260 into adalimumab eow, and 257 into adalimumab weekly). Overall, 155 patients in the randomized eow group, 129 in the randomized weekly group, and 95 in the randomized placebo group were receiving eow dosing at their last visit. The remainder of patients in the randomized eow and weekly groups were receiving open-label adalimumab weekly at their last visit. For patients originally randomized to placebo, 67 were on placebo and 99 were on open-label adalimumab weekly at the last visit. The mean exposure to adalimumab (i.e. average time receiving adalimumab therapy) was 378 days in the group of patients initially randomized to placebo, 517 days in the adalimumab eow group, and 602 days in the adalimumab weekly group.
A total of 467 patients entered ADHERE (Figure 1). Of these patients, 239 were still receiving blinded therapy in CHARM (42 of those randomized to placebo and 95 and 102 of those randomized to adalimumab eow and weekly respectively) and thus received open-label adalimumab eow at the start of ADHERE. A majority of patients who received open-label adalimumab eow at the start of ADHERE remained on that dosage throughout the study (267 of 349 patients, 76.5%). Of the 260 patients randomized to adalimumab eow (the starting maintenance dose), 71 (27.3%) changed to weekly dosing during year 1 and an additional 34 (13.1%) changed to weekly dosing during year 2.
Demographics and clinical characteristics at CHARM baseline were similar between randomized treatment groups (Table 1). Corticosteroids were used by 43.6% of patients at CHARM baseline, immunosuppressants were used by 46.9% of patients and 49.9% of patients were anti-TNF-naive. At the beginning of ADHERE, 201 patients were receiving immunosuppressants (43.0% of the total patients enrolled in ADHERE). Six (2.6%) of the 233 patients who were not receiving immunosuppressants at baseline initiated immunosuppressants during ADHERE (three adalimumab eow; three adalimumab weekly; and zero placebo).
|Placebo (N = 261)||Adalimumab eow (N = 260)||Adalimumab weekly (N = 257)|
|Age (years), mean, (s.d.)||36.9 (11.4)||36.8 (11.5)||37.8 (12.1)|
|Men, n (%)||99 (37.9)||97 (37.3)||100 (38.9)|
|White, n (%)||246 (94.3)||245 (94.2)||231 (89.9)|
|Body weight (kg), mean (s.d.)||71.1 (18.4)||70.5 (16.9)||71.0 (18.5)|
|Disease duration (years), median||8.4||7.9||7.9|
|Baseline CDAI score, median||306||302||299|
|Baseline IBDQ score, median||125||124||122|
|Smoking status, n (%) smokers||96 (36.8)||92 (35.4)||89 (34.6)|
|Presence of fistula, n (%)||47 (18.0)||30 (11.5)||40 (15.6)|
|Steroid use, n (%)||114 (43.7)||109 (41.9)||116 (45.1)|
|CRP concentration >1.0 mg/dL, n (%)||131 (50.2)||118 (45.7)||133 (44.0)|
|Immunosuppressant use, n (%)||133 (51.0)||112 (43.1)||121 (47.1)|
|5-Aminosalicylate use, n (%)||116 (44.4)||96 (36.9)||94 (36.6)|
|Prior TNF-antagonist exposure, n (%)||130 (49.8)||133 (51.2)||127 (49.4)|
Maintenance of remission and clinical response. The median CDAI scores over time are provided in Figure 2a (see Figure S1a, for as-observed data). By week 116 from CHARM baseline, which represents up to 2 years of therapy, 37.6% of patients randomized to placebo, 41.9% of patients randomized to adalimumab eow and 49.8% of patients randomized to adalimumab weekly achieved clinical remission (LOCF analysis) (Figure 2b). Week-116 CDAI remission rates for patients who were being treated with immunosuppressant therapy at baseline of ADHERE [64.2% (129 of 201)] were similar to those for patients who were not receiving immunosuppressant therapy at baseline [60.2% (160 of 266)]. Using an as-observed analysis (Figure S1b), 66.7% of patients randomized to placebo, 67.8% of patients randomized to adalimumab eow, and 68.2% of patients randomized to adalimumab weekly achieved clinical remission at week 116.
For all groups, remission rates were sustained from the end of CHARM through week 116 in ADHERE. Of the 145 patients randomized to adalimumab who were in remission at the end of CHARM and who entered ADHERE, 122 (84.1%) maintained remission at week 116. The percentages of patients achieving CR-70 and CR-100 responses are shown in Figure 2c and d respectively. Results for as-observed analyses are provided in the online supplement (Figures S1c and S1d).
Remission in patients whose dosage was reduced from weekly in CHARM to eow in ADHERE. Of the 75 patients on blinded, weekly adalimumab therapy for the duration of CHARM who were in remission at the time of entry to ADHERE, at which point they received open-label adalimumab eow, 85% (64 patients) maintained the eow dosage in ADHERE. Of the 64 patients who remained on adalimumab eow in ADHERE, 54 and 50 maintained clinical remission at weeks 92 and 116 respectively. Eleven patients returned to weekly dosing because of flare or loss of response. Of these patients, 10 (91%) regained response and nine (82%) regained remission.
Health-related quality of life
Baseline IBDQ scores indicated substantially impaired quality of life in this patient population (Table 1). Median IBDQ scores over time are provided in Figure 3a. For all groups, improvements in HRQOL were observed starting at week 26 in CHARM and were sustained for the duration of therapy in ADHERE. Figure 3b summarizes IBDQ remission over time. The week-116 IBDQ remission rates (IBDQ ≥170) were 62.7% (126 of 201) and 53.4% (142 of 266) for patients who used immunosuppressants at the baseline of ADHERE and patients not receiving immunosuppressants respectively.
Hospitalization and surgery
Reductions in hospitalization risk for the first 56 weeks of adalimumab vs. placebo treatment in CHARM have been previously reported.12 During adalimumab therapy in ADHERE, 23, 33 and 17 patients were hospitalized in the eow, weekly and placebo treatment groups respectively (Table 2). Over the full 2 years of study, 60, 69 and 80 patients were hospitalized in the eow, weekly and placebo groups respectively. A majority of hospitalizations occurred in patients who had not been hospitalized in the past year and were related to complications of CD. Ten major surgeries were identified in ADHERE (4, 5 and 1 respectively for randomized eow, weekly and placebo). A total of 14, 14 and 22 major surgeries for the eow, weekly and placebo groups respectively were reported during the 2-year study period.
|All-cause hospitalization||CD-related hospitalization|
|Total number of hospitalized patients during ADHERE||Patients with new hospitalizations*||2-year overall number of patients hospitalized||Total number of hospitalized patients during ADHERE||Patients with new hospitalizations*||2-year overall number of patients hospitalized|
The Kaplan–Meier curves presented in Figures 4a and b are survival analyses of all-cause hospitalization and CD-related hospitalization by originally randomized groups in CHARM. At week 116, the cumulative estimated rates of all-cause hospitalization were 30%, 32% and 36% respectively for the randomized eow, weekly and placebo treatment groups (Figure 4a). The week-116 cumulative estimated rates for CD-related hospitalization were 23%, 19% and 25% for the eow, weekly and placebo treatment groups respectively (Figure 4b). The shape for all parameters in the Weibull model for hospitalization rates was less than 1, indicating decreasing hazard rates for both all-cause hospitalization [0.85; 95% confidence interval (CI), 0.75–0.96] and CD-related hospitalization (0.80; 95% CI, 0.69–0.93) over time.
A total of 854 patients received at least one injection of adalimumab in CHARM or ADHERE, representing 1072.1-PYs of adalimumab exposure. Adverse events were reported in 94.1% of patients, 25.6% of patients experienced a serious adverse event and 20.4% of patients discontinued from the study because of an adverse event (Table 3). Adverse events of interest with anti-TNF agents include serious infections, malignancies, opportunistic infections (including tuberculosis) and demyelinating disorders. In CHARM/ADHERE, 56 (6.6%) adalimumab-treated patients developed a serious infection during the study. The most common serious infections were abdominal abscesses (eight patients) and anal abscesses (seven patients). Three patients developed tuberculosis (described previously21) and two patients developed nondisseminated herpes zoster.
|Adverse event||N = 854 n (%)||PYs = 1072.1 events (Events/100-PYs)|
|Any adverse event||804 (94.1)||7644 (713.0)|
|At least possibly related to study drug||475 (55.6)||1637 (152.7)|
|Serious adverse event||219 (25.6)||357 (33.3)|
|Leading to discontinuation of study drug||174 (20.4)||212 (19.8)|
|Fatal adverse event*||2 (0.2)||2 (0.2)|
|Adverse events of interest|
|Infection||510 (59.7)||1424 (132.8)|
|Serious infection†||56 (6.6)||69 (6.4)|
|Malignancy‡||11 (1.3)||12 (1.1)|
|Injection-site pain||54 (6.3)||59 (5.5)|
|Opportunistic infection§||19 (2.2)||22 (2.1)|
|Tuberculosis||3 (0.4)||3 (0.3)|
|Congestive heart failure||0||0|
|Demyelinating disorder||2 (0.2)||3 (0.3)|
Opportunistic infections, all of which were nonserious, nonsystemic candidiasis, were reported in 19 (2.2%) patients. Eleven patients (1.3%) were diagnosed with a malignancy during the 2-year study period. Two adalimumab-treated patients developed demyelinating disorders. Treatment-emergent adverse events occurring in >5% of patients are summarized in the online supplement (Table S1). The most common adverse events during the 2-year study period included CD, nasopharyngitis, arthralgia, abdominal pain, headache, nausea, influenza and pyrexia. The reasons for premature discontinuation during ADHERE are provided in Figure 1. Adverse events were the most common reason for withdrawal (11 patients randomized to placebo, 11 patients randomized to adalimumab eow, 15 patients randomized to adalimumab weekly).
At present, consensus guidelines for the treatment of CD22, 23 include the use of TNF antagonists for induction and maintenance treatment of patients with moderate-to-severe CD who have not responded to a steroid or immunosuppressive agent or in whom corticosteroids are contraindicated or not desired. Recently, there has been growing interest in introducing biological therapy earlier in the treatment of CD, based on studies that have shown greater rates of remission with TNF-antagonist therapy for patients with a shorter duration of CD.9, 10, 24 Evidence-based decisions to prescribe TNF antagonists earlier and long-term in the treatment of CD require additional long-term data supporting their safety, efficacy, cost effectiveness and ability to modify the disease course and prevent the complications that can lead to hospitalization and/or surgery.25
This analysis of CHARM and its open-label extension (ADHERE) includes patients treated with adalimumab for up to 2 years and represents the first large, prospective, multicentre analysis of the 2-year efficacy and safety of a biological therapy for the treatment of CD. Importantly, the clinical remission rates and improvements in HRQOL that were achieved in adalimumab-treated patients at week 26 in CHARM6 were sustained for up to 2 years of adalimumab therapy. For this analysis, we used an intention-to-treat approach whereby results were summarized by the original randomized group regardless of the actual therapy received at the time points analysed. Therefore, it is not possible to draw conclusions about the small differences observed in rates of remission or HRQOL between dosing groups at year 2. At the last visit, 74% of patients randomized to placebo were receiving adalimumab (36% and 38% were receiving eow and weekly therapy respectively). Of the patients randomized to adalimumab eow and weekly, 60% and 50% were receiving eow dosing at the last visit respectively.
In a previous analysis of the 1-year CHARM data, adalimumab-treated patients had lesser 1-year risks of hospitalization and surgery compared with placebo-treated patients.12 Likewise, the Weibull model results presented here demonstrated a general trend of decreasing hazard rate for hospitalization over the 2-year period. The observed differences in hospitalization between the original adalimumab arms and placebo were driven primarily by differences in CHARM, in which some of the patients randomized to placebo were exposed only to the two-dose induction regimen of adalimumab; hence, these findings suggest that early maintenance therapy may be an important factor in reducing hospitalization rates in patients with CD. As might be expected after receiving adalimumab therapy in year 2, the rates of hospitalization and surgery in the randomized placebo group decreased from year 1 to year 2.
Data evaluating the clinical and quality-of-life outcomes of patients treated with other TNF antagonists beyond 1 year are limited by small numbers of patients and retrospective designs.26–29 Other prospective studies have demonstrated that TNF-antagonist therapy is associated with a decrease in the number of CD-related hospitalizations and hospitalization days,11 including patients with fistulizing CD,30 but long-term effectiveness data through 2 or more years of therapy are limited. A recent single-centre, observational study evaluated the long-term clinical outcome and hospitalization and surgery rates of 614 patients treated with infliximab over a median of 4.6 years and found that infliximab maintained improvement in a real-life cohort of patients with CD; the analysis included only 70 patients who received scheduled infliximab for 2 years or longer.31 This study also found that scheduled therapy with infliximab led to a lesser rate of hospitalization than episodic use. Patients with an initial and sustained clinical benefit with infliximab also had a significant reduction in the need for major abdominal surgery compared with the number of surgeries required prior to initiation of infliximab.
With respect to the long-term safety of adalimumab, an analysis summarizing more than 10 years of clinical observation indicates that the safety profile of adalimumab in moderate-to-severe CD is comparable to that of other immune-mediated inflammatory disorders.32 Moreover, a recent publication of safety data following up to 5 years of adalimumab therapy in six global clinical trials for CD concluded that the risk of mortality in this population was not different from what would be expected in the general population.21 While 94% of the patients experienced an adverse event, patients with moderate-to-severe CD experience a substantial rate of adverse events because of the nature of their disease. This is evidenced by the adverse event rates observed for placebo recipients in the controlled induction trials for adalimumab, CLASSIC-I and GAIN, which were 74% and 73% respectively by week 4, as compared with the 57–75% overall rate for patients receiving adalimumab. In addition, in both trials, the rates of serious adverse events were similar or greater for placebo patients compared with adalimumab patients.4, 5 Moreover, the overall adverse event rates for placebo and adalimumab patients during the double-blind portion of the CHARM trial were similar. In fact, placebo patients experienced a greater rate of serious adverse events and serious infections than did patients receiving blinded adalimumab.6 Results of the present study support that longer term adalimumab therapy is associated with low incidences of adverse events of interest, such as serious infection, opportunistic infection/tuberculosis, malignancy/lymphoma and demyelinating disorders. Physicians should weigh the risks associated with any therapy for CD with the potential benefits of the therapy in symptom reduction, quality-of-life improvements and avoidance of hospitalizations and surgeries.
Although cost data were not evaluated in the present study, the clinical benefits in terms of sustained remission and decreased hospitalization are consistent with a previous cost-effectiveness analysis that used data from CHARM, in which maintenance therapy with adalimumab eow was associated with significantly lesser costs in a lifetime model compared with conventional, nonbiological therapy, particularly for patients with severely active CD.14 In addition, a Markov cohort analysis using patient-level data from the United Kingdom showed that continuous adalimumab therapy for 2 years was cost-effective for initial responders compared with standard care for CD.33 Finally, in a recently published study that used a cost-utility framework from a private-payer perspective over a 56-week time horizon, drug and administration costs, drug waste and hospitalization rates for patients with moderate-to-severe CD were lower with adalimumab therapy than with infliximab therapy.34
This study is the first to look at rates of dosage intensification with adalimumab beyond 1 year. An analysis of the rates and benefits of weekly dosing in CHARM has been presented previously, concluding that the rate of change to weekly dosing for flares or nonresponse was low in year 1.35 The overall rate of dosage intensification in this study is greater than what has been observed in real-life settings,36, 37 but is less than what has been reported in long-term real-life evaluations of infliximab dosage intensification.38, 39 This analysis is also the first to report on the response of patients who moved from blinded weekly to open-label eow therapy. Although the design of CHARM/ADHERE did not allow patients who moved from eow to open-label weekly dosing to resume eow dosing, a majority (85%) of the patients who were in remission on blinded weekly therapy at the end of CHARM were able to maintain on eow therapy only during ADHERE, with substantial rates of sustained remission on the lower dose.
Clinicians may be faced with considering discontinuation of anti-TNF therapy for patients with CD after a period of maintenance therapy, based on safety or economic concerns or patients’ requests. The greater rates of remission6 and quality of life,19 and the lower rates of hospitalization and surgery12 observed for adalimumab-treated patients in 1-year controlled trials support the notion that, at least through the first year of therapy, anti-TNF therapy withdrawal is an inferior strategy. Recent data from the STORI trial40 provide some preliminary evidence that it may be possible to withdraw anti-TNF therapy in a highly select group of patients, but this strategy needs further evaluation in a controlled clinical trial. Another study provides some preliminary evidence that it might be possible to withdraw azathioprine therapy for patients receiving combination therapy.41 However, given the decreasing anti-TNF drug concentrations in the azathioprine withdrawal group and the low statistical power of the study for non-inferiority, additional controlled data are needed.41 Therefore, withdrawal of concomitant medication might be another approach to improving safety results without sacrificing the benefits of continuous use of biologics. While more research is needed in the long-term treatment of biologics, physicians should carefully consider the risks and benefits of different approaches in the long-term use of biologics. The relative risks of therapy vs. disease progression because of inadequate therapy of different treatment strategies, including long-term combination therapy, withdrawal of concomitant immunomodulator therapy and withdrawal of anti-TNF therapy after 1 year, need further evaluation.
In summary, these analyses demonstrate that adalimumab therapy can induce and sustain remission and improve quality of life for up to 2 years, as well as reduce long-term hospital admission rates in patients with moderate-to-severe CD.
Declaration of personal interests: Remo Panaccione, Jean-Frédéric Colombel, William Sandborn, Paul Rutgeerts and Geert D’Haens have served as study investigators and consultants for Abbott Laboratories and have participated in continuing medical education events supported by unrestricted educational grants from Abbott. Anne Robinson, Jingdong Chao, Parvez Mulani and Paul Pollack are Abbott employees and own shares of Abbott. The authors thank the following individuals for their contributions to this manuscript: Naijun Chen, MS, of Abbott Laboratories, for assistance with data management support; Winnie Lau, BS, of Abbott Laboratories, for her contributions to the analysis of the adverse event data; and Kathleen G. Lomax, MD, for her contributions to the analysis of the hospitalization and surgery data. In addition, the authors thank Cathryn M. Carter, MS, of Arbor Communications, Inc., and Michael A. Nissen, ELS, formerly of Abbott, for medical writing and editing support in the development and revision of the manuscript. Declaration of funding interests: Medical writing support was funded by Abbott. The CHARM and ADHERE studies and the analysis reported in this article were supported by research grants from Abbott Laboratories, Abbott Park, Illinois. The analyses reported in this paper are based on a Phase III trial and its open-label extension and pertain to a commercial product [HUMIRA (adalimumab)].