ORIGINAL ARTICLE
Early Infliximab Trough Levels Are Associated with Persistent Remission in Pediatric Patients with Inflammatory Bowel Disease Namita Singh, MD,* Casey J. Rosenthal, BS,* Gil Y. Melmed, MD,†,‡ James Mirocha, MS,§ Sharmayne Farrior, BSN,* Silvia Callejas, BSN,* Bhavna Tripuraneni, BS,* Shervin Rabizadeh, MD,* ,‡ and Marla C. Dubinsky, MD* ,‡
Background: Low infliximab (IFX) trough levels and high anti-infliximab antibodies (ATI) levels are associated with loss of response to IFX. Optimizing IFX levels to maintain target concentrations before loss of response may improve long-term efficacy. We hypothesized that trough levels at week 14 are predictive of IFX durability.
Methods: A prospective observational cohort of pediatric patients with inflammatory bowel disease initiating IFX had IFX and ATI levels drawn at weeks 14 and 54. Primary outcome was week 54 persistent remission (PR), defined as clinical remission without IFX dose intensification. Univariate analyses tested associations of week 14 IFX (IFX14) and ATI (ATI14) levels, clinical and laboratory data with week 54 outcomes. Receiver operating curve analysis and positive and negative predictive values for IFX14 cut-off points were examined.
Results: Of 58 patients enrolled, 8 (13%) stopped IFX before week 14 and 4 discontinued IFX between weeks 14 and 54. IFX14 level (P ¼ 0.03),
baseline C-reactive protein (CRP) level (P ¼ 0.01), and week 14 CRP (CRP14) level (P ¼ 0.0001) were associated with PR. A model with IFX14 levels predicting PR had an area under the receiver operating curve of 0.68 and a model with both IFX14 level and CRP14 .1.0 mg/dL had an area under the receiver operating curve of 0.74. IFX14 cut points of .3, .4, and .7 mg/mL had positive predictive values of 64%, 76% and 100%, respectively, for predicting PR.
Conclusions: Both IFX levels and CRP at week 14 were significantly associated with week 54 efficacy. A model combining both CRP and IFX at week 14 may help predict remission at week 54. (Inflamm Bowel Dis 2014;20:1708–1713) Key Words: pharmacokinetics, infliximab level, anti-infliximab antibody, inflammatory bowel disease, pediatrics
I
nfliximab (IFX) is approved for use in children with both Crohn’s disease (CD) and ulcerative colitis (UC). The REACH trial reported that close to 90% of children responded to induction and close to 60% of children were in remission at 1 year.1 Almost half of these patients, however, lost response and required dose modification to recapture remission.1 A similar durability of IFX was reported outside the clinical trial arena, whereby 70% were still on IFX after 3 years of follow-up.2 However, half of these Received for publication May 27, 2014; Accepted June 10, 2014. From the Departments of *Pediatrics and †Medicine, Cedars-Sinai Medical Center, Los Angeles, California; ‡F. Widjaja Foundation Inflammatory Bowel and Immunobiology Institute, Cedars-Sinai Medical Center, Los Angeles, California; and § Department of Biostatistics, Cedars-Sinai Medical Center, Los Angeles, California. Supported in part by the National Institute/National Institute of Digestive Disease and Kidney (NIDDK) Grant DK084554 (M.C.D.), and The Abe and Claire Levine Chair in Pediatric IBD (M.C.D.). N. Singh and C. J. Rosenthal contributed equally to this work. M. C. Dubinsky: Consultant for Prometheus Laboratories Inc. and Janssen Pharmaceuticals Inc. G. Y. Melmed: Consultant for Janssen Pharmaceuticals Inc. The other authors have no conflicts of interest to disclose. Reprints: Marla C. Dubinsky, MD, 8635 West 3rd Street, Suite 1165W, Los Angeles, CA 90048 (e-mail: [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000137 Published online 22 August 2014.
1708
| www.ibdjournal.org
patients required dosing adjustments. For UC, close to 75% of pediatric patients respond to IFX induction and about 40% are in remission at 1 year.3 The maintenance of a sustained and durable remission with IFX has become an important clinical challenge. The etiology for secondary loss of response may be complex; however, a common factor is the pharmacokinetic profile of IFX and immunogenicity. There have been many cross-sectional studies demonstrating the positive association of IFX trough level and the negative association of anti-infliximab antibodies (ATI) with efficacy outcomes.4–6 A recent meta-analysis confirmed that the presence of ATI is associated with a significantly higher risk of loss of clinical response to IFX and lower serum IFX levels in patients with inflammatory bowel disease (IBD).7 The pooled risk ratio of loss of clinical response to IFX in patients with IBD who had ATIs was 3.2 (95% confidence interval, 2.0–4.9; P , 0.0001). Attempts have been made to define a minimal trough and antibody level that is best associated with improved IFX outcomes, such that being above a laboratory defined cutoff may not be sufficient. A trough level of 3 mg/mL or above has been shown to be independently associated with a lower C-reactive protein (CRP) and has been proposed as a cut point to maximize clinical efficacy of IFX.8 The current approach is to obtain an IFX trough Inflamm Bowel Dis Volume 20, Number 10, October 2014
Inflamm Bowel Dis Volume 20, Number 10, October 2014
and ATI level when a patient has lost response to IFX to seek an explanation as to why the patient is no longer responding to standard on-label dosing. The key, however, may be to determine the pharmacokinetic profile of patients before losing response and to intervene proactively and optimize to a target concentration. Akin to thiopurine monitoring, early measurement of drug levels during or at the end of induction may enable optimization of IFX dosing proactively and lead to less loss of response and less immunogenicity. Three retrospective studies conducted in adult patients with IBD have shown that higher IFX trough levels and lower ATI levels at the time of the week 14 infusion were associated with improved sustained disease response.9–11 In this prospective observational study, we hypothesized that week 14 IFX trough levels predicted week 54 IFX outcomes.
MATERIALS AND METHODS Patient Population and Ethical Considerations Pediatric patients with CD and UC younger than 21 years initiating IFX and enrolled in a prospective multicenter NIH funded study examining Genome Wide Association (GWAS) predictors of response to IFX were eligible. The decision to treat with IFX was at the discretion of the investigator. Patients received standard IFX induction regimen of 5 mg/kg infused at week 0, 2, and 6. No dose adjustments were made during the induction phase. Only those patients enrolled at Cedars-Sinai Medical Center (CSMC) who completed 54 weeks of follow-up were included in this substudy. This study was approved by the CSMC Institutional Review Board (IRB No. 19251).
Disease Activity Assessments At each infusion, a physical examination and standard of care laboratories, including complete metabolic panel, CRP level, erythrocyte sedimentation rate, complete blood count with differential, and iron panel, were performed. Disease activity was measured using the Pediatric Crohn’s Disease Activity Index for patients with CD aged ,20 years or the Crohn’s Disease Activity Index for patients with CD aged 20 years and above, and the Partial Mayo score for all patients with UC.
Outcome Measures and Definitions Our primary outcome was week 54 persistent remission (PR), which was defined as clinical remission based on the relevant disease activity index at week 54, in the absence of any dose intensification during IFX maintenance. Secondary outcomes included clinical remission that looked solely at whether a patient met the criteria of remission according to the appropriate disease activity index, and clinical and laboratory remission that consisted of a patient being in PR in addition to having a normal CRP level. Sustained durable remission (SDR) defined a patient who was in PR from week 14 through week 54 (SDR14) or from week 22 through week 54 (SDR22). Primary nonresponders were defined as patients who stopped IFX before the first maintenance infusion (week 14). Early terminators were patients who entered
Infliximab Levels Predict Remission in IBD
the maintenance phase at week 14, but stopped IFX therapy before the week 54 end point.
Serum Infliximab and Anti-infliximab Antibody Measurements IFX trough and ATI levels were drawn before the week 14 infusion and just before the week 54 infusion or at early termination and sent to Prometheus Laboratories (San Diego, CA) for testing. Samples collected before July 1, 2012, were tested by an enzyme-linked immunosorbent assay (ELISA).12 The lower limit of detection of this assay was 1.4 mg/mL for IFX and 1.69 mg/mL for ATI. Samples collected after July 1, 2012, were tested using a homogenous mobility shift assay (HMSA) with a lower limit of detection of 1.0 mg/mL for IFX and 3.1 U/mL for ATI.10 Fifty percent of the samples were tested using both assays during the transition to the new HMSA assay to confirm previous reports of a high correlation between ELISA and HMSA for IFX levels.10 We observed a correlation coefficient of 0.91 in this subset of samples. Given the high correlation, we elected to include those IFX levels tested with ELISA only in both the qualitative and quantitative analyses. However, our correlation coefficient between the 2 assays for ATI levels was only 0.75, thus we elected to only report ATI status qualitatively (detectable or not detectable), and not quantitatively. ATI levels are reported in mg/mL for the ELISA and U/mL for the HMSA.12
Statistical Analysis Numerical variables were summarized by median and range or interquartile range (IQR) and were compared across independent groups by the Wilcoxon rank sum test. Associations between numerical variables were assessed by the Spearman rank order correlation coefficient. Categorical variables were summarized by frequency and percent and were compared across independent groups by the Fisher’s exact test. Change in status on binary variables (e.g., detectable IFX at weeks 14 and 54) was assessed by McNemar’s test for related proportions. Positive (PPV) and negative predictive values were calculated in the standard fashion. The method of DeLong et al13 was used to compare area under the receiver operating characteristic curves (AUC) for related models. A 2-sided 0.05 significance level was used throughout. Statistical calculations were made using SAS Version 9.2 (SAS Institute, Cary, NC).
RESULTS Patient Population A total of 58 patients were enrolled. Eight patients (13%) were primary nonresponders and 50 entered the maintenance phase of IFX therapy (week 14). Four patients (8%) were early terminators (stopped IFX between week 14 and week 54) and 46 remained on IFX and completed the study through week 54 (Fig. 1). The median age at diagnosis for the cohort was 11.4 (range, 6.6–18.4) years, median disease duration was 3.6 (range, 0.2–93.0) months, 53% of patients were male, and 81% of all patients had a diagnosis of CD (Table 1). www.ibdjournal.org |
1709
Inflamm Bowel Dis Volume 20, Number 10, October 2014
Singh et al
week 14 and 54 (P . 0.99). IFX14 levels positively correlated with IFX54 levels (r ¼ 0.46, P ¼ 0.001). At week 14, 10% of patients had a detectable ATI level and 26% had a detectable ATI level at week 54 (McNemar’s P ¼ 0.02).
Week 14 Infliximab Levels and Week 54 Outcomes
Median IFX14 level was 4.7 mg/mL for patients in PR versus only 2.6 mg/mL for patients not in PR (P ¼ 0.03) (Fig. 2). The median IFX14 level was 5.1 mg/mL in patients in SDR22 versus 3.0 mg/mL in patients not in SDR (P ¼ 0.04). Higher median IFX14 levels were also observed for the other secondary outcomes (Table 2). Unlike median IFX14 levels, IFX14 status was not significantly associated with week 54 outcomes (Table 3). Sixty-nine percent of patients with a detectable IFX14 were in PR at week 54, as compared with only 40% without a detectable IFX14 level (P ¼ 0.11). In addition, 51% of patients with detectable IFX14 were in SDR22 versus only 20% of those without a detectable IFX14 (P ¼ 0.06).
FIGURE 1. Patient population.
Clinical and Biomarker Associations with PR at Week 54
Week 54 Clinical Efficacy Outcomes Of the 50 patients who entered IFX maintenance therapy, 60% (30/50), 70% (35/50), and 58% (29/50) were in PR, clinical remission, and clinical and laboratory remission at week 54, respectively. SDR from week 14 through week 54 (SDR14) and week 22 through week 54 (SDR22) was observed in 38% (19/50) and 42% (21/50), respectively.
Infliximab and Anti-infliximab Antibody Status at Weeks 14 and 54 IFX and ATI status was defined as detectable (above assay level of detection) or undetectable. A total of 71% had a detectable IFX level at week 14 (median [IQR], 5.1 [3.1–7.5] mg/mL), and 69% had a detectable IFX level at week 54 (median [IQR], 5.2 [3.7–9.1] mg/mL). McNemar’s test showed no overall significant changes in the status of patients who had detectable IFX levels at
On univariate analysis, height, weight, and albumin levels at week 0 or at week 14 were not associated with week 54 outcomes. Forty-five percent of patients had an elevated CRP ($1 mg/dL) at baseline (CRP0) and 14% had an elevated CRP at week 14 (CRP14). CRP0 status (CRP ,1 mg/dL versus .1 mg/dL) was not associated with week 54 PR (P ¼ 0.23). However, median CRP0 level was significantly higher in those patients in PR (0.5 mg/dL [IQR, 0.08–1.3 mg/dL]) as compared with those not in PR (2.6 mg/dL [IQR, 0.2–3.6 mg/dL; P ¼ 0.01]). Both CRP14 status (P ¼ 0.0002) and median CRP14 level (0.05 mg/dL [IQR, 0.01–0.2 mg/dL] versus 0.7 mg/dL [IQR, 0.1–1.5 mg/dL];
TABLE 1. Baseline Patient Demographics Baseline Demographics Age at diagnosis, median (range), yr CD:UC, % Male:female, % Disease duration, median (range), mo Immunomodulator use, % PCDAI score, median (IQR) Partial Mayo score, median (IQR) Median CRP level, mg/dL CRP status (,1.0 mg/dL:.1.0 mg/dL), %
N ¼ 58 11.4 (6.6–18.4) 81:19 53:47 3.6 (0.2–93.0) 43 20 (12.5–32.5) 5 (4–7) 1.1 55:45
PCDAI, pediatric Crohn’s disease activity index.
FIGURE 2. Week 14 infliximab level and outcome.
1710
| www.ibdjournal.org
Inflamm Bowel Dis Volume 20, Number 10, October 2014
Infliximab Levels Predict Remission in IBD
TABLE 2. Week 14 Infliximab Levels and Outcomes Week 54 Outcome (Yes Versus No)
IFX14 Median Level, mg/mL
PR Clinical remission Clinical and laboratory remission SDR14 SDR22
4.7 3.2 4.2 5.5 5.1
Pa
versus 2.6 versus 2.2 versus 3.0 versus 3.1 versus 3.0
0.03 0.07 0.07 0.05 0.04
different than IFX14 levels alone (P ¼ 0.80). The AUC increased, albeit insignificantly, with the addition of week 14 body mass index (BMI14), CRP14 status, or delta CRP as compared with IFX14 alone (Table 4). As shown in Table 5, an IFX14 level of $4 mg/mL was associated with a PPV of 76% and AUC of 0.64 for week 54 PR. In raising the cut point from $3 mg/mL to $4 mg/mL, the PPV increased by 12%. An IFX14 level $7 mg/mL had a 100% PPV with an AUC of 0.67.
a P value: Wilcoxon rank sum test. Bold text indicates significant P values.
DISCUSSION
P ¼ 0.0001) were associated with week 54 PR. Delta CRP (CRP0 level 2 CRP14 level) was not associated with PR. Higher CRP14 levels correlated with lower IFX14 levels (r ¼ 20.64, P , 0.0001). Interestingly, a significant correlation was not observed between CRP54 and IFX54 levels (r ¼ 0.25, P ¼ 0.13).
Impact of Immunomodulators Immunomodulator use either at initiation of IFX (IMM0) or by week 14 (IMM14) was not associated with any of the week 54 efficacy outcomes. Both IFX14 and IFX54 median levels were numerically but not statistically higher in patients who initiated an immunomodulator at the time of IFX initiation (IFX14: 3.1 [IMM0 no] versus 6.8 [IMM0 yes] mg/mL; P ¼ 0.14 and IFX54: 3.45 [IMM0 no] versus 7.6 [IMM0 yes] mg/mL; P ¼ 0.29). IMM0 had no association with ATI14 status (10.3% [IMM0 no] versus 9.1% [IMM0 yes]; P . 0.99) or ATI54 (27.8% [IMM0 no] versus 18.2% [IMM0 yes]; P ¼ 0.70). IMM14 was not associated with ATI14 status (12.5% [IMM14 no] versus 7.7% [IMM14 yes]; P ¼ 0.66). However, IMM14 was found to be associated with ATI54 status (40.9% [IMM14 no] versus 12.0% [IMM14 yes]; P ¼ 0.04).
Predictive Value of Week 14 Infliximab Levels IFX14 level was found to have an AUC of 0.68. CRP14 status alone had an AUC of 0.66, which was not statistically
Early dose optimization, before loss of response due to subtherapeutic drug levels, may be an important strategy in the management of anti-tumor necrosis factor–treated patients. This is the first study to prospectively demonstrate that trough IFX levels, before the first maintenance infusion at week 14, are associated with therapeutic outcomes at week 54 in pediatric patients with IBD. Our data lend support to previous retrospective studies examining early predictors of week 54 clinical outcomes. Vande Casteele et al10 showed that a low week 14 IFX level (,2.2 mg/mL) predicted ATI development and subsequent IFX discontinuation (74% specificity and 82% sensitivity). Bortlik et al9 similarly found that IFX levels .3 mg/mL at week 14 to 22 predicted a sustained response. The specificity of $3 mg/mL in our cohort was only 50% with a PPV of only 64%. The PPV increased to 75% when using a cut point of 4 mg/mL or above and 100% for a cut point of at least 7 mg/mL. There is much discussion surrounding the impact of immunomodulators on trough and ATI levels. Bortlik et al9 found that in those with concomitant immunomodulator use, week 14 to 22 IFX levels were higher, and the frequency of ATIs was lower. Interestingly, in our study, the presence of baseline concomitant immunomodulator use was not associated with higher IFX levels at any time point. Concomitant immunomodulator use was also not associated with decreased week 14 ATI development. However, patients who initiated an immunomodulator by the week 14 infusion were found to have significantly less ATI measured at
TABLE 3. Week 14 Infliximab and Anti-infliximab Antibody Status and Outcomes IFX14 Status
ATI14 Status
Detectable
Detectable
Outcome
Yes (n ¼ 35)
No (n ¼ 15)
PR Clinical remission Clinical and laboratory remission SDR14 SDR22
69% 77% 66% 46% 51%
40% 53% 40% 20% 20%
(24/35) (27/35) (23/35) (16/35) (18/35)
(6/15) (8/15) (6/15) (3/15) (3/15)
a
P
0.11 0.11 0.12 0.12 0.06
Yes (n ¼ 5) 60% 60% 60% 20% 20%
(3/5) (3/5) (3/5) (1/5) (1/5)
No (n ¼ 45) 60% 71% 58% 40% 44%
(27/45) (32/45) (26/45) (18/45) (20/45)
Pa .0.99 0.63 .0.99 0.64 0.38
a
P-value: Fisher’s exact test.
www.ibdjournal.org |
1711
Inflamm Bowel Dis Volume 20, Number 10, October 2014
Singh et al
TABLE 4. Predictive Models for Week 54 PR Models
AUC (95% CI)
IFX14 level CRP14 status IFX14 level and BMI14 IFX14 level and delta CRP IFX14 level and CRP14 status
0.68 0.66 0.74 0.73 0.74
Pa — 0.80 0.27 0.15 0.06
(0.53–0.83) (0.55–0.76) (0.59–0.88) (0.59–0.88) (0.60–0.89)
a AUC of each model versus AUC IFX14 level. CI, confidence interval.
week 54. Hanauer et al11 found that early ATIs had a negative effect on serum IFX concentration, duration of response, and improvement in Crohn’s disease activity index. Thus, antibodies are an important determinant of IFX efficacy, and efforts should be made to optimize the dosing of IFX and minimize ATI formation. Vande Casteele et al10 reported that an ATI level .9 U/mL at week 14 was negatively associated with IFX efficacy. In our cohort, 10% of our patients had already developed ATIs by week 14 and 29% of patients had IFX levels below the lower limit of detection. Week 14 may be an opportune time to determine the pharmacokinetic and immunogenicity profile of the patient, as it is the first 8-week interval postinduction. If a patient has low IFX levels already at week 14, there is a high likelihood that this pharmacokinetic profile will persist through subsequent 8-week intervals. Thus, early dose adjustment may result in better outcomes. It has been reported that therapeutic IFX levels and sustained remission rates are associated with lower CRP levels.9 Schnitzler et al14 found that during IFX induction, CRP normalization was associated with better outcomes. We similarly observed that higher CRP levels correlated with lower IFX levels at week 14, and lower
TABLE 5. Clinical Utility of Week 14 IFX Level Cut Points IFX14 Cut PPV, Point, mg/mL Sensitivity Specificity % $1 $2 $3 $4 $5 $5.5 $6 $7
80 70 60 53 50 47 37 33
45 45 50 75 85 90 95 100
69 66 64 76 83 88 92 100
NPV, negative predictive values; CI, confidence interval.
1712
| www.ibdjournal.org
NPV, % 60 50 46 52 53 53 50 50
AUC (95% CI) 0.63 0.58 0.55 0.64 0.68 0.68 0.66 0.67
(0.49–0.80) (0.44–0.72) (0.41–0.69) (0.51–0.78) (0.55–0.80) (0.57–0.80) (0.56–0.76) (0.58–0.75)
baseline and week 14 CRP levels were significantly associated with clinical outcomes. In our study, IFX levels or an elevated CRP at week 14, individually, were moderately predictive (AUC 0.68) of PR at week 54. There was, however, an increase in the AUC to 0.74 when our predictive model combined IFX levels with CRP elevation at week 14. Larger studies are needed to validate our findings and improve the predictive power and utility of our model. We did not actively use IFX trough levels to adjust the dose in this study. However, based on our data, this may be clinically important. The TAXIT study optimized the IFX dose in patients already doing well on maintenance IFX to attain a trough level between 3 and 7 mg/mL.15 The CRP and activity indices did significantly improve in patients with CD when the dose was optimized. In the second part of TAXIT, patients were randomized to clinically based versus IFX level–based dose adjustments. Although both groups resulted in the same proportion of patients achieving clinical and biologic remission (69% versus 72%, P ¼ 0.7), those patients treated based on IFX levels did have less ATI formation and higher IFX levels than those treated clinically.16 The study does suggest that occasional monitoring may be warranted in patients doing well on maintenance therapy. It remains unknown whether higher trough levels are needed at the end of induction or at the start of maintenance in patients with active disease, as compared with patients whose disease activity is controlled during maintenance. Currently, it is not clinically indicated to measure an IFX trough level at every infusion. We hypothesize that proactively adjusting IFX dose based on the week 14 level may result in sustained therapeutic trough levels throughout maintenance and result in better longterm outcomes. Alternatively, or in addition to dose adjustment, if a patient is on monotherapy and already has a low IFX trough level at week 14, adding a concomitant immunomodulator may be prudent to decrease drug clearance. This has yet to be proven in a large number of patients; however, a number of small studies do support this intervention.17,18 A potential limitation to our study is that not all patients were tested with the same assay, given the timing of the study and Prometheus changing their assay technique as discussed in the Materials and Methods section. However, like other studies, we confirmed the strong correlation of IFX levels (r2 ¼ 0.91) between the ELISA and the HMSA assays.10 It is important to recognize that there may be subtle differences in optimal cut points predicting efficacy outcomes between different assays. The bigger issue is the fact that the ELISA uses microgram per milliliter and the HMSA uses units per milliliter to measure levels of ATI, and we did not feel confident using a conversion factor, as our correlation coefficient for ATI levels was not as strong as that for IFX levels. If we had used the HMSA assay for all samples, we would have been able to more confidently assess the association between ATI levels and week 54 outcomes and not just ATI status. The HMSA enables the measurement of ATI, irrespective of IFX levels, and thus is considered to be the preferred assay for ATI quantification. Another potential limitation to our study is that endoscopic remission was not our primary end point, which may have improved the robustness of our week 14
Inflamm Bowel Dis Volume 20, Number 10, October 2014
data. Not all patients had an ileocolonosocpy performed at week 54, and thus clinical and laboratory end points were measured. This is the first prospective study to confirm that the trough IFX level before entering the maintenance phase of therapy is associated with 1-year remission outcomes. Furthermore, the combination of IFX level and CRP elevation at week 14 improved the predictive capability of PR at 1 year. The replication of this study in a larger cohort is important and will lead to stronger predictive models. A clinical trial randomizing patients to dose and/ or frequency adjustment based on levels obtained postinduction versus clinically driven dose intensification is warranted. The success of IFX therapy is heavily dependent on individualizing the dosing strategy for each patient and establishing the pharmacokinetic profile of the patient early in the treatment course to optimize clinical efficacy and patient outcomes.
REFERENCES
1. Hyams J, Crandall W, Kugathasan S, et al. Induction and maintenance infliximab therapy for the treatment of moderate-to-severe Crohn’s disease in children. Gastroenterology. 2007;132:863–873. 2. Hyams JS, Lerer T, Griffiths A, et al. Long-term outcome of maintenance infliximab therapy in children with Crohn’s disease. Inflamm Bowel Dis. 2009;15:816–822. 3. Hyams J, Damaraju L, Blank M, et al. Induction and maintenance therapy with infliximab for children with moderate to severe ulcerative colitis. Clin Gastroenterol Hepatol. 2012;10:391–399. 4. Maser EA, Villela R, Silverberg MS, et al. Association of trough serum infliximab to clinical outcome after scheduled maintenance treatment for Crohn’s disease. Clin Gastroenterol Hepatol. 2006;4:1248–1254. 5. Seow CH, Newman A, Irwin SP, et al. Trough serum infliximab: a predictive factor of clinical outcome for infliximab treatment in acute ulcerative colitis. Gut. 2010;59:49–54. 6. Vermeire S, Noman M, Van Assche G, et al. Effectiveness of concomitant immunosuppressive therapy in suppressing the formation of antibodies to infliximab in Crohn’s disease. Gut. 2007;56:1226–1231.
Infliximab Levels Predict Remission in IBD
7. Nanda KS, Cheifetz AS, Moss AC. Impact of antibodies to infliximab on clinical outcomes and serum infliximab levels in patients with inflammatory bowel disease (IBD): a meta-analysis. Am J Gastroenterol. 2013;108: 40–47. 8. Feagan BG, Singh S, Lockton S, et al. Novel infliximab (IFX) and antibody-to-infliximab (ATI) assays are predictive of disease activity in patients with Crohns Disease. Gastroenterology. 2012;8S:12–14. 9. Bortlik M, Duricova D, Malickova K, et al. Infliximab trough levels may predict sustained response to infliximab in patients with Crohn’s disease. J Crohns Colitis. 2013;7:736–743. 10. Vande Casteele N, Gils A, Singh S, et al. Antibody response to infliximab and its impact on pharmacokinetics can be transient. Am J Gastroenterol. 2013;108:962–971. 11. Hanauer SB, Wagner CL, Bala M, et al. Incidence and importance of antibody responses to infliximab after maintenance or episodic treatment in Crohn’s disease. Clin Gastroenterol Hepatol. 2004;2:542–553. 12. Guerra I, Chaparro M, Bermejo F, et al. Utility of measuring serum concentrations of anti-TNF agents and anti-drug antibodies in inflammatory bowel disease. Curr Drug Metab. 2011;12:594–598. 13. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44:837–845. 14. Schnitzler F, Fidder H, Ferrante M, et al. Long-term outcome of treatment with infliximab in 614 patients with Crohn’s disease: results from a singlecentre cohort. Gut. 2009;58:492–500. 15. Vande Casteele N, Gils N, Ballet V, et al. Individualised infliximab treatment using therapeutic drug monitoring: a prospective controlled trough level adapted infliXImab treatment (TAXIT) trial. J Crohn’s Colitis. 2012; 6:S6. 16. Vande Casteele N, Gils N, Ballet V, et al. Randomised controlled trial of drug level versus maintenance therapy in IBD: final results of the TAXIT study. United European Gastroenterol J. 2013;1:A1–A134. 17. Ben-Horin S, Waterman S, Kopylov U, et al. Addition of an immunomodulator to infliximab therapy eliminates antidrug antibodies in serum and restores clinical response of patients with inflammatory bowel disease top of form. Clin Gastroenterol Hepatol. 2013;11: 444–447. 18. Ong DE, Kamm MA, Hartono JL, et al. Addition of thiopurines can recapture response in patients with Crohn’s disease who have lost response to anti-tumor necrosis factor monotherapy. J Gastroenterol Hepatol. 2013;28:1595–1599.
www.ibdjournal.org |
1713
Early Infliximab Trough Levels Are Associated with Persistent Remission in Pediatric Patients with Inflammatory Bowel Disease Namita Singh, MD,* Casey J. Rosenthal, BS,* Gil Y. Melmed, MD,†,‡ James Mirocha, MS,§ Sharmayne Farrior, BSN,* Silvia Callejas, BSN,* Bhavna Tripuraneni, BS,* Shervin Rabizadeh, MD,* ,‡ and Marla C. Dubinsky, MD* ,‡
Background: Low infliximab (IFX) trough levels and high anti-infliximab antibodies (ATI) levels are associated with loss of response to IFX. Optimizing IFX levels to maintain target concentrations before loss of response may improve long-term efficacy. We hypothesized that trough levels at week 14 are predictive of IFX durability.
Methods: A prospective observational cohort of pediatric patients with inflammatory bowel disease initiating IFX had IFX and ATI levels drawn at weeks 14 and 54. Primary outcome was week 54 persistent remission (PR), defined as clinical remission without IFX dose intensification. Univariate analyses tested associations of week 14 IFX (IFX14) and ATI (ATI14) levels, clinical and laboratory data with week 54 outcomes. Receiver operating curve analysis and positive and negative predictive values for IFX14 cut-off points were examined.
Results: Of 58 patients enrolled, 8 (13%) stopped IFX before week 14 and 4 discontinued IFX between weeks 14 and 54. IFX14 level (P ¼ 0.03),
baseline C-reactive protein (CRP) level (P ¼ 0.01), and week 14 CRP (CRP14) level (P ¼ 0.0001) were associated with PR. A model with IFX14 levels predicting PR had an area under the receiver operating curve of 0.68 and a model with both IFX14 level and CRP14 .1.0 mg/dL had an area under the receiver operating curve of 0.74. IFX14 cut points of .3, .4, and .7 mg/mL had positive predictive values of 64%, 76% and 100%, respectively, for predicting PR.
Conclusions: Both IFX levels and CRP at week 14 were significantly associated with week 54 efficacy. A model combining both CRP and IFX at week 14 may help predict remission at week 54. (Inflamm Bowel Dis 2014;20:1708–1713) Key Words: pharmacokinetics, infliximab level, anti-infliximab antibody, inflammatory bowel disease, pediatrics
I
nfliximab (IFX) is approved for use in children with both Crohn’s disease (CD) and ulcerative colitis (UC). The REACH trial reported that close to 90% of children responded to induction and close to 60% of children were in remission at 1 year.1 Almost half of these patients, however, lost response and required dose modification to recapture remission.1 A similar durability of IFX was reported outside the clinical trial arena, whereby 70% were still on IFX after 3 years of follow-up.2 However, half of these Received for publication May 27, 2014; Accepted June 10, 2014. From the Departments of *Pediatrics and †Medicine, Cedars-Sinai Medical Center, Los Angeles, California; ‡F. Widjaja Foundation Inflammatory Bowel and Immunobiology Institute, Cedars-Sinai Medical Center, Los Angeles, California; and § Department of Biostatistics, Cedars-Sinai Medical Center, Los Angeles, California. Supported in part by the National Institute/National Institute of Digestive Disease and Kidney (NIDDK) Grant DK084554 (M.C.D.), and The Abe and Claire Levine Chair in Pediatric IBD (M.C.D.). N. Singh and C. J. Rosenthal contributed equally to this work. M. C. Dubinsky: Consultant for Prometheus Laboratories Inc. and Janssen Pharmaceuticals Inc. G. Y. Melmed: Consultant for Janssen Pharmaceuticals Inc. The other authors have no conflicts of interest to disclose. Reprints: Marla C. Dubinsky, MD, 8635 West 3rd Street, Suite 1165W, Los Angeles, CA 90048 (e-mail: [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000137 Published online 22 August 2014.
1708
| www.ibdjournal.org
patients required dosing adjustments. For UC, close to 75% of pediatric patients respond to IFX induction and about 40% are in remission at 1 year.3 The maintenance of a sustained and durable remission with IFX has become an important clinical challenge. The etiology for secondary loss of response may be complex; however, a common factor is the pharmacokinetic profile of IFX and immunogenicity. There have been many cross-sectional studies demonstrating the positive association of IFX trough level and the negative association of anti-infliximab antibodies (ATI) with efficacy outcomes.4–6 A recent meta-analysis confirmed that the presence of ATI is associated with a significantly higher risk of loss of clinical response to IFX and lower serum IFX levels in patients with inflammatory bowel disease (IBD).7 The pooled risk ratio of loss of clinical response to IFX in patients with IBD who had ATIs was 3.2 (95% confidence interval, 2.0–4.9; P , 0.0001). Attempts have been made to define a minimal trough and antibody level that is best associated with improved IFX outcomes, such that being above a laboratory defined cutoff may not be sufficient. A trough level of 3 mg/mL or above has been shown to be independently associated with a lower C-reactive protein (CRP) and has been proposed as a cut point to maximize clinical efficacy of IFX.8 The current approach is to obtain an IFX trough Inflamm Bowel Dis Volume 20, Number 10, October 2014
Inflamm Bowel Dis Volume 20, Number 10, October 2014
and ATI level when a patient has lost response to IFX to seek an explanation as to why the patient is no longer responding to standard on-label dosing. The key, however, may be to determine the pharmacokinetic profile of patients before losing response and to intervene proactively and optimize to a target concentration. Akin to thiopurine monitoring, early measurement of drug levels during or at the end of induction may enable optimization of IFX dosing proactively and lead to less loss of response and less immunogenicity. Three retrospective studies conducted in adult patients with IBD have shown that higher IFX trough levels and lower ATI levels at the time of the week 14 infusion were associated with improved sustained disease response.9–11 In this prospective observational study, we hypothesized that week 14 IFX trough levels predicted week 54 IFX outcomes.
MATERIALS AND METHODS Patient Population and Ethical Considerations Pediatric patients with CD and UC younger than 21 years initiating IFX and enrolled in a prospective multicenter NIH funded study examining Genome Wide Association (GWAS) predictors of response to IFX were eligible. The decision to treat with IFX was at the discretion of the investigator. Patients received standard IFX induction regimen of 5 mg/kg infused at week 0, 2, and 6. No dose adjustments were made during the induction phase. Only those patients enrolled at Cedars-Sinai Medical Center (CSMC) who completed 54 weeks of follow-up were included in this substudy. This study was approved by the CSMC Institutional Review Board (IRB No. 19251).
Disease Activity Assessments At each infusion, a physical examination and standard of care laboratories, including complete metabolic panel, CRP level, erythrocyte sedimentation rate, complete blood count with differential, and iron panel, were performed. Disease activity was measured using the Pediatric Crohn’s Disease Activity Index for patients with CD aged ,20 years or the Crohn’s Disease Activity Index for patients with CD aged 20 years and above, and the Partial Mayo score for all patients with UC.
Outcome Measures and Definitions Our primary outcome was week 54 persistent remission (PR), which was defined as clinical remission based on the relevant disease activity index at week 54, in the absence of any dose intensification during IFX maintenance. Secondary outcomes included clinical remission that looked solely at whether a patient met the criteria of remission according to the appropriate disease activity index, and clinical and laboratory remission that consisted of a patient being in PR in addition to having a normal CRP level. Sustained durable remission (SDR) defined a patient who was in PR from week 14 through week 54 (SDR14) or from week 22 through week 54 (SDR22). Primary nonresponders were defined as patients who stopped IFX before the first maintenance infusion (week 14). Early terminators were patients who entered
Infliximab Levels Predict Remission in IBD
the maintenance phase at week 14, but stopped IFX therapy before the week 54 end point.
Serum Infliximab and Anti-infliximab Antibody Measurements IFX trough and ATI levels were drawn before the week 14 infusion and just before the week 54 infusion or at early termination and sent to Prometheus Laboratories (San Diego, CA) for testing. Samples collected before July 1, 2012, were tested by an enzyme-linked immunosorbent assay (ELISA).12 The lower limit of detection of this assay was 1.4 mg/mL for IFX and 1.69 mg/mL for ATI. Samples collected after July 1, 2012, were tested using a homogenous mobility shift assay (HMSA) with a lower limit of detection of 1.0 mg/mL for IFX and 3.1 U/mL for ATI.10 Fifty percent of the samples were tested using both assays during the transition to the new HMSA assay to confirm previous reports of a high correlation between ELISA and HMSA for IFX levels.10 We observed a correlation coefficient of 0.91 in this subset of samples. Given the high correlation, we elected to include those IFX levels tested with ELISA only in both the qualitative and quantitative analyses. However, our correlation coefficient between the 2 assays for ATI levels was only 0.75, thus we elected to only report ATI status qualitatively (detectable or not detectable), and not quantitatively. ATI levels are reported in mg/mL for the ELISA and U/mL for the HMSA.12
Statistical Analysis Numerical variables were summarized by median and range or interquartile range (IQR) and were compared across independent groups by the Wilcoxon rank sum test. Associations between numerical variables were assessed by the Spearman rank order correlation coefficient. Categorical variables were summarized by frequency and percent and were compared across independent groups by the Fisher’s exact test. Change in status on binary variables (e.g., detectable IFX at weeks 14 and 54) was assessed by McNemar’s test for related proportions. Positive (PPV) and negative predictive values were calculated in the standard fashion. The method of DeLong et al13 was used to compare area under the receiver operating characteristic curves (AUC) for related models. A 2-sided 0.05 significance level was used throughout. Statistical calculations were made using SAS Version 9.2 (SAS Institute, Cary, NC).
RESULTS Patient Population A total of 58 patients were enrolled. Eight patients (13%) were primary nonresponders and 50 entered the maintenance phase of IFX therapy (week 14). Four patients (8%) were early terminators (stopped IFX between week 14 and week 54) and 46 remained on IFX and completed the study through week 54 (Fig. 1). The median age at diagnosis for the cohort was 11.4 (range, 6.6–18.4) years, median disease duration was 3.6 (range, 0.2–93.0) months, 53% of patients were male, and 81% of all patients had a diagnosis of CD (Table 1). www.ibdjournal.org |
1709
Inflamm Bowel Dis Volume 20, Number 10, October 2014
Singh et al
week 14 and 54 (P . 0.99). IFX14 levels positively correlated with IFX54 levels (r ¼ 0.46, P ¼ 0.001). At week 14, 10% of patients had a detectable ATI level and 26% had a detectable ATI level at week 54 (McNemar’s P ¼ 0.02).
Week 14 Infliximab Levels and Week 54 Outcomes
Median IFX14 level was 4.7 mg/mL for patients in PR versus only 2.6 mg/mL for patients not in PR (P ¼ 0.03) (Fig. 2). The median IFX14 level was 5.1 mg/mL in patients in SDR22 versus 3.0 mg/mL in patients not in SDR (P ¼ 0.04). Higher median IFX14 levels were also observed for the other secondary outcomes (Table 2). Unlike median IFX14 levels, IFX14 status was not significantly associated with week 54 outcomes (Table 3). Sixty-nine percent of patients with a detectable IFX14 were in PR at week 54, as compared with only 40% without a detectable IFX14 level (P ¼ 0.11). In addition, 51% of patients with detectable IFX14 were in SDR22 versus only 20% of those without a detectable IFX14 (P ¼ 0.06).
FIGURE 1. Patient population.
Clinical and Biomarker Associations with PR at Week 54
Week 54 Clinical Efficacy Outcomes Of the 50 patients who entered IFX maintenance therapy, 60% (30/50), 70% (35/50), and 58% (29/50) were in PR, clinical remission, and clinical and laboratory remission at week 54, respectively. SDR from week 14 through week 54 (SDR14) and week 22 through week 54 (SDR22) was observed in 38% (19/50) and 42% (21/50), respectively.
Infliximab and Anti-infliximab Antibody Status at Weeks 14 and 54 IFX and ATI status was defined as detectable (above assay level of detection) or undetectable. A total of 71% had a detectable IFX level at week 14 (median [IQR], 5.1 [3.1–7.5] mg/mL), and 69% had a detectable IFX level at week 54 (median [IQR], 5.2 [3.7–9.1] mg/mL). McNemar’s test showed no overall significant changes in the status of patients who had detectable IFX levels at
On univariate analysis, height, weight, and albumin levels at week 0 or at week 14 were not associated with week 54 outcomes. Forty-five percent of patients had an elevated CRP ($1 mg/dL) at baseline (CRP0) and 14% had an elevated CRP at week 14 (CRP14). CRP0 status (CRP ,1 mg/dL versus .1 mg/dL) was not associated with week 54 PR (P ¼ 0.23). However, median CRP0 level was significantly higher in those patients in PR (0.5 mg/dL [IQR, 0.08–1.3 mg/dL]) as compared with those not in PR (2.6 mg/dL [IQR, 0.2–3.6 mg/dL; P ¼ 0.01]). Both CRP14 status (P ¼ 0.0002) and median CRP14 level (0.05 mg/dL [IQR, 0.01–0.2 mg/dL] versus 0.7 mg/dL [IQR, 0.1–1.5 mg/dL];
TABLE 1. Baseline Patient Demographics Baseline Demographics Age at diagnosis, median (range), yr CD:UC, % Male:female, % Disease duration, median (range), mo Immunomodulator use, % PCDAI score, median (IQR) Partial Mayo score, median (IQR) Median CRP level, mg/dL CRP status (,1.0 mg/dL:.1.0 mg/dL), %
N ¼ 58 11.4 (6.6–18.4) 81:19 53:47 3.6 (0.2–93.0) 43 20 (12.5–32.5) 5 (4–7) 1.1 55:45
PCDAI, pediatric Crohn’s disease activity index.
FIGURE 2. Week 14 infliximab level and outcome.
1710
| www.ibdjournal.org
Inflamm Bowel Dis Volume 20, Number 10, October 2014
Infliximab Levels Predict Remission in IBD
TABLE 2. Week 14 Infliximab Levels and Outcomes Week 54 Outcome (Yes Versus No)
IFX14 Median Level, mg/mL
PR Clinical remission Clinical and laboratory remission SDR14 SDR22
4.7 3.2 4.2 5.5 5.1
Pa
versus 2.6 versus 2.2 versus 3.0 versus 3.1 versus 3.0
0.03 0.07 0.07 0.05 0.04
different than IFX14 levels alone (P ¼ 0.80). The AUC increased, albeit insignificantly, with the addition of week 14 body mass index (BMI14), CRP14 status, or delta CRP as compared with IFX14 alone (Table 4). As shown in Table 5, an IFX14 level of $4 mg/mL was associated with a PPV of 76% and AUC of 0.64 for week 54 PR. In raising the cut point from $3 mg/mL to $4 mg/mL, the PPV increased by 12%. An IFX14 level $7 mg/mL had a 100% PPV with an AUC of 0.67.
a P value: Wilcoxon rank sum test. Bold text indicates significant P values.
DISCUSSION
P ¼ 0.0001) were associated with week 54 PR. Delta CRP (CRP0 level 2 CRP14 level) was not associated with PR. Higher CRP14 levels correlated with lower IFX14 levels (r ¼ 20.64, P , 0.0001). Interestingly, a significant correlation was not observed between CRP54 and IFX54 levels (r ¼ 0.25, P ¼ 0.13).
Impact of Immunomodulators Immunomodulator use either at initiation of IFX (IMM0) or by week 14 (IMM14) was not associated with any of the week 54 efficacy outcomes. Both IFX14 and IFX54 median levels were numerically but not statistically higher in patients who initiated an immunomodulator at the time of IFX initiation (IFX14: 3.1 [IMM0 no] versus 6.8 [IMM0 yes] mg/mL; P ¼ 0.14 and IFX54: 3.45 [IMM0 no] versus 7.6 [IMM0 yes] mg/mL; P ¼ 0.29). IMM0 had no association with ATI14 status (10.3% [IMM0 no] versus 9.1% [IMM0 yes]; P . 0.99) or ATI54 (27.8% [IMM0 no] versus 18.2% [IMM0 yes]; P ¼ 0.70). IMM14 was not associated with ATI14 status (12.5% [IMM14 no] versus 7.7% [IMM14 yes]; P ¼ 0.66). However, IMM14 was found to be associated with ATI54 status (40.9% [IMM14 no] versus 12.0% [IMM14 yes]; P ¼ 0.04).
Predictive Value of Week 14 Infliximab Levels IFX14 level was found to have an AUC of 0.68. CRP14 status alone had an AUC of 0.66, which was not statistically
Early dose optimization, before loss of response due to subtherapeutic drug levels, may be an important strategy in the management of anti-tumor necrosis factor–treated patients. This is the first study to prospectively demonstrate that trough IFX levels, before the first maintenance infusion at week 14, are associated with therapeutic outcomes at week 54 in pediatric patients with IBD. Our data lend support to previous retrospective studies examining early predictors of week 54 clinical outcomes. Vande Casteele et al10 showed that a low week 14 IFX level (,2.2 mg/mL) predicted ATI development and subsequent IFX discontinuation (74% specificity and 82% sensitivity). Bortlik et al9 similarly found that IFX levels .3 mg/mL at week 14 to 22 predicted a sustained response. The specificity of $3 mg/mL in our cohort was only 50% with a PPV of only 64%. The PPV increased to 75% when using a cut point of 4 mg/mL or above and 100% for a cut point of at least 7 mg/mL. There is much discussion surrounding the impact of immunomodulators on trough and ATI levels. Bortlik et al9 found that in those with concomitant immunomodulator use, week 14 to 22 IFX levels were higher, and the frequency of ATIs was lower. Interestingly, in our study, the presence of baseline concomitant immunomodulator use was not associated with higher IFX levels at any time point. Concomitant immunomodulator use was also not associated with decreased week 14 ATI development. However, patients who initiated an immunomodulator by the week 14 infusion were found to have significantly less ATI measured at
TABLE 3. Week 14 Infliximab and Anti-infliximab Antibody Status and Outcomes IFX14 Status
ATI14 Status
Detectable
Detectable
Outcome
Yes (n ¼ 35)
No (n ¼ 15)
PR Clinical remission Clinical and laboratory remission SDR14 SDR22
69% 77% 66% 46% 51%
40% 53% 40% 20% 20%
(24/35) (27/35) (23/35) (16/35) (18/35)
(6/15) (8/15) (6/15) (3/15) (3/15)
a
P
0.11 0.11 0.12 0.12 0.06
Yes (n ¼ 5) 60% 60% 60% 20% 20%
(3/5) (3/5) (3/5) (1/5) (1/5)
No (n ¼ 45) 60% 71% 58% 40% 44%
(27/45) (32/45) (26/45) (18/45) (20/45)
Pa .0.99 0.63 .0.99 0.64 0.38
a
P-value: Fisher’s exact test.
www.ibdjournal.org |
1711
Inflamm Bowel Dis Volume 20, Number 10, October 2014
Singh et al
TABLE 4. Predictive Models for Week 54 PR Models
AUC (95% CI)
IFX14 level CRP14 status IFX14 level and BMI14 IFX14 level and delta CRP IFX14 level and CRP14 status
0.68 0.66 0.74 0.73 0.74
Pa — 0.80 0.27 0.15 0.06
(0.53–0.83) (0.55–0.76) (0.59–0.88) (0.59–0.88) (0.60–0.89)
a AUC of each model versus AUC IFX14 level. CI, confidence interval.
week 54. Hanauer et al11 found that early ATIs had a negative effect on serum IFX concentration, duration of response, and improvement in Crohn’s disease activity index. Thus, antibodies are an important determinant of IFX efficacy, and efforts should be made to optimize the dosing of IFX and minimize ATI formation. Vande Casteele et al10 reported that an ATI level .9 U/mL at week 14 was negatively associated with IFX efficacy. In our cohort, 10% of our patients had already developed ATIs by week 14 and 29% of patients had IFX levels below the lower limit of detection. Week 14 may be an opportune time to determine the pharmacokinetic and immunogenicity profile of the patient, as it is the first 8-week interval postinduction. If a patient has low IFX levels already at week 14, there is a high likelihood that this pharmacokinetic profile will persist through subsequent 8-week intervals. Thus, early dose adjustment may result in better outcomes. It has been reported that therapeutic IFX levels and sustained remission rates are associated with lower CRP levels.9 Schnitzler et al14 found that during IFX induction, CRP normalization was associated with better outcomes. We similarly observed that higher CRP levels correlated with lower IFX levels at week 14, and lower
TABLE 5. Clinical Utility of Week 14 IFX Level Cut Points IFX14 Cut PPV, Point, mg/mL Sensitivity Specificity % $1 $2 $3 $4 $5 $5.5 $6 $7
80 70 60 53 50 47 37 33
45 45 50 75 85 90 95 100
69 66 64 76 83 88 92 100
NPV, negative predictive values; CI, confidence interval.
1712
| www.ibdjournal.org
NPV, % 60 50 46 52 53 53 50 50
AUC (95% CI) 0.63 0.58 0.55 0.64 0.68 0.68 0.66 0.67
(0.49–0.80) (0.44–0.72) (0.41–0.69) (0.51–0.78) (0.55–0.80) (0.57–0.80) (0.56–0.76) (0.58–0.75)
baseline and week 14 CRP levels were significantly associated with clinical outcomes. In our study, IFX levels or an elevated CRP at week 14, individually, were moderately predictive (AUC 0.68) of PR at week 54. There was, however, an increase in the AUC to 0.74 when our predictive model combined IFX levels with CRP elevation at week 14. Larger studies are needed to validate our findings and improve the predictive power and utility of our model. We did not actively use IFX trough levels to adjust the dose in this study. However, based on our data, this may be clinically important. The TAXIT study optimized the IFX dose in patients already doing well on maintenance IFX to attain a trough level between 3 and 7 mg/mL.15 The CRP and activity indices did significantly improve in patients with CD when the dose was optimized. In the second part of TAXIT, patients were randomized to clinically based versus IFX level–based dose adjustments. Although both groups resulted in the same proportion of patients achieving clinical and biologic remission (69% versus 72%, P ¼ 0.7), those patients treated based on IFX levels did have less ATI formation and higher IFX levels than those treated clinically.16 The study does suggest that occasional monitoring may be warranted in patients doing well on maintenance therapy. It remains unknown whether higher trough levels are needed at the end of induction or at the start of maintenance in patients with active disease, as compared with patients whose disease activity is controlled during maintenance. Currently, it is not clinically indicated to measure an IFX trough level at every infusion. We hypothesize that proactively adjusting IFX dose based on the week 14 level may result in sustained therapeutic trough levels throughout maintenance and result in better longterm outcomes. Alternatively, or in addition to dose adjustment, if a patient is on monotherapy and already has a low IFX trough level at week 14, adding a concomitant immunomodulator may be prudent to decrease drug clearance. This has yet to be proven in a large number of patients; however, a number of small studies do support this intervention.17,18 A potential limitation to our study is that not all patients were tested with the same assay, given the timing of the study and Prometheus changing their assay technique as discussed in the Materials and Methods section. However, like other studies, we confirmed the strong correlation of IFX levels (r2 ¼ 0.91) between the ELISA and the HMSA assays.10 It is important to recognize that there may be subtle differences in optimal cut points predicting efficacy outcomes between different assays. The bigger issue is the fact that the ELISA uses microgram per milliliter and the HMSA uses units per milliliter to measure levels of ATI, and we did not feel confident using a conversion factor, as our correlation coefficient for ATI levels was not as strong as that for IFX levels. If we had used the HMSA assay for all samples, we would have been able to more confidently assess the association between ATI levels and week 54 outcomes and not just ATI status. The HMSA enables the measurement of ATI, irrespective of IFX levels, and thus is considered to be the preferred assay for ATI quantification. Another potential limitation to our study is that endoscopic remission was not our primary end point, which may have improved the robustness of our week 14
Inflamm Bowel Dis Volume 20, Number 10, October 2014
data. Not all patients had an ileocolonosocpy performed at week 54, and thus clinical and laboratory end points were measured. This is the first prospective study to confirm that the trough IFX level before entering the maintenance phase of therapy is associated with 1-year remission outcomes. Furthermore, the combination of IFX level and CRP elevation at week 14 improved the predictive capability of PR at 1 year. The replication of this study in a larger cohort is important and will lead to stronger predictive models. A clinical trial randomizing patients to dose and/ or frequency adjustment based on levels obtained postinduction versus clinically driven dose intensification is warranted. The success of IFX therapy is heavily dependent on individualizing the dosing strategy for each patient and establishing the pharmacokinetic profile of the patient early in the treatment course to optimize clinical efficacy and patient outcomes.
REFERENCES
1. Hyams J, Crandall W, Kugathasan S, et al. Induction and maintenance infliximab therapy for the treatment of moderate-to-severe Crohn’s disease in children. Gastroenterology. 2007;132:863–873. 2. Hyams JS, Lerer T, Griffiths A, et al. Long-term outcome of maintenance infliximab therapy in children with Crohn’s disease. Inflamm Bowel Dis. 2009;15:816–822. 3. Hyams J, Damaraju L, Blank M, et al. Induction and maintenance therapy with infliximab for children with moderate to severe ulcerative colitis. Clin Gastroenterol Hepatol. 2012;10:391–399. 4. Maser EA, Villela R, Silverberg MS, et al. Association of trough serum infliximab to clinical outcome after scheduled maintenance treatment for Crohn’s disease. Clin Gastroenterol Hepatol. 2006;4:1248–1254. 5. Seow CH, Newman A, Irwin SP, et al. Trough serum infliximab: a predictive factor of clinical outcome for infliximab treatment in acute ulcerative colitis. Gut. 2010;59:49–54. 6. Vermeire S, Noman M, Van Assche G, et al. Effectiveness of concomitant immunosuppressive therapy in suppressing the formation of antibodies to infliximab in Crohn’s disease. Gut. 2007;56:1226–1231.
Infliximab Levels Predict Remission in IBD
7. Nanda KS, Cheifetz AS, Moss AC. Impact of antibodies to infliximab on clinical outcomes and serum infliximab levels in patients with inflammatory bowel disease (IBD): a meta-analysis. Am J Gastroenterol. 2013;108: 40–47. 8. Feagan BG, Singh S, Lockton S, et al. Novel infliximab (IFX) and antibody-to-infliximab (ATI) assays are predictive of disease activity in patients with Crohns Disease. Gastroenterology. 2012;8S:12–14. 9. Bortlik M, Duricova D, Malickova K, et al. Infliximab trough levels may predict sustained response to infliximab in patients with Crohn’s disease. J Crohns Colitis. 2013;7:736–743. 10. Vande Casteele N, Gils A, Singh S, et al. Antibody response to infliximab and its impact on pharmacokinetics can be transient. Am J Gastroenterol. 2013;108:962–971. 11. Hanauer SB, Wagner CL, Bala M, et al. Incidence and importance of antibody responses to infliximab after maintenance or episodic treatment in Crohn’s disease. Clin Gastroenterol Hepatol. 2004;2:542–553. 12. Guerra I, Chaparro M, Bermejo F, et al. Utility of measuring serum concentrations of anti-TNF agents and anti-drug antibodies in inflammatory bowel disease. Curr Drug Metab. 2011;12:594–598. 13. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44:837–845. 14. Schnitzler F, Fidder H, Ferrante M, et al. Long-term outcome of treatment with infliximab in 614 patients with Crohn’s disease: results from a singlecentre cohort. Gut. 2009;58:492–500. 15. Vande Casteele N, Gils N, Ballet V, et al. Individualised infliximab treatment using therapeutic drug monitoring: a prospective controlled trough level adapted infliXImab treatment (TAXIT) trial. J Crohn’s Colitis. 2012; 6:S6. 16. Vande Casteele N, Gils N, Ballet V, et al. Randomised controlled trial of drug level versus maintenance therapy in IBD: final results of the TAXIT study. United European Gastroenterol J. 2013;1:A1–A134. 17. Ben-Horin S, Waterman S, Kopylov U, et al. Addition of an immunomodulator to infliximab therapy eliminates antidrug antibodies in serum and restores clinical response of patients with inflammatory bowel disease top of form. Clin Gastroenterol Hepatol. 2013;11: 444–447. 18. Ong DE, Kamm MA, Hartono JL, et al. Addition of thiopurines can recapture response in patients with Crohn’s disease who have lost response to anti-tumor necrosis factor monotherapy. J Gastroenterol Hepatol. 2013;28:1595–1599.
www.ibdjournal.org |
1713
