MULTICENTER COHORT STUDY OF INFLIXIMAB PHARMACOKINETICS AND THERAPY RESPONSE IN PEDIATRIC ACUTE SEVERE ULCERATIVE COLITIS

Abstract

Background and Aims:

We aimed to model infliximab (IFX) pharmacokinetics (PK) in pediatric acute severe ulcerative colitis (ASUC) and assess the association between PK parameters, including drug exposure, and clinical response.

Methods:

We studied a multicenter prospective cohort of hospitalized children initiating IFX for ASUC or IBD-unclassified. Serial IFX serum concentrations over 26 weeks were used to develop a PK model. We tested the association of PK parameter estimates with day 7 clinical response, week 8 clinical remission, week 26 corticosteroid-free clinical remission (CSF-CR) (using the Pediatric Ulcerative Colitis Activity Index) and colectomy-free survival.

Results:

38 participants received IFX (median initial dose 9.9 mg/kg). Day 7 clinical response, week 8 clinical remission, and week 26 CSF-CR occurred in 71%, 55%, and 43%, respectively. Albumin, C-reactive protein, white blood cell count, platelets, weight, and antibodies to IFX were significant covariates incorporated into a PK model. Week 26 non-remitters exhibited faster IFX clearance than remitters (P=.013). However, cumulative IFX exposure did not differ between clinical response groups. One (2.7%) and four (10.8%) participants underwent colectomy by week 26 and two years, respectively. Day 3 IFX clearance > .02 L/h was associated with colectomy (HR 58.2, 95% CI: 6.0–568.6, P < .001).

Conclusions:

At median higher-than-label IFX dosing for pediatric ASUC, baseline faster IFX CL was associated with colectomy and, at week 26, with lack of CSF-CR. IFX exposure was not predictive of clinical outcomes. Higher IFX dosing may sufficiently optimize early outcomes in pediatric ASUC. Larger studies are warranted to determine if sustained intensification can overcome rapid clearance and improve later outcomes.

ClinicalTrials.gov Identifier:

NCT02799615

Graphical Abstract

Acute severe ulcerative colitis (ASUC) affects 15% of pediatric patients with ulcerative colitis (UC), warrants hospitalization, and is associated with high risk for colectomy.^1–4 In children, ASUC is defined by a pediatric ulcerative colitis activity index (PUCAI) score 65 or greater due to severe diarrhea, abdominal pain, hematochezia, nocturnal awakening, and limitation of activities.⁵ Infliximab (IFX), a chimeric monoclonal antibody to tumor necrosis factor (TNF)-α, is the most common treatment for corticosteroid-refractory pediatric ASUC. However, outcomes after IFX in ASUC are not reflected by the pediatric randomized trial, which excluded hospitalized patients.⁶ The largest prospective study of pediatric ASUC, published 12 years ago, found that 24% of children treated with IFX at conventional dosing of 5 mg/kg undergo a colectomy prior to discharge and 52% by 1 year.⁴

These suboptimal outcomes may be related to altered IFX pharmacokinetics (PK), including rapid clearance (CL), leading to reduced drug exposure in ASUC. Proposed mechanisms for the high degree of inflammation in ASUC leading to rapid IFX CL include high TNF antigen burden, upregulation of reticuloendothelial elimination, and proteolytic degradation and leakage of the drug through inflamed bowel.⁷ Prospective studies of IFX PK in pediatric ASUC are needed to inform whether primary non-response is due to high CL and inadequate exposure (PK failure), or inflammation driven by pathogenic mechanisms not targeted by IFX (pharmacodynamic [PD] failure).

Population PK modeling with Bayesian estimation is a technique that incorporates clinical and laboratory covariates established to be associated with drug PK and prior measured drug concentrations to predict future individual PK parameters.^{8, 9} PK models have been developed for adult moderate to severe UC and adult and pediatric Crohn’s disease, but none exist for pediatric UC.^{9, 10}

We performed this prospective, multicenter, pilot and feasibility cohort study to establish a contemporary estimate of clinical response to IFX in pediatric ASUC, estimate and model IFX PK parameters and their relationship to baseline factors, and assess associations between IFX PK parameters and clinical response as proof of concept.

METHODS

Study Population

The Anti-TNF Therapy for Refractory Colitis in Hospitalized Children (ARCH) study was a prospective cohort study conducted at seven centers in the United States and Canada. Patients ages 4 to 18 years old hospitalized with UC or inflammatory bowel disease-unclassified (IBD-U) (diagnosed by established criteria¹¹) initiating treatment with IFX were enrolled between May 2016 and August 2018 and followed for 26 weeks from their first infusion. To be included, anti-TNF naïve patients must have had severe disease as defined by PUCAI ≥ 65 upon admission and ≥45 at IFX initiation (Detailed inclusion/ exclusion criteria are in Supplementary Table 1). The study was approved by the institutional review board at each site.

Procedures

The treating gastroenterologist determined IFX dosing. Clinical disease activity was assessed by PUCAI and partial Mayo scores.⁵ Colectomy status was recorded during the 26-week follow-up period and thereafter ascertained from the medical record. Laboratory testing obtained at baseline, Day 7, and prior to each infusion included complete blood count, C-reactive protein (CRP), erythrocyte sedimentation rate, and albumin.

Serum was obtained serially for measurement of IFX concentrations and antibodies to IFX (ATI) as detailed in Figure 1. IFX and ATI concentration measurements were performed by Prometheus Laboratories (San Diego, CA, USA) using a drug tolerant homogenous mobility shift assay. The lower limit of detection for ATI for this assay is 3.13 U/mL. Plasma TNF and fecal calprotectin (fCal) were obtained at enrollment, the earlier of Day 7 or day of discharge, and at the final infusion. Plasma TNF and fCal were measured centrally by enzyme-linked immunosorbent assays (ELISA) (fCal: Bühlmann, Switzerland; TNF: Human TNF-alpha Quantikine ELISA, R&D Systems, Minneapolis, MN, USA).

Figure 1.

Diagram of study protocol highlighting clinical outcomes and IFX concentrations. The inpatient hospitalization following infusion 1 included intensive IFX and ATI measurements. Following discharge, IFX and ATI concentrations were obtained prior to each infusion with a peak obtained following the penultimate infusion. The clinical outcomes included Day 7 clinical response, Week 8 colectomy-free clinical remission, and Week 26 corticosteroid-free colectomy-free clinical remission. The treating clinician determined the IFX dose and interval.

Central endoscopic and histopathologic scoring (Supplementary Methods)

Outcomes

The primary endpoint was the relationship between IFX exposure, measured by area under the IFX concentration-time curve (AUC), and Day 7 clinical response, defined as a PUCAI < 35 without a colectomy. Secondary endpoints included Week 8 colectomy-free clinical remission on IFX (CR, PUCAI <10), Week 26 corticosteroid-free, colectomy-free clinical remission on IFX (CSF-CR), and colectomy-free survival.

Population Pharmacokinetic Model Development (Supplementary Methods)

Sample Size Estimate and Statistical Analysis

We estimated requiring 36 participants to have >90% power to estimate PK parameters per FDA guidelines and clinical response at Day 7 within < 7.8% standard error.¹² The study was designed as a pilot and feasibility study. Assuming a clinical response rate between 60–70% and effect size of 0.75–1 SD, we estimated having a 54–78% power to detect an association between IFX exposure and Day 7 clinical response. Differences in PK parameters between response groups were assessed by the non-parametric Mann-Whitney U Test. Univariable logistic or cox regression was used to test associations between variables and the clinical endpoints. The small number of patients precluded the use of multivariable analysis. Given the exploratory nature of the regression analyses we did not apply a multiple comparisons correction to P-values, and instead, report false discovery rate as Q-values. To limit missing data related to IFX cessation, Week 26 cumulative IFX exposure and trough concentrations were estimated for 2 participants who discontinued IFX between 16 and 26 weeks using the PK model to simulate 1–2 additional infusions at the same dose and interval.

RESULTS

Baseline characteristics and IFX dosing

We enrolled 38 participants hospitalized with UC (n=36) or IBD-U (n=2) initiating IFX, 37 of whom completed the 26-week follow-up (Supplementary Figure 1). Neither participant with IBD-U exhibited small intestinal inflammation: one had macroscopic patchiness and the other exhibited more severe involvement proximally. Participant baseline demographic and clinical characteristics are shown in Table 1. 87% had extensive or pancolitis and 66% had a severe endoscopic Mayo score (Mayo 3).

Table 1.

Demographic and clinical characteristics of study population.

Characteristic	N=38
Age, years	14.5 (13.3, 16.8)
Sex
Female	19 (50)
Male	19 (50)
Ethnicity
Hispanic or Latino	0 (0)
Not Hispanic or Latino	38 (100)
Race
White	32 (84)
Non-white	6 (16)
Weight, kg	52.6 (42.4, 64.7)
Weight Z-score	0.4 (−0.5, 1.0)
Height, cm	163.8 (158.9, 172.5)
Height Z-score	0.3 (−0.2, 0.8)
BMI, kg/m2	19.1 (16.9, 24.3)
BMI Z-score	0.02 (−1.0, 1.0)
Family History of IBD	2 (5)
Current smoker	0 (0)
Cigarette Smoke Exposure	5 (13)
Previous Appendectomy	0 (0)
Concomitant Medications at IFX Initiation
Steroids	34 (89)
Oral 5-aminosalicyclic acid	9 (24)
6-Mercaptopurine	0
Methotrexate	1 (2.6)
Antibiotics	6 (16)
Diagnosis
Ulcerative Colitis	36 (95)
IBD-U	2 (5)
Disease Duration, days	5 (0.3, 102.7)
Disease Extent
E2: left sided, distal to splenic flexure	3 (8)
E3: extensive, hepatic flexure	6 (16)
E4: pancolitis, proximal to hepatic flexure	25 (66)
Incomplete Assessment	4 (11)
Partial Mayo Score	8 (8, 8.8)
Albumin (g/dL)	3 (2.5, 3.3)
Hemoglobin (g/dL)	10 (9.1, 11.6)
C-reactive protein (mg/dL)	2.2 (0.4, 4.0)
Platelets	405 (337, 461)
White blood cell count	13.5 (10.4, 16.6)
Fecal calprotectin (μg/g) (n = 35)	4680 (3319, 8071)
TNF (pg/mL) (n = 36)	1.9 (1.3, 3.1)
Baseline PUCAI	75 (70, 80)
Infusion 1 PUCAI	65 (55, 75)
Initial Infusion Dose (mg)	500 (400, 600)
Initial Infusion Dose (mg/kg)	9.9 (9.3, 10.3)
Mayo Endoscopic Score (n=31)	3 (2, 3)
2 (Moderate)	10 (32)
3 (Severe)	21 (68)
Total Mayo Score (n=31)	11 (10, 11)
Histology Scores (n=32)
Total Robarts Histopathology Index	8.5 (6, 13)
Rectal Biopsy Peak Eosinophil Count per high power field	22 (11.2, 44)

Quantitative variables are expressed as medians (quartile 1, quartile 3), and dichotomous variables are shown as n (%). BMI, body-mass index; IBD, inflammatory bowel disease; IBD-U, IBD-unclassified; TNF, tumor necrosis factor; PUCAI, Pediatric Ulcerative Colitis Activity Index

The median initial IFX dose was 9.9 (9.3, 10.3) mg/kg and 89% received an IFX dose greater than 7 mg/kg (Figure 2A). Sixteen percent of patients received a second IFX infusion between days 4 and 6 (median 9.3 mg/kg). Median induction and maintenance dosing ranged between 9–10 mg/kg (Figure 2B). Approximately two thirds of participants were maintained at dosing intervals ≤ 6 weeks (Figure 2C).

Figure 2.

Infliximab dosing over the study period. (A) Histogram showing relative percentage of subjects’ initial IFX dosing in mg/kg increments. (B) Growth curve modeling the mean IFX dose over the 26-week study period (blue regression line as quadratic term, with dotted green line as 95% CI) overlying a spaghetti plot of IFX doses over time for each individual. (C) Growth curve modeling the mean dosing interval (time between doses) in weeks between IFX infusions across all dosing intervals during the study period overlying a spaghetti plot of dosing intervals for each individual over all infusion intervals. The outlier low dose in (B) and short interval in (C) are due to a partial second infusion, stopped because of an infusion reaction, with a subsequent dose the following day, prior to IFX discontinuation.

Clinical Outcomes

71.1% of participants achieved Day 7 clinical response, 55.3% Week 8 CR, and 43.2% Week 26 CSF-CR (Figure 3A). Of the 76% (28/37) who had a fCal at Week 26, 21.4% (6/28) attained CSF-CR and fCal < 250 μg/g. Of the 13 participants still on IFX at Week 26 but not attaining CSF-CR, 2 were on corticosteroids (1 on budesonide 9 mg and prednisone 5 mg; 1 on prednisone 10 mg).

Figure 3.

Clinical endpoints including day 7 clinical response, week 8 clinical remission, week 26 corticosteroid-free clinical remission, and colectomy-free survival. (A) Bar graph demonstrating percent of patients that achieved each endpoint. (B) Kaplan-Meier curve of colectomy-free survival.

IFX was stopped prior to week 26 in 8 of 37 (21.6%) participants. Of the 8 participants that stopped IFX, 1 underwent colectomy, 1 had an infusion reaction and switched to adalimumab, and 6 did not respond to IFX and transitioned to vedolizumab.

The incidence of colectomy was overall low. No participants underwent colectomy prior to discharge and 2.7% (1/37) by Week 26. Charts were reviewed for incidence of colectomy after the 26-week prospective study period. Median follow-up was 15.0 (10.3, 22.4) months. By 12 and 24 months, 2.7% (1/37) and 10.8% (4/37), respectively, required colectomy (Figure 3B).

Pediatric ASUC PK model

We analyzed 312 IFX concentrations (171 trough, 73 peak, and 68 serial concentration after initial hospital infusions) to develop the final PK model, which included weight, albumin, white blood cell count (WBC), ATI, CRP, and platelets (PLT) as covariates (Figure 4A). PK parameters were estimated with acceptable precision (standard error < 30% around all covariate coefficients) except for ATI due to small number of positive samples (Supplementary Table 2). Visual predictive check (Figure 4B) and goodness-of-fit plots (Supplementary Figure 2) showed that the model-predicted IFX concentrations reflected the observed distribution without obvious bias.

Figure 4.

Pharmacokinetic modeling for pediatric ASUC. (A) The equation for the final PK model for pediatric ASUC includes albumin, ATI, WBC, weight, CRP, and PLT as covariates. (B) Visual predictive check performed using 1000 simulations to visually examine if stimulated concentrations using the final model are representative of the distribution of the observed concentrations. (C) Scatter plots showing covariates that contribute to variability seen in CL, volume of distribution, and inter-compartmental CL. The parameters are adjusted for other significant covariates.

Specific covariate-PK parameter associations are depicted in Figure 4C. Faster CL was associated with presence of ATI, higher WBC, and lower albumin. Higher central (V1) and peripheral (V2) volumes of distribution were associated with increased weight, and higher V2 was associated with increased PLT. 36.8% of participants were treated with a concomitant immunomodulator during the study (26.3% methotrexate, 13.1% thiopurine), but this was not a significant covariate for PK parameters. Analyses of covariate impact on steady state exposure and relationship of body size measures with PK parameters are detailed in the Supplementary Results and Supplementary Figure 3 and 4.

Relationship of IFX PK parameters to Clinical Response and Remission

The number of participants with sufficient IFX treatment duration to be included in PK-outcome relationship analyses at each timepoint is detailed in Supplementary Table 3. Median Day 7 IFX exposure was numerically but not significantly higher in Day 7 non-responders (25345 [18090, 29124] mg·h/L) compared to responders (19868 [16173, 23956] mg·h/L, P = .12, Figure 5A) and estimated Day 7 IFX concentrations were significantly higher in non-responders (104 [74,157]) compared to responders (66 [46,95], P = .03, Supplementary Figure 5). These findings were driven by more early second infusions prior to Day 7 in non-responders (45%) compared to responders (4%). In fact, there was no difference in exposure at Day 3, prior to additional IFX doses, between Day 7 responders and non-responders (Supplementary Figure 6). Exploratory analyses using a reduction in PUCAI of ≥ 20 points as an alternate definition of Day 7 clinical response yielded similar results (Supplementary Figure 7).

Figure 5.

Relationship between IFX PK and clinical outcomes in pediatric ASUC. Graphs comparing IFX exposure (A) and CL (B) between response/remission groups at each time point. (C) Graph comparing IFX CL across study time points. Dots represent individual exposure (AUC) or CL estimates. Box and whiskers represent median, interquartile range, and range of estimates. (D) Kaplan Meier Curve comparing colectomy-free survival between above and below a Day 3 CL threshold of .02 L/h.

Similarly, there were no significant differences in cumulative IFX exposure between those attaining and not attaining Week 8 CR or Week 26 CSF-CR (Figure 5A). Although there were no significant differences in IFX exposure between response/remission groups, there was an interesting shift in the relationship between exposure and response over the study. Median IFX exposure was numerically higher in non-responders/non-remitters compared to responders/remitters at the earlier time points of Day 7 (P = .124) and Week 8 (P = .164), whereas by Week 26, median IFX exposure became numerically higher in remitters compared to non-remitters (P = .248; Figure 5A). A similar pattern was observed for IFX concentrations (Supplementary Figure 5).

IFX CL across the cohort was significantly faster at Day 7 than that at Week 8 (P = .002) or Week 26 (P < .001) indicating CL slows with treatment (Figure 5C). There were no differences in IFX CL at Day 7 or Week 8 between response/remission groups. At Week 26, however, CL was significantly faster in non-remitters compared to remitters (P = .011, Figure 5B). Accordingly, participants attaining Week 26 CSF-CR had a numerically, but not significantly, higher median trough IFX concentration nearest Week 26 (19.5 [13.6, 30.3] μg/ml) compared to those not attaining Week 26 CSF-CR (14.2 [6.0, 21.3] μg/ml; P = .254; Supplementary Figure 5).

Relationship of IFX PK parameters to Drug Discontinuation (Supplementary Results and Supplementary Figure 8)

Clinical Outcome Prediction

We applied univariable logistic and Cox regression to assess for predictors of Day 7 clinical response and Week 26 CSF-CR (Supplementary Table 4). Lower Day 3 PUCAI trended towards an association with Day 7 clinical response (P = .058). Day 3 PUCAI had an area under the receiver operating characteristic curve (AUROC) of .747 (95%CI .559-.935; P = .033) for predicting Day 7 clinical response. A Day 3 PUCAI cutoff ≤ 45 was predictive of Day 7 clinical response with an 66.7% sensitivity, 80.0% specificity, 85.7% positive predictive value (PPV), and 57.1% negative predictive value (NPV). Also, lower predicted Day 7 IFX concentration was associated with Day 7 clinical response (P = .043), likely due to higher frequency of early second infusion in non-responders.

Lower Day 7 PUCAI was associated with Week 26 CSF-CR (P = .021). Day 7 PUCAI had an AUROC of .733 (95%CI .570-.899; P = .016) for Week 26 CSF-CR. A Day 7 PUCAI cutoff ≤ 15 was predictive of Week 26 CSF-CR with 75.0% sensitivity, 71.4% specificity, 66.7% PPV, and 78.9% NPV.

Rapid CL at Day 3 was strongly associated with risk of colectomy with a HR of 1.302 (95% CI 1.068–1.588) for each ml/h increase in CL. CL cut off greater than .02 L/h was associated with a HR for colectomy of 58.2 (95% CI 6.0–568.6; P < .001; Figure 5) and exhibited 100% sensitivity, 91.2% specificity, 57.1% PPV, and 100% NPV.

DISCUSSION

We performed, to our knowledge, the largest prospective study of IFX treatment for pediatric ASUC and developed the first IFX population PK model for this patient population. Most patients received higher than label 5 mg/kg IFX dosing. Although the majority (71%) of patients exhibited an early clinical response by Day 7, just under half (43.2%) achieved Week 26 CSF-CR. The observed incidence of colectomy was low. Patients who did not achieve Week 26 CSF-CR exhibited faster CL at that time point. Furthermore, early rapid CL was strongly associated with colectomy.

We observed a lower colectomy rate than that reported in prior prospective and retrospective pediatric cohorts with similar characteristics treated with IFX 5 mg/kg (Supplementary Table 5).^{4 13 14} The low colectomy rate observed in our study may have been related to the overall higher initial IFX doses used. Prior retrospective studies of adult and pediatric ASUC patients have supported lower colectomy rates with intensified IFX induction regimens when compared to standard induction regimens.^{13, 15} The availability of vedolizumab may also have contributed to a lower colectomy rate in our study compared to pediatric studies conducted before vedolizumab approval.

We found faster IFX clearance at Day 3 (> .02 L/h) was significantly associated with risk of colectomy. Although there was a wide confidence interval around this effect estimate due to the low number of colectomies, our finding is consistent with a retrospective study of IFX for adult ASUC in which a very similar baseline CL threshold of 0.026 L/h was predictive of colectomy.¹⁶ Despite our observed strong association of baseline clearance with colectomy, we found no signal of higher early IFX exposure with early clinical response. This suggests that, at the higher than standard dosing used in this study, poor early response was not due to low peripheral drug exposure, i.e., PK failure. Lack of an association of peripheral exposure with early outcomes does not preclude the possibility that local tissue exposure is inadequate in non-responders. In fact, one study found that the association of serum and tissue anti-TNF biologic concentrations broke down with worsening inflammation.¹⁷ Local factors within the tissue, such as high metalloproteinase activity, may limit the local concentration or binding of IFX, thus limiting drug efficacy.¹⁸ Finally, the lack of association of IFX exposure with early clinical response may be reflective of disease driven by mechanisms other than TNF, i.e. PD failure.

We observed a slowing of IFX clearance over time, consistent with treatment itself improving drug PK, likely through reducing inflammation. Nonetheless, at Week 26, non-remitters exhibited faster clearance than remitters. A retrospective adult study similarly observed a relationship between IFX CL and clinical remission.¹⁹ Consistent with the observed association of CL with Week 26 outcomes, we found non-statistically significant lower cumulative IFX exposure and estimated trough concentrations in Week 26 non-remitters compared to remitters. In fact, whereas the early signal at Day 7 and Week 8 was for higher exposure in non-responders, likely due to intensified dosing in response to poor clinical status, by Week 26 the observed relationship was reversed with higher exposure in remitters compared to non-remitters. These data raise the possibility that 10 mg/kg dosing during induction may prevent early PK failure, but as infusions spread apart during the maintenance phase, patients may be at risk for later PK failure.

We developed a novel accurate pediatric ASUC IFX PK model, which highlights the relationship of severe inflammation in ASUC with drug PK. Similar to adult models, we confirmed the strong relationships of ALB and ATI with IFX clearance in pediatric ASUC.¹⁰ Interestingly, we now identify WBCs and PLTs to predict IFX PK (CL and V1, respectively) in pediatric ASUC. A recent large cohort study found WBC and PLTs to be two baseline standard laboratory values highly associated with disease severity in pediatric UC.²⁰ Increased WBCs, specifically neutrophils, may participate in the elimination of IFX through Fcγ receptors that bind IgG antibodies and internalize the antibody, or through the local expression of proteases that degrade the monoclonal antibody.^{21, 22} In the future, this new pediatric UC IFX PK model could be incorporated into dosing dashboards, such as those developed for CD, that inform precision dosing.^{9, 23, 24}

Strengths of our study include its prospective and multicenter design, and the frequency of measured IFX concentrations, particularly early after the first infusion. Although this is the largest prospective contemporary cohort of pediatric ASUC patients treated with IFX, and the only study with extensive pharmacokinetic data, the study was underpowered to detect associations between drug exposure and clinical response. Another limitation of the observational design was the lack of dose standardization. Since most patients were treated with greater than 7 mg/kg IFX, we are unable to draw conclusions regarding how exposure relates to early response after doses less than 7 mg/kg. Although PUCAI score strongly correlates with endoscopic Mayo score, the lack of endoscopic outcomes is another limitation.⁵ This was an observational study and follow-up clinical flexible sigmoidoscopy within 26 weeks is not routine standard of care for pediatric UC. Lastly, only a randomized design could definitively determine comparative efficacy of IFX dosing regimens.

In conclusion, IFX clearance is associated with colectomy and CSF-CR in pediatric ASUC. In this contemporary cohort treated with higher than standard IFX dosing, we observed a lower colectomy rate than that reported previously. Initial 10 mg/kg dosing may be sufficient to overcome rapid drug clearance and optimize early outcomes in pediatric ASUC; however, further investigation is warranted to determine if sustained intensification of maintenance regimens can improve longer term outcomes. Standard laboratory assessments including ALB, PLTs, ATI, CRP and WBCs, in addition to weight, can be incorporated into a predictive model that accurately estimates IFX PK in pediatric ASUC.

What You Need to Know.

• Background

• The reported 1-year colectomy rate after standard dose infliximab for pediatric acute severe UC is 52%. Outcomes may be related to altered drug pharmacokinetics related to disease severity.

• Findings

• In this prospective cohort study of children treated with an average 10 mg/kg infliximab for acute severe UC, drug pharmacokinetics were not associated with early clinical response. The 1-year incidence of colectomy was 2.7% and associated with early rapid clearance.

• Implications for patient care

• The higher than standard infliximab dosing used in this study for pediatric acute severe UC may sufficiently optimize early clinical outcomes and results in lower colectomy rates than that previously reported after standard dosing.

Abbreviations:

(ASUC)

acute severe ulcerative colitis

(UC)

ulcerative colitis

(PUCAI)

pediatric ulcerative colitis activity index

(IFX)

infliximab

(TNF)

tumor necrosis factor

(PK)

pharmacokinetics

(CL)

clearance

(PD)

pharmacodynamics

(IBD-U)

inflammatory bowel disease-unclassified

(CRP)

C-reactive protein

(ATI)

antibodies to IFX

(fCal)

fecal calprotectin

(ELISA)

enzyme-linked immunosorbent assays

(AUC)

area under the IFX concentration-time curve

(CR)

colectomy-free clinical remission on IFX

(CSF-CR)

corticosteroid-free, colectomy-free clinical remission on IFX

(WBC)

white blood cell count

(PLT)

platelets

(BMI)

body mass index

(V)

volume of distribution

(AUROC)

area under the receiver operating characteristic curve

(PPV)

positive predictive value

(NPV)

negative predictive value