WO2022187657A1 - Methods of selecting sub-populations of crohn's disease patients amenable to il23 antagonist therapy - Google Patents

Abstract

The disclosure provides pharmacometric methods based on models incorporating pharmacokinetic and pharmacodynamic data to identify those inflammatory bowel disease patients, such as Crohn's disease patients, responsive to Interleukin 23 antagonist therapy by measuring the Interleukin 22 and/or C-reactive Protein biomarker levels in patient samples. Also provided are prognostic methods for refining the identification of such patients by distinguishing therapy-driven effects from placebo effects based on Crohn's Disease Activity Index changes resulting from IL23 antagonist therapy.

Description

METHODS OF SELECTING SUB-POPULATIONS OF CROHN'S DISEASE PATIENTS AMENABLE TO IL23 ANTAGONIST THERAPY

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the priority benefit under 35 U.S.C. §119(3) of U.S.

Provisional Patent Application No. 63/157,363, filed March 5, 2021, which is incorporated herein by reference in its entirety.

INCORPORATION BY REFERENCE OF MATERIAL SUBMITTED

ELECTRONICALLY

[0002] This application contains, as a separate part of the disclosure, a Sequence Listing in computer-readable form which is incorporated by reference in its entirety and identified as follows: Filename: 56042A_Seqlisting.txt; Size 11,099 bytes; Created: March 4, 2022.

FIELD

[0003] The disclosed subject matter relates generally to methods and compositions for the medical treatment of disease and more particularly to the treatment of inflammatory bowel diseases such as Crohn's disease.

BACKGROUND

[0004] Crohn's disease (CD), in addition to ulcerative colitis, is one of the chronic inflammatory bowel diseases (IBD) affecting both adults and children. It can cause inflammation in any part of the gastrointestinal tract and clinical presentations include abdominal pain, non- bloody diarrhea and the formation of fistulas, strictures and abscesses ¹. Treatment options include chemical steroids, immunosuppressives, such as azathioprine and methotrexate, and biologies such as infliximab, adalimumab, certolizumab, ustekinumab and vedolizumab, which have been shown to be effective for moderate to severe CD by targeting specific pathways implicated in CD ².

[0005] Patients start the treatment of CD using either a “bottom up” (set up) or “top down” approach, and therapy is individualized based on clinical presentation, medical history, prior treatments, severity of disease and the like ^3-5. When patients fail to respond or are intolerant to the chemical agents, those patients are switched to the alternative treatment option of biologies ^{3- 5}. However, there can be 10-30% of primary non-responders and about 30-40% of secondary non-responders that fail to achieve sustained clinical response ⁶. The cause of the non-response can be due to drug-related factors, such as neutralizing antibodies, altered clearance of drug or possibly biological escape mechanisms, or factors unrelated to the drug, including absence of active inflammation, concurrent infection or septic complications ⁶.

[0006] Brazikumab (MEDI2070, AMG139) a human monoclonal antibody (mAh) under clinical development for CD, selectively targets the pl9 subunit of interleukin-23 (IL-23), a pro- inflammatory cytokine implicated in the pathogenesis of CD^{14, 15}. IL-23 induces the downstream production of IL-22, which is significantly increased in CD and correlates with disease activity¹⁶.

[0007] C-reactive protein (CRP) is an acute phase protein produced by hepatocytes in response to inflammatory conditions, and its production is stimulated by active IBD-associated cytokines ¹⁷. The use of BCRP to predict treatment response has produced conflicting results. BCRP levels were found higher in primary non-responders than sustained responders following the treatment with infliximab ¹⁰. However, in a different study of infliximab in treatment of patients with CD, more subjects with high BCRP (>= 0.7 mg/L) showed maintained remission through 54 weeks of infliximab therapy than subjects with low BCRP (< 0.7 mg/L) (44.8% vs 22.0%, p-value = 0.012) ¹¹. Similarly, higher response to the treatment with ustekinumab was also observed with patients with high BCRP levels (>= 10 mg/L) ¹³. Besides therapeutic response, an inverse relationship between BCRP levels and placebo effect was found among patients with only placebo treatment in the studies of certolizumab and ustekinumab ^{l 2}' ¹³.

[0008] Thus, a need continues to exist in the art for methods of selecting inflammatory bowel disease patients amenable to treatment with IL23 antagonists, such as Crohn's disease patients amenable to treatment with anti-IL23 antibodies, as well as methods of treating such patients with these therapeutics.

SUMMARY

[0009] Pharmacokinetic/Pharmacodynamic (i.e., PK/PD) modeling was used to investigate the utility of PK and biomarkers in predicting the efficacy (Crohn's Disease Activity Index, CDAI) of IL23 antagonists such as the anti-IL23 monoclonal antibody brazikumab in treating Crohn's disease using CDAI as an exemplary efficacy measure. The methodology provides a data-driven framework for precision therapy for Crohn's disease. In a phase 2a trial in patients with moderate to severe Crohn's disease, treatment with brazikumab was associated with clinical improvement. Brazikumab treatment effect was determined to be dependent on the baseline IL- 22 (BIL22) or baseline C-reactive protein (BCRP) as predictive biomarkers, and placebo effect was found to be correlated with the baseline CDAI (BCDAI) as a prognostic biomarker. A maximal inhibition (I_max) on CDAI input function of 50.6% and 42.4% was predicted for patients with extremely high BIL22 or BCRP, compared to an inhibition of 20.9% and 17.8% for patients with extremely low BIL22 or BCRP, respectively, which were mainly due to the placebo effect. We demonstrated that model-derived IB₅₀ (baseline biomarker level that would achieve 50% of I_max ) of 22.8 pg/mL and 8.03 mg/L for BIL22 and BCRP as the cutoff can effectively identify sub-group subjects with higher response with improved separation of responders when compared to using the median value as the cutoff. The experiments disclosed herein exemplify the utility of pharmacometrics to quantify biomarker-driven responses in biologic therapies and distinguish between predictive and prognostic biomarkers, complementing clinical efforts to identify subpopulations with higher likelihood of response to brazikumab.

[0010] Accordingly, one aspect of the disclosure is drawn to a pharmacometric method of identifying a sub-population of Crohn's disease patients amenable to treatment with an IL23 antagonist comprising: (a) measuring a baseline level of a biomarker in a plurality of Crohn's disease patients; (b) administering an IL23 antagonist to the plurality of Crohn's disease patents;

(c) determining a baseline biomarker level associated with 50% of maximal inhibition (IB5o);and

(d) identifying the patient as amenable to treatment with an IL23 antagonist if the biomarker level of the patient is greater than the IB₅₀. In some embodiments, the patient is identified as amenable to treatment with the IL23 antagonist if the biomarker level of the patient is greater than 110% of the IB₅₀. In some embodiments, the method further comprises continued administration of the IL23 antagonist to the patient if the patient has a biomarker level greater than the IB ₅₀, such as continuing administration of the IL23 antagonist to the patient if the patient has a biomarker level greater than 110% of the IB₅₀.

[0011] In some embodiments, the IL23 antagonist is an anti-IL23 antibody. In some embodiments, the anti-IL23 antibody is brazikumab, ustekinumab, briakinumab, guselkumab, risankizumab, tildrakizumab, risankizumab, NeutraKine™ IL23 p40 antibody, BI-655066, LY- 3074828, or an antigen-binding fragment thereof. In some embodiments, the anti-IL23 antibody specifically binds to the pl9 subunit of IL23 (SEQ ID NO:l), the p40 subunit of IL23 (SEQ ID NO:2), or both subunits. In some embodiments, the anti-IL23 antibody or antigen-binding fragment thereof comprises (i) a variable region (VH) comprising or consisting of SEQ ID NO: 3 (SIN5) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 4 (SIN6), or (ii) a variable region (VH) comprising or consisting of SEQ ID NO:5 (SIN43) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 6 (SIN44). In some embodiments, the anti-IL23 antibody or antigen-binding fragment thereof comprises at least one complementarity determining region selected from SEQ ID NOS: 7-12 (SIN31-36) or SEQ ID NOS: 13-18 (SIN45-50). In some embodiments, the anti-IL23 antibody or antigen-binding fragment thereof is brazikumab. In some embodiments, the inhibition comprises a change in the magnitude of at least one symptom of Crohn's disease. In some embodiments, the inhibition is measured using the Crohn's Disease Activity Index (CDAI). In some embodiments, the biomarker levels are measured using a patient sample selected from the group consisting of whole blood, blood serum, plasma, saliva, sputum, bronchoalveolar lavage fluid, cerebrospinal fluid, pleural fluid, pericardial fluid, ascites, synovial fluid, epithelial cells, urine, stool, skin, tissue biopsy, and a combination thereof, such as wherein the sample is a blood serum sample.

[0012] In some embodiments, the biomarker is Interleukin 22 (IL22). In some embodiments, the IL22 IB₅₀ is about 22.8 pg/ml, including embodiments wherein the IL22 IB₅₀ is 22.8 pg/ml.

In some embodiments, the biomarker is C-reactive Protein (CRP). In some embodiments, the CRP IB₅₀ is about 8.03 mg/L, including embodiments wherein the CRP IB₅₀ is 8.03 mg/L. In some embodiments, the patient is switched to an alternative Crohn's disease therapy if the biomarker level of the patient is less than 110% the IB₅₀, such as wherein the patient is switched to an alternative Crohn's disease therapy if the biomarker level of the patient is less than the IB₅₀.

[0013] Another aspect of the disclosure is directed to a pharmacometric method of identifying a sub-population of Crohn's disease patients having a positive prognosis upon administration of IL23 antagonist therapy comprising: (a) measuring a baseline level of Interleukin 22 (BIL22) and/or a baseline level of C-reactive Protein (BCRP) in a plurality of Crohn's disease patients;

(b) determining a Crohn's Disease Activity Index (CDAI) score for the Crohn's disease patients;

(c) administering an IL23 antagonist to the Crohn's disease patents; (d) determining a maximal inhibition of CDAI score after administering the IL23 antagonist in the Crohn's disease patients; and (e) identifying a Crohn's disease patient as having a positive prognosis upon IL23 antagonist therapy if the percent reduction in CDAI score for the patient is greater than 20.9% for IL22 or greater than 17.8% for CRP, or both. In some embodiments, the percent reduction in CDAI score for the patient is about 50.6% for IL22 and/or about 42.4% for CRP. In some embodiments, the percent reduction in CDAI score for the patient is at least 50.6% for IL22 and/or at least 42.4% for CRP. In some embodiments, the IL23 antagonist is an anti-IL23 antibody. In some embodiments, the anti-IL23 antibody is brazikumab, ustekinumab, briakinumab, guselkumab, risankizumab, tildrakizumab, risankizumab, NeutraKine™ IL23 p40 antibody, BI-655066, LY-3074828, or an antigen-binding fragment thereof. In some embodiments, the anti-IL23 antibody specifically binds to the pl9 subunit of IL23 (SEQ ID NO:l), the p40 subunit of IL23 (SEQ ID NO:2), or both subunits. In some embodiments, the anti-IL23 antibody or antigen-binding fragment thereof comprises (i) a variable region (VH) comprising or consisting of SEQ ID NO: 3 (SIN5) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 4 (SIN6), or (ii) a variable region (VH) comprising or consisting of SEQ ID NO:5 (SIN43) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 6 (SIN44). In some embodiments, the anti-IL23 antibody or antigen- binding fragment thereof comprises at least one complementarity determining region selected from SEQ ID NOS: 7-12 (SIN31-36) or SEQ ID NOS: 13-18 (SIN45-50). In some embodiments, the anti-IL23 antibody or antigen-binding fragment thereof is brazikumab. In some embodiments, the inhibition comprises a change in the magnitude of at least one symptom of Crohn's disease. In some embodiments, the biomarker levels are measured using a patient sample selected from the group consisting of whole blood, blood serum, plasma, saliva, sputum, bronchoalveolar lavage fluid, cerebrospinal fluid, pleural fluid, pericardial fluid, ascites, synovial fluid, epithelial cells, urine, stool, skin, tissue biopsy, and a combination thereof. In some embodiments, the sample is a blood serum sample. In some embodiments, the patient is switched to an alternative Crohn's disease therapy if the percent reduction in maximal inhibition of CDAI score is no greater than 20.9% for patients with IL22 levels greater than 22.8 pg/ml. In some embodiments, the patient is switched to an alternative Crohn's disease therapy if the percent reduction in maximal inhibition of CDAI score is no greater than 17.8% for patients with CRP levels greater than 8.03 mg/L.

[0014] Other features and advantages of the disclosure will be better understood by reference to the following detailed description, including the drawing and the examples.

BRIEF DESCRIPTION OF THE DRAWING

[0015] Figure 1. PK and efficacy raw data profiles of subjects in Phase lb and 2a trials. a) PK Raw Data Profiles of Healthy Volunteers and Crohn's Disease Patients in Phase lb and 2a Trials. HV: healthy volunteers; CD: Crohn's disease patients; IV: intravenous. The PK data are stratified by the subject populations and the route of drug administration. The data of HV are all from the phase lb study with three doses in total, administered once every four weeks ( i.e ., Q4W), at a large dose range (70-700 mg). All of the data of CD patients with intravenous (IV) infusion at 210 mg and part of the data of CD patients with IV infusion at 700 mg are also from the phase lb study. All data of CD patients with subcutaneous (SC) injection at 210 mg and most of the data of CD patients with SC injection at 700 mg were from the phase 2a study, which investigated only 2 IV doses at week 0 and week 4 during the double -blinded period and 26 SC Q4W doses between week 12 and week 112 during the open-label period; b) PK Elimination Profiles of Healthy Volunteers and Crohn's Disease Patients in Phase lb Trial. CD: Crohn's disease; c) Efficacy Raw Data Profiles of Crohn's Disease Patients in Phase 2a Trial Double-blinded Period. The PD data are stratified by the treatment arms. IV: intravenous.

[0016] Figure 2. Schematic Structure of Final PK and Biomarker/Efficacy Model a) BIL22- dependent efficacy model structure b) BCRP-dependent efficacy model structure. The PK can be described using a two-compartment model and is linked to an indirect response model where the total inhibitory effect accounts for the contribution from a placebo effect and drug effect.

The drug effect, in turn, depends on the baseline level of the serum IL-22 or CRP biomarker described by a hill function are the amount of brazikumab present in the

central and peripheral compartments, respectively. Similarly, V_c and V_p are the volumes of distribution of the central and peripheral compartments, respectively. A_sc represents the amount of brazikumab administered at the subcutaneous injection site and k_a and F are the first-order absorption rate constant and bioavailability, respectively. CL is the clearance of brazikumab from the central compartment while Q is the intercompartmental clearance k_in is input function for the CDAI scores while k_out is the elimination rate. CL is dependent on the gender and disease status of the subjects and V_c is dependent on the gender of the subjects.

[0017] Figure 3. Statistical correlation between baseline Crohn's Disease Activity Index

(BCDAI) and placebo effect of patients with brazikumab treatment or placebo in the final BIL22-dependent efficacy model.

[0018] Figure 4. Visual predictive checks for the final population PK model and biomarker/efficacy models a) PK model visual predictive check; b) BIL22 -dependent PD model visual predictive check; c) BCRP-dependent PD model visual predictive check. The PK plots are stratified by study populations and the routes of drug administration. The efficacy plots are stratified by the treatment arms. [0019] Figure 5. Re-sampled Relationship between the Decrease in CDAI Scores as a Function of the Baseline level of IL-22 (a) and CRP (b) based on the Efficacy Model, a) BIL22- dependent inhibitory effect; b) BCRP-dependent inhibitory effect. The placebo effect is shown as the red ribbon (red line representing the typical placebo effect and shaded area representing the 90% confidence interval), while the total inhibitory effect from both the placebo effect and drug effect is shown as the blue ribbon (blue line representing the typical total inhibitory effect and shaded area representing the 90% confidence interval of the drug effect together with the typical placebo effect). This confidence interval is derived from 5000 simulations using the parameters in the covariance matrix output from NONMEM.

[0020] Figure 6. Statistical analysis of the difference in changes of CDAI scores at week 8 between high and low biomarker groups, a) BIL22-dependent efficacy response; b) BCRP- dependent efficacy response. The p-values are compared between the t-tests using biomarker median and model-based estimate (IB₅₀) as the cutoff for BIL22 and BCRP individually. Dash lines are the means of the changes of CDAI scores at week 8 of each biomarker subgroup.

[0021] Figure 7. Statistical correlation between CDAI score at baseline and placebo effect of patients with brazikumab treatment or placebo in the final BCRP-dependent efficacy model.

[0022] Figure 8. Shows patient disposition up to Week 24. The chart summarizes the number and disposition of enrolled and randomized patients in a phase 2a study (clinicaltrials.gov identifier: NCT01714726) evaluating MEDI2070 in patients having moderate or severe CD. mITT, Modified Intent-to-Treat Population; PP, per protocol; OL, Open Label.

[0023] Figure 9. Figure 9A shows clinical efficacy at Week 8 in the Modified Intent-to-Treat Population as indicated by CDAI clinical effect, CDAI remission, and CR100. The rate of CDAI clinical effect (defined as a CDAI score less than 150 or reduction from baseline in CDAI score of at least 100 points) at week 8 was significantly higher in the MEDI2070 group versus the placebo group (49.2% vs. 26.7%, respectively; P=0.01). CDAI remission (defined as a CDAI score less than 150) rates were 27.1% with MED 12070 and 15.0% with placebo (P=0.10). CR100 rates (defined as at least a 100-point CDAI score reduction) were 45.8% and 25.0% in the MEDI2070 and placebo groups, respectively (P=0.02). Figure 9B shows clinical efficacy at Week 8 in the Modified Intent-to-Treat Population as indicated by composite end points. A significantly greater proportion of patients receiving MED 12070 achieved the composite end points of CDAI effect and 50% reduction in FCP or CRP versus baseline (P<0.001), and CDAI remission and 50% reduction in FCP or CRP versus baseline (P=0.02). CDAI, Crohn's Disease Activity Index; Cl, confidence interval; CRP, C-reactive protein; FCP, fecal calprotectin; p, p- value.

[0024] Figure 10. Figure 10A shows efficacy at various time points through Week 24 in the Open-Label Population using nonresponder imputation as indicated by Crohn's Disease Activity Index effect rate (from baseline to week 24). Sample size of open-label period: placebo/MEDI2070 group (n=52); MEDI2070 group (n=52). Figure 10B shows efficacy at various time points through Week 24 in the open-label population using nonresponder imputation as indicated by Crohn's Disease Activity Index remission rate (from baseline to week 24). Sample size of open-label period: placebo/MEDI2070 group (n=52); MEDI2070 group (n=52).

[0025] Figure 11. Shows percent of patients with CDAI- 100 response over time by baseline IL22 levels. IL22 LO, baseline IL22<15.6 pg/mL; IL22 HI, baseline IL22>5.6 pg/mL. Cl, confidence interval. Median IL22 concentration across study population: 15.6 pg/mL. 116 patients had evaluable baseline IL22 values. Patients with baseline serum IL22 levels >15.6 pg/mL had a statistically significant increased CDAI- 100 response compared to placebo at week 8.

[0026] Figure 12. Shows percent of patients with CDAI- 100 response over time by baseline LCN2 Levels. LCN2 LO, baseline LCN2 less than 215 ng/mL; LCN2 HI, baseline LCN2 of at least 215 ng/mL. Cl, confidence interval. Median LCN2 concentration across study population: 215 ng/mL. Patients with baseline serum LCN2 levels of at least 215 ng/mL had a statistically significant increased CDAI- 100 response compared to placebo at week 8.

[0027] Figure 13. Figure 13A-C show the differential clinical response rate between subjects treated with MED 12070 compared to placebo as measured by: (Figure 13A) the difference between the percentage (%) of subjects achieving a CDAI score less than 150 or a reduction in CDAI score of greater than 100); (Figure 13B) the difference between the percentage (%) of subjects achieving a 100-point improvement in CDAI score; or (Figure 13C) the difference between the percentage (%) of subjects achieving a CDAI score less than 150 or a reduction in CDAI score greater than 100, + also achieving a greater than 50% reduction in either FCP or CRP compared to baseline FCP or CRP, respectively, between subjects treated with MED 12070 and placebo at Week 8 as a function of baseline IL22 and LCN2 levels. Clinical Response rates are reported as the delta or difference between the rate observed for MEDI2070-treated subjects and the rate observed for subjects treated with placebo ("Response Rate Differential"). For the set of baseline values of IL22 or LCN2 across the entire study population, each distribution was divided into 10 levels, or deciles, such that each of the 11 analyte cut-offs (noted as "quantile" in Figure 13A-C) progressively segmented the study population into groups with 10% less of the total study population. For example, at 4th decile (shown as the 0.4 quantile), 40% of the total study population had a baseline IF-22 or LCN2 serum level less than a particular analyte level while 60% of the total study population had a baseline IF-22 or LCN2 serum level greater than or equal to the particular analyte level. As shown in Figure 13A-C, CD patients treated with MEDI2070 having increasingly higher levels of baseline IL22 or LCN2 achieved higher clinical response rates (as measured using three different clinical response measurements) compared to placebo, illustrating that MEDI2070 induced better clinical responses in patients with high baseline IL22 or LCN2 serum levels at week 8. Subjects with high levels of IL22 or LCN2 (including, e.g., subjects with IL22 or LCN2 levels at the 5th, 6th or 7^th deciles (0.5, 0.6 or 0.7 quantiles)) had greater clinical response rate differences from placebo (irrespective of which of the three different clinical response measurements was used) compared to the IL22 or LCN2 low subjects (including, e.g., subjects with IL22 or LCN2 levels at the 1^st or 2^nd deciles (0.1 or 0.2 quantiles)). These results underscore the predictive value of high or elevated IL22 and/or LCN2 serum levels in identifying patients responsive to treatment with MEDI2070. A summary of the baseline IL22 or LCN2 serum levels corresponding to each of the deciles described in Figure 13A-C is provided in Example 6 (see Table 6 (IL22) or Table 7 (LCN2)).

[0028] Figure 14. Shows percent reduction from baseline in serum IL22 levels in MEDI2070 and placebo groups. Error bars represent median absolute deviation (MAD). Compared to treatment with placebo, treatment with MED 12070 was associated with a significantly greater drop in serum IL-22 levels by Week 8. "N" corresponds to the number of subjects at week 0.

[0029] Figure 15. Compares the differential clinical response rate (defined as the difference in the percentage (%) of subjects achieving a CDAI-100 response between treatment and placebo ("CDAI Response Delta vs. Placebo")) (y axis) and the differential clinical remission rate (defined as the difference in the percentage (%) of subjects achieving a total CDAI score of below 150 points between treatment and placebo ("CDAI Remission Delta vs. Placebo")) (x axis) for patients treated with various molecules administered for 4-10 weeks, as shown. Patients treated with MED 12070 for 8 weeks who had a baseline CRP of at least 5 mg/L (n=85); baseline IL-22 of at least 11.3 pg/mL (n=81); baseline IL-22 of at least 15.6 pg/mL (n=58); or baseline IL-22 of at least 11.3 pg/mL + CRP of at least 5 mg/L (n=62) (as measured by IL22 immunoassay) had increased CDAI-100 response rates and/or CDAI remission rates. Both the CDAI-100 response rates and the CDAI remission rates of patients treated with MED 12070 for 8 weeks who had a baseline CRP of at least 5 mg/L; baseline IL-22 of at least 11.3 pg/mL; baseline IL-22 of at least 15.6 pg/mL; or baseline IL-22 of at least 11.3 pg/mL + CRP of at least 5 mg/L (as measured by IL22 immunoassay) were greater than the reported CDAI-100 response rates and/or the CDAI remission rates for other molecules currently approved or under development to treat CD patients including: Ustekinumab (after 6 weeks or 8 weeks of treatment with a 6 mg/kg dose as reported in Ligure 1 of Sandborn et al., N Engl J Med. 2012 Oct. 18; 367(16): 1519-28.); Vedolizumab (after 6 weeks or 10 weeks of treatment as reported in Ligure 3 of Sands et. al., Gastroenterology. 2014 September; 147(3):618-627); or A dalimumab (after 4 weeks of treatment in patients who are secondary failures to infliximab as reported in Sandbom et. al, Ann Intern Med. 2007; 146:829-838). The overall clinical response and clinical remission rates for all patients treated with MEDI2070 in the Phase 2a study, irrespective of biomarker status (mITT (n=119)), was similar to the response rates of other molecules currently approved or under development. mITT, Modified Intent-to-Treat Population. Table 8 summarizes the CDAI-100 response rate differential and the CDAI remission rate differential for each of the MEDI2070-treated subgroups plotted in Ligure 15.

DETAILED DESCRIPTION

[0030] Investigation of factors that influence response or failure would help identify the most appropriate biological regimen for individual CD patients ⁶. Among all the factors, clinical or biological factors that correlate with the disease prognosis or the response to therapy have highlighted the need for further investigation. Disclosed herein are experimental investigations of the baseline Crohn's Disease Activity Index (BCDAI) and the baseline C-reactive protein (BCRP) in Crohn's disease patients ^7-13.

[0031] A phase 2a clinical study was conducted to investigate the efficacy of brazikumab, where an association between baseline serum IL-22 levels (BIL22) and the therapeutic effect of brazikumab was found: higher BIL22 levels were associated with an increased likelihood of response compared to placebo, suggesting that BIL22 may be a potential predictive biomarker to select the patients who most likely will benefit from brazikumab treatment ¹⁴. [0032] As a measure of Crohn's disease treatment efficacy, a Crohn's Disease Activity Index (CDAI) was used. CDAI is a score that combines weighted scores of clinical and laboratory variables to estimate disease severity¹⁸. It is composed of eight subjective and objective variables of disease, including diarrhea, abdominal pain, general well-being, extra-intestinal manifestations, anti-diarrheal use, abdominal mass, hematocrit, and weight¹⁸. CDAI scores of less than 150 indicate a clinical remission or inactive disease and scores over 450 indicate severely active disease. In a previous study of certolizumab pegol, patients with high BCDAI (>292) had low response in both treatment and placebo arms (p-value <0.05)⁷. In another retrospective study of ustekinumab, BCDAI was significantly lower in the subject group that achieved clinical remission than in the group that failed to achieve such remission (p-value = 0.03)⁸. Also, a higher BCDAI was associated with lack of treatment response to adalimumab, and the need for a more frequent dosing of adalimumab was discussed⁹. Overall, the underlying relationship of BCDAI and treatment response has not been quantitatively studied. A quantitative relationship between BCDAI and therapeutic effect and/or disease progression is required to demonstrate the role of BCDAI as a predictive or prognostic factor.

[0033] Although various efforts have been made to identify and validate biomarkers for treatment response and disease prognosis, the clinical usage of biomarkers has been challenging due to the lack of well-characterized, quantitative relationships between biomarker level, treatment response and disease progression. In the work disclosed herein, we used integrated population PK/biomarker/efficacy models (pharmacometrics) to investigate both the relationship between baseline biomarkers and efficacy, and the cutoff values of BIL22 and BCRP to identify patients more likely to respond to brazikumab. With this approach, we characterized the interplay between pharmacokinetics, biomarkers and efficacy of brazikumab, and we quantified the impact of the biomarkers on overall response, thus laying a data-driven framework for precision medicine in the treatment of Crohn's disease (i.e., CD).

[0034] Disclosed herein is an integrated model-based framework for analyzing the clinical data by developing a population PK model and two biomarker/efficacy models to investigate the utility of biomarkers, specifically BIL22 and BCRP, in predicting response to brazikumab. This framework allows for a quantitative description of the relationship between biomarker levels and efficacy response caused by drug effect and/or placebo effect, as well as inter-subject variability in the model parameters. This work exemplifies the value of pharmacometrics to define the cutoffs of BIL22 and BCRP for identifying patients with higher response to anti-IL23 treatment and enable biomarker-based precision medicine.

[0035] The pharmacokinetic (i.e., PK) data of brazikumab obtained from both phase lb and 2a studies were best described using a two-compartment model with first-order absorption and elimination (Figure la-b, Figure 2). The PK model showed a 43.9% slower clearance in healthy volunteers, compared with CD patients of the same gender. Fecal loss of proteins due to the disruption of the mucosal lining of the gut wall (“leaky gut”) was postulated as the major factor contributing to increased drug clearance, as was shown in patients with severe IBD undergoing infliximab treatment¹⁹. Compared to female subjects, male subjects had a faster clearance as well as a higher volume of distribution of the central compartment. The latter observation implies that body weight is another potential covariate on V_c, which was identified with slightly lower statistical significance in the screening of covariates.

[0036] The efficacy data from the phase 2a study was described in two indirect response models, of which the placebo effect was correlated with BCDAI and the drug effect was dependent on BIL22 or BCRP in a sigmoidal function (Figure lc, Figure 2). In the relationship between BIL22 and drug effect, a value of γaround 20 can be easily affected in the estimation step and thus results in a large uncertainty in the initial estimation, and it was fixed at 20 in the final biomarker/efficacy model. This large γvalue indicates a steep relationship, which approximates a dichotomous function for the biomarker-dependent drug effect, consistent with the position that, at the dose of 700 mg IV, brazikumab would lead to roughly 50.6% inhibition for patients with BIL22 level above the IB₅₀ (22.8 pg/mL), while for patients with BIL22 level below 22.8 pg/mL, brazikumab had minimal inhibition of 20.9% (Figure 5a). This is consistent with the phase 2a study showing that patients with higher BIL22 tended to have a higher remission in terms of the reduction in CDAI scores, of which the median of BIL22 (15.6 pg/mL) was used arbitrarily as the cutoff¹⁴. The t-test result with IB₅₀ as the cutoff for subgrouping the subjects showed a more significant difference in the absolute change of CDAI score at week 8 between the high and low biomarker groups, in comparison to the one with median value as the cutoff (p-value: 0.0026 vs 0.29; Figure 6a). Therefore, our model-based analysis provides a data-driven cutoff value for BIL22 that distinguishes the high-responders from the low- or non- responders to brazikumab. In the BCRP-dependent efficacy model, although the value of γin the initial estimation is moderate, it still caused large uncertainty in the estimation of the steepness of biomarker/drug effect relationship. As such, the value of γwas fixed at the initial estimate of 2.07 (Table 2). The BCRP-dependent drug effect implies that brazikumab could lead to a total inhibition between 17.8% and 42.4% depending on the individual BCRP level at the dose of 700 mg IV (Figure 5b). This relationship is consistent with the findings of ustekinumab, which also targets IL-23, where subjects with high BCRP (≥10 mg/L) had a high response to the treatment¹³. Although there is not a steep relationship between BCRP and drug effect in our model-based analysis, the t-test using IB₅₀ (8.03 mg/L) as the cutoff showed that the difference between the high and low biomarker groups had similar significance to that found between the groups when using the median (15.7 mg/L) as the cutoff (p-value: 0.009 vs 0.0084; Figure 6b). In the phase 2a study, a lower number of subjects (17 vs. 23) were in the BIL22 high biomarker group when using IB₅₀ of 22.8 pg/mL as the cutoff than when using the median of 15.6 pg/mL as the cutoff. However, more subjects (38 versus 23) were in the BCRP biomarker high group when using IB₅₀ of 8.03 mg/L as the cutoff than when using the median of 15.7 mg/L as the cutoff. Patient selection using this approach should also consider other factors , such as the risk benefit profile for the drug in the particular patient population.

[0037] By using a pharmacometrics approach, we were able to identify a negative correlation between BCDAI and placebo effect, and the latter has a typical value of 20.9% and 17.8% in the final BIL22-dependent and BCRP-dependent efficacy models, respectively (Figure 3, Figure 7, Table 2). The negative correlations between BCDAI and placebo effect are supported by the findings from several studies where patients with higher BCDAI had lower clinical response in the treatment arm and/or placebo arm^7-9. The placebo effect and the sigmoidal relationships of biomarker-dependent drug effect identified in the integrated models imply that for those patients with extremely low baseline biomarker levels of BIL22 or BCRP, the efficacy is mainly driven by the placebo effect (Figure 5). Indeed, for a subpopulation that had BIL22 or BCRP below the IB₅₀ estimates, the mean reduction of CDAI score from baseline was only 72.3 or 29.9, contrasted with 169.8 or 131.5 for the subpopulation that had BIL22 or BCRP above the IB₅₀ estimates, respectively (Figure 6).

[0038] Five challenges in IBD research, including preclinical human IBD mechanisms, environmental triggers, novel technologies, precision medicine and pragmatic clinical research, were addressed in 2019²⁰. Precision medicine, as one of the five challenges to IBD research, is designed to utilize specific clinical and biological characteristics to predict the course of disease development and treatment outcome in order to optimize clinical care²⁰. The precision medicine approach involves stratifying patients into distinct subgroups using the clinical or biological characteristics and determining the optimal treatment strategies¹⁵. Therefore, identifying such clinical or biological characteristics and determining the optimal cutoff to stratify patients can be critical in the process of drug development and clinical practice. In this work, we utilized a pharmacometrics approach to screen a series of clinical or biological characteristics, identify predictive or prognostic biomarkers to understand the relationships between these characteristics and clinical response, and predict the course of disease development and treatment outcomes. The advantages of pharmacometrics over traditional statistical analysis commonly used during clinical development provided us with a tool having greater power to detect the biomarker- dependent effect by integrating the longitudinal data over the time course while accounting for the variability of multiple sources, including drug effect and placebo effect. In addition, this tool was also able to differentiate between predictive biomarkers (BIL22 and BCRP) and prognostic biomarkers (BCDAI) by quantifying biomarker-dependent effect associated with placebo treatment and drug treatment, respectively. The model-estimated cutoffs of BIL22 and BCRP to identify patients with higher response will help in the design of subsequent phase 2b and phase 3 clinical trials with a selected subpopulation of patients and increase the power to detect the efficacy leading to a greater probability of success for the clinical development of brazikumab. This knowledge is expected to improve the identification of patient populations more likely to benefit from brazikumab.

[0039] In summary, this work demonstrates that pharmacometrics can be used to quantitatively evaluate potential predictive and prognostic biomarkers for biologic therapies for CD. Optimization of the biomarker cutoffs by pharmacometric analysis potentially will increase the power of detecting the efficacy of brazikumab, improve the probability of clinical trial success and accomplish the precision medicine objective of identifying the patients likely to benefit from treatment.

EXAMPLES

Example 1

Material and Methods

Study Population and Trial Design

[0040] The analysis data was derived from a phase lb study and a phase 2a study, which were registered at clinicaltrial.gov (NCT01258205 and NCT01714726, respectively)^{14, 15}. The phase lb study was a randomized, double-blind, placebo-controlled, sequential multiple- ascending- dose design in 34 subjects. The study was designed to evaluate the safety, tolerability, pharmacokinetics, pharmacodynamics, and effects on disease activity of brazikumab. Investigational product doses ranged from 70 mg to 700 mg; routes of administration were subcutaneous (SC) injections or intravenous (IV) infusion. Extensive sampling was conducted after the first and third dose.

[0041] The phase 2a study was a two-part study comprising a 12-week, double-blind, placebo- controlled, treatment period followed by a 100- week, open-label, treatment period to evaluate the short-term efficacy and the short- and long-term safety of brazikumab in 119 patients with moderate to severe, active CD who have failed or are intolerant to anti-TNFoc therapy. Placebo or 700 mg of brazikumab was administered as an IV infusion over a period of at least 60 minutes on day 1 and day 29 during the 12-week induction period, while subjects enrolled in the open- label period received SC injections of 210 mg brazikumab once every four weeks (i.e., Q4W) from week 12 through week 112. Sparse sampling was conducted before or after the dose during both the double-blinded and open-label periods^{14, 15}.

Modeling and Data Analysis

[0042] A nonlinear, mixed-effects modeling approach was used for all the PK and biomarker/efficacy analysis using NONMEM (ICON Development Solutions; version 7.3), PsN (Perl-speaks-NONMEM) (Lindbom et al., Comput Methods Programs Biomed. 79(3):241-57 (2005), Lindbom et al., Comput Methods Programs Biomed. 75(2):85-94 (2004)), pirana and the first-order conditional estimation method with interaction (FOCE-I). R (R Core Team (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria) was used to assemble the NONMEM data file, run simulations and create graphical outputs ^21-23.

PK Model Development

[0043] Using all PK data that is derived from the phase lb and 2a studies, including the open- label period, one- and two-compartment models with first-order absorption from the dosing site of SC administration of brazikumab and linear or nonlinear elimination from the central compartment were explored to describe the pharmacokinetics of brazikumab. Efficacy Model Development

[0044] The population efficacy model was built using collected data from the double-blind placebo-controlled period of the phase 2a trial. An indirect response (IDR) model with inhibition or stimulation was investigated as a potential model to describe the longitudinal CDAI score data. The temporal relationship of the placebo effect was explored using a linear, step or mono-exponential decay function. A sigmoidal function was used to investigate the potential biomarker level-dependent drug effect.

Model Evaluation and Selection

[0045] During the model-building process, successful minimizations, the likelihood Objective Function Value (OFV) and goodness-of-fit plots were used to evaluate each iteration of the model before progression to the next stage. These include plots of population and individual predictions versus observed data, and conditional weighted residuals versus time. Visual Predictive Checks (VPCs) were also used at the end of every key stage to compare the observed data with model-simulated concentration-time profiles and the associated medians and 95% prediction intervals.

Use of CDAI Score as PD Endpoint

[0046] The Crohn's Disease Activity Index (CDAI) is the main efficacy endpoint used in the PK/biomarker/efficacy model. CDAI is a widely used measurement to evaluate CD activity since it was developed in the late 1970s⁸. Clinicians will give a weighted score on 8 aspects of an individual patient, such as whether the patient had abdominal pain or used anti-diarrheal drugs and hematocrit and body weight and the total score - the CDAI score - indicates the severity of disease in the patient. The following cut-offs have been used: Below 150 for CD patients who are in remission, 150-219 for mild activity, 220-450 for moderate activity, and above 450 (to about 600) as severe disease. Despite some drawbacks and the development of other clinical activity indices, CDAI had been the gold standard for evaluating disease activity and most frequently used in clinical trials⁹ until recently where patient-reported outcomes are reportedly being increasingly favored by the US Food and Drug Administration (FDA) as primary endpoints in clinical trials of IBD¹⁰. PK Assay for brazikumab

[0047] PK samples were assayed using a validated ECL immunoassay method with a lower limit of quantitation (i.e., LLOQ) of 12.5 ng/mL. The quantitative range was 12.5 to 12,800 ng/mL. Briefly, study samples were added to a plate that had been passively coated with a mouse anti-brazikumab monoclonal antibody. After capture of brazikumab to the immobilized antibody, unbound materials were removed by a wash step. Ruthenium-labelled mouse anti- brazikumab monoclonal antibody was added for detection of captured brazikumab. After another wash step, a tripropylamine read buffer (MSD^®, MD) was added to the buffer to enhance the electro-chemi-luminescent signals (ECL counts). The ECL counts were directly proportional to the amount of brazikumab bound by the capture reagent. The conversion of ECL counts for study samples to concentrations was performed using Watson LIMS (v7.3.0.01, Thermo Lab Systems, Inc., PA, USA) data reduction software.

Biomarker Assay for IL22 and CRP

[0048] IL-22 samples were assayed using a quantitative ELISA-based immunoassay method with quantifiable range from 10 pg/mL-800 p/mL in 100% serum. The assay is reproducible and precise. Briefly, A mouse monoclonal antibody specific to human IL-22 was pre-coated onto a microplate. Standards, controls, and samples were added to the wells and allowed to incubate. Following a wash step, horseradish peroxidase-conjugated (i.e., HRP-conjugated) monoclonal antibody specific to human IL-22 was added to each well and allowed to incubate. IL-22 protein was bound by the immobilized antibody and the HRP-conjugated monoclonal antibody. After washing away any unbound substances, an enzyme substrate solution was added to the wells and allowed to incubate. The intensity of the color generated was directly proportional to the amount of bound IL-22. Color development was quenched by addition of an acid solution. The resulting color change was quantified with a microplate reader, which measured the optical density of the contents of each well. The optical density was used to quantify the IL-22 concentration in the test sample relative to a standard curve produced using recombinant E. coli- derived IL-22 and which was run on each plate.

[0049] Serum CRP concentration were measured using the immunoturbidimetric high sensitive assay (Roche Integra) with upper limit of normal of 3.00 mg/L. This method has been approved by the PDA for clinical use and the details of the assay method²⁴ are incorporated herein by reference. Modeling Inter-individual Variabilities

[0050] The inter-individual variabilities for PK parameters were assumed to be log-normally distributed, i.e. θ_ij = θ_{TV j} ■ e^ηi,j , where θ_ij is the j^th parameter for the i^th individual, θ_TVj is the typical value of the j^th parameter for the population, and ηi,j is assumed to follow a normal distribution of mean zero and variance of for the j^th parameter. The constant drug effect

(EFF) in the intermediate model was also assumed to be log-normally distributed. The placebo effect, however, is modeled to have an additive variability, i.e., PLAC_i = PLAC_TV + η_i, with the assumption that the placebo effect in any individual may be higher or lower than the typical value of the placebo effect by an equal probability.

[0051 ] A proportional residual error model was found to best fit the pharmacokinetic data and can be described as Cobs(t)_i = Cpred(t)_i(1+ε_i) where Cobs_i(t) and Cpred_i(t) are the observed and predicted serum concentration of the i^th individual at time t, respectively, and the residual error ε_i is assumed to follow a normal distribution with mean zero and variance of σ².

[0052] A combined model of proportional and additive residual error was found to best fit the pharmacodynamic data and can be described as Aobs(t)_i = Apred(t)_i(1+ε1i)+ε2_i , where Aobs_i(t) and Apred_i(t) are the observed and predicted CDAl scores of the i^th individual at time t, respectively, and the residual error ε1_i and ε2_i are assumed to follow a normal distribution with mean zero and variance of α1² and α2², respectively.

Example 2

Demographics and Characteristics of Study Subjects

[0053] The population PK analysis included data pooled from a total of 34 subjects from the phase lb study and 119 subjects from the phase 2a study, whereas the biomarker/efficacy analysis included only data from the phase 2b study of brazikumab. The demographics of the study population are summarized in Table 1. Briefly, among all 34 subjects (4 females and 30 males) in the phase lb study, 4 subjects had mild to severe CD, while the remaining 30 subjects were healthy volunteers (HV). For the phase 2a study, all the subjects (74 females and 45 males) had moderate to severe, active CD who failed or were intolerant to anti-TNFa therapy. The median values for BCD Al, BIL22 and BCRP at baseline were 317, 15.6 pg/mL, and 15.7 mg/L, respectively (as measured at day 0 before treatment). The other disease-related biological characteristics of the patients in the phase 2a study are summarized in Table 1.

18 Table 1. Demographic Summary of the Subjects in Phase lb and Phase 2a Trials

N: Number of subjects; %: percent of the total subjects; NA: not available; ND: not determined; FCP: Fecal Calprotectin; LCN2: Lipocalin-2; MIP3A: Macrophage inflammatory protein-3

Example 3

Population PK/PD Modeling

[0054] By combining the data from both phase lb and phase 2a studies, the PK profiles are presented depending on study populations and administration routes (Figure la). Based on the raw PK data collected in the phase lb study, we observed a linear elimination in HV within the dose range of brazikumab (70-700 mg IV). Additionally, we were able to identify that CD patients had a faster clearance (CL) than HV, as reported in other studies^14-16 (Figure lb).

[0055] The PK data were adequately described by a two-compartment model with first-order absorption from the dosing site of subcutaneous (i.e., SC) administration of brazikumab and linear CL from the central compartment within the dose range investigated (Figure la-b, Figure 2, Eq. 1-6). In the final PK model with covariate analysis, HV had a 44% lower CL compared to CD patients of the same gender (0.135 vs 0.24 L/d for female subjects). Gender had an independent effect on CL, with male subjects having a CL 23.3% higher than female subjects in the same health condition. Male subjects also had a higher volume of distribution for the central compartment (Vc) than female subjects (4.00 vs 3.23 L). All three covariate relationships were identified with statistical significance (p-value <0.01). The final PK model parameters are shown in Table 2.

Table 2: Final Population PK and Biomarker/Efficacy Model Parameter Estimates

^a Typical value for CL=0.24 x (1-0.439 x health status) x (1+0.233 x gender) x exp (η1); for health status, CD patients = 0, and healthy subjects = 1; for gender, female = 0, and male = 1 ^b Typical value for V_c=3.23 x (1+0.239 x gender) x exp (η2); for gender, female = 0, and male = 1

[0056] For SC administration:

[0057] For IV administration:

A_c and A_p are the amount of brazikumab present in the central and peripheral compartments, respectively;

V_c and V_p are the volumes of distribution of the central and peripheral compartments, respectively; Asc represents the amount of brazikumab administered at the subcutaneous injection site; k_a and F are the first-order absorption rate constant and bioavailability respectively;

CL is the clearance of brazikumab from the central compartment;

Q is the intercompartmental clearance.

[0058] Because brazikumab is a monoclonal antibody specifically binding to IL-23 and blocking the binding of IL-23 to its receptor, an indirect response (IDR) model with inhibition of k_in was chosen rather than an IDR model with stimulation of k_out, to reflect the mechanism of action and better describe the observed longitudinal CDAI scores (Figure lc, Eq. 1-4). Data from the open-label period were not included due to potential influence of unblinding on the clinical readout from the open-label period. A placebo effect was first estimated using the placebo arm data (Figure lc right). The placebo effect was found to be a constant value as the best fit over time in the double-blind period (Eq. 4).

(Eq. 1)

CDAI is the CDAI score over time; k_in is zero-order input rate; k_out is first-order elimination rate;

I _total is total inhibitory effect accounting for the placebo effect and/or the drug effect, for the data from only placebo arm, I_total is fixed as 0;

HL is the half-life to describe the remission rate.

[0059] The observed efficacy data (longitudinal CDAI scores) of the brazikumab treatment arm were included to establish a PK/efficacy relationship (Figure lc left). Because only one dose level (700 mg intravenous (i.e., IV)) was investigated in the phase 2a study, a continuous concentration-response relationship could not be established with statistical significance. Therefore, drug effect was considered independent of drug concentration but dependent on predictive biomarker levels in this phase 2a study. Statistical correlations and mathematical functions, such as sigmoidal function, were used to explore the relationships between the biomarkers of interest listed in Table 2 and drug and/or placebo effect. A minimal objective function value (OFV) of the model and pharmacologically meaningful parameter estimates were considered as the criteria to select the final model.

[0060] Among all the biological or clinical biomarkers, BIL22, BCRP and BCDAI were the three biomarkers identified based on the screening criteria. Final models included BCDAI- correlated placebo effect plus BIL22- or BCRP-dependent drug effect (Figure 2). A linear function was used to describe the relationship between BCDAI and model-estimated placebo effect (p-value < 0.01, Figure 3, Figure 7), while a sigmoidal function was used to describe the relationships between BIL22 or BCRP and drug effect (Eq. 5).

I_max is the maximum inhibition of baseline biomarker level on the CDAI production rate caused by the drug effect;

IB₅₀ is the baseline level of biomarker that would achieve 50% of l _max', γis the hill coefficient to describe the steepness of the biomarker-dependent drug effect;

CB is the baseline biomarker level of BIL22 or BCRP.

[0061] In the final BIL22- and BCRP-dependent efficacy models, the placebo effect was estimated to inhibit k_m by 20.9% and 17.8% and the correlation between BCDAI and placebo effect was -0.999 (p-value < 0.01) and -1.00 (p-value < 0.05), respectively (Table 2, Figure 3 and Figure 7). Drug effect was found to be significantly related to BIL22 (p-value < 0.01) or BCRP (p-value < 0.05). In the final BIL22-dependent efficacy model, the typical values of maximal inhibition (I_max) on k_in caused by drug effect was 29.7%, the biomarker level achieving 50% maximal inhibition {IB ₅₀) was 22.8 pg/mL and the steepness of the sigmoidal curve ( γ) was 20 (fixed in the final model due to large uncertainty in the initial model estimation). In the final BCRP-dependent PD model, the typical values of I_max, IB₅₀ and γwere 24.6%, 8.03 mg/L and 2.07 (fixed in the final model due to large uncertainty in the initial model estimation), respectively.

[0062] The final PK model and biomarker/efficacy models were evaluated based on goodness- of-fit plots. Prediction-corrected visual predictive check for population PK model was conducted and stratified based on subject populations and routes of drug administration in the phase lb and 2a trials (Figure 4a). Similarly, visual predictive checks were conducted for BIL22-dependent and BCRP-dependent efficacy models individually and stratified based on study arms in the phase 2a trial (Figure 4b-c).

[0063] The biomarker/efficacy relationship parameters in the final models, including BCDAI- dependent placebo effect and BIL22- or BCRP-dependent drug effect, were estimated with good precision, with relative standard error (RSE) of no more than 30%. Re-sampled parameters using the final model estimates demonstrate that for subjects with low BIL22 (< 22.8 pg/mL), the population mean value of the maximal inhibition (Imax) is 20.9%, much lower compared to 50.6% of I_max for subjects with high BIL22 (>= 22.8 pg/mL; Figure 5a). Similarly, the treatment of brazikumab could lead to an inhibition as low as 17.8% for subjects with low BCRP (< 8.03 mg/L), compared to 42.4% for subjects with high BCRP (>=8.03 mg/L; Figure 5b).

[0064] To investigate whether the cutoff estimated through modeling could effectively identify sub-group subjects with higher response, a t-test was conducted to compare the change of CDAI score from week 0 to week 8 between sub-group subjects (biomarker high vs. biomarker low) using model-estimated IB₅₀ or median in the phase 2a study (Figure 6)¹⁴. Using the IB₅₀, comparing the adjusted mean value between subgroups, the absolute change of CDAI scores from week 0 to week 8 shows a statistically significant difference between high and low biomarker groups for both BIL22 (high group vs low group: -169.8 (60% relative standard deviation (RSD)) vs -72.3 (137% RSD), p-value < 0.01) and BCRP (high group vs low group: - 131.5 (71% RSD) vs -29.9 (396% RSD), p-value < 0.01). When using the median, the difference between high and lower biomarker groups for the absolute change of CDAI score from week 0 to week 8 is not significant for BIL22 (high group vs low group: -123.0 (100% RSD) vs -89.2 (108% RSD), p-value = 0.29), but remains significant for BCRP (high group vs low group: - 147.7 (66% RSD) vs -69.7 (154% RSD), p-value < 0.01). The difference in drug effect between high- and low-biomarker groups for BIL22 is more significant when using model-derived IB₅₀ as the cutoff than when using the median (p = 0.0026 vs. p = 0.29). The difference in drug effect between high- and low-biomarker groups for BCRP is comparable when using IB₅₀ and the median (p = 0.009 vs. p = 0.0084) as cutoff point. This comparison demonstrated that using pharmacometrics has the potential for improving the power to detect the therapeutic effect of brazikumab by separating the high-responders from the low- or non-responders.

Example 4

Phase 2a Clinical Study Protocol

[0065] A Phase 2a study was conducted to evaluate the efficacy and safety of multiple IV doses of MED 12070 (700 mg) or placebo administered on Week 0 (Day 1) and Week 4 (Day 29) during an initial 12-week treatment period in subjects with moderate to severe CD who have failed or are intolerant to anti-TNFa therapy. The following clinical trial protocol was carried out in human patients. Results of the clinical trial are presented in Example 5.

I. Clinical Study Summary

[0066] The Phase 2a study evaluated the efficacy and safety of multiple IV doses of MEDI2070 (700 mg) or placebo administered on Week 0 (Day 1) and Week 4 (Day 29) during an initial 12-week treatment period in subjects with moderate to severe CD who failed or are intolerant to anti-TNFa therapy. A 100-week, open-label, treatment period was included to allow evaluation of the long-term safety of MED 12070 (administered at 210 mg SC Q4W) and to provide information on PK and efficacy data.

[0067] The primary objective of this study was to evaluate the efficacy of MEDI2070 versus placebo to induce a clinical effect (defined as at least a 100-point reduction in CDAI from baseline) or remission (defined as CDAK150)) at Week 8 in subjects with moderate to severe CD.

[0068] Secondary objectives of this study included evaluating: the efficacy of MEDI2070 versus placebo in achieving CDAI remission at Week 8; the effect of MEDI2070 versus placebo in achieving at least a CDAI 100-point reduction from baseline at Week 8; the effect of MEDI2070 versus placebo in achieving at least a CDAI 70-point reduction from baseline at Week 8; the effect of MEDI2070 versus placebo in achieving CDAI remission or at least a CDAI 100-point reduction from baseline at Week 12; the effect of MEDI2070 versus placebo on the change from baseline in CDAI at Week 8; the safety and tolerability of MEDI2070; and the PK and immunogenicity (IM) of MED 12070.

[0069] Exploratory objectives of this study included evaluating: the effect of MEDI2070 on other measures of clinical effect including but not limited to CDAI response, change from baseline in CDAI at other timepoints, and sustained CDAI response; the effect of treatment on inflammatory markers in blood and stool; the predictive value of blood or fecal biomarkers with respect to subject response to MED 12070; and the effect of MEDI2070 on Patient Reported Outcomes (PROs).

II. Study Design Overview

[0070] This was a two-part Phase 2a study comprising a 12-week, double-blind, placebo- controlled, treatment period followed by a 100- week, open-label, treatment period to evaluate the short-term efficacy, and the short- and long-term safety of MEDI2070 in subjects with moderate to severe, active CD who failed or are intolerant to anti-TNFa therapy as determined by the investigator. Subjects were stratified based on the number of prior anti-TNFa agents that they failed (1 versus more than 1). Subjects at various centers worldwide were randomized in a 1:1 ratio to initially receive a fixed IV dose of MED 12070 (700 mg) or placebo on Week 0 (Day 1) and Week 4 (Day 29) during the 12-week, double-blind, placebo-controlled, treatment period. At the completion of the double-blind, placebo-controlled, treatment period (Week 12), subjects had the option to enter a 100- week, open-label, treatment period where they received open-label MEDI2070 (210 mg SC) Q4W (Week 12 through Week 112) as described in Figure 8 .

[0071] MEDI2070 or placebo was administered as an IV infusion over a period of at least 60 minutes via an infusion pump on Week 0 (Day 1) and Week 4 (Day 29). Subjects who completed the 12-week, double-blind, placebo-controlled, treatment period and entered the 100- week, open-label, treatment period received MEDI2070 at a fixed dose of 210 mg SC injection Q4W for 26 doses (through Week 112). The primary analysis of the study was conducted after the last subject in the study completed the 12-week, double-blind, placebo-controlled, treatment period or was withdrawn from the study prior to completing the 12-week, double-blind, placebo- controlled, treatment period.

[0072] The doses of MED 12070 in this study were 700 mg IV Q4W in the 12-week, double- blind, placebo-controlled, treatment period (Weeks 0 and 4) and 210 mg SC Q4W in the 100- week, open-label, treatment period (Week 12 to Week 112).

[0073] Pharmacokinetic/PD modeling of MEDI2070 predicted a greater than 99% mean suppression of plasma IL23 throughout the duration of the study for both the 700 mg IV and 210 mg SC dosing regimens. Furthermore, (potency-corrected) serum concentrations of MEDI2070 at these dosing regimens were predicted to be higher than those of the anti-IL12/23p40 antibody ustekinumab shown to be efficacious in CD (Sandbom et al. Gastroenterol. 135:1130-41 (2008)). Administration of 700 mg IV was predicted to rapidly achieve steady-state target suppression.

[0074] Previous clinical data (Sandborn et al. Gastroenterol. 135:1130-41 (2008)) indicate that changes from placebo in the response parameters after administration of IV ustekinumab maximized between 6-8 weeks after start of ustekinumab administration; therefore, the proposed placebo-controlled, double-blind period of 12 weeks is sufficient to characterize the effect of MEDI2070 in this patient population. Regulatory guidance recommends assessing the primary endpoint for induction of effect between 4-8 weeks (or after 2 cycles of dosing), thus the assessment of the primary endpoint at Week 8 is appropriate. In addition, Week 12 was assessed to observe the potential extended time course of effect for MEDI2070.

[0075] All subjects received investigational product (700 mg MEDI2070 or placebo) by IV infusion over a minimum period of 60 minutes during the 12-week, double-blind, placebo- controlled, treatment period. MED 12070 (700 mg) or placebo was delivered in 5.0% w/v dextrose in a volume of 100 mL over a minimum of 60 minutes using an infusion pump. Before and after each IV infusion, the IV access was flushed with 5% w/v dextrose.

[0076] MEDI2070 was administered by subcutaneous (SC) injection during the 100-week, open-label, treatment period. Each SC dose was administered to the subject's anterior abdominal wall by an experienced and qualified staff member. Each SC injection was no more than 1.0 mL in volume (i.e., 3x1.0 mL injections for the 210 mg SC dose). As the MEDI2070 dosage volume exceeds 1.0 mL, the dose was equally divided in 3 syringes and administered as multiple SC injections on alternating (left or right) sites on the subject's anterior abdominal wall with no more than approximately 30 seconds of time between each injection and at a distance of at least 2 cm apart.

III. Disease Evaluation and Methods

[0077] CDAI: The CDAI is the oldest and most widely used of several multi-item instruments that have been developed and is validated for use in clinical studies to measure disease activity in CD (Best et al. Gastroenterology 70:439-44 (1976); Sands et al. N Engl J Med. 350(9):876-85 (2004)). The CDAI measures the severity of active disease using symptom scores that are monitored over the previous week and includes patient-reported symptoms, physician-assessed signs, and laboratory markers. The CDAI score is calculated by summing weighted scores for subjective items (number of liquid or very soft stools, abdominal pain and general well-being) recorded by a diary during a 1-week period, and objective items (associated symptoms, taking anti-diarrheals such as loperamide/opiates, abdominal mass, hematocrit, daily morning temperature, and body weight). The CDAI scores range from 0 to 600, with higher scores indicating greater disease activity. Subjects with scores less than 150, 150 to 219, and 220 to 450 represent remission, mild disease, and moderate to severe disease, respectively, whereas subjects with scores greater than 450 have very severe disease (Buxton et al. Value Health. 10:214-20 (2007)). The CDAI was calculated at the site in order to determine the eligibility for the study. For statistical analysis, CDAI for all visits was also calculated.

[0078] Patient Reported Outcomes: The study assessed patient reported outcomes (PROs) using the Inflammatory Bowel Disease Questionnaire (IBDQ). The IBDQ was completed by subjects using a paper questionnaire at study visits, scheduled at Weeks 0, 4, 8, and 12 and was administered before any other assessments. [0079] Inflammatory Bowel Disease Questionnaire (IBDQ): The IBDQ is a validated, disease- specific PRO instrument that measures health-related quality of life in patients with IBD (Guyatt et al. Gastroenterology. 96:804-10 (1989)). The IBDQ covers the following dimensions: bowel symptoms (10 items), systemic symptoms (5 items), emotional function (12 items), and social function (5 items). Items are scored on a 7 -point Likert scale, yielding a global score in the range 32 to 224 (with higher scores indicating better quality of life). The IBDQ has been frequently used in drug approval applications to assess treatment efficacy in IBD. The IBDQ was designed to be self-administered and completed in 5 minutes.

IV. Endpoints

[0080] The primary endpoint of the study was CDAI response at Week 8, defined by either a CDAI score greater than 150 or a CDAI reduction from baseline of at least 100 points. Baseline was defined as the latest non-missing observation prior to first administration of the investigational product.

[0081] Missing data were imputed by non-responder imputation approach; i.e., any subject with missing information on primary endpoint was assumed to be a non-responder. In addition, subjects with a clinically meaningful increase in steroid use were assumed to be a non-responder for the primary analysis perspective.

[0082] For the primary endpoint, comparisons between treatment arms were based on the mITT population and were performed using a logistic regression model with treatment and stratification factor as covariate. The stratification factor for this study was defined by the number of prior anti-TNFa agents a subject had received (1 vs more than 1). The significance of treatment effect was tested using the two-sided test at a=10%.

[0083] The following sensitivity analyses were performed: (1) A sensitivity analysis was performed by adjusting for baseline CDAI score and/or other baseline covariates by extending the model planned above for subjects in mITT population. (2) A mixed effect longitudinal logistic regression model was implemented on the observed responses. This logistic random- effect model includes a random intercept to account for the variability between subjects. Fixed categorical effects include stratification factor, treatment, visit, and treatment-by-visit interaction, as well as the continuous fixed covariates of baseline score. (3) A sensitivity analysis was performed by adjusting for baseline CDAI score and/or other baseline covariates by extending the model planned above for subjects in the PP Population. [0084] Secondary Endpoints included: 1) CDAI remission at Week 8, as defined by a CDAI score less than 150; 2) A reduction of at least 100 points from baseline in CDAI at Week 8; 3) A reduction of at least 70 points from baseline in CDAI at Week 8; 4) CDAI response (either remission defined by CDAI less than 150 or a CDAI reduction from baseline of at least 100 points from baseline) at Week 12. Secondary endpoints 1, 2, 3, and 4 were analyzed in a similar way to the primary endpoint. In addition, a sensitivity analysis was performed by carrying forward the last response on or before the Week 8 visit to Week 12 for subjects who had increased steroid (defined as 5 mg/day prednisolone, or equivalent, or 3 mg/day budesonide). 5) Change from baseline CDAI at Week 8. Secondary endpoint 5 was analyzed by using the inverse probability weighting generalized estimating equations method adjusting for prior anti- TNFa use. Also a sensitivity analysis was performed using an ANCOVA model after missing data were imputed using the Last Observation Carried Forward (LOCF) approach through the end of the double-blind, placebo-controlled, treatment period adjusting for prior anti-TNFa use. 6) To evaluate the safety and tolerability of MED 12070, the safety and tolerability endpoints included AEs including SAEs, significant changes in laboratory values and vital signs. All safety-related endpoints for the 12-week, double-blind, placebo-controlled, treatment period were reported based on safety population using the actual treatment received. Subjects randomized to placebo with at least one dose of MEDI2070 were included in the active arm; also the subjects randomized to the MED 12070 arm without any active dose were included in the placebo arm. The Medical Dictionary for Regulatory Activities (MedDRA) was used to code all AEs. Treatment-emergent AEs were defined as any AE with onset on or after the administration of the first dose of investigational product up to and including 36 weeks post-last dose. 7) To evaluate the PK and immunogenicity (IM) of Multiple Doses of MED 12070, descriptive statistics of serum MED 12070 concentration data were provided by visit. Individual and mean serum concentration-time profiles of MEDI2070 were generated. For PK data analysis, time zero was defined as the beginning of infusion. The presence of anti-drug antibodies (ADAs) to MEDI2070 in serum was also assessed. Example 5

Phase 2a Clinical Study Results

I. Study Design and Patients

[0085] The phase 2a study (clinicaltrials.gov identifier: NCT01714726) described in Example 4 included a 12- week, double-blind, placebo-controlled treatment period followed by a 100- week open-label treatment period to evaluate short-term efficacy and short- and long-term safety of MEDI2070 in patients with active moderate to severe Crohn's disease who failed prior anti- TNFa therapy. Conducted at 60 centers worldwide, the study included adults aged 18 to 65 years diagnose with ileal, ileo-colonic, or colonic Crohn's disease for at least 6 months.

[0086] Patients had to have moderate to severe active Crohn's disease (defined as a Crohn's Disease Activity Index [CDAI] score of 220-450), with evidence of active inflammation (baseline C-reactive protein [CRP] mg/L, fecal calprotectin [FCP] 250 pg/g, or endoscopic evidence of inflammation [photographic documentation of at least 3 nonanastomotic ulcerations, each more than 0.5 cm in diameter, or 10 aphthous ulcerations involving at least 10 cm of contiguous intestine within 12 weeks before screening). Patients must have received at least one anti-TNFa therapy at approved doses for Crohn's disease, with primary nonresponse (signs and symptoms of active disease, despite at least 1 induction regimen, including at least 2 doses of anti-TNF-a therapy at least 2 weeks apart) or secondary nonresponse (recurrence of symptoms of persistently active disease during maintenance anti-TNFa therapy following initial clinical benefit) or intolerance (including but not limited to infusion-related reaction, demyelination, congestive heart failure, or infection). Exclusion criteria included having an allogeneic bone marrow transplant or cell-based transplantation; short-bowel syndrome; obstructive stricture within 3 months of study; bowel surgery within 12 weeks of study; ileostomy or colostomy; a clinically significant concomitant disease; or prior treatment with a biologic agent targeting IL12 or IL23. Washout of other prior therapies was required, including infliximab for 8 weeks before study, adalimumab or certolizumab for 10 weeks, natalizumab for 12 weeks; cyclosporine, mycophenolate mofetil, sirolimus, thalidomide, or tacrolimus for 4 weeks; intravenous or intramuscular corticosteroids for 2 weeks; or topical mesalamine or rectal corticosteroids for 2 weeks. Concomitant use of 5-aminosalicylates and glucocorticosteroids was permitted if doses were stable for at least 2 weeks before randomization and remained stable through week 8. Similarly, concomitant use of immunomodulators, e.g., azathioprine, was permitted if the dose remained stable from 8 weeks before randomization through week 8. Use of oral antibiotics for Crohn's disease, probiotics, and antidiarrheals were allowed.

[0087] All patients provided written informed consent. The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines and each site's local institutional review board approved the protocol.

II. Treatment

[0088] For the double-blind period, an interactive voice or web-based response system was used for randomization to treatment arms, assigning unique randomization codes to each patient. Randomization was stratified based on the number of prior failed anti-TNFα agents (1 vs more than 1). Using blinded randomization with concealed allocation based on a permutation block algorithm, patients were randomized 1:1 within each stratum to receive MEDI2070700 mg or placebo intravenously over 60 minutes at week 0 (day 1) and week 4 (day 29). Patients, investigators, and the sponsor were blinded to treatment until the last patient reached week 12, when the primary analysis was conducted. During the open-label period (weeks 12 through 112), all patients received MEDI2070210 mg subcutaneously every 4 weeks (26 doses over 100 weeks).

III. Study Assessments

[0089] The primary outcome measure was the proportion of patients achieving clinical effect, defined as at least a 100-point CDAI score reduction from baseline or CDAI score less than 150 at week 8. Secondary measures included the proportion of patients achieving CDAI remission (defined as CDAI score lower than 150) at week 8, at least a 100-point CDAI score reduction (CR100) from baseline at week 8, at least a 70-point CDAI score reduction (CR70) from baseline at week 8, clinical effect at week 12, CDAI remission at week 12, and the safety and tolerability of MEDI2070, including adverse events, serious adverse events, and significant changes in laboratory values and vital signs. Serum MEDI2070 concentrations and the presence of ADAs were evaluated at baseline, week 4, week 8, week 12, week 24, and end of study.

[0090] Exploratory outcome measures included inflammatory markers in the blood and stool (CRP and FCP), responses of other biomarkers and their predictive value for clinical effects, sustained clinical effect between weeks 8 and 24, and sustained CDAI remission assessed at weeks 8 and 24. Biomarker assessments included changes in IL22 serum levels (expressed in high levels in Crohn's disease patients and a marker of disease activity [Schmechel et al., Inflamm Bowel Dis, 14:204-212 (2008)]) from baseline after treatment with MEDI2070 versus placebo at weeks 8 and 12; association of change from baseline in IL22 levels with clinical effect and CDAI remission after treatment with MEDI2070 versus placebo at weeks 8 and 12; and assessment of higher IL22 serum levels at baseline as a predictor of clinical effect and CDAI remission after treatment with MEDI2070 compared with placebo at weeks 8 and 12.

IV. Statistical Analysis

[0091] Assuming a CDAI clinical effect rate of 20% in the placebo group, approximately 54 patients per treatment arm were required to provide 87% power to detect a 25% difference in CDAI clinical effect rates at week 8 between MEDI2070700 mg and placebo, using a two-sided test at significance level of a=0.1. Assuming an approximately 10% dropout rate per treatment arm adjustment, approximately 60 patients were to be randomized.

[0092] The modified intent-to-treat population included all randomized patients who received at least one dose of study treatment during the double-blind period. The safety population comprised patients who received any study treatment during the double-blind period. The per- protocol population included patients who received all treatment doses and had no major protocol deviations. The open-label population included all patients who received at least 1 dose of MEDI2070 during the open-label period.

[0093] For the primary and secondary measures with binary outcomes, comparison between treatment arms was based on the modified intent-to-treat population, using a logistic regression model [Ge et al., Drug Inf J, 45:481-493 (2011)], with treatment and stratification factor (number of prior anti-TNFa agents of 1 vs >1) as covariates. The significance of the treatment effect was tested using the 2-sided test at a=10%. A sensitivity analysis was conducted by extending the planned logistic regression model to adjust for baseline CDAI score and/or other baseline covariates. Missing data for dichotomous end points were imputed using non-responder imputation. A patient imputed to be a non-responder before week 8 was considered a non- responder for all subsequent visits. Additionally, patients with a clinically meaningful increase in steroid use were assumed to be nonresponders. Clinically meaningful increase in steroid dose was defined as an increase of at least 5 mg/day for at least 3 days of prednisone or equivalent, or an increase of at least 3 mg/day for at least 3 days of budesonide. Missing data for continuous measures were handled using the inverse probability weighting generalized estimating equations method, adjusting for prior anti-TNFa agent use. [0094] Pharmacokinetic data were summarized. Exploratory analyses were performed for patients in the open-label period. Efficacy data from the double-blind and open-label periods up to week 24 for these patients were combined and reported by treatment arms in the double-blind period. Change from baseline in CRP and FCP were analyzed using a mixed-effects repeated measures model, adjusting for prior anti-TNFa use and baseline, for the open-label population. Comparisons between MED 12070210 mg at week 24 and MED 12070700 mg or placebo at week 12 were performed.

V. Patient Results

[0095] In total, 121 patients were randomized; 119 of these received double-blind treatment (Figure 8). In the double-blind period, 52 of 59 (86.7%) and 52 of 60 (85.2%) patients completed week 12 in the MEDI2070 and placebo groups, respectively. Baseline patient characteristics generally were balanced between treatment groups (Table 3), except CDAI and CRP levels were numerically higher in the MEDI2070 group. Results from weeks 8, 12, and 24 are reported here; the study is ongoing for long-term follow-up.

Table 3

^*Three patients had missing fecal calprotectin assessments. VI. Efficacy

( i ) Week 8

[0096] For the primary outcome measure, the rate of clinical effect at week 8 was significantly higher in the MEDI2070 group versus the placebo group (49.2% vs. 26.7%, respectively;

P=0.01; point estimate: 0.225 [90% confidence interval (Cl): 0.083 to 0.368]; Figure 9A). CDAI remission rates were 27.1% with MED 12070 and 15.0% with placebo (P=0.10; point estimate: 0.122 [90% Cl: 0.000 to 0.243]). CR70 rates were 52.5% and 46.7% in the MEDI2070 and placebo groups, respectively (P=0.52), and CR100 rates were 45.8% and 25.0% in the MEDI2070 and placebo groups, respectively (P=0.02). A significantly greater proportion of patients receiving MEDI2070 achieved the composite end points of CDAI effect and 50% reduction in FCP or CRP versus baseline (P<0.001), and CDAI remission and 50% reduction in FCP or CRP versus baseline (P=0.02; Figure 9B). MEDI2070 treatment also resulted in significantly greater reductions in FCP versus placebo (least squares mean change -153.5 vs. - 49.7; least squares mean difference - 105.6 [90% Cl: -184.0 to -27.2]; P=0.027) and significantly reduced CRP levels (least squares mean change: - 12.6 vs. 5.1 with placebo; least squares mean difference - 17.6 [90% Cl: -28.3 to -6.9]; P=0.007).

(ii) Week 12

[0097] At week 12, the rate of clinical effect was greater with MEDI2070 (37.3% vs 28.3%; P=0.29), as was the rate of CDAI remission (20.3% vs. 13.3%; P=0.31); however, the differences were not statistically significant. Rates of CR100 at week 12 were 37.3% with MED 12070 and 28.3% with placebo (P=0.288). At week 12, significant reductions in FCP and CRP with MEDI2070 were maintained (FCP: least squares mean change -179.6 vs -55.1; least squares mean difference -124.6 [90% Cl: -221.2 to -27.9]; P=0.034; CRP: least squares mean change: - 13.4 vs -2.6 with placebo; least squares mean difference -10.8 [90% Cl: - 17.6 to -4.1]; P=0.008).

(iii) Week 24

[0098] Clinical effect and CDAI remission rates were maintained in the MED 12070 group at 24 weeks. The proportion of patients receiving placebo then open-label MEDI2070 (placebo/MED 12070 group) who achieved clinical effect and CDAI remission was similar to those receiving MEDI2070700 mg then open-label MEDI2070 (MED 12070/MEDI2070) at week 24 (Figures 10A and 10B). Over the 24- week period, the MEDI2070/MEDI2070 group continued to achieve the composite end points of CDAI effect plus 50% reduction in FCP or CRP versus baseline, and CDAI remission plus 50% reduction in FCP or CRP versus baseline (Table 4). The proportion of patients in the placebo/MED 12070 group achieving both composite end points at week 24 was similar to those in the MEDI2070/MEDI2070 group. In the MEDI2070/MED 12070 group, changes from baseline in FCP and CRP levels were maintained between weeks 12 and 24; in the placebo/MED 12070 group, decreases in FCP and CRP levels were observed between weeks 12 and 24.

Table 4

VII. MEDI2070 Exposure-Response Relationship

[0099] At the 700-mg intravenous dose of MEDI2070, no definitive relationship between exposure and efficacy was established. Serum MEDI2070 concentrations at the week 4 trough and at weeks 8 and 12 varied over a tenfold range and were not related to CDAI response or nonresponse at week 8. These findings are consistent with dosing at the plateau of the dose- response curve.

VTTT. Biomarkers

[0100] Patients in the MEDI2070 group had greater reductions in serum IL22 levels versus those in the placebo/MEDI2070 group (Figure 14). Baseline serum IL22 levels greater than or equal to the median value of 15.6 pg/mL were associated with an increased likelihood of clinical effect in the MEDI2070 group. Patients in the MEDI2070 group with baseline IL22 levels less than 15.6 pg/mL had a CR100 response rate similar to that in the placebo group (Figure 11). Clinical response was not a strong function of IL22 in the placebo group.

IX. Safety

(i) Week 12

[0101] At the end of the double-blind period, the proportion of patients with treatment- emergent adverse events was similar between the MED 12070 and placebo groups (67.8% vs. 68.3%, respectively), as were the proportion of patients with grade 3 or greater adverse events (10.2% vs. 11.7%, respectively) and serious adverse events (8.5% vs. 8.3%, respectively). Treatment-related adverse events were observed in 13.6% of patients receiving MEDI2070 and 21.7% of patients receiving placebo; those observed in at least 5% of patients in either group are shown in Table 5. Serious adverse events occurred in five patients in the MED 12070 group and five in the placebo group; events included Crohn's disease (three events in two patients), colonoscopy-associated colon perforation (n=l), pyrexia (n=l), and cellulitis (n=l) in the MEDI2070 group, and anemia (n=l), Crohn's disease (two events in two patients), diarrhea (n=l), gastrointestinal hemorrhage (n=l), and abdominal abscess (n=l) in the placebo group. Table 5

[0102] Infections that were serious or at least grade 3 in severity, or that required oral or parenteral antimicrobial therapy, occurred in four patients with four events in the MED 12070 group, and in seven patients with 11 events in the placebo group. Treatment discontinuation attributable to treatment-emergent adverse events occurred in 8.5% of MEDI2070 patients (including gastrointestinal disorders [6.8%] and infection [1.7%]) and in 10.0% of placebo patients (including gastrointestinal disorders [6.7%], infection [1.7%], and eye disorder [1.7%]).

(ii) Week 24

[0103] At week 24, treatment-emergent adverse events were observed in 67.3% of patients in the MED 12070 group and 65.4% of those in the placebo/MED 12070 group, with treatment- related adverse events observed in 25.0% and 21.2% of patients, respectively. Treatment- emergent adverse events of at least grade 3 severity occurred in 13.5% of patients in the MEDI2070 group and 3.8% of patients in the placebo/MEDI2070 group; serious adverse events were observed in 15.4% and 7.7% of patients, respectively. Treatment discontinuation attributable to treatment-emergent adverse events occurred in 9.6% of MEDI2070 patients (including Crohn's disease flare [3.8%], anal fistula [1.9%], anemia and lymphopenia [1.9%], and anal abscess [1.9%]), and in 3.8% of placebo/MED 12070 patients (including pelvic abscess [1.9%] and kidney stones [1.9%]). The number of infection events that were serious, at least grade 3 in severity, or required oral or parenteral antimicrobial therapy were equal in the MEDI2070 and placebo/MED 12070 groups (13 events).

X. Immunogenicity

[0104] Anti-drug antibodies were detected in two of 119 patients. One patient receiving MEDI2070 had anti-drug antibodies at baseline, but not in subsequent assessments. The other patient, who received placebo during the double -blind period, had anti-drug antibodies at week 24.

XI. Discussion

[0105] In patients with moderate to severe Crohn's disease who failed prior anti-TNFa therapy, MED 12070 resulted in a significantly greater rate of clinical effect, a composite of a 100-point reduction in CDAI and CDAI remission, at 8 weeks compared with placebo (49.2% vs. 26.7%, respectively; P=0.01), meeting the primary study end point. Clinical efficacy was consistent with the biologic effects observed with MEDI2070. Patients in the MEDI2070 group had greater decreases from baseline in FCP and CRP relative to placebo. The beneficial effects of MEDI2070 on FCP and CRP are remarkable given that this patient population was previously heavily treated with anti-TNF-a therapies; more than 65% of patients had received two or more prior anti-TNFa agents.

[0106] At week 12, the rate of clinical effect was greater in the MED 12070 group, as was the rate of CDAI remission; however, these differences were not significant. Significant reductions in FCP and CRP with MEDI2070 were maintained at week 12. At week 24, clinical effect, CDAI remission, and changes from baseline in levels of FCP and CRP were maintained between weeks 12 and 24 in the MEDI2070 group. Among patients in the placebo/MED 12070 group, the proportion achieving clinical effect and CDAI remission at week 24 was similar to those who received MED 12070 in both periods, and significant decreases in FCP and CRP levels were observed between weeks 12 and 24. At the 700-mg intravenous dose of MED 12070, no definitive relationship between exposure and efficacy was established. This finding is consistent with dosing at the plateau of the dose-response curve.

[0107] MEDI2070 was well tolerated, with rates of treatment-emergent adverse events, grade 3 or greater adverse events, serious adverse events, and discontinuations owing to treatment- emergent adverse events similar to those of placebo at 12 weeks. Rates of serious or severe infections were less than 10% in the MED 12070 group at week 12. At week 24, overall rates of treatment-emergent adverse events were similar between the MEDI2070 and placebo/MEDI2070 groups; rates of grade 3 or greater adverse events and serious adverse events were numerically greater in the MEDI2070 group versus the placebo/MEDI2070 group (13.5% vs. 3.8% and 15.4% vs. 7.7%, respectively), as were rates of discontinuations attributable to treatment- emergent adverse events (9.6% vs. 3.8%, respectively). Rates of serious or severe infections were the same in both treatment groups at 24 weeks (13 events). One patient had anti-drug antibodies at baseline; presumably, this was a false-positive result and was not detected on follow-up assessment.

[0108] Although specific inclusion criteria, primary end points, and timing of assessments vary, the results of our study suggest that the short-term CDAI remission rate achieved with MEDI2070 (27.1% at 8 weeks) generally compares favorably with those of other biologic therapies evaluated in patients with Crohn's disease who failed prior anti-TNFa therapy. These include ustekinumab (26% [7/27] at 8 weeks) (Sandbom et al. Gastroenterology 135: 1130-41 (2008)), vedolizumab (10.5% [n=105] at 6 weeks), (Sandbom et al. N Engl J Med 369:711-21 (2013)) and adalimumab (21% [34/159 with secondary nonresponse to infliximab] at 4 weeks) (Sandborn et al. Ann Intern Med 146:829-38 (2007)). The safety profile of MEDI2070 also compares favorably with other biologies available for patients with Crohn's disease. A recent meta-analysis of studies evaluating the dual IL23/IL12 inhibitors ustekinumab and briakinumab found an increased risk of major adverse cardiovascular events in psoriasis patients treated for 12 to 20 weeks [Tzellos et al., J Eur Acad Dermatol Venereol, 27:1586-1587 (2013)]. No patients in our study experienced a major adverse cardiovascular event up to 24 weeks of follow-up. MEDI2070 is specific for IL23 and does not inhibit IL12.

[0109] We evaluated serum IL22 levels at baseline and following MEDI2070 treatment. IL22, expressed at high levels in Crohn's disease, is an effector cytokine that supports mucosal barrier integrity and is an indicator of IL23 axis activity [Schmechel et al., Inflamm Bowel Dis, 14:204- 212 (2008)]. Serum IL22 levels were reduced in the MEDI2070 group compared with the placebo group. Additionally, baseline serum IL22 levels greater than or equal to 15.6 pg/mL were associated with an increased likelihood of clinical effect in the MEDI2070 group, whereas MEDI2070-treated patients with baseline IL22 levels less than 15.6 pg/mL had CR100 responses similar to those in the placebo group. Figure 11, In contrast, in a study by Dige et al., IL22 levels were not reduced in Crohn's disease patients effectively treated with adalimumab [Dige et al., J Crohns Colitis, 7:248-255 (2013)]. This study is among the first to incorporate novel biomarkers to further the understanding of the pathogenesis and improve the treatment of Crohn's disease.

[0110] In conclusion, MED 12070 treatment demonstrated consistently robust efficacy, with an acceptable safety profile in patients with Crohn's disease who failed prior anti-TNFa therapies.

Example 6

Identification ofIL23 Pathway Biomarkers (IL22 and/or LCN2) as Predicted Biomarkers for Treatment of IL23 -Mediated Diseases with an Anti-IL23 Antibody

[0111] IL23 is expressed primarily from activated dendritic cells and macrophages (see Gaffen et al (2014) Nature Revs Immunol 14: 585-600; Oppmann et al (2000) Immunity 13: 715-251) and acts directly on a variety of hematopoietic cell types including Thl7, Th22, yo T cells and innate lymphoid cells (ILCs) to induce cytokines including IL22, IL21, IL17A, IL17F, IF17A/F, TNF alpha and GM-CSF (see, e.g., Gaffen et al (2014) Nature Revs Immunol 14:585-600; Zheng et al (2007) Nature 445: 648-51; El-Behi et al (2011) Nature Immunol 12: 568-575). These effector and regulatory cytokines can in turn act on a variety other cell types expressing the appropriate cognate receptors. IL23-induced IL22, for example, can stimulate IL22-receptor expressing epithelial cells and keratinocytes to secrete antimicrobial proteins such as LCN2 (Sonnenberg et al (2010) Adv Immunol 107: 1-29; Stallhofer et al (2015) Inflamm Bowel Dis 2015 Aug. 7; Behnsen et al (2014) Immunity 40:262-73).

[0112] In the Phase 2a study described in Examples 4 and 5, CD patients with elevated baseline serum IL22 or LCN2 levels of greater than or equal to 15.6 pg/mL or 215 ng/mL, respectively (as measured using the immunoassays described in Example 3), had an increased likelihood of clinical effect in the MED 12070 group, whereas MEDI2070-treated patients with baseline IL22 levels less than 15.6 pg/mL or 215 ng/mL, respectively, had CR100 responses similar to those in the placebo group. Figures 11 and 12. In particular, patients with baseline serum IL22 levels 15.6 pg/mL or patients with baseline serum LCN2 levels 215 ng/mL were observed to have statistically significant increased CDAI-100 responses when treated with MEDI2070 compared to placebo at week 8. Figures 11 and 12.

[0113] To further understand the relationship between baseline serum IL22 and/or LCN2 levels and response to treatment with MED 12070, the set of baseline values of either IL22 or LCN2 across the entire study population was divided into 10 levels, or deciles, such that each of the 11 analyte cut-offs progressively segmented the study population into groups with 10% less of the total study population. The differential clinical response rate between MED 12070 and placebo exposed subjects as a function of baseline IL22 and LCN2 serum levels at each decile cut-off is provided in Figure 13A-C and the individual IL22 and LCN2 serum decile values are summarized below in Table 6 (IL22) and Table 7 (LCN2). For example, as reported in Table 6, at the 4^th decile, 40% of the total study population had a baseline IL-22 level of less than 12.7 pg/mF and 60% of the total study population had a baseline IF-22 level of at least 12.7 pg/mF. The CDAI response rate at week 8 (as measured by the percentage (%) of subjects achieving a CDAI score less than 150 or a reduction in CDAI score of greater than 100) in those subjects exposed to MEDI2070 and with baseline IF-22 levels of at least 12.7 pg/mF, i.e., above the 4th decile, was 58.3%. 21 subjects exposed to MEDI2070 and with baseline IL22 levels above the 4^th decile for the study population (12.7 pg/ml) were found to be CDAI responders at week 8. The function in R called 'quantiles' was used to determine decile values. The numbers of CDAI responder and non-responders in the MEDI2070 and placebo-exposed groups, and the CDAI response rates at week 8 of the study in subjects with baseline IL22 or LCN2 values greater than or equal to each decile cut are also indicated in Tables 6 and 7, and differences between treatment and placebo response rates for each decile cut are provided in Figure 13A. Two additional measurements of clinical response, i.e., the difference between the percentage (%) of subjects treated with MEDI2070 versus those treated with placebo achieving a 100-point improvement in CDAI score at week 8 (Figure 13B), and the difference between the percentage (%) of subjects treated with MED 12070 versus those treated with placebo achieving a CDAI response (CDAI score less than 150 or a reduction in CDAI score of greater than 100) + also achieving a greater than 50% reduction in either FCP or CRP compared to baseline FCP or CRP, respectively, at week 8 (Figure 13C) as a function of baseline IL22 and LCN2 levels/deciles described in Table 6 or 7 were also performed.

[0114] As shown in Figure 13A-C and reported in Table 6 below, CD patients treated with MEDI2070 having increasingly higher levels of baseline IL22 achieved higher CDAI response rates at week 8 compared to placebo (as measured using any of the three different clinical response measurements shown in Figure 13A-C), illustrating that MED 12070 induced better clinical responses in patients with high baseline IL22 serum levels. Notably, subjects with high levels of IL22 (including, e.g., subjects with IL22 levels at the 5th, 6th or 7^th deciles (0.5, 0.6 or 0.7 quantiles)) had greater clinical response rate differences from placebo (irrespective of which of the three different clinical response measurements was used) compared to the IL22 low subjects (including, e.g., subjects with IL22 levels at the 1st or 2^nd deciles (0.1 or 0.2 quantiles)). See Figure 13A-C. These IL22-high subjects also had increased CDAI response rates compared to all comers treated with MED 12070 (see, e.g., Figure 15).

[0115] Importantly, the CDAI response rates and CDAI remission rates observed in IL22-high subjects treated with MEDI2070 in the Phase 2a study are amongst the highest clinical response rates to biologies therapy for CD reported to date. For example, as shown in Figure 15, the CDAI- 100 response rate differential (defined as the difference in the percentage (%) of subjects achieving a CDAI- 100 response between treatment and placebo) and/or CDAI remission rate differential (defined as the difference in the percentage (%) of subjects achieving a reduction in total CDAI score to below 150 points between treatment and placebo) achieved in patients having elevated baseline serum IL22 treated with MED 12070 for 8 weeks were highly increased compared to the published CDAI- 100 response and/or CDAI remission rates of patients treated with a number of other compounds currently approved or under development to treat CD including: Ustekinumab (response rates after 6 weeks or 8 weeks of treatment with a 6 mg/kg dose as reported in Figure 1 of Sandbom et al., N Engl J Med. 2012 Oct. 18; 367(16): 1519-28.); Vedolizumab (response rates after 6 weeks or 10 weeks of treatment as reported in Figure 3 of Sands et. al., Gastroenterology. 2014 September; 147 (3): 618-627); or Adalimumab (response rates after 4 weeks of treatment in patients who are secondary failures to infliximab as reported in Sandborn et. al, Ann Intern Med. 2007; 146:829-838). For example, both the CDAI-100 response rate differential ("CDAI Response Delta vs. Placebo") and the CDAI remission rate differential ("CDAI Remission Delta vs. Placebo") achieved in patients treated with MED 12070 for 8 weeks who had a baseline CRP of at least 5 mg/L; baseline IL-22 of at least 11.3 pg/mL; baseline IL-22 of at least 15.6 pg/mL; or baseline IL-22 of at least 11.3 pg/mL + CRP of at least 5 mg/L (as measured by IL22 immunoassay) were greater than the reported CDAI-100 response rate differential and/or the CDAI remission rate differential for Ustekinumab, Vedolizumab and Adalimumab reported in Figure 15. The overall clinical response and remission rates for all patients treated with MEDI2070 in the Phase 2a study, irrespective of biomarker status, was similar to the response rates of other molecules currently approved or under development. Table 8 summarizes the CDAI-100 response rate differential and the CDAI remission rate differential for each of the MEDI2070-treated subgroups plotted in Figure 15. These results further underscore the surprising and unexpected predictive value of high or elevated IL22 serum levels (alone or in combination with other biomarkers disclosed herein) in identifying patients having an IL23 -mediated disease or disorder responsive to treatment with an IL23 antagonist (including, e.g., an anti-IL23 antibody or fragment thereof such as MEDI2070).

[0116] Similarly, as shown in Figure 13A-C and reported in Table 7, CD patients treated with

MEDI2070 having increasingly higher levels of baseline LCN2 achieved higher clinical response rates (as measured using any of the three different clinical response measurements shown in Figure 13A-C) at week 8 compared to placebo, supporting the position that MEDI2070 induced better clinical responses in patients with high baseline LCN2 serum levels. Notably, LCN2-high subjects (including, e.g., subjects with LCN2 levels at the 5th 6th or 7th deciles (0.5, 0.6 or 0.7 quantiles)) had greater clinical response rate differences from placebo (irrespective of which of the three different clinical response measurements was used) compared to the LCN2 low subjects (including, e.g., subjects with LCN2 levels at the 1st or 2^nd deciles (0.1 or 0.2 quantiles)). These results further demonstrate the surprising and unexpected predictive value of high or elevated LCN2 serum levels in identifying patients having an IL23-mediated disease or disorder responsive to treatment with an IL23 antagonist (including, e.g., an anti-IL23 antibody or fragment thereof such as MED 12070).

[0117] The relationship between increasing clinical response rates and increasing baseline biomarker levels was not always as evident in patients in the 8^th, 9^th, or 10^th deciles. At these IL22 or LCN2 levels, too few, if any, patients treated with MED 12070 or placebo were available for analysis. See, e.g., Figure 13A-C; Tables 6 and 7. Given a larger sample size, however, increased clinical response rates for patients identified as having IL22 or LCN2 levels at the 8th,

[0118] 9th or 10th deciles are expected.

[0119] Taken together, these results support that high IL22 serum levels and/or high LCN2 serum levels (including, e.g., the median baseline IL22 and/or LCN2 serum levels identified in the study or serum IL22 levels between 7.9 pg/mL and 31.4 pg/mL and/or serum LCN2 levels between 142.8 ng/mL and 261.1 ng/mL) can be used to identify a patient having an IL23- mediated disease or disorder suitable for treatment with an IL23 antagonist (including, e.g., an anti-IL23 antibody or fragment thereof such as MED 12070). Table 6

Table 7

Table 8

[0120] As noted previously, IL23-induced IL22 induces cells to secrete LCN2. Thus, the observation that baseline serum levels of two separate IL23 pathway members ( i.e ., elevated IL22 or LCN2) each were predictive of patient clinical response to MEDI2070 (e.g., CDAI-100 Response Rate at week 8) strongly suggests that other IL23 pathway biomarkers may also predict an increased likelihood of clinical effect in response to treatment with MEDI2070. Accordingly, to the extent that serum baseline levels of IL22 and/or LCN2 or any other IL23 pathway analyte (including, e.g., CCL20, IL17F, IL17A/F, IL23R, IL12B, IL6, IL21, TNF, CCR6, CCL22,

IL1R1, IFN-y, S100A12, DEFB-2, DEFB-4, IF1, SERPINB3, PI3/Elafin, LL37, RORy, RORyT, IL26, S100A7, DEFB103B, or GM-CSF) reflects increased IL23 axis activity, a patient determined to have increased IL23 pathway activity (as determined by measuring one or more IL23 pathway biomarkers) is expected to be more likely to benefit from treatment with an IL23 antagonist (including, e.g., an anti-IL23 antibody or an antigen-binding fragment thereof such as MEDI2070).

REFERENCES

1. Harlan, W. R.; Meyer, A.; Fisher, J., Inflammatory Bowel Disease: Epidemiology, Evaluation, Treatment, and Health Maintenance. N C Med J 2016, 77 (3), 198-201. 2. Neurath, M. F., Cytokines in inflammatory bowel disease. Nat Rev Immunol 2014, 14 (5), 329-42.

3. Mould, D.; Upton, R.; Wojciechowski, J., Pharmacokinetics-Based Dosing Strategies for Therapeutic Proteins in Inflammatory Bowel Disease. In Quantitative Pharmacology and Individualized Therapy Strategies in Development of Therapeutic Proteins for Immune-Mediated Inflammatory Diseases, Zhou, H.; Mould, D., Eds. 2019.

4. Dignass, A.; Van Assche, G.; Lindsay, J. O.; Lemann, M.; Soderholm, J.; Colombel, J. F.; Danese, S.; D'Hoore, A.; Gassull, M.; Gomollon, F.; Hommes, D. W.; Michetti, P.; O'Morain, C.; Oresland, T.; Windsor, A.; Stange, E. F.; Travis, S. P.; (ECCO), E. C. s. a. C.

O., The second European evidence-based Consensus on the diagnosis and management of Crohn's disease: Current management. J Crohns Colitis 2010, 4 (1), 28-62.

5. Travis, S. P.; Stange, E. F.; Lemann, M.; Oresland, T.; Chowers, Y.; Forbes, A.; D'Haens, G.; Kitis, G.; Cortot, A.; Prantera, C.; Marteau, P.; Colombel, J. F.; Gionchetti, P.; Bouhnik, Y.; Tiret, E.; Kroesen, J.; Starlinger, M.; Mortensen, N. J.; Organisation, E. C. s. a. C., European evidence based consensus on the diagnosis and management of Crohn's disease: current management. Gut 2006, 55 Suppl 1, i 16-35.

6. D'Haens, G. R.; Panaccione, R.; Higgins, P. D.; Vermeire, S.; Gassull, M.; Chowers, Y.; Hanauer, S. B.; Herfarth, H.; Hommes, D. W.; Kamm, M.; Lofberg, R.; Quary, A.;

Sands, B.; Sood, A.; Watermeyer, G.; Watermayer, G.; Lashner, B.; Lemann, M.; Plevy, S.; Reinisch, W.; Schreiber, S.; Siegel, C.; Targan, S.; Watanabe, M.; Feagan, B.; Sandbom, W. J.; Colombel, J. F.; Travis, S., The London Position Statement of the World Congress of Gastroenterology on Biological Therapy for IBD with the European Crohn's and Colitis Organization: when to start, when to stop, which drug to choose, and how to predict response? Am J Gastroenterol 2011, 106 (2), 199-212; quiz 213.

7. Hanauer, S. B.; Panes, J.; Colombel, J. F.; Bloomfield, R.; Schreiber, S.; Sandbom, W. J., Clinical trial: impact of prior infliximab therapy on the clinical response to certolizumab pegol maintenance therapy for Crohn's disease. Aliment Pharmacol Ther 2010, 32 (3), 384-93.

8. Miyazaki, T.; Watanabe, K.; Kojima, K.; Koshiba, R.; Fujimoto, K.; Sato, T.; Kawai, M.; Kamikozuru, K.; Yokoyama, Y.; Hida, N.; Nakamura, S., Efficacies and Related Issues of Ustekinumab in Japanese Patients with Crohn's Disease: A Preliminary Study. Digestion 2020, 101 (1), 53-59.

9. Sandbom, W. J.; Colombel, J. F.; Schreiber, S.; Plevy, S. E.; Pollack, P. F.; Robinson, A. M.; Chao, J.; Mulani, P., Dosage adjustment during long-term adalimumah treatment for Crohn's disease: clinical efficacy and pharmacoeconomics. Inflamm Bowel Dis 2011, 17 (1), 141-51.

10. Magro, F.; Rodrigues-Pinto, E.; Santos-Antunes, J.; Vilas-Boas, F.; Lopes, S.; Nunes, A.; Camila-Dias, C.; Macedo, G., High C-reactive protein in Crohn's disease patients predicts nonresponse to infliximab treatment. J Crohns Colitis 2014, 8 (2), 129-36.

11. Reinisch, W.; Wang, Y.; Oddens, B. J.; Link, R., C-reactive protein, an indicator for maintained response or remission to infliximab in patients with Crohn's disease: a post-hoc analysis from ACCENT I. Alimentary Pharmacology & Therapeutics 2012, 35 (5), 568-576.

12. Schreiber, S.; Rutgeerts, P.; Fedorak, R. N.; Khaliq-Kareemi, M.; Kamm, M. A.; Boivin, M.; Bernstein, C. N.; Staun, M.; Thomsen, O. 0.; Innes, A., A Randomized, Placebo- Controlled Trial of Certolizumab Pegol (CDP870) for Treatment of Crohn's Disease. Gastroenterology 2005, 129 (3), 807-818.

13. Toedter, G. P.; Blank, M.; Lang, Y.; Chen, D.; Sandbom, W. J.; de Villiers, W. J. S., Relationship of C-Reactive Protein With Clinical Response After Therapy With Ustekinumab in Crohn's Disease. American Journal of Gastroenterology 2009, 104 (11), 2768-2773. 14. Sands, B. E.; Chen, J.; Feagan, B. G.; Penney, M.; Rees, W. A.; Danese, S.; Higgins, P. D. R.; Newbold, P.; Faggioni, R.; Patra, K.; Fi, J.; Klekotka, P.; Morehouse, C.;

Pulkstenis, E.; Drappa, J.; van der Merwe, R.; Gasser, R. A., Efficacy and Safety of MEDI2070, an Antibody Against Interleukin 23, in Patients With Moderate to Severe Crohn's Disease: A Phase 2a Study. Gastroenterology 2017, 153 (1), 77-86.e6.

15. Ma, C.; Jairath, V.; Khanna, R.; Feagan, B. G., Investigational drugs in phase I and phase II clinical trials targeting interleukin 23 (IL23) for the treatment of Crohn's disease. Expert Opin Investig Drugs 2018, 1-12.

16. Schmechel, S.; Konrad, A.; Diegelmann, J.; Glas, J.; Wetzke, M.; Paschos, E.; Lohse, P.; Goke, B.; Brand, S., Finking genetic susceptibility to Crohn's disease with Thl7 cell function: IL-22 serum levels are increased in Crohn's disease and correlate with disease activity and IL23R genotype status. Inflammatory Bowel Diseases 2007, 14 (2), 204-212.

17. Iskandar, H. N.; Ciorba, M. A., Biomarkers in inflammatory bowel disease: current practices and recent advances. Transl Res 2012, 159 (4), 313-25.

18. Best, W. R.; Becktel, J. M.; Singleton, J. W.; Kern, F., Jr., Development of a Crohn's disease activity index. National Cooperative Crohn's Disease Study. Gastroenterology 1976, 70 (3), 439-44.

19. Brandse, J. F.; van den Brink, G. R.; Wildenberg, M. E.; van der Kleij, D.; Rispens, T.; Jansen, J. M.; Mathot, R. A.; Ponsioen, C. Y.; Löwenberg, M.; D'Haens, G. R., Foss of Infliximab Into Feces Is Associated With Fack of Response to Therapy in Patients With Severe Ulcerative Colitis. Gastroenterology 2015, 149 (2), 350-5. e2.

20. Denson, F. A.; Curran, M.; McGovern, D. P. B.; Koltun, W. A.; Duerr, R. H.; Kim, S. C.; Sartor, R. B.; Sylvester, F. A.; Abraham, C.; de Zoeten, E. F.; Siegel, C. A.; Burns, R. M.; Dobes, A. M.; Shtraizent, N.; Honig, G.; Heller, C. A.; Hurtado-Forenzo, A.; Cho, J. H., Challenges in IBD Research: Precision Medicine. Inflamm Bowel Dis 2019, 25 (Suppl 2), S31- S39.

21 Keizer, R. J.; Karlsson, M. O.; Hooker, A., Modeling and Simulation Workbench for NONMEM: Tutorial on Pirana, PsN, and Xpose. CPT Pharmacometrics Syst Pharmacol 2013,

2, e50.

22. Jonsson, E. N.; Karlsson, M. O., Xpose— an S-PFUS based population pharmacokinetic/pharmacodynamic model building aid for NONMEM. Comput Methods Programs Biomed 1999, 58 (1), 51-64.

23. Keizer, R. J.; van Benten, M.; Beijnen, J. H.; Schellens, J. H.; Huitema, A. D., Pirana and PCluster: a modeling environment and cluster infrastructure for NONMEM. Comput Methods Programs Biomed 2011, 101 (1), 72-9.

24. Folekha, P. H.; Chittamma, A.; Roberts, W. L.; Sritara, P.; Cheepudomwit, S.; Suriyawongpaisal, P., Comparative study of two automated high-sensitivity C-reactive protein methods in a large population. Clin Biochem 2005, 38 (1), 31-5

[0121] All publications and patents mentioned in the application are herein incorporated by reference in their entireties or in relevant part, as would be apparent from context. As but one example, the description of experiments relating to clinical trials lb and 2a provided in US Pat. Pub. No. 2018/0252728 are incorporated herein by reference. Various modifications and variations of the disclosed subject matter will be apparent to those skilled in the art without departing from the scope and spirit of the disclosure. Although the disclosure has been described in connection with specific embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Various modifications of the described modes for making or using the disclosed subject matter that are obvious to those skilled in the relevant field(s) are intended to be within the scope of the following claims.

Claims

Claims What is Claimed is:

1 . A pharmacometric method of identifying a sub-population of Crohn's disease patients amenable to treatment with an IL23 antagonist comprising:

(a) measuring a baseline level of a biomarker in a plurality of Crohn's disease patients;

(b) administering an IL23 antagonist to the plurality of Crohn's disease patients;

(c) determining a baseline biomarker level associated with 50% of maximal inhibition ( IB₅₀); and

(d) identifying the patient as amenable to treatment with an IL23 antagonist if the biomarker level of the patient is greater than the IB₅₀.

2. The method of claim 1 wherein the patient is identified as amenable to treatment with the IL23 antagonist if the biomarker level of the patient is greater than 110% of the IB₅₀.

3. The method of claim 1 further comprising continued administration of the IL23 antagonist to the patient if the patient has a biomarker level greater than the IB₅₀.

4. The method of claim 2 further comprising continued administration of the IL23 antagonist to the patient if the patient has a biomarker level greater than 110% of the IB₅₀.

5. The method of claim 1 wherein the IL23 antagonist is an anti-IL23 antibody.

6. The method of claim 5 wherein the anti-IL23 antibody is brazikumab, ustekinumab, briakinumab, guselkumab, risankizumab, tildrakizumab, risankizumab, NeutraKine™ IL23 p40 antibody, BI-655066, LY-3074828, or an antigen-binding fragment thereof.

7. The method of claim 5 wherein the anti-IL23 antibody specifically binds to the p19 subunit of IL23 (SEQ ID NO:1 ), the p40 subunit of IL23 (SEQ ID NO:2), or both subunits.

8. The method of claim 5 wherein the anti-IL23 antibody or antigen-binding fragment thereof comprises (i) a variable region (VH) comprising or consisting of SEQ ID NO: 3 (SIN5) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 4 (SIN6), or (ii) a variable region (VH) comprising or consisting of SEQ ID NO:5 (SIN43) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 6 (SIN44).

9. The method of claim 5 wherein the anti-IL23 antibody or antigen-binding fragment thereof comprises at least one complementarity determining region selected from SEQ ID NOS: 7-12 (SIN31-36) or SEQ ID NOS:13-18 (SIN45-50).

10. The method of claim 5 wherein the anti-IL23 antibody or antigen-binding fragment thereof is brazikumab.

11 . The method of claim 1 wherein the inhibition comprises a change in the magnitude of at least one symptom of Crohn's disease.

12. The method of claim 1 wherein the inhibition is measured using the Crohn's Disease Activity Index (CDAI).

13. The method of claim 1 wherein the biomarker levels are measured using a patient sample selected from the group consisting of whole blood, blood serum, plasma, saliva, sputum, bronchoalveolar lavage fluid, cerebrospinal fluid, pleural fluid, pericardial fluid, ascites, synovial fluid, epithelial cells, urine, stool, skin, tissue biopsy, and a combination thereof.

14. The method of claim 13 wherein the sample is a blood serum sample.

15. The method of claim 1 wherein the biomarker is Interleukin 22 (IL22).

16. The method of claim 15 wherein the IL22 IB₅₀ is about 22.8 pg/ml.

17. The method of claim 16 wherein the IL22 IB₅₀ is 22.8 pg/ml.

18. The method of claim 1 wherein the biomarker is C-reactive Protein (CRP).

19. The method of claim 18 wherein the CRP IB₅₀ is about 8.03 mg/L.

20. The method of claim 18 wherein the CRP IB₅₀ is 8.03 mg/L.

21 . The method of claim 1 wherein the patient is switched to an alternative Crohn's disease therapy if the biomarker level of the patient is less than 110% the IB₅₀.

22. The method of claim 2 wherein the patient is switched to an alternative Crohn's disease therapy if the biomarker level of the patient is less than the IB₅₀.

23. A pharmacometric method of identifying a sub-population of Crohn's disease patients having a positive prognosis upon administration of IL23 antagonist therapy comprising: (a) measuring a baseline level of Interleukin 22 (BIL22) and/or a baseline level of C-reactive Protein (BCRP) in a plurality of Crohn's disease patients;

(b) determining a Crohn's Disease Activity Index (CDAI) score for the Crohn's disease patients;

(c) administering an IL23 antagonist to the Crohn's disease patients;

(d) determining a maximal inhibition of CDAI score after administering the IL23 antagonist in the Crohn's disease patients; and

(e) identifying a Crohn's disease patient as having a positive prognosis upon IL23 antagonist therapy if the percent reduction in CDAI score for the patient is greater than 20.9% for IL22 or greater than 17.8% for CRP, or both.

24. The method of claim 23 wherein the percent reduction in CDAI score for the patient is about 50.6% for IL22 and/or about 42.4% for CRP.

25. The method of claim 23 wherein the percent reduction in CDAI score for the patient is at least 50.6% for IL22 and/or at least 42.4% for CRP.

26. The method of claim 23 wherein the IL23 antagonist is an anti-IL23 antibody.

27. The method of claim 26 wherein the anti-IL23 antibody is brazikumab, ustekinumab, briakinumab, guselkumab, risankizumab, tildrakizumab, risankizumab, NeutraKine™ IL23 p40 antibody, BI-655066, LY-3074828, or an antigen-binding fragment thereof.

28. The method of claim 26 wherein the anti-IL23 antibody specifically binds to the p19 subunit of IL23 (SEQ ID NO:1 ), the p40 subunit of IL23 (SEQ ID NO:2), or both subunits.

29. The method of claim 26 wherein the anti-IL23 antibody or antigen-binding fragment thereof comprises (i) a variable region (VH) comprising or consisting of SEQ ID NO: 3 (SIN5) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 4 (SIN6), or (ii) a variable region (VH) comprising or consisting of SEQ ID NO:5 (SIN43) and/or a light chain variable region (VL) comprising or consisting of SEQ ID NO: 6 (SIN44).

30. The method of claim 26 wherein the anti-IL23 antibody or antigen-binding fragment thereof comprises at least one complementarity determining region selected from SEQ ID NOS: 7-12 (SIN31-36) or SEQ ID NOS:13-18 (SIN45-50).

31 . The method of claim 26 wherein the anti-IL23 antibody or antigen-binding fragment thereof is brazikumab.

32. The method of claim 23 wherein the inhibition comprises a change in the magnitude of at least one symptom of Crohn's disease.

33. The method of claim 23 wherein the biomarker levels are measured using a patient sample selected from the group consisting of whole blood, blood serum, plasma, saliva, sputum, bronchoalveolar lavage fluid, cerebrospinal fluid, pleural fluid, pericardial fluid, ascites, synovial fluid, epithelial cells, urine, stool, skin, tissue biopsy, and a combination thereof.

34. The method of claim 33 wherein the sample is a blood serum sample.

35. The method of claim 23 wherein the patient is switched to an alternative Crohn's disease therapy if the percent reduction in maximal inhibition of CDAI score is no greater than 20.9% for patients with IL22 levels greater than 22.8 pg/ml.

36. The method of claim 23 wherein the patient is switched to an alternative Crohn's disease therapy if the percent reduction in maximal inhibition of CDAI score is no greater than 17.8% for patients with CRP levels greater than 8.03 mg/L.