22-09-08 TNF antagonists: ADA and PK Theo Rispens Immunogenicity therapeutic antibodies Infliximab Mouse Chimeric 30% mouse Immunogenicity: Adalimumab Humanized 3-5% mouse Human measurement of ADA Antigen binding test (ABT) prot. A - capture IgG from serum (Prot. A Sepharose) - wash out unbound material - incubate with radiolabeled adalimumab F(ab’)2 - wash out unbound label - measure radioactivity Prevent background of anti-hinge antibodies IVIG F(ab’)2 IFX F(ab’)2 anti-IFX protein A Sepharose Rispens et al., JIM 2011 anti-hinge Rispens et al., JIM 2011 Development of Antidrug Antibodies Against Adalimumab and Association With Disease Activity and Treatment Failure During Long-term Follow-up Bartelds et al., JAMA April 2011 Long term clinical effects of ADA 272 adalimumab treated RA patients 3 year follow-up Long term measurement of adalimumab levels, ADA and disease activity. Bartelds et al., JAMA April 2011 baseline characteristics Total Patient with AAA Patients without AAA n=272 n=76 n=196 Age, years 54 ± 12 53 ± 13 54 ± 11 Female, no. (%) 219 (81) 62 (82) 157 (80) RF, no. (%) 196 (72) 57 (75) 139 (71) Prior DMARDs 3.1 ± 1.4 3.4 ± 1.5* 3.0 ± 1.3* MTX use, no. (%) 202 (74) 41 (54)* 161 (82)* MTX dose (mg/wk) 25 (15‒25) 18 (10‒25)* 25 (15‒25)* No DMARD, no. (%) 51 (19) 28 (37)* 23 (12)* Disease duration (years) 8 (3‒17) 12 (5‒18)* 8 (3-16)* Erosive disease, no. (%) 201 (74) 63 (83)* 138 (70)* ESR (mm/h) 23 (11‒42) 35 (18‒60)* 21 (11‒39)* CRP (mg/L) 12 (5‒29) 19 (7‒46)* 11 (4‒22)* DAS28 5.2 ± 1.2 5.5 ± 1.1* 5.1 ± 1.3* Bartelds GM, et al. JAMA. 2011;305:1460‒1468 Adalimumab: Cohort 272 RA patienten During 156 weeks of follow-up anti-adalimumab antibodies were detected in 76 (28%) of patients 67% of ADA+ patients is positive already after 28 weeks 30 % ADA+ 25 20 15 Antibodies against adalimumab (%) 10 5 0 0 28 56 Bartelds et al., JAMA April 2011 84 112 140 Remission (DAS28 < 2.6) is associated with absence of anti-drug antibodies Remission probability 0.5 AAA - 0.4 0.3 p<0.0001 0.2 0.1 0.0 AAA+ 0 50 100 150 200 Time in weeks Gecorrigeerd voor MTX dosis, bezinking en CRP (HR:3.6; 95%CI:1.8-7.2, p<0.0001) Bartelds et al., JAMA April 2011 Anti-drug antibody formation to adalimumab and infliximab measured with ABT is correlated with diminished clinical efficacy Drug interference Assay comparison: Drug interference - Discrepancies found between assays: - In IgG4-specific assay, higher frequency of positive patients compared to similar IgG-specific assay - Therefore, systematic comparison: - 216 RA patients - sera after 28 weeks of treatment - Comparison between: - bridging elisa - IgG-specific antigen binding test - IgG4-specific antigen binding test Anti-adalimumab measured in bridging ELISA and antigen binding test different numbers of ADA+ patients in different assays prot. A 7% Hart et al., JIM 2011 13% α-IgG4 22% Relation between drug levels and ADA bridging elisa: detects ADA only if no adalimumab is detected prot. A α-IgG4 Hart et al., JIM 2011 Inhibition by adalimumab - serum with high anti-adalimumab levels - inhibition by adding increasing amounts of adalimumab Hart et al., JIM 2011 Rationalization of drug interference prot. A α-IgG4 prot. A Natalizumab: IgG4 • Treatment of Multiple Sclerosis • Directed against a4-integrin (VLA-4) • Interferes with lymphocyte migration across blood-brain barrier • IgG4 antibody: Fab arm exchange anti-lambda natalizumab x serum IgG4 VLA-4 Labrijn et al. Nature Biotechnol. 2009 Antibodies to natalizumab at 12, 24, and 52 weeks Vennegoor et al., Mult. Sclerosis 2013 Drug interference: mono vs bispecific IgG4 125I natalizumab F(ab’)2 ADA protein A Sepharose Rispens et al. J Pharm Biomed Anal 2013 Drug interference differs between assay formats Moreover, drug interference may depend differentially on drug characteristics Dynamics of ADA formation to adalimumab and drug-tolerant assays Measurement of ADA in complex with drug - Can we increase recovery by acid treatment? - Prevention of re-association upon neutralization: block adalimumab with rabbit anti-idiotype F(ab) adalimumab F(ab’)2 anti-adalimumab prot. A Ph-shift anti-Idiotype ABT (PIA) adalimumab patient anti-adalimumab acid rabbit anti-adalimumab add excess rabbit anti-ada Fab neutralize prot. A Can we increase recovery using acid treatment and blocking agent? Model system: Rabbit anti-idiotype ADA ADA + acid treatment ADA + drug + F(ab) + acid t. ADA + drug + F(ab) ADA + drug • Add adalimumab to make complexes • Acid to dissociate complexes • Excess of rabbit anti-idiotype F(ab) to keep complexes dissociated. Recovery is more efficient with a combination of acid treatment and rabbit anti-idiotype F(ab) Schouwenburg et al., JIM 2010 prot. A Long term measurement with PIA - 3 year follow up for 99 adalimumab-treated patients 47% of the patients develop ADA >90% of ADA+ patients develop ADA in the first 6 months van Schouwenburg et al. ARD 2013 Results gradients Prot A adalimumab, no ADA (in ABT) adalimumab, low ADA (in ABT) Most sera contain complexes Complexes are small These complexes are not rapidly cleared no adalimumab, high ADA (in ABT) Tolerance Long term follow-up with PIA allows research on transient ADA production Of the 53 patients positive for ADA at any time point, 17 are transiently positive for ADA. van Schouwenburg et al. ARD 2013 • Drug tolerant ADA assays using acid dissociation • >90% of ADA+ patients develop ADA in the first 6 months • Responses are diverse and include ‘transient’ and ‘persistent’ cases • Drug tolerant ADA assays may provide immunological insight, but offer limited clinical insight Characterization of anti-adalimumab antiodies Making human monoclonal antibodies Schouwenburg et al., ARD 2012 All monoclonal antibodies are derived from different precursor B-cells data unpublished, available upon request (t.rispens@sanquin.nl) Mabs underwent extensive somatic hypermutation A VH gene usage Vκ gene usage VH CDR3 length data unpublished, available upon request (t.rispens@sanquin.nl) FR CDR FR CDR All mAbs bind to an overlapping epitope ADA-bt ADA ADL data unpublished, available upon request (t.rispens@sanquin.nl) Competition of adalimumab binding: ADA from patient sera A single monoclonal anti-adalimumab inhibits binding of all serum antiadalimumab (50 pt tested so far) + IVIG F(ab) + anti-Id F(ab) 1.2 Schouwenburg et al., Ann Rheum Dis. 2012 mAbs neutralize adalimumab TNF IL-8 TNF + ADL TNF + ADL + ADA no IL-8 IL-8 data unpublished, available upon request (t.rispens@sanquin.nl) TNF + ADL + ADA Summary I - Antibody response to adalimumab: - antigen-driven: - extensive somatic hypermutation and affinity maturation - broad: - broad range of unrelated clones - restricted: - >98% of antibody response targets overlapping epitopes Neutralization by ADA Adalimumab: TNF inhibition 125I adalimumab Fab Adalimumab data unpublished, available upon request (t.rispens@sanquin.nl) Infliximab: potential epitopes TNF VH VL infliximab non-human germ-line non-mouse germ-line adapted from Liang et al. JBC 2013 Infliximab: TNF inhibition 125I infliximab Fab Infliximab data unpublished, available upon request (t.rispens@sanquin.nl) certolizumab: TNF inhibition 125I certolizumab Certolizumab data unpublished, available upon request (t.rispens@sanquin.nl) correlation WEHI assay – anti-infliximab RIA data unpublished, available upon request (t.rispens@sanquin.nl) • ADA to anti-TNF antibodies are predominantly neutralizing: > 98% for adalimumab (16 pt) > 90% for infliximab (25 pt) > 97% for certolizumab (9 pt) • paratope is immunodominant? • No need for bioassays assessing neutralizing capacity Drug level testing PK assay Anti-idiotype adalimumab TNF Anti-TNF Serum adalimumab concentrations no AAA Median adalimumab concentration (MG/L) AAA 13-100 AU/ml AAA >100 AU/ml 14 12 10 * 8 6 4 2 0 t=0 t=16wk Bartelds et al., JAMA April 2011 t=40wk t=78wk t=130wk p<0.0001* Drug levels vs anti-drug antibodies: a balance 16 or 28 weeks after start adalimumab ABT vs PK PIA vs PK Detection of anti-drug-antibodies (ADA) Anti-TNF antibody levels ADA production Free anti-TNF agent ADA-drug complexes Free ADA ADA detection method ELISA – – – + ABT – – +/– + PIA – +/– + + ++ + +/– – Pharmacokinetic assay (TNF capture) van Schouwenburg PA, et al. Nat Rev Rheumatol. 2013;9:164‒172 Concentration-effect curve (adalimumab/RA) Each dot is mean of 20 patients Pouw et al. ARD 2013 in press The next step towards personalized medicine Smart testing for health benefits and cost-effective therapy. biologicals@sanquin.nl | http://www.sanquin.nl/en/products-services/biologicals/ August 7, 2014 | 52 Drug levels (PK) • Functional adalimumab levels: balance between drug levels and ADA levels • functional drug levels are correlated with clinical efficacy • Some non-responders with normal/high drug levels • Optimal concentrations of adalimumab for RA are ca. 5-8 mg/L Take home • neutralization leads to non-response • Quantity of ADA is relevant: balance ADA/drug • ADA assays are difficult to standardize • Drug-tolerant assays not useful for clinic • Drug level testing can be useful for clinical management Acknowledgements Sanquin Research Pauline van Schouwenburg Margreet Hart Reade Gertjan Wolbink Simone Kruithof Margret de Koning Els de Groot Charlotte Krieckaert Marieke van Ham Margret Bartelds Mieke Pouw Gertjan Wolbink Mike Nurmohamed Diana Wouters Lucien Aarden Karin van Schie Sanquin Diagnostics Services Genmab Desiree van der Kleij Henk de Vrieze Rob de Jong Astrid van Leeuwen Steven Stapel Esther van Buren Tom Vink