RSABE - BEBAC • Consultancy Services for Bioequivalence

Reference-Scaled Average Bioequivalence (NTIDs)NTIDs might be problematicsteep/flat PK/PD-curvesresp. × 2response × 20πεχεπ Pharma Edge10 100concentr. × 2Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 20133 • 21

Reference-Scaled Average Bioequivalence (NTIDs)NTIDs (FDA)NTIDs from ANDAs reviewed by FDA/OGDwithin 1996 – 2008 (89 studies)πεχεπ Pharma EdgeDrugWarfarinLevothyroxineCarbamazepineLithium carbonateDigoxinPhenytoinTheophyllineStudies29915165123Mean5.79.38.07.821.79.217.9AUC 0-tRange MeanC maxRange3.3 – 11.0 12.7 7.7 – 20.13.8 – 15.5 9.6 5.2 – 18.64.4 – 19.4 8.7 5.2 – 17.64.5 – 14.0 13.5 6.4 – 24.413.1 – 32.24.1 – 18.612.8 – 24.221.014.918.2Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201314.3 – 26.17.4 – 20.011.8 – 25.8LX YuApproaches to Demonstrate Bioequivalence Critical Dose DrugsAdvisory Committee for Pharmaceutical Science and Clinical Pharmacology, April 13, 2010http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/AdvisoryCommitteeforPharmaceuticalScienceandClinicalPharmacology/UCM209319.pdf4 • 21

Reference-Scaled Average Bioequivalence (NTIDs)NTIDs (FDA)πεχεπ Pharma EdgeFor NTIDs 20% fluctuation in plasmaconcentrations might be clinically relevantNTIDs often have low variability; CIs of twogenerics might be 85–90% and 115–120%.Switchability? Potential Approaches:AUC: PE ⊂ 90–111%AUC: PE ⊂ 95–105%AUC: CI ⊂ 90–111% (like EMA)AUC: CI ⊂ 90–111% and includes 100% (like Denmark)AUC: CI ⊂ 95–105%Reference Scaled Average Bioequivalence (RSABE)Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 20135 • 21

Reference-Scaled Average Bioequivalence (NTIDs)NTIDs (FDA)Percentage of ANDAs passing tighter criteria(89 studies)MethodCI includes 100%CI ⊂ 90–111%CI ⊂ 90–111% and includes 100%PE ⊂ 90–111%RSABEAUC 0-tC max84.3 69.786.5 60.777.5 50.6100.0 95.5not assessedπεχεπ Pharma EdgeTighter AR ensures smaller differences in mean BADifferences in variability between products are notaddressedRSABE suggestedLX Yu 2010Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 20136 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Statistical modelπεχεπ Pharma EdgeFully replicated TRTR | RTRT designABE model−θ ≤ µ − µ ≤ + θA T R ASABE modelµT− µR−θS≤ ≤ + θSσW Regulatory regulatory switching condition θ based onregulatory constant σ 00.1 and ∆ 1.11111 (=1/0.9, theupper BE limit)2⎛ ln ∆ ⎞θ ≡ ⎜ ⎟⎝ σ0 ⎠Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 20137 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Evaluationπεχεπ Pharma EdgeSABEMixed effects model (SAS Proc MIXED, PhoenixLinear Mixed Effects).Determine 95% upper confidence limit for( ) 2 2Y −Y −θ⋅ sT R WRby Howe’s method (like in SABE for HVDPs).Bioequivalent if 95% upper CL ≤0.ABEMixed effects model.Bioequivalent if 90% CI ⊂ 80.00–125.00%.Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 20138 • 21

Reference-Scaled Average Bioequivalence (NTIDs)EvaluationComparison of σ WT with σ WRMixed effects model of intra-subject contrast T 1–T 2and R 1–R 2by sequence.Comparison based on s WTand s WR(the estimates ofσ WTand σ WR).s WRis already available from SABE (R 1–R 2); similarsetup for T 1–T 2to obtain s WT.Determine 90% confidence interval of σ WT/σ WRass s s sWT WR WT WR,FF( ν , ν ) 1 −α( ν , ν )α2 1 2 2 1 2πεχεπ Pharma EdgeAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 20139 • 21

Reference-Scaled Average Bioequivalence (NTIDs)EvaluationComparison of σ WT with σ WRs WTis the estimate σ WTwith ν 1degrees of freedom(ν 1= n 1– 2 in the fully replicate).s WRis the estimate σ WRwith ν 2df.Probability of risk type I α = 0.1.F α ⁄ 2(ν1 ,ν 2 ) is the value of the F-distribution with ν 1(numerator) and ν 2(denominator) degrees offreedom and a probability of α ⁄ 2.πεχεπ Pharma EdgeF 1–α ⁄ 2(ν1 ,ν 2 ) is the value of the F-distribution with ν 1and ν 2df and a probability of 1 – α ⁄ 2.Bioequivalent if 95% upper CL of σ WT/σ WR≤2.5.Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201310 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Consequences of ScalingAt σ WR 0.1 (CV 10.03%) the expanded AR is90.00–111.11%140%FDA 2012 (AUC, C max )140%130%130%Bioequivalence limits120%110%100%90%111.11%90.00%10.03%21.42%120%110%100%90%CV WR510152025L – U94.87 – 105.4190.02 – 111.0885.35 – 117.0281.17 – 123.2077.15 – 129.6280%80%3073.40 – 136.25πεχεπ Pharma Edge70%0% 5% 10% 15% 20% 25% 30%CV WRAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201370%11 • 21

Reference-Scaled Average Bioequivalence (NTIDs)NTIDs (FDA)πεχεπ Pharma EdgeAs a consequence of scaling the AR fors WR >0.21179 (CV WR >21.42%) will be widerthan the conventional 80.00–125.00%.Possible ‘ways out’1. Cutoff on s WRand switch to conventional unscaled ABE2. A “Must Pass Both” criterion: RSABE + ABE Both methods maintain the patient’s risk

Reference-Scaled Average Bioequivalence (NTIDs)NTIDs (FDA)Both methods preserve the patient’s risk0.06cutoff=0.21179must pass bothactual level of significance0.050.040.030.02πεχεπ Pharma EdgeDJ Schuirmann 20110.15 0.17 0.19 0.21 0.23 0.25 0.27 0.29sigmaWRAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201313 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Exampleπεχεπ Pharma EdgeCNS drug from BEBAC’s filesRTRT | TRTR full replicate, 18 subjects, balanced,complete FDA1. critbound: –0.0098283 ≤0 (CV WR12.49%, CV WT5.58%)2. ABE: 90% CI 93.90–103.35% ⊂ AR3. upper 95% CL of s WT/s WR0.68427 ≤2.5 EMAAR 90.00–111.11%ABE: 90% CI 93.90–103.35% ⊂ AR(CV WR15.86%, CV WT5.73%)Data set in Excel 2000 format:http://bebac.at/downloads/NTID.xlsAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201314 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Example120%120%114.00%Bioequivalence limits110%100%111.11%110%100%90%87.72%90.00%90%πεχεπ Pharma Edge80%FDA EMAAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201380%15 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Thank You!Reference-ScaledAverage Bioequivalence (Part II)Open Questions?Helmut SchützBEBACConsultancy Services forBioequivalence and Bioavailability Studies1070 Vienna, Austriahelmut.schuetz@bebac.atπεχεπ Pharma EdgeAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201316 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Fully replicated 4-way designSAS code (FDA)data test1;set test;if (seq=1 and per=1) or (seq=2 and per=2);lat1t=lauct;run;data test2;set test;if (seq=1 and per=3) or (seq=2 and per=4);lat2t=lauct;run;data ref1;set ref;if (seq=1 and per=2) or (seq=2 and per=1);lat1r=lauct;run;πεχεπ Pharma Edgedata ref2;set ref;if (seq=1 and per=4) or (seq=2 and per=3);lat2r=lauct;run;Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201318 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Fully replicated 4-way design (cont’d)data scavbe;merge test1 test2 ref1 ref2;by seq subj;ilat=0.50.5*(lat1t+lat2t-lat1r-lat2r);dlat=lat1r-lat2r;run;SAS code (FDA)proc mixed data=scavbe;class seq;model ilat =seq/ddfm=satterth;estimate 'average' intercept 1 seq 0.5 0.5/e cl alpha=0.10.1;ods output CovParms=iout1;ods output Estimates=iout2;ods output NObs=iout3;title1 'scaled average BE';title2 'intermediate analysis - ilat, mixed';run;pointest=exp(estimate);x=estimate**2–stderr**2;boundx=(max((abs(lower)),(abs(upper))))**2;πεχεπ Pharma EdgeAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201319 • 21

Reference-Scaled Average Bioequivalence (NTIDs)Fully replicated 4-way design (cont’d)SAS code (FDA)proc mixed data=scavbe;class seq;model dlat=seq/ddfm=satterth;estimate 'average' intercept 1 seq 0.5 0.5/e cl alpha=0.10.1;ods output CovParms=dout1;ods output Estimates=dout2;ods output NObs=dout3;title1 'scaled average BE';title2 'intermediate analysis - dlat, mixed';run;s2wr=estimate/2;dfd=df;theta=((log(1.111111.11111))/0.10.1)**2;y=-theta*s2wr;boundy=y*dfd/cinv(0.950.95,dfd);sWR=sqrt(s2wr);critbound=(x+y)+sqrt(((boundx-x)**2)+((boundy-y)**2));πεχεπ Pharma EdgeAdvanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201320 • 21

Reference-Scaled Average Bioequivalence (NTIDs)πεχεπ Pharma EdgeUnscaled 90% BE confidence intervalsSAS code (FDA)PROC MIXEDdata=pk;CLASSES SEQ SUBJ PER TRT;MODEL LAUCT = SEQ PER TRT/ DDFM=SATTERTH;RANDOM TRT/TYPE=FA0(2) SUB=SUBJ G;REPEATED/GRP=TRT SUB=SUBJ;ESTIMATE 'T vs. R' TRT 1 -1/CL ALPHA=0.10.1;ods output Estimates=unsc1;title1 'unscaled BE 90% CI - guidance version'; title2 'AUCt';run;data unsc1;set unsc1;unscabe_lower=exp(lower);unscabe_upper=exp(upper);run;RSABE if1. critbound ≤0 and2. 90% CI of ABS within 0.8000 and 1.2500 and3. 95% upper CL of sWT/sWR ≤2.5.Advanced concepts of IVIVC through case studies & Biostatistical aspects of Referenced scaled &Two Stage Designs: A regulatory perspective | Mumbai, 25 – 27 January 201321 • 21