Minimizing treatment duration and doses Jordan J. Feld MD MPH Toronto Centre for Liver Disease Sandra Rotman Centre for Global Health University of Toronto
Disclosures Research: Abbvie, Gilead, Janssen, Merck Consulting: Abbvie, Gilead, Janssen, Merck,
Outline Do we need shorter therapy? Can we predict the minimum duration? Do the data agree with the theories how short can we go with DAAs? Alternative approaches
Outline Do we need shorter therapy? Can we predict the minimum duration? Do the data agree with the theories how short can we go with DAAs? Alternative approaches
Outline Do we need shorter therapy? Can we predict the minimum duration? Do the data agree with the theories how short can we go with DAAs? Alternative approaches
Pros & Cons of shortening Pros Easier for patients Reduced toxicity Increased adherence Simplicity Preference Easier for treaters Capacity Monitoring Cheaper Fewer doses = fewer $ Cons (or at least non-pros) Therapies are easy & safe Overtreatment is safe Toxicity could increase Adherence may not change Could increase complexity Therapies are easy & safe Limited or even no monitoring Limited impact on capacity DAAs are cheap to produce New drugs could actually increase costs Relapse Always a risk
So should we aim for shorter therapy? All things being equal, yes But all things being equal is a big question mark Assuming: Similar or identical SVR rate Similar or better safety profile Similar complexity to treatment simplicity may trump short
Where is short therapy most important? Harder to reach populations Homeless PWID Incarcerated Limited capacity Low/middle income countries Rural/remote regions (?)
Outline Do we need shorter therapy? Can we predict the minimum duration? Do the data agree with the theories how short can we go with DAAs? Alternative approaches
Log HCV RNA (IU/mL) How long do we need to treat? 6 5 4 3 2 1 10-4.22 0 1 st Phase Block of viral production 2nd Phase Clearance of infected cells AND/OR Cure of infected cells LLOD - serum LLOD body (Cure boundary) 0 2 4 6 8 10 12 16 24 Weeks of therapy Affected by: Host: Cirrhosis, TE,?IL28B Virus: Resistant variants Drug: Single/multiple MOA, RBV ~5 log Perelson Nat Rev Gastro Hep 2015
Probability to eliminate the last viral particle SVR12 (%) What does the model predict? Theory Telaprevir monotherapy 100% Adherence Partial Adherence Treatment duration - weeks Assume: 100% adherence No resistance ie really for combo DAA Clinical Trial: ION-3 SOF/LDV 100 80 60 40 20 0 94 202/ 215 Model predicts fairly well - ~95% with 8 weeks 93 201/ 216 95 206/ 216 S/L S/L/R S/L 8 wks 12 wks Guedj Hepatology 2011, Kowdley NEJM 2014
Log HCV RNA (IU/mL) But it s never that simple SYNERGY: SOF/LDV x 12w vs SOF/LDV/9660 (NNI) or SOF/LDV/9451 (PI) x 6w Results: - Based on kinetics: predict SVR of 7% in B & 26% in C - EOT A 0/20, B 12/20 and C 10/20 PCR +ve only 1 relapse! - Detectable virus in serum = 5-6 log/virus in the body! - SVR means: 1. Immune clearance 2. Detectable HCV RNA non-infectious SOF/LDV (A) SOF/LDV/9451 (PI) (C) SOF/LDV/9669 (NNI) (B) Time (days) Kohli Lancet 2015
Log HCV RNA (IU/mL) Not as simple as the slope LLOQ serum Immune boundary Cure boundary Days on treatment Weeks on treatment Throws off predictions poor understanding/prediction of immune boundary Means kinetics alone are probably not enough to predict Perelson Nat Rev Gastro Hep 2015
Log HCV RNA (IU/mL) Over-treatment of most patients 6 5 4 3 2 1 0 Shorter durations will still cure some people but not everyone Little things will matter: Fibrosis/cirrhosis Adherence, IL28b/IFNL4 Baseline resistance Treatment Duration LLOD - serum Immune boundary LLOD body (Cure boundary) 0 2 4 6 8 10 12 16 24 Weeks of therapy Perelson Nat Rev Gastro Hep 2015
SVR12 (%) What happens with effective therapy? ION 1, & 2: SOF/LDV +/- RBV x 12-24w Naïve Prior Trt (incl PI) Failures 100 99 97 98 99 94 96 99 99 80 60 40 20 0 211/ 214 S/L 211/ 217 S/L/R 12 wks 212/ 217 S/L 215/ 217 S/L/R 24 wks 102/ 109 S/L 107/ 111 S/L/R 12 wks 108/ 109 S/L 110/ 111 S/L/R 24 wks Nearly all patients are cured with 12 weeks Treatment-experience +/- cirrhosis need longer (or RBV) Afdahl NEJM 2014, Afdahl NEJM 2014, Kowdley NEJM 2014
SVR12, % SVR12 % Not just in clinical trials 100 LDV/SOF LDV/SOF+RBV 97 97 95 97 92 100 PrOD 95 80 80 60 60 40 40 20 0 150/ 154 607/ 627 153/ 161 8 Weeks 12 Weeks 24 Weeks 12 Weeks 24 Weeks 86/ 89 12/ 13 20 0 412/ 432 G1 LDV/SOF LDV/SOF+RBV Highly effective data in the real-world Terrault AASLD 2015, Zuckerman EASL 2016
We are almost certainly overtreating most people We re not that good! To get such excellent results, most people do not NEED 12 weeks Significant variability in viral kinetics We are treating to the cure the slow responders (and curing the fast responders in the process)
So if 12 weeks is over-treatment, how short can we go?
Approaches to shorten DAAs alone DAAs in selected populations Baseline On-treatment response DAAs + Non-DAA
Approaches to shorten DAAs alone DAAs in selected populations Baseline On-treatment response DAAs + Non-DAA
SVR12 (%) How short can we go? Grazoprevir (PI) + Elbasvir (NS5A) + Sofosbuvir (Nuc) x 4 8 weeks in G1 with or without cirrhosis 100 80 4 Wks 87 6 Wks 80 8 Wks 94 60 40 20 n/n = 0 33 10/30 26/30 16/20 17/18 No cirrhosis Cirrhosis Very potent regimen 4 weeks clearly inadequate Gane EASL 2015
SVR12 (%) 6 weeks seems to be the edge of the cliff for DAAs SOF/VEL (NS5A) /VOX (PI) x 4 or 6 weeks 100 80 60 93 6 Wks 87 67 4 Wks 40 20 n/n = 0 14/15 13/15 20/30 Tx Naive, No Cirrhosis Tx Naive, Cirrhosis Txt Exp d, +/- Cirrhosis 27 4/15 Tx Naive, No Cirrhosis Very potent regimen 4 weeks clearly inadequate Gane EASL 2015
Approaches DAAs alone DAAs in selected populations Baseline On-treatment response DAAs + Non-DAA
SVR12 (%)/relapse Can we identify those who will do well with shorter therapy? 100 P=0.76 <6 M IU/mL >6 M IU/mL 97 96 89 94 80 60 40 20 0 252/ 261 126/ 131 139/ 146 8 12 8 80/ 85 12 P=0.01 <6 M IU/mL >6 M IU/mL 1.9 5/260 1.5 2/131 8.9 15/ 169 8 12 8 Baseline HCV RNA measure of virus & immune equilibrium Lower baseline shorter therapy more likely to work 1.2 1/85 12 FDA Analysis of ION-3 Study
Other ways to select 8 weeks Baseline factors predict 8 weeks Female sex IL28B/IFNL4 genotype Fibrosis (?) Arguably better predictors HCV RNA highly variable +/- 0.25 log on the SAME sample! 6,000,000 = 6.78 log Range 3.37 M to 10.67 M O Brien, Feld Hepatology 2015
SVR12 (%) Overall 8 weeks looks good! GT 1: LDV/SOF 8 weeks 97% 95% 97% 100% 99% 119 123 242 254 150 154 43 44 103 103 69 70 Phase 3 Studies TRIO Cohort HCV- TARGET 172 202 IFI GECCO Multiple real-world cohorts showing HCV RNA approach works
Can we predict who needs only 4 weeks? High baseline viral load predicts failure But low baseline viral load nor IL28B CC predict success Lawitz AALSD 2014 LB
Choosing the right population: A very provocative trial! SVR12 100% 80% PWID 4 weeks of SOF/LDV/RBV +/- Peg G1 or 3 (and 1 G2), TN, on OST, F0-2 75% 94% 60% 40% 20% 0% n=16 Relapse 1 LTFU 3 LDV/SOF + RBV 4 weeks n=16 Relapse 0 LTFU 1 LDV/SOF + PegIFN + RBV 4 weeks Single centre study but interesting? Partial immunity from repeated exposures? Overhuis AASLD 2016 Abstract 921
Back to Response-guided therapy? Response-guided therapy with 3 DAAs (SOF/PI/NS5A) ultra-short (3 week) therapy for those with rapid response Lau AASLD 2015 LB-23
HCV RNA<LLOQ Successful approach Outcome: 13 of 18<500 IU/mL by d2 All SVR with 3 weeks of tx Others SVR with 12w of SOF/LDV Overall 100% SVR Critical factors: G1b, Asian, IL28B CC Weigh complexity vs shorter therapy cost saving for patients buying their own medications (China) Lau AASLD 2015 LB-23
SVR12 (%) SVR12 (%) Alternative: Over-treat, but 1 size fits all 100 SOF + Velpatasvir (GS-5816) (NS5A) x 12 wks in G1, 2, 4, 5, 6 Naïve/Experienced +/- cirrhosis 80 99 98 99 100 100 97 100 SOF/VEL x 12 vs SOF/VELVOX x 8 100 80 9 6 9 8 60 60 40 20 0 1 relapse 2 lost to follow-up 1 withdrew consent 1 relapse 1 death 618 624 206 210 117 118 104 104 116 116 34 35 41 41 Total 1a 1b 2 4 5 6 Genotype Adding drugs to shorten doesn t always work 40 0 21 relapse 3 rel 124/ 132 SOF/ RBV 133/ 134 SOF/ VEL Feld NEJM 2015, Jacobson AASLD 2016 31
One size fits all - pangenotypic regimens PI + NS5A + Nuc Voxilaprevir(PI) + Velpatasvir (NS5A) + SOF (Nuc) 12w Grazoprevir (PI) + Ruzasvir (NS5A) + MK-3682 (Nuc) 12w Sovaprevir (PI) + Odalasvir (NS5A) + ACH-3422 (Nuc) -?8w PI + NS5A Glecaprevir (PI) + Pibrentasvir (NS5A)?8w Promising combinations 8-12 weeks for all patients incl post-daa
Approaches DAAs alone DAAs in selected populations Baseline On-treatment response DAAs + Non-DAA
Is there another way forward?reconsidering hosttargeting agents.
Many potential host targets Li PNAS 2009
mir122 A host target Janssen NEJM 2013
Miravirsen Miravirsen SC injection weekly Janssen NEJM 2013
A better mir122 inhibitor Increases liver uptake Increased mir122 inhibition Van der Ree EASL 2015
HCV RNA level (log IU/mL) age, 619 209 pixels) Single injection data:image/png;base64,ivborw0kggoaaa 4 mg/kg RG-101 Placebo 4 mg/kg HCV RNA BLOQ (n=4) Time (Weeks) 9 of 12 patients with undetectable HCV RNA with single dose 4 negative out to 20 weeks Van der Ree EASL 2015
A potential 4-week option SVR12 100% (27/27) 96% (26/27) 92% (22/24) SVR12 SVR48 Surprisingly successful with even single DAAs On hold for toxicity jaundice unclear future Horvath AASLD 2016 - withdrawn
shrna copes per cell One injection for cure? shrna Day 50 shrna-22 shrna-19 shrna-6 Single shot to cynomologous monkeys HCV Replicon 3 shrnas targeting conserved regions of HCV Long-lasting expression (180 d) of all 3 shrnas at levels needed to suppress replicon in non-human primates Suhy Mol Ther 2012 Dose TT-034 vg/kg 1.25 x 10 11 1.25 x 10 12 6.25 x 10 12
Summary Shorter therapy is potentially attractive but ONLY if equally safe, effective and simple Cost alone should not be the primary reason to shorten pricing should be per course of Tx Modeling predicts most will need 8 weeks or less To go below 8 weeks with DAAs alone baseline or on-treatment response not one size fits all is complexity worth the effort? To get to 4 weeks or shorter, will likely need more than DAAs