Translating TB Treatment Response

1

Translating TB Treatment Response

Rada Savic, Professor

Dept .of Bioengineering and Therapeutic SciencesDiv. of Pulmonary and Critical CareUCSF TB Center

1

Exciting Times for TB Drug Development

2

First ever 4 month Regimen for DS-TB

First ever 6 month Regimen for XDR/MDR -TB

Class Compound, Developer Development

Oxazolidinone

ribosome inhibitors

Sutezolid; Sequella, TB Alliance IIC

TBI-223; TB Alliance I

Delpazolid; LegoChem IIA

DprE1 inhibitors OPC-167832; Otsuka IIA

BTZ-043; University of Munich, DZIF IIA

TBA-7371; TB Alliance, Gates MRI IIA

QcrB inhibitor Telacebec, Q203; Qurient, Infectex IIA

Cholesterol-dependent

M.tb inhibitor

GSK286; Glaxo-Smith-Kline I

Diarylquinoline , ATP

synthase inhibitors

TBAJ587; TBAJ876; TB Alliance I

Gyrase B inhibitor SPR-720; Spero, Gates MRI I

New drugs/new MoA in the pipeline

New NGO/Pharma/Academia Partnerships

• BMGF PanTB Collaboration• UNITE4TB- EU• ACTG - US

2

Optimization Combinatorics a.k.a. Too many options

3

Drugs Doses FrequenciesDuration

2. Optimizing Multidrug TB Regimen

Going forward (Joint Regimen Development Program)

• 8-12 novel compounds (first in human completed) • (4 major Pharma, BMGF, Academia)• >1 million (2-5 drug combinations) possible

Immune deficient

nude mice

Rabbit lesion model

Kramniklesion model

• Immune impact

• Lesion specific site of disease

Variety of Models and Tools,however they are not quantitative and predictive yet

Hollow fiberPKPD

Balb/C

In vitro combinations

4

3

Quantitative Translation Toolbox

Dose-rangingPK

First in human safe dosing

Rabbit lesionPK

Clinical lesionPK

New New

Existing

Existing

BALB/c & Nude mice baseline

ClinicalPhase 2a

Preclinical:

Clinical:

BALB/c PKPD(mono)

New

Existing

Control

True drug effectin combination

with immune kill

Existing

M,Z RIF, INH, PZA, MFX,LZD, RPT, BDQ,

PMD, DLM

New

BALB/c PKPD(multi)

New

Existing

Clinical Phase 2& beyond

Existing NewHRZ, JPM, JPMZ,

LJP …

BALB/c resistance

Existing

Clinical MDR

H & R

Existing

BALB/c CFU

ClinicalTTP, TTCC,

CFU

Describe the interactions

between drugs

5

(I) Empirical ML/AI TOOL for regimen ranking in a mouse

Probability of Relapse

Ranking compounds based on 2 /partial 3 drug combo dataPredicting long term relapse from early (1 month) responseFor the known:

• Incubation time• Baseline• Regimen composition• CFU change at Month 1

Strydom et al, TB Science 2020

4

7

(II) Lung Resection Studies and Lesion coverage Tool

Ernest et al, 2020, Strydom et al, PLOS Medicine 2019, Prideaux et al, Nat Medicine 2016, Rifat et al, Sci Transl Med 2018

Time after dose (h)

Method

Result

Quantitative prediction ofPhase 2B/Phase 3 results

Synergy/Additivity/Antagonism assessment

Strydom et al, PLOS Comp Biol 2020, Bartelink et al, Clin Transl Sci 2017

(III) Integrated immunology/pharmacology models: PREDICTIVE tools

5

(IV) Novel Early Trial Designs linked to Translational Tools

Prioritized regimens (1) Prioritized regimens (2)

(V) Data Sharing and Modeling Tools – TB ReFLECT

Only “hard-to-treat” phenotypes require 6 month duration.; 5 item risk score can help stratify patients (Cavitation, Baseline severity, HIV, malnutrition, sex)

10

6

Integration of Computational Tools and Data Science Approaches

Acknowledgements

UCSFTB Translational Team• Jackie Ernest • Natasha Strydom • Nan Zhang • John Fors • Amelia Deitchamn • Imke Bartelink• Chores Wang

TB Clinical Team• Marjorie Imperial• Leah Jarlseberg• Craig Shafer• Payam Nahid• Patrick Phillips

• Veronique Dartois

• Eric Nuermberger• Kelly Dooley