Contemporary Imaging Endpoints with Locoregional Therapies: - - PowerPoint PPT Presentation

Contemporary Imaging Endpoints with Locoregional Therapies: Response!

Riad Salem MD MBA FSIR Professor of Radiology, Medicine and Surgery Vice-Chair, Image-Guided Therapy Chief, Vascular and Interventional Radiology Robert H Lurie Comprehensive Cancer Center Northwestern University Chicago, Illinois

Disclosures

Consultant

• BTG UK Ltd
• Boston Scientific
• BMS
• Dova

Scientific Advisory Board

• BTG UK Ltd
• Cook
• Eisai
• Exelixis

Grant/Research Support

• BTG UK Ltd

Outline

1-Differentiating RECIST 1.1 and mRECIST response 2-Investigator review and central review differences

• Why are outcomes different?

3-Making clinical decisions based on response

Point 1: Differentiating RECIST 1.1 and mRECIST

COMPETING METHODS OF ASSESSING RESPONSE

• 1. Uni-dimensional
• Response Evaluation Criteria in Solid Tumors

(RECIST)

• 2. Bi-dimensional
• World Health Organization (WHO)
• 3. Volumetric
• 4. European Association for the Study of the

Liver (2D EASL, mRECIST)

• 5. European Association for the Study of the

Liver (3D EASL)

• 6. Functional Diffusion-Weighted (DW) MR

Imaging (ADC)

Bruix et al J Hep 2001

• EASL (necrosis) Guidelines measure

change in the amount of enhancing (viable) tissue only

• Tissue Viability (Necrosis)

– Assumes enhancement is tumor – Did not describe methods to measure necrosis – Qualitative observation

• Number of tumors required to assess

response decreased to:

– Maximum of 5 – Maximum 2/organ

Eisenhauer et al Eur J Cancer 2009

Lencioni et al Semin Liver Dis 2010

• Reviews and clarifies imaging methodology
• New lesions
• Portal vein thrombosis-non measurable
• LN > 2 cm (malignant)
• Pleural effusions/ascites
• Retrospective adjudication of extrahepatic metastases

mRECIST versus RECIST v1.1

mRECIST RECIST v1.1

Complete response Disappearance of any intratumoral arterial enhancement in all target lesions Disappearance of all target lesions Partial response Decrease of ≥ 30% in the sum of diameters of viable (enhancement in the arterial phase) target lesions, taking as reference the baseline sum of the diameters of target lesions Decrease of ≥ 30% in the sum of diameters of target lesions, taking as reference the baseline sum of the diameters of target lesions Progressive disease Increase of ≥ 20% in the sum of the diameters of viable (enhancing) target lesions, taking as reference the smallest sum of the diameters of viable (enhancing) target lesions recorded since treatment started Increase of ≥20% in the sum of the diameters of target lesions, taking as reference the smallest sum of the diameters of target lesions recorded since treatment started

Eisenhauer et al Eur J Cancer 2009, Lencioni et al Sem Liv Dis 2010

Edeline et al Cancer 2011

Patterns of Progression-Enhancement-mRECIST

Patterns of Progression-Portal Vein Thrombosis

Point 2: Investigator review and central review differences Why are outcomes different?

Site versus Central review

Site Review

• No protocol training
• May not be liver MR/CT

specialized

• MD possibly biased
• Aware of clinical circumstance
• Prospectively performed
• Influenced by patient care
• Affects treatment decisions

Central Review

• Protocol training + charter
• Specialist in liver MR/CT
• Competency is tested
• No MD bias
• Blinded to clinical circumstance
• Retrospectively performed
• Intended for statistical analyses

and temporally unrelated to patient care

• Per patient discordance in

treatment-response finding

• Reader reviews all patient scans
• 2nd, 3rd readerà adjudication

Normal reader variability adjudication

Point 3: Making clinical decisions based on imaging Response matters!

Response versus Non Response: Interpret with caution

1-Survival analyses by simple R vs NR biases responders

à Misinterpretation that response provides survival benefit

2-Guarantee-time bias

à Better biology patients live longer and show R à Worse biology live insufficient time to show R, hence NR

3-Two methods to correct for this bias

à Landmark-most commonly used à Mantel-Byar

Anderson et al JCO 1983

Responder Non Responder

Landmark Analyses

1-Evaluates survival by tumor response 2-Addresses guarantee-time bias via the selection of fixed time points after treatment onset 3-Patients at the landmark time points are stratified by best response and survival analyses are performed 4-Patients who die before the landmark time point are excluded 5-Multivariate Cox regression models with OR as a time- dependent covariate

• Adjusts OS outcome corrected for confounding factors

(liver function, burden)

• guarantee/immortal-time bias

Memon et al Gastroenterology 2011, Lencioni et al JHEP 2017, Meyer et al Liv Int 2017, Kudo et al ASCO 2018

Kudo et al ASCO GI 2019

• Time-dependent Analysis
• Landmark Analyses confirm responders>non responders

Time since start of treatment

DOS

Tumor load at Baseline Lethal tumor load

PFS +20% from nadir Tumor nadir

• 30%

Why does response matter? Related to PFS

• for any given PFS, highest RR

yields lowest tumor burden

Llovet et al JHEP 2019

PFS surrogate of survival

Conclusions

• mRECIST favored over RECIST 1.1 in HCC
• Site vs central read will show differences
• Landmark analyses critical when assessing R vs NR survival
• For any given PFS, better RR beneficial for patient
• PFS as potential surrogates
• With advent of IO, response (and duration) has become prime

endpoint of interest