Short summary

Mid-therapy FDG-PET/CT in patients with lymphoma permits informed treatment-related decisions and more accurate prognostication. Currently, assessments are made visually. The semi-quantitative framework PERCIST may offer significant advantages that may improve the ability to predict which patients have good vs. bad prognoses. Accurate predictions are necessary for timely implementation of more aggressive therapies and reducing unnecessary side effects. Our investigations aims to make a direct comparison between these methods using survival measures as primary endpoints.

Rationale

Diffuse large B-cell lymphoma (DLBCL) is the most common form of non-Hodgkin lymphoma with approximately 450 new cases every year in Denmark. The Danish annual incidence rate for Hodgkin lymphoma (HL) is approximately 150. DLBCL and HL are both aggressive lymphomas with high metabolisms which make them stand out clearly on FDG-PET/CT scans. This FDG-avidity is reflected in the current European guidelines, which recommend FDG-PET/CT for staging, mid-treament and end-of-treatment response evaluation of these lymphomas. The Lugano Classification is the recommended tool for assessing treatment response. When evaluating treatment response with the Lugano Classification, response is assessed visually and graded on the Deauville Scale.

A different set of semi-quantitative response criteria known as PERCIST may offer advantages over the Lugano Classification. PERCIST utilizes the quantitative measure of SULpeak and provides guidelines for a standardized and reproducible scanning procedure. This enables PERCIST to grade treatment response with a continuous measure providing more information and higher consistency between observers and institutions while having the potential for implementation in future automated response evaluations.

To our knowledge, PERCIST and the Lugano Classification have only been compared for response assessment of DLBCL in two previous studies. These studies failed to investigate the prognostic information carried with each response category and importantly, they failed to investigate the ability to identify poor responders at interim.

In a retrospective clinical study at our institution, we aim to investigate and compare the ability to predict overall survival (OS) and progression free survival (PFS) with PERCIST and the Lugano Classification in a population of patients with DLBCL and HL, respectively. Secondary outcomes include 1) Agreement; 2) Optimal cut-off for percentage change in SULpeak for distinguishing responders from non-responders.

Description of the cohort

Men and women from the age of 18 and up who were diagnosed with DLBCL or HL and began first-line treatment at OUH between January 2010 (DLBCL) or January 2015 (HL) and July 2019. All included patients had baseline, mid-therapy and end-of-treatment FDG-PET/CT scans.

Data and biological material

For this retrospective study, relevant information from patient journals, baseline blood samples, histopathological verification of diagnoses and FDG-PET/CT scans are retrieved and stored.

Collaborating researchers and departments

Consultant, Head of research, Ph.D., Associate Professor Malene Grubbe Hildebrandt

Department of Nuclear Medicine, Odense University Hospital

• Consultant, Head of Research, Ph.D., Associate Professor, Malene Grubbe Hildebrandt
• Consultant, Associate Professor, Anne Lerberg Nielsen
• Biostatistician, Ph.D., Professor, Oke Gerke

Department of Clinical Research, University of Southern Denmark

Department of Hematology, Odense University Hospital

• Consultant, Ph.D., Associate Professor, Thomas Stauffer Larsen
• Specialist Registrar, Ph.D., Karen Juul-Jensen