TOwards Personalized multi-modality esophageal cancer treatment using machine learning-based Quantitative MRI

Esophageal carcinoma is amongst the world*s most prevalent and fatal
malignancies, with an overall 5-year survival rate of less than 20%. Since the
CROSS trial, most patients with resectable esophageal cancer are treated with
chemotherapy and radiotherapy, so-called, neoadjuvant chemoradiotherapy (nCRT),
followed by surgery. However, the median overall survival after nCRT + surgery
is only four years. Furthermore, long-lasting (10+ years) deterioration in the
quality of life has been described after curative esophageal cancer surgery.
Conversely, 29% of patients who underwent esophageal cancer surgery had a
pathological complete response (pCR) after nCRT. These complete responders risk
serious complications from extensive surgery that they do not require.
Therefore, other treatments, such as definitive CRT, are investigated.
Treatments could become substantially more efficient if we can personalize
treatment and identify which patients benefit from a more intensive treatment
regime (to improve survival) and which patients could safely be treated with a
less intensive treatment regime (to improve quality of life). Current imaging
techniques (CT, PET-CT, EUS) have substantial limitations for predicting
response. Thus, novel biomarkers are urgently needed to guide and personalize
treatment. Quantitative magnetic resonance imaging (MRI) is a versatile
technique that measures tumour pathophysiology non-invasively. MRI can play a
major role in patients with esophageal cancer, particularly for treatment
response prediction and monitoring. Several studies at expert centres,
including the Dutch PRIDE study, are investigating the use of quantitative MRI
for response assessment after nCRT. These studies show promising results for
DWI, DCE and T2* MRI for identifying responding patients at an early treatment
stage. Most studies highlighted the need for improving quantitative imaging
techniques and adding motion management. Reasons for poor image quality are the
presence of cardiac and respiratory motion. Furthermore, the esophagus and the
esophageal have a small diameter, necessitating high-resolution imaging. We
hypothesize that improved quantitative MRI technologies might be able to
accurately visualize tumour pathophysiology in esophageal cancer and thereby
aid to substantially improve patient tailored treatment of esophageal cancer.

Primary:
Sub-study I:
- To develop high-quality fast quantitative MRI protocol for
multi-contrast imaging of tumour pathophysiology at 3T in esophageal cancer
patients

Sub-study II:
- To determine inter- & intra-session test-retest characteristics of
optimized quantitative MRI in esophageal cancer patients

Sub-study III:
- validate optimized quantitative MRI for prediction of the presence of
residual tumor in the ex vivo surgical specimen in individual esophageal cancer
patients


Secondary:
Sub-study II
- Determine typical baseline characteristics of quantitative MRI parameters in
esophageal cancer patients
Sub-study III
- Perform spatial correlation of quantitative MRI parameters to histopathology
of ex vivo surgical specimens
- Characterize the response of quantitative MRI parameters to neoadjuvant
therapy on individual esophageal cancer patients
- To investigate whether there is a relation between the baseline quantitative
MRI parameters and treatment efficacy (MRI1)
- To investigate whether there is a relation between the change in quantitative
MRI parameters and treatment efficacy (MRI1, MRI2 & MRI3)
- To investigate the relation between the quantitative MRI parameters and
18-F-FDG-PET-CT parameters

Prospective single-center cohort study

Participation in this study does not lead to an immediate advantage for the
participant. The
overall aim of the TOP-QMRI project is to develop a quantitative MRI technique
that could provide further insight into treatment efficacy for esophageal
cancer patients.

The risks of this study are expected to be low and contains the risk associated
with the infusion of Dotarem (Gadoteric acid) contrast injection
(hypersensitivity reactions or hypotension). The decision to participate in the
study or not will not delay or otherwise influence the treatment trajectory for
patients. Where possible, we will aim to plan scans on days that the patients
are already visiting the Amsterdam UMC for regular treatment.

Sub-study I: 1 MRI. Time invested by the patient for participating in this
study will approximate (MRI-scan time 75 minutes) 1 hour and 15 minutes.
Sub-study II: 2 MRI at 2 different moments in time. Both MRIs will take place
before neoadjuvant treatment. Time invested by the patient for participating in
this study will approximate (MRI-scan time 75 minutes) 2 hours and 30 minutes.
Sub-study III: 3 MRIs at 3 different moments in time. The first MRI will take
place before neoadjuvant treatment, the second MRI two weeks after the start of
neoadjuvant treatment and the third MRI before esophagectomy. Time invested by
the patient for participating in this study will approximate (MRI-scan time 75
minutes) 3 hours and 45 minutes.