Whole Body MRI With DWI for Monitoring Patients Treated for Testicular Cancer Stage II-III

AIMS AND HYPOTHESES

The overall aim of this study is to reduce the risk of radiation-induced secondary cancers in patients operated for testicular cancer (TC) by replacing computed tomography (CT) as a follow-up imaging method with non-ionizing whole body magnetic resonance imaging (WB-MRI) including diffusion weighted imaging (DWI). Under the hypotheses that MRI can detect relapse in stage I patients, that MRI is not inferior to CT for diagnosing and monitoring metastatic disease in stage II-III patients, and that the scan time of MRI with DWI can be reduced in order to facilitate the clinical and practical feasibility, the trial has two specific aims leading to two substudies:

1. To study the ability of WB-MRI with DWI to replace standard CT in TC stage II-III patients in a prospective, non-inferiority study.
2. To evaluate if scan time in the WB-MRI can be significantly reduced in the TC stage II-III group by using multiband DWI technique and by evaluating the minimum necessary image quality while maintaining sufficient diagnostic accuracy in order to improve clinical applicability of the technique.

BACKGROUND TC affects approximately 1% of Danish men with 301 new cases in 2015 and is the most common cancer in men aged 15-35 years. It is the most curable solid cancer type with a 5-year survival rate of 90-95%; In 2013 97% of Danish patients were alive two years after diagnosis. According to Danish and European guidelines, staging and follow-up of these patients involve 5-10 CT scans of each patient, imposing a significant radiation burden. Based on the risk estimates of BEIR VII and the Danish guidelines, approx. 3-5 of the 300 Danish patients presenting with TC each year are expected to develop a radiation-induced secondary cancer in their remaining life time, half of which are expected to be fatal.

MRI is rapidly developing with higher availability and higher sensitivity for cancer detection though the use of DWI and new WB-MRI-technologies that can cover large parts of the body in clinically relevant scan times. With this development, it is within reach to almost eliminate this radiation burden by substituting the large amount of CT scans with MRI scans in TC. MRI is without any known risk of long-term side effects. Despite these obvious benefits of MRI, limited data exist on MRI used in follow-up of TC.

In a joint effort between Depts. of Radiology and Oncology at AUH we already introduced MRI for the follow-up of non-disseminated TC (stage I) in 2008. We now want to evaluate the results of this unique cohort of patients followed by a prospective trial to evaluate if the newest WB-MRI techniques can replace CT in patients with disseminated TC (stage II-III).

CLINICAL DATA in order to establish a reference for relapse, a database with all relevant data will be established using REDCap database software at Aarhus University. Clinical data incl. age, blood tests (AFP, hCG and LDH), histopathological reports, prior treatment, clinical notes, and additional scans will be evaluated for the entire study period.

This study does not include material for biobanks and data will not be retrieved from or transferred to other institutions or countries.

RECRUITMENT AND INFORMED CONSENT The patients will be recruited when referred to Dept. of Oncology, AUH. The patients will be approached by and asked to participate in the study by the oncologist at their first visit to the department. The patients are encouraged to bring an assessor. The information will be given in an undisturbed room and written information will be handed out. The timeframe from referral to treatment is governed by the "pakkeforløb" set by the Danish Health and Medicines Authority and will not be influenced by participation in the current investigation. Because of this short time frame the patients are given a relatively short period of 24 hours for consideration and can request further information from the principal investigator. Written informed consent will be obtained.

Participation in this study is covered by the Patient Compensation Association (Patienterstatningen).

SUBSTUDY A Can WB-MRI (chest, abdomen, pelvis) including DWI with or without CT of the chest replace the standard contrast enhanced CT for monitoring patients with TC stage II-III?

Design:

Prospective comparison of WB-MRI with DWI to standard CT in a non-inferiority paired design with binary outcome. The non-inferiority design is used as the aim is to evaluate if MRI can replace CT and because although considered the "golden standard", the sensitivity of CT for detecting relapse in patients with TC has been reported to be 70-90%.

Imaging:

MRI will be performed before treatment (chemotherapy or radiotherapy) is initiated and will be compared to the standard CT used for staging prior to initiating treatment. At present the standard of care is a contrast enhanced CT of the chest, abdomen and pelvis. Each of these patients will receive one control scan with both MRI and CT in order to assess the ability of MRI to monitor effect of treatment. Since MRI is known to be less sensitive for detecting lung metastases than CT, it will furthermore be evaluated if a combination of MRI with a CT of the chest performs better than MRI alone.

MRI will be performed on a 1.5 T Philips Ingenia scanner equipped with dedicated surface coils for WB-MRI scanning and a new prototype Multiband DWI technique capable of reducing DWI scan time with approx. 50% as well as the standard DWI sequences. The total acquisition time for T1-weighted, T2-weighted, and DWI sequences for chest, abdomen, and pelvis is approx. 40-45 min. No contrast agents will be administered for the MRI scans.

Analysis:

Three datasets will be analysed for comparison:

Contrast enhanced CT of the chest, abdomen and pelvis WB-MRI with DWI with CT of the chest reused from above scan WB-MRI with DWI MRI will be evaluated independently and blinded by two experienced MRI radiologists and CT will be evaluated independently and blinded by two experienced CT radiologists. Lesions will be characterised by localisation and greatest axial diameter and characterised as

1. definitely benign
2. probably benign
3. can be either benign or malignant
4. probably malignant
5. definitely malignant For binary analysis 1-2 will be considered benign and 3-5 will be considered malignant. Histopathology will be recorded if available. Any lesion detected at follow-up that can be retrospectively detected on at least one of the initial imaging modalities and shows either increase or decrease (due to treatment) in size will be classified as a missed metastasis.

Interreader agreement will be calculated.

Statistical analysis:

This study will be conducted as a non-inferiority paired design with a binary outcome (metastasis or no metastasis on a lesion basis). The patients are their own controls as they undergo both imaging modalities.

The study is designed to claim non-inferiority of WB-MRI with DWI or WB-MRI with DWI plus CT of the chest compared to standard of care contrast enhanced CT of the chest, abdomen and pelvis if the confidence interval (CI) of the difference is within a of 10%. Contrast enhanced CT of the chest, abdomen and pelvis will be considered the golden standard to which WB-MRI with DWI with or without CT of the chest is compared.

Interreader agreement will be calculated. Sensitivity, specificity, PPV, and NPV will be calculated for both WB-MRI with and without CT of the chest and standard of care CT for detecting lesions. Receiver operating characteristics (ROC) curves will be calculated.

Dept. of Biostatistics will be consulted.

SUBSTUDY B Can scan time for WB-MRI be significantly reduced in the TC stage II-III group by using multiband DWI technique and by evaluating the minimum necessary image quality while maintaining sufficient diagnostic accuracy in order to improve clinical applicability of the techniques.

Design:

A prospective comparison of Multiband DWI vs time optimized standard DWI for evaluating patients with TC stage II-III using CT as reference.

Imaging:

Imaging from substudy A will be reused, no further imaging is obtained. For both standard DWI scans and multiband DWI scans, raw DWI data will be saved on the 1.5 T Philips Ingenia MRI scanner used for the study.

Analysis:

The DWI raw data will be reconstructed with different number of signal averages (NSA) directly corresponding to DWI scans obtained with different acquisition times. In this way, the effect of decreasing scan time can be studied without acquiring multiple datasets. The reconstructed "scans" with acceptable image quality will then be studied in a blinded fashion and compared to CT. In the same way, images with and without Multiband DWI sequences will be compared to CT.

MRI will be evaluated independently and blinded by two experienced MRI radiologists as described for substudy A and compared to the contrast enhanced CT of the chest, abdomen and pelvis evaluated in substudy A. To reduce recall bias, the radiologists will evaluate the different reconstructions in a randomised fashion two months apart.

Statistical analysis:

Sensitivity, specificity, PPV, and NPV will be calculated for each reconstructed dataset to determine the fastest scan with acceptable diagnostic accuracy.

Dept. of Biostatistics will be consulted.