MRI-Guided Radiation Therapy for the Treatment of Early-Stage Kidney Cancer, the MRI-MARK Trial

Patients will undergo split function renal scans to determine glomerular filtration rate (GFR) at baseline, 12 months, and 24 months. Change in GFR will be reported by diving the difference in GFR at baseline and at the respective follow up visit, by baseline GFR. The change of GFR at follow up visits and baseline will be evaluated by paired t-test. Mixed model will be used to explore the change of GFR over time after taking the within-subject correlation.

The frequency of non-laboratory non-fatigue adverse events will be determined, defined as a grade 3+ event following SBRT, attributed to SBRT, and affecting the ipsilateral kidney or an adjacent organ (contralateral kidney, liver, small bowel, large bowel, stomach, spleen, or nerves) as per Common Terminology Criteria for Adverse Events (CTCAE) version 5.0.

Patients will undergo MRI of the abdomen at regular intervals (i.e. at 6, 12, 18, and 24 months) permitting analysis of changes in functional imaging biomarkers evaluating tumor and renal perfusion and diffusion, which can be correlated to clinical outcome and toxicities. Mixed model will be used to explore the change of tumor characteristics over time after taking the within-subject correlation. Multi b-value diffusion-weighted imaging (DWI) derived diffusion and perfusion changes will be evaluated over the course of therapy to assess their utility as predictors of renal insufficiency following therapy.

All tumors will be biopsied at baseline. When safe and feasible, patients will undergo re-biopsy at 24 months following SBRT. Viability and percent necrosis will be determined by hematoxylin and eosin (H\&E) staining. Other stains such as ki67 may be included. Taking pathologic evaluation as gold standard for no viable tumor cells following SBRT, will correlate tumor biopsy results with local control by imaging. McNemar test will also be used to evaluate the agreement between local control (yes versus [vs] no) and pathologic evaluation (responder vs none). Will also calculate sensitivity and specificity of our primary endpoint of local control (no growth by computed tomography [CT] scan at 24 months).

Will estimate no progression Response Evaluation Criteria in Solid Tumors (RECIST) at 24 months (less than 20% growth in largest tumor dimension) along with the 95% confidence interval (CI).

Will estimate the probabilities of patients remaining free from distant metastases utilizing the Kaplan Meier method with corresponding 95% CIs. The log rank test will be used to compare the survival difference between important subgroups (such as tumor categories, etc.). Cox model will be conducted to examine the clinical factor effect on time-to-event variables.

Will estimate the probabilities of patient survival utilizing the Kaplan Meier method with corresponding 95% CIs. The log rank test will be used to compare the survival difference between important subgroups (such as tumor categories, etc.). Cox model will be conducted to examine the clinical factor effect on time-to-event variables.

Assessed using the validated ENRICH questionnaire. Questionnaire scores will be calculated based on standardized manual associated with the instrument. Incomplete surveys (in the event that a participant declines to answer one or more questions) will still be considered valid and will be used for study purposes. Descriptive statistics will be used to summarize scores of the questionnaire. The distribution of each continuous variable will be summarized by its mean, standard deviation, median, and range. The distribution of each categorical variable will be summarized in terms of its frequencies and percentages. We will compare results at each time point to estimate financial toxicity of treatment. The association between patient clinical factor (such as T-stage or histology result) and clinical measurement (e.g., local control, pathologic response, tumor characteristics by MRI) will be examined by Wilcoxon's rank sum test or Fisher's exact test, as appropriate.

Will compare planned doses to true total doses delivered to the target and adjacent normal structures.

Will generate target structure and normal tissue dose volume histograms (DVHs) based on the simulation imaging as per routine treatment planning. Verification MR images will then be obtained immediately prior to each delivered fraction, which may or may not prompt an adaptive plan change. Will use the verification MR images and the delivered plan to generate additional target structure and normal tissue DVHs that will approximate "true" delivered dose. Will generate a cumulative true DVH based on each individual fraction, and compare this to the original planning DVH. Normal tissue toxicity will be assessed in light of both the planned and the true DVH. To assess the impact of intrafraction target and normal structure movement, MR images may also be obtained at specified time intervals during an individual treatment fraction.

The frequency of non-laboratory non-fatigue adverse events will be determined per CTCAE version 5.0. These events will be correlated to the true delivered dose received by the associated organ(s). The number of cases in which the prescribed isodose line failed to cover the entirety of the iGTV will be determined, and correlated to outcomes.