PH Sensitive MRI Based Resections of Glioblastoma

Current standard of care therapy and all FDA approved adjuvant therapy for glioblastoma continue to provide less than 12 months of progression free survival (PFS) and less than 24 months of overall survival (OS). There is an extreme need for any novel therapy against glioblastoma that increases progression free survival and overall survival in patients diagnosed with this invasive form of cancer. A significant reason for such a poor prognosis is the infiltrative nature of this tumor in non-enhancing regions (NE) beyond the central contrast-enhancing (CE) portion of tumor, which is difficult to visualize and treat with surgical, medical, or radiotherapeutic means.

Since tumor cells exhibit abnormal metabolic behavior leading to extracellular acidification, we theorize a newly developed pH-sensitive MRI technique called amine chemical exchange saturation transfer echoplanar imaging (CEST-EPI) may identify infiltrating NE tumor beyond what is clear on standard MRI with gadolinium contrast. This phase I safety study will use use intraoperative CEST-EPI guided resections in glioblastoma at increasing distances from areas of CE tumor to test whether this technique is safe and can remove additional areas of infiltrative NE tumor.

The primary objective of this study is to assess the safety of pH-sensitive amine CEST-EPI guided resections for glioblastoma.The secondary objectives of this study include:

1. A preliminary efficacy analysis of CEST-EPI guided resections in extending progression free and overall survival.
2. To confirm that resected tissue obtained from pH-sensitive amine CEST-EPI guided resections contain infiltrating NE tumor.

The primary endpoint for this study will be safety of resecting "CEST positive", acidic regions within T2 hyperintense regions of glioblastoma thought to contain active NE tumor at increasing distances from contrast enhancing tumor with development of a recommended maximal tolerated resection.

1. At the maximal tolerated resection, a preliminary efficacy study with endpoints of progression free survival (as defined by RANO Resect 2.0) 1 and overall survival.
2. Quantitation of infilitrating tumor burden on CEST-EPI resected tissue using immunohistochemical staining.

12 patients up to 24 patients based on resection limiting toxicities with potential expansion of up to 16 patients at the maximum tolerated resection.

Inclusion Criteria:

1. Must be able to provide written informed consent
2. Male or female > 18 years of age
3. Karnofsky Performance Scale (KPS) > 70 (indicating good performance status).
4. Individuals with suspected, newly diagnosed or recurrent IDH wild type WHO IV glioblastoma (intraxial, expansile contrast-enhancing mass without evidence of metastatic disease. This will be reviewed by UCLA neuroradiology to only include patients with high likelihood of GBM)

Exclusion Criteria:

1. Pediatric patients
2. Diagnostic uncertainty (reviewed by UCLA neuroradiology history extracranial malignancy or autoimmune disease)
3. Medical conditions that make patients a poor candidate for anesthesia and/or surgery (decision for surgery will follow standard pre-operative clearance guidelines and will not differ for this specific study from standard of care treatment plan)
4. Involvement of eloquent areas (as defined by MRI signal clearly involving areas that would lead to a qualifying neurologic deficit as defined in surgical limiting toxicity - this will specifically include: 1) primary motor cortex, 2) primary sensory cortex, 3) sensorimotor fibers as defined on pre-operative diffusion tensor imaging, 4) primary language areas (Broca, Wernicke), 5) arcuate fasiculus as defined on pre-operative diffusion tensor imaging Pre-operative: Standard of care pre-operative MRI including perfusion and pH-weighted amine CEST-EPI (which will add up to 15 minutes of scan time) for a single pre-operative exam prior to surgery.

Surgery: 1 day (subjects to be admitted to the hospital) Follow-up: inpatient stay (1-3 days), 2 week clinical assessment (outpatient post-op clinic visit). MRI and clinical assessment at 4 weeks (end of resection limited toxicity window). Following this, there will be standard of care follow up with MRI and clinical assessment starting at 8 weeks +/- 4 weeks (per RANO 2.0). 1

Total study duration for recruitment, enrollment, and study completion of all subjects is up to 2 years.

Single-arm, surgical resection escalation safety trial with a preliminary efficacy study at the maximal tolerated resection

This safety evaluation will mimic a phase 1 dose escalation safety study using a rule based approach on based on a i3+3 design.2 Using standard of care resection of contrast enhancement as the baseline, we will begin with 3 subjects with maximal resection + "CEST positive" areas 0.7 cm from the contrast enhancing boundary within areas of T2 hyperintensity. If there is not > 1 pre-determined resection limiting toxicity (RLT, defined below) in this cohort, the r

This is a single center, single arm, phase 1 surgical dose escalation clinical trial with a preliminary efficacy study at the recommended safe maximal resection. Total duration of subject participation will be two years. Total duration of the study is expected to be 5 years.

Patients with suspected newly diagnosed or recurrent isocitrate dehydrogenase wild-type (IDH-WT) glioblastoma will be enrolled in this study if within the selection criteria (see inclusion and exclusion criteria below). Prior to surgery, patients will undergo: 1) standard pre-operative medical clearance 2) standard of care pre-operative magnetic resonance imaging (MRI) 34 with gadolinium contrast with thin cut T1 with and without contrast and diffusion tensor imaging and 3) amine chemical exchange saturation transfer echoplanar (CEST-EPI) MRI. 3,18,19,33 Pre-operative post-processing of CEST-EPI regions to identify areas of potential infiltrating tumor will be carried out as previously described. 3 In addition, resection constraints will be added based (starting at 0.7cm and increasing in 0.7cm intervals over 4 intervals to a maximum constraint of 2.8cm). This maximum constraint will reach the theorized threshold for significant infiltrating tumor cell burden based on preliminary data. 3,14 This data will be converted into a intraoperative object through our intraoperative neuronavigation system (BrainLab Surgical Navigation System, Munich, Germany).

On day of surgery, subjects will undergo standard of care procedures for craniotomy for brain tumor resection at Ronald Reagan UCLA Medical Center. Upon initial tumor resection, a biopsy specimen will be sent to neuro-pathology to confirm diagnosis. 5-aminolevulinic acid (5-ALA), a safe FDA approved fluorescence molecule used to increase extent of resection in glioma surgery will be administered for comparision with CEST imaging in surgery. 17 Areas of 5-ALA positivity and areas of CEST-EPI positivity will be biopsied for further comparision as described in the secondary objective. This will be followed by resection using the intraoperative CEST-EPI identified region with resection constraint using the intraoperative neuronavigation object. After completion of surgery, standard of care procedures for craniotomy will be completed.

To assess for short and long term complications related to surgery alone, patients will be admitted post-operatively with standard post-op care. There will be a immediate within 24-hr MRI while patient is inpatient, followed by 2 week post-operative visit alongside a 4 week MRI 35 to evaluate for rapid early progression. This will be followed with standard of care 8 week MRI and initiation of standard of care (temolozolomide and radiotherapy). The window of monitoring for surgery related complications will be the 4 week MRI and office visit as the resection limiting toxicities (described below) all fall within this window.

Given this protocol, safety constraints and analysis will follow a pre-determined statistical and methodological protocol. Using standard of care resection of contrast enhancement as the baseline, we will begin with 3 subjects with maximal CEST positive resection 0.7 cm from the contrast enhancing edge. The resection levels up to (0.7cm, 1.4 cm, 2.1cm or 2.8cm) from areas of CE tumor ("doses") escalation within each cohort will occur following a rule-based approach based on a i3+3 design 2 with a target resection level limiting toxicity (RLT) probability of 0.25 and an equivalence interval (EI) of 0.20 to 0.30, meaning that the RLT rate must not exceed the upper bound of the equivalence interval (EI) in order to extend the distance from contrast enhancing edge or for a distance to be selected for validation/expansion in "dose" escalation (if already at the highest dose level). Given that variability in the data can be large in a small cohort, the i3+3 design requires 2 criteria are met when determining whether to stay at the current distance or to reduce. First, the RLT rate must exceed the upper bound of the EI (ie, ≥1 RLTs occur in 3 patients). Next, if subtracting 1 from the number of RLTs results in a RLT rate within the EI, then the i3+3 rules state to stay at the current level and enroll 3 additional patients (Note: if a cohort in Dose Escalation is expanded to 6 patients and the adjusted RLT rate remains within the EI, this dose may be further expanded/validated in dose expansion following discussion with the Independent Data Monitoring Committee [IDMC]/Safety Monitoring Committee [SMC]). However, if the adjusted RLT rate still exceeds the EI, the rules state to enroll 3 patients at the next lower level.

Decision Rule Table (source https://i3design.shinyapps.io/i3plus3/) :

Following the completion of resection level escalation, a recommended maximum resection (MR) will be calculated using all available RLT data. The recommended MR will be calculated using an isotonic regression. Under this framework, the MR is the resection for which the isotonic estimate of the RLT rate is closest to the target RLT rate of 0.25. The recommended MR or a lower recommended level will be selected as the primary recommendation or resection.

As a secondary objective, if the MR is greater than the standard of care resection, (the DSMB recommendation is that a certain interval of CEST-EPI resection is safe) the MR will be used for a single-arm preliminary efficacy analysis compared to propensity matched historical controls of resection of IDH-WT GBM. If this expansion is carried out, we will carry out CEST-EPI resections at the MR in 16 patients using the same protocol and follow up as described above. Follow up will additionally include interval MRI as recommended by RANO 2.0 1 to calculated PFS and OS. Comparisions with historical controls will be carried out with cox regression analyses.

As additional secondary data, an objective measure of tumor burden will be evaluated. Tumor tissue will be collected from each resection. The secondary endpoint of this study is to determine the dependence of pH-sensitive amine CEST-EPI contrast on "active" tumor density. This will be measured through immunohistochemical staning of tumor (via H&E, anti-EGFR for EGFR amplified tumors), and proliferative tumor (anti-Ki67) which will be reviewed by blinded neuro-pathologist. Comaprisions of tumor burden across CEST-EPI and 5-ALA metrics will be compared with paired sample chi square tests and linear regression analysis. A study timeline is shown below.

5 CRITERIA FOR EVALUATION 5.1 Primary Efficacy Endpoint The primary endpoint for this study will be the safety of resecting CEST-EPI positive regions of glioblastoma.

5.2 Safety Evaluations This is an additional risk study and all safety evaluations will be reviewed as described below by JCCC DSMB and at multidisciplinary brain tumor board including neurosurgeons, neuro-oncologists, neuro-radiologists, and neuro-pathologists. Safety will be evaluated by "resection limited toxicities" (RLTs) and are pre-defined and are listed in detail below.