Concurrent immunotherapy and re‐irradiation utilizing stereotactic body radiotherapy for recurrent high‐grade gliomas

Abstract Background Clinical trials evaluating immune checkpoint inhibition (ICI) in recurrent high‐grade gliomas (rHGG) report 7%–20% 6‐month progression‐free survival (PFS), while re‐irradiation demonstrates 28%–39% 6‐month PFS. Aims We evaluate outcomes of patients treated with ICI and concurrent re‐irradiation utilizing stereotactic body radiotherapy/fractionated stereotactic radiosurgery (SBRT) compared to ICI monotherapy. Methods and Results Patients ≥18‐years‐old with rHGG (WHO grade III and IV) receiving ICI + SBRT or ICI monotherapy between January 1, 2016 and January 1, 2019 were included. Adverse events, 6‐month PFS and overall survival (OS) were assessed. Log‐rank tests were used to evaluate PFS and OS. Histogram analyses of apparent diffusion coefficient maps and dynamic contrast‐enhanced magnetic resonance perfusion metrics were performed. Twenty‐one patients with rHGG (ICI + SBRT: 16; ICI: 5) were included. The ICI + SBRT and ICI groups received a mean 7.25 and 6.2 ICI cycles, respectively. There were five grade 1, one grade 2 and no grade 3–5 AEs in the ICI + SBRT group, and four grade 1 and no grade 2–5 AEs in the ICI group. Median PFS was 2.85 and 1 month for the ICI + SBRT and ICI groups; median OS was 7 and 6 months among ICI + SBRT and ICI groups, respectively. There were significant differences in pre and posttreatment tumor volume in the cohort (12.35 vs. 20.51; p = .03), but not between treatment groups. Conclusions In this heavily pretreated cohort, ICI with re‐irradiation utilizing SBRT was well tolerated. Prospective studies are warranted to evaluate potential therapeutic benefits to re‐irradiation with ICI + SBRT in rHGG.

Immune checkpoint inhibitors (ICIs) provided impressive results in melanoma, nonsmall cell lung cancer, 12 and in untreated brain metastases secondary to these malignancies. 13 While several clinical trials evaluating ICI in HGGs are ongoing, results are disappointing to date.
Prior studies evaluating salvage re-irradiation report 28% to 39% 6-month progression-free survival (PFS) rates, with a 26% median 1-year overall survival (OS) rate. 11,[14][15][16][17][18] Radiotherapy (RT) may improve ICI efficacy through several mechanisms, including altering tumor cell surface proteins, and enhancing the availability and assortment of intracellular peptide pools. These effects, in conjunction with inducing MHC class I expression, provides a larger repertoire of antigenic targets to elicit an immune response. RT induces major histocompatibility complex (MHC) class I expression through upregulating interferon-γ, promoting T cell recruitment, 19 conferring increased survival compared with either modality alone in mouse models. 20 Additionally, stereotactic body radiotherapy / fractionated stereotactic radiosurgery (SBRT), entailing conformally delivering higher RT doses in fewer treatments, may be preferable over conventional RT delivered over several weeks with regard to augmenting immune responses, 21 while minimizing the impact on circulating lymphocytes.
Combining ICI with concurrent RT may increase the therapeutic ratio in rHGG, however, there are no prospective studies evaluating toxicities and outcomes of concurrent re-irradiation with ICI + SBRT.
We report treatment-related adverse events (AE) in patients with rHGG treated with concurrent ICI + SBRT. PFS, OS and changes in tumor volume and perfusion characteristics after treatment were also evaluated.

| Patient selection
This study was an institutional review board approved retrospective review (19- Figure 1). Patients were excluded if they did not receive ICI monotherapy, ICI concurrently with SBRT (n = 333), or if they had outside imaging during treatment that was not available for analysis (n = 2), leaving 21 patients for analysis.

| Treatment
Patients treated with nivolumab received intravenous infusions of 3 mg/kg, administered on days 1 and 14 of a 28-day cycle. Patients treated with pembrolizumab received intravenous infusions of 2 mg/kg, administered on days 1 and 21 of a 21-day cycle. A subsequent ICI cycle was delayed until recovery of grade 3 or higher hematologic or grade 2 or higher nonhematologic toxicities.
RT planning volumes were created by a radiation oncologist and neurosurgeon. Dose and fractionation were determined on the basis of lesion size, prior radiotherapy, and meeting dose constraints for adjacent organs at risk. For treatment planning, high-resolution 1 mm magnetic resonance T1 sequences with contrast were rigidly fused to CT simulation scans. Treatment was delivered using volumetric modulated arc therapy (VMAT) of 3-4 noncoplanar arcs and either 6Â or 10Â flattening filter-free beams ensuring 95% of the planning target volume received the prescribed dose. RT plans were generated using Eclipse v15.6 (Varian Medical Systems, Palo Alto, CA) with AAA or AcurosXB planning algorithms. Organ at risk constraints for SBRT plans adhered to TG101 guidelines. RT was delivered on a Novalis (BrainLab, Munich, Germany) Truebeam STX linac (Varian Medical Systems, Palo Alto, CA), with multileaf collimator leaf width of 2.5 mm.

| Imaging
All patients underwent brain MRI on 1.5 or 3 Tesla systems (Skyra, Aera, Biograph mMR, Siemens Healthcare; Discovery 750 w, Signa HDxt, GE Healthcare, Milwaukee, WI), pre and posttreatment per institutional standards as previously described. 22 Olea Medical 3.0 software (La Ciotat, France) was used for DCE perfusion MRI processing and histogram analysis. The volumes-ofinterest encompassing all voxels with enhancing tumor, pre and posttreatment were including in histogram analysis to produce bloodbrain barrier permeability metrics, including median, mean, and 90% of the plasma volume (Vp) and volume transfer constant (Ktrans). Diffusion metrics, including median, mean, and 10% of the ADC were also evaluated. All values were normalized utilizing the contralateral normal white matter.

| Evaluation
Routine testing included weekly laboratory tests including complete blood counts and basic metabolic panel, examinations at every clinical visit, and contrast-enhanced brain MRI every 4 weeks.
Neuroradiologic response following treatment was determined by response assessment in neuro-oncology (RANO) criteria. 23  PFS and OS were defined as time from the day one of ICI until disease progression and death, respectively. This calculation assumes PFS/OS proportion estimates of 50% to conservatively maximize the width of the obtained confidence intervals.

| Adverse events
Among the ICI + SBRT cohort, there were four instances of grade 1 fatigue and one instance of grade 1 thrombocytopenia (Table 3).
There was one instance of grade 2 fatigue, two instances each of grade 1 fatigue and grade 1 constipation among the ICI monotherapy cohort. There were no grade 3-5 AE, radiographic findings consistent with radiation necrosis on follow-up imaging, or treatment-related deaths in either cohort. No patients discontinued ICI due to toxicity.

| Response
Among the ICI + SBRT cohort, there were no CR, 2 PR, 8 SD and 6 PD at one month ( Figure 1).

| DISCUSSION
There is a paucity of therapeutic options and no validated standard of care for rHGG. While awaiting results of ongoing prospective trials ( One putative explanation for suboptimal immune response is the high rate of lymphopenia observed in HGG patients, with one group reporting that T cells are available in this population but are sequestered in the bone marrow. 29 The aforementioned studies provide a rationale for optimizing immunotherapeutic efficacy through its implementation in an immunologically favorable setting, such as priming the immune system to tumor-specific antigens. RT may improve ICI effects by increasing the availability and diversity of intracellular peptides, increasing MHC class I expression, and promoting T cell recruitment and infiltration. 20,21,30,31 Technological advancements in the delivery of SBRT allow for highly conformal treatments that substantially reduce AE associated with re-irradiation in other disease sites. 16 Several studies show function status improvements and decreased reliance on corticosteroid following SBRT monotherapy with a low risk of late central nervous system toxicity. 14,15,17,18,32 Additionally, SBRT dosefractionation schemes may be more effective than conventionally fractionated RT with regard to augmenting immune responses. 21,33 This option also allows RT completion within one to five treatments, which is convenient for patients.
A closer look at this cohort notes several limitations that could be considered in future studies geared towards optimizing a response. The optimal treatment approach for patients with rHGG continues to be an area of ongoing investigation. This small retrospective study suggests ICI can be safely given concurrently with re-irradiation using SBRT for patients with rHGG. These initial findings support evaluating whether optimizing conditions for combinatory ICI + SBRT approaches may lead to favorable clinical responses, or whether  attention should be turned to other therapeutic avenues to address this unmet need in neuro-oncology.