Hypofractionated Stereotactic Re-irradiation and Anti-PDL1 Durvalumab Combination in Recurrent Glioblastoma: STERIMGLI Phase I Results

Abstract

Background: Hypofractionated stereotactic radiotherapy (hFSRT) is a salvage option for recurrent glioblastoma (GB) which may synergize anti-PDL1 treatment. This phase I study evaluated the safety and the recommended phase II dose of anti-PDL1 durvalumab combined with hFSRT in patients with recurrent GB.

Methods: Patients were treated with 24 Gy, 8 Gy per fraction on days 1, 3, and 5 combined with the first 1500 mg Durvalumab dose on day 5, followed by infusions q4weeks until progression or for a maximum of 12 months. A standard 3 + 3 Durvalumab dose de-escalation design was used. Longitudinal lymphocytes count, cytokines analyses on plasma samples, and magnetic resonance imaging (MRI) were collected.

Results: Six patients were included. One dose limiting toxicity, an immune-related grade 3 vestibular neuritis related to Durvalumab, was reported. Median progression-free interval (PFI) and overall survival (OS) were 2.3 and 16.7 months, respectively. Multi-modal deep learning-based analysis including MRI, cytokines, and lymphocytes/neutrophil ratio isolated the patients presenting pseudoprogression, the longest PFI and those with the longest OS, but statistical significance cannot be established considering phase I data only.

Conclusion: Combination of hFSRT and Durvalumab in recurrent GB was well tolerated in this phase I study. These encouraging results led to an ongoing randomized phase II. (ClinicalTrials.gov Identifier: NCT02866747).

Keywords: Durvalumab; deep learning; hypofractionated stereotactic re-irradiation; phase I clinical trial; recurrent glioblastoma.

Figure 1.

(A) Swimmer plot describing the follow-up of each patient. Duration of treatment (hFSRT and Durvalumab infusion), duration of steroid treatment, and clinical outcome (PR: partial response; PD: progression disease). (B) Lymphocytes count follow-up for each patient.

Figure 2.

Study design. Patients began Durvalumab on day 5 of the hypofractionated stereotactic radiotherapy and then monthly for up to 12 months. hFSRT: hypofractionated stereotactic radiotherapy; PD-L1: programmed death ligand-1; MRI: magnetic resonance imaging.

Figure 3.

(A) Deep learning-based analysis data preparation. (B) Unsupervised contrastive learning using FLAIR and Gd-T1 MRI tumor slices. A triplet loss was used to compare the 64 features vectors Z1 FLAIR (anchor), Z1 Gd-T1 (positive sample), and Z2 FLAIR (negative sample). Note that Z2 FLAIR came either from another patient or from the same patient but at a different date.

Figure 4.

(A) MRI data only, diagram for 3 clusters. Each dot represents one patient (at different visits). Patients #02 and #05 with the longest overall survivals were clearly separated from the rest of the cohort. (B) Multimodal feature vectors (MRI; cytokine level; neutrophil/lymphocyte ratio) for all patients. Diagram for 6 clusters. Each dot represents one patient denoted by the first number and the visit number. Each color corresponds to a cluster as computed using k-means clustering. Patient #04’s second visit was also clustered with patient #05’s intermediate visits. Patient #04 also showed significant pseudoprogression.

Figure 5.

2D UMAP representations of multimodal feature vectors (MRI; cytokine level; neutrophil/lymphocyte ratio) for all patients. Each dot represents one patient denoted by the first number and the visit number. Each color corresponds to a cluster as computed using k-means clustering. Patient #05’s (longest PFI) visits were all clustered (A1). Increasing the number of clusters to 3 leads to a clustering of patient #01’s visits (A2). Patients #01, #03, and #05’s visits separated into distinct clusters (B1), ranging from the shortest to the longest PFI. At 5 clusters (B2), patient #05’s intermediate visits (5.2; 5.3), during which clinical pseudoprogressions were identified, segregated into a distinct cluster.