A phase I trial on the intra- and post-operative intracranial administration of ipilimumab and nivolumab in patients with recurrent high-grade glioma.

Background: Intracerebral (iCer) administration (admin) of ipilimumab (IPI) and nivolumab (NIVO) plus IV NIVO following resection of recurrent high-grade glioma (rHGG) was well tolerated and showed encouraging overall survival (OS) (J. Duerinck et al. JITC 2021). The safety of additional postoperative (postop) bi-weekly intracavitary (iCav) admin of NIVO or NIVO + IPI (: first in human intracranial CTLA-4 blockade) was investigated in a phase I trial (3+3 design with cohort expansion). Methods: Within 24h prior to surgery, 10 mg NIVO IV was admin, followed by a maximal safe resection and injection of the brain tissue lining the resection cavity with 5 mg IPI + 10 mg NIVO, and positioning of a catheter in the resection cavity connected to an Ommaya reservoir. Only in patients (pts) receiving postop iCav IPI, 10 mg NIVO and 5 mg IPI were admin via the Ommaya at the end of surgery. Postop 1, 5, or 10 mg NIVO was admin iCav as a single agent. In subsequent pts, postop 10 mg iCav NIVO was combined with 1, 5 or 10 mg iCav IPI. All postop iCav admin were combined with NIVO 10 mg IV, and repeated Q2w (< 24w). On-treatment CSF samples were used for cytology, chemical analysis, and measurements of NIVO/IPI concentrations. Results: 43 pts (32 male) initiated treatment, all receiving the predefined pre- and intraop doses of IV and iCer IPI/NIVO. No unexpected AE related to the intraop treatment occurred. Postop treatment was initiated in 39 pts, all receiving 10 mg NIVO IV Q2w. Postop NIVO IV was combined with iCav NIVO 1, 5 or 10 mg in 3, 4, and 9 pts. The median number of postop IV/iCav NIVO admin was 7 (3-7), 4.5 (0-11), and 2 (1-11), resp. Next, postop iCav NIVO 10 mg was combined with iCav IPI 1, 5 or 10 mg in 10, 6, and 11 pts. The median number of postop IV/iCav NIVO + IPI admin was 3 (0-12), 4 (1-11), and 4 (0-12), resp. Dose limiting toxicity consisted of transient grade 3 aseptic neutrophilic pleocytosis with pyrexia and neurological deterioration in 1 and 3 pts treated with 5 and 10 mg IPI iCav, resp. Most frequent TRAEs were fatigue (n=24), headache (n=19), fever (n=17), and bacterial Ommaya colonization (n=11). No grade 5 AE occurred. At database lock, all pts were off study treatment, 1 pt stayed progression-free, and 5 were alive (mFU 80w (31-140)). OS compared favorably against a Belgian historical control cohort (469 pts; log rank p: 0.010) with an improved 1 and 2y OS rate (33 vs. 18.6% and 11.7 vs. 5.7%, resp.). Adding iCav IPI postop did not significantly alter PFS or OS. There was an elevated protein level and lymphocytic pleocytosis in >90% of CSF samples and no evidence for NIVO accumulation in the CSF (IPI under evaluation). Conclusions: In this first in human phase I trial on intracranial CTLA-4/PD-1 blockade in pts with rHGG amenable for resection, intraop iCer and postop iCav admin of NIVO+/- IPI was found to be feasible and safe up to a bi-weekly postop iCav dose of 1 mg IPI + 10 mg NIVO; with encouraging OS results.