342O - Intracranial administration of CTLA-4 and PD-1 immune checkpoint blocking monoclonal antibodies in recurrent glioblastoma (rGB): A multi-cohort adaptive phase I clinical trial

Background

Intracerebral (iCE) administration (admin) of ipilimumab (IPI) and nivolumab (NIVO) plus IV admin of NIVO following resection of rGB was well tolerated and resulted in encouraging overall survival (OS)(Schwarze et al, JITC 2021). In 4 subsequent cohorts (A to D), the safety of intratumoral (iTU) admin of NIVO and IPI, followed by repeated intracavitary (iCA), or intrathecal (iTH) admin of NIVO +/- increasing doses of IPI was investigated.

Methods

Within 24h prior to surgery, NIVO (10mg) was administered IV in all patients (pts), followed by a maximal safe resection (B+C), or biopsy (A+D), followed by iCE (B+C) or iTU (A+D) admin of IPI (5 mg) plus NIVO (10 mg). At the end of surgery, 10 mg NIVO plus 1, 5 or 10 mg of IPI was administered iCA or iTH (: Ommaya reservoir) in C and D respectively. iCA and iTH admin of NIVO with/-out IPI, as well as IV admin of NIVO (10 mg) were repeated Q2w (max 24w). On-treatment CSF samples were used for cytology, chemical analysis and measurements of NIVO/IPI concentrations.

Results

39 pts (27 male) were enrolled (A: n=16, B: n=16, C: n=4, D: n=3; enrolment ongoing in C+D). At database lock, all pts in A+B were off study treatment. All pts received the predefined dose of iCE/iTU IPI/NIVO. The median of iCE/IV NIVO admin was 5 (1-12) in A and 4(1-12) in B. Most frequent AEs were fatigue (n=30), headache (n=19), confusion (n=14), dysphasia (n=13), fever (n=10), and bacterial colonization of the Ommaya reservoir (n=5). On-treatment neurological deterioration due to subacute symptomatic cerebral edema requiring corticosteroids occurred in 6 pts. There were no G5 AEs. irAEs were infrequent. In A+B, PD was diagnosed in all pts, and 27 pts have died. Median PFS and OS were numerically inferior for pts treated in A vs. B (median PFS 5w (95% CI 1-8) vs. 13w (95% CI 7-19); median OS 23w (95% CI 0-53) vs. 42w (95% CI 30-54); 6m-OS-rate 50% (95% CI 26-75) vs. 68.8% (95% CI 46-91)). An elevated protein level and lymphocytic pleocytosis were observed in >90% of CSF samples. There was no evidence for accumulation of NIVO/IPI in the CSF.

Conclusions

Intracranial admin of NIVO +/- IPI in pts with rGB was found to be feasible, safe, and associated with encouraging OS in pts amenable to resection.

Clinical trial identification

NCT03233152.

Editorial acknowledgement

Legal entity responsible for the study

Department of Medical Oncology, Universitair Ziekenhuis Brussel.

Funding

Has not received any funding.

Disclosure

B. Neyns: Financial Interests, Personal, Invited Speaker: Roche; Financial Interests, Personal, Invited Speaker: Bristol-Myers Squibb; Financial Interests, Personal, Invited Speaker: Merck Sharp & Dohme; Financial Interests, Personal, Invited Speaker: Novartis; Financial Interests, Personal, Invited Speaker: AstraZeneca; Financial Interests, Institutional, Funding: Pfizer; Financial Interests, Institutional, Funding: Novartis; Financial Interests, Institutional, Funding: Roche; Financial Interests, Institutional, Funding: Merck-Serono. J.K. Schwarze: Non-Financial Interests, Personal, Other: Amgen; Financial Interests, Personal, Invited Speaker: Novartis; Non-Financial Interests, Personal, Other: MSD Oncology. G. Awada: Financial Interests, Personal, Invited Speaker: Novartis; Financial Interests, Personal, Advisory Board: Novartis; Non-Financial Interests, Personal, Other: MSD Oncology; Non-Financial Interests, Personal, Other: Astellas Pharma; Non-Financial Interests, Personal, Other: Novartis; Non-Financial Interests, Personal, Other: Pfizer. All other authors have declared no conflicts of interest.