Phase 1b/2a study of galunisertib, a small molecule inhibitor of transforming growth factor-beta receptor I, in combination with standard temozolomide-based radiochemotherapy in patients with newly diagnosed malignant glioma

Summary Purpose Galunisertib, a TGF-β inhibitor, has demonstrated antitumor effects in preclinical and radiographic responses in some patients with malignant glioma. This Phase 1b/2a trial investigated the clinical benefit of combining galunisertib with temozolomide-based radiochemotherapy (TMZ/RTX) in patients with newly diagnosed malignant glioma (NCT01220271). Methods This is an open-label, 2-arm Phase 1b/2a study (N = 56) of galunisertib (intermittent dosing: 14 days on/14 days off per cycle of 28 days) in combination with TMZ/RTX (n = 40), versus a control arm (TMZ/RTX, n = 16). The primary objective of Phase 1b was to determine the safe and tolerable Phase 2 dose of galunisertib. The primary objective of Phase 2a was to confirm the tolerability and pharmacodynamic profile of galunisertib with TMZ/RTX, and the secondary objectives included determining the efficacy and pharmacokinetic (PK) profile of galunisertib with TMZ/RTX in patients with glioblastoma. This study also characterized the changes in the major T-cell subsets during TMZ/RTX plus galunisertib treatment. Results In the Phase 2a study, efficacy results for patients treated with galunisertib plus TMZ/RTX or TMZ/RTX were: median overall survival (18.2 vs 17.9 months), median progression-free survival (7.6 vs 11.5 months), and disease control rate (80% [32/40] vs 56% [9/16] patients) respectively. PK profile of galunisertib plus TMZ/RTX regimen was consistent with previously published PK data of galunisertib. The overall safety profile across treatment arms was comparable. Conclusion No differences in efficacy, safety or pharmacokinetic variables were observed between the two treatment arms. Electronic supplementary material The online version of this article (10.1007/s10637-020-00910-9) contains supplementary material, which is available to authorized users.


Immunohistochemistry
Tumor specimens from patients' original diagnostic tumor sample were formalin fixed, and then paraffin embedded to prepare 5-µm sections for staining. The most recent tumor specimens were used for the central pathology review. Tissue staining and examinations were performed in a blinded fashion. The pathological results were entered into a database for subsequent comparison to the clinical data. Tissue histological analysis was performed as previously described [1].
Briefly, Hematoxylin and Eosin (H&E) staining was performed to assess the general anatomical phenotype of tumor samples. The baseline expression of tissue biomarkers such as glial fibrillary acidic protein (GFAP), Ki67, CD3, phospho-SMAD2 (pSMAD2), and isocitrate dehydrogenase 1 (IDH1) R132H was evaluated by IHC staining and scoring method developed at the neuropathology laboratory at the University Clinic of Heidelberg, Germany.

Assessment of tumor response
Assessment of tumor response was based on Response Assessment in Neuro-Oncology (RANO) criteria [2]. The overall response and clinical benefit rates based on RANO criteria were estimated for each treatment arm by dividing the total number of confirmed responders (CR and PR), or patients experiencing benefit by the number of patients who received at least 1 dose of study treatment. A patient with a confirmed response, partial response (PR), or stable disease (SD) was considered to have received a benefit from the treatment.

Dose reductions
During the first 3 Cycles of treatment in Phase 2a, 6 patients from the galunisertib plus radiochemotherapy cohort had dose reductions in galunisertib due to AEs such as wound infection (2 patients; Grade 3), vomiting (Grade 1), constipation (Grade 2), hydrocephalus (Grade 3), seizures (Grade 3), and alanine aminotransferase increased (Grade 3) (data not shown). The most common AE causing a dose reduction in TMZ was a decrease in platelet count in both the galunisertib plus radiochemotherapy arm and the radiochemotherapy alone arm (data not shown).

Assessment of tumor response
Among the 40 patients treated with galunisertib plus radiochemotherapy, 3 patients had a best overall response (BOR) of complete response (CR), and none had partial response (PR) representing an overall response rate of 7.5% (90% CI: 2.1, 18.3%) (Supplemental Table 1

Pharmacokinetics (PK) for Phase 1b and 2a
The PK parameters of galunisertib from Phase 1b Cycles 1, 2, and 3 were determined by a standard noncompartmental methods of analysis using WinNonlin (supplemental Fig. 4a-c). The PK analysis was conducted on each patient who had completed at least 2 days of sampling.
Individual plasma concentration versus time profiles and individual and summary statistics of PK parameters for Phase 1b patients were carried out (Supplemental Fig. 4). To determine a potential difference in galunisertib effect in combination with radiochemotherapy (Cycle 1 and 2) or alone (Cycle 3), a PK non-compartmental analysis was performed for 18 patients (10 patients treated with 160 mg/day and 8 patients with 300 mg/day) in Phase 1b (Supplemental
Thrombocytopenia was attributed mainly to the TMZ administration, while not excluding some possible (not likely) association with LY2157299. There were no dose-limiting toxicities (DLT) and no clinically meaningful cardiotoxicities observed. None of the patients met the definition of DLT, per protocol and investigator/sponsor decision. Given that TMZ in combination with radiation carries a risk of approximately 19% grade 3/4 thrombocytopenia, and 13% grade 2 rash, the observed event rate in Phase 1b was not unexpectedly high [3,4].
The PK data of Phase 1b were consistent with the previously described PK analysis of galunisertib suggesting that the PK profile of galunisertib was not altered when combined with TMZ and radiation.
To conclude, based on overall toxicity and PK information from Phase 1b, the dose of 300 mg/day of galunisertib was selected for the Phase 2a part of the study.