Abstract
Background
Glioblastoma relapse is associated with activation of phosphatidylinositol 3-kinase (PI3K) signalling pathway. In preclinical studies, the pan-PI3K inhibitor buparlisib showed antitumour activity in glioma models.
Methods
This was a two-part, multicentre, phase Ib/II study in patients with recurrent glioblastoma pretreated with radiotherapy and temozolomide standard of care. Patients received buparlisib (80 mg or 100 mg once daily) plus carboplatin (area under the curve (AUC)=5 every 3 weeks), or buparlisib (60 mg once daily) plus lomustine (100 mg/m2 every 6 weeks). The primary endpoint was to determine the maximum tolerable dose (MTD) and/or recommended phase II dose of buparlisib plus carboplatin or lomustine.
Results
Between 28 February 2014 and 7 July 2016, 35 patients were enrolled and treated with buparlisib plus carboplatin (n=17; buparlisib (80 mg) plus carboplatin, n=3; and buparlisib (100 mg) plus carboplatin, n=14), or buparlisib (60 mg) plus lomustine (n=18). The MTD of buparlisib was determined to be 100 mg per day in combination with carboplatin at an AUC of 5 every 3 weeks. The MTD of buparlisib in combination with lomustine could not be determined as it did not satisfy the MTD criteria per the Bayesian logistic regression model.
Conclusion
The overall safety profile of buparlisib remained unchanged, and no new or unexpected safety findings were reported in this study. Preliminary assessment for both combinations did not demonstrate sufficient antitumour activity compared with historical data on single-agent carboplatin or lomustine.
Trial registration number
Keywords: buparlisib, BKM120, recurrent glioblastoma, rGBM
Key questions.
What is already known about this subject?
Glioblastoma (GBM) is the most common and most aggressive malignant primary brain tumor in adults, with poor survival rates. Bevacizumab, an anti–vascular endothelial growth factor antibody, has improved progression-free survival in recurrent GBM (rGBM), but without any prolongation of overall survival.
A high unmet medical need in rGBM treatment still remains, and the molecular basis of the recurrence process in GBM is still poorly understood.
What does this study add?
Here, we present results of the phase Ib/II, open-label, multicenter, randomized study of buparlisib plus carboplatin or lomustine in patients with recurrent glioblastoma.
Preliminary assessment for both combinations did not demonstrate sufficient anti-tumor activity compared with historical data on single-agent carboplatin or lomustine.
How might this impact on clinical practice?
The modest outcomes observed in the current study are consistent with those reported for rGBM and further highlight the challenges of treating rGBM.
Introduction
Glioblastoma (GBM) is the most common and most aggressive malignant primary brain tumour in adults, with poor survival rates.1–5 The current standard of care (SoC) for patients with newly diagnosed GBM includes tumour resection followed by radiotherapy (RT) and chemotherapy (CT; temozolomide (TMZ)).6–8
GBM has an unfavourable prognosis mainly due to its high propensity for tumour recurrence, with a median survival of 12–15 months.9 10 Recurrence is common, with 75% of patients with GBM experiencing disease progression within 2 years of diagnosis and less than 10% surviving for 5 years after diagnosis.1 2 11 12 Bevacizumab, an antivascular endothelial growth factor antibody, has improved progression-free survival (PFS) in recurrent GBM (rGBM), but without any prolongation of overall survival.13 A high unmet medical need in rGBM treatment remains, and the molecular basis of the recurrence process in GBM is still poorly understood.14 Preclinical data suggest that activation of the phosphatidylinositol 3-kinase (PI3K) signalling pathway is one of the key factors contributing to GBM relapse.15 The PI3K pathway was found to be frequently altered in GBM, with up to 90% of GBM tumours having an activated PI3K pathway.16–18 Therefore, GBM represents a disease with a compelling biological rationale for treatment with PI3K inhibitors. We hypothesise that combining chemotherapeutic agents used for GBM treatment with a PI3K inhibitor may confer a clinical benefit to patients with rGBM.
Buparlisib is a potent and highly specific oral pan-class I PI3K inhibitor of all class 1 isoforms.19 Buparlisib has also been shown to cross the blood–brain barrier, accumulate in the brain tissue of non-tumour-bearing rats, and efficiently downregulate tissue phospho-S6 and phospho-AKT.20 Buparlisib has also shown preclinical efficacy in various PI3K pathway-hyperactivated cancer models, including GBM.21–24
Here, we report the maximum tolerated dose (MTD) and the recommended phase II dose (RP2D) of buparlisib in combination with carboplatin or lomustine and the safety and preliminary antitumour activity of these combinations in patients with rGBM.
Patients and methods
Study design and participants
This was a two-part, multicentre, phase Ib/II study in patients with rGBM pretreated with RT and TMZ SoC. In the phase Ib part of the study, approximately 15–22 evaluable patients per treatment arm were planned to be enrolled to determine the MTD and/or RP2D of oral buparlisib administered once daily in combination with carboplatin or with lomustine based on dose-limiting toxicities (DLTs) using a Bayesian logistic regression model (BLRM) with overdose control (figure 1).
Figure 1.
Study design. MTD, maximum tolerated dose; qd, once daily; q3w, once every 3 weeks; q6w, once every 6 weeks; RP2D, recommended phase II dose.
Eligible patients were aged 18 years or older with rGBM pretreated with RT and TMZ SoC. Patients must have had at least one measurable and/or non-measurable lesion as per the Response Assessment in Neuro-Oncology (RANO) criteria,25 must have recovered (to grade ≤1) from all clinically significant toxicities related to prior antineoplastic therapies, must have had a Karnofsky Performance Status (KPS) ≥70%, and must have had adequate organ and bone marrow functions. PI3K expression status was not an inclusion criterion and testing isocitrate dehydrogenase (IDH) mutations or O6-methylguanine-DNA methyltransferase (MGMT) methylation status was not planned.
Patients were excluded if they had received previous treatment with PI3K pathway inhibitors: lomustine or carboplatin for newly diagnosed GBM or rGBM, or antineoplastic treatment for rGBM. Patients who had received more than one line of cytotoxic CT, patients with a history of or having an active cardiac disease, patients with grade ≥3 anxiety, patients with a Generalized Anxiety Disorder-7 (GAD-7) mood scale score ≥15, and patients with a score ≥12 on the Patient Health Questionnaire-9 (PHQ-9) were not permitted.
Written informed consent was obtained from all patients.
Outcomes
The primary endpoint in phase Ib (dose escalation part) was to determine the MTD and/or RP2D of buparlisib plus carboplatin/lomustine combinations in patients with rGBM whose disease had progressed after SoC (RT with TMZ in combination and adjuvant), regardless of PI3K pathway activation status. The secondary endpoints were to evaluate the safety and tolerability of buparlisib plus carboplatin/lomustine combinations and to assess the preliminary antitumour activity of buparlisib plus carboplatin/lomustine combinations in patients with rGBM whose disease had progressed after SoC, regardless of PI3K pathway activation status.
Assessments
Contrast MRI scans and clinical factors were evaluated based on investigator assessment at baseline within 28 days before the start of treatment and subsequently every 6 weeks from treatment start until progression of disease or until start of another antineoplastic treatment or death, whichever occurred earlier. Tumour response and progression were assessed using the RANO working group-updated response assessment criteria for high-grade gliomas.25
Safety was monitored by occurrence of adverse events (AEs), physical examination, evaluation of vital signs, weight, KPS, ECG and cardiac imaging, and laboratory evaluations including glucose monitoring. Patient-rated mood was assessed using GAD-7 and PHQ-9.
Statistical considerations
Determination of the MTD/RP2D of each combination treatment was based on a synthesis of all relevant data available from all dose levels evaluated continuously during the phase Ib part, including safety information, data on DLTs, and all Common Terminology Criteria for Adverse Events grade ≥2 toxicity data during cycle 1 from evaluable patients.
The full analysis set consisted of all patients who received at least one dose of study treatment (buparlisib and/or carboplatin for the buparlisib plus carboplatin arm; buparlisib and/or lomustine for the buparlisib plus lomustine arm). The safety set (SS) included all patients who had received at least one dose of study treatment and had at least one postbaseline safety assessment. All patients from the SS who either met the minimum exposure criteria and had sufficient safety evaluations or had experienced a DLT during the first cycle were included in the dose-determining set and contributed to dose escalation decisions.
For the buparlisib plus carboplatin phase Ib arm, the minimum exposure criteria during the first cycle were at least 16 of the 21 full daily planned doses of buparlisib and the full carboplatin dose of 5 AUC (area under the curve) for 21 days. For the buparlisib plus lomustine phase Ib arm, the minimum exposure criteria during the first cycle were at least 32 of the 42 full daily planned doses of buparlisib and at least one dose of 100 mg/m2 of lomustine.
An adaptive BLRM dose escalation with overdose control was used to guide the recommended dose for the next cohort of patients. MTD and/or RP2D was defined as the highest drug dosage that did not cause, with a posterior probability greater than 25%, a DLT in more than 35% of the treated patients during the first cycle of treatment.
Results
Patient characteristics
Between 28 February 2014 and 7 July 2016, 35 patients were enrolled and treated with either the combination of buparlisib plus carboplatin (n=17; buparlisib (80 mg) plus carboplatin (AUC 5), n=3; and buparlisib (100 mg) plus carboplatin (AUC 5), n=14), or buparlisib plus lomustine (n=18; buparlisib (60 mg) plus lomustine 100 mg/m2). The primary reason for end of treatment in both the buparlisib plus carboplatin arm and the buparlisib plus lomustine arm was disease progression (88.2% and 77.8%, respectively) (table 1).
Table 1.
Patient disposition by treatment arm
| Disposition reason, n (%) | Buparlisib+carboplatin | Buparlisib (60 mg) + lomustine | ||
| Buparlisib (80 mg)+carboplatin | Buparlisib (100 mg)+carboplatin | All patients | ||
| Patients enrolled, n | 3 | 14 | 17 | 18 |
| Treated | 3 (100) | 14 (100) | 17 (100) | 18 (100) |
| Patients treated | ||||
| End of treatment | 3 (100) | 14 (100) | 17 (100) | 18 (100) |
| Primary reason for end of treatment | ||||
| Adverse event | 0 | 2 (14.3) | 2 (11.8) | 3 (16.7) |
| Death | – | – | – | 1 (5.6) |
| Progressive disease | 3 (100) | 12 (85.7) | 15 (88.2) | 14 (77.8) |
| Post-treatment follow-up after end of treatment | ||||
| Not applicable* | 3 (100) | 12 (85.7) | 15 (88.2) | 15 (83.3) |
| Patients no longer being followed for post-treatment follow-up | 0 | 2 (14.3) | 2 (11.8) | 3 (16.7) |
| Primary reason for discontinuation of post-treatment follow-up | ||||
| Progressive disease | 0 | 2 (14.3) | 2 (11.8) | 1 (5.6) |
| Death | – | 1† (7.1) | 1 (5.9) | 1 (5.6) |
| Withdrawal of consent | – | – | – | 1 (5.6) |
Percentage is based on ‘n’.
*Patients who withdrew consent from the study, died or decided not to attend the post-treatment efficacy follow-up at the end of treatment evaluation.
†Deaths occurred up to 30 days after last dose of study treatment.
The median age of the patients was 55 years in the buparlisib plus carboplatin arm and 58 years in the buparlisib plus lomustine arm. The majority of the patients in both arms were Caucasians, reflecting the countries that participated in this study. Overall, 76.5% of patients in the buparlisib plus carboplatin arm vs 83.3% of patients in the buparlisib plus lomustine arm had a KPS ≥80%. At baseline, 58.8% of patients in the buparlisib plus carboplatin arm and 50% of patients in the buparlisib plus lomustine arm had only T1 target lesions. In both arms, majority of the patients had their first recurrence/progression within 12 months of initial diagnosis. Time since the most recent relapse/progression was less than 3 months in 94.1% of patients in the buparlisib plus carboplatin arm and 94.4% of patients in the buparlisib plus lomustine arm (table 2).
Table 2.
Patient demographics and other baseline characteristics by treatment arm
| Characteristics | Buparlisib (80 mg)+carboplatin (n=3) | Buparlisib (100 mg)+carboplatin (n=14) | All patients (buparlisib+carboplatin) (n=17) | Buparlisib (60 mg)+lomustine (n=18) |
| Age, median (range), years | 54.0 (35.0–55.0) | 57.0 (29.0–67.0) | 55.0 (29.0–67.0) | 58.0 (35.0–73.0) |
| <65 years, n (%) | 3 (100) | 11 (78.6) | 14 (82.4) | 13 (72.2) |
| ≥65 years, n (%) | 0 | 3 (21.4) | 3 (17.6) | 5 (27.8) |
| Sex, n (%) | ||||
| Male | 3 (100) | 10 (71.4) | 13 (76.5) | 13 (72.2) |
| Female | 0 | 4 (28.6) | 4 (23.5) | 5 (27.8) |
| Race, n (%) | ||||
| Caucasian | 2 (66.7) | 12 (85.7) | 14 (82.4) | 14 (77.8) |
| Asian | 1 (33.3) | 0 | 1 (5.9) | 1 (5.6) |
| Other | 0 | 1 (7.1) | 1 (5.9) | 1 (5.6) |
| Unknown | 0 | 1 (7.1) | 1 (5.9) | 2 (11.1) |
| Karnofsky Performance Status, n (%) | ||||
| 100 | 1 (33.3) | 2 (14.3) | 3 (17.6) | 1 (5.6) |
| 90 | 1 (33.3) | 2 (14.3) | 3 (17.6) | 12 (66.7) |
| 80 | 1 (33.3) | 6 (42.9) | 7 (41.2) | 2 (11.1) |
| 70 | 0 | 4 (28.6) | 4 (23.5) | 3 (16.7) |
| Primary site of cancer, n (%) | ||||
| Brain | 3 (100) | 12 (85.7) | 15 (88.2) | 17 (94.4) |
| CNS: supratentorial | 0 | 2 (14.3) | 2 (11.8) | 1 (5.6) |
| Type of lesions at baseline, n (%) | ||||
| T1 target only | 0 | 10 (71.4) | 10 (58.8) | 9 (50.0) |
| Non-target T1/T2/FLAIR only | 0 | 0 | 0 | 2 (11.1) |
| Both | 3 (100) | 4 (28.6) | 7 (41.2) | 7 (38.9) |
| Time from initial diagnosis to first recurrence/progression, n (%) | ||||
| <6 months | 1 (33.3) | 3 (21.4) | 4 (23.5) | 2 (11.1) |
| 6 to <12 months | 2 (66.7) | 5 (35.7) | 7 (41.2) | 12 (66.7) |
| 12 to <24 months | 0 | 4 (28.6) | 4 (23.5) | 1 (5.6) |
| ≥24 months | 0 | 2 (14.3) | 2 (11.8) | 3 (16.7) |
| Time since most recent relapse/progression, n (%) | ||||
| <3 months | 3 (100) | 13 (92.9) | 16 (94.1) | 17 (94.4) |
| 3 to <6 months | 0 | 1 (7.1) | 1 (5.9) | 1 (5.6) |
CNS, central nervous system; FLAIR, fluid attenuated inversion recovery.
Efficacy and MTD
The starting dose of buparlisib was 80 mg once daily in combination with carboplatin (at an AUC of 5 every 3 weeks); the starting dose of buparlisib was reduced to 60 mg once daily for potential increased myelotoxicity in combination with lomustine (at 100 mg/m2 every 6 weeks).
The MTD of buparlisib in combination with carboplatin at an AUC of 5 every 3 weeks was confirmed to be 100 mg per day. No dose of buparlisib in combination with lomustine arm satisfied the criteria for confirmation of MTD.
Assessment for both combinations (buparlisib plus carboplatin or lomustine) demonstrated no enough antitumour activity compared with historical data with single-agent carboplatin or lomustine. The median PFS was 1.4 months (95% CI 1.1 to 1.6) in the buparlisib plus carboplatin arm and 1.3 months (95% CI 1.2 to 1.4) in the buparlisib plus lomustine arm. In the 80 mg buparlisib dose level plus carboplatin, the overall response rate and disease control rate (DCR) were 33.3% (95% CI 0.8 to 90.6) each; one patient had a long-lasting partial response (PR; 15.1 months). In the 100 mg buparlisib dose level plus carboplatin, the DCR was 21.4% (95% CI 4.7 to 50.8); no patient had a complete response (CR) or PR. In the 60 mg buparlisib dose level plus lomustine, the DCR was 11.1% (95% CI 1.4 to 34.7) and no patient had a CR or PR (table 3).
Table 3.
Best overall response as per local investigator assessment per RANO criteria, by treatment
| Buparlisib (80 mg)+carboplatin (n=3) | Buparlisib (100 mg)+carboplatin (n=14) | Buparlisib (60 mg)+lomustine (n=18) | |
| Patients with measurable enhancing T1 lesion at baseline, n (%) | 3 (100) | 14 (100) | 16 (88.9) |
| Patients with non-measurable, non-enhancing T2/FLAIR lesion at baseline, n (%) | 2 (66.7) | 3 (21.4) | 6 (33.3) |
| Best overall response, n (%) | |||
| Complete response (CR) | 0 | 0 | 0 |
| Partial response (PR) | 1 (33.3) | 0 | 0 |
| Stable disease (SD) | 0 | 3 (21.4) | 2 (11.1) |
| Progressive disease (PD) | 2 (66.7) | 11 (78.6) | 14 (77.8) |
| Unknown | 0 | 0 | 2 (11.1) |
| Overall response rate (ORR)*, n (%) (95% CI) | 1 (33.3) (0.8 to 90.6) | 0 (0.0 to 23.2) | 0 (0.0 to 18.5) |
| Disease control rate (DCR)†, n (%) (95% CI) | 1 (33.3) (0.8 to 90.6) | 3 (21.4) (4.7 to 50.8) | 2 (11.1) (1.4 to 34.7) |
*ORR includes CR+PR
†DCR includes CR+PR+SD.
FLAIR, fluid attenuated inversion recovery; RANO, Response Assessment in Neuro-Oncology.
Preliminary assessment indicated no enough antitumour activity compared with historical data with single-agent carboplatin or lomustine. Based on the overall challenging safety profile and no enough antitumour activity observed in this phase of the study, no additional patients were enrolled and the phase II part of the study was not conducted.
Safety
The median duration of exposure to study treatment was 47 days (range: 42–798 days) in the 80 mg buparlisib dose level plus carboplatin and 42 days (range: 21–259 days) in the 100 mg buparlisib dose level plus carboplatin. The median duration of exposure to study treatment was 42 days (range: 15–231 days) in the 60 mg buparlisib plus lomustine arm.
In the buparlisib plus carboplatin arm, DLTs were reported in three patients, all in the 100 mg buparlisib dose level, including one event of decreased neutrophil count (grade 4), one event of anxiety disorder (grade 4) and one event of suicidal ideation (grade 2). In the buparlisib plus lomustine arm, five patients reported DLTs, including one event of thrombocytopaenia (grade 4), one event of fatigue (grade 3), one event of decreased platelet count (grade 4), two events of depression (one grade 2 and one grade 3) and one event of pneumonitis (grade 3).
The most common any-grade AEs in the buparlisib plus carboplatin arm were headache (52.9%), fatigue and nausea (47.1% each), and thrombocytopaenia (35.3%). The most common grade 3/4 AEs were decreased neutrophil count and decreased platelet count (17.6% each) and hypokalaemia and lymphopaenia (11.8% each). In the buparlisib plus lomustine arm, the most common any-grade AEs were fatigue, nausea, and decreased platelet count (44.4% each) and anaemia (33.3%). The most common grade 3/4 AEs were thrombocytopaenia (22.2%) and increased gamma-glutamyltransferase, hypokalaemia, decreased lymphocyte count, neutropaenia, decreased platelet count and decreased white cell count (11.1% each) (table 4).
Table 4.
Most frequent on-treatment adverse events, regardless of study treatment relationship (all grade incidence ≥15% in any arm)
| Buparlisib+carboplatin arm | ||||||
| Preferred term, n (%) | Buparlisib (80 mg)+carboplatin (n=3) | Buparlisib (100 mg)+carboplatin (n=14) | All patients (buparlisib+carboplatin) (n=17) | |||
| All grades | Grades 3/4 | All grades | Grades 3/4 | All grades | Grades 3/4 | |
| Headache | 2 (66.7) | 0 | 7 (50.0) | 0 | 9 (52.9) | 0 |
| Fatigue | 3 (100) | 0 | 5 (35.7) | 1 (7.1) | 8 (47.1) | 1 (5.9) |
| Nausea | 2 (66.7) | 0 | 6 (42.9) | 0 | 8 (47.1) | 0 |
| Thrombocytopaenia | 0 | 0 | 6 (42.9) | 0 | 6 (35.3) | 0 |
| Constipation | 1 (33.3) | 0 | 4 (28.6) | 0 | 5 (29.4) | 0 |
| Depression | 1 (33.3) | 0 | 4 (28.6) | 1 (7.1) | 5 (29.4) | 1 (5.9) |
| Diarrhoea | 0 | 0 | 5 (35.7) | 0 | 5 (29.4) | 0 |
| Insomnia | 2 (66.7) | 0 | 3 (21.4) | 0 | 5 (29.4) | 0 |
| Neutropaenia | 0 | 0 | 5 (35.7) | 1 (7.1) | 5 (29.4) | 1 (5.9) |
| Decreased platelet count | 1 (33.3) | 0 | 4 (28.6) | 3 (21.4) | 5 (29.4) | 3 (17.6) |
| Decreased appetite | 0 | 0 | 4 (28.6) | 0 | 4 (23.5) | 0 |
| Hyperglycaemia | 1 (33.3) | 0 | 3 (21.4) | 0 | 4 (23.5) | 0 |
| Hypertension | 1 (33.3) | 0 | 3 (21.4) | 1 (7.1) | 4 (23.5) | 1 (5.9) |
| Decreased neutrophil count | 1 (33.3) | 1 (33.3) | 3 (21.4) | 2 (14.3) | 4 (23.5) | 3 (17.6) |
| Vomiting | 1 (33.3) | 0 | 3 (21.4) | 0 | 4 (23.5) | 0 |
| Decreased white cell count | 1 (33.3) | 0 | 3 (21.4) | 1 (7.1) | 4 (23.5) | 1 (5.9) |
| Anxiety | 1 (33.3) | 0 | 2 (14.3) | 0 | 3 (17.6) | 0 |
| Hiccups | 1 (33.3) | 0 | 2 (14.3) | 0 | 3 (17.6) | 0 |
| Pruritus | 2 (66.7) | 0 | 1 (7.1) | 0 | 3 (17.6) | 0 |
| Somnolence | 1 (33.3) | 0 | 2 (14.3) | 0 | 3 (17.6) | 0 |
| Dehydration | 1 (33.3) | 1 (33.3) | 1 (7.1) | 0 | 2 (11.8) | 1 (5.9) |
| Hypokalaemia | 0 | 0 | 2 (14.3) | 2 (14.3) | 2 (11.8) | 2 (11.8) |
| Lymphopaenia | 0 | 0 | 2 (14.3) | 2 (14.3) | 2 (11.8) | 2 (11.8) |
| Myalgia | 1 (33.3) | 0 | 1 (7.1) | 0 | 2 (11.8) | 0 |
| Seizure | 1 (33.3) | 0 | 1 (7.1) | 1 (7.1) | 2 (11.8) | 1 (5.9) |
| Urinary tract infection | 1 (33.3) | 0 | 1 (7.1) | 0 | 2 (11.8) | 0 |
| Buparlisib+lomustine arm | ||
| Preferred term, n (%) | Buparlisib (60 mg)+lomustine (n=18) | |
| All grades | Grades 3/4 | |
| Fatigue | 8 (44.4) | 1 (5.6) |
| Nausea | 8 (44.4) | 0 |
| Decreased platelet count | 8 (44.4) | 2 (11.1) |
| Anaemia | 6 (33.3) | 1 (5.6) |
| Confusional state | 5 (27.8) | 0 |
| Depression | 5 (27.8) | 1 (5.6) |
| Headache | 5 (27.8) | 0 |
| Anxiety | 4 (22.2) | 0 |
| Hypertension | 4 (22.2) | 1 (5.6) |
| Hypokalaemia | 4 (22.2) | 2 (11.1) |
| Insomnia | 4 (22.2) | 0 |
| Neutropaenia | 4 (22.2) | 2 (11.1) |
| Thrombocytopaenia | 4 (22.2) | 4 (22.2) |
| Decreased white cell count | 4 (22.2) | 2 (11.1) |
| Increased alanine aminotransferase | 3 (16.7) | 1 (5.6) |
| Asthenia | 3 (16.7) | 0 |
| Constipation | 3 (16.7) | 0 |
| Diarrhoea | 3 (16.7) | 0 |
| Gastro-oesophageal reflux disease | 3 (16.7) | 0 |
| Hyperglycaemia | 3 (16.7) | 1 (5.6) |
| Memory impairment | 3 (16.7) | 0 |
| Decreased neutrophil count | 3 (16.7) | 1 (5.6) |
| Seizure | 3 (16.7) | 1 (5.6) |
| Somnolence | 3 (16.7) | 0 |
Buparlisib dose was reduced due to AEs in one (33.3%) patient in the 80 mg buparlisib plus carboplatin dose level, three (21.4%) patients in the 100 mg buparlisib plus carboplatin dose level, and one (5.6%) patient in the buparlisib plus lomustine arm. Buparlisib dose interruptions due to AEs occurred in one (33.3%), six (42.9%) and seven (38.9%) patients in the 80 mg buparlisib plus carboplatin dose level, the 100 mg buparlisib plus carboplatin dose level, and the buparlisib plus lomustine arm, respectively (table 5).
Table 5.
Dose reductions and interruptions of study drug
| Buparlisib* | Carboplatin† | Buparlisib* | Lomustine‡ | |||
| Buparlisib (80 mg) + carboplatin (n=3) |
Buparlisib (100 mg) + carboplatin (n=14) |
Buparlisib (80 mg) + carboplatin (n=3) |
Buparlisib (100 mg) + carboplatin (n=14) |
Buparlisib (60 mg) + lomustine (n=18) |
Buparlisib (60 mg) + lomustine (n=18) |
|
| Reductions, n (%) | ||||||
| Number of patients requiring dose reduction§ | ||||||
| 0 | 2 (66.7) | 11 (78.6) | 2 (66.7) | 13 (92.9) | 17 (94.4) | 18 (100) |
| 1 | 1 (33.3) | 3 (21.4) | 1 (33.3) | 1 (7.1) | 1 (5.6) | 0 |
| 2 | 0 | 0 | 0 | 0 | 0 | 0 |
| ≥3 | 0 | 0 | 0 | 0 | 0 | 0 |
| Number of patients with at least one dose reduction by reason¶ | 1 (33.3) | 3 (21.4) | 1 (33.3) | 1 (7.1) | 1 (5.6) | 0 |
| Adverse event | 1 (33.3) | 3 (21.4) | 1 (33.3) | 1 (7.1) | 1 (5.6) | 0 |
| Interruptions, n (%) | ||||||
| Number of patients with dose interruption§ | ||||||
| 0 | 2 (66.7) | 8 (57.1) | 2 (66.7) | 13 (92.9) | 11 (61.1) | 17 (94.4) |
| 1 | 0 | 4 (28.6) | 1 (33.3) | 1 (7.1) | 5 (27.8) | 1 (5.6) |
| 2 | 0 | 0 | 0 | 0 | 1 (5.6) | 0 |
| ≥3 | 1 (33.3) | 2 (14.3) | 0 | 0 | 1 (5.6) | 0 |
| Number of patients with at least one dose interruption by reason¶ | 1 (33.3) | 6 (42.9) | 1 (33.3) | 1 (7.1) | 7 (38.9) | 1 (5.6) |
| Adverse event | 1 (33.3) | 6 (42.9) | 1 (33.3) | 1 (7.1) | 7 (38.9) | 1 (5.6) |
| Dosing error/technical problems | 1 (33.3) | 0 | 0 | 0 | 1 (5.6) | 0 |
| Number of patients with permanent discontinuation by reason, n (%) | 3 (100) | 14 (100) | 3 (100) | 14 (100) | 18 (100) | 18 (100) |
| Adverse event | 0 | 3 (21.4) | 0 | 2 (14.3) | 3 (16.7) | 3 (16.7) |
| Physician decision | 0 | 0 | 1 (33.3) | 0 | – | – |
| Death | – | – | – | – | 1 (5.6) | 1 (5.6) |
| Progressive disease | 3 (100) | 11 (78.6) | 2 (66.7) | 12 (85.7) | 14 (77.8) | 14 (77.8) |
Percentage is based on ‘n’.
*Denotes the study drug buparlisib
†Denotes the study drug carboplatin
‡Denotes the study drug lomustine.
§A patient with multiple occurrences of the reason for dose reduction or interruption is counted only once in that category
¶A patient with multiple reasons for dose reduction or interruption is counted only once in the total row (ie, ‘number of patients with at least one dose reduction/interruption by reason’)
Two on-treatment deaths due to disease progression were reported, one each in the buparlisib plus carboplatin arm and in the buparlisib plus lomustine arm.
Patient-reported mood assessments
In both the buparlisib plus carboplatin arm and the buparlisib plus lomustine arm, the majority of the patients reported a shift from baseline to the mild depression severity category post baseline (score: 5–9) on PHQ-9. No patient reported a score in the severe depression severity category (score: 20–27) (table 6).
Table 6.
Shift from baseline to worst postbaseline depression severity on PHQ-9
| Treatment | Baseline category | Worst postbaseline category | |||||
| n (%) | None | Mild | Moderate | Severe | Missing | ||
| n (%) | n (%) | n (%) | n (%) | n (%) | |||
| Buparlisib (80 mg)+ carboplatin (n=3) | None | 3 (100) | 1 (33.3) | 2 (66.7) | 0 | 0 | 0 |
| Mild | 0 | 0 | 0 | 0 | 0 | 0 | |
| Moderate | 0 | 0 | 0 | 0 | 0 | 0 | |
| Severe | 0 | 0 | 0 | 0 | 0 | 0 | |
| Total | 3 (100) | 1 (33.3) | 2 (66.7) | 0 | 0 | 0 | |
| Buparlisib (100 mg)+ carboplatin (n=14) | None | 8 (57.1) | 2 (25.0) | 2 (25.0) | 4 (50.0) | 0 | 0 |
| Mild | 4 (28.6) | 1 (25.0) | 1 (25.0) | 1 (25.0) | 0 | 1 (25.0) | |
| Moderate | 2 (14.3) | 0 | 2 (100) | 0 | 0 | 0 | |
| Severe | 0 | 0 | 0 | 0 | 0 | 0 | |
| Total | 14 (100) | 3 (21.4) | 5 (35.7) | 5 (35.7) | 0 | 1 (7.1) | |
| Buparlisib (60 mg)+ lomustine (n=18) | None | 10 (55.6) | 4 (40.0) | 4 (40.0) | 2 (20.0) | 0 | 0 |
| Mild | 5 (27.8) | 0 | 3 (60.0) | 2 (40.0) | 0 | 0 | |
| Moderate | 3 (16.7) | 0 | 2 (66.7) | 0 | 0 | 1 (33.3) | |
| Severe | 0 | 0 | 0 | 0 | 0 | 0 | |
| Total | 18 (100) | 4 (22.2) | 9 (50.0) | 4 (22.2) | 0 | 1 (5.6) | |
Grades based on severity: normal (severity: none, score: 0–4), mild (score 5–9), moderate (score 10–19) and severe (score 20–27).
PHQ-9, Patient Health Questionnaire-9.
For PHQ-9 question 9 regarding suicidal thoughts, all patients in the 80 mg buparlisib dose level plus carboplatin (100%), most of the patients in the 100 mg buparlisib plus carboplatin dose level (78.6%), and in the buparlisib (60 mg) plus lomustine arm (72.2%) had a worst postbaseline score of 0 (not at all), which was the same as their baseline score. One patient reported a worst postbaseline score of 1 (several days) and one patient reported a score of 2 (more than half the days) in the 100 mg buparlisib plus carboplatin dose level. Three patients reported a score of 1 and one patient reported a score of 3 (nearly every day) in the buparlisib plus lomustine arm.
For the GAD-7 questionnaire, in the buparlisib plus carboplatin arm, eight patients had no shift from baseline (seven patients remained at none and one at mild). Eight patients shifted to a worse category (one patient in the 80 mg buparlisib dose level from none to mild and seven in the 100 mg buparlisib dose level: four from none to mild, one from none to moderate, one from mild to moderate, and one from mild to severe). One patient in the 100 mg buparlisib dose level shifted from mild at baseline to none post baseline. In the buparlisib plus lomustine arm, seven patients remained at the same category as at baseline (five patients at none and two at mild), six patients shifted to a worse category (five patients from none to mild, and one from none to moderate), three patients shifted from mild at baseline to none post baseline, and one from severe to mild (table 7).
Table 7.
Shift from baseline to worst postbaseline anxiety severity on GAD-7
| Treatment | Baseline category | Worst postbaseline category | |||||
| n (%) | None | Mild | Moderate | Severe | Missing | ||
| n (%) | n (%) | n (%) | n (%) | n (%) | |||
| Buparlisib (80 mg)+ carboplatin (n=3) | None | 3 (100) | 2 (66.7) | 1 (33.3) | 0 | 0 | 0 |
| Mild | 0 | 0 | 0 | 0 | 0 | 0 | |
| Moderate | 0 | 0 | 0 | 0 | 0 | 0 | |
| Severe | 0 | 0 | 0 | 0 | 0 | 0 | |
| Total | 3 (100) | 2 (66.7) | 1 (33.3) | 0 | 0 | 0 | |
| Buparlisib (100 mg)+ carboplatin (n=14) | None | 11 (78.6) | 5 (45.5) | 4 (36.4) | 1 (9.1) | 1 (9.1) | 0 |
| Mild | 3 (21.4) | 1 (33.3) | 1 (33.3) | 1 (33.3) | 0 | 0 | |
| Moderate | 0 | 0 | 0 | 0 | 0 | 0 | |
| Severe | 0 | 0 | 0 | 0 | 0 | 0 | |
| Total | 14 (100) | 6 (42.9) | 5 (35.7) | 2 (14.3) | 1 (7.1) | 0 | |
| Buparlisib (60 mg)+ lomustine (n=18) | None | 12 (66.7) | 5 (41.7) | 5 (41.7) | 1 (8.3) | 0 | 1 (8.3) |
| Mild | 5 (27.8) | 3 (60.0) | 2 (40.0) | 0 | 0 | 0 | |
| Moderate | 0 | 0 | 0 | 0 | 0 | 0 | |
| Severe | 1 (5.6) | 0 | 1 (100) | 0 | 0 | 0 | |
| Total | 18 (100) | 8 (44.4) | 8 (44.4) | 1 (5.6) | 0 | 1 (5.6) | |
Grades based on severity: normal (severity: none, score: 0–4), mild (score: 5–9)
GAD-7, Generalized Anxiety Disorder-7.
Discussion
The purposes of this study were to determine the MTD and/or RP2D for the combination of buparlisib plus carboplatin or lomustine and to assess the tolerability and the preliminary antitumour activity of the combinations in patients with rGBM. The MTD of buparlisib was established as 100 mg per day in combination with carboplatin at an AUC of 5 every 3 weeks. The MTD of buparlisib in combination with lomustine 100 mg/m2 every 6 weeks could not be determined as it did not satisfy the MTD criteria per the BLRM. Overall, rapid disease progression was observed with both the combinations, thus limiting study treatment exposure. The median PFS was 1.4 months in the buparlisib plus carboplatin arm and 1.3 months in the buparlisib plus lomustine arm. Only one patient (in the buparlisib plus carboplatin arm) had a PR lasting 15.1 months, and most patients had progressive disease as the best response. All except one patient in the study had AEs suspected to be related to the study treatment. Based on the overall safety profile and preliminary antitumour activity observed in this study, the phase II part of the study was not conducted. Moreover, due to the small number of patients, low response rate and short duration of treatment, no conclusion can be drawn on the relationship between response and PI3K activation status.
The modest outcomes observed in the current study are consistent with those reported for rGBM and further highlight the challenges of treating rGBM.
Several pan-PI3K inhibitors, including buparlisib, pictilisib and CLR457, have been evaluated for their clinical efficacy in a variety of tumours. However, most of the studies assessing these agents have demonstrated a modest improvement in efficacy outcomes accompanied by a challenging toxicity profile. In a phase III trial in patients with hormone receptor-positive breast cancer, buparlisib in combination with fulvestrant showed that the addition of a PI3K inhibitor improved outcomes; however, toxicities in the buparlisib arm limited the treatment duration and intensity.26 27 Similarly, pictilisib also demonstrated a modest efficacy improvement in combination with an aromatase inhibitor in hormone receptor-positive breast cancer, but the clinical activity was limited by the challenging toxicity profile.28 CLR457, evaluated in solid tumours, demonstrated limited antitumour activity and poor tolerability, leading to termination of its clinical development.29 Overall, data from the current study and from previous studies of pan-PI3K inhibitors highlight the challenges of achieving a meaningful clinical benefit when targeting all class I PI3K isoforms.
The current study demonstrated that, with the dose levels studied, neither buparlisib plus carboplatin nor buparlisib plus lomustine significantly improved antitumour activity compared with historical data on single-agent carboplatin or lomustine.
Acknowledgments
We thank the patients who participated in this trial and their families, as well as the staff members at individual trial centres who provided support. We also thank Avinash Yerramsetti (Novartis Healthcare) for providing medical editorial assistance with this manuscript.
Footnotes
Contributors: The study was designed by the investigators and the sponsor. Design and conduct of the study were undertaken by the sponsor in collaboration with the investigators. The study investigators and their respective research teams collected the data. Novartis Pharmaceuticals Corporation compiled the data for summation and analysis. All authors were responsible for data interpretation. The article was prepared by MR in conjunction with all the authors, including employees of the sponsor. The corresponding author had final responsibility for the decision to submit the manuscript for publication.
Funding: The study was initiated, funded and sponsored by Novartis Pharmaceuticals Corporation.
Competing interests: PMC reports grants from AstraZeneca and personal fees from AbbVie, AstraZeneca, BMS, Daiichi Sankyo, Leo Pharma, Merck, MSD and Vifor, outside the submitted work. MC acted in an advisory role for Novartis, Sanofi, Pierre Fabre, Lilly and AstraZeneca, outside the submitted work. MJG-G has received honoraria from Novartis, Pfizer and Roche and acted in an advisory role for Daiichi Sankyo, Novartis and Pfizer. OC reports grants from Roche, personal fees and non-financial support from Roche and AbbVie, personal fees from Ipsen and Celdex, and non-financial support from Servier, outside the submitted work. AI reports grants and travel funding from Carthera (September 2019), research grants from Transgene, Sanofi and Air Liquide, and travel funding from Leo Pharma, outside the submitted work. HG reports grants from AbbVie, personal fees and non-financial support from AbbVie and Ignyta, personal fees from BMS, MSD, Eisai and Merck Serono, outside the submitted work. JR reports grants from Novartis during the conduct of the study. PYW reports research support from Agios, AstraZeneca, BeiGene, Eli Lilly, Genentech/Roche, Karyopharm, Kazia, MediciNova, Merck, Novartis, Oncoceutics, Sanofi Aventis and VBI Vaccines, participated on advisory boards for AbbVie, Agios, AstraZeneca, Blue Earth Diagnostics, Eli Lilly, Genentech/Roche, Karyopharm, Kiyatec, Puma, Vascular Biogenics, Taiho, Deciphera, VBI Vaccines and Tocagen, and speaker for Merck and Prime Oncology. EP reports scientific consultancy role for Celgene. VD is an employee of Novartis Pharma SAS and holds Novartis stock options. DM is an employee of Novartis Pharma and holds Novartis stock options. ME-H is an employee of Novartis Pharmaceuticals Corporation and holds Novartis stock options.
Patient consent for publication: Not required.
Ethics approval: This study was conducted in accordance with the Good Clinical Practice guidelines and the Declaration of Helsinki. An independent ethics committee and institutional review boards approved the study protocol at each participating centre.
Provenance and peer review: Not commissioned; externally peer reviewed.
Data availability statement: Data are available upon reasonable request. Novartis is committed to sharing with qualified external researchers, access to patient-level data and supporting clinical documents from eligible studies. These requests are reviewed and approved by an independent review panel on the basis of scientific merit. All data provided are anonymised to respect the privacy of patients who have participated in the trial in line with applicable laws and regulations. Trial data availability is according to the criteria and process described at www.clinicalstudydatarequest.com.
References
- 1.Davis ME. Glioblastoma: overview of disease and treatment. Clin J Oncol Nurs 2016;20:S2–8. 10.1188/16.CJON.S1.2-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hanif F, Muzaffar K, Perveen K, et al. Glioblastoma multiforme: a review of its epidemiology and pathogenesis through clinical presentation and treatment. Asian Pac J Cancer Prev 2017;18:3–9. 10.22034/APJCP.2017.18.1.3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Rock K, McArdle O, Forde P, et al. A clinical review of treatment outcomes in glioblastoma multiforme--the validation in a non-trial population of the results of a randomised Phase III clinical trial: has a more radical approach improved survival? Br J Radiol 2012;85:e729–33. 10.1259/bjr/83796755 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ozdemir-Kaynak E, Qutub AA, Yesil-Celiktas O. Advances in glioblastoma multiforme treatment: new models for nanoparticle therapy. Front Physiol 2018;9:170–70. 10.3389/fphys.2018.00170 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sun Y, Alberta JA, Pilarz C, et al. A brain-penetrant Raf dimer antagonist for the noncanonical BRAF oncoprotein of pediatric low-grade astrocytomas. Neuro Oncol 2017;19:774–85. 10.1093/neuonc/now261 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bleeker FE, Molenaar RJ, Leenstra S. Recent advances in the molecular understanding of glioblastoma. J Neurooncol 2012;108:11–27. 10.1007/s11060-011-0793-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Stupp R, Taillibert S, Kanner AA, et al. Maintenance therapy with Tumor-Treating fields plus temozolomide vs temozolomide alone for glioblastoma: a randomized clinical trial. JAMA 2015;314:2535–43. 10.1001/jama.2015.16669 [DOI] [PubMed] [Google Scholar]
- 8.Villà S, Balañà C, Comas S. Radiation and concomitant chemotherapy for patients with glioblastoma multiforme. Chin J Cancer 2014;33:25–31. 10.5732/cjc.013.10216 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Nandeesh BN, Naskar S, Shashtri AH, et al. Recurrent glioblastomas exhibit higher expression of biomarkers with stem-like properties. J Neurosci Rural Pract 2018;9:086–91. 10.4103/jnrp.jnrp_417_17 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol 2009;10:459–66. 10.1016/S1470-2045(09)70025-7 [DOI] [PubMed] [Google Scholar]
- 11.Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005;352:987–96. 10.1056/NEJMoa043330 [DOI] [PubMed] [Google Scholar]
- 12.Ostrom QT, Cioffi G, Gittleman H, et al. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2012-2016. Neuro Oncol 2019;21:v1–100. 10.1093/neuonc/noz150 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wick W, Gorlia T, Bendszus M, et al. Lomustine and bevacizumab in progressive glioblastoma. N Engl J Med 2017;377:1954–63. 10.1056/NEJMoa1707358 [DOI] [PubMed] [Google Scholar]
- 14.van Linde ME, Brahm CG, de Witt Hamer PC, et al. Treatment outcome of patients with recurrent glioblastoma multiforme: a retrospective multicenter analysis. J Neurooncol 2017;135:183–92. 10.1007/s11060-017-2564-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Westhoff M-A, Karpel-Massler G, Brühl O, et al. A critical evaluation of PI3K inhibition in glioblastoma and neuroblastoma therapy. Mol Cell Ther 2014;2:32. 10.1186/2052-8426-2-32 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mao H, Lebrun DG, Yang J, et al. Deregulated signaling pathways in glioblastoma multiforme: molecular mechanisms and therapeutic targets. Cancer Invest 2012;30:48–56. 10.3109/07357907.2011.630050 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fan Q-W, Weiss WA. Targeting the RTK-PI3K-mTOR axis in malignant glioma: overcoming resistance. Curr Top Microbiol Immunol 2010;347:279–96. 10.1007/82_2010_67 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cancer Genome Atlas Research Network Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 2008;455:1061–8. 10.1038/nature07385 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Maira S-M, Pecchi S, Huang A, et al. Identification and characterization of NVP-BKM120, an orally available pan-class I PI3-kinase inhibitor. Mol Cancer Ther 2012;11:317–28. 10.1158/1535-7163.MCT-11-0474 [DOI] [PubMed] [Google Scholar]
- 20.de Gooijer MC, Zhang P, Buil LCM, et al. Buparlisib is a brain penetrable pan-PI3K inhibitor. Sci Rep 2018;8:10784. 10.1038/s41598-018-29062-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Burger MT, Pecchi S, Wagman A, et al. Identification of NVP-BKM120 as a potent, selective, orally bioavailable class I PI3 kinase inhibitor for treating cancer. ACS Med Chem Lett 2011;2:774–9. 10.1021/ml200156t [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Koul D, Fu J, Shen R, et al. Antitumor activity of NVP-BKM120--a selective pan class I PI3 kinase inhibitor showed differential forms of cell death based on p53 status of glioma cells. Clin Cancer Res 2012;18:184–95. 10.1158/1078-0432.CCR-11-1558 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Speranza M-C, Nowicki MO, Behera P, et al. BKM-120 (Buparlisib): a Phosphatidyl-Inositol-3 kinase inhibitor with anti-invasive properties in glioblastoma. Sci Rep 2016;6:20189. 10.1038/srep20189 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wen PY, Touat M, Alexander BM, et al. Buparlisib in patients with recurrent glioblastoma harboring phosphatidylinositol 3-kinase pathway activation: an open-label, multicenter, Multi-Arm, phase II trial. J Clin Oncol 2019;37:741–50. 10.1200/JCO.18.01207 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wen PY, Chang SM, Van den Bent MJ, et al. Response assessment in neuro-oncology clinical trials. J Clin Oncol 2017;35:2439–49. 10.1200/JCO.2017.72.7511 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Baselga J, Im S-A, Iwata H, et al. Buparlisib plus fulvestrant versus placebo plus fulvestrant in postmenopausal, hormone receptor-positive, HER2-negative, advanced breast cancer (BELLE-2): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2017;18:904–16. 10.1016/S1470-2045(17)30376-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Di Leo A, Seok Lee K, Ciruelos E, et al. Abstract S4-07: BELLE-3: a phase III study of buparlisib + fulvestrant in postmenopausal women with HR+, HER2–, aromatase inhibitor-treated, locally advanced or metastatic breast cancer, who progressed on or after mTOR inhibitor-based treatment. Cancer Research 2017;77. [Google Scholar]
- 28.Krop IE, Mayer IA, Ganju V, et al. Pictilisib for oestrogen receptor-positive, aromatase inhibitor-resistant, advanced or metastatic breast cancer (FERGI): a randomised, double-blind, placebo-controlled, phase 2 trial. Lancet Oncol 2016;17:811–21. 10.1016/S1470-2045(16)00106-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Harding JJ, Bauer TM, Tan DSW, et al. Characterization and phase I study of CLR457, an orally bioavailable pan-class I PI3-kinase inhibitor. Invest New Drugs 2019;37:271–81. 10.1007/s10637-018-0627-4 [DOI] [PMC free article] [PubMed] [Google Scholar]