Phase 1 study of mTORC1/2 inhibitor sapanisertib (TAK-228) in advanced solid tumours, with an expansion phase in renal, endometrial or bladder cancer

Background This Phase 1 dose-escalation/expansion study assessed safety/tolerability of sapanisertib, an oral, highly selective inhibitor of mTORC1/mTORC2, in advanced solid tumours. Methods Eligible patients received increasing sapanisertib doses once daily (QD; 31 patients), once weekly (QW; 30 patients), QD for 3 days on/4 days off QW (QD × 3dQW; 33 patients) or QD for 5 days on/2 days off QW (QD × 5dQW; 22 patients). In expansion cohorts, 82 patients with renal cell carcinoma (RCC), endometrial or bladder cancer received sapanisertib 5 mg QD (39 patients), 40 mg QW (26 patients) or 30 mg QW (17 patients). Results Maximum tolerated doses of sapanisertib were 6 mg QD, 40 mg QW, 9 mg QD × 3dQW and 7 mg QD × 5dQW. Frequent dose-limiting toxicities (DLTs) included hyperglycaemia, maculo-papular rash (QD), asthenia and stomatitis (QD × 3dQW/QD × 5dQW); expansion phase doses of 5 mg QD and 30 mg QW were selected based on tolerability beyond the DLT evaluation period. One patient with RCC achieved complete response; nine experienced partial responses (RCC: seven patients; carcinoid tumour/endometrial cancer: one patient each). Sapanisertib pharmacokinetics were time-linear and supported multiple dosing. Pharmacodynamic findings demonstrated treatment-related reductions in TORC1/2 biomarkers. Conclusions Sapanisertib demonstrated a manageable safety profile, with preliminary antitumour activity observed in RCC and endometrial cancer. Clinical trial registration ClinicalTrials.gov, NCT01058707.

mTORC2 inhibition, can result in the activation of AKT through a negative feedback mechanism, which may limit rapalog efficacy by accelerating tumour progression. 12 Preclinical data have demonstrated that inhibition of AKT activity, through mTORC2 inhibition, may block tumour progression. 13 Subsequently, a new generation of mTORC1/2 inhibitors have been developed, as dual inhibition may offer an advantage over mTORC1 inhibitors by targeting at least three key enzymes (PI3K, AKT and mTOR). 14 Proof of this concept was demonstrated in preclinical models for several epithelial malignancies. [15][16][17][18] Sapanisertib is an investigational, oral and highly selective adenosine triphosphate-competitive mTOR kinase inhibitor that suppresses both mTORC1 and mTORC2. Preclinical models of efficacy, safety and pharmacokinetics (PK) of sapanisertib have largely relied on daily dosing. [19][20][21][22][23] Investigation of different dosing schedules in preclinical efficacy models has demonstrated commensurate tumour growth inhibition with administration of sapanisertib in daily or intermittent dosing schedules. [19][20][21][22][23] These preclinical data generally indicated that efficacy was related to total exposure (i.e. area under the curve) and independent of dosing schedule. However, the clinical tolerability was expected to be different for different schedules. Therefore, we evaluated a range of dosing schedules in this first-in-human Phase 1 study (NCT01058707) of sapanisertib in patients with advanced solid tumours, with an expansion phase including patients with RCC, endometrial or bladder cancer.
Definitions for dose-limiting toxicities (DLTs), as predetermined in the study protocol, were: any grade ≥3 non-haematologic toxicity (except inadequately treated grade 3 nausea and/or vomiting and grade 3 diarrhoea [all patients should have received optimal antiemetic and/or antidiarrhoeal prophylaxis treatment], grade 3 hyperglycaemia lasting ≤14 days [all patients should have received optimal anti-glycaemic treatment, including insulin) and grade 3 rash lasting ≤3 days [all patients should have received topical steroid treatment, oral antihistamines and pulse oral steroids, if necessary]); grade 4 neutropenia lasting >7 days in the absence of growth factor support; grade 4 neutropenia of any duration accompanied with fever ≥38.5°C and/or systemic infection; any other grade ≥4 haematologic toxicity.
Sapanisertib dosing was withheld for Grade ≥ 3 treatmentrelated toxicities. If the event resolved to Grade ≤ 1 or baseline values within 28 days of interrupting therapy, the patient could resume study treatment at a ≥25% dose reduction or, for patients in the dose-escalation phase, at the next lower dose level with the sponsor's approval. If dose modification was required for subjects receiving ≤4 mg QD, then the dosing frequency was decreased to 5 days per week, instead of decreasing the daily dose administered. If sapanisertib dosing was delayed for >28 consecutive days for treatment-related toxicity, despite supportive treatment per standard clinical practice, or more than 2 dose reductions of sapanisertib were required in a patient, sapanisertib therapy was stopped, the patient was discontinued from the study and the follow-up visit was completed within 30 days of the last administration of sapanisertib.
The maximum tolerated dose (MTD) was to be determined in the QD schedule before enrolment of initial single-patient cohorts for the QW and QDx3d QW schedules. If a grade ≥ 2 AE was observed in any single-patient cohort, an additional 2-5 patients were assigned to that cohort. Subsequent dose cohorts for that schedule included 3-6 patients. Dose cohorts for the QDx5dQW schedule enrolled three patients with an additional three patients enrolled if a DLT was observed per a standard 3 + 3 design. Once the MTD was identified for each of the dosing schedules, an additional six patients were enrolled to obtain further PK and safety data prior to expansion.
Based on data collected during dose-escalation, the expansion phase evaluated the safety and efficacy of QD and QW sapanisertib in patients with RCC, endometrial or bladder cancer. Patients could continue receiving sapanisertib for ≤1 year (or beyond, if the investigator and sponsor agreed) in the absence of disease progression or unacceptable toxicity.
The primary objective was to determine the MTD and DLTs for each sapanisertib dosing schedule and evaluate the safety and tolerability of sapanisertib in both phases. Secondary objectives were to evaluate preliminary antitumour activity, PK in peripheral blood, and pharmacodynamics (PD) of sapanisertib, as measured by modulation in phosphorylation of S6, 4EBP1 and NDRG1 in surrogate tissue (skin) and tumour.
The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice. Institutional review boards approved all aspects of the study. All participants provided written informed consent.

Patients
Eligible patients were aged ≥18 years with locally advanced or metastatic solid tumours who had failed standard-of-care therapy, had an Eastern Cooperative Oncology Group performance status of 0-1, and had adequate bone marrow, hepatic, renal and metabolic (fasting serum glucose ≤ 130 mg/dL and fasting triglycerides ≤ 300 mg/dL) function.
Patients with locally advanced or metastatic brain tumours were eligible if their brain metastases had been treated (without evidence of progression or haemorrhage post-treatment), and if they had not taken dexamethasone 4 weeks prior to the first study drug administration and with no ongoing requirement for dexamethasone or antiepileptic drugs. In the expansion phase, eligible patients had measurable disease per the Response Evaluation Criteria in Solid Tumors (RECIST), v1.1.
Patients could be considered for enrolment in one of three disease-specific cohorts: (1) advanced or recurrent endometrial adenocarcinoma with disease progression following ≥1 prior chemotherapy regimen; (2) advanced/metastatic urothelial cancer (carcinoma of the bladder, ureter and/or renal pelvis), progressive after ≥1 prior therapy in the metastatic/unresectable setting; (3) advanced RCC after failure of ≥1 prior antivascular endothelial growth factor therapy with no prior TORC1 inhibitor therapy, or progressed on treatment with TORC1 inhibitor therapy.
Patients who had received prior cancer therapy within 2 weeks, systemic corticosteroid therapy within 1 week prior to the first dose of study drug, or bisphosphonates within 30 days prior to the first sapanisertib dose were not eligible. Additionally, patients with impaired cardiac function or significant active cardiovascular disease were also excluded. In the expansion phase, patients who had received prior AKT, PI3K, dual PI3K/TORC1/2 or TORC1/2 inhibitors were excluded.

Assessments
Patients who received ≥1 dose of sapanisertib were included in the safety population. The response-evaluable population included patients who received ≥1 dose of sapanisertib, had measurable disease at baseline, and had ≥1 postbaseline assessment. In the dose-escalation phase, the evaluable Phase 1 study of mTORC1/2 inhibitor sapanisertib (TAK-228) in advanced. . . MH Voss et al. population included patients who received ≥75% of the planned sapanisertib doses in cycle 1 or experienced a DLT. Response was assessed according to the RECIST v1.1 24 after every two treatment cycles. Adverse events (AEs) were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events, v4.0. Patients were provided with a home blood-glucose meter to monitor their fasting predose blood-glucose measurements to assess hyperglycaemia as an on-target AE and PD marker.
Peripheral blood was collected serially to quantify plasma levels of sapanisertib for PK analysis via validated liquid chromatography tandem mass spectrometry with an assay range of 1-1000 ng/mL (MicroConstants, San Diego, CA), during both dose-escalation and expansion phases. Blood samples were collected before each dose and 0.5, 1, 2, 4, 8 and 24 h after dosing on cycle 1, day 1 and cycle 2, day 1, in the dose-escalation phase. In the expansion phase, samples were collected predose and 2, 4 and 6 h postdose on cycle 1, day 1, and then predose and 2 h postdose on day 8, 15 or 22 of cycle 1.
The tissue PK and PD of sapanisertib were investigated in skin and tumour biopsies. During dose escalation, 3-mm core skin biopsies were collected predose and on any day from 8-15 (3 h postdose) of cycle 1 for immunohistochemistry (IHC) assessment of mTORC1 downstream effects (p4EBP1, pS6) and pNDRG1 (Mosaic Laboratories, Lake Forest, CA). IHC for p-AKT did not pass validation and is not reported. Archival tumour tissues for assessment of prognostic markers were collected at baseline.
Statistical analyses were primarily descriptive and graphical in nature, with no formal statistical hypothesis testing.  Table 1.

Patients
DLTs and MTD determination Dose escalation, DLTs and MTDs are summarised in Table 2 (14) 1 (3)  4 (13) 10 (9) 22 (56)  10 (59)  13 (50)  45 (55) RCC, TORC1i naïve - Number of prior treatment regimens, median (range) QD once daily, QD × 3dQW QD for 3 days on/4 days off QW, QD × 5dQW QD for 5 days on/2 days off QW, QW once weekly, TORC1i target of rapamycin complex 1 inhibitor therapy. Based on biochemical, PK and tolerability data, and early signs of antitumour activity, the QD and QW schedules were selected for expansion. Sapanisertib 5 mg QD was selected rather than the MTD of 6 mg, because the 6 mg dose was poorly tolerated beyond the DLT evaluation period in 10 evaluable patients. The QW 40 mg dose was determined as the MTD based on DLT criteria and was initially selected for expansion. However, follow-up of the first 21 patients enrolled at this dose revealed that several patients required dose modifications or discontinued treatment due to AEs; the recommended expansion dose was thereafter reduced to 30 mg QW.
Treatment exposure and safety During dose escalation, patients received a median of two treatment cycles (range, 1-58; Supplementary Table 1). Across all doses, 97% of patients experienced treatment-related AEs (TRAEs), and 11% experienced at least one treatment-related grade ≥ 3 AE (Supplementary Table 1). The most common TRAEs were hyperglycaemia, nausea, stomatitis, vomiting, decreased appetite and diarrhoea (Table 3). Hyperglycaemia was the most common grade ≥ 3 TRAE (Supplementary Table 2), which was generally well controlled with metformin and home glucose monitoring. Serious AEs occurred in 43% of patients; 24% discontinued treatment due to an AE (Supplementary Table 1).
During expansion, patients received a median of two treatment cycles (range, 1-26; Supplementary Table 1). All patients in this phase experienced at least one TRAE; 46% experienced at least one grade ≥ 3 TRAE (Supplementary Table 1). TRAEs and all-cause AEs in the expansion phase are summarised in Table 3 and Supplementary  Table 3, respectively. The most common TRAEs according to schedule were fatigue (56%) and nausea (51%) at 5 mg QD, fatigue (88%), nausea and hyperglycaemia (77% each) at 40 mg QW, and hyperglycaemia (71%), nausea and fatigue (65% each) at 30 mg QW (Table 3). Supplementary Table 2 and Supplementary Table 4 summarise grade ≥ 3 TRAEs and all-cause AEs, respectively, in the expansion phase. Hyperglycaemia was also the most common grade ≥ 3 TRAE in the expansion phase. Serious AEs occurred in 41% of patients and 15% discontinued due to an AE.
AEs resulting in treatment discontinuation during the expansion phase occurred in 18% of patients receiving 5 mg QD or 30 mg  Table 1). There were 7 on-study deaths within 30 days of last dose. Four patients died during dose-escalation: 1 due to ventricular fibrillation and cardiac arrest (6 mg QD considered possibly related to sapanisertib by the investigator, in a patient with preexisting cardiovascular risk factors), 3 from progressive cancer; 2 gastric (10 mg QW [n = 1], 9 mg QD × 3dQW [n = 1]) and 1 breast (12 mg QD × 3dQW). Three patients died during expansion due to disease progression (30 mg QW [n = 2], 40 mg QW [n = 1]).
Antitumour activity At data cut-off (02 October 2015), two patients, one with RCC receiving 15 mg QW (dose escalation; 58 cycles) and one with TORC1 inhibitor−naïve RCC receiving 30 mg QW (expansion; 26 cycles), were still receiving study drug. Tumour response across both study phases is summarised in Table 4 and Supplementary  Fig. 1. Supplementary Fig. 1 also shows study duration and best percentage change for the expansion phase. During doseescalation, two patients with RCC receiving 15 mg QW and 40 mg QW achieved a partial response (PR) maintained for >32 months for an objective response rate (ORR) of 8%, and one with a carcinoid tumour receiving 10 mg QD × 5dQW achieved a PR maintained for > 32 months for an ORR of 6%. Five patients in dose-escalation maintained stable disease (SD) for ≥ 6 months. In the expansion phase, one patient with TORC1 inhibitor−naïve RCC receiving 40 mg QW achieved a complete response (CR) maintained for 16 months. Five additional patients with RCC achieved PRs: three TORC1 inhibitor-naïve (5 mg QD, 30 mg QW, and 40 mg QW) for an ORR of 22%, and two with TORC1 inhibitor failure (both 5 mg QD) for an ORR of 9%. There was also one PR in a patient with endometrial cancer receiving 5 mg QD (ORR of 6%). Four of the seven objective responses reported in the expansion phase persisted for > 16 months: three patients with TORC1 inhibitor-naïve RCC (30 mg QW [n = 1], 40 mg QW [n = 2]) and one with TORC1 inhibitor failure RCC (5 mg QD). Six patients with RCC maintained SD for ≥6 months. Among patients with bladder cancer (n = 13), no objective responses were achieved, and 1/5 patients with SD maintained SD for ≥ 6 months.

Pharmacokinetics
Cycle 1, day 1, PK data were available from 112 patients and cycle 2, day 1, PK data were available from 70 patients. Single-dose (cycle 1, day 1) and multiple-dose (cycle 2, day 1) PK data are summarised in Fig. 1, Supplementary Fig. 2 and Supplementary Table 5. Sapanisertib exhibited rapid oral absorption, with a median T max ranging from 1.0 − 2.8 h across all doses (Supplementary Table 5). Sapanisertib plasma concentrations generally increased in a dosedependent manner within the 2-40 mg dose range; mean plasma half-life ranged from 5.9 to 9.4 h, with PK variability percentage coefficient of variation (%CV) values for area under the curve from time zero to 24 h postdose (AUC 0-24h ) ranging from~26 to 87%.  QD once daily, QD × 3dQW once daily for 3 days on and 4 days off each week, QD × 5dQW once daily for 5 days on and 2 days off each week, QW once weekly. a Includes oropharyngeal pain and mucosal inflammation.
Sapanisertib did not accumulate in plasma to any appreciable extent with repeat dosing in any of the schedules (Fig. 1). Sapanisertib PK were generally consistent during both assessment periods (cycle 1, day 1, and cycle 2, day 1), indicating a lack of a time-dependent accumulation with repeat dosing.

Pharmacodynamics
During dose-escalation, the PD effect of sapanisertib on downstream effectors of TORC1 (p4EBP1 and pS6) and TORC2 (pPRAS40 and pNDRG1) was measured in skin-tissue biopsies from 88 patients collected at baseline and days 8-15 (3 h postdose) of cycle 1. Treatment-related decreases in p4EBP1, pS6, pPRAS40 and pNDRG1 were consistent with dual TORC1/2 inhibition at sapanisertib doses of ≥4 mg (Supplementary Fig. 3). There was a trend toward a dose-dependent PD effect between 2-6 mg sapanisertib; however, no noteworthy differences in PD marker trends were observed at doses ≥6 mg or across the different dosing schedules and dose levels. PD data in tumour tissue were limited due to the small number of samples. For the five patients with paired tumour samples available for review (biopsies at baseline and cycle 1, week 2), the levels of p4EBP1, pS6, pPRAS40 and pNDRG1 decreased after sapanisertib treatment (Supplementary Fig. 4). An integrated PK/PD analysis correlated the plasma concentration of sapanisertib at a single time point with PD findings in skin and demonstrated a concentration-related change in pS6, p4EBP1, pNDRG1 and pPRAS40 ( Supplementary Fig. 5).

DISCUSSION
In this open-label, Phase 1, first-in-human study, the safety profile of sapanisertib was characterised and shown to be manageable and consistent with the toxicity profiles of other mTOR inhibitors. Sapanisertib showed preliminary antitumour activity in patients with RCC (TORC1 inhibitor-naïve and exposed) and in endometrial cancer.
The sapanisertib MTDs were determined as 6 mg QD, 40 mg QW, 9 mg QD × 3dQW and 7 mg QD × 5dQW. In a previous study in advanced haematological malignancies, the sapanisertib MTD for QD dosing was slightly lower (4 mg) and the MTD for QD × 3dQW was the same (9 mg). 25 Consistent with the lower MTD determined for QD dosing, 5 mg QD (expansion phase) was selected for further investigation, as 6 mg QD was poorly tolerated beyond the DLT evaluation period. DLTs varied across the dosing schedules: maculo-papular rash and hyperglycaemia, considered on-target toxicities for this class of agents, were the most frequently reported DLTs with QD dosing, and stomatitis and asthenia with the QD × 3dQW and QD × 5dQW schedules. Sapanisertib DLTs previously reported in patients with haematologic malignancies, were stomatitis, urticaria, blood creatinine elevation, fatigue, nausea and vomiting with QD dosing, and erythematous rash, fatigue, asthenia, mucosal inflammation and thrombocytopenia with QD × 3dQW dosing. 25 The safety profile of sapanisertib in this Phase 1 study was generally manageable across all schedules, and tolerability was greater with increased intermittence of dosing. Common AEs related to sapanisertib across all schedules in both phases included hyperglycaemia, nausea and vomiting. Hyperglycaemia is a known side effect of PI3K pathway inhibition reported in other studies investigating dual mTORC inhibitors. 26,27 Hyperglycaemia was found to be dose-dependent in the QD dosing schedule, occurring more frequently at the higher QW dose. Hyperglycaemia in both phases was generally grade 1-2 in severity, consistent with episodes previously reported, 25 and easily controlled with metformin therapy. Otherwise, there were no consistent dosedependent trends in AEs related to sapanisertib among the schedules, and AEs reported for sapanisertib 5 mg QD and 30 mg QW in the expansion phase were generally manageable. In the expansion phase, the sapanisertib dose for QW dosing was reduced from 40 mg (MTD determined during dose escalation) to 30 mg due to high rates of nausea and vomiting.
Preliminary antitumour activity was observed with sapanisertib across different schedules in patients with RCC and endometrial cancer. One patient receiving sapanisertib 40 mg QW had a CR and nine patients achieved PR, with response maintained in two patients for >32 months and four for >16 months. This is consistent with efficacy reported in other phase 1/2 studies of dual TORC1/2 kinase inhibitors in patients with advanced malignancies, 25,[27][28][29] and the ORR of 22% reported in patients with RCC naïve to a TORC1 inhibitor compares favourably to ORRs observed with rapalogs in patients with RCC (ORR 8.6% with temsirolimus 8 and 1-5% with everolimus). 9,7,30,31 However, a recent Phase 2 study of sapanisertib 30 mg QW versus everolimus 10 mg QD in patients with refractory clear-cell RCC reported no responses with single-agent sapanisertib versus an ORR of 13% with everolimus (NCT02724020). 32 The PK profile showed that single doses of oral sapanisertib were rapidly absorbed and concentrations increased in a dose-dependent  manner between 2 and 40 mg; the PK of sapanisertib was timelinear and supported the use of multiple dosing. Following repeat dosing, sapanisertib did not accumulate in plasma to any appreciable extent in any of the dosing schedules. Ghobrial et al. 25 reported a similar PK profile, supporting the administration of both daily and intermittent schedules. Therefore, QW and QD dosing schedules were selected for expansion based on their apparent similar preliminary antitumour activity and better tolerability. PD findings demonstrated treatment-related reductions in TORC1/2 biomarkers (p4EBP1, pS6, pPRAS40 and pNDRG1), which supports dual TORC1/2 inhibition of sapanisertib in doses ≥4 mg. In conclusion, sapanisertib had a manageable safety profile across the various schedules studied. Recommended Phase 2 doses, based on the data presented here, include 30 mg QW and 5 mg QD. Preliminary antitumour activity was observed in RCC and endometrial cancer. Phase 2 studies of sapanisertib (QW dosing) in patients with previously treated, metastatic clear-cell RCC (NCT03097328) and endometrial cancer (NCT02725268) have completed enrolment. Findings from these studies and those presented here contribute to sapanisertib dosing decisions in ongoing Phase 1/2 studies as a single agent or in combination with standard-of-care therapy in multiple malignancies.

ACKNOWLEDGEMENTS
We thank all the patients who participated in these studies and their families, as well as all the investigators and site staff who made these studies possible. We also acknowledge writing support from Helen Wilkinson, PhD (FireKite, an Ashfield company, part of UDG Healthcare plc, which was funded by Millennium Pharmaceuticals, Inc.) and editorial support from Janice Y. Ahn, PhD

ADDITIONAL INFORMATION
Ethics approval and consent to participate The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice. Institutional review boards (listed in the Supplementary Information) approved all aspects of the study. All participants provided written informed consent.
Data availability The datasets, including the redacted study protocol, redacted statistical analysis plan and individual participants data supporting the results reported in this article, will be made available within three months from initial request, to researchers who provide a methodologically sound proposal. The data will be provided after its de-identification, in compliance with applicable privacy laws, data protection and requirements for consent and anonymisation.