Durvalumab and Tremelimumab in Resectable HCC

INTRODUCTION This is a phase II, open label single arm, multicentre study of Tremelimumab in combination with Durvalumab (IMFINZI) in preoperative patients with resectable hepatocellular carcinoma (HCC) followed by adjuvant Durvalumab.

Disease background: HCC represents the third most common cause of cancer-related death and accounts for over 80% of primary liver cancers worldwide. Treatments are numerous and are selected based on tumour extent, patient comorbidities, and liver dysfunction. Surgical resection is recommended in patients with resectable tumours with otherwise well-compensated liver disease in the absence of clinically significant portal hypertension (hepatic vein to portal system gradient >10 mmHg or a platelet count ≤100,000/ml). Patients with macrovascular invasion, multifocal disease (within the same lobe), or large tumours that would require vascular resections and reconstructions are not typically considered for surgical resections according to the current clinical management guidelines. They are instead generally recommended to undergo systemic or locoregional therapies for palliation if they are not otherwise candidates for liver transplantation. In more favorable conditions of liver-confined disease and an absence of macrovascular invasion, still, many patients are deemed unresectable. Within this setting, strategies that can downstage tumors to allow resection to be performed, and help decrease the postoperative disease recurrence risk, are urgently needed. Moreover, when surgically resected, as many as 70% of the patients develop disease recurrence. Unfortunately, there are few clinical variables available that can aid in predicting postoperatively recurrence risk, many of which are only available on the pathology specimen after the liver resection has been performed, such as microvascular invasion or the presence of satellite lesions. A preoperative tumour biopsy, may identify biomarkers to immune checkpoint inhibitor response that can further help to refine treatment selection and potentially a subset of patients where neoadjuvant immune checkpoint inhibitor therapy use may be particularly beneficial. There is typically no discrepancy between the number of patients intended to receive liver resection and those who can undergo resection (i.e., no dropout while awaiting a liver resection). Consequently, given that the risk of tumour progression while awaiting surgical resection is low, the addition of neoadjuvant therapy during this wait is unlikely to lead to patients being unable to undergo their intended liver resection.

Study Rationale In advanced HCC combination immunotherapeutic strategies are rapidly becoming a new standard of care. The rationale for the use of immunotherapy in HCC is supported by the high expression of immunosuppressive cells, including PD-L1 expression, which is a poor prognostic factor. Therefore targeting immune checkpoints may improve outcomes, and several adjuvant trials are currently accruing patients.

Tremelimumab is a human IgG2 monoclonal antibody (mAb) directed against CTLA-4. The CTL-associated antigen 4 (CTLA-4) is an inhibitory receptor for B7, acting as an immune checkpoint resulting in a downregulation in T-cell activation through the competition with CD28 for B7 binding. Notably different to its IgG1 mAb counterparts, Tremelimumab may have less potential for antibody dependent cell mediated toxicity. Tremelimumab has previously been evaluated in hepatitis C-virus related HCC in a phase 2 trial .Twenty patients were included in the study, which reported a partial response rate of 17.6% and a disease control rate of 76.4%. Time to progression was 6.48 months (95% CI 3.95-9.14). In a phase 1/2 trial, evaluated the use of Tremelimumab in combination with ablation in patients with advanced HCC. Nineteen of the thirty-two patients enrolled were evaluable. Of these, five (26.3%, 95% CI 9.1-51.2%) achieved a partial response. In patients in whom a clinical benefit was observed, tumor biopsies at six weeks demonstrated an increase in CD8+ T cells. Six and 12-month tumor progression-free survival probabilities were 57.1% and 33.1%, respectively. The median OS was 12.3 months (95% CI 9.3-15.4 months). No dose-limiting toxicities were observed.

Durvalumab is a human immunoglobulin G1 kappa monoclonal antibody which blocks the interaction between the programmed cell death ligand 1 (PD-L1) (a member of the B7 family) with PD-1 (CD279). PD-L1 is expressed on several cells, including dendritic cells, macrophages, and many human cancer cells. PD-1 upregulation has been demonstrated on effector CD8+ T-cells in HCC. Moreover, both PD-1 and PD-L1 upregulation has been found to portend a poor prognosis in patients with HCC following surgical resection. In an interim analysis of a phase 1/2 trial of Durvalumab as monotherapy in HCC, out of forty patients, treatment-related adverse events occurred in 80.0% . Grade 3-4 treatment-related adverse events were reported in 20% of patients. Four patients achieved a partial response, and the objective response rate (complete response + partial response) was 10.3% (95% CI 2.9-24.2). Median OS was 13.2 months (95% CI 6.3-21.1). Following this, initial results of Study 22 a phase 1/2 randomized study were reported on Durvalumab and Tremelimumab in unresectable HCC (NCT02519348). In an updated analysis of Study-22 presented at the ASCO annual meeting longest survival in the advanced HCC population was reported with priming dose Tremelimumab 300mg x 1 dose (T300) in combination with Durvalumab 1500mg every four weeks. The overall response rate was 24% and median time to response 1.86 months. Median OS in this arm (18.7 months) was superior to single agent ICI and to low dose Tremelimumab in combination with Durvalumab. Eight patients or 10.8% stopped study due to treatment related adverse events. The HIMALAYA study, a randomized, multicenter phase 3 study of Durvalumab and Tremelimumab (300mg priming dose) as first-line therapy in unresectable HCC is closed to accrual.

Neoadjuvant therapy is commonly used in other solid-organ malignancies for three main reasons: downstage patients with advanced disease and to target micrometastatic disease, ensure patients are appropriately selected for curative treatment, and to evaluate tumour response to treatment before surgery. The role of neoadjuvant therapy in potentially resectable HCC is not clearly defined. Nonetheless, the aggressive biology of HCC, with high rates of tumour recurrence and de novo tumours after curative-intent resection, along with high immunosuppressive cell expression and CTLA-4 and PD-1 immune checkpoint upregulation, an investigation of the role of neoadjuvant immunotherapy in patients who present with resectable disease is warranted. Such treatment sequencing has demonstrated encouraging results in other lethal malignancies, such as non-small cell lung cancer. For example, evaluated the use of neoadjuvant PD-1 blockade (Nivolumab) in resectable (stage I, II, or IIIA) lung cancer (non-small-cell) in a pilot study. Nivolumab in the neoadjuvant setting had a side-effect profile that was acceptable and did not delay surgery. Complete surgical resection was obtained in 20 of the 21 tumors that were resected. Nine out of these twenty tumors (45%) demonstrated a major pathological response, which was correlated to the pre-treatment tumor mutational burden. Moreover, treatment with PD-1 blockade resulted in the expansion of mutation-associated, neoantigen-specific T-cell clones in peripheral blood.

Research hypothesis

• Pre-operative Durvalumab and Tremelimumab is safe in the pre surgical setting for upfront resectable HCC.

• Combination Durvalumab and Tremelimumab pre-operatively will result in changes in immune and molecular characteristics within the tumour microenvironment.

STUDY DESIGN This is a phase II, open label single arm, multicentre study of single dose Tremelimumab (T300) in combination with Durvalumab preoperatively in patients with upfront resectable HCC followed by adjuvant Durvalumab.

This study will enroll 28 patients at three academic institutions in Canada and Europe. Patients will receive 1 dose Tremelimumab (300 mg) with Durvalumab (1500mg) at cycle 1 (4W) and 1 further cycle of Durvalumab (1500mg) pre surgical resection. Post-surgical resection patients will begin adjuvant Durvalumab (1500mg Q4W) to complete 13 cycles of treatment (or 11 post operatively) in total.

Eligibility will be limited to patients with resectable HCC and preserved liver function. Prior systemic treatment for HCC will not be allowed although prior curative intent surgery or RFA is permitted.

All participants will be treated until progressive disease or unacceptable toxicity or withdrawal of consent or another discontinuation criterion is met. All participants will be followed for survival until the end of study.

No dose reductions of Durvalumab or Tremelimumab will be allowed. Dose modification and toxicity management for Durvalumab and Tremelimumab -related toxicity, including management of immune-mediated reactions, infusion-related reactions, and non-immune-mediated reactions will be provided in the Toxicity Management Guidelines (Appendix 1).

If the second cycle of Durvalumab is omitted due to toxicity, patients can resume adjuvant Durvalumab at the investigator's discretion and in accordance with protocol.

All patients enrolling will be required to undergo a pre-enrolment biopsy and for resected specimens will be required to provide consent to supply a sample of their tumor.

Scientific rationale for study design Rationale for primary endpoint The primary objective of this study is to assess the safety of Tremelimumab (T300) and Durvalumab pre-operatively in patients with resectable HCC followed by adjuvant Durvalumab (Q4W) for up to a maximum of 13 cycles or 11 post-operatively.Rationale for other secondary and exploratory endpoints The key secondary objectives are to evaluate feasibility, pathological response rates and overall response rates to pre-operative treatment.

Feasibility will be determined by the number of patients who had a delay in planned surgical resection (measured in days) due to TRAEs.

Major pathological response rates were recently demonstrated with the combination of CTLA-4 and PD-1 inhibition in resectable lung cancer. Furthermore, dynamic changes in immune correlates and microbiota composition were also documented.

Blood and tissue samples will be used to explore potential biomarkers in residual biological samples that may influence the progression of cancer and/or identify patients likely to have treatment benefit.

PD-L1 expression will be assessed and correlated with efficacy as a secondary objective. PD L1 is the target of Durvalumab, and it is hypothesized that its expression correlates with clinical efficacy.

Justification for dose Dose rationale for combination regimen of Durvalumab 1500mg Q4W plus Tremelimumab T300 The Durvalumab + Tremelimumab doses and regimen selected for this study are based on the goal of selecting an optimal combination dose of Durvalumab and Tremelimumab that would yield sustained target suppression (sPD-L1), demonstrate promising efficacy, and have an acceptable safety profile.

Rationale for 1 cycle of combination therapy with high dose Tremelimumab followed by Durvalumab monotherapy A summary of the existing PK and pharmacodynamic data has been utilized to guide the regimen selection for the combination of Durvalumab 1500 mg plus single dose of Tremelimumab 300 mg. Clinical Study Protocol Drug Substance Durvalumab (MEDI4736) and Tremelimumab Study Code D419CC00002 Version 1.0 Date 09 August 2017 39(219) Pharmacokinetics/pharmacodynamics data. The supporting data for this regimen are based on PK and pharmacodyamic data from regimens that used Tremelimumab doses of greater than 1 mg/kg from Study D4190C00006. An approximate dose-proportional increases in PK exposure (maximum plasma concentration and area under the plasma drug concentration-time curve from time 0 to Day 28 post-dose) was observed with increasing doses of Tremelimumab (1, 3, and 10 mg/kg). An exploratory pharmacodynamic analysis bioanalytically evaluated the effects of Tremelimumab on proliferating T-cells from NSCLC patients who received Tremelimumab (1, 3, or 10 mg/kg) and Durvalumab (15 or 20 mg/kg) combination treatment. Monotonic increases in pharmacodynamic activity with the combination (increased activation/ proliferation markers on CD4 and CD8 T-cells in periphery) were observed with increasing doses of Tremelimumab (1, 3, 10 mg/kg). The peak increase (%) from baseline of CD4+Ki67+ T-cells was observed 8 days post administration, and the peak level was significantly increased (p ≤ 0.05) as increasing dose of Tremelimumab in the range of 1 to 10 mg/kg. Study data also suggested that higher peak exposure (Cmax) of Tremelimumab is related to a higher maximum pharmacodynamic effect in the NSCLC patient population. Overall, the PK/pharmacodynamic data suggest that Tremelimumab of dose greater than 1 mg/kg with a higher peak exposure may be associated with a higher pharmacodynamic effect. Additionally, based on simulation data, the Cmax (78 µg/mL) post single dose administration of Tremelimumab 4 mg/kg is approximately 4-fold higher than the predicted Cmax (19 µg/mL) post the first dose of Tremelimumab 1 mg/kg, and is 3-fold higher than the predicted Cmax (25 µg/mL) post the fourth dose of Tremelimumab 1 mg/kg in a Q4W×4 doses setting.