KEYNOTE-991: pembrolizumab plus enzalutamide and androgen deprivation for metastatic hormone-sensitive prostate cancer

Prostate cancer is a leading global cause of cancer-related death among men [1,2]. Patients with localized disease have a favorable prognosis, with a 99% 10-year survival rate if treated early [1]. However, approximately 30% of men have recurrent disease after initial treatment [3]. For patients who develop metastatic disease, treatment options have evolved considerably; overall survival rate has improved; however, 5-year survival rates are poor at 30–50% for the subgroup of patients with de novo or high-volume disease [4–6].

Patients with metastatic hormone-sensitive prostate cancer (mHSPC) traditionally have been treated with androgen deprivation therapy (ADT) alone (i.e., luteinizing hormone-releasing hormone [LHRH] therapy with an agonist or antagonist or bilateral orchiectomy). ADT has shown substantial initial clinical benefits; however, over several years the disease tends to progress [4,7,8]. Recent developments have led to combination therapies being considered the standard-of-care for mHSPC, which now involves ADT plus a taxane-based chemotherapy such as docetaxel or a next-generation hormonal agent (NHA) such as enzalutamide, apalutamide or abiraterone acetate [7].

The ARCHES and ENZAMET trials in particular were designed to evaluate the addition of enzalutamide to ADT for the treatment of mHSPC. In ARCHES, enzalutamide plus ADT reduced the risk of radiographic progression-free survival (rPFS) by 61% (hazard ratio [HR]: 0.39; 95% CI: 0.30–0.50; p < 0.001) and extended overall survival (OS) (HR: 0.66; 95% CI: 0.53–0.81; p < 0.0001) compared with ADT plus placebo [9,10]. Likewise, adding enzalutamide to ADT offered a significant survival benefit compared with adding on a standard nonsteroidal anti-androgen (bicalutamide, flutamide or nilutamide) in the ENZAMET trial (HR: 0.67; 95% CI: 0.52–0.86; p = 0.0002) [11].

Adding apalutamide to ADT for men with mHSPC also improved 24 month OS rate to 82.4%, compared with 73.5% in the placebo plus ADT group in the TITAN trial, which translated to a 33% lower risk of death for men in the apalutamide group (HR: 0.67; 95% CI: 0.51–0.89; p = 0.005) [12]. Abiraterone acetate plus ADT also showed prolonged OS versus ADT alone in men with mHSPC in the STAMPEDE (HR: 0.63; 95% CI: 0.52–0.76; p < 0.001) and LATITUDE (HR: 0.62; 95% CI: 0.51–0.76; p < 0.001) trials [13].

In the phase III PEACE-1 trial, ADT combined with abiraterone and docetaxel with or without radiotherapy significantly improved rPFS (HR: 0.50; 95% CI: 0.40–0.62; p < 0.0001) and OS (HR: 0.75; 95% CI: 0.59–0.96; p = 0.021) compared with ADT plus docetaxel in men with mHSPC [14,15].

The optimal treatment sequence for mHSPC is unknown; most experts favor enzalutamide or abiraterone acetate in combination with ADT as first-line treatment for patients who experienced relapse or had low-volume disease, and a taxane as initial treatment for high-volume disease or second-line treatment at progression to metastatic castration-resistant prostate cancer (mCRPC), whereas evolving standards suggest that combining ADT/docetaxel with an androgen receptor inhibitor may further improve survival rates in some men with high-volume disease [15–18]. In addition, radiation to the prostate improved survival rates selectively in men with newly diagnosed M1 disease defined by standard imaging based on the STAMPEDE trial [14]. However, mHSPC invariably progresses, and all patients will eventually develop mCRPC, with a poor prognosis of 1–5 years depending on prognostic factors such as pain, patterns of spread, prior therapy and prostate-specific antigen (PSA) levels [19]. Therefore, there is a significant need for more effective treatments that provide durable PFS and OS benefits for patients with mHSPC to prolong the time to mCRPC.

KEYNOTE-991

We describe the rationale and design of the phase III KEYNOTE-991 (ClinicalTrials.gov, NCT04191096) study, which will be conducted to evaluate the efficacy and safety of pembrolizumab plus enzalutamide plus ADT compared with placebo plus enzalutamide plus ADT in NHA-naive participants with mHSPC.

Background & rationale

The programmed death 1 (PD-1) receptor–ligand interaction is a major immunomodulatory pathway that is often used by tumors to suppress antitumor immune activity [20]. Although the expression of PD-1 ligands 1 and 2 (PD-L1 and PD-L2) on prostate tumor cells is highly variable, the use of enzalutamide can drive further upregulation of PD-L1 expression in the tumor microenvironment [21,22]. The combination of antiandrogen therapy and immunotherapy targeting the PD-1/PD-L1 axis has been associated with potentially enhanced and durable response rates in men with mCRPC that has not responded to enzalutamide and in previously untreated patients [23,24]. Patients whose disease progressed during enzalutamide therapy also had a significantly increased number of PD-L1-/PD-L2-positive dendritic cells in their blood than treatment-naive participants and participants who continued to respond to enzalutamide [21]. Therefore, blocking the PD-1/PD-L1 interaction in the treatment of mHSPC warrants further evaluation because it may enhance the immune response against enzalutamide-resistant cells that emerge in response to enzalutamide treatment.

Pembrolizumab is a humanized monoclonal antibody that binds to PD-1 and prevents interaction with its ligands. The safety and the efficacy of pembrolizumab monotherapy in men with advanced prostate cancer have been demonstrated in two early phase trials. The results of the phase Ib KEYNOTE-028 trial (ClinicalTrials.gov, NCT02054806) reported that, in a cohort of 23 patients with heavily pretreated, PD-L1-positive mCRPC, four (17%) had an objective response, with a mean duration of response of 13.5 months. Pembrolizumab was found to have an acceptable safety profile in this trial; four patients (17%) experienced grade 3 or 4 treatment-related adverse events (TRAEs), and there were no treatment-related discontinuations or deaths [25]. In the phase II KEYNOTE-199 study (ClinicalTrials.gov, NCT02787005), patients with mCRPC who developed resistance to enzalutamide after a prior response were enrolled into cohorts 4 (measurable disease by Response Evaluation in Solid Tumors [RECIST], version 1.1; n = 81) and 5 (bone metastases or bone-predominant disease; n = 45) and treated with a combination of pembrolizumab and enzalutamide [26]. The objective response rate in cohort 4 was 12%, and disease control rate after more than 30 months of follow-up was 53% in cohort 4 and 51% in cohort 5 per RECIST v1.1 criteria. These responses were observed regardless of PD-L1 expression status. Of the 126 patients enrolled in cohorts 4 and 5, 35 (28%) experienced at least one grade ≥3 TRAE, and 18 (14%) discontinued pembrolizumab therapy because of a TRAE. Two patients died of TRAEs (myasthenic syndrome and Guillain–Barré syndrome).

The ongoing KEYNOTE-365 phase Ib/II trial (ClinicalTrials.gov, NCT02861573) is designed to assess the safety and efficacy of pembrolizumab combination therapy in patients with mCRPC. In cohort C (N = 102) of this study, the effects of pembrolizumab plus enzalutamide were evaluated in patients with mCRPC who were previously treated with abiraterone acetate [27]. In this cohort, the objective response rate in patients with RECIST-measurable disease was 10.5%. Median OS and rPFS were 20.1 and 6.0 months, respectively. In addition, of the 101 patients with baseline PSA measurements, 24 (24%) had a confirmed PSA decrease of at least 50% over a median time of 40 months from enrollment to data cut-off. Grade ≥3 TRAEs occurred in 43 (42%) patients, and one patient died of a TRAE (cause unknown) [27]. These results suggest that combining pembrolizumab and enzalutamide may be an effective treatment strategy in this setting, and the phase III KEYNOTE-641 study (ClinicalTrials.gov, NCT03834493) of pembrolizumab plus enzalutamide compared with enzalutamide alone in patients with mCRPC with or without prior abiraterone acetate treatment is ongoing [28].

Although prior studies of pembrolizumab therapy focused on patients with mCRPC, because of the long-term sequelae associated with antitumor immune response, there is an opportunity to improve the therapeutic efficacy of pembrolizumab if given at the initiation of ADT in patients with mHSPC. A pilot study of 12 patients with previously untreated mHSPC was conducted to assess the safety and feasibility of combination therapy with six cycles of pembrolizumab, short-term ADT (8 months) and prostatic cryoablation. After a median follow-up of 31.3 months, median PSA PFS was 14 months and median systemic therapy-free survival was 17.5 months. Five patients (42%) had PSA levels of 0.6 ng/ml or less at 1 year after treatment, and the median time to development of mCRPC was not reached. All reported adverse events (AEs) during the study period were grade 1 or grade 2 [29]. Therefore, these data from early phase trials may suggest a potential effect of pembrolizumab combination therapies in patients with mHSPC and warrant further evaluation.

Study design

KEYNOTE-991 is a randomized, placebo-controlled, parallel-group, multi-site, double-blind, phase III study of pembrolizumab in combination with enzalutamide and ADT versus placebo plus enzalutamide and ADT in participants with mHSPC (Figure 1). Approximately 1232 participants across 29 countries will be randomly assigned 1:1 to receive either pembrolizumab 200 mg iv. or placebo once every 3 weeks (Q3W), plus enzalutamide 160 mg orally once daily and ADT. There is no planned crossover between treatment arms. Patients and investigators will be blinded to pembrolizumab or placebo assignment, but the administration of enzalutamide will be open label. The dose and frequency of ADT administration will be determined per local product labeling and kept consistent throughout the study. Before random assignment, participants will be stratified by prior treatment with docetaxel (yes or no) and the presence of high-volume disease (yes or no).

Treatment with pembrolizumab/placebo may continue for up to 35 cycles (~2 years starting from the first infusion in cycle 1) or until study discontinuation or a lack of clinical stability is reached (defined as unacceptable toxicity, signs or symptoms indicating clinically significant disease progression, decline in performance status, rapid disease progression or threat to vital organs or critical anatomical sites necessitating urgent alternative medical intervention). Treatment with enzalutamide will proceed continuously from day 1 of cycle 1 in both arms. If pembrolizumab/placebo treatment is completed or discontinued for reasons other than progressive disease, participants still receiving enzalutamide will continue to receive enzalutamide and ADT until study discontinuation or a lack of clinical stability is reached.

Eligibility criteria

Key eligibility criteria are described in Table 1. Briefly, men ≥18 years of age with mHSPC (at least two bone lesions and/or visceral disease) are eligible for enrollment. Participants must maintain continuous ADT with an LHRH agonist or antagonist during study treatment or have a history of bilateral orchiectomy. Participants will be permitted to undergo up to 3 months of ADT before randomization, up to six cycles of docetaxel plus ADT completed 2 months or less before random assignment, one course of palliative radiation or surgical therapy at least 4 weeks before random assignment or one course of definitive radiation therapy for participants with low-volume metastatic disease completed at least 4 weeks before random assignment. Participants will be excluded if they received prior NHA, anti-PD-1, anti-PD-L1 or anti-PD-L2 therapy.

End points

The dual primary end points are OS, defined as the time from randomization to death from any cause and rPFS, defined as the time from randomization to radiographic progression per Prostate Cancer Working Group (PCWG) modified RECIST v1.1 as determined by blinded independent central review (BICR) or death from any cause (whichever occurs first) [30]. Soft tissue will be assessed per RECIST v1.1 modified to follow a maximum of ten target lesions and a maximum of five target lesions per organ, and bone disease will be assessed per PCWG criteria.

The secondary efficacy end points include time to initiation of the first subsequent anticancer therapy (TFST), defined as the time from randomization to TFST or death, whichever occurs first; PSA response rate, defined as a decline of at least 50% from baseline, measured twice, at least 3 weeks apart; objective response rate and duration of response per PCWG3 modified RECIST v1.1 as assessed by BICR; and time to PSA progression, defined as at least a 25% increase and at least 2 ng/ml above the nadir, confirmed by a second value at least 3 weeks later if a PSA decline from baseline occurs or at least a 25% increase and at least 2 ng/ml increase from baseline beyond 12 weeks if PSA decline from baseline does not occur. Time to pain progression, based on the Brief Pain Inventory-Short Form (BPI-SF) item three, opiate analgesic use based on analgesic quantification algorithm score, time to radiographic soft tissue progression and time to first symptomatic skeletal related event are also secondary efficacy end points. Safety and tolerability also will be assessed as a secondary end point.

Study procedures

Tumor imaging and assessment of disease will be evaluated using PCWG modified RECIST v1.1 by BICR. Initial tumor imaging at screening will be performed within 42 days before the date of random assignment. On-study imaging assessments will be performed every 12 weeks from the date of random assignment until study discontinuation due to disease progression per PCWG-modified RECIST v1.1 by BICR, the start of a new anticancer treatment, death, withdrawal of consent or the end of the study, whichever occurs first.

Chest, abdomen and pelvis imaging will be required for all participants at screening to determine eligibility and on study for all scheduled imaging visits; computed tomography (CT) with iv. and oral contrast is preferred, but MRI can be used if CT is contraindicated. Bone imaging (e.g., bone scintigraphy, radionuclide bone imaging) of the whole body will be required for all participants at screening and on study for all scheduled imaging visits. Disease progression in bone lesions should be confirmed by subsequent bone imaging at least 6 weeks after site-assessed first radiographic evidence of disease progression. Brain imaging will be required only as clinically indicated, both at screening and on study. If brain imaging is performed, MRI is strongly preferred, but CT with contrast is acceptable if MRI is contraindicated.

PSA levels will be centrally assessed to monitor disease status Q3W, from random assignment through week 12, then every 12 weeks thereafter. PSA screening will be performed within 10 days before random assignment.

Patient-reported outcomes, as assessed by the BPI-SF, Functional Assessment of Cancer Therapy-Prostate (FACT-P) and EuroQol 5-dimensions 5-level (EQ-5D-5L) questionnaires, will be administered on an electronic device at the site in the order listed every 21 days (on day 1 of every cycle) through cycle eight, then every two cycles through cycle 24, then every four cycles thereafter until discontinuation of study drug and at the safety follow-up visit. Only the BPI-SF will be administered at screening.

AEs will be monitored throughout the study and for 30 days after the end of treatment (90 days for serious AEs or 30 days if the patient initiates a new anticancer therapy) and will be graded using the National Cancer Institute Common Terminology Criteria for Adverse Events, version 5.0.

PD-L1 biomarker analysis will be assessed in a central laboratory before random assignment using the PharmDx assay per manufacturer’s instructions [31,32]. PD-L1 expression will be measured using combined positive score, which is the number of PD-L1-staining cells (tumor cells, lymphocytes and macrophages) divided by the total number of viable tumor cells, multiplied by 100. To identify novel biomarkers, the following biospecimens to support exploratory analyses of cellular components (e.g., protein, RNA, DNA and metabolites) and other circulating molecules will be collected from all participants: archival or newly obtained tumor tissue and blood for genetic, RNA, serum and plasma biomarkers and circulating tumor DNA analyses.

Statistical analysis

Primary efficacy analyses will be performed in the intention-to-treat population (all randomly assigned patients). Safety will be assessed in the all-participants-as-treated population, defined as all randomly assigned patients who received study drug, and will be analyzed by treatment received. At least one laboratory, vital sign or electrocardiographic measurement obtained after at least one dose of study intervention is required for inclusion in the analysis of the respective safety parameter.

The treatment difference for the primary end points (OS and rPFS) and selected secondary end points (time to initiation of the first subsequent anticancer therapy, time to symptomatic skeletal-related event and time to pain progression) will be evaluated by comparing the pembrolizumab and placebo study arms using a stratified log-rank test. The HR will be estimated using a stratified Cox proportional hazards regression model with the Efron method of handling ties. Event rates over time will be estimated within each treatment group using the non parametric Kaplan–Meier method.

Safety and efficacy interim analyses are planned. An external data monitoring committee will routinely review safety and efficacy interim results and will make recommendations for study discontinuation or protocol modifications.

Conclusion

Previous findings from the KEYNOTE-199 and KEYNOTE-365 studies suggest that pembrolizumab in combination with enzalutamide may be an active and well-tolerated treatment option for mCRPC. The phase III KEYNOTE-991 study is designed to assess the efficacy and safety of pembrolizumab or placebo in combination with enzalutamide plus ADT in NHA-naive participants with mHSPC. The results from this study will help define the role of combination therapy with pembrolizumab in an earlier setting for patients with mHSPC – a population for whom prolonging PFS to prevent the development of castration-resistant disease is crucial – and identify whether treatment with pembrolizumab and enzalutamide translates into a clinical benefit.