CrossMark
Pembrolizumab in paediatric patients with advanced melanoma or a PD-L1-positive, advanced, relapsed, or refractory solid tumour or lymphoma (KEYNOTE-051): interim analysis of an open-label, single-arm, phase 1-2 trial
Birgit Geoerger, Hyoung Jin Kang, Michal Yalon-Oren, Lynley V Marshall, Catherine Vezina, Alberto Pappo, Theodore W Laetsch, Antonio S Petrilli, Martin Ebinger, Jacek Toporski, Julia Glade-Bender, Wayne Nicholls, Elizabeth Fox, Steven G DuBois, Margaret E Macy, Susan L Cohn, Kumudu Pathiraja, Scott J Diede, Scot Ebbinghaus, Navin Pinto
Summary
Background Pembrolizumab is approved for the treatment of advanced cancer in adults; however, no information is available on safety and efficacy in paediatric patients. We aimed to establish the recommended phase 2 dose of pembrolizumab and its safety and antitumour activity in advanced paediatric cancer.
Methods KEYNOTE-051 is an ongoing phase 1-2 open-label trial. In this interim analysis, children aged 6 months to 17 years were recruited at 30 hospitals located in Australia, Brazil, Canada, France, Germany, Israel, Italy, South Korea, Sweden, the UK, and the USA. Patients with melanoma or a centrally confirmed, PD-L1-positive, relapsed or refractory solid tumour or lymphoma, and a Lansky Play/Karnofsky Performance status score of 50 or higher, received intravenous pembrolizumab at an initial dose of 2 mg/kg every 3 weeks. Pharmacokinetics and dose-limiting toxicities were used to establish the recommended phase 2 dose, and the safety and antitumour activity of this dose were assessed. Primary endpoints were determination of dose-limiting toxicities at the maximum administered dose, safety and tolerability, and the proportion of patients with objective response to pembrolizumab for each tumour type according to the Response Evaluation Criteria in Solid Tumours version 1.1 or the International Neuroblastoma Response Criteria. Safety and efficacy were assessed in all treated patients who received at least one dose of pembrolizumab. Separate reporting of the cohort of patients with relapsed or refractory classical Hodgkin lymphoma was a post-hoc decision. The data cutoff for this interim analysis was Sept 3, 2018. This trial is still enrolling patients and is registered with ClinicalTrials.gov, number NCT02332668.
Findings Of 863 patients screened between March 23, 2015, and Sept 3, 2018, 796 had tumours that were evaluable for PD-L1 expression (278 [35%] were PD-L1-positive); 155 eligible patients were enrolled and 154 had at least one dose of pembrolizumab. The median age of the enrolled patients was 13 years (IQR 8-15). Median follow-up was 8.6 months (IQR 2.5-16-4). No dose-limiting toxicities were reported in phase 1, and pembrolizumab plasma concentrations were consistent with those previously reported in adults; the recommended phase 2 dose was therefore established as 2 mg/kg every 3 weeks. Of the 154 patients treated, 69 (45%) experienced grade 3-5 adverse events, most commonly anaemia in 14 (9%) patients and decreased lymphocyte count in nine (6%) patients. 13 (8%) of the 154 patients had grade 3-5 treatment-related adverse events, most commonly decreased lymphocyte count in three (2%) patients and anaemia in two (1%) patients. 14 (9%) patients had serious treatment-related adverse events, most commonly pyrexia (four [3%]), and hypertension and pleural effusion (two [1%] each). Four patients (3%) discontinued treatment because of treatment-related adverse events, and two (1%) died (one due to pulmonary oedema and one due to pleural effusion and pneumonitis). Of 15 patients with relapsed or refractory Hodgkin lymphoma, two had complete and seven had partial responses; thus, nine patients achieved an objective response (60-0%; 95% CI 32- 3-83.7). Of 136 patients with solid tumours and other lymphomas, eight had partial responses (two patients each with adrenocortical carcinoma and mesothelioma, and one patient each with malignant ganglioglioma, epithelioid sarcoma, lymphoepithelial carcinoma, and malignant rhabdoid tumour); the proportion of patients with an objective response was 5.9% (95% CI 2·6-11-3).
Interpretation Pembrolizumab was well tolerated and showed encouraging antitumour activity in paediatric patients with relapsed or refractory Hodgkin lymphoma, consistent with experience in adult patients. Pembrolizumab had low antitumour activity in the majority of paediatric tumour types, and responses were observed in only a few rare PD-L1-positive tumour types, suggesting that PD-L1 expression alone is not sufficient as a biomarker for the selection of paediatric patients who are likely to respond to PD-1 checkpoint inhibitors. Final results of KEYNOTE-051, expected by September, 2022, with the possibility for extension, will report further on the activity of pembrolizumab in Hodgkin lymphoma, microsatellite instability-high tumours, and melanoma.
Lancet Oncol 2019
Published Online December 4, 2019 https://doi.org/10.1016/ S1470-2045(19)30671-0 See Online/Comment https://doi.org/10.1016/ S1470-2045(19)30803-4 Department of Paediatric and Adolescent Oncology, Gustave Roussy Cancer Campus, Université Paris-Saclay, Villejuif, France (B Geoerger MD); Department of Pediatrics, Seoul National University College of Medicine, Seoul National University Cancer Institute, Seoul National University Children’s Hospital, Jongno-gu, Seoul, South Korea (H J Kang MD); Pediatric Hemato-Oncology Department, Sheba Medical Center at Tel Hashomer, Ramat Gan, Israel (M Yalon-Oren MD); Paediatric and Adolescent Oncology Drug Development, The Royal Marsden Hospital and the Institute of Cancer Research, Sutton, UK (LV Marshall PhD); Pediatric Oncology, McGill University Health Centre, Montréal, QC, Canada (C Vezina MD); Oncology Department, St Jude Children’s Research Hospital, Memphis, TN, USA (A Pappo MD); Department of Pediatrics and Harold C Simmons Comprehensive Cancer Center, University of Texas Southwestern Medical Center/Children’s Health, Dallas, TX, USA (T W Laetsch MD); Federal University of São Paulo, Pediatric Oncology Institute (GRAACC-UNIFESP), São Paulo, Brazil (Prof A S Petrilli MD); Department of Pediatric Hematology and Oncology, Children’s University Hospital, Tübingen, Germany
(M Ebinger MD); Department of Pediatrics, Skåne University Hospital, Lund, Sweden (J Toporski MD); Division of Pediatric Hematology and Oncology and Stem Cell Transplantation, Columbia University Medical Center, New York, NY, USA (J Glade-Bender MD); Oncology Department, Lady Cilento Children’s Hospital, South Brisbane, QLD, Australia (W Nicholls MBChB); Developmental Therapeutics, Division of Oncology, Children’s Hospital of Philadelphia, Philadelphia, PA, USA (E Fox MD); Pediatric Hematology and Oncology, Dana-Farber/Boston Children’s Cancer and Blood Disorders Center, Boston, MA, USA (S G DuBois MD); Center for Cancer and Blood Disorders, Children’s Hospital Colorado, Aurora, CO, USA (M E Macy MD); Department of Pediatrics, The University of Chicago Medicine, Chicago, IL, USA (S L Cohn MD); Department of Medical Oncology, Merck & Co, Kenilworth, NJ, USA (K Pathiraja MA, S J Diede MD, S Ebbinghaus MD); and Hematology and Oncology, Seattle Children’s Hospital, Seattle, WA, USA (N Pinto MD)
Correspondence to: Dr Birgit Geoerger, Department of Paediatric and Adolescent Oncology, Gustave Roussy Cancer Campus, Université Paris-Saclay, 94805 Villejuif Cedex, France birgit.geoerger@ gustaveroussy.fr
Funding Merck Sharp & Dohme, a subsidiary of Merck & Co.
Copyright @ 2019 Elsevier Ltd. All rights reserved.
Introduction
Despite advances in elucidating the molecular mechanisms that underlie paediatric cancer, it remains the most fatal childhood disease in developed countries, and treatment typically includes non-specific, cytotoxic chemotherapy.1,2 Advances made in the past few years in adult oncology have led to increased use of immuno- therapy for the treatment of a range of tumour types. In particular, checkpoint inhibitors that block T-cell inhibitory molecules, such as cytotoxic T-lymphocyte- associated protein 4 (CTLA-4), PD-1, and PD-L1, have substantially improved outcomes in adults with cancer.1,3 As in adult patients, T-cell infiltrates have been found in the tumours of some children with cancer, supporting the rationale for immune checkpoint pathways as therapeutic targets for advanced paediatric cancers.4
Few studies of checkpoint inhibition have reported on the paediatric population. In phase 1 and phase 2 studies of advanced solid tumours, treatment with the CTLA-4 inhibitor ipilimumab showed similar pharmacokinetics and toxicity in children and adults.5,6 In a phase 2 study, which enrolled patients aged 12 to 17 years with advanced melanoma, ipilimumab was also shown to be safe and effective in some patients (at 1 year, three of four patients receiving ipilimumab 3 mg/kg and five of eight patients
receiving ipilimumab 10 mg/kg were alive; two patients on 10 mg/kg had partial response, and one on 3 mg/kg had stable disease; and at 3 years, one patient had achieved a durable partial response without further treatment).6 Based on these results, in 2017, the US Food and Drug Administration (FDA) expanded the indication for ipilimumab to include the treatment of unresectable or metastatic melanoma in patients aged 12 years and older.7 Until 2015, inhibitors of PD-1 and PD-L1 had not been explored prospectively in clinical trials involving paediatric patients. The present study and two other paediatric trials of PD-18 and PD-L19 inhibitors in children with advanced cancer should provide much-awaited initial insights into the role of immune checkpoint inhibition in these patients. Both of these parallel studies showed the good safety profiles of PD-1 and PD-L1 inhibitors, but low activity in classical paediatric cancers except Hodgkin lymphoma.
Pembrolizumab is a highly selective, humanised, IgG4 monoclonal antibody designed to block the binding of PD-1 to its ligands, PD-L1 and PD-L2. PD-1 forms a complex with either PD-L1 or PD-L2,10 and, under normal conditions, this complex functions to terminate the response of activated T cells.11 However, in tumour cells, the PD-1 pathway can be altered to overcome active T-cell
Research in context
Evidence before this study
We searched PubMed for studies published in any language from inception to May 31, 2019, which were on PD-1, PD-L1, and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) inhibitors in advanced paediatric cancer using the primary search terms “paediatric” OR “childhood” AND “PD-1” OR “PD-L1” OR “pembrolizumab” OR “MK-3475” OR “lambrolizumab” OR “CTLA-4” OR “ipilimumab” OR “nivolumab” OR “BMS-936558” OR “atezolizumab” OR “MPDL3280A” OR “durvalumab” OR “MEDI4763” OR “avelumab” OR “MSB0010718C”. We found two articles on ipilimumab in paediatric patients with cancer. A phase 1 study in paediatric patients with advanced solid tumours showed tolerability of ipilimumab, but no antitumour activity as a single drug. A phase 2 study of ipilimumab in adolescents with unresectable stage III or IV melanoma showed tolerability and antitumour activity. We searched PubMed again with the same search specifications using the search terms “paediatric” AND “PD-1” and narrowed our searches by selecting for clinical trial articles. We found 16 articles, of which none reported a clinical trial in paediatric patients with cancer. Congress abstracts from annual oncology meetings (including the American Society of Clinical Oncology) were also included. The final reference list was generated based on relevance to the scope of this paper.
Added value of this study
Data from our interim analysis of KEYNOTE-051 showed that PD-L1 positivity in tumour tissue was not a predictor of response to pembrolizumab in paediatric patients. The antitumour activity of pembrolizumab that we observed was primarily in relapsed or refractory Hodgkin lymphoma and in a few rare tumour types (most notably, mesothelioma and adrenocortical carcinoma). The treatment with pembrolizumab at our recommended phase 2 dose of 2 mg/kg every 3 weeks was well tolerated by the 154 enrolled paediatric patients across various tumour types. Final results of KEYNOTE-051 will report further on the activity of pembrolizumab in Hodgkin lymphoma, microsatellite instability-high tumours (MSI-H), and melanoma. Final results are expected by September, 2022, with the possibility for extension given the rarity of MSI-H tumours in children.
Implications of all the available evidence
The original adult dose of pembrolizumab approved by the US Food and Drug Administration (2 mg/kg every 3 weeks) can be safely administered to paediatric patients with advanced cancer and results in equivalent drug exposure. This dose shows antitumour activity in relapsed or refractory Hodgkin lymphoma and a few rare tumour types in paediatric patients.
immune surveillance and facilitate tumour progression.12 High expression of PD-L1 on tumour cells correlates with poor prognosis and poor survival in adult patients with various cancers.13 In paediatric solid tumours, low expression of PD-1, PD-L1, and PD-L2 mRNA and protein has been detected by NanoString nCounter analysis and immunohistochemistry.14 However, PD-L1 expression varies between tumour types, and its expression is not always associated with response to PD-1 inhibitors such as pembrolizumab in adult cancers.15
Pembrolizumab has shown robust antitumour activity and a favourable safety profile in several tumour types and is currently approved in multiple countries for one or more advanced malignancies, including several solid tumour types in adults (eg, melanoma, lung cancer, head and neck squamous cell carcinoma, urothelial cancer, gastric cancer, and microsatellite instability-high [MSI-H] cancer).10 In 2017, the FDA granted accelerated approval for pembrolizumab for the treatment of adults and children with relapsed or refractory Hodgkin lymphoma; relapsed or refractory primary mediastinal large B-cell lymphoma; unresectable or metastatic, microsatellite instability-high or mismatch repair-deficient solid tumours that progressed after treatment; and recurrent or metastatic Merkel cell carcinoma.10
In this Article, we report interim results from KEYNOTE-051, which, to our knowledge, is the first prospective clinical study of pembrolizumab in children with advanced cancer. We evaluated the safety of pembrolizumab, established the recommended phase 2 dose for paediatric advanced cancer, and evaluated the antitumour activity of this dose in a spectrum of tumours that were screened as PD-L1-positive.
Methods
Study design and participants
KEYNOTE-051 is an ongoing multicentre, non- randomised, open-label, single-arm, phase 1-2 trial. Patients were recruited from 30 hospitals located in Australia, Brazil, Canada, France, Germany, Israel, Italy, South Korea, Sweden, the UK, and the USA (appendix pp 1-2). Eligible patients were those aged between 6 months and 17 years with histologically or cytologically confirmed advanced melanoma or a PD-L1-positive, advanced, relapsed or refractory solid tumour or lymphoma with measurable disease according to the Response Evaluation Criteria in Solid Tumours (RECIST) version 1.1.16 Patients with neuroblastoma who did not have measurable disease per RECIST 1.1, but had meta- iodobenzylguanidine (MIBG)-positive evaluable disease were eligible. Under protocol amendments, patients with relapsed or refractory Hodgkin lymphoma (Dec 14, 2016; appendix p 7) and patients with MSI-H tumours (Aug 28, 2017; appendix p 10) were enrolled regardless of PD-L1 status (Aug 22, 2018; appendix p 14).
Apart from cases in which PD-L1 expression was not essential, patients were required to have tumour samples
(archival or freshly obtained) available for PD-L1 assessment. All patients required a performance score of 50 or higher on the Lansky Play scale for patients 16 years or younger or on the Karnofsky Performance scale for patients older than 16 years. Patients were required to have adequate haematological function (absolute neutrophil count ≥1000 cells per uL [≥750 cells per uL for patients with known bone marrow metastatic disease], platelets ≥100000 per uL [≥50000 per uL for patients with known bone marrow metastatic disease], and haemoglobin ≥8 g/dL for all patients), renal function (creatinine clearance ≥70 mL/min per 1-73 m2), and hepatic function (total bilirubin ≤1.5xupper limit of normal for age, alanine aminotransferase ≤110 U/L, and serum albumin ≥2 g/dL). Patients with active brain metastases and those who had received previous therapy with an anti-PD-1, anti-PD-L1, or anti-CTLA-4 drug were excluded. For additional eligibility criteria, see appendix (pp 20-24) and the protocol.
All participants or legally acceptable representatives provided written, informed consent. The study protocol and all amendments were approved by the institutional review board or independent ethics committee (or both) of each participating institution. The trial complies with the Declaration of Helsinki, International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use, and all local laws and regulations. The full protocol is available in the appendix.
Procedures
In phase 1, a modified 3+3 design (dose-finding) and a modified toxicity probability interval approach (dose- confirmation) were used to determine the paediatric recommended phase 2 dose.17 The initial dose in phase 1 was 2 mg/kg every 3 weeks, which is the originally approved clinical dose for adults.18 The dose could be de-escalated to a minimum of 1 mg/kg every 3 weeks on the basis of the proportion of patients with dose- limiting toxicities (target toxicity probability of approximately 25% for the maximum tolerated dose; evaluated during the first 3 weeks of treatment) and could be escalated to a maximum of 10 mg/kg every 3 weeks, per protocol. Doses every 2 weeks could have been explored depending on pharmacokinetic and pharmacodynamic data; however, 2 mg/kg every 3 weeks met the pharmacokinetic and pharmacodynamic criteria (outlined in the protocol; exposure <50% of that observed in adults19). Dose-limiting toxicity was defined as toxicity that met prespecified criteria (National Cancer Institute Common Terminology Criteria for Adverse Events [CTCAE] version 4.0), occurring during the dose-limiting toxicity evaluation period (the first cycle of treatment [21 days] for every 3-week dosing and the first two cycles of treatment [28 days] for every 2-week dosing, and was considered by the investigator to be related to study treatment. Further definition of
See Online for appendix
dose-limiting toxicity is provided in the appendix (pp 23-24) and the protocol. If the starting dose was generally safe and well tolerated and met the pharmacokinetic targets, the same dose was explored in phase 2 without de-escalation or escalation. Adaptive enrolment using a sequential monitoring technique (outlined in the protocol) was used in phase 2 (tumour expansion cohorts were determined by indication), and patients were treated with the paediatric recommended phase 2 dose as determined in phase 1. Dose reduction was not allowed in phase 2. In phase 2, patients received pembrolizumab at 2 mg/kg intravenously every 3 weeks.
In all patients, baseline PD-L1 expression in tumour samples (positivity defined as any staining of the stroma or PD-L1 expression on at least 1% of tumour cells) was assessed centrally at QualTek Molecular Laboratories (Goleta, CA, USA) using a prototype immunohisto- chemistry assay.20 The pharmacokinetic sample collection schedule is presented in the appendix (p 24).
Tumour imaging was done every 8 weeks for the first 6 months, then every 12 weeks thereafter by CT, MRI, or MIBG imaging. Response was assessed according to RECIST 1.1 or International Neuroblastoma Response Criteria (INRC) score21 by the local investigator at each site. Immune-related RECIST (irRECIST)22 was also used for making treatment decisions in cases of tumour progression. Laboratory safety assessments (haematology, chemistry, urinalysis, and other tests) were done weekly during the first cycle, then before each dose, as specified in the study protocol. Adverse events, including serious adverse events, were monitored throughout the study and for 30 days thereafter (90 days for serious adverse events) and graded according to the National Cancer Institute CTCAE 4.0. Adverse events were designated as treatment- related by the treating physician. Serious adverse events were defined as any event that resulted in death, was life threatening, resulted in persistent or significant disability or incapacity, resulted in or prolonged an existing inpatient hospitalisation, or any other important medical event. Additionally, a new cancer that was not a condition of the present study and any event associated with an overdose were considered as serious by the sponsor for collection purposes. Immune-mediated adverse events were derived from a list of terms specified by the sponsor, and events on this list were reported as immune-mediated irrespective of whether they were considered immune-related or treatment-related by the investigator.
Immune function analysis involved memory B-cell and T-cell profiling and measurement of vaccinated antibody titres at cycle 1 (pretreatment) and cycle 4 (post-treatment); these assays were selected to assess the effect of pembrolizumab on the immunological competence of paediatric patients. Whole blood samples collected in Cyto-Chex BCT tubes (Streck, Omaha, NE, USA) before dosing in cycle 1 (pretreatment) and before dosing in cycle 4 (post-treatment) were used for immune
function analysis. Memory B cells and T cells were profiled with two validated flow cytometry assays provided through Merck & Co (Kenilworth, NJ, USA). Cytotoxic T cells were identified by CD8 expression and helper T cells by CD4 expression. Memory CD4 and CD8 T cells were defined by the expression of CD45RO. Absolute numbers of memory B cells (defined as CD19+CD27+) and memory CD4 (CD3+CD4+) and CD8 (CD3+CD8+) T cells were calculated with supporting clinical Trucount assays (BD Biosciences, San Jose, CA, USA). Vaccinated antibody titres of tetanus toxoid and Streptococcus pneumoniae were measured with standard clinical grade assays. Summary statistics, change from baseline, and distributions of the change from baseline were recorded for memory B cells and T cells and vaccinated antibody titres.
Treatment continued until disease progression, defined by irRECIST,22 unacceptable toxicity, if the patient or legal representative withdrew consent, if the investigator decided to stop treatment, or after 24 months of treatment. Patients who had an investigator-confirmed complete response per RECIST 1.1 or INRC could stop treatment after receiving at least 24 weeks of pembrolizumab if they had received at least two treatments with pembrolizumab beyond the initial complete response. These patients and those who discontinued pembrolizumab after 24 months and achieved stable disease or better were eligible for an additional 1 year of treatment on radiographical disease progression if they met the criteria for retreatment (available in the protocol) and if the study was still ongoing. For patients who had confirmed disease progression or started a new anticancer therapy, survival status was collected every 12 weeks post- treatment until death, withdrawal of consent, or the end of the study.
Outcomes
The primary endpoint of the phase 1 part of the study was determination of dose-limiting toxicities during cycle 1 at the maximum administered dose of pembrolizumab, and the safety and tolerability of pembrolizumab. The primary endpoints of phase 2 were safety and tolerability, and the antitumour activity of pembrolizumab, measured as the proportion of patients achieving an objective response (complete or partial response) within each tumour type according to RECIST 1.1, or INRC for patients with MIBG-positive disease, as assessed by investigator review.
Secondary endpoints were characterisation of pharma- cokinetics (Cmax> Tmax, AUC, and Ctrough), duration of response, the proportion of patients with disease control, progression-free survival, and overall survival. Duration of response was defined as the time from first response to disease progression or death in patients who achieved a partial response or complete response. The proportion of patients with disease control was defined as those
863 patients provided a tumour sample for analysis of PD-L1 expression 796 had samples that were evaluable for PD-L1 expression 67 had samples that were non-evaluable for PD-L1 expression
64 patients with non-evaluable samples excluded
796 patients with known PD-L1 status 518 with PD-L1-negative tumours (5 with melanoma) 278 with PD-L1-positive tumours
3 patients with unknown PD-L1 status (2 with Hodgkin lymphoma and 1 with astrocytoma)
513 patients with PD-L1-negative tumours other than melanoma excluded
286 patients were potentially eligible
131 patients not enrolled in the study*
155 patients enrolled (12 in phase 1, 143 in phase 2) 147 with PD-L1-positive tumours
5 with PD-L1-negative melanoma
3 with missing PD-L1 status (2 with Hodgkin lymphoma and 1 with astrocytoma)
patients with a best overall response of complete response, partial response, or stable disease. Progression- free survival was defined as the time from first dosing to first-documented disease progression according to RECIST 1.1, or death from any cause, whichever occurred first. Overall survival was defined as the time from first dosing to death from any cause. Duration of follow-up was defined as the time from the first dose to death or to the database cutoff if the patient was still alive.
Other secondary endpoints were the measured changes in vaccinated antibody titres (tetanus and pneumococcus) and memory B cells (CD27+CD19+) and memory T cells (CD4+CD8+) with pembrolizumab treatment.
Additional secondary endpoints were proportion of patients who had an objective response, duration of response, proportion of patients with disease control, and progression-free survival per irRECIST; assessment of at least five of the following markers: NRAS, BRAF, MEK, KIT, PDGF, TP53, RB1, BRCA1, Akt phos- phorylation, IL-17, and PD-L1 in tumour tissue samples at baseline and in available samples at disease progression; evaluation of the relation between baseline tumour PD-L1 status and activity; and PD-L1 and PD-L2 expression and other biomarkers in responders versus non-responders. The results of some of these additional secondary endpoints, and pharmacodynamic and gene analyses, will be presented elsewhere.
| All enrolled and treated patients (n=154) | |
|---|---|
| Age | |
| Median | 13 (8-15) |
| 6 months to <2 years | 3 (2%) |
| 2-5 years | 22 (14%) |
| 6-9 years | 25 (16%) |
| 10-17 years | 104 (67%) |
| Sex | |
| Male | 80 (52%) |
| Female | 74 (48%) |
| Primary diagnosis | |
| CNS primary tumours | 38 (25%) |
| High-grade glioma* | 22 (14%) |
| Atypical teratoid rhabdoid tumour | 4 (3%) |
| Ependymoma | 4 (3%) |
| Medulloblastoma | 2 (1%) |
| Pilocytic astrocytoma | 2 (1%) |
| Low-grade astrocytoma not otherwise specified | 1 (<1%) |
| Other CNS primary tumourt | 3 (2%) |
| Other non-CNS solid tumours | 63 (41%) |
| Neuroblastoma | 11 (7%) |
| Melanoma# | 8 (5%) |
| Adrenocortical carcinoma | 4 (3%) |
| Hepatoblastoma§ | 5 (3%) |
| Hepatocellular carcinoma | 3 (2%) |
| Wilms' tumour | 3 (2%) |
| Renal cell carcinoma | 2 (1%) |
| Renal medullary carcinoma | 3 (2%) |
| Other solid tumour IT | 10 (6%) |
| Other carcinoma|| | 14 (9%) |
| Sarcoma | 33 (21%) |
| Osteosarcoma | 10 (6%) |
| Other non-rhabdomyosarcoma soft tissue sarcoma ** | 10 (6%) |
| Rhabdomyosarcoma not otherwise specified | 1 (<1%) |
| Embryonal rhabdomyosarcoma | 4 (3%) |
| Alveolar rhabdomyosarcoma | 2 (1%) |
| Inflammatory myofibroblastic tumour | 4 (3%) |
| Malignant rhabdoid tumour | 2 (1%) |
| Lymphoma | 20 (13%) |
| Hodgkin lymphoma | 18 (12%) |
| Diffuse large B cell lymphoma | 1(<1%) |
| Precursor T lymphoblastic lymphoma | 1 (<1%) |
| Lansky Play score/Karnofsky performance score | |
| 100 | 61 (40%) |
| 90 | 38 (25%) |
| 80 | 24 (16%) |
| 70 | 16 (10%) |
| 60 | 7 (5%) |
| 50 | 6 (4%) |
| Missingtt | 2 (1%) |
| (Table 1 continues on next page) |
| All enrolled and treated patients (n=154) | |
|---|---|
| (Continued from previous page) | |
| Overall staging | |
| I | 2 (1%) |
| IA | 2 (1%) |
| IB | 1 (<1%) |
| II | 5 (3%) |
| IIA | 4 (3%) |
| IIB | 2 (1%) |
| IIE | 1 (<1%) |
| III | 16 (10%) |
| IIIA | 3 (2%) |
| IIIB | 3 (2%) |
| IV | 87 (56%) |
| IVA | 1 (<1%) |
| IVB | 5 (3%) |
| Missing## | 22 (14%) |
| Brain metastases present | |
| Yes | 15 (10%) |
| No | 138 (90%) |
| Unknown | 1 (<1%) |
| Previous anticancer treatment | |
| Treatment naive | 7 (5%) |
| Systemic chemotherapy | 137 (89%) |
| Number of previous lines of systemic therapy | |
| 1 | 51 (33%) |
| 2 | 41 (27%) |
| ≥3 | 45 (29%) |
| Radiotherapy | |
| YesSS | 88 (57%) |
| No | 66 (43%) |
Data are median (IQR) or n (%). Percentages do not always equal 100% due to rounding. * High-grade glioma included glioblastoma (14 patients); grade 3 glioma and gliomatosis cerebri (one patient with each); and malignant glioma, anaplastic astrocytoma, and high-grade astrocytoma not otherwise specified (two patients with each). +Other CNS primary tumour included craniopharyngioma, meningioma, and paraganglioma (one patient with each). Melanoma included cutaneous melanoma, spitzoid melanoma, primary CNS melanoma, and ulcerated nodular melanoma (one patient with each), and melanoma not otherwise specified (four patients). §Hepatoblastoma included epithelial hepatoblastoma, epithelial mixed foetal and embryonal type, and mixed epithelial and mesenchymal hepatoblastoma with teratoid features (one patient with each), and hepatoblastoma not otherwise specified (two patients). ITOther solid tumour included hepatic malignancy not otherwise specified, mesoblastic nephroma, and malignant histiocytosis (one patient with each), mesothelioma and neuroendocrine tumour (two patients with each), and chordoma (three patients). ||Other carcinoma included clear cell soft tissue carcinoma, lymphoepithelial carcinoma, nasopharyngeal carcinoma, thymic cell carcinoma, adenocarcinoma, colloid carcinoma, colon cancer adenocarcinoma, lymphoepithelial carcinoma of the mediastinum, myoepithelial carcinoma of the thymus, mixed acinar-endocrine carcinoma, myoepithelial carcinoma, non-small-cell carcinoma, small cell carcinoma, and squamous cell carcinoma (one patient with each). ** Other non-rhabdomyosarcoma soft tissue sarcoma included angiosarcoma, epithelioid sarcoma, spindle cell sarcoma, and alveolar soft part sarcoma (one patient with each), and sarcoma not otherwise specified (six patients). ttDocumented as a protocol violation, but not discovered until after the start of treatment and so the patients were retained on the study. ##Overall staging was not required for diagnoses without standard staging systems. §§Includes ten patients who had radiotherapy only without chemotherapy.
Table 1: Baseline characteristics of enrolled and treated patients
Statistical analysis
For each tumour type, a minimum of ten patients were planned to be enrolled at the recommended phase 2 dose, including those already enrolled in phase 1. After enrolment of the first ten patients who, at the recom- mended phase 2 dose, had at least one post-baseline
response assessment of confirmed or unconfirmed complete response or partial response, per RECIST 1.1 (until enrolment of fewer than 25 patients), a sequential monitoring procedure was used to simulatenously evaluate efficacy and futility according to predefined rules outlined in the protocol. We did multiple interim analyses, suitable because of the sequential design of the trial, to monitor the proportion of patients with an objective response among enrolled patients for each indication. Based on the predefined futility guidelines and the entire safety and efficacy dataset (all enrolled patients who received at least one dose of study drug) across indications, enrolment to an indication could be stopped before a maximum of 25 patients was reached. This report is based on an interim analysis following a protocol amendment (Aug 22, 2018) to stop enrolment for most solid tumour indications (except of patients with microsatellite instability-high tumours, adolescent patients with melanoma aged ≥12 years to ≤18 years, and patients with relapsed or refractory Hodgkin lymphoma) because of poor signals of efficacy per futility rules (ie, that enrolment would be stopped if there were no responses in the first ten patients with at least one post-baseline response assessment. All changes made to the trial protocol after the start of the study are summarised in the protocol. Separate reporting of the cohort of patients with relapsed or refractory classical Hodgkin lymphoma was a post-hoc decision. The study had 84% power to show that the best overall response to pembrolizumab was greater than 10% at an overall one-sided 8% a level, if the true best overall response within an indication was 35%.
The dose-response relationship was estimated with all dose-limiting toxicity data from the trial using Bayesian pooling of adjacent violators.17 Safety and activity were assessed in all treated patients who received at least one dose of pembrolizumab. All patients with evaluable baseline and post-baseline scans were assessed for response, excluding patients who achieved stable disease less than 42 days after the first dose of pembrolizumab, as per the KEYNOTE clinical development programme. For the proportion of patients with an objective response or disease control, we determined the 95% CI using the exact binominal method. We used the Kaplan-Meier method to estimate duration of response, progression-free survival, and overall survival, and for post-hoc analyses of progression-free survival and overall survival at 6 and 12 months. We calculated summary statistics for Cmax (post-dose) and Ctrough (pre-dose) on the basis of nominal time after the first dose. The pharmacokinetic analysis was done with Phoenix WinNonlin software (version 6.3.0.395; Certara USA, Princeton, NJ, USA). We also did descriptive analyses of memory B-cell and T-cell counts and vaccinated antibody titres in patients who received at least one pembrolizumab dose and had available measurements at cycles 1 and 4. SAS software (version 9.4) was used for all analyses.
| Grade 1-2 | Grade 3 | Grade 4 | Grade 5 | |
|---|---|---|---|---|
| Total number of patients* | 74 (48%) | 11 (7%) | 0 | 2 (1%) |
| Fatigue | 12 (8%) | 0 | 0 | 0 |
| Anaemia | 10 (6%) | 2 (1%) | 0 | 0 |
| Pyrexia | 11 (7%) | 0 | 0 | 0 |
| Decreased lymphocyte count | 8 (5%) | 3 (2%) | 0 | 0 |
| Increased aspartate aminotransferase | 8 (5%) | 1 (<1%) | 0 | 0 |
| Hypothyroidism | 8 (5%) | 0 | 0 | 0 |
| Nausea | 8 (5%) | 0 | 0 | 0 |
| Rash, maculopapular | 8 (5%) | 0 | 0 | 0 |
| Diarrhoea | 8 (5%) | 0 | 0 | 0 |
| Abdominal pain | 7 (5%) | 0 | 0 | 0 |
| Increased alanine aminotransferase | 7 (5%) | 0 | 0 | 0 |
| Asthenia | 7 (5%) | 0 | 0 | 0 |
| Hyperthyroidism | 6 (4%) | 0 | 0 | 0 |
| Decreased white blood cell count | 6 (4%) | 0 | 0 | 0 |
| Decreased appetite | 4 (3%) | 0 | 0 | 0 |
| Pruritus | 3 (2%) | 1 (<1%) | 0 | 0 |
| Rash | 4 (3%) | 0 | 0 | 0 |
| Decreased platelet count | 4 (3%) | 0 | 0 | 0 |
| Arthralgia | 4 (3%) | 0 | 0 | 0 |
| Erythema | 4 (3%) | 0 | 0 | 0 |
| Colitis | 1 (<1%) | 1 (<1%) | 0 | 0 |
| Gastric ulcer | 0 | 1 (<1%) | 0 | 0 |
| Decreased neutrophil count | 2 (1%) | 1 (<1%) | 0 | 0 |
| Pleural effusion | 1 (<1%) | 0 | 0 | 1 (<1%) |
| Pneumonitis | 1 (<1%) | 0 | 0 | 1 (<1%) |
| Pulmonary oedema | 0 | 0 | 0 | 1 (<1%) |
| Hypertension | 2 (1%) | 1 (<1%) | 0 | 0 |
| Photosensitivity | 0 | 1 (<1%) | 0 | 0 |
| reaction | ||||
| Dyspnoea | 1 (<1%) | 1 (<1%) | 0 | 0 |
Data are n (%). Events of grades 3-5 occurring in at least one patient and corresponding grade 1-2 events are shown. * Some patients presented with more than one event at each grade, with the highest grade in each patient presented.
Table 2: Treatment-related adverse events
The data cutoff date was Sept 3, 2018. A per-protocol cohort of patients with relapsed or refractory Hodgkin lymphoma who were enrolled under a protocol amendment (Dec 14, 2016; appendix p 7) were not assessed for response because their response assessment was planned per the International Working Group (IWG) criteria.23
KEYNOTE-051 is registered with ClinicalTrials.gov, NCT02332668.
Role of the funding source
Funding for this research was provided by Merck Sharp & Dohme, a subsidiary of Merck & Co. The sponsor collaborated with academic advisers to design the study
| Grade 1-2 | Grade 3 | Grade 4 | Grade 5 | |
|---|---|---|---|---|
| Hypothyroidism | 13 (8%) | 0 | 0 | 0 |
| Hyperthyroidism | 6 (4%) | 0 | 0 | 0 |
| Infusion-related reaction | 5 (3%) | 0 | 0 | 0 |
| Thyroiditis | 2 (1%) | 0 | 0 | 0 |
| Adrenal insufficiency | 1(<1%) | 0 | 0 | 0 |
| Severe skin reactions | 0 | 1 (<1%) | 0 | 0 |
| Colitis | 1 (<1%) | 1 (<1%) | 0 | 0 |
| Pneumonitis | 2 (1%) | 0 | 0 | 1 (<1%) |
| Some patients presented with more than one grade in each patient presented. | event or reaction, with | the highest | ||
| Table 3: Immune-mediated adverse events and infusion reactions | ||||
and collect, analyse, and interpret the data. The sponsor funded medical writing and editorial assistance for this report. All authors had full access to all the study data and approved the decision to submit the manuscript for publication. The corresponding author had full access to all the data and had final responsibility for the decision to submit for publication.
Results
Between March 23, 2015 and Sept 3, 2018, 863 patients were screened for PD-L1 expression in tumour samples; 796 (92%) patients had evaluable tumour samples. Of these 796 patients with PD-L1-evaluable tumours, 518 patients had PD-L1-negative tumours (and were therefore not eligible for inclusion in the study) and 278 (35%) had PD-L1-positive tumours (figure 1 and appendix p 25-28). 155 eligible patients were enrolled at 30 sites in 11 countries (figure 1 and appendix pp 1-2).
Per protocol, 12 patients were enrolled in the phase 1 part of the study (six in the dose-finding part and six in the dose-confirmation part) and they were also included in phase 2. In phase 2, 143 patients were enrolled (dose expansion), giving a total 155 eligible patients across both phases. One patient enrolled to phase 2 did not receive study treatment because of rapid disease progression after enrolment; thus, a total of 154 patients received at least one dose of pembrolizumab.
At data cutoff (Sept 3, 2018), two (1%) patients had completed study treatment, 22 (14%) were still on study treatment, and 130 (84%) had discontinued treatment, mainly because of disease progression (as determined by imaging; in 93 [60%]), clinical progression (as determined clinically when no imaging was done; 19 [12%]), adverse events (seven [5%]), or physician or guardian decision (11 [7%]). At data cutoff, the median follow-up was 8.6 months (IQR 2.5-16.4). The median time on pembrolizumab was 8.3 months (4.9-18.3) for the Hodgkin lymphoma population (n=18 patients) and 1.4 months (0.7-3.5) for the solid tumour or other lymphoma population (n=136).
The median age of the treated population was 13 years (IQR 8-15; table 1). Patients had a range of primary
A Relapsed or refractory Hodgkin lymphoma
100
80
60
40
Change from baseline (%)
20
0
-20
-40
-60
-80
-100
B Any tumour type except Hodgkin lymphoma
100
Inflammatory myofibroblastic tumour
Hepatoblastoma
80
Wilms’ tumour
Atypical teratoid rhabdoid tumour
60
Non-rhabdomyosarcoma soft-tissue sarcoma
Osteosarcoma Other carcinoma
40
Change from baseline (%)
20
0
-20
-40
Medulloblastoma
Ependymoma
High-grade glioma
Malignant histiocytosis
☐ Non-Hodgkin lymphoma
Rhabdomyosarcoma
-60
Neuroblastoma
Mesoblastic nephroma
Glioblastoma
Neuroendocrine tumour
Chordoma
Hepatocellular carcinoma
Melanoma
Renal cell carcinoma Epithelioid sarcoma
-80
Adrenocortical carcinoma
Malignant rhabdoid tumour
Malignant ganglioglioma
Other CNS primary tumour
Mesothelioma
Lymphoepithelial carcinoma
-100
Patients
Shown are best percentage changes from baseline in the sum of the longest diameter of target lesions in patients with at least one post-baseline assessment and (A) relapsed or refractory Hodgkin lymphoma (15 patients; ) or (B) any tumour type except Hodgkin lymphoma (107). 29 patients with other tumour types discontinued before their first planned imaging assessment. Three of the 18 patients with Hodgkin lymphoma were pending response assessment per IWG criteria. The bars represent individual patients. Percentage changes of more than 100% were truncated at 100%. Dashed lines at 20% increase and 30% decrease represent the thresholds for consideration of progressive disease and partial response, respectively, according to RECIST 1.1 criteria. RECIST=Response Evaluation Criteria in Solid Tumours.
diagnoses. Most patients (n=147) had received previous treatment (137 had previously received systemic chemotherapy with or without radiotherapy; ten had received only radiotherapy). Of the seven (5%) patients who were treatment-naive, five had a primary diagnosis
of melanoma and two had other non-CNS solid tumours. Consistent with the study entry criterion of PD-L1 positivity for all tumour types except melanoma, 146 (95%) patients had PD-L1-positive tumours; five of the eight patients with melanoma had PD-L1-negative tumours. Two patients with Hodgkin lymphoma and one with anaplastic astrocytoma had missing PD-L1 status.
In phase 1, no dose-limiting toxicities were reported in patients treated with the initial dose of pembrolizumab at 2 mg/kg every 3 weeks. The pharmacokinetic para- meters of pembrolizumab following intravenous administration in paediatric patients is summarised in the appendix (p 29). The observed plasma concentrations at this dose (appendix pp 45) were consistent with those reported in adults administered the same dose.19 Therefore, the paediatric recommended phase 2 dose of pembrolizumab was determined to be 2 mg/kg every 3 weeks, and no additional doses were evaluated.
Because all the patients enrolled across phases 1 and 2 received the same dose of pembrolizumab, the safety results of the study phases were combined. Adverse events of any cause occurred in 149 (97%) of the 154 treated patients, but were mostly mild to moderate (grades 1 and 2 according to CTCAE 4.0); grade 3-5 adverse events occurred in 69 (45%) patients, most commonly anaemia in 14 (9%) patients and decreased lymphocyte count in nine (6%) patients (appendix pp 30-39). The only grade 3 or worse adverse events in phase 1 was grade 3 decreased lymphocyte count in two (1%) patients. Treatment was interrupted in 18 (12%) of 154 patients because of adverse events, most commonly increased alanine aminotransferase (n=three patients; [2%]; appendix pp 40-41). Seven (5%) patients discon- tinued treatment because of adverse events (appendix pp 42-43); four (3%) of these were considered treatment- related (grade 3 aspartate aminotransferase increase, grade 3 hypertension, grade 5 pleural effusion and grade 5 pneumonitis [in the same patient; pneumonitis being the reason for discontinuation], and grade 5 pulmonary oedema). Six (4%) patients had one or more adverse events that resulted in death (one patient with each of gastric adenocarcinoma, increased blood creatinine, malignant ependymoma, pulmonary oedema, sepsis, and one with pleural effusion and pneumonitis). Two of these deaths were considered potentially treat- ment related; a 15-year-old boy with chest sarcoma had pneumonitis at day 13 and pleural effusion at day 14 of treatment, and a 14-year-old girl with renal medullary carcinoma had pulmonary oedema at day 21 in the setting of sepsis following the first pembrolizumab administration.
Treatment-related adverse events of any grade occurred in 87 (56%) patients, with the most frequent being anaemia (in 12 [8%] patients), fatigue (12 [8%]), decreased lymphocyte count (11 [7%]), pyrexia (11 [7%]), increased serum aspartate aminotransferase (nine [6%]), and diarrhoea, hypothyroidism, nausea, and maculopapular
| Hodgkin lymphoma (n=15)*+ | Any other tumour type (n=136)* | |
|---|---|---|
| Objective response | 60.0% (32-3-83-7) | 5.9% (2-6-11-3) |
| Disease control+ | 80-0% (51.9-95-7) | 26.5% (19-3-34-7) |
| Best overall response | ||
| Complete response | 2 (13%) | 0 |
| Partial response | 7 (47%) | 8 (6%)S |
| Stable disease IT | 3 (20%) | 28 (21%) |
| Progressive disease | 3 (20%) | 74 (54%) |
| Non-evaluable|| | 0 | 2 (1%) |
| No assessment ** | 0 | 24 (18%) |
Data are n (%) or % (95% CI). * Only confirmed responses by Response Evaluation Criteria in Solid Tumours version 1.1 (RECIST v1.1) or International Neuroblastoma Response Criteria (INRC) are included; patients pending assessment have been excluded. tThree patients in a per-protocol Hodgkin lymphoma cohort were pending response assessment at data cutoff because their response was scheduled to be assessed with International Working Group criteria. Includes patients with complete response, partial response, and stable disease. §Patients with partial response had the following tumour histologies: adrenocortical carcinoma and mesothelioma (two patients with each), malignant ganglioglioma, epithelioid sarcoma, lymphoepithelial carcinoma, and malignant rhabdoid tumour (one patient with each). “[Patients had to have attained stable disease ≥42 days from the first dose of pembrolizumab. ||Two patients achieved stable disease on days 19 and 32, respectively, after the first dose of pembrolizumab and were therefore considered as non-evaluable per the KEYNOTE programme. ** 29 patients in the any other tumour type group with no assessments discontinued pembrolizumab before their first scheduled imaging assessment.
Table 4: Antitumour activity
rash (eight [5%] each; table 2). Grade 3-5 treatment- related adverse events occurred in 13 (8%) patients, most frequently decreased lymphocyte count (three [2%]) and anaemia (two [1%]; table 2). Treatment-related serious adverse events occurred in 14 (9%) patients, most commonly pyrexia (four [3%]), and hypertension and pleural effusion (two [1%] each). Other serious treatment-related adverse events were gastric ulcer, gastro-oesophageal reflux disease, tumour flare, pulmonary oedema, photosensitivity reaction, and pruritus (one patient with each). Two patients had multiple drug-related serious events: one had peripheral oedema, pyrexia, and enterocolitis infectious; and one had dyspnoea, pneumonitis, adrenal insufficiency, and pleural effusion.
Immune-mediated adverse events and infusion reactions, regardless of relation to pembrolizumab, occurred in 28 (18%) patients, the most common being grade 1-2 hypothyroidism (13 [8%]), hyperthyroidism (six [4%]), and infusion-related reactions (five [3%]; table 3). Treatment was modified (ie, interrupted or withdrawn) in four (3%) patients because of an immune- mediated adverse event (thyroiditis, pneumonitis, hyperthyroidism, or colitis). Median time to onset of immune-mediated adverse events and infusion-related reactions ranged from 13 days to 164 days after the initiation of treatment (appendix p 44).
Of the 154 treated patients, 29 (19%) in the other tumours cohort discontinued the study before their first planned imaging assessment. Three (2%) patients with
A
100
90
Progression-free survival (%)
80
70
60
50
40
30
20
10
0
0
4
8
12
16
20
Number at risk (number censored)
15 (1)
11 (3)
7 (0)
5 (0)
4 (2)
0 (0)
B
100
90
80
Overall survival (%)
70
60
50
40
30
20
10
0
0
5
10
15
20
25
30
Time since first study treatment (months)
Number at risk (number censored)
15 (4)
11 (3)
8 (1)
7 (0)
7 (6)
1 (1)
0 (0)
relapsed or refractory Hodgkin lymphoma were enrolled to the dedicated Hodgkin lymphoma cohort following protocol amendment, and their response assessment was planned to be done according to IWG criteria23 instead of RECIST 1.1 (results to be included in the final analysis report). 122 (79%) patients had at least one post-baseline assessment of target lesions according to RECIST 1.1, including the initial 15 patients with Hodgkin lymphoma. 45 (37%) of the 122 patients showed a reduction from baseline in the size of their target lesion, including the 15 patients with Hodgkin lymphoma, although some of these reductions were small (figure 2).
For the patients with Hodgkin lymphoma, the decrease in target lesion size was generally sustained over time in most cases (appendix p 47). Nine of the 15 patients with Hodgkin lymphoma achieved an objective response (60.0% [95% CI 32.3-83.7]); the best overall responses were two complete responses and seven partial responses; an unconfirmed partial response was reported for one additional patient. All 15 patients showed a reduction from baseline in the size of their target lesion (although
A
100
90
Progression-free survival (%)
80
70
60
50
40
30
20
10
0
0
5
10
15
20
25
30
Number at risk (number censored)
136 (4)
25 (3)
14 (4)
8 (5)
2 (1)
1 (1)
0 (0)
B
100
90
80
Overall survival (%)
70
60
50
40
30
20
10
0
0
8
16
24
32
40
Time since first study treatment (months)
Number at risk (number censored)
136 (8)
66 (21)
29 (14)
10 (4)
3 (3)
0 (0)
reduction was <30% reduction in three patients; figure 2A). 12 of 15 patients with Hodgkin lymphoma achieved disease control (80.0%; 95% CI 51-9-95-7; table 4). The median time to response in the nine confirmed responders with Hodgkin lymphoma was 1.9 months (IQR 1.8-1.9; appendix p 48). At data cutoff, four of the nine responders with Hodgkin lymphoma had ongoing responses and were still on treatment; the median duration of response was 17.3 months (95% CI 6.5-17.5; appendix p 50). Nine patients with Hodgkin lymphoma had progression-free survival events (disease progression, death, or started new anticancer therapy) by data cutoff; the median progression-free survival was 12.2 months (95% CI 2.1-19.4; figure 3A). In post-hoc analyses, progression-free survival at 6 months was 72.7% (95% CI 42.5-88.8) and at 12 months was 51.9% (95% CI 21-2-75-8). At data cutoff, no deaths were reported in the Hodgkin lymphoma population; thus, median overall survival was not reached, and estimated overall survival at 6 and 12 months (post-hoc analysis) was 100% (figure 3B).
In the cohort of 136 patients with any other tumour type, best overall responses were eight confirmed partial responses (5.9% [95% CI 2.6-11.3]). These eight partial responses were achieved in two patients with adrenocortical carcinoma, the only two patients with mesothelioma, and a single patient with each of malignant ganglioglioma, epithelioid sarcoma, lymphoepithelial carcinoma, and malignant rhabdoid (table 4). Overall, 30 patients with other tumour types showed a reduction from baseline in the size of their target lesion (figure 2B). In addition to the eight confirmed partial responses, some tumour reduction (<30% decrease) was seen in patients with high-grade glioma, embryonal rhabdomyosarcoma, and other non-rhabdomyosarcoma soft-tissue sarcoma (two patients with each), and glioblastoma, chordoma, ependymoma, hepatoblastoma, neuroblastoma, ganglio- neuroblastoma, mesoblastic nephroma, malignant histio- cytosis, inflammatory myofibroblastic tumour, and other carcinoma (one patient with each; figure 2B). The proportion of patients with solid tumours or other lymphomas who achieved disease control was 26.5% (95% CI 19. 3-34-7; table 4). The median time to response in the eight confirmed responders with other tumours was 1.9 months (IQR 1.9-3.6; appendix p 49). Four of the eight responders with other tumours or lymphomas had ongoing responses at data cutoff; the median duration of response was not reached (95% CI 7.1 months to not reached; appendix p 51), and five (63%) of the eight patients were estimated to have a response lasting at least 9 months. No patient showed a response after initial progression (suspected pseudoprogression).
118 (87%) of 136 patients with other tumour types had progression-free survival events by data cutoff; the median progression-free survival was 1.9 months (95% CI 1·8-1.9; figure 4A). In a post-hoc analysis, progression-free survival was estimated to be 18.7% (95% CI 12.6-25.8) at 6 months and 12.9% (7.8-19.4) at 12 months. At data cutoff, 86 (63%) of 136 patients with other tumour types had died; the median overall survival was 9.0 months (95% CI 6.2-14.5; figure 4B). Overall survival was estimated to be 59.1% (95% CI 50-2-66-9) at 6 months and 45.8% (37.0-54.2) at 12 months (post-hoc analysis), in patients with other tumour types.
Pretreatment and post-treatment blood samples were available to assess memory B cells in 50 patients, and to assess memory T cells in 49 of these 50 patients (including 11 patients with Hodgkin lymphoma). Absolute counts of memory B cells (CD19+CD27+) and memory T cells (CD45RA-CD45RO+ and CD8+CD45RA- CD45RO+) seemed to generally increase after pembro- lizumab treatment (appendix pp 52-55). Pretreatment and post-treatment vaccinated antibody titres were available for tetanus toxoid for 61 patients (11 with Hodgkin lyphoma) and for S pneumoniae vaccines for 59 patients (10 with Hodgkin lyphoma). Pembrolizumab treatment led to minimal changes in the antibody titres against both vaccines (appendix pp 56-58).
Discussion
In this phase 1-2 study of paediatric patients with advanced cancer, the pharmacokinetics and toxicity of pembrolizumab at 2 mg/kg every 3 weeks were similar to those reported in adult patients with cancer given the same dose.19 No dose de-escalation or escalation was needed, and pembrolizumab 2 mg/kg every 3 weeks was established as the recommended phase 2 dose. Previously, exposure of pembrolizumab at 2 mg/kg every 3 weeks19 was shown to be similar in terms of pharamacokinetics to that of the approved dose of 200 mg every 3 weeks in adult patients with advanced cancer.24 The recommended paediatric dose of pembrolizumab approved by the FDA is 2 mg/kg (up to 200 mg) every 3 weeks in patients with classical Hodgkin lymphoma, primary mediastinal large B-cell lymphoma, micro- satellite instability-high or mismatch repair-deficient cancers, and Merkel cell carcinoma.10
To assess whether or not pembrolizumab treatment had any effect on the developing immune system, we did immune function analyses in a subset of the paediatric patients. No major untoward effects on the developing immune system were detected. Although the immune function analyses were limited by the lack of assessment at multiple post-treatment timepoints, pembrolizumab treatment did not have a major effect on memory B-cell and T-cell profiles or on vaccinated antibody titres to the tetanus toxoid protein vaccine or S pneumoniae glycoprotein vaccine.
Nine (60%) of 15 evaluable children with relapsed or refractory Hodgkin lymphoma had confirmed objective responses; these responses were durable, consistent with the high proportion of objective responses and durations of response (median of 16.5 months) in adult patients with relapsed or refractory Hodgkin lymphoma.25 Based on the high frequency of responses, enrolment of these patients is ongoing, in accordance with a protocol amendment to add a cohort of patients with relapsed or refractory classical Hodgkin lymphoma to KEYNOTE-051. These patients will be assessed for response with revised IWG criteria23 instead of RECIST 1.1, used in this initial signal-finding study. Ultimately, data for this separate Hodgkin lymphoma cohort (three patients) will be presented alongside those for ongoing cohorts of paediatric patients with melanoma (ages 12-18 years) or MSI-H tumours (ages 6 months to 18 years). Antitumour activity with atezolizumab at 15 mg/kg (maximum dose 1200 mg) every 3 weeks has been reported in two of five patients with Hodgkin lymphoma in a study involving children and young adults (median age, 14 years).9 One explanation for these findings is that Hodgkin lymphoma is associated with 9p24.1 amplification, which drives PD-L1 and PD-L2 expression.26 Additional trials of check- point blockade in paediatric Hodgkin lymphoma (the iMATRIX9 trial and NCT023044588) will help to clarify the role of immune checkpoint inhibition monotherapy in Hodgkin lymphoma and other solid tumours.
Interestingly, in the ongoing CheckMate 744 trial (NCT02927769) involving patients aged 5-30 years with relapsed or refractory Hodgkin lymphoma, preliminary results showed that combination therapy with nivolumab and brentuximab vedotin led to 16 (64%) of 25 patients achieving a complete metabolic response, and four (16%) patients with a partial metabolic response.27 These trials are challenging the current treatment approach and might allow patients who respond well to checkpoint inhibitors and other targeted therapies to avoid treatment with radiotherapy and haemopoietic transplantation.
In our study, the two patients with mesothelioma both showed partial responses. Both patients had developed mesothelioma as a secondary malignancy after previous treatment for neuroblastoma and leukaemia. Further- more, two out of four treated patients with adrenocortical carcinoma achieved partial responses. However, although pembrolizumab is approved for adult patients with unresectable or metastatic melanoma, no objective responses were recorded in the eight paediatric patients with cutaneous melanoma enrolled in KEYNOTE-051. This lack of response might partially be explained by the fact that four of these patients were aged 3 years or younger, and the biological drivers of melanoma in this very young age group might be distinct from adolescent melanoma;28 as such the cutaneous melanoma cohort is still open for enrolment for adolescent patients with melanoma.
We included patients with PD-L1-positive, advanced, relapsed or refractory solid tumour or lymphoma because PD-L1 expression in tumours has been asso- ciated with improved clinical response to pembro- lizumab in some advanced relapsed or refractory tumour types.10 However, efficacy was low in patients with common paediatric solid tumours, despite an enriched population of patients with PD-L1-positive tumours (≥1% staining in tumour cells and adjacent immune cells as assessed by immunohistochemistry). This finding is consistent with clinical trials investi- gating nivolumab and atezolizumab as monotherapies in patients not screened for PD-L1 tumour expression.8,9 Plausible reasons underlying the poor activity of PD-1 or PD-L1 inhibition in paediatric tumours include low mutation rate,29 low expression of major histocom- patibility complex (which might impair a T-cell response to any neoantigen),30 the immaturity of the immune system in young children,31 and the immaturity of or alterations in the gut microbiome. 32,33 Nonetheless, emerging reports8,9 suggest further investigation of immune checkpoint inhibition is warranted in some childhood cancers. Given that this study followed a single-arm, open-label, phase 1-2 design, inherent limitations exist, such as patients and investigators not being masked to the drug. Additional limitations were a reliance on tumour PD-L1 status for patient selection, the short duration of follow-up, and the absence of an analysis of biological correlates of immune function
and response. Analysis of secondary endpoints not included within this investigation might provide further insight.
In adults, malignancies with high proportions of patients that respond to checkpoint inhibition (eg, melanoma and non-small-cell lung cancer34) have high mutational burdens, which might serve as neoantigens for a T-cell response and make these tumour types amenable to checkpoint inhibition.1 By contrast, only 6% of paediatric cancers are hypermutated.35 This result is consistent with a study that found a higher mutational load in biallelic mismatch repair-deficient glioblastoma multiforme than in sporadic paediatric and adult gliomas and other brain tumours; in the same study, two paediatric patients with germline biallelic mismatch repair-deficiency responded to nivolumab.36 Clinical trials are now underway to examine combined therapy with different checkpoint inhibitors (nivolumab and ipilimumab; NCT02304458) or with brentuximab vedotin (NCT02927769) in children with relapsed or refractory solid tumours and Hodgkin lymphoma.
The current study and others have shown that immune checkpoint inhibition is generally safe in paediatric patients. However, the small proportion of patients with an overall response highlights the challenges involved in treating this patient population, and suggests that tumour expression of PD-L1 alone (at least at a positivity level of 1%) is not sufficient as a biomarker to identify children who are likely to respond. Nevertheless, the preliminary results suggesting antitumour activity of pembrolizumab in Hodgkin lymphoma and some rare paediatric tumour types warrants further exploration. Future trials should help to define the patient populations and tumour types that are appropriate for immune checkpoint inhibition. Ongoing trials designed to evaluate combination therapies with other checkpoint inhibitors or immunomodulators also hold promise as treatments for challenging paediatric tumours. Final results of KEYNOTE-051 will report further on the activity of pembrolizumab in Hodgkin lymphoma, MSI-H tumours, and melanoma. Final results are expected by September, 2022, with the possibility for extension, given the rarity of MSI-H tumours in children.
Contributors
BG, EF, KP, SJD, and SE designed the study. BG, HJK, LVM, CV, ME, MY-O, TWL, JT, JG-B, WN, EF, SGD, MEM, SLC, ASP, and SJD collected data. JG-B oversaw the protocol. JT was responsible for the management of patients. BG, HYK, LVM, AP, KP, SE, and NP analysed the data. BG, HJK, LVM, AP, TWL, JG-B, EF, MEM, KP, SJD, SE, and NP interpreted the data. All authors participated in drafting the manuscript and reviewing iterations of the manuscript, and approved the final draft for submission.
Declaration of interests
TWL reports grants to his institution from Merck in support of the submitted work; grants and personal fees from Pfizer; and personal fees from Novartis, Eli Lilly, and Loxo Oncology, outside the submitted work. AP reports personal fees from Loxo Oncology, Bayer, and AbbVie, outside the submitted work. LVM reports advisory board honoraria from Bristol-Myers Squibb and Tesaro, outside the submitted work. ME reports receiving travel fees from Novartis and Jazz Pharmaceuticals, outisde the submitted work. JG-B reports grants to her institution from Merck and the National Institutes of Health (Children’s Oncology Group-National
Cancer Institute) in support of the submitted work; and support to her institution from Merck, Amgen, Eli Lilly, Pfizer, Bristol-Myers Squibb, Celgene, Eisai, Novartis Pharmaceuticals, and Ignyta, outside the submitted work. SGD reports grants from Merck, personal fees and travel support from Loxo Oncology, and travel support from Roche, outside the submitted work. MEM reports grants to her institution from Merck outside of the submitted work, and holds stock in Johnson & Johnson. SLC reports equities from Merck, outside of the submitted work. KP, SJD, and SE are employees of Merck Sharp & Dohme, a subsidiary of Merck, and SJD and SE hold stock in Merck. All other authors declare no competing interests.
Data sharing
The data sharing policy of Merck Sharp & Dohme, including restrictions, is available at http://engagezone.msd.com/ds_ documentation.php. Requests for access to the clinical study data can be submitted through the EngageZone site or via email to dataaccess@merck.com.
Acknowledgments
Funding for this study was provided by Merck Sharp & Dohme, a subsidiary of Merck & Co. We thank the patients and their families and caregivers and all primary investigators and site personnel for participating in the study. We also thank the following individuals from Merck & Co: Qing Zhao, Raluca A Budiu, and Richard Wnek (for immune function analyses), and Anne Chain (for pharmacokinetic analyses). Medical writing and editorial assistance was provided by Doyel Mitra, of the ApotheCom pembrolizumab team (Yardley, PA, USA); this assistance was funded by Merck Sharp & Dohme Corp., a subsidiary of Merck & Co.
References
1 Majzner RG, Heitzeneder S, Mackall CL. Harnessing the immunotherapy revolution for the treatment of childhood cancers. Cancer Cell 2017; 31: 476-85.
2 Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin 2018; 68: 7-30.
3 Majzner RG, Simon JS, Grosso JF, et al. Assessment of programmed death-ligand 1 expression and tumor-associated immune cells in pediatric cancer tissues. Cancer 2017; 123: 3807-15.
4 Vakkila J, Jaffe R, Michelow M, Lotze MT. Pediatric cancers are infiltrated predominantly by macrophages and contain a paucity of dendritic cells: a major nosologic difference with adult tumors. Clin Cancer Res 2006; 12: 2049-54.
5 Merchant MS, Wright M, Baird K, et al. Phase I clinical trial of ipilimumab in pediatric patients with advanced solid tumors. Clin Cancer Res 2016; 22: 1364-70.
6 Geoerger B, Bergeron C, Gore L, et al. Phase II study of ipilimumab in adolescents with unresectable stage III or IV malignant melanoma. Eur J Cancer 2017; 86: 358-63.
7 Bristol-Myers Squibb. YERVOY® (ipilimumab) injection, for intravenous use. September, 2019. https://packageinserts.bms.com/ pi/pi_yervoy.pdf (accessed Nov 18, 2019).
8 Davis KL, Fox E, Reid JM, et al. ADVL1412: initial results of a phase I/II study of nivolumab and ipilimumab in pediatric patients with replased/refractory solid tumors: a COG Study. Proc Am Soc Clin Oncol 2017; 35 (suppl): 10526.
9 Geoerger B, Zwaan CM, Marshall LV, et al. Atezolizumab for children and young adults with previously treated solid tumours, non-Hodgkin lymphoma, and Hodgkin lymphoma (iMATRIX): a multicentre phase 1-2 study. Lancet Oncol 2019; published online Nov 25. https://doi.org/10.1016/S1470-2045(19)30693-X.
10 KEYTRUDA® (pembrolizumab) for injection, for intravenous use. September, 2019. https://www.merck.com/product/usa/pi_ circulars/k/keytruda/keytruda_pi.pdf (accessed Nov 18, 2019).
11 Topalian SL, Drake CG, Pardoll DM, et al. Targeting the PD-1/B7- H1(PD-L1) pathway to activate anti-tumor immunity. Curr Opin Immunol 2012; 24: 207-12.
12 Thompson RH, Dong H, Lohse CM, et al. PD-1 is expressed by tumor-infiltrating immune cells and is associated with poor outcome for patients with renal cell carcinoma. Clin Cancer Res 2007; 13: 1757-61.
13 McDermott DF, Atkins MB. PD-1 as a potential target in cancer therapy. Cancer Med 2013; 2: 662-73.
14 Pinto N, Park JR, Murphy E, et al. Patterns of PD-1, PD-L1, and PD-L2 expression in pediatric solid tumors. Pediatr Blood Cancer 2017; 64: e26613.
15 Patel SP and Kurzrock R. PD-L1 expression as a predictive biomarker in cancer immunotherapy. Mol Cancer Ther 2015; 14: 847-56.
16 Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009; 45: 228-47.
17 Ji Y, Liu P, Li Y, Bekele BN. A modified toxicity probability interval method for dose-finding trials. Clin Trials 2010; 7: 653-63.
18 Chuk MK, Chang JT, Theoret MR, et al. FDA approval summary: accelerated approval of pembrolizumab for second-line treatment of metastatic melanoma. Clin Cancer Res 2017; 23: 5666-70.
19 Patnaik A, Kang SP, Rasco D, et al. Phase I study of pembrolizumab (MK-3475; anti-PD-1 monoclonal antibody) in patients with advanced solid tumors. Clin Cancer Res 2015; 21: 4286-93.
20 Dolled-Filhart M, Locke D, Murphy T, et al. Development of a prototype immunohistochemistry assay to measure programmed death ligand-1 expression in tumor tissue. Arch Pathol Lab Med 2016; 140: 1259-66.
21 Park JR, Bagatell R, Cohn SL, et al. Revisions to the International Neuroblastoma Response Criteria: a consensus statement from the National Cancer Institute Clinical Trials Planning Meeting. J Clin Oncol 2017; 35: 2580-87.
22 Seymour L, Bogaerts J, Perrone A, et al. iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics. Lancet Oncol 2017; 18: e143-52.
23 Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol 2007; 25: 579-86.
24 Freshwater T, Kondic A, Ahamadi M, et al. Evaluation of dosing strategy for pembrolizumab for oncology indications. J Immunother Cancer 2017; 5: 43.
25 Chen R, Zinzani PL, Fanale MA, et al. Phase II study of the efficacy and safety of pembrolizumab for relapsed/refractory classic Hodgkin lymphoma. J Clin Oncol 2017; 35: 2125-32.
26 Green MR, Monti S, Rodig SJ, et al. Integrative analysis reveals selective 9p24.1 amplification, increased PD-1 ligand expression, and further induction via JAK2 in nodular sclerosing Hodgkin lymphoma and primary mediastinal large B-cell lymphoma. Blood 2010; 116: 3268-77.
27 Harker-Murray P, Leblanc T, Mascarin M, et al. Response-adapted therapy with nivolumab and brentuximab vedotin (BV), followed by BV and bendamustine for suboptimal response, in children, adolescents, and young adults with standard-risk relapsed/refractory classical Hodgkin lymphoma. Blood 2018; 132: 927.
28 Lu C, Zhang J, Nagahawatte P, et al. The genomic landscape of childhood and adolescent melanoma. J Invest Dermatol 2015; 135: 816-23.
29 Gröbner SN, Worst BC, Weischenfeldt J, et al. The landscape of genomic alterations across childhood cancers. Nature 2018; 555: 321-27.
30 Haworth KB, Leddon JL, Chen CY, Horwitz EM, Mackall CL, Cripe TP. Going back to class I: MHC and immunotherapies for childhood cancer. Pediatr Blood Cancer 2015; 62: 571-76.
31 Simon AK, Hollander GA, McMichael A. Evolution of the immune system in humans from infancy to old age. Proc Biol Sci 2015; 282: 20143085.
32 Routy B, Le Chatelier E, Derosa L, et al. Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science 2018; 359: 91-97.
33 Park JA, Cheung NV. Limitations and opportunities for immune checkpoint inhibitors in pediatric malignancies. Cancer Treat Rev 2017; 58: 22-33.
34 Yarchoan M, Hopkins A, Jaffee EM. Tumor mutational burden and response rate to PD-1 inhibition. N Engl J Med 2017; 377: 2500-01.
35 Campbell BB, Light N, Fabrizio D, et al. Comprehensive analysis of hypermutation in human cancer. Cell 2017; 171: 1042-56.e10.
36 Bouffet E, Larouche V, Campbell BB, et al. Immune checkpoint inhibition for hypermutant glioblastoma multiforme resulting from germline biallelic mismatch repair deficiency. J Clin Oncol 2016; 34: 2206-11.