Pembrolizumab and Lenvatinib in Patients With Brain Metastases From Melanoma or Renal Cell Carcinoma

This is a phase 2, Simon's 2-stage designed study with 2 cohorts of anti-PD-1/PD-L1 experienced patients with untreated brain metastases: 1) melanoma and 2) renal cell carcinoma (RCC). The primary endpoint of this study is to determine the best intracranial response of combined pembrolizumab and lenvatinib in patients with untreated brain metastases from melanoma or RCC who are anti-PD-1/PD-L1--experienced. Secondary endpoints include best overall objective response (combined intracranial and extracranial response), progression-free survival (PFS), overall survival (OS), duration of intracranial response, and rate of adverse events. Exploratory/correlative endpoints will include evaluation of pre-treatment tumor tissue (either intracranial or extracranial) for immunohistochemical markers (PDL-1, tumor infiltrating lymphocytes, and angiogenic factors) and genetic analysis of tumor mutations or mutational burden. Pre-treatment and on-treatment blood samples will be collected and evaluated for biomarkers of response by cytokine profiling and transcriptomic analysis. Patients must have at least one evaluable asymptomatic intracranial lesion, no smaller than 5mm and no larger than 3 cm. Patients may have prior radiation to or surgical resection of a symptomatic brain metastasis as long as at least one untreated lesion or unequivocally growing lesion is present for response assessment. Pembrolizumab 200mg IV every 3 weeks will be administered in combination with lenvatinib 20 mg PO daily for up to 2 years. Patients must have received at least 2 doses of an anti-PD-1/PD-L1 mAb at some point in their treatment course and must have unequivocal intracranial progression. Intracranial progression in patients who are anti-PD-1/PD-L1 experienced is defined as either development of a new brain lesion(s) or unequivocal progression in a previously irradiated or resected brain lesion(s) site. Patients can be deemed to have progression after discussion and group consensus of the case at tumor board. Secondary imaging assessments to confirm intracranial progression are not required. Patients who are enrolled from sites outside of the Sponsor (Yale) must have patient eligibility approved by the Sponsor. Archival tissue from extracranial and/or CNS metastases will be obtained, if available, for correlative studies. A baseline pre-treatment fresh biopsy will also be required from an accessible metastasis unless there is not an easily accessible site to biopsy or if a biopsy is determined to be unfeasible by the treating physician after discussion with the study PI. The tumor tissue will be studied retrospectively for PD-L1 expression, TIL characteristics, and other immune and angiogenic markers that may predict sensitivity to this drug combination. First response assessment will be performed at 6 weeks and will include MRI of the brain and CT body scans (or other clinically indicated body imaging such as MRI or PET CT) to assess systemic disease. Brain metastasis response will be determined using modified RECIST (mRECIST) 1.1 criteria, and extracranial disease response will be determined using RECIST 1.1. Responses will be confirmed with repeat imaging done at 12 weeks. Subsequent imaging will be performed at 12-week intervals thereafter and include an MRI of the brain along with body imaging, which may include CT, MRI, or PET/CT, as clinically indicated. Patients will discontinue treatment for disease progression, unacceptable toxicity, patient withdrawal from the study, termination of study, or death. Patients may be treated beyond progression of intracranial metastases after consultation with the study PI, provided symptomatic lesions are treated with stereotactic radiosurgery (SRS) or surgery. Dose reduction of pembrolizumab for immune-related toxicities is not permitted. Dose-reduction of lenvatinib for related toxicities is permitted. Number of Patients: A total of up to 62 eligible patients will be enrolled (27 patients with melanoma with allowance for 3 additional patients if any are unevaluable and 29 patients with RCC with allowance for 3 additional patients if any are unevaluable). Either cohort can be stopped for futility according to Simon's two-stage design. The study will accrue for approximately 36 months and will be open for approximately 24 additional months as patients on study are followed.