I-SPY 2 Trial

I-SPY 2 Trial
Name	I-SPY 2 Trial
Other names	I-SPY 2
Phase	II/III adaptive
Condition	Breast cancer
Sponsor	QuantumLeap Healthcare Collaborative
Start date	2010
Status	ongoing

I-SPY 2 Trial The I-SPY 2 Trial is a multi-center platform trial for neoadjuvant breast cancer that uses adaptive randomization and biomarker-driven arms to evaluate investigational agents, linking institutions such as Dana–Farber Cancer Institute, Memorial Sloan Kettering Cancer Center, University of California, San Francisco, Massachusetts General Hospital, and networks including National Cancer Institute and Foundation Medicine. The study integrates molecular assays developed by groups like Agilent Technologies, Nanostring Technologies, and Foundation Medicine while collaborating with pharmaceutical partners such as Roche, Pfizer, Novartis, Merck & Co., and Amgen to accelerate drug development and regulatory decision-making with agencies including the Food and Drug Administration and European Medicines Agency.

Introduction

I-SPY 2 began as an evolution of earlier efforts at adaptive oncology trials following models from Cancer and Leukemia Group B, EORTC, National Surgical Adjuvant Breast and Bowel Project, NSABP, and consortia like ACOSOG and SWOG. The trial targets high-risk, locally advanced breast cancers characterized by markers used in classifications such as HER2, BRCA1, BRCA2, TP53, PIK3CA, and transcriptional signatures analogous to those from PAM50 and projects linked to The Cancer Genome Atlas. Designed to shorten timelines seen in traditional randomized controlled trials like HERceptin (trastuzumab) and trials informing adjuvant chemotherapy standards, the platform leverages resources from biobanks and cooperative groups including All of Us Research Program and Broad Institute collaborations.

Trial Design and Methodology

The I-SPY 2 design is an adaptive, Bayesian, phase II/III platform trial informed by statistical methods developed at institutions like Harvard University, Stanford University, University of Pennsylvania, and by statisticians who contributed to innovations used in trials such as BATTLE and FIRE-3. Patients are stratified by molecular subtypes defined by assays associated with Oncotype DX, MammaPrint, and expression profiles used in studies by Dana-Farber Cancer Institute teams, then randomized adaptively with Bayesian predictive probability rules similar to methods from MD Anderson Cancer Center and Johns Hopkins University. Endpoints include pathologic complete response (pCR) measured post-surgery, event-free survival modeled per conventions from NSABP B-18 and GeparSixto, and exploratory correlative endpoints leveraging sequencing pipelines from Broad Institute and Sequence Ontology resources.

Treatments and Adaptive Randomization

The platform tests multiple investigational agents and combinations contributed by companies including Genentech, Eli Lilly and Company, Bristol Myers Squibb, AstraZeneca, Celgene, and academic drug-development groups from University of California, San Diego and Yale University. Adaptive randomization reallocates enrollment probabilities toward regimens demonstrating higher posterior probabilities of success within biomarker-defined subgroups, an approach influenced by designs used in I-SPY 1, BATTLE-2, and oncology platforms like NCI MATCH and LUNG-MAP. Arms have included inhibitors targeting pathways involving PI3K, AKT1, mTOR, and immune checkpoints such as PD-1 and CTLA-4, with companion agents like chemotherapies derived from regimens developed at MD Anderson Cancer Center and targeted agents whose development histories intersect with trials like CLEOPATRA and KEYNOTE-522.

Key Outcomes and Findings

I-SPY 2 has reported multiple arms graduating to phase III based on predicted efficacy signals, with agents advancing to confirmatory trials similar to those leading to approvals in trials like IMpassion130 and SOLAR-1. The trial demonstrated that adaptive platform strategies can identify signals of increased pCR rates in biomarker-defined cohorts, influencing later-phase trials at centers such as Vanderbilt University Medical Center and Cleveland Clinic. Publications by investigators affiliated with University of California, San Francisco, Stanford University, and Dana–Farber Cancer Institute have documented operational outcomes, statistical performance, and examples where Bayesian decision rules guided discontinuation or expansion of arms comparable to decisions seen in NCI MATCH and I-SPY 1 follow-ups.

Biomarkers and Companion Diagnostics

I-SPY 2 emphasizes predictive biomarkers and companion diagnostics sourced from technologies pioneered by Agilent Technologies, Nanostring Technologies, Foundation Medicine, Guardant Health, and researchers linked to The Cancer Genome Atlas program. Biomarker strata include hormone receptor status involving ESR1 profiling, HER2 status derived from assays influenced by HercepTest developments, genomic alterations such as PIK3CA mutations, and immune-related signatures akin to those studied in TCGA immunogenomic analyses. The trial's biomarker framework parallels companion diagnostic strategies that led to regulatory approvals in cases like trastuzumab emtansine and alpelisib.

Regulatory Impact and Implementation in Practice

I-SPY 2 has engaged regulatory bodies including the Food and Drug Administration, European Medicines Agency, and national health technology assessment organizations to explore adaptive approval pathways and expedited development strategies similar to programs like Breakthrough Therapy designation and Accelerated approval. Lessons from I-SPY 2 inform policy discussions in forums attended by stakeholders from National Cancer Institute, American Society of Clinical Oncology, European Society for Medical Oncology, and payer groups such as Centers for Medicare & Medicaid Services, influencing how platform data may support labeling, coverage, or design of confirmatory phase III trials analogous to precedents set by KEYTRUDA approvals.

Criticisms and Limitations

Critiques of the platform echo concerns raised in debates around adaptive trials in literature from New England Journal of Medicine, The Lancet Oncology, and methodological critiques by statisticians at Columbia University and University of Oxford about potential operational bias, multiplicity, and generalizability to broader populations treated at centers like Mayo Clinic and Mount Sinai Health System. Limitations include reliance on pCR as a surrogate endpoint debated in policy forums involving FDA Oncology Center of Excellence and outcomes researchers from Johns Hopkins University Bloomberg School of Public Health, challenges in assay harmonization across vendors such as Agilent Technologies and Nanostring Technologies, and logistic complexity when coordinating industry partners like Roche and cooperative groups like SWOG.

Category:Clinical trials