Comprehensive Genomic Profiling Enables Efficient “N of 1” Trials

Wayne Kuznar

June 2015, Vol 6, No 5 - AVBCC 2015 5th Annual Conference


Washington, DC—Genomically driven medicine has created an opportunity for value in oncology by substantially improving outcomes and quality of life for patients with advanced cancer, said Vincent A. Miller, MD, Chief Medical Officer, Foundation Medicine, Cambridge, MA, at the Fifth Annual Conference of the Association for Value-Based Cancer Care.

Two premises behind the genomic characterization of cancer are:

  • Cancer is a disease of the genome, with each profile unique to the patient
  • Genomically driven treatment with targeted agents is generally safer, more efficacious, and more cost-­effective than traditional systemic treatment applied universally across a population with advanced cancer, said Dr Miller.
As our understanding of cancer complexity and the pathways through which cancers are activated is increasing, a comprehensive approach is required to best inform pathway-driven cancer care. “Comprehensive genomic profiling is an optimized solution to improve outcomes and manage utilization costs,” Dr Miller said.

“Our approach at Foundation Medicine is to get the data,” he said. “To get the data, one needs a rigorous, well-­validated genomic characterization of the patient’s tumor, and a way to extract the data in a believable, accurate, and understandable format from which multiple stakeholders can ask their questions and assess the value for them.”

Types of Genomic Testing

Three types of genomic testing are done in clinical practice today:

  • Category 1 tests are single gene markers
  • Category 2 tests are hot spot next-­generation sequencing panels that focus on a narrow subset of genes
  • Category 3 tests comprise comprehensive genomic profiling.
Category 3 tests sequence the entire coding region of cancer genes, not only the high-probability common regions. Genomic profiling based on comprehensive next-generation sequencing focuses on all exons of the 315 known clinically and biologically relevant cancer genes, and the introns of 28 genes that are frequently rearranged in human cancer.

This type of profiling has validated high accuracy achieved by high, uniform coverage of >99.5% of exons covered more than 100 times. It sequences from all clinically relevant specimen types, including those from fine-needle aspiration.

The Value of Comprehensive Genomic Profiling

Non–small-cell lung cancer (NSCLC) has proved to be fertile ground for establishing the value of comprehensive genomic profiling in patients with cancer, said Dr Miller. He illustrated the impact of comprehensive genetic profiling by showcasing a female patient with NSCLC with a rare mutation who received targeted therapy that significantly extended her survival.

The woman presented in mid-2010 with NSCLC and visual changes, and was found to have bilateral retinal metastases. She received systemic chemotherapy and did well for 18 months before her disease progressed with bony metastases. A tumor specimen sent to Foundation Medicine was profiled and was found to harbor a never-before-seen TRIM33-RET fusion, for which there were no approved therapies. Foundation Medicine documented that a clinical trial of cabozantinib (Cometriq) for patients with RET fusions was available. The woman had symptomatic and radiographic improvement for 7 months, her vision improved, and her bone pain resolved with the anti-RET therapy.

One year after identifying RET gene fusions, Cancer Discovery published interim results of the clinical trial (Drilon A, et al. 2013;3:630-635), when 3 of the first 6 patients with RET fusion–positive lung adenocarcinoma had responded to cabozantinib. “We went from a discovery to an N of 1,” said Dr Miller.

Initial profiling of lung adenocarcinomas using next-generation sequencing is an efficient and sensitive strategy. Among a series of 47 patients with no or light smoking history and “pan-negative” lung adenocarcinomas, next-generation sequencing was successfully performed in 31 patients and uncovered ?1 genomic alterations in 94% of patients. A total of 96 genomic alterations were identified.

An actionable genomic alteration with a targeted agent based on National Comprehensive Cancer Network guidelines was found in 26% of patients, and a targeted agent in a clinical trial in 39%. A genomic alteration that was not previously detected in lung cancers (SHC2-ERBB2) was found in 1 patient.

Some of the challenges of comprehensive genomic profiling were also laid out. “Much of the testing is done in patients with advanced disease who are unlikely to have multiple opportunities for additional systemic therapy,” Dr Miller said. Many times, tissue available for testing has been depleted. In addition, unpredictable genomic alterations are often detected, and the likelihood of a molecular match is proportional to the number of trials available. Finally, because the patients have advanced disease, they are mostly eligible for phase 1 trials that often have “starts and stops” because of the toxicities that are discovered.

Umbrella Trials versus Basket Trials

The 2 recently introduced clinical trial designs—umbrella trials and basket trials—can quickly evaluate novel targeted agents in patients with cancer and specific mutations.

Umbrella trials test the impact of different drugs on various genetic mutations in a single type of cancer. Basket trials test the effect of drugs on a single genetic mutation in a variety of cancer types.

An example of an umbrella trial is MASTER LUNG-1, which is being conducted in squamous-cell lung cancer. Patients undergo biomarker profiling, are grouped according to the type of genomic alteration found, and are randomized to a targeted therapy or standard of care (usually chemotherapy). The trial has multiple phase 2 and phase 3 arms, with “rolling” opening and closure.

The basket approach is exemplified by the neratinib HER2 mutation basket study, in which a HER2 mutation is identified in patients with bladder, colon, endometrial, gastric, ovarian, or other solid tumor cancer and is then treated with neratinib until disease progression or the development of intolerable side effects.

These “N of 1” trials could save millions of dollars that are spent on inappropriate interventions and conventional phase 3 trials, said Dr Miller. For example, in the classical pathway approach, for every patient that the 10 most frequently prescribed drugs help, they fail to improve the condition of 3 to 24 patients. The aggregated results of many “N of 1” trials will offer information about how to better treat subsets of the population, or even the population at large, Dr Miller argued.