The Evolving Role of Precision Medicine in Clinical Practice
Genetic alterations in molecular pathways are involved in tumor development, survival, and progression. Precision cancer medicine is about using the cancer genome to guide treatment decisions, according to Christine M. Walko, PharmD, BCOP, Personalized Medicine Pharmacologist, Personalized Medicine Clinical Service, and Chair, Clinical Genomic Action Committee, Moffitt Cancer Center, Tampa, FL.
“We know genetic alterations in molecular pathways make tumors grow,” she said at the 2019 Hematology/Oncology Pharmacy Association Annual Conference. “We have the technology to profile it, and we have drugs—on-label, off-label and in clinical trials—that we can utilize to target these profiles.”
Using the Cancer Genome to Guide Treatment
“It’s not just about the gene; it’s about the mutation that could be driving the cancer,” said Dr Walko.
Therefore, it is important to distinguish between the gene mutation that could be driving the cancer and the gene mutations that should not be targeted (ie, passenger mutations and benign germline mutations).
According to Dr Walko, tumor genomic interpretation should address several questions:
- Does the gene mutation provide information about prognosis?
- Does the gene mutation provide information that can be used to predict response or resistance to therapy?
- What are the patient characteristics and strength of evidence that need to be considered for treatment prioritization?
Not all mutations in known oncogenes are pathogenic, but the gene combined with the mutation could be pathogenic. For example, BRCA2 is a known tumor suppressor gene that is pathogenic when inactivated (eg, BRCA2 C1159), but some alterations can be benign (eg, BRCA2 K3326). “The big things that we’re looking for are driver mutations: tumor suppressor genes that are getting turned off and oncogenes that are getting turned on,” Dr Walko said. “We want to separate those out from passenger mutations.” Passenger mutations do not have an effect on a cell’s growth advantage and are less clinically relevant.
Types of Genomic Variants
Prognostic variants represent the underlying biology of the tumor, and provide information about cancer outcomes independent of treatment. For example, TP53 in chronic lymphocytic leukemia is associated with poor progression-free survival.
Predictive variants provide information about response to a specific treatment. For example, activating EGFR mutations are associated with increased response to EGFR inhibitors, such as erlotinib and gefitinib.
However, some biomarkers can be prognostic and predictive if they predict both outcomes and response to therapy.
“We’re looking for clinical actionability,” Dr Walko said. Actionability can be characterized by a genetic alteration that predicts a response to a certain therapy, or provides diagnostic or prognostic information, or by a clinically relevant germline alteration that informs disease risk. She explained that actionability may also mean investigating a clinical trial that is available for a patient’s particular alteration.
Off-Label Drug Acquisition
According to Dr Walko, the success of a patient receiving off-label drug therapy depends heavily on his or her insurance.
“We try to write our patient consults so that we can use them to get off-label drug therapy by including information about the genetic mutation, an explanation of the human data and the citations,” she said. “They want to see human data, even if it’s just one case. Then it’s about having the patience and the time to go back and forth with the insurance company and convince them.”
“We do see germline mutations bleed into our tumor interpretations,” Dr Walko said. If a tumor is analyzed with matched normal, healthy tissue, the alterations found in the normal tissue can be subtracted out. But commercial tests do not compare the 2 types of tissue, and if normal tissue is not analyzed, it is much more difficult to distinguish between cancer cells and healthy cells.
“So everything occurring in the germline gets pooled in with the cancer cells,” she explained. “In some cases, like with a germline BRCA mutation, we do want to target it, but most of the time we’re just trying to subtract those out.”
Establishing Clinical Evidence
Historically, precision medicine trials occurred in large subsets of the population (eg, HER2, BCR-ABL, and BRAF mutations). This enabled enrollment and completion of large-scale prospective studies, and demonstrated the clinical utility of the diagnostic marker (eg, BRAF V600E) and companion diagnostic tests, as well as the clinical efficacy of targeted drugs.
However, the continued growth of precision medicine in oncology is faced with challenges. In rare mutations, diverse levels of evidence support the value of a genetic biomarker being associated with a response to a targeted agent, and rare mutations across tumor types make enrolling enough patients on a randomized study logistically difficult.
“This is why with ROS1 or MET, we’re looking at a phase 1 trial. Even if we tried, we couldn’t do a randomized trial with enough power, unless we had 20 years to do it, and by then, we’ve already missed the boat,” Dr Walko said. “So we have to figure out right now, how to treat these patients and how to get this evidence.”
Basket and Umbrella Trials
Potential answers to these particular challenges include basket trials, which treat patients with specific agents targeting aberrant molecular pathways, independent of tumor origin (eg, the NCI MATCH trial), and umbrella trials, which assign patients with particular tumor histologies to treatment regimens specifically designed to target the tumor’s oncogenic molecular pathway (eg, the BATTLE-2 trial in lung cancer).
Dr Walko noted that these types of trials often produce lower levels of evidence. However, in general, precision medicine trials have shown benefit.
A multivariate analysis of 570 phase 2 novel therapy trials that enrolled more than 32,000 patients showed that treatment allocated by personalized approach consistently and independently correlated with higher response rates, longer median progression-free survival, and longer overall survival.
“So we do know that these work, we likely just need better drugs for the targets that we’re assessing,” Dr Walko said.
Molecular diagnostics are at the center of personalized healthcare. In July 2012, a rapid molecular test for Kirsten RNA associated rat sarcoma 2 virus gene (KRAS) mutations—the therascreen KRAS RGQ PCR Kit (therascreen KRAS test)—was the first to receive approval from the US Food and Drug Administration (FDA) as a [ Read More ]
At Johns Hopkins Hospital, each specialist in my practice sees approximately 8 to 10 patients with nonmetastatic NSCLC per month, some of whom are not candidates for surgery based on physiologic parameters. In most cases, we follow the NCCN Guidelines or ASCO clinical practice guidelines in our management of patients with early-stage NSCLC, except in clinical scenarios where the patient may not fit into a particular category within the guidelines, or when we enroll a patient in a clinical trial. For example, we may determine that a neoadjuvant clinical study is appropriate for a patient with stage IB NSCLC, whereas this recommendation is not concordant with the NCCN Guidelines. There are also instances in which we apply recently published clinical study data when managing our patients—even before the NCCN Guidelines have been updated to reflect the most recent findings.