OR WAIT null SECS
The road to precision medicine as a fixture in healthcare delivery is still a journey, but recent advances and new understandings in gene profiling, biomarker development, and AI and data analytics are steadily bringing these therapies closer to the individual patient every day.
Personalized medicines are increasingly being approved by FDA. In 2021, the agency cleared 17 such therapies—designed to treat specific anomalies or mutations in patients with genetic susceptibility to associated diseases. According to a recent study1, approvals for personalized medicines rose from under 10% to more than a quarter of all approvals in the last decade.
These drugs use biomarkers or mutations to identify alternative treatment plans for certain patients, while also considering patients’ social, economic, and environmental conditions.
In her work as vice president of clinical operations at AccessDx Laboratory, Houda Hachad, PharmD, identifies certain biomarkers to predict the functions of certain medications, and tailors therapies to patients’ specific needs. By doing so, she, like other healthcare providers (HCPs), can create treatment plans that have a higher likelihood of working for that particular patient.
“This allows us to provide the most effective therapy or the most appropriate medication without resorting to a trial-and-error approach, which most prescribers use,” she tells Pharm Exec.
Hachad describes tumor profiling as one of the most effective applications of precision medicine. Using new technologies, she and other researchers have been able to study specific tumors in individual patients and identify a variety of drivers that cause these cells to multiply. She explains that two patients may have the same cancer, but the underlying reasons that their tumors are growing may be different.
“By studying these alterations and drivers, companies were able to design medications to reverse these processes,” says Hachad. “The FDA has approved a number of medications in cancer that are targeted therapies because they really target a specific tumor that is driven by a specific alteration.”
She describes a shift in cancer treatment away from traditional approaches, such as chemotherapy, which is designed to kill all of the tumor cells regardless of what drivers are involved. By using precision medicine, doctors may be able to target those drivers while avoiding the toxicity associated with chemotherapy.
In late June, FDA approved a treatment for patients with metastatic solid tumors that have a specific type of mutation that can trigger an increase in tumor growth and the spread of cancer cells. The BRAF V600E mutation is known to produce such a scenario in more than 20 different tumor types.
This particular BRAF mutation is present in a small number of patients, however. For example, only 10% to 12% of metastatic colorectal patients express this mutation, although it can cause serious complications in treatment.
On June 23, FDA granted accelerated approval for Novartis drugs Tafinlar (dabrafenib) and Mekinist (trametinib) as treatments for adults and children over 6 years old who have tumors with the BRAF V600E mutation. This type of therapy is considered a tumor-agnostic treatment, meaning that the drugs treat the molecular or genetic makeup of the cancer, as opposed to fighting it based on the location or type of cancer. These two therapies demonstrated positive results during clinical trials in treating a wide variety of cancers, such as gastrointestinal cancers, glioma, and biliary tract cancer. As long as the BRAF V600E mutation was present, the drugs tended to yield positive results.
From a precision medicine perspective, the duo shows promise. As opposed to treating the cancers based on traditional methods, doctors can test patients for this mutation and treat them based on their specific genetics. The downside of these and similar therapies is they don’t offer much help to patients that don’t fit the genetic makeup. For example, Tafinlar and Mekinist did not yield strong response in tumors with other forms of BRAF mutations.
Recent technological innovations and trends have contributed to the growth of personalized medicine. The biggest challenge physicians and clinical teams now face with this field of treatment is data collection. While this is important in all areas of healthcare, it takes on even more significance when therapies and treatment plans are being tailored to specific individuals.
Two advancements have helped significantly in this area. First, the rise of wearable smart devices has made it much easier for medical professionals to collect data and monitor patients. More importantly, the development and refinement of artificial intelligence (AI) technologies have made it easier for researchers to organize this data and track more specific trends and data points. Pharma companies can use these technologies for more accurate identification of biomarkers and better recognition of behavioral patterns in patients that may need to be addressed.
New tools for data collection allow researchers and doctors to tailor treatment plans and medications to fit the specific needs of each patient. This reaches beyond genetics and takes patients’ lifestyles and circumstances into account as well.
Dr. Rich Christie, MD, PhD, a neuroscientist and chief medical officer at AiCure, uses AI to help physician investigators and site teams keep track of patients during clinical trials. Since the program utilizes patients’ smartphone cameras and microphones, it also provides physicians more opportunity to analyze the state of the patient. Using this information, researchers can detect signs of certain conditions.
“If you think about things like depression, schizophrenia, or Parkinson’s disease, each of those patient states are things that a physician could observe or listen for just sitting across from a patient,” explains Christie.
While the program can’t be used to diagnose diseases, it could prove valuable in identifying certain warning signs or symptoms that can trigger a physician to call the patient in for an examination. These are known as digital biomarkers, Christie points out.
Amy Brown is the founder and CEO of Authenticx, a conversational intelligence company that helps healthcare organizations listen to everyday customer conversations using AI and algorithm-based technology that can analyze calls, chats, and emails based on pharma-specific criteria. For example, if a customer calls a doctor’s office or clinic’s billing department to explain that they can’t afford the treatment and won’t be coming in for follow-up appointments, that call can be flagged to the care team so that they can explore alternative solutions.
Brown says that this is one example of systemic and long-rooted reasons why some patients may not start or continue on therapy. Driving factors can include the complexity of the healthcare system and economic and cost concerns. For healthcare companies, identifying the reasons preventing patients from beginning or continuing therapy is vastly important for creating and implementing potential solutions.
“Listening to the patient population helps leaders understand—in a very human way—the people that they’re serving,” says Brown. “And from that process, they’re able to deliver care in a more personalized way.”
These types of data help HCPs to understand the barriers that prevent patients from obtaining the treatment or therapy they need. Armed with such insights, HCPs can better equip themselves to build treatment plans tailored to the specific needs of their patients and the resources available, for example, caregivers or healthcare funding initiatives.
The AI programs used by Authenticx and AiCure leverage mobile channels to gather specific data. Christie says that, in his experience, patients have been very open to incorporating their phones into their treatment plans. While there are inherent privacy concerns associated with data collected from people’s homes, patients, according to Christie, seem to respond well to using their phones to collect it themselves.
“What exactly do you use to collect the data?” asks Christie. “Is it something like a cell phone that many people are very comfortable using and is integrated into people’s lives in ways that few technologies are? The amount of time that many people spend with their phones is remarkable.”
This data can facilitate conversations between patients and physicians. The use of personalized medicine plans requires doctors to know as much about their patients as possible, from genetic, behavioral, and situational perspectives.
FDA granted recognition to the first tumor mutation database in October 2021, based on a partial listing of the Memorial Sloan Kettering Cancer Center’s Oncology Knowledge Base. The database is considered a valid source for scientific evidence for Level 2 and 3 biomarkers and will be included in the Public Human Genetic Variant Database. With this update, developers can use this data in tumor profiling tests and evaluations for premarket submissions.
FDA recently approved several new indications for existing personalized medicine therapies. These include a new indication for Tecartus (brexucabtagene autoleucel), first approved to treat mantle cell lymphoma. The new indication covers use in patients with acute lymphoblastic leukemia.
“With any type of new technology that is offered and is brought to patient care, we need to establish a full framework of applicability,” says Hachad. “Because this relates to medications, there is a large community that the regulators are a part of that is establishing the framework. When I say ‘defining the framework,’ it’s applying and using the information that is available to us in a transparent manner.”
AI platforms and technologies being used to aid in personalized medicine regimens are designed to be HIPAA- and Payment Card Industry Data Security Standard (PCI DSS)-compliant. According to Brown, the goal of these AI efforts is to help HCPs better listen to their patients and derive insights from those conversations. The goal is not to build another system for housing protected personal health information.
“We have capabilities within our platform to redact personally identifiable information or obfuscate or modulate voice so that it’s not identifiable,” says Brown. “Healthcare organizations are still getting the value of the insights out of it.”
The platform is also designed to pick up on the reporting of any treatment side effects, which can be a difficult task under normal circumstances. Patients, for example, may not realize that mentioning a headache is considered a side effect, and may only bring it up in passing; doctors, however, are still required to report it.
“If you listen to any medication or therapy commercial on TV, you’ll hear the voice talking about side effects,” says Brown. “If patients happen to mention side effects when they are talking to a pharmaceutical company, that company is required to identify it and bring that information to the FDA. For example, a patient may call up a nurse line and mention in passing that the therapy is going well, but they had a small headache. That nurse representative is required to report that.”
AI algorithms are designed to catch the way that patients talk about side effects and pull out the conversations where these are mentioned, even if only a small percentage of patients are reporting them. In cases when individuals are placed on a personalized therapy, which may vary from patient to patient, the platform can still locate the mentions of side effects and ensure that they are reported appropriately.
Personalized medicines offer new and innovative opportunities to treat certain patient populations with more effective treatments. By studying biomarkers and other genetic issues, drug developers and biopharma manufacturers can approach treatment from new directions. Novel therapies for rare or difficult-to-treat diseases may focus on the root cause, while others may target specific mutations.
Hachad notes that as the technology used to study biomarkers and genetics becomes more advanced, researchers will be able to ascertain more information about a patient using even fewer biomarker signals than they rely on now.
“One single point can allow you to pinpoint different subtypes of diseases or processes that are altered,” she says. “Often, some of these points can be merged with other information or factors about patients that aren’t necessarily genetics. You can identify responders and non-responders much sooner during the care journey for that patient. Basically, you’re getting a better picture about that individual much sooner than what has been traditionally done.”
The hope is as precision medicine and therapy approaches continue to advance and evolve, stories about patients waiting months or even years to find out why initial treatments aren’t working will become fewer and fewer.
Mike Hollan is Pharm Exec’s Editor. He can be reached at email@example.com.