Harsha Rajasimha, PhD, discusses current challenges leading to clinical trial delays, such as the “valley of death.”
Pharmacy Times® interviewed Harsha Rajasimha, PhD, founder and CEO of Jeeva Informatics, on the potential to accelerate patient recruitment and solve clinical trial delays. Currently, the research, development, and approval process of a new drug by the FDA can take between 12 to 15 years. However, studies have shown patient recruitment issues are the cause of 85% of clinical trial delays.
Pharmacy Times: What are the different phases of the drug approval process by the FDA?
Harsha Rajasimha: Typically, drug approval process involves a preclinical phase, which enables first-in-human clinical trial for a new drug or entity. And that goes through phase 1, phase 2, phase 3 trials, typically. And sometimes there can be phase 1 and 2, or phase 2 and 3 combined, and stuff like that, and different types of trial designs these days, but typically phase 1, phase 2, phase 3, where phase 1 is entirely looking at safety of the drug in humans, because it's only tested in animals. Phase 2, looking at safety and efficacy at very low doses to make sure there is some drug effect. And phase 3 is full-fledged in a larger population, to assess the safety and efficacy profile across the range of doses in a range of patients in the target population. And after that, all of this data gets wrapped into an FDA submission for review. And the whole process from phase 1 to phase 3 and the review process can take about 7 years, typically, sometimes longer. And so, it's a very expensive and time-consuming process. And this process is often called the “Valley of Death,” because only 1 in 9, 1 out of 10 drugs actually make it through the regulatory approval. Most of them will fail at phase 1, or phase 2, or phase 3, about 30% at each phase. So, very few drugs make it to market.
Pharmacy Times: Could you discuss issues pertaining to blinding during clinical trials?
Rajasimha: There is single blinding and double blinding where in single blinding, only the patients are not informed or aware whether they're actually getting a drug or a placebo. And in double blinding, both the investigator and the patient are unaware which patient gets the placebo and which patient gets the actual drug. Now, there are open-label clinical trials where everybody knows who is getting what, and that's also getting increasingly common. And so the gold standard in assessing the safety and efficacy of new drugs, first-time drugs, and not in a repurposing trial where the safety and efficacy is already established in one indication, say arthritis, and now it's being repurposed for a rare indication or a different disease is something that happens direct to phase 3 trials, because the safety is already established and even efficacy, to a certain extent, in other indications. So, blinding may or may not be employed in a specific clinical trial, but for a brand-new drug, randomized controlled clinical trials have—they’re called RCTs—that’s the gold standard, even today. And that involves double blind, most of the time, where there is a group of patients who may get the drug and a group of patients who may not get the drug. They may not be the same group size, as often they want to minimize the number of patients who do not get the drug and they get the placebo, just because it's both not fair on the patients that they are denied a treatment and two, it's big, we want to minimize that number to the extent possible.
Pharmacy Times: Could you discuss some challenges around using placebo versus standard of care, and why might placebo be a more frequent option?
Rajasimha: I think, you know, I wouldn't claim to be an expert specifically on that question. But in general, if it's a vaccine trial, for example, a placebo is used, that should not be a problem. But if it is a severe life-threatening disease, then it's usually the standard of care. Say, some stage 3 or stage 4 cancer, it's not ethical to deny the patient of any treatment, and so the patient should at least get a standard of care. And so everybody gets a minimum standard of care, and some may get standard of care plus the drug, or no standard of care but only the new drug in the clinical trial, and stuff like that. So, essentially, it's balanced by ethical guidelines. You know, the Belmont Report and the Helsinki declaration, and the international Harmonization on the Good Clinical Practice, which defines the guidelines on whether a placebo is ethical to use in a clinical trial or if the standard of care should be used, with the interest of patient outcomes being at the heart of these choices.
Pharmacy Times: Could you discuss the “valley of death” further, and what some of the specific challenges are in bringing new drugs to market?
Rajasimha: Absolutely. You know, this whole process where only one or 12% of the candidate drugs making it through this regulatory review process and getting approval, that's what is called the “valley of death,” and the majority of them do not make it there for a given medical indication or the target population. And so not to be that that drug may not work for any other indication—if it's proven to be safe in humans, maybe it might be effective in certain indications, but not other indications. So, a drug-repurposing trial in the future might actually make these drugs approved for other indications. And so, this whole “valley of death” is complex and complicated by various challenges. And we spent the last 4 years interviewing various stakeholders of clinical trials asking, why does it take 10 to 12 years from when a candidate drug is discovered in the laboratory, to when it is approved by the FDA,? It can be a total of 10 to 12 years with about 7 years in the clinical development phase, and then it can cost over 2 and a half billion dollars to bring 1 successful product to market. That's because they have to underwrite the cost of the 9, every 9 failures for every successful drug approval. And so the burden of getting that 1 successful drug means you have to go through 9 potential failures before that happens.
Pharmacy Times: What are some persistent challenges faced in clinical trials during drug testing?
Rajasimha: The number one challenge we heard firsthand from various biopharmaceutical sponsors and medical device sponsors is patient recruitment. You know, 30% of clinical trials terminate because they're unable to enroll the target number of patients in that clinical trial. The 70% of the trials that do get past the recruitment barrier, majority of them are still delayed. They don’t enroll patients in a timely fashion. So, every day that a clinical trial gets delayed, will be costing the biopharma sponsor anywhere from $600,000 to $8 million in lost revenue of the drug or device, because these drugs and devices have very finite patent life, during which that return on investment has to be recovered of all the R&D costs, which can cost hundreds of millions of dollars before a drug or device can get approved. And hence, the barrier of patient recruitment is huge, and that has to be overcome. Those that do get past the patient recruitment barrier, about 30% of the patients on an average across all disease areas and all phases of clinical trial drop out during the course of a clinical trial, post enrollment. And that can be a significant burden as well.
Now, beyond patient recruitment, retention, we heard protocol complexity as a major barrier or challenge. Increasingly, the number of biomarkers, the number of endpoints, the data points that are captured in a clinical study is significantly growing. And the use of wearable devices and sensors offers additional challenges to ensure continuous streaming data is captured on an ongoing basis, not just episodic data capture, when patients visit the clinical site on a periodic basis. And hence, if—usually, a clinical trial goes through 2 to 3 protocol amendments during the course of a trial, which means pretty much every clinical trial has to implement multiple protocol amendments in the middle of a trial. They decide to change the protocol because they learned something from the early patients, either enrolling or not enrolling, how they can modify the protocol for better. And FDA encourages that and sometimes may even require changes in protocol, depending on how the trial is progressing. And so continuous real-time monitoring of the study as it progresses has become necessary. And hence, that’s another barrier.
The fourth one is around the use of multiple digital platforms and electronic databases and tools. We heard literally, the clinical researchers have to use 30 different tools to make 1 clinical trial successful. And when the clinical trial completes, for the next trial, they have to rebuild the whole infrastructure all over again. It's like every trial piece like the first ever trial undertaken by mankind, because each protocol is so unique. And so, the same infrastructure that was used for another trial cannot be used for this trial. And that is costing significant time, effort, burden on researchers at the investigator site-level, but also sponsors, CROs and the whole complexity of human collaboration. Ultimately, we focused around the patients, and the patients have additional burden of having to travel to the sites and participate. The travel burden can be a huge depending on the location of the site, with respect to the patient, as well as the frequency of visits to the site, as well. So, taking all these into account and the data quality—and mind you this is a highly regulated space, so the insuring compliance, those are all table stakes, but the regulations are often changing. Every year, the new set of guidelines and updates to the existing guidances that FDA issues and other regulatory agencies might issue, which all adds to the burden of how complex it becomes and expensive and time consuming to execute 1 clinical trial.
Pharmacy Times: What are some potential solutions to solving clinical trial delays in relation to patient recruitment?
Rajasimha: This is something the whole industry has been battling and my perspective at Jeeva is that, after listening to various stakeholders on what solutions they have been trying, patient recruitment involves largely 2 major phases. One is identification of potential patients, then screening them to get eligible patients enrolled into the study. So, identification of patients has largely been left to investigator sites, and some of them may be academic medical centers with electronic medical records and databases of patients with a sizeable volume of patients in the database that they could search and do some data mining, AIML, to match patients to specific inclusion exclusion criteria of a given clinical trial. And in some instances, now, there are a growing number of AIML-based platforms where the identification of patients can be more based on real world data, at least the preliminary screening or preliminary identification of patients.
And then this next process of rapidly screening these patients to meet the inclusion exclusion criteria can be quite challenging, because not all data is available in digital format. Especially if you go outside of the United States, the electronic medical records are not either available or not standardized. Even in the US, there is like anywhere from 50 to 200 electronic medical record systems now. And although majority of large hospital systems might be on Epic, and Cerner, and platforms like that, there is a large number of tertiary care centers, which may have different types of electronic medical record systems and so on. So, even beyond medical records, lab reports, insurance claims data, and other prescription drug information and other types of data may all be necessary to accurately match a particular patient to a particular clinical trial. With protocols being increasingly complex, we may need even genetic sequencing data and follicular signatures, additional complex laboratory data to really confirm a patient is actually eligible or not. And so there is a large number of screening failures during this process and that adds to the patient recruitment burden.
And so, what we have seen work is a platform approach, which can connect the dots and which can streamline and even automate a lot of the manual processes. And ultimately, when a patient has to choose whether to enroll or not in a clinical trial, they also consider the burden of the protocol throughout the clinical trial process, not just before enrollment and screening, but also post-enrollment. How many times do I have to visit a site? Is it close to where I live? Or do I have to travel to another out of state, on a plane? How frequently? And is travel reimbursed or not? And parking, availability and so on and so forth. So, depending on how painful the patient already is with the existing medical conditions, the additional burden of this travel and logistics can be significant. And on top of that, if they need to bring a family member along with them because they need their assistance as a caregiver, that bit makes it even more harder to make sure the caregiver is willing to take time off from their work and travel with the patient to the site.
So, that's why patient recruitment remains a huge and the biggest barrier and problem in clinical trials, operation wise and logistics wise. And not just recruitment, but also retention, which means it takes more than just platform technology, which is human-centric, and which can address all these barriers and bring the logistics of 30 different logins into a single login so both patients, investigator site, sponsor CROs, all stakeholders can collaborate on a single login. That can make a big difference to clinical trial, operations wise, and minimize the burden on all stakeholders involved. But that's not enough, you know. It takes staff training. People, process and technology, right. So, humans and technology, a platform with minimal logins and maximum automation, minimal manual repetitive tasks—all those are critical, but also needs to fit into the workflows of various actors that have to collaborate in a clinical trial.