The National Institutes of Health’s Recombinant DNA Advisory Committee earlier this month expressed a number of concerns about how Nucleonics has designed its proposed phase I trial for its expressed RNAi hepatitis B therapy, NucB 1000.
Chief among the issues raised by the RAC reviewers are the doses Nucleonics proposes to test in the clinical study, which the committee believes could potentially overwhelm patients’ natural RNAi processes, and the company’s decision to evaluate NucB 1000 only in mouse models in its preclinical research.
“The doses being put forward in [the trial’s protocol] represent quite a large dose in relationship to previously conducted gene-therapy trials,” Natasha Caplen, one of the RAC reviewers and head of the National Cancer Institute’s Gene Silencing Section (see RNAi News, 1/29/2004), said during the meeting. “One of our large concerns is that, when exploiting the RNAi mechanism, we don’t interfere with its natural role in regulating endogenous gene expression."
In the trial’s protocol, “we are now trying to extrapolate from mice — because all of the studies to date are in mice — the human situation,” she added. “There may be a requirement for additional models.”
The committee has no authority beyond making comments and recommendations.
During a relatively short question-and-answer session between the reviewers and members of Nucleonics’ management, the company defended its proposed clinical trial, noting that its dosing schedule begins very low and that its protocols include numerous safety reviews.
Additionally, Nucleonics argued that there are few good animal models for hepatitis B since chimpanzees are no longer available due to animal protection rules.
This week, Nucleonics President and CEO Bob Towarnicki told RNAi News that his company would review and respond to the RAC’s recommendations once they have been formally submitted, adding that Nucleonics’ plans to file an investigational new drug application for NucB 1000 before the end of the year remain unchanged.
On the RAC
The NIH established the RAC in 1974 to address concerns over the safety of using recombinant DNA techniques to manipulate genetic material. Among the RAC’s duties is to provide the NIH with “advice concerning various advances in recombinant DNA technology, as well as ethical and safety considerations associated with novel or possibly risky forms of recombinant DNA research,” according to the NIH.
Because NucB 1000 uses a plasmid DNA to express shRNA, Nucleonics was required to submit its phase I protocols to the RAC for review.
This month’s meeting came after Nucleonics submitted preclinical data and its phase I protocols to the RAC, after the committee issued its questions and comments back to the company, and after the company returned its formal reply.
According to Nucleonics, its proposed phase I trial will include about 18 treatment-naïve chronic hepatitis B patients who will receive a single intravenous dose of NucB 1000. Specifically, the trial will evaluate five dose levels: 5 mg, 15 mg, 50 mg, 150 mg, and 350 mg.
Once the first 50 mg cohort is treated, the company will conduct a safety review. If no problems are observed, patients in a new cohort will receive a single 50 mg dose of the drug, followed by another three weeks later. After the 350 mg cohort is treated, patients in a new cohort will receive a single administration of this highest dose, followed by another three weeks later.
Towarnicki told RNAi News this week that in each cohort, “only one patient will be dosed at a time with a minimum of one week between each subject. After the third patient in a cohort is dosed, a complete safety analysis [lasting] 6 weeks will be performed before initiating the next dose level.
“We have proposed a very conservative and carefully monitored dose-escalation study,” he added.
Based on their comments during this month’s meeting, however, it appears that the members of the RAC’s review team did not all agree.
“I really have fundamental reservations about the preclinical data that support this protocol,” Stephen Dewhurst, a researcher at the University of Rochester Medical Center and RAC member, said during the meeting. “I have fundamental reservations about the nature of the animal models, [and] I have fundamental reservations about the expressed reasons to have not used more biologically relevant models.”
Noting that “350 milligram dosing by the IV route in a human being would be by far the most plasmid DNA ever delivered,” Dewhurst said that Nucleonics’ proposed study “is a trial that really attempts to push the envelope, and in that context I think the standard needs to be quite high to proceed.
“The obvious concern that overarches everything in my mind is [that] while this is a safety trial, there is the fundamental, underlying issue of risk-benefit,” he added. “My concern is that the promise of benefit in my mind has not been established convincingly by the preclinical data.”
According to Dewhurst, “if we’re going [up] to a dose of 350 milligrams IV in humans, I am very uneasy about moving directly from mouse to human.”
Jake Liang, a National Institute of Diabetes and Digestive and Kidney Diseases researcher and RAC reviewer, commented that he views “the [mouse] model used [by Nucleonics in its preclinical package as] a glorified in vitro model,” adding that he was surprised the company did not conduct additional work in another animal model such as a transgenic mouse.
Dewhurst added that “there really does need to be some other animal model. I understand … that one would like a virally infected model, but if that’s not doable, I think an uninfected model would at least give us a feel for what the outcome might be in a worst-case scenario.”
Citing a recent Nature paper in which Mark Kay and colleagues from Stanford University showed that sustained, high-level shRNA expression in murine livers can result in severe toxicity and death, possibly as a result of interference with endogenous microRNA processing machinery, the RAC’s Caplen cautioned during the meeting that at the dosing levels proposed by Nucleonics, an RNAi-based drug such as NucB 1000 has the potential to “affect the entire RNAi mechanism … in a saturating sense.”
Although Nucleonics provided to the RAC preclinical data demonstrating that NucB 1000 had no affect on the function of miR-122, an miRNA expressed at high levels in the liver, “my concern … is that the assay was done by looking at a perfect target for miR-122,” she said. “It’s not looking at what is probably closer to the endogenous role of the human miR-122, which is unlikely to relate to its perfect alignment with a sequence; it’s more likely to be related to partial sequence relationships, which may or may not induce cleavage or … protein translation.
“So [Nucleonics’ example is] still a very weak surrogate I feel for looking at the activity,” Caplen said.
According to Caplen, Nucleonics had “suggested that it would not be appropriate to try to use some sort of microarray expression profiling or some other consideration of the effect of introducing these shRNAs into human cells.”
She conceded that there are limitations to such analyses, but said that “commercial array platforms that are now available are relatively consistent, and I think it will be very important to do this kind of experimentation in both the presence and the absence of the target.
“I stress that I think there needs to be some additional experimentation to address the potential effects on the microRNA pathway,” Caplen added. “While it’s clear that RNAi has a huge potential to be exploited as a therapeutic, we need to make sure that we’re trying wherever we have knowledge as to how it is controlling endogenous gene expression that we use all the types of assays that are available to us to ensure that we’re minimizing that effect.”
Following a comment period by the RAC’s reviewers, members of Nucleonics management attempted to address the key issues that were raised during the meeting.
Tackling the reviewers’ concerns about the animal models used, Nucleonics Co-Founder and Vice President of Research and Development Chandrasekhar Satishchandran said that the company did “try to use a transgenic [mouse] model … [they proved to be] inconsistent in their ability to produce viral particles.”
Cathy Pachuk, a company co-founder and vice president of preclinical research, added that “there are only two animal models for HBV … mouse models and variations of mouse models … and … a chimp model which is no longer available.”
“I have fundamental reservations about the nature of the animal models, [and] I have fundamental reservations about the expressed reasons to have not used more biologically relevant models.”
She said that “at least one, if not more than one, HBV small-molecule therapy went into the clinic without going into an animal model except afterwards,” and that “there are other indications such as [hepatitis C for which] there are no animal models, period. Basically, you do your safety studies and do your cell culture in vitro studies.”
As for conducting miRNA microarray analyses, Pachuk said that while these experiments could be done, “the question is, ‘What does the data mean?’
“We don’t know how to interpret the data at this time, [or know] what cells do we do the analysis on,” she said. “We could do the experiment and file the data away for a later time when more data is collected so that it can be interpreted in a meaningful way, but right now … the data can’t be interpreted. All I can say is it’s something we could do.”
Addressing the RAC reviewers’ safety concerns, Pachuk added that Nucleonics does have “safety data on [a] single-dose product in mice, on multi-dose, [and] on numerous cell lines including human-derived hepatocytes that do not express the target. … We have hydrodynamic safety data, as well, which would address the safety of the drug component as well as the shRNAs that would be produced at very high levels in the liver and don’t see toxic responses.”
In an interview with RNAi News this week, Towarnicki said he was “frustrated” with the way the RAC meeting was organized, particularly the amount of time allotted to Nucleonics to respond to questions and concerns.
“Typically … they have someone raise a question or a group of questions then give the company an opportunity to respond,” he said. “In this case, they kind of broke that protocol and let … their reviewers critique and gave us a limited time to respond to selected questions. Unfortunately, what that does in the public record is allow a lot of misunderstanding to stand.”
Additionally, Towarnicki said that there continued to be “misunderstanding on the part of several of the reviewers that we will correct in our written response” to the issues raised during the meeting.
”When we receive the letter [from the RAC outlining their final] recommendations, we will respond appropriately back to them … and let them know what we adopted,” he said. “If we haven’t adopted a recommendation, [we will inform them] why.”
He added that after several pre-IND meetings with the US Food and Drug Administration, Nucleonics is confident in its preclinical data and phase I study design, and “we still intend to file the IND by the end of the month.”