Paul S. Larson, MD, paul.larson@ucsf.edu, is professor and vice-chair of Neurological Surgery at the University of California San Francisco, chief of Neurosurgery at the San Francisco Veterans Affairs Medical Center, and surgical director of the Parkinson’s Disease Research, Education and Clinical Center.

Paul Larson

While pharmaceutical companies spend billions of dollars developing groundbreaking gene therapies and drugs to address complex diseases of the central nervous system (CNS) (1), the method of delivery is often relegated to an afterthought. With interest and investment in this area continuing to grow at a rapid rate, it is increasingly important to maximize the therapeutic efficacy of both the drug and the delivery method.

 Neurodegenerative disorders and brain cancers are two of the most devastating human diseases, and the world’s aging population continues to increase the frequency of these disorders, making rapid and successful drug development more important than ever. Yet, CNS drug development faces extraordinary hurdles and typically takes much longer than non-CNS therapies. Gene therapy clinical trials for CNS diseases are particularly challenging and protracted due to complexities of the brain, potential side effects, the need for direct and often targeted infusion, and delivery complications. Ultimately, the impact of these life changing CNS drugs is being diminished by the “physics of infusion”. Improved delivery methods must be considered and adopted early in the drug development process, or we risk negating their potential therapeutic benefits (2).

The following discusses the current limitations of traditional drug delivery into the brain, reviews recent advancements and benefits of real time interventional MRI (iMRI)-guided gene therapies, describes some ongoing clinical trials, and provides advice for pharmaceutical companies on how to proceed with intracranial therapies and delivery.

It is well known that local delivery of therapeutic agents into the central nervous system offers advantages, including circumventing the blood-brain barrier, the potential for minimizing systemic toxicity, and the ability to achieve higher concentrations of agents at the target site (3). Intracranial delivery has been performed using traditional neurosurgical techniques and tools, including stereotactic frames and standard cannulas or catheters. These methods lack the ability to confirm accurate catheter placement. Another major challenge has been the inability to visualize the drug delivery in real-time—thus making it impossible to evaluate the accuracy and coverage of drug delivery within target stuctures (4). This traditional delivery method has become known as “blind” infusion; surgeons would infuse the prescribed amount of the drug to the target area but could not ascertain whether the drug covered the target as intended. In addition, traditional blind delivery with a simple hand injection using a syringe does not allow for infusing at a consistent flow rate, making measurement of delivery rates inaccurate and variable across procedures (4).

Gene therapy trials using traditional intracranial delivery methods began almost 15 years ago, but early results showed limited initial success (4). For example, the PRECISE trial, a large-scale Phase III trial delivering interleukin 13 (IL13)-pseudomonas exotoxin for recurrent glioblastoma, failed to show clinical benefit of survival. Post hoc analysis of the trial patients revealed that less than half of the implanted catheters were in the optimal position—in fact, only 68% of catheters were positioned in accordance with the specified protocol. Additionally, the investigators described “a fundamental lack of infusion distribution physics” as a major reason for efficacy failures. In the years since, similar outcomes have been noted in other trials (3).

Traditional surgical image guidance in the operating room is not real time, but rather dependent on co-registration to a previously-acquired MRI. In addition to the errors that co-registration can introduce with regards to device placement, conventional neuro-navigation platforms offer no means of visualizing infusions in real-time. Without live image guidance, a surgeon has no way to ensure accurate catheter placement and monitor the pattern of distribution of the infusions within the intended target (3).

Initially, surgeons believed that an infusion cannula could be placed into the center of a brain target and the injected fluid would behave in a predictable fashion, similar to an injection of fluid into a beaker of homogenous gelatinous material. The assumption was that the infusion would grow as a uniform spherical ball, increasing in size proportional to an increase in infusion volume. Surgeons now know this is not the case. For example, infusions in the putamen can result in spread of infusate outside the putamen along perivascular spaces, reflux along the inserted cannula, and, to a lesser extent, leakage into the vacated catheter tract upon its removal. Utilizing real-time imaging has enabled surgeons to observe these delivery complications and test new techniques and devices to optimize results.

The advent of real time iMRI-guided infusion has enabled larger volumes of delivered therapeutics and tactics, such as varying infusion rates and adjusting cannula depth/position to maximize coverage of the target area—and to minimize the degree of vector spread away from the intended target. This has significant implications for the success of clinical trials, as evidence suggests that adequate coverage of the target area may have a dramatic impact on patient outcomes (3).

For pharmaceutical and biologics companies, the inability to predict, visualize, and confirm infusate coverage of target regions has proven to be a major hurdle in clinical development—and a cause for therapeutic inefficacy. Without intraoperative monitoring, it has not been possible to determine how differences in surgical approaches, cannula designs, vector volumes, and dosing affect coverage within the brain (4).

One platform that allows for iMRI-guided catheter placement and real-time visualization of delivery of therapeutic agent is ClearPoint NeuroNavigation, developed by US-based medical device company, ClearPoint Neuro, which is currently enabling a range of clinical trials involving neuro-oncology, neurodegenerative disorders, rare genetic disorders, and movement disorders. It is an integrated hardware and software platform that communicates with the MRI scanner console to provide high-resolution imaging for trajectory planning, entry point selection, device placement, and infusion monitoring. The navigation system has been used in conjunction with the ClearPoint Neuro SmartFlow Cannula for precise dosage of drugs or gene therapies in approved clinical trials. 

The introduction of iMRI-guided drug delivery uncovered many injection challenges associated with targeted infusion within the brain. Based upon the diffusion behaviors and characteristics observed during early iMRI monitoring in non-human primates, many protocols have now adopted a technique of either progressive advancement from proximal to distal for the duration of infusion, or a “pull-back” method for cell delivery where the first deposit is delivered to the most distal region. This technique of administering multiple infusions along the trajectory path was found to increase the surgeon’s ability to achieve infusions that cover the putamen or other target area as uniformly as possible, with the added benefit of providing the ability to move the cannula tip away from problematic perivascular spaces. In addition, innovations in cannula design have resulted in more reliable, efficient, and increased distribution of agents. Because image-guided systems provide real-time visualization of the infusion, they allow for intraoperative troubleshooting and real time modification to key infusion parameters, such as flow rate, positioning of the catheter, or terminating the infusion when the desired coverage has been attained (3).

This latter advancement is particularly noteworthy as it introduces the potential for 

, a game-changing capability in treating CNS disorders or brain cancer. Historically, physicians have strictly followed a standardized trial protocol that specifies uniform infusion parameters across all patients, including the dose (volume) of therapeutic delivered. Evidence suggests that many early trials administered dosages that were insufficient to achieve desired clinical outcomes. iMRI-guided surgery offers the possibility of using much higher volumes, and adjusting the volume to focus on patient specific coverage, rather than a pre-determined amount. This concept of patient specific dosing is new and exciting in its potential to improve clinical outcomes—and it is currently being applied right now in an ongoing Parkinson’s Disease (PD) trial.

Finally, these technological advancements afford an improved level of safety. iMRI provides surgeons with a significantly higher degree of confidence that the cannula is following a safe trajectory and being placed in the desired target location. This is particularly important when targeting deep structures, where safe corridors through the brain can be only 2–3 millimeters wide.

Ongoing clinical trials of iMRI-guided drug delivery

At the University of California San Francisco, Paul Larson, MD, a professor and the vice-chair of Neurological Surgery at the university, is involved in three active human clinical programs utilizing iMRI-guided gene therapy delivery. The first is an amino acid decarboxylase program (AADC), which is an enzyme replacement gene therapy strategy for PD that uses adeno associated virus (AAV) to deliver a gene for AADC. The program, sponsored by Neurocrine Biosciences and Voyager Therapeutics, is now a Phase II multi-center study. The Phase I results indicate that, as coverage of the putamen improved, there was a significant increase in PET activity and a concurrent decrease in PD medications (5).

The second is a Phase Ib investigation sponsored by Brain Neurotherapy Bio to evaluate the safety and potential clinical effect of AAV2-GDNF delivered to the putamen in subjects with earlier stage PD. Growth factors are an exciting approach as they have the potential to be a disease-modifying therapy.

The third is the world’s first intracranial gene therapy study for Huntington’s Disease (HD). Sponsored by uniQure Biopharma, it is a Phase I/II study of AMT-130 delivered to the striatum of patients with early manifest HD and is designed to establish safety and proof-of-concept. The gene product is designed to interrupt production of the aberrant Huntington protein.

Beyond these studies, iMRI-guided delivery is being used in many other ongoing or planned Phase I and Phase II clinical trials for indications, including glioblastoma, PD, Huntington’s Disease, Amyotrophic Lateral Sclerosis (ALS), and pediatric indications, including AADC deficiency, Freidreich’s Ataxia, Angelman Syndrome, and Sanfillippo Syndrome. iMRI-guided delivery with the SmartFlow cannula is already cleared for Cytarabine injection into the ventricles to treat leukemia and non-Hodgkin's lymphoma, and numerous other programs are in development involving gene therapy, cell transplants, and other therapeutics.

Although it may be premature to assign hard savings to iMRI-guided drug delivery, there are significant potential benefits from such an approach. Given the tremendous R&D costs to develop a gene therapy as well as the cost of a single dose and the infusion procedure, any increase in the likelihood of a successful procedure would translate into considerable savings for the healthcare industry at large—as well as accelerating adoption of pharmaceutical gene therapies. The potential for iMRI guidance to ensure safe and accurate placement of the catheter, administer patient-specific dosages, minimize leakage, and optimize coverage addresses many of the challenges faced by early gene therapy trials. When considering the costs of past negative phase two trials, the business case for leveraging iMRI is apparent.

The advantages of real-time MRI-guided gene therapy delivery are clear. However, one challenge for pharma companies is the number of neurosurgery centers that have implemented iMRI at their facilities. The ClearPoint NeuroNavigation system is currently installed in more than 60 of the leading neurosurgical centers across the United States, and is expanding into Europe. While the number of institutions adopting iMRI is growing, there is a need for more pharmaceutical companies and neurosurgeons to embrace this improved delivery method for intracranial therapeutics. To do otherwise is to potentially jeopardize billions of dollars in R&D investments, put study participants in harm’s way with little chance to show benefit, and run the risk of obviating cures for life-threatening diseases.

1. J.P. Fuhr, “The Possibility of Encouraging Innovation and Competition by Erwin A. Blackstone.” 

 8 (6) 1481–1493 (2018).
3. S.J. Han, et al., 

 16 (6) 635–639 (2016).
4. R.M. Richardson, et al., 

91, 1210–1218 (2020).
5. C.W. Christine, et al., 

Paul S. Larson*, MD, paul.larson@ucsf.edu, is professor and vice-chair of Neurological Surgery at the University of California San Francisco, chief of Neurosurgery at the San Francisco Veterans Affairs Medical Center, and surgical director of the Parkinson’s Disease Research, Education and Clinical Center.

*To whom all correspondence should be addressed.

Articles

Leveraging real-time MRI guidance for intracranial gene therapy administration potentially improves efficacy and outcomes.