By Chris Mason, M.D., Ph.D., Chief Scientific Officer, AVROBIO
Lentiviral gene therapy, involving the genetic modification of stem cells outside of the body, requires a conditioning regimen in order to create the necessary space for engraftment of gene-modified cells in the bone marrow. Busulfan is an established conditioning agent, building on decades of use in bone marrow transplants as well as the treatment of hundreds of patients in lentiviral gene therapy clinical studies. Clinical evidence continues to emerge that points to busulfan’s unique capabilities, for example, its ability to support potential microglial network reconstitution with gene-modified microglia and its potential to address central nervous system (CNS) manifestations of disease. Since busulfan’s discovery in 1951, the transplant community has continually refined its use, striving to maximize its potential.
As we continue to seek out ways at AVROBIO to enhance the patient experience and outcomes with lentiviral gene therapy, I’m pleased to share another exciting development in busulfan’s continued journey of optimization.
Thanks to a recently announced collaboration between Saladax Biomedical, a leading diagnostics provider, and AVROBIO, a leader in lentiviral gene therapy, hospitals and clinics may soon have access to a fully automated, nanoparticle immunoassay kit to simplify and streamline personalized conditioning, enabling and simplifying state-of-the-art precision dosing with this important drug.
Let me explain why this new assay kit has the potential to be so transformative, and how it came to be. It all comes back to the one and only time I’ve been pleased to have a research idea scooped up by someone else!
Not Too Much, Not Too Little: Why Total Exposure Has to be Just Right
Busulfan’s pharmacokinetics is highly challenging in that it is metabolized differently from person to person, and even from one day to the next within the same person. Furthermore, it has a narrow therapeutic window as a conditioning agent for patients undergoing lentiviral gene therapy. Put simplistically, it conforms to the Goldilocks Principle: too little exposure and the chances of optimal engraftment drop; too much exposure risks out-of-range toxicities, including the potential for serious liver damage, called veno-occlusive disease (VOD).
We know from the retrospective analysis of hundreds of patients (Bartelink et al., The Lancet Haematology, (2016)) that the optimal cumulative exposure over four days to busulfan as a conditioning agent – that is, when it’s clearing space in the bone marrow for transplanted cells to engraft – is a cumulative area under the curve (AUC) of 78 to 101 mg.hr/L.
AVROBIO is pioneering a personalized medicine approach which aims to precisely target the exposure to hit the midpoint of that “Goldilocks” range. Our target is a cumulative AUC of 90 mg.hr/L, a regimen we refer to in shorthand as Bu90. When successfully implemented, such an approach can maximize the potential for long-term engraftment while avoiding toxicities associated with out-of-range exposure.
The challenge is how to hit that target exposure precisely for each individual patient.
The process of intravenously infusing a busulfan dose takes three hours. To evaluate how the patient is metabolizing it, blood is taken at intervals (e.g., one, three, five and seven hours) after dosing is complete. These samples must then be sent to one of just a small number of labs that are equipped to run this particular analysis, which may be geographically distant from the gene therapy dosing site. That can mean starting the patient’s busulfan drip in the middle of the night so all blood samples can be collected and shipped by early morning, ensuring arrival at the lab during the hours a technician is on duty.
Once the samples get to the specialized lab, analyzing the blood via gas chromatography-mass spectrometry (GC-MS) takes two to three hours – and you can process just one sample at a time. All the while, the clock is ticking. Only after all the samples have been evaluated can the patient’s rate of metabolizing busulfan (for that day), and hence tissue exposure, be calculated. Armed with this knowledge, the next day’s dose can be determined to build toward achieving the desired final cumulative target exposure. Only then can the pharmacist make up the appropriate next dose. All this needs to happen before the next scheduled infusion. There are only 24 hours between the start of infusions, making such monitoring extremely challenging.
Therapeutic Drug Monitoring (TDM) typically targets a range, hence at the outer edges of the range there is the potential to increase the risk of out-of-range toxicity (upper end) or to reduce engraftment (lower end). Furthermore, because of the time constraints with this process, in some cases precision dosing for busulfan conditioning may be less than precise, or often not attainable. As noted above, patients may break down busulfan at different speeds from one day to the next, and if you’re not continuously monitoring, you may have trouble hitting that “Goldilocks” cumulative exposure. This is why further refining TDM, through a combination of readily available point-of-care busulfan plasma concentration measurements and targeting the mid-point of the range, promises to achieve a more consistent targeted exposure.
It’s important to note that even with this personalized medicine approach, in our clinical trials to date we have observed anticipated side effects from busulfan, including nausea, mucositis, fever, rash and hair thinning/loss, which typically came on quickly with a peak of three to five days and resolved quickly. Busulfan, indicated to be dosed in combination with cyclophosphamide to treat Chronic Myeloid Leukemia (CML), may cause temporary or permanent infertility when injected prior to allogeneic (or donor) HSC transplantation. In our lentiviral gene therapy trials, we use personalized busulfan as a single agent over a single cycle with the patient’s own HSCs cells, and the potential risk of infertility due to busulfan in this setting is still being studied.
At AVROBIO, we’ve spent quite a lot of time organizing logistics and supply chains that we anticipate will enable robust four-day personalized dosing to precisely target a busulfan cumulative exposure of 90 mg.hr/L in patients receiving our investigational gene therapies. We’re proud of what we’ve accomplished, but long term we believe there is significant opportunity in bringing innovation to this delivery infrastructure for a better patient experience, clinical process and reduced burden on healthcare sites.
A Quest to Bring Busulfan Blood Concentration Monitoring On-Site
Last year, I set a goal for myself — identify a way to enable any hospital or clinic to carry out quick and simple busulfan concentration measurement. This became a major preoccupation for me. The drug is such an effective and important conditioning agent, especially since it uniquely enables gene-corrected cells to be distributed throughout the entire body, including the CNS. I knew that if such an assay could be created, it could potentially have a substantial impact on patient safety, as well as therapeutic outcomes for bone marrow transplants and lentiviral gene therapy.
The obvious solution would be to create a test that would work on machines found in every hospital or clinic -- an immunochemistry assay being the most obvious as it is quick and easy to perform anywhere. But unfortunately, busulfan is not a good candidate for such an assay because it is too reactive to be used for the routine calibration and control samples that are essential prerequisites for this type of test.
To get around this issue, I decided to see if we could chemically engineer a more stable version of the busulfan molecule in the lab, so I could use such a busulfan derivative in a much simpler drug concentration assay than conventional GC-MS. Before embarking on this effort, and hoping that someone might have done this already, I trawled the patent databases just in case.
And that was the game-changing moment.
Onto my screen popped a pleasant surprise: a hit for Saladax Biomedical. I was thrilled someone else had already done the hard work of invention. I reached out immediately to their President and CEO, Salvatore Salamone, Ph.D., a highly respected and hugely enthusiastic leader in the field of diagnostics.
Dr. Salamone explained that more than a decade earlier, his team had developed a high-speed busulfan assay for patients with CML. Yet before they could get it on the market, a new therapy for CML was approved and the busulfan assay was deemed irrelevant. The Saladax team put it on the shelf. When I asked if they’d consider reviving it for use with lentiviral gene therapy, they were delighted.
Saladax had come up with a very elegant solution to the Bu90 precision-dosing challenge. They made an isoelectronic substitution of nitrogen for oxygen in the busulfan molecule. This created a busulfan analogue that was stable for assay calibration. In order to create an immunoassay, they engineered a busulfan antibody fused to a specific nanoparticle. The antibodies bind to busulfan in the patient’s blood sample, enabling a technician to assess the concentration of busulfan by evaluating the scatter of light off the nanoparticles.
This assay has several advantages over the traditional approaches to drug concentration monitoring. First and foremost, it can be run in minutes — not hours — on automated analytical devices already commonly used in hospitals and clinics. That means each sample can be processed when it’s fresh and busulfan activity is most representative of what’s occurring in the patient’s body. Further, the assay requires just a small blood sample — microliters. We believe that will make for a much better patient experience than frequent milliliter venous blood draws.
We are excited about the potential of this game-changing assay and glad to be partnering with Saladax, which has a global reputation for excellence in diagnostics that enable precision dosing. Under the terms of our collaboration, AVROBIO will fund development, validation and regulatory approval of the busulfan assay kit. Saladax plans to file with certain regulatory authorities in 2021. Our goal is to have the busulfan assay kit commercially available as soon as possible, and we expect this will greatly expand access to personalized conditioning with busulfan.
We believe this assay will be a major benefit to patients enrolled in our clinical trials, as it will enable personalized dosing at any clinical trial site, eliminating a major logistical challenge, and freeing the limitation of site selection to specialized GC-MS access. This will be especially important to us once we are at the commercial stage. Yet we felt strongly that we did not want to keep this advancement exclusive to ourselves, but rather give it to the world. The agreement with Saladax stipulates that they will seek regulatory approval in multiple jurisdictions and make the assay commercially available following approval, in countries around the world.
In other words, while we anticipate it will help our programs, this collaboration is not about giving AVROBIO a competitive advantage. It’s about the potential to improve busulfan conditioning for all gene therapy and HSC transplant patients worldwide. We have high hopes that the assay’s arrival will take the uncertainty out of busulfan dosing and further elevate its desirability for personalized conditioning deploying state-of-the-art precision dosing.
We have had strong positive feedback from transplant physicians who see a great opportunity to improve outcomes and treatment uniformity for patients receiving conditioning for a range of indications (not just lentiviral gene therapy). Easy access to rapid, real-time adjustment to dosing is clearly an obvious step-up — and long overdue, in these experts’ view. We view this collaboration as the potential start of a new chapter in busulfan’s remarkable history.