|
|
The Most Interesting Things Take Time
The world’s first approved product from monoclonal antibody technology was Muromonab-CD3 (Orthoclone OKT3), launched in 1986. It’s a prime example of “mAb version 1.0,” unleashed before a wave of newer, more manufacturable and better engineered products would begin to enter the market in the late-90s. Later molecular and process iterations based on “mAb 1.0” would take several decades to finally ramify into diverse platforms such as CAR-T, Bi-specific T-cell engagers (BiTEs), diagnostics, and biosensors. As a cellular technology platform, mesenchymal stem/stromal cells (MSCs) are proceeding with a story of convergent evolution. Here, we’ll explain how these versatile cells are enabling a bright future for advanced therapeutics, not merely as standalone cellular doses but rather as essential components of engineered advanced therapeutics for “MSC 2.0” and beyond. Look no further than your nearest Science, 1, 2, 3 Nature, 4 or Nature Communications 5 papers to catch a glimpse of MSC-enabled biomedical breakthroughs. Paralleling the path navigated by CARTs for immuno-oncology as an emergent branch of mAb-tech, these cells are increasingly a sine qua non for synthetic biology, 6 co-cultured cells and tissues, 7 bioprinting, 8 or cell-derived “nano-machines” of targeted and controlled drug delivery. 9, 10
Administered as cell therapies, MSCs’ cumulative clinical trial numbers (>1000) and planned patient enrollments (>60,000) are in a close second place only behind T cells for the years spanning 2017 to 2021. Moreover, at least a dozen MSC cell therapy products are globally approved, including Mesoblast’s Prochymal/RYONCIL® (remestemcel-L) for acute GvHD in Canada and the U.S., TEMCELL in Japan for GvHD, and Takeda’s Alofisel (Darvadstrocel) for Crohn’s disease anal fistulas in the EU, Israel, Switzerland, Serbia, United Kingdom and Japan. There are presently several cell-infused bone or tissue matrices that may contain live MSCs on the market (e.g., via Allosource, NuVasive, Orthofix, RTI Surgical, Smith & Nephew, Stryker Therapeutics, and Zimmer Biomet). Despite this progress, only one MSC-based cell therapy has been approved in the United States to date, Mesoblast’s RYONCIL® on December 18th, 2024.
Although their clinical use is growing globally, press reporting from recent setbacks 11, 12, 13, 14 have suggested that predictable efficacy and consistent CMC—not safety—hinders more widespread adoption of marketed MSC products. As victims of their own early successes that reported dramatic efficacy in early animal studies and human trials, MSCs may trend along the Gartner Hype Cycle™, 15 a familiar topography navigated by many breakthrough technology platforms. Some conventional wisdom has perceived that MSCs map onto the Cycle’s “Trough of Disillusionment,” 16 where both silicon and bio-based industries incubate in the doldrums of product prototypes and their early-adopters, gliding along with academic-oriented DIY and artisanal production methods. Many of these MSC trial materials conformed to what has been dubbed “MSC 1.0,” 17 meaning that they are “First Generation Products Composed of Neat or Minimally Modified MSCs.”
We’ve seen this lull before in other technology arenas. “Television 1.0” first emerged as a kludgy prototype by the mid-1920’s before finally entering mass-produced products after WWII. TV would later combine with transistors, of course, in astonishing ways observed today. “Transistor 1.0,” invented in the late-40s, found its way into portable radios in the 1950s; yet its microprocessor chip progeny did not spawn a revolution in personal computers until the late-70s and smartphones after the year 2000. Just as transistors are more commonly viewed as components of cell phones and not products by themselves, MSCs are transforming into foundational componentry for advanced cell therapy “devices” that may also team up with novel attachment substrates, 18 synthetic genes, 19, 20 priming treatments, 21 co-cultured cell types, 22, 23 engineered tissues, cell-based ‘microbots,’ 24 and secretion of natural nano-sized lipoparticles. 25
In 1990, the late Dr. Arnold Caplan 26 outlined the nascent field of MSCs 27 and was among the first visionaries to describe these cells as a technology platform. Incidentally, Windows 3 began to appear on desktop computers that year. While computational power and GUIs are parsecs ahead in 2024, many clinical therapies that leveraged MSCs would continue to use rudimentary prototype cellular “devices” or “operating systems” without much (or any) functionalization or “versioning.” These tended to exhibit outstanding safety and often signaled good efficacy from IND-enabling pre-clinical data through Phase II trials. However, a robust tally of BLAs from Phase III trials has proved to be far more elusive. Such studies require much larger patient groups, increased dose numbers from multiple manufacturing lots, and multiple clinical translational centers. Perhaps overly-ambitiously, MSC “Version 1.0” was often aimed at broad indication diseases with complex etiologies, where a winning trial design to carefully enroll a cadre of “responders” may have once escaped our diligence. Despite this, MSCs today continue to generate positive headlines as well as promising early-Phase 28 and preclinical data. 29
MSC 1.0’s Rollout Hits Speedbumps
It’s been proposed 30 that advanced MSC-based medicines are affected by phenotypic heterogeneity. Donor type, isolation conditions, and expansion processes prior to dosing may yield populations that are functionally different between patients, ramifying into variable therapeutic efficacy and trial endpoint p-values. Mechanisms of action (MOAs) via MSCs may be highly complex and are not fully understood 31 in the context of widely differing pathologies. That is, MSCs are believed to exert beneficial effects in many ways: via their secretome components (paracrine factors and extracellular vesicles), their ability to migrate toward and plug sites of vascular leakage, their intercellular signaling via direct contact or post-apoptotic engulfment macrophages, or their propensity to differentiate into functional bone, fat or cartilage. 32 Among other properties, MSCs can release anti-microbial peptides for favorable wound healing, 33 promote angiogenesis for restoration of blood vessels, secrete factors or vesicles to resolve inflammatory milieu, and “donate” healthy mitochondria for enhanced bioenergetics in sites of tissues damaged by hypoxia. 34, 35, 36
Today, MSC products are only evaluated according to individual quality attributes. Moreover, universal standards have not fully solidified to selectively bridge their potency markers across varied disease indications. Hence, there have been efforts to leverage a consistent bioprocess and donor type in pursuit of a more homogeneous MSC product. However, this “generic” MSC phenotype (if it even exists) could conceivably water down favorable MOAs that are disease- or patient-specific. MSC material control will thus demand renewal of trial designs to stratify cohorts according to their “omics”—combined with expertise in precision medicine—to (1), find the right MSCs for the patients and (2), find the right patients for the MSCs. 37
MSC 1.0’s major recent calibration event was on October 1, 2020, when the FDA sent Mesoblast a Complete Response Letter in response to the BLA for remestemcel-L to treat severe GvHD in children. This is despite consistent signals of efficacy, no alternative treatment options, and a 9:1 recommendation by the Oncologic Drugs Advisory Committee in support of approval.
The FDA’s comments spotlighted the “…Need for further scientific rationale to demonstrate the relationship of potency measurements to the product’s biologic activity” and that the quality attributes do not evince a “demonstrated relationship to the clinical performance of specific lots,” which “may not be sufficient to ensure the manufacturing process consistently produces remestemcel-L lots of acceptable quality.”
To put it mildly, FDA as the global regulatory “gold standard” didn’t instruct MESO to step on the gas and manufacture hundreds of thousands of remestemcel-L doses. Quite the opposite! The Agency’s message was aimed clearly not just at Mesoblast, but for other plucky cell therapy developers, even those touting outstanding preclinical data. And yet, this is not to be a death knell so much as a reveille.
What is underappreciated is that MSC bioproduction is becoming democratized via a sea change brought about by subtle upstream and downstream process innovations. Now that there are tens of thousands of patients dosed over the past 20+ years, MSC GMP manufacturing has become standardized and implementable at both small translational center scales, as well as within industry at 10s of Billions of cell lot sizes (with scalability set to take this to 10-100X lot size increases as demand increases). In addition, with the GMP supply chain for cellular raw materials now readily available and off the shelf, it has dramatically lowered the bar for anyone to test MSC 1.0 products in new ways and for new indications. With a base foundation for new product innovation laid, acceleration towards a long awaited MSC “system upgrade” is inevitable. The real-world historical example via T cells informs us that this may also be imperative.
T Cell 2.0 – From Fits and Starts to TRUCKs and CAR-Ts
Adoptive cell therapy (ACT; or adoptive cell transfer) 38 is defined as the ex vivo identification and selection of lymphocytes with specific anti-tumor activity, which are then re-infused, perhaps along with systemic administration of lymphocyte growth factors like IL-2. The earliest demonstrations of this concept have been practiced in vivo since the 1960s. By 1988, an experimental therapy 39 for human metastatic melanoma was devised that combined isolated tumor-infiltrating lymphocytes (TILs) with IL-2 after a lymphodepletion with cyclophosphamide. Objective if not transient regressions were observed in 9 of 15 (60%) patients who had not been previously treated with IL-2. Despite this initially promising result, early adoptive transfer therapies were often constrained by both the persistence of the TILs and limited applicability beyond melanoma, 40 which is a cancer type uniquely prone to abundant expression and presentation of diverse neoantigen targets. 41
Fortunately, the technology landscape of immuno-oncology did not lie fallow. New “software” toolsets from molecular biology upgraded T cell “hardware” to retarget an immune response to specific cancer antigens. The first prototype CAR-Ts were invented 42 separately 43 by Dr. Yoshikazu Kurosawa’s group 44 in 1987 and Dr. Zelig Eshhar and coworkers in 1989, 45 which chimerized specific antibody variable light and heavy chains with C-terminal TCR alpha and beta chains and endowed the cells with antibody-like specificity. These early chimeric antigen receptors (“CARs”) needed to be encoded on separate retroviral vectors and thus were impractical due to low co-transfection efficiency. However, so-called first-generation CARs yielded preclinical breakthroughs by 1994, which fused a single chain antibody with sequence from CD3ζ or FcϵRIγ. This configuration could reliably be applied in vivo to shrink ERB2+ xenografted tumors 46 in athymic nude mice.
By the late 1990s, first-generation (Gen 1) CARs were injected into human HIV and colorectal cancer patients via the efforts of Margo Roberts, Kristen Hege and coworkers at Cell Genesys, Inc. Although these early trials were deemed to be safe, they lacked efficacy. Nevertheless, their manufacturing protocol was robust due to the use of anti-CD28 beads for co-stimulation, along with the nominal anti-CD3 and IL-2. Around this same time, Dr. Michel Sadelain 47 of Memorial Sloan-Kettering (MSKCC) and Dr. Helene Finney and team at Celltech Therapeutics Ltd. devised second-generation CARs, 48 which added a costimulatory intracellular domain from CD28, together with the CD3ζ segment. Such designs performed with highly amplified responses in vivo. Other Gen 2 CAR-T formats utilized 4-1BB/CD137 for CARs in lieu of CD28. Compared with CD28 domain CARs, the 4-1BB versions appeared to exhibit pharmacodynamic properties of increased persistence and less aggressive cytokine release syndrome 49 in some contexts. Gen 2 CARs rapidly entered human trials in the 2000s against both solid and liquid tumors.
Trial-and-error taught this (initially) small group of investigators some valuable lessons. Notably, the molecular targeting of by CARs 50 or cloned TCRs 51 on solid tumors could be more challenging than expected—even deadly—due to on-target off-tumor (OOT) effects and aggressive cytokine release syndromes. 52 These results rallied considerable focus of CAR design onto a single target: CD19 on B cell lymphomas and leukemias. In the watershed year 2013, powerful efficacy in human trials from CARs designed by groups led by Drs. Carl June 53 and Michel Sadelain 54 were reported with their respective Gen 2 therapies, which led to the tisagenlecleucel’s (Kymriah®) approval as the first gene therapy in the United States in August 2017. The current list of five other FDA-approved CAR-Ts (Kymriah, Yescarta, Tecartus, Breyanzi, Abecma, and Carvykti) are also second-generation designs containing CD28 or 4-1BB, and target CD19 or BCMA—all B cell related cancers. By May 2022, Dr. Carl June announced that 15,000 patients have already been dosed with CAR-T therapies.
CD19 continues to be a benchmark target for new CAR design, which has evolved through third, fourth, and even fifth generations. 55 Gen 3 CARs combine co-stimulatory CD28 and 4-1BB fragments within a single signaling domain for more potency and persistence. Gen 4 CARs (aka “TRUCKS” or “armored CARs”) utilize increasingly complex transgene designs encoding both the CAR as well as a cytokine (e.g., IL-12 or IL-15) that potentiate T cell viability serial killing without exhaustion. So-called Gen 5 CARs leverage synthetic biology principles 56 for combinatorial regulatory sequences and/or gated logic targeting functions and rheostatic, promoter-controlled bioactivity, as well as suicide switches for increased safety. With Gen 5, it’s commonly asserted that CAR-Ts can finally broaden their scope of molecular targeting to finally include solid tumors, infectious diseases, or even chronic conditions like heart disease. CAR engineering is no longer limited to T cells, but also is in development for NK cells, and was even demonstrated 6 for in vivo effectiveness with MSCs. 57
Next Gen MSC Applications Rev Up for High Performance Alongside the CARs
Hard-won knowledge earned along the “long road” 58 of CAR-T clinical development is directly applicable to MSCs. Obviously, T cells could not assert their full prowess until investigators learned how to harness them with fine-tuned controls on their potency, viability, targeting, and safety via genetic engineering. A new model system, “T Cell 2.0,” was essential. With CAR-T, “T Cell 1.0” (adoptive cell transfer) has been transformed into a controlled drug delivery modality whereupon its primary active ingredient is a synthetic, programmable gene network. The term “MSC 2.0” 17 was similarly coined to encompass applications of MSCs that are a leap beyond traditional uses where MSCs once comprised the sole ingredient. The MSC 2.0 concept evinces a template for custom engineering by genetic modification, bioproduction of secreted products, priming by unique physical and media culture additions, or use as co-cocultured material alongside other cells in tissue engineering. To wield a different analogy, the global $286 billion dollar satellite industry doesn’t generate most of its wealth from the launch systems; it’s the payloads delivered to float in orbit for useful tasks that are key to ROI. Likewise, for both MSCs and T cells, we might better imagine them as “vehicles,” as means rather than ends. MSC 2.0 concepts therefore unfetter the clinical developer to be deterministic about the MSCs’ MOAs—to properly stage each new cell therapy cure for each new mission and clinical indication.
Unlike CD8+ T-cells, the main job description for MSCs isn’t killing virus-infected cells or tumors. Instead, MSCs are closely related 59 to vascular pericytes that help remodel a hypoxic and/or inflammatory wound milieu after infection or physical injury. 60, 61, 62 Like T-cells, MSCs have been in clinical trials for decades. In contrast, however, MSCs can be safely administered as an allogeneic therapy with no prior immunosuppression. 63 MSCs even appear to be tolerogenic in some settings, showing clinical signals of immune system cooling and efficacy against diseases like GvHD or autoimmune indications like arthritis. 64
The talent of MSCs for so-called “regenerative medicine” and allogenic applications lowered barriers that would otherwise block their mass expansion in bioreactors. 65 With ongoing accelerating process innovations quietly reshaping the opportunity space, MSC culture expansion systems now accommodate higher cell numbers than T cells by two orders of magnitude. These instrument platforms are demonstrated to exceed 100-liter batch volumes 66 and cell harvests in the 10s of billions. 67 Depending on bioprocess or indication, this may equate to hundreds of trillions of extracellular vesicles 68 (EVs)—or sufficient doses for 10s of trial patients per run. 17, 68, 69 Nevertheless, for MSCs to be anointed with a “Kymriah moment” within the USA, allo-MSCs may need to be manufactured and administered even more consistently and with stringent control of the CQAs 70 and CPPs, 71 applying an industrial mentality to right-scaled applications. 72, 73, 74, 75 Hence, what’s been termed “Life Science Industrials” by Dynamk Capital is sure to play an important ongoing role for the pioneers of CAR-T and MSCs that demand a solid bridge between cell-drug discovery and biomanufacturing.
On closer inspection, it’s now clear that most technology drivers of the MSC opportunity space have migrated from the “Trough of Disillusionment” to the “Slope of Enlightenment.” 16 Obstacles were stumbled on in the past. Yet investigators got back up, dusted themselves off and left these behind for a long hike ahead to the summit. This pattern is described as when:
“…More instances of how the technology can benefit the enterprise start to crystallize and become more widely understood. Second- and third-generation products appear from technology providers. More enterprises fund pilots; conservative companies remain cautious.” 15
Tell Us What You Think!