Can Early Failure of Cartilage Implants Be Detected with Ferumoxytol Labeling?

See also the article by Theruvath and Nejadnik et al in this issue.

Introduction

Ferumoxytol is a second-generation superparamagnetic iron oxide (SPIO) nanoparticle noted for its “ultrasmall” size (ultrasmall SPIOs are defined as having a hydrodynamic diameter of < 50 nm) and associated long serum half-life (10–14 hours, compared with 10–12 minutes for regular SPIOs) (1). Ferumoxytol was originally designed as an MRI contrast agent but is predominately now used as intravenous iron supplementation for severe anemia (2). It has garnered much interest recently as an alternative contrast agent to gadolinium chelates, which are being flagged amid increasing concerns about nephrogenic sclerosis and gadolinium brain deposition.

In this issue of Radiology, Theruvath and colleagues (3) report a different application of ferumoxytol in which they labeled matrix-associated stem cell implants (MASIs) ex vivo as a way to track MASI implants in porcine knees in vivo with MRI. Although previous experiments have tested iron oxide nanoparticle labeling of therapeutic cells, to our knowledge, this is the only published study to test the safety profile of ferumoxytol-labeled MASIs in a large-animal model.

MASIs are autologous mesenchymal stem cell–based cartilage implants derived from bone marrow. Because very few (<0.01%) true stem cells are derived from bone marrow aspirate, the term connective tissue progenitor cells is sometimes preferred. For the purpose of MASIs, progenitor cells are selected from bone marrow aspirate, expanded ex vivo, filled into bioengineered matrices, and then implanted arthroscopically, typically into full-thickness chondral defects greater than 2 cm² (4). In this environment, cells in a successful implantation encourage chondral repair, either through direct differentiation into chondrocytes or indirectly by releasing hormones/cytokines that promote the differentiation of neighboring cells.

MASI is desirable compared with currently approved repair techniques like autologous chondrocyte implantation (ACI), which requires two separate surgeries: one for chondrocyte harvesting and another for implantation. Various forms of ACI have been developed since 1994 (5), including matrix-associated ACI, which received U.S. Food and Drug Administration approval in 2016 for use in full-thickness chondral defects of the knee (6). In addition to simplifying the repair process, connective tissue progenitor cells may also produce better hyaline-to-fibrocartilage ratios in the regenerated tissue compared with chondrocyte implantation (7). Despite the potential utility of mesenchymal stem cell–based therapeutic implants, few studies using bone marrow or adipose tissue cells have progressed to phase III clinical trials, in part because of the time required for product development, as well as uncertainty about clinical effectiveness, safety, and product differences (6,8).

In our experience (9) and that of others (4), conventional MRI assessment of cartilage implants is difficult within the first several months after implantation. Implants often appear diffusely hyperintense relative to native cartilage on water-sensitive images, lacking the expected grayscale stratification (reflecting collagen orientation) that is typically seen in healthy cartilage until the late postoperative period (approximately 12 months after implantation). Cartilage grafts may also show hypertrophy in more than 25% of cases (4,10). It can therefore be challenging to determine if the chondral implant will properly integrate and mature. If the chondral repair technique fails at a very early stage, rescue interventions at a much later stage may be ineffective. A noninvasive objective method to detect early failure would be advantageous for improving overall MASI outcomes by facilitating earlier intervention to rescue failed grafts and could also help inform future technologic improvements to MASI and other cartilage repair procedures.

In their study, Theruvath and colleagues (3) demonstrated that ferumoxytol labeling of MASIs prior to implantation in porcine knees was effective for qualitative and quantitative T2 mapping MRI assessment of graft integrity at 2 weeks after surgery. Standard MRI revealed incomplete cartilage repair of apoptotic MASI only after 24 weeks. The authors employed four groups of cells to assess the safety and utility of ferumoxytol labeling: (a) ferumoxytol-labeled MASIs versus unlabeled controls to test whether ferumoxytol hinders chondrogenesis and (b) viable versus apoptotic MASIs to test whether ferumoxytol labeling can be utilized to assess engraftment outcomes.

The authors found that labeled, viable MASIs had significantly shorter T2 relaxation times than unlabeled controls at week 1 (22.2 msec ± 3.2 vs 27.9 msec ± 1.8; P < .001), but by week 4, the iron had metabolized and cartilage signal and T2 relaxation times did not differ (26.1 msec ± 4.0 vs 27.0 msec ± 0.3; P = .34). Therefore, MR observation of cartilage repair tissue, or MOCART, scores at 4 and 12 weeks in both viable cohorts showed no differences. There was also no difference in joint function or histopathologic analysis of cartilage defect repair outcomes, suggesting that ferumoxytol labeling of MASIs does not inhibit chondrogenesis. By the 2nd week following implantation, ferumoxytol-labeled apoptotic MASIs exhibited loss of iron signal, manifesting as T2 prolongation (26.6 msec ± 4.9), while ferumoxytol-labeled viable MASIs demonstrated significantly lower T2 values (20.8 msec ± 5.3; P = .001). Loss of iron at 2 weeks in the apoptotic MASI group correlated with incomplete chondral repair, supported by histopathologic findings at 12–24 weeks.

A potential study limitation was the way in which apoptosis was induced—that is, with the use of mitomycin. Mitomycin-induced apoptosis is an artificial process that likely differs, with regard to the rate and mechanism, from naturally occurring apoptosis after cell implantation. An alternative method would be placement of fibrin glue without cells or with frozen nonviable cells, but presumably the authors’ goal was to model what might ensue following implantation of viable cells that then undergo apoptosis. The authors also mention that apoptotic cells of MASIs are cleared by macrophages, but do not demonstrate any macrophages in the histologic results provided. Macrophages likely do play a central role in tissue repair and regeneration, and excessive macrophage accumulation is associated with cytokine production that leads to fibrosis. It is possible that impaired healing in those groups receiving MASIs with apoptotic cells was due more to an adverse effect of macrophages rather than to the loss of implanted cells.

Overall, this study provides supportive safety and efficacy data that could translate to a human study of ferumoxtyol-labeled MASIs. It also provides impetus to the labeling of stem cells to enhance repair and regeneration of other orthopedic tissues (eg, stem cell augmentation of rotator cuff repair). One consideration is to determine whether the type of microenvironment (eg, “proinflammatory” [such as in arthritic knees] compared with the relatively “sterile” knees used in this experiment) would affect the stability of ferumoxytol-labeled tissue. Considerations will also have to be made to determine the optimal labeling dose and technique, which may vary between tissue types. As the authors allude to in the article, alternative quantitative techniques such as R2* mapping may be more sensitive for deeper layers of articular cartilage that have rapid T2 decay values and for the presence of iron, which may facilitate the tracking of ferumoxytol-labeled MASIs for longer than the 2 weeks that the authors were able to follow MASIs in their experiment. Following MASIs for a longer time period may be important, as the “critical” time when failure or poor biologic incorporation occurs is not precisely known. These techniques may also be able to provide increased specificity that could determine the percentage of viable cells rather than simply define a cartilage implantation as a “failure” or “success.”