Safety Studies for Use of Adipose Tissue-Derived Mesenchymal Stromal/Stem Cells in a Rabbit Model for Osteoarthritis to Support a Phase I Clinical Trial

Abstract

Adipose-derived mesenchymal stem cells (AMSCs) offer potential as a therapeutic option for clinical applications in musculoskeletal regenerative medicine because of their immunomodulatory functions and capacity for trilineage differentiation. In preparation for a phase I clinical trial using AMSCs to treat patients with osteoarthritis, we carried out preclinical studies to assess the safety of human AMSCs within the intra-articular joint space. Culture-expanded human AMSCs grown in human platelet-lysate were delivered via intra-articular injections into normal healthy rabbit knees and knees at risk for the development of osteoarthritis after bilateral medial anterior hemimeniscectomy. Treatment outcomes and safety were evaluated by assessing the general health, function, and behavior of the animals. Joint tissues were analyzed by x-ray, magnetic resonance imaging, and histopathology. Intra-articular AMSC therapy was well tolerated in this study. We did not observe adverse systemic reactions, nor did we find evidence of damage to intra-articular joint tissues. Thus, the data generated in this study show a favorable safety profile for AMSCs within the joint space in support of a phase I clinical trial evaluating the clinical utility of AMSCs to treat osteoarthritis. Stem Cells Translational Medicine 2017;6:910-922.

Keywords: Cell therapy; Osteoarthritis; Rabbit; Regenerative medicine; Stem cell; Transplant.

Figure 1

Study design. An experimental timeline showing the procedures for cohorts A and B. Top: Cohort A: Injection of right and left rabbit knees at time zero (left) followed by sacrifice and tissue harvest at either 7 days after injection (center) or 14 days after injection (right). The left knee of each rabbit was injected with lactated Ringer’s while the right knees were treated with either lactated Ringer’s, 2 × 10⁶ cells, or 12 × 10⁶ cells. Bottom: Cohort B: Injection of right and left rabbit knees at 4 weeks after bilateral anterior hemimeniscectomies. The left knee of each rabbit was injected with lactated Ringer’s, while the right knees were treated with lactated Ringer’s, 2 × 10⁶ cells, or 6 × 10⁶ cells. Animals in this cohort were sacrificed 4 weeks after injection (8 weeks after partial meniscectomy). ƚ, One rabbit in cohort A receiving lactated Ringer’s injections in both knees developed knee swelling and systemic illness consistent with septic arthritis and was euthanized 4 days after injection. ∗, A rabbit scheduled to receive a lactated Ringer’s injection into its right knee was sacrificed before receiving injection therapy because of postoperative complications related to the surgical meniscectomy procedure. Abbreviation: R, right.

Figure 2

Rabbit meniscectomy. Surgical procedure for medial anterior hemimeniscectomy. (A): Rabbit knee after shaving and anatomical marking. (B): A vertical incision made along the medial aspect of the knee joint reveals the patellar tendon and medial collateral ligament. (C): An incision is made along the joint line to expose the meniscus. (D): The infrapatella fat pad is retracted, exposing the lateral insertion of the medial meniscus. (E): The lateral insertion site is released. (F): The meniscus is reflected medially out of the joint. (G): Its attachment at the medial collateral ligament is released. (H): The resected medial meniscus. (I): The fascia and skin are then closed with sutures. (J): The rabbit knee after surgery.

Figure 3

Injection of AMSCs into anesthetized rabbits. Intra‐articular knee injection procedure after shaving, with the injection of stem cells into the rabbit knee joint under ultrasound image guidance, is shown. (A): The rabbit is anesthetized and positioned in the supine position on the operating table. (B): A 19‐gauge needle is inserted into the intra‐articular space along the joint line under ultrasound guidance. (C): Ultrasound imaging shows the needle (arrowhead) passing between the patellar tendon and the femur positioned within the trochlear groove. Abbreviations: F, femur; PT, patellar tendon.

Figure 4

Cohort A histology. Representative H&E‐stained histological samples of cohort A rabbit right knees injected with 12 × 10⁶ AMSCs and the corresponding control left leg (lactated Ringer’s) are shown. Tissues include the femur, tibia, synovium, infrapatellar fat pad, and meniscus. (a–e): No significant pathology was observed in left leg tissues. (f–j): On the right side, mild to moderate aggregation of lymphoplasmacytic cells was observable in all tissues. No tissue injury/damage associated with the infiltrate was observed. Scale bar = 5 mm. Abbreviations: AMSC, adipose‐derived mesenchymal stem cell; H&E, hematoxylin and eosin staining.

Figure 5

Gross morphology of articular surface after meniscectomy (cohort B). Images of rabbit tibias and femurs extracted 8 weeks after meniscectomy depicting gross morphology. (A): Left tibia, lactated Ringer’s. (B): Left tibia, 2 × 10⁶ cells. (C): Right tibia, lactated Ringer’s. (D): Right tibia, 6 × 10⁶ cells. (E): Right femur, 6 × 10⁶ cells. (F): Left femur, lactated Ringer’s. (G): Right femur, 6 × 10⁶ cells. (H): Right femur, lactated Ringer’s. Arrowhead, signs of roughness. Abbreviation: M, medial.

Figure 6

Cohort B histology: 6 × 10⁶ AMSC. H&E‐stained histological samples of cohort B rabbit right knees injected with 6 × 10⁶ AMSCs and the corresponding control left legs (lactated Ringer’s) are shown. Tissues include the tibia, synovium, infrapatellar fat pad, and meniscus. (a–g): No significant pathology was observed in left leg tissues (a–e) or in the right tibia samples (f, g). (h–j): The synovium (h), infrapatellar fat pad (i), and meniscus (j) had mild inflammation and aggregation of lymphohistiocytic cells, with signs of mild subintimal edema in (h) and (j). Scale bar = 5 mm. Abbreviations: AMSC, adipose‐derived mesenchymal stem cell; Ant.Tibia, anterior tibial tubercle side; H&E, hematoxylin and eosin staining; Post.Tibia, posterior tibial fibular side.

Figure 7

Rabbit magnetic resonance imaging. Magnetic resonance T2‐weighted maps were generated for each rabbit knee, and the average T2 relaxation times in milliseconds for the articular cartilage of the lateral and medial femoral condyles and tibial plateaus were measured. (A): No significant differences were detectable between left and right legs for different lateral tissues (lateral femur, p = .904; lateral tibia, p = .965), whereas the right medial femur and right medial tibia had significantly higher (p = .004) T2 times compared with the left sides. (B): No significant differences in T2 were observed between lactated Ringer’s or 2 × 10⁶ or 6 × 10⁶ AMSC‐injected right legs (lateral femur, p = .792; lateral tibia, p = .427; medial femur, p = .078; medial tibia, p = .743). ∗∗, p < .005. Abbreviations: AMSC, adipose‐derived mesenchymal stem cell; L., left; MRI, magnetic resonance imaging; R., right.