**1. Introduction**

Cartilage defects are highly prevalent joint disorders and leading causes of disability and chronic pain in the world [1]. Articular cartilage defects are mainly attributed to trauma, osteoarthritis, osteonecrosis, and osteochondritis dissecans. Chondral lesions were found in around 60% of the patients undergoing knee arthroscopy [2–4].

Back in 1743, Hunter described the challenge of cartilage repair, stating that "once the cartilage is destroyed, it never recovers [5,6]." His observation still holds today. The avascular characteristics of cartilage constrain its self-regeneration from injury. If left untreated, the damaged cartilage gradually progresses into severe osteoarthritis. Through decades of effort, multiple surgical treatments have been developed to promote cartilage healing, such as abrasion arthroplasty [7], microfracture [8,9], and mosaicplasty [10,11]. However, these surgical approaches are usually associated with fibrocartilage formation [12,13], limited tissue sources [14], and donor-site morbidity [15], and its long-term efficacy remains controversial [16].

Cell therapy for cartilage repair was proposed in the 1980s by Robert Langer and Charles Vacanti using the approaches of tissue engineering. They clearly defined that a fine reconstruction of cartilage defect must include selected cells for transplantation, excipients, which are seeded with selected cells, and functional restoration of defect areas as before [17]. The cellular therapeutic innovation was realized in 1994. Autologous chondrocyte implantation (ACI) was introduced to treat cartilage defects in the knee [18]. In the past two decades, ACI has demonstrated its efficacy for knee osteoarthritis. Two systematic reviews concluded that, relative to microfracture and mosaicplasty, ACI may be the best option for large defects in active young patients who have had the symptoms for a short period and have not undergone a chondral surgery before [13,19]. In addition, a scaffoldbased ACI, matrix-induced autologous chondrocyte implantation (MACI), outperformed microfracture in a 2-year randomized study [20]. Still, only a few products are available on the market: Carticel (FDA-approved in 1997), Chondron (South Korea MFDS-approved in 2001), and MACI (FDA-approved in 2016). The scarcity implies that some limitations remain, such as the limited source of chondrocytes, donor site morbidity, uncertain hyaline cartilage formation, low recovery of the recipient site, and questionable longevity of these implants or their derivative tissues [21].

The discovery of adult stem cells aroused a paradigm shift in regenerative medicine. The features of self-renewal and multipotency of stem cells make them ideal cell sources for cellular therapy. A variety of stem cell-based therapeutic innovations have been developed using mesenchymal stem cells (MSCs) derived from bone marrow [22], adipose tissue [23], synovium [24], peripheral blood [25], or periosteum [26]. Several clinical trials have demonstrated the safety and therapeutic efficacy of autologous MSC implantation for cartilage repair [27–30]. However, the implantation of undifferentiated MSCs cannot guarantee certain chondrogenesis in vivo, which might lead to heterogeneity of regenerated tissues.

The chondrogenesis of MSCs can be guided using growth factors [31] or biophysical/biomechanical stimuli [32] to improve the functional properties of the derived neocartilage tissues, including mature matrix formation [29,33]. Our previous study identified a unique population of chondrocyte precursors (CPs) derived from bone marrow mesenchymal stem cells (BMSCs) during chondrogenic induction [34]. These atelocollagenembedded CPs (Kartigen®) can secrete glycosaminoglycan (GAG) and collagen type II but without lacunae formation.

A variety of biomaterials have been used for cartilage tissue engineering, including collagen [35], alginate [36], poly-lactic-glycolic acid [37], and tri-copolymer [38]. In this study, we selected atelocollagen because it is a low-immunogenic derivative of collagen obtained by the removal of N– and C–terminal telopeptide components [39]. Atelocollagen has been broadly applied in the regeneration of cartilage [34,40,41], intervertebral disc [42], cornea [43], periodontal tissues [44], and skin [45] to serve as a carrier for cell delivery and to provide an appropriate microenvironment for tissue regeneration.

A 9-year follow-up trial demonstrated that Kartigen® integrated with the host tissue and resulted in the formation of hyaline-like cartilage, thereby improving the impaired knee functions [34]. Due to the lack of a randomized control group in the previous study, we initiated this controlled and randomized trial to evaluate the safety and efficacy of Kartigen for repairing cartilage defects in the weight-bearing site of medial femoral condyles through the comparison with the microfracture treatment.
