**4. Discussions**

Despite the rapid progress in cartilage engineering, *in vitro* engineering of cartilage with a fine controlled 3D structure, such as human nose, remains a great challenge due to the lack of appropriate scaffolds. PGA has proven to be one of the most successful scaffolds for cartilage regeneration. However, for *in vitro* engineering of a cartilage with a precise shape, PGA unwoven fibers (the most widely used physical form) still have some drawbacks, such as the difficulties in controlling an accurate shape.

To achieve this, a negative mold corresponding to the desired shape is required. CAD/CAM, as a novel technique, has been widely used for the fabrication of anatomically accurate 3D models [20- 23]. Particularly, this method can accurately perform complicated manipulations of the original 3D data, including Boolean operations, mirror imaging, and scaling [24- 26]. CAD/CAM technique was therefore used in the current study for the production of the negative mold for a human nose. Using this mold, PGA fibers were able to be accurately prepared into the nose-shaped scaffold.

The mechanical strength of PGA scaffold alone is not sufficient for the shape maintenance, and thus PLA coating was used to strengthen its mechanical properties as reported [10, 11, 27]. However, a high amount of PLA in the scaffold would negatively affect cartilage formation because of poor cell compatibility [14]. Therefore, an appropriate PLA content in the scaffold is important for both shape maintenance and biocompatibility. In the current study, we evaluated the effects of six PLA contents on the scaffolds' biocompatibility. According to the current results, although the mechanical strength of the scaffolds increased with increasing PLA content, 20% is an acceptable PLA amount for preparing the scaffolds in terms of cell seeding efficiency.

Finally, aided by CAD/CAM technique, the PGA fibers were prepared into the accurate shape of a human nose. Furthermore, by coating with PLA, the scaffold could obtain sufficient mechanical strength to retain the original shape. These results may provide useful information for future nose reconstructions by *in vitro* engineered cartilage as well as for the engineering of other tissues with complicated 3D structures.

In summary, this study established a method to precisely engineer a PGA/PLA scaffold with the shape of human nose. In future studies, we will also investigate the fate of these scaffolds after cell seeding, especially subcutaneous implantation in an immunocompetent animal model.
