The five year outcome of a clinical feasibility study using a biphasic construct with minced autologous cartilage to repair osteochondral defects in the knee

International Orthopaedics (2020) 44:1745–1754 https://doi.org/10.1007/s00264-020-04569-y ORIGINAL PAPER The five year outcome of a clinical feasibility study using a biphasic construct with minced autologous cartilage to repair osteochondral defects in the knee Tzu-Hao Tseng 1 & Ching-Chuan Jiang 1 & Howard Haw-Chang Lan 2 & Chun-Nan Chen 3 & Hongsen Chiang 1 Received: 14 January 2020 / Accepted: 6 April 2020 / Published online: 4 May 2020 # SICOT aisbl 2020 Abstract Purpose Autologous minced cartilage has been used to repair cartilage defects. We have developed a biphasic cylindrical osteochondral construct for such use in human knees, and report the five year post-operative outcomes. Methods Ten patients with symptomatic osteochondral lesion at femoral condyles were treated by replacing pathological tissue with the osteochondral composites, each consisted a DL-poly-lactide-co-glycolide chondral phase and a DL-poly-lactide-coglycolide/β-tricalcium phosphate osseous phase. A flat chamber between the two phases served as a reservoir to house double-minced (mechanical pulverization and enzymatical dissociation) autologous cartilage graft. The osteochondral lesion was drill-fashioned a pit of identical dimensions as the construct. Graft-laden construct was press fit to the pit. Post-operative outcome was evaluated using Knee Injury and Osteoarthritis Outcome Score (KOOS) up to five years. Regenerated tissue was sampled with arthroscopic needle biopsy for histology at one year, and imaged with magnetic resonance at one, three, and five years to evaluate the neocartilage with MOCART chart. Subchondral bone integration was evaluated with computed tomography at three and five years. Results Nine patients completed the five-year follow-up. Post-operative mean KOOS, except that of the “symptom” subscale, had been significantly higher than pre-operation from one year and maintained to five years. The change of MOCRAT scores of the regenerated cartilage paralleled the change of KOOS. The osseous phase remained mineralized during the five-year period, yet did not fully integrate with the host bone. Conclusions This novel construct for chondrocyte implantation yielded promising mid-term outcome. It repaired the osteochondral lesion with hyaline-like cartilage durable for at least five years. Keywords Cartilage repair . Clinical feasibility test . Polylactic-co-glycolic acid . Tricalcium phosphate . Autologous cartilage implantation Introduction The defects of articular cartilage had been regarded “once destroyed, never repairs” in the natural setting since such discovery by Dr. John Hunter two centuries ago. These lesions * Hongsen Chiang hongsen@ntu.edu.tw 1 Department of Orthopaedic Surgery, National Taiwan University Hospital, 7 Chungsan South Road, Taipei 10002, Taiwan 2 Department of Medical Imaging and Radiological Sciences, College of Health Sciences, Central Taiwan University of Science and Technology, Taichung, Taiwan 3 Exactech Taiwan Co., Ltd., Hsinchu, Taiwan deteriorate the function of joints in young, working population, with symptoms comparable to severe osteoarthritis (OA) [29]. If left untreated, these defects progress and eventually evolve into advanced OA [46]. The treatment of focal chondral and osteochondral defects has challenged orthopaedic surgeons with their limited healing after operation. Therefore, the surgical strategies have advanced from traditional reparative approaches to more ambitious regenerative procedures, aiming to restore the articular surface with hyaline-like cartilage instead of fibrocartilage [6, 14]. Autologous cartilage implantation (ACI) is one option of cartilage regeneration and yields promising clinical outcome. The durability of ACI has also been reported in previous studies [42, 48]. The success of original methods of ACI further encourages subsequent research and evolution of the technique. 1746 Originally, the ACI placed cultivated autologous chondrocytes in a defect sealed with a periosteum patch [10]. Because of the high incidence of subsequent graft hypertrophy, the periosteum patch was substituted with collagen membrane in the revised model of ACI [27]. However, both original and revised models had surgical and biologic drawbacks including a technically demanding process, uneven cell distribution, and cell leakage [44]. The matrix-assisted autologous chondrocyte implantation (MACI) was developed to overcome these problems [6, 25], using biodegradable materials to serve as the temporary scaffolds and to maintain the implanted chondrocytes at the recipient sites. The MACI is preferable for its easier techniques and comparable clinical results to ACI [6]. All these chondrocyte implantation techniques have a huge demand of cells for implantation; therefore, surgically harvested cells require a culture-expansion process in the laboratory before they can be implanted in the second surgery. Recent in vivo studies reported that lower seeding density on selected biomaterial resulted in good cartilage synthesis [12, 15], given that “optimal growth-conditions” were provided. These findings had made possible the convenience by one-stage surgery of cartilage regeneration. Because the isolated chondrocytes have been deprived of all their natural housing which may be critical for their presentations of building neocartilage, implantation of minced cartilage may be more promising for cartilage regeneration. Chondrocytes in the minced cartilage stay in a more natural environment that may better support their original phenotype. Minced cartilage embedded into appropriate polymeric scaffolds for grafting has been reported to be a feasible design for one-step cartilage regeneration [36, 39]. One clinical randomized trial showed promising shortterm outcome at two years [18]. Osteochondral lesions are more complex to repair than pure cartilage defects, because they involve extensive pathology at the subchondral bone that supports the cartilage graft [26, 47]. Combining MACI with autologous bone grafts to replace the subchondral bone can be a solution [45], but the surgical procedures are complicated. Alternatively, using a composite osteochondral matrix to support simultaneous regeneration of both cartilage and subchondral bone would be more easily applicable with the benefit of developing a secure osteochondral interface within the graft. Various osteochondral scaffolds have been fabricated for such use [47], but only a few of them have been tested clinically with short-term results being reported [7, 9, 11, 14, 19, 20, 24, 30, 32, 41]. One example was the clinical use of the TruFit Plug [7, 20, 30, 41, 49], an osteochondral scaffold designed for MACI. It had been withdrawn from global market due to the poor growth of subchondral bone with formation of cysts and sclerosis two years after implantation [49]. Another multilayered nano-composite biomaterial has also been used clinically for osteochondral regeneration; however, it is a cell-free scaffold and reports about the surgical outcome exist conflicting results [11, 16, 32]. International Orthopaedics (SICOT) (2020) 44:1745–1754 Wrapping up these suggestions in the literature, we had developed a composite osteochondral construct for the implantation of double-minced autologous cartilage with a onestage surgery. We conducted a clinical feasibility study using this novel construct to replace osteochondral lesion on the femoral condyles of the knee joint and reported the two-year surgical outcome [14]. We kept following these patients to testify the safety and durability of the treatment until five years. This article reports the clinical assessment of knee function and the evaluation of regenerated tissues with magnetic resonance images (MRI) and computed tomography (CT) during the mid-term follow-up. Methods Biomaterial The biphasic construct was fabricated with a modified solvent merging/particulate leaching method as previously described [34]. The porous cylindrical construct consisted of a chondral and an osseous phase. The chondral part was one-sixth of its height and made of polylactic-coglycolic acid (PLGA). The rest of the cylinder was the osseous part made of PLGAtricalcium phosphate (TCP) composite (Fig. 1). A flat chamber of 6.5 mm diameter and 1 mm height lay between the chondral and osseous parts and served as the reservoir for double-minced autologous cartilage graft. The final product was a cylindrical construct 8.5 mm diameter and 8.5 mm high. All the biphasic constructs in this study were fabricated in a good manufacturing practice (GMP) compliant laboratory. Toxicology tests and animal experiments had been done for preclinical evaluation [34, 35]. The efficacy of the construct for cartilage regeneration had also been testified with a porcine model [13, 15]. Fig. 1 The biphasic osteochondral construct. The asterisks and bars (1.5 mm) indicate the chondral phases, and the rest height of the constructs is its osseous phase. The construct is a barrel-and-plug constitution, facilitating an easy loading of the minced cartilage from the opening at the osseous side. After loading of cartilage graft, the plug is assembled to conceal a flat chamber housing the graft between the plug and the chondral phase of the construct International Orthopaedics (SICOT) (2020) 44:1745–1754 Patient recruitment and baseline evaluation This clinical feasibility study was a prospective cohort trial conducted after its protocol being approved by the Institutional Review Board of the corresponding author’s hospital and the national ministry of health, ensuring the full compliance to the national legal regulation on clinical trial and the Helsinki Declaration. We invited consecutive patients who fulfilled the following criteria in our clinic to participate in this study: (1) age of 18– 60 years, (2) skeletally mature with closure or absence of physeal plate at distal femur on the plain roentgenography, (3) having a symptomatic osteochondral lesion on the femoral condyle of knee joint, as diagnosed by plain roentgenography and MRI, (4) the lesion with largest dimension not exceeding 20 mm on the MRI, and (5) the lesion of grades three or four by the Dipaola’s classification [21]. The exclusion criteria were (1) any unhealed fracture in either lower limb, (2) diffuse degenerative change of the object knee, (3) serologically confirmed rheumatism, (4) metabolic arthropathy such as gouty arthritis, (5) knee stiffness for any reason with a flexion contracture exceeding 20° and/or maximal flexion less than 130°, and (6) pregnant patients. The lesions were re-evaluated with arthroscopy during the surgery; only those of grade three or four by Guhl’s classification [28] would proceed to the osteochondral implantation. Twelve patients were invited and ten of them were included, involving six men and four women with the average age of 27.6 years. Every participating patient signed an informed consent. As described before [14], the MRI findings of all these lesions were uniform, featuring well-defined osteochondral area with intermediate signal in all sequences that clearly delineate the pathological tissue from the host bone. Three patients had undergone operations in the object knees: one had anterior cruciate ligament reconstruction two years earlier, one had arthroscopic debridement for the osteochondral lesion one year prior, and one had repair of the osteochondral lesion with an autologous periosteal patch one year prior. None of the procedures resulted in successful cartilage regeneration in the osteochondral defects. All the patients signed an informed consent to participate in this study. Pre-operative knee function was evaluated with the Knee Injury and Osteoarthritis Outcome Score (KOOS) questionnaire, plus the pain score reported by the patient on a 100mm visual analog scale (VAS) during sitting down and standing up from a chair. Surgical procedures and post-operative protocol The details of surgical procedures and post-operative protocol had been described previously [14]. In brief, we first inspected the knee joints with arthroscope to locate the lesions. Depending on their location, the lesions were approached 1747 via a longitudinal mini-arthrotomy along the medial or lateral border of patellar tendon. An 8-mm diameter area was defined with a circular punch to fully contain a smaller lesion. In case of a larger lesion, a maximum of three adjacent circles should be able to contain it or the case would be excluded. Cartilage within the defined area was excised and collected as part of the autograft. Additional autograft was curette-harvested from non-articulating margin of the affected condyle until the total packed volume of 0.15 cm3 cartilage was collected. The autograft was immersed immediately in sterile saline and morselized with a power-driven tissue pulverizer to particles smaller than 1000 mm, as being filtered through a sieve in the pulverizer. Cartilage particles were further dissociated with Collagenase (Librase, Roche Diagnostics, Mannheim, Germany) at 37 °C for 20 min. The collagenase was removed by copious rinses with saline. The processed graft was collected and transferred to the flat chamber in the biphasic construct. The recipient sites were prepared by drilling the subchondral to form a round pit identical to the shape and size of the construct. Adjacent pits might be prepared for larger lesions as being previously defined by the circular punch. The prepared osteochondral construct was manually press fit into a pit, so that its chondral surface was flush with the articular surface (Fig. 2a, b). Among the patients, one required two constructs, another one required three constructs, and the others needed a single construct to repair the lesions. After the operations, the knees were protected with a brace to restrict its range of motion from full extension to 90° flexion for six weeks. Ambulation with partial weight-bearing on the operated leg was allowed the same day after operation. These patients would refrain from vigorous sports activities until six months postoperation. Follow-up evaluation As the protocol of this feasibility study, second-look arthroscopies were done to harvest a sample of regenerated tissue at 12 months post-operatively. A sample of regenerated tissue was collected from the centre using a 1-mm diameter biopsy trephine driven perpendicular to surface of the repair to about 10 mm deep. We had reported the short-term outcomes at 12 months post-operation, including function scores, plain roentgenography, MRI, second-look arthroscopic appearance of the repaired lesion (Fig. 2c), and histologic evaluation of regenerated tissue sampled by needle biopsy during the arthroscopy [14]. We kept following these patients for five years. MRI was acquired from a 1.5-T machine (GE Sigma EXCITE) at three and five years, including sequences of fast spin-echo (T2-FSE), three-dimensional fat-suppressed gradient-echo (3D-GE-FS), and others. The images of the grafted sites at 12 months, three years, and five years postoperation were interpreted by an independent radiologist to evaluate the quality of regenerated tissue using the revised 1748 International Orthopaedics (SICOT) (2020) 44:1745–1754 symptoms (Symptoms), activities of daily living (ADL), sports and recreational activities (Sport/Rec), and quality of life (QoL) was analyzed separately. The total score from a subscale was normalized and transformed according to the KOOS manual to obtain the final score as a percentage reversely proportional to the severity of problem, with 100 indicating no problem at all. Statistical analysis Fig. 2 An example of replacing an osteochondral lesion with the biphasic construct loaded with double-minced autologous cartilage. a The lesion on the lateral femoral condyle was identified through a mini-osteotomy. b The lesion was replaced with the construct, with its surface being flush with the native articular level. c Arthroscopic appearance of the repaired lesion at one year after the surgery. Regenerate cartilage filled the original lesion with a surface flush with the surrounding joint surface. MOCART (Magnetic resonance Observation of Cartilage Repair Tissue) chart [37, 38]. CT scans were taken at three and five years; the images were interpreted by the same radiologist to evaluate the integration of the construct to the surrounding subchondral bone at the recipient sites. First, percentage of disappeared implant-bone interface along the entire boundary of the implant’s osseous phase was estimated. Second, percentage of the total osseous phase that homogenized with the host subchondral bone, as being indicated by the homogenization of mineral density on the image, was estimated. Subjective functions of the operated knees were evaluated clinically with KOOS and VAS score at two, three, and five years post-operation. The score from each of the five subscales of KOOS questionnaire, including pain (Pain), other Mean KOOS and mean VAS scores at each time of follow-up were compared with the pre-operative value respectively using Wilcoxon signed rank test. Mean KOOS in respective subscale at 12 months, 24 months, three years, and five years were compared using Kruskal-Wallis test with Dunn’s test as the post hoc analysis. The MOCART scores of regenerate cartilage at one, three, and five years post-operation were compared using Cochran’s Q test with McNemar test as the post hoc test. The CT findings of subchondral condition at three and five years were compared using McNemar test. Before the analysis, we simplified the MOCART scores to only two grades for each variable: complete or incomplete “filling of the defect,” complete or incomplete “integration to border zone,” intact or brock “surface,” isointensive or non-isointensive “signal intensity.” The score from each variable was analyzed respectively without being summed up. Similarly, we simplified the CT findings to only two grades: > 50% or < 50% “visible implant-bone interface,” and < 40% or > 40% “homogenization of implant with host bone.” Statistic calculations were exercised with SPSS software (version 20.0.0, SPSS Inc., Chicago, IL) on a Microsoft Windows-based computer. For each statistical comparison, the p < 0.05 was regarded a significant difference. Results Totally nine patients completed the five-year clinical follow-up. There was no minor or major adverse effect since postoperative three months. The mean KOOS of all subscales at both three and five years were significantly higher than the pre-operative values, respectively (Fig. 3a). Contrarily among the postoperative scores, the mean KOOS in any subscale at 12 months, 24 months, three, or five years were comparable with one another time of acquisition. The mean three-year and five-year VAS scores at chair rise and the mean five-year VAS scores at squatting were both lower than the preoperative value (Fig. 3b). All of the nine patients had MRI taken at 12 months. One of them disagreed to take further image study, whereas of the other eight patients, one missed the three-year MRI and another one missed the five-year MRI. The mean scores in every, International Orthopaedics (SICOT) (2020) 44:1745–1754 1749 Fig. 3 Clinical outcome. a Mean KOOS for subscales (see text for abbreviations). The asterisks and the pond signs indicate a value significantly higher and lower, respectively, than the pre-operative. The double-cross (‡) signs indicate comparable KOOS scores of the same subscale among various post-operative times. b Mean VAS score. Asterisks indicate a significantly lower value except the “effusion,” subscale of the MOCART chart were comparable among times throughout the follow-up period, i.e., the post-operative 12-month, three-year, and five-year (Table 1). The post hoc analysis of the “effusion” score, however, did not identify a significant pairwise difference among times (p = 0.063 for 12-month versus one-year or three-year and p = 1.000 for three-year versus five-year). All of the eight patients agreeing image studies beyond 12 months took the CT scan at three years, whereas one missed the scan at five years (Table 2). None of the implants had the implant-bone interfaces completely disappeared at the end of five years; more than 50% total contour of the interface remaining visible in more than half of the implantations. However, there was a trend that the visible interface decreased with time (p = 0.083). Similarly, none of the implants had the content in the osseous phase completely homogenized with that of the host subchondral bone in five years. The higher radiopacity made the implants 1750 Table 1 International Orthopaedics (SICOT) (2020) 44:1745–1754 MRI evaluation at one, thee, and five years post-operatively by MOCART scale Number of patients 1 year 3 years 3 4 1 4 3 1 Incomplete, < 50% thickness of adjacent cartilage Subchondral bone exposed 2. Integration to border zone Complete Incomplete with split-like border Incomplete, visible defect < 50% length of repair tissue 0 0 0 0 7 1 0 6 2 0 Incomplete, visible defect > 50% length of repair tissue 3. Surface of repair tissue Intact Damaged < 50% depth of repair tissue Damaged > 50% depth of repair tissue/total degeneration 4. Structure of repair tissue 0 0 8 0 0 7 1 0 5 3 1. Degree of defect repair/filling of defect Complete Hypertrophy Incomplete, > 50% thickness of adjacent cartilage Homogeneous Inhomogeneous or cleft formation 5. Signal intensity of repair tissue on dual T2-FSE sequence Isointense Moderately hyperintense Markedly hyperintense 6. Subchondral lamina Intact Not intact 7. Subchondral bone Intact Not intact 8. Adhesion Yes No 9. Effusion Yes No 1 year 5 years 4 4 0 4 2 1 0 0 1 0 7 1 0 6 1 1 0 0 1.000 8 0 0 6 1 1 0.500 6 2 1.000 5 3 7 1 0.625 5 3 0 5 3 0 0.250 5 3 0 5 3 0 1.000 7 1 5 3 0.500 7 1 3 5 0.125 6 2 8 0 0.500 6 2 8 0 0.500 0 8 0 8 1.000 0 8 0 8 1.000 4 4 8 0 0.063 4 4 8 0 0.063 visually distinguishable from the native cancellous bone at five years, although the high-radiopaque area tended to shrink with time. None of the eight implants at three years and only two at five years had more than 40% total osseous phase depict a homogenous radiopacity with the surrounding bone, indicating that cancellous bone with normal mineralization was hardly rebuilt by the implants (Fig. 4). Histological examinations of the regenerate tissue at 12 months post-operatively revealed regenerated cartilaginous tissue in the chondral phase without residual biomaterial. The regenerate tissue was hyaline in nature as indicated by the positive staining with Alcian blue and with immunohistological staining for collagen type II. We have p 1.000 1.000 P 1.000 1.000 previously demonstrated example histological pictures of the regenerate tissue [14], whereas another example is illustrated in Fig. 5. Discussion The primary purpose of this extension of the original feasibility study was to affirm the mid-term safety of this biphasic construct. Minor adverse effects such as knee swelling or soreness were recorded at post-operative six weeks and disappeared at three-month follow-up without treatment [14]. No recurrence of these adverse effects, nor any serious adverse effect had been recorded during the five-year period. We could International Orthopaedics (SICOT) (2020) 44:1745–1754 Table 2 1751 CT evaluation at 3 years and 5 years Number of patients Post-operative time Visible implant-bone interface Invisible < 50% total boundary > 50% total boundary Visible throughout the boundary Homogenization of implant with host bone Complete > 70% 40~69% < 40% None Fig. 4 Image evaluation. a oneyear MRI. b three-year MRI. c five-year MRI (sagittal acquisition of T2-FSE sequence). Neocartilage (indicated by the triangle arrows) integrates well to neighbouring native cartilage with similar thickness. The subchondral area (indicated by arrows) maintained without oedema, granulation tissue, cysts, or sclerosis. d three-year CT. e fiveyear CT showed incomplete subchondral bone integration 3 years 5 years 0 0 3 5 0 3 2 2 0 0 0 7 1 0 1 1 4 1 therefore affirm the mid-term safety of the repair surgery. The implants in this study contain PLA, which is degraded to lactic acid by hydrolysis. Although there is inherent inflammation in any degradation process, the severity is associated with the amount of the monomer released at one time. For example, polyglycolide (PGA), a rapidly degrading material, leads to marked synovial inflammation due to its large release of acidic breakdown products [1, 22]. In contrast, slowly degrading lactide-containing implants, such as PLLA or PLGA/β-TCP, do not cause clinically significant synovitis [17]. This may be the reason why most of the patients had mild radiographic effusion but all of them were asymptomatic in the present study. The secondary outcome measure was the durability about the improvement of knee function post-operatively. The subjective outcome had remained constant since post-operative one year despite that the osseous phase did not completely homogenize with the surrounding native bone. All our patients has had osteochondral lesions, in which the integrity of subchondral bone is important to maintain the overlying 1752 International Orthopaedics (SICOT) (2020) 44:1745–1754 Fig. 5 a A 49-year-old man had had an osteochondral lesion at the lateral femoral condyle in his right knee. b Second-look arthroscopy at 12 months post-operatively showed full-filling of the original defect with regenerate tissue. Tissue biopsy with a trephine was done at centre of the regenerate tissue. (Inlay) An osteochondral tissue sample, 1 mm diameter and 10 mm long, was harvested. Asterisk indicates the cartilaginous part. c The low power and d high power histological pictures of the cartilaginous part of the regenerate tissue after haematoxylin and eosin staining show lacunated cells (arrow) scattering among a homogenous ground matrix articular cartilage [26]. When the joint bears weight, insufficient support from abnormal subchondral bone raises strain within the cartilage, promotes degradation of cartilaginous matrix, and subsequently causes failure of the regenerated cartilage. The clinical experience using TruFit™ Plug [7, 20, 30, 41, 49] indicated that the plugs filled the cartilaginous defects well and incorporate with the adjacent native cartilage in the short term [7, 20, 30]; however, the regenerated cartilage failed to grow into the subchondral bone and subsequently deteriorated [30, 49]. The osseous part of the TruFit™ Plug is made of calcium sulfate and PGA. Both two substances have been widely used in biodegradable orthopaedic implants [23, 31, 50]; they degrade quickly after implantation with the former dissolving in weeks [31, 33] and the latter decomposing in few months [17, 43]. Such material property makes the implants lose their mechanical strength rapidly. In the cases of the osteochondral implants, the quick loss of osseous support in the construct would negatively affect the survival of regenerated cartilage at the chondral phase. In order to maintain a longer support for the regenerated cartilage, we made the osseous phase of the osteochondral implant with a PLGA/β-TCP composite which had been reported to have a controlled rate of absorption, few unfavourable effects, and good osteoconductivity [3, 4, 40]. Interferential screws made of such composite induced joint effusion in only 5% patients and bone cysts around implants in 4% [5]. We observed “intact subchondral bone” while making scores on the MOCART chart in all samples at both three-year and fiveyear postoperation (Fig. 4a–cc). Such long-lasting integrity of subchondral bone may contribute to the maintenance of regenerated cartilage. At post-operative five years, the chondral defects were complete-filled in 62.5% of patients while the thickness of regenerated cartilage was more than half thickness of the adjacent native cartilage in 87.5% of patients. The score in every respective subscale except “effusion” of the MOCART chart held constant from one- to five-year post-operation. The MRI-based MOCART chart has high sensitivity detecting oedema, granulation tissue, cysts, and sclerosis at the subchondral bone; however, it does not evaluate the integration of the implants with the adjacent native bone. Contrarily, CT has good resolution about the bone and has been preferred for the evaluation of the subchondral bone growth after osteochondral repair [2, 16]. In the present study, seven patients had CT images at five-year post-operation and none had the implants fuse completely with the host bone (Fig. 4d, e), whereas five of these patients had less than 40% fusion. Besides, normal cancellous bone did not regenerate well at the osseous phase of the implants in any of these images. Joint-function improvement after osteochondral repair surgery with various osteochondral scaffolds has been reported in the literature [16, 30, 41]. However, such initial improvement usually declines in as short as one or two years presumably because the symptoms alleviate or disappear by the removal of the pathological bone instead of the regeneration of healthy bone [16]. Therefore, the post-operative improvement degrades with the decomposing implants without the unsuccessful regeneration of subchondral bone or cartilage, or both [30, 47, 49]. We observed the same immediately post-operative and short-term improvement about KOOS and VAS score among the present cohort of patients. The maintenance of such improvement to post-operative five years in the present study might be attributed to the successful regeneration of cartilage that was supported by the sustained construction at the osseous phase in the implants, primarily from the delayed resorption of the TCP content. On the CT images, the osseous phase of the implants constantly depicted a radiopacity higher than that of the surrounding host bone, indicating a higher mineral International Orthopaedics (SICOT) (2020) 44:1745–1754 content in the implants. Whether this mineral content is longexisting, resolves eventually, or successfully be replaced with normal cancellous bone will need a longer observation. The KOOS has been validated to define severe knee disabilities; the QoL is the most specific among the five subscales to evaluate the problems caused by focal chondral lesions [8]. We also found that QoL score was the most-improved postoperatively among all subscales, and such superiority was noted at all times of follow-up. Overall, the scores in each specific subscale of KOOS among all post-operative times were comparable with one another. Mean VAS scores at chair rises and squatting remained significantly lower than the preoperative value until five years. The sustained post-operative improvement both subjectively and objectively for as long as five years indicated that this osteochondral construct would be a feasible regenerative technique for the repair of osteochondral lesions. Further follow-up is required to find its long-term durability. There were several limitations in this study. First, the case number was small because this was a clinical feasibility study for a novel device. Type II error may occur during the statistical exercise. For example, the “effusion” score in MOCART chart had a trend to increase in the post hoc analysis when comparing the 12-month with three-year or five-year results, although all the effusions were mild and asymptomatic. Second, the missing data lead to selection bias, although the follow-up rate was as high as 90% for function scores. Third, the evaluations for the regenerated cartilage at three- and five-year were based on MRI observation without a tissue proof. The only histological data of the regenerated cartilage were collected at one year in the secondlook arthroscopic biopsy; we had previously reported the tissue samples as a hyaline cartilage containing viable cells distributing in a mixed columnar-cluster pattern [14]. Whether such histology was durable with time was not verified in this study, because the ethical concern about the repetitive surgical harvest of tissue samples. We concluded that the biphasic cylindrical construct with the double-minced autologous cartilage would replace the osteochondral defects with repair cartilage that remained for as long as five years and parallelly provided improvement about knee function. The subchondral bone kept intact to support the regenerated cartilage during this period. This novel construct might be a feasible treatment for the osteochondral lesions in the knee joints. Funding information This study was funded by Exactech Taiwan Co., Ltd. 1753 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Compliance with ethical standards 16. Conflict of interest Author, Chun Nan Chen, is an employee of BioGend Therapeutics Co., Ltd.; the corporation acquired Exactech Taiwan in 2018. Athanasiou KA, Agrawal CM, Barber FA, Burkhart SS (1998) Orthopaedic applications for PLA-PGA biodegradable polymers. Arthroscopy 14:726–737 Barber FA, Dockery WD (2011) A computed tomography scan assessment of synthetic multiphase polymer scaffolds used for osteochondral defect repair. Arthroscopy 27:60–64 Barber FA, Dockery WD, Cowden CH 3rd (2013) The degradation outcome of biocomposite suture anchors made from poly L-lactideco-glycolide and beta-tricalcium phosphate. Arthroscopy 29:1834– 1839 Barber FA, Dockery WD, Hrnack SA (2011) Long-term degradation of a poly-lactide co-glycolide/beta-tricalcium phosphate biocomposite interference screw. Arthroscopy 27:637–643 Barber FA, Spenciner DB, Bhattacharyya S, Miller LE (2017) Biocomposite implants composed of poly(lactide-co-glycolide)/beta-tricalcium phosphate: systematic review of imaging, complication, and performance outcomes. Arthroscopy 33:683–689 Bartlett W, Skinner JA, Gooding CR, Carrington RW, Flanagan AM, Briggs TW et al (2005) Autologous chondrocyte implantation versus matrix-induced autologous chondrocyte implantation for osteochondral defects of the knee: a prospective, randomised study. J Bone Joint Surg (Br) 87:640–645 Bekkers JE, Bartels LW, Vincken KL, Dhert WJ, Creemers LB, Saris DB (2013) Articular cartilage evaluation after TruFit plug implantation analyzed by delayed gadolinium-enhanced MRI of cartilage (dGEMRIC). Am J Sports Med 41:1290–1295 Bekkers JE, de Windt TS, Raijmakers NJ, Dhert WJ, Saris DB (2009) Validation of the knee injury and osteoarthritis outcome score (KOOS) for the treatment of focal cartilage lesions. Osteoarthr Cartil 17:1434–1439 Berruto M, Delcogliano M, de Caro F, Carimati G, Uboldi F, Ferrua P et al (2014) Treatment of large knee osteochondral lesions with a biomimetic scaffold: results of a multicenter study of 49 patients at 2-year follow-up. Am J Sports Med 42:1607–1617 Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L (1994) Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 331: 889–895 Brix M, Kaipel M, Kellner R, Schreiner M, Apprich S, Boszotta H et al (2016) Successful osteoconduction but limited cartilage tissue quality following osteochondral repair by a cell-free multilayered nano-composite scaffold at the knee. Int Orthop 40:625–632 Chaipinyo K, Oakes BW, Van Damme MP (2004) The use of debrided human articular cartilage for autologous chondrocyte implantation: maintenance of chondrocyte differentiation and proliferation in type I collagen gels. J Orthop Res 22:446–455 Chiang H, Kuo TF, Tsai CC, Lin MC, She BR, Huang YY et al (2005) Repair of porcine articular cartilage defect with autologous chondrocyte transplantation. J Orthop Res 23:584–593 Chiang H, Liao CJ, Hsieh CH, Shen CY, Huang YY, Jiang CC (2013) Clinical feasibility of a novel biphasic osteochondral composite for matrix-associated autologous chondrocyte implantation. Osteoarthr Cartil 21:589–598 Chiang H, Liao CJ, Wang YH, Huang HY, Chen CN, Hsieh CH et al (2010) Comparison of articular cartilage repair by autologous chondrocytes with and without in vitro cultivation. Tissue Eng Part C Methods 16:291–300 Christensen BB, Foldager CB, Jensen J, Jensen NC, Lind M (2016) Poor osteochondral repair by a biomimetic collagen scaffold: 1- to 3-year clinical and radiological follow-up. Knee Surg Sports Traumatol Arthrosc 24:2380–2387 1754 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. International Orthopaedics (SICOT) (2020) 44:1745–1754 Cobaleda Aristizabal AF, Sanders EJ, Barber FA (2014) Adverse events associated with biodegradable lactide-containing suture anchors. Arthroscopy 30:555–560 Cole BJ, Farr J, Winalski CS, Hosea T, Richmond J, Mandelbaum B et al (2011) Outcomes after a single-stage procedure for cellbased cartilage repair: a prospective clinical safety trial with 2year follow-up. Am J Sports Med 39:1170–1179 Delcogliano M, Menghi A, Placella G, Speziali A, Cerulli G, Carimati G, et al. (2014) Treatment of osteochondritis dissecans of the knee with a biomimetic scaffold. A prospective multicenter study. Joints 2:102–108 Dhollander AA, Liekens K, Almqvist KF, Verdonk R, Lambrecht S, Elewaut D et al (2012) A pilot study of the use of an osteochondral scaffold plug for cartilage repair in the knee and how to deal with early clinical failures. Arthroscopy 28:225–233 Dipaola JD, Nelson DW, Colville MR (1991) Characterizing osteochondral lesions by magnetic resonance imaging. Arthroscopy 7:101–104 Edwards DJ, Hoy G, Saies AD, Hayes MG (1994) Adverse reactions to an absorbable shoulder fixation device. J Shoulder Elb Surg 3:230–233 Faour O, Dimitriou R, Cousins CA, Giannoudis PV (2011) The use of bone graft substitutes in large cancellous voids: any specific needs? Injury 42(Suppl 2):S87–S90 Filardo G, Kon E, Di Martino A, Busacca M, Altadonna G, Marcacci M (2013) Treatment of knee osteochondritis dissecans with a cell-free biomimetic osteochondral scaffold: clinical and imaging evaluation at 2-year follow-up. Am J Sports Med 41: 1786–1793 Filardo G, Kon E, Roffi A, Di Martino A, Marcacci M (2013) Scaffold-based repair for cartilage healing: a systematic review and technical note. Arthroscopy 29:174–186 Gomoll AH, Madry H, Knutsen G, van Dijk N, Seil R, Brittberg M et al (2010) The subchondral bone in articular cartilage repair: current problems in the surgical management. Knee Surg Sports Traumatol Arthrosc 18:434–447 Gooding CR, Bartlett W, Bentley G, Skinner JA, Carrington R, Flanagan A (2006) A prospective, randomised study comparing two techniques of autologous chondrocyte implantation for osteochondral defects in the knee: periosteum covered versus type I/III collagen covered. Knee 13:203–210 Guhl JF (1982) Arthroscopic treatment of osteochondritis dissecans. Clin Orthop Relat Res:65–74 Heir S, Nerhus TK, Rotterud JH, Loken S, Ekeland A, Engebretsen L et al (2010) Focal cartilage defects in the knee impair quality of life as much as severe osteoarthritis: a comparison of knee injury and osteoarthritis outcome score in 4 patient categories scheduled for knee surgery. Am J Sports Med 38:231–237 Joshi N, Reverte-Vinaixa M, Diaz-Ferreiro EW, DominguezOronoz R (2012) Synthetic resorbable scaffolds for the treatment of isolated patellofemoral cartilage defects in young patients: magnetic resonance imaging and clinical evaluation. Am J Sports Med 40:1289–1295 Kelly CM, Wilkins RM, Gitelis S, Hartjen C, Watson JT, Kim PT (2001) The use of a surgical grade calcium sulfate as a bone graft substitute: results of a multicenter trial. Clin Orthop Relat Res:42– 50 Kon E, Delcogliano M, Filardo G, Pressato D, Busacca M, Grigolo B et al (2010) A novel nano-composite multi-layered biomaterial for treatment of osteochondral lesions: technique note and an early stability pilot clinical trial. Injury 41:693–701 Kumar CY, Nalini KB, Menon J, Patro DK, Banerji BH (2013) Calcium sulfate as bone graft substitute in the treatment of osseous bone defects, a prospective study. J Clin Diagn Res 7:2926–2928 Liao CJ, Chen CF, Chen JH, Chiang SF, Lin YJ, Chang KY (2002) Fabrication of porous biodegradable polymer scaffolds using a 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. solvent merging/particulate leaching method. J Biomed Mater Res 59:676–681 Liao CJ, Lin YJ, Chiang H, Chiang SF, Wang YH, Jiang CC (2007) Injecting partially digested cartilage fragments into a biphasic scaffold to generate osteochondral composites in a nude mice model. J Biomed Mater Res A 81:567–577 Lu Y, Dhanaraj S, Wang Z, Bradley DM, Bowman SM, Cole BJ et al (2006) Minced cartilage without cell culture serves as an effective intraoperative cell source for cartilage repair. J Orthop Res 24:1261–1270 Marlovits S, Singer P, Zeller P, Mandl I, Haller J, Trattnig S (2006) Magnetic resonance observation of cartilage repair tissue (MOCART) for the evaluation of autologous chondrocyte transplantation: determination of interobserver variability and correlation to clinical outcome after 2 years. Eur J Radiol 57:16–23 Marlovits S, Striessnig G, Resinger CT, Aldrian SM, Vecsei V, Imhof H et al (2004) Definition of pertinent parameters for the evaluation of articular cartilage repair tissue with high-resolution magnetic resonance imaging. Eur J Radiol 52:310–319 McCormick F, Yanke A, Provencher MT, Cole BJ (2008) Minced articular cartilage—basic science, surgical technique, and clinical application. Sports Med Arthrosc Rev 16:217–220 Ntagiopoulos PG, Demey G, Tavernier T, Dejour D (2015) Comparison of resorption and remodeling of bioabsorbable interference screws in anterior cruciate ligament reconstruction. Int Orthop 39:697–706 Pearce CJ, Gartner LE, Mitchell A, Calder JD (2012) Synthetic osteochondral grafting of ankle osteochondral lesions. Foot Ankle Surg 18:114–118 Peterson L, Vasiliadis HS, Brittberg M, Lindahl A (2010) Autologous chondrocyte implantation: a long-term follow-up. Am J Sports Med 38:1117–1124 Sheikh Z, Najeeb S, Khurshid Z, Verma V, Rashid H, Glogauer M (2015) Biodegradable materials for bone repair and tissue engineering applications. Materials (Basel) 8:5744–5794 Sohn DH, Lottman LM, Lum LY, Kim SG, Pedowitz RA, Coutts RD et al (2002) Effect of gravity on localization of chondrocytes implanted in cartilage defects. Clin Orthop Relat Res:254–262 Steinhagen J, Bruns J, Deuretzbacher G, Ruether W, Fuerst M, Niggemeyer O (2010) Treatment of osteochondritis dissecans of the femoral condyle with autologous bone grafts and matrixsupported autologous chondrocytes. Int Orthop 34:819–825 Strauss EJ, Goodrich LR, Chen CT, Hidaka C, Nixon AJ (2005) Biochemical and biomechanical properties of lesion and adjacent articular cartilage after chondral defect repair in an equine model. Am J Sports Med 33:1647–1653 Tamaddon M, Liu C (2018) Enhancing biological and biomechanical fixation of osteochondral scaffold: a grand challenge. Adv Exp Med Biol 1059:255–298 Vasiliadis HS, Danielson B, Ljungberg M, McKeon B, Lindahl A, Peterson L (2010) Autologous chondrocyte implantation in cartilage lesions of the knee: long-term evaluation with magnetic resonance imaging and delayed gadolinium-enhanced magnetic resonance imaging technique. Am J Sports Med 38:943–949 Verhaegen J, Clockaerts S, Van Osch GJ, Somville J, Verdonk P, Mertens P (2015) TruFit plug for repair of osteochondral defects— where is the evidence? Systematic review of literature. Cartilage 6: 12–19 Walsh WR, Morberg P, Yu Y, Yang JL, Haggard W, Sheath PC et al (2003) Response of a calcium sulfate bone graft substitute in a confined cancellous defect. Clin Orthop Relat Res. https://doi.org/ 10.1097/01.blo.0000030062.92399.6a228-236 Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.