PURPOSETo document clinical, radiologic, and cellular data of a prospective patient series treated by a tri-layer collagen-hydroxyapatite biomimetic osteochondral scaffold (CHAS) intra-operatively seeded with cultivated autologous chondrocytes (AC) or with filtered bone marrow stem/stromal cells (fBMSC) to address chronic osteochondral knee lesions.METHODSThirty-six consecutive patients (15 to 59 years) with chronic osteochondral lesions (1.8-10 cm2) in the condylar or patellofemoral knee surfaces were enrolled. Lesions were covered with CHAS fixed with a fibrin glue. The superficial layer of CHAS was intra-operatively injected with active cells: in initial five patients, ACs were put directly onto dry CHAS (dry-AC); next, eight AC patients had CHAS moistened with cell culture media (media-AC), while the tourniquet was released allowing blood soaking of CHAS in the rest (14 blood-AC, 9 blood-fBMSC). Seventeen (50%) patients required different concomitant procedures. All patients were followed for serious adverse events (SAE) or graft failures; clinical, radiographic, and MRI evaluation was conducted. Cellular data on the injected cells were assessed.RESULTSAt a follow-up of 39 months (16-81), 17 patients required an additional surgical intervention: seven graft-related SAE (early post-operative synovitis and/or arthrofibrosis) were registered (3 dry-AC, 3 media-AC, 1 blood-fBMSC). There were two graft failures (1 dry-AC, 1 blood-fBMSC) for secondary reasons. All clinical scores significantly improved from pre- to post-operative values: IKCD subjective 44 to 65; IKDC examination (9/17/5/5) to (20/10/5/1); KOOS (P61/S59/ADL67/Sp32/QoL31) to (P79/S75/ADL84/Sp55/QoL51); Tegner activity scale 3.3 to 4.4. There was evidence of radiographic osteoarthritis progression-Kellgren-Lawrence 1.0 to 1.5. MOCART scores at the final follow-up averaged 71 (10 to 95). Graft-type analysis demonstrated an increased rate of graft-related SAE in dry-AC and media-AC, but their final outcomes were equivalent. Cellular data of AC at the implantation were as follows: cells in suspension 9.2 × 106, viability 95%. In blood-fBMSC group, a cell suspension with 87% viability was injected, which contained 1156 CFU-Fs.CONCLUSIONCHAS with intra-operative seeding of active cells, either AC or fBMSC, led to an overall successful outcome for the treatment of chronic osteochondral lesions in the knee. Blood soaking of CHAS in situ before cell seeding significantly decreased early post-operative adverse events, such as synovitis and arthrofibrosis.