Biodegradation of 3D-Printed PCL/β-TCP Scaffolds Versus Collagen Membranes for Bone Augmentation: In Vitro Thickness and Weight Loss Over Six Months

Purpose: This controlled in vitro repeated-measures study compared the biodegradation behavior of a commercially available collagen membrane with a three-dimensionally printed polycaprolactone scaffold containing 25% beta-tricalcium phosphate (PCL/β-TCP). The primary outcome was change in specimen thickness over six months; secondary outcomes included weight change, time to structural failure, and surface morphology. Methods: Seventy-two 10 x 10 mm specimens were randomized by material and degradation medium into phosphate-buffered saline, simulated body fluid (SBF), 0.25% trypsin, or SBF supplemented with collagenase (n = 9 per subgroup). Specimens were incubated at 37°C and evaluated from 3 hours to 6 months. Thickness and weight were measured using standardized positioning and calibrated instruments. Scanning electron microscopy was used to characterize surface changes. Linear mixed-effects modeling, Bonferroni-adjusted post hoc testing, effect-size reporting, and Kaplan-Meier survival analysis were used. Because this was a laboratory in vitro study with no human participants or animals, clinical trial registration was not applicable. Results: Collagen membranes showed rapid reductions in thickness and weight, especially in enzymatic media, with complete structural loss in several subgroups within weeks. PCL/β-TCP scaffolds degraded more slowly and persisted longer structurally. At the final evaluation, material type significantly affected weight (F = 22.95, p < 0.001, partial ηp2 = 0.264) and thickness (F = 100.78, p < 0.001, partial ηp2 = 0.611). Significant material-by-solution interactions were also observed for weight (F = 15.59, p < 0.001, partial ηp2 = 0.422) and thickness (F = 15.33, p < 0.001, partial ηp2 = 0.418). Kaplan-Meier analysis showed shorter median survival for collagen than for PCL/β-TCP in most media, with the most rapid failure in collagen exposed to trypsin. The 0.25% trypsin findings represent an accelerated, nonphysiological challenge and do not predict in vivo degradation timelines. Conclusions: Within the limitations of this static in vitro model, collagen degraded rapidly across media, whereas PCL/β-TCP demonstrated prolonged structural stability and surface changes consistent with slower, surface-mediated degradation. These findings support further biologic, mechanical, and clinical evaluation of PCL/β-TCP scaffolds for guided bone regeneration procedures requiring extended space maintenance.