Comparison of a polyethylene glycol membrane and a collagen membrane for the treatment
of bone dehiscence defects at bone level implants—A prospective, randomized, controlled,
multicenter clinical trial
Abstract Objectives The aim of the present randomized, controlled clinical multicenter
trial was to compare a polyethylene glycol (PEG) and a native collagen membrane (BG)
for simultaneous guided bone regeneration at bony dehiscence-type defects around bone
level titanium implants. Material and methods The study enrolled 117 patients requiring
implant treatment in the posterior maxilla or mandible with expected buccal bony dehiscence-type
defects at the placed titanium implants. According to a parallel groups design, defects
were filled with a synthetic bone filler and randomly assigned to either PEG or BG
membrane. As primary parameter, the relative vertical bone fill was assessed at baseline
and at re-entry after 6 months of healing. As secondary parameters, the marginal bone
level (MBL) was assessed radiographically and soft tissue conditions were recorded
up to 18 months postloading. Results Both groups showed comparable vertical bone fill
revealing a relative change in defect height of 59.7% (PEG) and 64.4% (BG). The absolute
mean reduction in defect size was 2.5 mm in the PEG group and 3.2 mm in the BG group.
Although both groups revealed a statistically significant mean defect reduction (p
< .001), a comparison between the two groups did not show statistical significances.
The non-inferiority test with inferiority limit of ?5% could not be rejected, based
on the 90% confidence interval of the differences of the two means with lower limit
?15.4%. After 18 months, an MBL increase of 0.45 ± 0.43 mm in the PEG group and 0.41
± 0.81 mm in the BG group was detected (p < .001). Soft tissue complications were
observed in both groups without showing statistical significance. Conclusions Both
membranes supported bone regeneration at dehiscence-type defects and obtained vertical
bone fill with a relative change in defect height of 59.7% (PEG) and 64.4% (BG); however,
the non-inferiority of PEG could not be shown.