Patient Characteristics

The experimental population consisted of 175 patients with one GTR treated defect included. Average age was 44.5 ± 10.2 (range 18 to 76 years), 43% were males, and 32% were current smokers. Before GTR surgery, the baseline full-mouth plaque and bleeding scores (FMPS, FMBS) were 13.9% ± 6.5% and 10.5% ± 3.7%, respectively; 1 year later they were 9.6% ± 3.7% (FMPS range 1.6% to 19%) and 7% ± 3.3% (FMBS range 0% to 17.4%). Defect characteristics at baseline and the 1-year treatment outcomes are shown in Table 1. One year following GTR treatment, an average increase in CAL of 4.6 ± 2 mm was observed; residual PD was 2.8 ± 1 mm.

Table 1. Baseline characteristics of intrabony defects following infection control therapy and 1 year outcomes following regenerative therapy

After the 1-year evaluation, this population was regularly examined every 2 years for a period up to 16 years (average 8 ± 3.4 years). During this period, 66.9% of the 175 patients fully complied with the maintenance program.

Survival Analysis

Tooth loss

Six GTR treated teeth were lost during the follow-up period; all patients were smokers and five did not participate in the periodontal maintenance program.

Table 2. Characteristics of GTR-treated sites associated with tooth loss

Four of the teeth had been treated with non-resorbable expanded polytetrafluoroethylene (ePTFE) membranes and the remaining two with a combination of an absorbable membrane supported by a filler material. The characteristics of the sites that resulted in tooth loss are displayed in Table 2. The conditional probability of tooth loss was low, ranging from 0% to 1.4% for each 2-year period (Table 3). A Cox proportional hazard model was constructed that indicated the significant association of the type of treatment and participation in the periodontal maintenance program with the risk of tooth loss following GTR treatment (Table 4). Since all patients who lost teeth were smokers, the effect of this variable could not be statistically assessed.

Table 3. Life-table analysis for tooth loss

CAL loss ≥2 mm

The first analysis was performed to establish survival with reference to the baseline (pre-GTR) CAL. In this context, an event was defined as having lost ≥2 mm from the original CAL (pre-GTR). No further CAL loss compared to baseline presentation was observed in 92% of cases. The conditional probability of an event was low, ranging from 0.7% to 3.5% for each 2-year period (Table 5). A multiple Cox regression model was fit to analyze the impact of different patient associated variables. Following backward elimination of non-significant factors, the highly significant model indicated that cigarette smoking, participation in a periodontal maintenance program, age at baseline, and the type of regenerative treatment utilized (non-resorbable ePTFE membranes, absorbable membranes, or combination of a barrier membrane with a filler) were significantly associated with the risk of disease progression from baseline (Table 4).

Table 4. Cox proportional hazard models (N = 175)

Table 5. Life-table analysis for attachment loss ≥ 2 mm with respect to baseline CAL

A subsequent analysis focused on loss of CAL compared to the CAL observed at the examination 1 year after GTR treatment. Figure 1 displays the Kaplan-Meier survival curve for the population: 66.2% of the population did not present CAL loss ≥2 mm over the 16-year observation period. The conditional probability of CAL loss is described for each 2-year interval in Table 6.

Table 6. Life-table analysis for attachment loss ≥ 2 mm with respect to clinical attachment measured 1 year after regenerative therapy

Probabilities were similar in the various time periods and ranged from 3.4% to 7.7% for a 2-year interval. A multiple Cox proportional hazard regression analysis indicated that being a smoker significantly increased the probability of losing 2 mm or more of the regenerated CAL, while having been regularly on a periodontal maintenance program decreased that risk (Table 4). Figure 2 displays the survival curves of subjects on periodontal SPC or on community care, while the survival curves of smokers and non-smokers are promised teeth compares favorably with reports of the illustrated in Figure 3.

Subpopulation characterized according to IL-1 genotypes.

Thirty-two of the 86 patients for whom IL-1 genotype status had been determined were positive for the described IL-1 composite genotype. IL-1 positive subjects were followed for an average of 7.5 ± 2.1 years, while the follow-up of the IL-1 negative subjects was 8.7 ± 3.1 years. A proportional hazard model constructed to assess the impact of patient factors in the subpopulation characterized in terms of IL-1 genotypes is presented in Table 7. No significant effect of IL-1 genotype status was observed.

Table 7. Cox Proportional Hazard Model for Subset With IL-1 Gene Polymorphism Data (N = 86)