Participant’s characteristics
Because one of the participating surgeons had a long-term illness, 12 participants received no MI in the study period (Fig 3). The recruitment period ranged between January 2014 and July 2015. The study ended in July 2018. The new or optimized RPDs of the 79 study jaws (31 maxillae) were either double crown-retained (n= 52), retained by double crowns and clasps (n= 15), clasp-retained (n= 10), or precision attachment-retained (n=2) which were worn for at least two months. The median of the residual teeth was two in the maxilla (1st quartile = 2; 3rd quartile = 4) and three in the mandible (1st quartile = 2; 3rd quartile = 5). Participants with study maxillae showed on the opposing jaw: RPDs (n=19), fixed dental prostheses (n=9) or natural dentition (n=3). Participants with study mandibles showed on the opposing jaw: complete dentures (n=10), RPDs (n=30) or fixed dental prostheses (n=8). A total of 112 maxillary (median 3 per jaw) and 120 mandibular (median 2 per jaw) MIs were placed. Three participants received MIs in the maxilla and mandible. As seen in the flow chart (Fig 3), each 38 participants were allocated to group A (mean age 66.4 years, 22 women) and B (mean age 65.4 years, 25 women). In group A, 12 maxillary and 22 mandibular RPDs were primarily soft relined because the insertion torque of at least one MI was < 35 Ncm.
The MIs of three maxillae and three mandibles were immediately loaded with housings (insertion torque ≥ 35 Ncm). The MIs of 16 maxillae and 23 mandibles in group B were delayed loaded. A total of six participants in group A (16%) and seven participants in group B (18%) were lost to follow-up in the whole study period, among them each two per group until the second year. Figure 4 shows the distribution by tooth site and MI diameter. In the maxilla, MIs with a diameter of 2.4 mm were predominantly placed whereas in the mandible the diameter of 2.1mm and 1.8mm predominated.
Mini-implants Stability
The median of maximum insertion torque of 222 MI was 25 Ncm (1st quartile: 20 Ncm, 3rd quartile 35 Ncm). The different number of stability measurements and the higher number of PTVs compared with RFA values (total number 1270 versus 897), especially after the first year, has several reasons (Table 1). First, the response rate for visits of the participants varied between follow-up points. Second, just after surgery (T0), stability values were available in the university hospital exclusively. Third, if the mucosa covered the MI insertion square a safe fixation of the SmartPeg and subsequent measurement was sometimes impossible. Fourth, during the second year of the study, the manufacturer transferred the fabrication of the MI system to another company and supplies of SmartPegs were stopped.
Table 1 Median and quartiles for all mini-implant stability measurements at insertion and follow-ups
|
|
|
Periotest
|
|
Resonance frequency analysis
|
Time point
|
|
|
Na
|
1th Qb
|
Median
|
3rd Q
|
|
N
|
1st Q
|
Median
|
3rd Q
|
Insertion
|
t0
|
|
72
|
0.6
|
1.6
|
3.6
|
|
77
|
41
|
44
|
49
|
2 weeks
|
t1
|
|
222
|
1.7
|
3.6
|
6.3
|
|
217
|
32
|
39
|
43
|
4 months
|
t2
|
|
205
|
2.8
|
4.5
|
7.8
|
|
203
|
30
|
37
|
43
|
4.5 months
|
t3
|
|
193
|
2.6
|
4.5
|
7.3
|
|
195
|
32
|
39
|
43
|
1 year
|
t4
|
|
209
|
2.0
|
3.5
|
6.2
|
|
161
|
35
|
41
|
44
|
2 years
|
t5
|
|
196
|
2.0
|
3.5
|
6.6
|
|
36
|
39
|
43
|
46
|
3 years
|
t6
|
|
173
|
1.8
|
3.4
|
5.6
|
|
|
|
|
|
Total
|
|
|
1270
|
2.0
|
3.6
|
6.6
|
|
889
|
32
|
39
|
44
|
anumber of measurements
bquartile
According to table 1, primary PTV and RFA values from the university hospital suggest the highest MI stability at the day of placement.
The adjusted values and 95% CIs of the follow-ups by groups in figures 5 and 6 correspond with women, age of 65 years, mandible, first premolar site, and first study center. Assumed a tolerance limit of +1 for PTVs, the immediately loaded MI of group A showed no inferiority compared with delayed loaded MIs of group B (Fig 5), and the differences between groups are statistically not significant (P=0.788). The 95% CI after three years is compatible with effects between -1.4 (in behalf of group B) and 1.9 (in behalf of A). Similar results were obtained for the RFA values (Fig 6). The 95% CI after 2 years is compatible with effects between -5.2 (in behalf of group B) and 2.0 (in behalf of A) without a statistically significant difference (P= 0.390). However, in the first year of observation PTVs tended to be higher and RFA values tended to be lower with wide overlapping 95% CIs for the immediately loaded MIs of group A compared to group B. The values of the groups converged after the first year.
The Spearman correlation of rs = -0.83 between PTV und RFA values (n = 887) suggest a high negative correlation on condition of independent observations. The negative correlation of the maximum insertion torque with PTVs was weak (rs = -0.15), whereas the positive correlation with RFA values was moderate (rs = 0.24) as seen in figure 7. Figure 8 shows the PTVs by MI diameter and length in the maxilla and mandible at the 1-year follow-up.
The diameters of 1.8 and 2.1 mm are pooled because of the low number of the thinnest MIs. The stability of 2.4 mm MIs was higher (mean PTV for all lengths together: 4.1 in the maxilla and -0.3 in the mandible) than the stability of MIs with lower diameters (mean PTV for all lengths together: 9.8 in the maxilla and 4.7 in the mandible). The differences between MI lengths were negligible.
Mini-implants Survival Rate
A total of 8 failures were registered in three participants of group A versus 5 failures in two participants of group B, which resulted in cumulative 3-year survival rates of 92% versus 95% (Fig 9). In group
A, one maxillary MI was misplaced and could not be covered by the RPD because of its buccal malposition. This MI was immediately removed. All other MI failed due to lost osseointegration. Each three losses in three maxillae, two losses in one other maxilla, the malpositioned MI and another loss in one mandible yield MI survival rates of 87% for maxillary and 99% for mandibular MIs (Fig 10).
According to Cox regression analyses considering dependent observations within one person, the 95% CI for the difference between the groups was 0.1 to 3.4 (Hazard ratio [HR]= 0.6, P=0.554). The 95% CI for the difference between the jaws was 1.6 - 142.2 (HR= 15.3; P= 0.017). Participants with MI losses wore either RPDs with multiple abutments (n=3) or had natural dentitions (n=2) in the opposing jaw. One patient with three MI losses was a current heavy smoker, the other patients with MI losses were never smokers. The lost mandibular MI was placed in a fresh extraction socket contrary to all other MIs. MI failures were either spontaneous losses during removing the RPD at home (n=6) or MI loosenings (n=6) with subsequent removal without any severe inflammatory reaction of the peri-implant tissue or pronounced bone defects.
Periotest Values and Implant Survival
The majority of PTVs for failed MIs (8 out of 12) showed an increase between two weeks after placement (T1) and the follow-up before the loss (Table 2). According to Cox regression analyses, plus one unit PTV increases the failure risk (HR=1.23; 95% CI=1.15-1.31; P< 0.001). An increase of 5, 10 and 15 units PTV result in HRs of 1.8, 7.9, and 22.0. Transformed in curves, the 3-year survival rate probability decreased from 97% in a PTV difference of 5 to 92% in a difference of 10 units (Fig 11).
Table 2 Periotest values of failed mini-implants and the duration up to loss
Noa
|
Group
|
FDIb tooth site
|
Periotest values at
|
Loss after
|
2 weeks
|
4 Moc
|
4.5 Mo
|
1 yrd
|
2 yrs
|
1
|
B
|
22
|
0.3
|
27.5
|
|
|
|
4 Mo
|
23
|
1
|
4.6
|
6.3
|
13
|
24
|
32 Mo
|
2
|
A
|
13
|
-0.8
|
3.1
|
4.4
|
9.2
|
|
18 Mo
|
23
|
1.9
|
25
|
|
|
|
4 Mo
|
24
|
5.9
|
10.7
|
12.5
|
19
|
|
17 Mo
|
3
|
A
|
21
|
2.5
|
5.4
|
5
|
3.6
|
6.4
|
35 Mo
|
22
|
3.8
|
6
|
7.2
|
9.2
|
28
|
28 Mo
|
23
|
5.5
|
24.5
|
23
|
22
|
|
17 Mo
|
4
|
A
|
42
|
1
|
|
|
|
|
1 Mo
|
5
|
B
|
11
|
32
|
28
|
32
|
|
|
5 Mo
|
12
|
12.5
|
11
|
-
|
8.2
|
|
24 Mo
|
13
|
5.3
|
7.3
|
-
|
12.5
|
|
24 Mo
|
aParticipant number with mini-implant loss
bFederation Dentaire Internationale tooth numbering system
cmonths
dyear