A total of 92 patients (38 men, 54 women; mean age 54.3 years, range 20–78) received implant-supported prostheses (at the seven private practices) between 1984 and 1990. The distribution of patients by age and gender is presented in Table 1. Fifty-seven patients (140 implants) were considered dropouts due to the fact that no data were obtained at the endpoint, but 25 years had passed since the prosthetic treatment delivery of one dropout patient (with four implants). The Kaplan-Meier estimate shows the censorings (Fig. 1), and they present unbiased throughout the observation period. The dropout reasons for 42 patients were illness, moved away, and not showing up for check-ups; another 15 patients had passed away before the present analysis.

A total of 223 implants were placed in 15 fully edentulous patients and 77 partially edentulous patients. The distributions of implants by diameter, length, and location are presented in Tables 2, 3, and 4. Twenty-four implants were placed in the maxilla (10.8%), and 199 implants were placed in the mandible (89.2%). Only two implants were applied to the maxillary anterior region, whereas 152 implants (68.2%) were applied to the mandibular posterior region (Tables 2 and 3). Regarding the sizes of the implants, 4.1-mm dia. and 10-mm length were the most frequently used implant dimensions (70.9 and 39%, respectively).

Four implants of one patient were inserted into the re-constructed mandible with iliac bone, due to an ameloblastoma. Additional soft tissue managements were applied to 96 implants. Free gingival graft was used for 86 implants, and frenectomy and vestibular extension were applied to 15 and 13 implants, respectively.

Sixteen implants were lost during the observation period. The Kaplan-Meier cumulative survival rates were 97.4, 95.4, and 89.8% at 10, 15, and 25 years after the prosthesis setting, respectively (Fig. 1). After stepwise backward selection, implant position in the mandibular vs. the maxilla showed the significant difference in the cumulative survival rate (Table 5, Fig. 2c). The gender, implant type, additional soft tissue management, and width of keratinized mucosa did not provide significant differences with respect to the survival of the evaluated implants in this study (Fig. 2a, b, d, and e). The reasons for late failure were peri-implant infection (14 implants) and unknown (two implants).