A retrospective analysis of autotransplanted teeth including an evaluation of a novel surgical technique

For the present study, all consecutive patients who had undergone at least one tooth autotransplantation at the Department for Oral Surgery and Stomatology, University of Bern, between January 2000 and February 2018, were found by thorough electronic and hand search of the patient’s records of the department. Thereafter, they were contacted by phone. Inclusion criterion was a minimal time in function of at least 12 months. Pregnant women and subjects unable to provide informed consent as well as patients with missing or incomplete records were excluded from the study. All patients contacted were provided detailed information about the present study and were invited for a follow-up examination between November 2018 and February 2019. Informed consent was obtained from all participants by the principal investigator (PI; CR) at the beginning of the visit. The study protocol was approved by the standing ethics review board (Ethics Committee of the State of Bern, approval number 2018-01597), and the investigation was performed in accordance with the Declaration of Helsinki (2013).

All patients were asked to provide information about the status of the transplant upon phone inquiry, i.e., if the tooth was still in place, if any additional treatment had been performed so far, and if any symptom was present in rest or function.

The records were screened with regard to the following pre- and perioperative variables: stage of eruption (erupted, partially erupted, impacted), indication for transplantation, donor and recipient site, modification of the root (i.e., root-end resection), use of EMD during surgery, and type and duration of splinting.

Patients who agreed to attend a follow-up examination were evaluated clinically and radiographically by the principal investigator (CR), who was not involved in the treatment of any of the included patients.

The following clinical parameters were assessed at the transplanted tooth and its adjacent teeth: presence of redness, swelling or fistula, periodontal probing depths (in mm) and bleeding on probing (yes/no according to Lang 1986 [27]) at both three buccal and three oral aspects, gingival recession (class I to IV according to [28], sensitivity to CO₂-snow, pain on percussion, mobility (grades 0 to III according to Miller 1950[29], presence of occlusal contacts, and, if applicable, type of restauration.

A periapical radiograph (PA; Soredex Minray, Helsinki, Finland) using a film holder for parallel technique (XCP film holder; Dentsply Sirona, Bensheim, Germany) was taken. The images were assessed using Digora (Soredex, Helsinki, Finland) on a Flexscan monitor (Eizo, Hakusan, Japan) for the following parameters: integrity of the periodontal space, radiolucencies (periapical, intra-, para-, and periradicular) and signs of pulp canal obliteration when compared to the postoperative PA. Root formation stage prior to surgery was classified according to [20] on the preoperative PA. The stages represent the relation of the actual to the estimated final root length of the transplanted tooth. Thus, the root length was classified as R\( \frac{1}{4} \), R\( \frac{2}{4} \), R\( \frac{3}{4}, \) or complete. Teeth with complete root length were additionally allocated according to the status of the apex: open (R\( \frac{4}{4} \)), half-closed (A\( \frac{1}{2} \)), or closed apex (Ac). In the present analysis, teeth with completed root length (R\( \frac{4}{4} \), A\( \frac{1}{2}, \) or Ac) were classified as mature and R\( \frac{1}{4} \)–R\( \frac{3}{4} \) as immature from the perspective of the surgical protocol.

Additionally, cone beam computed tomography (CBCT) using a low-dose protocol (180° rotation at 90 kV, 5 mA, 9 s) and a limited field of view (FOV) of 4 × 4 cm (Accuitomo 170; JMorita MFG Corp., Kyoto, Japan) was performed in patients aged 16 years or older. The CBCT images were assessed for the same parameters as the 2D radiographs using the device-specific image processing software i-Dixel (JMorita MFG Corp.). In order to avoid bias by additional information from the CBCT data, PAs were assessed first. In cases of inconclusive findings, the images were evaluated by a second examiner (SJ) and discussed until consensus was reached.

All transplanted teeth still in situ and not associated with pain upon phone inquiry were considered as surviving. Teeth associated with symptoms or having been removed were classified as failure at this first stage of the analysis.

For teeth undergoing the follow-up examination, the additional category “success” was defined by fulfilling the following criteria (adapted and modified from [30]):

Teeth presenting with one or more questionable clinical and/or radiographic criteria at the follow-up visit were classified as teeth with potentially adverse findings, i.e., allocated to the surviving teeth. Teeth deemed as a failure upon phone inquiry but with the corresponding patient not attending the visit as well as the failures with clinical examination were both considered as failures in the success analysis, assuming that most patients having had their transplant extracted would refuse a visit.

Patient data were first analyzed descriptively. Survival probabilities were estimated using the Kaplan-Meier method. Between-group differences (mature versus immature transplants) were compared using the log-rank test. Potential factors influencing survival were analyzed using a Cox proportional hazards model.

A p value of 0.05 was considered as statistically significant. All statistical analyses were performed using R and RStudio (version 1.2.1335, www.​r-project.​org).

Thirty-five (22 female, 13 male) patients underwent one (32 patients) or two (3 patients) tooth transplantations, resulting in a total of 38 autotransplantations. Three teeth in three patients had missing records and were excluded from the investigation. The median age at the time of surgery was 13 years (range 8–28 years). The most frequent indications were aplasia (n = 18) and failure of orthodontic extrusion of an impacted tooth (n = 14). Six of the donor teeth were fully erupted, seven were partially erupted, and 22 were impacted. Twenty five of the 35 transplants were immature at the time point of surgery (R\( \frac{1}{4} \) (n = 4), R\( \frac{2}{4} \) (n = 10), and R\( \frac{3}{4} \) (n = 11)) and were therefore transplanted without any manipulation of the root. The 10 remaining teeth showed complete root length (R\( \frac{4}{4} \) (n = 2) and Ac (n = 8)) prior to transplantation, and all underwent an intraoperative root-end resection. Thus, it was decided to regard the pairs immature/unmodified as well as mature/root-end resected as synonyms in the further analyses. Relevant patient information and the donor and recipient sites of the transplants are shown in Table 1. Treatment was performed by a total of 8 senior surgeons, i.e., trained and board-certified oral surgeons of the department. The initial postoperative healing was uneventful in all cases.

All of the 32 included patients provided information about the subjective condition of the transplant via phone; i.e., all their 35 transplanted teeth were available for survival analysis. Of these 32 included patients, 27 with 30 transplants (of which one failed) attended the follow-up examination. The two patients who reported a failure upon phone inquiry but were not willing to attend the examination are included in the success analysis (Fig. 1). Figure 2 shows representative clinical and radiographic images at surgery and at follow-up of immature and mature transplants, respectively.

Thirty-two transplants in 30 patients are asymptomatic, and three patients report the loss of one transplant each (Table 1). For all failed transplants, the relevant data until the last follow-up visit was available in the patient records. Two of the 25 unmodified (i.e., immature) teeth were extracted, one after 4.1 years due to inflammatory root resorption (a right maxillary second premolar) and the other after 8.4 years due to combined replacement and invasive cervical root resorption (a left maxillary second molar), respectively. One of the 10 root-end resected (i.e., mature) teeth was extracted after 4.8 years due to combined replacement and invasive cervical root resorption (a left mandibular second premolar). Based on the phone inquiry, an overall survival probability of 91.4% (32 of 35 teeth) after a median observation period of 3.4 years was calculated. A more detailed analysis exhibited a 92% survival probability (23 of 25 teeth) for unmodified teeth after a median observation period of 4.8 years and 90% (9 of 10 teeth) for root-end resected teeth after a median observation period of 2.5 years. The Kaplan-Meier 5-year survival probability estimate for immature transplanted teeth was 92.9% (95% CI: 80.3–100%). For mature transplanted teeth, the corresponding 5-year survival probability estimate is 50% (95% CI: 12.5–100%; Fig. 3). The log-rank test for comparison of the Kaplan-Meier survival estimates was not statistically significant (p = 0.18). The calculated failure rate is approximately three times higher in the mature group as compared to the immature group, but the difference is not statistically significant (p = 0.52; Table 2). In a Cox proportional hazards model, none of the potential predictors (patient’s age, gender, smoking habit, number of roots, stage of root development and eruption, use of EMD, use of antibiotics, type and duration of splinting) was significantly associated with survival probabilities.

Twenty-seven patients with 29 transplanted teeth in function and one failed transplanted tooth (R\( \frac{4}{4} \), root-end resected left mandibular second premolar) attended the follow-up visit (Fig. 1). On average, the 27 patients are seen 3.4 years (1–16.1 years; Table 1) following autotransplantation.

In the cohort of unmodified teeth, two (a left maxillary first premolar with R\( \frac{3}{4} \) and a right mandibular second premolar with R\( \frac{3}{4} \)) required a root canal treatment 4 and 17 months after transplantation. In the cohort of root-end resected teeth, the need for a root canal treatment came up in one case (a right mandibular second molar) 4 months after surgery.

PAs were taken in all patients undergoing the follow-up examination (29/29 transplants). The resulting 2D radiographic parameters are presented in Table 3. Postoperative root growth is present in 10 of 23 unmodified teeth, seven of which seemed to achieve the expected root length (Table 1). Interestingly, three teeth with root formation stage of R\( \frac{1}{4} \) and one tooth with root formation stage of R\( \frac{2}{4} \) show ingrowth of bone-like tissue to the pulp chamber with formation of a periodontal space to the pulpal dentin interface (Fig. 4). Eight of the 29 PAs were deemed inconclusive by the PI regarding root resorptions (n = 8) or width of the periodontal ligament space (n = 1) and were discussed between the examiners to reach an agreement.

Nineteen patients with 20 transplants agreed to undergo an additional CBCT examination. The 3D radiographic parameters for all the surviving transplants are presented in Table 3. Surprisingly, CBCT revealed intraradicular radiolucency at the level of the CEJ in two teeth and an interruption of the periodontal space with replacement resorption in three teeth, which was not visible in PAs. Four of the 20 transplants presented with insufficient radiographic evidence for the buccal or oral alveolar bone walls. Three of these had been fully impacted canines, transplanted to their intended position, whereas one was a third molar transplanted to a premolar site. Nine of the 20 CBCTs were deemed inconclusive by the PI regarding width of the periodontal ligament space (n = 4), root resorptions (n = 3), or the presence of buccal bone walls (n = 2) and were discussed between the examiners to reach an agreement. When compared to PA, CBCT yielded additional information for the diagnosis of root resorptions and missing buccal/oral bone.

Five of the unmodified teeth (21.7%) showed clinical and/or radiographic findings: two were in an infraposition, one presented a perio-endo lesion and two an invasive cervical resorption, which was detectable only on the CBCT scan. Interestingly, five out of seven transplanted molars developed postoperative findings or failed, whereas this was the case only in two of 15 premolars. The single canine in this group exhibited no abnormality. Details are shown in Table 1.

Four of the root-end resected teeth (44.4%) showed postoperative findings, clinically and/or radiographically: the only transplant with preoperative signs of hypercementosis presented with a perio-endo lesion and all three multi-rooted teeth (FDI 18, 46, and 47) showed missing signs of ongoing pulp canal obliteration of which two additionally had increased periodontal pocket depths (8-9 mm). Interestingly, three out of four transplanted molars developed postoperative findings or failed, whereas this was the case in only one of two premolars and one of three canines, i.e., the one with hypercementosis. Details are shown in Table 1.

Finally, the overall success probability after a median follow-up of 3.4 years was 62.5% (20 of 32 transplants) or 69.6% (16 of 23) for transplantations of unmodified teeth after a median observation period of 4.8 years and 44.4% (4 of 9) for root-end resected teeth after a median observation period of 2.5 years.