Determination of risk factors for recurrence of the deformity after plate removal in patients WHO had tension band plating for deformity correction.

Hemiepiphysiodesis is a widely used method in the correction of childhood long bone deformities. Temporary hemiepiphysiodes which allows to guide the growth of an extremity is a very simple surgical method with a low risk of growth plate injury and implant failure. The implant used for temporary hemiepiphysiodesis is removed after the goal is achieved. Once the implant has been removed, the growth plate can continue to grow normally. Therefore, patients should be kept under close follow-up in terms of recurrence of deformity. In the literature, it has been shown that the previous deformity may recur (rebound phenomenon-RP) after removing the tension band plate, but the risk factors for RP have not been investigated in detail Objective The aim of this study is to investigate the risk factors of RP in patients who had hemiepiphysiodes using a tension band plate to correct long bone deformities. Material and Method Patients who underwent temporary hemiepiphysiosis in the Department of Orthopedics and Traumatology of Hacettepe University Hospital due to lower extremity coronal plan deformity between 2010 and 2018were retrospectively rewieved. Total of 217 segments of 103 patients were evaluated. Patients who had standart erect leg lower extremity radiographs (before tension band plate application; at the time of implant removal and at the end of follow-up) and patients with at least one year follow-up after plate removal was included in the study. The study was carried out on 94 segments of 43 patients. Patients' radiographs were evaluated by two surgeons. Measurements were made on the PACS (Centricity Universal Viewer Version 6.0) system. Mechanical lateral distal femoral angle (mLDFA), mechanical medial proximal tibial angle (mMPTA), mechanical axis zone, direction of deformity (varus / valgus), length of the segment (mm) were measured in each radiograph. The age of implant application (months), the duration of the implant in the patients (months), the age of implant removal (months) were noted. The rate of correction of the joint orientation angles (degrees / month) and the bone growth (mm / month) after the implant was removed was calculated. Increase of 5⁰ or more in the joint orientation angles (mLDFA, mMPTA) in the direction of the initial deformity was considered as a RP. The relationship between the patient parameters and deformity recurrence was evaluated and risk factors for RP were determined. Results In total, 94 segments (55 femurs, 39 tibia) of 43 patients (15 male, 28 female) were included in the study. In pre-operative period, 77.7% (n=73) of the segments had valgus, deformity and 23.3% (n=21) had varus deformity. The mean age at implant application was 88.9 ± 40.5 months and the mean age at implant removal was 106 ± 43.0 months. Three different groups were determined in terms of the course of joint orientation angles after implant removal: Bone segments with joint orientation angle change within 5⁰ (ReP(-) group) (n=39); those with recurrence of deformity 5⁰ and above, (ReP(+) group) (n = 46) and segments in which the hemiepiphysiodesis effect persists despite plate removal (n = 9). Parameters thought to be effective on RP were evaluated separately for femur/tibia and valgus/varus deformities. RP was observed in 56.2% (n = 41) of all valgus deformities (n = 73), and in 23.8% (n = 5) of varus deformities (n = 21). The RP rate was 52,7% in all femurs. RP rate was 60% (n = 27) and 20% (n=2) in femoral segments with preoperative valgus (n=45) and varus (n=10) deformities, respectively. The difference between groups were significant (p=0.03). When all tibias were evaluated together RP rate was 43,6%. Tibial segments with properative valgus deformity had %50 (n=14) of RP; tibias with pre-operative varus deformity had %27,3 (n=3) of RP. Although the RP rate was higher in tibias with valgus deformity the difference was not significant (p=0.19). When gender, etiology and sides were compared, there was no statistically significant difference between ReF(+) and ReP(-) groups both in femur and tibia. The age at implant application was found to be 101,9 ± 45,3 and 86,0 ± 40,2 months in ReP(-) and ReP(+) femurs respectively (p = 0,25). The age at implant application was found to be 87,5 ± 30,1 and 84,5 ± 42,7 months in ReP(-) and ReP(+) tibias respectively (p=0,40), The age of implant removal was 123.6±49.1 and 102,5±39,5 months, respectively, in ReP(-) and ReP(+) femurs, (p = 0,09) and 108,8 ± 37,0 and 101,7 ± 42,1 months in ReP(-) and ReP(+) tibias (p=0,60). Correction rate of mLDFA in ReP(-) ve ReP(+) femurs was 0,7±0,4⁰ / month and 1 ± 0,3⁰/month respectively (p = 0,01). The mMPTA correction rate was 0,6 ± 0.3⁰ / month and 0.7 ± 0.3⁰ / month in ReP(-) ve ReP(+) tibias respectively (p = 0,7) . No significant correlation was found between the preoperative mLDFA severity of femurs undergoing hemiepiphysiodesis due to valgus deformity and recurrence of deformity [ReP(-) group = 80,2±4,1⁰, ReP(+) grup 78,9 ± 4,4⁰ p=0.26)]. The age at implant application in the same patient group was found to be 120±40,9 months in ReP(-) group and 85,7±39,6 months in ReP(+) group (p=0,008). The age at implant removal was 139,5±45,9 months in ReP(-) group and 102,4±38,6 months in ReP(+) group (p=0,005). There was a significant relationship between the preoperative mMPTA severity of the tibia treated with hemiepiphysiodes due to valgus deformity and deformity recurrence (ReP(-) group = 102,1±5,4⁰; ReP(+) group = 96,2±4,5⁰; p=0.004). The RP rate in 34 segments in which the mechanical axis was not overcorrected was %52,9 (n=18) whereas it was %59 (n=23) in 39 overcorrected segments (p=0.60). When the overcorrection was evaluated based on joint orientation angles it was found that the overcorrection rate was %77,8 (35 in 45 femurs) and %71,4 of those femurs (n=25) developed RP. Overcorrection was not observed in 10 femurs and only two of them (%20) developed RP. There was significant difference between two groups (p=0.008). Overcorrection did not effect the RP rate in tibial segments. Conclusion Guided growth treatment of pediatric long bone deformities utilizing tension band plating is effective and safe. The risk of RP after plate removal is high. The risk of RP was found to be higher in patients with younger age at application and removal of the plate. Moreover, the RP rate was especially high in femoral valgus deformities. Younger age at plate application and removal, higher correction rate in deformity was significant risk factors for RP in those patients. It was found that intentional overcorrection also increased the risk of RP.