Bhatia and Leighton [3] described natural growth of the maxilla at the age of ten to twelve years by an increase of SNA angle of 0.5° and natural growth of the mandibula at that age by an increase of SNB angle of 0.6°. Modifying facial growth using orthopedic forces has been of special interest in orthodontists for ages, since maxillary protraction was pioneered by Delaire in the 1970s [14]. Forward movement of the maxilla by 1–2 mm, a 3° increase in SNA, 1.02° decrease in SNB and a reduction in ANB of -2.43° were described with wide variations [5, 8, 10, 14, 18]. Improved dental arch relationships mostly because of dentoalveolar compensation were the results of maxillary protraction with facemask [11]. In literature, the ideal treatment timing for orthopedic treatment of class III malocclusion with maxillary retrognathia was at the age of five to eight years during deciduous and early mixed dentition [14, 23]. Six months after maxillary protraction, the maxilla showed forward movement, but also proclination of upper incisors and extrusion of maxillary molars resulting in increased lower face height and the mandibula rotating in posterior direction. Correction of the class III malocclusion was due to forward movement of the maxilla, but also clockwise rotation of the mandibula. Therefore, correction of the overjet was because of dental and skeletal changes [19, 23]. Unwanted dental and skeletal side effects, such as proclination of upper incisors and clockwise rotation of the mandibula, were also seen in patients with facemask of our study. Success and failure of class III malocclusion treatment depends on the potential of growth and treatment is requiring a long-term period, making patient´s motivation difficult in the long run [23, 30]. Therefore, knowing the ideal timing for facemask therapy to obtain better treatment results is indispensable. Takada et al. [30] described the maximum peak of maxillary growth between ten to twelve years of age for girls and twelve to thirteen years for boys. This claim is contentious. Other than that, Alexander et al. suggested that the maximum peak of maxillary growth occurs during the prepubertal period [1]. Orthopedic class III malocclusion treatment shows the best results when the facemask is applied before the pubertal growth spurt, because the suture´s adaptability and response to maxillary protraction decreases with age [11, 24, 30]. Dibbets and van der Weele [15] investigated the treatment with the facemask and its forces regarding temporomandibular joint dysfunction. They reported no causal relationship of facemask treatment with temporomandibular joint dysfunction even with a 500 g force on each side. In our study, neither patients with bone anchors nor patients with facemask showed symptoms of temporomandibular joint disfunction before or after treatment. To reduce unwanted side effects of facemask therapy, titanium bone anchors were used some years later for maxillary protraction being well tolerated by the patients [6, 10, 11]. Maxillary protraction was performed using a rigid external distractor, a facemask after Le Fort I corticotomy in patients with a cleft or a combination of skeletal anchorage in the upper jaw and facemask [10]. Liu et al. [22] described a technique using bone anchored hooks combined with facemask and additional sutural distraction for correction in four patients at the age of six to twelve years or Le Fort III osteotomy in four patients older than twelve years with and without cleft lip and palate. No complications concerning surgery process or loosening of the bone-born hooks for distraction occurred. The midface advancement was 8 mm in patients with sutural distraction and 10 mm in patients with Le Fort III osteotomy with remarkable changes in face contour and normal occlusion after the treatment. There was no relapse described after a six months follow-up. Kircelli and Pektas [21] used miniplates on the lateral nasal wall of the maxilla in six patients at the mean age of 11.8 ± 1.1 years and attached them to a facemask for 10.8 ± 2.4 months. The infraorbital region moved 3.3 ± 1.1 mm forward. Point A moved 4.8 ± 2.0 mm forward. The results remained stable over the 15.2 ± 0.9 months follow-up period. They concluded that skeletal anchorage combined with a facemask leads to remarkable advancement of the midface and soft-tissue profile in the late mixed-dentition period. Bone anchors and class III elastics without additional corticotomy or osteotomy was pioneered by De Clerck et al. in the 2000s [11]. Extraoral facemask was no longer needed with this approach and elastics can be worn all over the day. In this study, De Clerck et al. described the treatment of three female patients with maxillary deficiency and a concave soft tissue profile at the age of ten to eleven years. Anterior crossbite was corrected in all three patients after treatment and the soft tissue profile improved. The cephalometric radiographs showed an improvement of ANB, Wits appraisal and facial convexity. Upper incisor inclination remained stable during treatment, lower incisors were proclined afterwards. The class III correction was stable from the end of treatment to a 11- and 38-months follow-up. In a later study, De Clerck et al. [13] treated twenty-five Class III patients at the mean age of 11.10 ± 1.1 years with bone anchors and Class III elastics and took cone-beam computed tomography images before elastic wear and after treatment. They reported a posterior displacement of the mandibula after the treatment in all patients. The displacement of the posterior ramus was 2.74 ± 1.36 mm, of the condyles, 2.07 ± 1.16 mm and of the chin − 0.13 ± 2.89 mm. Even remodeling of the mandibular fossa at its anterior eminence was 1.38 ± 1.03 mm and resorption of bone of the posterior region was − 1.34 ± 0.06 mm. Cevidanes et al. [4] compared 21 patients with bone anchors at the mean age of 11 years 10 months ± 1 year 10 months and 34 patients with facemask at the mean age of 8 years 3 months ± 1 year 10 months after one year of treatment. Maxillary advancement and midfacial length were about 2.5–3.0 mm larger in patients with bone-anchors. There were no differences between sagittal growth and position of the mandibula between the two groups. Maxillomandibular divergency was decreased of about 3° in patients with bone anchors, slight counterclockwise rotation of the mandibula was noted in patients with bone anchors and clockwise rotation of the mandibula was seen in patients with facemask. Patients with bone anchors did not show the same amount of lingual inclination of lower incisors as patients with facemask did. Nguyen et al. [26] reported their results of twenty-five Class III patients at the mean age of 11.10 ± 1.1 years treated with bone anchors and Class III elastics. Cone-beam computed tomography images before elastic wear and after treatment showed a mean forward displacement of the maxilla of 3.7 mm and of the zygomas of 4.3 mm, but also incisors came forward 3.7 mm. De Clerck and Swennen [10] described the success rate of miniplates concerning stability with 97% in twenty-five patients with mean age 12.0 ± 1.2 years, but during elastic use five miniplates out of hundred showed signs of mobility. Two miniplates were stable again, after the patients stopped using elastics for two months. The other three miniplates were removed and replaced after three months of healing. Contrary to expectations Cornelis et al. [7] concluded in their systematic review of 28 full-text articles concerning bone-anchored maxillary protraction that the level of evidence available for supporting maxillary protraction effect using bone anchors was low. They remarked identical samples in publications reporting results that tended to suggest positive results using bone anchors for class III malocclusion treatment. They even questioned clinical significance concerning the differences in sagittal correction between bone anchors and facemask and recommended long-term follow-up results. In a newer study Kamel et al. [20] reported their results of seventeen Class III patients during late mixed or early permanent dentition treated with a hybrid hyrax expander and class III elastics to a bone-supported bar in the mandibula for about one year compared to thirteen patients without treatment. SNA angle showed an increase in the treated group of 4.64 ± 0.95° and in the control group of 0.42 ± 0.21°. SNB angle showed a decrease in the treated group of -0.25 ± 0.47° and in the control group an increase of 1.03 ± 0.59°. ANB angle showed an increase in the treated group of 4.90 ± 1.31° and a decrease in the control group of -0.61 ± 0.55°. Wits appraisal showed an increase in the treated group of 5.27 ± 1.07° and in the control group of 0.28 ± 0.45°. The lower face height increased in the treated group and the mandibula showed a clockwise rotation with closure of gonial angle. Maxillary and mandibular incisors showed proclination in the treated group. Mesialization and extrusion of upper molars were seen in the treated group. Nienkemper et al. [27] reported their results of 16 growing class III children (mean age 9.5 ± 1.6 years) treated with a hybrid hyrax-facemask combination using pre- and posttreatment cephalograms compared with a control group of 16 untreated Class III subjects. The mean treatment duration was 5.8 ± 1.6 months. The results showed significant improvement in SNA (2.4°), SNB (-1.7°) and Wits appraisal (4.5 mm). Comparison of the treatment and the control group showed a larger gonial angle in the control group. All mini-implants in the treatment group remained stable during treatment. Compared to the results of our study, these values were higher than in our facemask group, since we did not use mini-implants. Ngan et al. [25] compared 20 class III patients (mean age 9.8 ± 1.6 years) with tooth-borne rapid palatal expansion appliance and facemask and 20 class III patients (mean age 9.6 ± 1.2 years) with bone-anchored rapid palatal expansion appliance and facemask. The tooth-borne facemask group showed more proclination of maxillary incisors (2.12 mm), the bone-anchored facemask group showed less downward movement of Point A (-0.4 mm) than the tooth-borne facemask group (1.2 mm) and less opening of the mandibular plane in the bone-anchored facemask group (-0.25°) than in the tooth-borne facemask group (2.76°). The results of the tooth-borne facemask group were comparable to the results of our study. Regarding failure rate, the hybrid-hyrax with palatal mini-screws could be an alternative to the bone anchors used in our study. However, treatment with mini-screws is not covered by the statutory health insurance (GKV) applying to our patients. Since most of those were dependent on treatment modalities covered by the statutory health insurance, they decided on bone anchors as described in our study.

The observed failure of the bone anchors was mainly due to poor oral or hand hygiene, which led to infections and subsequent loosening of the bone anchors. Therefrom, better hygiene could lower the failure rate.

Early treatment of class III patients, as shown in our study at the age of 6.74 ± 1.15 years for facemask patients, aimed to effect maxillary growth at sutures articulating with the frontal, zygomatic, ethmoid and palatal bones. Later treatment of class III patients, as shown in our study at the age of 11.00 ± 1.76 years for bone anchor patients, aimed to effect apposition processes over all surfaces, as this is the predominant growth mechanism after the end of suture activity at around seven years of age [16].

In both groups, changes at t1 showed active treatment outcomes of the facemask or bone anchors, including possible growth effects that may occur between six and eleven years of age. Nonetheless, the maxilla moved more forward in patients with bone anchors (2.30 ± 1.18°) than in patients with facemask (1.22 ± 2.28°). Contrary to expectations, the mandibula moved more forward in patients with bone anchors (0.71 ± 1.28°) than in patients with facemask (-0.81 ± 1.35°) as well. Accordingly, the ANB angle was smaller in patients with bone anchors (0.08 ± 1.56°) than in patients with facemask (2.20 ± 2.56°) after treatment. A clockwise rotation of the maxilla was more expressed in patients with bone anchors (0.90 ± 2.24°) than in patients with facemask (-0.64 ± 1.88°). The clockwise rotation of jaws increases the ANB angle as well. Clockwise rotation of the mandibula was less in patients with bone anchors (-1.01 ± 2.48°) than in patients with facemask (1.05 ± 1.51°). Divergency of the maxilla and the mandible was less in patients with bone anchors (-2.10 ± 3.73°) than in patients with facemask (1.68 ± 1.77°). Mandibular angle change was almost the same in patients with bone anchors (0.41 ± 4.00°) and in patients with facemask (0.40 ± 2.80°) resulting in a horizontal growth pattern in patients with bone anchors and in a vertical growth pattern in patients with facemask. Forward movement of upper incisors was less in patients with bone anchors (1.69 ± 5.62°) than in patients with facemask (3.35 ± 6.18°). Forward movement of lower incisors was greater in patients with bone anchors (2.03 ± 4.81°) than in patients with facemask (-4.73 ± 5.28°), since the chin cap part of the facemask influences the lower incisor inclination in terms of backward movement of the incisors. Movement of the incisors influences the anterior region of the upper and lower jaw and influences ANB angle as well. Nasolabial angle got larger in patients with bone anchors (4.92 ± 17.59°) than in patients with facemask (-2.58 ± 9.47°), mainly because the patients with bone anchors presented greater initial proclination of the upper incisors already prior to treatment. The change of the occlusal plane inclination was only significant in patients with facemask (2.63 ± 2.43°), whereas it was almost indistinguishable in patients with bone anchors (1.00 ± 3.05°). Wits was smaller but improved more during treatment in patients with bone anchors (2.01 ± 2.65 mm) than in patients with facemask (1.85 ± 4.09 mm), depending on the changes of the maxilla and the mandibula and incisor inclination of both jaws.

The forward movement of the upper lip was less in patients with bone anchors (-0.58 ± 2.07 mm) than in patients with facemask (0.95 ± 1.90 mm), depending on the movement of the maxilla and upper incisors. Backward movement of the lower lip was less in patients with bone anchors (-1.42 ± 2.02 mm) than in patients with facemask (0.42 ± 1.39 mm), depending on the movement of the mandibula and lower incisors.

All patients with a facemask presented a change of the deciduous to the permanent upper and lower incisors during treatment. This change influenced upper and lower incisor inclination as well. Due to the age discrepancy of patients with facemask and with bone anchors, a clear comparison of the two maxillary protraction protocols is limited, but nevertheless of significant clinical interest, since patients requiring treatment for a skeletal class III are referred to the orthodontist at different ages. The ideal control for both groups of our study would consist of untreated growing class III patients with corresponding age. However, the ALARA principle prohibits X-rays in patients without appropriate treatment. Our study comprised cephalometric radiographs in patients with immediate treatment need. Apart from this, growing patients with a skeletal class III for whom treatment is indicated should not be left untreated for ethical reasons.

Finally, certain clinical aspects of the two treatment approaches of our study must be considered. The extraoral facemask is more bulky and less tolerated than intraoral bone anchors and class III elastics. The amount of facemask use per day is smaller than for bone anchors. Two surgical procedures are required for the bone anchors, that is, the insertion and removal of the miniplates. After protraction of the upper jaw, however, the bone anchors can be used for anchoring or distalisation during subsequent orthodontic treatment. In addition, bone anchors are a useful treatment approach for patients whose facemask therapy had not been successful, and as an attempt to avoid or at least to decrease the amount of later orthognathic surgery, especially in patients that have been too old for facemask treatment.