Six databases were investigated: PubMed, Cochrane Library, Google Scholar, Scopus, LILACS, and Web of Science. The electronic search was conducted until September 2020, using keywords and MeSH terms connected by the Boolean operator “AND.” The PubMed search combined the following term sequence: “orthognathic surgical procedures”[Mesh] AND “condylar resorption” AND “systematic review” and “condylar resorption” AND “orthognathic surgery” AND “systematic review.” The other search databases combined the following term sequence: “orthognathic surgery” AND “condylar resorption” AND “systematic review.” Additionally, a manual search was conducted in the reference lists of the selected SRs. No restriction of language or publication date was imposed.

The electronic search was independently conducted by two investigators (SB and GC), screening titles and abstracts in parallel to evaluate the reviews for eligibility. In case of missing information, full-text reading was necessary for a final decision. A third author (AG) discussed and resolved any discrepancies between the two authors.SRs and MAs were included, which allowed for the extraction of data on CROS. Narrative reviews, overview of reviews, duplicate articles, studies with no evaluation of condylar morphology, and studies on orthognathic surgery of syndromic patients or cleft lip and palate patients were excluded. The Cohen's kappa coefficient (κ) was calculated to determine the inter-rater agreement between the two investigators (SB and GC).

The data extraction from the eligible reviews was independently performed by the same two investigators (SB and GC), recording the following information: author, publication date, study design (SR or MA), number of included studies, number of included patients, dentoskeletal malocclusion, type of intervention, methodological data, quality assessment of primary studies, outcomes, results of reviews, and author’s conclusion.

The methodological quality of each SR was independently assessed by the two investigators (SB and GC) using the latest version of A Measurement Tool to Assess Systematic Reviews (AMSTAR-2).¹³ AMSTAR-2 includes 16 domains, 9 non-critical items, and 7 critical items that strongly influence the final score. The quality assessment can range from high to critically low.¹³

A total of 62 records were selected from the electronic search of six different databases (PubMed, n = 9; Cochrane Library, n = 0; Google Scholar, n = 25; Scopus, n = 13; LILACS, n = 1; Web of Science, n = 14), and no studies were added by manual search. After excluding the duplicates, 25 potentially significant studies were found. After screening titles and abstracts, 23 full-text articles were screened for eligibility, and 2 studies were excluded.^5,14 After full-text reading, 13 reviews were excluded because they did not meet the inclusion criteria (Table 1). Figure 1 shows the flow diagram of the search strategy and SR selection: 10 SRs were included for the qualitative analysis.^7,8,15-22 Quantitative analysis could not be conducted because no MAs were performed. The inter-rater agreement coefficient was κ = 0.93.

Table 2 summarizes all data extracted from the SRs. Each SR included a different number of studies (ranging from 6 to 76), and none performed an MA. The publication year ranged from 2008 to 2019. Regarding the study design, most of the SRs included observational retrospective and prospective studies. Non-randomized controlled trial (non-RCT) was the most common study type, with the others being case control study, cohort study, case series, and case report.^7,8,15-22 Only two SRs described the results of RCTs.^19,20 Dentoskeletal Class II was the most frequent malocclusion analyzed in the included SRs.^7,8,15-22 Other studies evaluated patients with dentofacial malformations such as skeletal Class III, skeletal asymmetry, and skeletal open bite.^7,17,20,21 The most common surgical procedure included BSSO with or without Le Fort I osteotomy.^7,16-22 Five reviews reported the outcomes after Le Fort I osteotomy alone, unilateral sagittal split osteotomy, and intraoral vertical ramus osteotomy.^7,16-18,22 Post-surgical changes of condylar morphology were analyzed on two-dimensional (2D) radiographic examinations (orthopantomography and/or lateral cephalogram), three-dimensional (3D) radiographic examinations (computed tomography, cone beam computed tomography, and magnetic resonance imaging), or combined 2D and 3D examinations.^7,8,15-22 The follow-up period ranged between 12 and 60 months, with a minimum and a maximum follow-up period of 3 and 192 months, respectively.^7,8,15-22 The number of patients included in the SRs ranged between 180 and 3,777, with a mean age of 26.6 years.^7,8,15-22 Post-orthognathic CR was the primary outcome in the included studies evaluating the morphological changes of the mandibular condyle after surgery.^7,15-22 Two SRs reported the effectiveness of CR management.^8,16

The methodological quality of each SR was determined using AMSTAR-2.¹³ Most of the included SRs had a low methodological quality.^7,8,15,19-22 Three SRs had a critically low methodological quality.^16-18 The most critical items of the AMSTAR-2 checklist were the absence of details for excluded studies and the absence of a comprehensive search strategy.

Nine of the included SRs aimed to describe the main characteristics and risk factors of CROS (Table 3).^7,15-22 Jędrzejewski et al.¹⁹ reported that CR manifestations usually ranged from 6 months to 2 years after surgical treatment, but Catherine et al.¹⁶ extended this time to 6 years post-operatively. Although with different diagnostic methods, CR could be identified by both radiographic and clinical signs.^8,16,20 Further, CR was recognized by 2D radiological examinations if there was a reduction of ramus or condylar height, with a minimum vertical decrease of 2 mm.²⁰ A 3D analysis of the condylar profile assessed that a reduction of at least 17% of the condylar volume was considered as the cut-off value for the diagnosis of CR.⁸ Clinical signs of bilateral CR included an anterior open bite or a Class II malocclusion with retrognathia.¹⁶ On the contrary, an active process of unilateral CR could identify mandibular asymmetry with an ipsilateral Class II malocclusion and a contralateral open bite.¹⁶

There is no agreement among the included SRs about the role of sex or age in CROS.^7,15-22 Five SRs reported that females showed a higher risk for CR than males, with an approximate female-to-male ratio of 5:1.^16,17,19-21 Two SRs stated that young patients may be more prone to CR, while de Moraes et al.¹⁷ concluded that age could not be correlated with post-operative CR.^16,17,21 In terms of pre-surgical evaluation, it was mandatory to exclude an active process of condylar atrophy, the presence of severe temporomandibular disorders, and an intra-articular inflammatory damage.^8,16-18

Risk factors for CR were evaluated both in the sagittal and vertical planes.^7,15-22 Angle Class II with mandibular hypoplasia was the most common malocclusion in which CR occurred after a significant mandibular advancement (incidence rate: 1.4–31%).^7,15-22 An orthognathic plan with advancements greater than 10 mm may be a risk factor for CR.¹⁶ Furthermore, in Angle Class II malocclusions, micrognathia often involved condylar morphology characterized by a small volume, a reduced adaptive capacity to external load, and an increased risk of CR.⁸ Only one SR reported a more prone CR in Angle Class III malocclusion after a surgical mandibular set-back greater than 6 mm.⁸ On the vertical plane, a hyperdivergent skeletal pattern with an anterior open bite, a low posterior-to-anterior facial height, and an increased mandibular plane angle (MPA) could cause CR.^7,15-22 However, the entity to define the critical clockwise rotation of the mandible was not unanimously accepted, but it could be described when the MPA was greater than 40°.^16,18 In patients with a dolichofacial profile, a pre-surgical posterior condylar inclination was exposed to CR because the antero-superior surface and the latero-superior area were subjected to an excessive load on the sagittal and coronal planes, respectively.^7,15-22

Six of the included SRs evaluated the type of fixation in relation to CR.^8,16-19,21 Wire fixation, rigid fixation (bicortical miniscrew and miniplates), and intermaxillary fixation (IMF) are reported as the possible options.^16,17 The incidence of CROS or surgical relapse was not affected by the fixation method in 6.3–13% of the cases.^16,17,21 However, a majority of the authors recorded a high risk of CROS after a prolonged IMF rather than a rigid fixation, while no difference was found between wire and rigid fixation.^8,18,19 Although no studies have compared the two types of rigid fixation, Jędrzejewski et al.¹⁹ reported that bicortical miniscrews could determine inferior alveolar nerve damage.

Two included SRs reported the main therapeutic approaches for CR.^8,16 Despite the fact that there are no precise guidelines on the management of CR to date, several suggestions were made, differentiating the pre- and post-surgical phases of evaluation.⁸ In the pre-operative phase, the aim was to annul or limit the risk factors.⁸ In the post-surgical phase, the objectives included the treatment of inflammation and pain, as well as the correction of skeletal deformities and occlusal instability.^8,16 Conservative or surgical treatment were proposed.^8,16 Conservative therapy managed to avoid CR progression, stabilizing the pathologic conditions without improving dentofacial deformities.^8,16 It included anti-inflammatory drugs, occlusal splints, physiotherapy, and in some cases, orthodontic or restorative treatments.^8,16 Surgical treatment should be performed at least 6 months after orthognathic surgery, considering the severity of CR, the surgeon’s experience, and the patient’s intentions.⁸ Disc repositioning, condylectomy, chondro-costal graft, or complete prosthetic TMJ reconstruction were more invasive therapeutic options that were reported as possible treatments for CR.^8,16