Acetabular cup position and risk of dislocation in primary total hip arthroplasty: A systematic review of the literature

selected as relevant n = 27 Full-text articles eligible n = 26 Studies included in the systematic review n = 28 Additions following review references (title, abstract, and full text) n = 2 Excluded Non-pertinent titles n = 461 Excluded Exclusion criteria met n = 61 Excluded Full text unavailable n = 1 Figure 1. PRISMA fl ow diagram of search strategy and review of literature (Moher et al. 2009). Figure 2. Radiographic cup inclination (I) measured on AP pelvic radiographs (Jolles et al. 2002). Figure 3. Radiographic cup anteversion as calculated using an AP radiograph (Abdel et al. 2016). d: short axis of the ellipse of the acetabular component; D: long axis of the ellipse of the acetabular component. Anteversion (A) is calculated as: A = sin-1 (d/D). Figure 4. Radiographic cup anteversion (A) from a lateral shootthrough radiograph (Jolles et al. 2002). Acta Orthopaedica 2017; 88 (1): 10–17 13 3-dimensional analysis of cup positioning may be performed by CT, which will produce angles corresponding to anatomical inclination and anteversion, as shown in Figure 5. Anatomical inclination is defi ned as the angle between the acetabular axis and the longitudinal axis; anteversion is defi ned as the angle between the transverse axis and the acetabular axis when projected onto the transverse plane (Murray 1993). These angles are not affected by pelvic positioning, a known confounder in radiological measurements (Wan et al. 2009, Kanawade et al. 2014). 4 studies used CT analysis to determine cup anteversion. To calculate inclination, either AP radiographic reconstructions from CT were generated (Kim et al. 2009) or supplementary AP radiographs were used (Pierchon et al. 1994, Nishii et al. 2004, Fujishiro et al. 2016). 2 studies did not describe the methods by which cup anteversion and inclination were measured (Dudda et al. 2010, Moskal and Capps 2011). 2 studies described the use of intraoperative tools in assisting the placement of the acetabular component: a goniometer (Li et al. 1999) and computer-assisted surgery (CAS) (McLawhorn et al. 2015). Anteversion and inclination angles The distribution of cup anteversion and inclination angles was calculated in 19 of the 28 articles and compared between dislocating and non-dislocating hips. Paterno et al. (1997) looked only at inclination. 3 of those 19 studies found a statistically signifi cant difference between mean inclination values when comparing dislocating hips and non-dislocating hips. Biedermann et al. (2005) (anterolateral approach) found that hips that dislocated anteriorly were more abducted, and those that dislocated posteriorly were less abducted than non-dislocating hips. Kim et al. (2009) (posterolateral approach) identifi ed hips with posterior dislocation to have higher inclination angles compared to non-dislocating hips. Garcia-Rey and Garcia-Cimbrelo (2016) (posterolateral) found that dislocating hips in general had greater angles of inclination than nondislocating hips. 6 of 8 studies found that mean anteversion angles differed statistically signifi cantly between dislocating hips and nondislocating hips. 4 of these studies used the posterolateral approach (Lewinnek et al. 1978, Nishii et al. 2004, Kim et al. 2009, Fujishiro et al. 2016), 1 the posterior approach (Masaoka et al. 2006), and 1 the anterolateral approach (Biedermann et al. 2005). 4 studies analyzed anterior and posterior dislocations separately; 3 found that anterior dislocations were associated with increased anteversion (Lewinnek et al. 1978, Biedermann et al. 2005, Fujishiro et al. 2016), and 2 found that posterior dislocations were less anteverted compared to non-dislocators (Masaoka et al. 2006, Fujishiro et al. 2016). Nishii et al. (2004) found that anteversion was less in dislocating hips that dislocated posteriorly than in non-dislocating hips. 1 study did not state whether dislocations were anterior or posterior, but the dislocations were found to be less anteverted in this case (Kim et al. 2009). A signifi cant degree of variability between articles was found when comparing mean angles of anteversion and inclination. As shown in Table 6 (see Supplementary data), most of the articles did not fi nd statistically signifi cant differences in mean inclination values (16/19) and mean anteversion values (12/18) between dislocating and non-dislocating THAs. Target zone for cup placement Lewinnek et al. (1978) proposed a safe zone of 30–50 degrees of inclination and 5–25 degrees of anteversion as a means of minimizing postoperative dislocation. Given the association between excessive inclination and an increased rate of wear and edge loading, Callanan et al. (2011) recommended that an inclination range of 30–45 degrees was more ideal. 16 studies identifi ed a range of angles of anteversion and inclination to guide safe placement of the acetabular component; these new target zones differed from the range of values fi rst proposed by Lewinnek et al. (1978), and are summarized in Table 7 (see Supplementary data). Biedermann et al. (2005) (transgluteal approach) found a statistically signifi cant reduction in dislocation risk for 35–55 degrees of inclination and 5–25 degrees of anteversion. Rittmeister and Callitsis (2006) (76% anterolateral THA) found no difference for this target range. For the posterolateral approach, Fujishiro et al. (2016) assessed 10–30 degrees of anteversion and found a statistically signifi cant reduction in dislocation risk. 4 studies found a statistically signifi cant reduction in cup dislocation for differing target ranges of anteversion and inclination. Biedermann et al. (2005) identifi ed 35–55 degrees of inclination and 5–25 degrees of anteversion, Grammatopoulos et al. (2015) 27–57 degrees of inclination and −3 to 27 degrees of anteversion, and Danoff et al. (2016) identifi ed 30–50 degrees of inclination and 10–25 degrees of anteversion Garcia-Rey and Garcia-Cimbrelo (2016) identifi ed 2 “safe windows”: (1) 35–50 degrees of inclination and 5–25 degrees of anteversion, and (2) 35–50 degrees of inclination and 15–25 degrees of anteversion. The surgical approaches Figure 5. Anatomical cup anteversion using CT imaging (Kim et al. 2009). A: anatomical anteversion with respect to the sagittal plane measured on CT transverse images. 14 Acta Orthopaedica 2017; 88 (1): 10–17 were as follows: posterolateral (Garcia-Rey and Garcia-Cimbrelo 2016), posterior (Danoff et al. 2016), 74% anterolateral (Grammatopoulos et al. 2015), and anterolateral (Biedermann et al. 2005). While the target range identifi ed by Grammatopoulos et al. (2015) was broader (± 15 degrees) than the other ranges detailed above, it nevertheless identifi ed that extreme outliers of cup positioning from the target zone have a higher risk of dislocation. 11 studies compared the combined values of anteversion/ inclination to their placement inside or outside the safe zone as described by Lewinnek et al. (1978). These fi ndings are summarized in Table 8. Only 2 of these studies (Biedermann et al. 2005, Danoff et al. 2016) found a statistically signifi cant reduction in postoperative dislocation by placing cups in the Lewinnek safe zone. This was compared to the 4 articles that determined that there was no signifi cant difference between dislocating and non-dislocating hips (Leichtle et al. 2013, Grammatopoulos et al. 2015, McLawhorn et al. 2015, Opperer et al. 2016). The other articles listed in Table 8 did not verify the differences between dislocators and non-dislocators in the Lewinnek safe zone with statistical analysis. 4 studies found that there were more dislocating hips outside the Lewinnek safe zone than inside (Lewinnek et al. 1978, Masaoka et al. 2006, Grammatopoulos et al. 2015, Danoff et al. 2016). This can be compared to 7 studies that identifi ed more dislocating THAs in the Lewinnek safe zone than there were outside it (Biedermann et al. 2005, Minoda et al. 2006, Leichtle et al. 2013, Esposito et al. 2015, McLawhorn et al. 2015, Opperer et al. 2016, Abdel et al. 2016). 1 study (McLawhorn et al. 2015) identifi ed that a greater proportion of dislocating hips were in the safe zone (83%) compared to the non-dislocating group (53%). Moskal and Capps (2011) compared navigated and nonnavigated placement of the acetabular component with regard to positioning and dislocation rate. 81% of cups placed with the navigated method landed in the Lewinnek safe zone, as compared to 63% in the non-navigated group (p < 0.001). The navigated group also had less dislocations (8/779) than the non-navigated group (17/684) (p = 0.03). Mean angles of anteversion and inclination were similar between the 2 groups (Moskal and Capps 2011).