- Email: [email protected]
Correlation between coxa profunda and morphological parameters of acetabular coverage in a Japanese cohort: A CT study
Journal of Orthopaedic Science 21 (2016) 667e672
Contents lists available at ScienceDirect
Journal of Orthopaedic Science journal homepage: http://www.elsevier.com/locate/jos
Original article
Correlation between coxa profunda and morphological parameters of acetabular coverage in a Japanese cohort: A CT study Tomohiro Goto*, Kazuaki Mineta, Keizo Wada, Yasuaki Tamaki, Daisuke Hamada, Tomoya Takasago, Kosaku Higashino, Koichi Sairyo Department of Orthopedics, Institute of Biomedical Sciences, Tokushima University Graduate School, 3-18-15 Kuramoto, Tokushima 770-8503, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 December 2015 Received in revised form 22 May 2016 Accepted 26 June 2016 Available online 21 July 2016
Background: Coxa profunda is generally considered an indicator of acetabular overcoverage. However, recent studies have opposed this view. The correlation between coxa profunda and acetabular overcoverage thus remains controversial. The purpose of this study was to investigate the prevalence of coxa profunda and the association between coxa profunda and acetabular coverage based on sex in Japanese subjects using computed tomography. Methods: We reviewed the computed tomography scans of 151 Japanese consecutive patients (302 hips) aged <50 years who underwent abdominopelvic computed tomography for symptoms unrelated to hip disease. Coxa profunda was diagnosed when the floor of the acetabular fossa touched or was medial to the ilioischial line. We measured the lateral center edge angle, acetabular roof obliquity angle, acetabular version, and anterior and posterior acetabular sector angles as parameters of acetabular coverage. Results: Coxa profunda was observed in 45.7% of all subjects, and was more common in females (63.0%) than in males (29.5%). We found no significant differences in the incidence of coxa profunda among subjects with dysplasia, overcoverage, and normal coverage. Males with coxa profunda showed significantly greater lateral center edge angle, anterior acetabular sector angle, and lower acetabular roof obliquity angle, whereas anterior and posterior acetabular sector angles were greater in all subjects and females with coxa profunda. Coxa profunda showed poor sensitivity (57.7%) and specificity (56.8%) for detecting overcoverage in all subjects. Conclusions: Coxa profunda is a common radiographic feature in Japanese patients, especially in women. Coxa profunda may be less useful for diagnosing acetabular overcoverage because of its common occurrence with various hip morphologies and its poor specificity for detecting acetabular overcoverage. However, men with coxa profunda have greater acetabular coverage, so coxa profunda may contribute to acetabular coverage in men. We should take into account that coxa profunda may have different implications between the sexes. © 2016 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
1. Introduction Femoroacetabular impingement (FAI) has recently been widely recognized as a cause of hip pain in young adults. The pathological condition of FAI involves abnormal, repetitive abutment of the femur and acetabulum during hip motion and can cause labrum and cartilage injuries [1,2]. FAI is caused by morphological abnormalities such as small headeneck offset or aspherical femoral head in cam impingement, whereas acetabular overcoverage is present in
* Corresponding author. fax: þ81 88 633 0178. E-mail address: [email protected] (T. Goto).
pincer impingement [3]. To establish the appropriate therapeutic strategy for FAI, a comprehensive evaluation including clinical history, physical examination, and above all, radiographic evaluation, is essential. Coxa profunda, defined as the floor of the acetabular fossa touching or medial to the ilioischial line in the AP view of the pelvis, has been considered one of the diagnostic radiographic findings for acetabular overcoverage [2,4]. In conjunction with proximal femoral morphological parameters such as antetorsion, neck-shaft angle, and concavity, this may lead to pincer impingement [2e8]. However, in recent years, several reports have opposed a positive relationship between coxa profunda and acetabular overcoverage [8e11], suggesting that coxa profunda is a common radiographic finding in
http://dx.doi.org/10.1016/j.jos.2016.06.008 0949-2658/© 2016 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
668
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
normal hips and may often be observed even in dysplastic hips [9,10,12]. Thus, whether coxa profunda can serve as a reliable radiographic indicator of acetabular overcoverage remains controversial. Furthermore, there are racial differences in hip morphology. It has been reported that primary hip osteoarthritis (OA) is common in Western countries; however, in Japan, the predominant type of hip OA is secondary OA due to hip dysplasia [13e15]. Thus, it is possible that the pathological basis of FAI differs in various ethnic populations, since FAI is based on morphological abnormalities and is noted to be a possible cause of early primary hip OA [1e3]. We believe that it is necessary to conduct detailed evaluations in each ethnic population to elucidate the pathogenesis of FAI, based on the morphological abnormalities, in hip disorders. In this study, we verified the prevalence of coxa profunda and the association between the presence of coxa profunda and radiographic parameters of acetabular coverage according to sex using computed tomography (CT) in Japanese subjects. Moreover, the sensitivity, specificity, positive predictive value, and negative predictive value of coxa profunda for acetabular overcoverage were assessed. 2. Materials and methods 2.1. Patients This retrospective study was approved by the institutional review board at our institution. We reviewed the CT scans of 155 Japanese consecutive patients (310 hips) under 50 years of age who underwent abdominal and pelvic CT for reasons unrelated to hip symptoms in our institution from September 2010 to June 2011. The 50-year-old age limit was set so that the study focused on relatively younger subjects who are susceptible to FAI and less influenced by degenerative changes in the hip joint. We excluded patients with avascular necrosis, hip OA, and/or previous hip surgery. OA was defined as the presence of at least one of the following: joint space narrowing, osteophytes, subchondral bone sclerosis, and subchondral bone cysts. After exclusion, 151 patient records (302 hips) were available for further analyses. We analyzed 156 male hips and 146 female hips, with a mean patient age of 37.8 years (range, 21e49 years). 2.2. Computed tomography evaluations All CT scans were ordered from the departments of general surgery, urology, and gynecology. Each CT was performed using a
16-detector-row CT scanner (Aquilion 16; Toshiba Medical Systems, Tokyo, Japan), and the acquired images were reconstructed into CT images. Observation was performed using a 1-mm slice thickness multiplanar reconstruction (MPR) image with Aquarius NET Server (TeraRecon, Inc., San Mateo, CA, USA) under standard bone settings. All measurements were performed with the patient in the supine position and 3DCT images were used to confirm the pelvic rotation by adjusting bilateral anterior superior iliac spine in the sagittal view. A single experienced orthopedic doctor reviewed all patient images one time. After that, he re-reviewed a subset of 50 subjects, which were selected randomly, and the other experienced orthopedic doctor evaluated those 50 subjects twice to verify the reliability of the radiographic analysis. They reviewed the images for the presence of coxa profunda and the measurement of radiological parameters pertaining to acetabular coverage. The diagnosis of coxa profunda was made when the floor of the acetabular fossa touched or was medial to the ilioischial line [2] on the coronal transparent three-dimensional model of the whole pelvis (simulating pelvis radiograph) (Fig. 1). We could obtain high contrast images from the coronal transparent 3D model by which accurate diagnosis of coxa profunda could be made. Recently, it has been more common using these images for morphological studies [16]. We measured the following parameters for evaluation of acetabular coverage: lateral center edge angle (LCE) [4,17], acetabular roof obliquity angle (ARO) [4], acetabular version (AV) [18], anterior acetabular sector angle (AASA), and posterior acetabular sector angle (PASA) [19]. We used these parameters to conduct a detailed analysis from multiple directions: LCE represents the lateral acetabular coverage, ARO shows the lateral opening of the acetabulum and lateral inclination of the weight-bearing surface, AV shows the anterior opening of the acetabulum, and AASA and PASA indicate the acetabular coverage at anterior and posterior aspects. LCE was determined by measuring the angle between two lines: a line through the center of the femoral head perpendicular to the transverse axis of the pelvis and a line connecting the center of the femoral head to the acetabular lateral margin (Fig. 2a). ARO was the angle formed by a horizontal line referenced on the pelvis (the line connecting the inferior aspects of the bilateral teardrops) and a line connecting the medial point of the sclerotic zone to the lateral edge of the sourcil (Fig. 2b). LCE and ARO were measured on a reformatted coronal image passing through the center of the femoral head. AV, AASA, and PASA are measurements of the anteroposterior acetabular coverage in an axial reformat. AV was obtained by measuring the angle between
Fig. 1. The coronal transparent three-dimensional model of the hip. Images of negative coxa profunda (a) and positive coxa profunda (b) are shown (white dotted line: the floor of the acetabular fossa, white line: the ilioischial line).
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
669
Fig. 2. The measurements of the lateral center edge angle (LCE) (a) and the acetabular roof obliquity angle (ARO) (b).
two lines: a line connecting the anterior and posterior acetabular rims and a line perpendicular to the coronal plane of the pelvis that extended anteriorly from the posterior wall (Fig. 3). Two AV measurements were made at the two separate transverse sections: central AV, a plane through the bilateral femoral heads (Fig. 3a); and cranial AV, a plane corresponding to a point 5 mm distal to the acetabular roof (Fig. 3b). We measured AASA and PASA by examining the angle made by a line connecting the center of the bilateral femoral head and lines from the center of the femoral heads to the anterior and posterior margins of the acetabulum (Fig. 4). We defined acetabular overcoverage as LCE > 40 and/or ARO < 0 , and acetabular dysplasia as LCE < 25 and/or ARO > 15 [4,6,8 20e22]. Other cases were defined as normal acetabular coverage.
of 39.8 years (range, 21e49 years) who were selected randomly. Measurements of all the radiological parameters were repeated two times each by two different observers, who were blinded to the results reported by the other observer, and the intraclass/interclass correlation coefficients (ICC) were assessed. Kappa coefficient was also performed for checking reproducibility in the diagnosis of coxa profunda. The intraclass/interclass correlation coefficients were almost in perfect agreement for all radiographic measurements (ICC: 0.88e0.97) (Table 1) and reproducibility in the diagnosis of coxa profunda was excellent (kappa value: 0.94). 3. Results 3.1. The prevalence of coxa profunda in Japanese subjects
2.3. Statistical analysis Statistical analyses were performed using SPSS software, version 22.0 (SPSS Inc., Chicago, IL, USA). The ManneWhitney Utest, chi-square tests applying Yates correction, and Fisher's exact test were used to compare patients with and without coxa profunda, male and female subjects, and to examine the prevalence of coxa profunda for statistical analysis. Statistical differences were considered significant when p < 0.05. The sensitivity and specificity of coxa profunda with respect to acetabular overcoverage were reported with 95% binomial confidence intervals. The reliability of the radiographic analysis was assessed by determining the inter- and intra-observer reproducibility in 50 patients; 22 male hips and 28 female hips, with a mean patient age
Coxa profunda was observed in 45.7% (138 of 302 hips) of all the subjects. Coxa profunda was much more common in females (63.0%) than in males (29.5%) (p < 0.001) (Table 2). Of the 302 hips, 39 hips showed dysplasia (defined as LCE < 25 and/or ARO > 15 ), 60 hips had overcoverage (defined as LCE > 40 and/or ARO < 0 ), and 204 hips showed normal coverage (Table 2). We found no significant differences in the incidence of coxa profunda between each of these three cohorts (p ¼ 0.431). However, in each cohort, significant differences were noted for overcoverage and normal coverage (p ¼ 0.029 and <0.001, respectively) with respect to the prevalence of coxa profunda between the sexes. In the dysplasia group, the incidence of coxa profunda tended to be high in females; however, no statistical significance was noted. The highest
Fig. 3. Acetabular version (AV) measurements. Two measurements were made at the two separate transverse sections: (a) central AV, a plane through the bilateral femoral heads; (b) cranial AV, a plane corresponding to a point 5 mm distal to the acetabular roof.
670
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
Fig. 4. The measurements of the anterior acetabular sector angle (AASA) and the posterior acetabular sector angle (PASA).
Table 1 Reliability of the radiographic parameters. Parameters
Examiner
Intra-observer reliability ICC (1,2)
Inter-observer reliability ICC (2,2)
Lateral center edge angle
TG KM TG KM TG KM TG KM TG KM TG KM
0.96 0.90 0.97 0.96 0.94 0.93 0.95 0.96 0.88 0.93 0.92 0.92
0.94
Acetabular roof obliquity Central anteversion angle Cranial anteversion angle Anterior acetabular sector angle Posterior acetabular sector angle
0.97 0.96 0.97 0.91 0.91
prevalence of coxa profunda was 78.6% in females with acetabular overcoverage (Table 2). 3.2. The association between the presence of coxa profunda and radiographic parameters of acetabular overcoverage by sex using computed tomography In all the subjects, although there were no significant differences in age, LCE, ARO, or AV (both central and cranial) between the hips Table 2 Prevalence of coxa profunda.
Total subjects (N ¼ 302) Dysplasia (N ¼ 39) Overcoverage (N ¼ 60) Normal coverage (N ¼ 204) Comparison among three cohorts (p value) a b c
3.3. The sensitivity, specificity, positive predictive value, and negative predictive value of coxa profunda for acetabular overcoverage The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of coxa profunda as an indicator of acetabular overcoverage are shown in Table 5. In all the subjects, the sensitivity, specificity, and PPV were relatively low but NPV was high. The values were different between the sexes; coxa profunda showed a relatively high specificity but low PPV for acetabular overcoverage in males, whereas it showed high sensitivity and NPV in females. 4. Discussion
Prevalence of coxa profunda (%) All hips
with and without coxa profunda, the AASA and PASA in hips with coxa profunda were significantly greater than in those without coxa profunda (Table 3). Further, significant sex differences in several radiographic parameters were noted (Table 4). Males with hips with coxa profunda showed significantly greater LCE (p ¼ 0.004) and AASA (p < 0.001) but lower ARO (p < 0.001), whereas in females, AASA and PASA in hips with coxa profunda were greater than those in hips without coxa profunda. According to CT measurement values, males showed relatively greater acetabular coverage compared with females. Although the AASA in females with coxa profunda was greater than in those without coxa profunda, its mean value was lower than the AASA in males without coxa profunda (Table 4).
Males
Females
Comparison between sexes (p value)
45.7% (138/302) 29.5% (46/156) 63.0% (92/146) <0.001a 43.6% (17/39)
23.1% (3/13)
53.8% (14/26)
0.093b
53.8% (28/52)
42.1% (16/38)
78.6% (11/14)
0.029b
44.1% (93/211)
25.7% (27/105) 63.2% (67/106) <0.001a
0.431a
0.232c
Chi-square test. Fisher's exact test. Yates continuity corrected Chi-square test.
0.484c
In this study, we identified coxa profunda in 138 of 302 hips (45.7%), and large differences were observed in its incidence between males (29.5%) and females (63.0%) (p < 0.001). Moreover, we found no significant differences in the prevalence of coxa profunda between the three cohorts; further, no significant differences were noted in dysplasia (43.6%), overcoverage (53.8%), and normal acetabular coverage (44.1%) in all hips (p ¼ 0.431), and in the hips of males (p ¼ 0.232) or females (p ¼ 0.484). Our results suggested that coxa profunda is commonly observed in various types of acetabular coverage as well as in normal acetabular coverage, and its occurrence is higher in females in Japanese subjects. There were several previous reports that discussed the prevalence of coxa profunda. According to studies from countries outside of Asia (Europe, North and South America) involving FAI subjects, the prevalence of coxa profunda ranges from 27% to 77.4% [2,5,8e10,23]. The prevalence rates in our study were indeed within this range, thus, it seems that
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672 Table 3 CT measurement values between hips with and without coxa profunda.
Age (years) Lateral center edge angle ( ) Acetabular roof obliquity ( ) Central anteversion angle ( ) Cranial anteversion angle ( ) Anterior acetabular sector angle ( ) Posterior acetabular sector angle ( )
Coxa profunda (N ¼ 138) mean ± SD (range)
Without coxa profunda (N ¼ 164) mean ± SD (range)
p Value
37.8 ± 8.4 (20e49) 33.1 ± 7.4 (17.0e53.3) 5.3 ± 6.5 (11.5 to 24.3) 19.1 ± 6.4 (0.7e33.9) 15.3 ± 9.8 (16.5 to 44.3) 60.5 ± 8.3 (39.3e79.8) 98.2 ± 7.9 (75.6e116.1)
37.9 ± 8.2 (22e49)
0.957
31.6 ± 7.2 (13.5e54.6)
0.172
6.6 ± 6.6 (11.6 to 36.6)
0.053
18.6 ± 6.6 (2.1e35.8)
0.556
14.0 ± 9.4 (14.2 to 38.2)
0.147
58.2 ± 9.1 (15.3e87.3)
0.041
95.2 ± 9.0 (72.3e120.6)
<0.001
there were no ethnic differences in the prevalence of coxa profunda. Anderson et al. (a report from USA) found a prevalence of 34% in asymptomatic football players (male patients); 56% in patients undergoing periacetabular osteotomy; and 58% in patients with surgical dislocation [9]. Nepple et al. (a report from USA) estimated the prevalence of coxa profunda in four cohorts, namely, those with dysplasia (41%), FAI (64%), Perthes deformities (31%), and asymptomatic individuals (76%, highest); the average prevalence was 55% [10]. They also observed a strong sex difference, being 70% in females and 24% in males in the cohorts. These values were similar to our results. Although the patients' backgrounds are different among the studies, it seems that there were little ethnic differences in the prevalence of coxa profunda; a considerable high prevalence of coxa profunda may be observed in both USA and Japanese populations. Our study revealed a nearly two times higher prevalence of coxa profunda in females. Coxa profunda is a radiographic character defined as fovea projection relative to the ilioischial line in the AP view but the actual coronal plane of the fovea and the ilioischial line is different in three dimensions. Therefore, the relative positional relationship between the fovea and ilioischial line can be changed by anatomical variations such as the width of the pelvis (ischium). We attribute the sex difference to the relatively wider pelvis in females than in males. In terms of the relationship between the presence of coxa profunda and acetabular depth, several authors have found no significant differences in the LCE and ARO between cohorts with and without coxa profunda among asymptomatic and symptomatic subjects [3,10]. On the other hand, Allen et al. showed significantly
671
greater LCE in hips of FAI subjects with coxa profunda (average, 38.7 ) than in those without coxa profunda (average, 33.6 ) [24]. Another report demonstrated significant associations for LCE and ARO with the presence of coxa profunda (LCE, 36.2 ; ARO, 2.8 with coxa profunda; and LCE, 32.6 ; ARO, 5.7 without coxa profunda) [25]. In the present study, we observed statistically significant differences in several radiographic parameters: AASA and PASA in all subjects and females as well as LCE, ARO, and AASA in males. These results indicate that cases of coxa profunda tended to have greater acetabular coverage, particularly in male subjects, and yet the average of each LCE, ARO, and AASA of men with coxa profunda was still within normal range. This is because our subjects were extracted from the general population who underwent CT examination unrelated to hip disorders. Thus, most of our patients showed normal hip morphology. Although our results showed greater AASA and PASA in female subjects, there were no differences in LCE and ARO, which are commonly used to define acetabular coverage; moreover, the mean AASA in females with coxa profunda was lower than that in males without coxa profunda. Thus, the correlation of coxa profunda with acetabular coverage may be weak in female subjects. Fujii et al. retrospectively reviewed Japanese dysplastic hips and suggested that hips with increased AV and an inwardly rotated pelvis tended to have coxa profunda [12]. Their results differ partially from ours. We assume that the severity of acetabular dysplasia lead to these differences, because severe dysplasia results in dramatic changes in pelvic morphology, including rotation of the innominate bone that influences AV [26]. The dysplasia in the abovementioned study [12] was far more severe (average LCE, 8.6 ; ARO, 23 ) than in our study (average LCE, 21.2 ; ARO, 13.8 ). We found that coxa profunda showed poor sensitivity (57.7%) and specificity (56.8%) in determining acetabular overcoverage in all subjects in our study. With respect to sex differences, coxa profunda in males showed relatively high specificity (74.8%) but low PPV (37.0%). PPV does not always synchronize with specificity when the morbidity rate is low, and the number of our male subjects of acetabular overcoverage with coxa profunda was small. On the other hand, in females, coxa profunda had high sensitivity (78.9%) and NPV (92.6%) but poor specificity (39.9%). Thus, coxa profunda cannot be an indicator of acetabular overcoverage; however, the high sensitivity and NPV in females indicate that the use of coxa profunda for excluding acetabular overcoverage may be feasible in female patients. Our study has certain limitations. First, we had no information regarding the patients' backgrounds, including clinical symptoms or medical history. This was intended to exclude any biases; therefore,
Table 4 CT measurement values between sexes with and without coxa profunda.
Number of hips Age (years) Lateral center edge angle ( ) Acetabular roof obliquity ( ) Central anteversion angle ( ) Cranial anteversion angle ( ) Anterior acetabular sector angle ( ) Posterior acetabular sector angle ( )
Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female
Coxa profunda (N ¼ 138) mean ± SD (range)
Without coxa profunda (N ¼ 164) mean ± SD (range)
p Value
46 92 37.2 ± 9.4 (22e49) 38.1 ± 7.9 (20e49) 36.0 ± 7.2 (20.6e53.3) 31.6 ± 7.1 (17e50.6) 1.8 ± 5.2 (6.9 to 16) 7.0 ± 6.5 (11.5 to 24.3) 15.7 ± 5.0 (6.1e28.9) 20.7 ± 3.4 (0.7e33.9) 12.6 ± 9.1 (8.1 to 31.8) 16.7 ± 9.9 (16.5 to 44.3) 65.0 ± 7.8 (49.0e79.8) 58.3 ± 7.7 (39.3e78.6) 96.2 ± 8.1 (75.6e110.8) 99.2 ± 7.7 (76.4e116.1)
110 54 38.4 ± 8.1 (22e49) 37.0 ± 8.4 (22e49) 32.5 ± 8.1 (16.3e54.6) 29.6 ± 6.5 (13.5e44.7) 5.6 ± 6.9 (11.6 to 36.6) 8.6 ± 5.5 (4.2 to 25) 17.8 ± 6.7 (2.1e35.8) 20.3 ± 6.1 (5.4e33.2) 12.8 ± 10.0 (14.2 to 38.2) 16.2 ± 7.9 (3.9e32.3) 59.7 ± 9.5 (15.3e87.3) 55.1 ± 7.1 (38e72.2) 94.9 ± 9.3 (72.3e120.6) 95.7 ± 8.5 (76e112.2)
0.610 0.320 0.004 0.230 <0.001 0.092 0.065 0.727 0.835 0.612 <0.001 0.017 0.216 0.009
672
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
Table 5 Sensitivity, specificity, PPV, and NPV of coxa profunda for acetabular overcoverage. [5] Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI)
All hips
Males
Females
57.7% (0.45e0.69)
41.5% (0.30e0.54)
78.9% (0.58e0.91)
56.8% (0.54e0.59)
74.8% (0.71e0.79)
39.4% (0.36e0.41)
21.7% (0.17e0.26)
37.0% (0.26e0.48)
16.3% (0.12e0.19)
86.6% (0.83e0.90)
78.2% (0.74e0.83)
92.6% (0.85e0.97)
[6]
[7]
[8]
PPV: positive predictive value, NPV: negative predictive value. [9]
all assessments were independent of the patients' radiographic data in a consecutive Japanese cohort. Second, we selected subjects under 50 years of age; therefore, we could not evaluate the association between the occurrence of coxa profunda and age. This is because we focused our study on younger patients who are more likely to be affected by FAI and to avoid the influence of degenerative change in the hips. Third, most of our subjects showed normal acetabular coverage, and the cohorts of dysplasia and acetabular overcoverage consisted of a small number of patients with mild deformities, resulting in a narrow distribution of measurement values. However, we consider that these cohorts were adequate for accomplishing our main purpose, that is, to survey a variety of hip morphological features among the general Japanese population. Fourth, we defined dysplasia as LCE < 25 and/or ARO > 15 , and acetabular overcoverage as LCE > 40 and/or ARO < 0 . Although several alternative cut-off values of dysplasia and overcoverage have been reported, no obvious consensus has been reached. These cut-off values have been commonly used for defining hip morphology types and are considered the most reasonable at present. In conclusion, our data revealed the prevalence of coxa profunda in Japanese subjects based on sex. The prevalence of coxa profunda is higher in women than in men. Coxa profunda may be less useful for diagnosing acetabular overcoverage because of its common occurrence with various hip morphologies and its poor specificity for detecting acetabular overcoverage. The presence of coxa profunda indicated significant greater acetabular coverage, especially in males. We should take into account that coxa profunda may have different implications between the sexes.
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
Conflicts of interest None. References €tzli H, Siebenrock KA. Femo[1] Ganz R, Parvizi J, Beck M, Leunig M, No roacetabular impingement: a cause for osteoarthritis of the hip. Clin Orthop Relat Res 2003 Dec;417:112e20. [2] Beck M, Kalhor M, Leunig M, Ganz R. Hip morphology influences the pattern of damage to the acetabular cartilage: femoroacetabular impingement as a cause of early osteoarthritis of the hip. J Bone Jt Surg Br 2005 Jul;87(7):1012e8. [3] Ganz R, Leunig M, Leunig-Ganz K, Harris WH. The etiology of osteoarthritis of the hip: an integrated mechanical concept. Clin Orthop Relat Res 2008 Feb;466(2):264e72. PE, Kim Y-J, Trousdale RT, Sierra RJ, Leunig M, [4] Clohisy JC, Carlisle JC, Beaule Schoenecker PL, Millis MB. A systematic approach to the plain radiographic
[22]
[23]
[24]
[25]
[26]
evaluation of the young adult hip. J Bone Jt Surg Am 2008 Nov;90(Suppl. 4): 47e66. Philippon MJ, Weiss DR, Kuppersmith DA, Briggs KK, Hay CJ. Arthroscopic labral repair and treatment of femoroacetabular impingement in professional hockey players. Am J Sports Med 2010 Jan;38(1):99e104. Tannast M, Siebenrock KA, Anderson SE. Femoroacetabular impingement: radiographic diagnosis e what the radiologist should know. AJR Am J Roentgenol 2007 Jun;188:1540e52. Weir A, de Vos RJ, Moen M, Holmich P, Tol JL. Prevalence of radiological signs of femoroacetabular impingement in patients presenting with long-standing adductor-related groin pain. Br J Sports Med 2011 Jan;45(1):6e9. Diesel CV, Ribeiro TA, Coussirat C, Scheidt RB, Macedo CAS, Galia CR. Coxa profunda in the diagnosis of pincer-type femoroacetabular impingement and its prevalence in asymptomatic subjects. Bone Jt J 2015 Apr;97-B(4): 478e83. Anderson LA, Kapron AL, Aoki SK, Peters CL. Coxa profunda: is the deep acetabulum overcovered? Clin Orthop Relat Res 2012 Dec;470(12):3375e82. Nepple JJ, Lehmann CL, Ross JR, Schoenecker PL, Clohisy JC. Coxa profunda is not a useful radiographic parameter for diagnosing pincer-type femoroacetabular impingement. J Bone Jt Surg Am 2013 Mar 6;95(5):417e23. Nicholls AS, Kiran A, Pollard TCB, Hart DJ, Arden CPA, Spector T, Gill HS, Murray DW, Carr AJ, Arden NK. The association between hip morphology parameters and nineteen-year risk of end-stage osteoarthritis of the hip: a nested case-control study. Arthritis Rheum 2011 Nov;63(11):3392e400. Fujii M, Nakamura T, Hara T, Nakashima Y, Iwamoto Y. Does radiographic coxa profunda indicate increased acetabular coverage or depth in hip dysplasia? Clin Orthop Relat Res 2015 Jun;473(6):2056e66. Hoaglund FT, Shiba R, Newberg AH, Leung KY. Diseases of the hip. A comparative study of Japanese Oriental and American white patients. J Bone Jt Surg Am 1985 Dec;67(9):1376e83. Jingushi S, Ohfuji S, Sofue M, Hirota Y, Itoman M, Matsumoto T, Hamada Y, Shindo H, Takatori Y, Yamada H, Yasunaga Y, Ito H, Mori S, Owan I, Fujii G, Ohashi H, Iwamoto Y, Miyanishi K, Iga T, Takahira N, Sugimori T, Sugiyama H, Okano K, Karita T, Ando K, Hamaki T, Hirayama T, Iwata K, Nakasone S, Matsuura M, Mawatari T. Multiinstitutional epidemiological study regarding osteoarthritis of the hip in Japan. J Orthop Sci 2010 Sep;15(5):626e31. Yoshimura N, Campbell L, Hashimoto T, Kinoshita H, Okayasu T, Wilman C, Coggon D, Croft P, Cooper C. Acetabular dysplasia and hip osteoarthritis in Britain and Japan. Br J Rheumatol 1998 Nov;37(11):1193e7. ry C, Larbi A, Lecouvet FE, Vande Berg BC, Omoumi P. Lepage-Saucier M, Thie Femoroacetabular impingement: normal values of the quantitative morphometric parameters in asymptomatic hips. Eur Radiol 2014 Jul;24(7):1707e14. Wiberg G. Shelf operation in congenital dysplasia of the acetabulum and in subluxation and dislocation of the hip. J Bone Jt Surg Am 1953 Jan;35-A(1): 65e80. Kopydlowski NJ, Tannenbaum EP, Bedi A, Smith MV, Sekiya JK. An increase in cranial acetabular version with age: implications for femoroacetabular impingement. J Arthroplasty 2014 Sep;29(9):1741e4. Anda S, Terjesen T, Kvistad KA, Svenningsen S. Acetabular angles and femoral anteversion in dysplastic hips in adults: CT investigation. J Comput Assist Tomogr 1991 JaneFeb;15(1):115e20. €nnis D, Heinecke A. Acetabular and femoral anteversion: relationship with To osteoarthritis of the hip. J Bone Jt Surg Am 1999 Dec;81(12):1747e70. Kang ACL, Gooding AJ, Coates MH, Goh TD, Armour P, Rietveld J. Computed tomography assessment of hip joints in asymptomatic individuals in relation to femoroacetabular impingement. Am J Sports Med 2010 Jun;38(6):1160e5. Chakraverty JK, Sullivan C, Gan C, Narayanaswamy S, Kamath S. Cam and pincer femoroacetabular impingement: CT findings of features resembling femoroacetabular impingement in a young population without symptoms. AJR Am J Roentgenol 2013 Feb;200(2):389e95. Corten K, Ganz R, Chosa E, Leunig M. Bone apposition of the acetabular rim in deep hips: a distinct finding of global pincer impingement. J Bone Jt Surg Am 2011 May;93(Suppl. 2):10e6. PE, Ramadan O, Doucette S. Prevalence of associated deAllen D, Beaule formities and hip pain in patients with cam-type femoroacetabular impingement. J Bone Jt Surg Br 2009 May;91(5):589e94. Boone G, Pagnotto MR, Walker JA, Trousdale RT, Sierra RJ. Radiographic features associated with differing impinging hip morphologies with special attention to coxa profunda. Clin Orthop Relat Res 2012 Dec;470(12):3368e74. Fujii M, Nakashima Y, Sato T, Akiyama M, Iwamoto Y. Pelvic deformity influences acetabular version and coverage in hip dysplasia. Clin Orthop Relat Res 2011 Jun;469(6):1735e42.
Contents lists available at ScienceDirect
Journal of Orthopaedic Science journal homepage: http://www.elsevier.com/locate/jos
Original article
Correlation between coxa profunda and morphological parameters of acetabular coverage in a Japanese cohort: A CT study Tomohiro Goto*, Kazuaki Mineta, Keizo Wada, Yasuaki Tamaki, Daisuke Hamada, Tomoya Takasago, Kosaku Higashino, Koichi Sairyo Department of Orthopedics, Institute of Biomedical Sciences, Tokushima University Graduate School, 3-18-15 Kuramoto, Tokushima 770-8503, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 December 2015 Received in revised form 22 May 2016 Accepted 26 June 2016 Available online 21 July 2016
Background: Coxa profunda is generally considered an indicator of acetabular overcoverage. However, recent studies have opposed this view. The correlation between coxa profunda and acetabular overcoverage thus remains controversial. The purpose of this study was to investigate the prevalence of coxa profunda and the association between coxa profunda and acetabular coverage based on sex in Japanese subjects using computed tomography. Methods: We reviewed the computed tomography scans of 151 Japanese consecutive patients (302 hips) aged <50 years who underwent abdominopelvic computed tomography for symptoms unrelated to hip disease. Coxa profunda was diagnosed when the floor of the acetabular fossa touched or was medial to the ilioischial line. We measured the lateral center edge angle, acetabular roof obliquity angle, acetabular version, and anterior and posterior acetabular sector angles as parameters of acetabular coverage. Results: Coxa profunda was observed in 45.7% of all subjects, and was more common in females (63.0%) than in males (29.5%). We found no significant differences in the incidence of coxa profunda among subjects with dysplasia, overcoverage, and normal coverage. Males with coxa profunda showed significantly greater lateral center edge angle, anterior acetabular sector angle, and lower acetabular roof obliquity angle, whereas anterior and posterior acetabular sector angles were greater in all subjects and females with coxa profunda. Coxa profunda showed poor sensitivity (57.7%) and specificity (56.8%) for detecting overcoverage in all subjects. Conclusions: Coxa profunda is a common radiographic feature in Japanese patients, especially in women. Coxa profunda may be less useful for diagnosing acetabular overcoverage because of its common occurrence with various hip morphologies and its poor specificity for detecting acetabular overcoverage. However, men with coxa profunda have greater acetabular coverage, so coxa profunda may contribute to acetabular coverage in men. We should take into account that coxa profunda may have different implications between the sexes. © 2016 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
1. Introduction Femoroacetabular impingement (FAI) has recently been widely recognized as a cause of hip pain in young adults. The pathological condition of FAI involves abnormal, repetitive abutment of the femur and acetabulum during hip motion and can cause labrum and cartilage injuries [1,2]. FAI is caused by morphological abnormalities such as small headeneck offset or aspherical femoral head in cam impingement, whereas acetabular overcoverage is present in
* Corresponding author. fax: þ81 88 633 0178. E-mail address: [email protected] (T. Goto).
pincer impingement [3]. To establish the appropriate therapeutic strategy for FAI, a comprehensive evaluation including clinical history, physical examination, and above all, radiographic evaluation, is essential. Coxa profunda, defined as the floor of the acetabular fossa touching or medial to the ilioischial line in the AP view of the pelvis, has been considered one of the diagnostic radiographic findings for acetabular overcoverage [2,4]. In conjunction with proximal femoral morphological parameters such as antetorsion, neck-shaft angle, and concavity, this may lead to pincer impingement [2e8]. However, in recent years, several reports have opposed a positive relationship between coxa profunda and acetabular overcoverage [8e11], suggesting that coxa profunda is a common radiographic finding in
http://dx.doi.org/10.1016/j.jos.2016.06.008 0949-2658/© 2016 The Japanese Orthopaedic Association. Published by Elsevier B.V. All rights reserved.
668
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
normal hips and may often be observed even in dysplastic hips [9,10,12]. Thus, whether coxa profunda can serve as a reliable radiographic indicator of acetabular overcoverage remains controversial. Furthermore, there are racial differences in hip morphology. It has been reported that primary hip osteoarthritis (OA) is common in Western countries; however, in Japan, the predominant type of hip OA is secondary OA due to hip dysplasia [13e15]. Thus, it is possible that the pathological basis of FAI differs in various ethnic populations, since FAI is based on morphological abnormalities and is noted to be a possible cause of early primary hip OA [1e3]. We believe that it is necessary to conduct detailed evaluations in each ethnic population to elucidate the pathogenesis of FAI, based on the morphological abnormalities, in hip disorders. In this study, we verified the prevalence of coxa profunda and the association between the presence of coxa profunda and radiographic parameters of acetabular coverage according to sex using computed tomography (CT) in Japanese subjects. Moreover, the sensitivity, specificity, positive predictive value, and negative predictive value of coxa profunda for acetabular overcoverage were assessed. 2. Materials and methods 2.1. Patients This retrospective study was approved by the institutional review board at our institution. We reviewed the CT scans of 155 Japanese consecutive patients (310 hips) under 50 years of age who underwent abdominal and pelvic CT for reasons unrelated to hip symptoms in our institution from September 2010 to June 2011. The 50-year-old age limit was set so that the study focused on relatively younger subjects who are susceptible to FAI and less influenced by degenerative changes in the hip joint. We excluded patients with avascular necrosis, hip OA, and/or previous hip surgery. OA was defined as the presence of at least one of the following: joint space narrowing, osteophytes, subchondral bone sclerosis, and subchondral bone cysts. After exclusion, 151 patient records (302 hips) were available for further analyses. We analyzed 156 male hips and 146 female hips, with a mean patient age of 37.8 years (range, 21e49 years). 2.2. Computed tomography evaluations All CT scans were ordered from the departments of general surgery, urology, and gynecology. Each CT was performed using a
16-detector-row CT scanner (Aquilion 16; Toshiba Medical Systems, Tokyo, Japan), and the acquired images were reconstructed into CT images. Observation was performed using a 1-mm slice thickness multiplanar reconstruction (MPR) image with Aquarius NET Server (TeraRecon, Inc., San Mateo, CA, USA) under standard bone settings. All measurements were performed with the patient in the supine position and 3DCT images were used to confirm the pelvic rotation by adjusting bilateral anterior superior iliac spine in the sagittal view. A single experienced orthopedic doctor reviewed all patient images one time. After that, he re-reviewed a subset of 50 subjects, which were selected randomly, and the other experienced orthopedic doctor evaluated those 50 subjects twice to verify the reliability of the radiographic analysis. They reviewed the images for the presence of coxa profunda and the measurement of radiological parameters pertaining to acetabular coverage. The diagnosis of coxa profunda was made when the floor of the acetabular fossa touched or was medial to the ilioischial line [2] on the coronal transparent three-dimensional model of the whole pelvis (simulating pelvis radiograph) (Fig. 1). We could obtain high contrast images from the coronal transparent 3D model by which accurate diagnosis of coxa profunda could be made. Recently, it has been more common using these images for morphological studies [16]. We measured the following parameters for evaluation of acetabular coverage: lateral center edge angle (LCE) [4,17], acetabular roof obliquity angle (ARO) [4], acetabular version (AV) [18], anterior acetabular sector angle (AASA), and posterior acetabular sector angle (PASA) [19]. We used these parameters to conduct a detailed analysis from multiple directions: LCE represents the lateral acetabular coverage, ARO shows the lateral opening of the acetabulum and lateral inclination of the weight-bearing surface, AV shows the anterior opening of the acetabulum, and AASA and PASA indicate the acetabular coverage at anterior and posterior aspects. LCE was determined by measuring the angle between two lines: a line through the center of the femoral head perpendicular to the transverse axis of the pelvis and a line connecting the center of the femoral head to the acetabular lateral margin (Fig. 2a). ARO was the angle formed by a horizontal line referenced on the pelvis (the line connecting the inferior aspects of the bilateral teardrops) and a line connecting the medial point of the sclerotic zone to the lateral edge of the sourcil (Fig. 2b). LCE and ARO were measured on a reformatted coronal image passing through the center of the femoral head. AV, AASA, and PASA are measurements of the anteroposterior acetabular coverage in an axial reformat. AV was obtained by measuring the angle between
Fig. 1. The coronal transparent three-dimensional model of the hip. Images of negative coxa profunda (a) and positive coxa profunda (b) are shown (white dotted line: the floor of the acetabular fossa, white line: the ilioischial line).
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
669
Fig. 2. The measurements of the lateral center edge angle (LCE) (a) and the acetabular roof obliquity angle (ARO) (b).
two lines: a line connecting the anterior and posterior acetabular rims and a line perpendicular to the coronal plane of the pelvis that extended anteriorly from the posterior wall (Fig. 3). Two AV measurements were made at the two separate transverse sections: central AV, a plane through the bilateral femoral heads (Fig. 3a); and cranial AV, a plane corresponding to a point 5 mm distal to the acetabular roof (Fig. 3b). We measured AASA and PASA by examining the angle made by a line connecting the center of the bilateral femoral head and lines from the center of the femoral heads to the anterior and posterior margins of the acetabulum (Fig. 4). We defined acetabular overcoverage as LCE > 40 and/or ARO < 0 , and acetabular dysplasia as LCE < 25 and/or ARO > 15 [4,6,8 20e22]. Other cases were defined as normal acetabular coverage.
of 39.8 years (range, 21e49 years) who were selected randomly. Measurements of all the radiological parameters were repeated two times each by two different observers, who were blinded to the results reported by the other observer, and the intraclass/interclass correlation coefficients (ICC) were assessed. Kappa coefficient was also performed for checking reproducibility in the diagnosis of coxa profunda. The intraclass/interclass correlation coefficients were almost in perfect agreement for all radiographic measurements (ICC: 0.88e0.97) (Table 1) and reproducibility in the diagnosis of coxa profunda was excellent (kappa value: 0.94). 3. Results 3.1. The prevalence of coxa profunda in Japanese subjects
2.3. Statistical analysis Statistical analyses were performed using SPSS software, version 22.0 (SPSS Inc., Chicago, IL, USA). The ManneWhitney Utest, chi-square tests applying Yates correction, and Fisher's exact test were used to compare patients with and without coxa profunda, male and female subjects, and to examine the prevalence of coxa profunda for statistical analysis. Statistical differences were considered significant when p < 0.05. The sensitivity and specificity of coxa profunda with respect to acetabular overcoverage were reported with 95% binomial confidence intervals. The reliability of the radiographic analysis was assessed by determining the inter- and intra-observer reproducibility in 50 patients; 22 male hips and 28 female hips, with a mean patient age
Coxa profunda was observed in 45.7% (138 of 302 hips) of all the subjects. Coxa profunda was much more common in females (63.0%) than in males (29.5%) (p < 0.001) (Table 2). Of the 302 hips, 39 hips showed dysplasia (defined as LCE < 25 and/or ARO > 15 ), 60 hips had overcoverage (defined as LCE > 40 and/or ARO < 0 ), and 204 hips showed normal coverage (Table 2). We found no significant differences in the incidence of coxa profunda between each of these three cohorts (p ¼ 0.431). However, in each cohort, significant differences were noted for overcoverage and normal coverage (p ¼ 0.029 and <0.001, respectively) with respect to the prevalence of coxa profunda between the sexes. In the dysplasia group, the incidence of coxa profunda tended to be high in females; however, no statistical significance was noted. The highest
Fig. 3. Acetabular version (AV) measurements. Two measurements were made at the two separate transverse sections: (a) central AV, a plane through the bilateral femoral heads; (b) cranial AV, a plane corresponding to a point 5 mm distal to the acetabular roof.
670
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
Fig. 4. The measurements of the anterior acetabular sector angle (AASA) and the posterior acetabular sector angle (PASA).
Table 1 Reliability of the radiographic parameters. Parameters
Examiner
Intra-observer reliability ICC (1,2)
Inter-observer reliability ICC (2,2)
Lateral center edge angle
TG KM TG KM TG KM TG KM TG KM TG KM
0.96 0.90 0.97 0.96 0.94 0.93 0.95 0.96 0.88 0.93 0.92 0.92
0.94
Acetabular roof obliquity Central anteversion angle Cranial anteversion angle Anterior acetabular sector angle Posterior acetabular sector angle
0.97 0.96 0.97 0.91 0.91
prevalence of coxa profunda was 78.6% in females with acetabular overcoverage (Table 2). 3.2. The association between the presence of coxa profunda and radiographic parameters of acetabular overcoverage by sex using computed tomography In all the subjects, although there were no significant differences in age, LCE, ARO, or AV (both central and cranial) between the hips Table 2 Prevalence of coxa profunda.
Total subjects (N ¼ 302) Dysplasia (N ¼ 39) Overcoverage (N ¼ 60) Normal coverage (N ¼ 204) Comparison among three cohorts (p value) a b c
3.3. The sensitivity, specificity, positive predictive value, and negative predictive value of coxa profunda for acetabular overcoverage The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of coxa profunda as an indicator of acetabular overcoverage are shown in Table 5. In all the subjects, the sensitivity, specificity, and PPV were relatively low but NPV was high. The values were different between the sexes; coxa profunda showed a relatively high specificity but low PPV for acetabular overcoverage in males, whereas it showed high sensitivity and NPV in females. 4. Discussion
Prevalence of coxa profunda (%) All hips
with and without coxa profunda, the AASA and PASA in hips with coxa profunda were significantly greater than in those without coxa profunda (Table 3). Further, significant sex differences in several radiographic parameters were noted (Table 4). Males with hips with coxa profunda showed significantly greater LCE (p ¼ 0.004) and AASA (p < 0.001) but lower ARO (p < 0.001), whereas in females, AASA and PASA in hips with coxa profunda were greater than those in hips without coxa profunda. According to CT measurement values, males showed relatively greater acetabular coverage compared with females. Although the AASA in females with coxa profunda was greater than in those without coxa profunda, its mean value was lower than the AASA in males without coxa profunda (Table 4).
Males
Females
Comparison between sexes (p value)
45.7% (138/302) 29.5% (46/156) 63.0% (92/146) <0.001a 43.6% (17/39)
23.1% (3/13)
53.8% (14/26)
0.093b
53.8% (28/52)
42.1% (16/38)
78.6% (11/14)
0.029b
44.1% (93/211)
25.7% (27/105) 63.2% (67/106) <0.001a
0.431a
0.232c
Chi-square test. Fisher's exact test. Yates continuity corrected Chi-square test.
0.484c
In this study, we identified coxa profunda in 138 of 302 hips (45.7%), and large differences were observed in its incidence between males (29.5%) and females (63.0%) (p < 0.001). Moreover, we found no significant differences in the prevalence of coxa profunda between the three cohorts; further, no significant differences were noted in dysplasia (43.6%), overcoverage (53.8%), and normal acetabular coverage (44.1%) in all hips (p ¼ 0.431), and in the hips of males (p ¼ 0.232) or females (p ¼ 0.484). Our results suggested that coxa profunda is commonly observed in various types of acetabular coverage as well as in normal acetabular coverage, and its occurrence is higher in females in Japanese subjects. There were several previous reports that discussed the prevalence of coxa profunda. According to studies from countries outside of Asia (Europe, North and South America) involving FAI subjects, the prevalence of coxa profunda ranges from 27% to 77.4% [2,5,8e10,23]. The prevalence rates in our study were indeed within this range, thus, it seems that
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672 Table 3 CT measurement values between hips with and without coxa profunda.
Age (years) Lateral center edge angle ( ) Acetabular roof obliquity ( ) Central anteversion angle ( ) Cranial anteversion angle ( ) Anterior acetabular sector angle ( ) Posterior acetabular sector angle ( )
Coxa profunda (N ¼ 138) mean ± SD (range)
Without coxa profunda (N ¼ 164) mean ± SD (range)
p Value
37.8 ± 8.4 (20e49) 33.1 ± 7.4 (17.0e53.3) 5.3 ± 6.5 (11.5 to 24.3) 19.1 ± 6.4 (0.7e33.9) 15.3 ± 9.8 (16.5 to 44.3) 60.5 ± 8.3 (39.3e79.8) 98.2 ± 7.9 (75.6e116.1)
37.9 ± 8.2 (22e49)
0.957
31.6 ± 7.2 (13.5e54.6)
0.172
6.6 ± 6.6 (11.6 to 36.6)
0.053
18.6 ± 6.6 (2.1e35.8)
0.556
14.0 ± 9.4 (14.2 to 38.2)
0.147
58.2 ± 9.1 (15.3e87.3)
0.041
95.2 ± 9.0 (72.3e120.6)
<0.001
there were no ethnic differences in the prevalence of coxa profunda. Anderson et al. (a report from USA) found a prevalence of 34% in asymptomatic football players (male patients); 56% in patients undergoing periacetabular osteotomy; and 58% in patients with surgical dislocation [9]. Nepple et al. (a report from USA) estimated the prevalence of coxa profunda in four cohorts, namely, those with dysplasia (41%), FAI (64%), Perthes deformities (31%), and asymptomatic individuals (76%, highest); the average prevalence was 55% [10]. They also observed a strong sex difference, being 70% in females and 24% in males in the cohorts. These values were similar to our results. Although the patients' backgrounds are different among the studies, it seems that there were little ethnic differences in the prevalence of coxa profunda; a considerable high prevalence of coxa profunda may be observed in both USA and Japanese populations. Our study revealed a nearly two times higher prevalence of coxa profunda in females. Coxa profunda is a radiographic character defined as fovea projection relative to the ilioischial line in the AP view but the actual coronal plane of the fovea and the ilioischial line is different in three dimensions. Therefore, the relative positional relationship between the fovea and ilioischial line can be changed by anatomical variations such as the width of the pelvis (ischium). We attribute the sex difference to the relatively wider pelvis in females than in males. In terms of the relationship between the presence of coxa profunda and acetabular depth, several authors have found no significant differences in the LCE and ARO between cohorts with and without coxa profunda among asymptomatic and symptomatic subjects [3,10]. On the other hand, Allen et al. showed significantly
671
greater LCE in hips of FAI subjects with coxa profunda (average, 38.7 ) than in those without coxa profunda (average, 33.6 ) [24]. Another report demonstrated significant associations for LCE and ARO with the presence of coxa profunda (LCE, 36.2 ; ARO, 2.8 with coxa profunda; and LCE, 32.6 ; ARO, 5.7 without coxa profunda) [25]. In the present study, we observed statistically significant differences in several radiographic parameters: AASA and PASA in all subjects and females as well as LCE, ARO, and AASA in males. These results indicate that cases of coxa profunda tended to have greater acetabular coverage, particularly in male subjects, and yet the average of each LCE, ARO, and AASA of men with coxa profunda was still within normal range. This is because our subjects were extracted from the general population who underwent CT examination unrelated to hip disorders. Thus, most of our patients showed normal hip morphology. Although our results showed greater AASA and PASA in female subjects, there were no differences in LCE and ARO, which are commonly used to define acetabular coverage; moreover, the mean AASA in females with coxa profunda was lower than that in males without coxa profunda. Thus, the correlation of coxa profunda with acetabular coverage may be weak in female subjects. Fujii et al. retrospectively reviewed Japanese dysplastic hips and suggested that hips with increased AV and an inwardly rotated pelvis tended to have coxa profunda [12]. Their results differ partially from ours. We assume that the severity of acetabular dysplasia lead to these differences, because severe dysplasia results in dramatic changes in pelvic morphology, including rotation of the innominate bone that influences AV [26]. The dysplasia in the abovementioned study [12] was far more severe (average LCE, 8.6 ; ARO, 23 ) than in our study (average LCE, 21.2 ; ARO, 13.8 ). We found that coxa profunda showed poor sensitivity (57.7%) and specificity (56.8%) in determining acetabular overcoverage in all subjects in our study. With respect to sex differences, coxa profunda in males showed relatively high specificity (74.8%) but low PPV (37.0%). PPV does not always synchronize with specificity when the morbidity rate is low, and the number of our male subjects of acetabular overcoverage with coxa profunda was small. On the other hand, in females, coxa profunda had high sensitivity (78.9%) and NPV (92.6%) but poor specificity (39.9%). Thus, coxa profunda cannot be an indicator of acetabular overcoverage; however, the high sensitivity and NPV in females indicate that the use of coxa profunda for excluding acetabular overcoverage may be feasible in female patients. Our study has certain limitations. First, we had no information regarding the patients' backgrounds, including clinical symptoms or medical history. This was intended to exclude any biases; therefore,
Table 4 CT measurement values between sexes with and without coxa profunda.
Number of hips Age (years) Lateral center edge angle ( ) Acetabular roof obliquity ( ) Central anteversion angle ( ) Cranial anteversion angle ( ) Anterior acetabular sector angle ( ) Posterior acetabular sector angle ( )
Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female
Coxa profunda (N ¼ 138) mean ± SD (range)
Without coxa profunda (N ¼ 164) mean ± SD (range)
p Value
46 92 37.2 ± 9.4 (22e49) 38.1 ± 7.9 (20e49) 36.0 ± 7.2 (20.6e53.3) 31.6 ± 7.1 (17e50.6) 1.8 ± 5.2 (6.9 to 16) 7.0 ± 6.5 (11.5 to 24.3) 15.7 ± 5.0 (6.1e28.9) 20.7 ± 3.4 (0.7e33.9) 12.6 ± 9.1 (8.1 to 31.8) 16.7 ± 9.9 (16.5 to 44.3) 65.0 ± 7.8 (49.0e79.8) 58.3 ± 7.7 (39.3e78.6) 96.2 ± 8.1 (75.6e110.8) 99.2 ± 7.7 (76.4e116.1)
110 54 38.4 ± 8.1 (22e49) 37.0 ± 8.4 (22e49) 32.5 ± 8.1 (16.3e54.6) 29.6 ± 6.5 (13.5e44.7) 5.6 ± 6.9 (11.6 to 36.6) 8.6 ± 5.5 (4.2 to 25) 17.8 ± 6.7 (2.1e35.8) 20.3 ± 6.1 (5.4e33.2) 12.8 ± 10.0 (14.2 to 38.2) 16.2 ± 7.9 (3.9e32.3) 59.7 ± 9.5 (15.3e87.3) 55.1 ± 7.1 (38e72.2) 94.9 ± 9.3 (72.3e120.6) 95.7 ± 8.5 (76e112.2)
0.610 0.320 0.004 0.230 <0.001 0.092 0.065 0.727 0.835 0.612 <0.001 0.017 0.216 0.009
672
T. Goto et al. / Journal of Orthopaedic Science 21 (2016) 667e672
Table 5 Sensitivity, specificity, PPV, and NPV of coxa profunda for acetabular overcoverage. [5] Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI)
All hips
Males
Females
57.7% (0.45e0.69)
41.5% (0.30e0.54)
78.9% (0.58e0.91)
56.8% (0.54e0.59)
74.8% (0.71e0.79)
39.4% (0.36e0.41)
21.7% (0.17e0.26)
37.0% (0.26e0.48)
16.3% (0.12e0.19)
86.6% (0.83e0.90)
78.2% (0.74e0.83)
92.6% (0.85e0.97)
[6]
[7]
[8]
PPV: positive predictive value, NPV: negative predictive value. [9]
all assessments were independent of the patients' radiographic data in a consecutive Japanese cohort. Second, we selected subjects under 50 years of age; therefore, we could not evaluate the association between the occurrence of coxa profunda and age. This is because we focused our study on younger patients who are more likely to be affected by FAI and to avoid the influence of degenerative change in the hips. Third, most of our subjects showed normal acetabular coverage, and the cohorts of dysplasia and acetabular overcoverage consisted of a small number of patients with mild deformities, resulting in a narrow distribution of measurement values. However, we consider that these cohorts were adequate for accomplishing our main purpose, that is, to survey a variety of hip morphological features among the general Japanese population. Fourth, we defined dysplasia as LCE < 25 and/or ARO > 15 , and acetabular overcoverage as LCE > 40 and/or ARO < 0 . Although several alternative cut-off values of dysplasia and overcoverage have been reported, no obvious consensus has been reached. These cut-off values have been commonly used for defining hip morphology types and are considered the most reasonable at present. In conclusion, our data revealed the prevalence of coxa profunda in Japanese subjects based on sex. The prevalence of coxa profunda is higher in women than in men. Coxa profunda may be less useful for diagnosing acetabular overcoverage because of its common occurrence with various hip morphologies and its poor specificity for detecting acetabular overcoverage. The presence of coxa profunda indicated significant greater acetabular coverage, especially in males. We should take into account that coxa profunda may have different implications between the sexes.
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
Conflicts of interest None. References €tzli H, Siebenrock KA. Femo[1] Ganz R, Parvizi J, Beck M, Leunig M, No roacetabular impingement: a cause for osteoarthritis of the hip. Clin Orthop Relat Res 2003 Dec;417:112e20. [2] Beck M, Kalhor M, Leunig M, Ganz R. Hip morphology influences the pattern of damage to the acetabular cartilage: femoroacetabular impingement as a cause of early osteoarthritis of the hip. J Bone Jt Surg Br 2005 Jul;87(7):1012e8. [3] Ganz R, Leunig M, Leunig-Ganz K, Harris WH. The etiology of osteoarthritis of the hip: an integrated mechanical concept. Clin Orthop Relat Res 2008 Feb;466(2):264e72. PE, Kim Y-J, Trousdale RT, Sierra RJ, Leunig M, [4] Clohisy JC, Carlisle JC, Beaule Schoenecker PL, Millis MB. A systematic approach to the plain radiographic
[22]
[23]
[24]
[25]
[26]
evaluation of the young adult hip. J Bone Jt Surg Am 2008 Nov;90(Suppl. 4): 47e66. Philippon MJ, Weiss DR, Kuppersmith DA, Briggs KK, Hay CJ. Arthroscopic labral repair and treatment of femoroacetabular impingement in professional hockey players. Am J Sports Med 2010 Jan;38(1):99e104. Tannast M, Siebenrock KA, Anderson SE. Femoroacetabular impingement: radiographic diagnosis e what the radiologist should know. AJR Am J Roentgenol 2007 Jun;188:1540e52. Weir A, de Vos RJ, Moen M, Holmich P, Tol JL. Prevalence of radiological signs of femoroacetabular impingement in patients presenting with long-standing adductor-related groin pain. Br J Sports Med 2011 Jan;45(1):6e9. Diesel CV, Ribeiro TA, Coussirat C, Scheidt RB, Macedo CAS, Galia CR. Coxa profunda in the diagnosis of pincer-type femoroacetabular impingement and its prevalence in asymptomatic subjects. Bone Jt J 2015 Apr;97-B(4): 478e83. Anderson LA, Kapron AL, Aoki SK, Peters CL. Coxa profunda: is the deep acetabulum overcovered? Clin Orthop Relat Res 2012 Dec;470(12):3375e82. Nepple JJ, Lehmann CL, Ross JR, Schoenecker PL, Clohisy JC. Coxa profunda is not a useful radiographic parameter for diagnosing pincer-type femoroacetabular impingement. J Bone Jt Surg Am 2013 Mar 6;95(5):417e23. Nicholls AS, Kiran A, Pollard TCB, Hart DJ, Arden CPA, Spector T, Gill HS, Murray DW, Carr AJ, Arden NK. The association between hip morphology parameters and nineteen-year risk of end-stage osteoarthritis of the hip: a nested case-control study. Arthritis Rheum 2011 Nov;63(11):3392e400. Fujii M, Nakamura T, Hara T, Nakashima Y, Iwamoto Y. Does radiographic coxa profunda indicate increased acetabular coverage or depth in hip dysplasia? Clin Orthop Relat Res 2015 Jun;473(6):2056e66. Hoaglund FT, Shiba R, Newberg AH, Leung KY. Diseases of the hip. A comparative study of Japanese Oriental and American white patients. J Bone Jt Surg Am 1985 Dec;67(9):1376e83. Jingushi S, Ohfuji S, Sofue M, Hirota Y, Itoman M, Matsumoto T, Hamada Y, Shindo H, Takatori Y, Yamada H, Yasunaga Y, Ito H, Mori S, Owan I, Fujii G, Ohashi H, Iwamoto Y, Miyanishi K, Iga T, Takahira N, Sugimori T, Sugiyama H, Okano K, Karita T, Ando K, Hamaki T, Hirayama T, Iwata K, Nakasone S, Matsuura M, Mawatari T. Multiinstitutional epidemiological study regarding osteoarthritis of the hip in Japan. J Orthop Sci 2010 Sep;15(5):626e31. Yoshimura N, Campbell L, Hashimoto T, Kinoshita H, Okayasu T, Wilman C, Coggon D, Croft P, Cooper C. Acetabular dysplasia and hip osteoarthritis in Britain and Japan. Br J Rheumatol 1998 Nov;37(11):1193e7. ry C, Larbi A, Lecouvet FE, Vande Berg BC, Omoumi P. Lepage-Saucier M, Thie Femoroacetabular impingement: normal values of the quantitative morphometric parameters in asymptomatic hips. Eur Radiol 2014 Jul;24(7):1707e14. Wiberg G. Shelf operation in congenital dysplasia of the acetabulum and in subluxation and dislocation of the hip. J Bone Jt Surg Am 1953 Jan;35-A(1): 65e80. Kopydlowski NJ, Tannenbaum EP, Bedi A, Smith MV, Sekiya JK. An increase in cranial acetabular version with age: implications for femoroacetabular impingement. J Arthroplasty 2014 Sep;29(9):1741e4. Anda S, Terjesen T, Kvistad KA, Svenningsen S. Acetabular angles and femoral anteversion in dysplastic hips in adults: CT investigation. J Comput Assist Tomogr 1991 JaneFeb;15(1):115e20. €nnis D, Heinecke A. Acetabular and femoral anteversion: relationship with To osteoarthritis of the hip. J Bone Jt Surg Am 1999 Dec;81(12):1747e70. Kang ACL, Gooding AJ, Coates MH, Goh TD, Armour P, Rietveld J. Computed tomography assessment of hip joints in asymptomatic individuals in relation to femoroacetabular impingement. Am J Sports Med 2010 Jun;38(6):1160e5. Chakraverty JK, Sullivan C, Gan C, Narayanaswamy S, Kamath S. Cam and pincer femoroacetabular impingement: CT findings of features resembling femoroacetabular impingement in a young population without symptoms. AJR Am J Roentgenol 2013 Feb;200(2):389e95. Corten K, Ganz R, Chosa E, Leunig M. Bone apposition of the acetabular rim in deep hips: a distinct finding of global pincer impingement. J Bone Jt Surg Am 2011 May;93(Suppl. 2):10e6. PE, Ramadan O, Doucette S. Prevalence of associated deAllen D, Beaule formities and hip pain in patients with cam-type femoroacetabular impingement. J Bone Jt Surg Br 2009 May;91(5):589e94. Boone G, Pagnotto MR, Walker JA, Trousdale RT, Sierra RJ. Radiographic features associated with differing impinging hip morphologies with special attention to coxa profunda. Clin Orthop Relat Res 2012 Dec;470(12):3368e74. Fujii M, Nakashima Y, Sato T, Akiyama M, Iwamoto Y. Pelvic deformity influences acetabular version and coverage in hip dysplasia. Clin Orthop Relat Res 2011 Jun;469(6):1735e42.