髋关节撞击合并临界髋臼发育不良的手术治疗失败的预测放射学参数（2023）Predictive radiological parameters of failure following surgical management of femoroacetabular impingement associated with borderline acetabular dysplasia
Foissey C, Abid H, Martinot P, Cazor A, Thaunat M. Predictive radiological parameters of failure following surgical management of femoroacetabular impingement associated with borderline acetabular dysplasia[J]. Orthop Traumatol Surg Res, 2023,109(4): 103349.
转载文章的原链接1：
https://pubmed.ncbi.nlm.nih.gov/35688380/
转载文章的原链接2：
https://www.sciencedirect.com/science/article/pii/S1877056822001761?via%3Dihub
Abstract
Introduction
The role of arthroscopic treatment of femoroacetabular impingement (FAI) in mild or borderline hip dysplasia (lateral center edge angle = 18–25) is controversial. It is recommended to combine capsular plication with femoral neck osteoplasty and suture repair of the labrum. Few studies have investigated which radiological parameters are associated with failure of this procedure.
关节镜治疗轻度或临界髋关节发育不良（外侧中心边缘角= 18-25）的髋关节撞击（FAI）的作用是有争议的。建议将关节囊plication与股骨颈成形术及盂唇缝合修复相结合。很少有研究调查了哪些放射学参数与手术失败有关。
Hypothesis
A larger number of radiological signs of mild or borderline hip dysplasia than radiological signs of FAI negatively influence the outcomes of arthroscopic treatment.
轻度或临界髋关节发育不良的放射学征象多于FAI的放射学征象，对关节镜治疗的结果有负面影响。
Methods
This was a retrospective study done with data collected prospectively after a minimum of 2 years’ follow-up in patients who underwent arthroscopic treatment of FAI due to cam impingement combined with mild to moderate hip dysplasia. Patients with hip osteoarthritis graded as ≥ Tönnis 2 were excluded. Functional outcome scores (mHHS and NAHS) were determined along with the need for reoperation. Three groups were defined based on the outcomes: success (mHHS ≥ Patient Acceptable Symptomatic State [PASS]); moderate improvement (improvement in mHHS + mHHS ＜ PASS) and failure (worsening of the mHHS). Demographic data, radiographic parameters (LCEA, Tönnis angle, alpha angle, offset, FEAR index, Shenton‘s line, Cliff sign, anterior wall index [AWI], posterior wall index [PWI]) and intraoperative findings were compared between these three groups. The area under the curve (AUC) for the most relevant data identified during the univariate analysis were then modeled to define which factors were the best at predicting failed arthroscopic treatment a posteriori.
这是一项回顾性研究，对因凸轮撞击合并轻度至中度髋关节发育不良而接受关节镜治疗的FAI患者进行了至少2年的前瞻性随访。分级≥Tönnis 2的髋关节骨关节炎患者被排除在外。功能结局评分(mHHS和NAHS)与再次手术的需要一起确定。根据结果分为三组:成功(mHHS≥患者可接受症状状态[PASS]);中度改善(mHHS改善+ mHHS ＜ PASS)和失败(mHHS恶化)。比较三组患者的人口学资料、影像学参数(LCEA、Tönnis角、α角、offset、FEAR指数、Shenton线、Cliff征、前壁指数[AWI]、后壁指数[PWI])及术中表现。然后对在单变量分析中确定的最相关数据的曲线下面积(AUC)进行建模，以确定哪些因素最能预测后验关节镜治疗失败。
Results
The study analyzed 39 patients. The mean mHHS was 76 ± 15 [40–92]. The “success” group consisted of 21 patients (56%, 21/39), the “moderate improvement” group of 12 patients (31%, 12/39) and the failure group of 6 patients (13%, 6/39) (2 subsequently underwent total hip replacement, 4 underwent shelf acetabuloplasty). The FEAR index was significantly higher and the AWI and VCE were significantly lower in the “failure” group. The FEAR index was the best predictive factor; an index ≥ 4° detected 100% of failures with 96% specificity. The patients in the “moderate improvement” group were significantly older (37 years ± 8 [18–45] ＜ p= 0.04) with a significantly higher offset (2 mm ± 3 [−5–5] ＜ p = 0.004); an offset ≥ 2 mm could detect 73% of them with a specificity of 72%.
该研究分析了39名患者。平均mHHS为76±15[40-92]。“成功”组21例(56%，21/39)，“中度改善”组12例(31%，12/39)，失败组6例(13%，6/39)(2例随后行全髋关节置换术，4例行髋臼置换术)。“失败”组的FEAR指数显著升高，AWI和VCE显著降低。FEAR指数为最佳预测因子;指数≥4°检出率为100%，特异性为96%。“中度改善”组患者年龄明显增加(37岁±8岁[18-45]＜ p= 0.04)，偏移量明显增加(2 mm±3[−5-5]＜ p= 0.004); offset≥2 mm可检出73%，特异性为72%。
Conclusion
A FEAR index ≥ 4° and offset ≥ 2 mm is best able to detect patients at risk of failure and unsatisfactory results, respectively, in the mild clinical cam-type impingement in a population of FAI patients.
在FAI患者群体mild clinical cam-type impingement中，FEAR指数≥4°和offset≥2 mm分别最能检测出有失败风险的患者和结果不满意的患者。
Level of evidence
IV, retrospective series.
Keywords
Femoroacetabular impingement，Micro-instability，Borderline dysplasia， FEAR index，Anterior wall index
1. Introduction
The role of arthroscopy for the treatment of hip pain in the context of mild or borderline acetabular dysplasia (lateral center edge angle [LCEA] = 18–25° [1] ) is still controversial. While the clinical and radiographic findings provide evidence in favor of femoro-acetabular impingement (FAI) (Fig. 1 ), the challenge is in distinguishing between the contribution of micro-instability (which causes lesions of the chondrolabral junction, from inside to out) and a cam effect (which causes lesions of the chondrolabral junction, from outside to in) [2].
关节镜在轻度或临界髋臼发育不良（外侧中心边缘角[LCEA] = 18-25°[1]）髋部疼痛治疗中的作用仍存在争议。虽然临床和影像学结果提供了支持髋关节撞击（FAI）的证据（图1），但难点在于如何区分微不稳定性（由内到外导致关节软骨连接处病变）和凸轮效应（由外到内导致关节软骨连接处病变） [2]
Fig. 1. Patients who has cam-type impingment in their right hip with moderate dysplasia (LCEA=18̊).
Micro-instability at the hip is a recently described pathology that combines to varying amounts insufficient acetabular coverage, excessive femoral anteversion and ligamentous hyperlaxity [3]. The arthroscopic treatment consists of doing capsular plication [1] but the results are controversial [4]. The role of a surgical procedure to correct the insufficient bone coverage (shelf acetabuloplasty, periacetabular osteotomy) as the first-line procedure or after failed arthroscopic treatment has yet to be defined [5], [6].The femoral cam‘s contribution to the spectrum of symptoms is difficult to identify, especially in patients whose range of motion in flexion and internal rotation are less affected than in FAI of a hip with normal radiological angles [7]. Along with potentially disappointing outcomes, the risk of arthroscopic surgery in this indication [8], [9], [10] is that it will destabilize the hip joint [11].
髋关节微不稳定是最近被描述的一种病理，它结合了不同程度的髋臼覆盖不足、股前倾过度和韧带过度松弛[3]。关节镜下治疗包括囊膜plication [1]，但结果存在争议[4]。矫正骨覆盖不足（髋臼成形术、髋臼周围截骨术）的外科手术作为一线手术或在关节镜治疗失败后的作用尚未明确[5]，[6]。股骨凸轮对症状谱的影响难以确定，特别是在屈曲和内旋运动范围受影响程度小于正常放射角度髋关节FAI的患者中[7]。除了可能令人失望的结果外，关节镜手术治疗该适应症的风险[8]，[9]，[10]是它会使髋关节失稳[11]。
Our clinical experience appears to confirm these mixed results. Initially, we wanted to analyze the outcome (function and pain) of hip arthroscopy for FAI in patients with borderline hip dysplasia. At that time, the radiographic criteria used to differentiate between micro-instability and FAI were based on standard measurements of the lateral-center edge (LCE) angle for acetabular coverage and the Tönnis angle (aka acetabular index). We wanted to analyze “after the fact” to what extend the radiographic parameters that were recently introduced to differentiate micro-instability [12] could help us better select candidate patients to improve our results.
我们的临床经验似乎证实了这些混杂的结果。最初，我们想分析伴有临界髋关节发育不良的FAI患者的髋关节镜检查结果(功能和疼痛)。当时，用于区分微不稳定性和FAI的放射学标准是基于髋臼覆盖的外侧中心边缘(LCE)角和Tönnis角(又称髋臼指数)的标准测量。我们想分析“事后”如何扩展最近引入的用于区分微不稳定性的放射学参数[12]，以帮助我们更好地选择候选患者以改善我们的结果。
The first hypothesis was that our overall results were worse than those described in the FAI literature for hips without dysplasia. The second hypothesis was that failures would be related to mild signs of FAI (lower volume femoral cam effect) and/or a predominance of radiological signs of micro-instability.
第一个假设是，对于没有发育不良的髋关节，我们的总体结果比FAI文献中描述的要差。第二种假设是，失败可能与轻度FAI症状(lower volume股骨凸轮效应)和/或微不稳定的放射学症状有关。
2. Material and methods
2.1. Patients
This was a descriptive, retrospective, single-center, single-surgeon (MT) study of data collected prospectively. All patients who underwent arthroscopic treatment of cam-type FAI between January 2016 and February 2019 and who had borderline hip dysplasia (LCE = 18–25°) [12] were included if they had at least 2 years of follow-up. The surgical indication was made based on the combination of clinical signs of impingement (groin pain found during FADIR [Flexion Adduction Internal Rotation] and/or FABER [Flexion Abduction External Rotation] maneuvers) and radiographic signs of impingement (modified head–neck offset [mHNO] ＜5 mm [13], alpha angle of Nötzli ＞55° [12]). Patients who had hip osteoarthritis graded ≥ Tönnis 2, who did not have hip dysplasia (LCE ＞25°) or who had severe hip dysplasia (LCE ＜18°) were excluded.
[13] Espié A, Elia F, MurgierJ, Chiron P, Chaput B. Modified head-neck offset for diagnosing anterior femoro-acetabular impingement. Int Orthop 2016;40:687–95
这是一项描述性、回顾性、单中心、单外科医生(MT)的前瞻性数据研究。所有在2016年1月至2019年2月期间接受关节镜治疗的cam型FAI患者，并且伴有临界髋关节发育不良(LCE = 18-25°)[12]，如果随访时间至少为2年。手术指征是根据撞击的临床症状(FADIR[屈曲内收内旋]和/或FABER[屈曲外展外旋]操作中发现的腹股沟疼痛)和撞击的影像学症状(改良头颈offset [mHNO] ＜5 mm [13]， α角Nötzli ＞55°[12])确定的。排除髋关节骨关节炎分级≥Tönnis 2、未发生髋关节发育不良(LCE ＞25°)或严重髋关节发育不良(LCE ＜18°)的患者。
2.2. Data collection
Demographics, pain intensity (visual analog scale, VAS) and functional outcome scores (modified Harris Hip Score [mHHS] [14], Non Arthritic Hip Score [NAHS] [15]) were collected during the preoperative consultation. An experienced hip surgeon evaluated the criteria described by McClincy et al. [12] (Fig. 2) on AP views of the pelvis: LCE angle, angle of Tönnis, FEAR (Femoro-Epiphyseal Acetabular Roof) index, anterior wall index (AWI) and posterior wall index (PWI) [16], Cliff sign [17] and Shenton‘s line [18]. The alpha angle (Nötzli) was measured on a 45° Dunn view along with the mHNO (a negative value corresponds to a cervical bump) [13]. CT arthrography or MRI was requested routinely to look for labral or cartilaginous lesions.
[16] Hiza E, Dierckman BD, Guanche C, Applegate G, Shah D, Ryu JH. Reliability of the tönnis classification and its correlation with magnetic resonance imaging and intraoperative chondral damage. Arthroscopy 2019;35:403–8
[17] PackerJD, CowanJB, Rebolledo BJ, Shibata KR, Riley GM, Finlay AK, et al. The cliff
sign: a new radiographic sign of hip instability. Orthop J Sports Med 2018;6
[13] Espié A, Elia F, MurgierJ, Chiron P, Chaput B. Modified head-neck offset for diagnosing anterior femoro-acetabular impingement. Int Orthop 2016;40:687–95
Fig. 2. Radiographic parameters measured for this study.
Intraoperatively, data collected consisted of the presence of cartilage lesions and chondrolabral lesions (Beck classification [19]), along with suturing of the labrum.
Postoperative radiographs were taken and used to measure the alpha angle and mHNO. All patients were called at the final assessment to collect information about complications, reoperations, pain intensity (VAS), mHHS, NAHS, satisfaction level (where 0 is very disappointed and 10 is extremely satisfied) and to complete missing data.
2.3. Surgical technique
The surgical indication was made after at least 3 months of conservative treatment had failed, which consisted of physiotherapy, modified activities, intra-articular corticosteroid injection and a course of anti-inflammatory drugs. The surgical procedure was carried out using the “capsulotomy first” technique [20]. A step-by-step longitudinal capsulotomy in the same direction as the fibers was done. The first intra-articular step consisted of exploring the peripheral compartment followed by the central compartment under traction. Traction was then released to carry out resection of the femoral cam using a motorized burr, while checking for the absence of residual impingement in flexion and internal rotation with fluoroscopy. Capsular plication was done at the end of the procedure; the margins of the capsule were trimmed back 5-7 mm then closed using non-absorbable suture to retension the soft tissues (Fig. 3).
Fig. 3. Capsuloplasty for retensioning: standard capsulotomy along the fiber axis at the start of the procedure (A) then resection of the capsular margins by “orange” segments using a coagulation probe (B), side to side suture with Scorpion™ suture passer using 2 to 4 throws of non-absorbable suture (D). (C=capsule, FH=femoral head, L=labrum).
2.4. Rehabilitation protocol
All patients were operated in the ambulatory setting and were provided rehabilitation by a physiotherapist. Full weight bearing was allowed right away, with crutches used as needed to control pain. The patients were prescribed non-steroidal anti-inflammatory drugs for 15 days to prevent heterotopic ossifications. Active hip flexion was prohibited for the first 2 weeks to prevent the development of psoas tendinopathy. Also, prohibited were flexion ≥ 80° and forceful hip extension for the first 2 weeks, internal rotation for the first 3 weeks and external rotation for the first 6 weeks. Gradual return to sport was allowed after 3 months.
2.5. Statistics
For postoperative pain, the PASS (Patient Acceptable Symptomatic State) was set at 2 [21], the MCID (Minimal Clinically Important Difference) was an improvement of more than 2 points [22]. For the mHHS, the PASS was set at 74 [23], [24] and the MCID at 18 [25], [26]. For the NAHS, the MCID and PASS were calculated by our facility‘s biostatistics unit: PASS of 69, MCID of 19. Failure was defined as postoperative mHHS ＜ 74. Three groups of patients were defined based on the improvement in mHHS: the “success” group had an improvement in the mHHS + mHHS post-op ≥ PASS, the “moderate improvement” group had an improvement in the mHHS + mHHS post-op ＜ PASS and the “failure” group had worsening of the mHHS.
Continuous variables were described by their mean value, standard deviation and minimum and maximum values. Their distribution was evaluated using a Shapiro-Wilk test. The following statistical tests were used for these variables: Student‘s t test when the variable of interest was independent and normally distributed; paired Student‘s t test when the variable of interest was paired, independent and normally distributed; the Mann-Whitney test for independent variables that were not normally distributed, and the Wilcoxon signed rank test for paired variables that were not normally distributed. Qualitative variables were summarized as percentages. A Chi2 test or Fisher test was used for these.
The area under the curve (AUC) for the parameters found to be relevant during the univariate analysis were modeled to evaluate their ability to detect failure or conversion. For each parameter, the AUC and the best threshold using the Youden method [27] were calculated. Parameters with AUC ＜0.70 were considered as inaccurate, between 0.70 and 0.90 as having moderate accuracy and ＞0.90 as having very good accuracy [28]. The Bonferroni method was used to compare the various AUC values [29].
The significance threshold was set at 5%. The statistical analysis was performed using XLSTAT™ software (version 2021, AddInsoft, Paris, France).
2.6. Ethics
All procedures done for the study that involved human participants followed the ethical standards of the national and/or institutional review board and the 1964 Helsinki declaration and its subsequent modifications or comparable ethical standards.
3. Results
During the study period, 195 hips were operated with arthroscopy for FAI. Among them, 43 hips (41 patients) met the inclusion criteria. Four patients (9%, 4/41) were lost to follow-up and excluded from the analysis. The mean follow-up was 34 months ± 9.7 [24–57]. The demographic data are provided in Table 1
.
Table 1. Demographics, preoperative clinical features, preoperative radiographic measurements and intraoperative findings in this study population.
Empty Cell Total (n = 39)
Preoperative demographics
Sex (women) 12 (31%)
Age (years) 31 ± 10 [15–47]
Left/Right 24 (62%)/15 (38%)
BMI 23 ± 2 [20–27]
Preoperative clinical data
FADIR + 39 (100%)
FABER + 36 (92%)
Preoperative imaging
Preop offset (mm) 1.5 ± 2.6 [−5–5]
LCE angle (°) 23 ± 3 [18–25]
Tönnis (°) 9.2 ± 3 [5–15]
Alpha angle (°) 65 ± 10 [55–90]
Femoral neck–shaft angle (°) 132 ± 5 [124–146]
FEAR index (°) −9.8 ± 11 [−28–8.9]
AWI (%) 34 ± 11 [10–53]
PWI (%) 84 ± 14 [53–111]
Cliff sign 5 (13%)
Intraoperative findings
Acetabular cartilage lesions:
Number 36 (92%)
Classification 2.2 ± 1 [0–4]
Labral damage:
Number 29 (74%)
Classification 1.6 ± 1.2 [0–4]
Femoral cartilage lesions:
Number 2 (5%)
Classification 0.2 ± 0.7 [0–4]
Repair of labrum 22 (56%)
BMI:body mass index; LCE: lateral-center edge; FEAR: Femoro-Epiphyseal Acetabular Roof; AWI: anterior wall index; PWI: posterior wall index
During the intraoperative assessment, 36 hips (92%, 36/39) had acetabular cartilage lesions, 2 (5%, 2/39) had femoral head cartilage lesions and 29 (74%, 29/39) had labral tears, of which 22 (75%, 22/39) were repaired by suturing (Table 1). The postoperative radiographs found an alpha angle ＜55° in all patients with a mean of 46 ± 4 [38–50].
At the final review, all the outcome scores (NAHS, mHHS, pain) had improved significantly (Table 2). Twenty-one hips (54%, 21/39) had a mHHS at the final review ≥ to the PASS and were allocated to the “success” group. Conversely, 12 hips (31%, 12/39) had an improvement in the mHHS that was less than the PASS, thus were considered in the “moderate improvement” group. In six hips (15%, 6/39) the symptoms were worse; they represented the “failure” group. All six failures underwent revision surgery: four underwent shelf acetabuloplasty with a satisfactory outcome; two underwent conversion to total hip replacement (THR) at another facility. The mean time to reoperation was 29 months ± 11 [14-38]
.
Table 2. Comparison of pre- and postoperative outcome scores.
Empty Cell Preoperative (n = 39) Postoperative (n = 39) p value
NAHS [26]
Mean ± SD [min-max] 63 ± 15 [35–73] 78 ± 15 [50–99] 0.0003
Number ≥ PASS 22 (56%)
Number ≥ MCID 13 (33%)
mHHS [14]
Mean ± SD [min-max] 55 ± 14 [19–72] 76 ± 15 [40–92] ＜0.0001
Number ≥ PASS 21 (54%)
Number ≥ MCID 23 (59%)
Pain (VAS)
Mean ± SD [min-max] 7.1 ± 1.6 [1–9] 2.8 ± 2.2 [0–7] ＜0.0001
Number ≤ PASS 18 (46%)
Number ≥ MCID 29 (74%)
Bold values indicate significant differences.
NAHS: Non-Arthritic Hip Score; mHHS: modified Harris hip score; VAS: visual analog scale
Compared to the “success” group, the “failure” group had a higher proportion of women (66%, p = 0.04) and had more parameters suggestive of micro-instability (Table 3): LCE = 20° ± 2 [18-21] (p = 0.002), FEAR index= 6.9 ± 1 [6-8.9] (p = 0.0003) and AWI= 0.21 ± 0.08 [0.1-0.33] (p = 0.003). The angle of Tönnis and the Cliff sign were not significantly related to the risk of failure, and none had a break in Shenton‘s line. The FEAR index was the most accurate with an AUC of 0.99 (Fig. 4a) and an ideal threshold of 4°, which corresponded to a Se= 100% and Sp= 96%. The AWI was the second most accurate parameter with an AUC of 0.92, and threshold of 0.23 (Se = 83%, Sp = 89%). The LCE angle had a moderate accuracy with AUC of 88% and threshold of 21° (Se = 100%, Sp = 78%). There was no significant difference between these curves.
Table 3. Comparison of the “failure” and “moderate improvement” groups with the “success” group.
Empty Cell Success (mHHS≥74) (n = 21) Failure
(n = 6) p value Moderate improvement
(n = 12) p value
Demographics
Sex (women) 4 (19%) 4 (66%) 0.04 4 (33%) 0.42
Age (years) 29 ± 10 [15–45] 28 ± 11 [15–47] 0.72 36.8 ± 8 [18–45] 0.04
BMI 23 ± 2 [20–27] 23 ± 1 [20–24] 0.43 24 ± 2 [20–27] 0.76
Preoperative scores
NAHS [26]
64 ± 14 [38–73] 56 ± 16 [40–70] 0.007 64 ± 17 [35–72] 0.85
mHHS [14]
59 ± 13 [47–72] 45 ± 4 [40–49] ＜.0002 52 ± 16 [19–70] 0.195
Pain (VAS) 6.5 ± 1.9 [1–8] 7.5 ± 0.5 [7,8] 0.326 7.8 ± 0.7 [6–9] 0.01
Preoperative imaging
Tönnis stage
Stage 0 16 (76%) 3 (50%) 0.63 6 (50%) 0.29
Stage 1 7 (24%) 3 (50%) 6 (50%)
Preop offset (mm) 0.7 ± 2 [2–5] 3 ± 2 [0–5] 0.128 2 ± 3 [–5–5] 0.004
LCE (°) 23 ± 2 [18–25] 20 ± 2 [18–21] 0.002 23 ± 3 [18–25] 0.97
Tönnis (°) 9.4 ± 3 [5–15] 11 ± 2 [10–15] 0.21 9.5 ± 4 [5–16] 0.99
Alpha angle (°) 66 ± 10 [55–90] 63 ± 12 [55–80] 0.52 64 ± 8 [55–78] 0.65
Femoral neck–shaft angle (°) 132 ± 5 [125–146] 132 ± 4 [126–137] 0.83 131 ± 5 [124–139] 0.77
FEAR index (°) −13.6 ± 8 [−28-2.1] 6.9 ± 1 [6-8.9] 0.0003 −13 ± 8 [−22-6.5] 0.88
AWI 0.34 ± 0.09 [0.21–0.49] 0.21 ± 0.08 [0.1-0.33] 0.003 0.41 ± 0.11 [0.24-0.53] 0.14
PWI 0.82 ± 0.14 [0.53-1.02] 0.85 ± 0.14 [0.73-1.11] 0.83 0.87 ± 0.17 [0.58-1.06] 0.31
Cliff sign 3 (14%) 0 (0%) 1 2 (16%) 1
Intraoperative findings
Acetabular cartilage lesions:
Number 21 (100%) 6 (100%) 1 9 (75%) 0.04
Classification 2.5 ± 0.8 [1–4] 2 ± 0 [2] 0.66 1.6 ± 1.2 [0–3] 0.75
Labral damage:
Number 16 (76%) 5 (83%) 1 8 (67%) 0.45
Classification 1.8 ± 1.4 [0–4] 1.8 ± 1.2 [0–3] 0.42 1.2 ± 1 [0–3] 0.03
Femoral cartilage lesions:
Number 1 (5%) 0 1 1 (8%) 1
Classification 0.2 ± 0.9 [0–4] 0 0.8 0.2 ± 0.6 [0–2] 0.18
Repair of labrum 10 (48%) 3 (50%) 1 9 (72%) 0.16
Satisfaction (/10) 8 ± 1 [6–9] 1 ± 1 [0–2] ＜0.0001 6 ± 1 [2–7] ＜0.0001
Bold values indicate significant differences. * relative to the “success” group.
mHHS: modified Harris Hip Score; BMI: body mass index; LCE: lateral-center edge angle; FEAR: Femoro-Epiphyseal Acetabular Roof; AWI: anterior wall index; PWI: posterior wall index
Fig. 4. ROC curves with the area under the curve (AUC) and the best thresholds according to the Youden method and the associated sensitivity (Se), specificity (Sp), positive predictive value (PPV) and negative predictive value (NPV): A. For the detection of conversions (FEAR index, LCEA and AWI). B. For the detection of failures (modified head-neck offset).
Compared with the “success” group, the “moderate improvement” group included older patients (37 years ± 8 [18-45] p ＜ 0.04) and the preoperative mHNO was larger (2 ± 3 [−5–5] p = 0.004). The value of the alpha angle was not related to these disappointing results. The mHNO value provided moderate accuracy to distinguish these hips, with an AUC of 0.70 and threshold of 2 mm (Se = 73%, Sp = 72%) (Fig. 4b).
4. Discussion
With a 56% success rate, the results of femoral cam resection combined with capsular plication for the treatment of hip impingement presenting with mild to moderate acetabular dysplasia are considered as inadequate. The good outcomes achieved after shelf acetabuloplasty in these patients tend to confirm that the origin of the residual pain is not related to the typical causes of failure of arthroscopic treatment of hip impingement (intra-articular adhesions, advanced chondral lesions, insufficient resection of the femoral cam) [30-32] but actually to the hip‘s micro-instability component, requiring a corrective bone procedure. Recent advances in the preoperative assessment of these patients, particularly the description of new radiological parameters (Fear Index, AWI), will help to better define the risk factors for failure of arthroscopic treatment in this population. Our study confirms the relevance of the FEAR index, which was able to detect 100% of patients who needed another bone procedure beyond the initial femoral neck osteoplasty.
股骨凸轮切除联合capsular plication治疗伴有轻至中度髋臼发育不良的髋关节撞击的成功率为56%，但结果被认为是不充分的。这些患者髋臼置换术后取得的良好结果表明，残留疼痛的起源与关节镜治疗髋关节撞击失败的典型原因(关节内粘连、软骨病变晚期、股骨凸轮切除不足)无关[30-32]，而实际上与髋关节的微不稳定成分有关，需要进行骨矫正手术。这些患者术前评估的最新进展，特别是对新的放射学参数(Fear Index，AWI)的描述，将有助于更好地确定这类人群中关节镜治疗失败的危险因素。我们的研究证实了Fear Index的相关性，该指数能够100%地检测出在最初的股骨颈成形术之外需要另一次骨手术的患者。
A recent meta-analysis of four articles comparing the treatment of FAI in patients with and without borderline acetabular dysplasia found few differences in the postoperative scores or the reoperation rate between the two groups at less than 2 years’ follow-up [33]. The mean mHHS was 81, similar to the one in this study; however the PASS and MCID were not disclosed, which makes it difficult to interpret the relative success and failure of that study. Conversely, the reoperation rate was lower than in our study (2.2%), likely due to better preoperative patient selection that excluded patients with pure micro-instability. Other studies had much higher reoperation rates: Fukui et al. had a conversion rate of 25% [34] and Hatakeyama et al. of 24% [35]. After the fact, each of these studies found a predominance of signs suggestive of micro-instability (break in Shenton‘s line, VCA ≤ 17°, angle of Tönnis ≥ 15°) or advanced lesions (residual joint space＜ 2 mm, Tönnis grade 1 hip osteoarthritis, severe intraoperative lesions).
最近的一项荟萃分析对4篇比较伴有和不伴有临界髋臼发育不良患者FAI治疗的文章进行了分析，发现在随访不到2年的情况下，两组患者的术后评分或再手术率几乎没有差异[33]。平均mHHS为81，与本研究相似;然而，PASS和MCID没有披露，这使得很难解释该研究的相对成功和失败。相反，再手术率低于我们的研究(2.2%)，这可能是由于术前患者选择更好，排除了单纯微不稳定的患者。其他研究的再手术率要高得多:Fukui等的转换率为25% [34]，Hatakeyama等的转换率为24%[35]。事实上，这些研究都发现了提示微不稳定(Shenton线断裂，VCA≤17°，Tönnis角度≥15°)或晚期病变(残留关节间隙＜ 2mm, Tönnis级髋关节骨关节炎，严重术中病变)的优势迹象。
Failures were predicted with high precision and consisted of patients who had predominantly micro-instability. The FEAR index was first described in 2017 par Wyatt et al. [36]. They defined a threshold of 5° to best detect hips with micro-instability in patients who have borderline dysplasia (Se = 78%, Sp = 80%). Other teams subsequently analyzed this index; all agreed on the major benefit of using it but each proposed a different threshold: −5° (Se = 29%, Sp = 92%) for Truntzer et al. [37], −1.3° (Se = 90%, Sp = 89%) for Smith et al. [38] and 2° (Se = 85%, Sp = 95%) for Batailler et al. [39]. The variation in these thresholds is in part due to the population studied, along with the definition of the gold standard. In our study, the 4° threshold (Se = 100%, Sp = 96%) is close to the threshold in Wyatt‘s original study [36]. The accuracy parameters found here must only be interpreted in the context of a population in which FAI is present concurrently with borderline dysplasia to detect patients who may not improve with arthroscopic treatment.
失败的预测精度很高，主要由微不稳定的患者组成。2017年，par Wyatt等人首次描述了FEAR指数[36]。他们定义了5°的阈值，以最好地检测临界发育不良患者的微不稳定髋关节(Se = 78%， Sp = 80%)。其他团队随后分析了这个指数;所有人都同意使用它的主要好处，但每个人都提出了不同的阈值:Truntzer等人的- 5°(Se = 29%， Sp = 92%) [37]， Smith等人的- 1.3°(Se = 90%， Sp = 89%)[38]和Batailler等人的2°(Se = 85%， Sp = 95%)[39]。这些阈值的变化部分是由于所研究的人口以及definition of the gold standard。在我们的研究中，4°阈值(Se = 100%， Sp = 96%)接近Wyatt原始研究中的阈值[36]。此处发现的准确性参数必须仅在FAI与临界发育不良同时存在的人群中进行解释，以检测可能无法通过关节镜治疗改善的患者。
The AWI was first described in 2012 par Siebenrock et al. [40] who found a mean value of 0.28 in patients with hip dysplasia. Since then, it has mainly been used to characterize the anterior coverage in patients who have severe dysplasia ＜18° with a view to acetabular reorientation surgery [41], [42] but not very often to characterize micro-instability in a hip with moderate dysplasia. The 0.23 threshold proposed here could also be a relevant criterion for the need to perform a bone procedure to correct the insufficient coverage, in addition to the femoral neck osteoplasty for FAI.
[40] Siebenrock KA, Kistler L, Schwab JM, Büchler L, Tannast M. The acetabular wall index for assessing anteroposterior femoral head coverage in symptomatic patients. Clin Orthop Relat Res 2012;470:3355–60,
[41] Stetzelberger VM, Leibold CS, Steppacher SD, Schwab JM, Siebenrock KA,Tannast M. The acetabular wall index is associated with long-term conversion to THA after PAO. Clin Orthop Relat Res 2021;479:1052–65,
[42] Wylie JD, Ferrer MG, McClincy MP, Miller PE, Millis MB, Kim YJ,et al. What is the reliability and accuracy of intraoperative fluoroscopy in evaluating anterior lateral, and posterior coverage during periacetabular osteotomy? Clin Orthop Relat Res 2019;477:1138–44,
2012年par Siebenrock等人首次描述了AWI[40]，他们发现髋关节发育不良患者的AWI平均值为0.28。从那时起，该方法主要用于鉴定＜18°严重发育不良患者髋臼再定位手术的前覆盖范围[41]，[42]，但不常用于鉴定中度发育不良患者髋关节的微不稳定性。除了股骨颈成形术治疗FAI外，本文提出的0.23阈值也可能是需要进行骨手术以纠正覆盖不足的相关标准。
The “moderate improvement” group was harder to predict; an offset of more than 2 mm in our population was able to predict 73% of this outcome. This is a lower threshold than the one found by Espié et al. [13] which seems logical in a population that has acetabular dysplasia. It is also interesting to note that offset was a better predictor of moderate improvement than alpha angle. This finding echoes other studies that found that using the offset provided better diagnostic accuracy [13] and that its value had better correlation with clinical findings [43] than that of the alpha angle. As for age, Fukui et al. [34] had already recommended being very careful with patients over 35 years of age in this indication, which corresponds to the mean age of patients in the “moderate improvement” group in our study.
“中度改善”组更难预测;在我们的人群中，超过2毫米的offset能够预测73%的结果。这一阈值低于espieet al.[13]发现的阈值，这在髋臼发育不良人群中似乎是合乎逻辑的。同样有趣的是，offset比α角更能预测中度改善。这一发现与其他研究相呼应，这些研究发现，使用offset提供了更好的诊断准确性[13]，其值与临床表现的相关性优于α角[43]。年龄方面，Fukui等[34]已经建议35岁以上患者在该适用证中要非常小心，这与我们研究中“中度改善”组患者的平均年龄相对应。
While there were no major complications in our patients, especially true iatrogenic postoperative instability [44], it appears that the arthroscopic capsule re-tensioning procedure that we performed in the direction of the fibers does not provide adequate treatment of the micro-instability component of these hips, which remain painful despite treatment of the femoral cam. For this indication, this brings into question whether a procedure to retension the iliofemoral ligament via an interportal capsulotomy as described by Domb et al. could be effective [45]. With this technique, the authors do not prohibit use of arthroscopy in the context of hip dysplasia; they recommend combining it when doing a bone procedure on the acetabular side to diagnose and treat any intra-articular lesions, and to address the femoral cam [46,47].
虽然我们的患者没有出现重大并发症，特别是真正的医源性术后不稳定[44]，但我们在纤维方向进行的关节镜下囊再张紧术似乎不能充分治疗这些髋关节的微不稳定成分，尽管治疗了股骨cam，但仍存在疼痛。对于这一适应症，Domb等人所描述的通过interportal capsulotomy来保留髂股韧带的手术是否有效成为了一个问题[45]。有了这项技术，作者不禁止在髋关节发育不良的情况下使用关节镜;他们建议在髋臼侧进行骨手术时将其结合起来诊断和治疗任何关节内病变，并解决股骨CAM问题[46,47]。
The main limitation of our study is the small number of patients, which may have led to a lack of statistical power to detect certain predictive factors such as the Cliff sign [17]. However, our sample size is in the range of other published studies on this topic; two literature reviews included studies with a mean of 52 patients with hip dysplasia [33], [48]. The low statistical power did not affect our ability to detect differences during univariate testing but limited our ability to compare ROC curves. The second limitation was that we did not analyze certain data that have previously been shown to be relevant in this indication: hypermobility based on the Beighton score, clinical maneuvers to test for instability (hip dial test, PART [Prone Apprehension Relocation Test]) and the anterior coverage angle [3]. Unfortunately, these parameters were not always recorded in the study period. The third limitation is the measurement error for the parameters of interest; however, the main studies on these parameters have found high reliability and satisfactory reproducibility [36], [40]. Lastly, although it was the method most often used in the literature, the measurements were made on plain radiographs, not on imaging by slice, which some authors have stated is more accurate [3], [49]
. 我们研究的主要局限性是患者数量较少，这可能导致在检测某些预测因素(如Cliff征)方面缺乏统计能力[17]。然而，我们的样本量在这个主题的其他已发表研究的范围内;两篇文献综述平均纳入了52例髋关节发育不良患者[33,48]。低统计能力不影响我们在单变量检验中检测差异的能力，但限制了我们比较ROC曲线的能力。第二个限制是，我们没有分析先前被证明与该适应症相关的某些数据:基于Beighton评分的过度活动、测试不稳定性的临床操作(髋关节刻度测试、PART[俯卧忧虑移位测试])和前关节覆盖角[3]。不幸的是，这些参数在研究期间并不总是被记录下来。第三个限制是感兴趣参数的测量误差;然而，对这些参数的主要研究都发现了较高的可靠性和令人满意的重现性[36,40]。最后，尽管这是文献中最常用的方法，但测量是在X线平片上进行的，而不是在切片成像上进行的，一些作者认为切片成像更准确[3,49]
5. Conclusion
A FEAR index ≥ 4° and offset ≥ 2 mm are best able to detect patients at high risk of instability-related failure or moderate improvement, respectively, due to a mild cam-type impingement in the population with FAI. Routinely measuring these parameters will help to improve the overall results in this indication.
在FAI患者中，FEAR指数≥4°和offset≥2 mm分别最能检测出不稳定相关failure或中度改善的高风险患者，这是由于轻微的凸轮型撞击造成的。常规测量这些参数将有助于改善该适适用证的总体结果。