Improved preoperative planning and arthroscopic technique in the treatment of femoroacetabular impingement

Full Text

INTRODUCTION

Femoroacetabular impingement (FAI) is a significant cause of pain, functional limitation, and reduced quality of life in young and middle-aged patients [1, 2]. The morphological basis of this syndrome is abnormal contact between the proximal femur and the acetabular rim, resulting in damage to the acetabular labrum, articular cartilage, and other structures of the hip joint [1, 2]. Hip arthroscopy is one of the main surgical treatment options for FAI and provides clinical improvement in a substantial proportion of patients [3, 4, 5]. However, the outcomes of arthroscopic treatment for FAI remain variable [5]. This variability may be related not only to patient-specific characteristics but also to differences in the surgical techniques used during the key stages of the procedure. Consequently, further improvement of arthroscopic treatment for FAI has been associated with the optimization of these critical operative stages [6].

Among the most debated aspects of hip arthroscopy are the type of capsulotomy per-formed, the method of acetabular labral fixation, and the accuracy of cam deformity correction [7, 8, 9, 10, 11]. Excessive capsulotomy may result in capsular defects, persistent postoperative pain, and reduced hip joint stability [9, 10]. The choice of labral treatment strategy may influence both morphological and clinical outcomes [8]. Cam deformity correction also remains challenging, as achieving target quantitative para-meters does not always reflect the restoration of the optimal head-neck offset geometry, while both under-resection and over-resection may adversely affect treatment outcomes [6, 12].

Despite the presence of literature addressing capsulotomy, labral refixation, and cam deformity correction, data regarding the morphological and clinical outcomes associated with simultaneous modification of several key stages of arthroscopic surgery remain limited.

The aim of the study — to compare morphological and clinical outcomes of arthroscopic treatment in patients with femoroacetabular impingement managed with standard versus modified surgical strategy.

METHODS

Study design

A single-center comparative cohort study was conducted, including two consecutive clinical cohorts of patients with FAI who underwent hip arthroscopy at the Vreden National Medical Research Center of Traumatology and Orthopedics between February 2018 and February 2025.

Patients

The study included 88 patients who underwent a total of 94 hip arthroscopic procedures. The retrospective cohort consisted of 61 patients (64 hips) who underwent surgery between February 2018 and March 2022. The prospective cohort included 27 patients (30 hips) who underwent surgery between November 2022 and February 2025. Bilateral procedures were per-formed in three patients in each cohort.

The inclusion and exclusion criteria were identical for both groups. Patients with clinically significant symptoms of FAI and cam deformity located in the anterosuperior head-neck junction of the femur on imaging studies were eligible for inclusion [13, 14]. Exclusion criteria included acetabular retroversion with a retroversion index greater than 30%, acetabular dysplasia with a Wiberg angle of less than 20°, avascular necrosis of the femoral head, Legg-Calvé-Perthes disease, previous hip surgery, and signs of degenerative joint disease. Degenerative changes were defined as grade III-IV articular cartilage damage according to the Outerbridge classification based on intraoperative assessment and/or MRI findings, acetabular cystic changes identified on X-ray, CT, or MRI scans, joint space narrowing ≤ 3 mm, and/or joint incongruity. An additional exclusion criterion was the absence of patient-reported outcome measures obtained either preoperatively or at the 12-month postoperative follow-up.

Surgical technique

In all cases, the operations were performed through standard arthroscopic portals [15]. In the retrospective cohort, joint access was obtained primarily through an interportal capsulotomy; in two cases, the capsulotomy was extended to a T-shaped configuration. Following the completion of the main surgical procedure, capsular closure was performed in all patients. In the presence of an acetabular labral tear, acetabular rim preparation and labral refixation were carried out using both knotted and knotless suture anchors. Cam deformity correction was performed using a conventional technique aimed at achieving target postoperative values of the alpha angle and offset ratio [11].

In the prospective cohort, joint access was obtained exclusively through a longitudinal capsulotomy, followed by routine capsular closure. When a labral tear was present, labral refixation was performed using knotless suture anchor fixation only. Cam deformity correction was carried out according to the developed method of preoperative planning and intraoperative execution that has been submitted as a patent application (priority certificate obtained, re-gistration No 2026102357). Using a Dunn 45° radiographic view, target values of the alpha angle and head-neck offset ratio were determined preoperatively. The planned resection zone was then defined graphically, and key parameters for its intraoperative implementation were calculated (Figure 1). During surgery, a reference landmark was created at the apex of the cam deformity, and stepwise contouring osteoplasty was performed toward the medial and lateral synovial folds until a smooth, spherical transition between the femoral head and neck was restored [12, 13, 14].

 

Figure 1. Application of the improved algorithm for cam deformity correction: a — Dunn 45° X-ray with lines corresponding to the preoperative and target values of the alpha angle and offset ratio; b — determination of the planned resection zone; c — postoperative Dunn 45° X-ray demonstrating alpha angle and offset ratio lines within the reference range, with restoration of the spherical contour of the femoral head

 

Outcome assessment

Radiographic assessment

Hip X-rays in the Dunn 45° position were obtained preoperatively and on the first post-operative day. Preoperative and postoperative X-rays were used to assess the alpha angle and the head-neck offset ratio [16]. The extent of resection following femoroplasty was classified according to the criteria proposed by Y. Mansor et al. as under-resection, adequate resection, or over-resection [12]. Over-resection was defined as positioning of the bony contour within the resection zone inside the ideal circle constructed from the contour of the femoral head, with a depth of over-resection exceeding 5% of the femoral head diameter. Under-resection was defined as persistence of the bony contour outside the ideal circle proximal to the point corresponding to an alpha angle of 50°. In addition, the shape of the femoroplasty was evaluated. The contour was classified as spherical when a constant curvature and parallel relationship to the ideal circle were maintained. Conversely, it was classified as non-spherical when local flattening, asymmetry, straight segments, angular irregularities, or regions of reverse curvature were present (Figure 2).

 

Figure 2. Examples of a non-spherical contour after femoroplasty: a — an area of reverse curvature with localized concavity; b — localized flattening/step-off of the contour

 

Magnetic resonance imaging

Follow-up MRI examinations were performed in both groups without contrast enhancement using 3.0-T scanners according to a standardized imaging protocol [17, 18]. The median time to follow-up MRI was 13.9 [13.2; 19.8] months in the retrospective group and 15.3 [13.9; 22.9] months in the prospective group (p = 0.397).

Capsular integrity was assessed on proton density fat-suppressed (PD-FS) images with focused evaluation of the capsulotomy site [18, 19]. Three categories of capsular healing were identified: capsular defect, partial healing, and complete healing. A capsular defect was defined as persistent discontinuity of the capsule with slit-like separation of the capsular edges and extension of synovial fluid into the defect. Partial healing was defined as incomplete restoration of capsular fiber continuity. Complete healing was defined as restoration of uninterrupted capsular continuity throughout the entire postoperative region without the evidence of a residual defect. For additional analysis, cases were grouped into two categories: intact capsule (complete healing) and non-intact capsule (partial healing or capsular defect).

Changes associated with acetabular labral anchors were evaluated specifically at the an-chor fixation sites. Such changes were recorded when focal areas of increased signal intensity on PD-FS images and/or fluid-signal collections were present within the paralabral soft tissues, as well as in cases demonstrating thickening and heterogeneity of the capsulolabral complex surrounding the fixation zone [17, 18].

Clinical assessment

Clinical outcomes were assessed preoperatively and at 12 months after surgery. Patient-reported outcome measures (PROMs) included the International Hip Outcome Tool-33 (iHOT-33) and the Hip Outcome Score (HOS), with addi-tional analysis of the HOS-ADL (Activities of Daily Living) subscale, reflecting functional per-formance in everyday activities, and the HOS-Sport subscale, reflecting sports participation and higher levels of physical activity [20, 21]. The achievement of the minimal clinically important difference (MCID) and substantial clinical benefit (SCB) was determined according to previously published threshold values for the respective outcome measures. For the iHOT-33, the MCID was 6.1 points and the SCB was 24.5 points. For the HOS-ADL, the MCID and SCB thresholds were 9.0 and 10.0 points, respectively. For the HOS-Sport, the MCID was 6.0 points and the SCB was 29.9 points [22, 23].

Postoperative management

All patients were mobilized on the first postoperative day with restrictions on weight-bearing through the operated limb. The use of crutches was recommended with partial weight-bearing limited to approximately 15-20% of body weight during the first 3 weeks, followed by gradual progression to full weight-bearing over the subsequent 2 weeks.

Statistical analysis

Preliminary data processing was performed in Microsoft Excel (Microsoft Corp., USA). Statistical analyses were conducted using SPSS Statistics v. 26.0 software (IBM, USA), as well as the Python 3.13.5 programming language with the use of the SciPy 1.17.0, statsmodels 0.14.6, and NumPy 2.3.5 libraries for specialized analyses. All analyses were performed at the level of individual hips/procedures (n = 94). Quantitative variables are presented as median and interquartile range (Me [Q₁; Q₃]), whereas qualitative variables are presented as n (%; 95% confidence interval, CI). Exact 95% CI for proportions were calculated using the Clopper-Pearson method.

The distribution of quantitative variables was assessed using the Shapiro-Wilk test and by visual inspection of histograms and Q-Q plots. Between-group comparisons of quantitative variables were performed using the Mann-Whitney U test, while within-group pre- and postoperative comparisons were conducted using the Wilcoxon signed-rank test. For paired changes, the Hodges-Lehmann estimator (HL) and its 95% CI were additionally calculated. Categorical variables in 2×2 contingency tables were compared using the Fisher’s exact test, whereas contingency tables containing three or more categories were analyzed using the chi-square test. When statistically significant differences were identified in tables with three or more categories, post-hoc pairwise comparisons of proportions were performed with Bonferroni correction.

For continuous clinical outcomes, reduced linear regression models were constructed. The dependent variables were the 12-month improvements in iHOT-33, HOS-ADL, and HOS-Sport scores. Three prespecified morphological predictors were included in the models: non-intact capsular status, non-spherical femoro-plasty morphology, and the presence of at least one MRI-detected abnormality in the region of acetabular labral anchors. Models were additionally adjusted for the corresponding preoperative score. Heteroscedasticity-robust HC3 standard errors were used to calculate 95% CI and p-values. To identify factors associated with clinically meaningful binary outcomes, multivariable logistic regression models were applied. Failure to achieve the respective clinically meaningful threshold was considered the event of interest. Given the limited number of events and the risk of model overfitting, logistic regression analyses were performed only for outcomes with at least 30 events. Multicollinearity among predictors was assessed during model development. Results of linear regression analyses are reported as β coefficients, whereas logistic regression re-sults are presented as adjusted odds ratios (aOR) with 95% CI. An aOR > 1 indicated a higher likelihood of failing to achieve the corresponding clinically meaningful outcome threshold. All statistical tests were two-sided, and a p-value < 0.05 was considered statistically significant.

RESULTS

Preoperative characteristics of the groups

Prior to surgery, the patient groups did not differ in most demographic, clinical, and radiological characteristics; a statistically significant inter-group difference was found only in the HOS-ADL scale, the value of which was lower in the prospective group (Table 1).

 

Table 1. Preoperative characteristics of the groups

Parameter	Retrospective group (n = 64)	Prospective group (n = 30)	p
Cam-type FAI	14 (21.9%; 12.5-34.0)	9 (30.0%; 14.7-49.4)	0.445
Mixed-type FAI	50 (78.1%; 66.0-87.5)	21 (70.0%; 50.6-85.3)	0.445
Age at surgery, years	37.8 [28.8; 40.3]	36.9 [31.9; 44.8]	0.394
BMI, kg/m²	22.86 [20.14; 25.46]	23.20 [21.27; 24.92]	0.508
Wiberg angle preoperatively, deg.	32.0 [27.6; 37.8]	31.2 [27.0; 34.8]	0.340
Alpha angle (Dunn 45°) preoperatively, deg.	60.1 [47.9; 73.1]	57.2 [52.6; 64.9]	0.635
Offset ratio preoperatively	0.07 [0.03; 0.13]	0.07 [0.05; 0.10]	0.842
iHOT-33 preoperatively, points	54.8 [38.5; 63.0]	45.0 [39.9; 57.9]	0.293
HOS-ADL preoperatively, points	79.4 [67.7; 87.1]	61.6 [55.1; 77.6]	0.009
HOS-Sport preoperatively, points	54.2 [32.6; 66.0]	42.3 [30.0; 52.1]	0.264

Quantitative variables are presented as Me [Q₁; Q₃], qualitative — n (%; 95% CI).

 

Characteristics of the performed interventions

Surgical tactics differed between the groups. In the retrospective group, joint was accessed via interportal capsulotomy in 62 (96.9%) cases and via T-shaped capsulotomy in 2 (3.1%) cases; in the prospective group, longitudinal capsulotomy was performed in all 30 (100.0%) cases. Statistical comparison of capsulotomy types was not performed, as this parameter was a component of the surgical tactics being compared.

Labral repair was performed in 54 (84.4%) cases of the retrospective group and in 22 (73.3%) cases of the prospective group; no statistically significant difference in the frequency of labral repair was found between the groups (p = 0.262). The number of fixators used per surgery did not differ significantly; however, only knotless fixators were used in the prospective group (Table 2).

 

Table 2. Surgical procedure characteristics

Parameter	Retrospective group (n = 64)	Prospective group (n = 30)	p
Labral repair	54 (84.4%; 73.1-92.2)	22 (73.3%; 54.1-87.7)	0.262
Without refixation	10 (15.6%; 7.8-26.9)	8 (26.7%; 12.3-45.9)	0.262
Total number of fixators per surgery	3.0 [2.0; 3.0]	2.0 [0.2; 3.0]	0.373
Knotless fixators per surgery	0.0 [0.0; 2.0]	2.0 [0.2; 3.0]	–
Knot-type fixators per surgery	0.0 [0.0; 2.2]	0.0 [0.0; 0.0]	–

Quantitative variables are presented as Me [Q₁; Q₃], qualitative — n (%; 95% CI).

 

Clinical outcomes

In both groups, by the final follow-up, an increase in the values of all used PROMs was recorded. Pre- and postoperative values, as well as the magnitude of improvement, are presented in Table 3.

 

Table 3. Dynamics of PROMs scores in clinical groups

Scale	Group	Preoperative, Me [Q1; Q3]	Postoperative, Me [Q1; Q3]	Δ, Me [Q1; Q3]; HL (95% CI)	p
iHOT-33	Retrospective	54.8 [38.5; 63.0]	80.4 [71.8; 90.8]	23.5 [10.3; 39.5] HL 26.1 (21.9-31.4)	< 0.001
	Prospective	45.0 [39.9; 57.9]	90.4 [81.3; 96.2]	38.3 [29.3; 50.3] HL 39.4 (33.8-44.6)	< 0.001
HOS-ADL	Retrospective	79.4 [67.7; 87.1]	91.9 [88.3; 95.6]	11.8 [5.5; 23.2] HL 14.0 (10.3-19.9)	< 0.001
	Prospective	61.6 [55.1; 77.6]	94.9 [88.7; 98.2]	32.5 [13.2; 42.6] HL 27.9 (22.8-36.0)	< 0.001
HOS-Sport	Retrospective	54.2 [32.6; 66.0]	77.8 [68.7; 86.1]	20.8 [5.6; 45.1] HL 23.6 (18.1-31.9)	< 0.001
	Prospective	42.3 [30.0; 52.1]	80.6 [72.2; 93.8]	37.5 [25.0; 53.5] HL 36.1 (27.2-44.7)	< 0.001

 

In intergroup comparison, the postoperative iHOT-33 score was statistically significantly higher in the prospective group. Postoperative HOS-ADL and HOS-Sport scores were also higher in the prospective group; however, the differences in these scales did not reach statis-tical significance.

Intergroup differences in the magnitude of improvement were found for all three scales: iHOT-33 — 38.3 [29.3; 50.3] vs 23.5 [10.3; 39.5] points (p = 0.003), HOS-ADL — 32.5 [13.2; 42.6] vs 11.8 [5.5; 23.2] points (p = 0.005), and HOS-Sport — 37.5 [25.0; 53.5] vs 20.8 [5.6; 45.1] points (p = 0.038) (Figure 3).

 

Figure 3. Median improvement in PROM scores with interquartile ranges in the retrospective and prospective groups

 

The rate of achieving SCB for iHOT-33 was higher in the prospective group, amounting to 26/30 (86.7%; 95% CI 69.3-96.2) vs 28/64 (43.8%; 95% CI 31.4-56.7) in the retrospective group (p < 0.001). The prospective group also achieved MCID and SCB for HOS-ADL more frequently: 23/30 (76.7%; 95% CI 57.7-90.1) vs 34/64 (53.1%; 95% CI 40.2-65.7), respectively (p = 0.041). According to Fisher's exact test, no intergroup differences were found in the rate of achieving MCID for iHOT-33, MCID for HOS-Sport, or SCB for HOS-Sport (Table 4).

 

Table 4. Rate of reaching MCID and SCB thresholds across clinical groups

Scale		Retrospective group (n = 64)	Prospective group (n = 30)	p
iHOT-33	MCID	54 (84.4%; 73.1-92.2)	29 (96.7%; 82.8-99.9)	0.165
	SCB	28 (43.8%; 31.4-56.7)	26 (86.7%; 69.3-96.2)	< 0.001
HOS-ADL	MCID	34 (53.1%; 40.2-65.7)	23 (76.7%; 57.7-90.1)	0.041
	SCB	34 (53.1%; 40.2-65.7)	23 (76.7%; 57.7-90.1)	0.041
HOS-Sport	MCID	46 (71.9%; 59.2-82.4)	27 (90.0%; 73.5-97.9)	0.064
	SCB	24 (37.5%; 25.7-50.5)	16 (53.3%; 34.3-71.7)	0.182

Qualitative data are presented as n (%; 95% CI).

MCID — minimal clinically important difference; SCB— substantial clinical benefit.

 

Quality of cam deformity correction

The postoperative alpha angle in the 45° Dunn view was 41.1 [39.4; 43.5]° in the retrospective group and 38.0 [37.0; 40.8]° in the prospec-tive group (p < 0.001). The postoperative offset ratio was 0.17 [0.15; 0.19] and 0.19 [0.17; 0.20], respectively (p = 0.002). An alpha angle of less than 50° was achieved in all cases in both groups. An offset ratio ≥ 0.17 was recorded in 46 of 64 cases (71.9%; 95% CI 59.2-82.4) in the retrospective group and in 26 of 30 cases (86.7%; 95% CI 69.3-96.2) in the prospective group (p = 0.190).

Upon the qualitative assessment of femoro-plasty results, a spherical shape was achieved in 22 of 64 cases (34.4%; 95% CI 22.4-47.8) in the retrospective group and in 29 of 30 cases (96.7%; 95% CI 82.8-99.9) in the prospective group (p < 0.001). According to the criteria of Y. Mansor et al., adequate resection was recorded in 34 of 64 (53.1%; 95% CI 40.2-65.7) and 29 of 30 (96.7%; 95% CI 82.8-99.9) cases, respectively; over-resection was recorded in 26 of 64 (40.6%; 95% CI 28.5-53.6) and 1 of 30 (3.3%; 95% CI 0.1-17.2). The differences in the distribution across the three resection quality categories were statistically significant (p < 0.001); upon post-hoc comparison with Bonferroni correction, the differences were significant for over-resection (p < 0.001) and adequate resection (p < 0.001), but not for under-resection (p = 0.485) (Figure 4).

 

Figure 4. Distribution of the quality of cam deformity correction according to the criteria of Y. Mansor et al. in the retrospective and prospective groups

 

Capsular status according to MRI data

According to follow-up MRI, an intact capsule was identified in 32 of 64 cases (50.0%) in the retrospective group and in 23 of 30 cases (76.7%) in the prospective group. Partial healing was recorded in 20 of 64 (31.2%) and 6 of 30 (20.0%) cases, respectively, and capsular defect in 12 of 64 (18.8%) and 1 of 30 (3.3%). The differences in the distribution across the three capsular status categories were statistically significant (p = 0.031). Upon post-hoc comparison with Bonferroni correction, the difference in the frequency of complete healing was significant (p = 0.043), whereas the differences in partial healing (p = 0.767) and capsular defect (p = 0.131) did not reach statistical significance (Figure 5).

 

Figure 5. Distribution of capsular status according to follow-up MRI findings in the retrospective and prospective groups

 

In the pooled sample, intact capsular status on follow-up MRI was associated with more favorable clinical outcomes: patients with an intact capsule showed greater improvement in iHOT-33, HOS-ADL, and HOS-Sport, as well as a higher rate of achieving clinically significant thresholds across all analyzed scales (Table 5).

 

Table 5. Dependence of clinical outcomes on capsule status in the pooled sample

Parameter	Intact capsule (n = 55)	Non-intact capsule (n = 39)	p
Δ iHOT-33	39.1 [24.3; 50.4]	22.4 [10.3; 29.7]	< 0.001
Δ HOS-ADL	22.1 [13.2; 42.6]	7.3 [4.4; 17.6]	< 0.001
Δ HOS-Sport	41.7 [21.7; 56.9]	16.7 [0.0; 27.5]	< 0.001
iHOT-33 SCB	40 (72.7%; 59.0-83.9)	14 (35.9%; 21.2-52.8)	< 0.001
HOS-ADL MCID	42 (76.4%; 63.0-86.8)	15 (38.5%; 23.4-55.4)	< 0.001
HOS-ADL SCB	42 (76.4%; 63.0-86.8)	15 (38.5%; 23.4-55.4)	< 0.001
HOS-Sport MCID	48 (87.3%; 75.5-94.7)	25 (64.1%; 47.2-78.8)	0.012
HOS-Sport SCB	32 (58.2%; 44.1-71.3)	8 (20.5%; 9.3-36.5)	< 0.001

Quantitative data are presented as median [Q₁; Q₃], qualitative — as n (%; 95% CI).

 

MRI findings in the area of the fixators

Among procedures in which labral repair was performed, any MRI changes in the area of the fixators were identified in 22 of 54 cases (40.7%; 95% CI 27.6-55.0) in the retrospective group and in 4 of 22 cases (18.2%; 95% CI 5.2-40.3) in the prospective group (p = 0.068) (Table 6).

 

Table 6. MRI changes around the fixators after labral repair on follow-up MRI

Parameter	Retrospective group (n = 54)	Prospective group (n = 22)	p
Any MRI changes around fixators	22 (40.7%; 27.6-55.0)	4 (18.2%; 5.2-40.3)	0.068
Changes around knotless fixators	6 (11.1%; 4.2-22.6)	4 (18.2%; 5.2-40.3)	0.462
Changes around knot-type fixators	16 (29.6%; 18.0-43.6)	Not used	–

Qualitative variables are presented as n (%; 95% CI).

 

In an analysis at the level of individual fixators, MRI signs of changes were identified in 21 of 84 knot-type fixators (25.0%; 95% CI 16.2-35.6) and in 16 of 127 knotless fixators (12.6%; 95% CI 7.4-19.7; p = 0.026). Among procedures in which only one type of fixator was used, changes were identified in 14 of 24 cases (58.3%; 95% CI 36.6-77.9) when only knot-type fixators were used and in 12 of 46 cases (26.1%; 95% CI 14.3-41.1) when only knotless fixators were used (p = 0.010) (Table 7).

 

Table 7. Frequency of MRI changes depending on fixator type

Parameter	Knot-type fixators	Knotless fixators	p
MRI changes at the level of individual fixators, n/N (%), 95% CI	21/84 (25.0) 16.2-35.6	16/127 (12.6) 7.4-19.7	0.026
MRI changes in procedures using only one type of fixator, n/N (%), 95% CI	14/24 (58.3) 36.6-77.9	12/46 (26.1) 14.3-41.1	0.010

 

In univariate analysis, among procedures with labral refixation, the presence of MRI changes in the area of the fixators was not associated with the magnitude of improvement in iHOT-33, HOS-ADL, and HOS-Sport, nor with the rate of achieving clinically significant MCID and SCB thresholds (Table 8).

 

Table 8. Analysis of the clinical significance of MRI changes around the fixators

Parameter	Changes detected (n = 26)	No changes detected (n = 50)	p
Δ iHOT-33	22.23 [10.36; 46.59]	30.53 [21.37; 46.88]	0.243
Δ HOS-ADL	13.24 [7.35; 31.68]	13.33 [5.89; 26.47]	0.943
Δ HOS-Sport	25.00 [5.56; 32.30]	24.45 [9.03; 44.34]	0.926
iHOT-33 SCB	12 (46.2%; 26.6-66.6)	29 (58.0%; 43.2-71.8)	0.344
HOS-ADL MCID	15 (57.7%; 36.9-76.6)	28 (56.0%; 41.3-70.0)	1.000
HOS-ADL SCB	15 (57.7%; 36.9-76.6)	28 (56.0%; 41.3-70.0)	1.000
HOS-Sport MCID	18 (69.2%; 48.2-85.7)	39 (78.0%; 64.0-88.5)	0.416
HOS-Sport SCB	7 (26.9%; 11.6-47.8)	20 (40.0%; 26.4-54.8)	0.318

Quantitative data are presented as median [Q₁; Q₃], qualitative — as n (%; 95% CI).

 

Multivariate analysis of predictors associated with clinical outcomes

Multivariate analysis was performed for three continuous outcomes (improvement in iHOT-33, HOS-ADL, and HOS-Sport) and three binary outcomes: failure to achieve iHOT-33 SCB, failure to achieve HOS-ADL MCID/SCB, and failure to achieve HOS-Sport SCB. Linear models for continuous outcomes were additionally adjusted for the preoperative value of the respective scale (Table 9).

 

Table 9. Multivariate analysis of predictors associated with clinical outcomes

A. Continuous outcomes, β (95% CI); p
Predictor	Δ iHOT-33	Δ HOS-ADL	Δ HOS-Sport
Non-intact capsule	–8.93 (–13.08; –4.78) p < 0.001	–3.95 (–6.77; –1.13) p = 0.007	–9.12 (–15.01; –3.23) p = 0.003
Non-spherical femoroplasty	–6.19 (–10.73; –1.65) p = 0.008	–1.28 (–4.18; 1.63) p = 0.386	–6.74 (–13.60; 0.13) p = 0.054
≥ 1 MRI change around fixators	3.26 (–0.89; 7.42) p = 0.122	1.41 (–1.17; 3.99) p = 0.282	5.42 (–0.89; 11.73) p = 0.092
B. Binary outcomes, aOR (95% CI); p
Predictor	Failure to achieve iHOT-33 SCB (40/94)	Failure to achieve HOS-ADL MCID/SCB (37/94)	Failure to achieve HOS-Sport SCB (54/94)
Non-intact capsule	4.54 (1.85-11.11) p < 0.001	5.20 (2.06-13.13) p < 0.001	5.17 (1.98-13.48) p < 0.001
Non-spherical femoroplasty	1.72 (0.70-4.21) p = 0.239	2.53 (1.00-6.40) p = 0.050	1.24 (0.50-3.07) p = 0.649
≥ 1 MRI change around fixators	1.57 (0.58-4.25) p = 0.372	0.84 (0.30-2.38) p = 0.749	2.32 (0.80-6.72) p = 0.120

β — regression coefficient; aOR — adjusted odds ratio.

 

In the linear models, non-intact capsule was associated with less improvement in iHOT-33 and HOS-Sport (p = 0.003). Non-spherical femoroplasty was associated with less impro-vement in iHOT-33 (p = 0.008), whereas the association for HOS-Sport was borderline (p = 0.054) and was not statistically significant for HOS-ADL. The presence of MRI changes around the fixators was not independently associated with improvement in PROMs. In the logistic models, non-intact capsule was the only predictor that maintained a consistent independent association with all analyzed binary outcomes: it increased the probability of failing to achieve iHOT-33 SCB (p < 0.001), HOS-ADL MCID/SCB (p < 0.001), and HOS-Sport SCB (p < 0.001).

DISCUSSION

In the present study, the refinement of arthro-scopic surgical technique was associated with more favorable morphological and clinical out-comes compared with the standard approach. In the prospective group, the postoperative iHOT-33 score was higher than in the retrospective group, the improvement across all used scales was greater, and the rate of achieving substantial clinical benefit for iHOT-33 was 86.7% compared with 43.8%. For comparison, in a large analysis from the UK Non-Arthroplasty Hip Registry, which included 4963 arthroscopic procedures for FAI, clinical improvement was also maintained at 12 months of follow-up; however, the MCID for iHOT-12 was achieved by approximately two-thirds of patients, and SCB by less than half [5]. Thus, the SCB rate for iHOT-33 in the retrospective group of our study is comparable to the data from the large registry, whereas in the prospective group it was higher. Given the significance of the deformity type and baseline patient characteristics [24], the observed differences between the groups, which were otherwise comparable, are likely attributable to the comprehensive optimization of the surgical technique.

On follow-up MRI, an intact capsule was more frequently identified in the prospective group, whereas capsular defect was less common. In the pooled sample, intact capsular status was associated with greater improvement in iHOT-33, HOS-ADL, and HOS-Sport scores, as well as with a higher rate of achieving clinically significant thresholds. In the linear models, non-intact capsule was associated with less improvement in all three PROMs, and in the logistic models, with an increased probability of failing to achieve clinically significant thresholds. Similarly, in the study by F. Yang et al., patients with a non-healed or partially healed capsule after arthroscopy had lower postoperative HOS-ADL (75.1 vs 83.5; p = 0.007), HOS-SSS (64.5 vs 77.1; p = 0.005), and iHOT-12 (56.1 vs 70.2; p = 0.006) scores, as well as a lower rate of achieving MCID for HOS-ADL (52% vs 80%; p = 0.037) [25]. In the present study, the direction of the effect was similar: with a non-intact capsule, the probability of failing to achieve clinically significant thresholds was higher for iHOT-33 SCB, HOS-ADL MCID/SCB, and HOS-Sport SCB. At the same time, data on capsular healing after different surgical approaches remain heterogeneous [19, 25]. The lower rate of incomplete capsular healing with longitudinal capsulotomy compared with transverse interportal capsulotomy may be related to the biomechanical characteristics of the anterior capsuloligamentous complex [26]. During hip extension, tension on the anterior capsule may increase stress on the suture line after transverse capsulotomy, whereas with a longitudinal incision orientation, the conditions for approximating the capsule edges may be more favorable.

Another important component of the modified surgical technique was cam deformity correction. In the prospective group, a spherical femoroplasty shape (96.7% vs 34.4%) and adequate resection according to the criteria of Y. Mansor et al. (96.7% vs 53.1%) were achieved more frequently, whereas the rate of over-resection was significantly lower (3.3% vs 40.6%). Data from revision arthroscopy show that residual deformity re-mains one of the main causes of re-operations [6], while excessive resection is associated with worse functional outcomes [12]. In the study by Y. Mansor et al., which included 130 revision arthroscopies, over-resection was identified in 20 (15.4%) cases; after revision surgery, the modified Harris Hip Score (mHHS) was lower in the over-resection group than in the under-correction group (66.7±19.8 vs 81.0±14.5; p = 0.031), and conversion to total hip arthro-plasty was more frequent (30% vs 0%; p = 0.024) [12]. These findings confirm that the assessment of femoroplasty quality should not be limited solely to the alpha angle value. Although some studies have considered the postoperative alpha angle as a prognostically significant parameter, with a threshold of 48.3° associated with achieving an acceptable symptomatic state at 5 years [27], other data do not support its independent association with clinical outcome [28]. This is consistent with our data and supports an approach focused not only on quantitative correction but also on restoring the spherical geometry of the femoral head-neck junction.

Differences related to labral refixation were less pronounced. In the prospective group, the rate of any MRI changes in the area of the fixators after labral repair was 18.2% compared with 40.7% in the retrospective group, and in the analysis at the level of individual fixators, changes were more frequently identified around knot-type constructs than around knotless ones (25.0% and 12.6%, respectively). According to a systematic review, knotless and knot-type fixators provide comparable clinical outcomes, although knotless fixation may be associated with less postoperative pain and less pronounced reactive tissue changes [29]. In the systematic review by W.L. Johns et al., which included 47 studies and 6185 patients, the postoperative mHHS was comparable with knotless and knot-type fixators — 86.3±2.8 and 88.8±5.5, respectively; the re-operation rate was 8.1% and 6.2%, and the conversion rate to total hip arthroplasty was 4.4% and 1.9%, respectively [29]. Our data do not allow us to assert a standalone clinical superiority of abandoning knot-type fixators, as univariate analysis showed no association with PROMs. Nevertheless, the more favorable morphological profile of knotless fixation appears to be a practically important observation.

Study limitations

Study limitations include the single-center, non-randomized design, the relatively small size of the prospective group, and the simultaneous implementation of multiple technical components. Nevertheless, the combined assessment of radiographic, MRI, and clinical outcomes allows us to consider the proposed approach as a practically significant direction for further improvement of arthroscopic treatment in patients with FAI. The practical significance of such refinements is particularly relevant given the expanding use of hip arthroscopy in Russia [30].

CONCLUSION

The refinement of arthroscopic surgical tech-nique in patients with femoroacetabular impingement was associated with more favorable morphological and clinical outcomes com-pared with the standard approach. The modified technique group demonstrated higher quality of cam deformity correction, more frequent complete capsular healing on follow-up MRI, a lower rate of changes around labral fixators, and greater improvement in iHOT-33, HOS-ADL, and HOS-Sport scores. Comprehensive optimization of surgical tactics appears to be a promising direction for improving treatment outcomes and requires further research in larger studies.

DISCLAIMERS

Author contribution

Filonov P.V. — data acquisition, analysis and interpretation, statistical data processing, drafting the manuscript.

Bogopolskiy O.E. — data acquisition, analysis and interpretation, editing the manuscript.

Tikhilov R.M. — study concept and design, editing the manuscript, scientific guidance.

All authors have read and approved the final version of the manuscript of the article. All authors agree to bear responsibility for all aspects of the study to ensure proper conside-ration and resolution of all possible issues related to the correctness and reliability of any part of the work.

Funding source. This study was not supported by any external sources of funding.

Disclosure competing interests. The authors declare that they have no competing interests.

Ethics approval. The study was approved by the local ethics committee of the Vreden National Medical Research Center of Traumatology and Orthopedics (extract No 3 from protocol No 3, 12.10.2023).

Consent for publication. Written consent was obtained from the patients for publication of relevant medical information and all of accompanying images within the manuscript.

Use of artificial intelligence. No generative artificial intelligence technologies were used in the preparation of this manuscript.

About the authors

Pavel V. Filonov

Vreden National Medical Research Center of Traumatology and Orthopedics

Author for correspondence.
Email: drpavelfilonov@gmail.com
ORCID iD: 0000-0001-7758-0128
SPIN-code: 2498-0191
Russian Federation, St. Petersburg

Oleg E. Bogopolsky

Vreden National Medical Research Center of Traumatology and Orthopedics

Email: 9202211@gmail.com
ORCID iD: 0000-0002-4883-0543
SPIN-code: 9938-4303

Cand. Sci. (Med.)

Russian Federation, St. Petersburg

Rashid M. Tikhilov

Vreden National Medical Research Center of Traumatology and Orthopedics

Email: rtikhilov@gmail.com
ORCID iD: 0000-0003-0733-2414
SPIN-code: 3602-4912

Dr. Sci. (Med.), Professor, Corresponding Member of the RAS

Russian Federation, St. Petersburg

References

Supplementary files