The Effect of Pelvic Incidence on Outcomes After Hip Arthroscopy for Femoroacetabular Impingement and Acetabular Labral Tears

Abstract

Background:

In the setting of femoroacetabular impingement (FAI), decompression osteoplasties reconcile deleterious loading patterns caused by cam and pincer lesions. However, native variations of spinopelvic sagittal alignment may continue to perpetuate detrimental effects on the labrum, chondrolabral junction, and articular cartilage after hip arthroscopy.

Purpose:

To evaluate the effect of pelvic incidence (PI) on postoperative outcomes after hip arthroscopy for acetabular labral tears in the setting of FAI.

Study Design:

Cohort study; Level of evidence, 3.

Methods:

A retrospective query of prospectively collected data identified patients ≥18 years of age who underwent primary hip arthroscopy for FAI and acetabular labral tears between February 2014 and January 2022, with 3-, 6-, 12-, and 24-month follow-ups. Measurements for PI, pelvic tilt (PT), sacral slope (SS), and acetabular version were obtained via advanced diagnostic imaging. Patients were stratified into low-PI (<45°), moderate-PI (45°≤ PI ≤ 60°), and high-PI (>60°) cohorts. Patient-reported outcome measures (PROMs), clinically meaningful outcomes (ie, minimal clinically important difference, Patient Acceptable Symptom State, substantial clinical benefit, and maximal outcome improvement), visual analog scale (VAS) pain scores, and patient satisfaction were compared across cohorts.

Results:

A total of 74 patients met eligibility criteria and were stratified into low-PI (n = 28), moderate-PI (n = 31), and high-PI (n = 15) cohorts. Correspondingly, patients with high PI displayed significantly greater values for PT (P = .001), SS (P < .001), acetabular version (P < .001), and acetabular inclination (P = .049). By the 12- and 24-month follow-ups, the high-PI cohort was found to have significantly inferior PROMs, VAS pain scores, rates of clinically meaningful outcome achievement, and satisfaction relative to patients with moderate and/or low PI. No significant differences were found between cohorts regarding rates of revision arthroscopy, subsequent spine surgery, or conversion to total hip arthroplasty.

Conclusion:

After hip arthroscopy, patients with a high PI (>60°) exhibited inferior PROMs, rates of achieving clinically meaningful thresholds, and satisfaction at 12 and 24 months relative to patients with low or moderate PI. Conversely, the outcomes of patients with low PI (<45°) were found to match the trajectory of those with a neutral spinopelvic alignment (45°≤ PI ≤ 60°). These findings highlight the importance of analyzing spinopelvic parameters preoperatively to prognosticate outcomes before hip arthroscopy for acetabular labral tears and FAI.

When left untreated, femoroacetabular impingement (FAI) syndrome poses an elevated risk for the early progression of degenerative joint disease. ²¹ As a result, the use of hip arthroscopy has dramatically increased to address the maladaptive effects of FAI and prevent early conversion to total hip arthroplasty (THA). ⁵³ Indeed, the literature has highlighted the significant benefits of femoroacetabular osteoplasties in conjunction with labral preservation to improve symptoms and long-term outcomes. ²¹ However, independent of FAI, studies have found sagittal alignment to be closely correlated with health-related quality of life measures, and recent investigations have further revealed that abnormal spinopelvic anatomy may be innately linked to pathologies of the spine and hip.^{2,14,16,19,23,24,47}

Pelvic incidence (PI), pelvic tilt (PT), and sacral slope (SS) are common spinopelvic parameters used to evaluate sagittal alignment and gauge the adaptability of the pelvis to postural change. ²⁰ Specifically, PT and SS are dynamic parameters that vary with body position and may evolve with pathologic and/or compensatory changes over time. Conversely, PI is a fixed morphological parameter that remains constant over the course of an individual's adult life.^18,47 For any given position, SS and PT function as complementary angles that, when summed, yield a constant PI value for any given patient (PI = SS + PT). ²⁹ Furthermore, PI inherently dictates the degree of lumbar lordosis needed to maintain sagittal alignment, a relationship that has since been pragmatically estimated to be lumbar lordosis = PI ± 9°. ⁴⁵ Therefore, changes in PI predictably affect the compensatory ranges of spinopelvic motion (ie, PT, SS, and lumbar lordosis), with studies also reporting associations with acetabular version and inclination (ie, Tönnis angle). ¹⁸ As such, PI has generally been regarded as the primary axis of sagittal balance, indicative of the body's ability to transduce force and maintain balance within the conus of energy economics.^20,42

Perturbations in sagittal alignment, specifically high degrees of PI, have been shown to increase vertebral endplate loading that mechanistically offloads to the pelvic girdle (ie, pelvic rotation) as the primary mechanism of compensation. ⁴² Over time, degenerative stiffening of the lumbar spine may demand further extremes of hip/pelvis mobility to maintain native alignment, with hip-specific literature citing associations with the development of acetabular labral tears, femoral head osteonecrosis, premature osteoarthritis, and even inferior outcomes after THA.^{2,14,19,23,24,47} Overall, the misappropriation of load absorption and force transduction imposed by spinopelvic malalignment can contribute to accelerated arthrosis of both adjacent facets and neighboring large joints.^27,45 Thus, while osteoplasties may mitigate degenerative loading patterns caused by FAI, native variations in lumbosacral alignment may continue to pose deleterious implications in the context of hip preservation surgery.

Currently, there is growing interest regarding the influence of variations in sagittal alignment on postoperative outcomes after hip arthroscopy to elucidate new drivers of patient selection. ⁵¹ Notably, the prognostic implications of spinopelvic parameters remain uncertain for patients undergoing FAI decompression and acetabular labral repair. Therefore, the purpose of this study was to investigate the effect of PI on outcomes after hip arthroscopy for FAI and acetabular labral tears and, secondarily, track rates of subsequent surgeries (ie, revision hip arthroscopy, lumbar spine surgery, and THA). We hypothesized that patients with low (PI, <45°) and/or high (PI, >60°) PI would display inferior outcomes compared with those with moderate PI (45°≤ PI ≤ 60°) after hip arthroscopy.

Methods

Study Population and Design

After institutional review board (IRB) approval (Mass General Brigham IRB Nos. 2019P002191 and 2013P001442), this retrospective study identified patients from a prospectively collected, single-surgeon registry tracking outcomes of patients after primary hip arthroscopy. All included patients were ≥18 years of age, underwent primary hip arthroscopy for the treatment of FAI and symptomatic labral tears between February 2014 and January 2022, completed baseline patient-reported outcome measures (PROMs), and had replete PROMs for ≥3 of the 4 follow-up time points (3, 6, 12, and 24 months) postoperatively (ie, minimum 12-month follow-up). Additionally, all patients were required to have advanced, preoperative supine lumbosacral imaging (ie, magnetic resonance imaging [MRI] or computed tomography [CT]) to allow for measurement of spinopelvic parameters. Exclusion criteria consisted of labral debridement, previous surgery on the ipsilateral hip/leg, previous spine surgery (eg, decompression laminectomy or discectomy, kyphoplasty, and single- and/or multilevel fusion), absence of advanced (MRI or CT) preoperative supine lumbosacral imaging, radiographic evidence of hip dysplasia (lateral center-edge angle, <20°), advanced hip osteoarthritis (Tönnis grade >1), or confounding hip/spine pathologies (ie, hip: fractures, inflammatory arthropathies, and avascular necrosis; spine: congenital scoliosis, retrolisthesis, and lumbosacral transitional vertebra) that may independently influence outcomes or alter spinopelvic parameters. ⁵⁴ Last, patients identified with morphologic abnormalities (ie, global retroversion, coxa profunda, etc) necessitating surgical dislocation and/or periacetabular osteotomy were routinely referred to an affiliated specialist and were not included in this study.

All patients underwent a preoperative physical examination consisting of provocative testing of the labrum and assessment of impingement-related symptoms. Those with positive findings on clinical examination (eg, pain and/or limited range of motion with flexion, adduction, and internal rotation or flexion, abduction, and external rotation) subsequently underwent a magnetic resonance arthrogram to confirm the presence of labral pathology coupled with a diagnostic and therapeutic intra-articular injection (eg, combined local anesthetic with a low-dose corticosteroid). Finally, all patients trialed nonoperative management, including formal physical therapy, activity modification, and/or oral anti-inflammatory medications, for ≥3 months preoperatively. ³⁷ Those with persistent symptoms, despite nonoperative management, were offered hip arthroscopy.

Abbreviated Surgical Technique

All hip arthroscopies were performed by a single, fellowship-trained surgeon (S.D.M.) using previously published techniques.^{9,22,32,33,38,40} Briefly, patients were administered general anesthesia and positioned supine on a hip distraction table (Advanced Supine Hip Positioning System; Smith & Nephew) with a well-padded perineal post. Intra-articular access was facilitated via a puncture capsulotomy technique to avoid biomechanical disruption of the iliofemoral ligament and zona orbicularis. Using intra-articular fluid distention, we first established the anterolateral portal with the aid of fluoroscopic guidance, and subsequent portals (ie, anterior, midanterior, and Dienst) were placed under direct arthroscopic visualization. ⁹ Procedures were performed with the sparing use of intermittent traction and an emphasis on chondrolabral junction preservation.^38,40

A thorough diagnostic survey was performed to assess and grade the extent of damage to the labrum, chondrolabral junction, and articular cartilage surfaces. For pincer deformities, acetabular recession and acetabuloplasty were performed as clinically indicated. ⁴⁰ To address labral lesions, repair alone was performed if adequate tissue was amenable to suture fixation. Conversely, if the labrum was deemed irreparable because of hypoplasia (width, <5 mm), complex tearing, or frank degeneration of native tissue, labral augmentation was carried out with capsular autograft.^22,38 Chondral lesions (ie, full-thickness chondral flaps, focal Outerbridge grade ≥2 lesions, and chondrolabral junction breakdown) were addressed using a standardized method of bone marrow aspirate concentrate augmentation. ³³ With traction released, a final survey from the peripheral compartment was performed to ensure restoration of the hip suction seal and confirm an in-round labral repair.^38,40 If indicated, a decompression femoroplasty was carried out with the hip flexed to approximately 45°. Last, a dynamic range of motion examination ensured the adequacy of decompressions and that incisions were closed.

Postoperative Rehabilitation

All patients underwent a standardized 5-phase, patient-guided rehabilitation program and were prescribed deep vein thrombosis prophylaxis with daily aspirin (81 mg) for 3 weeks postoperatively. Patients were permitted immediate weightbearing as tolerated using a flat-foot gait and crutches. By 6 weeks, patients began the use of a stationary bicycle without resistance. At 10 weeks, patients transitioned to an elliptical trainer with low resistance and/or swimming while avoiding intense flutter kicks. At 4 months, patients were cleared for light strengthening exercises (ie, short-arc leg presses and reverse hamstring curls). Finally, at 6 months, patients were permitted to resume impact loading as tolerated, with clearance for high-pivoting activities beginning no sooner than 9 to 12 months postoperatively. ³⁷

FAI and Spinopelvic Imaging Parameters

Standing anteroposterior and Dunn lateral radiographs were used to obtain alpha angle, lateral center-edge angle, limb-length discrepancy, Tönnis grade, and acetabular inclination. ⁷ Additionally, while PI is traditionally extrapolated from standing lateral radiographs, recent studies have shown conventional supine, advanced imaging modalities (ie, MRI and CT) to be more reliable assessments of sagittal parameters.^6,34 As such, the present study employed previously validated methodologies to obtain spinopelvic measurements exclusively from available MRI or CT scans.^{16,28,36,49,52}

Specifically, using separate sagittal cuts, the center points of circles outlining the bilateral femoral heads were identified and a line was subsequently drawn to connect these center points. The midpoint of this line (ie, center of the bicoxofemoral axis) was marked on a sagittal cut that incorporated the center of the sacral endplate. PI was then derived as the angle produced by a perpendicular line from the midpoint of the sacral endplate (S1) to the center of the bicoxofemoral axis. ⁵² PT was calculated as the angle between the line connecting the center of the bicoxofemoral axis to the center of the sacral endplate and the vertical axis. Next, SS was measured as the angle between a line drawn along the superior border of the sacral endplate and the horizontal axis (Figure 1). ¹⁶ As previously described, by integrating multiple views (ie, coronal, sagittal, and axial cuts) to determine the level of maximum congruence between the femoral heads/acetabulum at the 3-o’clock position, acetabular version was derived from the angle formed by a line that passed through the anterior and posterior bony edges of the acetabular rim and a line perpendicular to an interischial line that connects both posterior margins of the ischium.^{15,36,49,50,52} Intermeasurement reliability analyses were performed on available study patients with both MRI and CT imaging to assess the cross-modality reliability for measurements of PI, PT, and SS.^{16,28,36,49,52} On the basis of the renewed Roussouly sagittal profile classification, patients were stratified into low-PI (<45°), moderate-PI (45°≤ PI ≤ 60°), and high-PI (>60°) cohorts. ²⁵

Figure 1.

Spinopelvic measurements using (A-C) computed tomography scans or (D-F) magnetic resonance imaging scans. The center points of the femoral heads were marked on sagittal slices (A, C, D, F). These positions were translated onto the sagittal cut that incorporated the center of the sacral endplate (B, E). A line was then drawn connecting the center of the femoral heads, and the midpoint was identified to yield the center of the bicoxofemoral axis. Using this orientation, we then measured spinopelvic parameters, which included pelvic incidence, pelvic tilt, and sacral slope.

Data Collection and Functional Outcome Evaluation

Patient and descriptive data, including age, sex, laterality, body mass index (BMI), subjective patient reporting of low back pain, and the aforementioned radiologic parameters, were collected preoperatively. Intraoperative variables of interest included the most severe Outerbridge grade observed on the acetabulum or femoral head, arthroscopic procedures performed, presence of full-thickness chondral lesions/flaps, and degree of injury to the labrum/chondrolabral junction based on validated classification scales. ⁴ Primary outcomes included the modified Harris Hip Score (mHHS), Hip Outcome Score–Activities of Daily Living (HOS-ADL), Hip Outcome Score–Sports Subscale (HOS-SS), Non-Arthritic Hip Score (NAHS), International Hip Outcome Tool–33 (iHOT-33), and visual analog scale (VAS) for pain. Secondary outcomes included clinically meaningful outcome achievement, assessed by calculating the percentage of patients who achieved previously published threshold values for minimal clinically important difference,^39,46 Patient Acceptable Symptom State,^39,46 substantial clinical benefit,^39,46 and maximum outcome improvement.^31,41 Finally, rates of patient satisfaction, revision hip arthroscopy, subsequent spine surgery, and conversion to THA were tracked.

Statistical Analysis

An a priori power analysis was performed based on previous literature to identify the number of patients necessary to achieve 80% power for our primary analysis. Using an estimated standard deviation in iHOT-33 scores of 17.0, an expected mean difference equivalent to the 2-year minimal clinically important difference threshold published by Nwachukwu et al, ³⁹ and an estimated 40:40:20 ratio of patients with high, moderate, and low PI, we determined that a sample size of 66 patients was needed to achieve 80% power (minimum: 26.4 patients with low PI, 26.4 with moderate PI, and 13.2 with high PI). ³³ This arbitrary ratio of patients with high, moderate, and low PI was used based on previous claims that a high PI may be protective against symptomatic FAI.^26,51

Descriptive statistics were tabulated as mean ± SD or frequency (percentage), while parameter estimates were presented as mean (95% CI). Categorical variables were compared between cohorts using chi-square or Fisher exact tests, as appropriate. Statistically significant chi-square or Fisher exact tests were further evaluated using direct cell-to-cell comparisons to determine which groups were significantly different from each other at the level of P < .05. ¹² Normality and equal variances of continuous data were confirmed using Shapiro-Wilk tests and F tests, respectively. Continuous data were then compared between cohorts via 2 methods. First, 1-way analysis of variance tests with Tukey post hoc analyses were used to compare data between groups. Next, linear mixed-effects models, including time and random by-participant intercepts, were implemented to assess longitudinal changes in PROM and VAS pain scores. ³ These models provide greater statistical power by linking observations for each participant, accounting for variability between patients, and incorporating all available data rather than excluding participants missing a single follow-up time point.^3,44 Finally, associations between radiologic measurements (eg, PI, PT, and acetabular version) were analyzed using linear regression analysis. Analyses were performed using Microsoft Excel (Microsoft) and R Version 4.2.1 (R Foundation for Statistical Computing), and P < .05 was considered significant.

Results

Patient Characteristics, Intraoperative Findings, and Procedures Performed

A total of 74 patients met eligibility criteria and were included in our analyses (Figure 2).

Figure 2.

Flowchart detailing patient selection criteria. LCEA, lateral center-edge angle.

After obtaining spinopelvic measurements, patients were subsequently stratified into low-PI (<45°; n = 28), moderate-PI (45°≤ PI ≤ 60°; n = 31), and high-PI (>60°; n = 15) cohorts. The mean length of follow-up for the entire study cohort was 25.16 ± 2.91 months. No significant differences in age (P = .159), sex (P = .450), BMI (P = .098), or self-reported back pain (P = .663) were observed between groups. Excluding spinopelvic parameters, no significant differences in patient characteristics or preoperative variables were observed (Table 1). Additionally, no significant differences were identified regarding distributions of intra-articular pathologies or arthroscopic procedures performed (Table 2).

Table 1.

Patient Characteristics by Cohort ^a

	Low PI, n = 28	Moderate PI, n = 31	High PI, n = 15	P
Age, y	36.39 ± 10.4	41.45 ± 11.7	35.8 ± 13.3	.159
Sex				.450
Male	10 (35.7)	16 (51.6)	6 (40)
Female	18 (64.3)	15 (48.4)	9 (60)
BMI	24.15 ± 3.1	25.83 ± 4.3	26.44 ± 3.4	.098
Laterality				.123
Right	20 (71.4)	17 (54.8)	6 (40)
Left	8 (28.6)	14 (45.2)	9 (60)
LLD, mm	5.84 ± 3.4	5.28 ± 3.8	6.43 ± 5.3	.653
PI, deg	37.96 ± 5.1	52.58 ± 4.3	63.47 ± 4.1	<.001 b
Pelvic tilt, deg	7.79 ± 6.9	13.06 ± 8.6	17 ± 7.2	<.001 b
Sacral slope, deg	30.14 ± 8.9	39.26 ± 8.9	46.47 ± 7.1	<.001 b
Acetabular version, deg	14.71 ± 5.5	17.48 ± 5.6	23.07 ± 3.8	<.001 b
Alpha angle, deg	56.82 ± 18.3	52.94 ± 15.4	56 ± 18.2	.665
LCEA, deg	36.04 ± 6.7	36.55 ± 6.9	31.67 ± 6.3	.060
Tönnis angle, deg	2.54 ± 4.4	1.97 ± 5.9	6.07 ± 5.8	.049 b
Tönnis grade				.918
0	13 (46.4)	14 (45.2)	6 (40)
1	15 (53.6)	17 (54.8)	9 (60)
Type of FAI				.546
Pincer deformity	14 (50)	20 (64.5)	9 (60)
Cam deformity	2 (7.1)	1 (3.2)	2 (13.3)
Cam and pincer	12 (42.9)	10 (32.3)	4 (26.7)
Patient-reported low back pain	18 (64.3)	17 (54.8)	10 (66.7)	.663
Follow-up, mo	23.29 ± 5.5	24.88 ± 3.6	21.36 ± 6.0	.077

^aData are presented as n (%) or mean ± SD. BMI, body mass index; FAI, femoroacetabular impingement; LCEA, lateral center-edge angle; LLD, limb-length discrepancy; PI, pelvic incidence.

^bA significant difference between groups.

Table 2.

Intraoperative Findings and Procedures Performed by Cohort ^a

	Low PI, n = 28	Moderate PI, n = 31	High PI, n = 15	P
Labral condition				.227
Normal	00 (0)	00 (0)	00 (0)
Degeneration	16 (57.1)	14 (45.2)	6 (40)
Full-thickness tear	4 (14.3)	12 (38.7)	4 (26.7)
Detachment	8 (28.6)	5 (16.1)	5 (33.3)
Ossification	00 (0)	00 (0)	00 (0)
Most severe Outerbridge grade				.875
0	00 (0)	00 (0)	00 (0)
1	2 (7.1)	1 (3.2)	00 (0)
2	6 (21.4)	7 (22.6)	2 (13.3)
3	16 (57.1)	19 (61.3)	9 (60)
4	4 (14.3)	4 (12.9)	4 (26.7)
Chondrolabral junction				.478
Normal	00 (0)	1 (3.2)	1 (6.7)
Malacia	9 (32.1)	7 (22.6)	1 (6.7)
Debonding	5 (17.9)	5 (16.1)	3 (20)
Cleavage	11 (39.3)	15 (48.4)	6 (40)
Defect	3 (10.7)	3 (9.7)	4 (26.7)
Chondral flap	12 (42.9)	12 (38.7)	3 (20)	.314
Labral management				.060
Repair alone	1 (3.6)	6 (19.4)	4 (26.7)
Augmentation with capsular autograft	27 (96.4)	25 (80.6)	11 (73.3)
Osteoplasty performed				.546
Acetabuloplasty	14 (50)	20 (64.5)	9 (60)
Femoroplasty	2 (7.1)	1 (3.2)	2 (13.3)
Femoroacetabular osteoplasty	12 (42.9)	10 (32.3)	4 (26.7)

^aData are presented as n (%). PI, pelvic incidence.

As expected, significant differences were observed between cohorts when assessing PT (P = .001), SS (P < .001), acetabular inclination (P = .049), and functional acetabular version (P < .001) (Table 1). Notably, separate linear regressions were performed to reveal PI (F[1,72] = 20.925; P < .001; R² = 0.225) to be a stronger predictor of functional acetabular version than PT (F[1,72] = 5.492; P = .022; R² = 0.071). ⁵ Intermeasurement reliability analyses performed on study patients (n = 6) with both MRI and CT imaging revealed excellent cross-modality reliability for measurements of PI, PT, and SS (intraclass correlation coefficient values: 0.98 (P = .001), 0.90 (P = .016), and 0.94 (P = .005), respectively).^{16,28,36,49,52}

Patient-Reported Outcome Measures

Comparing preoperative baseline PROM and VAS pain scores between cohorts revealed no significant differences; however, HOS-ADL and HOS-SS trended toward near significance (P = .066) (Table 3). By the 3-month follow-up, the high-PI cohort displayed significantly inferior PROMs relative to both the moderate- and the low-PI cohorts in mHHS (P = .013), HOS-ADL (P = .010), and NAHS (P = .006). This trend persisted at the 6- and 12-month follow-ups, with the high-PI cohort exhibiting significantly lower scores relative to both the moderate- and the low-PI cohorts across all collected outcome measures (Table 3). At the 24-month follow-up, the high-PI cohort had significantly inferior HOS-SS (P = .001) and NAHS (P = .007) scores relative to the low-PI cohort and worse outcomes across all metrics compared with the moderate-PI group (Table 3, Figure 3).

Table 3.

Mean Patient-Reported Outcome Scores ^a

	Mean Scores (95% CI)				Post Hoc P Values
	Low PI	Moderate PI	High PI	P	Low vs Moderate	High vs Low	High vs Moderate
Baseline
mHHS	58.17 (52.9-63.4)	61.43 (56.1-66.8)	59.01 (50.5-67.5)	.669	.658	.982	.853
HOS-ADL	69 (61.6-76.4)	70.81 (63.9-77.7)	57 (46.4-67.6)	.066	.930	.127	.062
HOS-SS	44 (35.8-52.2)	37.32 (28.7-45.9)	27.33 (16.2-38.5)	.066	.477	.053	.323
NAHS	61.35 (53.4-69.3)	62.38 (56.7-68.1)	56.54 (43.7-69.4)	.613	.976	.714	.592
iHOT-33	35.27 (29.4-41.2)	37.82 (32.7-43)	38.79 (27.9-49.7)	.737	.810	.765	.979
VAS pain	5.84 (4.9-6.8)	5.56 (4.7-6.5)	6.01 (4.8-7.2)	.814	.896	.973	.821
3-mo
mHHS	75.6 (70.8-80.4)	75.69 (70.6-80.7)	61.5 (47.7-75.3)	.013 b	.999	.018 b	.017 b
HOS-ADL	78.19 (72.8-83.5)	81.36 (76.7-86)	66.83 (56.2-77.5)	.010 b	.661	.048 b	.008 b
HOS-SS	38.41 (27.3-49.5)	46.07 (33.6-58.6)	23.5 (5.9-41.1)	.099	.611	.328	.081
NAHS	75.16 (71-79.3)	77.24 (72.9-81.6)	63.52 (53-74)	.006 b	.795	.022 b	.004 b
iHOT-33	57.33 (51.7-63)	64.32 (57.2-71.4)	46.65 (34.3-59)	.014 b	.290	.182	.011 b
VAS pain	3.11 (2.3-3.9)	2.47 (1.7-3.2)	3.86 (2.1-5.6)	.143	.489	.551	.133
6-mo
mHHS	85.31 (80.2-90.4)	83.34 (77.9-88.8)	62.88 (53.1-72.7)	<.001 b	.859	<.001 b	<.001 b
HOS-ADL	85.96 (81.6-90.3)	87.65 (83.5-91.8)	68.33 (55-81.7)	<.001 b	.887	.001 b	<.001 b
HOS-SS	62.5 (52-73)	68.46 (58-78.9)	36.75 (17.7-55.8)	.004 b	.701	.020 b	.003 b
NAHS	81.9 (77.1-86.7)	85.7 (81.2-90.2)	69.69 (57.9-81.5)	.003 b	.565	.026 b	.002 b
iHOT-33	67.52 (60.9-74.1)	71.48 (63.1-79.9)	50.58 (36.4-64.7)	.010 b	.742	.039 b	.008 b
VAS pain	2.28 (1.5-3)	2.07 (1.3-2.8)	4.62 (3.2-6)	.001 b	.911	.002 b	.001 b
12-mo
mHHS	84.92 (79.9-89.9)	89.22 (85.8-92.6)	71.31 (61.1-81.6)	<.001 b	.415	.004 b	<.001 b
HOS-ADL	90.2 (86.2-94.2)	91.36 (88.4-94.3)	71.93 (59.6-84.2)	<.001 b	.938	<.001 b	<.001 b
HOS-SS	71.4 (61.7-81.1)	77.71 (70.3-85.1)	43.57 (28.1-59)	<.001 b	.568	.001 b	<.001 b
NAHS	85.92 (81.1-90.7)	89.91 (86.7-93.1)	71.54 (59.4-83.7)	<.001 b	.500	.004 b	<.001 b
iHOT-33	73.45 (65.1-81.8)	79.91 (72.9-86.9)	55.03 (42.7-67.4)	.001 b	.461	.017 b	.001 b
VAS pain	2.22 (1.3-3.2)	1.39 (0.9-1.9)	3.81 (2.5-5.1)	.001 b	.272	.043 b	.001 b
24-mo
mHHS	85.07 (77.7-92.4)	87.34 (82.8-91.9)	72.38 (63.4-81.3)	.021 b	.833	.057	.016 b
HOS-ADL	87.29 (80.5-94)	91.5 (88.1-94.9)	78.6 (68.6-88.6)	.028 b	.462	.180	.021 b
HOS-SS	75.04 (63.4-86.7)	83.37 (78.3-88.4)	52.3 (32.2-72.4)	.001 b	.362	.022 b	.001 b
NAHS	84.64 (76.6-92.7)	88.19 (84.7-91.7)	68.91 (51.5-86.3)	.007 b	.674	.032 b	.005 b
iHOT-33	75.12 (64.2-86.1)	78.73 (72.2-85.2)	57.01 (41.3-72.7)	.034 b	.812	.086	.026 b
VAS pain	2.43 (1.4-3.5)	1.83 (1.1-2.6)	4.01 (2.0-6.0)	.050 b	.609	.188	.039 b

^aHOS-ADL, Hip Outcome Score–Activities of Daily Living; HOS-SS, Hip Outcome Score–Sports Subscale; iHOT-33, International Hip Outcome Tool–33; mHHS, modified Harris Hip Score; NAHS, Non-Arthritic Hip Score; PI, pelvic incidence; VAS, visual analog scale.

^bA significant difference between groups.

Figure 3.

Mean patient-reported outcome measures (error bars denote SE) stratified by low (<45°), moderate (45°≤ PI ≤ 60°), and high (>60°) pelvic incidence (PI) over time. HOS-ADL, Hip Outcome Score–Activities of Daily Living; HOS-SS, Hip Outcome Score–Sports Subscale; iHOT-33, International Hip Outcome Tool–33; mHHS, modified Harris Hip Score; NAHS, Non-Arthritic Hip Score; VAS, visual analog scale.

Mixed-effects models incorporating time as a continuous variable revealed similar trends in PROMs and VAS pain scores between cohorts across the study period. Specifically, given moderate PI as the reference cohort, weighted differences in mean scores for the low-PI cohort did not reach statistical significance for any measured outcome; however, consistent with our primary analysis, the high-PI group was significantly inferior across all PROMs and VAS pain (Table 4).

Table 4.

Weighted Differences in Mean PROM Scores Between PI Groups Across the Study Period ^a

	Weighted Difference (95% CI) b	P
mHHS
Low PI	−0.98 (−5.74 to 3.78)	.690
High PI	−12.65 (−18.50 to −6.79)	<.001 c
HOS-ADL
Low PI	−1.92 (−7.35 to 3.51)	.490
High PI	−15.00 (−21.62 to −8.37)	<.001 c
HOS-SS
Low PI	−3.52 (−13.02 to 5.98)	.470
High PI	−23.25 (−34.83 to −11.67)	<.001 c
NAHS
Low PI	−2.19 (−7.84 to 3.46)	.450
High PI	−12.63 (−19.61 to −5.65)	<.001 c
iHOT-33
Low PI	−3.93 (−10.96 to 3.11)	.280
High PI	−14.80 (−23.43 to −6.18)	.0013 c
VAS pain
Low PI	0.45 (−0.25 to 1.15)	.210
High PI	1.65 (0.79 to 2.52)	<.001 c

^aHOS-ADL, Hip Outcome Score–Activities of Daily Living; HOS-SS, Hip Outcome Score–Sports Subscale; iHOT-33, International Hip Outcome Tool–33; mHHS, modified Harris Hip Score; NAHS, Non-Arthritic Hip Score; PI, pelvic incidence; PROM, patient-reported outcome measure; VAS, visual analog scale.

^bReference: moderate-PI group.

^cA significant difference between groups.

Clinically Meaningful Outcomes

When assessing rates of clinically meaningful outcome achievement, a significantly smaller proportion of patients in the high-PI cohort achieved threshold values for minimal clinically important difference, patient acceptable symptom state, substantial clinical benefit, and maximal outcome improvement at 12- and 24-month follow-ups. Relative to the high- and low-PI cohorts, the moderate-PI cohort achieved threshold values at the highest rates (Table 5).

Table 5.

Frequency of Patients Achieving Clinically Significant Thresholds ^a

	12 Mo					24 Mo
	Threshold	Low PI, %	Moderate PI, %	High PI, %	P	Threshold	Low PI, %	Moderate PI, %	High PI, %	P
MCID
mHHS	Δ >6.9	84.6	89.3	50.0	.009 c	Δ >9.2	87.5	90.0	80.0	.688
HOS-ADL	Δ >8.8	68.0	71.4	71.4	.957	Δ >9.7	54.2	66.7	80.0	.331
HOS-SS	Δ >13.9	72.0	92.9	57.1	.023 b	Δ >14.3	66.7	86.7	60.0	.118
NAHS	Δ >9.1	73.9	88.9	46.2	.014 b	Δ >8.3	66.7	86.2	75.0	.259
iHOT-33	Δ >15.1	76.0	89.3	35.7	.001 c	Δ >13.9	83.3	90.0	55.6	.091
PASS
mHHS	>84.8	46.2	53.6	21.4	.137	>83.3	66.7	73.3	30.0	.044 c
HOS-ADL	>89.7	60.0	64.3	14.3	.006 c	>88.2	66.7	80.0	20.0	.002 c
HOS-SS	>72.2	44.0	67.9	14.3	.004 b	>76.4	70.8	73.3	20.0	.007 c
NAHS	>81.9	73.9	88.9	30.8	.001 c	>85.6	66.7	65.5	25.0	.092
iHOT-33	>69.1	52.0	71.4	21.4	.009 c	>72.2	70.8	70.0	22.2	.021 c
SCB
mHHS	>86.9	46.2	53.6	21.4	.137	>85.8	54.2	56.7	10.0	.030 c
HOS-ADL	>89.7	60.0	64.3	14.3	.006 c	>91.9	50.0	70.0	20.0	.019 b
HOS-SS	>78.1	40.0	53.6	7.1	.014 c	>77.9	70.8	73.3	20.0	.007 c
NAHS	>91.9	30.4	37.0	15.4	.376	>94.4	28.6	24.1	0.00	.286
iHOT-33	>72.6	52.0	64.3	14.3	.009 c	>76.8	62.5	60.0	22.2	.093
MOI
mHHS	54.8%	53.9	71.4	21.4	.009 c	54.8	66.7	66.7	10.0	.004 c
HOS-SS	44.6%	56.0	75.0	7.1	<.001 c	44.6	75.0	80.0	30.0	.009 c
NAHS	52.5%	65.2	85.2	30.8	.003 c	52.5	57.1	75.9	25.0	.032 b
iHOT-33	55.8%	66.7	60.7	14.3	.004 c	55.8	66.7	70.0	22.2	.029 c
VAS pain	55.5%	66.7	71.4	35.7	.068	55.5	58.3	63.3	44.4	.600

^aHOS-ADL, Hip Outcome Score–Activities of Daily Living; HOS-SS, Hip Outcome Score–Sports Subscale; iHOT-33, International Hip Outcome Tool–33; MCID, minimal clinically important difference; mHHS, modified Harris Hip Score; MOI, maximal outcome improvement; NAHS, Non-Arthritic Hip Score; PASS, Patient Acceptable Symptom State; PI, pelvic incidence; SCB, substantial clinical benefit; VAS, visual analog scale.

^bHigh-PI group is significantly different from the moderate-PI group only.

^cHigh-PI group is significantly different from both the moderate- and the low-PI groups.

Patient Satisfaction and Subsequent Surgeries

Evaluating patient satisfaction via a binary metric (ie, “Are you satisfied with the treatment you received? Yes/No”) revealed significant differences between cohorts. Specifically, when comparing satisfaction between cohorts, Fisher exact tests showed a significant association between PI cohort and patient satisfaction at 12-month (P = .013) and 24-month (P = .001) follow-ups. Post hoc cell-to-cell comparisons revealed that the high-PI cohort was significantly less likely to be satisfied with their treatment relative to the moderate-PI cohort alone at the 12-month follow-up (adjusted standardized residual, −2.7; P < .05) and relative to both the moderate- and the low-PI cohorts at the 24-month follow-up (adjusted standardized residual, −3.6; P < .05).

Using the maximum available follow-up, we tracked subsequent surgeries and found that no patients underwent revision hip arthroscopy, 8 had subsequent spine surgery, and 7 were converted to THA. No significant differences were found between cohorts in terms of rates of subsequent spine surgery (P = .268), which included 1 (3.57%; lumbar fusion), 5 (16.13%; 4 lumbar decompression with fusion, 1 lumbar fusion), and 2 (13.33%; 2 lumbar laminectomy) patients for the low-, moderate-, and high-PI cohorts, respectively, at a mean time of 47.67 ± 18.65 months. Rates of subsequent THA were also not found to be significantly different (P = .055), with 7.14% (n = 2), 3.23% (n = 1), and 26.67% (n = 4) converting in the low-, moderate-, and high-PI cohorts, respectively, at a mean time of 47.76 ± 22.04 months.

Discussion

The present study demonstrates that patients with high PI (>60°) experienced inferior outcomes relative to the low-PI (<45°) and moderate-PI (45°≤ PI ≤ 60°) cohorts with respect to PROMs, rates of achieving clinically meaningful thresholds, and satisfaction at 12 and 24 months after hip arthroscopy for acetabular labral repair and FAI. Interestingly, low PI was not found to portend worse outcomes, as data trends revealed no significant differences in PROMs relative to the moderate-PI cohort. Overall, these results partially support our hypothesis and may lend credence toward the prognostic utility of assessing spinopelvic parameters during the preoperative workup for hip arthroscopy.

While FAI is the primary predisposing factor to acetabular labral tears, our results indicate that spinopelvic sagittal imbalance may play a prominent role in the development of intra-articular hip pathologies. This notion was previously validated when Mascarenhas et al ³⁴ reported that symptomatic patients could be effectively differentiated from asymptomatic volunteers using 3-dimensional MRI scans to assess FAI morphology in combination with spinopelvic parameters. Furthermore, the current study is among the first to correlate the effect of spinopelvic parameters in the context of clinical outcomes after primary hip arthroscopy. The only previous study, performed by Knapik et al, ²⁰ used a similar sample size (N = 61) but reported that spinopelvic parameters did not significantly influence PROMs when stratified by PI, PI–lumbar lordosis mismatch, or PT. Unfortunately, differences in PI allocation cutoffs, procurement of spinopelvic parameters from conventional radiographs, and omission of intra-articular characteristics (ie, labral damage, chondrolabral junction injury, chondral grading, etc) preclude any comparative analyses.^7,20,21,52

Currently, the 2 basic competing mechanisms of FAI offer an overly simplistic explanation of this condition, evidenced by the significant asymptomatic population possessing FAI deformities. ⁴⁵ Specifically, studies have estimated the prevalence of morphologic features of FAI in asymptomatic patients to be as high as 37% and 67% for cam and pincer deformities, respectively.^11,18,45 As such, increasing attention has been given to the hip-spine relationship in the nonarthritic population, as studies have reported that a high percentage (79%) of patients evaluated for low back pain additionally have positive findings when tested with provocative hip maneuvers.^34,43 While individuals with the morphological features of FAI may remain asymptomatic if the spine is sufficiently mobile to maintain sagittal alignment, limited range of motion in the spine can alter sagittal alignment and subsequently necessitate compensatory alterations in hip biomechanics that can provoke symptoms. ¹⁰ However, the etiopathogenesis of hip pain has been proposed to vary depending on the native spinopelvic orientation of the patient.

Previous literature has described patients with low PI as “hip users,” making them more reliant on hip motion because of their less intrinsic lumbopelvic motion. When these patients have FAI morphology, their increased reliance on hip motion makes them prone to experience symptoms arising from a more dynamic conflict at the acetabular rim.^42,48 As a result, patients with low PI often attempt to maintain sagittal alignment and a normal degree of lumbar lordosis by anteriorly tilting the pelvis (ie, decreasing PT).^23,52,54 This anterior tilt of the pelvis may impose overcoverage of the anterosuperior acetabulum and reduced flexion at the hip joint, a hallmark of symptomatic FAI. Although the etiopathogenesis of FAI remains highly debated, studies have gone further to suggest that cam deformities may be directly associated with decreased PI. Specifically, the extremes of hip motion required by hip users may lead to increased pathologic femoroacetabular contact that incites osseous remodeling and stress reaction bone formation. ⁵² In an osteological study, Gebhart et al ¹³ found decreased PI to be associated with an alpha angle >55° and acetabular version <15°. Similarly, a cadaveric study by Morris et al ³⁵ reported that pelvises with low PI had a significantly lower acetabular version. Along with the findings of the present study, we propose that patients with low PI may in fact be the most likely candidates to benefit from hip arthroscopy, as targeted decompression osteoplasties may effectively alleviate osseous impingement and restore native hip biomechanics to reconstitute compensatory range of motion without pain. ⁸

Conversely, patients with high PI have been previously labeled as “spine users.” This subset of patients has been hypothesized to possess more baseline lumbopelvic mobility, making them less dependent on hip motion and less prone to impingement-specific symptoms even with radiographic evidence of FAI. ⁴² However, individuals with a high PI inherently rely heavily on lumbar lordosis to maintain sagittal balance. This high degree of lordosis transduces significant mechanical stress to posterior articular joints, resulting in accelerated arthrosis and stiffening of the lumbosacral spine.^2,23,52,54 Additionally, spine users may compensate by posteriorly tilting their pelvis; however, this maladaptive postural response effectively causes functional undercoverage of the femoral head anteriorly, thereby creating a more verticalized articular loading surface of the acetabulum.^{13,23,24,27,48} Even in the absence of FAI, previous literature has suggested that the deleterious loading of the anterior acetabulum may be the impetus for acetabular labral tears, nontraumatic femoral head collapse, and accelerated rates of osteoarthritis.^14,23,24,54 Thus, impingement may not be the primary driver of symptoms for all patients with FAI, but rather the resulting compensatory adjustments seen in the setting of high PI could underlie the relatively worse functional outcomes observed in the present study.

Differences in pathophysiology among PI groups may seemingly imply different patterns of FAI (ie, higher rates of cam morphology in patients with low PI) and intra-articular pathology (ie, chondral delamination); however, such differences were not uniformly observed among cohorts. Although patients with a high PI may experience a similar pathophysiology as those with acetabular dysplasia, recent studies have found they are not significantly associated. ¹⁷ While outside the scope of the present study, the contributions of acetabular and/or femoral version are similarly becoming increasingly recognized as important factors in patients with FAI.^1,30 Cumulatively, these findings indicate that the hip should not be treated in isolation, but rather in conjunction with the entire spinopelvic complex.^16,18,45,50 Thus, given the host of growing literature highlighting the complex interplay between the hip and variations in sagittal alignment, future, high-level, prospective studies are needed to fully elucidate the clinical implications of spinopelvic parameters on outcomes after hip arthroscopy.

Limitations

The present study leverages a robust single-surgeon database to complement the limited data that exist evaluating spinopelvic parameters and outcomes after hip arthroscopy. All cohorts received similar and comprehensive preoperative workups; failed ≥3 months of nonoperative treatment, including physical therapy; underwent hip arthroscopy by a single, high-volume, fellowship-trained orthopaedic surgeon using a uniform surgical approach; and progressed through the same protocol postoperatively. However, despite the consistency of these results tracked across multiple validated PROMs, our results should be interpreted within the context of certain limitations. First, this was a retrospective investigation with limitations that are commensurate with other observational studies. As such, a causal relationship between spinopelvic parameters and inferior PROMs cannot be posited. Second, our sample size was limited given that dedicated spinopelvic imaging is not standard during the preoperative evaluation of FAI and/or labral pathologies. However, mixed-effects models treating time as a continuous variable were used to increase statistical power by linking observations for each participant.^3,44 Third, as is common practice, an a priori power analysis was conducted for our primary analysis, but additional power analyses for our secondary analysis of clinically meaningful outcome achievement (ie, categorical analysis) were not carried out. It is likely that the higher rate of conversion to THA observed in the high-PI cohort would have been significant with additional power. Fourth, despite the enhanced reliability attributed to obtaining spinopelvic measurements from CT and MRI, the use of 2 imaging modalities may have introduced variability in our recorded measurements. However, our secondary analysis revealed excellent cross-modality intermeasurement reliability.^7,52 Fifth, a substantial portion of potentially eligible patients was excluded because of lack of appropriate imaging or inadequate follow-up, leading to the possibility of selection bias. Finally, all procedures were performed by a single, high-volume, fellowship-trained surgeon, using unique techniques, which limits the external validity of these study results.^22,38

Conclusion

After hip arthroscopy, patients with a high PI (>60°) exhibited inferior PROMs, rates of achieving clinically meaningful thresholds, and satisfaction at 12 and 24 months relative to patients with low or moderate PI. Conversely, the outcomes of patients with low PI (<45°) were found to match the trajectory of those with a neutral spinopelvic alignment (45°≤ PI ≤ 60°). These findings highlight the importance of analyzing spinopelvic parameters preoperatively to prognosticate outcomes before hip arthroscopy for acetabular labral tears and FAI.