Hip arthroscopy enables classification and treatment of precollapse subchondral insufficiency fracture of the femoral head associated intra-articular pathology

Subchondral insufficiency fracture of the femoral head (SIFFH) is recognized as an uncommon cause of acute hip pain and is differentiated from osteonecrosis of the femoral head. A recent study reports SIFFH occurring in military recruits and patients with osteoporosis [21]. Patients usually suffer from the acute onset of pain following minor injury and prolonged untreated SIFFH can result in rapidly progressive osteoarthritis of the hip joint [5, 21]. However, the pathomechanism of SIFFH still remains unknown.

In its precollapse stage, the diagnosis of SIFFH can be established with magnetic resonance imaging (MRI).

Magnetic Resonance Imaging (MRI) may distinguish SIFFH from osteonecrosis, osteochondritis dissecans (OCD) and transient osteoporosis of the hip (TOH). On MRI, SIFFH displays diffuse bone marrow oedema pattern on STIR and a low-intensity band on T1-weighted images. The shape of the low-intensity band in SIFFH is typically discontinuous, irregular, serpentine and parallel to the articular surface. This differentiates SIFFH from other hip pathologies, which may have different prognoses (e.g. osteonecrosis often progresses to femoral head collapse and degenerative osteoarthritis with relatively poor prognoses, whereas TOH often resolves with conservative treatment) [10]. In contrast, Pape et al. [15] have shown in a review article that SIF may be the initiating factor of what was formerly believed to be spontaneous osteonecrosis of the knee (SPONK). It is possible that SIF and osteonecrosis can be stages of the same disease even in hip joint.

SIFFH often progresses to subchondral collapse of femoral head predisposing to secondary osteoarthritis [22]. Several reports demonstrate the surgical treatment of SIFFH. Yamamoto et al. [21] described trochanteric rotational osteotomy as effective in the treatment of younger patients with SIFFH since their fractures are typically in the anterosuperior region of the femoral head. Several studies also have reported that the majority of older patients with subchondral collapse have been treated with total hip arthroplasty (THA) for secondary osteoarthritis [9, 21].

The early phase of SIFFH was typically managed with conservative treatment including non-weight bearing with crutches to minimize adverse mechanical stress against injured femoral head and non-steroidal anti-inflammatory agents [21]. Although some patients improve with conservative treatment, many cases progress to femoral head collapse despite even prolonged non-weight bearing. A minimally invasive treatment option for precollapse recalcitrant SIFFH has not been reported. As little is known of SIFFH, hip arthroscopy may aid its assessment as well as detection of associated intra-capsular pathology.

Moreover, hip arthroscopy enables minimally invasive internal fixation of osteochondral lesions and fractures of the femoral head. Arthroscopic fragment fixation using hydroxyapatite poly-lactate acid (HA/PLLA) composite offers a less invasive and effective treatment for intra-articular osteochondral lesions in a variety of joints. Uchida et al. [19] demonstrated that excellent healing and clinical outcomes for the patients with osteochondritis dissecans of the elbow were demonstrated following arthroscopic fragment fixation with HA/PLLA threaded pins. We hypothesize that hip arthroscopy is a useful tool for assessing and treating patients with precollapse SIFFH.

The purposes of this study were to investigate (1) the clinical, radiographic and arthroscopic presentation of patients with SIFFH and (2) the outcomes following arthroscopic treatment with internal fixation using hydroxyapatite poly-lactate acid threaded pins and concomitant arthroscopic treatment of associated findings.

The clinical records of 623 patients who underwent hip arthroscopic surgeries between 2009 and 2016 were retrospectively reviewed. Inclusion criteria were adult patients with precollapse SIFFH concurrent with borderline developmental dysplasia of the hip (BDDH), FAI and labral tears. We performed arthroscopic antegrade internal fixation with HA/PLLA pins, labral repair and femoral osteoplasties. No patients required open procedures. The diagnosis SIFFH was based on a comprehensive history of hip pain that began without any history of major trauma, normal radiographs with regard to the femoral head (Fig. 1a), a fracture line parallel to the articular surface on computed tomogram (CT) (Fig. 1b), diffuse femoral head bone marrow oedema pattern on MRI and subchondral low-signal-intensity band on Proton or T1-weighted MRI that was irregular, serpiginous and parallel to the articular surface (Fig. 1c). Oblique sagittal or coronal views also showed diffuse femoral head bone marrow oedema pattern (Fig. 1d) and acetabular labral tear (Fig. 1e) [21]. Patients with osteonecrosis and osteochondritis dissecans of the femoral head were excluded.

One patient was diagnosed as SIFFH at 12 months due to persistent pain after internal fixation of a femoral neck fracture and was excluded because of prior ipsilateral hip surgery. Nine patients met the inclusion criteria at minimum 12-month follow-up (median, 30.0 months, range 12–56).

Regarding other bony abnormalities associated with SIFFH, including FAI and BDDH, clinical inclusion criteria were groin pain for more than 3 months, restricted hip ROM (flexion < 105° and/or restricted internal rotation in flexion < 20°), positive flexion abduction external rotation test (FABER) and a positive anterior impingement test. The hip dial test was utilized to detect anterior capsular laxity. It is performed with the patient supine and the hip in neutral extension. The leg is internally rotated and then released to allow for relaxed external rotation. A positive test is defined as greater external rotation of the affected hip than that of the contralateral hip [4].

Radiographic evidence of a cam deformity (alpha angle > 55° or head–neck offset < 8 mm on at least one radiographic view or computed tomography CT or magnetic resonance imaging MRI), [2], and/or pincer deformity, defined as lateral CE angle (LCEA) from well-centred AP pelvic radiograph > 25° plus crossover sign, coxa profunda, or posterior wall sign, and/or prominence of the ischial spine sign. Patients with LCEA ranging from 20° to 24° were diagnosed as BDDH. Physical examination was also performed to evaluate impingement and instability of the hip joint. Exclusion criteria included patients who had only non-surgical treatment, femoral head osteonecrosis or follow-up period less than 1 year.

Nine patients met the inclusion criteria at minimum of 12-month post-operative follow-up. No patients had prior history of hip surgery.

The median age was 49 years old (range 43–65) at the time of surgery. There were five female patients and four male patients. There were five right hips and four left hips. The median BMI was 24.3 kg/m² (range, 20.1–31.8). The median period from symptom onset to surgery was 6.0 months (range 3–40 months). Relevant preoperative hip examination findings are summarized in Table 2. Preoperative radiographic findings are summarized in Table 3. Of note, mixed FAI was present in six patients, borderline developmental dysplasia (BDDH) in two patients and pincer FAI in one patient. All radiographic data are shown in Table 3.

Intra-articular findings including SIFFH lesions and status of acetabular articular cartilage, labrum and ligamentum teres are shown in Table 4. Acetabular labral tears were present in all patients. Acetabular cartilage delamination was found in all patients (MAHORN grade I in two cases, grade II in one case and grade III in three cases). Importantly, the cartilage lesions were classified by the new SIFFH classification scheme based on arthroscopic observation (Grade 1A in two cases, Grade IB in one case, Grade 1C in five cases, Grade 2A in one case). The median MHHS significantly improved from preoperatively (67.1, range 36.3–78.0) to post-operatively (96.8 range 82.5–100; p = 0.001). The median NAHS significantly improved from preoperatively (34.0, range 17–63) to post-operatively (78.0 range 61–80; p = 0.001). No radiographic osteoarthritic progression was observed by Tonnis grade. Post-operative MRI at 1 year after surgery showed bone marrow oedema completely resolved and SIFFH lesion successfully healed in all patients (Fig. 3b, c). There were no complications, revision surgeries or conversion arthroplasties following initial hip arthroscopic surgeries.

The main findings of this study are that the arthroscopic internal fixation of precollapse SIFFH with HA/PLLA thread pins in concert with arthroscopic treatment of associated bony deformities and labral tears provided clinical benefit. Union of SIFFH with resolution of bone oedema occurred in all patients. Moreover, none progressed to discernable femoral head collapse or worsening of radiographic osteoarthritis for the duration of the study. Hence, our hypothesis was supported.

All patients had labral tears, associated with the underlying bone deformities (2 BDDH, six mixed type FAI and one pincer type FAI). A pilot study by Ishihara demonstrated that SIFFH is associated with DDH [8]. A recent case report has shown that SIFFH occurred in patients with overcoverage of acetabular rim that can be implicated with pincer type of impingement [10]. A recent study has shown that eight of nine patients with rapidly destructive hip osteoarthritis associated with SIFFH had inverted labral tears [5]. In our study, our arthroscopic and radiographic findings led us to consider that labral tears associated with underlying bone abnormality can lead to micro-instability, a result of concurrent pathologies including cartilage damage and SIFFH of the femoral head in this case series. It is conceivable that labral tears may contribute to the pathomechanism of SIFFH; however, 100% association in this small series does not establish causation.

For a precollapse SIFFH lesion, non-surgical treatment is usually initially recommended. Several studies have shown that some of SIFFH patients can be managed by non-surgical treatment with survival rate from about 40 to 50% with a high (50–60%) THA conversion rate [6, 9, 14].

Previous studies evaluated the MRI findings of just SIFFH lesion without commenting on concomitant intra-articular pathologies including labral tear, cartilage damage and ligamentum teres injury. In this study, an acetabular labral tear and bony deformities such as FAI and BDDH were found in all patients.

Three of nine patients without fragment fixation could be treated non-operatively for the SIFFH but had extensive labral tears; therefore, arthroscopic labral repair with suture anchors was performed with clinical benefit. Our preferred arthroscopic management for SIFFH lesion with an associated labral tear is a labral repair with FAI correction and capsular closure.

Arthroscopic fixation of SIFFH with HA/PLLA threaded pins is indicated for early phase or precollapse SIFFH. It remains unknown whether this surgery can benefit patients with post-collapse SIFFH or patients with more advanced osteoarthrosis (Tonnis 2 or 3). In this study, the cartilage lesions were classified during diagnostic arthroscopy (Grade 1A in two cases, grade IB in one case, Grade 1C in five cases and Grade 2A in one case). A previous study demonstrated that arthroscopic fragment fixation with HA/PLLA threaded pins can provide excellent clinical outcomes and osteochondral healing for treating elbow osteochondritis dissecans [19]. In fact, HA/PLLA composite materials are widely used for osteochondral injuries with bone defects and used as a strong filler material having osteoconductive properties [18]. A recent case report describes that core decompression with void filter provided excellent clinical outcome for SIFFH [16]. An advantage of this procedure is that post-operative pin removal is not required. Arthroscopic fixation of SIFFH with HA/PLLA threaded pins may facilitate fracture healing via fragment stabilization and/or stimulation of autogenous biologic healing.

In this study, all patients had undergone and failed an initial trial of non-surgical treatment including rest, non-weight-bearing protocol and physiotherapy. Failure was defined as the persistence of symptoms and bone marrow oedema pattern on T2-weighted MRI. Several histological studies have shown that a bone marrow oedema pattern on T2-weighted MRI is frequently found in either SIFFH or osteonecrosis of the femoral head, indicative of fluid exudation due to proliferating fibroblasts and chronic inflammatory cells [11, 23]. Interestingly, the preoperative bone marrow oedema pattern dramatically diminished at one month post-operatively in all cases.

It is suggested that SIFFH in the context of hip instability may be due to the presence of an acetabular labral tear, which is associated with bone deformities. Both open and arthroscopic fragment stabilization allows bone healing and union and improves symptoms of femoral head OCD [13] [17]. Consistent with those studies, osseous union and significant clinical improvement was obtained in all of our patients that underwent arthroscopy with pin stabilization of unstable lesions in precollapse SIFFH.

From the aforementioned evidence, arthroscopic internal fixation of SIFFH with HA/PLLA thread pins along with arthroscopic treatment of associated chondrolabral and bony deformities can enhance the healing process of SIFFH. Thus, the internal fixation of the SIFFH lesion is recommended as part of the comprehensive arthroscopic treatment of recalcitrant precollapse SIFFH afflicted with this condition if the SIFFH lesion is unstable. Fragment fixation is indicated for symptomatic SIFFH lesion grades IB, IC and IIA.

In this study, preoperative physical examinations of all patients showed a positive anterior impingement sign and a positive dial test. MRI as well as arthroscopic findings also showed not only SIFFH, but also acetabular labral tears and bone abnormalities in all patients. Hence, SIFFH was associated bony deformities and acetabular labral tears (see in ``Conclusion’’). Moreover, radiographic union and clinical benefit was observed in all six patients with unstable lesions undergoing arthroscopic surgery including pin stabilization. Our findings indicate that fragment fixation combined with FAI correction, labral preservation and capsular closure can be effective for precollapse SIFFH with unstable lesions, whereas FAI correction, labral preservation and capsular closure procedure without pin fixation can be effective for precollapse SIFFH with stable lesions.

Despite its uncommon prevalence, this study provides clinically relevance by educating health providers about SIFFH as an uncommon, but distinct (via specific MRI criteria) hip condition that may cause recalcitrant symptoms potentially amenable to arthroscopic surgical intervention. The presence of SIFFH should alert the surgeon to the high likelihood of associated bony and labral pathology which may be treated with hip arthroscopy. Moreover, arthroscopic stabilization of unstable precollapse lesions with HA/PLLA threaded pins may yield clinical benefit and bony union.

This study has several limitations. First, this was a retrospective case series study without a control group. Second, a limited number of patients were included in this study. However, SIFFH is a very uncommon condition. Third, short-term follow-up limits conclusions as to the durability of the observed successful outcomes.

SIFFH is associated with bony deformities and labral tears. Promising outcomes are reported for arthroscopic treatment of precollapse SIFFH with bioabsorbable thread pin stabilization of unstable lesions and treatment of intra-capsular associated pathology in stable ones.