Published online Aug 19, 2022.
https://doi.org/10.4055/cios22044
Factors Affecting the Clinical Course of Subchondral Fatigue Fracture of the Femoral Head
Abstract
Background
Subchondral fatigue fracture of the femoral head (SFFFH) is a rare disease, and its disease entity is established in recent decades. Although there are a few studies on SFFFH, most of them are case series involving around 10 cases, and the clinical course of SFFFH is still not well known. This study analyzed the factors affecting the clinical course of SFFFH.
Methods
Patients who visited our institution from October 2000 to January 2019 were retrospectively evaluated. Of eligible cases, 89 hips (80 patients) were diagnosed with SFFFH and non-surgical treatment outcomes were analyzed. Radiographs and medical charts were reviewed for following factors: the degree of femoral head collapse, the interval between the onset of hip pain and the first hospital visit, hip dysplasia, osteoarthritic changes, sex, and age.
Results
Hip pain decreased in 82 cases (92.1%) through non-surgical treatment, and 7 cases (7.9%) underwent surgery. Patients with good results of non-surgical treatment had improvement 2.9 months on average after the treatment. All cases without a collapsed femoral head (55 cases) had hip pain alleviation through non-surgical treatment. Cases with femoral head collapse of 4 mm or less and non-surgical treatment within 6 months from the onset of hip pain (22 cases) all had hip pain alleviation. Among 8 cases with femoral head collapse of 4 mm or less and non-surgical treatment after 6 months or more from the onset of hip pain, 3 underwent surgery and 1 had persistent hip pain. Those with femoral head collapse of over 4 mm (3 cases) all underwent surgery. The osteoarthritic changes, dysplastic hip, sex, and age were not statistically related to the success of non-surgical treatment.
Conclusions
The success of non-surgical treatment for SFFFH can be affected by the degree of femoral head collapse and the timing of non-surgical treatment.
Subchondral stress fracture of the femoral head (SSFFH) is a rare disease, and its disease entity was established in recent decades.1, 2) SSFFH consists of subchondral insufficiency fracture of the femoral head (SIFFH) and subchondral fatigue fracture of the femoral head (SFFFH). SIFFH occurs in patients whose bone quality is abnormal due to osteoporosis or other causes, and SFFFH occurs in healthy adults with normal bone quality.3) Although there are a few studies on the clinical outcomes of SFFFH, most of them are case series involving around 10 cases,2, 4, 5, 6) and the clinical course of SFFFH is still not well known. Patients with SFFFH without collapse recover well without surgical treatment.2, 4, 5) For the cases of collapsed SFFFH, however, the treatment is still controversial. Some studies have reported that patients with collapsed SFFFH recovered well without surgical treatment,5) whereas others have reported that patients eventually required surgical treatment.7)
The present study analyzed the factors affecting the clinical course of SFFFH. This study investigated whether the following factors could affect the success of non-surgical treatment of SFFFH: the degree of femoral head collapse, the interval between the onset of hip pain and the first visit to a hospital, the presence of hip dysplasia, osteoarthritic (OA) changes of the hip joint, sex, and age.
METHODS
Ethical Statements
The study was performed after obtaining approval from the Institutional Review Board of Seoul National University Hospital (No. H-2005-097-1123). Informed consent was waived by the Board and all methods were carried out in accordance with relevant guidelines.
Study Population
We evaluated the patients who had visited the hip clinic of the orthopedic department of Seoul National University Hospital from October 2000 to January 2019. The patients who met the following conditions were diagnosed with SFFFH:6, 8) (1) normal bone quality, (2) radiographs of hip anteroposterior and frog-leg views with the collapsed femoral head, (3) a subchondral low signal intensity band on T1-weighted magnetic resonance imaging (MRI), (4) a bone marrow edema (BME) pattern surrounding the aforementioned subchondral band, and (5) an absence of additional abnormal signal intensity band outside the subchondral fracture line. Those who satisfied the following criteria were considered to have normal bone quality: (1) age of 60 years or under, (2) without underlying disease affecting the normal bone quality such as secondary osteoporosis and osteogenesis imperfecta, and (3) with a hip Singh index score9) of 5 or 6. Those with one of the following conditions were excluded: (1) having undergone surgery for SFFFH before visiting our institution or (2) follow-up loss before the disappearance of hip pain or undergoing an operation. Finally, 89 hips in 80 patients were evaluated: 51 men and 29 women with a mean age of 35.4 years (range, 18–60 years) (Fig. 1). The mean follow-up period was 15.5 months (range, 0.3–191 months) (Table 1).
Fig. 1
Study population. MRI: magnetic resonance imaging, BME: bone marrow edema, SFFFH: subchondral fatigue fracture of the femoral head.
Table 1
Patient Characteristics
Treatment
Initially, non-surgical treatment, that is, reducing weight-bearing with a crutch gait and administering pain medications, was provided. Patients were asked to walk with a load of 30% to 50% of their body weight on the affected limb using crutches. Initial follow-up was performed with intervals of not exceeding 1 month. Then, the interval was adjusted according to the severity of hip pain. Non-surgical treatment was considered successful if the patient returned to normal daily life without hip pain. Surgery was performed under the following two conditions: (1) there was no pain relief with non-surgical treatment and arthritic changes progressed on serial radiographs; or (2) femoral head collapse was so severe (> 4 mm) that hip pain was not expected to improve with non-surgical treatment. Femoral head collapse exceeding 4 mm was considered severe according to the Association Research Circulation Osseous (ARCO) classification system.9) The type of surgery was determined based on the existence of arthritis. Total hip arthroplasty was performed if there were arthritic changes on the hip joint, and reduction with a bone graft was performed if there were no arthritic changes.
Radiological and Clinical Evaluation
Radiographs of anteroposterior and frog-leg views were taken at each visit. The radiographs were reviewed for the presence of femoral head collapse, a dysplastic hip, and OA changes. If there was a collapse, the maximum depression of the collapsed femoral head was measured using concentric circles.10) The patients were then grouped based on the degree of collapse with modification of the ARCO classification system:10) without collapse, collapse of 4 mm or less, and collapse exceeding 4 mm. The lateral center-edge angle (LCEA) was measured following Wiberg method.11) The hips with LCEAs of < 20° were defined as dysplastic hips, those with LCEAs of > 25° were defined as normal, and the others were defined as borderline dysplastic hips.11) OA changes included definite osteophytes, joint space narrowing, and subchondral cyst.
MRIs were obtained at an average interval of 3.1 months (range, 0.5–39 months) after the onset of hip pain. The location of the subchondral fracture and the extent of BME were analyzed using MRI. The location of the subchondral fracture was evaluated on the coronal, sagittal, and axial views of MRI with the longest fracture line in each image. From the medical chart, the interval between the onset of hip pain and the first hospital visit was assessed and grouped into two categories: under 6 months and 6 months and longer.
Statistical Analysis
Univariate logistic regression analysis was performed to examine possible correlations of the clinical course of SFFFH with the degree of femoral head collapse, the interval between the onset of hip pain and the first visit to a hospital, the presence of hip dysplasia, OA changes of the hip joint, sex, or age. A p-value of < 0.05 was considered significant. Statistical analysis was performed using IBM SPPS ver. 25.0 (IBM Corp., Armonk, NY, USA).
RESULTS
Overall Results
The patients with SFFFH complained of the hip joint pain after military drill, vigorous running, climbing, and martial arts. The subchondral fracture was located in the superior portion of the femoral head in all cases, except for 1 case with a posteroinferior subchondral fracture. Among the cases with fractures located in the superior portion, 48 cases (53.9%) had fractures in the anterosuperior area, 19 (21.3%) in the mid-superior area, and 12 (13.5%) in the posterosuperior area. Eight patients (9.0%) had subchondral fractures extending from the anterosuperior to mid-superior area. One patient (1.1%) had a subchondral fracture across the entire superior portion of the femoral head.
Forty-eight patients (53.9%) had BME within half of the femoral head, 19 (21.3%) in more than half of the femoral head but not exceeding the head, and 22 (24.7%) beyond the femoral head. Of all cases, 82 (92.1%) had their hip pain relieved by non-surgical treatment, and 7 (7.9%) underwent surgery. Patients with good results after non-surgical treatment had hip pain alleviation in 2.9 months after treatment (range, 0.3–12 months) (Fig. 2). Non-surgical treatment was successful in all patients who had femoral head collapse of 4 mm or less and visited the hospital within 6 months after the onset of hip pain (77 patients, 100%).
Fig. 2
A case of successful non-surgical treatment. A 28-year-old man visited the hospital 1 month after the onset of pain in his left hip. (A) The initial radiograph showed a slightly flattened femoral head. The degree of collapse was measured through the concentric circle drawn along the edge of the femoral head and centered on the hip center. The length of the largest collapse area was measured. (B) Fat-suppressed T2-weighted spin-echo magnetic resonance imaging (MRI) of the patient showed subchondral fracture (arrowheads) and bone marrow edema of the entire femoral head and neck. With non-surgical treatment, the hip pain disappeared in 4 months. (C) The 5-month follow-up radiograph showed that the femoral head remained flattened with sclerotic changes. (D) The 15-month follow-up MRI showed the healed fracture line and decreased bone marrow edema of the femoral head.
Of the 7 patients who underwent surgery, 4 underwent surgery due to their large femoral head collapse (> 4 mm) (Fig. 3A-C) at the first visit, and 3 underwent surgery following non-surgical treatment for 1, 12, and 18 months, respectively (Fig. 3D-F). None of the patients progressed from a collapse of 4 mm or less to a collapse greater than 4 mm through the conservative treatment. Reduction with a bone graft was performed in 4 cases, and total hip arthroplasty was performed in 3 cases. One of the patients who had a reduction with a bone graft showed progression of arthritic changes on serial radiographs and had persistent pain. The patient eventually underwent total hip arthroplasty 3 years after the first operation.
Fig. 3
Cases of surgical treatment. A 21-year-old man visited the hospital 1 month after the onset of pain in his left hip. (A) The initial radiograph showed large (5 mm) femoral head collapse. The degree of collapse was measured through the concentric circle drawn along the edge of the femoral head and centered on the hip center. The length of the largest collapse area was measured. (B) Reduction with bone graft surgery was done immediately after the first visit. (C) The patient has been able to perform daily life without hip pain for 4 years since the operation. (D) A 37-year-old man visited the hospital because of pain in his right hip that had started 12 months ago. The initial radiograph showed a flattened femoral head. Non-surgical treatment through reduced weight-bearing was performed for 12 months. (E) However, the hip pain did not decrease much, and the radiograph showed progressed osteoarthritic changes. (F) Finally, total hip arthroplasty was done.
Three cases failed non-surgical treatment and underwent a surgical procedure. The first case was a 37-year-old man who started to feel discomfort in his right hip after running vigorously. After 12 months of discomfort, the patient visited our institution and was diagnosed with SFFFH. Radiographs of the first visit showed a slightly flattened femoral head with minimal joint space narrowing. Although non-surgical treatment was provided, the patient felt persistent hip pain, and OA changes of the affected hip worsened. After 12 months of non-surgical treatment, total hip arthroplasty was performed (Fig. 3D-F). The second case was a 60-year-old woman who complained of right hip pain after missing a step. After 43 months of discomfort, the patient visited our institution and was diagnosed with SFFFH. A slightly flattened femoral head with joint space narrowing and osteophytes were shown on the radiographs. Non-surgical treatment was maintained for 18 months. However, the femoral head became more collapsed and OA changes progressed. Furthermore, hip pain was not alleviated; therefore, total hip arthroplasty was performed. The third case was a 21-year-old man who visited our institution 7 months after the onset of hip pain. The pain in the hip started during military drill. Radiographs of the first visit showed a collapsed femoral head with 3-mm depression. After 1 month of non-surgical treatment, the patient decided to undergo surgery due to persistent hip pain. Reduction with a bone graft was done. At 3 years from the surgery, however, OA changes progressed and pain in the hip joint continued. Eventually, revision surgery, total hip arthroplasty, was done for the patient.
On univariate logistic regression, the degree of femoral head collapse and the interval between the onset of hip pain and the first visit to a hospital showed statistical relation with the non-surgical treatment outcome; however, the presence of hip dysplasia, OA changes of the hip joint, sex, and age did not (Table 2).
Table 2
Univariate Logistic Regression for the Factors Affecting Treatment Outcomes
Degree of Femoral Head Collapse
Fifty-five cases without femoral head collapse had reduced hip pain after non-surgical treatment. Among the others, 30 had femoral head collapse of 4 mm or less, and 4 had collapse exceeding 4 mm. Of the 30 cases with femoral head collapse of 4 mm or less, 27 had hip pain disappearance after non-surgical treatment, and 3 underwent surgery: 2 cases underwent total hip arthroplasty after non-surgical treatment for 12 and 18 months each (Fig. 3D-F) and 1 case underwent reduction with a bone graft after non-surgical treatment for 1 month. All 4 cases with femoral head collapse exceeding 4 mm underwent surgery after the first visit (Fig. 4).
Fig. 4
The clinical course of subchondral fatigue fracture of the femoral head observed in this study. Op: operation.
The Interval between the Onset of Hip Pain and the First Visit
Of the 89 cases, 71 visited our institution within 6 months from the onset of hip pain, and 18 visited six months after the onset. Of the 71 cases, 46 had no collapsed femoral head at the first visit, 22 had femoral head collapse of 4 mm or less, and 3 had collapsed femoral heads exceeding 4 mm. All cases except for those with femoral head collapse exceeding 4 mm had alleviated hip pain through non-surgical treatment.
Of the 18 cases that visited our institution 6 months after the onset of hip pain, 9 had no collapsed femoral heads at the first visit, 8 had femoral head collapse of 4 mm or less, and 1 had femoral head collapse exceeding 4 mm. All 9 cases without collapse lessened their hip pain through non-surgical treatment. Of the 8 cases with femoral head collapse of 4 mm or less, 3 underwent surgery and 1 was recommended surgery for persistent pain but refused. Immediate surgery was done in 1 case with femoral head collapse exceeding 4 mm (Fig. 4).
Other Parameters
Based on LCEA, 74 cases had normal hip morphology, 12 cases had borderline dysplastic hips, and 3 cases had dysplastic hips. Surgery was performed in 4 cases of normal hips, 2 cases of borderline dysplastic hips, and 1 case of dysplastic hip. Whether hip was dysplastic or not was not associated with the success of non-surgical treatment in univariate analysis (borderline dysplastic hip, p = 0.527; dysplastic hip, p = 0.102). Eleven patients had OA changes at the first visit. Of them, 2 underwent surgery. OA changes did not show statistically significant association with the success of non-surgical treatment (p = 0.872). Sex and age were not associated with the success of non-surgical treatment in univariate analysis (p = 0.190 and p = 0.146, respectively).
DISCUSSION
In this study, hip pain was alleviated in most of the patients with SFFFH (92.1%) through non-surgical treatment in approximately 3 months. Three months seemed the general fracture-healing period required for subchondral fracture healing through reduced weight-bearing. However, non-surgical treatment was not effective in all cases, and surgery was inevitable in some cases. Based on the results of this study, the success of non-surgical treatment for SFFFH can be predicted from the degree of femoral head collapse and the interval between the onset of hip pain and the first visit to a hospital. Non-surgical treatment was successful in all patients who had femoral head collapse of 4 mm or less and visited the hospital within 6 months after the onset of hip pain. Therefore, we carefully suggest a treatment decision algorithm for SFFFH (Fig. 5).
Fig. 5
Treatment decision algorithm for subchondral fatigue fracture of the femoral head (SFFFH). Tx: treatment, Sx: symptom.
In all patients without femoral head collapse, hip pain was alleviated through non-surgical treatment. On the other hand, the patients with femoral head collapse over 4 mm immediately underwent surgery. It cannot be denied that there was a possibility of obtaining good results through non-surgical treatment for these cases. However, for reducible fractures that occurred within 3 months, we judged that providing non-surgical treatment only was not ethically right.
For the patients with femoral head collapse of 4 mm or less, the treatment outcomes differed according to the interval between the onset of hip pain and the first visit to a hospital. The period from the onset of hip pain to the first hospital visit can also be considered the period from the onset of hip pain to the start of non-surgical treatment. Non-surgical treatment was effective when started within 6 months from the onset of hip pain for the patients with femoral head collapse of 4 mm or less (22 of 22 cases, 100%) (Fig. 4). However, non-surgical treatment was not effective when started 6 months or later from the onset of hip pain for this type of patients: surgery was performed in 3 of the 8 cases and there was also a case with persistent hip pain although the patient refused to undergo surgery. In addition, the remaining 4 cases had already recovered from pain at the time of their first visit.
Previous studies have shown results consistent with those of the current study. Song et al.2) have reported that of 5 cases with SFFFH, 3 without collapsed femoral heads recovered well, and 2 with marked collapse underwent surgical management. According to Kim et al.,4) all 5 cases with SFFFH (4 without femoral head collapse and 1 with femoral head collapse of less than 4 mm) had hip pain reduction with non-surgical management. All 5 patients had visited the hospital within 1 week to 3 months from the onset of hip pain. Yamamoto et al.8) have reported that among 14 men diagnosed with SFFFH, 12 recovered well and 2 underwent surgery. However, the degree of collapse and the time from the onset of the hip pain to the hospital visit were not described.
Lee et al.5) have reported that all 9 SFFFH cases with collapsed femoral heads improved with non-surgical treatment, and surgical treatment was not necessary. However, 3 of the 9 cases had undergone surgery before visiting the institution. Besides, only 1 case had femoral head collapse exceeding 4 mm. The study did not specify which case had undergone surgery before the first visit; however, it can be presumed that the case with femoral head collapse of more than 4 mm had undergone surgery before the visit. This explains why all cases in this study showed good results with non-surgical treatment despite their collapsed femoral heads.
Kim et al.7) have reported the fate of 34 patients with SFFFH. They suggested a novel staging system for SFFFH. In patients with collapsed femoral heads and depressed articular margins (stage IIb), surgical treatment was eventually required. However, for patients with collapsed femoral heads but without depressed articular margins (stage IIa), non-surgical treatment was sufficient. The need for surgical treatment in all cases with collapsed femoral heads and depressed articular margins seems to be a conflict with Lee’s results.5) However, the degree of collapse with a depressed articular margin is likely more than 4 mm. In addition, the period from the onset of hip pain to the first visit in the study was relatively long with an average of 4.4 months (range, 0.5–16 months). These may have influenced the author’s conclusion that collapsed femoral heads required surgical treatment eventually. In addition, Iwasaki et al.6) and Ikemura et al.12) have reported 5 and 15 cases of SSFFH, respectively. However, it was difficult to say that those cases had true SFFFH because the bone quality was abnormal6) or the age of the patient group ranged from 36 to 83 years.12)
Biomechanical studies have shown that the joint contact pressure of dysplastic hips was concentrated on the lateral edge of the acetabulum.13, 14, 15) This can worsen the collapse and decrease the therapeutic effect of non-surgical treatment.15) However, hip dysplasia did not affect the success of non-surgical treatment with statistical significance in the current study. The result may be due to the sample size of our study was not sufficient to show the statistical difference.
Patients with OA changes from the first visit may have a higher risk of non-surgical treatment failure because arthritis is an irreversible progressive change. However, of the 11 cases with OA changes, only 2 (18.2%) experienced non-surgical treatment failure. Besides, OA changes did not show statistically significant association with recovery after non-surgical treatment.
Analysis of 89 cases can be insufficient to describe the true clinical course of SFFFH. However, the population size of the current study is far larger than that of other studies, and considering the rareness of the disease entity, the number of cases can be considered sufficiently large to evaluate the factors affecting the success of non-surgical treatment. Another limitation is the lack of body mass index (BMI) information in this study. Most of the BMI data of the included cases were not available because the BMI data of outpatients were not routinely measured in our institution. The relation between the progression of OA and the success of non-surgical treatment was not revealed in this study due to the short follow-up period. Arthritis is a disease progressing slowly over a long period; therefore, it is necessary to observe the results of non-surgical treatment through long-term follow-up. In addition, further research is needed on whether more arthritic changes occur in patients with SFFFH in long-term follow-up than the normal population. Finally, patients who had SFFFH but recovered well without the treatment would not have visited the hospital. Including the interval between the onset of hip pain and the first visit to the hospital as a parameter can cause a selection bias. Nevertheless, it is worth noting that patients with SFFFH who came to the hospital after more than 6 months from hip pain had poor results with conservative treatment. In conclusion, the success of non-surgical treatment for SFFFH can be affected by the degree of femoral head collapse and the timing of non-surgical treatment.
CONFLICT OF INTEREST:No potential conflict of interest relevant to this article was reported.
References
-
Wiberg G. The anatomy and roentgenographic appearance of a normal hip joint. Acta Chir Scand 1939;83 Suppl 58:7–38.
-
Cited by
MeSH Terms
Figures
1 / 5
Tables
1 / 2
