Is there a role for controlled repositioning and mini-open primary osteoplasty in the management of unstable slipped capital femoral epiphysis?

ABSTRACT

The management of unstable slipped capital femoral epiphysis is controversial with variable rates of avascular necrosis (AVN). Treatment options include in-situ stabilization, gentle/positional reduction and screw fixation and modified Dunn’s procedure (MDP). We present a technique of controlled repositioning (CRP) of the epiphysis to pre-acute slip stage, screw fixation and primary osteoplasty. Between 2015 and 2020, 38 unstable slips were treated in our institution. Of these, 14 underwent successful CRP and the rest were treated with MDP. All the 14 patients who had CRP and completed 1-year follow-up were included for this study. The head–neck angle (HNA) was measured at presentation and alpha angle, head–neck offset and AVN were assessed during follow-up. The average age was 14 years (9–18) and mean follow-up was 17.7 months (12–43). The average intraoperative flexion internal rotation before osteoplasty was −18.5° (−40° to −5°) which improved to +22.1° (+15° to +30°). The average preoperative HNA was 48.7° (34.1° to 70.7°) which improved to 18.4° (1.8° to 35.7°) post-operatively. At final follow-up, the average alpha angle and head–neck offset were 46.4° (30.9° to 64.6°) and 0.22 (0.09 to 0.96), respectively. The AVN rate in the CRP group was 7.1% compared with 20.8% in the MDP group, which was not significant (P = 0.383). Two patients had screw breakage. CRP, screw fixation and mini-open primary osteoplasty is a feasible treatment option in a subgroup of patients with unstable SCFEs. The limitation with this technique is that the final decision is made intraoperatively, and hence the patient and parents need to be counselled and consented appropriately. Level of evidence: Level IV—Case series.

INTRODUCTION

An unstable slipped capital femoral epiphysis (SCFE) by definition is a slip in which the child is unable to walk even with crutch support, irrespective of the duration of symptoms [1]. Unstable SCFEs constitute 10–20% of all slips and are usually acute or acute-on-chronic [1–4]. The sequelae or complications of SCFE are slip progression, avascular necrosis (AVN), femoroacetabular impingement (FAI), osteoarthritis and chondrolysis. AVN is the most feared complications of an unstable SCFE, and the reported rates vary between 10% and 60% [2, 5–8]. The AVN in an unstable SCFE may be attributed to the vascular insult at the time of the acute slip or iatrogenic due to the reduction manoeuvre or surgical approach [9].

The treatment of unstable SCFE is quite controversial, and various options have been described in the literature. Closed reduction and screw fixation, Parsch technique, in-situ fixation and secondary corrective osteotomy and MDP have all been described for unstable SCFE [5]. Yet, there is no consensus on the ideal management of unstable SCFEs. We have been treating unstable SCFEs with a trial of controlled repositioning (CRP), screw fixation and mini-open primary osteoplasty since 2015. All those patients in whom the CRP failed were taken up for the MDP. In this paper, we report our technique of CRP, screw fixation and primary osteoplasty for unstable SCFEs and present the complications and radiological outcomes.

MATERIALS AND METHODS

We are a tertiary care referral centre for orthopaedics and on average treat 25 children with SCFEs every year. Ethics committee approval was obtained for reviewing data of all children with SCFEs operated between January 2015 and June 2020. During this period, we had surgically managed 133 hips with SCFE, of which 95 were stable and 38 were unstable. All 38 unstable SCFEs were taken up for a trial of CRP and 14 of them in whom the repositioning was successful were proceeded for screw fixation and osteoplasty, and the rest were treated by MDP. These 14 unstable SCFEs formed the study cohort for this paper. Their demographic and intraoperative data were obtained from the hospital information system. Fisher’s exact test and point biserial correlation were done for AVN rates between groups and the correlation between continuous and categorical variables, respectively. Statistical analysis was done with SPSS v 25, and a P-value of 0.05 was considered significant.

All procedures were performed by a single surgeon with clinical fellowship and cadaveric laboratory training for safe surgical dislocation and MDP. The institutional protocol is shown in the flow chart (Fig. 1). The head–neck angle (HNA) and anterior physeal separation (APS) were measured to quantify the severity of the slip [10]. HNA was calculated as the angle subtended between the long axis of the neck and the axis of the epiphysis (perpendicular line to the line joining the ends of the physis) in the cross-table lateral radiograph. All these patients were taken up for surgery as an emergency procedure in the next available operating theatre. The decision of definitive procedure was made based on the slip angle and degree of slip on the cross-table lateral radiograph after the trial of CRP.

Fig. 1.

Institutional treatment flow chart for unstable SCFEs.

Rationale of CRP

The unstable slip is usually an acute event of slippage of the capital femoral epiphysis through a potentially weak physis similar to a traumatic separation. Many of these patients have long-standing weakness of the physis with gradual slippage, leading to chronic stable SCFE, which is often mild. In these patients, there is a sub-periosteal new bone formation on the posterior aspect of the neck and the retinacular flap is intact (Fig. 2A). In the event of acute slippage, the epiphysis grossly slips posteriorly and inferiorly. The retinacular flap becomes lax with this acute slippage (Fig. 2B). The damage to the retinacular flap and the crucial vessels in this flap could possibly happen only if there is a violent force causing displacement. The rationale of the CRP is bringing back epiphysis to pre-acute slip position by gentle internal rotation (IR), which restores the length and tension of the retinacular flap. The general rule in orthopaedics is that any acute traumatic displacement is amenable for a closed reduction unless there are specific restraints in the form of fragment button-holing, soft tissue interposition or locked displacements. Hence, we believe that in a certain proportion of patients with unstable SCFEs, it is possible to safely reposition the capital femoral epiphysis to the pre-acute slip stage and convert the severe SCFEs to mild SCFEs. The mild SCFEs could then be treated with screw fixation and mini-open osteoplasty to relieve impingement (Fig. 2C).

Fig. 2.

The rationale of controlled repositioning and osteoplasty for unstable SCFEs. (A) Pictorial representation of a mild stable SCFE with new bone formation on the posterior aspect. (B) Diagrammatic representation of an acute-on-chronic unstable SCFE. (C) Pictorial representation of CRP, screw fixation and primary osteoplasty.

Technique

The patient was positioned supine on a radiolucent table. Maintaining the limb in the same amount of external rotation during transfers is of paramount importance. Hence, an assistant was instructed to hold the affected limb to avoid any undue IR (Fig. 3A). Regional anaesthesia was given for all of these patients, and anteroposterior (AP) and cross-table lateral view of the affected hip was obtained before any attempted manipulation (Fig. 3B and C). Following this, a gentle attempt to reposition the affected limb was made by gentle, controlled IR of the affected limb. This manoeuvre would be stopped at the first point of resistance, and forced IR beyond this point was never done (Fig. 3D). In acute-on-chronic slips, we did not attempt to reduce the stable part of the slip. The hip was then gently flexed to 90° and extended back, following which a cross-table lateral view of the affected hip was checked on the image intensifier to assess the severity of slip (Fig. 3E).

Fig. 3.

(A) Patient positioned on a radiolucent table; note the external rotation position of the left lower limb. (B, C) Image intensifier view of the affected hip in anteroposterior and cross-table lateral projection showing severe slip. (D) The clinical picture following controlled repositioning (CRP). Note the neutral position of the limb with the patella and foot pointing upwards. (E) Image intensifier view of the affected hip in cross-table lateral projection after CRP showing mild slip. The residual slip, potential to cause impingement, is marked with a black dotted line. (F) Image intensifier view of the affected hip in cross-table lateral projection after screw fixation and anterior osteoplasty, AP view shown in the inset.

The repositioning was considered successful if the HNA was <30° and the slip grade was <33% after the above manoeuvre. The CRP was attempted only once, and if acceptable repositioning was not achieved, the patient was planned and positioned for open capital realignment (Online resource 1). For those patients in whom CRP was successful, the slip was stabilized with one or two 6.5 mm fully threaded cancellous screws (Fig. 3F). The flexion internal rotation (FIR) was checked with the hip in 90° flexion and an anterior mini-open osteoplasty was performed if the FIR was <15°. Through a 5 cm bikini-incision, the hip capsule was exposed through Heuter approach. Capsulotomy was done and hemarthrosis was evacuated to expose the head–neck junction (Fig. 4A and B). The torn anterior periosteum was elevated and the anterosuperior neck bump was shaved using a high-speed burr until an FIR of 15° was achieved (Fig. 4C).

Fig. 4.

(A) Intraoperative picture showing the exposure of hip through Heuter approach. (B) Closer look at the head–neck junction showing the torn anterior retinacular flap and the prominent metaphysis off the epiphysis. (C) Intraoperative picture after osteoplasty showing the head–neck offset.

Post-operative protocol

All the patients were advised strict, non-weight-bearing mobilization for 6 weeks, followed by partial weight-bearing for another 4 weeks. Hip abductor strengthening exercises were started at 6 weeks. Full weight-bearing was allowed at 10 weeks post-operatively. Follow-up radiographs were taken at 6 weeks, 10 weeks, 3 months, 6 months and yearly, until complete physeal closure [11].

Assessment

The patients were followed up clinically and radiographically for a minimum of 1 year. At each follow-up, presence of pain, limp and FIR was assessed. Radiographs at each follow-up included AP, frog-leg lateral and Dunn’s view of pelvis with both hips. The follow-up radiographs were evaluated for slip progression, AVN and implant-associated problems. FAI was assessed at the final follow-up by the presence of FIR restriction, alpha angle and anterior head–neck offset ratio (Fig. 5) [12, 13]. An alpha angle of more than 55° and a head–neck offset ratio less than 0.21 ± 0.03 were considered abnormal [14]. Two patients who had implant failure were excluded from final radiographic measurements.

Fig. 5.

(A) Measurement of α-angle in the frog-leg lateral view (40° in this example). (B) Measurement of the head–neck offset in the lateral view.

RESULTS

Of the 38 children with unstable SCFEs, CRP was successful in 14 (36.8%) patients. All these patients underwent screw fixation followed by osteoplasty, as all of them had FIR <15° (Table I). There were 12 boys and 2 girls with an average age of 14 years (9–18). The slips were equally distributed on the right side and left side. Eight were acute and six were acute-on-chronic. Past medical history was notable for pan-hypopituitarism in one patient. The precipitating event was a ground-level fall in nine, with no identifiable traumatic event in five patients.

Table I.

Details of patients treated with controlled repositioning, screw fixation and osteoplasty

No.	Age	Sex	Side	Type	APS	HNA (preoperative)	HNA (final)	Alpha angle (final)	Head–neck offset ratio (final)	FIR (before osteoplasty)	FIR (after osteoplasty)	AVN	Follow-up (months)
1	15	Male	Left	Acute	6.2	70.7	30.3	NA	NA	−30	+30	No	12
2	15	Male	Right	Acute on chronic	3.6	39.5	14.7	53	0.09	−40	+25	No	12
3	13	Male	Left	Acute	0	34.1	7.3	59.5	0.14	−20	+20	No	12
4	13	Female	Right	Acute on chronic	NA	42.6	35.7	31.7	0.15	−5	+30	No	12
5	18	Male	Right	Acute on chronic	5.9	57.8	24.1	64.6	0.2	−20	+20	No	25
6	17	Male	Right	Acute	NA	41.8	22.3	48.3	0.15	−20	+30	No	43
7	14	Male	Left	Acute on chronic	5.9	47.1	1.8	30.9	0.25	−30	+20	No	12
8	13	Male	Right	Acute on chronic	2	47	8.7	39	0.2	−15	+15	No	27
9	15	Male	Left	Acute	3.5	37.5	19.6	33.8	0.96	−10	+15	No	27
10	9	Female	Left	Acute	5.5	55	25.7	47.3	0.11	−10	25	Yes	12
11	13	Male	Left	Acute	5.6	48.8	13.5	NA	NA	−15	20	No	18
12	13	Male	Right	Acute	6.4	55.1	16.2	48.5	0.12	−20	20	No	12
13	14	Male	Left	Acute	8.3	55.2	14.2	55.3	0.15	−15	15	No	12
14	14	Male	Right	Acute on chronic	3	50.6	24.2	45.8	0.19	−10	25	No	12

The average time from acute event to surgery was 84 h (48–168). The average APS was 4.6 mm (0–8.3). The average intraoperative FIR before osteoplasty was −18.5° (−40° to −5°) which improved to +22.1° (+15° to +30°) after osteoplasty, with an average gain of 40.6°. Pre- and post-operative average HNAs were 48.7° (34.1° to 70.7°) and 18.4° (1.8° to 35.7°), respectively, with a decrease of 30.3° (6.9° to 45.3°). The average duration of follow-up was 17.7 months (12–43). At the final follow-up, the average alpha angle and head–neck offset ratio were 46.4° (30.9° to 64.6°) and 0.22(0.09–0.96), respectively. A case illustration is shown in Fig. 6.

Fig. 6.

(A) Radiograph of the pelvis with both hips in AP projection showing an acute SCFE on the right hip. (B) AP view of the right hip in AP projection at 12 months follow-up following CRP and osteoplasty showing normal hip joint with no evidence of AVN. (C) Frog-leg lateral view of the pelvis with both hips at 12 months follow-up following CRP and osteoplasty showing α-angle of 31.7° and head–neck offset ratio of 0.15.

In the CRP group, one patient (7.1%) developed mild segmental AVN of an anterosuperior quadrant of the epiphysis, noted at 4 months post-operatively. She was treated with screw removal and intralesional ibandronate injection at 7 months post-index procedure (Online resource 2). In the MDP group, five patients (20.8%) developed complete AVN. The occurrence of AVN was found to be independent of the type of procedure (CRP versus MDP) by the Fisher’s exact test (P = 0.383) (Table II). None of the patients had chondrolysis or arthritic changes at the final follow-up.

Table II.

Analysis of AVN rates between CRP and MDP

	CRP	MDP	Total
No AVN	13 (92.86%)	19 (79.17%)	32 (84.21%)
AVN	1 (7.14%)	5 (20.83%)	6 (15.79%)
Total	14	24	38

Fisher’s exact statistic value is 0.3829 (not significant at P < 0.05).

Two patients stabilized with single screws had screw breakage with slip progression due to early weight-bearing against advice. One was noted at 14 weeks and the other at 10 weeks follow-up. The first patient was treated with screw removal and MDP, and at 1-year follow-up, he showed no evidence of AVN (Online resource 3). The second patient was also advised for MDP but was lost to follow-up at 3 months. Except for this one patient, all had a minimum follow-up of 1 year. None of the other patients had slip progression following single screw fixation. All patients were able to return to their normal activity level.

DISCUSSION

The management of unstable SCFEs is controversial, and there are no clear guidelines available [15–17]. The main objectives of treatment in SCFE are: (i) stabilization of epiphysis and (ii) achievement of an impingement free range of motion close to normal. While screw fixation avoids further slippage, gaining an impingement free movement entirely depends on the severity of the slip. In-situ fixation with osteoplasty for mild-to-moderate stable SCFEs is well documented in the literature [18]. We believe that impingement in mild-to-moderate SCFEs can be tackled with osteoplasty, whereas the severe ones would need repositioning of the capital femoral epiphysis. With CRP, we convert a severe SCFE to mild (position prior to the acute event), which can be treated with screw fixation and osteoplasty. This is the basis for our method of treatment of unstable SCFEs, which is not described in the literature so far. Also, intraoperative assessment of femoral head perfusion is carried out by drilling 1.8 mm K-wire into the non-weight-bearing portion of the femoral head which is found to be a reliable indicator [19]. The safe limit of osteoplasty of the neck in SCFEs is not known, although biomechanical studies have shown reduction in strength and energy to failure after osteoplasty [20]. Osteoplasty was not done for those with HNA 30° and translation of more than 33% and protected the neck with screws to avoid fracture neck of femur.

AVN is the most significant complication of SCFE, leading to a poor outcome. Loder commented that AVN in unstable SCFEs would never be a ‘never’ event [8]. The reported AVN rates in unstable SCFE range from 10% to 80%, with most large series reporting approximately 25% [21]. The aetiology and mechanism of AVN are, however, poorly understood. The three important factors that could influence AVN incidence are slip severity, reduction manoeuvre and timing of intervention.

Slip severity and AVN

Southwick slip angle is reported to be a predictor of the development of AVN and secondary arthritis [2, 22, 23]. We did not calculate Southwick slip angle as we could not take frog-leg lateral view in our patients due to severe pain. The APS is proposed to be another indicator of the severity of acute slips, which is a sensitive and specific predictor of AVN [10]. Although APS was present in 11 out of 14 cases, with an average of 4.6 mm (0–8.3), segmental AVN was noted in only one patient with 5.5 mm of APS. There was no statistically significant association between the presence of APS and development of AVN in our study (P = 0.719).

Repositioning manoeuvre and AVN

Except for a few anecdotal case series, there are no large studies on outcomes of unreduced unstable SCFE [24, 25]. Hence, the incidence of AVN in unreduced SCFE is not known. Unacceptable rates of AVN (25–100%) have been reported in some studies and some authors have warned against reducing unstable SCFE [24–26]. On the other hand, some reports claim gentle reduction as a safe method for managing unstable SCFEs [22, 27, 28]. Maeda et al. with the help of angiography demonstrated that blood supply could return following gentle reduction and hence concluded that the reduction manoeuvre does not necessarily contribute to AVN [29]. It is to be noted that the senior-most member of the team should preferably do this manoeuvre with utmost caution. We prefer to use the terminology ‘controlled repositioning’ because, by and large, orthopaedic surgeons are known for forced manipulation of fracture whenever it comes to reduction. The crux of CRP is in accentuating the fact that the ultimate aim is to reposition the epiphysis back to the position which existed before the acute event and not to achieve anatomical reduction. In truly acute slips, one might be able to reposition it to a completely anatomical position. In contrast, in those with acute-on-chronic slips, it would be repositioned to the pre-acute slip state.

The CRP manoeuvre was successful only in 14 of the 38 unstable SCFEs in our series and the rest 24 were decided for MDP. Essentially, 37% of the unstable SCFEs were amenable for repositioning who would have otherwise been planned for MDP, which is a much more extensive surgery with high rates of AVN. In our study, the AVN rate with CRP was 7.1% compared with 20.8% with MDP. However, the combined AVN rates with CRP and MDP for unstable slips in our study was 15.7%, which is comparable to the reported AVN rates for unstable slips (10–80%) [21]. Although it is debatable that the trial of CRP could have caused increased AVN in MDP patients, we would like to emphasize that 80% of the patients who had MDP following CRP did not have AVN. Thus, the 20% AVN is fortuitous and would have happened due to unstable SCFE. Also, the fact that these patients had failed CRP implies that these could have been severe enough to cause tear of the retinacular flap.

Time to intervention and AVN

The time to reduction since acute slip is accepted to be an important risk factor for AVN [30]. Low rates of AVN have been reported in few studies in which intervention was done within 24 h of presentation [31–33]. Sankar et al. in their study on 70 unstable SCFEs did not find any relationship between the timing of surgery and the incidence of AVN [3]. However, they recommended an urgent reduction in all unstable SCFEs. Kalogrianitis et al. proposed the ‘window’ of opportunity and recommended treating unstable SCFEs within 24 h of presentation [11]. None of the children in our series underwent surgery within 24 h, and the average time from acute event to surgery was 84 h (48–168). Hence, we cannot recommend for emergency intervention within 24 h in acute unstable SCFEs.

The typical pistol grip deformity following in-situ pinning of SCFEs causes FAI, a precursor of osteoarthritis [32, 33]. Leunig et al. noted changes in the acetabular labrum and cartilage even in mild SCFEs and have suggested osteoplasty in addition to in-situ fixation [34]. All the children in our series had FIR <15˚ after CRP and screw fixation and hence underwent osteoplasty. Eight of the 14 SCFEs were truly acute similar to acute traumatic separation, and one would expect to get a perfect anatomical reduction in this situation. But we could not get an anatomical reduction in these possibly because all these eight could have had completely asymptomatic mild chronic SCFE. The average alpha angle, at final follow-up in our series, was 46.4° (30.9° to 64.6°). Only two of the 12 patients had a high alpha angle of >55° and five patients had an offset ratio of <0.15. However, none of them had symptomatic FAI.

We initially used a single 6.5 mm cancellous screw to fix unstable SCFEs. Following the experience of early fixation failure in two of our cases (although it was against our advice), we now prefer to use two screws whenever possible [11].

More surgeons now prefer anatomical reduction of the femoral epiphysis via Ganz’s MDP. This procedure is technically demanding, is highly surgeon dependent, requires specialized training, has a steep learning curve and has produced variable rates of AVN in different centres [5]. Unlike stable SCFEs, most unstable SCFEs are usually moderate to severe, and in-situ pinning for all of them may not be ideal, given the higher incidence of FAI and secondary osteoarthritis with increasing severity of slip [23]. In our institution, MDP was reserved for patients wherein we failed to reduce severe SCFEs to mild ones by CRP. Of the 24 SCFEs who underwent MDP in our study, five developed AVN (20.8%).

The management of unstable SCFEs is controversial and the controversy persists due to lack of strong evidence. In-situ fixation is the commonly practised procedure by most orthopaedic surgeons, and MDP is reserved for hip enthusiasts. Although few reports recommended gentle reduction or positional reduction, its role in unstable SCFEs is not clearly established. Based on our experience, we believe that there is definitely a role for a trial of CRP in all unstable SCFEs. This gentle manoeuvre is unlikely to increase the risk of AVN when performed with great caution and understanding of the ‘pathobiology’ of unstable SCFE. To the best of our knowledge, there is no report on the use of mini-open osteoplasty for unstable SCFEs. The limitations of our study are the small sample size and the short follow-up. A longer follow-up of these patients would be needed to see if they develop symptomatic AVN in the long term.

CONCLUSION

Controlled repositioning, screw fixation and primary osteoplasty is a feasible treatment option in a subgroup of patients with unstable SCFEs. All the patients in our series had good short-term outcome and only one had mild segmental AVN without any functional disability. The limitation with this technique is that the final decision is made intraoperatively, and hence the patient and parents need to be counselled and consented appropriately.

CONFLICT OF INTEREST STATEMENT

On behalf of all the authors, the corresponding author declares no conflict of interest with this manuscript. All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript. The authors have no financial or proprietary interests in any material discussed in this article.

FUNDING

The authors have no relevant financial or non-financial interests to disclose.