The Perils of Sleeper Plates in Multiple Hereditary Exostosis: Tibial Deformity Overcorrection Due to Tether at Empty Metaphyseal Hole

Background:

Hemi-epiphysiodesis is the mainstay of treatment for angular deformities at the knee in children with multiple hereditary exostosis (MHE). Upon deformity correction, the metaphyseal screw may be removed from the hemi-epiphysiodesis plate, the sleeper plate technique, with anticipated reimplantation of the metaphyseal screw should the original deformity recur. The aim of the present study is to compare the incidence of complications with the sleeper plate technique with complete plate removal in an MHE cohort.

Methods:

Patients under the age of 18 with MHE who underwent hemi-epiphysiodesis of the proximal tibia and/or distal femur between February 1, 2016, and February 6, 2022 with a minimum 2-year follow-up or follow-up to skeletal maturity were identified via ICD-10 codes. Patient charts and radiographic images were reviewed to assess for the bone(s) treated, the use of sleeper plates, and whether any complication occurred, including overcorrection from bony ingrowth at the empty holes or deformity recurrence.

Results:

In 13 patients, 19 knees underwent hemi-epiphysiodesis at 30 sites; 13 distal femoral and 17 proximal tibial. Of 30 plates, 18 (60%) were removed completely upon deformity correction and 3 (10%) did not require removal due to skeletal maturity. Four of 13 (30.8%) femoral plates and 5 of 17 (29.4%) tibial plates were left as sleeper plates. All 5 tibial sleeper plates developed bony ingrowth into the empty metaphyseal screw hole, which led to unintended progressive deformity overcorrection. In the majority of cases, the deformity was addressed by the removal of the plate and exophytic bone and hemi-epiphysiodesis on the other side of the affected proximal tibia with subsequent resolution of the deformity.

Conclusions:

All tibial sleeper plates developed bony ingrowth into the screwless metaphyseal hole. The bony ingrowth functioned as a tether, resulting in progressive deformity overcorrection. Sleeper plates should be avoided at the proximal tibia in patients with MHE, and extreme caution should be exercised when considering this technique at the distal femur or other sites.

Levels of Evidence:

Level III—retrospective comparative study.

Multiple hereditary exostosis (MHE) is a condition in which patients develop numerous osteochondromas, with 70% to 98% of patients with MHE having osteochondromas involving the distal femur or proximal tibia.^1–4 Such lesions may lead to leg length discrepancy and/or angular deformity at the knee.⁵ Up to one-third of patients with MHE develop genu valgum and one-sixth develop genu varum secondary to osteochondromas near the distal femoral/proximal tibial physes.³

Hemi-epiphysiodesis is the most common treatment for angular deformities at the knee in children with MHE. Surgical implants used for hemi-epiphysiodesis include staples, transphyseal screws, and tension band plating.^6–8 Sleeper plates are guided growth tension band plates in which the metaphyseal screw has been removed following correction, while the plate and epiphyseal screw are left in situ in cases where angular deformities are likely to recur following the initial correction and plate removal.⁹ Should the deformity recur, the metaphyseal screw can theoretically be reinserted to complete the tension band and effect further correction.

In a prior retrospective review of 133 limbs treated with guided growth tension band plates and screws,¹⁰ the outcomes of hemi-epiphysiodesis in a non-MHE population were reported. The authors cautioned that due to the risk of bar formation, growth arrest, and need for subsequent removal of the sleeper plate due to inadequate positioning for screw reinsertion, the practice of sleeper plates should be abandoned. Another study of 25 sleeper plates in a non-MHE population demonstrated four cases of thethering (16%) all of which improved to same degree after complete plate removal. These instances of tethering and unintended overcorrection occurred with underlying diagnosis of fibular deficiency and klinefelter syndrome. Studies on hemi-epiphysiodesis in MHE are limited, but they indicate the condition leads to a greater risk of rebound phenomenon and a slower response to guided growth than an idiopathic group.^12,13 The present study aims to assess correction and complications associated with hemi-epiphysiodesis in an MHE population. The primary outcome was the prevalence of overcorrection between sleeper plates and sites with plates removed or left in situ at maturity. Secondary outcomes were the prevalence of infection, recurrence of original deformity, skin irritation, and need for revision hemi-epiphysiodesis.

METHODS

This IRB-approved retrospective cohort study took place at a single tertiary care center. ICD-10 codes for multiple congenital exostoses or multiple hereditary osteochondromas (Q78.6, D16.9) were used to identify potential patients with MHE under the age of 18 who were treated surgical between February 1, 2016, and February 6, 2022. Charts were reviewed to identify patients with MHE who underwent hemi-epiphysiodesis of the distal femur and/or proximal tibia. Patients were included if they had at least 2 years of follow-up or if they reached skeletal maturity before this 2-year cutoff. Patient charts and radiographic images were reviewed to assess for the bone(s) treated, whether guided growth plates were left in situ, converted to sleeper plates, or entirely removed, and whether any complication occurred, including overcorrection from bony ingrowth at the empty metaphyseal holes. Descriptive statistics included means and standard deviations. A comparison of demographics and complications between sleeper and nonsleeper plates was performed with χ² and independent samples t tests. Statistics were performed using SPSS version 22.0 (IBM Corp., Armonk, NY).

RESULTS

Eighteen patients with MHE underwent guided growth during the study period. Five were excluded due to insufficient follow-up <2 years of thirteen patients 4 (30.8%) were female, with 4.1±1.2 y follow-up, (Table 1) 19 knees with 30 hemi-epiphysiodesis sites were included, 13 distal femoral and 17 proximal tibial.

TABLE 1.

Demographics

Characteristic (by patient unless specified)	Sleeper (N=3)	Nonsleeper (N=10)	P
Sex, n (%)			1.000
Female	1 (33.3)	3 (30)
Race, n (%)			0.879
White	2 (67)	5 (50)
Black or African American	1 (33)	2 (20)
Asian	0	1 (10)
American Indian or Alaskan Native	0	1 (10)
Declined	0	1 (10)
Ethnicity, n (%)			0.423
Hispanic or Latino	1 (33)	1 (10)
Follow-up, mean ± SD, y	4.3 ± 1.2	4.0 ± 1.9	0.828

Counts, percentages, and means with SDs reported.

Of the 30 plates, 18 (60%) were removed completely upon deformity correction and 3 (10%) did not require removal due to skeletal maturity. Four of 13 (30.8%) femoral plates and 5 of 17 (29.4%) tibial plates (244) were left as sleeper plates, with the metaphyseal screws removed. All 5 tibial sleeper plates developed bony ingrowth into the empty metaphyseal screw hole (Fig. 1). This led to unintended progressive deformity overcorrection (Figs. 2A–E) in all 5 tibial sleeper plates, compared with none of the nonsleeper tibial plates (P<0.001). Three of 5 overcorrected knees underwent tibial sleeper plate removal, debridement of exophytic bone, and hemi-epiphysiodesis on the other side of the affected knee with subsequent correction of deformity of the affected knee (Figs. 2F, G). The other 2 overcorrected knees underwent removal of hardware and debridement of bone with one overcorrection returning to neutral alignment and the other remaining in acceptable minimal varus alignment.

FIGURE 1.

Computed tomography (CT) coronal (A), CT axial (B) and AP x-ray (C) views of a right knee demonstrating exophytic bone growing into the screwless metaphyseal hole of the proximal tibia sleeper plate.

FIGURE 2.

Standing hip-to-ankle radiographs demonstrating right genu varus preoperatively (A) and lateral distal femur and lateral proximal tibia hemi-epiphysiodesis, at 3 (B), 6 (C), and 8 (D) months postoperatively. One year and 10 months after screw removal (E), progressive valgus deformity was noted. Four months after medial hemi-epiphysiodesis (F), the patient’s valgus deformity is improving and not yet fully corrected with medial hemi-epiphysiodesis plates. G, Valgus deformity fully corrected 1-year postmedial hemi-epiphysiodesis.

None of the 4 distal femoral sleeper plates developed bony ingrowth at the empty metaphyseal hole.

There were no complications, specifically tethering/overcorrection, rebound deformity, revision hemi-epiphysiodesis, or infection, related to the 18 plates which were completely removed upon deformity correction nor related to the three plates left in situ due to physeal closure at skeletal maturity.

DISCUSSION

In this small cohort of patients with MHE whose knee deformities were treated with tension band hemi-epiphysiodesis, a critical complication was identified related to sleeper plates. All tibial plate screwless holes developed ingrowth tethering the growth plate and leading to overcorrection (Fig. 1). None of the femoral plates in the sleeper plate cohort developed bony ingrowth, but they were all removed concomitantly with ipsilateral tibial plate removal. No patients with plates removed at the time of correction or left in situ due to maturity experienced deformity overcorrection or other complications requiring additional surgery. For the sleeper plate group, treatment with removal of hardware, removal of the exophytic bone, and hemi-epiphysiodesis on the contralateral side of the tibiae led to correction of alignment when necessary.

The sleeper plate technique theoretically allows for a potentially smaller procedure should the deformity recur. However, it should not be anticipated to reduce the total number of procedures required, and it has previously been recognized that 75% of sleeper plates are no longer at an adequate position for simple metaphyseal screw replacement and as such many sleeper plates require removal and replacement of the entire plate.¹⁰ Of note, although physeal bar tethering has previously been described with sleeper plates,¹⁰ this occurred at a rate of 3.6%-16% in non-MHE population. Although not specific to MHE, in a wide spectrum of conditions treated with sleeper plates, recurrence of deformity requiring repeat hemi-epiphysiodesis with reinsertion of the metaphyseal screw was reported at only 20%.¹⁰

In a retrospective cohort study of patients with idiopathic versus MHE-related genu valgum treated with guided growth with staples there was no difference in the percentage of patients that achieved acceptable correction (neutral or overcorrection), although patients with MHE experienced slower deformity correction.¹³ In the current series, bony tethering occurred in all tibial sleeper plates with exophytic bone tethering at the empty metaphyseal screw hole. There were no occurrences of tethering through an empty screw hole at the distal femoral plate. The cause for this difference is unknown.

This study has a number of limitations. First, the retrospective nature limits our analysis to data that were captured in the clinical encounter and documented in the chart. Second, utilization of the sleeper plate technique was surgeon-dependent and not randomized. Third, this study took place during COVID when routine follow-up visits were delayed for various reasons resulting in suboptimal surveillance following guided growth; some deformities may have been identified sooner if follow-up had been more frequent. Finally, the sample size was small related to a single center, a relatively short study period and stringent inclusion criteria. However, despite the small study size, the critical nature of this complication was clearly identified.

In conclusion, bony ingrowth to the empty metaphyseal hole should be anticipated with sleeper plates used in the setting of MHE. Sleeper plates should be avoided at the proximal tibia in patients with MHE, and extreme caution should be exercised when considering this technique at the distal femur or other sites.