Reliability and stability of three common classifications for Legg-Calvé-Perthes disease

An AI based classifier model for lateral pillar classification of Legg-Calve-Perthes

We intended to compare the doctors with a convolutional neural network (CNN) that we had trained using our own unique method for the Lateral Pillar Classification (LPC) of Legg-Calve-Perthes Disease (LCPD). Thousands of training data sets are frequently required for artificial intelligence (AI) applications in medicine. Since we did not have enough real patient radiographs to train a CNN, we devised a novel method to obtain them. We trained the CNN model with the data we created by modifying the normal hip radiographs. No real patient radiographs were ever used during the training phase. We tested the CNN model on 81 hips with LCPD. Firstly, we detected the interobserver reliability of the whole system and then the reliability of CNN alone. Second, the consensus list was used to compare the results of 11 doctors and the CNN model. Percentage agreement and interobserver analysis revealed that CNN had good reliability (ICC = 0.868). CNN has achieved a 76.54% classification performance and outperformed 9 out of 11 doctors. The CNN, which we trained with the aforementioned method, can now provide better results than doctors. In the future, as training data evolves and improves, we anticipate that AI will perform significantly better than physicians.

Reliability of Goldberg Scoring System in the Radiographic Evaluation of Bony Union after Bone Grafting

Background

Evaluation of bony union after bone grafting is very important in orthopedic surgery. The aim of this study was to verify inter- and intraobserver reliability of the Goldberg scoring system for radiographic evaluation of bony union after bone grafting in various situations of animal models.

Methods

Twenty-seven male C57/BL6 mice, which lack the ability to synthesize galactose-alpha-1,3-galactose (GalT KO mice), and 9 C57/BL6 mice carrying a wild-type gene were used as animal models. We divided the mice into four groups. In group 1, syngenic bone grafting and intramedullary fixation were performed (9 wild type C57BL/6 mice). In group 2, allogenic bone grafting was performed (9 GalT KO mice). In group 3, an alpha-galactosidase-treated porcine xenograft was transplanted into the femur to reduce the antigenicity (9 GalT KO mice). In group 4, a non-treated porcine xenobone grafting was performed (9 GalT KO mice). The level of radiographic bony union (Goldberg method) was assessed by three orthopedic surgeons. Intra- and interobserver reliability for radiographic evaluation was assessed.

Results

In the Goldberg scoring system, most of the radiographic measurements showed substantial to almost perfect intraobserver reliability. The total score showed substantial intraobserver reliability. The kappa coefficient (κ) of the first examiner was 0.603, the κ of the second examiner was 0.790, and the κ of the third examiner was 0.758. The scoring system showed substantial interobserver reliability. The κ of the first session was 0.641 and the κ of the second session was 0.649.

Conclusions

The Goldberg scoring system is a reliable tool for radiographic evaluation of bony union after bone grafting.

Dynamic deformation of the femoral head occurs on weightbearing in Legg-Calves-Perthes disease: a translational pilot study

Aims

Though the pathogenesis of Legg-Calve-Perthes disease (LCPD) is unknown, repetitive microtrauma resulting in deformity has been postulated. The purpose of this study is to trial a novel upright MRI scanner, to determine whether any deformation occurs in femoral heads affected by LCPD with weightbearing.

Methods

Children affected by LCPD were recruited for analysis. Children received both standing weightbearing and supine scans in the MROpen upright MRI scanner, for coronal T1 GFE sequences, both hips in field of view. Parameters of femoral head height, width, and lateral extrusion of affected and unaffected hips were assessed by two independent raters, repeated at a one month interval. Inter- and intraclass correlation coefficients were determined. Standing and supine measurements were compared for each femoral head.

Results

Following rigorous protocol development in healthy age-matched volunteers, successful scanning was performed in 11 LCPD-affected hips in nine children, with seven unaffected hips therefore available for comparison. Five hips were in early stage (1 and 2) and six were in late stage (3 and 4). The mean age was 5.3 years. All hips in early-stage LCPD demonstrated dynamic deformity on weightbearing. Femoral head height decreased (mean 1.2 mm, 12.4% decrease), width increased (mean 2.5 mm, 7.2% increase), and lateral extrusion increased (median 2.5 mm, 23% increase) on standing weightbearing MRI compared to supine scans. Negligible deformation was observed in contra-lateral unaffected hips, with less deformation observed in late-stage hips. Inter- and intraclass reliability for all measured parameters was good to excellent.

Conclusion

This pilot study has described an effective novel research investigation for children with LCPD. Femoral heads in early-stage LCPD demonstrated dynamic deformity on weightbearing not previously seen, while unaffected hips did not. Expansion of this protocol will allow further translational study into the effects of loading hips with LCPD.

Cite this article: Bone Joint Open 2020;1-7:364–369.

Comparison of surgical and nonsurgical containment methods for patients with Legg-Calvé-Perthes disease of the onset ages between 6.0 and 8.0 years: Salter osteotomy versus a non-weight-bearing hip flexion-abduction brace

Since bone healing potential decreases with age, patients with Legg-Calvé-Perthes disease should receive treatment appropriate to their age group. Nonsurgical treatment is commonly applied to patients under 6.0 years of age at the onset and surgical treatment is recommended for those over 8.0 years of age, but it remains unclear which is better for those between 6.0 and 8.0 years. The aim of this retrospective study was to compare outcomes of Salter osteotomy and a non-weight-bearing brace in this age group. Inclusion criteria were unilateral Legg-Calvé-Perthes disease patients who were 6.0-8.0 years of age at the onset, who had more than 50% femoral head involvement without hinge abduction, and who underwent either Salter osteotomy (n = 35) or a non-weight-bearing hip flexion-abduction brace (n = 18). Radiological and clinical outcomes at skeletal maturity were compared between the two groups. The mean follow-up durations were 9.4 years in the Salter osteotomy group and 10.0 years in the brace group. There was no significant difference in the modified Waldenström classification at the beginning of treatment and the Catterall and modified lateral pillar classifications evaluated at the fragmentation stage between the groups. At skeletal maturity, the Stulberg classification, the sphericity deviation score, femoral head overgrowth, and the articulo-trochanteric distance were similar between the groups, but the Salter osteotomy group showed significantly smaller lateralization of the femoral head and better acetabular shape and coverage than the brace group: femoral head lateralization (P < 0.001), acetabular depth-to-width ratio (P = 0.002), Sharp angle (P < 0.001), lateral acetabular shape (P = 0.027), acetabular head index (P < 0.001). There was no significant difference in hip pain and motion between the groups. In this age group, Salter osteotomy provides better femoral head position and acetabular shape and coverage than a non-weight-bearing brace.

Radiographic classifications in Perthes disease

Background and purpose - Different radiographic classifications have been proposed for prediction of outcome in Perthes disease. We assessed whether the modified lateral pillar classification would provide more reliable interobserver agreement and prognostic value compared with the original lateral pillar classification and the Catterall classification. Patients and methods - 42 patients (38 boys) with Perthes disease were included in the interobserver study. Their mean age at diagnosis was 6.5 (3-11) years. 5 observers classified the radiographs in 2 separate sessions according to the Catterall classification, the original and the modified lateral pillar classifications. Interobserver agreement was analysed using weighted kappa statistics. We assessed the associations between the classifications and femoral head sphericity at 5-year follow-up in 37 non-operatively treated patients in a crosstable analysis (Gamma statistics for ordinal variables, γ). Results - The original lateral pillar and Catterall classifications showed moderate interobserver agreement (kappa 0.49 and 0.43, respectively) while the modified lateral pillar classification had fair agreement (kappa 0.40). The original lateral pillar classification was strongly associated with the 5-year radiographic outcome, with a mean γ correlation coefficient of 0.75 (95% CI: 0.61-0.95) among the 5 observers. The modified lateral pillar and Catterall classifications showed moderate associations (mean γ correlation coefficient 0.55 [95% CI: 0.38-0.66] and 0.64 [95% CI: 0.57-0.72], respectively). Interpretation - The Catterall classification and the original lateral pillar classification had sufficient interobserver agreement and association to late radiographic outcome to be suitable for clinical use. Adding the borderline B/C group did not increase the interobserver agreement or prognostic value of the original lateral pillar classification.

Legg-Calvé-Perthes disease: classifications and prognostic factors

INTRODUCTION

Legg-Calvé-Perthes Disease (LCPD) represents idiopathic avascular necrosis of femoral head in pediatric population. Indications for treatment depend mostly on prognosis about femoral head sphericity and hip congruence at the end of growth. The aim of this review is to highline prognostic factors of LCPD.

METHODS

Bibliographic search in PubMed allowed selection of 33 articles concerning prognostic factors and/or classification of LCPD.

CONCLUSION

Clinical factors of poor prognosis are overweight, female sex, age exceeding 6 years old, and lack of hip abduction. Radiologically, Herring's classification is consensual because of its high prognostic value and very good reproducibility. The other signs of femoral head "at-risk" and the assessment of the reduction in abduction of the femoral head in the acetabulum are also prognostic of late evolution. MRI seems to be a future tool in assessing the fate of hips in LCPD. It is likely that a better understanding of LCPD etiology would precise the prognosis of this disease.

Comparison of the results from simple radiography, from before to after Salter osteotomy, in patients with Legg-Calvé-Perthes disease

Objectives

To determine whether the clinical variables and preoperative classification of patients with Legg-Calvé-Perthes disease (LCPD) who undergo Salter osteotomy correlate with the radiographic result at the time of skeletal maturity.

Methods

In this retrospective cohort study, 47 individuals with LCPD who were treated using Salter osteotomy (1984–2004) were evaluated. The patients were evaluated according to sex, skin color, side affected and age at which osteotomy was performed. The preoperative radiographs were analyzed in accordance with the classifications of Waldenström, Catterall, Laredo and Herring. The radiographs obtained at the time of skeletal maturity were classified using the Stulberg method.

Results

The mean age at the time of surgical treatment was 82.87 months (6.9 years). The age presented a statistically significant correlation with the Stulberg grades at skeletal maturity (p < 0.001). Patients over the age of 6.12 years tended to present less favorable results. The variables of sex, skin color and side affected did not present any statistically significant correlation with the prognosis (p = 0.425; p = 0.467; p = 0.551, respectively). Only the Laredo classification presented a statistically significant correlation with the final result given by the Stulberg classification (p = 0.001). The other classifications used (Waldenström, Catterall and Herring) did not present any correlation between the time at which surgery was indicated and the postoperative result.

Conclusions

The age at which the patients underwent surgical treatment and the Laredo classification groups were the only variables that presented significant correlations with the Stulberg classification.

Interobserver and intraobserver reliability of the modified Waldenström classification system for staging of Legg-Calvé-Perthes disease

BACKGROUND

The absence of a reliable classification system for Legg-Calvé-Perthes disease has contributed to difficulty in establishing consistent management strategies and in interpreting outcome studies. The purpose of this study was to assess interobserver and intraobserver reliability of the modified Waldenström classification system among a large and diverse group of pediatric orthopaedic surgeons.

METHODS

Twenty surgeons independently completed the first two rounds of staging: two assessments of forty deidentified radiographs of patients with Legg-Calvé-Perthes disease in various stages. Ten of the twenty surgeons completed another two rounds of staging after the addition of a second pair of radiographs in sequence. Kappa values were calculated within and between each of the rounds.

RESULTS

Interobserver kappa values for the classification for surveys 1, 2, 3, and 4 were 0.81, 0.82, 0.76, and 0.80, respectively (with 0.61 to 0.80 considered substantial agreement and 0.81 to 1.0, nearly perfect agreement). Intraobserver agreement for the classification was an average of 0.88 (range, 0.77 to 0.96) between surveys 1 and 2 and an average of 0.87 (range, 0.81 to 0.94) between surveys 3 and 4.

CONCLUSIONS

The modified Waldenström classification system for staging of Legg-Calvé-Perthes disease demonstrated substantial to almost perfect agreement between and within observers across multiple rounds of study. In doing so, the results of this study provide a foundation for future validation studies, in which the classification stage will be associated with clinical outcomes.

Rotational open-wedge osteotomy improves treatment outcomes for patients older than eight years with Legg-Calve-Perthes disease in the modified lateral pillar B/C border or C group

PURPOSE

Since 2003, we have been performing rotational open-wedge osteotomy (ROWO) for older children with severe Legg-Calve-Perthes disease (LCPD). We compared the treatment outcomes before and after the introduction of ROWO for patients aged ≥ eight years, classified in the modified lateral pillar B/C border or C group.

METHODS

Before the introduction of ROWO, conservative treatment (non-weight bearing [NWB]) was performed for all cases between 1986 and 2002, whereby there were 34 hips (NWB group). The median patient age, at the first visit, was 9.1 years. According to the lateral pillar classification, 21 hips were categorized into the B/C border group and the others into the C group. Meanwhile, 14 consecutive hips underwent ROWO combined with NWB between 2003 and 2008 (ROWO+NWB group). The median patient age, at the first visit, was 9.1 years. Six hips were in the lateral pillar B/C border group and the others were classified into the C group. There were no significant differences in pre-treatment patients' demographic data.

RESULTS

At the latest follow-up, according to the Stulberg classification, there were 19 class II hips, 13 class III hips, and two class IV hips in the NWB group. In the ROWO+NWB group, there were 12 class II hips, two class III hips, and no class IV hips.

CONCLUSIONS

The application of ROWO combined with NWB in the treatment of older children with severe LCPD improved their outcomes, compared to those obtained by NWB treatment alone.

What is the usefulness of the fragmentation pattern of the femoral head in managing Legg-Calvé-Perthes disease?

Background

Within the lateral pillar classification of the Legg-Calvé-Perthes (LCP) disease, hips seem quite variable in the pattern of fragmentation as seen in radiographs. The purpose of this study was to determine: if it is possible to reliably subdivide the lateral pillar groups into femoral head fragmentation patterns, and if such a subdivision of the lateral pillar groupings is clinically useful in managing LCP disease.

Methods

Two hundred and ninety-three anteroposterior radiographs taken at the maximal fragmentation stage (189 lateral pillar B, 57 B/C border, and 47 C hips; mean bone/chronologic age at the time of first visit, 6.2/7.9 years) and at skeletal maturity (mean age, 16.6 years) were analyzed. We distinguished 3 fragmentation patterns in each pillar group based on the region of major involvement. We tested the inter- and intraobserver reliability of our classification system and analyzed the relationships between the fragmentation patterns and the Stulberg outcomes as well as other factors such as surgical treatment and age.

Results

Inter- and intraobserver consistency in fragmentation pattern assignments was found to be substantial to excellent. A statistically significant trend (p = 0.001) in the proportion of Stulberg III or IV outcomes in comparison with Stulberg I and II was only found for the different fragmentation patterns in our lateral pillar B patients: fragmentation patterns having mainly lateral-central necrosis led to poor outcomes. No significant association was found between fragmentation patterns and Stulberg outcomes in pillar groups B/C border and C.

Conclusions

Our results are consistent with the lateral pillar classification itself. Therefore, fragmentation patterns in each lateral pillar classification did not provide clinical usefulness in the management of LCP disease.

Unfavorable hip stress distribution after Legg-Calvé-Perthes syndrome: a 25-year follow-up of 135 hips

To study the effect of hip and pelvis geometry on development of the hip after Perthes disease, we determined the resultant hip force and contact hip stress distribution in a population of 135 adult hips of patients who had been treated for Perthes disease in childhood. Contra-lateral hips with no record of disease were taken as the control population. Biomechanical parameters were determined by mathematical models for resultant hip force in one-legged stance and for contact hip stress, which use as an input the geometrical parameters assessed from anteroposterior radiographs. The mathematical model for stress was upgraded to account for the deviation of the femoral head shape from spherical. No differences were found in resultant hip force and in peak contact hip stress between the hips that were in childhood subject to Perthes disease and the control population, but a considerable (148%) and significant (p < 0.001) difference was found in the contact hip stress gradient index, expressing an unfavorable, steep decrease of contact stress at the lateral acetabular rim. This finding indicates an increased risk of early coxarthritis in hips subject to Perthes disease.

How reliable are standard radiographic measures of the foot and ankle in children with achondroplasia?

BACKGROUND

Radiographic measurements are typically used in achondroplasia (ACH) during correction of lower limb alignment. However, reliabilities for the measurements on weightbearing radiographs of the foot and ankle in patients with ACH have not been described, and the differences between the ACH population and subjects without ACH likewise have not been well characterized; these issues limit the use of studies on this subject.

QUESTIONS/PURPOSES

We proposed (1) to measure the inter- and intraobserver reliability of a number of radiographic measures of ankle and foot alignment in an achondroplastic cohort of patients; and (2) to compare our radiographic measurement values with age-matched literature-based normative values.

METHODS

Ten radiographic measurements were applied to foot and ankle radiographs of 20 children (40 feet) with ACH (mean age, 10 years; range, 8-16 years). Interobserver and intraobserver reliabilities of these radiographic measurement methods were obtained and expressed by intraclass correlation coefficients (ICCs). The mean values were calculated and compared with the literature-based values.

RESULTS

The interobserver reliability was excellent for eight measurements with ICCs ranging from 0.801 to 0.962, except for lateral talo-first metatarsal angle and mediolateral column ratio, which were much lower. The intraobserver reliability was excellent for all 10 radiographic measurements with ICCs ranging from 0.812 to 0.998. Compared with existing literature-based values, all 10 measurements had a significant difference (p < 0.01).

CONCLUSIONS

We suggest tibiotalar angle, calcaneal pitch angle, tibiocalcaneal angle, talocalcaneal angle, naviculocuboid overlap, talonavicular coverage angle, metatarsal stacking angle, and AP talo-first metatarsal angle with excellent interobserver and intraobserver reliabilities should be considered preferentially in analysis of foot and ankle alignment in children with ACH.