Radiological comparison of acetabular morphology following Dega and Pemberton osteotomies in the management of developmental dysplasia of the hip

INTRODUCTION

Developmental dysplasia of the hip (DDH) is characterized by two main pathological features: A flattened acetabular socket and femoral head malposition.^[1] In the event of a shallow acetabulum, surgical intervention by pelvic osteotomy is often required to improve femoral head coverage. Pelvic osteotomy is usually recommended for the correction of acetabular dysplasia after 18 months of age.^[2] The age of DDH patients and the severity of the deformity determine the choice between the various pelvic osteotomies. Several pelvic osteotomies have been described in the literature for children with open triradiate cartilage.^[3,4] The Salter innominate osteotomy (SIO) is a re-directional osteotomy, whereas the Pemberton osteotomy (PO) and Dega osteotomy (DO) reshape the acetabulum (acetabuloplasty) and are the most frequently used techniques. PO and DO are chosen based upon the age, acetabular obtusity, and the surgeon’s assessment of Catterall’s stability criteria.^[5] Ponseti emphasized the significance of interstitial development in determining acetabular depth, as well as the functional interplay between the femoral head and the acetabulum.^[6] A shallow acetabulum, as seen in DDH, reduces the load-bearing surface area, thereby increasing contact pressure and the risk of osteoarthritis.^[7]

PO, an incomplete transiliac osteotomy, involves a controlled cut through the ilium. The acetabular roof is rotated anteriorly and laterally, utilizing the ilioischial limb of the triradiate cartilage as the pivot point, thereby enhancing coverage of the femoral head.^[8] The PO can correct the acetabular dysplasia by adjusting the acetabular direction, correcting the deformity, and increasing the acetabular depth. This procedure reduces AI, stabilizes the hip joint, and improves the coverage of the femoral head.

The DO technique originated in Dega’s 1964 German publication, which preserved the medial iliac cortex and emphasized the importance of avoiding its perforation. On the contrary, the modifications proposed by Dega in 1974 involved partial cuts to the inner anterior and middle cortices while maintaining an intact sciatic notch. Dega described two types of incomplete transiliac osteotomies but did not clearly distinguish between them, which led to confusion. A 1964 German publication briefly referenced his first osteotomy, which later became known as the “supra-acetabular semicircular osteotomy” in the Polish literature.^[9-11] Moreover, the author did not highlight the substantial differences between the two techniques, mystifying an exact understanding of the evolutionary history of these techniques.

Quantitative metrics such as the acetabular index (AI),^[12] the acetabular depth ratio (ADR),^[12] and the Wiberg center-edge angle (CEA)^[13] are routinely used to evaluate acetabular morphology and hip congruency, thereby aiding the assessment of the severity of dysplasia. To improve understanding of hip morphology, ADR was included in the quantitative matrices, which evaluate acetabular morphology by assessing both its width and depth [Figure 1]. Several studies have reported improvements in AI and CEA angle following both SIO and PO; however, data for direct comparison between PO and DO are scarce in the literature. This study was conducted to assess radiographic acetabular development in patients with DDH who underwent PO or DO. The study clarifies the ambiguities surrounding Dega’s surgical technique, using illustrations. The study also examines how various acetabular morphological parameters, including ADR, AI, and CEA, affect surgical outcomes.

MATERIALS AND METHODS

This retrospective cohort review examined the medical records of 119 hips of 96 DDH patients who underwent PO or DO between January 2008 and September 2014. Of these, 59 (49.6%) hips were treated by PO and 60 (50.4%) underwent DO. The author performed all surgeries during a period of 10 years. Patients with a minimum post-operative follow-up period of 6 years were included in the study. Patients who underwent teratological dislocations, patients with neuromuscular disorders, hip dislocations due to cerebral palsy, and patients with less than 6 years of follow-up were excluded from the study.

Study population

PO group

This group of patients included 59 hips of 48 patients with DDH, who underwent PO. The group comprised 41 (85.4%) girls and 7 (14.6%) boys. The mean age of patients in the PO group at the time of surgery was 24.83 ± 8.0 months. The number of hips afflicted on the right side was 19 (32.2%) and on the left side were 40 (67.8%). Femoral shortening was performed in 15 (25.4%) hips.

DO group

This group comprised 60 hips in 48 patients, with 40 (67%) girls and 8 (13%) boys. Bilateral involvement was recorded in 12 (20.0%) patients. The mean age at surgery was 24.54 ± 8.1 months. Affected hips on the left side were 39 (65%). Femoral shortening was performed in 11 (18.3%) hips.

Close

Figure 1:

(A) Calculation of acetabular depth ratio after Dega osteotomy pre-operative radiograph, (B) postoperative radiograph. Acetabular width: Is the distance between the lateral sourcil edge to the lowest point of the acetabular teardrop (Yellow line). Contour of the inner acetabulum (Red line). Acetabular depth is the perpendicular distance from the midpoint of the acetabular width to the acetabular roof (Blue line).

Export to PPT

RESULTS

A total of 96 patients with a mean age of 24.69 ± 8.0 months were included in the study. Eighty-one (84.4%) were females, and 15 (15.6%) were males. Of the 96 patients, 119 hips were evaluated; 59 (49.6%) were treated with PO, and 60 (50.4%) underwent DO. The PO group had a mean follow-up period of 10.61 ± 3.1 years, which was significantly longer (P < 0.006) than that of the DO group (9.35 ± 1.5 years) [Table 1]. There was no significant difference in age at surgery between the two groups (24.83 ± 8.0 vs. 24.54 ± 8.1 years; P = 0.859). Bilateral dislocations were observed in 11 (18.6%) hips of the PO group and 12 (20.0%) hips in the DO group (P = 0.851). There was no significant difference in the affected hip between the two groups (P = 0.747). However, the Tonnis grades in the PO significantly leaned toward grades 3 and 4 compared with grades 1 and 2 in the DO group (P = 0.029) [Table 2].

Close

Figure 2:

(A-I) Schematic illustration of sequential steps of the osteotomy techniques. Left column: Dega osteotomy; Right column: Pemberton osteotomy. (a1 and a2): A simulated model of dysplastic acetabulum frontal view showing the triradiate cartilage. (b1 and b2), Osteotomy trajectory overview - frontal view (b1) • Beginning superior to the anterior inferior iliac spine • Curvilinear, backward and downward • Ends at ilioischial limb of triradiate cartilage (b2) • Curvilinear osteotomy starting above anterior inferior iliac spine• Curves gently cephaled and posteriorly.• Reaches point superior to the acetabulum midpoint.• Ends 1–1.5cm in front of the sciatic. • Directed obliquely medially toward the triradiate cartilage and stops just above it. The medial cortex is not resected. (c1 and c2) Osteotomy trajectory overview - outer pelvic view (c1) Extends backward and downward to the ilioischial limb of triradiate cartilage. For a greater lateral extension, make the outer cut more superior. For a less lateral extension, make the outer cut more parallel to the inner cut (c2). Starting above the anteroinferior iliac spine. Curving gently cephalad and posteriorly. Reaching the midpoint of acetabulum. Ends 1–1.5 cm in front of the sciatic notch. Leaves the posterior quarter intact. A higher starting point and steeper osteotomy angle enhance lateral coverage. Proximity to the acetabulum results in a thinner and more flexible acetabular fragment improved reshaping. (d1 and d2) Osteotomy trajectory: Inner view of pelvis (d1) Marking osteotomy line begins medial cut line 1–1.5 cm above anteroinferior iliac spine Curve inferiorly and posteriorly, aiming at sciatic notch. Parallel to the lateral cortex osteotomy line (d2) marking osteotomy line osteotomy is typically restricted to the anterior third in specific cases, the middle third may also be included osteotomy is typically restricted to the anterior third In specific cases, the middle third may also be included. Increasing medial iliac table involvement improves anterior converge (e1 and e2). Osteotomy frontal view (e1). Fluoroscopy guided for beginners sciatic notch starting from front to create medial and lateral cut lines ending at ilioischial limb of triradiate cartilage use wider curved osteotomes. Begin anteriorly follow medial and lateral cut lines (e2). Fluoroscopy-guided sciatic notch location. Use osteotomes from lateral cortex, starting from lateral cortex directed medially and inferiorly perform osteotomy at the inner corner of tri-radiate cartilage (TRC) will preserve the TRC. Curved osteotomes are utilized for osteotomy procedures that target the inner corner of TRC, thereby preserving the cartilage. (f1 and f2) Controlled leverage of distal fragments to open osteotomy (f1). Utilizing an osteotome or a laminar spreader distal part is tilted outward, downward, or forward. The distal fragment is hinging on the TRC (f2). Uses osteotome or laminar spreader potential greenstick fracture due to outer cortical cut spreading into greater sciatic notch. The direction of required coverage is determined intraoperatively by direct observation of the hip joint. (g1 and g2) Insertion of bone graft: Frontal view (g1). Utilized tricortical iliac crest autograft. Resected femoral segment from femoral shortening. Graft Configuration Parabola-like triangular or wedge-shaped. Inserted straddling the cortical part of inner iliac table (g2). Utilized tricortical iliac crest autograft. Resected femoral segment from femoral shortening. Graft Configuration: Wedge or triangular-shaped. Main part is positioned anteriorly. Smaller portion wedged posteriorly in front of sciatic notch. Ideally, base flush with the outer table. (h1 and h2) Graft insertion: Inner view of pelvis (h1). Preserving pelvic stability continuously, undisturbed usually, the osteotomy bone graft is stable and there is no need for internal fixation. If the bone graft is not stable, fixation with one or two Kirschner wires may be necessary (h2). Preserving pelvic stability more medial cortex continuous, undisturbed; however, since the posterior portion of the inner cortex is still intact, the outer cortical greenstick fracture does not weaken the recoil and stability at the osteotomy site. (i1 and i2) Graft Insertion: Outer view of pelvis (i1) Hinge points TRC (i2) Classic Dega osteotomy hinge points (Black Star: Sciatic notch, Red Star: Posterior pelvic cortex, Yellow Star: TRC horizontal limb in i2, Blue Star, in i1: Symphysis pubis.) Purple arrow = Direction of displacement of distal fragment and center of rotation, Blue component = Osteotome

Export to PPT

The pre-operative mean AI was comparable for the two groups, measuring 40.03° ± 7.0° for the PO group and 39.60° ± 7.2° for the DO group (P = 0.740). However, the PO group posted a significantly higher mean post-operative AI than the DO group (21.31 ± 8.5 vs. 18.07 ± 4.9, P = 0.012), higher mean AI at 1-year follow-up (20.31 ± 6.2 vs. 18.00 ± 5.0, P = 0.027), as well as a higher mean latest AI (22.53 ± 11.0 vs. 14.12 ± 4.2, P < 0.001), respectively. Nonetheless, both surgical groups showed statistically significant improvement in the AI from the pre-operative period to the final follow-up (P < 0.001), with the DO cohort demonstrating a greater degree of correction [Table 3].

There was no statistically significant difference in the pre-operative ADR between the DO group and the PO groups (16.72 ± 6.4 vs. 18.17 ± 7.8, P = 0.267). However, the mean post-operative ADR was significantly higher in the DO group compared to the PO group (43.31 ± 7.8 vs. 36.05 ± 8.6, P < 0.001). The mean ADR increase in the DO group was +26.59, which was significantly higher than the mean change of +17.88 in the PO group (P < 0.001). CEA was not measured preoperatively in all cases because of hip displacement. The CEA at 1 year in the PO group was 23.19 ± 6.1°, compared with 22.87 ± 7.9° in the DO group (P = 0.804). The final CEA was not statistically different between the two groups (27.31 ± 6.6 vs. 28.50 ± 7.7°; P = 0.365) [Table 3]. The mean change in CEA from 1 year to the latest was not statistically significant between the two groups (P = 0.200). The PO group exhibited an increase in the CEA by 4.12°, whereas the DO group increased by 5.63° (P = 0.200) [Table 3].

The final radiological examination was performed at the final follow-up visit, and the radiographs were assessed according to the Severin classification.^[21] Grade I outcomes were excellent in 83.1% of hips in the PO group, compared with 75% in the DO group. Grade II showed good results in 13.6% of hips in the PO group, compared with 16% in the DO group. Two percent of PO-treated hips and 4% of DO-treated hips had fair outcomes (Grade III). None of the hips in the PO group were classified as Grade IV, whereas one hip in the DO group was classified as Grade IV. None of the hips were assigned group V or VI [Table 4].

The total number of AVN observed in this study until the latest follow-up was 28 (23.5%) [Table 5]. In the PO group, AVN was identified in 10 (16.9%) hips, and in the DO group in 18 (30%); however, the difference was statistically insignificant [Table 5]. Type 1 AVN is generally considered a temporary, irregular ossification that tends to normalize without deformity and is therefore ignored. There was no occurrence of infection in the immediate post-operative period. Autograft dislodgment occurred in two patients and was successfully treated by extending the broomstick cast, followed by a nocturnal abduction brace until a normal AI was achieved. The study revealed nine post-operative complications following osteotomies that required surgery. There were four valgus overgrowths (captus valgum), two excessive varus deformities, and three excessive internal rotations (intoeing) that were managed with surgical intervention.

DISCUSSION

Despite several studies evaluating PO and SIO, and DO versus SIO, there is a dearth of research specifically addressing the comparative efficacy of the PO and DO. Both are classified as volume-reducing procedures and are considered incomplete osteotomies.

In the current study, post-operative radiographs demonstrated that AI in both groups had comparable corrections. AI improved over the 1^st year and at the latest follow-up; however, the DO group achieved greater improvement. This improvement may be attributable to increased coverage of the femoral head within the weight-bearing area of the acetabulum, which, in turn, promotes acetabular growth and development. In addition, the data confirm that the acetabulum’s remodeling potential persists until age 6.^[21,22] The study found that the AI variation between the two study groups can be attributed to a more stable graft positioning of DO compared to PO, leaving the medial wall of the ileum intact. It covers more of the anterior or lateral femoral head by adjusting graft size and positioning, unlike PO, which uses a single graft rather than two, as in DO.^[23-25]

DO and PO produced improvements in AI values comparable to those in prior studies.^[26-30] López-Carreño et al.^[22] observed that DO resulted in greater improvements in acetabular indices. Alassaf^[30] suggested that there is no significant difference in AI correction capacities between DO and PO; however, the precision of the corrective AI is more crucial than the type of osteotomy. ADR, like AI, is a radiographic measure reflecting acetabular development.^[31,32] A recent study by Danişman et al.^[32] indicates that ADR values have returned to normal at the latest follow-up, suggesting that DO does not reduce acetabular volume [Figure 3]. ADR measurements returned to normal levels in both groups, and the present study supports this conclusion. The key concept is that the depth-to-width ratio is altered by increasing the socket’s depth and decreasing its width, while the final proportions remain unchanged.^[26,33,34] This finding is in accordance with the results of the current study, as depicted in this illustration [Figure 3].

The CEA is another radiological parameter often used to assess the degree of femoral head coverage in DDH after PO and DO. Both osteotomies aim to enhance acetabular coverage in patients with DDH.^[13] Studies^[30,35] suggest that POs generally produce a slightly greater increase of 15–25° in CEA than DOs, 10–20°, particularly in cases of anterior acetabular deficiency. The current study found that the mean CEA was 27.31 ± 6.6 for PO and 28.50 ± 7.7 for DO (P = 0.365). The two groups showed no statistically significant difference in final CEA values. A study by Sarikaya et al., comparing POs and DOs in children aged 4–8 years, reported improved acetabular coverage, with DOs achieving better CEA outcomes than POs.^[26] According to patient-specific 3D-printed pelvic models, the PO provides greater anterior and superior coverage, which may increase CEA in some circumstances, whereas the DO provides uniform coverage in the superior, superior-anterior, and anterior areas.^[36] Consistent with the findings of Sarikaya et al.,^[26] the current study demonstrated a greater mean improvement in CEA after DO, even though the difference was not statistically significant.

The association between AVN of the femoral head and pelvic osteotomies has been debated in orthopedic research. There is contradictory data in the literature regarding whether pelvic osteotomies increase the risk of femoral head AVN. Some studies suggest that because of the extensive nature of the intervention, pelvic osteotomies, regardless of type, may increase the risk of AVN.^[37,38] On the contrary, authors^[27,39,40] evaluated the relationship between pelvic osteotomies and AVN and found no significant difference between patients who had open reduction versus open reduction with pelvic osteotomy. In the current study, all patients received the same combination of treatment comprising open reduction and pelvic osteotomy, with or without femoral osteotomy. Data analysis found no statistically significant differences between the PO and DO groups. It is well known that after PO, the distal fragment may exert pressure on the femoral head, thereby increasing the risk of AVN.^[41,42] The findings presented here demonstrate the value of achieving adequate acetabular correction during surgical procedures and warrant careful execution of the technique. Excessive correction can result in adverse outcomes, including re-dislocation, femoroacetabular impingement, or AVN. Figure 4 presents a case study of a patient with persistent hip subluxation and AVN due to coxa valga. In contrast, Figure 5 illustrates another case of a patient who developed a severe varus deformity following osteotomy, resulting from excessive acetabular correction, which necessitated subsequent procedures, including slotted acetabular augmentation and capsular reefing. Therefore, careful preoperative imaging, particularly volumetric and morphological assessments, is crucial for planning and achieving optimal femoral head coverage. Despite these potential challenges, pelvic osteotomies are useful because they improve hip stability and reduce the need for future surgeries.

Close

Figure 3:

This schematic illustration with radiographs represents how reorienting the acetabulum narrows the entrance of the socket while deepening it, comparable to tilting the upper half of a bowl. (A) In developmental dysplasia of the hip, the acetabulum undergoes morphological changes, including a loss of normal spherical shape and a shallower acetabular fossa. (B) Bending the upper portion of a bowl inward enhances coverage and biomechanics, while the inferior acetabulum, a fixed bottom edge, ensures stability.

Export to PPT

Severin’s radiographic classification enables the evaluation of the medium- and long-term outcomes of DOs and POs. In the current study, the radiographic findings for POs and DOs were 83.1% and 75.0% for type 1, and 16.7% and 13.6% for type 2, respectively. These findings were comparable between the two techniques, with the majority receiving excellent results. El-Sayed et al.^[43] with 88% (types I and II); Carvalho Filho et al.^[44] achieved better results, with 81% of hips in classifications I and II; Bhuyan^[45] with 83.3%; and Yagmurlu et al.^[46] with 74% showed satisfactory results.

Comparative analysis of both osteotomies revealed that neither PO nor DO affected the pelvic ring of the birth canal.^[47] PO is more complex, and an inappropriate technique compared to DO may lead to complications such as iatrogenic fracture of the posterior pelvic column, loss of correction, damage to Y-shaped cartilage, and premature closure of triradiate cartilage.^[48] The complication may require additional hip joint reconstructions, which can adversely affect both the hip joint condition and the child’s overall health. Iatrogenic fracture of the posterior pelvic column could compromise the primary fixation of the autograft, potentially leading to loss of surgical correction. In our series, this complication occurred in two patients. PO significantly decreases intraoperative radiation exposure compared to DO.

DO, by preserving the medial wall of the acetabulum, offers greater intrinsic stability compared to the PO. This advantage may eliminate the need for additional fixation with a K-wire. Furthermore, the method is highly adaptable, enabling a skilled surgeon to tailor the acetabular correction through the use of various-sized triangular grafts. This adaptability allows for precise targeting of the most severe acetabular deficiency.^[30] Analysis indicates that blood loss and operative time metrics are nearly equivalent between the two procedures.

The main limitation of this study was its retrospective design, which lacked follow-up clinical outcomes. Not all patients included in the study were followed until the attainment of skeletal maturity. All measurements were taken by a single author, which may introduce observer bias. Moreover, these measurements were obtained from two-dimensional radiographs, which may differ from the three-dimensional imaging assessments performed during surgery using state-of-the-art equipment.

Close

Figure 4:

Serial radiographs of a 26-month-old girl operated for left hip dislocation: (A) Preoperative radiograph with dislocation Tönnis, grade 4, (B) Immediate postoperative radiograph after Dega osteotomy demonstrating overcorrection of the left hip (C) Radiograph after 6 years, displaying a valgus appearance of the proximal femur, a shorter lateral femoral neck length, and a broken Shenton’s line. (D) Radiograph after 7 years of developing progressive valgus deformity (E) Radiograph after 9 years and, (F) after 11 years of surgical intervention demonstrating mild femoral neck deformity with coxa valga. The joint dysplasia was classified as Severin Group II.

Export to PPT

Close

Figure 5:

Serial radiographs of a 37-month-old girl diagnosed with left dysplasia of the hip. She was treated by open reduction, Dega osteotomy, and femoral shortening osteotomy. (A) Preoperative radiograph with dislocation Tönnis, grade 4, (B) Immediate postoperative radiographs of the left hip reveal overcorrection. (C) At 4 years of age, there is evident stiffness of hip. (D) At 5.5 years, signs of central physeal tethering and growth inhibition, sclerosis, and (E) At 17 years, shortening of the neck indicates coxa brevis, and at (F) 18 years, the joint dysplasia was classified as Severin Group III.

Export to PPT

CONCLUSION

The pelvic osteotomy, performed using DO and PO techniques, achieves AI and CEA values close to age norms, resulting in increased ADR and improved acetabular morphology. However, there is no “ideal” pelvic osteotomy for hip dysplasia treatment of varying severities, necessitating further analysis of surgical interventions. Consequently, it will be possible to develop an algorithm for a differentiated approach to selecting a surgical method for correcting a dysplastic acetabulum. The choice should be tailored to the desired coverage pattern and the specific acetabular architecture; both osteotomies are effective, although the DO may result in greater AI and ADR improvement in some patients and provides broader coverage.