|Home | About | Journals | Submit | Contact Us | Français|
Protrusio acetabuli is identified on anteroposterior (AP) radiographs of the pelvis with an acetabular line projecting medial to the ilioischial line. We documented this radiographic sign and additional radiographic parameters in 19 patients (29 hips) with protrusio and compared the parameters to those of 29 older patients (29 hips) with advanced primary osteoarthritis (OA) but no protrusio and 12 younger patients (22 hips) with protrusio but no advanced OA. A negative acetabular roof angle and particularly large acetabular fossa were more apparent in younger patients; these hips suggest the destruction of a protrusio hip begins less in the medial joint area and more in the posteroinferior joint, and the mechanism is driven less by excessive medially directed forces but by a pincer impingement. While the indication for joint-preserving surgery currently consists primarily of a valgus femoral osteotomy based on AP radiographs and patient age, modern decision making also relies on cartilage evaluation and requires advanced surgical techniques. We conclude joint-preservation surgery must be tailored to the individual hip morphology.
Level of Evidence: Level III, prognostic study. See Guidelines for Authors for a complete description of levels of evidence.
The pathomorphology of an acetabulum protruding into the true pelvis was first described by Otto in 1816 . During the next 100 years, the interest in hip protrusion was focused on etiology and classification, while treatment recommendations remained limited. In 1935, Overgaard  presented the first useful classification into primary and secondary protrusio, which was later modified by Gilmour  into the most commonly used classification today. It distinguishes between cases explained by an underlying disease (secondary) and those without such etiological explanation (primary). Today, the list of causes of secondary protrusion ranges widely from infection, trauma, and underlying metabolic, neoplastic, and genetic diseases . Bilateral manifestation with a female predominance has typically been described for primary protrusio [1, 12]. Heredity  and racial influence  have also been reported. A number of morphological arguments including accelerated epiphyseal growth and premature fusion of the triradiate cartilage have suggested a developmental etiology  of primary protrusion; however, these have not yet been substantiated.
The consequence of acetabular protrusio is secondary osteoarthritis (OA) and has been characterized by a loss of medial joint space, while the craniolateral (superior) joint space initially remains largely unaltered. The mechanism has been explained by higher load transmission through the medial aspect of the joint [5, 18]. Even in minimal primary protrusion the femoral head has been observed to migrate medially over time. We presume the acetabular morphology of the protrusion with OA differs from the classic OA patients.
The present standard surgical treatment in middle and older age has been total joint replacement, while resection arthroplasty and even arthrodesis were historical treatment options. Anterior acetabuloplasty was first performed by Smith-Petersen  to increase motion in older patients with marked stiffness. Surgical closure of the triradiate cartilage has been proposed for the skeletally immature hip ; however, this approach has not been widely adopted, in part because of the inability to predict which hips will undergo progressive protrusion. Valgus intertrochanteric osteotomy has also been recommended in the young adult under the age of 40 years without the presence of arthritic changes [7, 19, 24].
The aims of this paper were: (1) to compare the morphology using preoperative radiographs in an arthroplasty database of a cohort of patients with acetabular protrusion with osteoarthritis to an age- and gender-matched cohort with classic osteoarthritis; (2) to characterize the morphology of young patients with acetabular protrusion and to compare these patients to the patients with acetabular protrusion with osteoarthritis; and (3) to retrospectively review the clinical and radiographic information of the young protrusion patients who underwent joint-preserving surgery.
We retrospectively compared the radiographic parameters in 19 patients (29 hips) with acetabular protrusio who underwent total hip replacement for secondary OA (protrusio OA group) to those in two groups of patients: (1) 29 patients (29 hips) with OA (but no protrusio) who were age-, gender-, and Tönnis scale  matched (OA Control group), and (2) 12 young (< 43 years of age) patients (22 hips) with protrusio being evaluated for possible joint-preserving surgery (protrusio joint preservation group). The 19 patients with protrusio included all those patients between 35 and 81 years who had undergone total hip arthroplasty. Protrusio was defined by an acetabular line crossing the ilioischial line by 3 mm (male) or 6 mm (female) on the anteroposterior (AP) view (Fig. 1A–C) [2, 13, 31].
The patient demographic information was recorded including the age, gender, hip involvement (unilateral or bilateral), type (primary or alternate diagnosis), and subsequent surgical procedures.
All patients of both groups had a standard AP radiograph of the pelvis with the patient in a supine position and cross-table lateral views. All radiographs were performed with the coccyx positioned in the midline, about 1 cm above the pubic symphysis (neutral tilt) and the obturator foramina and the greater trochanter symmetrical (neutral rotation) . All radiographs were graded using tools to measure length (mm) and angles (°) provided by the PACS by one orthopaedic surgeon (SJN). The joint space was measured at two points, medial and superior. The following parameters were measured: Tönnis angle , Sharp’s angle , the lateral center edge (LCE) angle of Wiberg , femoral head extrusion index, neck-shaft angle, cross-over sign, posterior wall sign , ilioischial line relative to acetabular fossa, center of rotation of femoral head relative to the top of the trochanter .
According to availability, additional information from the radiographic studies was also recorded including false profile radiographs , computer tomography, and magnetic resonance (MR) arthrography . These studies have been used to determine whether pathomorphological aspects can be better visualized than with AP pelvic radiographs alone and have currently been used to guide treatment for joint-preservation surgery [19, 21, 22]. Furthermore, the radiographic findings of previously operated hips from this group have been correlated to the intraoperative pathology observed during routine surgical hip dislocation . False profile views to show the posteroinferior joint space  were available from seven hips. Computer tomography was performed in four patients and MR arthrography was available from seven hips.
At the time of surgery, the amount and location of cartilage injury was recorded. Except for one hip with reversed periacetabular osteotomy only, the first step of surgery was surgical dislocation , followed by global or localized trimming of the acetabular rim, and/or relative neck lengthening , and/or reversed periacetabular osteotomy (PAO)  and/or valgus intertrochanteric osteotomy. In a reversed PAO, the protrusio acetabulum was reoriented with flexion and internal rotation of the acetabular fragment to allow impingment-free motion in flexion and internal rotation.
Histograms were performed to determine whether the data were normally distributed. Superior joint space and the medial joint space measurements were the only measurements that did not follow a normal distribution and for these we used nonparametric tests. We determined differences in superior joint space and medial joint space between all groups using the Kruskal-Wallis tests and performed post-hoc analysis with Tukey test to compare between protrusio OA and control OA cohorts and between protrusio OA and protrusio joint preservation cohorts. We determined differences in radiographic morphology (Tönnis angle, Sharp’s angle, LCE angle, FH extrusion, neck-shaft angle) between protrusio OA and control OA cohorts using one-way ANOVA with Bonferroni post-hoc analysis. We also determined differences in radiographic morphology (Tönnis angle, Sharp’s angle, LCE angle, FH extrusion, neck-shaft angle) between protrusio OA and protrusio joint preservation cohorts using one-way ANOVA with Bonferroni post-hoc analysis. Analysis was performed using SPSS software (SPSS, Inc., Chicago, IL).
We observed a number of radiographic differences between protrusion hips and the OA control group. Although the degree of joint degeneration was similar between the two cohorts (Table 1), the pattern of joint space narrowing differed markedly (Tables 2, ,3).3). In the protrusio group, the medial joint space was decreased and the superior joint space was increased when compared to the OA control group. The acetabular morphology also differed between the two groups. All hips in the protrusio group had an ilioischial line lateral to the acetabular fossa, whereas the opposite was observed in the hips of the OA control group. The posterior rim was lateral to the center of rotation in a greater percentage (p = 0.004) of the hips in the protrusion group than in the OA control group (100% versus 24%, respectively). There was only one hip in the OA control group with a positive crossover sign. Parameters that measure lateral coverage, center edge angle, and Sharp’s angle were greater in the protrusion group compared to the OA control group. The mean Tönnis angle in the protrusion group differed (p = 0.0001) from that in the OA control group and was negative (−14°) compared to (+ 2°) in the OA control group. On the femoral side, there were distinguishing radiographic characteristics between the two groups. The center of rotation of the femoral head was lower than the tip of the greater trochanter in 93% of cases in the protrusion group. The OA control group, in contrast, had a center of rotation of the femoral head lower than the tip of the trochanter in 69% of the hips. The neck-shaft angle of the protrusio group (121°) was substantially less than the OA control group (130°).
The radiographic characteristics in the group of young patients differed slightly from those with protrusion and advanced OA (Table 4). However, four hips in two young patients had a substantially negative roof angle, which we interpreted as primary and not due to bone apposition of the rim (Figs. 2, ,33A–B).
Although the superior joint space on the false profile views was still normal with little osteophytic reaction, the posteroinferior joint space was clearly narrowed in four hips (Fig. 4A), an argument against joint-preserving surgery. Computer tomography in the four patients confirmed the findings of the false profile radiography (Fig. 4B). MR arthrography showed cartilage narrowing in the posteroinferior joint from minimal to substantial in all seven hips. In addition, the acetabular fossa subjectively seemed larger than normal in all cases, extending high over the fovea capitis into the weight-bearing zone and resulting in a rather narrow cartilaginous roof in the four hips with marked negative roof angle (Fig. 3B). All seven hips had the typical MR-signs of pincer impingement with bone apposition at the acetabular rim and the femoral neck in various degrees. In one hip we observed considerable cartilage destruction seen on MR posteriorly but also anterosuperiorly.
In all 11 hips having surgical dislocation, cartilage damage from minimal to substantial could be seen at the posteroinferior acetabulum and/or posterior contour of the femoral head (Fig. 5), while the cartilage damage of the medial contour of the head was less severe. We recorded the short-term results of this individualized surgical approach (Table 5).
Because each case is unique, we selected case examples to illustrate the clinical presentation and subsequent surgical procedure. A 24-year old female patient had symptomatic post-traumatic protrusio acetabuli after nonoperative treatment of a complex acetabular fracture (Fig. 6A). We used an ilioinguinal approach to periacetabular osteotomy that allowed manipulation of the acetabular fragment such that the acetabulum was lateralized with anatomic correction of the pubic ramus with a 16-hole pelvic reconstruction plate (Fig. 6B). Less than a year after surgery, the patient had a successful vaginal delivery. One 25-year old female had osteogenesis imperfecta with limited ROM until a fatigue fracture of the acetabular fossa (Fig. 7A). We originally planned a surgical dislocation with circumferential rim trimming and relative lengthening of the femoral neck. MR arthrography revealed mild cartilage injury around the medial contour of the head. Therefore we performed a reversed periacetabular osteotomy. Three-year postoperative radiographs demonstrated a lateralized acetabulum with a horizontal roof and healing of the fatigue fracture; however, the craniomedial joint space remained narrow (Fig. 7B). At that time the patient reported pain only after strenuous activities with marked improvement in ROM and femoroacetabular clearance. A 16-year old male had bilateral protrusio acetabuli and coxa vara with right-sided medial joint space narrowing and marked restriction of motion. The patient complained of only occasional pain but desired improvement of his limited ROM. We therefore planned a staged joint preserving surgery for the right hip followed by the left hip in 6 months. The preoperative radiographs demonstrated a pincer acetabulum with a horizontal roof, and therefore, an acetabular reorientation would not address the underlying pathomorphology (Fig. 8A). We performed surgical dislocations with circumferential rim trimming with labral refixation and valgus intertrochanteric osteotomies to reduce the medializing forces. At two-year followup the right hip had a slight increase in medial joint space (Fig. 8B). The 30° valgus osteotomy improved the joint clearance with markedly improved ROM.
We describe the morphologic characteristics in hips with protrusio acetabuli with an acetabular line measuring 3 mm in males and 6 mm in females medial to the ilioischial line [1, 2, 7, 13, 19]. In the protrusio acetabuli OA group, other parameters, especially the lateral center edge angle, the roof angle, and the extrusion index but also the topography of joint space narrowing, differed from those in the control OA group; however, such pathological values have not necessarily correlated with protrusion and may occur in a normal population . Our observations suggest this constellation of abnormal radiographic measurements can be observed in protrusion even in the setting of advanced OA. The same abnormal radiographic measurements were observed in a younger group of protrusion hips but to a lesser degree; however, the absence of osteophytic reactions clearly illustrated important morphological particularities such as an excessively negative roof angle combined with an acetabular fossa extending high into the weight-bearing zone. These unique findings have never been reported in a case series, and the only other study reported on a single case by Dunlop and colleagues .
There are a number of limitations in the present study. We retrospectively identified patients from our protrusion database rather than from a prospective cohort comparison. Due to the rarity of the condition, a prospective study of protrusion acetabuli patients would not be practical due to expense and long study period. The OA control group was obtained from the hip arthroplasty database using age, gender, and Tönnis scale matched to the OA protrusion group, and the investigators attempted to create a cohort that would allow a comparison of the hip morphology of these two groups. Other morphologic characteristics (i.e., crossover sign) were not considered in the selection criteria for the OA control group, and therefore may be limit the comparison to protrusio OA group. The study did not utilize a validated, hip-specific outcome instrument and range of motion data to describe the functional outcome of the protrusion joint preservation patients. There were a limited number of patients with early- to midterm followup. Given the infrequent presentation of a young patient with protrusion acetabuli, the present series was the largest series of patients undergoing joint preservation surgery for acetabuli protrusion.
When joint-preserving surgery is being considered, abnormal acetabular morphology requires a more individual approach (Fig. 9). A careful individual and family history should be obtained, and the etiology should be determined with appropriate medical consultation. Complete physical examination including the hip range of motion, strength, and anterior and posterior impingement testing should be performed as well as an exam of the contralateral extremity. A plain radiograph with an adequate AP view provides the diagnosis. Further evaluation of protrusion candidates for joint-preserving surgery revealed the false profile radiographic view  was a very specific view to exclude hips with excessive posterior cartilage destruction in an otherwise minimally affected joint. The false profile projection has been part of the routine radiographic evaluation for acetabular dysplasia, and Lequesne  has demonstrated its value in acetabular protrusion. However, the false profile view has not found wider acceptance for imaging acetabular abnormalities. MR arthrography in the early stage of degeneration revealed that protrusio hips have reactive changes along the acetabular rim and on the head-neck junction that have typically been observed in pincer impingement [3, 11]. Dynamic gadolinium MRI of cartilage may also be useful to evaluate the degree of cartilage degeneration, but these studies may not be available at most institutions. Together with intraoperative findings of surgical hip dislocation in protrusion hips, we propose that that chronic impingement, especially its contre coup destruction of the cartilage in the posterior joint, initiates the osteoarthritic process of the protrusio hip rather than the pathologic medializing forces as discussed in the literature [7, 18, 19, 22]. The precise amount of posterior cartilage injury is not known, but additional prospective studies will attempt to determine the effect of cartilage destruction on the clinical outcome.
We recommend the majority of cases begin the procedure with a surgical dislocation of the femoral head . Open surgical dislocation will reduce the depth of the acetabulum by trimming of the acetabular rim, and labral refixation may be performed when appropriate . Some cases may require osteochondroplasty of the head-neck junction and relative lengthening of the femoral neck in order to increase the femoroacetabular clearance . In these cases, a valgus intertrochanteric osteotomy is indicated to lateralize the femoral head relative to the acetabulum.
When the MR arthrogram demonstrates an abnormality of the articular cartilage with extension of the acetabular fossa in the weight-bearing zone, trimming of the rim may be too late. Osteochondroplasty of the acetabular rim has rarely been sufficient for the treatment of the complex pathomorphology of an individual case. According to the given morphology, a femoral-sided osteotomy (ie, valgus intertrochanteric osteotomy) or a pelvis-sided osteotomy (ie, reversed periacetabular osteotomy ) may be appropriately performed simultaneously. However, the reversed periacetabular osteotomy for protrusion has been clearly more demanding than the classic periacetabular osteotomy for the dysplastic acetabulum . In reverse PAO, manipulation of the fragment for reversed PAO is more difficult because the abduction correction should be executed around the center of the femoral head. Long-term followup is necessary and may provide further information on the efficacy of the procedure to maintain hip function and, perhaps, delay the degenerative process.
We thank Thomas Guggi from the Schulthess Clinic for his help assessing the protrusio data bank.
Each author certifies that he or she has no commercial associations (eg, consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article.
Each author certifies that his institution either has waived or does not require approval for the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.