Preoperative Accuracy of Acetabular Bone Defect Assessment in Complex Hip Revision Arthroplasty Based on 3D Life‐Sized Model: A Case Series

1. Introduction

The number of total hip arthroplasty (THA) procedures performed annually is projected to rise 71% by the year 2030, and thus the revision burden is also predicted to grow [1].

The indications for Revision Total Hip Arthroplasty (R-THA) are varied and some of the common causes are aseptic loosening (50%), instability (16%), infection (15%), debilitating pain, periprosthetic fractures, or component failure. In revision cases, massive acetabular osteolysis with Paprosky type IIC and III defects can often be found. In these situations stable fixation of the acetabular component can be very challenging, necessitating structural bone grafting, metal augments and even custom implants. Careful preoperative planning is crucial for the proper reconstruction of hip biomechanics. A comprehensive understanding of the abnormal bony anatomy is required to accurately evaluate acetabular bone defects and manage them appropriately.

The conventional 2D radiographic image often does not allow an accurate evaluation of bone defect, because the 3D extension of the defect cannot be evaluated as well as the bone stock at the level of anterior and posterior columns as well as pelvic disjunction. Computer tomography (CT) is more accurate in evaluating bone loss and fracture lines, but metal hardware in the hip usually causes significant scatter of the images. In addition, CT evaluation on a 2D screen still poses some difficulty in anatomy evaluation. This brings about increased difficulty for the surgeon, who frequently is forced to change implant type during the intervention increasing time of intervention and related complications.

Three-dimensional (3D) printed models based on CT data can aid the surgeon in planning complex acetabular reconstruction. Regarding hip pathology, currently, the major fields of application of 3D printing planning are: treatment of proximal femoral fractures [3], treatment of acetabular fractures [4], hip arthroplasties [5], and reconstructive oncological surgery [6]. According to the current literature, 3D planning facilitates surgical planning and fracture reduction, allows for reduction of the associated soft tissue trauma, it is associated with reduced intraoperative time and blood loss, it grants minimal handling of the plate, and reduced fluoroscopic screening times. In addition, when used in combination with patient specific instruments, it resulted in more accurate acetabular cup placement than traditional planning and instrumentation.

The purpose of this study is to compare the accuracy of bony defect assessment in severe acetabular defects surgery with 3D printing planning as compared to conventional 2D radiographic planning and CT.

2. Materials and Methods

In the present study we included patients who underwent total or partial revision surgery due to septic or aseptic implant loosening classified as Paprosky IIC and III, in the period between April 2019 and December 2021. Each patient signed an informed consent form prior to surgery.

We collected patients’ characteristics, including gender, age, BMI, diagnosis, and follow up. The type of implant utilized in each patient was also recorded. For each patient included we developed a 3D printed model of the femur and pelvis based on their computed tomographic (CT) scan. Clinical and radiographic evaluations were performed before and immediately after surgery and subsequently in the various outpatient clinical checks at 1 month, 3 months, 6 months and after 1 year. Clinical evaluation was performed by an orthopedic surgeon, blinded by type of surgery, and included physical examination with measurement of Harris Hip Score (HHS) and analog scale visual (VAS) for pain. ROM evaluation was performed with a universal goniometer according to standard measurement guidelines [7,8].

2.1. Radiographic evaluation and design of 3D models

All patients had conventional anteroposterior (AP) radiographs of the pelvis and Lowenstein's lateral hip projection in both preoperative and follow-up evaluations. Two independent testers, a radiologist experienced in musculoskeletal radiology and a senior orthopedic surgeon who was not involved in the surgery of these patients, were enrolled to evaluate acetabular bone defects. Disagreements between the two testers were resolved using the intraoperative description of the defect. According to the Paprosky classification [9], acetabular defects were classified on preoperative AP radiographs using four bone landmarks: radiological tear defect, Kohler's line integrity, ischial lysis, and superior defect with cranial migration of the center of rotation of the hip.

Preoperative and postoperative digital radiographs in AP view were used to assess limb length discrepancy (LLD) with Hip Artroplasty Templating 2.4.3 software running with 64-bit OsiriX v.5.8.1. Measurements were performed using the ischial tuberosities and the lesser trochanter as reference points. Preoperative and postoperative center of rotation were also measured on the vertical axis of the radiological tear and on the horizontal axis from a line passing between the radiological tears [10] (Figure 1).

The making of the 3D models arises from a CT examination of the pelvis and lower limbs. All CT images and DICOM data were sent to an external company (Medics srl) to develop a three-dimensional digital model of the pelvis and lower limbs using HA3D ™ technology. Using a computer, the 3D digital model of the pelvis and lower limbs can be rotated by any angle and axially. In addition, a certain amount of bone tissue can be eliminated by the image processing function. Therefore, in this way we were able to identify and separate the different anatomical structures by observing for example the relationship between the acetabulum and the femoral head, the anatomical characteristics of the nerve bundles, and the presence of any implants from multiple angles and directions. It was also possible to obtain various useful information for the purposes of the surgical intervention, including bone thickness, defined on the basis of chromatic variations in a color scale ranging from red (<1 mm) to green (>10 mm); the presence of bone gaps, on which measurements are made in the cranio-caudal, antero-posterior and medial-lateral direction first on the 3D model and subsequently in the operating room; and finally, the presence of osteophytes and the size and orientation of the implants to be revised (Figure 2).

Postoperative radiographs were evaluated for the presence of radiolucent lines adjacent to the acetabular implant, including any augments, according to DeLee and Charnley [11], indicating mechanical loosening of the components. The prosthetic implant was considered unstable if a radiolucent line with at least 1 mm in width crossed all three acetabular zones, upper third, medial or lower third according to DeLee and Charnley, or if a migration of components could be observed (Figure 3). The fibrous stability of the cup was characterized by a radiolucent line less than 1 mm in width crossing two of the three acetabular zones. The whole construct was considered stable in the presence of bone growth and if the prosthetic components were in close contact with the pelvic bone in the absence of radiolucent lines in at least two of the three acetabular areas [12]. Loosening was defined by the presence of a change greater than 10° in the component's abduction angle or whether the entire cup moved 6 mm or more on the vertical plane, or on the horizontal plane from its postoperative position [13]. The presence of heterotopic ossification was evaluated according to Brooker classification [14].

2.2. Statistical analysis

The analyses of all the patients were performed using SPSS for Mac (version 16.0, SPSS Inc, Chicago, Illinois). The categorical data collected were gender, preoperative diagnosis, type of acetabular bone defect, post-operative patient satisfaction, radiological loosening of the acetabular components, radiolucent lines adjacent to the acetabular implant, and/or heterotopic augmentation and ossification. The numerical data collected for the study were age, BMI, HHS values, hip ROM, LLD, and cranial migration of the hip center of rotation. The descriptive statistic was calculated. Statistical significance of improvement in HHS, hip ROM, and LLD values ​​was assessed using the Wilcoxon Two-Tailed Signed Rank Test for the paired sample because the data were not normally distributed. A P value of less than 0.05 was considered significant.

3. Results

In the study period a total of 303 hip arthroplasty revision surgeries were performed in our institution. The study group included 25 patients (20 females - 5 males), mean age 62.9 ± 10.8 years, mean BMI 26.3 ± 5.3 kg. Patients were followed for an average of 13.8 months (± 2.2 months) following the intervention (Table 1).

All surgeries were performed with a posterolateral approach with patient in lateral decubitus. Mean time of surgery was 206.1 ± 39.5 minutes, blood loss was 463 ± 176.9 ml, hospital stay was 3.7 ± 0.6 days in our operating unit and 13.6 ± 6.5 days in the dedicated rehabilitation therapy department and postoperative complications included 4 dislocations and 2 fistulas (Table 2).

During the surgery, the implants placed on the 3D model were faithfully replicated on the patient.

Of these 25 patients, 9 underwent revision of both the cup and the stem. Stem removal was performed transfemorally and subsequent synthesis of the greater trochanter using metal cerclages and screws was performed. The remaining 16 patients underwent revision of the cup alone.

A modular multi-hole tantalum cup (Zimmer, Warsaw, Indiana) was used in all patients, with a median diameter of 58.8 (range, 52-66). In 8 patients a TM augment (Zimmer, Warsaw, Indiana) was added, fixed to the pelvic bone with a median number of 3 screws (range, 2 - 6).

In 14 patients a vitamin E-infused Highwall polyethylene insert (E-Poly, Biomet) was cemented in the cup, in 10 a dual mobility metal cup (Avantage, Zimmer) was cemented in the cup, was cemented in the cup and in one patient a Constrained liner (Zimmer, Warsaw, Indiana) was required (Table 3).

Functional results were evaluated using the Harris Hip Score (HHS), the visual analog scale (VAS); and the Limb Length Discrepancy (Table 4).

The mean preoperative HHS was 35.9 ± 8.4 points and the postoperative HHS at the final follow-up was 81.2 ± 13.2 points (p < 0.001). Mean VAS decreased from 6.7 ± 1.4 points before surgery to 2.4 ± 1.0 points at final follow-up (p < 0.001). Mean LLD of the operated leg was 1.95 ± 1.23 cm of shortening before surgery and 0.57 ± 0.6 cm of shortening postoperatively (p < 0.001).

The mean vertical position of the center of rotation from the line joining the radiological tears changed from 3.5 ± 1.7 cm before surgery to 1.97 ± 0.65 cm after surgery (P< 0.05). The mean horizontal position of the center changed from 3.9 ± 1.5 cm before surgery to 3.2 ± 0.5 cm after surgery (P< 0.05). The mean acetabular component abduction angle changed 59.7° ± 29.6° before 46° ± 3.9 after surgery (P< 0.05) (Table 5).

Heterotopic ossification was found in 5 (20%) of the 25 hips. Of these, two showed a Grade I according to Brooker, two a Grade II and one a Grade III. No patient required additional surgery to remove the ossification. The number of patients with lameness decreased from 25 cases before surgery to 4 (low grade) after surgery. The preoperative and postoperative clinical and radiological parameters are listed in Table 5.

Among the 25 patients operated, 4 presented instability. Three among these required reintervention and one was reduced non surgically. Two patients developed an infection within one-year from the intervention, followed by removal of the mobile prosthetic components (insert and head) with preservation of the osteo-integrated and/or cemented prosthetic components, and subsequent targeted long-term antibiotic therapy.

The 4 reference bone landmarks (Kholer line, Tear drop, Vertical Migration, Ischiatic Lysis) were measured on the RX, CT and 3D model for all 25 cases, and subsequently compared with surgical reports. On the x-ray, medial wall (Kholer line / tear drop) was erroneously reported as affected in 3 cases. Vertical migration was miscalculated in 4 cases and ischiatic lysis was inaccurately reported in 5 cases. On the CT, medial wall was miscalculated in 5 cases, vertical migration in 2 cases, and ischiatic lysis in 4 cases. On the 3D model, the medial wall was erroneously reported as intact in 1 case and misreported as affected in 2 cases. Vertical migration was miscalculated in 2 cases and ischiatic lysis was incorrectly reported in 1 case (Table 6, Table 7 and Table 8).

5. Conclusions

From our observations, we believe that the use of the 3D model allows to accurately assess, in the preoperative phase, the type of defect, and consequently, the best surgical strategy for complex hip arthroplasty revisions. An improved preoperative evaluation allows better postoperative results, a reduction in the duration of the surgery, with a consequent reduction in infectious complications, in blood loss and also in costs for each single operation. For this reason, we believe that the benefits brought by this technique have a greater impact as compared to the costs, which, in addition, are going to decrease with the spread of this methodology.