1 Introduction

Catastrophic pelvic osteolysis typically results from excessive polyethylene wear, leading to macrophage-mediated osseous resorption. Compromise of the bone stock can cause the acetabular component of a total hip arthroplasty (THA) to shift superomedially or superolaterally.¹ This erosion may lead to pelvic discontinuity, a complication that occurs in less than 5% of all revision cases.² These cases may not be adequately addressed by standard revision components, which normally require at least 40–60% of viable bone stock for osteointegration.³ The Paprosky classification of acetabular osteolysis provides nomenclature for describing these lesions as well as potential treatment guidance.⁴ Patients with less bone loss (Type IA or IB) may benefit from standard revision components, such as a cementless acetabular cup, and those with greater bone loss may be treated with multi-hole jumbo cups.⁵ However, hips that require a triflange component typically have severe osteolysis with limited viable bone stock and are commonly characterized as Paprosky IIIA or IIIB, with IIIB representing pelvic discontinuity.^1,6

Multiple treatment modalities other than custom triflange implants have been described for these lesions, including cup-cage constructs, antiprotrusio cages, acetabular distraction with porous cups, plating with hemispherical shells, or trabecular metal augments and shells.^7,8

The goals of revision THA involving catastrophic osteolysis are to reduce pain and to preserve function, in part by restoring the biomechanical center of the hip and bridging the discontinuity, if present.^9,10 The variety of surgical options for acetabular-deficient revision procedures underscores its complexity. While there is currently no gold standard of treatment, recent literature suggests custom-triflange implants may be a favorable solution.^2,5,7–13

A triflange implant addresses acetabular lytic defects by providing multiple points of fixation outside of the acetabulum. Thin-cut CT scans are used to create a three-dimensional (3D) implant model with custom iliac, ischial, and pubic flanges that match the native anatomy.^5,6,9 Additional standard measures to increase stability can also be implemented, such as lateralized or constrained liners, as well as femoral reconstruction with larger heads.^9,14 Additional pores and hydroxyapatite coating are applied to improve ingrowth.¹⁰

There are limited studies describing long-term outcomes of the custom triflange acetabular component. A recent systematic review reported only 579 procedures in the literature with a mean follow-up time of approximately 4.10 years.¹⁰ This study examines seven cases of custom triflange implants while providing one of the longest documented mean follow-up periods of 7.39 years.

2 Methods

Seven patients with catastrophic pelvic osteolysis underwent triflange revision at a single academic institution by a single fellowship-trained adult reconstruction orthopaedic surgeon between 2002 and 2015. After appropriate sIRB approval, electronic medical records (EMRs) were retrospectively reviewed for patient demographic data and all relevant perioperative and postoperative documentation.

Paprosky classification was determined by two independent observers to stratify the degree of each patient's acetabular osteolysis. The presence or absence of pelvic discontinuity was also recorded. Indications for triflange revision surgery in addition to catastrophic osteolysis were investigated and included aseptic loosening, instability, and dislocation. All procedures were performed via the extended posterior Kocher-Langenbeck approach. Intraoperative and postoperative complications were documented as well as each patient's long-term clinical and radiographic statuses. The key intraoperative and immediate postoperative variables examined included blood loss requiring transfusion, intensive care unit admission, deep vein thrombosis, and pulmonary embolism.

This study's primary measured outcome was revision-free survivorship. Secondary outcomes included patient-reported pain and ambulatory statuses. Complications such as non-revision reoperation, gait disturbance, dislocation, and infection were also recorded. Descriptive statistics were calculated with IBM SPSS (Statistical Package for Social Sciences) Version 26.0.0.1 for Mac (Armonk, NY).

3 Results

Four patients (57.1%) were classified as Paprosky IIIA, and one patient each (14.3%) was classified as Paprosky IIIB, Paprosky IIC, and Paprosky IIB; one (14.3%) of the seven patients presented with pelvic discontinuity. Patients were predominantly female (57.1%), and one patient (14.3%) had a history of severe rheumatoid arthritis (Table 1). One patient (14.3%) had a history of Ewing's sarcoma involving the pelvis (Table 2), which necessitated his index THA (Table 3).

Primary indications for the index THA were diverse. All patients undergoing triflange revision surgery had catastrophic osteolysis; the main secondary indication was aseptic loosening (4/7, 57.1%). At the time of triflange implantation, patients were on average 51.1 years old (range: 31.6–66.2). Minimum follow-up time was 1.61 years, and the mean follow-up time was 7.39 years (median: 7.50 years, range: 1.61–16.8 years) (Table 1).

All patients required blood transfusions (Table 4). One patient (14.3%) experienced a deep vein thrombosis that led to a non-fatal pulmonary embolus (Table 5). Three of seven patients (42.9%) required admission into the intensive care unit (Table 4) for hemodynamic monitoring.

Revision-free survivorship was 85.7% (6/7), as only one patient underwent revision due to recurrent dislocations (Table 4). This patient ultimately required three revisions, involving acetabular liners and femoral components, but not the triflange itself, and also showed evidence of aseptic loosening and screw breakage at a recent follow-up visit (Table 5). A second patient experienced two dislocations, which both required reoperation via closed reduction under anesthesia (Table 5). This patient did not undergo revision because she was subsequently stable in an abduction brace; however, this patient sustained a long-term gait disturbance (Table 4). Another patient underwent an incision and drainage reoperation for an acute postoperative hematoma but did not have further deep infection.

4 Discussion

Although this study's cohort experienced a high frequency of intraoperative and immediate postoperative complications, their long-term results provide further support that a custom triflange implant is a reasonable treatment for cases of catastrophic pelvic osteolysis. After a brief review of the literature we identified 615 hips that underwent revision to a custom triflange acetabular component and summarized their findings in Table 6.^6,9,12–27 Thus, despite the small number of patients (N = 7) in this study, our cohort adds over one percent to all case data for this technique, including a substantial number of hips classified as Paprosky IIB, IIC, and IIIA. Reports on outcomes of IIB and IIC are particularly scarce.¹⁰ Kieser et al. and Berend et al. previously reported and included the use of a triflange component in Paprosky IIA, IIB, and IIC hips in the analysis of their studies.^12,18 In their analysis Kieser et al. mention that a IIA and a IIB hip that underwent triflange revision failed to show osteointegration, according to the Moore criteria.^12,28 However, they reported that the patients had good clinical outcomes that did not require reoperation. Other than Kieser et al.‘s brief mention in the analysis, neither of the studies provides further information on the outcomes based on Paprosky classification.

Long-term primary outcomes in this study were similar to those in the literature, specifically revision-free survivorship (85.7% compared to 82.7%).¹⁰ The overall complication rate in this cohort was slightly higher (42.9% compared to 29%). Additionally, dislocation was the most common complication as reported in other studies.^10,21 Because of the high dislocation rate, some authors advocate for combining the triflange implant with dual mobility cups, constrained liners, or simultaneous femoral reconstruction using a large femoral head size.^7,9,24,29,30 Revision for a constrained liner was a frequent occurrence among the studies. Joshi et al. had one patient experience dislocation in the immediate postoperative period early in the study, and as a safeguard used constrained liners during the index arthroplasty in the remaining 25 patients. As a result, none of these patients experienced dislocations.²⁶

One of our patients who dislocated postoperatively had a history of developmental dysplasia of the hip (DDH). The bony hypoplasia and concomitant abductor insufficiency inherent to DDH is a risk factor for dislocation.³¹ The patient's instability and postoperative Trendelenburg gait may have been exacerbated by the proximal extension of the surgery endangering the superior gluteal nerve.^9,19,25,32 The surgical approach, which requires a gluteus medius split or subperiosteal elevation, can directly damage the abductor mechanism through surgical trauma or by placing the superior gluteal nerve at risk. Multiple authors have recommended to condense the iliac flange and to perform a greater trochanteric osteotomy to increase visualization and to relieve strain on the superior gluteal nerve.^19,24,25 Surprisingly, no patients experienced deep infections or clinically significant heterotopic ossification, which are common complications when undergoing revision³³ with massive exposure and hardware.

Secondary outcomes included ambulation status and subjective pain. At their final follow-up appointments, all patients in this study were able to ambulate; two were ambulating without assistance (Fig. 1A–B). One patient walked with the occasional assistance of a cane when outside of the house (Fig. 1C–D). Ambulation status was not frequently reported in the literature. Nevertheless, Christie et al. as well as Holt & Dennis, reported that a high frequency of patients were able to ambulate without assistance after their surgery.^25,27 In Christie et al.‘s cohort, 46.2% ambulated without assistance, 38.5% ambulated with assistance of a cane or crutches, 13.8% ambulated with assistance of a walker, and 1.5% remained non-ambulatory.²⁷ Similarly, Holt & Dennis reported that 69% ambulated without assistance, 19% ambulated with assistance of a cane or crutches, 8% ambulated with assistance of a walker, and no patient remained non-ambulatory.²⁵

Close

Fig. 1 A–D: (A) Anteroposterior X-ray of patient seven pre-triflange revision, (B) Anteroposterior X-ray of patient seven post-triflange revision at most recent follow-up of 4.54 years, (C) Anteroposterior X-ray of patient three pre-triflange revision, (D) Anteroposterior X-ray of patient three post-triflange revision at most recent follow-up of 7.50 years.

Export to PPT

While analog pain scales were not administered during our study, more than half of this cohort (4/7, 57.1%) did not subjectively report pain at their latest follow-up. This is a better outcome than described in a previous cohort²¹ and represents a key factor in patient satisfaction.³⁴ Other secondary findings included radiographic complications: one patient experienced both asymptomatic screw breakage and aseptic loosening at the most recent follow-up visit; however, the triflange component had not displaced at this time. When comparing treatment options for pelvic discontinuity in revision THAs, one meta-analysis demonstrated that the custom triflange was similar to cup-cage constructs and had better mechanical failure outcomes than antiprotrusio cages and porous metal.⁸

While long-term outcomes of triflange implants are promising, the procedure does not come without risks. The cohort in this study all experienced some form of intraoperative or immediate postoperative complications. Operations in this cohort were complicated by blood loss requiring transfusion in all cases and required postoperative intensive care unit admissions in three patients. It is well-established that revision surgeries and increased blood loss are both risk factors for intensive care unit admissions³⁵; patients on average lose over 1 L of blood intra-operatively.³⁶ While short term surgical outcomes are favorable according to reports in the literature, other studies corroborate the immediate postoperative risks that come from this procedure. Friedrich et al. reports that 50% of their cohort required ICU admissions for postoperative monitoring, with a mean stay of 1.8 days.¹⁴ Concurrently, reports from Berasi et al., Joshi et al., and DeBoer et al. mention average of 696 mL (154–1400 mL),⁶ 2,200 ml (1500–4000 mL),²⁶ and 2060 mL (600–6000 mL),¹¹ respectively. While they make no mention on whether any of their patients required ICU admissions, it is reasonable to infer from the blood loss ranges, that some required further monitoring in the ICU, if not most.

Previous authors have described the triflange revision as technically challenging,²⁶ and one group experienced technical problems in 41% of their cases.²¹ One patient experienced a particularly difficult course. This may be partly explained by her age, higher BMI of 45.3 kg/m², and history of fracture and infection. Aside from her severe pelvic osteolysis, she was at an increased risk for recurrent dislocations due to her history of avascular necrosis²⁹ and multiple surgical revisions prior to the triflange implant.³⁷ Kosashvili et al. reported that revisions posed a significantly higher risk of dislocation and infection in patients who underwent four or more revision surgeries,³⁷ which is consistent with this patient's history.

Another potential disadvantage of the triflange implant is its manufacturing process. The process of converting imaging to creation and implantation of the titanium triflange component takes approximately six weeks, and reported costs range from $5200 to $12,500.^5,18,27 However with continued innovations in 3D printing, these limitations may become diminished in the future.³⁸

The primary limitation of this study was the small sample size, which lends to low statistical power and reduces the generalizability to broader populations. Nevertheless, adding detailed long-term outcome data on an additional seven hips still significantly enhances the currently sparse literature. Additionally, the rarity of these cases leads to varying interpretations of radiographic images and subsequent treatment options. The Paprosky classification has variable intraobserver (κ = 0.14 to 0.75) and interobserver (κ = 0.02 to 0.79) reliability.⁴ Still, it is likely the best existing means by which to classify and treat these lesions.

While a previous meta-analysis compared different salvage procedures for the treatment of pelvic discontinuity cases,⁸ future studies could potentially prospectively compare the triflange to other treatment modalities, such as the cup-cage technique or the acetabular distraction technique.^39,40 However, this may prove difficult as cases requiring these implants occur with relative rarity, and therefore acquisition of an adequate sample size to reveal significance will remain difficult in the near future. As the population ages and the number of total hip arthroplasties performed increases, the number of cases will also rise, potentially leading to appropriately powered studies that determine the best treatment options for catastrophic pelvic osteolysis.

5 Conclusion

Despite the complex preoperative planning, cost, and potentially difficult perioperative period, the custom triflange acetabular implant is a viable option for patients with catastrophic pelvic osteolysis. Revision was necessary in only one patient in this cohort, and no patients required removal of the triflange acetabular implant. Our experience matched those reported in a recent systematic review.¹⁰ Most patients experienced positive outcomes regarding stability, pain, and ambulation, but need to be counseled on the high complication rate of this salvage procedure.