Well-fixed cementless acetabular components may be revised in cases of infection, malposition, or extensive osteolysis.1 Bone preservation is critically important for subsequent acetabular reconstruction. Traditionally, special curved osteotomes, also called Aufranc acetabular gouges (Figure 1), have been used to disrupt the bone-implant interface when removing well-fixed acetabular components with some potential for bone loss.2,3 Other techniques have been described, including the use of a pneumatic impact wrench, an angled punch, and a reciprocating saw; however, all have been associated with an increased risk for bone loss or other complications, such as pelvic fracture.3–6 More recently, the Explant Acetabular Cup Removal System (Zimmer, Warsaw, Indiana) (Figure 2) was introduced to minimize bone loss by using relatively thin, curved blades specific for the outer diameter of the acetabular component to disrupt the bone-implant interface. Mitchell et al7 reported that, in their experience with this system, no more host bone was lost than the thickness of the blade, approximately 2 mm. Modified techniques have also been described using this system for the removal of large monoblock acetabular components used in hip resurfacing.8–12
Curved osteotomes, or Aufranc acetabular gouges.
The Explant System (Zimmer, Warsaw, Indiana) consists of curved blades corresponding with the component's outer diameter and attached to a rotating handle device with a centering head matching the inner diameter of the polyethylene liner, which stabilizes the device (A). A short blade (top) is used to initiate the development of the bone-implant interval, followed by a longer, full-radius blade (bottom) that completes the dissociation of the component from the bone (B).
Although the Explant System has been favored, both anecdotally and in the literature,7–12 to the current authors' knowledge there have been no comparative data to support claims that it is a superior method for removing well-fixed cementless acetabular components. The purpose of this study was to compare bone loss following the use of 2 different techniques— Aufranc gouges and the Zimmer Explant Acetabular Cup Removal System—for the removal of well-fixed cementless hemispherical acetabular components.
Materials and Methods
With the approval of the institutional review board, 623 patients who underwent revision total hip arthroplasty (THA) for any reason at the authors' institution between 2002 and 2013 were identified through the joint replacement registry, which houses prospectively gathered data. All procedures were performed by 1 of 3 fellowship-trained arthroplasty surgeons (S.B.G., W.J.M., J.I.H.). Operative reports were reviewed to identify those who underwent revision of well-fixed cementless acetabular components for any reason except infection; cases in which the component was identified by the surgeon as being grossly loose or having only fibrous ingrowth were excluded. In addition, because the Explant System was designed for the removal of cementless hemispherical acetabular components, only failed hemispherical or subhemispherical components were included. Furthermore, revisions for failed cementless elliptical acetabular components, failed cemented acetabular components, failed cup-cage constructs, and failed hip resurfacing were also excluded. Of the cases reviewed, 94 hips in 89 patients met these criteria.
After further review, cases were excluded if the outer diameter of the acetabular component removed was not available in the medical record or institutional registry (n=24) or if multiple methods were used to remove the component (ie, both Aufranc gouges and the Explant System) (n=13). In addition, patients in whom a cup-cage construct was used for reconstruction (n=2) or in whom the final acetabular component was smaller than the initial component (n=4) were excluded. Thus, 51 hips in 47 patients remained for the final analysis.
In all patients, once the acetabular component was found to be well fixed and indicated for removal, the method of acetabular component removal was selected based on surgeon preference. Two surgeons used the Explant System exclusively. In selected cases, when the rim of the acetabular component was not easily accessible, a rongeur was used to remove overlying bone prior to using the Explant System. The third surgeon used either Aufranc gouges or the Explant System on a case-by-case basis: the Explant system was used in cases where the peripheral rim of the acetabular component was easily accessible; otherwise, the Aufranc gouges were used. The polyethylene liner and screws, if present, were removed. In cases where Aufranc gouges were used, a series of them were used sequentially to develop the interval between the acetabular component and surrounding bone until the component was completely loosened.
The Explant System was used according to the manufacturer's technique. The system was originally designed for the removal of modular cementless hemispherical implants and requires an intact polyethylene liner or trial liner and knowledge of the outer diameter of the component being removed. Once screws are removed, the liner is replaced. Curved blades that correspond to the component's outer diameter are attached to a rotating handle with a centering head matching the inner diameter of the polyethylene liner, which stabilizes the device. First, a short blade initiates the development of the interval as the surgeon moves the handle circumferentially around the acetabular component. Next, a longer, full-radius blade completes the dissociation of the component from the bone.
Regardless of the method used, once the component was removed, the remaining bone stock was assessed and the appropriate component type was selected for reconstruction. The acetabulum was underreamed by 0 to 2 mm, using hemispherical reamers, based on bone quality and defect morphology. If necessary, as determined by the surgeon, impaction bone grafting was performed with acetabular reamings and/or allograft cancellous bone chips. In all cases, the component was impacted into position and stabilized with multiple 6.5-mm screws.
Demographic information was obtained from the medical record. Prerevision anteroposterior radiographs of the pelvis were reviewed to classify acetabular bone defects using the American Academy of Orthopaedic Surgeons (AAOS) classification for acetabular bone defects. Operative reports were reviewed for surgical details, including the reason for revision, type of revision performed, method of acetabular component removal (Aufranc gouges or Zimmer Explant Acetabular Cup Removal System), the outer diameters of the acetabular component removed, the final reamer used, the acetabular component implanted, type and geometry of the acetabular component implanted, and the use of augments or bone grafting. Given that there is no validated, direct method for the quantification of bone loss, the authors elected to use multiple surrogates: (1) the difference between the initial acetabular component outer diameter and the diameter of the final reamer; (2) the difference between the initial acetabular component outer diameter and the new acetabular component diameter; and (3) the use of impaction bone grafting. Although previous studies have used the difference between the initial acetabular component outer diameter and the diameter of the final reamer,7–10,12 it is unclear whether it is better than using the new acetabular component outer diameter as an endpoint, given that both values may be affected by the desired amount of press-fit of the new component. Therefore, the authors elected to report both values. They also reported the use of bone grafting as another surrogate for bone loss, understanding that it may be influenced by other factors (eg, surgeon preference, preexisting bony defects) in addition to the surgical technique used for acetabular component removal. Although none of these measures individually is an ideal surrogate for bone loss, combined they may provide a more complete representation.
Patients were divided into 2 groups based on the method of acetabular component removal. Demographics were compared between the 2 groups. Differences in age and duration of acetabular component implantation between the 2 groups were assessed with a 2-tailed Wilcoxon-Mann-Whitney test. Differences in sex, preoperative diagnosis, preoperative AAOS defect, component morphology (elliptical vs hemispherical), and use of bone grafting were assessed with the chi-square test. A 2-tailed Wilcoxon-Mann-Whitney test was performed to assess the difference in bone loss between the 2 groups, via the differences between the outer diameters of the removed acetabular component and both the final reamer and new acetabular component. A post-hoc power analysis was performed to confirm that the sample size was sufficient. With an SD of 2.3, a minimum of 21 patients in each group would be required to detect a 2-mm difference with an alpha of 0.05 and power of 0.80.
The study group comprised 17 men (18 hips) and 30 women (33 hips) who underwent revision at a mean age of 65 years (range, 28–88 years). The most common reasons for revision were metal-on-metal bearings with or without associated adverse local tissue reaction (18 hips; 35%) and instability (14 hips; 27%). There were 45 (88%) isolated acetabular revisions and 6 (12%) combined acetabular and femoral revisions. The median duration of implantation of these acetabular components was 5 years (range, 0.5–21 years) prior to revision. Most hips had no preoperative acetabular bony defects (34 hips; 67%). Seventeen (33%) hips had cavitary defects (AAOS type II).
Twenty-seven (53%) acetabular components were removed with the Explant System, whereas 24 (47%) were removed with Aufranc gouges (Table 1). All surgeries in the Aufranc group were performed by a single surgeon, whereas surgeries in the Explant group were performed by 3 different surgeons (9 of these surgeries were performed by the same surgeon as that in the Aufranc group). There was no difference between the groups in age (P=.10), sex (P=.14), preoperative diagnosis (P=.32), preoperative AAOS defect (P=.55), duration of acetabular component implantation (P=.20), or revision type (P=.11). The median difference between the initial acetabular component outer diameter and the diameter of the final reamer was 5 mm (range, 0–10 mm) in the Aufranc group and 3 mm (range, 0–8 mm) in the Explant group (P=.004) (Table 2). The median change in acetabular component outer diameter in the Aufranc group was 6 mm (range, 0–10 mm) and 4 mm (range, 0–8 mm) in the Explant group (P=.002) (Table 2). All hips were reconstructed with standard hemispherical or elliptical acetabular components. Elliptical acetabular components were used in 63% (15 of 24 cases) of hips in the Aufranc group and 56% (15 of 27 cases) of hips in the Explant group (P=.61). No metal augments were used in any hip. Hips in the Aufranc group were more likely than those in the Explant group to have bone grafting at the time of acetabular reconstruction (54% vs 26%; P=.04). Among cases performed by the single surgeon who used both devices, there was no detectable difference in bone grafting (54% vs 43%; P=.59).
Comparison of Aufranc and Explant Group Data
Comparison of Bone Loss
In this study, the difference between the outer diameter of the initial acetabular component and the diameter of the final reamer, as well as the difference between the outer diameters of the initial and new acetabular components, was smaller in revisions using the Zimmer Explant Acetabular Cup Removal System, suggesting less bone loss than when the more traditional Aufranc gouges are used. In addition, the use of Aufranc gouges was associated with a significantly greater use of bone grafting at the time of reconstruction.
The authors' results with the Zimmer Explant Acetabular Cup Removal System compare favorably with other reported experiences. Mitchell et al7 reported their initial experience using the Explant System in 31 acetabular revisions. They reported bone loss as the difference between the outer diameter of the acetabular component removed and the diameter of the final reamer; the median difference was 4 mm. However, they also presented the size of the new acetabular component for each patient; therefore, the difference between the initial and new acetabular component diameters could also be calculated. The median difference between the initial and final acetabular components was 6 mm in their study.7 A report on the use of the Explant System for the removal of monoblock acetabular components used in hip resurfacing showed that the final reamer was 1 to 2 mm larger than the acetabular component removed.8 The calculated median difference between the outer diameters of the acetabular component removed and the new one placed was 2 mm. Other case reports on the removal of nonmodular acetabular components report this difference to be between 2 and 4 mm.9,10,12
The current study has some important limitations. First, it is a single-institution study in which the procedures were performed by experienced, fellowship-trained arthroplasty surgeons at a referral center. Thus, their experience may not be generalizable to the average practice. Second, although the use of multiple surgeons who used different implants may have influenced the results, doing so increased the feasibility of studying a relatively rare occurrence and sheds some light on the results that could be expected from a larger sample of surgeons. The results of this study may also be biased by the fact that the method of acetabular component removal was based on surgeon preference. Because of this, there may be other surgeon-based differences, including surgical technique, indications for bone grafting, and final choice for implant size, that could affect the results of this study. For instance, bone grafting was found to occur more commonly in the Aufranc group, which only had 1 surgeon. However, when evaluating only cases in both groups performed by the same surgeon, there was no detectable difference in bone grafting. It is unclear whether there was no true difference in the use of bone grafting or whether the sample size is too small to demonstrate a difference. It is therefore possible that the difference on bone grafting seen between the 2 groups overall is a result of this surgeon's preference for bone grafting. Third, the results of this study may also be biased by the fact that the method of acetabular component removal was based on surgeon preference. Fourth, both hemispherical and elliptical acetabular components were used. Although the use of elliptical, rather than hemispherical, acetabular components in some cases may imply greater defects that could not be adequately filled with a hemisphere, hemispherical reamers were used in all cases, so the measures used to quantify bone loss should be unaffected. The rims of the elliptical components used were 2 mm larger than the rims of the true hemispherical components of the same outer diameter. Furthermore, because the groups did not differ statistically in the use of elliptical shells (63% vs 56%; P=.61), the study was not powered specifically for this variable. Finally, there is no direct method for measuring bone loss, either in this study or others previously published. The current authors elected to closely replicate the methodology of previous studies—using the difference between the initial acetabular component outer diameter and the diameter of the final reamer—to allow for the comparison of their results to those in the literature. They also added the difference between the outer diameters of the initial and revised acetabular components as a second measure, which did not differ from the first measure. Even with these methods to estimate bone loss, there is no way to quantify the amount of bony overgrowth that may have been removed to facilitate the use of both the Explant System and the Aufranc gouges.
The results of this study suggest that less bone loss is associated with the use of the Zimmer Explant Acetabular Cup Removal System to remove well-fixed cementless acetabular components when compared with the use of Aufranc gouges. Although the potential bone preservation afforded by this technique may become important in future acetabular reconstructions, the exact clinical significance of these findings is unclear.
- Blaha JD. Well-fixed acetabular component retention or replacement: the whys and the wherefores. J Arthroplasty. 2002; 17(4)(suppl 1):157–161. doi:10.1054/arth.2002.33301 [CrossRef]
- Masri BA, Mitchell PA, Duncan CP. Removal of solidly fixed implants during revision hip and knee arthroplasty. J Am Acad Orthop Surg. 2005; 13(1):18–27. doi:10.5435/00124635-200501000-00004 [CrossRef]
- Mallory TH, Lombardi AV Jr, Fada RA, Adams JB, Kefauver CA, Eberle RW. Noncemented acetabular component removal in the presence of osteolysis: the affirmative. Clin Orthop Relat Res. 2000; 381:120–128. doi:10.1097/00003086-200012000-00015 [CrossRef]
- Lachiewicz PF, Anspach WE III, . Removal of a well-fixed acetabular component: a brief technical note of a new method. J Bone Joint Surg Am. 1991; 73(9):1355–1356.
- Daum WJ, Calhoun JH. Removal of the acetabular component minimizing destruction of the bone bed. J Arthroplasty. 1988; 3(4):379–380. doi:10.1016/S0883-5403(88)80041-X [CrossRef]
- Pierson JL, Jasty M, Harris WH. Techniques of extraction of well-fixed cemented and cementless implants in revision total hip arthroplasty. Orthop Rev. 1993; 22(8):904–916.
- Mitchell PA, Masri BA, Garbuz DS, Greidanus NV, Wilson D, Duncan CP. Removal of well-fixed, cementless, acetabular components in revision hip arthroplasty. J Bone Joint Surg Br. 2003; 85(7):949–952. doi:10.1302/0301-620X.85B7.13593 [CrossRef]
- Preiss RA, Patil S, Meek RM. The use of modular femoral head trials to centre the Explant blade facilitates retrieval of well-fixed acetabular components with minimal bone loss. Arch Orthop Trauma Surg. 2011; 131(7):1003–1006. doi:10.1007/s00402-010-1229-6 [CrossRef]
- Rawal JS, Soler JA, Rhee JS, Dobson MH, Konan S, Haddad FS. Modification of the explant system for the removal of well fixed hip resurfacing sockets. J Arthroplasty. 2010; 25(7):1170. doi:10.1016/j.arth.2009.11.003 [CrossRef]
- Olyslaegers C, Wainwright T, Middleton RG. A novel technique for the removal of well-fixed cementless, large-diameter metal-on-metal acetabular components. J Arthroplasty. 2008; 23(7):1071–1073. doi:10.1016/j.arth.2008.04.006 [CrossRef]
- Taylor PR, Stoffell KK, Dunlop DG, Yates PJ. Removal of the well-fixed hip resurfacing acetabular component: a simple, bone preserving technique. J Arthroplasty. 2009; 24(3):484–486. doi:10.1016/j.arth.2008.01.305 [CrossRef]
- Blumenfeld TJ. Removing a well-fixed non-modular large bearing cementless acetabular component: a simple modification of an existing removal device. J Arthroplasty. 2010; 25(3):498. doi:10.1016/j.arth.2009.01.017 [CrossRef]
Comparison of Aufranc and Explant Group Data
|Variable||Aufranc Group||Explant Groupa||P|
|Age, mean (range), y||62.2 (28–88)||67.9 (49–82)||.09|
|Preoperative diagnosis, No.|
|Preoperative acetabular defect, No.|
| AAOS type II||7||10|
|Duration of implantation, median (range), y||8.6 (0.5–21)||6.5 (0.5–17)||.17|
|Type of revision, No.|
| Acetabular and femoral||1||5|
|Shape of new acetabular component, No.|
Comparison of Bone Loss
|Variable||Aufranc Group||Explant Groupa||P|
|Difference between initial acetabular component size and final reamer size, median (range), mm||5 (0–10)||3 (0–8)||.004|
|Difference between initial and new acetabular component size, median (range), mm||6 (0–10)||4 (0–8)||.002|
|Bone grafting, No.||13 (54%)||7 (26%)||.04|