Short stem total hip arthroplasty with the direct anterior approach demonstrates suboptimal fixation

Short femoral stems might be an attractive option in young patients, as they preserve bone stock and supposedly allow more physiological loading on the proximal femur [1]. Short stem designs have been also developed to facilitate their insertion through less-invasive surgical exposures [2], including the direct anterior approach (DAA).

The AMIStem-H® (Medacta, Switzerland) is a cementless short femoral stem with a rectangular triple tapered design and hydroxyapatite (HA) coating. Based on straight rectangular tapered stems, it has been specifically designed for facilitating stem introduction through the DAA by reducing overall dimensions by 33% as well as the lateral flare of the stem. The AMIStem-H® is a type IIIB short stem according to the classification of Feyen et al. [3]. It can also be classified as a type 4 stem according to Khanuja et al. [4], or as a trochanter harming type according to Falez et al. [1]. In our institution, both the AMIStem-H® and the DAA were introduced simultaneously in 2008.

The purpose of this study was to assess the five year outcomes of the AMIStem-H® implanted through the DAA. We determined (1) the occurrence of short-term complications including revisions; (2) the clinical and patient-reported outcomes (PROs) based on the Harris Hip Score (HHS), the Short-Form health survey (SF-12®), the Western Ontario McMaster Universities (WOMAC) score, and the visual analogue scale (VAS) for evaluation of patients’ satisfaction; and (3) the radiological results with a focus on femoral stem fixation, bone ingrowth, and stress shielding; and (4) we evaluated the possible relation between patient demographics, femoral morphology, surgeon experience, and the primary fixation of the femoral stem.

Overall, 698 THAs met the inclusion criteria and were included for the final analysis. For comparison, during the same period, a total of 1878 THAs (mean age 67.5 years, SD 13.1, 56% in women) were performed in our department for the same indications. For the current series, the mean age was 65.7 (SD 12.6, range, 18–96), 53% were women, and pre-operative BMI was 27.14 (SD 5, range, 17.3–48.9) (Table 1). The mean follow-up was 6.2 years (SD 1.8, range, 2–9.7 years). Over the entire follow-up period, 59 (8.5%) patients died and 45 (6.4%) were lost to follow-up (Fig. 1).

In the present study, we found a 4.2% revision rate, which was mainly due to aseptic loosening of the femoral component, and a high rate of patients presenting radiological features of poor stem stability. Our findings are comparable with the 2018 Australian implant registry data showing a 3.5% revision rate at five years for the AMIStem-H® combined with the same cup [9], but are higher than those reported for other short stems implanted through the DAA. In a retrospective review of 247 consecutive THAs with the Tribute® short stem, Attenello et al. [10] reported a femoral stem subsidence of more than 5 mm in four hips, and none required revision surgery at 27-month follow-up. In a retrospective review of 899 consecutive patients followed up to 24 months, Cidambi et al. [11] found a 1.3% revision rate for femoral aseptic loosening of a short, mediolaterally tapered stem. Interestingly, they found no cases of aseptic loosening in patients who received a standard-length collared, fully HA-coated stem. However, in a recent meta-analysis including 4280 patients with long stems and 2545 with short stems from 34 studies, Panichkul et al. [12] failed to find any difference in terms of revision for aseptic loosening between long and short femoral stems implanted through the DAA.

We found a periprosthetic femoral fracture rate of 1.7%, which is comparable to fracture rates reported with short stems implanted through the DAA [13]. In a case series of 640 patients, Dietrich et al. [14] reported a 1.6% femoral fracture rate when a short stem was used (130 Fitmore® stems and 53 hips AMIStems-H®) vs. 6.8% when a standard-length stem was implanted (457 Quadra® stems). The impact of prosthetic design and stem length on intra-operative femoral complications remains, however, unclear in the literature. Using a cementless standard-length femoral component, Jewett et al. [15] reported a high rate of intra-operative complications including 2.3% of trochanteric fractures, and 0.49% of femoral perforations or fractures. In contrast, in a retrospective series of 686 patients (851 hips), the use of short cementless tapered-wedge stems was associated with an increased 2.7% rate of fracture and was the only predictor for peri-operative periprosthetic complications [16].

Clinical results were globally very good with evident improvement of the HHS and WOMAC scores, and a high satisfaction level. However, comparing the satisfaction rate clinical scores and PROs between two and five years, we found a substantial increase in the proportion of unsatisfied patients, a decrease in the proportion of pain-free patients as well as lower WOMAC and SF-12 physical component scores five years after surgery.

Because radiological signs are so far the best predictor of implant survival, much of our attention was given to radiological results. We found femoral stem subsidence > 2 mm in 42 hips (12.9%), proximal radiolucencies in 101 hips (31.5%), cortical thickening in 52 (16.0%), and a pedestal in 219 (67.6%). To further evaluate femoral stem fixation, we used the Engh score and found a clear correlation between Engh’s stability categories and aseptic loosening. The calculated risk of stem revision for aseptic loosening was significantly higher for the hips with an Engh score between − 10 and 0 (category “suboptimal but stable”) when compared to hips with an Engh score > 0. We also found a correlation between reported pain and Engh score < 0, which suggests that these patients have not only an increased risk for aseptic loosening but also inferior clinical results. In a retrospective study including 100 primary THAs, Maier et al. [17] reported the results of the Fitmore® stem at a mean of 3.3-year follow-up. Radiographic analysis was performed for 79 THAs, with 20 hips (25%) demonstrating radiolucencies < 2 mm, 50 hips (63%) demonstrating cortical hypertrophy, and 13 hips (17%) demonstrating both proximal radiolucencies and distal cortical hypertrophy. The authors concluded that in these cases, the stem was probably stabilized by a distal fixation. Ishi et al. [18] showed that poor radiographic outcomes of uncemented stems were associated with a high canal flare index (CFI) and insufficient proximal filling, especially in Japanese women with narrow femoral canals. Similarly, D’Ambrosio et al. [19] evaluated the influence of femoral morphology or femoral component filling on radiological outcomes following THA using a fully HA-coated femoral stem. In a series of 183 THAs, they found that femurs with either proximally flared or distally narrowed canals, or insufficient proximal filling, tend to have less favourable radiological outcomes. Successful osseointegration was obtained with a filling threshold greater than 70% at 2 cm below the lesser trochanter. The authors also found a correlation between a low CFR measured 2 cm above the lesser trochanter and high CFI, which is more frequently associated to Dorr type A morphology [20]. In the present study, we did not evaluate the CFR, but we examined femoral morphology according to the Dorr classification. Interestingly, we did not find any correlation between the Dorr type and the occurrence of radiological changes or the Engh score category.

Finally, we found a clear correlation between patient activity and the Engh score: high activity, younger age, and lower ASA score were all associated with an Engh score < 10 (suboptimal fixation). In a series of 725 hip with signs of femoral stem loosening, Munger et al. [21] concluded that increased activity in younger patients with unrestricted mobility was an important factor of aseptic loosening of the femoral component. Whether young, healthy and more active patients, through the advantage of a less traumatic DAA, undergo a more rapid rehabilitation process, which can result in micromotion at the bone-implant interface, cannot be excluded nor confirmed.

The current study has several limitations. First, the concomitant introduction of a new stem and the DAA is clearly a limiting factor for this study, whereas it is not possible to determine if the stem design or the surgical approach or both are the cause of our findings. To our knowledge, there are no reports describing the use of the AMIS-H stem through other surgical approaches, which could have been useful for comparison purposes. Several authors have raised concern about the DAA being a potential risk factor for early aseptic loosening of the stem, especially in the early experience [22]. However, we did not find in our study any relation between surgeon experience and aseptic loosening or radiological signs. Second, we present a nonrandomized, observational study, in which there might be residual confounding factors and potential selection bias. The DAA has gained popularity among surgeons because it spares the abductors and allows faster recovery with shorter length of stay. Thus, it could have been proposed to younger and active patients who wanted to recover quicker. However, the mean age was 65.7 and 47% of the patients were male in our series, which is comparable to other studies [23], and to the age and gender of all THAs performed during the same period in our department. Third, despite an active strategy to increase response rates (phone calls, letters, emails when available), 96 patients (18.2%) refused the five year follow-up and were thus unavailable at the time of the analysis. We acknowledge that missing data might impact the robustness of our results. Finally, the appropriateness of the Engh score could also be debated, as it was rarely used to evaluate shorter stems [16]. In our study, there was a moderate-to-good agreement between the reviewers regarding the Engh score, with an inter-rater reliability that was actually higher than previously published with this scale [24].

In conclusion, short bone-conserving femoral stems in THA have been designed to preserve proximal bone stock for potential revision surgery, but also to facilitate less-invasive surgical exposures and revision procedures. However, little data are available on their design rationale, fixation features, and clinical outcomes. We found a high incidence of hips presenting radiological signs of suboptimal fixation in young and active patients, with an increased risk of revision for early aseptic loosening. The short stem evaluated in this study has been subsequently abandoned in all patients and replaced by a conventional length stem.