Radiological stability of a quadrangular uncemented femoral component compared with its shortened version in primary total hip arthroplasty

Abstract

Aims

Cementless femoral fixation is widely used in total hip arthroplasty (THA), with increasing interest in shorter femoral component designs. These femoral components aim to preserve proximal bone, promote physiological load transfer, and facilitate minimally invasive approaches. Among these, ‘shortened’ femoral components, derived from conventional stems and retaining a diaphyseal extension, may offer advantages in terms of implant positioning. However, clinical evidence regarding their mechanical reliability remains limited. This study aimed to compare the two-year radiological stability of a shortened uncemented quadrangular femoral component with that of its conventional-length counterpart in patients undergoing THA for osteoarthritis.

Methods

We conducted a prospective, single-blind, randomized, noninferiority trial between January 2018 and December 2021. Patients undergoing primary THA were randomly assigned to receive either the conventional-length quadrangular femoral component or its shortened version. The primary outcome was femoral component migration at two years, assessed with Ein Bild Roentgen analyze - femoral component analysis (EBRA-FCA). Secondary outcomes included intraoperative and postoperative complication rates, as well as the 12-item Oxford Hip Score (OHS) at 12 months.

Results

A total of 143 hips were randomized, with 139 (70 conventional-length, 69 shortened) included in the clinical analysis. Per-protocol radiological analysis was carried out on 98 hips (49 hips per group). At two years, mean femoral component migration was 0.94 mm (SD 0.49) in the shortened-femoral component group and 0.86 mm (SD 0.52) in the conventional-length group. The upper bound of the 90% CI (–0.09 to 0.25) was below the prespecified noninferiority margin of 1 mm. No significant differences were observed in complication rates or functional outcomes (p > 0.05).

Conclusion

At two years, radiological outcomes of the shortened quadrangular femoral component were consistent with noninferiority compared with the conventional-length version. Although the planned sample size was not fully reached, the findings remained within the prespecified noninferiority margin.

Cite this article: Bone Jt Open 2026;7(4):465–472.

Introduction

Cementless femoral components, particularly straight designs with a quadrangular cross-section, are widely used in total hip arthroplasty (THA), with consistently good clinical outcomes and proven long-term survivorship.¹ However, they may be associated with postoperative thigh pain and proximal stress shielding.^2,3

To address these limitations, shorter femoral components have been developed over recent decades. Their initial goal was to preserve femoral bone stock and facilitate potential future revisions. Their reduced length may also simplify implantation, particularly when using minimally invasive approaches such as the direct anterior approach (DAA).⁴ By promoting a more physiological load transfer to the proximal femur, these implants are expected to reduce stress shielding and the incidence of thigh pain.^5,6

Consequently, numerous short femoral components with distinct designs and biomechanical concepts are now available. Khanuja et al⁷ proposed a classification to better characterize these implants. Among them, type 4 ‘shortened femoral components’, derived from conventional designs, offer a promising compromise: their design closely resembles that of clinically proven standard femoral components while providing the advantages of a shorter femoral implant. Their retained diaphyseal extension may facilitate implantation, and help reduce the risk of frontal malalignment reported with some short femoral component designs.⁸

Despite their increasing use, evidence supporting the mechanical safety and clinical equivalence of shortened quadrangular femoral components remains limited. It has yet to be demonstrated that shortening the femoral component does not compromise mechanical reliability, clinical outcomes, or complication rates. To date, no randomized controlled trial (RCT) has directly compared a quadrangular cementless femoral component with its shortened counterpart.

The aim of this RCT was to compare the two-year radiological stability of a shortened quadrangular femoral component with that of its standard-length version in patients undergoing primary THA for osteoarthritis (OA) by assessing femoral component migration. We hypothesized that the shortened femoral component would be noninferior to the standard femoral component regarding two-year axial migration.

Methods

Study design and population

This was a two-arm single-centre, prospective, randomized, single-blinded noninferiority trial conducted between January 2018 and December 2021. The study included adult patients aged ≥ 18 years undergoing THA for primary or secondary OA, with symptoms unresponsive to medical treatment and the ability to provide written informed consent. Exclusion criteria were prior femoral surgery preventing implantation of either femoral component, contraindications to anesthesia, active infection or tumour, or any social, psychological, or geographical factors likely to compromise follow-up. The study adhered to CONSORT guidelines for rigorous reporting. Patients were randomized on the day of surgery to receive either a standard-length Hype femoral component (Std group) or its shortened version (Mini group) (Société d’Études Recherche Fabrication (SERF), France). Randomization was stratified by surgeon and conducted using variable block sizes. Randomization lists were managed by an independent randomization centre, and group assignment was revealed using sealed, opaque envelopes opened in the operating room on the morning of surgery, under the supervision of the SERF clinical research department. Patients were blinded until the three-month follow-up. The study was approved by the institutional ethics board.

Implant migration was assessed using Ein Bild Roentgen analyze - femoral component analysis (EBRA-FCA), a method detailed below. Migration values were assessed longitudinally using standardized radiographs obtained at six weeks, three months, 12 months, and two years. A sample size of 69 hips per group was calculated based on a prespecified noninferiority margin of 1 mm in mean axial migration, with 90% power and a one-sided α of 5%, assuming a pooled standard deviation of 2 mm.^9,10 To account for dropouts and nonanalyzable radiographs, target enrolment was 160 hips.

A total of 153 hips from 127 patients were randomized. Ten hips were excluded preoperatively due to surgery cancellations or a decision to use a cemented femoral component, leaving 143 hips operated. One patient received a cemented femoral component, one patient died from a non-surgery-related cause, one underwent bipolar revision without Hype femoral component implantation for surgical-site infection, and one patient was lost to follow-up before the first postoperative follow-up visit; these cases could therefore not be included in the clinical or radiological analyses. In total, 139 hips (70 Std, 69 Mini) therefore remained available for clinical and radiological follow-up. One patient in each group ultimately received the opposite femoral component at the time of surgery. Both crossovers occurred for logistical reasons related to implant and instrumentation availability at the time of surgery; analyses were therefore performed according to the femoral component implanted. Baseline patient characteristics were balanced between the two groups (Table I). At the two-year timepoint, 12 hips (seven Mini, five Std) were lost to follow-up. Among the remaining hips, 19 (eight Mini, 11 Std) had radiographs obtained outside the predefined two-year time window and could therefore not be included in the two years per-protocol analysis, which requires careful adherence to the protocol schedule to avoid time-related bias. In addition, ten hips (five Mini, five Std) had radiographs rejected by the EBRA-FCA because of insufficient comparability. The per-protocol analysis therefore included 98 hips (49 Std, 49 Mini) (Figure 1).

Table I.

Population characteristics.

Variable	Std group (n = 70)	Mini group (n = 69)	SMD
Mean age, yrs (SD)	61.73 (10.35)	62.39 (12.52)	0.06
Mean weight, kg (SD)	73.05 (1.41)	71.85 (17.16)	0.08
Sex, n (%)			0.13
Female	49 (70)	44 (64)
Male	21 (30)	25 (36)
Cause of osteoarthritis, n (%)
Primary	61 (87)	56 (81)	0.16
Avascular osteonecrosis	4 (5.7)	3 (4.3)	0.06
Post-traumatic	0 (0)	1 (1.4)	0.16
Inflammatory/rheumatic	0 (0)	1 (1.4)	0.17
Dysplasia	5 (7.1)	8(12)	0.17
Surgical approach, n (%)
DAA	58 (83)	52 (75)	0.20
Röttinger	12 (17)	15 (22)	0.13
Dall	0 (0)	2 (2.9)	0.24
Dorr classification, n (%)
A	22 (31)	19 (28)	0.07
B	47 (67)	47 (68)	0.02
C	1 (1.4)	3 (4.3)	0.17
Mean preop Oxford Hip Score (SD)	39.1 (9.8)	40.4 (8.8)	0.15
Femoral component, n (%)			0.33
Standard offset (SCC/SCC mini)	69 (99)	63 (91)
Lateralized (SCL/SCLA mini)	1 (1)	6 (9)

DAA, direct anterior approach; SCC, standard component with a collar; SCL, standard component lateralized; SCLA, standard component lateralized mini; SMD, standardized mean difference; Std, standard.

Fig. 1.

Flowchart. EBRA-FCA, Ein Bild Roentgen analyze - femoral component analysis.

Surgical technique and implants

All procedures were performed by two senior surgeons (PA, DB) using three surgical approaches: DAA, Röttinger anterolateral, or Dall modified anterolateral.¹¹ Acetabular components were implanted based on preoperative planning and included cemented and uncemented options depending on bone quality and surgeon preference. Femoral broaching matched the component’s geometry, and the allocated implant was inserted accordingly. Perioperative management followed standard protocols, including antibiotic and thromboprophylaxis, early mobilization, and standardized rehabilitation pathways.

The Hype femoral component is a straight, cementless quadrangular femoral component made of TA6V titanium alloy with a dual coating: a titanium spray on the metaphyseal region and a hydroxyapatite (HA) layer covering the entire femoral component. Offset options include standard versions without collar (SCS, 130°) and with a collar (SCC, 130°), as well as lateralized (SCL), high-offset (SCHO) and coxa vara (SCV, 120°) variants, available in 11 sizes (125 to 180 mm). The Hype Mini (SCC mini) shares the same design, but with a reduced length (sizes 2 to 7; 104 to 128 mm) and is manufactured only as a collared version; a lateralized Mini variant (SCLA mini) is also available. For standard-offset femoral components, the collared SCC version was used in all cases (Figure 2).

Fig. 2.

Mini standard-offset Hype femoral component (size 6) and conventional standard-offset Hype femoral component (size 4).

Follow-up

Patients were followed up at six weeks, three months, 12 months, and two years postoperatively. At each follow-up visit, radiological assessment included anteroposterior views of the pelvis and the operated hip, as well as a lateral view.

Primary outcome measurement

The primary outcome was femoral component migration at two years, measured with EBRA-FCA. This validated software allows accurate assessment of axial migration based on 19 reference points, using successive anteroposterior radiographs of the hip overtime (Figure 3). In the present study, the day-one postoperative radiograph acquired in our department was used as the reference image for all migration measurements. It offers a specificity of 100% and a sensitivity of 78% for detecting implant migration greater than 1 mm, compared with the gold-standard radiostereometric analysis (RSA) technique.^12,13 Interobserver repeatability of EBRA-FCA measurements has been previously validated, with good to excellent agreement between observers (Cronbach’s α = 0.84).¹² When a radiograph lacks sufficient comparability, EBRA-FCA classifies it as nonanalyzable, leading to a missing value in the dataset. As demonstrated by Krismer et al EBRA-FCA allows the prediction of femoral fixation failure with a sensitivity of 69% and a specificity of 79%, using a migration threshold of 1.5 mm over two years.¹⁴ In addition to migration, other radiological signs of mechanical failure, including radiolucent lines, cysts, and femoral component-bone interface changes, were assessed qualitatively when migration exceeded predefined thresholds. Although the primary endpoint of the study was migration at two years, patients were followed clinically and radiologically for up to five years.

Fig. 3.

Diagram and example of the 19 reference points marked around the femoral implant in Ein Bild Roentgen analyze - femoral component analysis (EBRA-FCA) migration analysis.

Secondary outcomes

Secondary endpoints included the occurrence of intraoperative and postoperative complications, functional outcomes assessed by the 12-item Oxford Hip Score (OHS)^15,16 at 12 months.

Statistical analysis

This study was designed as a noninferiority trial with a prespecified margin of 1 mm for axial femoral stem migration. The primary analysis compared mean axial migration at two years between groups, expressed as the difference Mini minus Std. Noninferiority of the Mini femoral component was concluded if the upper bound of the two-sided 90% CI for this difference remained below 1 mm, corresponding to a one-sided α level of 5%.

Continuous variables were compared using independent-samples t-tests. Categorical variables were compared using chi-squared or Fisher’s exact tests, as appropriate. The primary noninferiority analysis was performed on a per-protocol basis. Secondary outcomes were analyzed with a two-sided α of 0.05.

Although not all patients had analyzable radiographs at the two-year endpoint, the vast majority had analysable radiographs at earlier postoperative time points. To address the missing data at this endpoint, a sensitivity analysis was conducted using multiple imputation by chained equations (MICE). This approach used available intermediate follow-up radiographs to inform the imputed values, ensuring that the imputation was grounded in observed data from earlier time points. The MICE package in R (R Project for Statistical Computing, Austria) was used, and linear regression models were performed on both datasets to confirm the robustness of the results. Statistical significance was set at p < 0.05.

Results

Primary outcome

At two years, mean axial migration was 0.94 mm (SD 0.49) for the Mini group and 0.86 mm (SD 0.52) for the Std group, yielding a mean difference of 0.08 mm (Table II). The upper bound of the two-sided 90% CI (–0.09 to 0.25) remained below the prespecified noninferiority margin of 1 mm. In the sensitivity analysis, mean axial migration was 0.95 mm (SD 0.50) for the Mini group and 0.97 mm (SD 0.57) for the Std group (Table III). Results of the sensitivity analysis were consistent with the primary analysis (90% CI –0.19 to 0.15) (Table IV).

Table II.

Axial femoral component migration over time assessed by Ein Bild Roentgen analyze - femoral component analysis.

Timepoint	Group	N*	Mean migration (SD)	Median migration, mm (range)	Mean FU, mnths (SD)	FU range, mnths	p-value*
6 weeks	Mini	42	0.25 (0.38)	0.21 (0.01 to 1.49)	1.18 (0.24)	0.66 to 1.64	0.766
	Std	35	0.27 (0.28)	0.25 (0.01 to 0.82)	1.27 (0.31)	0.56 to 2.16
3 months	Mini	44	0.43 (0.45)	0.41 (0.01 to 1.64)	3.29 (0.59)	2.56 to 5.54	0.173
	Std	46	0.32 (0.35)	0.25 (0.01 to 1.17)	3.35 (0.63)	2.66 to 5.38
12 months	Mini	41	0.66 (0.46)	00.72 (0.06 to 2.16)	12.83 (1.27)	10.43 to 15.11	0.659
	Std	41	0.62 (0.50)	0.50 (0.07 to 1.83)	12.22 (1.59)	9.11 to 15.25
2 years	Mini	49	0.94 (0.49)	0.91 (0.06 to 2.31)	24.26 (2.12)	18.95 to 29.48	0.434
	Std	49	0.86 (0.52)	0.75 (0.03 to 2.15)	25.31 (2.22)	19.38 to 29.02

^*p-values refer to between-group comparisons of axial migration.

FU, follow-up; Std, standard.

Table III.

Mean migration values in mm over time, with data imputation.

Time	Std, n	Mean Std (SD)	Mini, n	Mean Mini (SD)
6 weeks	70	0.27 (0.4)	69	0.28 (0.4)
3 months	70	0.45 (0.5)	69	0.44 (0.4)
12 months	70	0.74 (0.6)	69	0.71 (0.5)
24 months	70	0.97 (0.6)	69	0.95 (0.5)

Std, standard.

Table IV.

Sensitivity analysis results comparing original and imputed data.

	Original data				Imputed data
Variable	N	β	90% CI	p-value	N	β	90% CI	p-value
Total group	98				139
Mini		—	—			—	—
Standard		0.08	(-0.09 to 0.25)	0.434		-0.02	(-0.19 to 0.15)	0.831

Nine femoral components (9.1%) exhibited migration > 1.5 mm at two years (five Mini, four Std; p = 1.000). Migration occurred mainly during the first postoperative year (mean 1.56 mm at 12 months and 1.85 mm at two years). Maximum migration was 2.31 mm (Mini group). Mean weight among patients with migration > 1.5 mm was 65 kg. Only one patient had a femur classified as Dorr C. A non-progressive periprosthetic radiolucent line involving less than 50% of the interface was observed in one patient in the Std group. A distal pedestal was present in two patients in the Mini group and in one patient in the Std group.

Secondary outcomes

Five intraoperative complications were recorded. One diaphyseal periprosthetic fracture occurred in the Mini group and required plate and cable fixation. Three greater trochanter fractures were observed (two Std, one Mini), none requiring fixation. One transverse acetabular fracture occurred in the Mini group after cup impaction and did not require surgery. Overall, the intraoperative complication rates were 2.8% for the Std group and 4.3% for the Mini group (p = 0.681).

Four postoperative complications occurred, all unrelated to the femoral component. One aseptic loosening of the acetabular component in the Mini group required revision. A single dislocation (Mini, Röttinger approach) was treated nonoperatively. Another Mini patient underwent reoperation for a painful cementophyte. One Std patient (DAA) developed persistent neuropathic pain. Mean OHS at 12 months was 17.1 (SD 7.5) in the Std group and 18.3 (SD 8.8) in the Mini group (p = 0.387).

Discussion

Shortened femoral components derived from conventional femoral designs aim to combine the benefits of short femoral components with the proven reliability and effectiveness of conventional implants. Among these, short quadrangular femoral components are increasingly used, despite the lack of prospective comparative studies confirming their clinical efficacy and absence of mechanical failure risk. In this prospective RCT, radiological stability at two years, assessed by axial migration, was consistent with noninferiority of the Hype Mini femoral component compared with the standard Hype femoral component.

Our study is the first to examine the migration profile of a shortened, quadrangular femoral component design within a comparative, prospective, and randomized study. Girardot et al¹⁷ investigated complications occurring within the first 60 days after implantation of the Hype standard or Symbol femoral components (Dedienne Santé, France), the latter being a shortened quadrangular implant, including migration exceeding 3 mm.¹⁷ However, their study was retrospective, had a short follow-up period, and did not report migration values. Our findings for the standard Hype femoral component align with previous studies. Notably, Imagama et al³ reported a mean axial migration of 1.3 mm at 18 months for the collared Corail femoral component (DePuy Synthes, USA), which shares a very similar design as the standard Hype femoral component, using EBRA-FCA.¹⁸

Nine femoral components exhibited radiological migration greater than 1.5 mm, a threshold identified by Krismer et al¹⁴ as a risk factor for long-term mechanical failure. However, this threshold should be interpreted with caution, as it was based on analyses that did not distinguish between cemented and uncemented femoral components using EBRA-FCA. Most migration occurred during the first postoperative year, with a subsequent plateau in the migration curve. Although this finding may initially seem concerning, it aligns with rates reported in other series. In their series of 109 hips implanted with the Corail stem, Dammerer et al¹⁸ reported that 23.9% exceeded the 1.5 mm threshold at two years. Cemented femoral components generally show lower migration than uncemented femoral components measured with EBRA-FCA, which may explain why Krismer’s threshold appears more stringent for uncemented designs.^9,10 Streit et al¹⁹ evaluated the ability of EBRA-FCA to predict long-term aseptic mechanical failure of the cementless CLS Spotorno stem (Zimmer Biomet, USA), and proposed a migration threshold of 2.7 mm for uncemented femoral components. In this study, no femoral component exceeded 2.7 mm of axial migration at two years.

Several risk factors have been identified in relation to femoral component migration after implantation. Body weight is among the most frequently cited, with overweight and obese patients tending to show greater levels of migration.²⁰ The Dorr morphological classification may also play a role, with type B femora appearing more prone to higher migration than type A femora.²¹ Independently of the Dorr classification, insufficient endomedullary filling of the femur by the femoral component could contribute to increased axial migration.²² However, current evidence remains inconclusive, and no consensus has been reached on this matter. In our sample, these risk factors were not over-represented among patients who experienced migration greater than 1.5 mm. Finally, collared femoral components appeared to demonstrate reduced migration compared with collarless femoral components.¹⁸ As femoral components used in this study were collared, this factor did not influence the results.

Another notable observation is that femoral components continued to migrate beyond three months postoperatively. It could be anticipated that initial migration occurs mainly during the first three months after implantation, reflecting early osseous remodelling, with femoral components generally stabilizing thereafter as osseointegration progresses. Callary et al²³ studied the collarless Corail femoral component using RSA and found a mean migration of 0.63 mm at six years in a prospective cohort of 30 patients, with most migration occurring in the first six months and becoming negligible afterward. Our results differ, but this phenomenon of early subsidence with later stabilization is not unknown and was already described by Krismer et al.¹⁴ Similarly, Freitag et al²⁴ reported that migration of the Fitmore stem (Zimmer Biomet, USA), as assessed by EBRA-FCA, continued up to 24 months, with no significant changes observed between 24 and 60 months. Long-term monitoring of femoral component migration up to five years in our patients would provide valuable insights.

Short femoral components have long been considered at higher risk of intraoperative fracture, particularly metaphyseal-anchored designs.^7,25 Four fractures occurred in our series (three greater trochanter, one diaphyseal), with no clear group difference, corresponding to a 2.9% rate among the 139 clinically analyzed patients. Reported fracture rates with uncemented femoral components vary from 1% to 20%, depending on femoral component design. A Danish registry study found a 1.3% intraoperative fracture rate with uncemented implants in primary THA excluding femoral neck fractures.²⁶ Increasing evidence is available for shortened femoral components: Molli et al²⁷ reported fracture rates of 3.1% and 0.4% for TaperLoc compared with MicroPlasty stems in nearly 600 patients, and Girardot et al¹⁷ reported 2.8% and 0.3% for Hype Std compared with Symbol femoral components.¹⁷

OHS at 12 months were very similar between groups, with no significant difference. Recent data on type 4 femoral components support this finding.²⁵ Similarly, Girardot et al¹⁷ found no significant differences in functional scores between the Hype Std and Symbol femoral components.

The main strength of this study lies in its methodology. It is a comparative, prospective, randomized trial. To our knowledge, it is the first of this scale to compare radiological migration of a conventional quadrangular femoral component and its shortened version. More broadly, very few publications have compared standard and type 4 femoral components of identical design within a prospective randomized study.

The main limitation of this study is the reduced number of hips with radiographs obtained within the predefined two-year time window and meeting EBRA-FCA comparability criteria, which resulted in fewer complete observations for the primary analysis. Patient follow-up was conducted partly during the COVID-19 pandemic, which occasionally led to delays in scheduled follow-up visits. Radiographs obtained substantially later than two years were not used to estimate two-year migration, as this would have introduced time-related bias. Despite this limitation, sensitivity analyses using multiple imputation yielded consistent results. Moreover, axial migration showed lower variability than anticipated.

In conclusion, in this RCT, two-year radiological stability, assessed by axial migration, was consistent with noninferiority of the shortened Hype Mini femoral component compared with the conventional Hype Standard femoral component. Although the planned sample size was not fully reached, these findings are consistent with noninferiority of the shortened quadrangular femoral component in cementless THA.

Take home message

- The early radiological stability of a shortened uncemented quadrangular stem was consistent with non-inferiority compared to its conventional-length counterpart.

- These findings suggest no increased mechanical risk, with comparable functional outcomes and complication rates for both designs.

- Shortened quadrangular stems thus represent a legitimate option for primary total hip arthroplasty in patients with osteoarthritis.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.