The Effect Of Taper Angle and Spline Geometry on the Initial Stability of Tapered, Splined Modular Titanium Stems

Jeffery L. Pierson, MD
Joint Replacement Surgeons of Indiana Foundation, Inc.

Scott R. Small, MS
Joint Replacement Surgeons of Indiana Foundation, Inc.

Jose A. Rodriguez. MD
Lenox Hill Hospital

Michael N. Kang, MD
Insall-Scott-Kelly Institute for
Orthopaedics and Sports Medicine

Andrew H. Glassman, MD
The Ohio State University
There have been significant advances in revision total hip arthroplasty (THA) over the past two decades. In femoral revision surgery, tapered, splined modular titanium stems (TSMTSs) have emerged as a particularly effective option. In short to midterm follow-up, TSMTSs have reported minimum five year survivorship of 85-94 percent 1-5 and have exhibited improved quality of life measures, fewer intraoperative fractures, and better ability to reproduce leg length and offset.5-13 While these results are encouraging, component subsidence continues to be a cause for early mechanical failure and a cause for re-revision (with increased subsidence associated with severity of bone defect).5 Bohm and Bischel14 reported an average migration of 5.9mm in 149 TSMTSs at a mean 4.8 year follow-up, with 26 hips exhibiting more than 10mm of migration. Rodriguez et al.15 reported a 6.2 percent rate of subsidence up to 10mm at 6.2 years follow-up, while Park et al.5 reported 5 percent of cases with 10-20mm of subsidence at 1.6 years. In many of these cases, secondary stability was reported without re-revision, attributed to the tapered design of the stem. Interestingly, the design rationale for the degree of taper angle and spline geometry is not well documented in the literature.

The purpose of this study was to evaluate the effect of two major design elements of a TSMTS (the degree of taper angle and spline geometry design) on the initial mechanical stability of the implant, as measured by implant subsidence and torsional resistance.


 In order to perform a comparative analysis of axial and torsional stability of revision stem spline designs, custom stem samples were manufactured from wrought titanium (Ti-6Al-4V) per ASTM Standard F1472-08. Experimental groups consisted of two spline configurations (Narrow and Broad) with five taper angle groups per spline configuration (2.5°, 3.0°, 3.5°, 4.0°, 5.0°), for a total of 10 distinct sample experimental groups. Three specimens were included in each spline configuration and taper angle combination group. All test specimens consisted of a 102mm of tapered spline, 18mm in 38mm smooth cylindrical stub 13mm in diameter to integrate with the test fixture. Stems included 10 longitudinal splines spaced circumferentially at an increment of 36°.The spline geometry in the narrow configuration had a 0.4-0.5mm wide spline, whereas the broad configuration had a 0.9-1.0mm wide spline. These spline widths were chosen as they reflected the range of spline widths observed in stems that are currently on the market and available for inspection. Both configurations had a spline height of 1.9mm (Figure 1). Solid rigid polyurethane foam blocks (0.64 g/cc, Sawbones, Inc., Vashon, WA) of size 50mm x 50mm x 20mm were used as the test substrate and were reamed utilizing standard manufacturer-supplied reamers, matching stem taper angles, on a digitized mill. Roughing, followed by finishing passes were performed to a diameter at which the center 20mm of the test specimen was engaged into the foam block.

Figure 1. Mechanical drawings and three dimensional views of broad and narrow spline geometries showing taper angle and spline geometries as variables of interest in this study.


Axial and torsional mechanical testing was conducted utilizing a biaxial electrodynamic load frame (ElectroPuls E10,000 A/T, lnstron, Norwood, MA). Proximally, specimens were gripped in the upper pneumatic grip of the load frame. Distally, reamed foam specimens were placed within a hollow support chamber enabling stem insertion and rotation, while allowing for free x-y translation and constraining foam rotation. Axial tests were performed by inserting the test specimen at a displacement controlled rate of 1mm/s until a maximum specimen displacement of 15mm was reached. Maximum compressive load was calculated from the axial output of the load frame. Axial resistance was calculated utilizing a data analysis package (LoggerPro 3.8.5, Vernier Software & Technology, Beaverton, OR) as the slope of the linear region of the displacement-force curve in each trial. Specimens were cleaned and inspected for damage after each trial, with tests repeated five times per specimen in a new foam block for each trial.

The purpose of this study was to evaluate the effect of two major design elements of a TSMTS (the degree of taper angle and spline geometry design) on the initial mechanical stability of the implant, as measured by implant subsidence and torsional resistance.

Following axial testing, torsional resistance was quantified for each spline configuration and taper angle using the same experimental fixturing. Preliminary testing of axial insertion forces incrementally from 0 to 1000 N revealed an axial force of 400 N as the minimum threshold for spline penetration into the surface of the reamed foam. For this reason, a constant compressive axial preload of 400 N was applied to the stem specimen during all torsional tests as means to simulate slight cortical bone engagement at the onset of torque application. Each stem specimen was rotated within the foam block at a rate of 0.5° per second until a peak rotation of 10° was reached. As with axial testing, repeated measures were conducted for a total of five trials per specimen. Torque, rotation and axial load data were collected from the load frame controller. Peak torque at 1° of stem rotation was quantified. Peak torsional resistance was measured as the linear slope of the rotation-torque curve within the first 0.2° of stem rotation within the foam block.

Figure 2. A)Least square mean values of peak load in axial testing. * Indicates statistically significant difference in peak load between broad and narrow spline geometries at the specified taper angle (P < 0.05). B) Least square mean axial resistance. * Indicates statistically significant difference between broad and narrow spline geometries (P < 0.05).

The mean diameter of the 3.5° taper at the most proximal point of implantation in our foam model is 15.5mm. By calculating the length of the arc created at the perimeter during component rotation around the central axis, it was determined that a 1° rotation of the test specimen equates to a 140μm relative micromotion between the spline and the foam model. Because micromotion at the stem-bone interface greater than 150μm at any aspect in the component may inhibit bone ongrowth and proper biological fixation, 16 evaluating the rotational stability at this small rotational increment was deemed helpful in order to identify any subtle differences in stability between designs at the level of micro-motion critical to bone ongrowth.

In summary, we evaluated two metrics of both axial and rotational stability as derived from load frame displacement, load and torque transducers. Higher observed values in maximum axial load, the load required to generate 15mm of subsidence, along with the axial resistance, the load required per 1mm of subsidence, serve to indicate a more axially stable construct. Likewise, higher observed values in peak torque, the torque required to induce 1° of stem rotation, and axial resistance, the torque required per degree of rotation at the first 0.2° of rotation, serve to indicate increased rotational stability in the stem design.

Figure 3. A) Mean peak torque required to generate 1.0° of rotational displacement. In the repeated-measures multivariate linear regression, spline design was a significant covariate with the narrow spline averaging a 9 percent greater peak torque across all taper angles (P = 0.0005). B) Mean calculated rotational resistance as observed in torsional testing. No statistically significant difference was detected between broad and narrow spline geometries at any taper angle tested.

Statistical analysis was performed utilizing repeated-measures multivariate linear regression techniques. For both axial and rotational tests the spline geometry, taper angle, and the interaction between each were analyzed for covariance. Least square means were derived for each test response for comparison between combinations of spline geometry and taper angle. A P-value or less than 0.05 was considered statistically significant. Our study was adequately powered (with power = 0.80, two-sided alpha = 0.05, and beta [probability of type II error] = 0.20) to detect differences between designs in means of 600 N maximum axial load and 100 N/mm of axial resistance in axial tests, as well as differences of means of 0.2 Nm or peak torque and 0.8 Nm/deg of axial resistance.


The broad spline design produced significantly higher maximum compressive loads than the narrow spline design taper angles of 3.5°, 4°and 5°, representing an increase in axial stability over narrow splines of 33 percent, 42 percent and 32 percent respectively (P <0.0001) (Figure 2A). Within both the broad and narrow spline configuration, the smallest maximum compressive load was observed within stems with taper angles of 2.5° and 3°. Overall, stem specimens with broad spline configurations and a 5° taper angle exhibited 21-137 percent greater axial stability than the other spline combinations (P < 0.0001 ). There was found to be an overall greater than additive effect when considering the interaction between design and taper angle (P < 0.0001). Simply put, a greater difference in maximum axial stability between broad and narrow design was observed at higher taper angles. As a second measure of axial stability, axial resistance closely corresponded with maximum compressive load data (Figure 2B). In this measure of axial stability, the broad splines exhibited significantly higher axial resistance than the narrow splines at 3.5°, 4° and 5° of 37, 51, and 56 percent respectively (P < 0.0017). Overall, the 5° broad stem designs exhibited a 36-269 percent greater axial resistance than the geometry and taper angle combinations tested (P < 0.0001). In the axial resistance model, as with the axial compressive load model, a greater than additive effect was observed when considering the interaction between spline geometry and taper angle, with increasing influence of design at greater taper angles (P = 0.0001).

 Rotational stability was determined as the greatest torque recorded during testing at a stem rotation of less than 1.0°, as well as the peak rotational resistance during the initial stem rotation. Maximum torque required to rotate each spline 1.0° is shown in Figure 3A. In the repeatedmeasures multivariate linear regression accounting for all taper angles, the narrow spline design demonstrated a small, yet statistically significant, trend of 4-15 percent higher maximum torque required to generate 1.0° of stem rotation (P = 0.0018). Torsional resistance, representing a relative initial stiffness of the spline seating interface, is shown in Figure 3B. Neither taper angle nor spline geometry exhibited a significant overall effect on torsional resistance with the exception of an small outlier, the 4.0° tapered specimens, which exhibited slightly lower torsional resistance in both broad and narrow splines than the other taper angles (P < 0.0001).

Stem A Stem B Stem C Stem D Stem E Stem F
Taper angle 2o17 2o18(measured template)

2o19(measured template)

2.5o20 3o21 3.5022
Taper length 104mm17

(inferred: 85-190mm)23


(inferred:101- 251mm)24



Spline quantity 817

Size 12:8
Size 25:10
(measured specimen)

Size 17:8(measured

Size 14-ISmm:6
Size 16-21mm: 8
Size 22-31mm: 1024




Spline height 1-2.9mm

(sizes 19-25mm
eight increases fro proximal to

(1.5mm on size 14)27

1.5mm(measured specimen)

Spine width

(measured specimen)

0.5mm(measured specimen)




 Initial mechanical stability is a critical factor in achieving bone ongrowth and long-term success of revision THA. TSMTSs have been developed to promote high initial stability in cases of proximal bone deficiency, while reducing risk of thigh pain and intraoperative fracture when compared with other revision stem designs and philosophies. The literature contains very little of the design rationale for key elements in the design features of TSMTS. In particular, the design elements of taper angle and different spline geometries have not been well studied and vary significantly in the TSMTS currently on the market (Table 1).The purpose of the current study was to evaluate the effect of different taper angles and spline geometry on the initial mechanical stability of the implant. Specifically, we asked if differences in spline geometry and stem taper angle result in differences in (1) axial implant stability, and (2) stability to torsional stresses in a simulated revision THA with significant femoral bone loss.

 We recognize limitations associated with the methodology and clinical extrapolation of our findings. First, we utilized a non-physiological polyurethane foam model as a substitute for a dynamic in vivo environment. This high-density foam is the basis for mechanical bench testing for cortical screw pull-out testing, and is commonly used in the literature as a bone substitute, but has not been validated as behaving similarly to cortical bone in the femoral diaphysis. Nevertheless, the use of a foam model enables tight control of interspecimen variability, a common problem with cadaveric testing, providing a uniform medium to compare relative differences in stem design factors while holding all other variables constant. Second, a single material density, length of engagement, reaming and insertion procedure was used in this study, while clinical bone density, degree or fixation and interference fit vary widely clinically. Surgical technique in terms of canal reaming and implant insertion also varies in revision surgery and is dictated primarily by bone quality and surgeon preference. Homogenization of these factors provides a baseline for comparative study of the key parameters while minimizing uncontrolled ancillary interactions. Third, we performed quasi-static mechanical tests evaluating peak rotational stability while in vivo, stem loading is predominately dynamic and cyclical. Aseptic loosening and subsidence are clinically related to micromotion at the bone-implant interface, and as such, the quantification of initial axial and rotational stability generates a baseline of overall mechanical resistance to component migration for each stem design.

Initial axial stability is a key factor in long-term success of revision total hip arthroplasty, particularly in the scenario of poor proximal bone quality. Subsidence in TSMTS has been reported in a number of follow-up studies and can become more likely as bone quality diminishes. 5, 14, 15, 28-30 In a recent study, Van Houwelingen et al.30 reported a five to 10 year survivorship, or 90 percent, for ZMRR (Zimmer, Warsaw, IN) TSMTS implanted in severe femoral defects. In seven of the 65 stems implanted, component subsidence of a mean 12.3mm was measured; however, all of those implants retained secondary stability and did not require re-revision. The authors of the study hypothesized that the reason for this secondary stability to be the 3.5° stem taper angle, compared to the common 2° taper angle in other TSMTS designs. Bolstering that hypothesis, Park et al5 reported subsidence in five of 59 TSMTS implantations of the 2° tapered Lima modular femoral stems (Lima-Lto, Udine, Italy), with three re-revisions resulting from subsidence of 10-20mm. Similarly, in the 2° tapered Wagner stem (Zimmer, Warsaw. IN) a subsidence of greater than 10mm has been reported from1528 to 20 percent,14,16 resulting in re-revision rates of 616 to 10 percent28 in those subsided stems. However, it is important to remember that there are numerous factors that determine implant stability, with taper angle being one important factor. No taper angle will accommodate for poor surgical technique where intimate cortical contact is not achieved.31 The current study reports substantially increased axial resistance and stability with increased taper angle. Though numerous factors are present which lend to implant stability, these data support the hypothesis that increased taper angle is associated with decreased subsidence and improved implant stability. The current literature lacks any discussion of TSMTS spline geometry; however, our results tend to show a paired increase in axial stability between taper angle and broad spline geometry.

The current study reports substantially increased axial resistance and stability with increased taper angle.

In addition to axial resistance to subsidence, rotational stability is a key aspect to resisting aseptic loosening in total hip arthroplasty. A few biomechanical studies have documented the impact of distal stem geometry on rotational stability in comparisons between cylindrical and fluted stems. In a cadaveric study, Kirk et al. observed a 27 percent greater resistance to component micromotion in the Link MP TSMTS compared to a cylindrical cobalt stem, however micromotions in both designs were well below the threshold at which osteointegration would be inhibited. 32 Jakubowitz et al.33 quantified primary rotational stability between conical and cylindrical revision stems, with the Wagner-SL (Zimmer, Warsaw. IN) and MRP (Peter Brehm GmbH, Weisendorf, Germany) conical stems generating up to 96 percent less mean overall movement in synthetic femurs with type III defects. Likewise, in an early biomechanical study, Kendrick et al. observed significantly higher torsional stability in cementless fluted stems over porous-coated , finned and slotted finned designs.34 While these studies have demonstrated increased stability of TSMTSs over their counterparts, no differentiation in the literature has been made investigating spline design and taper angle in relation to rotational stability. The current study indicates that spline geometry does minimally influence rotational stability in some aspects, but not nearly to the degree that it affects axial subsidence. The narrow spline seems better able to “dig” into the substrate and slightly improve resistance to rotational instability, however the increase in peak torque resistance of less than 0.5 Nm observed in this study is unlikely to generate a clinically significant result.

The authors have considerable experience with cylindrical, extensively porous coated implants in revision total hip arthroplasty. Our experience, as well as the literature, has been very good with this revision strategy, with the notable exception of dealing with Paprosky types 3B and 4 femora. In these highly damaged bones, cylindrical extensively porous coated implants are technically demanding to implant and have a high failure rate due to lack of osseointegration. Because of this experience, we began to use tapered, splined stems in these challenging cases with much more favorable results. Having successfully used this reconstructive strategy on these highly damaged femora, we began to expand our use of tapered, splined stems on more revisions, including those femora with less damage (Paprosky 1-3A), also with excellent results. Our experience seems to parallel many other surgeons as the use of tapered, splined implants has grown significantly in the market. In fact, some of the authors now use tapered, splined stems for all revision femoral reconstructions. We think it is highly likely that there will be continued increase use of tapered, splined stems in revision total hip arthroplasty as more surgeons gain experience with them and the design of these implants is optimized. In conclusion, our results demonstrate that taper angle and spline geometry are important variables in achieving initial mechanical stability as measured by resistance to stem subsidence, particularly with respect to axial stability. Specifically, higher degrees of taper angle (5° taper angle) and a broad spline geometry are superior to lower taper angles and a narrower spline geometry. In terms of rotational stability, taper angle demonstrates no repeatable influence, while a narrow spline geometry exhibits minimal improvement in tolerance to peak torsional when compared to broad spline designs. Differences in axial stability between taper angles may explain some of the clinical differences (and stem subsidence rates) that are reported in the literature with TSMTS. Additionally, these data are helpful when evaluating revision stems of this general type that are currently available to surgeons and provide guidance on the development of future tapered, splined modular titanium stems. •

Reprinted from The Journal of Arthroplasty, Volume 30, Number 7, July 2015, Pages 1254-1259, with permission from Elsevier.


  1. Koster G, Walde TA. Wiiiert HG. Five-to-10 year results using a noncemented modular revision stem without bone grafting. J Anhroplasty 2008:964.
  2. D, Eliaz N, Levi O, Backstein D, Kosashvili Y, Safir O, Gross AE. Fracture of cementless femoral stems at the mid-stem junction in modular revision hip arthroplasty systems, J Bone Joint Surg 20 11;93-A:57
  3. Munro JT, Garbuz OS, Masri BA, et al. Role and results of tapered fluted modular titanium stems in revision hip arthroplasty. J Bone Joint Surg 2012:94-B:58.
  4. Palumbo BT, Morrison KL, Baumgarten AS, et al. Results of revision total hip arthroplasty with modular, titanium-tapered femoral stems in severe proximal metaphyseal and diaphyseal bone loss. J Arthroplasty 2013;28:690.
  5. Park VS, Lee JH, Park JH, et al. A distal fluted, proximal modular femoral prosthesis in revision hip arthroplasy. J Arthroplasty 2010;25.932.
  6. Garbuz OS, Toms A, Masri BA et aI. Improved outcomes in femoral revision arthroplasty with tapered fluted modular titanium stems. Clin Orthop Relat Res 2006;453:199.
  7. McInnis OP, Home G, Devane PA. Femoral revision with a fluted, tapered, modular stem: seventy patients followed for a mean 39 years. J Anhroplasty 2006;21:372.
  8. Ovesen O,Emmeluth C, Hofbauer C, et al. Revision total hip arthroplasty using a modular tapered stem with distal fixation: good shortterm results in 125 revisions. J Arthroplasty 2010;25 :348
  9. Park Y, Moon Y, Lim S. Revision total hip arthroplasty using a fluted and tapered modular distal fixation stem with and without extended trochanteric osteomy. J Arthroplasty 2007;22(7): 993.
  10. Restrepo C, Mashadi M, Parvizi J. et al. Modular Femoral sterns in revision total hip arthroplasty. Clin Orthop Relat Res 2011;469(2):476.
  11. Richards CJ, Duncm CP, Masri BA, et al. Femoral revision hip arthroplasty: a comparison of two stem designs. Clin Orthop Relat Res 2010;468:491.
  12. Rodriguez JA, Fada R, Murphy SB, et al. Two-year to five-year follow-up of femoral defects in femoral revision treated with the Link MP modular stem. J Arthroplasty 2009; 24(5):751
  13. . Schuh A, Werber S, Holzwarth U, et al. Cementless modular hip revision arthroplasty using the MRP Titan Revision Stem: outcome of 79 hips after an average of 4 years’ follow-up. Arch Orthop Trauma Surg 2004;124(5):306.
  14. Böhm P, Bischel O. Femoral revision with the Wagner SL revision stem. J Bone Joint Surg 2001;83A:I023.
  15. Rodrguez JA, Deshmukh AJ, Klauser WU, et al. Patterns of osseointegration and remodeling in femoral revision with bone loss using modular, tapered, fluted, titanium stems. J Arthroplasty 2011;26(8):1409.
  16. Pilliar RM, Lee JM, Maniatopoulos C. Observations on the effect of movement on bone ingrowth into porous surface implants. Clin Orthop Relat Res 1986;208:108.
  17. Zimmer, Inc. Wagner SL RevisionR Hip Stem Surgical Technique; 2009.
  18. Kwong LM. Surgical Technique with the MPDI Hip Stem for Revision of a Failed Femoral Hip Stem. Link Orthopaedics 2002.
  19. Stryker Orthopaedics. RestorationR Modular Revision Hip System: Surgical Protocol; 2005.
  20. DePuy Orthopaedics. ReclaimR Revision Hip System Design Rationale; 2011.
  21. Biomet, Inc. ArcosR Modular femoral Revision Stem. Undated; 2011.
  22. Inc. ZMRR Hip System: 2009.
  23. Link Orthopaedics.MPR Reconstruction Prosthesis: Cementless & Cementable Implants & Instruments; 2011.
  24. DePuy Orthopaedics. ReclaimR Revision Hip System Operating Room Reference Guide: 2011.
  25. Biomet, Inc.. ArcocR Modular Femoral Revision System: Surgical Technique; 2011.
  26. Inc. ZMRR Revision Taper Hip Prosthesis: Surgical Technique for Revision Hip Arthroplasty; 2002.
  27. Link Orthopaedics. The MPR Hip Stem Design Rationale; 2009.
  28. Grunig R. Morscher E, Ochsner PE. Three- to 7-year results with the uncemented SL femoral revision prosthesis. Arth Onhop Trauma Surg 1997;116:187.
  29. Kolsd K, Adalberth G, Mallmin H, et al. The Wagner revision stem for severe osteolysis. 31 hips followed for 1.5-5years. Acta Orthop Scand 1996;67:541.
  30. Van Houwelingen AP, Duncan CP, Masri BA, et al. High survival of modular tapered stems for proximal femoral bone defects at 5 to 10 years follow up. Clin Orthop Relat Res 2013;471:454.
  31. Patel PO, Klika AK. Murray TG, et al. Influence of technique with distally fixed modular stems in revision total hip arthroplasty. J Arthroplasty 2010;251(6):926.
  32. Kirk KL, Potter BK, Lehman Jr RA, et al. Effect of distal stem geometry on interface motion in uncemented revision total hip prostheses. Am J Orthop 2007 :36(10):545.
  33. Jakubowitz E, Bitsch RG, Heisel C, et al. Primary rotational stability of cylidrical and conical revision hip stems as a function of femoral bone defects: an in vitro comparison. J Biomech 2008;41.3078.
  34. Kendrick JB, Noble PC, Tullos HS. Distal stem design and the torsional stability of cemetless femoral stems. J Arthroplasty 1995;10(4):463.