Secondary Logo

Journal Logo

Original Contributions

Biomechanical Evaluation of a Double-Threaded Pedicle Screw in Elderly Vertebrae

Mummaneni, Praveen V.*†; Haddock, Sean M.*†; Liebschner, Michael A. K.; Keaveny, Tony M.†‡; Rosenberg, William S.*

Author Information
Journal of Spinal Disorders & Techniques: February 2002 - Volume 15 - Issue 1 - p 64-68
  • Free


During the past decade, the use of pedicle screws in spinal stabilization has dramatically increased (1,2). Failure of implanted pedicle screws to provide adequate stabilization can necessitate additional surgical procedures to achieve spinal fusion. One mode of failure of these implants is screw pullout (3–7). It is commonly believed that design factors of the screw that can increase pullout strength should lead to the improved clinical performance of pedicle screws (3–7).

Factors previously shown to be associated with increased pedicle screw pullout strength include higher insertional torque, higher bone mineral density, increased screw length, and increased screw outer thread diameter (3,5,6,8–13). Other variables, including the shape of the screw tooth profile and small changes (<1 mm) in the minor diameter of the screw, do not alter pullout strength (14). In addition, cylindrical and conical screw shapes, with or without self-tapping threads, have similar pullout strengths. Even variations of up to 20° in the angle of screw insertion have not been shown to significantly alter screw pullout strength (10,11,15).

Screw pullout strength may be enhanced by adding an additional thread to the screw shaft. Bone screws with two different threads on opposite ends have previously been shown to have a compressive effect between the two threaded components that pulls displaced fracture fragments together (16,17). These screws have been used successfully to treat displaced scaphoid and metatarsal fractures that were otherwise difficult to reduce (16,17). Pedicle screws with two different nonparallel threads at opposite ends of the screw have been used to reduce and stabilize lumbosacral spondylolisthesis (18,19). A “double-threaded” pedicle screw that has two parallel threads of differing heights throughout the full length of the screw shaft is yet another design possibility. There are no published reports on the biomechanical performance of any screws with two threads, nor are there any reports of their application in vertebral bodies in the elderly.

Osteoporotic vertebrae in older individuals have very low pedicle screw pullout strengths (8,9). Consequently, any screw design that increases pullout strength could improve spinal fusions in elderly patients. The goal of our study was to test the hypothesis that a double-threaded pedicle screw placed in donor vertebrae from the elderly could exploit the compressive effect reportedly generated by two different threads to increase bone purchase and thereby increase pullout strength. The specific objectives of this study were the following: 1) to test the insertional torque, pullout strength, energy-to-failure, and stiffness of a new double-threaded pedicle screw (Orthopedic Biosystems Ltd., Scottsdale, AZ, U.S.A.) in elderly cadaver pedicles, and 2) to compare its performance to that of a single-threaded pedicle screw with similar dimensions that is currently in clinical use (Texas Scottish Rite Hospital [TSRH] pedicle screw, Medtronics Sofomor Danek, Memphis, TN, U.S.A.).


Thirty-two human vertebral bodies (levels T12–L5) were harvested from 12 embalmed cadavers (7 female, 5 male; mean age ± SD: 85 ± 7.3 years, range: 73–94 years), soft tissues were removed, and the bones were stored in plastic bags at −20°C until testing. The vertebral bodies were divided in half in the sagittal plane using an electric bone saw to provide two paired bone samples per vertebral body (9) (Fig. 1). The advantage of using paired bone samples is that we were then able to use contralateral pedicles of the same vertebral body for placement of different screw types, thereby controlling for any interbody variations in bone mineral density. Radiographs were taken to exclude specimens that were damaged or had pedicles <7 mm in diameter. Eleven of the 32 specimens were excluded after this process, resulting in a total of 21 paired specimens for testing.

FIG. 1.
FIG. 1.:
Artist's depiction of specimen division, screw insertion, specimen mounting, and screw pullout. [Adapted from: Hirano T, Hasegawa K, Takahashi HE, et al. Structural characteristics of the pedicle and its role in screw stability. Spine 1997;22:2504–10; with permission.]

Standard, single-threaded TSRH pedicle screws (Medtronics Sofamor Danek, Memphis, TN, U.S.A.) and double-threaded pedicle screws (Orthopedic Biosystems Ltd., Scottsdale, AZ, U.S.A.) of similar dimensions were selected for testing (Fig. 2). Both screw types had identical outer diameters (6.5 mm) and outer thread pitch (2.8 mm), but the double-threaded screws had a second parallel thread of lower height (5.2-mm diameter). The inner diameters of the two screws differed by 0.0 to 0.7 mm because of the tapered nature of the TSRH screw (Table 1). Because published data suggest that such small changes of inner diameter are not significant in pullout testing (14), the potential confounding effect of this difference was considered negligible compared with the advantage of making a biomechanical comparison between the new double-threaded screw and a standard clinically used single-threaded screw.

Dimensions of the screws tested
FIG. 2.
FIG. 2.:
Photograph of the double-threaded screw (top) and the single-threaded screw (bottom) used in our experiment.

Before screw placement, an awl was used to create a pilot hole in the pedicle, but no drilling or tapping of the holes was performed (10). All screws were inserted by a single surgeon (P.V.M.) to a length of 30 mm according to the standard surgical method of Magerl (20). All screws were inserted parallel to the long axis of the pedicle, and no screws penetrated either the cortex of the pedicle or the cortex of the anterior vertebral body (Fig. 1). Insertional torque was recorded during screw placement using a modified torque screwdriver (Transducer Techniques, Temecula, CA, U.S.A.).

After the two screw types were placed into the 21 paired specimens, radiographs were again taken to confirm proper screw placement. The 42 vertebral body halves were wrapped in latex and embedded in a plastic cup with bone cement (Bosworth Fastray, Skokie, IL, U.S.A.) with the tops of the screws protruding out of the cups (21). Care was taken to prevent the bone cement from coming into contact with the protruding pedicle screws. The cups were then mounted in a vise on a servohydraulic testing frame (MTS, Eden Prairie, MN, U.S.A.). The screws were gripped by a chuck attached to the hydraulic actuator. The long axis of each screw was aligned co-linearly with the hydraulic actuator to minimize bending. Each screw was then pulled fully from the pedicle using a constant displacement rate of 1 mm/s) (Fig. 1) (6,10). Load was measured using a 1,115 N (112.5 kg) load cell (AMTI, Watertown, MA, U.S.A.), and displacement was measured internally from the linear variable displacement transducer (MTS).

Three output parameters were defined from the resulting force-deformation curves: maximum pullout force, energy-to-failure (area under the force-deformation curve), and stiffness (slope of the initial portion of the force-deformation curve) (Fig. 3) (11). These output parameters were then compared statistically between the two screw types using a paired t test. Regression analysis was also performed for the two screw's pullout strengths (Microsoft Excell 98, Microsoft Corporation, Redmond, WA, U.S.A.).

FIG. 3.
FIG. 3.:
Sample force-deformation plot for the double-threaded screw. The stiffness, maximum pullout strength, and energy-to-failure are labeled. Force-deformation plots for the single-threaded screw were similar.


The biomechanical performance of the double-threaded screw was evaluated. The mean ± SD values of insertional torque for the double-threaded screw was 6.02 ± 2.69 Nm [85.3 ± 38.1 inch pounds (in-lb)], ranging from 1.41 to 11.44 Nm (20–162 in-lb). For maximum pullout strength, the mean ± SD values was 567 ± 238 N, ranging from 128 to 980 N. For energy-to-failure, the mean ± SD values were 1.04 ± 1.12 J, ranging from 0.07 to 3.72 J. Finally, the stiffness values for the double-threaded screw had a mean ± SD of 292 ± 121 N/mm, ranging from 110 to 563 N/mm (Table 2).

Characteristics of screw performance

Four parameters of the performance of the two screw types were then compared. First, the mean ± SD values of insertional torque for the single-threaded screw was 6.89 ± 3.35 Nm [97.6 ± 47.5 in-lb), ranging from 1.48 to 13.70 Nm (21–194 in-lb). The insertional torque values were higher for the single-threaded screw in 15 of the 21 matched pairs, and this difference was statistically significant (p = 0.04) (Table 2). Second, for the single-threaded screw, the mean ± SD values for maximum pullout strength was 614.67 ± 261.77 N, ranging from 168 to 1,042 N. The maximum pullout strength values were higher for the single-threaded screw in 14 of 21 of the matched pairs. This difference was not statistically significant (p = 0.12) (Table 2). Regression analysis of the maximum pullout strengths of the single-threaded screw versus the double-threaded screw revealed that the intercept of the regression line was not statistically different from zero (p > 0.34) and the slope was not significantly different from unity (p > 0.5) (Fig. 4). Third, the energy-to-failure for the single-threaded screw had mean ± SD values of 0.88 ± 0.96 J, ranging from 0.11 to 4.30 J. The energy-to-failure was higher for the double-threaded screw in 13 of 21 cases, but this difference was not statistically significant (p = 0.29) (Table 2). Finally, mean ± SD stiffness for the single-threaded screw was 317.05 ± 165.18 N/mm, ranging from 72 to 588 N/mm. Stiffness was higher for the single-threaded screw in 11 of the 21 paired specimens, but this too was not statistically significant (p = 0.54) (Table 2).

FIG. 4.
FIG. 4.:
Direct comparison of maximum screw pullout strength of the single-threaded pedicle screw versus the double-threaded screw. The intercept of the regression line was not statistically different from zero (p > 0.34), and the slope was not significantly different from unity (p > 0.5).


The goal of this study was to test the hypothesis that a double-threaded pedicle screw would have greater bone purchase and pullout strength in elderly bone than would a typical clinically used single-threaded screw. During the pullout tests, both screw types pulled straight out of the pedicle, stripping the interface between the screw threads and the trabecular bone in a similar fashion. The major diameter of the screw determines the amount of bone sheared off during pullout (14). The minor screw diameter has been shown in previous studies to be a relatively less important variable for screw pullout (14), and we have found that a second lower-height inner thread is also a less significant variable. A post-hoc statistical power analysis indicated that we could detect a 15% statistically significant difference between the groups in maximum pullout strengths using our 21 paired specimens. We found that the mean maximum pullout strength was 7% higher for the single-threaded screw, but this difference was too small to be statistically significant. We conclude, therefore, that the second smaller thread on the double-threaded screw is not likely to significantly increase either bone purchase or pullout strength in elderly vertebrae.

Regardless of any possible systematic errors in our experiment, because we implanted both screw types in the same fashion into paired pedicles, the interscrew comparison should remain valid. Because we used paired pedicles for our comparison, we did not need to perform dual X-ray absorptiometry to assess for interspecimen variation in bone mineral density. Prior studies have shown that there is minimal variation in pullout strength because of differences in bone mineral density between paired pedicles (13).

When interpreting our results, it is appropriate to realize a number of caveats. First, we chose to perform only transverse axial screw pullout in our test because the change in thread design was most likely to affect transverse shear across the screw surface. However, pedicle screws in vivo are exposed to complex multiaxial cyclic loads (7,12). To more closely simulate in vivo loads, some researchers prefer to perform axial pullout tests after first cyclically loading the screws in the sagittal plane, whereas others subject pedicle screws to simultaneous sagittal and axial cyclic loads to create a rotational moment resulting in screw pullout (7,12). It remains unclear how closely these various methods of cyclic loading actually simulate in vivo fatigue loads on the pedicle screws. Cyclic loading of the two screw types was not performed in our study, and it remains to be determined whether a significant difference exists between the two screws' performances after cyclic loading. However, it is unlikely that a major difference in screw performance will exist during complex cyclic loading when no significant difference can be found during simple axial screw testing. Second, our experiment was conducted on embalmed elderly spine specimens. Because of lower costs and greater availability, embalmed spines have been used to conduct several recent pedicle screw pullout studies (9,22–25). McElhaney et al. (26) have shown that embalmed beef long bones have decreased compressive strength but unchanged tensile strength when compared with fresh beef long bones. However, in the pullout performance of an implant, any embalming effects appear to be minor. Prior studies have shown that screws placed into embalmed human spines have pullout strengths similar to those placed into osteoporotic fresh human spines (24). Consistent with this, our pullout values are similar to those found for 6.5-mm Steffee pedicle screws implanted in fresh elderly osteoporotic vertebrae (13). Consequently, we believe that embalmed spines are an adequate testing platform for biomechanical screw testing.

Our study is unique in that it addressed the biomechanical performance of a double-threaded screw with two parallel threads of differing heights throughout the screw shaft. Prior studies focused only on the clinical applications of screws with two, different, nonparallel threads on opposite ends of the screw either to reduce displaced bone fragments or to reduce lumbar spondylolisthesis (16–19). Future alterations in pedicle screw designs will merit further biomechanical testing of new screw types in the years to come.


A second, smaller, inner thread on a double-threaded pedicle screw does not translate into either increased bone purchase or higher pullout strengths in elderly vertebral bone.


The authors thank Pamela Derish for editorial assistance. We also thank Lippincott Williams & Wilkins (Philadelphia, PA, U.S.A.) for permission to modify and use Figure 1. This work was supported in part by an Unrestricted Research Grant from Medtronics Sofamor Danek (Memphis, TN, U.S.A.). The authors have no financial interest in the materials or devices described in this article. This work was supported by University of California, San Francisco; University of California, Berkeley; and an Unrestricted Research Grant from Medtronics/Sofamor Danek.


1. Masferrer R, Gomez CH, Karahalios DG, et al. Efficacy of pedicle screw fixation in the treatment of spinal instability and failed back surgery: a 5-year review. J Neurosurg 1998; 89:371–7.
2. McCormick PC. Selection criteria for degenerative lumbar spine instability. Clin Neurosurg 1997; 44:29–39.
3. Daftari TK, Horton WC, Hutton WC. Correlations between screw hole preparation, torque of insertion, and pullout strength for spinal screws. J Spinal Disord 1994; 7:139–45.
4. Moran JM, Berg WS, Berry JL, et al. Transpedicular screw fixation. J Orthop Res 1989; 7:107–14.
5. Soshi S, Shiba R, Kondo H, et al. An experimental study on transpedicular screw fixation in relation to osteoporosis of the lumbar spine. Spine 1991; 16:1335–41.
6. Willett K, Hearn TC, Cuncins AV. Biomechanical testing of a new design for Schanz pedicle screws. J Orthop Trauma 1993; 7:375–80.
7. Zdeblick TA, Kunz DN, Cooke ME, et al. Pedicle screw pullout strength. Correlation with insertional torque. Spine 1993; 18:1673–6.
8. Halvorson TL, Kelley LA, Thomas KA, et al. Effects of bone mineral density on pedicle screw fixation. Spine 1994; 19:2415–20.
9. Hirano T, Hasegawa K, Takahashi HE, et al. Structural characteristics of the pedicle and its role in screw stability. Spine 1997; 22:2504–10.
10. Kwok AW, Finkelstein JA, Woodside T, et al. Insertional torque and pull-out strengths of conical and cylindrical pedicle screws in cadaveric bone. Spine 1996; 21:2429–34.
11. Skinner R, Maybee J, Transfeldt E, et al. Experimental pullout testing and comparison of variables in transpedicular screw fixation. A biomechanical study. Spine 1990; 15:195–201.
12. Wittenberg RH, Lee KS, Shea M, et al. Effect of screw diameter, insertion technique, and bone cement augmentation of pedicular screw fixation strength. Clin Orthop 1993; 296:278–87.
13. Zindrick MR, Wiltse LL, Widell EH, et al. A biomechanical study of intrapeduncular screw fixation in the lumbosacral spine. Clin Orthop 1986; 203:99–112.
14. Krag MH, Beynnon BD, Pope MH, et al. An internal fixator for posterior application to short segments of the thoracic, lumbar, or lumbosacral spine. Design and testing. Clin Orthop 1986; 203:75–98.
15. Schatzker J, Sanderson R, Murnaghan JP. The holding power of orthopedic screws in vivo. Clin Orthop 1975; 108:115–26.
16. Herbert TJ, Fisher WE. Management of the fractured scaphoid using a new bone screw. J Bone Joint Surg [Br] 1984; 66:114–23.
17. Sammarco GJ, Scioli MW. Metatarsal osteotomy using a double-threaded compression screw: an adjunct to revision forefoot surgery. Foot Ankle 1989; 10:129–39.
18. Aota Y, Kumano K, Hirabayashi S, et al. Reduction of lumbar spondylolisthesis using a CDI pedicle screw system. Arch Orthop Trauma Surg 1995; 114:188–93.
19. Sijbrandij S. Reduction and stabilisation of severe spondylolisthesis. A report of three cases. J Bone Joint Surg [Br] 1983; 65:40–2.
20. Magerl FP. Stabilization of the lower thoracic and lumbar spine with external skeletal fixation. Clin Orthop 1984; 189:125–41.
21. Pfeiffer M, Gilbertson LG, Goel VK, et al. Effect of specimen fixation method on pullout tests of pedicle screws. Spine 1996; 21:1037–44.
22. Hasegawa K, Takahashi HE, Uchiyama S, et al. An experimental study of a combination method using a pedicle screw and laminar hook for the osteoporotic spine. Spine 1997; 22:958–62.
23. Lynn G, Mukherjee DP, Kruse RN, et al. Mechanical stability of thoracolumbar pedicle screw fixation. The effect of crosslinks. Spine 1997; 22:1568–72.
24. Seybold EA, Baker JA, Criscitiello AA, et al. Characteristics of unicortical and bicortical lateral mass screws in the cervical spine. Spine 1999; 24:2397–403.
25. Xu R, Ebraheim NA, Ou Y, et al. Anatomic considerations of pedicle screw placement in the thoracic spine. Roy-Camille technique versus open-lamina technique. Spine 1998; 23:1065–8.
26. McElhaney J, Fogle J, Byars E, et al. Effect of embalming on the mechanical properties of beef bone. J Appl Physiol 1964; 19:1234–6.

Biomechanics; Lumbar spine; Pedicle screws; Pullout strength

© 2002 Lippincott Williams & Wilkins, Inc.