ORIGINAL ARTICLE

Does the Angle of the Nail Matter for Pertrochanteric

Fracture Reduction? Matching Nail Angle and Native
Neck-Shaft Angle
Joshua A. Parry, MD,* Ian Barrett, MD,* Bradley Schoch, MD,† Brandon Yuan, MD,*
Joseph Cass, MD,* and William Cross, MD*

Level of Evidence: Therapeutic Level III. See Instructions for

Objectives: To determine whether fixation of pertrochanteric hip Authors for a complete description of levels of evidence.
fractures with cephalomedullary nails (CMNs) with a neck-shaft
angle (NSA) less than the native NSA affects reduction and lag (J Orthop Trauma 2018;32:174–177)
screw cutout.

Design: Retrospective comparative study. INTRODUCTION

The most frequent mode of failure after cephalomedul-
Setting: Level I trauma center.
lary nail (CMN) fixation of pertrochanteric femur fractures is
Patients/Participants: Patients treated with a CMN for unstable lag screw cutout, occurring on average in 3.5% of cases.1–3
pertrochanteric femur fractures (OTA/AO 31-A2.2 and 31-A2.3) The screw typically penetrates into the anterior–superior
between 2005 and 2014. femoral head as the fracture collapses into varus.4 A tip–
apex distance (TAD) .25 mm and a varus reduction have
Intervention: CMN fixation. been shown to increase the risk of cutouts.4–7
Main Outcome Measurement: NSA reduction and lag screw Pajarinen et al8 performed a randomized study compar-
cutout.
ing CMN and sliding hip screw (SHS) fixation of unstable
peritrochanteric hip fractures and found that the femoral neck-
Results: Patients fixed with a nail angle less than their native shaft angle (NSA) collapsed an average of 4 degrees of varus
NSA were less likely to have good reductions [17% vs. 60%, during the first 6 weeks in both groups. They recommended
95% confidence interval (CI), 263% to 218%; P = 0.0005], a reduction in slight valgus to more accurately restore the
secondary to more varus reductions (41% vs. 10%, 95% CI, native hip mechanics.
9%–46%; P = 0.01) and more fractures with $4 mm of displace- At our institution, the nail angle is not typically
ment (63% vs. 35%, 95% CI, 3%–49%; P = 0.03). The cutout matched to that of the native NSA while treating pertrochan-
was not associated with the use of a nail angle less than the teric fractures, with the 125 degree CMN being the most
native NSA (60% vs. 76%, 95% CI, 256% to 18%; P = 0.5), commonly used implant. The average native femoral NSA is
varus reductions (60% vs. 32%, 95% CI, 213% to 62%; P = 0.3), 127 6 7 degrees.9 Using a nail angle that is less than the
or poor reductions (20% vs. 17%, 95% CI, 224% to 44%; native NSA in the setting of unstable pertrochanteric fractures
P = 1.0). may affect the NSA reduction, the lag screw TAD, and lag
screw sliding, all of which may increase the risk of cutout.9,10
Conclusions: The fixation of unstable pertrochanteric hip fractures To our knowledge, this relationship has not been previously
with a nail angle less than the native NSA was associated with more studied. We hypothesize that using a nail angle that is less
varus reductions and fracture displacement but did not affect the lag than that of the native NSA will increase the risk of varus
screw cutout. malreduction and lag screw cutouts.
Key Words: hip fracture, neck-shaft ankle, cephalomedullary nails
PATIENTS AND METHODS
Accepted for publication November 10, 2017. After institutional review board approval, a trauma
From the *Department of Orthopaedic Surgery, Mayo Clinic, Rochester, MN; database from a single level-1 trauma center identified 957
and †Department of Orthopaedic Surgery, University of Florida, Gaines- pertrochanteric femur fractures treated with CMNs between
ville, FL.
The authors report no conflict of interest. 2005 and 2014. From this cohort, 154 multifragmentary
Presented in part at the American Academy of Orthopaedic Surgeons Annual pertrochanteric femur fractures (OTA/AO11 31-A2.2 and
Meeting, March 2, 2016, Orlando, FL. 31-A2.3) treated with the Gamma-III CMN (short or long)
Reprints: William Cross, MD, Department of Orthopaedic Surgery, Mayo (Stryker, Mahwah, NJ) were identified. Inclusion criteria were
Clinic, Gonda 14, 200 1st St SW, Rochester, MN 55905 (e-mail: Cross.
[email protected]).
as follows: OTA/AO 31-A2.2 and 31-A2.3 fractures, available
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. preoperative and postoperative anterior–posterior (AP) and
DOI: 10.1097/BOT.0000000000001096 lateral hip radiographs, and a minimum of 6-month follow-up

174 | www.jorthotrauma.com J Orthop Trauma  Volume 32, Number 4, April 2018

Copyright Ó 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Orthop Trauma  Volume 32, Number 4, April 2018 Does the Angle of the Nail Matter?

or until fixation failure. Exclusion criteria were the inability to postoperative NSA (ICC, 0.55) with a mean difference of 1.5
measure the NSA due to gross hip rotation and/or flexion on degrees. All analyses were performed using the JMP statis-
the radiographs, an altered native NSA due to previous injury tical software package (Version 8; SAS Institute Inc, Cary,
or surgery, and less than 6 months of follow-up with no cutout. NC). A P value less than 0.05 was considered statistically
Seventy-one of the 154 fractures were excluded from the anal- significant.
ysis: 23 were without a measurable native NSA, 22 died within
6 months, and another 26 had less than 6 months of follow-up
and did not experience a cutout. A total of 83 patients were left RESULTS
for analysis, including 46 (55%) OTA/AO 31-A2.2 and 37 The average postoperative NSA was less than the native
(44%) 31-A2.3 fractures. Average radiographic follow-up NSA (127 6 5 degrees and 129 6 6 degrees, respectively,
was 3.5 years (range, 6 months to 8.5 years). The average 95% CI, 21.3 to 24.1 degree; P = 0.0002). The average nail
age at the time of surgery was 81 6 11 years. Sixty-six percent angle differed from the native NSA by 24 6 5 degrees (95%
were women. Average follow-up was 45 6 29 months (range, CI, 25 to 23 degrees; P = 0.0001), with 67 (80%) patients
6–112 months). Long and short nails were used in 74 (89%) being fixed with a nail angle less than their native NSA. The
and 9 (11%) of the fractures, respectively. Short nails were average native NSA of fractures fixed with a nail angle of 130
only used in OTA/AO 31-A2.2 fractures, per surgeon prefer- degrees (n = 11) and 125 degrees (n = 70) was similar (131 6
ence. Nail angles used in this study, as identified by the man- 5 degrees and 129 6 5 degrees, respectively, 95% CI, 22 to
ufacturer, included 120 degrees (n = 2), 125 degrees (n = 70), 5; P = 0.4), and both were greater than the NSA of fractures
and 130 degrees (n = 11). fixed with a nail angle of 120 degrees (n = 2) (120 6 0.3
The TAD was measured as previously described by degrees; P = 0.02). The 130 degree, 125 degree, and 120
Baumgartner et al4 by measuring the distance from the tip of degree CMN achieved a similar average postoperative NSA
the lag screw to the apex of the femoral head on both AP and of 129 6 5 degrees, 126 6 4 degrees, and 122 6 3 degrees,
lateral films, correcting for magnification by dividing the respectively (P = 0.1).
diameter of the lag screw by its actual diameter. The NSA Patients fixed with a nail angle less than their native
was measured on the immediate postoperative AP radiograph NSA were less likely to have good reductions (17% vs. 60%,
with the leg held in 15 degrees of internal rotation, as the 95% CI, 263% to 218%; P = 0.0005) secondary to more
angle between a line through the center of the femoral head varus reductions (41% vs. 10%, 95% CI, 9%–46%; P = 0.01)
and the center of the femoral neck and the anatomic axis of and more fractures with $4 mm of displacement (63% vs.
the femur.12 The native NSA was measured on an AP radio- 35%, 95% CI, 3%–49%; P = 0.03). Varus reductions were in
graph of the uninjured hip. Rotation on postoperative radio- an average of 10 6 4 degree varus. Using a nail angle less
graphs was controlled by multiplying the known angle of the than the native NSA was not associated with an increased
nail by the quotient of the measured NSA over the measured TAD (16 6 4 mm vs. 16 6 4 mm, 95% 23 to 1.7 mm;
nail angle to determine the corrected postoperative NSA. The P = 0.6) or lag screw cutout (5% vs. 10%, 95% CI, 222%
difference between the corrected postoperative NSA and the to 8%; P = 0.6).
measured native NSA was calculated. Reductions were con- Fracture reduction was considered good in 23 (28%),
sidered good, acceptable, or poor based on a modified crite- acceptable in 45 (54%), and poor in 15 (18%). Acceptable
rion that was originally established by Baumgaertner et al.4 and poor reductions were secondary to varus reductions in 28
Good reductions had less than 4 mm of fragment displace- (34%) and/or $4 mm fracture displacement in 36 (43%). The
ment, less than 20 degree angulation on the lateral x-ray, and OTA/AO 31-A2.3 fractures were not more likely to have
a neutral or slightly valgus NSA (less than 5 degrees of varus a varus reduction (27% vs. 39%, 95% CI, 231% to 8%;
or 20 degrees of valgus).8 Acceptable reductions met the P = 0.3), a poor reduction (16% vs. 19%, 95% CI, 219%
criteria for alignment or displacement, but not both. Poor to 13%; P = 0.7), or a cutout (0% vs. 10%, 95% CI, 220% to
reductions met neither of the criteria. NSAs were measured 0.6%; P = 0.06).
independently by 2 orthopaedic surgeons (J.A.P., I.B.). There were 5 (6%) lag screws that cutout at an average
Continuous variables are presented as mean 6 SD. of 2 6 1 months (range, 1–3 months) postoperatively. Spe-
Nonparametric statistical tests were used because of nonnor- cific details of the 5 cutouts are provided in Table 1. All
mally distributed data as determined by the Shapiro–Wilk W occurred in OTA/AO 31-A2.2 fractures. After cutout, 1
test. The Wilcoxon signed rank test was used to evaluate patient was revised to hemiarthroplasty and 4 were revised
differences of paired continuous variables. The Wilcoxon to total hip arthroplasty. There were 2 additional cutouts that
rank sum test was used to compare unpaired continuous occurred late, one at 8 months secondary to nonunion and
variables between treatment groups. The x2 test and the fisher another at 33 months secondary to a traumatic fall, both of
exact test were used to evaluate differences between cate- which were excluded from the cutout analysis.
gorical variables. The two sample t test for proportions was Lag screw cutout was associated with an increased
used to estimate the 95% confidence interval (CI) for the TAD (23 6 7 mm vs. 16 6 4 mm, 95% CI, 2–10 mm; P =
difference between proportions. The interrater reliability of 0.001) and the use of a short nail (80% vs. 6%, 95% CI, 29%–
the NSA measurement was quantified by calculating the in- 97%; P = 0.0003). The cutout was not associated with the use
traclass correlation coefficient (ICC). The interrater reliability of a nail angle less than the native NSA (60% vs. 76%, 95%
of the native NSA was determined to be good (ICC, 0.63) CI, 256% to 18%; P = 0.5), varus reductions (60% vs. 32%,
with a mean difference of 1.4 degrees and fair for the 95% CI, 213% to 62%; P = 0.3), poor reductions (20% vs.

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. www.jorthotrauma.com | 175

Copyright Ó 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Parry et al J Orthop Trauma  Volume 32, Number 4, April 2018

TABLE 1. Cutout Details

Postoperative
Nail Angle NSA
Mismatch Mismatch
With Native With Native Lag Screw Time to
Age OTA/AO Nail NSA NSA Position in TAD Cutout
# (y) Sex Code Length (degrees) (degrees) Femoral Head Reduction (mm) (mo) Treatment
1 85 F 31-A2.2 Short 29 29 Center–center Acceptable 28 2 THA
2 88 F 31-A2.2 Short 29 6 Superior–center Good 28 1 HA
3 90 F 31-A2.2 Short 7 1 Center–center Good 11 3 THA
4 57 F 31-A2.2 Long 210 212 Center–center Poor 24 2 THA
5 77 F 31-A2.2 Short 2.5 27.5 Center–center Acceptable 23 2 THA
HA, hemiarthroplasty; THA, total hip arthroplasty.

17%, 95% CI, 224% to 44%; P = 1.0), age (79 6 13 vs. 81 Hypothetically, a mismatched nail angle may affect the
6 12, 95%, 213 to 9; P = 0.7), female sex (100% vs. 71%, ability to obtain a satisfactory TAD because of lag screw
95% CI, 213% to 42%; P = 0.3), or body mass index (24 6 4 positioning. However, Walton et al9 showed that implant
vs. 26 6 7, 95% CI, 28 to 5; P = 0.7). angle mismatch with the native NSA did not affect the ability
to obtain a satisfactory TAD, with an implant angle of 130
degrees obtaining a similar TAD when used with a native
DISCUSSION NSA ,125 and .125 degrees. Our study supports these
Lag screw cutout is the most frequent mode of failure findings, as the TAD obtained was not affected by the use
after CMN fixation of unstable pertrochanteric femur frac- of a nail angle less than the native NSA.
tures (OTA/AO 31-A2).1–3 A varus reduction has been linked The high rate of cutout among short nails was an
to an increased risk of cutout after pertrochanteric fracture unexpected finding for which we do not have a good
fixation; therefore, a slight valgus reduction has been advo- explanation. Short nails were only used in OTA/AO 31-
cated by some authors.2,3,13,14 We hypothesized that using A2.2 fractures at the discretion of the treating surgeons.
a nail angle that was less than the native NSA would increase Multiple studies have shown equivalent outcomes between
the risk of varus malreduction and lag screw cutout in these short and long nails for these fractures.17,18 A recent prospec-
fractures. Our study observed that patients treated with a nail tive randomized study of 212 patients comparing extramedul-
angle that was less than the native NSA were less likely to lary SHS and short CMNs for OTA/AO 31-A2 fractures
have good reductions secondary to a higher incidence of found a cutout rate of 2% (2/92) and 0.5% (1/112),
varus reduction and fracture displacement. Although this respectively.19
was statistically significant, poor and varus reductions were Strengths of this study include that it is a series of
not associated with lag screw cutout. However, our findings similar pertrochanteric fractures treated with 1 implant by
are limited by the low number of cutout events (n = 5). a heterogeneous group of treating surgeons (resident surgeons
Adruskow et al6 reviewed 235 patients with intertrochan- with trauma and nontrauma supervising staff), with an
teric fractures treated with gamma nail or SHS fixation and found average follow-up of 45 months. Weakness included its
no cases of cutout in those fractures reduced in 5 degrees–10 retrospective nature, large number of exclusions, and small
degrees of valgus, although this did not reach statistical signifi- numbers.
cance (P = 0.09). A varus malunion may affect functional out- A majority (80%) of patients in this study were treated
comes by causing limb shortening, abnormal abductor muscle with a nail angle less than the native NSA; thus, the
tension, and altered hip mechanics.15,16 A valgus reduction has association between nail angle mismatch and malreduction
the added benefit of bringing the nail closer to the mechanical may be secondary to sampling bias because of the small
axis of the limb, thereby decreasing the bending force placed on number of patients treated with a nail angle equal to or greater
the nail and may reduce shear forces across the fracture by ori- than the native NSA. With only 5 cutouts observed, this study
enting it in a more horizontal position.6,7,15,16 is also underpowered to determine the risk between cutouts
In this study, reduction quality was measured based on and the use of a nail angle less than the native NSA and/or
a modified criterion established by Baumgartner et al.4 In reduction quality.
their study of 109 unstable intertrochanteric fractures fixed The reliance on the radiographic measurement of the
with SHS, there were 32% good, 44% acceptable, and 23% NSA is another limitation of this study. In the setting of
poor reductions, and reduction quality was not associated comminuted intertrochanteric fractures, especially with a free
with cutouts. The overall cutout rate of unstable fractures lesser trochanteric fragment, ensuring correct rotation of the
was 12%. In comparison, our study had similar findings, with femur is difficult. The measurement of the NSA on radio-
28% good, 54% acceptable, and 18% poor reductions. The graphs has low interobserver and intraobserver reliability and
overall cutout rate was 6%. We also did not find an associa- is affected by rotation and flexion of the hip.20,21 Assessment
tion between the reduction quality and cutout. of proper rotation in a retrospective series is difficult. We did

176 | www.jorthotrauma.com Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

Copyright Ó 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Orthop Trauma  Volume 32, Number 4, April 2018 Does the Angle of the Nail Matter?

attempt to control for rotation using an algorithm to correct 7. Parker MJ. Valgus reduction of trochanteric fractures. Injury. 1993;24:
for rotation based on the known NSA of the nail, although to 313–316.
8. Pajarinen J, Lindahl J, Savolainen V, et al. Femoral shaft medialisation
our knowledge, this method has not been validated. Future and neck-shaft angle in unstable pertrochanteric femoral fractures. Int
studies are needed to validate this method. Orthopaedics. 2004;28:347–353.
Another important factor for pertrochanteric fracture 9. Walton NP, Wynn-Jones H, Ward MS, et al. Femoral neck-shaft angle in
reduction that was not considered in this study was the nail extra-capsular proximal femoral fracture fixation; does it make a TAD of
difference? Injury. 2005;36:1361–1364.
starting point. It has been demonstrated that an eccentric 10. Waddell JP. Fractures of the Proximal Femur: Improving Outcomes.
lateral starting point for trochanteric femoral nails may lead to Philadelphia, PA: Elsevier/Saunders; 2011.
varus malreduction, so this may limit our findings.22 11. Marsh JL, Slongo TF, Agel J, et al. Fracture and dislocation classification
At our institution, the nail angle of cephalomedullary compendium—2007: Orthopaedic Trauma Association classification,
implants is not typically matched to the native NSA when database and outcomes committee. J Orthop Trauma. 2007;21:S1–S133.
12. Paley D, Herzenberg JE, Tetsworth K, et al. Deformity planning for
treating intertrochanteric hip fractures. The authors are not frontal and sagittal plane corrective osteotomies. Orthop Clin North
aware of any literature that supports matching the nail angle Am. 1994;25:425–465.
with that of the native femur to prevent malreduction. We 13. Barton TM, Gleeson R, Topliss C, et al. A comparison of the long
hypothesized that a nail with an NSA less than that of the gamma nail with the sliding hip screw for the treatment of AO/OTA
31-A2 fractures of the proximal part of the femur: a prospective random-
native NSA would increase the likelihood of a varus reduc- ized trial. J Bone Joint Surg Am. 2010;92:792–798.
tion and cutout. This study found that varus reductions, and 14. Huang SG, Chen B, Zhang Y, et al. Comparison of the clinical effec-
subsequently poor reductions, were more common when tiveness of PFNA, PFLCP, and DHS in treatment of unstable intertro-
using a nail angle less than the native femoral NSA; however, chanteric femoral fracture. Am J Ther. 2017;24:e659–e666.
with the number of patients studied, this was not associated 15. Dhammi IK, Jain AK, Singh AP, et al. Primary nonunion of intertrochan-
teric fractures of femur: an analysis of results of valgization and bone
with cutouts. grafting. Indian J Orthopaedics. 2011;45:514–519.
16. Bartonícek J, Skála-Rosenbaum J, Dousa P. Valgus intertrochanteric
osteotomy for malunion and nonunion of trochanteric fractures. J Orthop
REFERENCES Trauma. 2003;17:606–612.
1. Yu J, Zhang C, Li L, et al. Internal fixation treatments for intertrochan- 17. Hou Z, Bowen TR, Irgit KS, et al. Treatment of pertrochanteric fractures
teric fracture: a systematic review and meta-analysis of randomized evi- (OTA 31-A1 and A2): long versus short cephalomedullary nailing. J
dence. Scientific Rep. 2015;5:18195. Orthop Trauma. 2013;27:318–324.
2. Aktselis I, Kokoroghiannis C, Fragkomichalos E, et al. Prospective rand- 18. Kleweno C, Morgan J, Redshaw J, et al. Short versus long cephalome-
omised controlled trial of an intramedullary nail versus a sliding hip dullary nails for the treatment of intertrochanteric hip fractures in patients
screw for intertrochanteric fractures of the femur. Int Orthop. 2014;38: older than 65 years. J Orthop Trauma. 2014;28:391–397.
155–161. 19. Reindl R, Harvey EJ, Berry GK, et al. Intramedullary versus extrame-
3. Bojan AJ, Beimel C, Taglang G, et al. Critical factors in cut-out com- dullary fixation for unstable intertrochanteric fractures: a prospective
plication after Gamma Nail treatment of proximal femoral fractures. randomized controlled trial. J Bone Joint Surg Am. 2015;97:1905–1912.
BMC Musculoskelet Disord. 2013;14:1. 20. Olsen M, Davis ET, Gallie PA, et al. The reliability of radiographic
4. Baumgaertner MR, Curtin SL, Lindskog DM, et al. The value of the tip- assessment of femoral neck-shaft and implant angulation in hip resurfac-
apex distance in predicting failure of fixation of peritrochanteric fractures ing arthroplasty. J Arthroplasty. 2009;24:333–340.
of the hip. J Bone Joint Surg Am. 1995;77:1058–1064. 21. Wilson JD, Eardley W, Odak S, et al. To what degree is digital imaging
5. Parker MJ. Cutting-out of the dynamic hip screw related to its position. J reliable? Validation of femoral neck shaft angle measurement in the era
Bone Joint Surg Br. 1992;74:625. of picture archiving and communication systems. Br J Radiol. 2011;84:
6. Andruszkow H, Frink M, Fromke C, et al. Tip apex distance, hip screw 375–379.
placement, and neck shaft angle as potential risk factors for cut-out 22. Ostrum RF, Marcantonio A, Marburger R. Critical analysis of the eccen-
failure of hip screws after surgical treatment of intertrochanteric fractures. tric starting point for trochanteric intramedullary femoral nailing. J Or-
Int Orthop. 2012;36:2347–2354. thop Trauma. 2005;19:681–686.

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. www.jorthotrauma.com | 177

Copyright Ó 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.