A BIOMECHANICAL EVALUATION OF THE GAMMA NAIL

SHEPHERD F. ROSENBLUM, JOSEPH D. ZUCKERMAN, FREDERICK J. KUMMER, BENJAMIN S. TAM

From the Hospitalfor Joint Diseases, New York

We examined the effect of the Gamma nail on strain distribution in the proximal femur, using ten
cadaver femora instrumented with six unidirectional strain gauges along the medial and lateral cortices. The
femora were loaded to 1800 N and strains were determined with or without distal interlocking screws before
and after experimentally created two-part and four-part fractures. Motion of the sliding screw and the nail
was also determined. Strain patterns and screw motion were compared with previously obtained values for a
sliding hip screw device (SHS).
The Gamma nail was shown to transmit decreasing load to the calcar with decreasing fracture stabifity,
such that virtually no strain on the bone was seen in four-part fractures with the posteromedial fragment
removed; increasing compression was noted, however, at the proximal lateral cortex. Conversely, the SHS
showed increased calcar compression with decreasing fracture stabifity. The insertion of distal interlocking
screws did not change the pattern of proximal femoral strain.
The Gamma nail imparts non-physiological strains to the proximal femur, probably because of its
inherent stiffness. These strains may alter bone remodelling and interfere with healing. Distal interlocking
screws may not be necessary for stable interfrochanteric fractures.

A sliding hip screw (SHS) is the most commonly used more medial location compared with the lateral cortical
device for the fixation of intertrochanteric hip fractures, fixation of the SHS. Secondly, the shorter lever arm of
providing secure fixation and controlled impaction the Gamma nail should decrease the tensile strain on the
(Clawson 1964 ; Harrington and Johnson 1973 ; Ecker, implant and reduce the risk of failure. Thirdly, the
Joyce and Kohl 1975; Jacobs et al 1976; Doppelt 1980; advantage of controlled impaction, characteristic of the
Jensen, Sonne-Holm and T#{248}ndevold 1980; H#{248}gh1982; SHS, is maintained. Finally, insertion ofthe Gamma nail
Heyse-Moore, MacEachern and Jameson Evans 1983; is a closed procedure which may require a shorter
MacEachern and Heyse-Moore 1983 ; Larrson, Elloy and operative time and less soft-tissue dissection.
Hansson l988a,b). An intramedullary device has recently Biomechanical studies of the SHS have shown that
been introduced as an alternative. The Gamma nail calcar compression and lateral tensile strain on the plate
(Howmedica, Rutherford, New Jersey) is a short intra- are dependent on the type and adequacy of fracture
medullary rod, which can be distally locked, combined reduction as well as the plate angle (Chang et al 1987;
with a sliding screw that inserts into the femoral head Meislin et al 1990). Calcar compression is significantly
and neck. increased after an anatomical reduction as compared
Itsdesign offers several theoretical advantages. with a medial displacement osteotomy (Chang et al
First, intramedullary fixation should provide more 1987). An anatomical reduction is also superior to medial
efficient load transfer through the calcar because of its displacement or other non-anatomical reduction when
tested to failure (Apel et al 1989 ; Walsh, Wilkinson and
Stother 1990). Comparison of plate angles in the SHS
varying from 130#{176}
to 150#{176} showed that the calcar
S. F. Rosenblum, MD, Resident compressive strain in both stable and unstable fractures
J. D. Zuckerman, MD, Chief, Geriatric Trauma Service was highest using a 1 30#{176} device (Meislin et al 1990).
B. S. Tam, BS, Research Extent
Department of Orthopaedic Surgery, Hospital for Joint Diseases, 301
At present, little information is available on the
East 17th Street, New York, NY 10003, USA. loading characteristics of the Gamma nail (Mahomed
F. J. Kummer, PhD, Associate Director 1990), and our study was designed:
Department of Bioengineering, Hospital for Joint Diseases, 301 East
17th Street, New York, NY 10003, USA.
1) to determine the proximal femoral strain distribution
Correspondence should be sent to Dr J. D. Zuckerman.
for both stable and unstable fractures fixed with the
Gamma nail;
© 1992 British Editorial Society ofBone and Joint Surgery
0301-620X/92/3317 $2.00 2) to determine the effects of the distal interlocking
J Bone Joint Surg [Br] 1992; 74-B : 352-7. screws on the proximal femoral strain distribution ; and

352 THE JOURNAL OF BONE AND JOINT SURGERY

 A BIOMECHANICAL EVALUATION OF THE GAMMA NAIL 353

3) to compare the calcar load transmission ofthe Gamma of force in one-legged stance (Walmsley 1933 ; Pauwels
nail with that of the SHS. 1980; Chang et al 1987). All femora were initially tested
intact to serve as their own controls.
Each femur was then subjected to vertically applied
compressive loads of 0, 600, 1200 and 1800 N using a
METHODS AND MATERIALS
Materials Testing System 410 servohydraulic testing
Ten fresh-frozen intact adult femora were radiographed mechanism (MTS, Minneapolis, Minnesota). For each
in anteroposterior and lateral projections to determine gauge, strain values were recorded at each load after a
the neck-shaft angle and the nail size, and also to ensure five-minute period to allow for equilibration.
that there were no abnormal specimens that would affect The femora were randomly assigned to either a
the results. On the day oftesting, the femora were thawed locked or unlocked group, to give five locked and five
in a warm water bath for several hours, all soft tissues unlocked tests. Each Gamma nail was inserted according
were removed, and the distal condyles resected. The to the manufacturer’s protocol and radiographs were
femoral shaft was potted vertically in a 20 mm metal tube obtained before testing. Femora were then tested intact
using polymethylmethacrylate. Each femur was then with the Gamma nail unlocked. For the locked group,
instrumented with unidirectional strain gauges (Type the two distal screws were then inserted, radiography
EA, Micro-Measurements, Raleigh, North Carolina) was repeated and the femur retested. Displacement of
using cyanoacrylic adhesive (Fig. 1). They were placed the sliding screw was measured after loading at 1800 N
along the axis of the nail and in the direction of the using vernier calipers.
principal femoral strains (Kal#{233}n1961 ; Ecker et al 1975; All the femora in both locked and unlocked groups
Chang et al 1987) as follows: were then tested under a series of conditions:
1) gauge A on the proximal lateral femur just below the 1) two-part fracture;
insertion point of the sliding screw; 2) four-part fracture anatomically reduced with cerclage-
2) gauge B medially just below the level of the lesser wire fixation of the posteromedial fragment ; and
trochanter; 3) four-part fracture made unstable by discarding the
3) gauge C laterally just proximal to the anticipated posteromedial fragment and the greater trochanteric
position of the proximal locking screw; fragment.
4) gauge D medially at the same level; and All bone cuts were made with a Gigli saw as shown
5) gauges E and F just distal to the level of the distal in Figure 2. After the four-part unstable fracture had
locking screw on thelateraland medialsides, respectively. been tested, all the femora that had been locked were
During testing, each femur was held in a steel vice tested unlocked, and those that had been unlocked were
at 25#{176}
adduction in the coronal plane and vertically in the tested locked. Radiographs were taken at the conclusion
sagittal plane, a position chosen to simulate the direction of testing and compared with those taken immediately

B C

Fig. 1 Fig. 2

Diagram to show the position of the Gamma nail and the Diagrams of the osteotomies used for testing : A) two-part
location ofthe strain gauges A to F in anterior and medial views. fracture; (B, C, D) anterior, posterior, and medial views of a
four-part fracture.

VOL. 74-B, No. 3, MAY 1992

354 S. F. ROSENBLUM, J. D. ZUCKERMAN, F. J. KUMMER, B. S. TAM

Resultant
Applied
I. Intact femur Load
II Femur with nail
III Two-part fracture
IV Four-part fracture
V Postercmedial fragment discarded

Curves

U Unlocked
- - . - - Locked

gauge A
500 0
gauge B
400
- 200
w
300

200 w -400
. - 600
100

0 i j -800
I II III IV V 1000
I H ifi IV V

400
S gauge C

300
-400
w
:1.
w -500
200
“S.
-600

I H III IV V I,,.., I I Fig. 3

I H III IV V
The average strain values for gauges
A to F at 1200 N for various
0 conditions, in both locked and un-
gauge E locked testing.
. 100 200
w 200 100
w

- 300
-100

I I I I I
-200
I II III IV V
I II III IV V

after insertion of the nail to identify any changes in its Table I. Sizes of Gamma nails
used to fix eight cadaveric fe-
position. mora
Strain values were found to be linear with applied
load ; those at 1200 N were analysed for each gauge Size Angle
(mm) (degrees) Number
(1200 N was chosen to allow for comparison with our
previous studies of the SHS) (Chang et al 1987; Meislin 12 130 4

et al 1990). The calcar strain data were calculated as a 12 135 2

percentage of the intact femoral strain before nail

14 130 2
insertion. Statistical testing was performed using Stu-
dent’s t-test.

assess the validity of the analysis at 1200 N, load versus

RESULTS
strain curves were obtained over the full range of applied
Of the ten femora, two had to be excluded because of loads. These showed an average correlation coefficient of
severe cracking that developed early in the testing 0.983 on linear regression analysis, confirming the
procedure, leaving four each in the locked and unlocked evaluations done at 1200 N. The average strain readings
groups. Their average Singh index was 5. 1 ± 0.83. The at 1200 N for each gauge in both locked and unlocked
sizes of the Gamma nails used are shown in Table I. To conditions are shown in Figure 3.

THE JOURNAL OF BONE AND JOINT SURGERY

 A BIOMECHANICAL EVALUATION OF THE GAMMA NAIL 355

Gauges A and B. The proximal femur showed a tendency

towards zero strain with decreasing fracture stability. C

The proximal lateral cortex (gauge A) demonstrated C,

0
progressively decreasing tensile load with decreasing 0.

fracture stability. The calcar (gauge B) experienced C

CD

progressively less compressive load to such an extent that In

ID
in a four-part fracture with the posteromedial fragment C,

CD
discarded, virtually no load was measured. These curves C,
C,
were very similar for both locked and unlocked femora. CD
C
In general, unlocked testing oflocked femora and locked
testing of unlocked femora in four-part unreduced
fractures did not show strain patterns in any gauges that
differed appreciably from those obtained in testing other 4-part fracture with anatomical reduction
femora in their respective groups.
Locked group: gauges C to F. Gauges C and D, just
proximal to the proximal interlocking screw site (gauge
C, lateral; gauge D, medial) showed load patterns almost
identical to those of gauges A and B, respectively,
decreasing tensile strain laterally and decreasing corn- C
CD

pressive strain medially with decreasing fracture stability. C,

CD
The patterns identified in gauges E and F, however, were 0.

C
quite different from those of gauges A and B. Gauge E ID

showed an initial lateral compressive strain, which Cl)

ID
increased with decreasing fracture stability. Gauge F, C,
ID
the distal medial strain gauge, was initially under a tensile C,

C,
load, but developed significant compression with inser- CD
C
tion of the nail. This compressive strain was seen
throughout subsequent testing, although it diminished
with decreasing fracture stability.
Unlocked group: gauges C to F. The strain patterns 4-part fracture-unreduced
obtained for the unlocked group were very similar to Fig. 4
those in the locked group. For these gauges, however,
Comparisons of the calcar strain, as a percentage of that for
there was a divergence of the strain pattern for stable the intact femur, in a typical sliding hip screw (SHS) and for
fractures, primarily at the two-part fracture stage. In the two test groups. Above ; an anatomically reduced four-part
fracture : below ; four-part fracture with the posteromedial
gauges E and F there was a reversal of strain pattern, fragment discarded.
while gauge C showed an exaggerated reversal in the
locked group as compared with the unlocked group. With
decreasing fracture stability, the locked and unlocked in the unlocked group and 1.1 mm in the locked group
curves returned to similar patterns. Gauge D had a for the four-part unreduced fractures. The 130#{176}
SHS
similar reversal pattern at the four-part anatomically showed an average sliding of 1.5 mm for the same
reduced fracture stage. fracture pattern (Meislin et al 1990). In all cases, these
Figure 4 shows the change in calcar strain as a differences were not significant (j, > 0.05). Comparison
percentage of that in the intact femur for the Gamma of radiographs taken before testing and at the conclusion
nail under locked and unlocked conditions compared of the trials showed a mean nail subsidence of 0.23 ±
with a typical SHS device. Data for the SHS device were 0.16 cm for both groups which were each tested unlocked
obtained using a 135#{176} four-hole plate (Chang et al 1987). as four-part unreduced fractures.
Four-part fractures with anatomical reduction and with
the posteromedial fragment discarded are represented.
DISCUSSION
For the four-part fracture anatomically reduced, there
was a decrease in calcar compression of an average of Perhaps the most surprising finding in this study was that
98% with the Gamma nail compared with a 55% increase the Gamma nail places the calcar under progressively
with the SHS. For the unstable four-part fracture, the more compression as the fracture becomes more unstable.
Gamma nail showed 93% less calcar compression, The proximal medial cortex experienced almost no load
whereas the SHS demonstrated 30% more compressive in both the stable and unstable fracture patterns. This
calcar strain. These differences were significant at contrasts with the standard SHS, which gave increasing
p < 0.001. calcarcompression with increasing instability. Theoretic-
Screw sliding with the Gamma nail averaged 0.9 mm ally, because of the shortened lever arm and intramedul-

VOL. 74-B, No. 3, MAY 1992

356 S. F. ROSENBLUM, J. D. ZUCKERMAN, F. J. KUMMER, B. S. TAM

lary fixation, it would be anticipated that the Gamma strain seen with decreasing fracture stability, particularly
nail would medialise load towards the calcar. This was at gauge A.
not the case, however, and the difference probably results The reversal of strain seen primarily in the two-part
from the inherent stiffness of the Gamma nail, particu- fractures when comparing the locked and unlocked
larly when compared with the SHS. femora was probably due to maintenance of postero-
The increased stiffness is due to several factors. medial continuity, which allowed more physiological
First, the Gamma nail has a large proximal diameter load transmission through the cortex rather than through
(17 mm), which makes it resistant to bending. Secondly, the implant. This cortical load transmission was not
the compression screw is larger than that used with the disturbed by the locking screws, which make the device
SHS (maximum diameter 12 mm vs 7.7 mm, respec- more load-bearing. The difference between locked and
tively). The increased dimensions of the Gamma nail unlocked load transmission underscores the importance
sliding screw give it a sixfold greater area moment of of maintaining the integrity of the posteromedial frag-
inertia than that of the SHS, resulting in a significantly ment for the transmission of normal forces. This was
greater resistance to bending. Thirdly, since the max- further supported by the finding that femora that were
imum deflection of the tip of the screw is inversely unlocked and retested, and unlocked femora that were
proportional to its moment of inertia and directly locked and retested showed no significantly different
proportional to its length (60 mm to the medial edge of strains adjacent to the holes or the screws in these
the nail, which is less than that seen in the typical SHS), unreduced four-part fractures.
the Gamma nail can be expected to show significantly Because two-part fractures are rotationally stable,
less deflection than a SHS. It would be anticipated and because the use of distal locking screws confers no
therefore that the sliding hip screw and side plate would significant advantage with respect to load distribution in
bend under loading, imparting compression medially, four-part fractures, the use of such screws may be
while the Gamma nail would resist these bending forces. unnecessary for intertrochanteric fracture fixation. In
With decreasing fracture stability the Gamma nail thus addition, even with unstable four-part fractures, minimal
becomes more load-bearing, thereby reducing calcar shortening (subsidence) occurred. Any consideration of
strain. These predictions are supported by a recent report the use ofthe distallocking screws must take into account
in which the Gamma nail showed a 30% greater load to the additional operative time and exposure needed for
failure than the SHS (Mahomed 1990). their insertion as well as their potential stress riser effect.
The increased stiffness of the Gamma nail causes The distal locking screws may be more appropriate for
load to be transmitted to the femur in a fashion analogous comminuted intertrochanteric-subtrochanteric fractures,
to that seen with a femoral prosthesis. Loading studies reverse obliquity fractures, and high subtrochanteric
performed with various prostheses have shown strain fractures, where rotational control of the distal fragment
distributions very similar to those with the Gamma nail. is more problematic and shortening is more likely to
There is a reversal of the normal strain patterns seen in occur.
the intact femur, with a near-zero compressive load along The decreased sliding seen with the Gamma nail’s
the calcar and maximum strain at the tip ofthe prosthesis compression screw compared with that of the SHS may
(Oh and Harris 1978). Other loading studies with the be due to its larger diameter, which creates a greater
Zickel nail also showed a reversal of normal strain screw-nail interface area. This decreased sliding, al-
distribution where there was loss of medial cortical though not statistically significant, is of concern because
continuity in subtrochanteric fractures (Cochran, Zickel it reduces the potential for impaction at the fracture, and
and Fielding 1980). We also found significant compres- could have an adverse effect on fracture healing.
sive loads at the end region of the Gamma nail, similar We examined only surface strains, but internal strain
to those reported with femoral implants. The conse- changes or the development of hoop stresses may also
quences of these load concentrations are uncertain, but affect the clinical response. Furthermore, the use of
they may be a factor in the occurrence of fractures at the additional gauges positioned distal to the nail tip would
tip of the nail (Calvert 1990). more completely describe the redistribution of femoral
Also noteworthy was the finding that the locking strains with changing fracture stability. Similar studies
screws had no significant effect on load distribution along of the SHS have been criticised for testing under static
the proximal femur. It may be that there is a more rather than dynamic loading conditions, which more
proximal load transfer from the nail to the femur, which accurately reflect clinical use. Dynamic loading reflects a
would also explain ourobservation that the nail undergoes biological condition, however, that cannot be reproduced
minimal subsidence, even when unlocked and loaded in in cadaver femora. Further, in living bone, cyclic loading
an unstable fracture pattern. This more proximal load occurs over a period of time and results in bone
transfer may be due to the larger proximal portion of the remodelling that may significantly change the load
nail abutting the isthmus, or to the sliding screw distribution. Although static testing is not ideal, we
impinging on the lateral cortex. The latter appears more believe that it can provide information that enhances our
likely, given the increasing lateral cortical compressive understanding of the loading characteristics of this new

THE JOURNAL OF BONE AND JOINT SURGERY

 A BIOMECHANICAL EVALUATION OF THE GAMMA NAIL 357

implant and allows comparison with the SHS when tested Harrington KD, JehnstenJO. The managementofcomminuted unstable
intertrochanteric fractures. J Bone Joint Surg [Am] 1973 ; 55-
under similar conditions. A:l367-76.
Conclusions. When used to fix either stable or unstable Heyse-Moore GH, MacEachern AG, Jameson Evans DC. Treatment of
intertrochanteric fractures the Gamma nail imparts non- intertrochanteric fractures of the femur : a comparison of the
Richards screw-plate with the Jewett nail-plate. J Bone Joint Surg
physiological loads along the femur; this is probably due
[Br] 1983; 65-B :262-7.
to its inherent stiffness. The use of distal locking screws Hegh J. Sliding screw in the treatment of trochanteric and subtrochan-
did not alter load distribution to the femur. The teric fractures. mnjury 1982; 14:141-5.
significance of these laboratory findings will be more Jacobs RR, Armstrong HJ, Whitaker HJ, Pazell J. Treatment of
intertrochanteric hip fractures with a compression hip screw and
clear when the results of clinical trials become available.
a nail plate. J Trauma 1976; 16:599-603.
This study was funded in part by the American Fracture Association JensenJS,Sonne-Holm S, Tendevold E. Unstable trochanteric fractures:
Howmedica Inc., and the Hospital for Joint Diseases Research and a comparative analysis of four methods of internal fixation. Acta
Development Foundation. Orthop Scand 1980; 51 :949-62.
Although none of the authors have received or will receive benefits
KaI#{235}D
R. Strains and stresses in the upper femur studied by the
for personal or professional use from a commercial party related directly
Stresscoat method. Acta Orthop Scand 1961 ; 31 :103-13.
or indirectly to the subject of this article, benefits have been or will be
received but are directed solely to a research fund, foundation, Larsson S, Elloy M, Hansson LI. Stability of osteosynthesis in
educational institution, or other non-profit institution with which one trochanteric fractures : comparison of three fixation devices in
or more of the authors is associated. cadavers. Acta Orthop Scand l988a; 59:386-90.
Larssen S, Elloy M, Hansson LI. Fixation of unstable trochanteric hip
fractures : a cadaver study comparing three different devices. Acta
REFERENCES Orthop ScaM l988b; 59 :658-63.
Apel D, Patwardhan A, Piuzur MS, DObOZi WR. Axial loading studies MacEachern AG, Heyse-Moere GH. Stable intertrochanteric femoral
of unstable intertrochanteric fractures of the femur. C/in Orthop fractures : a misnomer? J Bone Joint Surg [Br] 1983 ; 65-B :582-3.
1989; 246:156-64.
Mahemed N. Biomechanical comparison of the Gamma nail and the
Calvert PT. Use of the Gamma nail for fixation of proximal femoral sliding hip screw. Procs sixth annual meeting of the Orthopaedic
fractures. Seminars in Orthopaedics 1990 ; 5 :101-6. Trauma Association, Toronto, 1990:57.
Chang WS, Zuckerman JD, Kummer FJ, Frankel VH. Biomechanical Meislin RJ, ZuckermanJD, Kummer FJ, Frankel VH. A biomechanical
evaluation of anatomic reduction versus medial displacement analysis ofthe sliding hip screw : a question ofplate angle. JOrthop
osteotomy in unstable intertrochanteric fractures. C/in Orthop Trauma 1990;4:l30-6.
1987; 225:141-6.
Oh I, Harris WH. Proximal strain distribution in the loaded femur : an
aa DK. Trochanteric fractures treated by the sliding screw plate in vitro comparison of the distributions in the intact femur and
fixation method. J Trauma 1964; 4:753-6. after insertion of different hip-replacement femoral components. J
BoneJointSurg[Am] 1978; 60-A:75-85.
Cochran GV, ZickelRE, Fielding JW. Stress analysis of subtrochanteric
fractures : effect of muscle forces and internal fixation devices. In: Pauwels F. Biomechanics ofthe /ocomotor apparatus : contributions on the
UhthoffH, ed. Current conceptsofinterna/fixation offractures. New functiona/anatomy ofthe /ocomotor apparatus. Berlin, etc : Springer-
York, etc : Springer Verlag, 1980:211-27. Verlag, 1980.
Doppelt SH. The sliding compression screw - today’s best answer for Walmsley 1. The vertical axes of the femur and their relations : a
stabilization of intertrochanteric hip fractures. Orthop C/in North contribution to the study of the erect position. J Anat 1933 ; 67:
Am 1980; 11:507-23. 284-300.
Ecker ML, Joyce JJ III, Kohl EJ. The treatment of trochanteric hip Walsh ME, Wilkinson R, Stother 1G. Biomechanical stability of four-
fractures using a compression screw. J Bone Joint Surg [Am] 1975; part intertrochanteric fractures in cadaveric femurs fixed with a
57-A :23-7. sliding screw-plate. mnjury 1990; 21 :89-92.

VOL. 74-B, No. 3, MAY 1992