Archives of Orthopaedic and Trauma Surgery

https://doi.org/10.1007/s00402-018-2905-1

TRAUMA SURGERY

Non-prosthetic peri-implant fractures: classification, management

and outcomes
Lester W. M. Chan1 · Antony W. Gardner2 · Merng Koon Wong3 · Kenon Chua1 · Ernest B. K. Kwek1 · on behalf of the
Singapore Orthopaedic Research CollaborativE (SORCE)

Received: 8 September 2017

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Introduction  Non-prosthetic peri-implant fractures (NPPIFs) are an under-reported entity. Management is challenging
because of alterations in anatomy, the presence of orthopaedic implants and phenomena such as stress shielding, disuse
osteopenia and fracture remodeling. The aims of this paper were to review patterns of injury, management and outcomes
and to propose a classification system to aid further research.
Materials and methods  This study is a multi-centered retrospective case series. Patients were identified from the orthopaedic
department trauma databases of public hospitals in Singapore and individual surgeon case series of members of the Singapore
Orthopaedic Research Collaborative (SORCE) group.
Results  We collected a series of 60 NPPIFs in 53 patients. 38 fractures involved the femur, 12 the radius/ulna, 5 humeri, 3
tibia/fibula and 1 clavicle. 39 patients had fractures around plates and screws, 12 around nails, and 3 around screws. Fractures
were managed with a variety of surgical techniques. Six patients had surgical complications with refracture in four and non-
union in two cases. Two patients had multiple refractures (total 12 additional fractures). All surgical complications required
further surgery. Three patients had deep vein thrombosis and one patient died of post-operative pneumonia. Fractures were
classified according to the initial implant (plate or nail), the position of the new fracture relative to the original implant
(at the tip or distant) and the status of the original fracture (healed, not healed or failing). Surgical strategies for common
subtypes were reviewed.
Conclusions  This study represents the largest series in the literature. NPPIFs are a challenging clinical problem with a high
rate of post-operative complications. They are distinct from peri-prosthetic fractures and should be understood as a separate
entity. We, therefore, propose a novel classification system. Further research is needed to determine the optimal treatment
for the various subtypes.
Level of evidence  Therapeutic Level IV—case series.

Keywords  Peri-implant fracture · Peri-prosthetic fracture · Complication · Osteosynthesis · Non-prosthetic peri-implant

fracture · Classification

Introduction
Members of the “Singapore Orthopaedic Research CollaborativE We define a non-prosthetic peri-implant fracture (NPPIF) as
(SORCE)” are listed in the acknowledgements section.
a fracture in a bone with an existing non-prosthetic implant
* Ernest B. K. Kwek such as an extramedullary plate and screws or an intramedul-
ernest_kwek@ttsh.com.sg lary nail. NPPIFs are often reported together with peri-pros-
thetic fractures (PPFs) that occur around joint replacement
1
Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, implants [1, 2]; they are, however, a distinct clinical entity
Singapore 308433, Singapore
with different conceptual and practical considerations. PPFs
2
Ng Teng Fong General Hospital, 1 Jurong East Street 21, have been extensively studied and there are multiple classifi-
Singapore 609606, Singapore
cation systems and guidelines for treatment in the literature
3
Singapore General Hospital, Outram Road, [3–5]. In contrast, the literature addressing NPPIFs is sparse.
Singapore 169608, Singapore

13
Vol.:(0123456789)

The term NPPIF does not include failures of primary
fracture fixation such as sliding hip screw cut-outs and
implant breakages due to non-union. These have also been
extensively studied.
The aims of this study were to review patterns of injury,
management and outcomes and to attempt to classify
NPPIFs into a system that may aid further research and
study. Because of the low incidence of these fractures a
multi-centre collaboration was sought.
This study is the first research effort of the Singapore
Orthopaedic Research Collaborative (SORCE): A group
made up of orthopaedic trauma surgeons with representa-
tives from public sector hospitals managing adult trauma
in Singapore.
Materials and methods
Ethics approval for this study was obtained prior to com-
mencement (National Healthcare Group DSRB Ref:
2013/00893).
Cases were contributed by SORCE members from ortho-
paedic department databases and personal cases series.
Patients aged 18 years and older with a NPPIF in a long
bone were included. Fractures around prosthetic implants
and patients with pathological bone conditions (except oste-
oporosis) were excluded.
Suitable cases were identified and demographic, injury
and outcome data were collected from case records and
X-rays. Data were reviewed for patterns of injury, trends,
management strategies and outcomes.
Results
A total of 53 patients were identified with 60 NPPIFs. A
summary of cases is shown in Table 1.
42 of these fractures (39 patients) were of the lower limb.
18 fractures (14 patients) involved the upper limb. Of the
39 patients with lower limb injuries, 2 patients sustained
fractures above and below an implant in the same injury.
Two patients sustained double bone fractures (radius/ulna
or tibia/fibula) and one patient sustained a bilateral double
forearm fracture. These were regarded as separate NPPIFs.
Average follow-up was 16 months (range 0–75). Seven
patients were lost to follow-up or had less than 3 months
follow-up.
Demographics
The average age was 64 years (range 20–100). There were 31
male and 22 female patients. Demographics and anatomical
13
Archives of Orthopaedic and Trauma Surgery
distribution are shown in Fig. 1. A bimodal age distribution
was noted.
40 patients sustained their NPPIF in a low velocity fall,
12 had a road traffic accident (RTA) and 1 was the result
of the propagation of an unrecognized perforation of the
anterior femoral cortex by a long femur nail. All 12 RTA
patients were male and tended to be younger (mean age 47).
Previous injury details
12 fractures were around intramedullary nails, all of which
were in the femur, 45 fractures were around extramedullary
plates and screws and 3 fractures were around cancellous
screws of the proximal femur. Distribution of cases accord-
ing to original implant and location is shown in Fig. 2.
The average time from primary fracture fixation to peri-
implant fracture was 6.2 years (range 0–30 years). There was
trend toward fractures around nails occurring earlier than
those around plates (3.0 versus 7.6 years, p = 0.14).
Injury patterns
Based on our observations, we classified NPPIFs accord-
ing to the type of implant (nail or plate), the position of the
fracture relative to that implant (type 1—at the tip, type 2—
distant to the implant) and the healing status of the original
fracture (A—healed, B—not healed, C—failing). The pro-
posed classification is shown in Fig. 3 and illustrated with
case examples in Fig. 4.
65%(39/60) of NPPIFs were directly adjacent to the tip
of the implant (type 1). 33%(20/60) were distant to the tip
of the implant (type 2). One patient had an unusual injury
that was not classified by our system (no. 44). This patient
initially had a long cephalomedulary femur nail that was
inserted without distal locking screws and subsequently sus-
tained a spiral fracture around the shaft of the nail.
79%(46/58) of primary fractures were healed (subtype
A). 17%(10/58) were not healed but stable (subtype B) and
in 3% (2/58) the fracture fixation was failing (subtype C).
Management
Of 60 fractures, 57 were treated surgically. Three fractures
in two patients (no. 44 and 53) were treated conservatively.
One patient (no. 45) underwent minimal surgery with
removal of prominent screws without internal fixation due
to comorbidities.
Of the ten patients with intramedullary nails treated surgi-
cally, four had retention of the original nail and extramed-
ullary fixation around the nail, five had the original nail
removed and were fixed with either an extramedullary
implant or a longer nail. One had removal of reconstruction
screws alone.
 Table 1  Summary of cases
Case no Age Sex Side Bone Mechanism Time to Original Originala NPPIF type Subtype Implant New implant Follow up Complications
NPPIF implant removed? (months)
(months)

1 39 M L Femur RTA​ 122 Plate Femur shaft P2 A Y Cancellous 35 N

screws
L Femur 122 P1 A Y LISS plate— 35 N
distal femur
2 39 M R Femur RTA​ 91 IM nail Femur shaft N1 A Y Cancellous 75 N
screws
R Femur 91 N1 A Y LISS plate— 75 Non-union—
distal femur ICBG 10/1/11
Archives of Orthopaedic and Trauma Surgery

3 64 F R Femur Low energy 12 DHS ITa P2 A N LISS plate— 3 Died at 3

fall distal femur months—
pneumonia
4 81 M L Femur Low energy 57 Long gamma Subtrochan- N1 A Y LISS plates— 5 N
fall nail terica distal femur
5 89 F R Femur Low energy 34 DHS ITa P2 A N LISS plates— 5 DVT
fall distal femur
6 82 F L Femur Low energy 165 Distal femur Supracondylara P2 A Y Long PFNA 8 N
fall plate/TKR
7 77 F R Femur Low energy 75 DHS ITa P1 A Y Antegrade 36 N
fall femoral nail
8 88 F L Femur Low energy 113 DHS ITa P1 A Y Antegrade 24 N
fall femoral nail
9 28 M L Femur Low energy 2 LCDCP 10 Femur shaft P1 B Y Antegrade 10 Non-union:
fall hole femoral nail dynamised
10 41 M R Femur RTA​ 49 IM nail Femur shaft N1 B N Zimmer vari- 14 N
able angle
distal femur
plate
11 61 M R Femur Low energy 93 Distal femur Supracondylara P1 A Y LCP 16 N
fall plate
12 59 M R Femur Low energy 103 DCS ITa with ST P1 A Y Long DCS 6 N
fall extension/
femur shaft
fracture
13 90 M L Femur Low energy 243 Smith petersen ITa P2 A N LISS plates— 8 Refracture and
fall hip screw distal femur reoperation.
Both united at
8 months
14 84 M L Femur Low energy 31 DHSs—long IT fracture ? P2 A N Distal plate 12 N
fall Shaft fracture

13


Table 1  (continued)
Case no Age Sex Side Bone Mechanism Time to Original Originala NPPIF type Subtype Implant New implant Follow up Complications
NPPIF implant removed? (months)

13
(months)

15 43 M L Femur Low energy 17 LCDCP Femur shaft P2 A Y Long DHS 14 Peri-implant

fall with greater fracture twice
trochanter with reopera-
screw tions
16 100 F L Femur Low energy 17 CM nails— ITa N2 A N Zimmer vari- 6 N
fall short able angle
distal femur
plate
17 75 M R Humerus Low energy 90 LCP Humerus shaft P1 A Y Antegrade nail 3 N
fall
18 42 M L Tibia RTA​ 54 Proximal tibia Tibial plateau P2 A N Distal tibial 12 N
double plate locking plate
19 64 F R Femur Low energy 23 LISS plate Supracondylar P1 B N Antegrade nail 32 N
fall above TKR
20 86 F L Femur Low energy 3 Antegrade nail Femur shaft N1 B N LISS plates— 18 N
fall distal femur
21 77 F R femur Low energy 12 CM nails— ITa N1 B N LISS plates— 27 N
fall long distal femur
22 79 F L Femur Low energy 2 CM nails— ITa N1 C Y Reverse LISS 20 N
fall short plate
23 31 M R Femur Intra-op frac- 0 CM nail -short ITa N1 B Y Long PFNA 5 DVT
ture
24 33 M R Humerus RTA​ Unknown Plate Humerus shaft P1 A N Anterolateral 9 N
LCDCP
25 37 M R Humerus RTA​ 61 Narrow Humerus shaft P1 A Y Narrow 4.5 4 N
LCDCP LCDCP
26 73 M L Radius RTA​ 365 Old plate Radius shaft P1 A N 3.5 LCP 3 N
27 70 F L Radius Low energy 58 DR T plate DRa P2 A Y 3.5 LCDCP 22 N
fall
28 57 M L Radius Low energy 369 Partially Radius shaft P1 A N LCDCP 5 N
fall removed
plate
29 62 M R Femur Low energy 26 Cancellous NOFa S1 A Y DCS with cer- 36 N
fall screws clage wires
30 72 M R Tibia RTA​ 24 Tibial plate Tibia metaphy- P2 A N Lateral proxi- 3 N
seal fracture mal tibial
plate
Archives of Orthopaedic and Trauma Surgery
 Table 1  (continued)
Case no Age Sex Side Bone Mechanism Time to Original Originala NPPIF type Subtype Implant New implant Follow up Complications
NPPIF implant removed? (months)
(months)

31 47 F R Radius Low energy 68 LCDCP Radius shaft P1 A Y 3.5 LCDCP 84 Refractured

fall twice recquir-
ing revision
R Ulna Ulna shaft P1 A Y 3.5 LCDCP Refractured
twice recquir-
ing revision
L Radius Low energy 7 LCDCP Radius shaft P1 A Y 3.5 LCDCP 84 Refractured
fall twice recquir-
Archives of Orthopaedic and Trauma Surgery

ing revision
L Ulna Ulna shaft P1 A Y 3.5 LCDCP Refractured
twice recquir-
ing revision
32 74 F L Femur Low energy 13 DHS ITa P1 A N Retrograde 2 N
fall femoral nail
33 60 F L Femur Low energy 483 Old plate Femur shaft P1 A Y 4.5LCDCP 9 N
fall
34 20 M L Radius Low energy 2 DR locking DRa P1 A Y 2.4 mm DR 12 N
fall plate locking plate
35 78 F L Femur Low energy 122 DCSs—broken Distal femur P2 A N Hemiarthro- 4 N
fall fracture plastys—
Zimmer ML
taper
36 94 M L Femur Low energy 8 PFLPs—short ITa P2 A N Zimmer MDN 14 N
fall retrograde
nail
37 38 M R Tibia RTA​ 13 Low bend dis- Distal tib-fib P1 A Y Screw fixation 19 N
tal tib plate fracture
R Fibula 13 1/3 tubular Distal tib-fib P1 A Y LCDCP 19 N
fracture
38 53 M R Humerus Low energy 238 LCDCP Humerus shaft P1 B N LCDCP 5 N
fall
39 77 F L Femur Low energy 64 DHS ITa P2 A N Retrograde 13 N
fall femoral nail
40 95 M L Femur Low energy 1 Cancellous NOFa S1 A Y Cemented 3 DVT
fall screws bipolar with
cerclage
wires
41 38 F R Femur Low energy 142 Short retro- Supracondylar N1 A Y Long retro- 11 Refracture and
fall grade nail fracture grade nail reoperation

13


Table 1  (continued)
Case no Age Sex Side Bone Mechanism Time to Original Originala NPPIF type Subtype Implant New implant Follow up Complications
NPPIF implant removed? (months)

13
(months)

42 80 F R Femur Low energy 74 DHS ITa P2 A N Retrograde nail 10 N

fall
43 80 F L Femur Low energy 75 DHS ITa P1 A Y Long PFNA 8 N
fall
44 73 M R Femur Low energy 2 Long PFNA ITa NX B – Conservative Lost to follow N
fall up
45 62 M R Femur Low energy 21 Recon nail Femur shaft N1 A N Recon screws Lost to follow N
fall removed up
46 67 F L Femur Low energy 97 LISS plate Supracondylara P1 A N Long PFNA 2 N
fall
47 25 F R Ulna Low energy 6 LCP ulna shaft P1 B – Conservative Lost to follow N
fall up
48 82 F R Femur Low energy 75 Cancellous NOFa S2 C Y Long stem 20 N
fall screws bipolar with
cerclage
wires
49 85 M R Femur RTA​ 53 DCS ITa P2 A N LISS plate 15 N
50 35 M R Clavicle RTA—bicycle 22 3.5 mm locking Distal clavicle P1 A N Precontoured 48 N
recon shaft locking plate
51 27 M R Radius RTA​ 69 3.5 mm Radius shaft P1 A N 3.5 mm 6 N
LCDCP LCDCP
52 71 M R Humerus Low energy 161 Lateral 1/3 Supracondylara P1 A Y 3.5 mm distal 1 N
fall tubular plate humerus LCP
53 39 M R Radius Low energy 16 3.5 mm Radius shaft P2 A – Conservative 12 N
fall LCDCP
R Ulna 16 3.5 mm Ulna shaft P1 B – Conservative 12 Non-union after
LCDCP conservative
management-
plated

RTA​road traffic accident, IM intramedullary, CM cephalomedullary, DR distal radius, DHS dynamic hip screw, DCS dynamic condylar screw, LCP locking compression plate, LCDCP low con-
tact dynamic compression plate, LISS less invasive stabilization system (Synthes), PFNA proximal femoral nail antirotation (Synthes), IT intertrochanteric, NOF neck of femur, A2FN antegrade
femoral nail (Synthes)
a
 Fracture
Archives of Orthopaedic and Trauma Surgery
Archives of Orthopaedic and Trauma Surgery
Fig. 1  Demographics and distribution of fractures
Fig. 2  Distribution of cases according to original implants and loca-
tion
There were 40 patients with extramedullary plates and/
or screws treated surgically. These patients had 45 fractures
in 44 bones. 24 had removal of the orginal implants and 20
had retention of implants. The 45 fractures were fixed with
intramedullary nails in 12 patients, extramedullary implants
in 28, screws alone in 2 and hemiarthroplasty in 3 cases.
The management strategies employed for femoral NPPIFs
are shown in Table 2. Some homogeneity in management
strategy was observed in certain fracture patterns (N1 frac-
tures below long antegrade nails were all treated with distal
femur locking plates) while most patterns had multiple sur-
gical treatment options.
Complications
20% (11/53) of patients experienced complications. By
region 9/39 lower limb patients had complications compared
to 2/14 upper limb patients. A summary of complications is
shown in Table 3.
Four patients had further NPPIFs in the same bone. One
patient (case 15) refractured twice in different places in the
same femur and one patient (case 32) refractured bilateral
forearm bones on two further occasions (eight additional
fractures). There were, therefore, 12 further NPPIFs from
61 original NPPIFs.
Three patients had non-union requiring further surgery
including refixation and bone grafting. All patients went on
to successful union.
13
 Archives of Orthopaedic and Trauma Surgery

Fig. 3  Fractures were classified

according to the type of implant
(nail or plate), the position of
the new fracture (type 1—at
the tip, type 2—distant to the
implant) and the healing status
of the original fracture

Fig. 4  a Type P1A fracture—

well healed original fracture
with NPPIF at the tip of the
implant, b type P2A fracture—
well healed original fracture
with NPPIF distant to implant,
c type N1C fracture—NPPIF at
the tip of the implant with the
original fracture failing

Overall 11% (6/53) of patients experienced surgical com-
plications. All these patients required further surgery.
Discussion
NPPIFs are an under-reported entity. The orthopaedic litera-
ture has focused on fractures of the proximal femur. In the
early 1990s cephalo-medullary (CM) nails were introduced
for fixation of peri-trochanteric fractures; CM nails had
theoretical biomechanical advantages over extramedullary
implants such as the sliding hip screw (SHS), unfortunately,
these advantages were offset by a high rate of peri-implant
fracture at the distal end of the nail. The 1st generation
Gamma Nail (Howmedica Ltd) suffered from early peri-
implant fracture rates of up to 17% [6]. Subsequent improve-
ments in nail design markedly reduced the fracture rate; the
most recent data suggest a rate of 0.7–3.5% for CM nails
and 0.3–1% for the SHS [7–15]. Add muller et al., Frisch.
The literature regarding non femoral NPPIFs is extremely
sparce [16]. Studies reporting upper limb NPPIFs are
focused on the effect of removal of implants as this appears
to be associated with a significantly higher risk of re-fracture
compared to if they are retained [17].
This paper represents the largest series of NPPIFs in the
literature and we hope that the following observations and

13
Archives of Orthopaedic and Trauma Surgery

Table 2  Common management strategies for femoral fractures according to classification

Original implant Injury pattern (no.) Management strategy No.

Antegrade femur nail (long) N1 (4) Distal femur locking plate 4

Antegrade femur nail (short) N1 (2) Removal of Implant (ROI) and long antegrade nail 1
ROI and long plate 1
N2 (1) Distal femur locking plate 1
Proximal femur plate (DHS/DCS/PFLP) P1 (5) ROI and antegrade nail 3
Retention of plate and retrograde nail 1
Revision to extended plate 1
P2 (8) Retention of plate and distal femur plate 5
Retention of plate and retrograde nail 3
Shaft plate P1 (3) ROI and replated 2
ROI and nailed 1
P2 (2) ROI and replated 2
Distal femur plate P1 (3) Antegrade nail 2
ROI and replated 1
P2 (2) ROI and antegrade nail 1
Hemiarthroplasty 1

Table 3  Non surgical and surgical complications Patterns of injury and classification rational

Type of complica- Complication No. of patients Rate (%)
tion We attempted to construct a classification system based
upon our observations of patterns of injury and treatment
Non surgical DVT 3 6
strategies. Classifying NPPIFs is challenging because of the
Pneumonia/death 1 2
high degree of heterogeneity. The closest analogue to the
Non-union 1 –
(conservative
NPPIF is the peri-prosthetic fracture, for which the most
treatment) well known classification is the Vancouver system [4]. The
Surgical Refracture (NPPIF) 4 (12 fractures) 8 (20) system classifies proximal femur PPFs into groups (A) tro-
Non-union 2 4 chanteric avulsion fractures; (B) fractures around the shaft
of the implant; and (C) fractures distant to the tip of the
implant. Type B fractures are sub-classified according to
proposed classification system will stimulate further interest whether the implant is well fixed (B1), loose (B2) or with
in this topic. poor bone stock (B3). The more recently developed unified
classification system (UCS) expands upon the Vancouver
Demographics classification to include other anatomical locations and other
permutations such as fractures in one bone supporting two
Femur fractures represent the majority (63%) of our cases. replacements (D), or in two bones supporting one replace-
We attribute this predominance to a combination of demo- ment (E), or in the non-replaced joint surface, facing and
graphic, anatomic and biomechanical factors. Proximal articulating with an implant (F) [5].
femur fractures typically affect elderly osteoporotic patients PPFs occur around prosthetic joints that are implanted at
and the femur itself is susceptible to NPPIFs due to its per- fixed locations with few variations in design. As reflected in
culiar shape and eccentric loading. the Vancouver classification, the variables affecting manage-
We observed a bimodal age distribution. The majority ment are the location of the new fracture, stability of the
of cases were in elderly patients with low velocity injuries, original prosthesis and the adequacy of bone stock. In com-
however, there was a smaller subset of younger patients parison, NPPIFs are not restricted to fixed anatomical sites,
involved in high-velocity motor vehicle accidents. The pat- the configuration and healing status of the original fracture
terns of injury were different with the latter marked by more is variable and there are multiple types and configurations
complex segmental fracture patterns and association with of osteosynthesis implant.
polytrauma. Treatment strategies and clinical outcomes When considering these factors, there are clearly a huge
are likely to be different between these two groups and it is number of permutations and a fully comprehensive classi-
important to recognize this in future research. fication becomes unwieldy and unnecessary for everyday

13

use. For this reason, we did not aim to take into account all
possible variables but chose the most important variables
affecting treatment. In a given location, these were the type
of implant (intramedullary or extramedullary), the position
of the NPPIF relative to that implant (at the tip or distant)
and the status of the original fracture (healed, healing and
failing construct). We were aware that some fracture patterns
would not be classified by our system, however, we deemed
that these were either infrequently encountered fracture pat-
terns or fractures for which the management is relatively
straightforward (for example, fractures at the apophysis and
fractures around screws alone).
We made the following observations regarding the factors
in the classification.
Type of implant
The majority of NPPIFs were around plates and screws (48
fractures) with fractures around nails forming a smaller
group (13 fractures). Inferences regarding relative inci-
dence are limited as the prevalence of extramedullary ver-
sus intramedullary implants in the community is unknown.
Short and long cephalomedullary nails and medullary nails
were similarly represented in our series; again, we are unable
to draw firm conclusions regarding relative incidences as
the prevalence of long versus short CM nails is unknown,
however, in keeping with the existing literature there does
not appear to be an excess of fractures in either group [15,
18, 19].
The type of original implant (intramedullary/extramed-
ullary) affects the subsequent placement of implants. If the
implant is retained, other devices must be positioned around
it. If the implant is removed, bone defects and areas of
potential weakness must be taken into account and bridged
if necessary. In either case, areas of high stress concentra-
tion between implants or at screw holes must be avoided.
This can be achieved by implants that bridge all areas of
potential weakness or overlapping implants. Studies based
on peri-prosthetic fracture models have demonstrated the
biomechanical superiority of such strategies [20].
We did not classify fractures around screws alone as the
management of these injuries is usually straightforward.
Position of the NPPIF
Fractures at the tip and distant to the tip of the implant differ
in terms of the mechanism of fracture as well as the treat-
ment options available. Mechanically, in type 1 fractures, the
implant has acted as a stress riser and predisposed the bone
to further fracture. In type 2, the implant has not directly
contributed to the fracture, which has occurred through
a distant area of weakness (usually the far metaphysis).
In terms of treatment, a type 1 fracture may be treated by
13
Archives of Orthopaedic and Trauma Surgery
changing the implant to a longer implant, such as exchang-
ing a short proximal femoral nail for a long nail that bridges
the original fracture zone and the new NPPIF. This would
not usually be possible with a type 2 fracture, which is more
likely to require a second implant. In such a case, a decision
has to be made regarding whether the implants should over-
lap or if the distance between implants is of sufficient length
to avoid significant stress concentration.
Healing status of the original fracture
The status of the original fracture is the third determining
factor. The type A NPPIF with a healed original fracture may
be amenable to removal of the original implant, however, the
quality of the healing and the effect of empty screws holes
must be taken into account. Type B with a stable but not
healed original fracture usually steers management toward
retention of the original implant. Type C is the equivalent
of a double level fracture requiring revision of the original
fixation in addition to fixation of the NPPIF.
Management strategies
Surgical management strategies differed according to loca-
tion. We used the classification to identify management
strategies that were applied to the common fracture types in
the femur (see Table 2). Figure 5 shows the use of the classi-
fication in developing a management algorithm for proximal
femur-based implants. Similar algorithms can be developed
for other locations (e.g., fractures around distal femur-based
implants or humerus shaft implants).
Cases in our series were managed using standard osteo-
synthesis implants including locking plates, intramedullary
nails and cerclage wires. Two patients were treated with
variable angle locking plates. No patients had ‘piggyback’
type plates or specialized cable plate systems.
Complications
We noted a relatively high incidence of complications such
as non-union and further occurrence of an NPPIF. Higher
complication rates may be anticipated in the setting of revi-
sion surgery through scar tissue, requiring more extensive
dissection and soft tissue stripping. Securing stable fixa-
tion can be challenging because of the position of existing
implants, screws holes and stress risers. Loss of bone stock
from stress shielding and disuse osteopenia may also con-
tribute to the surgical difficulty. While not all factors are
modifiable, there were technical errors that increased the
risk of fracture in three of the four cases that went on to have
further NPPIFs. These included leaving a short unprotected
segment of bone between implants or ending an implant near
a previous screw hole (see Fig. 6).
Archives of Orthopaedic and Trauma Surgery

uncommon (< 1%) [7–10], however, NPPIFs after a previous

NPPIF occurred at a rate of 20% in our series.

Limitations

This is a retrospective series of non-consecutive cases con-

tributed by a number of surgeons. Inferences regarding inci-
dence or prevalence of NPPIFs are limited. The management
of patients was at the discretion of the surgeon in charge
and we were unable to compare the outcomes of different
operative strategies.

Areas for further research

NPPIFs are uncommon injuries, however, with the increas-

ing use of osteosynthesis and an aging population they are
certain to become an increasingly important clinical prob-
lem. Further studies are required to assess the outcomes of
different management strategies for various fracture types
and determine the parameters for what is acceptable fixa-
tion in the context of a NPPIF. By defining principles and
optimal management strategies we can hope to improve out-
comes and reduce complications.

Fig. 5  Management algorithm for femur fractures with proximally Conclusions

based implants. Suggested treatment options are based upon the
type of implant (N, nail; P, plate), position of NPPIF (1, at the tip
of implant; 2, distant to the implant) and status of the fracture (A,
NPPIF are an interesting yet under-researched entity. Surgi-
healed; B, not healed; C, fixation failing) cal management is challenging, however, adherence to sound
treatment principles can help to avoid failure. Further study
is required to determine the optimal treatment for different
Of note, one female patient had bilateral forearm NPPIFs subtypes. The number of such fractures will undoubtedly
that were complicated by two further occurrences of NPPIF increase and present an important clinical challenge.
(eight fractures). She was investigated for metabolic bone
disease but no underlying cause was found. Another patient
had two further NPPIFs of the femur. The cause of these
recurrent events is likely to be multifactorial, however, it
appears that patients with NPPIFs are at high risk of fur-
ther events. NPPIFs following primary fracture fixation are

Fig. 6  Type P2A fracture below

a proximal femur plate. This
was fixed with a distal femur
locking plate. A short unbridged
segment became an area of
stress concentration result-
ing in a new NPPIF. This was
fixed with an anterior plate and
cerclage wires

13

Acknowledgements  This manuscript was prepared by the Singapore
Orthopaedic Research CollaborativE (SORCE), c/o Lester W. M. Chan,
MBBS, FRCS(Ed) and Ernest B. K. Kwek MBBS, MRCS(Edin),
M.Med(Ortho), FAMS, FRCS(Edin)(Ortho). Principal Investiga-
tor: Lester W. M. Chan. Manuscript Preparation: Lester W. M. Chan,
Antony W. Gardner, Ernest B. K. Kwek, Merng Koon Wong, Kenon
Chua. Case contributors: Lester W. M. Chan (Tan Tock Seng Hospital/
Khoo Teck Puat Hospital), Antony W. Gardner (Ng Teng Fong Gen-
eral Hospital), Ernest B. K. Kwek (Tan Tock Seng Hospital), Merng
Koon Wong (Singapore General Hospital), Fareed Kagda (Ng Teng
Fong General Hospital/National University Hospital), Diarmuid Mur-
phy (National University Hospital), Kein Boon Poon (Changi General
Hospital).
Compliance with ethical standards  12. Matre K, Havelin L, Gjertsen J, Vinje T, Espehaug B, Fevang J
Funding  The authors have no sources of funding to declare.
Conflict of interest  All authors declare that they have no conflict of
interest.
Ethical approval  For this type of study formal consent is not required.
This article does not contain any studies with human participants or
animals performed by any of the authors.
Informed consent  Not required.
References
1. Ruchholtz S, El-Zayat B, Kreslo D, Bücking B, Lewan U, Krüger
A et al (2013) Less invasive polyaxial locking plate fixation in
periprosthetic and peri-implant fractures of the femur—a prospec-
tive study of 41 patients. Injury 44(2):239–248
2. Della R, Leung K, Pape H (2011) Periprosthetic fractures: epide-
miology and future projections. J Orthop Trauma 2:S66–S70
3. Giannoudis PV, Kanakaris NK, Tsiridis E (2007) Principles of
internal fixation and selection of implants for periprosthetic femo-
ral fractures. Injury 38(6):669–687
4. Duncan CP, Masri BA (1995) Fractures of the femur after hip
replacement. Instr Course Lect 44:293–304
5. Duncan CP, Haddad FS (2014) The Unified Classification System
(UCS): improving our understanding of periprosthetic fractures.
Bone Jt J 96-B(6):713–716
6. Butt M, Krikler S, Nafie S, Ali M (1995) Comparison of dynamic
hip screw and gamma nail: a prospective, randomized, controlled
trial. Injury 26(9):615–618
7. Norris R, Bhattacharjee D, Parker M (2012) Occurrence of sec-
ondary fracture around intramedullary nails used for trochan-
teric hip fractures: a systematic review of 13,568 patients. Injury
43(6):706–711
13
Archives of Orthopaedic and Trauma Surgery
8. Bhandari M, Schemitsch E, Jönsson A, Zlowodzki M, Haiduke-
wych G (2009) Gamma nails revisited: gamma nails versus com-
pression hip screws in the management of intertrochanteric frac-
tures of the hip: a meta-analysis. J Orthop Trauma 23(6):460–464
9. Yli-Kyyny T, Sund R, Juntunen M, Salo J, Kröger H (2012) Extra-
and intramedullary implants for the treatment of pertrochanteric
fractures—results from a Finnish National Database Study of
14,915 patients. Injury 43(12):2156–2160
10. Chirodian N, Arch B, Parker M (2005) Sliding hip screw fixation
of trochanteric hip fractures: outcome of 1024 procedures. Injury
36(6):793–800
11. Matre K, Havelin L, Gjertsen J, Espehaug B, Fevang J (2013)
Intramedullary nails result in more reoperations than sliding hip
screws in two-part intertrochanteric fractures. Clin Orthop Relat
Res 471(4):1379–1386
(2013) Sliding hip screw versus IM nail in reverse oblique tro-
chanteric and subtrochanteric fractures. A study of 2716 patients
in the Norwegian Hip Fracture Register. Injury 44(6):735–742
13. Liu M, Yang Z, Pei F, Huang F, Chen S, Xiang Z (2010) A meta-
analysis of the Gamma nail and dynamic hip screw in treating
peritrochanteric fractures. Int Orthop 34(3):323–328
14. Müller F, Galler M, Zellner M, Bäuml C, Marzouk A, Fücht-
meier B (2016) Peri-implant femoral fractures: The risk is more
than three times higher within PFN compared with DHS. Injury
47(10):2189–2194
15. Dunn J, Kusnezov N, Bader J, Waterman BR, Orr J, Belmont PJ
(2016) Long versus short cephalomedullary nail for trochanteric
femur fractures (OTA 31-A1, A2 and A3): a systematic review. J
Orthop Traumatol 17(4):361–367
16. Reyes-Cabrera J, González-Alconada R, García-Mota M (2013)
[Distal tibia peri-implant fracture with an intramedullary nail: a
case report]. Rev Esp Cir Ortop Traumatol 57(6):443–445
17. Yao C, Lin K, Tarng Y, Chang W, Renn J (2014) Removal of
forearm plate leads to a high risk of refracture: decision regard-
ing implant removal after fixation of the forearm and analy-
sis of risk factors of refracture. Arch Orthop Trauma Surg
134(12):1691–1697
18. Boone C, Carlberg K, Koueiter D, Baker K, Sadowski J, Wiater P
et al (2014) Short versus long intramedullary nails for treatment of
intertrochanteric femur fractures (OTA 31-A1 and A2). J Orthop
Trauma 28(5):e96–e100
19. Kleweno C, Morgan J, Redshaw J, Harris M, Rodriguez E,
Zurakowski D, Vrahas M, Appleton P (2014) Short versus long
cephalomedullary nails for the treatment of intertrochanteric
hip fractures in patients older than 65 years. J Orthop Trauma
28(7):391–397
20. Kubiak EN, Haller JM, Kemper DD, Presson AP, Higgins TF,
Horwitz DS (2015) Does the lateral plate need to overlap the
stem to mitigate stress concentration when treating Vancouver
C periprosthetic supracondylar femur fracture? J Arthroplasty
30(1):104–108