Received: 18 October 2022 Revised: 18 February 2023 Accepted: 2 April 2023

DOI: 10.1111/vsu.13963

EXPERIMENTAL LAB RESEARCH

3D-printed, patient-specific cutting guides improve femoral

and tibial cut alignment in canine total knee replacement

Agnieszka B. Fracka MSc, DVM 1 | Bill Oxley VetMB, DSAS(Orth) 2 |

Matthew J. Allen VetMB, PhD 1

1
Department of Veterinary Medicine,
University of Cambridge, Cambridge, UK Abstract
2
Vet3D Ltd., Kendal, UK Objectives: The purpose of this cadaveric study was to determine whether
patient-specific guides (PSGs) improve the accuracy of tibial and femoral cut
Correspondence
alignment in canine total knee replacement (TKR), as compared with generic
Matthew J. Allen, Department of
Veterinary Medicine, University of cutting guides.
Cambridge, Madingley Road, Cambridge Study design: Original research.
CB3 0ES UK.
Email: [email protected]
Animals: Sixteen pelvic limbs from skeletally mature medium- to large-breed
canine cadavers.
Methods: Specimens were randomly allocated to one of two groups (PSG or
Generic; N = 8/group). In the Generic group, femoral and tibial ostectomies
were made using the standard canine TKR femoral cutting blocks and tibial
alignment guide. In the PSG group, the cuts were made using a series of cus-
tom 3D-printed cutting guides. “Planned” and “actual” tibial and femoral cut
alignments were compared in the frontal and sagittal planes, and errors were
calculated by subtracting actual from planned values.
Results: Use of 3D-printed PSGs improved tibial cut alignment in the frontal
plane but not the sagittal plane. PSGs also improved the alignment of the cra-
nial and distal femoral ostectomies but did not impact varus-valgus alignment.
Conclusions: These findings support the use of PSGs for TKR in dogs. Clinical
trials are now needed to determine whether the benefits of PSGs translate into
measurable improvements in joint function and implant longevity.
Clinical significance: PSGs have the potential to improve femoral and tibial
component alignment in canine TKR.

1 | INTRODUCTION clinical success in TKR depends on optimal implant

positioning and alignment to ensure appropriate bone
Total knee replacement (TKR) is recognized as an effec- ingrowth–bone ongrowth and load transfer. Conven-
tive treatment option for dogs with end-stage stifle dis- tional planning for canine TKR is based on plain radio-
ease.1 As in total hip replacement (THR), long-term graphs and the surgery is performed with the aid of

Abbreviations: 3D, 3-dimensional; CT, computed tomography; ETAG, extramedullary tibial alignment guide; FCB, femoral cutting block; PMMA,
polymethyl methacrylate; PSG, patient-specific guide; STL, Standard Tessellation Language; THR, total hip replacement; TKR, total knee
replacement.

Presented in part at the American College of Veterinary Surgeons Congress; October 12-15, 2022; Portland, Oregon.

674 © 2023 American College of Veterinary Surgeons. wileyonlinelibrary.com/journal/vsu Veterinary Surgery. 2023;52:674–685.
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FRACKA ET AL. 675

generic cutting blocks/guides. This is the same approach euthanatized for reasons unrelated to this study. The ani-
as is used in human TKR, but questions remain as to mals had been donated for use in teaching and research,
how accurate these generic TKR instruments are,2 partic- with the informed consent of the owners. The age and
ularly with respect to accurate alignment with the bodyweight of the cadavers were not recorded; however,
mechanical axes of the femur and tibia.3 It is widely rec- they were all above 20 kg (the minimum bodyweight for a
ognized in human TKR that component alignment is a TKR candidate at our institution) to be included in this
key determinant of long-term implant survival,4 and it study. The hindlimbs were disarticulated at the hip joint
has been reported that up to 22% of primary TKR cases and stored in sealed bags at 20 C until the study was con-
performed using standard instrumentation have axial ducted. Specimens were thawed for at least 24 h and
malalignment of greater than 3 degrees.5 Generic instru- allowed to equilibrate to room temperature prior to being
mentation is also more challenging to use in knees with used in the experiment. The specimens were randomly
deformities.5 allocated to one of two groups (PSG or Generic; N = 8/
There is growing interest in the use of patient-specific group).
guides (PSGs), designed from computed tomography
(CT) scans of the patient's anatomy. Proposed advantages
of PSGs include greater accuracy in the bone cuts, 2.1.1 | Preoperative planning and
reduced operative time and the ability to account for measurement of femoral and tibial cut
unique pathologies (especially deformity) that can com- alignment: Generic group
plicate the use of generic cutting guides.3,6 PSGs are
becoming more common in human orthopedics, includ- Orthogonal radiographs (full-length mediolateral and
ing TKR surgery. PSGs simplify pin placement and the caudocranial stifle series, plus a full-length caudocranial
creation of individualized bone cuts that correspond to view of the femur) were obtained to confirm skeletal
the patient's unique mechanical axis, in contrast with maturity, rule out the presence of pre-existing orthopedic
generic cutting guides that rely on average values for disease, and allow for surgical templating. A 25-mm cali-
determining the mechanical axes of the femur and tibia. bration ball was included in all radiographs for presurgi-
On the tibial side, PSGs can also avoid the need for place- cal templating. Femoral and tibial components of an
ment of an intramedullary alignment rod, making the appropriate size were positioned over caudocranial and
approach less invasive for the patient. mediolateral radiographs (Figure 1A,B). All radiographic
In veterinary medicine, the use of PSGs has garnered measurements were made with a commercial radio-
significant attention in the last few years, with potential graphic planning suite (vPOPPRO; Llangollen, UK) and
application in the correction of antebrachial growth measurements were performed by the same investigator
deformities,7–9 the drilling of transcondylar screws in to eliminate interobserver variability. The individual
humeral fractures and humeral intracondylar fissures,10 identities of specimens within the groups were coded but
atlanto-axial stabilization,11,12 pedicle screw placement,13 it was not possible to blind the observer to group alloca-
and femoral osteotomy.14 There have been isolated tion since the PSG measurements were made on CT
reports on the use of PSGs with custom canine and feline rather than plain radiographs (see later).
TKR implants,15,16 but their utility and accuracy have not Tibial cut alignment in the frontal plane was mea-
been evaluated in a systematic manner to date. We sured on caudocranial radiographs as the angle formed
hypothesized that PSGs would improve the surgeon's between the fixation surface of the tibial component and
accuracy in delivering the planned implant alignment in the mechanical axis of the tibia (formed by a line con-
TKR surgery. The objective of this canine cadaveric study necting the midpoint of the distal tibia and the intercon-
was therefore to determine whether the use of PSGs in dylar eminences on the proximal tibia) (Figure 1C). The
canine TKR increases the accuracy of femoral and tibial target frontal plane alignment was at 90 degrees to the
cut alignment, as compared with results obtained with mechanical axis.1 Sagittal cut alignment was measured
the use of generic cutting guides. on mediolateral radiographs as the angle between the fix-
ation surface of the tibial component and the tibial axis,
formed by a line connecting the midpoint of the talus dis-
2 | MATERIALS AND METHODS tally and the tibial intercondylar eminences proximally
(Figure 1D). The target sagittal cut alignment was with
2.1 | Cadaveric specimens 6 degrees of proximal tibial caudal slope, as per published
guidelines for canine TKR.1
Pelvic limbs were harvested from 16 skeletally mature Frontal plane alignment of the distal femoral cut was
medium- to large- mixed breed dogs that were determined as the angle formed between a line drawn
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
676 FRACKA ET AL.

F I G U R E 1 Preoperative
planning for the Generic guide
group, showing optimized
positioning of the femoral and
tibial components on
preoperative caudocranial
(A) and mediolateral
(B) radiographs in vPOPPRO
planning software. Frontal plane
alignment of the tibial cut was
determined on the caudocranial
view by reference to the
mechanical axis of the tibia
(C) and sagittal plane alignment
of tibial cut was determined
from the mediolateral view as
the angle between the relevant
fixation surface and the axis of
the bone (D). The angles that
were measured are indicated
with orange semi-circles.

along the distal articular surface of the femoral compo- and distal femoral cuts was determined by measuring the
nent (line 5 in Figure 2A), and a line representing the angle formed between the anatomical axis in the sagittal
anatomical axis of the femur (line 1 in Figure 2A). The plane (line 1 in Figure 2B, defined using the interactive
anatomical axis was determined with an interactive mea- tool, as above), and lines drawn along the inside of the
suring tool (“anatomic axis”) within the vPOPPRO suite, relevant fixation surface of the femoral component (lines
which connects the midpoints of the femur at 25 and 50% 2 and 3 in Figure 2B). The third and final sagittal plane
of femoral length. Sagittal plane alignment of the cranial measurement was the closing angle for the femoral
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FRACKA ET AL. 677

F I G U R E 3 For the patient-specific guides (PSG) group,

preoperative planning was performed on 3D models based on
computed tomography (CT) scans. Femoral and tibial components
were virtually implanted in the models (A and B), and the
alignment of the components determined relative to the femoral
and tibial axes. For the femur, the alignment of the cranial cut
(C) and the distal cut (D) were measured, and the closing angle was
determined (E). The red circle indicates the angle that was
measured.

component, which reflects the angle formed between the

cranial fixation surface (line 2 in Figure 2B) and the cau-
dal fixation surface (line 4 in Figure 2B). The target for
the closing angle was 10 degrees since this is the internal
angle formed in the BioMedtrix femoral component to
ensure a locking fit onto the femur.

2.1.2 | Preoperative planning and

measurement of femoral and tibial cut
alignment: PSG group
F I G U R E 2 (A) Frontal plane alignment of the femoral CT scans were used to confirm skeletal maturity, screen
component was determined by measuring the angle formed
for stifle joint pathology, and for surgical planning and
between the anatomical axis of the femur (line 1, passing between
the design of custom PSGs. 3D computer models of the
best fit circles at 25 and 50% of femoral length) and the distal
bones were created in medical imaging software
surface of the femoral component (line 5). The angle is indicated
with a black semi-circle. (B) Sagittal plane alignment of the femoral
(MIMICS version 24.0; Materialise BV, Belgium) and
component was determined by measuring the angles formed exported as STL files into surgical planning software
between the anatomical axis of the femur (line 1, passing between (3-matic version 16.0; Materialise BV, Belgium). The opti-
best fit circles at 25 and 50% of femoral length) and the cranial and mal positions of the femoral and tibial components were
distal fixation surfaces of the femoral component (lines 2 and defined in three dimensions (Figure 3A,B) and their fixa-
3, respectively). The angles are indicated with black semi-circles. tion surfaces were used to define the cutting planes for
The locking angle of the femoral component was measured as the the femur (Figure 3C,D) and tibia. The alignment of
angle formed between lines 2 (cranial cut) and 4 (caudal cut). these cutting planes to the anatomical axis (femur)
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
678 FRACKA ET AL.

and the mechanical axis (tibia) was then determined in

both the frontal and sagittal planes. On the 3D models,
the anatomical axis of the femur was defined as a line
passing through the center of the femur at 25 and 50% of
the length of the bone. As in the Generic group, the
mechanical axis of the tibia was defined by a line con-
necting the midpoint of the distal tibia and the intercon-
dylar eminences on the proximal tibia. Alignment was
measured in the frontal and sagittal planes using the
same approach as described for the Generic group, lead-
ing to values for frontal plane alignment of the tibial cut
and the distal femoral cut, and sagittal plane alignment
of the tibial cut, and the cranial and distal femoral cuts.
The locking angle was measured as the angle between
the cranial and caudal femoral cuts (Figure 3E). All the
3D measurements were made using the angle measure-
ment feature of 3-matic software. Tibial and femoral
measurements were performed by the same investigator
to eliminate interobserver variability. The individual
identities of specimens were coded but it was not possible
to blind the observer to group allocation since CT, rather
than plain radiographs, were only used for preoperative
measurement in the PSG group.

F I G U R E 4 The use of patient-specific guides (PSGs) in canine

2.2 | Design of PSGs total knee replacement (TKR). (A) The initial guide is designed to
locate two reference pins for the subsequent guides that are used to
The cutting planes defined on the planning images were prepare the cranial cut, the chamfer cut (B), the distal cut, and the
merged with custom-designed baseplates that mirrored caudal cut (C). On the tibia, a single guide (D) is attached to the
the external contours of the femur or tibia, as appropri- proximal medial aspect of the tibia and used to resect the articular
surface.
ate. The result was a set of five femoral PSGs and one tib-
ial PSG (Figure 4). The definitive PSGs were designed
and manufactured from an autoclavable resin by stereo- block on the top, which guides the ostectomy of the tibial
lithography (Vet3D, Staveley, Cumbria, UK). The fit of plateau. The oscillating saw was placed on the top of the
each guide was verified using a 3D-printed plastic model cutting block and around 6 mm of proximal tibia (at the
of the femur and tibia to ensure accurate contouring, cut level of tibial plateau eminences) was removed. Any resid-
location and trajectory. ual irregularity in the bone surface was removed by careful
trimming with the oscillating saw or a pair of rongeurs.
Next, four femoral osteotomies were made using the
2.2.1 | Surgical procedure for the Generic selected femoral cutting block (FCB). Before cuts were
group made, the FCB was secured in position using three fixation
pins (cranial, medial, and lateral). The cuts were made
Surgery was performed through a medial parapatellar through the slots in the FCB, and they always followed the
approach, and the joint accessed through a medial arthrot- same sequence: cranial, caudal, distal then chamfer cut. A
omy. The patella was luxated laterally and the infrapatellar trial femoral component was applied to the distal femur
fat pad, cranial cruciate ligament, caudal cruciate ligament and its fit evaluated. A trial tibial component was placed on
and menisci were removed. Throughout the procedure care the cut surface of the tibia and the joint was reduced to
was taken to preserve medial and lateral collateral liga- allow for assessment of joint gap symmetry, joint laxity,
ments. The extramedullary tibial alignment guide (ETAG; range of motion, femoral/tibial component position and sti-
BioMedtrix, Whippany, New Jersey) was aligned along the fle joint tracking. The optimal position for the tibial compo-
tibial mechanical axis by eye, then secured in place with fix- nent was identified and a 7-mm hole was then drilled into
ation pins (BioMedtrix LLC, Whippany, New Jersey) in line the tibial bone bed using a drill guide screwed into the trial
with current clinical guidelines.1 The ETAG has a cutting tibial component. The cavity was enlarged with a #7 fluted
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FRACKA ET AL. 679

incision and arthrotomy. Following stifle joint exposure, the

first femoral PSG was positioned against the femur and
secured with two bicoretical Ellis pins (2 mm in diameter);
these pins served as references for all the remaining PSGs
(Figure 5A). After the pins had been drilled, the first femo-
ral PSG was removed, and the cranial cutting guide was
placed over the reference pins. The cranial ostectomy was
performed using an oscillating saw fitted with a 0.89-mm
thick saw blade. The cranial guide was then removed, leav-
ing the pins in place, and the PSG for the chamfer cut was
seated over the pins (Figure 5B). Once the chamfer cut had
been performed, the guide was removed and replaced with
the guide for the distal cut, then the guide for the caudal
cut (Figure 5C). When all femoral cuts were complete, pins
were removed from the distal femur.
For the tibia, the PSG was placed on the craniomedial
aspect of the proximal tibia. A small hook at the distal
end of the guide was positioned so that it engaged with
the most distal part of the tibial crest (Figure 5D). The
guide was secured with two parallel 2.0-mm Ellis pins
into the tibial metaphysis, and the tibial articular surface
was resected with the oscillating saw. Any minor imper-
fections in the cut tibial surface were removed with either
the oscillating saw or a pair of rongeurs, as appropriate.
On completion of all cuts, the tibial and femoral compo-
nents were implanted in the same way as for the Generic
F I G U R E 5 Intraoperative photographs illustrating the use of group.
patient-specific guides (PSGs) in canine total knee replacement
(TKR). An indexing guide is used for reference pin placement (A),
after which there is sequential use of custom guides for the chamfer
cut (B) and the caudal cut (C). A separate tibial guide is applied to
2.4 | Postoperative assessment of tibial
the medial tibial surface (D). and femoral cut alignment

reamer to provide space for the tibial keel and surrounding Mediolateral and caudocranial radiographs were obtained
bone cement mantle. The trial components were then post-operatively for each specimen. The cemented tibial
removed, and the bone beds cleaned and flushed. Poly- component was left in place for radiography. Tibial cut
methyl methacrylate (PMMA) bone cement (Teknivet; Bio- alignment was determined in both the frontal plane and
Medtrix LLC, Whippany, New Jersey) was injected into the sagittal plane (Figure 6A,B), as described previously, using
tibial keel hole and applied to the undersurface of an appro- the ostectomized surface of the proximal tibia as the refer-
priately sized 3D-printed plastic tibial component that con- ence surface.
tained an internal wire marker to allow for visualization on One the femoral side, the femoral component was
postoperative radiographs. When the cement in the tibia removed from the bone prior to imaging to improve the
had cured, an appropriately sized cementless cobalt- visualization of the cut surfaces. Cut alignment was deter-
chromium femoral component was impacted on the distal mined using the same approach as described for preopera-
femur. The patella was reduced, and stifle joint was again tive measurement in the Generic group, resulting in
evaluated for laxity, implant stability and range of motion. angular measurements for the following cuts: frontal plane
The joint was closed routinely. alignment of the distal femoral cut (Figure 7A), and sagit-
tal plane alignment of the cranial cut (Figures 7B) and dis-
tal femoral cut (Figure 7C). The locking angle was
2.3 | Surgical procedure for the PSG measured (Figure 7D) and the fit of the femoral compo-
group nent on the cut distal femur was estimated by comparing
the cranial-caudal width of the cut femur against the inter-
The surgical approach in the PSG group was the same as nal cranial-caudal dimensions of the femoral component.
for the Generic group, with a medial parapatellar skin Tibial and femoral measurements were performed by the
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
680 FRACKA ET AL.

same investigator to eliminate interobserver variability. 2.5 | Data handling and statistical
The individual identities of specimens within the groups analysis
were coded to remove identifying information.
Errors in tibial and femoral cut alignment (in degrees)
were calculated by subtracting the measured (“actual”)
values obtained postoperatively from the values mea-
sured during the planning process (“planned”). Error
values were normally distributed and comparisons
between the two groups were made using an unpaired
Student's t-test with significance set at p < .05.
Fit of the femoral component was determined by
calculating the difference (in millimeters) between the
cranial-caudal width of the cut femur and the cranial-
caudal inner dimensions of the femoral component.
Data from the two groups were compared using an
unpaired Student's t-test, with significance set
at p < .05.

3 | RESULTS
F I G U R E 6 Measurement of postoperative tibial component
alignment was performed using the same approach as for the There were no significant technical problems with using
preoperative measurements. Frontal plane (varus-valgus) (A) and the PSGs. Use of the five sequential femoral guides was
sagittal plane (tibial slope) (B) were measured for each specimen. straightforward. As seen in Figure 5A–D, the guides fit
Angles are highlighted by black quarter-circles. very well onto the surface of the bone, allowing for a

F I G U R E 7 Measurement of
postoperative femoral alignment.
Frontal plane alignment was
determined relative to the
anatomical axis of the femur (A).
Sagittal plane measurements
included alignment of the cranial
ostectomy (B), distal ostectomy
(C) and the closing angle (D). Angles
are highlighted by white quarter-
circles.
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FRACKA ET AL. 681

F I G U R E 8 Tibial component alignment in canine total knee replacement (TKR) performed with generic or patient-specific guides
(PSGs). (A) The use of PSGs was associated with a statistically significant reduction in alignment error in the frontal plane (varus-valgus
alignment). (B) Sagittal plane alignment was similar in the two groups. The red line represents 3-degree threshold for axial malalignment.5
Numbers above bars represent p-values (ns, not significantly different).

precise and definitive fit. Their lower profile and trans- not associated with any difference in varus-valgus align-
parency allow greater visibility of the femur or tibia, at ment of the distal cut (3.02 ± 2.14 vs. 2.44 ± 1.67 for
least as compared with the bulkier generic guides. None the Generic and PSG groups, respectively; 95% CI: 2.65
of the guides loosened during the cutting procedure, to 1.47, p = .533) (Figure 9A). In contrast, sagittal plane
something that has been an issue in clinical cases oper- alignment of both the distal cut (5.93 ± 3.33 vs. 2.3
ated with the generic guides. ± 1.89 for the Generic and PSG groups, respectively;
95% CI 6.53 to 0.72; p = .018) (Figure 9B) and the
cranial cut (6.34 ± 4.53 vs. 2.5 ± 1.82 for the Generic
3.1 | Tibial component alignment and PSG groups, respectively; 95% CI: 7.539 to 0.1357,
p = .043) (Figure 9C) were significantly improved by the
Comparisons of the alignment data revealed a statistically use of PSGs. There was no statistically significant differ-
significant reduction in frontal plane alignment error in ence between the accuracy of closing angle cuts in the
the PSG group (2.41 ± 1.40 vs. 1.03 ± 0.96 ) for the two groups (3.75 ± 2.01 vs 3.59 ± 1.92 for the Generic
Generic and PSG groups, respectively; 95% CI: 2.674 to and PSG groups, respectively; 95% CI: 2.269 to 1.944)
0.1011 ( p = .036) (Figure 8A). In humans, optimal tib- (p = .871) (Figure 9D). The fit of the femoral component
ial component alignment is within 3 degrees of neutral;17 on the distal femur was also not different, with mean (±
as shown in Figure 8A, none of the PSG knees fell out- SD) under-sizing of the cranial-caudal femoral bone stock
side 3 degrees, compared with two of eight knees in the by 0.34 ± 0.38 mm for the PSG group and 0.4 ± 0.25 mm
Generic group. for the generic guides (p = .68).
There was no significant difference between PSG and
Generic groups for sagittal slope (1.60 ± 1.17 vs. 1.13
± 0.85 for the Generic and PSG groups, respectively; 4 | DISCUSSION
95% CI: 1.573 to 0.6229, p = .37) (Figure 8B).
The results from this study support the utility of PSGs for
improving the accuracy of sagittal alignment of the femo-
3.2 | Femoral component alignment ral cuts and frontal plane alignment of the tibial cut in
canine TKR. PSGs did not improve femoral alignment in
A total of four measurements were made to assess the the frontal plane when compared to generic cutting
accuracy of the femoral ostectomies. The use of PSGs was blocks, but this was in normal rather than pathological
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
682 FRACKA ET AL.

F I G U R E 9 Femoral
component alignment in canine
total knee replacement (TKR)
performed with generic or
patient-specific guides (PSGs).
(A) In the frontal plane, varus-
valgus alignment of the femoral
component was similar in the
two groups. In the sagittal plane,
statistically significant
improvements in femoral
component alignment were seen
for the distal cut (B) and the
cranial cut (C). The closing
angle, important for ensuring a
locking fit for the femoral
component, was similar in the
two groups (D). The numbers
above bars represent p-values
(ns, not significantly different).

joints. PSGs were as effective as generic guides at preserv- commercial veterinary navigation system on the market,
ing the cranial-caudal width and the 10-degree locking these benefits cannot be realized at present. However,
angle at the distal femur, both of which are likely to be most veterinary referral centers have easy access to CT
important determinants of early mechanical stability in a scanning, and the clinical success of PSGs in other areas
press-fit femoral component. Our findings are consistent of veterinary orthopedics, including spine surgery and
with those from similar studies in human TKR, which long bone deformity correction,6–14 provides a strong
also demonstrated improved alignment with the use of rationale for exploring their utility in canine TKR. In this
PSGs.3,5,18,19 study, all the knees operated with the assistance of PSGs
In human TKR, frontal plane alignment of the tibial fell within 3 degrees of neutral in the frontal plane.
component is a key determinant of long-term implant There is a learning curve for any surgical procedure,
survival.4 It has been suggested that restoring the and total joint replacement is no exception. In the human
mechanical axis to within 3 degrees of neutral in the field, it has been estimated that the learning curve for
frontal plane is associated with a better long-term THR is 50 cases, while for TKR surgeons need around
outcome,17,20 and this has driven interest in the use of 80 cases for technical proficiency.24 Careful preparation
assistive technologies such as surgical navigation21,22 and and attention to detail is crucial to avoid complications
PSGs to optimize tibial component alignment. We have and errors that may lead to ligament damage or instabil-
previously reported that surgical navigation improves tib- ity, loosening of the implant and pain.4,14,15 In our experi-
ial component alignment in canine TKR,23 but without a ence, the tibial and femoral ostectomies are the most
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FRACKA ET AL. 683

technically challenging aspects of TKR surgery, and inac- based on case requirements or even surgeon preference,
curate cutting of the articular surfaces may contribute to is facilitated by the custom nature of the PSGs.
errors in tibial and femoral component position- The use of PSGs is not without drawbacks. Placement
ing.3,5,17,25,26 By virtue of their use of the patient's own of the PSG does require additional dissection on the cra-
anatomy as the reference, PSGs are likely to be much nial and medial surfaces of the femur (proximal to the
better guides for aligning the tibial and femoral osteo- trochlea, to seat the initial femoral guide) and a slightly
tomies. From a practical point of view, our data indicate more extensive dissection over the distal part of the tibial
that PSGs use in TKR is effective, feasible and may be crest to allow the tibial guide to seat properly under the
particularly helpful for surgeons with limited prior expe- crest. CT scans are needed for surgical planning and
rience with TKR surgery. Tibial and femoral PSGs are guide design; these scans, and the costs associated with
easy and quick to apply, require less manipulation during manufacture of the PSGs, must then be factored into the
surgery than conventional cutting blocks, and allow bet- overall cost of the procedure. Additionally, CT scans do
ter visualization of the anatomy following placement. Pin not provide any information about soft tissue laxity or
loosening, which is recognized as a problem when balancing, both of which are key determinants of success
generic guides are used, was not see in any of the PSG in TKR surgery. Surgeons undertaking TKR must still be
cases from this study; this likely reflects the fact that the trained in the fundamental principles of flexion/
same two pins are required to resist the forces experi- extension gap balancing and management of periarticular
enced during all four of the femoral cuts. Additionally, as soft tissues. It may be advantageous to consider designing
more and more bone is resected from the femur, the bone more than one tibial PSG, in the event that the initial cut
support available to the guide decreases, with a concomi- is not deep enough to allow for equal tension in flexion
tant increase in load transfer through the pins them- and extension. More work, much of it feasible in
selves. The fact that each of the PSGs is used for a single cadavers, will be needed to define the best approach to
cut, with full bone support, may well make these pins optimizing joint laxity in clinical cases.
less prone to loosening. Additional clinical experience In general, data for tibial cut alignment fell within a
will be needed to confirm the significance of this clinically acceptable range of 3 degrees of error
observation. (Figure 8A,B), but more extreme values were noted for
For surgeons doing only a few TKR cases in a year, two of the 16 specimens (12%). In human studies, outliers
the use of PSGs offers a very cost-effective alternative to (>3 ) have been reported in 11.1% of patients undergoing
investing capital in the purchase of a complete set of conventional TKR, but in only 4% in patients undergoing
generic TKR instruments. Moreover, many of the dogs navigated TKR.28 In the present study, possible reasons
that present for TKR have end-stage disease and varying for the presence of outliers include errors in preoperative
degrees of bone deformity and joint collapse that can planning or, more likely suboptimal application of the
complicate the use of generic cutting guides.27 PSGs offer PSG to the bone surface, leading to inaccurate cut posi-
a very attractive solution for these cases; however, pro- tioning and/or trajectory.29 As experience with PSGs
spective clinical studies are now needed to better define increases, it should be possible to define the most appro-
the value of PSGs, and to determine the impact of better priate locations for positioning PSGs directly against the
alignment on clinical function and implant survival in bone surface. In a clinical setting, it will also be very
canine TKR. Additionally, further studies may give important to ensure that the positioning of PSGs in not
insights into potential design improvements of the PSGs hindered by the presence of thickened periarticular soft
which could further enhance their efficacy. Although the tissues (e.g., medial buttress) or marginal osteophytes.
accuracy of guides with very similar trochlear contact The custom instrumentation used in this study was
footprints and guide plane dimensions has been previ- designed specifically for these cadaveric specimens. STL
ously demonstrated for distal femoral osteotomy,13 design models of the definitive commercial femoral and tibial
modifications such as the use of combined guides for >1 components were used in the planning steps to ensure
osteotomy, or slotted guides, may prove beneficial. Slot- accurate alignment of the femoral and tibial cuts. The
ted guides are used commonly in human TKR PSGs but use of STLs of CAD files is recommended as inaccuracies
may have drawbacks in smaller veterinary patients, most in implant sizing could result in clinically significant
notably the reduced visibility of the saw blade in the errors in the positioning and alignment of the bone cuts,
vicinity of the collateral ligaments and constrained angle leading to suboptimal component alignment and/or
of attack of the blade. Additionally, for optimal accuracy, stability.
the thickness of the blade that will be used must be As with any cadaveric project, there were a number
planned for in every case. The ability to easily trial guide of limitations in this study. These included the relatively
design variations, and indeed to create different designs small number of specimens operated within each group,
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
684 FRACKA ET AL.

the use of healthy stifle joints as opposed to joints with 2. Stulberg S. How accurate is current TKR instrumentation? Clin
osteoarthritis or deformity, and the use of specimens for Orthop Relat Res. 2003;416:177-184.
which complete recording of bodyweight was not always 3. Laskin RS. Instrumentation pitfalls: you just can't go on autopi-
lot! J Arthroplasty. 2003;18(3 Suppl 1):18-22.
available. The preoperative measurements of planned cut
4. Ritter MA, Faris PM, Keating EM, Meding JB. Postoperative
alignment were made using CT in one group and plain alignment of total knee replacement. Its effect on survival. Clin
radiographs in the other group, which may introduce the Orthop Relat Res. 1994;299:153-156.
possibility for variation in our data. Despite these limita- 5. Ng VY, DeClaire JH, Berend KR, Gulick BC, Lombardi AV Jr.
tions, the results of the study are clear and compelling. Improved accuracy of alignment with patient-specific position-
Additional work is now needed to determine whether the ing guides compared with manual instrumentation in TKA.
advantages identified in this cadaveric study can be trans- Clin Orthop Relat Res. 2012;470:99-107.
6. Carwardine DR, Gosling MJ, Burton NJ, O'Malley FL,
lated into improved alignment in clinical TKR cases. In
Parsons KJ. Three-dimensional-printed patient-specific osteot-
addition, a prospective clinical trial will be required to
omy guides, repositioning guides and titanium plates for acute
see whether improvements in component alignment lead correction of antebrachial limb deformities in dogs. Vet Comp
to measurable and clinically impactful improvements in Orthop Traumatol. 2021;34:43-52.
overall implant function and/or survival. 7. Longo F, Penelas A, Gutbrod A, Pozzi A. Three-dimensional
computer-assisted corrective osteotomy with a patient-specific
surgical guide for an antebrachial limb deformity in two dogs.
5 | C ON C L U S I ON S Schweiz Arch Tierheilkd. 2019;161:473-479. English translation.
doi:10.17236/sat00216
8. Lee HR, Adam GO, Kim SJ. Application of patient-specific
3D-printed patient specific guides improve the accuracy instrumentation in a dog model with antebrachial growth
of tibial cut alignment and should be considered a viable deformity using a 3-D phantom bone model. Vet Sci. 2022;
option for surgeons undertaking canine TKR. 9:157.
9. Easter TG, Bilmont A, Pink J, Oxley B. Accuracy of three-
A U T H O R C ON T R I B U T I O NS dimensional printed patient-specific drill guides for treatment
Fracka AB, MSc, DVM: Performed surgical procedures, of canine humeral intracondylar fissure. Vet Surg. 2020;49:
obtained radiographic images and reviewed data. Drafted, 363-372.
10. Toni C, Oxley B, Behr S. Atlanto-axial ventral stabilisation
revised, and approved the submitted manuscript. Oxley B,
using 3D-printed patient-specific drill guides for placement of
VetMB, DSAS(Orth): Assisted with the design and fabrica- bicortical screws in dogs. J Small Anim Pract. 2020;61:609-616.
tion of the 3D guides, provided images of the PSGs (Figure 4) 11. Kamishina H, Sugawara T, Nakata K, et al. Clinical application
and helped with development of the surgical workflow. of 3D printing technology to the surgical treatment of atlan-
Reviewed drafts and approved the final manuscript for sub- toaxial subluxation in small breed dogs. PLoS One. 2019;14:
mission. Allen MJ, VetMB, PhD: designed the study, e0216445.
assisted with all surgeries, and performed radiographic mea- 12. Toni C, Oxley B, Clarke S, Behr S. Accuracy of placement of
pedicle screws in the lumbosacral region of dogs using 3D-
surements and data analysis. Reviewed and revised drafts of
printed patient-specific drill guides. Vet Comp Orthop Trauma-
the manuscript and approved the final version for
tol. 2021;34:53-58.
submission. 13. Hall EL, Baines S, Bilmont A, Oxley B. Accuracy of patient-
specific three-dimensional-printed osteotomy and reduction
A C K N O WL E D G M E N T S guides for distal femoral osteotomy in dogs with medial patella
The authors thank Dr Paddy Mannion (Cambridge Radi- luxation. Vet Surg. 2019;48:584-591.
ology Referrals) and the diagnostic imaging team at the 14. Fitzpatrick N, Miraldo DC, Meswania J. Custom-built con-
Queen's Veterinary School Hospital for invaluable assis- strained uniaxial and rotating hinge total knee replacement in
cats: clinical application, design principles, surgical technique,
tance with the CTs and X-rays for this study. STL files of
and clinical outcome. Vet Surg. 2021;50:1283-1295.
the canine TKR implants were kindly supplied by 15. Fitzpatrick N. Custom and non-custom implants: a review of
BioMedtrix LLC. complications. ACVS Surgical Summit 2018, Phoenix, AZ.
16. Sariali E. Accuracy of the knee reconstruction after total knee
CONFLICT OF INTEREST STATEMENT arthroplasty. A comparison of custom cutting blocks based on
The authors declare no conflicts of interest related to this a three-dimensional computerised planning and a conventional
report. ancillary. EFORT Congress Sessions. Online 2014 http://efort.
conference2web.com/content/3568/details.
17. Abdel MP, Oussedik D, Parratte S, Lustig S, Haddad FS. Coro-
R EF E RE N C E S
nal alignment in total knee replacement. Historical review con-
1. Liska WD, Doyle ND. Canine total knee replacement: surgical temporary analysis and future direction. Bone Joint J. 2014;96-
technique and one-year outcome. Vet Surg. 2009;38:568-582. B:857-862.
 1532950x, 2023, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vsu.13963 by Mississippi State University, Wiley Online Library on [13/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FRACKA ET AL. 685

18. Lombardi AV Jr, Berend KR, Adams JB. Patient-specific navigation of VISIONAIRE patient-specific instrumentation for
approach in total knee arthroplasty. Orthopedics. 2008;31: total knee arthroplasty. J Arthroplasty. 2013;28:469-473.
927-930. 26. Ritter MA, Davis KE, Meding JB, Pierson JL, Berend ME,
19. MacDessi SJ, Jang B, Harris IA, Wheatley E, Bryant C, Malinzak RA. The effect of alignment and BMI on failure of
Chen DB. A comparison of alignment using patient specific total knee replacement. J Bone Joint Surg Am. 2011;93:1588-
guides, computer navigation and conventional instrumentation 1596.
in total knee arthroplasty. Knee. 2014;21:406-409. 27. Liska WD, Marcellin-Little DJ, Eskelinen EV, Sidebotham CG,
20. Jeffery RS, Morris RW, Denham RA. Coronal alignment after Harrysson OLA, Hielm-Björkman AK. Custom total knee
total knee replacement. J Bone Joint Surg Br. 1991;73:709-714. replacement in a dog with femoral condylar bone loss. Vet
21. Catani F, Digennaro V, Ensini A, Leardini A, Giannini S. Navi- Surg. 2007;36:293-301.
gation-assisted total knee arthroplasty in knees with osteoar- 28. Mason JB, Fehring TK, Estok R, Banel D, Fahrbach K.
thritis due to extra-articular deformity. Knee Surg Sports Meta-analysis of alignment outcomes in computer-assisted
Traumatol Arthrosc. 2012;20:546-551. total knee arthroplasty surgery. J Arthroplasty. 2007;22:
22. Tigani D, Masetti G, Sabbioni G, Ben AR, Filanti M, Fosco M. 1097-1106.
Computer-assisted surgery as indication of choice: total knee 29. Bäthis H, Perlick L, Tingart M, Perlick C, Lüring C, Grifka J.
arthroplasty in case of retained hardware or extra-articular Intraoperative cutting errors in total knee arthroplasty. Arch
deformity. Int Orthop. 2012;36:1379-1385. Orthop Trauma Surg. 2005;125:16-20.
23. Peters KM, Hutter E, Siston RA, Bertran J, Allen MJ. Surgical
navigation improves the precision and accuracy of tibial com-
ponent alignment in canine total knee replacement. Vet Surg.
2016;45:52-59. How to cite this article: Fracka AB, Oxley B,
24. Hoerlesberger N, Glehr M, Amerstorfer F, Hauer G, Allen MJ. 3D-printed, patient-specific cutting
Leithner A, Sadoghi P. Residents' learning curve of total knee guides improve femoral and tibial cut alignment in
arthroplasty based on radiological outcome parameters: a retro- canine total knee replacement. Veterinary Surgery.
spective comparative study. J Arthroplasty. 2021;36(1):154-159.
2023;52(5):674‐685. doi:10.1111/vsu.13963
25. Lustig S, Scholes CJ, Oussedik SI, Kinzel V, Coolican MR,
Parker DA. Unsatisfactory accuracy as determined by computer