There are two demographics in this group (typically these are the young adult suffering a high energy impact, and the elderly osteoporotic individual with an injury sustained in a fall).
The relative proportion of the two sub-types explains some of the variations in results and opinions. Both these sub-types experience problems, although the problems are not identical in each group.
Young patients with comminuted fractures due to high energy injuries can run into problems with nonunion/delayed union/malunion & stiffness in the short term. They have the time ahead of them to eventually develop Post -Traumatic Arthritis (PTA) .
Older patients run in to problems with failure of fixation in their osteoporotic bones, cut-out, stiffness & nonunion. Some may need joint replacement or even distal femoral replacement.
No-one these days disputes that these injuries should be treated by surgical stabilisation. Then the most obvious question is whether a locking plate or a retrograde intramedullary nail is the best implant. The answer probably depends on the type of fracture. Supracondylar fractures may be ideal for nailing, intercondylar fractures may be ideal for plating, but where on this spectrum is the point at which this binary decision flips over? The $64,000 question. Another question is the threshold at which a replacement option should be considered.
This being orthopaedics, most studies on the topic cannot be construed as high quality evidence. No definitive evidence-based answer to the above questions exists.
That said, let’s get into the papers.
A historical study but of interest nonetheless, Egund & Kolmert (1982) https://doi.org/10.3109/17453678208992856 reviewed their experience of distal femoral fractures (n=62) after a mean follow-up of 5 years. 35 had been treated non-surgically, 27 had been treated operatively. 31 had had a previous disease or fracture of the same limb (perhaps suggesting disuse osteoporosis).
Sixty-two patients treated for distal femoral fractures from 1969-1976 were
re-examined, after a mean follow-up period of 5 years, to study deformities,
gonarthrosis and function. These fractures occur mainly in elderly persons with bone fragility due to age and disease. The deformities were analysed from precisely defined radiographic projections. An anatomical classification into supracondylar, unicondylar and bicondylar fractures, with subdivisions for undisplaced and displaced fractures was used. A special group for transcondylar fractures was included.
Displaced bicondylar fractures mostly healed with varus and anterior angulation, media1 unicondylar fractures with varus and lateral unicondylar fractures with valgus angulation. Most of the healed supracondylar fractures showed varus angulation. Three patients developed arthrosis in both the femoro-tibial and patellar compartments, and eleven only in the patellar area. Intercondylar or transcondylar diastasis, or difference of level in the joint surface exceeding 3 mm, caused a significant degree of gonarthrosis.
Function was assessed using the Knee Disability Sheet prepared at the Hospital for Special Surgery. The necessity of radiographic examination of the patellar joint in the axial projection on admission was stressed. Accurate reduction and adequate stabilization of intra-articular fractures seem to be important for reducing the risk of gonarthrosis and later impairment of function.
Of the purely supracondylar fractures, 1/9 had post-traumatic arthritis (PTA) although 3/9 had 15 degrees of angulation.
Of the unicondylar fractures, 2/10 had developed PTA in the patellofemoral compartment, but none in the femoro-tibial compartment.
They commented: ‘The influence on arthrosis of angular deviation in distal femoral fractures was not confirmed in this material but the observation period (2-10 years) was probably too short to exclude the possibility of gonarthrosis occurring later’.
The conclusion that we might draw from this is that articular incongruity causes PTA in a relatively short time; if there is a significant effect of malalignment on PTA, it takes longer than this follow-up period of 2-10 years.
This would be consistent with evidence from observations related to fractures elsewhere in the body – damage with displacement of joint surfaces quickly causes PTA if unreduced or unreconstructable ; more slowly if at all when anatomic reduction is achieved; and very slowly (if at all) if there is malalignment of a fracture which does not involve the joint.
(There will be a separate post about shaft malunion and osteoarthritis later on).
The above paper illustrates the closest we are likely to find to a non-operative paper in the relatively recent past, so it is of historical interest. Nowadays, non-operative management of distal femoral fractures is almost unheard of, because of the problems with alignment, stiffness, and prolonged hospitalisation associated with non-operative management.
Rademakers et al, (2004) Journal of Orthopaedic Trauma; Apr2004, Vol. 18 Issue 4, p213-219, 7p reviewed 67 patients with intra-articular femoral fractures treated surgically. 32 were traced and followed up at a mean of 14 years; the mean range of motion was 118 degrees.
The Neer score showed good to excellent results in 84% of the patients, and the HSS knee score showed good to excellent results in 75% of the patients. Patients with isolated fractures scored significantly better functionally
(Neer/HSS 90 points) compared with those with multiple fractures.
The Ahlbäck score showed a moderate to severe development of secondary osteoarthritis in 36% of all patients. Seventy-two percent of these patients still scored a good to excellent functional result. Seven patients (10%) had local complications in the form of a deep wound infection. Five of these patients were treated successfully, whereas 2 had a chronic infection that subsequently led to an arthrodesis.
These authors concluded that surgical treatment of monocondylar and bicondylar femoral fractures gave good long-term results after open reduction and internal fixation; that knee function increases through time, though the range of motion does not increase after 1 year; that the presence of secondary osteoarthritis does not mean less favorable functional results in most patients.
Naturally one must have some reservations these conclusions. If less than half the patients were followed up in the long term, how representative were they of the group as a whole? Disgruntled patients, or patients with a poor outcome, or whose joint has been replaced, might be disinclined to attend follow-up, or think it pointless. So the patients who attended may not be representative of the whole group.
Markmiller et al, Clinical Orthopaedics & Related Research,Number 426, pp. 252–257, 2004. ‘Femur–LISS and Distal Femoral Nail for Fixation
of Distal Femoral Fractures’. Two groups of 16 patients documented prospectively, half treated by LISS, half by distal femoral intramedullary nail. Follow-up was short – 1 year. The plate cases flexed to an average 110 degrees, versus 103 degrees for the nails (not significant). Overall there were no significant differences in outcome parameters, including non-union. In my view, this is a somewhat re-assuring preliminary study, but such small groups don’t prove that there is no difference between the performance of the two implants. Over half the cases in each group followed low energy trauma to osteopaenic bone.
Schutz et al, Arch Orthop Trauma Surg (2005) 125: 102-108. ‘Use of the Less Invasive Stabilization System (LISS) in patients
with distal femoral (AO33) fractures: a prospective multicenter study’. 62 patients, followed for 1 year. Correct axial alignment in 49 (74%), 5-10 degrees in 13; 10-20 degrees deviation in 1 case. Sagittal alignment correct in 56 cases (85%); 5-10 degrees deviation in 5; 10-20 degrees in 2 cases. 85% had healed in 1 year. 14 underwent other operations (6= bone grafting, 3 revisions due to implant loosening, debridements for infection in 2 cases. No patient reported outcomes apparently assessed.
The authors concluded: ….’that all distal femoral fractures (type
AO33A—C) including all degrees of severity can be stabilized with the LISS.
– There is no need for primary cancellous bone grafting;
in situations with severe open fractures or bone loss,
secondary bone grafting may be necessary.
– When using a minimally invasive surgical technique, it
is especially important to apply careful fracture
reduction and to position the implant precisely with
exact intraoperative controls.
– These results presented here are in accordance with
the manufacturer’s recommendations: LISS should be
fixed to the proximal fragment by at least 4 monocortical
locking head screws. Fixation to the distal
fragment should be carried out with 5 long locking
head screws.
– During the course of the study, the application of
longer implants was found to be advantageous.
– There is a need for a prospective, comparative study..’
Thomson et al, 2008, Orthopedics. 2008;31(8)https://doi.org/10.3928/01477447-20080801-33748-750Long-term Functional Outcomes After Intra-articular Distal Femur Fractures: ORIF Versus Retrograde Intramedullary Nailing. Retrospective study supported by Synthes. Small series, low rate of follow up.23 fractures in 22 patients. Results better in IMN than in locked plates. 50% had significant radiological PTA at 7 years, few had yet had knee replacement. The authors noted that SF-36 data showed significant dysfunction, with a mean physical functioning score of 40.5±31.3 (about 2 standard deviations
below the US population mean score (84.2±23.3). The mental health component mean was 66.3±26.8, less than 1 standard deviation below the US mean (74.7±18.0).” There was no significant difference in any domain of the SF-36/SMFA/Iowa Knee Score between the nail & plate groups.
Smith et al, Injury, Int. J. Care Injured 40 (2009) 1049–1063. ‘The clinical and radiological outcomes of the LISS plate for distal femoral fractures: A systematic review.’ ‘Twenty-one studies assessing 663 patients with 694 fractures were reviewed. The findings suggest that the LISS system may be an appropriate fixation method for the management of distal femoral fractures. However, there remains a high incidence of loss of reduction (n = 134; 19%), delayed or non-union (n = 40; 6%) and implant failure (n = 38; 5%). On analysis, such complications were largely confined to articles published before 2005, therefore during the infancy of the widespread clinical application of this trauma system. On critical appraisal, the evidence-base remains limited by recruiting small, under-powered sample sizes and poorly accounting for confounding variables such as osteoporosis, diabetes, multi-trauma and fracture classification.
Conclusion: Further study is required to assess the outcomes of LISS fixation in specific patient populations, and to compare the outcome of this fixation method to condylar plates and intramedullary devices, to determine the optimal management strategy for this complex patient group’.
All I can do is echo the assessment above! This is worth reading in full (all papers here deserve a full read of course).
Kolb et al, Injury, Int. J. Care Injured 40 (2009) 440–448:”The condylar plate for treatment of distal femoral fractures: A long-term
follow-up study” This retrospective study presented the long-term functional and radiological outcomes of indirect reduction techniques and fixation with a condylar plate for treatment of distal femoral supracondylar or intracondylar femoral fractures.
The series included 24 men and 17 women, mean age 51 years, between March 1994 and April 1999. All fractures were AO type 33, and eight were open fractures. Primary iliac bone graft was used in five cases. In one case of severe osteoporosis, screw fixation was augmented with cement. There were three delayed unions, one non-union and two infections; four
participants required reoperation with bone grafts. Two (5%) participants developed a second varus deformity and three a second valgus deformity; correction osteotomy with bone grafts was necessary in these cases.
At 1-year follow-up, 37 (90%) were available; all had healed fractures. One participant had a chronic infection, but refused arthrodesis; 23 (62%) had no pain and were able to walk 1–5 km or more; 4 (11%) had intermittent pain; 5 had (14%) moderate pain, 3 (8%) experienced restricted function on stairs; and 2 (5%) used a walking aid. (It is refreshing to see some outcomes quoted that patients would find meaningful).
The mean range of motion was normal in 19 (51%) cases, 15 (41%) participants had a range of motion between 90 and 120 degrees and 3 (8%) between 70 and 90 degrees. There was a leg length difference >2 cm in four (11%) cases. Within 1 year, 17 (45%) participants were working again and 8 (22%) had had to change their job. The mean Neer score was 78 (50–100) points.
After a mean follow-up of 9.5 years, the mean Neer score was 82 points and indicated that function was excellent in 16, satisfactory in 9, unsatisfactory in 4 and poor in 2 cases. The mean Neer score in cases of isolated fracture was 89 points and in cases with additional injuries was 72 points. Thus the long-term results of indirect reduction techniques of distal femoral fractures treated with the condylar plate were good to excellent in 82% of cases.
Henderson et al, J Orthop Trauma 2011;25:S8–S14. ‘Locking Plates for Distal Femur Fractures: Is There a Problem With Fracture Healing?
I recommend reading this in full. 15 publications and 3 abstracts were reviewed.The rate of complications related to healing ranged from 0% to 32% in these studies. Implant failures occurred late with 75% of failures occurring after 3 months and 50% after 6 months.
Conclusions: Complications of healing including nonunion, delayed union, and implant failure are not infrequent and represent ongoing problems with distal femur fracture treatment. Quotes:
‘The combination of excellent stability provided by
locking plates and minimally invasive biologically friendly
insertion should lead to improved healing rates over previous
devices. Unfortunately, clinical experience and some reports in
the literature indicate that fracture healing may not be better than
that achieved with previous methods of fixation. Difficulties
with fracture healing in the distal femur may present clinically as
an established nonunion, delayed union, a need for secondary
procedures, implant failure, and late loss of alignment….
There have been multiple recommendations for plate
modifications to increase construct strength and stability.
These include limiting screw cannulations, increasing plate
thickness, limiting plate holes, and the use of longer
plates. Limiting postoperative weightbearing is also
often recommended. These modifications are directed at
strengthening the mechanical properties of the fracture–plate
construct…… it is possible but not certain that these types of changes would have helped the eight cases that failed within 3 months to
successfully heal. However, it is much less likely that the
remainder of cases that failed after 3 months would have
successfully healed with more robust implant. In these cases,
plate modifications that increase construct rigidity may in fact be counterproductive and efforts should focus on enhancing
the healing environment at the fracture site and promoting
earlier and more robust callus formation.’
Hoskins et al, Bone Joint J 2016;98-B:846–50.’Nails or plates for fracture of the distal femur? n=297. IMN = 102, locking plate 195. Retrospective registry data. There was a clinically relevant and significant difference in quality-of-life at six months in favour of fixation with an IMN (mean difference in EuroQol-5 Dimensions Score (EQ-5D) = 0.12; 95% CI 0.02 to 0.22; p = 0.025). There was weak evidence that this trend continued to one year (mean difference EQ-5D = 0.09; 95% CI -0.01 to 0.19; p = 0.073). There was a significant although very small reduction in angular deformity using an IMN (mean difference -1.02; 95% CI -1.99 to -0.06; p = 0.073). There was no evidence that there was a difference in any other outcomes at any time point.
Harvin et al, Injury, Int. J. Care Injured 48 (2017) 2597–2601. ‘Working length and proximal screw constructs in plate osteosynthesis
of distal femur fractures’. Ninety-six patients with distal femur fractures with a mean age 60 years met inclusion criteria. None of the clinical parameters were statistically significant indicators of union. Likewise, none of the following surgical technique parameters were associated with fracture union: plate metallurgy, the mean working length, screw density and number of proximal screws and screw cortices. However, diaphyseal
screw technique did show statistical significance. Hybrid technique had a statistically significant higher chance of union when compared to locking (p = 0.02). All proximal locking screw constructs were 2.9 times more likely to lead to nonunion.
Conclusions: Plating constructs with all locking screws used in the diaphysis when bridge-plating distal femur locking plates were 2.9 times more likely to incur a nonunion. However, other factors associated with more
flexible fixation constructs such as increased working length, decreased proximal screw number, and decreased proximal screw density were not significantly associated with union in this study. 54% of cases had alignment within 5 degrees, 45% had greater than 5 degrees: there was no significant difference in union rate.
Kiyono et al, Journal of Orthopaedic Surgery and Research (2019) 14:384
https://doi.org/10.1186/s13018-019-1401-9. ‘Clinical outcomes of treatment with locking compression plates for distal femoral fractures in a retrospective cohort’.
n=71. 26 simple, 45 comminuted. 7 non-union. The non-union rate was significantly higher in comminuted fractures with bone medial fracture
distance exceeding 5 mm.
As in the Henderson series, there were no plate length differences, or bridge-span lengths differences between union & nonunion groups.
The authors commented:
‘Bone union was observed in all cases with simple or comminuted fractures
with respective medial fracture distances of ≤2mm or ≤ 5 mm. Based on our findings, we concluded that bone fragment distance between fracture fragments is more important than bridge span length of the fracture site and the number of empty holes.
Reducing comminuted bone fragments is difficult when the inner portion of the fracture site also exhibits comminuted fracture. In such cases, shortening the fracture site to approximately 5–10mm would be acceptable
to decrease the distance between the proximal and distal bone fragments….
In distal femoral fractures, non-union sometimes may occur, despite improvements in implant and reduction techniques. Well-known risk factors include smoking, bone defects, and comminuted fractures, which were supported by this study’s results. The bone fracture distance on the medial side of the distal femur was related to bone union rather than screw position.These results will be helpful in treating distal femoral fractures with plates.
Based on our findings, we concluded that bone fragment distance between fracture fragments is more important than bridge span length of the fracture site and the number of empty holes. The number of empty holes near the fracture site and the rate of bone healing were not clearly related’.
My conclusion from the above studies is that some technical implant-related factors and techniques may not be as important as we thought they were, with the exception of all-proximal-locking-screws.
We can’t change the nature of the injury itself, but it has to be taken into account. The patient’s habits such as smoking, alcohol use, opiate use, and medications may be susceptible to change – or not.
Reading through the historical accounts of this injury up to the present day, I am struck that problems with union (nonunion, delayed union, fixation failure & revision) seem to assume increasing importance in recent years. Bone healing does not appear to be facilitated by our current technical ability to hold bones rigidly out to original length (until fixation fails), allied with the move away from primary bone grafting in comminuted fractures.
Should we allow comminuted fracture zones to shorten in order to heal, and lengthen elsewhere? Or should we go back to primary bone grafting in very comminuted cases?
The Hoffa fracture is a rare fracture that shears off the posterior part of the lateral femoral condyle, or less commonly the medial femoral condyle. Most publications are small case studies or individual variants. It can be part of an extensively comminuted articular fracture of the femur.
Chandrabose et al, Journal of Orthopaedics 20, 2020, 21-27, published ‘A computed tomography-based classification of Hoffa fracture: Surgical
treatment considerations and prognostic outcome with assessment of
reproducibility’. These authors reviewed 103 Hoffa fractures occurring over a ten year period. They described a CT classification that makes sense, and outcomes related to the classification. Inter- and intra-observer agreement showed substantial agreement. Useful reference list. Overall the results were very good. I wonder if a European or North American population would be able to achieve as much knee flexion as was achieved in this large series from India. Squatting, sitting & kneeling in a high knee flexion position is known to be part of Indian and other Asian cultures, and probably this assists rehabilitation. (As a young teacher in India decades ago, I remember my initial difficulty in sitting cross-legged on the floor for meals without my legs going numb – which caused great amusement to my Indian colleagues! Eventually you adapt. Wonderful memories.)
Senthilkumaran et al, European Journal of Trauma and Emergency Surgery (2019) 45:841–848. https://doi.org/10.1007/s00068-019-01103-7. ‘Total knee arthroplasty for distal femoral fractures in osteoporotic bone: a systematic literature review’. 14 papers used in quantitative and qualitative synthesis. n= 181. One single cohort study (level III); 13 papers were case series (level IV). Mean mortality rate = 3.34% (range 0–10) at 30 days and 18.4% (range 0–42) at 1 year. Mean revision rate = 3.43% (range 0–25) at 1 year. The mean time to mobilisation was 3.90 days (range 2.5–6) with a mean time to discharge being 16.6 days (range 8–33).
The evidence is limited in quality, but it suggests that there is a role for acute knee arthroplasty in some distal femoral fractures. Mortality and revision rates are not excessive, and TKR may lead to faster mobilisation and discharge.
Hart et al, Journal of Arthroplasty, 2017-01-01, Volume 32, Issue 1, Pages 202-206. ‘Open Reduction vs Distal Femoral Replacement Arthroplasty for Comminuted Distal Femur Fractures in the Patients 70 Years and Older’. Retrospective relatively small study (because of frequency of these injuries). Re-operation rates similar, around 10%. Approx 20% nonunion in ORIF group. 23% ORIF group wheelchair-bound; none were in DFR group. The authors argue for a prospective RCT for comminuted articular fractures in the elderly, and it is hard to disagree.