Management of Refractory Aseptic Subtrochanteric Non-union by Dual Plating

Abstract

Background

Subtrochanteric fractures are challenging to treat because of their anatomical and biomechanical behaviours. Non-unions of this region become much more difficult to treat because of the previous surgical scar, fibrosis, mal-reduction, presence of an implant, compromised soft tissue, and osseous vascularity, bone-mass loss etc. The aim is to provide a stable mechanical environment by PF-LCP, augmented by LCP (dual plating) where biology can work uneventfully. Biology is re-initiated by decortication (shingling) and autologous cancellous bone graft.

Methods

Twelve cases of failed aseptic subtrochanteric non-union either with intact or broken implant were included in this study in a period of 3 years from August 2016 to July 2019. The interposing fibrous tissue resected in patients with mal-aligned fragments. The mechanical stabilization is achieved by orthogonal dual plating. PF-LCP on lateral and 4.5 mm LCP anteriorly, decortication, and cancellous graft applied before applying for the anterior plate. Patients were encouraged for a toe-touch walk with walking-frame from 3rd post-operative day. Functional outcomes were assessed using Parker Mobility Score (PMS).

Results

All fractures united in 7 ± 1.53 months. ROM at the knee remained unchanged but improved at the hip after revision surgery. Average PMS improved to 7.58 from pre-revision 1.75 validating the efficacy of this protocol.

Conclusion

Adequate stability by dual-plate construct and re-initiation of cellular and biochemical processes by decortication and cancellous bone-graft reunited ununited subtrochanteric fractures. This particular combination of plates and decortication has not been employed earlier as per our review of the literature.

Aim

To offer a new paradigm for the management of surgically failed subtrochanteric non-unions.

Introduction

Non-union is an irreversible stage in the cascade of fracture healing devoid of all potentials of regeneration. Non-union of subtrochanteric fracture reflects either inadequate stability or poor biology or both. Favorable biological and mechanical environments are pre-requisites for uneventful and uncomplicated fracture healing. Subtrochanteric zone is the most highly stressed zone of the body, the level of stress can go up to six times of body weight [1]. Thick cortical bone, tenuous blood supply, cantilever anatomy of head and neck with high unequal biomechanical stress, compressive forces on posterio-medial cortices more by 20% as compared to the tensile surface creating varus tendency, cortico-cancellous junction make the fractures of this zone more susceptible to non-union [1, 2]. Implant failure is a consequence, not a cause of non-union. Failed post osteosynthesis non-unions following once/multiple surgical interventions is challenging due to previous fibrous and scar tissues, osteoporosis, compromised vascularity of bone and soft tissues, dormant infection, and presence of the failed implant. This study aims to propose a treatment protocol for refractory non-unions of subtrochanteric fractures with co-morbidities optimization.

The options for primary fixation in fresh fracture are cephalomedullary nails (CMN), blade plate, DCS, DHS, and PFLCP. CMNs are the most preferred implant. Post-surgical causes for nonunion are mainly loss of initial reduction—varus or procurvatum, misplaced implant, wrong entry point, excessive soft tissue stripping, and selection of the wrong implant. The various procedures tried to treat non-union after nailing are dynamisation of nail, exchange nailing, bone grafting, augmentative plating, nail conversion to the plate or prosthetic replacement. The length of the proximal fragment, deformity, bone defect, bone stock, age, presence of implant, and surgeon’s preference help in decision making [3].

The healing time is more in non-unions hence the implant has to bear stresses for a longer duration of time. Moreover, the earlier failed implant is indicative of a slower pace of healing than stress accumulation on the device leading to failure. Osteo-periosteal decortication and cancellous bone graft re-initiates and expedites the process of fracture healing. The “Diamond Concept” clearly gives importance to both mechanical and biological factors in addressing associated comorbidities. Physiology works in stable anatomical conditions otherwise it becomes pathological.

Materials and Methods

Twelve patients of failed osteosynthesis of subtrochanteric non-union were included in the study. The inclusion criterion was patients above 18 years of age, fractures with lapse of 9 months from index surgery without any radiological evidence of union or non-progressive reparative status for last three consecutive months, failed implant at any point of time, without any obvious infection on clinical, hematological and radiological examinations. Pathological, atypical and infected fractures were excluded from this study. The age ranged from 18 to 65 years the average being 42.83 ± 13.38 years. The average operating time was 128.33 ± 17 min. The average time between index and revision surgery was 12.5 ± 7.69 months ranging from 6 to 36 months. The necessary demographic observations are summarized in tabular form (Table 1).

Table 1.

Details of patients with their results at last follow-up

Patient ID	Age	Sex	Implant	Status	Co-morbidities	Operating time (MIN)	Time between index and revision surgery (months)	Tobacco user	Time to heal (months)	PMS pre-injury	PMS pre-revision	PMS last follow-up	Result	Follow up in months
P1	50	F	PFN	Broken	No	120	10	No	6	9	0	8	Good	12
P2	58	M	IMN	Broken	No	150	36	Yes	9	9	0	7	Fair	18
P3	18	M	IMN	Broken	No	130	6	No	4	9	0	9	Excellent	10
P4	47	F	REVERSE DF-LCP	Intact	No	140	18	No	8	9	2	9	Excellent	16
P5	56	F	DHS	Broken	HTN, DM	110	9	No	5	9	2	7	Fair	15
P6	36	M	CMN	Intact	No	130	12	Yes	8	9	4	8	Good	14
P7	38	M	CMN	Broken	No	150	11	Yes	8	9	1	8	Good	16
P8	32	M	CMN	Intact	No	110	11	Yes	6	9	4	8	Good	15
P9	52	F	CMN	Intact	No	110	10	No	7	9	0	8	Good	15
P10	28	F	PFN	Intact	No	120	9	No	6	9	4	9	Excellent	12
P11	65	F	CMN	Intact	CAD, HTN, COPD	110	12	No	9	4	0	2	Fair	15
P12	34	M	PF-LCP	Broken	NO	160	6	No	8	9	4	8	Good	13

The possibility of dormant infection is excluded by carefully detailed history and clinical, hematological (CBC, CRP, ESR) and radiological examinations (Lysis around screws, resorption of bone, sequestrum and involucrum formation). Per-operative tissue from the non-union site was sent for culture and histopathological examination for further confirmation of infection.

With all clearances, informed consents, and optimization of co-morbidities, all surgeries were done under spinal anesthesia on a traction table in the supine position. Tranexamic acid 15 mg/kg body weight half an hour before surgery is given routinely to every patient unless contraindicated and repeated 6 h after the initial dose. Local infiltration of lignocaine with adrenaline diluted to 1:600,000 ratio done in the proposed line of the incision with the intention to minimize blood loss and to have better post-surgical analgesia.

In all cases, the non-union site and implant approached through a previous surgical incision extending as per need. Implants removed, intervening fibrous tissue resected in five patients to bring fragments in alignment retaining maximum peri-osteal attachment. Resection of intervening fibrous tissue and freshening of bone ends was necessary for obtaining a stable end to end alignment in five patients, the remaining seven patients needed no resection and freshening of ends as stable alignment was secured after removal of the implant. Anatomic alignment obtained and fixed by PFLCP in compression mode. At this stage decortication 20 mm on either side of the fracture in all approachable areas done by a sharp and narrow blade chisel or osteotome till fresh bleeding from parent bone visualized. Sufficient cancellous bone graft taken from same iliac crest applied over the decorticated surface and all around the fracture. An orthogonal 4.5 mm LCP was applied on the anterior aspect again in compression mode. Alignment of fragments and placement of screws in head checked under fluoroscopy in more than two planes to avoid the possibility of error. Standard norms of dual plating were followed to avoid stress risers.

Postoperative follow up was on 1st, 2nd weeks and on monthly interval till fracture united based on radiological evidence and afterwards also. All patients were encouraged toe touch walking with a walking frame and to undergo physical therapy from 3rd post-operative day, and load-bearing gradually increased as per tolerance retaining the walking frame.

Healing of non-union assessed on clinical and radiological parameters. Painless full weight-bearing walking was a clinical parameter of the union. Reborne Bone Healing Score [4] for radiological consolidation was used and evaluated by two orthopaedic surgeons (Table 2).

Table 2.

Reborne Bone Healing Score

Cortical score	Stage
0	Non-interpretable/non-visible. Hidden by plate
1	Fracture unchanged
2	Callus present but not continuous
3	Callus continuous but fracture still visible
4	Callus with same density as cortex

Results

We classified the results as a. Excellent with no decline in Parker Mobility Score (PMS), b. Good with a decline of 1 point c. Fair with a decline of 2 points and d. Poor with a decline of 3 or more points as compared to their pre-injury values or requiring further revision surgery (Table 3).

Table 3.

Criteria for evaluation of result based on Parker Mobility Score

S. no.	Decline in Parker Mobility Score W.R.T Pre-injury level	Grading of results
1	No decline	Excellent
2	Decline by 1 point	Good
3	Decline by 2 points	Fair
4	Decline by 3 or more points, re-revision	Poor

All patients followed till osseous union and even afterwards, the range of follow-up was from 10 to 18 months the average being 14.25 ± 2.08 months. The average time to heal was 7 ± 1.53 months ranging from 4 to 9 months. 6 patients out of 12 (50%) presented with a broken implant and remaining six with intact implant with varus malalignment.

Parker Mobility Score (PMS) improved in every patient as compared to pre-revision status. On comparison to pre injury PMS Three patients regained full points, Five patients lost one point with PMS 8 had good functional results and three patients had decline of two points to pre-injury level were placed in fair category. None had poor results or required re-revision. Chart with average PMS at three different stages i.e. pre-injury, pre-revision, and last follow-up shows an encouraging result validating the success of this procedure (Fig. 1). A patient (P11) with low pre-injury PMS of 4 declined to zero before revision surgery but at last follow-up, her PMS was two labelling her recovery as fair. Four patients reported painless lurch, one patient with low pre-injury PMS continued her walking frame and two patients needed a walking stick for outdoor movements. The average pre-injury PMS was 8.58 which declined to 1.75 just before revision surgery and improved to 7.58 at last follow-up with 25% patients excellent, 50% good and 25% had fair results.

Fig. 1.

Graph indicating Parker Mobility score at Pre-injury, pre-revision and last follow-up

All culture reports were negative and histopathology did not reveal anything significant. The range of motion at the hip improved but the knee remained in the same range as before revision surgery. No correlation in result could be established with co-morbidity as they were optimized before and during the course of treatment. Tobacco chewing/smoking was discouraged during the course of treatment. There was no loss of reduction in any of our patients till bony union. No incidence of plate or screw breakage was observed. None of the patients till last follow-up presented sign of avascular necrosis of head of femur. (Figs. 2, 3, 4, 5 show pre-revision status of non-union with a failed implant, post revision and at follow up.)

Fig. 2.

A Broken PFN with atrophic non-union, eleven months after index surgery. B Immediate Post-operative-dual plate and bone graft. C Osseous union at 6 months, alignment maintained

Fig. 3.

A 2 years after index surgery, treated by reverse LCP non-union subtrochanteric with varus deformity. B Post Operataive dual plate fixation and bone graft. C Osseous union at 40 weeks, alignment maintained

Fig. 4.

A Patient presented subtrochanteric non-union with broken barrel plate 9 months after index surgery. B Immediate post-operative X ray showing dual fixation and bone graft. C Bony healing evident at 26 weeks

Fig. 5.

A Non-Union subtrochanteric fracture treated by long PFN showing loss of reduction. B Immediate post-operative showing fixation by dual plating and bone graft. C At 36 weeks, signs of radiological healing present

Discussion

An uneventful fracture healing requires favorable biological and mechanical environments, imbalance leads to failure. Mechanical instability initiates biological failure which leads to implant failure. Vice-versa, failing biology creates undue stress on fracture mechanics causing the implant to fail. No earlier author has used PFLCP and LCP combinations to provide mechanical stability. There are reports of exchange nailing with bone graft, blade plate, augmentation [5], DCS and LCP combinations [6] or PFLCP alone [7]. There are various publications to manage non-union of subtrochanteric fractures but the sample size in these reported publications is small [3]. The overall fixation failure rate was 12% with a reoperation rate of 6% by one year as reported by Parker et al. [8].

The Parker Mobility Score is a composite measurement of a patient’s mobility indoors, outdoors and during shopping [9]. We used it in our study like others [10] to measure mobility as an outcome of revision surgery. In our context, outdoor walking is like going to a neighborhood and shopping is like going to the vegetable market and alike.

Orthogonal dual plating is a strong and stiff construct. Torsional stress is completely obliterated by PF-LCP’s proximal multiplanar screws in head and neck at 95°, 120° and 135°. Axial and bending stiffness also increases many folds because of PFLCP and LCP act as a single beam construct which is 4 times stronger than conventional plating like DCS [11]. The anterior plate being nearer to the mechanical axis increases its endurance to stresses. Hence, a more favorable mechanical environment is created by reducing stress on the implant. PF-LCP and LCP provide tension band effect as both are applied on the convex surface of the femur.

Judet in 1963 pioneered osteoperiosteal decortication for non-union of diaphyseal fractures, claiming good results. Guyver et al. [12] published his article in 2012 stating excellent results by obtaining a 99% union rate yet this technique is not popularized. He further emphasized that osteoperiosteal decortication remains a highly effective surgical technique in the management of failed fracture healing. Some authors have also reported excellent results [13, 14]. Decortication is like ploughing a field, the harder you plough better you sow and reap.

Decortication up to a minimum of 20 mm on each side of the fracture, proximally and distally, provides a pink vascular bed for the cancellous graft to get vascularised by the process of angiogenesis. Decortication is a time tested procedure but ignored due to over-reliance on mechanical stabilisation and biological reactivation by bone graft alone. Decortication may help to physically interlock bone graft with recipient's bone with new bone formation [15]. The cancellous bone graft helps in obtaining osseous union by osteoconduction, osteoinduction, and osteogenesis. Decortication helps as viable raised osteoperiosteal flakes increase raw fracture area which stimulates fracture healing. The multiple mini fractures in the outer layer of cortex created by decortication, initiate a hypoxia driven signaling pathways. The decreased perfusion concurrent with increased metabolic demand of repair leads to hypoxia near fracture site [16]. As a result, there is the restoration of blood flow due to stimulated angiogenesis which is a key component in fracture healing. Cellular response releasing vascular endothelial growth factor (VEGF-A) which is a proangiogenic factor stimulates angiogenesis. Hematoma formation from bleeding periosteal and cortical vessels provides necessary biochemicals to stimulate progenitor cells to form osteoblasts which are essential in fracture healing cascade, converting an established non-union fracture to start afresh for healing. Tall et al. [17], could achieve union in all fifty patients of nonunion of diaphyseal fracture of long bones with osteoperiosteal decortication and internal fixation alone without cancellous bone graft even in atrophic non-unions.

Sebastian et al., reported re-revision surgery for the non-union subtrochanteric fracture to be 32.5% [18]. Haidukewych et al. reported bone stock from prior attempts can compromise stable fixation [3]. Dual plating overcomes this deficiency as proximal screws of PFLCP are multi-planar and multi-directional gaining purchase in the subchondral bone. Kang et al. [19], in his series of 19 patients showed an exchange of previous implant and removal of fibrous tissue was better than in those retained in the treatment of non-unions of subtrochanteric fractures, 10 versus 9. In the present study, the author removed fibrous tissue only when reduction is to be achieved. We believe, by avoiding resection of intervening fibrous tissue, time and blood-loss are saved, without affecting the ultimate healing process in fracture with stable end-to-end reduction.

The exchange nailing for non-union of diaphyseal fractures of femur/tibia supports this concept. The compression at the fracture site generated by Tension Band Effect of these orthogonal plates stimulates mesenchymal stem cells to proliferate and differentiate into osteoblasts due to Piezo-electric effect converting fibrous tissue into osseous tissue [20].

Giannoudis et al. [21], used blade plate with enhancing biological and mechanical preconditions on “diamond” concept treated fourteen subtrochanteric non-unions but in one case further revision surgery was required due to blade plate failure. In our study, no patient required revision surgery.

Dietmar et al. [22], in their study of 17 patients of non-union of subtrochanteric fractures concluded that all three factors for subtrochanteric non-union were mechanical parameters. They identified risk factors either result of increased load on the implant (varus mal-alignment), reduced intrinsic stability of fracture (lack of medial support) or reduced stability of overall mechanical construct (auto dynamization). It is therefore reasonable to assume there is a summation effect. The spectrum of stability: absolute stable/rigid fixation permits 2% strain, relative stable fixation permits 2–10% strain and unstable fixation permits beyond 10% strain which leads to non-union. Relative stable fixation is for acute fractures which have the potential to repair itself as micromotion helps in angiogenesis and subsequently osteogenesis but in the case of non-union all biological processes of healing come to a complete halt. They are reinitiated by decortication and bone grafting effectively only in a stable mechanical environment provided by rigid fixation which prevents micromotion across the fracture line. The failed implant is indicative of undisputable defeat in the race between process of fracture healing and stress concentration on a point of implant making it vulnerable to failure. Non-unions even after revision take a longer time to heal. The orthogonal plates provide a stronger and stiffer construct enduring larger stresses for a much longer time. Murray et al. [23], in 1964 started the concept of dual plating for non-unions. Egol et al. [24], raised a theoretical concern for the amount of soft tissue stripping and delayed revascularization jeopardizing vascularity. Although, this concern in the treatment of acute fractures is justifiable but not in non-unions as wide exposure and dissection are required. Rubel et al. [25], concluded that the addition of another plate doesn’t increase failure rate. Moreover, the locking plate provides fixation without undesirable effect on periosteal vascularization and mechanical drawbacks associated with conventional locking plates.

In revision surgeries, we believe more strong and rigid fixation is required, which is adequately provided by dual plating. The surgical technique for non-union surgery requires more extensive dissection, debridement of fibrous tissue, and shortening to bleeding bone.

This often results in irregular bone ends not providing enough stability because of poor contact. Longer healing times are also common, requiring implants to bear more stress for a longer period of time.

The limitation of our study is the small sample size and no control group but the strength of the study is that all surgeries performed by a single surgeon following a standard protocol.

Conclusion

Favorable mechanical and biological environments provided by orthogonal dual plating, decortication and autogenous cancellous bone graft reunited ununited subtrochanteric fractures. With this protocol, large number of patients should be studied before final verdict. We conclude, strict application of basic non-union principles can result in a successful salvage of subtrochanteric non-unions in patients with one or multiple failed prior surgical interventions, stability being the most important. Once a wise man said, “Stability is success”.

Compliance with Ethical Standards

Conflict of interest

The authors hereby, declare that there is no conflict of interest in this study. This study is not sponsored/funded by any institution. All patients were admitted and operated by the authors only.

Ethical standard statement

This article does not contain any studies with human or animal subjects performed by the any of the authors.

Informed consent

All patients were operated upon after taking their informed consent. Nowhere in this study, identity/ethnicity/gender of patient is disclosed.