Treatment of atrophic nonunion via autogenous ilium grafting assisted by vertical fixation of double plates: A case series of patients.

OBJECTIVE: To investigate the efficacy of the treatment of atrophic nonunion using structural autogenous ilium bone grafting in combination with vertical fixation of double plates.
METHODS: This retrospective study analysed the clinical data from consecutive patients with atrophic nonunion who underwent autogenous ilium grafting in combination with double-plate vertical fixation. The injury type and the bone affected by nonunion, the duration of nonunion and the outcomes following surgery were recorded for all patients.
RESULTS: The study enrolled 43 patients with atrophic nonunion of the upper and lower limbs: 17 patients with tibial nonunion, 21 with femoral nonunion, four with humeral nonunion and one with radial shaft nonunion. The mean duration of postoperative follow-up was 14.5 months (range, 8-28 months). A total of 43 of 43 patients (100%) achieved a healed nonunion fracture without the occurrence of complications such as infection, fracture of internal fixation or pain in the harvesting site. Comprehensive postoperative assessments of bone healing and function were observed to be good and/or excellent in all 43 patients.
CONCLUSION: Structural autogenous ilium grafting used in combination with double-plate vertical fixation can provide a stable structural environment for near optimal bone healing in patients with atrophic nonunion.

Introduction

Despite of rapid progress in orthopaedic diagnosis and treatment, there still remains a nearly 5% incidence of limb nonunion.[1] As a result of its long duration, high costs and poor long-term prognosis,[2] limb nonunion often greatly impacts on patients, physicians and the healthcare system as a whole. In 2014, the Department of Orthopaedic Surgery, Hong Hui Hospital, Xi'an Jiaotong University College of Medicine, Xi'an, Shaanxi Province, China proposed to treat limb nonunion using structural autogenous ilium grafting combined with a locking plate. This technique has proven to be a reliable treatment method, both maximizing osteoconductive, osteogenic and osteoinductive properties of one autogenous bone as well as improving the biological environment at nonunion sites, resulting in a healing rate of 97.1%.[3] However, bridge fixation of a single plate cannot provide a strong enough mechanical support, resulting in four cases suffering from breakage of the plates after full-weight bearing activities.[3] As a consequence, our department subsequently proposed using structural autogenous iliac bone grafting in combination with vertical fixation of double plates. This current study retrospectively analysed clinical data from patients with atrophic nonunion who received this treatment in order to investigate the advantages and disadvantages of this method and to provide suggestions for its clinical application.

Patients and methods

Study population

This retrospective study enrolled consecutive patients with atrophic nonunion who received treatment using structural autogenous iliac bone grafting in combination with vertical fixation of double plates between January 2014 and May 2015 in the Department of Orthopaedic Surgery, Hong Hui Hospital, Xi'an Jiaotong University College of Medicine, Xi'an, Shaanxi Province, China. The inclusion criteria were as follows: (i) patients that satisfied diagnostic standards of atrophic nonunion: i.e. an unhealed fracture within 9 months and X-rays in 3 consecutive months not showing signs of fracture growth,[4] as well as no callus formation at the fracture ends or bone resorption accompanied by a visible bone defect; (ii) patients that agreed to participate in this study and had complete follow-up data. The exclusion criteria were as follows: (i) patients with a previous history of smoking; (ii) patients with a history of diabetes mellitus, metabolic disorders, immune diseases or cardiovascular diseases;[5] (iii) patients with long-term use of immunosuppressive agents or nonsteroidal anti-inflammatory drugs;[5] (iv) patients that suffered from hypertrophic nonunion or infected nonunion.

This study was approved by the Ethics Committee of Hong Hui Hospital (no. 201601002 201601002 201601002201601002). All patients agreed to participate in the study and provided written informed consent.

Data collection

Demographic and clinical data for each patient were retrieved from the medical records, including age, sex, cause of injury, number of operations and time since injury. Patients were followed serially in the clinic at 2 weeks, 4 weeks and then monthly until fracture union was observed by X-ray imaging and computed tomography scans. The mean time to fracture union and the rates of complications were calculated.

Surgical methods

General anaesthesia was administered during the surgical procedures. Patients were in the supine position and different operative incisions were selected according to the individual cases of nonunion. The Judet periosteal stripping technique[6] was used to reveal nonunion sites, with the surrounding soft tissue and periosteum protected as distally as possible. Then, the original plate was removed, with scar tissue and hardened sequestrum clearly eliminated. Normal lines of force of the limbs were corrected and bridge fixation was performed using a locking compression plate and at least six layers of cortices fixed at each side using locking screws. Subsequently, a slot was opened at the side across the ends of the nonunion. Punctuate bleeding (Paprika sign) was achieved during the operation.[7] The length of the bone groove was measured and corresponding ilium with cortex was selected. Then, it was trimmed and loaded into the bone groove, while cancellous bone granules were loaded into any bony defects. In addition, a reconstructive plate was attached to the anterior side of the bone graft and at least four layers of cortices were fixed at each side using locking screws.

Postoperative treatment

A drainage tube was indwelled after the operation and unobstructed drainage was maintained for 24–48 h. Functional exercise using a continuous passive motion machine (Rimec CPM 3000E; Beijing Technology Company Limited, Beijing, China) was started on the second day after surgery. Active functional exercise and muscle contraction exercise were started on the third day after surgery. Fracture status was regularly reviewed by X-ray imaging. If the formation of an external callus was found, the patient was asked to start half-load exercise with crutches. If the X-ray showed healed bone, the patient was instructed to start full weight-bearing exercises without crutches.

Study outcomes

The outcome measures included: (i) bone healing time as determined using the healing standards of fractures that included no tenderness or longitudinal percussion pain and no abnormal activities in the affected limbs;[8] X-ray images showed a blurred fracture line or continuous calluses; the upper limbs could continually uphold 1 kg of weight for 1 min and the lower limbs could walk on a smooth floor for 3 min without support; (ii) the occurrence of complications, including infection, plate fractures and pain at the harvesting site; (iii) a comprehensive postoperative assessment,[9] including bone and functional assessments. Bone assessment included bone healing, presence of infection, angulation deformity and limb shortening. An excellent bone assessment was represented by bone healing and no infection, angulation deformity of less than 7 degrees and limb shortening less than 2.5 cm. Good bone assessment was represented by bone healing and any two the above three findings noted. Poor bone assessment was represented by bone healing and only one of the above three findings noted. Very poor bone assessment was represented by a fracture that did not heal.

The functional assessment included checking for claudication, joint ankyloses, muscular atrophy and general dysfunction. Excellent functional assessment was represented by free activity and none of the above four findings present. Good functional assessment was represented by free activity and any one or two of the above four findings noted. Poor assessment was represented by free activity and any three or four of the above four findings noted. Very poor assessment was represented by failure of autonomic activity.

Results

This retrospective study enrolled 43 patients (28 males and 15 females) with a mean ± SD age of 42.0 ± 3.8 years (range, 23–68 years). There were 17 patients with tibial nonunion, 21 with femoral nonunion, four with humeral nonunion and one with radial shaft nonunion. A total of 23 patients sustained an injury from tumbling, 16 from a traffic accident, three from falling and one from crashing (Table 1). Photographs and X-ray images of typical patients are shown in Figures 1 and 2. The mean ± SD duration of nonunion (i.e. time since injury) was 24.2 ± 2.4 months (range, 6–156 months). Eleven patients received repeated nonunion surgery (seven with plate + plate, two with external fixation + plate, one with intramedullary nail + plate, and one with plate + bone grafting) and 32 patients received one nonunion surgery (all with plates).

Table 1.

Demographic and clinical data of patients (n = 43) who underwent treatment using structural autogenous iliac bone grafting in combination with vertical fixation of double plates to correct atrophic nonunion of upper and lower limbs.

Patient	Age, years	Sex	Cause of injury	Bone affected by nonunion	Number of operations	Time since injury, months	Follow-up, months	Duration of bone healing, months
1	68	M	Traffic	Tibial	1 (P)	10	13	6
2	43	M	Falling	Tibial	1 (P)	14	10	7
3	29	M	Tumbling	Femoral	1 (P)	6	22	14
4	42	M	Tumbling	Tibial	1 (P)	14	14	9
5	55	M	Falling	Femoral	2 (P+P)	26	24	14
6	32	M	Traffic	Humeral	1 (P)	12	8	4
7	25	F	Traffic	Femoral	1 (P)	11	16	11
8	27	M	Traffic	Femoral	1 (P)	9	16	11
9	28	M	Tumbling	Tibial	1 (P)	24	10	6
10	42	M	Tumbling	Femoral	1 (P)	16	8	5
11	44	F	Traffic	Femoral	2 (E+P)	24	24	9
12	43	M	Tumbling	Tibial	1 (P)	9	14	7
13	63	F	Falling	Femoral	1 (P)	96	11	10
14	51	F	Tumbling	Tibial	1 (P)	12	13	8
15	36	F	Traffic	Tibial	2 (P+P)	24	11	8
16	38	M	Tumbling	Femoral	1 (P)	16	12	10
17	45	M	Tumbling	Femoral	2 (P+P)	48	13	9
18	39	M	Traffic	Tibial	1 (P)	9	14	6
19	52	F	Tumbling	Tibial	1 (P)	9	13	6
20	34	M	Tumbling	Tibial	1 (P)	14	10	6
21	56	M	Traffic	Tibial	1 (P)	36	14	8
22	49	F	Tumbling	Humeral	2 (P+P)	10	17	7
23	30	M	Traffic	Femoral	1 (P)	12	16	8
24	62	M	Tumbling	Humeral	2 (P+P)	120	11	5
25	32	M	Traffic	Tibial	2 (P+P)	36	12	8
26	48	F	Tumbling	Femoral	1 (P)	10	20	14
27	41	M	Tumbling	Femoral	1 (P)	11	11	6
28	46	F	Tumbling	Humeral	1 (P)	12	10	5
29	57	F	Tumbling	Radius	2 (P+P)	156	12	5
30	47	F	Tumbling	Femoral	2 (N+P)	24	18	14
31	23	M	Traffic	Femoral	1 (P)	10	20	11
32	38	M	Traffic	Tibial	2 (E+P)	16	13	10
33	43	M	Tumbling	Tibial	1 (P)	9	12	6
34	42	F	Traffic	Femoral	1 (P)	24	12	8
35	29	F	Tumbling	Femoral	1 (P)	9	28	9
36	45	M	Tumbling	Femoral	1 (P)	9	14	10
37	46	F	Traffic	Tibial	1 (P)	11	24	5
38	57	F	Tumbling	Femoral	1 (P)	9	14	10
39	55	M	Crashing	Femoral	2 (P+G)	72	9	6
40	51	M	Tumbling	Tibial	1 (P)	14	9	5
41	28	M	Traffic	Femoral	1 (P)	10	21	11
42	41	M	Tumbling	Tibial	1 (P)	11	10	07
43	23	M	Traffic	Femoral	1 (P)	9	24	10

M, male; P, plate; F, female; E, external fixation; N, intramedullary nail; G, grafting.

Figure 1.

Images of a male patient with femoral shaft nonunion. The preoperative X-ray showed that the fracture had not healed and was accompanied by plate fracture (A). The postoperative X-ray demonstrated that the double plates were vertically fixed and the bone graft was full (B). An X-ray at 10 months after the operation suggested that the fracture lines were blurred and had healed well (C). Postoperative motor assessments were 0 degrees on stretching (D). Postoperative motor assessments were 90 degrees at flexion (E). The colour version of this figure is available at: http://imr.sagepub.com.

Figure 2.

Images of a male patient with femoral shaft nonunion. The preoperative X-ray showed that the fracture had not healed (A). A postoperative X-ray suggested that the double plates were vertically fixed and the bone graft was full (B). An X-ray at 10 months after the operation showed that fracture lines were blurred and had healed well (C). Postoperative motor assessments were 0 degrees on stretching (D). Postoperative motor assessments were 120 degrees at flexion (E). The colour version of this figure is available at: http://imr.sagepub.com.

All patients received follow-up visits after the operation. The mean ± SD duration of follow-up was 14.5 ± 1.3 months (range, 8–28 months) (Table 1). A total of 43 patients (100%) achieved a healed nonunion fracture without the occurrence of complications such as infection, fracture of internal fixation or pain in the harvesting site (Table 2). The mean ± SD duration of bone healing was 8.2 ± 1.1 months (range, 4–14 months). After the operation, all patients achieved an excellent or good result for bone healing and function. However, there were seven patients with claudication, three with joint ankyloses in the ankle and 13 with significant muscular atrophy compared with the healthy side. However, all patients regained good autonomic function, so special treatments were not required.

Table 2.

Postoperative complications and comprehensive postoperative evaluation of patients (n = 43) who underwent treatment using structural autogenous iliac bone grafting in combination with vertical fixation of double plates to correct atrophic nonunion of upper and lower limbs.

Patient	Plate fractures	Pain	Limb shortening	Infection	Angulation deformity	Claudication	Joint ankyloses	Muscular atrophy	General dysfunction	Postoperative assessments[a]
1	−	−	−	−	−	−	−	−	−	E+E
2	−	−	−	−	−	−	−	−	−	E+E
3	−	−	−	−	−	+	−	+	−	E+G
4	−	−	−	−	−	+	−	−	−	E+G
5	−	−	−	−	−	−	−	+	−	E+G
6	−	−	−	−	−	−	−	−	−	E+E
7	−	−	−	−	−	−	−	+	−	E+G
8	−	−	−	−	−	−	−	+	−	E+G
9	−	−	−	−	−	−	−	−	−	E+E
10	−	−	−	−	−	−	−	−	−	E+E
11	−	−	−	−	−	−	−	−	−	E+E
12	−	−	−	−	−	−	−	+	−	E+G
13	−	−	−	−	−	−	−	+	−	E+G
14	−	−	−	−	−	−	−	−	−	E+E
15	−	−	−	−	−	+	+	−	−	E+G
16	−	−	−	−	−	−	−	−	−	E+E
17	−	−	−	−	−	−	−	+	−	E+G
18	−	−	−	−	−	−	+	−	−	E+G
19	−	−	−	−	−	−	−	−	−	E+E
20	−	−	−	−	−	−	−	−	−	E+E
21	−	−	−	−	−	−	−	−	−	E+E
22	−	−	−	−	−	−	+	−	−	E+G
23	−	−	−	−	−	−	−	−	−	E+E
24	−	−	−	−	−	−	−	−	−	E+E
25	−	−	−	−	−	−	−	+	−	E+E
26	−	−	−	−	−	−	−	−	−	E+G
27	−	−	−	−	−	−	−	−	−	E+E
28	−	−	−	−	−	−	−	−	−	E+E
29	−	−	−	−	−	−	−	−	−	E+E
30	−	−	−	−	−	+	−	+	−	E+G
31	−	−	−	−	−	+	−	−	−	E+G
32	−	−	−	−	−	+	−	+	−	E+G
33	−	−	−	−	−	−	−	−	−	E+E
34	−	−	−	−	−	−	−	−	−	E+E
35	−	−	−	−	−	−	−	−	−	E+E
36	−	−	−	−	−	−	−	+	−	E+G
37	−	−	−	−	−	−	−	−	−	E+E
38	−	−	−	−	−	+	−	−	−	E+G
39	−	−	−	−	−	−	−	−	−	E+E
40	−	−	−	−	−	−	−	−	−	E+E
41	−	−	−	−	−	−	−	−	−	E+E
42	−	−	−	−	−	−	−	+	−	E+G
43	−	−	−	−	−	−	−	+	−	E+G

Postoperative assessments for ‘bone healing’ + ‘function’; E, excellent; G, good.

Discussion

Despite considerable progress in orthopaedic technology, limb nonunion is still observed in the clinic, often affecting both the physical and mental well-being of the patient in addition to their finances. Nonunion is a result of the synergistic action of various influencers, including biological and mechanical factors.[10] Therefore, the treatment of nonunion also needs to address the various factors at play, such as enhancing the stability of the local mechanics and simultaneously improving the biological environment to promote healing. Well-recognized fixations for nonunion after intramedullary nailing include the exchange of intramedullary nailing and plate augmentation. Studies have reported the successful treatment of nonunion by exchanging intramedullary nails,[11] which avoids exposure of the ends of fractures, thus decreasing surgical trauma. Moreover, bone healing rates reached 76–98%.[12] However, a previous study reported 96 cases of nonunion in the tibial shaft when intramedullary nails were exchanged and the rate of nonunion reached 36.5%.[13] Furthermore, contraindications to this treatment modality include infection, bone defects of greater than 5 mm and atrophic non-union.[13] Therefore, plate augmentation could supplement current therapy regimens, because it not only increases the pressure between the fracture ends, but also corrects deformities.[14] Moreover, the bone healing rate can reach 100% when using plate augmentation.[14] In the poor areas of northwest China, plate fixation remains the preferred method of fracture treatment due to economic constraints and the level of medical care available. Insufficient lengths of embedded plate can undermine local soft tissue and blood supply recovery.[15] If comminuted fractures lack interior bone support, the rigidity of the plate fixation is greatly reduced, resulting in a higher occurrence of nonunion after the plate fixation.[15] Replacement of intramedullary nails is supported by the fact that it not only provides the necessary mechanical stability, but more importantly, it allows for bone grafting of cancellous bone at the fracture ends through reaming.[16] Replacement of intramedullary nails also improves the local biological environment to enhance bone healing.[16] However, the bone mass that can be obtained remains limited, and for nonunion accompanied by bone defects, it is not appropriate to replace the intramedullary nails.[13] Research advocating plate replacement indicates that although the plate has the disadvantage of secondary trauma during surgery, it can achieve the aim of radical debridement, correction of deformities and full bone grafting.[17] Our experience supports the use of a locking plate combined with structural autogenous ilium grafting to treat nonunion after plate installation because it maximizes osteoconductive, osteogenic and osteoinductive properties of autologous bone, resulting in bone healing rates reaching 97.1%.[3] However, some treatments fail due to the span of the graft being large and the exterior plate being unable to provide sufficient stability for anterior bone grafting.[3] Furthermore, a plate was fixed vertically to the anterior of the bone graft for structural support, which provided a strategy that definitively improved the bone healing rate.

Double-plate technology has been widely used in comminuted fractures around joints, such as fractures of the distal humerus and of the proximal ulna.[18,19] Research confirms that double plates can provide greater mechanical stability compared with other fixation methods.[20] This practice may be applied to the treatment of nonunion as only one stable internal environment can ensure normal osteoblast functions and revascularization.[21] However, double-plate vertical fixation can be considered as an enhanced version of the technique. In this study, patients received a dynamic locking compression plate that was fixed at the lateral nonunion area to ensure that at least three locking screws were fixed at each of the distal and proximal ends (six layers of cortices). Then the reconstruction plate was fixed on the anterior of the bone graft areas to ensure that at least two locking screws were fixed at each side of the distal and proximal ends (four layers of cortices). Finally, two plates and 10 screws were used to establish cross-vertical fixation, obtaining a more powerful strength of fixation in the axial and rotational planes, effectively controlling the separation of the broken ends and the tendency to rotate. A previous study treated 14 cases of femoral shaft nonunion using the double-plate vertical-fixation technique; seven cases of atrophic nonunion and seven cases of hypertrophic non-union (eight cases after plate surgery and six cases after intramedullary nailing).[22] All patients achieved bone healing with a mean ± SD healing time of 5.2 ± 1.3 months, similar to the results reported in the current study.[22] However, the following differences between the two studies should be noted:[22] (i) this current study applied the double-plate vertical-fixation technique to the nonunion of limbs. In addition, all patients in the current study received double-plate vertical fixation and the only difference was that the exterior locking plate was 4.5 mm in the lower limbs and 3.5 mm in the upper limbs; (ii) this current study only permitted the inclusion of patients that had experienced atrophic nonunion after plate surgery. In our opinion, intramedullary nail exchange remains the gold standard treatment for nonunion after intramedullary nailing, even if there are bone defects. The preferred approach remains the use of an additional plate combined with bone grafting. However, in patients with hypertrophic nonunion, which usually has good callus growth without bone defects, there would be no need to use a full plate bone graft. Furthermore, double-plate vertical fixation would increase surgical trauma.

The selection of a fixation method affects the mechanical environment and healing processes.[21] Anatomical reset is not needed for fixation of intramedullary nails and 0.2–1.0 mm fretting between fractured ends is acceptable to achieve relative stability and indirectly influence bone healing.[23] In contrast, plate fixation requires full contact between fractured ends, with a fretting of less than 0.15 mm, to achieve absolute stability and direct bone healing.[24] Absolute stability is difficult to achieve. Although an exterior single plate can produce an effective pressure between the ends of the nonunion through power holes or via a pressure device, it remains unable to satisfy the stability requirements of an anterior bone defect or graft, resulting in early displacement of the graft, increase of interior tension, stress concentration and finally fracture of the plate.[24] However, double-plate vertical fixation can provide the most appropriate, absolute and stable environment during the healing processes of nonunion fractures.[25] Based on the exterior plate, an additional anterior plate can effectively prevent displacement of the bone graft and simultaneously ensure sufficient contact between the graft and host bone tissue.[19] As the fractured ends of the nonunion require steps to ensure absolute stability, bone healing takes longer when compared with relative stabilization.[24] In this current study, the mean ± SD duration of bone healing for the 43 patients was 8.2 ± 1.1 months, which was longer than with intramedullary nail exchange (29.8 weeks), plate augmentation (7.2 months) and single-plate fixation (6.3 months).[3,26,27]

The choice of method used and the area selected for bone grafting is also important.[28-30] This current study used structural autogenous ilium grafts and a slot was created at the anterior cortex across the broken ends of the nonunion. A Paprika sign during the procedure was apparent. The length of the slot was measured and the corresponding ilium with cortex tissue was obtained. After trimming, the ilium tissue was loaded into the slot, while cancellous bone granules were loaded into the areas of the bone defect. This method proved to be advantageous in that the bone graft fully crossed the inactive area of bone formation by bridging the active areas of bone formation at both sides. The technique maximized osteogenesis, osteoinduction and osteoconduction of autogenous bones. The bone graft was closely connected to the active areas of bone formation at both sides, leading to a continuous facilitation of neovascularization of the host bone, with osteoblast and active factors transported to inactive areas of bone formation. The ilium graft with cortex provides beam support, as does the reconstructed interior cortex, which was beneficial for the local stability of the interior bone defect. A previous study described 21 cases of femoral shaft nonunion in whom they applied local callus filling bone graft combined with horizontal fixation of double plates, which resulted in a bone healing rate of 100%.[17] The authors believed that horizontal fixation could obtain stability in the axial and rotational directions, thus supporting the medial cortex.[17] However, in our opinion, the drawback of this method is the wide range of dissection required when exposing the interior cortex and the difficulty of secondary removal. In addition, structural bone grafting used in this current study also provided support for the medial cortex, with double-plate vertical fixation avoiding interior dissection.

The major concern of double-plate fixation is whether too much soft tissue dissection would impact on bone healing.[31] However, in this current study, all 43 patients with atrophic nonunion achieved bone healing, which was consistent with the results of other studies.[17,22] The current authors considered that: (i) during surgery, application of the Judet periosteal stripping technique to expose the nonunion ends can protect the surrounding soft tissue, prevent periosteum damage and would not cause secondary trauma; (ii) atrophic nonunion was mostly associated with deformity, hyperplastic scar tissue and bone defects. Sufficient exposure was required for correction of the deformity during radical debridement and open bone graft; (iii) all cases included in this study involved nonunion after plate surgery and extra dissection of the soft tissue was not needed after removing the original plate.

Although none of the patients in the current study experienced complications (infection, fractures of internal fixation, deformity, movement disorders, pain in the area of bone harvesting) and they all had excellent postoperative bone healing and functional assessments, seven patients had claudication, three had joint ankyloses in the ankle and 13 had significant muscular atrophy compared with the healthy side. In our opinion, the main cause of these issues was a longer healing time and considerable psychological trauma experienced by the patients, which subsequently affected their postoperative rehabilitation and the therapeutic efficacy of the treatment regimen. Therefore, the treatment indication should be carefully considered initially, taking into account any bone defects, significant deformities and atrophic nonunion so that excessive therapeutic measures can be avoided. In addition, double-plate fixation provides a firmly fixed structural support. Thus, after the operation, patients should be encouraged to perform early functional, weight-bearing exercises.

There were some limitations to this study. It was a retrospective descriptive study, and with this, the ability to fully evaluate all forms and causes of nonunions were limited. Furthermore, the case number was small.

In conclusion, structural autogenous ilium grafting used in combination with double-plate vertical fixation can provide a stable structural environment for near optimal bone healing, as well as facilitating the effects of autologous bone grafts, in patients with atrophic nonunion of the upper and lower limbs. While this proposed method appears to be a reliable method for treating atrophic nonunion, the limitations of the study mean that further verification of these findings are necessary in a larger patient population.