Prognostic factors of extracorporeal shockwave therapy in the treatment of nonunion in long bones: a retrospective study

Abstract

Background:

Nonunion of long bone fractures is a significant complication following surgical fixation, with an incidence ranging from 5 to 10%. Surgical intervention is the standard treatment for nonunions, but it may come with potential complications. Nonoperative approaches, such as Extracorporeal Shockwave Therapy (ESWT), have been advocated as alternatives.

Methods:

In the retrospective study conducted between January 2004 and January 2018, 91 patients who underwent ESWT for tibia or femur nonunions were included. Nonunion was defined based on radiographic criteria and clinical symptoms. The nonunion morphology was categorized as hypertrophic, oligotrophic, or atrophic. ESWT was administered using the OssaTron device in a single treatment session. Bony union was defined as the presence of a bridging callus over the fracture site with more than three-fourths of the circumference in both planes within the 12-month postoperative period.

Results:

The study included 91 patients, with an overall union rate of 62.6%. A higher healing rate was observed in trophic nonunion(69.9%) than in atrophic nonunion(33.3%). Multivariate analysis identified the number of surgeries, maximum fracture gap, and atrophic nonunion as independent factors influencing the risk of fracture nonunion after ESWT. The receiver operating characteristic curves were generated for these factors, providing more than one surgical intervention, and fracture gap greater than 3.94 mm as negative predictors of ESWT for long bone nonunions.

Conclusion:

The study’s primary findings suggest that ESWT is effective in achieving bony union for nonunions in long bones(62.6%). Despite the overall positive results, the study highlights that atrophic nonunions, larger fracture gaps of more than 3.94 mm, and multiple surgeries are associated with poorer outcomes.

Introduction

Highlights

• Extracorporeal shockwave therapy (ESWT) emerges as a viable alternative for treating long bone nonunions.

• ESWT demonstrates heightened efficacy, particularly in long bone nonunions with fracture gaps below 3.94 mm.

• ESWT exhibits a higher failure rate in atrophic nonunions and multiple surgeries.

Nonunion of long bone fractures is a significant complication following surgical treatment. The U.S. Food and Drug Administration (FDA) defines nonunion as a fracture that has not healed within 9 months or shows no callus formation for 3 consecutive months¹. The estimated incidence of nonunion for long bone fractures ranges from 5 to 10%, with variations depending on factors such as injury pattern, anatomical site of the fracture, open or closed fracture, and patient-specific characteristics^2,3. Surgical intervention is generally considered the gold standard treatment for long bone nonunion, with success rates of up to 96%^4,5. Weber and Cech et al. classified nonunions into three types based on radiographic appearance: hypertrophic, oligotrophic, and atrophic⁶. Hypertrophic nonunions exhibit an exuberant callus formation around the fracture site, indicating a viable biological environment but insufficient stability for solid bone healing. Oligotrophic nonunions result from inadequate fracture reduction and vascular supply, leading to incomplete callus formation. Atrophic nonunions, on the other hand, show a lack of callus formation and atrophy of surrounding fragments due to the loss of the biological environment around the fracture site. Treatment options for nonunion are typically determined by the type of nonunion and underlying factors, considering both mechanical and biological aspects. In 2007, Giannoudis et al.⁷ proposed ‘The Diamond concept’, which involves four essential factors (mechanical environment, growth factors, osteoconductive scaffolds, and osteogenic cells) to achieve successful bone healing. However, it is important to be aware that surgical procedures may come with potential complications, including superficial or deep infections, muscle fibrosis, pain, donor site morbidity, and even the risk of nonunion. These complications can further burden the social health system economically^8,9.

In addition to surgical treatment for fracture nonunions, nonoperative approaches have been advocated in the literature, such as combined magnetic field therapy¹⁰, low-intensity pulsed ultrasound¹¹, and extracorporeal shockwave therapy (ESWT)⁹. ESWT was initially developed to treat urinary stones about 40 years ago¹². It is characterized by producing acoustic mechanical waves into tissues with a compressive, shear, and tensile force to induce biological response¹³. The pulse from ESWT can be delivered in various ways, including electrohydraulic, electromagnetic, or piezoelectric forms, producing a rapid high peak pressure followed by a negative pressure phase. This physical energy can directly or indirectly activate several signal transduction pathways, leading to gene expression and protein synthesis, resulting in chondroprotective effects, anti-inflammation, neovascularization, and tissue regeneration^14–16. This process of transforming energy into a biological response is known as the mechanotransduction pathway, which forms the fundamental theory behind the clinical application of ESWT¹⁷.

In 1991, Valchanou and Michailov first reported an 85.4% healing rate in the application of ESWT for the treatment of fracture nonunions¹⁸. Subsequently, over the following two decades, numerous studies have demonstrated the efficacy of ESWT in treating fracture nonunions. The International Society for Medical Shockwave Treatment (ISMST) also approved the clinical indication for the use of ESWT in nonhealing fractures based on supportive evidence¹⁹. Cacchio et al.²⁰ conducted a randomized controlled trial comparing ESWT and surgical intervention in the treatment of hypertrophic nonunions, with union rates of 70.7 and 73% at 6 months, respectively. The ESWT group experienced no adverse events compared to the 7% complications in the surgical group. Additionally, Furia et al.²¹ and Notarnicola et al.²² both demonstrated comparable outcomes between ESWT and surgery, with minimal adverse effects observed in the ESWT group. However, despite the established effectiveness of ESWT in treating fracture nonunions, certain clinical scenarios continue to yield unsatisfactory outcomes. In a retrospective study, Kuo et al.²³ identified case-specific parameters, such as nonunion morphology and gap, which may influence treatment results. Nevertheless, the factors impacting outcomes after ESWT are not always consistent across literature²⁴. Despite advancement, there persists a nonunion rate of ~20–30% in this clinical setting. In this study, our objective is to assess negative prognostic factors influencing the effectiveness of ESWT in treating nonunions of long-bone fractures.

Methods

Patients

This retrospective study received approval from our institutional review board and the work has been reported in line with the strengthening the reporting of cohort, cross-sectional, and case–control studies in surgery (STROCSS) criteria²⁵. Between January 2004 and January 2018, patients who underwent ESWT for tibia or femur nonunions were systematically reviewed. Nonunion was defined in two ways: first, as a fracture that failed to achieve cortical continuity on three out of four cortices following the initial internal fixation for a period exceeding 6 months. Second, it was described as the absence of radiographic progress towards union for three consecutive months, accompanied by an inability to bear weight on the affected extremity, pain upon palpation, or motion at the fracture site after 6 months of the initial internal fixation²⁶. Additionally, the nonunion morphology was categorized as hypertrophic, oligotrophic, or atrophic based on the classification by Weber and Cech et al.⁶. Patients with chronic osteomyelitis, a history of postoperative infection, active malignancy, or coagulopathy were excluded from the study.

Treatment

Patients with the femur or tibia nonunions underwent ESWT utilizing the OssaTron device (SANUWAVE Health, Inc.) in a single treatment session operated by a certificated orthropedist. Under general anesthesia, the patient was placed on a fracture table with a fracture site localized with fluoroscopy, and shockwaves were administered in two planes at 45^o and 60^o angles relative to the longitudinal axis of the femur or tibia. Each plane received 3000 impulses at a 28 kV energy setting, resulting in a maximum energy output in the treatment zone of 0.62 mJ/mm² energy flux density for each patient. After ESWT, the patient would mobilized with crutches for partial weight bearing on the affected extremity for 4–6 weeks.

Patient demographic data include age, sex, BMI, number of surgeries, smoking status, and underline disease of thyroid disease, and diabetes mellitus were recorded. Fracture morphology, closed or open fracture, maximum fracture gap (measured by two senior orthopedist and one junior orthopedist) measured in anteroposterior and lateral views of radiography were collected. The patients underwent follow-up with radiographic assessments comprising anteroposterior and lateral views at 1, 3, 6, and 12 months following ESWT. Bony union was defined as the presence of bridging callus over fracture site with more than three-fourths of the circumference in both planes within the 12-month postoperative period. If the bony union was not achieved within this timeframe, surgical intervention involving autologous bone grafting and revisional fixation was applied. The ESWT applied only in one session.

Search strategy

An electronic database search of PubMed was conducted using keywords such as ESWT, nonunion, long bone, and musculoskeletal. Studies focusing on union rate and risk factors for failure after ESWT were included. Exclusion criteria comprised studies where (1) nonunions were not located on long bones and (2) protocols and sessions of ESWT were inadequately described. Additionally, supplementary searches were conducted by examining the reference lists of the included literature.

Statistical analysis

The baseline characteristics are presented as mean and SD. The Kolmogorov–Smirnov test was used to test the normality of the data. The possible risk factors for nonunion, such as age, BMI, sex, number of surgery, time to ESWT, maximum fracture gap, smoking status, fracture location, thyroid disease, diabetes mellitus, and type of nonunion, were analyzed by univariate binary logistic regression. Then, the predictors with a P-value <0.30 were analyzed with multivariate logistic regression using backward selection. To quantify the treatment performance, the area under curve (AUC) were obtained from the receiver operating characteristic (ROC) curve analysis and the ROC curve was used to calculate the optimal cutoff value. A P-value of less than 0.05 was considered significant. All statistical analyses were performed using SPSS software V.21 (SPSS Inc.).

Results

Patient demographic characteristics

Between January 2004 and January 2018, we conducted a retrospective review of 137 patients, from which 46 individuals were excluded based on predefined criteria. The study ultimately included 91 patients, comprising 36 females and 55 males, with an average age of 36.1 years (range, 18–84). The mean duration from injury to the initial ESWT treatment was 15.9 months (range, 3–60). The maximum fracture gap, as measured in anteroposterior and lateral views of plain film, was 3.87 mm (range, 1–9). Regarding nonunion morphology, 50.5% (46/91) presented hypertrophic nonunions, 29.7% (27/91) exhibited oligotrophic nonunions, and 19.8% (18/91) were characterized as atrophic nonunions.

Clinical results

At the latest follow-up, the complete bony union was achieved in 57 out of 91 patients, resulting in an overall union rate of 62.6% (Table 1) without complications such as ecchymosis or hematoma formation. Patients were classified into either the union or nonunion group based on their fracture healing status at the last follow-up. No significant differences were observed in age, sex, BMI, mean time from injury to the first ESWT, smoking status, type of fracture, or location of fracture between the two groups (Table 2). In the univariate logistic regression model, factors such as the number of surgeries (OR, 9.93; 95% CI: 1.919–16.834; P=0.006), maximum fracture gap (OR, 2.18; 95% CI: 1.210–2.845; P=0.021), and type of nonunion (OR, 8.34; 95% CI: 1.591–18.157; P=0.017) were found to be associated with fracture healing (Table 2).

Table 1.

The union rates subsequent to ESWT across various types of long bone nonunion.

	Hypertrophic (N=46)	Oligotrophic (N=27)	Atrophic (N=18)	P
Union rate (%)	65.2 (30/46)	77.8 (21/27)	33.3 (6/18)	0.017
Overall union rate	62.6%

Table 2.

Demographic characteristics and risk factors associated with failure of ESWT for nonunion.

			Logistic regression
			Univariate		Multivariate
Continuous variable	Union (N=56)	Nonunion (N=35)	Odds ratio (95% CI)	P	Odds ratio (95% CI)	P
Age (y)	34.93±15.15	37.97±16.53	1.001 (0.982–1.042)	0.965
BMI (kg/m2)	25.11±4.59	26.94±5.85	0.856 (0.975–1.180)	0.117
No. of surgery	1.18±0.39	1.46±0.56	9.934 (1.919–16.834)	0.006	3.7 (1.257–10.886)	0.018
Time to ESWT (m)	15.57±11.14	16.60±12.02	1.010 (0.951–1.041)	0.795
Max Fx gap (mm)	3.43±1.26	4.58±1.30	2.177 (1.210–2.845)	0.021	1.985 (1.250–3.153)	0.004
Nominal variable
Sex (M/F)
Female	18	18	2.235 (0.943–5.395)	0.068
Male	38	17
Smoking
Yes	19	10	0.363 (0.117–1.129)	0.080
No	37	25
Type of fx
Close	48	29	0.783 (0.203–2.184)	0.819
Open	8	6
Location
Femur	38	28	1.938 (0.385–3.248)	0.605
Tibia	18	7
Diabetes
Yes	1	2	24.01 (0.338–45.401)	0.275
No	55	33
Thyroid disease			–	–	–
Yes	2	0
No	54	35
Type of nonunion
Atrophic	6	13	8.341 (1.591–18.157)	0.017	4.423 (1.296–15.096)	0.018
Hypertrophic	50	22

No. of surgery: Number of surgery.

Max Fx gap: Maximum fracture gap.

95% CI: 95% confidence interval.

The multivariate logistic regression model revealed that the number of surgeries (OR, 3.7; 95% CI: 1.257–10.886; P=0.018), maximum fracture gap (OR, 1.99; 95% CI: 1.250–3.153; P=0.004), and type of nonunion (OR, 4.42; 95% CI: 1.296–15.096; P=0.018) were independently associated with the risk of fracture nonunion following ESWT (see Table 2). ROC curves were generated based on these independent predictors. The AUC for the maximum fracture gap was 0.743 (95% CI: 0.590–0.862; P≤0.001) with a cutoff value of 3.94 mm, and for the number of surgeries, it was 0.628 (95% CI: 0.569–0.790; P=0.01) with a cutoff value of more than one surgery (Fig. 1 and Table 3).

Figure 1.

Analyze receiver operating characteristic (ROC) curves to assess optimal cutoff values for the maximum fracture gap and the number of surgeries in relation to the failure of fracture healing following ESWT. (A) AUC value: 0.743 and cutoff value: 3.94 mm; (B) AUC value: 0.628 and cutoff value: more than one surgery.

Table 3.

Predictive factors for the ROC model associated with ESWT failure in nonunion.

	AUC	95% CI	Sensitivity	Specificity	Cutoff value	P
Predictor
Max Fx gap (mm)	0.743	0.590–0.862	0.46	0.83	3.94	<0.001
No. of surgery	0.628	0.569–0.790	0.50	0.85	2	0.01

No. of surgery: Number of surgery.

Max Fx gap: Maximum fracture gap.

95% CI: 95% confidence interval.

AUC, area under the ROC curve.

Discussion

The primary findings of this study reveal that ESWT exhibited an overall union rate of 62.6% for nonunions in long bones, particularly in cases involving multiple surgeries, a fracture gap exceeding 3.94 mm, and atrophic nonunion—identified as negative predictors for achieving bony union. The efficacy of ESWT in promoting bone healing has been substantiated through animal experiments involving both acute fractures and chronic nonunions^27,28. Clinical studies have also reported its success in fracture healing^18,20. In 1997, Haupt²⁹ proposed potential mechanisms for the reaction phases of ESWT, encompassing: 1. the physical phase, involving positive and negative pressure to induce effects such as cavitation and increased permeability of cell membranes; 2. the physicochemical phase, stimulating cells to release biomolecules; 3. the chemical phase, altering the function of ion channels in cell membranes; and 4. the biological phase, modulating angiogenesis, anti-inflammatory responses, wound healing, and bone healing. While ESWT demonstrates potential efficacy in long bone nonunion, patients exhibiting highlighted negative prognostic factors should consider alternative interventions, such as a bone graft procedure.

The application of high-energy ESWT for nonhealing fractures was initially proposed by Valchanou and Michailov in a retrospective study, resulting in a bony union rate of 85.4% (70 out of 82 patients)¹⁸. In our current study, 62.6% of patients achieved bony union, a figure lower than our prior investigation⁹, which reported an overall healing rate of 80% (44 out of 55 patients). Subgroup analysis of our study revealed healing rates of 65.2% in hypertrophic nonunion, 77.8% in oligotrophic nonunion, and 33.3% in atrophic nonunion. These findings align with a prospective, nonrandomized study involving 143 nonunion patients conducted by Vulpiani et al.³⁰, who reported a healing rate of 61.3% for trophic nonunions (hypertrophic and oligotrophic) and 29.2% for atrophic nonunions. A systematic review by Zelle et al.³¹ in 2011 also indicated an overall healing rate of 76%, ranging from 41% to 85%, with a union rate of 29% in atrophic nonunion and 76% in hypertrophic nonunions. Atrophic nonunion, characterized by radiographically absent callus due to poor biology and vascularization, poses a challenge for treatment. Various surgical approaches have been proposed to address atrophic nonunion, aiming to enhance both biology and mechanical stability. These approaches include autologous bone grafting, bone morphogenetic protein, demineralized bone matrix, and parathyroid hormone therapy^32,33 with an effort to repair the biological environment. ESWT has shown the capability to up-regulate and express various proangiogenic and pro-osteogenic growth factors^15,34. This ability may modify the nonunion environment, potentially resulting in the bony union. Surgical treatment for long bone nonunions carries a reported failure rate of 10%, with atrophic nonunions identified as a risk factor for bony union (odds ratio 0.23)³⁵. Alternatively, each surgical approach could potentially exacerbate the already precarious blood supply of long bone nonunions. In our study, we observed a lower union rate of 33.3% in atrophic nonunion compared to a union rate of 69.9% in trophic nonunions. Despite the demonstrated capability of ESWT to upregulate and express various proangiogenic and pro-osteogenic growth factors reported in animal studies by Wang et al.^15,34, atrophic nonunion and number of surgeries emerged as independent negative influencing factors on the outcomes of ESWT for long bone nonunions.

The presence of bone defects poses a critical challenge in the management of fracture fixation. Existing evidence categorizes defects with a size greater than 1–2 cm and a loss of more than 50% of the bone’s circumference as critical-sized bone defects^36,37. Fracture nonunion is characterized by the absence of bony bridging between fragments, resulting in a persistent fracture gap of varying sizes. The application of high-energy ESWT in the treatment of fracture nonunions has demonstrated effectiveness in bridging these gaps³⁸. However, Vulpiani et al.³⁰ reported 27 patients with nonunion who failed to achieve bony union after ESWT, with four of them exhibiting a large fracture gap (>2 cm) on pretreatment radiographs. In a review article, Moya et al.³⁸ identified a nonunion gap exceeding 5 mm as a risk factor for poor outcomes following ESWT treatment. In our study, we also observed that a nonunion gap exceeding 3.94 mm served as a negative factor for fracture healing after ESWT.

There are several limitations to consider in this study. Firstly, it was structured as a retrospective design lacking a comparative control group. Secondly, although the case numbers were sufficient to demonstrate significant risk factors affecting bony union after ESWT, the limited sample size might somewhat compromise the robustness of the results, potentially overlooking some factors. Thirdly, our analysis only encompassed certain factors that could influence bony union; variables like NSAID usage, osteoporosis, or postmenopausal status were not included for statistical evaluation, which might exhibit a significant impact on bone union after ESWT or interact with presented risk factors. Lastly, our examination focused solely on radiographic outcomes, lacking functional measurements to demonstrate the degree of functional improvement between patients with or without bony union. A further prospective, randomized study is warranted to establish the comprehensive and high-level evidence.

Nevertheless, our study showcased that ESWT stands as an alternative treatment for nonunions, exhibiting minimal complications. Trophic nonunions exhibited a notably higher union rate compared to atrophic nonunions (69.9 vs. 33.3%, P<0.05). Unfavorable outcomes are associated with a maximum fracture gap surpassing 3.94 mm, multiple surgeries (more than 1), and the presence of atrophic nonunions. However, future prospective studies are necessary to validate and expand our findings.

Ethical approval

A retrospective study is approved by the Institutional Review Board of Chang Gung Memorial Hospital (IRB number: 202301823B0). The information is in the Methods of the manuscript.

Consent

Written informed consent is obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.

Source of funding

None.

Author contribution

K.-T.W.: conceptualization, data curation, funding acquisition, investigation, methodology, supervision, validation, writing – original draft, and writing – review and editing; J.-H.C.: formal analysis, investigation, and writing – original draft; S.-W.J.: formal analysis, data curation, and writing – original draft; P.-C.C.: methodology, validation, and writing – original draft; C.-J.W.: data curation, methodology, validation, and writing – original draft; W.-Y.C.: conceptualization, writing – original draft, and writing – review and editing.

Conflicts of interest disclosure

The authors have declared that they did not receive any honoraria or consulting fees in writing this manuscript. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

Research registration unique identifying number (UIN)

Unique identifying number: researchregistry10159.

Guarantor

Kuan-Ting Wu, Jai-Hong Cheng, Shun-Wun Jhan, Po-Cheng Chen, Ching-JenWang, and Wen-Yi Chou.

Data availability statement

The dataset used and analyzed in this study were obtained from the corresponding authors upon reasonable request.

Provenance and peer review

Invited paper for the special issue ‘Shockwave treatment’ and the guest editor is Dr Kandiah Raveendran.