ESWT has less trauma, low risk, accurate curative effect, and is a good choice for fracture nonunion patients who do not want to undergo surgery. It has shown the ability to induce healing of bone such as tibia, femur, and wrist, et al. in some studies [16, 17]. In this study, forty-two of 72 fracture nonunion, are grade 3 and 4 healing, accounting for 58% of the total number of fracture nonunion. This result further proves that ESWT has a better therapeutic effect on nonunion.
However, some patients do not get satisfying efficacy after ESWT and delay time to choose more effective other treatments. In this study, thirty of the seventy-two fracture nonunion have a healing grade ≤ 2, indicating that doctors need a way to determine early whether ESWT is effective for fracture nonunion. Applying UE, we find that some factors may be correlated to the effect of ESWT on fracture nonunion.
Statistical analysis revealed that nonunion gap was significantly correlated with ESWT outcome (r = -0.723) and was a risk factor in nonunion healing (OR = 3.074) (Table 1 and Table 2). In our study, 18 nonunion had a gap of > 5 mm and only one (5.5%) achieved healing of grade 3. To further investigate the correlation between nonunion gap and outcome, we performed ROC analysis and calculated that the cut-off was 4.200 mm. It meant that the prognosis would be poor if the nonunion gap was > 4.200 mm (Fig. 1). Previous studies had reported that a nonunion gap of > 5 mm indicated a poor prognosis with ESWT[8], and our study supported this conclusion. In addition, we found that the results of the nonunion gap measured by ultrasound and X-ray were consistent, but ultrasound technology had the advantage of observing the nonunion gap from multiple angles. [18, 19].
Local blood supply had a strong effect on bone healing and nonunion frequently occur in fractures with poor blood supply[18, 19], such as in fractures of the scaphoid bone or distal tibia. Our study revealed that good local blood supply was a protective factor in nonunion (OR = 0.191), and was significantly correlated with ESWT outcomes (r = 0.611) (Table 1 and Table 2). Moreover, we found that fracture nonunion of the femoral neck and distal tibia had a poor blood supply and weak response to ESWT. In contrast, fracture nonunion of the femur and humerus had an abundant blood supply and showed a strong response to ESWT[18, 19].
ESWT is a form of “mechanotherapy”. Impulse energy is conducted to the nonunion callus and periosteum via its “mechano-transduction” effect[20]. The impulse energy causes formation of micro-fractures and hematoma, and thus reactivates the fracture-healing mechanism[21]. According to this mechanism, we hypothesized that: effective micro-fracture and hematoma formation induced by ESWT was very important; micro-fracture was related to the hardness of callus and impulse energy. ARFI was a technique already in use to estimate the hardness of thyroid nodules or atherosclerotic plaques[22, 23]. Here, we adopted it as an innovative approach to estimating the hardness of callus, and found it practicable. After linear-regression analysis and logistic regression analysis, we found that hardness of callus/impulse energy was a risk factor of ESWT outcome (OR = 19.942), and was significantly correlated to outcome (r = 0.510) (Fig. 4). To further analyze the correlation between hardness of callus/impulse energy and ESWT outcome, ROC analysis was used (cut-off point of hardness of callus/impulse energy = 2.555) and this revealed that the prognosis was poor if hardness of callus/impulse energy was > 2.555. Moreover, hematoma formation was also significantly correlated to outcome (r = 0.722) and acted as a protective factor of ESWT outcome (OR = 0.015) (Fig. 5).
To eliminate confounding factors, multiple linear regression analysis was performed and this revealed that the four ultrasound signs correlated significantly with ESWT outcome, with a significance rank of: fracture nonunion gap (t = -5.224) > hematoma formation (t = 4.288) > local blood supply (t = 3.268) > hardness of callus/impulse energy (t = -2.056). The above results indicated that some factors might corelated to the clinical efficacy of ESWT on fracture nonunion.
Our study still has several limitations. Firstly, the sample size of the present study is insufficient, which may be correlated to nonunion morbidity; secondly, there is currently no objective standard to identify local blood supply and the outcome of nonunion, which means our standard depends on the clinical experience of the doctor and may be subjective. The further expanded sample size and the unified standards in clinical can help to confirm the conclusion.
In summary, Nonunion gap larger than 4.200 mm and hardness of callus/impulse energy larger than 2.555 are risk factors of ESWT outcome, and good local blood supply and effective hematoma formation induced by ESWT are protective factors. By analyzing the related factors of extracorporeal shockwave therapy on fracture nonunion, clinicians may screen out the patients with fracture nonunion who are not suitable for ESWT, and reduce the risk of fracture nonunion.